BRAIN DRAIN FROM TURKEY:
AN EMPIRICAL INVESTIGATION OF THE DETERMINANTS OF SKILLED
MIGRATION AND STUDENT NON-RETURN
A THESIS SUBMITTED TO
THE GRADUATE SCHOOL OF SOCIAL SCIENCES
OF
MIDDLE EAST TECHNICAL UNIVERSITY
BY
N L DEMET GÜNGÖR
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
IN
THE DEPARTMENT OF ECONOMICS
DECEMBER 2003
Approval of the Graduate School of Social Sciences
_______________________
Prof. Dr. Sencer Ayata
Director
I certify that this thesis satisfies all the requirements as a thesis for the degree of
Doctor of Philosophy.
_______________________
Prof. Dr. Erol Çakmak
Head of Department
This is to certify that we have read this thesis and that in our opinion it is fully
adequate, in scope and quality, as a thesis for the degree of Doctor of Philosophy in
Economics.
_______________________
Prof. Dr. Aysıt Tansel
Supervisor
Examining Committee Members
Prof. Dr. Aysıt Tansel
_______________________
Prof. Dr. Erinç Yeldan
_______________________
Prof. Dr. Yusuf Ziya Özcan
_______________________
Assoc. Prof. Dr. Cem Somel
_______________________
Assoc. Prof. Dr. Hakan Ercan
_______________________
ABSTRACT
BRAIN DRAIN FROM TURKEY:
AN EMPIRICAL INVESTIGATION OF THE DETERMINANTS OF SKILLED
MIGRATION AND STUDENT NON-RETURN
Güngör, Nil Demet
Ph.D., Department of Economics
Supervisor: Prof. Dr. Aysıt Tansel
December 2003, 315 pages
This study deals with skilled migration from a developing country perspective. The
migration of skilled individuals from developing countries to developed countries is often
viewed as a costly subsidy from the poor nations to the rich, and a threat to their economic
development. The first part of the study brings up to date both the theoretical and the policy
debate on the impact of skilled migration on the sending economies. The second purpose of
the study is to take a closer look at the motivations for skilled emigration from Turkey.
The emigration of skilled individuals from Turkey has attracted greater attention in
recent years, particularly after the experience of back to back economic crises that have led
to increased unemployment among the highly educated young. A survey study was
undertaken during the first half of 2002 in order to collect information on various
characteristics of Turkish professionals and Turkish students residing abroad. Over 2000
responses were received from the targeted populations. The information from this survey
iii
was then used to determine the empirical importance of various factors on return intentions
by estimating ordered probit models for the two samples.
In the migration literature, wage differentials are often cited as an important factor
explaining skilled migration. The findings of the study suggest, however, that other factors
are also important in explaining the non-return of Turkish professionals. Economic
instability in Turkey is found to be an important push factor, while work experience in
Turkey also increases non-return. In the student sample, higher salaries offered in the host
country and lifestyle preferences, including a more organized and ordered environment in
their current country of study increase the probability of not returning. For both groups, the
analysis also points to the importance of prior intentions and the role of the family in the
decision to return to Turkey or stay overseas.
Keywords: Labor Economics, Skilled Migration, Brain Drain, Student Non-Return, Higher
Education.
iv
ÖZ
TÜRK YE’DEN YURT DI INA BEY N GÖÇÜ: YURT DI INDA
OKUYAN Ö RENC LER N VE YÜKSEK Ö REN ML GÜCÜNÜN DÖNME
N YETLER ÜZER NE AMP R K B R ÇALI MA
Güngör, Nil Demet
Doktora, Ekonomi Bölümü
Tez Yöneticisi: Prof. Dr. Aysıt Tansel
Aralık 2003, 315 sayfa
Çalı mada, yüksek e itimli i gücü göçü kalkınmakta olan ülkeler açısından
irdelenmektedir. Geli mekte olan ülkelerden geli mi ülkelere gerçekle en nitelikli i gücü
göçü, geli en ülkeler açısından yüksek maliyetli bir hibe olarak nitelendirilebilir.
Çalı manın ilk bölümünde bu göçün göçveren ülkeler üzerindeki etkisini tartı an yazın ele
alınarak tartı mada ula ılan son noktanın ortaya konulması amaçlanmaktadır. Çalı manın
di er amacı, Türkiye’den yurt dı ına gerçekle en nitelikli insan göçünü belirleyen etmenleri
inceleyerek, bu göçte en etkili olanları belirlemektir.
Türkiye’den yurt dı ına nitelikli i gücü göçü özellikle son dönemlerde pe pe e
ya anan ekonomik krizlerden sonra daha da önem kazanmaktadır, çünkü ekonomik
krizlerin ardından e itimli gençlerde i sizlik önemli bir ölçüde artmı tır. Çalı ma, 2002
senesinin ilk yarısında gerçekle tirilen anket uygulamasının sonuçlarına dayanmaktadır.
Anketin hedef kitlesi yurt dı ında ö renimlerini sürdüren lisans, yüksek lisans ve doktora
ö rencileri ile üniversite e itimli i gücü olarak belirlenmi tir. Bu iki gruba ayrı anket
soruları da ıtılmı ve 2000’in üzerinde yanıt toplanmı tır. Anketlerden elde edilen verilerle,
v
çalı an profesyonellerin ve ö rencilerin Türkiye’ye geri dönme olasılıkları ve nedenleri,
sıralı probit modelleriyle tahmin edilmi tir.
Literatürde, yüksek nitelikli i gücünün yurt dı ına göç etmesinde ekonomik
nedenlerin önemi vurgulanmaktadır. Yurt dı ında kazanılan yüksek maa lar, beyin göçünün
en önemli nedenlerinden biri olarak görülmektedir. Çalı ma sonuçlarına göre yurt dı ında
çalı anların Türkiye’ye geri dönmeme kararında ba ka etkenlerin etkili oldu u
anla ılmaktadır. Yurt dı ında çalı anların Türkiye’ye geri dönmeme kararındaki en önemli
itici nedenlerden birinin Türkiye’deki ekonomik ve siyasî istikrarsızlık oldu u anla ılmı tır.
Ö renci grubunda ise yurt dı ındaki yüksek gelirler ve yurt dı ındaki sistemli ve düzemli
ya am tarzı geri dönmeme niyetinde önemli bulunmu tur. Analizde, her iki grup için
Türkiye’ye geri dönme veya yurt dı ında kalma kararında gitmeden önceki dönme niyetleri
ve ailenin rolü önemli bulunmu tur.
Anahtar Kelimeler: Çalı ma Ekonomisi, Nitelikli
Ö retim.
vi
gücü Göçü, Beyin Göçü, Yüksek
Bu tez Türkiye Bilimler Akademisi Sosyal Bilimlerde Yurtiçi - Yurtdı ı Bütünle tirilmi
Doktora Burs Programı tarafından desteklenmi tir.
This thesis has received the financial support of the Turkish Academy of Sciences
Fellowship Programme for Integrated Doctoral Studies in Turkey and/or Abroad in the
Social Sciences and Humanities.
vii
To my parents Ay e and Kâni Güngör
and in memory of my grandparents
Nilüfer and emsettin Akço lu,
viii
Maksude and Ali Rıza Güngör
ACKNOWLEDGEMENTS
I am indebted to my supervisor, Prof. Dr. Aysıt Tansel, for her untiring support and
unwavering belief in me. Her persistent encouragement has made this study possible, which
has been for me a new and rewarding learning experience.
I am very grateful to the members of my examining committee, Prof. Dr. Erinç
Yeldan, Prof. Dr. Yusuf Ziya Özcan, Assoc. Prof. Dr. Cem Somel and Assoc. Prof. Dr.
Hakan Ercan, for their kind encouragement and constructive remarks, which have greatly
improved the final version of this dissertation.
I have been very fortunate to have the continuous support and encouragement of
Prof. Dr. Fikret
enses and Assoc. Prof. Dr. Cem Somel throughout the years of our
acquaintance. I am also very grateful to them for providing a careful reading of a
preliminary version of the survey questionnaire.
In addition, I would like to express my deepest appreciation to Prof. Dr. Fikret
Görün and Prof. Dr. Oktar Türel for the interest and encouragement they have shown me
during the course of my studies.
My warmest thanks go to Prof. Dr. Mehmet Tomak and lawyer Mr. Necati Aras for
their unhesitant support of my studies. The financial support of the Turkish Academy of
Sciences and METU Research Fund, coded AFP-2000-4-03-06, are also very much
appreciated.
I gratefully acknowledge the support of the Middle East Technical University
Computer Center, Director Ayla Altun and the assistance given by the Informatics Group
Director Cihan Yıldırım in working out the bugs in the web survey form. The efforts of
many undergraduate students in the Faculty of Economic and Administrative Sciences were
ix
essential in building an e-mail database for the survey application. I would like to thank in
particular Nazlı Konaç, Esen Güngör and Burcu ensoy for their extra efforts.
I am also very grateful to the survey participants for their generosity in taking the
time to answer the survey questions and for providing invaluable feedback on my thesis
topic. By sharing their stories and making helpful suggestions for improvements in survey
design, they have greatly added to the value of this study.
Many friends, teachers and fellow PhD candidates in the economics department and
elsewhere have shared their experiences and helped me feel less alone during this lengthy
period. I am thankful to all of them.
Last but not least, I acknowledge with utmost gratitude the unlimited patience and
support shown to me by my family during the course of my studies.
x
TABLE OF CONTENTS
ABSTRACT ................................................................................................................
iii
ÖZ ...............................................................................................................................
v
ACKNOWLEDGEMENTS ........................................................................................
ix
TABLE OF CONTENTS ............................................................................................
xi
LIST OF TABLES ......................................................................................................
xv
LIST OF FIGURES .....................................................................................................
xx
CHAPTER
1. INTRODUCTION .......................................................................................
1
2. BRAIN DRAIN AND ECONOMIC DEVELOPMENT:
COSTS AND BENEFITS ...........................................................................
6
2.1
Introduction .......................................................................................
6
2.2
The Early Debate on the Impact of Skilled Migration ......................
7
2.2.1 The “Internationalists” .......................................................
9
2.2.2 The “Nationalists” ..............................................................
13
New Perspectives on the Impact of Skilled Migration ......................
17
2.3.1 “Brain Circulation” and Scientific Diasporas ....................
17
2.3.2 Recent “Brain Drain” and “Brain Gain” Models ...............
19
Concluding Remarks .........................................................................
27
3. THEORETICAL MODELS OF THE BRAIN DRAIN ..............................
30
2.3
2.4
3.1
Introduction .......................................................................................
30
3.2
Human Capital Theory of Migration .................................................
31
3.3
Theoretical Models of the Brain Drain based on Human
Capital Theory ...................................................................................
31
xi
3.3.1 Information Asymmetry as a Cause of Brain Drain ..........
34
3.3.2 Increasing Returns to Scale in Advanced Education .........
37
3.3.3 On-the-Job Training as a Cause of Brain Drain .................
40
3.3.4 Incorporating Learning-by-Doing into a Theory of
Brain Drain .........................................................................
43
3.3.5 Migration Chains and Herd Effects ...................................
44
3.4
Other Considerations .........................................................................
46
3.5
Concluding Remarks .........................................................................
49
4. LABOR MARKET CONDITIONS IN TURKEY AND TURKEY’S
EXPERIENCE WITH SKILLED MIGRATION ........................................
52
4.1
Introduction ......................................................................................
52
4.2
Supply and Demand in Higher Education .........................................
54
4.3
A Closer Look at Non-Returning Students .......................................
58
4.3.1 Private Students .................................................................
58
4.3.2 Government-Sponsored Students .......................................
59
4.4
Output of the Higher Education System: Stock of Graduates ...........
63
4.5
Can Turkey Afford to Ignore Skilled Emigration? ...........................
65
4.6
Concluding Remarks .........................................................................
68
5. SURVEY METHODOLOGY .....................................................................
71
5.1
Introduction ......................................................................................
71
5.2
Review of Empirical Studies of the Brain Drain ...............................
72
5.3
Survey Design and Methodology ......................................................
77
5.3.1 The Survey Population Defined ........................................
77
5.3.2 Sampling and Distribution Strategies ................................
80
5.3.4 Questionnaire Design .........................................................
83
5.3.5 Survey Implementation .....................................................
84
Concluding Remarks .........................................................................
85
6. SURVEY RESULTS AND
PRELIMINARY DATA ANALYSIS ........................................................
92
5.4
6.1
Introduction ......................................................................................
92
6.2
Respondent Profiles ............................................................................
92
xii
6.3
6.4
6.2.1 Age, Gender and Marital Status .........................................
93
6.2.2 Stay Duration and Country of Residence ...........................
93
6.2.3 Parental Education Levels and Parental Occupations ........
94
6.2.4 Bachelor’s Degree Institutions and Fields of Study ..........
96
6.2.5 Reasons for Going Overseas ..............................................
98
6.2.6 Family Support ...................................................................
105
6.2.7 Initial and Current Return Intentions .................................
106
6.2.8 Reasons for Returning and the Time Frame of Return ......
108
6.2.9 General Assessments of Study, Work, Social and Living
Conditions ..........................................................................
111
6.2.10 Difficulties Abroad and Adjustment Factors .....................
113
6.2.11 Evaluation of Various Push and Pull Factors ....................
117
A Closer Look at Student Respondents .............................................
123
6.3.1 Current Program and Field of Study ..................................
123
6.3.2 Types of Financial Support ................................................
124
6.3.3 Reasons for Choosing Current Institution of Study ............
126
6.3.4 Work Intentions after Completion of Studies ....................
127
6.3.5 Types of Organizations and Activities at Work after
Completion of Studies ........................................................
128
6.3.6 Push-Pull Factors by Degree Program and by Return
Intentions ............................................................................
130
6.3.7 Compulsory Military Service as a Reason for Not
Returning ............................................................................
132
6.3.8 Views of National Scholarship Recipients ........................
135
A Closer Look at Overseas Professionals .........................................
137
6.4.1 Highest Degree Held and Field of Highest Degree ...........
137
6.4.2 Stay Duration and Return Intentions ..................................
138
6.4.3 Return Intentions according to Location of Highest
Degree ................................................................................
140
6.4.4 Return Intentions by Level of Highest Degree ..................
141
6.4.5 Respondents by Occupation and Job Activities .................
143
xiii
6.4.6 Work Experience and Overseas Training ..........................
145
6.4.7 Respondents by Type of Organization ...............................
146
6.4.8 Positive Contributions to Turkey During Stay ....................
148
Concluding Remarks .........................................................................
149
7. EMPIRICAL INVESTIGATION OF THE RETURN INTENTIONS OF
TURKISH STUDENTS AND PROFESSIONALS ....................................
151
6.5
7.1
Introduction .......................................................................................
151
7.2
Estimation Procedures and Model Selection ....................................
151
7.3
Empirical Specification of the Model: Explanatory Variables .........
157
7.3.1 Income Differentials ..........................................................
158
7.3.2 Explanatory Variables for Testing Specific Brain Drain
Theories ..............................................................................
159
7.3.3 Other Explanatory Variables ..............................................
161
7.4
Determinants of the Return Intentions of Turkish Professionals ......
166
7.5
Determinants of the Return Intentions of Turkish Students ..............
189
7.6
Concluding Remarks .........................................................................
201
8. CONCLUSIONS ..........................................................................................
205
REFERENCES ............................................................................................................
213
APPENDICES
A. SUPPLEMENTARY TABLES FOR CHAPTER SIX ..............................
227
B. SUPPLEMENTARY TABLES FOR CHAPTER SEVEN .........................
249
C. SURVEY LETTERS AND SURVEY QUESTIONNAIRES .....................
264
C.1
E-Mail Cover Letter (English and Turkish Versions) .......................
264
C.2
Courtesy Reply Message (English and Turkish Versions) .................
266
C.3
English Mail-out Version of Tertiary-Educated Workforce Abroad
Survey .................................................................................................
267
English Mail-out Version of Turkish Students Abroad Survey .........
287
D. TURKISH SUMMARY ...............................................................................
305
VITA ...........................................................................................................................
315
C.4
xiv
LIST OF TABLES
TABLE
3.1 Theoretical Models of the Brain Drain ...................................................
33
4.1 YÖK Scholarship Recipients by Status, 2002 ........................................
62
4.2 Demand and Supply Projections for Selected Occupations, 2005 ( 000)
65
4.3 School Expectancy in 2000 for Selected Countries ...............................
67
5.1 A Sample of Previous Brain Drain Studies ............................................
76
6.1 Respondents by Father’s Educational Attainment Level (%) ................
95
6.2 Respondents by Mother’s Educational Attainment Level (%) ...............
95
6.3 Bachelor’s Degree Disciplines by Survey Group and Gender (%) ........
98
6.4 Reasons for Going Abroad by Survey Type (%) ....................................
99
6.5 Reasons for Going Abroad by Gender (%) ............................................
99
6.6 Reasons for Going Abroad by Program of Study, Students (%) ............
103
6.7a Top Reasons for Going Abroad by Academic Program, Students .........
104
6.7b Top Reasons for Going Abroad by Highest Degree, Professionals .......
105
6.8 Family Support for the Decision to Study or Work Overseas and for
Settling Abroad Permanently (%) ..........................................................
106
6.9 Initial Return Intentions (%) ...................................................................
107
6.10 Initial and Current Return Intentions, Professionals (%) .......................
107
6.11 Initial and Current Return Intentions, Students (%) ...............................
107
6.12 Predicted Return Dates for Respondents with Return Intentions (%) ....
108
6.13 Future Plans for Overseas Stay, by Survey Type and Gender (%) ........
110
6.14 Respondents’ General Assessment of Social Conditions in their
Current Country of Residence versus in Turkey (%) .............................
111
6.15 Respondents’ General Assessment of the Standard of Living in their
Current Country of Residence versus in Turkey (%) .............................
111
xv
6.16 Turkish Professionals’ General Assessment of Work Conditions in
their Current Country of Residence versus in Turkey (%) .....................
112
6.17 Turkish Students’ General Assesment of Academic Conditions in their
Current Institution of Study versus in Turkey (%) .................................
112
6.18 Difficulties Faced Abroad by Gender and Survey Type (%) .................
113
6.19 Top Difficulties by Gender and Survey Type (%) .................................
114
6.20 Push and Pull Factors Viewed as Important by Professionals and
Students (%) ...........................................................................................
118
6.21 Fields of Study and Return Intentions, Students (%) .............................
124
6.22 Return Intentions and Compulsory Academic Service Requirement (%)
125
6.23 Work Destinations and Current County of Residence ...........................
128
6.24 Intended Work Destinations and Organizations Immediately after
Completing Studies (%) .........................................................................
129
6.25 Intended Organization Five Years after Completing Studies by Initial
Work Destination (%) .............................................................................
130
6.26 Program of Study and Push / Pull Factors Viewed as Important by
Students (%) ...........................................................................................
131
6.27 Return Intentions and Push and Pull Factors Viewed as Important by
Students (%) ...........................................................................................
133
6.28 Top Five Push and Pull Factors according to Return Intentions ............
134
6.29 Highest Degree Held and Field of Highest Degree (%) .........................
137
6.30 Highest Degree by Level and Country (%) ............................................
138
6.31 Broad Occupation Groups and Return Intentions ..................................
143
6.32 Occupation Categories Sorted by Return and Non-Return Intentions ...
143
6.33 Percentage of Time Spent on Various Job Activities .............................
144
6.34 Type of On the Job Training and Return Intentions (%) ........................
146
6.35 Type of Formal Training and Return Intentions (%) ..............................
146
6.36 Return Intentions by Whether Respondent is Working in an Academic
or Related Environment ..........................................................................
147
7.1 Dependent Variable: Return Intentions of Turkish Professionals ..........
152
7.2 Dependent Variable: Return Intentions of Turkish Students .................
152
7.3 Marginal Effect of Gender, Professionals ..............................................
167
7.4 Marginal Effect of Having No Work Experience in Turkey ..................
171
xvi
7.5 Marginal Effect of Working in Turkey Immediately after Completing
Overseas Studies .....................................................................................
172
7.6 Marginal Effect of Contributions to Turkey ...........................................
172
7.7 Marginal Effect of Initial Return Intentions, Professionals ....................
173
7.8 Marginal Effect of Family Support and Marital Status ..........................
175
7.9 Marginal Effects of the Initial Reasons for Going .................................
177
7.10 Marginal Effects of Social and Standard of Living Assessments ..........
179
7.11 Marginal Effect of Highest Degree being a PhD from a Turkish
University ...............................................................................................
180
7.12 Marginal Effects of Fields of Study, Professionals ................................
181
7.13 Marginal Effect of Organization-Specific Formal Training ...................
182
7.14 Marginal Effect of Working in Academia or a Research Institution ......
183
7.15 Marginal Effects of Various Push Factors ..............................................
183
7.16 Marginal Effects of Various Pull Factors ...............................................
186
7.17 Marginal Effects of Difficulties Faced Abroad and Adjustment Factors
187
7.18 Marginal Effect of the Last Visit to Turkey and of September 11 .........
188
7.19 Marginal Effect of Gender, Students ......................................................
189
7.20 Marginal Effects of Initial Return Intentions, Students ..........................
191
7.21 Marginal Effects of the Family Support Variables, Students .................
192
7.22 Marginal Effects of Social and Standard of Living Assessments,
Students ..................................................................................................
194
7.23 Marginal Effect of Turkish Student Association Membership, Students
194
7.24 Marginal Effects of the Reasons for Going Abroad, Students ...............
196
7.25 Marginal Effects of Difficulties Abroad and Adjustment Factors,
Students ..................................................................................................
198
7.26 Marginal Effects of Compulsory Academic Service and Plans for an
Academic Career, Students ....................................................................
198
7.27 Marginal Effects of Various Push Factors, Students ..............................
200
7.28 Marginal Effects of Various Pull Factors, Students ...............................
200
7.29 Marginal Effects of the Last Visit to Turkey and Sept. 11, Students .....
201
7.30 Factors that have the Greatest Negative Impact on Return Intentions,
Professionals ...........................................................................................
202
xvii
7.31 Factors that have the Greatest Positive Impact on Return Intentions,
Professionals ...........................................................................................
203
7.32 Factors that have the Greatest Negative Impact on Return Intentions,
Students ..................................................................................................
204
7.33 Factors that have the Greatest Positive Impact on Return Intentions,
Students ..................................................................................................
204
A.1 Respondents by Age and Gender (%) ....................................................
227
A.2 Marital Status of Respondents ................................................................
227
A.3 Stay Duration of Respondents by Gender (%) .......................................
228
A.4 Respondents by Country of Residence ...................................................
228
A.5 Respondents by Father’s Occupation .....................................................
230
A.6 Respondents by Mother’s Occupation ....................................................
231
A.7 Language of Instruction in High School, Science and Social Science
Courses (%) ............................................................................................
231
A.8 Bachelor’s Degree Institutions of Respondents .....................................
232
A.9 Detailed Undergraduate Fields of Students with Bachelor’s Degrees ...
233
A.10 Detailed Undergraduate Fields of Overseas Turkish Workforce ...........
235
A.11 Current Program of Study by Gender, Students .....................................
237
A.12 Highest Degree Planned by Gender, Students ........................................
237
A.13 Living Accommodations by Study Program, Students (%) ...................
237
A.14 Living On or Off Campus by Study Program (%) .................................
238
A.15 Current Field of Study by Gender, Students (%) ...................................
238
A.16 Students by Current Program and Field of Study (%) ............................
238
A.17 Field of Study and Compulsory Academic Service Requirement (%) ...
239
A.18 Work Destinations after Completion of Studies .....................................
239
A.19 Intended Organization Immediately after Completing Studies by Work
Destination ....................................................................................
240
A.20 Intended Organization Five Years after Completing Studies by Work
Destination ....................................................................................
241
A.21 Respondents by Standard Occupation Classification, Broad Groups ....
242
A.22 Respondents by Detailed Occupation Categories, SOC classification ...
243
A.23 Percentage of Time Spent on R&D Activities by Occupation ...............
246
xviii
A.24 Return Intentions and R & D Intensity of Job Activities (%) ...............
246
A.25 Full Time Jobs Held in Turkey (number) ...............................................
247
A.26 Full Time Jobs Held Abroad (number) ..................................................
247
A.27 Number of Years Worked Abroad .........................................................
247
A.28 Sector of Current Organization ..............................................................
248
A.29 Type of Organization ..............................................................................
248
A.30 Location Where Current Job was Found ................................................
248
B.1 Associations of Explanatory Variables with Return Intentions (y),
Professionals ...........................................................................................
249
B.2 Summary Statistics and Descriptions of the the Variables used in the
Final Model, Professionals .....................................................................
252
B.3a Estimation Results and Marginal Effects for Outcomes y = 1 and y = 2,
Ordered Probit Model, Professionals .....................................................
254
B.3b Estimation Results and Marginal Effects for Outcomes y = 3, 4 and 5,
Ordered Probit Model, Professionals .....................................................
256
B.4 Marginal Effects for the Multinomial Logit Model, Professionals ........
258
B.5 Summary Statistics and Descriptions of the the Variables used in the
Final Model, Students .............................................................................
261
B.6 Estimation Results and Marginal Effects for Each Outcome, Ordered
Probit Model, Students ...........................................................................
263
xix
LIST OF FIGURES
FIGURE
4.1 Private Overseas Students in 2000, by Program of Study and Gender ..
59
4.2 Government-Sponsored Students, 1963-1998 ........................................
60
4.3 Number of Research Assistants Sent Abroad on YÖK Scholarships .....
61
4.4 Graduates from Universities in Turkey, by Discipline 1970-1999
(number) .................................................................................................
64
4.5 Graduates from Universities in Turkey, by Discipline 1970-1999
(percentage) ............................................................................................
64
4.6 Mean Years of Schooling for Selected Countries, 1960-1999 ...............
66
4.6 Difference in Mean Years of Schooling between Turkey and Selected
Countries, 1960-1999 .............................................................................
66
4.8 Stock of Physicians and Inhabitants per Physician in Turkey, 1928-1999
69
4.9 Stock of Nurses and Inhabitants per Nurse in Turkey, 1928-1999 ........
69
4.10 Practicing Physicians per 1000 Population for Selected OECD
Countries .................................................................................................
70
5.1 Turkish Student Enrollments at US Universities (numbers) ..................
79
5.2a Homepage of the Brain Drain Survey (English Version) .......................
87
5.2b Homepage of the Brain Drain Survey (Turkish Version) ......................
87
5.3 Turkish Brain Drain Student Survey Sample Web Page – Top .............
88
5.4 Turkish Brain Drain Student Survey Sample Web Page – Middle ........
89
5.5 Turkish Brain Drain Student Survey Sample Web Page – End .............
90
5.6 “Thank You!” Page ................................................................................
91
6.1 Bachelor’s Degree Institutions of Turkish Students Abroad (n = 967) ..
97
6.2 Bachelor’s Degree Institutions of Turkish Workforce Abroad
(n = 1223) ...............................................................................................
97
6.3 Reasons for Choosing Current Institution of Study ...............................
101
6.4 Top Reasons for Choosing Current Institution .......................................
101
xx
6.5 Return Reasons for Turkish Professionals (%) ......................................
109
6.6 Return Reasons for Turkish Students (%) ..............................................
109
6.7 Adjustment Factors by Survey Type (%) ...............................................
116
6.8 Top Adjustment Factors by Survey Type (%) ........................................
116
6.9 Students Abroad by Type of Financial Support (%) ..............................
125
6.10 Reasons for Choosing Current Institution of Study (%) ........................
126
6.11 Most Important Reason for Choosing Current Institution (%) ...............
127
6.12 Correspondence Analysis of Initial and Current Return Intentions and
Stay Duration ..........................................................................................
139
6.13 Correspondence Analysis of Return Intentions: Student Non-Return
versus Professional Migration ................................................................
141
6.14 Correspondence Analysis of Return Intentions and Level of Highest
Degree: Student Non-Return versus Professional Migration .................
142
6.15 Channels for Finding First Full-Time Job Abroad (FFTJ) and Current
Job (CJ) (%) ............................................................................................
147
6.16 Positive Contributions to Turkey During Stay (%) ................................
149
7.1 Return Intentions of Turkish Professionals, Observed Frequencies
155
7.2 Return Intentions of Turkish Students, Observed Frequencies
155
7.3a Effect of Stay Duration on Return Intentions .........................................
169
7.3b Cumulative Probabilities: Stay Duration and Return Intentions ............
169
7.4 Effect of Work Experience in Current Country on Return Intentions ....
170
7.5 Cumulative Probabilities: Social Assessment of Life Abroad ...............
179
7.6 Cumulative Probabilities: Standard of Living Assessment of Life
Abroad ....................................................................................................
179
7.7 Effect of the Interaction between Age and Importance of Advanced
Training Opportunities on the Probability of Not Returning (y = 4 or 5)
185
7.8 Effect of Stay Duration on Return Intentions, Students .........................
190
xxi
CHAPTER 1
INTRODUCTION
This study deals with skilled migration from a developing country perspective. The
first part of the study brings up to date both the theoretical and the policy debate on the
impact of skilled migration on the sending economies. The second purpose of the study is
to take a closer look at the motivations for skilled emigration from Turkey. The focal group
consists of Turkish professionals and Turkish students residing abroad. A survey study is
undertaken to collect information on the background characteristics, return intentions and
various factors affecting the return intentions for these two groups. The second part of the
study, therefore, consists of an empirical analysis of the determinants of the return
intentions of Turkish professionals and Turkish students with a view to shedding light on
the reasons behind the brain drain from Turkey.
The persistent development gap between much of the developing world and the
advanced countries has cast doubt on the convergence prediction of the neoclassical theory
of growth. The history of the development of nations has shown that while some less
developed countries have been able develop and join the ranks of the advanced countries,
other developing countries appear destined to remain in an underdevelopment trap. The
importance of initial conditions in the relative endowment of various resources is frequently
emphasized in explaining the diverse development experiences of the developing countries.
Human capital—as endowed in the stock of skilled workers—continues to receive
increasing attention as a valuable resource in the development process, apart from the usual
resources included in traditional economic growth models. It is contended, for example,
that the post-World War reconstruction of Europe and Japan could not have proceeded at
the pace that it did without the expediting role of an educated workforce. Similarly, a prior
base of human capital is believed to have played a key role in the rapid economic progress
1
of certain developing countries—particularly in East Asia—that has set them apart from
other LDCs in the development path.
Given the significance of human capital in development, an important issue is the
extent that less developed countries are affected by the continuous transfer of their human
resources to developed countries at apparently little cost to the receiving countries. Much of
the debate in the 1960s involved the moral dilemma faced by the developed nations in
accepting educated immigrants from resource-poor developing countries. One approach
centering on the individual, referred to as the “internationalist” paradigm, dismissed the
notion of a loss to developing countries. Skilled migration—being based on a rational
welfare-enhancing decision process—necessarily made individual migrants better off. At
the extreme end, whether migration of educated workers helped or hurt those “left behind”
was a matter of irrelevance for some “individualists” since migration ensured the efficient
allocation of global resources and increased global output, which they claimed benefited all
countries. Advocates of the “nationalist” paradigm, on the other hand, maintained that the
losses to the less developed sending countries were indeed very real and proposed policy
measures to mitigate developing country losses, including the much discussed Bhagwati tax
in the 1970s. Chapter Two presents a synopsis of these early paradigms that have placed the
brain drain phenomenon within a “nationalist” and an opposing “internationalist” or
“individualist” context, and the corresponding policy implications that follow from these
views.
More recently, attention has shifted to the possibility of benefits from brain drain
for the source countries. The concepts of “brain gain” and “brain circulation” have become
recurring themes in this literature and are used to illustrate the possibility that human
capital movements may not net out to a loss for sending countries. On the contrary, these
studies contend that a more complex picture emerges when skilled migration is viewed as a
dynamic process whereby those going abroad return home, even temporarily, to teach or
work in some productive capacity. With the advances in technology and the widespread use
of communications technologies, it is even argued that the physical presence of individuals
is no longer as necessary as it once was for countries to benefit from the knowledge and
expertise of expatriate populations. It is thus suggested that less developed countries can
make use of these new communications channels to transfer the knowledge of their
expatriate population, without the need for them returning.
2
On the theoretical side, the growth in the dynamic endogenous growth literature has
also influenced the modeling of the effects of skilled migration. There is a clear departure
from the earlier literature based on the neoclassical framework to models that account for
education and knowledge externalities. One segment of this literature has reinforced a
negative outcome for developing countries, while another segment has introduced the
possibility that emigration can have a positive impact on source countries by increasing the
incentive to invest in education. The latter part of Chapter Two reviews this current strand
of the brain drain literature that considers the possibility of positive effects on the source
country in addition to the traditionally cited benefit from remittances.
While much of the theoretical work on the brain drain has focused on the
macroeconomic impact to the developing country from the loss of human capital, the
number of studies that examine the theoretical causes behind the decision to migrate has
been limited. Since aggregate migration is the result of a complex decision-making process
on the part of highly skilled individuals, modeling this process becomes important to
understanding its causes. Chapter Three provides a review of the theoretical contributions
to the modeling of the migration decision. Some of these include placing the brain drain
within a framework of asymmetric information, while others emphasize the role of high
premiums given to specialized skills formed in the host country. These theories provide the
theoretical framework for the empirical investigation presented in the remainder of the
study.
The motivation for migration is, in general, set within a framework wherein various
forces act on the individual’s migration decision. These forces are usually expressed in
terms of a set of “push” factors emanating from an individual’s current environment in the
source country and a set of factors external to this environment that serve to “pull” the
individual to a new location. Migration takes place when the individual, after weighing the
various alternatives before her, makes the assessment that her welfare will improve as a
result of the decision to migrate.
In economic models of the brain drain, the weighing of alternatives takes place
within a rational decision-making process in which individuals are assumed to be capable
of evaluating the total lifetime welfare to be derived from working in the native country and
compare it with the total welfare from working in a foreign country. Income differentials
3
are believed to weigh heavily in this decision and are often presented as the central reason
for why individuals with high levels of education choose to migrate abroad. The
expectation of a higher income stream to be received in the foreign country is thus believed
to act as an important trigger for migration. When presented with the opportunity, the
rational individual from a developing country is expected to migrate to where she will earn
a higher return than she can expect in her native country.
Given the central role of income differentials in theoretical work, an important
question is how significant income differentials actually are in the decision to migrate to a
foreign country. And what is the relative importance of other factors that are most often
cited as playing a role in this decision? The second part of the study is thus concerned with
determining the factors that are of greatest influence in the international migration decision
of educated workers. Specifically, the focus is on the brain drain from Turkey to the rest of
the world. With a view to understanding the reasons behind the migration decision of
highly skilled individuals from Turkey, a survey was conducted during the first half of
2002 to determine the characteristics and return intentions of Turkish professionals and
Turkish student residing overseas. The survey yielded over 2000 responses from the
targeted populations. Prior to presenting the survey methodology and findings, a general
background section on labor market conditions and skilled migration in Turkey is provided
in Chapter Four.
Chapter Five is devoted to a detailed discussion of the survey methodology. The
questionnaire results are then used to identify empirically the importance of the various
factors involved in the decision to stay (or leave). This decision is motivated in part by
“pull” factors such as favorable compensation packages, a world-class work environment,
better living conditions, active recruitment by employers and so on and in part by “push”
factors that originate in the home country that may include political instability, cost of
living/inflation, and the inability to find work. Chapters Six and Seven present the analysis
of the results for both the student and professionals samples. The information collected
from the survey is used to determine whether some of the theoretical reasons given for the
brain drain in Chapter Three hold for the targeted groups.
The main findings of the empirical analysis may be summarized as follows:
Economic instability and uncertainty appears to be an important push factor for the Turkish
professionals working overseas. In addition, respondents who have returned to Turkey to
4
work after completing their studies and then decided to go abroad a second time are among
the least likely to return. For the student sample, pull factors including higher income levels
appear to have greater importance in determining return intentions. Higher salaries
offered in the host country and lifestyle preferences, including a more organized and
ordered environment in their current country of study increase the probability of student
non-return. For both groups, the analysis also points to the importance of prior intentions
and the role of the family in the decision to return to Turkey or stay overseas.
5
CHAPTER 2
BRAIN DRAIN AND ECONOMIC DEVELOPMENT:
COSTS AND BENEFITS
2.1. Introduction
The term “brain drain” describes the migration of highly skilled individuals from
their countries of origin to countries and regions that offer them greater opportunities. The
early debate on the brain drain commenced in the 1960s, and focused initially on the
welfare consequences of skilled migration for the sending and receiving countries.
According to one group of analysts, international labor mobility provided a mechanism for
the efficient reallocation of resources across borders, and international labor movements
were viewed as an equilibrating force for labor demand and labor supply on a global scale.
Another group of analysts maintained that the migration of skilled workers left the sending
countries worse off—especially less developed countries with low levels of human capital.
For these economies, the loss of valuable human resources through the emigration of their
skilled populations was believed to be particularly damaging, given that human capital
investments are costly and skilled workers are difficult to replace.
The two different views on the effects of skilled migration have been labeled the
“internationalist” and “nationalist” paradigms, reflecting the particular vantage point of the
advocates of each. One of the aims of the present chapter is to provide an overview of the
evolution of the brain drain literature since the early 1960s, starting from the early welfaretheoretic analyses set within a neoclassical framework to the present-day studies inspired
by the human capital-based “endogenous” growth theories. The central role given to human
capital accumulation in the new growth literature has added to the relevance of the concern
for the loss of skill individuals through emigration. While some of these studies have
reinforced the negative results of the early analyses (Miyagiwa, 1991; Haque and Kim,
1994; Galor and Tsiddon, 1997), other studies have taken a different approach and claim
6
that sending countries may actually stand to benefit from allowing a certain amount of
skilled emigration to take place. These are the so-called “beneficial brain drain” models
(Mountford, 1997; Vidal, 1998; Beine, Docquier and Rapoport, 2001). The chapter
provides an assessment of the early debate and the newer perspectives on the migration of
skilled workers and discusses the costs and benefits of skilled migration within both
frameworks. The early debate on the impact of skilled migration is presented in Section 2.2.
This is followed, in Section 2.3, by a review of the more recent contributions to the “brain
drain” literature including some recent, initial attempts to test the validity of the “brain
gain” assumption.
2.2. The Early Debate on the Impact of Skilled Migration
The brain drain phenomenon has been widely investigated since the mid-1960s
both in academic circles and by policymakers. In the United States, for example, special
commissions were set up by the U.S. Congress (1967, 1968a, 1968b) to specifically
examine and produce policy suggestions for the brain drain problem. Skilled migration
from the less developed countries to the advanced countries was viewed as a serious threat
to the development of these countries. The policy of accepting highly skilled immigrants
and allowing talented foreigners to work in the United States thus appeared to fall in
contradiction to the spirit of the aid packages provided by the U.S. that were intended to
train local manpower in these countries (U.S. Congress 1968b: 14-15). It was suggested
that “to the maximum feasible degree” foreign technicians in the United States be
encouraged to return to their home countries.
Too much reliance on foreign skills was a concern for the United States since it
provided an “easy solution” to structural problems within the U.S. economy that prevented
skilled individuals from being trained nationally. One of the proposed solutions to the brain
drain problem was the adoption of policies that would enable the U.S. to produce the
needed skills “at home”. This included easing monopolistic restrictions that created
artificial internal barriers to certain professions, such as medicine. It was argued that the
U.S. had the resources to invest in producing the needed skills without the need to resort to
immigration to the scale that it did (U.S. Congress 1968a: Report by W. Adams: 60-61).
The brain drain, it seems, raised not only economic concerns, but ethical and moral
dilemmas as well for developed countries. While the advanced countries also experience
7
outflows of national talent, the flow of skilled labor among the developed nations (“northnorth migration”) is generally seen as less problematic, since these nations have a greater
amount of resources and policy options at their disposal for remedying the structural
insufficiencies within their economies that lead to the loss of their skilled manpower. The
ability of advanced countries to replace their own emigrants with skilled immigrants from
less developed countries also carries the implication that the consequences of manpower
losses are less severe for developed economies than for the LDCs. Consequently, developed
countries appeared to be facing a moral obligation to help in the economic development of
developing countries that suffer economic losses from the brain drain, either by extending
aid packages or applying more restrictive immigration policies.
Tied to the early policy debate, separate views on the welfare effects of skilled
labor movements emerged in the academic literature. The concern of much of this early
literature involved the distinction between the welfare of emigrants and the welfare of those
“left behind”. Voluntary migration—the category that skilled migration is presumed to fall
into—was, in general, viewed to be welfare-improving since it was based on the rational
choice of individuals acting on the desire to improve their personal well-being (under the
implicit assumption of no uncertainty in outcomes). The more pertinent welfare concern,
from the standpoint of the LDCs, then became the issue of whether non-emigrants were
affected by losses in skilled manpower.
Neoclassical economic theory provides the framework for this early discussion of
the effects of skilled migration on the economies of the source countries. Neoclassical
theory has clear predictions for the effect of factor movements (the migration of capital and
labor) on factor prices (the rental rate of capital and wages offered to workers). Capital and
labor will flow from locations where they are relatively abundant to locations where they
are relatively scarce. This is explained by the law of “diminishing marginal productivity”:
increases in the quantity of an input will eventually lead to a decline in the productivity of
each additional unit of the input, if every other factor of production remains constant.
Given the assumption that factors are paid the value of their marginal product, each
factor will elicit a higher return in locations where it is relatively scarce. Migration, by
altering the capital-labor ratios (the relative factor endowments) of the source and host
countries, leads to changes in the marginal productivities and rates of return to capital and
labor. In the source country, there will be a rise in the marginal productivity and wage level
8
of the factor that emigrates (labor) and a fall in the marginal productivity of the factor that
stays put (capital or unskilled labor depending on the treatment). In the host country, the
inflow of labor will lead to a fall in its level of productivity and in its rate of return. Labor
will continue to flow out of the source country as long as a wage differential exists between
the source and host countries, and will stop only when the returns are equalized in both
locations. Given the free movement of individuals and full flexibility of factor prices,
neoclassical theory predicts that income differences between countries will vanish in the
long run.
Within this framework, the early theoretical discussion of the effects of skilled
migration may be divided into two distinct views, labeled the “internationalist” and
“nationalist” paradigms. The focal point of the two views in the debate over the migration
of skilled workers is better understood by categorizing the first as an “individualistic”
approach and the latter as an approach that brings societal or national welfare to the
forefront. These views are discussed in turn along with their policy implications in the next
two sub-sections.
2.2.1 The “Internationalists”
The supporters of the internationalist paradigm claim that one of the positive
effects of skilled migration is the increase in overall output from greater worldwide
allocational efficiency. The reallocation of skilled individuals to areas that make better use
of their skills increases their productivity, and this has a positive effect on world output.
The increase in total output, in turn, is purported to benefit all economies including the
economy of the source country (Johnson, 1968). The “internationalist” paradigm involves
the belief that individuals should be free to move about as they fit in search of greater
opportunities and better lifestyles in order to improve their welfares. They place great
importance on the freedom to act as an individual and in the freedom to exercise personal
choice. Harry Johnson (1965, 1967 and 1968), who is representative of the internationalist
paradigm, suggests that the notion of a “nation” and that of “nationalistic ties” are outdated
concepts, and that individual well-being or welfare is what matters in the migration
decision, provided that the private gains from migration do not bring a social cost to the
world.
9
In general, the private gains to the skilled migrant are believed to be positive since
the migration decision is motivated by significant private welfare gains to the individual.
This is believed to be especially true for migrants from developing countries where income
differentials between the country of immigration and the country of origin are substantial.
Johnson (1968: 79) has argued that national ties bring an “artificial barrier” to migration
and the “efficient allocation of ... talents among countries” since individuals with strong ties
to their native countries will migrate only when it involves quite significant gains in their
private incomes. Given that there is “very little possibility” of a loss to the migrant, the
pertinent question, according to Johnson, was whether any social costs are incurred from
the migration. Migration is viewed as beneficial if the private gains of the migrant exceed
the net social loss to the world. Specifically, Johnson (1968: 80-81) maintains that “any
possibility of a world loss ... hinge[s] on a loss of externalities to the country of emigration,
unmatched by an offsetting gain of externalities to the country of immigration, and
quantitatively large enough to outweigh the private income gains to the migrants”. Thus, so
long as the private gains to the individual and the social gains to the country of emigration
are greater than the social loss to the country of origin, there is a net world gain.
This represents a more extreme position within the “internationalist” paradigm.
While other analysts associated with this paradigm do not necessarily take the position that
losses to the migrant’s country of origin are not important so long as gains everywhere else
outweigh them, they nonetheless minimize the extent of these losses. For example, Grubel
and Scott (1966a) have acknowledged the possibility of redistribution effects through
changes in the marginal products of the remaining population. Yet, they have also
maintained that these income redistribution effects are negligible because of the “small”
numbers involved in the migration of skilled workers from the less developed countries.
In general, it is claimed that within a free market, laissez-faire setting, and in the
absence of (significant) externalities, there will be no adverse consequences for source
countries (Grubel and Scott, 1966a). In a market economy, each person is paid the marginal
product of her services. Since the migrating individual takes both the value of her marginal
product and her share in national income with her when she leaves, the incomes of the
remaining population are unchanged. Although per capita income may be reduced, this is
labeled a “statistical phenomenon” with no real welfare costs to the remaining population.
This analysis holds for small or marginal movements of skilled labor, which is one of the
10
important assumptions underlying the internationalist analysis. Within the same framework,
however, Berry and Soligo (1967) have shown that for non-marginal flows, the welfare of
the remaining population will, in fact, be reduced. It is pertinent at this point to note that
skilled migration flows from less developed countries to the developed world have grown
substantially over the years (Cervantes and Guellec, 2002), which is to say that the claim of
“marginal” or “inconsequential” flows has become less and less convincing.
The “internationalists” or “individualists” who adopted the neoclassical framework
in investigating the welfare consequences of international migration movements in the
1960s and 1970s reached the conclusion that in a market economy any long run losses for
the countries involved would be small, and that benefits to individuals in the form of
increased incomes and benefits to the world in the form of an increase in world output
would be greater than losses to non-emigrants in the LDCs. Possible adverse consequences
could arise from short run delays in the structural adjustment of economies to migration
flows. Grubel and Scott (1966a) have claimed that welfare losses are more likely to occur
in planned or centralized economies where workers may not be fully compensated for their
contribution to output. The policy conclusion is that markets should be kept free of
distortions including subsidies to education and policies that prevent wages from adjusting
freely and quickly to market conditions. It is also suggested that the developing countries
should adopt a “laissez-faire” policy toward study and work abroad since “foreign
education and immigration [are seen] as a ‘private investment’ outside the sphere of
government interference” (Chang and Deng, 1992: 56). A laissez-faire approach is
purported to benefit the home country governments by eliminating the financial burden of
sponsoring overseas studies.
The “internationalist” viewpoint is also called the “cosmopolitan liberal position”
by Harry Johnson, who has been one of its staunchest supporters. This view may simply be
summarized as the position that when individuals take actions to better their personal
welfare, the end result will be an improvement in global welfare. Ellerman (2003) identifies
two major weaknesses of the cosmopolitan liberal argument. First, he argues that the
actions of individuals and groups should not be viewed as independent, since there are
myriad interdependencies among various actors in development that will affect the final
outcome of any single action they may take. To illustrate, Ellerman presents an interesting
view of development as a multi-person prisoners’ dilemma situation in which the gains to
11
the individual from migrating quickly vanish as more and more individuals migrate. This is
due to the assumption of “diminishing returns to migration”, which may be “interpreted as
a tightening of controls at the receiving end and thus a raising of the costs of migration.”
While it is always in the best interest of the marginal individual to migrate, the end result
when everyone migrates is that no one reaps the benefits of migration and no one reaps the
benefits of development. The dominant outcome of the “game” is a situation where
everyone cooperates (e.g., stays home to work for the development of their home country.)
Although this presents a very simplified model of the possible effects of migration on
development, it is nevertheless a useful conceptual device for recognizing that individual
actions combine to form social phenomenon that may have very significant aggregate
repercussions, which are then reflected back on the individual. As Schelling (1978: 24-25)
has pointed out, although “people may care how it all comes out in the aggregate, their own
decisions and their own behavior are typically motivated toward their own interests, and
often impinged on by only a local fragment of the overall pattern” implying that “there is
no presumption that the self-serving behavior of individuals should usually lead to
collectively satisfactory results.”
Ellerman also calls attention to another very pertinent criticism of the liberal view
of migration, which is that while exit restrictions by the home country are considered to be
a violation of the rights of the individual, developed countries justify the restrictive
immigration policies that they themselves impose on the ground that such policies are “the
‘proper’ exercise of national sovereignty” (Ellerman, 2003). Developed countries have been
quick to advocate trade liberalization in goods and in the services of certain types of highly
skilled individuals (e.g., high-level personnel transfer within multinational corporations and
the movement of personnel on exploratory business trips), but have been less enthusiastic
about increasing unskilled labor mobility1. If the liberal viewpoint is to be taken at face
value, then permitting the movement of unskilled and semi-skilled labor from the LDCs to
the developed countries should result in substantial world gains by allowing LDCs to
exploit their comparative advantage in less skilled labor.
1
See Mattoo and Carzaniga (2003) for recents discussions on the General Agreement on Trade in Services
(GATS), mode 4, which pertains to the temporary movement of service providers across borders.
12
2.2.2. The “Nationalists”
According to the “nationalists”, the long term indirect contributions to the source
country’s economy from an increase in world output are unlikely to counter the immediate
short term losses to the developing country that result from the absence of skilled workers
and their services (Watanabe, 1969). The supporters of the “nationalist” view, in contrast
to the internationalists, do not consider the migration of skilled individuals to developed
countries to be simply a matter of freedom of movement. The net loss to the developing
country matters since the consequence is a worsening global income distribution. The
developed countries, for their part, do not stay indifferent to the possibility of distributional
effects within their borders as a result of in-migration, especially when immigrants are
viewed as a threat to native jobs. Patinkin (1968: 101) argues that the nation-state gains
even greater significance when the “world welfare” perspective of the internationalists is
adopted because:
...whereas nation-states can and do carry out fiscal policies (progressive
income taxes, transfer payments and the like) to ‘correct’ the effects on the
distribution of income generated by a free market process within their
borders, there is no world government to do this on an international basis.
There are indeed flows of aid from one nation to another—but the relative
impact of such aid on the world distribution of income is surely much less
than that achieved by a nation-state within its borders.
Another criticism aimed at the internationalist approach is that, while it considers
the possibility of positive externalities from having an educated population for the less
developed country, these are judged to be too small to have important welfare
consequences and to warrant further attention. Positive externalities occur because the
social returns to education are greater than the private returns to the individual. Since
individuals do not take into account social returns when deciding on their investment in
education, they may obtain an amount of schooling that is less than socially optimal.
In the absence of externalities, the foremost cost of skilled worker emigration to the
sending country is believed to be the investment, both public and private, made in
educating the migrant. The total cost of education is the direct costs and the foregone
earnings from not participating in the labor market. In calculating the loss in national
income to the sending country, the present value of the emigrant’s expected future income
stream must be taken into account (Watanabe, 1969). Grubel and Scott (1966a) have argued
13
that the home country will not lose from the emigration if the migrant’s marginal
productivity equaled the income he/she received—in other words, if the costs
(remuneration) from his/her employment was fully compensated by his/her contribution.
This argument is flawed for the following reasons: 1) the difficulty in measuring marginal
productivities; 2) ignoring the replacement cost of the skilled worker2; 3) viewing income
as a “cost” to the national economy and overlooking the multiplier effects that this income
would have generated through spending; and related to this 4) ignoring the higher
propensity to save outcome of income of a more educated workforce.
The argument that the loss of skilled manpower is not welfare-reducing for
developing countries also hinges strongly on the distinction made between the short run and
the long run. The internationalist framework concentrates on the long run steady state
consequences of labor movements, and ignores the short to medium term “transition
dynamics” of the economy in the adjustment from the earlier non-migration steady state to
the new post-migration steady state. The transition involves a slow process of re-educating
and replacing skilled workers lost to the economy through migration.
The aggregate data on human capital movements even differentiated for the level of
education mask “quality” differences in the movements in and out of a country. It has been
pointed out that the loss of one key scientist or innovator may mean immeasurable losses to
the domestic country. On the other hand, if the key scientist is not provided a productive
environment (e.g. given the facilities or required materials to carry out high value-added
projects), then the domestic country may not stand to gain as much by keeping this
individual than would the receiving country. For less developed countries, some would
argue that the top priority may not be to raise scientists and innovators especially when the
basic education system is lagging behind in investments in infrastructure and improvements
in quality.
Shortages or surpluses within the source country for different types of skilled labor
are also important in determining the severity of the manpower loss for the national
economy. The “costs” in replacing the emigrants with less qualified individuals (the loss in
efficiency within the national economy) should be accounted for, as well as the loss in the
positive externalities that would have been created from having a greater pool of skilled
individuals together in the economy. Some have argued that the loss of even one key highly
2
These are labelled “frictional costs” by Grubel and Scott.
14
skilled individual may entail significant repercussions for the developing economy. An
example of this is given by Watanabe (1969: 410):
There may be cases … where, but for the emigration of highly trained
personnel, a new enterprise could be launched, absorbing a large number of
hitherto unemployed workers. In such cases, which may not be rare in
developing countries, the total impact on employment could be
considerable owing to the multiplier effect and, at the same time,
technological progress and consequent improvements in productivity would
be greatly retarded.
Thus, as the passage illustrates the possibility that potential benefits of retaining
key personnel may be compounded for the source country through the multiplier effect.
However, there is uncertainty involved in whether and to what extent such benefits will be
realized. This type of potential benefit that the skilled emigrants would have bestowed on
the home economy, if they had stayed, is difficult to incorporate into a measurement of
their marginal productivities. On the other hand, the presence of skilled individuals in an
economy increases the probability that new enterprises are launched. This is akin to the
“increasing returns” argument given in the “new growth literature” for the positive
externalities created by networks of firms or individuals. The host economy reaps the
benefits from this externality, while the source country because of her relative lack of
skilled workers in the first place suffers a loss.
According to Baldwin (1970), the losses to the source country will be less severe if
there is an abundance of surplus in the economy of the types of workers that are emigrating.
The expansion of the higher education system in many developing countries has brought
with it considerable increases in the number of college graduates in these countries.
Manpower surpluses in certain disciplines are usually the consequence of the education
system of the country. Surpluses or deficiencies in certain disciplines may be the result of
the joint influence of institutional factors stemming from the structure of the higher
education system and “prestige factors” that compel students to choose disciplines based on
the “points” allocated to them. It may be argued that the institutional inefficiencies of the
higher education system cause a mismatch between the supply of workers and jobs
available in various disciplines. Viewed from this perspective, skilled migration becomes a
means of eliminating the “structural mismatch” across labor markets on a global scale.
15
But why does such a mismatch occur in the first place? That there would be a
mismatch between the supply of graduates and the needs of the domestic labor market
suggests that manpower planning strategies may be appropriate to increase the
employability of students locally once they graduate.
Ironically the externalities argument, summarily dismissed as inconsequential by
some of the proponents of the internationalist view, is the centerpiece of the recent
“beneficial brain drain” studies. While the early challengers of the neoclassical theory of
migration were critical of the dismissal of the externalities created from an educated
population, their approach and the outcome they predicted for the effects of educational
externalities on the welfare of LDC residents were in direct opposition to that of the
beneficial brain drain (BDD) studies. (Section 2.3 takes a detailed look at the BDD studies.)
The early theoretical contribution by Grubel and Scott (1966a) to the brain drain
literature is set within a neoclassical framework of perfect competition, flexible wages and
the absence of unemployment. The implications that emerged from the framework,
developed by Grubel and Scott (1966a), Johnson (1967) and Berry and Soligo (1967) were
challenged more rigorously in the 1970s. One of the critical assumptions for the predictions
of the neoclassical theory of migration is that factor prices are flexible and adjust rapidly in
response to labor movements to bring about factor price equalization across countries.
However, once market distortions are introduced into these models, their welfare
implications may be altered. The work of Bhagwati and others in this area in the mid-1970s
has shown that market distortions in the form of wage rigidities and education subsidies
may significantly change the welfare consequences for the countries involved. Bhagwati
and Hamada (1975) abandoned the assumption of flexible wages in order to provide a more
realistic setting for studying the consequences of the brain drain on the economic growth of
the sending country. They adopted instead the assumption of “rigid” wages, which enables
the possibility of unemployment in the economy.
Given the possibility of significant economic losses for the source country, which is
purported to be the case for developing countries in particular, this raises the question of
finding appropriate policies to mitigate these losses. Should the flow of skilled workers be
stemmed through selective restrictions on migration or should there be some income
transfer between skilled migrants to those remaining in the home country as reparation for
the economic losses to the source economy. Policies that focus on compensation through
16
taxation and other income transfer schemes require accurate measurements of the losses
and benefits, both direct and indirect, that the move may entail, and this is a formidable
task. The much-debated tax proposed by Bhagwati has never been implemented.
2.3 New Perspectives on the Impact of Skilled Migration
Many of the theoretical studies of skilled migration in the 1960s and 1970s
concentrated on the negative repercussions of international migration on developing
countries. Advancements in technology since then have greatly improved communication
among remote places, lowered travel costs and, on the whole, have increased interactions
between countries, and between expatriates and local residents. More recent discussions
have placed greater emphasis on positive aspects of international migration, such as the
existence of feedback mechanisms between natives and “scientific diasporas” and on the
possibility of positive externalities, which was discussed in the early literature but not
formalized. The more formal treatment of positive externalities has emerged in parallel to
the new approaches that place technology and learning within an “endogenous growth”
framework (see, for example, Aghion and Howitt, 1998). Some of the recent endogenous
growth theories of the impact of brain drain on economic growth underscore the possibility
that skilled migration may create incentives for greater educational investment in the source
economy and thus lead to greater human capital formation than would have occurred
without the possibility of migration. The positive influence of the possibility of emigration
on educational incentive structures in the home country has been dubbed the “beneficial
brain drain” or “brain gain”. Section 2.3.1 gives a brief summary of the emerging literature
on the new complexities of international skilled migration including “brain circulation”,
“reverse brain drain” and the effects of scientific diasporas and networks on sending
countries. Section 2.3.2 outlines some of the new growth models that incorporate
externalities in the study of the effects of skilled migration on economic growth.
2.3.1 “Brain Circulation” and Scientific Diasporas
Lower transportation costs have made it easier to travel globally, and the overall
mobility of skilled labor has increased as a result. It is argued that this has contributed
positively to developing countries by facilitating the “return” of expatriates, even if for
short periods of employment. Those who return are believed to impart valuable knowledge
17
and skills, gained from overseas experience, to their colleagues and work environment in
their native countries. Recent evidence suggests, however, that while skilled migration
between advanced countries is often of a temporary nature, emigrants from developing
countries are less likely to return to their home countries (Cervantes and Guellec, 2002). A
recent study of the Italian brain drain (Becker, Ichino and Peri, 2003), for example, shows
that the outflows of skilled individuals during the 1990s did actually represent a “brain
drain” when compared to the number of returnees—the stock of foreign college graduates
in Italy. The human capital content of this outflow also appeared to be significant, since a
greater share of graduates from the best Italian universities were going abroad. Given that
Italy has one of the lowest shares of college graduates among the OECD countries
according to the 2000 OECD figures, this suggests a significant loss in human capital for
Italy. The extent that a country benefits from return migration is dependent on its level of
development, and the least developed of the advanced countries can also suffer from a brain
drain, as in the case of Italy.
Thus, advanced economies appear to benefit more extensively from what has been
called “brain circulation” or “brain exchange”—the return of skilled workers after a period
of study or work abroad—in contrast to underdeveloped countries, unless the LDCs take
specific policy measures to make the return option more attractive for their expatriate
populations. South Korea provides a good example of a model of state-led return migration
or repatriation. The reversal of brain drain in South Korea is described as being “not a
spontaneous phenomenon, but ... a concerted state activity, vigorously persued from the
early phase of Korea’s industrialization in the mid-1960s” (Yoon, 1992: 5). The success of
South Korea in repatriating its skilled elite is attributed to the strong commitment of the
Park regime to building a scientific and technical base and the use of “directive” measures,
including very active and deliberate recruitment policies, the setting up of the Korea
Institute of Science and Technology (KIST), and assigning a significant degree of power to
technicians and scientists, which was unprecedented in Korean society. According to Yoon
(1992), the success and leading role of the state in institution-building, increasing R&D
capacity and recruiting overseas personnel provided a strong example, which the private
sector readily followed.
The emerging literature on the “benefits of brain drain” also focuses attention on
interactions between expatriates and local residents as another means by which a source
18
country may gain from “brain drain”. Meyer and Brown (1999) investigate the recent rise
of “diaspora networks”, some of which have the specific aim of contributing to the
development of their home countries. A problem with diaspora networks, however, is that
they may face commitment problems and disband easily. Large developing countries, such
as China and India, produce large diasporas and are more likely to benefit from these
networks than smaller countries, which means that making use of diaspora networks cannot
be a viable option for all developing countries (Kapur, 2001). It may be too soon to reach a
conclusion about the success of these formal networks, although the question is not so
much whether these networks are beneficial, but whether they can be used effectively as a
serious policy option for less developed countries.
Technology has greatly increased the number of informal networks as well,
allowing for greater interactions among professionals in different countries. Local
professionals can make use of these networks to consult and collaborate with expatriate
colleagues. In the academic professions, for example, project and study collaborations with
overseas colleagues undoubtedly benefit individual researchers by increasing their research
productivity and helping them advance in their professions. Yet, one could ask whether
improvements in the individual productivity of academicians through such interactions are
as significant to the needs of the higher education systems of developing countries as would
be the actual returning of scholars. There is, for example, chronic understaffing in the state
universities of Turkey that would be eased by the return of academicians to university posts
in Turkey. India is another case in point. Although a massive state-led expansion of the
higher education system—in terms of the number of higher degree granting institutions and
affliated colleges—took place after India’s Independence in 1947, this was not matched by
an equal devotion to raising the quality of education and providing adequate funding for
building and updating facilities (World Bank, 2000: 40-41). As a result, India’s higher
education system has been unable to attract and retain qualified academic staff and has lost
many of its best graduates to overseas universities and to private enterprise, which offer
better pay and work conditions.
2.3.2 Recent “Brain Drain” and “Brain Gain” Models
This section considers in detail some recent models of “brain drain” and “brain
gain” that are set within a human capital-driven endogenous growth framework, which
became popular with the seminal article of Lucas (1988). Lucas’ study amended the
19
neoclassical theory of growth by giving a vital role to human capital accumulation in
explaining the income diversities that appear to exist and persist between countries. His
study inspired numerous theoretical studies that place emphasis on the endogenous
accumulation of human capital as a source of long run growth, be it through formal
education and training or through informal learning-by-doing.
The “beneficial brain drain” and traditional “brain drain” studies considered in this
section are set within an overlapping generations (OLG) framework. Overlapping
generations models provide a dynamic framework for analyzing macroeconomic
phenomenon based on the micro-level behavior of individuals and households. The OLG
framework is ideal for looking at the global or macro consequences of the saving,
education, workforce and bequest decisions of individuals. Heterogeneity of agents arises
naturally from generational differences, since at any given point in time there will be
individuals who differ in terms of the stage they are in their life-cycles, although they may
be identical in every other respect such as preferences and endowments.
“Brain Gain Models” within an OLG Framework
In this section, the focus is on two recent studies that carry the analysis of the
earlier brain drain literature to a dynamic, overlapping generations setting with emphasis on
endogenous human capital accumulation as a source of long run growth. The theoretical
models developed by Haque and Kim (1994) and Wong and Yip (1999) are consistent with
the view that skilled migration will have a negative impact on economic growth in
developing countries. Each study also examines the impact of tax-financed education
subsidies under “human capital flight” and draws conclusions for education policy.
The set-up of each model is similar, taking place within an OLG setting in which
individuals live for two periods and derive utility from consumption in both periods.
Individuals decide on the optimal amount of time to spend on educational and labor market
activities in order to maximize their utilities across periods. Since neither saving behavior
nor intergenerational altruism is considered, all incomes are consumed in the period in
which they are earned.
In the first period of life, each individual spends a fraction of her non-leisure time
on education and the remainder working as an unskilled laborer. Education is an investment
in human capital that rewards the individual with greater income in the next period. In
20
Wong and Yip’s model individuals differ only in terms of the generation in which they are
born. The members of each generation are identical in that they follow the same life-cycle,
have the same endowments and make identical decisions with respect to the time they
spend on education and work. Individual-level decisions for each generation can therefore
be depicted by a representative individual. Heterogeneity of individuals, the breakdown into
the unskilled and skilled categories, occurs because there are two generations (young and
old) in any given period. The “young” represent the unskilled population and the “old”
represent the skilled worker population.
Haque and Kim (1994), on the other hand, differentiate individuals not only in
terms of their generation (young versus old), but also in terms of their latent abilities,
denoted by a. “More able” individuals will invest in greater amounts of schooling because
the pay-off in terms of productivity and income will be greater for them than it is for “less
able” individuals. As a result, at any given time there will be a continuum of heterogeneous
individuals with differing ability and productivity levels, rather than two separate, but
otherwise homogeneous categories of workers.
In addition to direct investments in education, human capital accumulation also
occurs through an intergenerational human capital externality, denoted by h. Each
generation inherits the human capital accumulated by the previous generation. Long run
increases in growth are therefore possible through this intergenerational transfer of
knowledge, which is based in part on the previous generations’ decisions to invest in
schooling. This explicit modeling of the human capital externality is what sets these models
apart from the previous literature.
Haque and Kim interpret the intergenerational human capital externality, ht, as the
average human capital level in the economy. The accumulation of human capital is a linear
process that is linked to individual-specific ability, the amount of time spent on education
and the economy-wide human capital externality. Wong and Yip, on the other hand,
interpret ht as the general knowledge level, which is modeled as a positive function of the
individual investment in schooling (knowledge gained through one’s own effort), the
number of educators in the economy (knowledge gained from the research of the educators)
and the previous period’s level of accumulated knowledge. Unlike the Haque-Kim study,
human capital accumulation is a concave function of the time spent on education. This
21
means that the increments to human capital decrease with the time spent accumulating it
through education.
In Wong and Yip (1999), there are two inputs into production, skilled and unskilled
labor, while Haque and Kim (1994) consider a single input, “effective labor”, which varies
across individuals depending on the human capital they inherit and the human capital they
accumulate from education. These models thus differ from the other models considered in
this section in that physical capital does not enter into the production function. This means
that any interactions and complementary effects between physical and human capital,
which may have great pertinence in explaining skilled wage differentials between
developing and developed countries, are necessarily ignored. In Haque and Kim’s model,
the most able are the ones who actually migrate abroad. There is an ability threshold where
individuals that have higher abilities than this threshold will migrate.
Both studies also examine the impact of government education and tax policies on
human capital accumulation and economic growth. Education (at all levels) is provided free
by the home country government. As a result, the only cost of education is the foregone
earnings from participating in the labor market. The government is assumed to keep a
balanced budget so that its expenditures are exactly offset by its revenues in each period.
Educational expenditures by the government are financed by an ad valorem tax on income.
Haque and Kim (1994) consider direct subsidies to individual incomes where the education
subsidy, denoted Et, grows in proportion to the average level of human capital in the
economy. Wong and Yip (1999) consider a situation where the government hires skilled
individuals as educators to provide free education to students. The number of educators at
time t is given by EDt.
The main findings of these studies in terms of education policy are as follows. The
Haque-Kim study shows that tax-financed subsidies to education are beneficial for
economic growth in a closed economy through their effect on human capital formation. The
subsidies bring down the cost of education and induce individuals to invest in more
schooling than they otherwise would have. In an open economy setting, however, some of
the investments in education—particularly at the higher levels of schooling—are not reaped
by the domestic economy because of human capital flight. Those with higher levels of
ability invest in higher levels of education because they expect greater returns on their
investment. Migration abroad is a selective process in which those with abilities greater
22
than some threshold level will actually migrate, and others with lower levels of ability will
remain behind. Given this, the source country loses both its most able (or productive)
human assets and does not get to collect on its investment at the higher education levels.
The policy conclusion reached under this scenario is that subsidies should be directed
toward lower levels of education in order to increase the human capital of those most likely
to remain behind to ensure growth even under a brain drain.
In the Wong-Yip study there are two categories of workers in each period—skilled
and unskilled. All individuals of a given generation are identical and as a result their
schooling decisions are identical. The only way for the number of unskilled (skilled)
workers to increase is for the next generation to decide to collectively invest in less (more)
education than the previous generation. This decision is based on the returns to education in
the labor market.
The “Beneficial Brain Drain” Models
Recent studies looking at the relationship between brain drain and economic
growth have challenged the conventional view that skilled migration inevitably leads to a
“brain drain” with the implied adverse consequences for the economy of the sending
country (Mountford, 1997; Stark, Helmenstein and Prskawetz, 1997 and 1998; Vidal, 1998;
Beine, Docquier and Rapoport, 2001). In these models, opening a developing country’s
economy to the possibility of migration increases the incentive to acquire skills. The
prospect of earning higher wages abroad leads to an overall increase in the investment in
skills by individuals in the domestic economy, which has positive consequences for
economic growth in the source country. Beine et al. (2001: 276) summarize the rationale of
these models:
In a poor economy with an inadequate growth potential, the return to
human capital is likely to be low and hence, would lead to limited incentive
to acquire education, which is the engine of growth. However, the world at
large does value education and hence, allowing migration to take place
from this economy would increase the educated fraction of its population.
Given that only a proportion of the educated residents would emigrate, it
could well be that in fine, the average level of education of the remaining
population would increase.
In these models, uncertainty plays an important role in establishing the positive
growth effect for the source economy. Mountford (1997) develops an open economy
23
overlapping generations (OLG) model in which a brain drain is shown to improve the
aggregate productivity level in the source country. His model follows the intuition of the
study by Miyagiwa (1991), which emphasizes the importance of scale economies in
education for attracting skilled migrants to locations with greater concentrations of skilled
workers. The greater concentration of skilled individuals in the host country is believed to
increase productivity levels by facilitating interactions, idea exchange, and collaborations
among skilled individuals. Using similar intuition, Mountford shows that the scale effect of
education can also work for the benefit of the source country, since the possibility of
migration leads to greater investment in education, and an increase in the number of skilled
individuals in the population. The growth externality created by the scale effect in
education is brought into the model by linking technology / productivity improvements to
the share of educated workers in the source economy in the previous period. The
“beneficial brain drain” (BBD) result is achieved when the relative wage differential
between the source and host countries is sufficiently high, and if the probability of
emigration is sufficiently low. In other words, a brain drain will have positive growth
effects for the source economy if a small possibility of emigration (e.g., only the very
highly skilled individuals leave) combined with high returns from migration induces a
sufficiently large number of people to invest in education in the source country.
The model presented by Stark, Helmenstein and Prskawetz (1997) is similar to the
Kwok-Leland (1982) model in that asymmetric information plays an important role in
explaining emigration and return decisions. However, asymmetric information in this
instance refers to the inability of the host country to discern the ability/skill levels of
incoming immigrants. When immigrants first arrive, they are offered a wage rate that is
based on the average productivity of the group of migrants. Individual skills and
productivities can only be identified after the migrants have spent some time working in the
host country. Once the true productivities of the migrants are discovered, the wages are
adjusted accordingly: the high-skill group receives a wage increase while the low-skill
group experiences a reduction in its wage.
Stark et al. (1997) proceed by characterizing the situation in which a “brain drain”
would occur, which is defined as high-skill workers remaining abroad and low-skill
workers returning home after the wage adjustments. Under the “brain drain” assumption, a
“brain gain” is possible for the source country only through the accumulation of human
24
capital by low-skill workers, since all the high-skill workers migrate permanently to the
host country. Consequently, a country that has a relatively high share of low-skill workers
stands to benefit from the emigration of its high-skill workforce. The possibility of
receiving higher wages abroad leads to human capital accumulation of the low-skill
workers, who return in the following period. Although Stark et al. (1997) provide a model
whereby a “brain gain” in possible for the source country, their analysis does not directly
look at the consequences for economic growth. However, the outcome in which the source
country ends up with a higher average level of human capital when “brain drain” is allowed
can easily provide the motivation for a human capital driven-model of endogenous
economic growth. In the following section some of the key features of the BBD models are
looked at in greater detail.
Closer look at Mountford (1997) and Beine et al.(2001)
Mountford (1997) and Beine et al. (2001) differentiate among individuals in terms
of their latent abilities. These studies are closely related to Miyagiwa’s model of scale
economies in education (Miyagiwa, 1991). Apart from differences in innate abilities,
individuals are assumed to be the same. This means they have the same preferences and
access to the same technologies. The latent ability of an individual is given by ai. The
ability parameter ranges between a0 and a1, and its distribution, f(a), is assumed to be
independent of the abilities of parents (Figure 6.1). Although the distribution of abilities in
the general population is depicted as following a normal distribution in the figure below,
studies often make the simplifying assumption of a uniform distribution in which each
ability category consists of an equal number of individuals. Mountford differs slightly from
Miyagiwa by looking at the share of skilled individuals in the population, s, rather than
their absolute number, Ls. The distribution is normalized such that
a1
a0
f (a )da = 1 and
0 < s < 1.
Mountford considers individuals who live for three periods. In the first period of
life, each individual decides whether to invest in education. Education has a fixed cost,
ceduc, which is the same for everyone. Since individuals do not have any private resources,
such as personal savings and family bequests, they must borrow from the capital market to
finance their investment in education at an interest rate r. Individuals can work only in the
second period of life and they must also repay their debts in this period. Consumption takes
place only in the third period when individuals retire. An individual with an ability level of
25
will invest in schooling only when the skilled wage rate, ws, is greater than the sum of the
unskilled wage rate, wu, and the total cost of education, ceduc(1 + r), in the second period:
ws( ) > wu + ceduc(1 + r)
The determination of the skilled wage and unskilled wage levels are similar to that
in Miyagiwa’s model. Individuals who possess a level of ability that is greater than some
threshold ability level, a*, will choose to invest in education. This threshold level is set by
the relative returns to education for each ability level in relation to the returns from
participating as an unskilled worker in the labor market in the second period of life.
The studies of Mountford (1997), Vidal (1998) and Beine, Docquier and Rappoport
(2001) reach qualitatively similar results. All share the idea of agglomeration economies in
which there is a productivity externality associated with the number of educated individuals
in the economy.
Policy Implications of the Beneficial Brain Drain Models
A crucial feature of these models is that the ability of an individual determines
whether he/she will devote any time to education, since higher levels of ability will provide
higher returns to education in terms of future income levels. In all models, wage (or more
generally welfare) differentials provide the motive for migration from the small, open
developing economy to the advanced economies.
The human capital-inducing effect of a positive probability of migration implies
that a policy of allowing migration outflows to take place will be beneficial from the
developing country’s perspective. If, however, as the models above show, there are no
restrictions to emigration, then everyone would leave with detrimental consequences for the
source country economy. The study by Stark and Wang (1999, 2002a) shows that given the
positive incentive to accumulate human capital under the possibility of migration, the
developing economy can find an optimal restrictive emigration policy that allows some
individuals to leave and others to remain behind with a greater amount of human capital,
which would not be possible under a strictly restrictive policy. They, in fact, argue that an
optimal emigration policy could even replace education subsidies as a way of inducing
further human capital accumulation.
26
2.4 Concluding Remarks
The “brain gain” models suggest that allowing skilled emigration to take place can
be “good” for the source country economy because it will increase the overall incentive to
invest in schooling. These models, however, do not consider the important role of
motivation on the productivity of individuals. The failure to emigrate will undoubtedly lead
to frustration among individuals if the value they saw in the extra education they received
(beyond that they would have chosen to achieve within a closed economy setting) was
merely as a means of leaving the country for “greener pastures”. Given the disappointment
in not reaching their goal of finding overseas employment, these individuals are unlikely to
be productive in their current jobs and are likely to engage in job search activities in order
to find the “next opportunity” for overseas employment. This will, of course, be at the
expense of giving full attention to their current jobs in their native countries. Thus, reaching
a higher educational attainment level by itself is not sufficient to guarantee higher
productivity levels and growth levels for developing countries.
In general, the brain gain studies focus on increases in the average schooling level
as a means of promoting economic growth, which will occur through an increase in the
private demand for schooling. It is presumed that this increased private demand can be met
adequately with the current level of resources and infrastructure available to the developing
country. The reality in developing countries is that educational resources and opportunities
are both limited and unequally dispersed over the population. Lack of private demand for
schooling is a mistaken presumption of these models. As the experiences of India and
Turkey clearly show, there is a very high demand for education, which the existing
education system is unable to satisfy. Overseas study helps to relieve this pressure,
although there is no guarantee that students will return once they complete their studies.
A serious omission of the “new growth” models considered is the lack of attention
given to demographic factors, which are important in a developing country context. For the
sake of simplicity, it is usually assumed that there is no population growth, and when there
are groups of workers differentiated by education or skill levels, they are assumed to be of
equal size. Connecting the schooling attainment of individuals directly with their abilities—
and nothing else—fails to recognize the significance of unequal opportunities in the
determination of who proceeds to the upper levels of the education system. Empirical
evidence strongly suggests that family wealth and parental schooling levels are significant
27
determinants of the level of schooling attainment of individuals (in addition to ability) (see
Ermisch and Francesconi, 2001; Tansel, 2002). The degree of intergenerational social
mobility has important implications for whether the poorer households and their
descendents are increasingly marginalized in the development process. An important issue
is to determine what the effect of a brain drain (the emigration of the most educated
segment of society) will be on social mobility and thus on the distribution of income
between wealthy and poor households. To address this, a model with more realistic
assumptions about demographic conditions and the behavior of households endowed with
differing initial income levels is required. Future research on this is warranted.
The brain drain models abstract from financial markets and thus do not consider the
possible effects of differences in saving behavior among different groups of households
(e.g. wealthy households vs. poor households). While the consumption, saving and bequest
decisions of households with different wealth endowments have been modeled within an
overlapping generations context, the OLG models that examine the consequences of brain
drain for LDC economies have so far sidestepped this important issue. Many studies focus
on the growth effects of migration and ignore distributional issues, which are as important
as efficiency considerations within a developing economy context.
Another important shortcoming of the models examined is the full employment
framework they use. Bhagwati and others have examined the possibility of unemployment
within a static context. The dynamic, business cycle effects on the brain drain are yet to be
studied in detail within an overlapping generations framework. Perhaps of broader
significance from the LDC perspective is the real effects of financial crises brought about
by adherence to a strict program of liberalization advocated by international agencies. The
hasty liberalization of capital markets in some developing countries, for example, has
increased their vulnerability to global economic fluctuations. The economic crises of recent
experience have affected not only the unskilled labor force, but skilled workers as well. It
may be said that the instability of liberalizing economies, with frequent episodes of
“financial crises”, leads to great uncertainties with respect to production and employment
within these economies. This, in turn, sets the broader macroeconomic context to which
skilled migrants respond.
28
The next chapter examines the economic theories of skilled migration that take
human capital theory as their basis, and which aim to explain why a wage differential exists
between the sending and receiving countries.
29
CHAPTER 3
THEORETICAL MODELS OF THE BRAIN DRAIN
3.1. Introduction
While much theoretical work has been done to model the effects of skilled
migration on the economies of both the source and host countries, theoretical models of the
brain drain are comparatively fewer. Chapter Two provided a synopsis of the theoretical
literature on the effects of the brain drain. The present chapter turns to theoretical models
that attempt to explain why skilled migration occurs. In economic models of the brain
drain, income differentials between the receiving and sending regions provide the main
motivation for aggregate skilled labor movements. One explanation for this wage
differential focuses on the complementarity between physical capital and human capital,
and on differences in the physical capital stock levels in the host and source countries.
Complementarity implies that skilled workers will be more productive and thus receive
higher pay in locations that are more abundant in physical capital. This promise of a higher
wage level, in turn, results in the migration of skilled workers to more developed countries
and regions. The implication for policy is relatively simple: developing countries can attract
and keep skilled individuals by augmenting their physical capital bases through physical
capital accumulation.
Other explanations for the wage differential between developed and developing
countries may be found in the more recent skilled migration or “brain drain” literature. The
initial focus of the chapter is on economic theories of skilled migration in which wage
differentials play a prominent role in the decision to emigrate. These theories are based on
the human capital approach, which is presented in section 3.2. Section 3.3 summarizes the
economic theories based on this approach that aim to explain why skilled individuals
choose to migrate or fail to return to their home countries after a period of study abroad.
30
The chapter ends with a brief look at some alternative theories of migration that may also
have pertinence for skilled migrants.
3.2 Human Capital Theory of Migration
In many economic theories of internal and international migration, the decision to
migrate from one location to another is believed to be made on the basis of whether the
move will bring net economic benefits to the potential emigrant. Formal models formulate
the net economic gain from migration in terms of the difference between the present values
of the income streams from working in the destination location compared to the original
location. Relocation and “psychic” costs such as the cost of adjusting to a new environment
and being away from family and friends are subtracted from the wage differential to arrive
at the net gain. Specifically, the net gain from international relocation may be written as:
T
Net Gain from Migrating =
wtF − wtH
−C
t −1
t =1 (1 + r )
(3.1)
where wF and wH are the wages for a given skill level in the host country, denoted by F, and
the source country, denoted by H. The rate at which the future is discounted is given by r,
C represents the total of monetary expenses (e.g., travel and relocation costs) and nonmonetary “psychic” costs of migration, and T is the period of retirement. Migration takes
place only when this net gain is positive (Sjaastad, 1962).
This view of the individual as a rational decision-maker is also called the “human
capital” approach since each individual decides on the best location—the location that will
bring the highest returns—given her investment in education, health and skills.
3.3. Theoretical Models of the Brain Drain based on Human Capital Theory
Early studies conducted in the 1960s and 1970s focused on the consequences of
brain drain for the countries involved. The analysis of skilled migration outcomes was
based on the view that brain drain is the response of skilled individuals to wage
differentials in different locations, and that these differentials are the result of productivity
differences between countries. Productivity differentials, in turn, arose from the differences
in the physical capital stock base of the sending and receiving countries. The relative
31
abundance of physical capital in advanced countries increased the productivity of skilled
labor because of the assumption of complementarity between skilled labor and physical
capital. The current section provides a detailed review of various studies that give
alternative explanations of how the wage differential, the primary motivator for “brain
drain”, occurs. In these models, the decision to emigrate is based on a comparison of the
wage offered in the destination country (wF ) to the wage offered in the country of origin
(wH). The superscripts H and F denote the home (source) and foreign (destination) countries
respectively. Depending on the exposition, the decision rule for emigration may take either
the form
k wF
wH
0<k<1
(3.1a)
where k is a discount factor applied to the foreign wage to reflect lifestyle and cultural
preferences, or the following form
wF − cmig
wH
(3.1b)
where cmig is the initial cost of migration that includes both monetary and “psychic” of
moving. A common feature of the studies is the positive link between the productivity of
workers and the wages offered by firms. The studies differ mainly in explaining how the
productivity differences occur and how they are reflected in the wage level. Table 3.1
summarizes the emigration decision rule for each model as a guide to the detailed analysis
of each model provided in subsequent sections.
Section 3.2.1 presents the Kwok-Leland (1982) model in which wage differentials
are based not on country differences in physical capital, but on individual differences in
talent or ability. In their study, the phenomenon of student non-return is explained by
information asymmetry between host and source country employers concerning the true
“talent” of students studying abroad. The informational advantage of host country
employers allows them to offer students wages that are commensurate with their skills,
while source country employers can only offer a wage that equals the average productivity
of returning students. This wage gap results in the best students remaining abroad and the
less productive students returning home.
32
Table 3.1. Theoretical Models of the Brain Drain
Study
Emigration Decision
wF
wH
Kwok-Leland
(1982)
k wF(ai)
MP(ai)
APR(aR)
Miyagiwa (1991)
wF(ai, LsF) − cmig
wH(ai, LsH)
h(LsF) ai
h(LsH) ai
Chen-Su (1995)
wF(aFi, KF) − cmig
wH(aHi, KH)
Epstein (2002)
wF(Mt) − cmig(Mt-1)
wH(Mt)
wF
wH(aR)
KFai(ae, KF)
wF(Mt)
KHai(ae, KH)
wH(Mt)
= wage offered by employers in foreign country
H
w
= wage offered by employers in the home country
k
= a fraction reflecting possible disutility from working outside home country
ai
= ability (skill) level possessed by individual i
aR
= average ability (skill) level of returning individuals
MP
AP
R
= marginal productivity of workers
= average productivity of students who have returned to work in the home
country
cmig
= monetary and non-monetary “psychic” costs associated with emigration
LSF
= number of skilled individuals in the foreign country
LS
H
= number of skilled individuals in the home country
h(·)
= positive externality from the agglomeration of skilled individuals, h (Ls) > 0.
ae
= on-the-job skill accumulation by individual through own effort
K
M
= capital-dependent on-the-job skill accumulation
= number of migrants in new location
Positive externalities from the agglomeration of human capital in the host country
provides a further explanation for the existence of wage differentials. Section 3.2.2
examines the model proposed by Miyagiwa (1991) in which increasing returns to scale in
higher education is given as a cause of brain drain. Chen and Su (1995) extend the “human
capital agglomeration model” in an attempt to explain student non-return based on the
argument that on-the-job training received abroad, after completion of studies in the foreign
country, increases the productivity of individuals working abroad and amplifies wage
differentials between the foreign and domestic countries. Section 3.2.3 provides details of
Chen and Su’s model of on-the-job training as a cause of brain drain. Wong (1995)
33
incorporates learning-by-doing into a model of brain drain. His model is considered in
section 3.2.4.
In section 3.2.5, network externalities are considered as a possible cause of brain
drain. The migration chain model underlines the importance of migration networks in
perpetuating subsequent migration. In migration chain models, while migrants still respond
to wage differentials in different locations, the positive externalities created by migrant
networks in a particular location may be the deciding factor in choosing a migration
destination. Helmenstein and Yegorov (2000) model the dynamics of migration flows
within a stochastic two-country framework comprising the host and source countries in
which such “chain effects” are important. Section 3.2.5 also looks at “herd” models of
migration, which provide another explanation of why “ethnic” or “migrant clustering” may
occur. Some non-economic considerations, such as psychological factors and foreign
language instruction, are discussed briefly in section 3.2.6. A common feature of the
models presented in the latter sections of Chapter 3 is their focus on the endogeneity of
emigration cost rather than differences in wage and productivity levels in the host and
source countries.
3.2.1. Information Asymmetry as a Cause of Brain Drain
One explanation of the brain drain, which focuses on student non-return, relies on
the assumption of asymmetric information on the part of employers in the host and source
labor markets1. The Kwok-Leland model (1982) was constructed to explain why many
Taiwanese students have chosen not to return to Taiwan after finishing their studies in the
United States, given that labor markets in Taiwan appear to be competitive in terms of
employment opportunities and income levels. Unlike the traditional wage differential
explanation, with its emphasis on physical capital differences in the source and host
countries, the Kwok-Leland model highlights differences in individual talent or skills as
measured by the productivity of individuals. The brain drain occurs because individual
differences in ability are best assessed by employers of the host country, who are then able
to give the appropriate compensation for the level of productivity they observe.
1
This explanation of the brain drain draws on the work of Akerlof (1970) who theorized that
imperfect information plays an important role in market outcomes.
34
Kwok and Leland (1982) hypothesize that host country employers have greater
knowledge about the true skills of the students who study in their country than source
country employers. The informational advantage of host country employers is the result of
several factors. Host country employers have: 1) greater familiarity with the academic
system of their own country and the output of this system; 2) more experience in hiring
both domestic and foreign graduates from universities in their country; and 3) a system of
hiring that makes use of in-depth interviews that allows them to gain further information
about job candidates. This information allows host country employers to offer wages to
foreign students that reflect their true productivities, whereas source country employers,
lacking this knowledge, can only offer a wage rate that reflects the average productivity of
returning students.
The following equation represents the general condition for returning. It states that
in order for students to return home after their studies are completed, their home country
must offer a wage rate, wH, that is greater than or equal to a some fraction (k) of the wage
rate offered by host country employers, wF.
wH
k wF
(3.2)
The fraction k reflects the tendency for individuals to have a stronger preference for
working in their home countries. In the Kwok-Leland model, the wage offered by the host
country is equal to the true productivity of workers. In other words, an individual i with
ability ai will receive a wage of wiF = MPi = MP(ai) in the host country, where MP denotes
marginal productivity. Ability ranges from a0 to a1 (a
[a0, a1]) and is distributed over the
population according to the distribution function f(a). There is a continuum of
productivities associated with continuum of ability levels. In the host or foreign country,
workers are offered a range of wages equal to their abilities and productivity levels. The
source or home country, on the other hand, is unable to differentiate between more
productive and less productive workers among the returning students and offers each
returning student (based on previous experience with returnees) a wage, wH, that is equal to
the average productivity (AP) of all returning students, APR, where R indicates “returning”
students. In other words, due to the information asymmetry, all students are offered the
same wage rate by the home country regardless of their ability level. The general return
condition (3.2) may be rewritten in terms of these assumptions for individual i as follows
35
R
AP =
R
a ⋅ f ( a) da
R
where
R
f (a)da
k MP(ai) (= k wF )
(3.3)
integrates over the set of productivities associated with returning students. In the
average productivity expression above, the numerator equals the total productivity of the
returning students, and the denominator shows the total number of returning students. A
student with an ability or productivity level that is greater than the average productivity of
returning students will choose to stay in the host country since she will receive a higher
wage. Conversely, students with productivities that are lower than the average will choose
to return because they will earn a higher wage in their home country. Therefore, an
important consequence of this model is that “bright” students will choose not to return
while the “mediocre” students will choose to2.
One of the implications of the Kwok-Leland model of information asymmetry is
that brain drain can occur even when students prefer to work in their home country and
income differentials at home are favorable for given productivity levels. The problem is
that firms in the country of origin cannot assess individual productivities effectively. The
Kwok-Leland framework of asymmetric information for explaining the migration of skilled
workers has been criticized on the grounds that employers with imperfect information will
eventually learn the true productivities of returning students (Chen and Su, 1995). The
information asymmetry, therefore, should not be expected to persist in subsequent periods.
Lien (1987a) extends the Kwok-Leland model by introducing the possibility of
signaling. Although source country employers may not have information about the true
abilities or productivities of returning students, quality signals such as the ranking of the
university from which the student graduates may give them an idea about the abilities of
returning students. In Lien (1987b), the migration of skilled individuals is modeled as the
outcome of a multi-stage decision process. The stages considered are as follows: 1)
Students in developing countries decide whether to go abroad and pursue advanced level
2
Katz and Stark (1984) show using the Kwok-Leland model that it is also possible for less skilled
workers to emigrate when the information asymmetry works in the opposite direction. When the host
country employers have less information about the true skills of the immigrants to their country they
will offer wages based on the average productivity of these immigrants. Thus, students with higherthan-average productivities will choose to stay in the home country because they will be offered a
better wage while students with lower-than-average productivity levels will choose to emigrate.
36
studies; 2) students who go abroad to study and successfully complete the doctoral
program, must decide whether to immediately return to their home countries or whether to
work abroad for a period; and finally 3) those who decide to take jobs in their country of
study, must decide whether to continue working abroad or return home. As in the previous
models, students are differentiated by ability and the asymmetric information setup in favor
of employers in the foreign country is kept in this multi-stage decision process. Thus,
Lien’s study offers a greater degree of sophistication in modeling information asymmetries
while, at the same time, maintaining the main outcomes of the Kwok-Leland model.
3.2.2. Increasing Returns to Scale in Advanced Education
Recent studies have emphasized the importance of education and skills for
economic growth and development. Some of these studies place particular importance on
the positive externalities from advanced education (see, for example, Jaffe, 1989). The idea
is that a greater concentration of individuals with advanced degrees within a geographical
area increases the productivity of similar individuals in the area and can lead to significant
spillover effects in surrounding regions as well. The marginal productivity increase is the
result of the harmonizing of knowledge that is endowed separately in each individual with
the knowledge of others in the group. Physical proximity increases the sharing of ideas and
induces greater cooperation and collaboration on projects. This is the positive, scale effect
of education on productivity and aggregate output. Miyagiwa (1991) formally introduces
the scale effect in advanced education into a model of the brain drain. In his model,
increasing returns to advanced education is given as an explanation of skilled migration.
The greater concentration of skilled individuals in developed areas tend to attract skilled
individuals from developing areas because of the positive scale effect of advanced
education, which increases the productivity and incomes of the skilled individuals.
Miyagiwa’s model shares some similarities with the asymmetric information
models of section 3.2.1. Individuals are heterogeneous in that they differ in their
endowment of “ability” or “talent”. Ability is denoted by a and can take on any value in the
interval a0 and a1. Whereas the Kwok-Leland model looks at the migration decisions of
students holding advanced overseas degrees, Miyagiwa also models the education decision
of individuals. If individuals choose not to invest in advanced education, they work as
unskilled workers and receive a wage, wu, which is the same for all unskilled individuals
37
regardless of their level of ability. Those who invest in education work as skilled workers
and earn a wage that equals their productivity. Since more “talented” individuals are more
productive, they receive a higher wage than less “able” or “talented” individuals. The return
to higher education is given by the following relationship which links the productivity of
individuals with their ability and the scale effect in advanced education. The wage of an
educated individual with ability level
is
ws( ) = MP( ) = h(Ls)·
(3.4)
where ws refers to the wage received by skilled individuals (those who invest in advanced
education), Ls is the number of skilled workers in the economy, and h(.) represents the
positive externality from the agglomeration of skilled individuals, such that
h (s) = h(Ls)/ Ls > 0.
Each individual initially decides whether to invest in advanced education by
comparing the net returns from receiving advanced education to the returns from working
as an unskilled worker. Advanced education has a fixed cost3, ceduc, which is the same for
all individuals regardless of ability. Therefore, an individual with ability
will choose to
invest in advanced education only if ws( ) - ceduc > wu. Given this condition, there is an
ability level, a*, for which the net return to advanced education is equal to the unskilled
wage rate. Those with an ability level that is higher than this threshold ability level will
choose to invest in advanced education. The total labor force, L, is the sum of skilled
workers, Ls =
a1
a*
f (a )da and unskilled workers, Lu =
a*
a0
f (a )da , as determined implicitly
by the threshold ability level a*.
After the education investment decision is made, individuals acquiring advanced
education must decide whether to work in their home country or to emigrate and work in a
foreign country. In the model, the source (home) and host (foreign) countries are similar in
many respects, including the distribution of ability in the general population. The main
difference between the two countries is assumed to be population size; the host country has
a greater population than the source country (NF > NH). This size difference, given identical
distributions in ability, implies that a greater number of individuals in the workforce will
3
This cost probably refers to the direct, out-of-pocket expenses such as tuition, books, travel and
board.
38
receive advanced education in the host country (LsF > LsH). The scale effect of advanced
education, therefore, works in favor of the host country [h(LsF) > h(LsH) since h (Ls) > 0]. As
a result, the return to advanced education is greater at each ability level in the host country:
wF = h(LsF)·a > wH = h(LsH)·a. This, in turn, means that the threshold ability level for
receiving advanced education is also lower (aF* < aH*), which serves to reinforce the scale
effect. That is, the greater return to advanced education induces previously uneducated
individuals to invest in advanced education.
The host and source countries in Miyagiwa’s study are the United States and
Taiwan respectively, which makes the difference in population size a valid assumption. In
the case of India or China, however, two countries with very large populations, the current
model would predict wrongly that these countries would attract rather than lose individuals
with advanced education through emigration. To make the model work for countries with
relatively large populations and substantial skilled migration, an explanation based on
differences in the cost of education or the returns to education is suggested (Miyagiwa,
1991: footnote 14, p. 748). If the cost of education, ceduc, in the country of origin is
sufficiently high to increase the minimum (threshold) ability level for investing in
education, then the number of skilled workers will be lower compared to the destination
country even if population size is greater.
The “cost of education” explanation maintains the assumption that the education
decision is determined by ability. This implies that the “most able” will also be the ones
who can afford to invest in education. This is obviously unrealistic if different income
groups have the same ability distribution as the population at large (notwithstanding the
possible existence of ‘free’ public education, which is generally inadequate in reaching
targeted groups in both developed and developing countries). Empirical studies have shown
that household income and parental education levels are important in determining the
sorting of individuals into educational classes. Lower household income levels and lower
levels of parental education affect educational attainment negatively. If an “unequal
opportunities in education” perspective is adopted, Miyagiwa’s model can more easily be
reconciled in terms of the Chinese or Indian experiences.
To summarize, in Miyagiwa’s model greater population size in the host country
produces a greater number of individuals with advanced education. The scale advantage in
advanced education results in a wage differential between the host and source countries in
39
favor of the host country. This wage differential motivates skilled emigration abroad. An
individual will migrate if the wage offered in the foreign country net of the cost of
migration is greater than the wage offered in the home country. For an individual with an
ability level of , the emigration decision may be written
wF( ) − cmig
h(LsF)· − cmig
wH( )
h(LsH)·
(3.5)
where cmig is the cost of migration, assumed to be the same for every individual.
The focus of Miyagiwa’s model is on the emigration of individuals who receive
their advanced education in the home country, whereas the Kwok-Leland model is a model
of student non-return. The effect of scale economies in education on migration has been
incorporated into a model of economic growth by Mountford (1997), who shows the
possibility that brain drain can have positive consequences for the source economy through
its effect on the average level of human capital formation. Mountford’s model, which is
very similar to Miyagiwa’s, was examined in the latter part of Chapter Two.
3.2.3. On-the-Job Training as a Cause of Brain Drain
Chen and Su (1995) offer an explanation of student non-return based on the
argument that on-the-job training received abroad complements the education completed in
the foreign country and increases the productivity of individuals with advanced overseas
degrees working in the foreign country. This, in turn, magnifies the wage differentials
between the foreign and domestic countries, and increases the opportunity cost of returning
to the home country. In their model, the motivation behind the decision to stay is, again, the
promise of a greater future income stream in the foreign country arising from higher wage
levels4.
4
The wage differential, in the form of expected income streams over the period of the student’s work
life, may be expressed as follows:
T
0
w F e − rt dt −
T
0
w H e − rt dt = ( w F − w H ) (1 − e − rT ) r > 0 , where r is
the discount rate and t = 0, … , T is the work horizon facing the student. The discussion of the Chen
and Su model is simplified by ignoring the time dimension; this simplification does not lead to any
loss in the implications or understanding of the model.
40
In the traditional analysis, the marginal productivity of skills, whether obtained
through the education system or on-the-job training, depends on physical capital differences
between the home and source countries. The complementarity between the stock of
physical capital and skills implies that the marginal productivity and wages of skilled
individuals will be greater in the country with a greater physical capital base. Chen and Su
also incorporate this idea into their model, except that they refer to a broader social stock of
capital, which is the sum of both physical and human capital. The wage received by an
individual completing advanced studies is dependent on three factors: the social stock of
capital (K), the chosen profession of the individual, and the individual’s skill level (a). The
expected wage levels for an individual with ability
in the foreign and home countries are
expressed in multiplicative fashion as wF = KF and wH = KH , respectively, where
is
a positive parameter that varies with profession. By assumption, the social stock of capital
is greater in the foreign country than in the home country (KF > KH). This is the only source
of difference between the source and host countries. The greater stock of social capital
increases the returns to skills because of the complementarity between capital and skills.
Thus, the emigration decision for an individual with ability
wF( ) − cmig
KF − cmig
is the following:
wH( )
KH
(3.6)
Equating the above condition and solving for a gives the threshold level of ability
(skills) for emigration to take place: a* = cmig / (KF − KH) . If the level of skills acquired by
the student at the end of her studies is less than a*, then she will return. The ability
parameter has a distribution f(a) and a cumulative distribution F(a) = f(a)da = 1. The
probability of non-return (stay), then, may be expressed as Prob(Stay) = 1 − F(a*). It is easy
to show that Prob(Stay)/ K* > 0 and Prob(Stay)/ K < 0.5 This indicates that an increase
in the social capital of the foreign country increases the probability of non-return, while an
increase in the social capital of the home country lowers this probability.
The capital stock argument by itself, however, does not explain why student nonreturn is a more prevalent form of brain drain. To explain this, Chen and Su decompose the
5
Prob(Stay)/ KF = - ( F/ a*)( a*/ KF) = - f(a*)(-cmig / (KF - KH)) > 0 and
Prob(Stay)/ KH = - ( F/ a*)( a*/ KH) = - f(a*)(cmig / (KF - KH)) < 0
41
skills acquired through on-the-job training after graduation into capital-dependent and noncapital-dependent components. Students will possess a skill level of a0 after completing
their formal education. They are able to increase their skills beyond this base level when
they enter the workforce and receive on-the-job training. The maximum amount of skills
that an individual with advanced schooling can accumulate through on-the-job training is
the sum of the skills that can be obtained by the individual’s own effort (ae) and the skills
that are dependent on the existing social capital in the country of work. The maximum skill
levels that can be achieved through on-the-job training in the foreign and domestic
countries are given by
aFmax = ae + KF
and
aHmax = ae + KH
>0
where ae is the non-capital-dependent component of skill accumulation and KF ( KH)
represents the capital-dependent component. Chen and Su show that the probability of
staying in the foreign country increases with the relative importance of the capitaldependent component (i.e., as
increases), given that KF is greater than KH.
Chen and Su argue that to the extent that education received formally complements
training, this complementarity will be greater for individuals who receive training in the
same country as they receive their advanced education. Accordingly, the marginal
productivity of on-the-job training received in foreign firms is greater for those educated in
the foreign country than for those educated in their native countries. Those who receive
advanced foreign degrees and stay on to receive on-the-job training, therefore, have a lower
incentive for returning to their native countries. Some implications arising from this model
are 1) superior students stay in the foreign country while inferior students return, as was the
case in the Kwok-Leland model; 2) the number of returning overseas students will be lower
in disciplines where on-the-job training is important in gaining specialized skills.
42
3.2.4. Incorporating Learning-by-Doing into a Theory of Brain Drain
In Arrow’s classic article (1962), knowledge acquired through learning is the
product of experience, which is termed “learning by doing”. Empirical studies of the
determinants of aggregate production have shown that only a small part of total output
production can be explained by capital and labor inputs alone6, and a very large part can be
ascribed to an undetermined residual, which has been called “technological advance”. One
of the contributions of Arrow’s paper is to describe how an endogenous theory of technical
change or the advance of knowledge based on learning-by-doing (experience through
production) can be incorporated into an aggregate model of economic growth. Since
experience is gained by producing, learning-by-doing is constructed as a function of the
total or cumulative output produced7.
Wong (1995) incorporates the learning-by-doing framework into an analysis of
labor migration. He constructs a two-period overlapping generations model to explain how
young workers decide on whether to stay in the home country or emigrate. In the model,
Wong defines brain drain as “working at home when young and working abroad when old”,
which he distinguishes from “permanent immigration” or “working abroad when both
young and old”. To simplify the analysis, it is assumed that wage rates are stationary over
time. In other words, a wage differential in the first period will continue in the second
period. Wong’s model does not explicitly explain why the wage differential exists.
The main results of Wong’s model are the following. In the initial period, if both
the foreign wage and the foreign output levels are greater than the domestic levels of wages
and output, a worker has a double incentive to emigrate: one stemming from the wage
differential in favor of the foreign country, the other because the worker will gain greater
work experience (implied by the higher output level) than he would in the domestic
country. The greater work experience increases the worker’s productivity and the wage she
will earn in the next period. Accordingly, the worker will emigrate in the first period and
remain in the foreign country in the second period since wage levels remain higher than the
domestic levels. In the case where foreign wages are higher but foreign output is lower than
the domestic levels in the initial period, the worker will choose to work in the domestic
6
The most famous study is the pioneer work of Solow (1957) who estimated an aggregate
production function for the United States.
7
Arrow (1962) uses cumulative gross investment as an index of experience.
43
country in order to increase her marginal productivity by gaining work experience. This
experience she gains in the first period allows her to earn a greater wage in the next period
based on her experience. She will choose to emigrate in the second period because foreign
wage levels will be higher.
3.2.5. Migration Chains and Herd Effects
In the chain model of migration, migrants are not a homogeneous group, but differ
based on the time of their arrival to the destination country. Two groups, “single” and
“chain”8 migrants, are defined to distinguish between migrants who arrive initially and
migrants who arrive later. The initial or single migrants are motivated to migrate mainly as
a result of ‘push’ factors originating from their home country (such as poor economic or
social conditions), while chain migrants migrate because of the ‘pull’ of fellow countrymen
in the foreign country.
Chain migrants enjoy certain advantages that single migrants do not.
These
advantages include information exchanges with the settled population in the destination
country about labor market conditions, housing and other relevant information. They may
also be able to keep lodging costs down by sharing accommodations with their fellow
nationals. These network benefits reduce the overall cost of migration for the chain migrant,
with the implication that chain migrants will choose a settlement location not only on the
basis of the wages offered in the location, but also on the existence of a supportive network
(Helmenstein and Yegorov, 2000).
Helmenstein and Yegorov (2000) use the chain migration concept to construct a
stochastic dynamic model that explains how migration from a source to a foreign country
may accelerate following a small initial inflow of immigrants. While single migrants start a
chain reaction of subsequent migration, it is found that the volume of the chain migration is
sensitive to the phase of the business cycle in the host country. It is assumed that each
migrant residing in the host country exerts some constant capacity to “pull” a certain
number of migrants per period. In recessionary periods, since the initial inflow of migrants
is small, the capacity to “pull” new migrants is lesser, which serves to dampen the
multiplier effect on chain migration. The wage elasticity of the demand for labor, which is
8
The chain migration concept is based on the earlier studies by MacDonald and MacDonald (1964)
and Gurak and Caces (1992).
44
higher during periods of recession, is therefore important in determining the outcomes of
the Helmenstein-Yegorov model. However, they make no distinction between skilled and
unskilled migrants and their model ignores the fact that in general a lower elasticity of
demand exists for skilled labor.
Other studies that look at migration networks include Bauer, Epstein and Gang
(2000, 2002a). Their work has been important in showing that network-externality effects
are not linear, but follow an inverse U shape: increasing initially then declining. In
explaining Mexican migration to the United States, Bauer et al. use both aggregate (share
of total Mexican community) and village-specific (share of village-specific Mexican
community) measures of migration networks to explain how migrants choose migration
destinations. Their study, based on data from the Mexican Migration Project, shows that an
inverse U shape exists for the effect of the share of the Mexican community in the US
location on the probability of choosing that location. This means that initially immigrants
are attracted to locations with a large Mexican community, but as the size of this
community increases the probability of choosing the location declines. Wage decreases
from increased competition and native population objections to a large ethnic community
have been given as possible explanations for this outcome.
Bauer et al. (2002a) also believe that village networks / ties may be important in
choosing a migration destination. In other words, migrants coming from a certain locality /
village in Mexico would choose locations in the United States in which a high
concentration of their fellow villagers existed. They show empirically that the effect of
village networks on location decisions also follows an inverse U shape, although this effect
is less pronounced than for the total Mexican community share in US destination variable.
In the same study, an alternative explanation of “immigrant clustering” is given:
that of “herd behavior” based on Banerjee (1992) and Epstein (2002). The assumption
behind the theory of “herd behavior” is that although some locations offer better conditions
than other locations, the knowledge about which location is the best is limited or unknown
to the potential emigrant. The emigrant has imperfect private information about each
destination. Based on this private information, the emigrant may feel that some locations
are better than other locations. On the other hand, the potential emigrant may observe that
many people with similar attributes to him/her have been choosing a location that had not
seemed to have been the best location among the alternative locations according to the
45
private information. An emigrant that follows “herd behavior” disregards the private
information and chooses the location that everyone else is choosing based on the belief that
his/her private information is incomplete and that others must have better information.
Bauer et al. (2002a) have tested the herd theory empirically using the same Mexican
dataset9 and have shown that herd effects also exert a significant influence on the location
decision of Mexican migrants.
While the migration externalities (chain) and herd models of migration do not
distinguish between different types of migrants (e.g., skilled vs. unskilled migrants), the
reasons they put forth in explaining migration in general based on ethnic or cultural
networks or links may also be useful in motivating the causes behind skilled migration.
The theory of network externalities may be adapted to the student and professional
migrants’ situations. The technological advances in communications, especially within the
past decade, have made internet networks possible. Groups with similar interests come
together in these networks to share information and solve problems. There are both general
and discipline-specific alumni networks that bring together university graduates. The
existence of geographic-based alumni networks for universities such as Middle East
Technical University and stanbul Technical University work in the same way as the
migrant networks discussed above in creating externalities for those living abroad and for
potential migrants that take part in these networks. Joining the network provides many
benefits to the participant. Some examples include the sharing of information on visarelated issues, foreign job openings, choosing the best university or least costly location for
study abroad. The existence of fellow countrymen helps lower “psychic” costs in a
particular location and facilitates the process of adaptation to a new environment.
3.4 Other Considerations
In this section, non-economic factors that affect skilled migration are lumped
together into the category labeled “other considerations”. These considerations often refer
to psychological, social or institutional factors that either ease or impede the transition to a
new culture or society and affect the decision to migrate.
9
In their sudy, network effects are captured by the stocks of migrants from the same country in a
particular migration location at the time of the migration decision, while herd effects are proxied by
the flows of emigrants to a location in the year before a person migrates.
46
The Effects of Foreign Language Instruction.
Foreign language instruction in schools in Turkey has been suggested as an
important catalyst for the brain drain (Kaya, 2002). Since language acquisition is more
difficult later in life, students exposed to a foreign language early on in their education will
experience less difficulty in adjusting to a foreign-language environment. This has the
effect of lowering the non-pecuniary cost of migration. Language ability will also improve
the chances of being accepted as an immigrant in the host country and increase the potential
earnings of accepted immigrants.
Psychological and Sociological Factors.
Many of the previous models have emphasized the economic aspects of the
decision to migrate. Psychological factors may be as important as or more important than
economic factors in some cases. In sociological studies of the brain drain, for example, it
has been shown that the degree of “normlessness”, “powerlessness” and “anomie” felt by
individuals can be important psychological factors determining whether individuals return
to their home countries. These psychological attributes are discussed in Hekmati (1973:
27). Powerlessness and normlessness are described as two attributes of “alienation”.
Powerlessness refers to the lack of control or mastery that an individual may feel over
political and social events, while normlessness refers to the “expectancy of the necessity of
deviant behavior in attaining of economic and political goals”. Anomie, on the other hand,
refers to “an individual’s perception of his society and his place in it”. High levels of
anomie and alienation felt in the home country by migrants may partially explain why they
do not return. There is evidence that psychological factors are important in adapting to a
new environment.
Constraints to individual decision-making: the family and social context.
Another micro-level approach, dubbed the “new economics of migration”, extends
the view of the rational individual by placing the micro decision-making process within a
broader social or “family” context (see Stark and Bloom, 1985; Vogler and Rotte, 2000).
The motive for individual migration becomes more than the desire to increase personal
welfare, but the desire to improve the well-being of the household to which the individual
belongs by reducing the labor market risks for the emigrant’s family as a whole. For
example, individual migration may be supported financially by the emigrant’s family, and
the emigrant, in turn, is expected to support his household by sending back remittances.
47
There is empirical evidence in support for a “family investment model” of migration, in
which families pool risks by diversifying their labor across borders.
Another focus of the new economics of migration is on the relative income position
of an emigrant in terms of a reference group in the host country. In this case absolute
income differentials are irrelevant, and individuals or households make their decisions
based on their “relative deprivation” levels. As the empirical study on Turkish students and
professionals presented in the next two chapters will show, families and social networks
consisting of friends and acquaintances are important influences on the decisions to study
and to work abroad.
The role of institutions: screening.
Commander, Kangasniemi and Winters (2002) emphasize the screening element in
the policies that take place at the national level and at the firm level. In the United States,
for example, preferential visas are issued to potential migrants working in priority areas.
Many countries also have or are in the process of adopting similar migration policies. The
issuing of work permits in the US requires sponsorship from a firm. Above average living
and working conditions make advanced countries attractive locations for both skilled and
unskilled individuals. This allows developed countries to use immigration restrictions as a
policy tool to select immigrants based on their qualities. The use of selective immigration
policies in favor of skilled migrants has become an increasingly important strategy both in
the United States and in other Western countries as a way to meet the growing demand for
skilled workers. The need for technology workers has intensified over the years in
knowledge-based countries, such as the United States. Much of this demand is concentrated
in “industrial districts” or “competence blocs”10, such as the Silicon Valley in California.
The growing reliance of U.S. high-technology firms on skilled foreign workers is
corroborated by the introduction of the “Brain Act” in the U.S. Congress in August 199911.
This bill was introduced in order to extend 5-year work visas to foreigners who are recent
10
The term “industrial district” was used by Marshall to denote a geographic area in which the
activities of actors within this region as a whole bring about increases in total factor productivity,
whereas the activities of individuals or single firms alone suffer from diminishing returns. Thus, the
existence of a network of firms functioning together creates positive externalities for the individual
firm. Eliasson has developed this idea by introducing “competence bloc” theory. A competence bloc
is “the configuration of actors that together initiates and stimulates the growth of an industry”
(Eliasson, 2000: 220).
11
Information accessed from the website: http://www.techlawjournal.com/cong106/h1b/Default.htm.
48
graduates from U.S. universities in the fields of science, math, engineering, and computer
science, in order to compensate for shortages in high technology manpower in these areas.
3.6. Concluding Remarks
Various theories have been proposed to explain the phenomenon of “brain drain”.
These theories center on the assumption that wage differentials provide the primary
motivation for the migration of skilled workers to more developed countries. Higher wages
abroad attract educated individuals from all over the world much like higher rates of return
to physical capital (higher interest rates) attract inflows of capital. The traditional
(neoclassical) explanation for these wage differentials is the existence of differences in the
stocks of physical capital between the source and host countries. Since developed countries
have higher physical capital stocks than developing countries, the productivity and wages
of skilled workers are higher in the more advanced economies as a result of the
complementarity between physical and human capital.
Alternative theories of the wage differential between developing and developed
countries are found in the more recent literature. The Kwok-Leland model of asymmetric
information, for example, was constructed to explain skilled migration for a very specific
circumstance: the incidence of non-return among Taiwanese students completing their
studies in the United States. This phenomenon could not be explained by traditional
arguments since the Taiwanese economy is viewed to be competitive in many respects with
the United States economy. Kwok and Leland proposed informational asymmetries
between domestic and host employers as a possible explanation for the emigration of
Taiwanese students. They argued that host firms have an advantage over home firms in
terms of their knowledge about the true productivities of students completing their studies
in the host country, and are, therefore, able to give appropriate compensation. Home
countries, on the other hand, would only be willing to offer a wage based on the average
productivity of returning students. This model was criticized on the ground that information
asymmetries can only be temporary, since home country employees would eventually
discover the true productivities of returning students and compensate them accordingly.
The second model considered in the chapter is based on the idea of increasing
returns to advanced education or “agglomeration economies”. The basic idea behind the
agglomeration externalities argument is that the concentration of individuals—professionals
49
with similar interests—in the same area increases the productivity of professional work.
Consequently, professionals and scientists in less developed countries, facing limited
opportunities in their native countries to interact with colleagues who have similar interests
and research agendas, choose to move to countries where they can do so. As more and more
individuals who specialize in the same area get together, this creates further incentive for
others to do the same. This may be viewed as an extension of the traditional argument in
which skilled migration is the result of differences in the physical capital base of the source
and destination countries to include social (or human) capital stocks. Wong’s (1995) model
of brain drain based on learning-by-doing interprets the greater output level in the host
country as representing a cumulative base of experience. Foreign workers choosing to stay
in the host country are able to take advantage of the greater base of experience and increase
their productivities from learning-by-doing.
Chen and Su (1995) have extended Miyagiwa’s agglomeration economies
argument to account for why student non-return is a more prevalent form of brain drain.
Cross-country differences in the stock of capital (human and physical) may serve to explain
why a wage differential exists between the host and source countries. However, it does not
explain why the failure of foreign-educated students to return to their countries of origin is
more prevalent than the migration of professionals who are educated in the home country.
To explain this, they propose a model in which training received on the job is specific to the
social stock of capital of a country. This training, however, is more productive when it is
obtained in the country where the advanced education is received. It is, therefore, argued
that on-the-job training received abroad after the completion of academic studies
complements the education received in the foreign country and increases marginal
productivities and wages.
Theories that endogenize the cost of migration were also considered in the chapter.
Migrant networks reduce both the pecuniary and non-pecuniary costs associated with the
initial move. Herd behaviour, on the other hand, arises as a result of uncertainty. Although
the theories of migrant networks and herd behavior are more generalized theories that do
not distinguish between skilled and unskilled migrants, they can be easily adapted into a
theory of skilled migration. The last section considered other non-economic factors that
may be important in the decision to migrate, such as language ability and psychological
50
factors. Difficulties in adapting to new circumstances may also be thought of being part of
the “psychic” costs of migration.
In the next chapter, background information on labor market conditions in Turkey
as well as a brief outline of Turkey’s experience with skilled migration is provided. From
this analysis, it appears that institutional, demographic and political factors are as
prominent in determining labor mobility patterns as purely economic reasons. In fact, each
set of factors are related and should be considered together to provide an understanding of
why skilled migration from Turkey takes place. The theories of on-the-job training (or
specialized training) and learning-by-doing are tested empirically in the econometric study
of the Turkish brain drain presented in Chapter Seven.
51
CHAPTER 4
LABOR MARKET CONDITIONS IN TURKEY AND TURKEY’S EXPERIENCE
WITH SKILLED MIGRATION
4.1 Introduction
Migration, both internal and across borders, is nothing new for Turkey. A
significant amount of rural-to-urban migration continues to take place within Turkey’s
borders, and is driven in large part by the greater employment and educational opportunities
available for families in urban areas. Paralleling this, a significant number of highly
educated individuals from Turkey have chosen to take advantage of overseas employment
opportunities. A significant number of them have also gone through a period of training and
education in their country of destination, reflecting in part the lack of opportunities for
specialized study within the higher education system in Turkey. The focus of the current
chapter is on labor market conditions and the higher education system in Turkey, which
will provide the background for the exploratory and empirical study of return intentions
presented in the latter part of the thesis.
Some of the factors that have been cited as important for skilled migration include
political instability, lower salaries and lack of employment opportunities in the home
country, as well as a preference to live abroad. In addition to these factors, several other
features of Turkey’s political economy are considered to be important in explaining the
Turkish brain drain. These include the lack of a national research and development strategy,
distortions in the education system and foreign language instruction in schools, all of which
have important labor market consequences (Kaya, 2002). Turkey’s first “brain drain” wave
began in the 1960s, with doctors and engineers among the first group of emigrants. During
that period, Europe was the most popular destination for Turkish professionals and
academicians (Kaya, 2002). Political instability and crisis, followed by the military coup in
1960 are believed to have instigated this initial exodus of highly skilled individuals. In
52
recent years, attention has shifted to young university graduates who are seriously
contemplating starting their careers abroad as a result of the current economic crisis.
Postgraduate studies overseas provide the first step for many in fulfilling this goal. Another
serious problem is that of non-returning government-sponsored research assistants who
have been sent abroad as an investment toward filling academic positions in the expanding
Turkish higher education system.
The brain drain issue has received considerable attention from the Turkish media as
a serious economic and social problem, particularly in the aftermath of the economic crises
of November 2000 and February 2001. In the earlier 1994 crisis, Turkey’s GNP had
declined by 6.1 percent. Although this was a record contraction at the time, the economy
recovered quickly in the following year and recorded a positive growth rate of 8.0 percent.
The 2001 economic crisis, however, was much more severe and GNP contracted by 9.4
percent, which is the worst growth performance in the history of the Turkish Republic1. The
recent crisis has been both prolonged and widespread in its repercussions compared to the
previous crises, affecting also university graduates on a much wider scale (I ı ıçok, 2002).
Even graduates of the prestigious universities in Turkey, who usually face better than
average prospects in the labor market, were affected. The perception of the brain drain as a
serious problem has increased following each crisis, and has also attracted the attention of
national authorities. In 2000, the Turkish government decided to form a joint task force of
experts from the Turkish Atomic Energy Agency, the Turkish Academy of Sciences
(TÜBA) and the Scientific and Technical Research Council (TÜB TAK), in order to
investigate Turkey’s brain drain problem (Cumhuriyet Gazetesi, 2000).
The chapter begins with a background on the economic and social conditions
prevailing in Turkey, and thus presents a setting within which to evaluate the migration
decisions of skilled individuals from Turkey. Section 4.2 reviews the conditions within the
higher education system in Turkey that may have promoted the exodus of tertiary level
students and exacerbated Turkey’s brain drain problem.
1
Figures were obtained from the State Institute of Statistics website:
http://www.die.gov.tr/ieyd/milhes/page27.html. Görün (1996) also indicates that in economic
downturns university graduates increasingly replace the positions that were previously filled by high
school graduates, and this is said to lead to deskilling of the work force with university education.
The tertiary level graduates who work below their appropriate skill level is also seen as an important
problem.
53
4.2 Supply and Demand in Higher Education
The higher education system may be thought of as the intermediary that produces
individuals with special skills and proficiencies that form the human capital required in
producing a more sophisticated range of final products. While there is considerable demand
pressure on the Turkish higher education system, improvements on the supply side have
been slow to take place.
Empirical studies indicate that investment in higher education, compared to the
other schooling levels, earns a very high private rate of return for both men and women in
Turkey (Dayıo lu and Kasnako lu, 1997; Tansel, 1994, 1999). Furthermore, these studies
also point to significant regional differentials in the rates of return to education at all levels.
While university education provides a high private rate of return in all regions, both
developed and underdeveloped, the highest returns are, not surprisingly, found in
industrialized districts where the three metropolises, stanbul, Ankara and zmir serve as
centers of attraction. The regional disparities in the private gains from education as well as
the greater educational opportunities have created a massive rural-to-urban exodus. This
has, in turn, exacerbated the regional disparities within Turkey, creating squatter
settlements with high levels of poverty. While unskilled workers show a high degree of
mobility within the domestic economy, highly educated workers show a high degree of
international mobility. The uneven development of the Turkish economy with disparities at
many levels including education, wages, and employment has led to both unskilled internal
migration and brain drain to other countries.
Economic development and rapid population growth have increased enrolments at
the primary and secondary levels of schooling, which, in turn, has generated a growing
social demand for higher education. According to a recent Higher Education Council
report, the high schools in Turkey, which currently take three years to complete, do not
provide adequate labor market preparation for their students2. The report indicates that “the
2
Indeed, there is informal evidence that suggests high school education is also inadequate in
preparing students for university education. To improve their chances of getting into a quality
university, many urban high school students go to after-school and week-end private tutorial schools
that have sprung up to profit from the enormous competition created by the nation-wide placement
exam. It may be reasonable to suggest that, ironically, the formal secondary education system has
been overshadowed by the preparations for the university placement exam. A graduate from an
Ankara high school, for example, admitted that students in their final year of high school spend most
of their in-class time solving exam questions, and that “teachers pretty much stay out of the way
54
main reason for the demographic pressures exerted on the Turkish tertiary system is the fact
that high school graduates who are unable to get into college or university lack the
knowledge and skills necessary to earn a livelihood” (YÖK, 2001: 30). The lack of in-firm
training programs on a wide scale is also believed to aggravate this problem. As a result,
university education is seen as an important means for training students and imparting the
skills that are critical for securing jobs.
In response to demand pressures, the number of universities in Turkey increased
from a total of eight prior to 1970 to seventy-one at the beginning of 1998. The expansion
of public and private universities is continuing at a rapid pace today. The Higher Education
Law (Yüksek Ö retim Kanunu), enacted in 1981, brought about a major reorganization of
the higher education system in Turkey. In 1982, with the establishment of the new
constitution, the Council for Higher Education (Yüksek Ö retim Kurulu – YÖK from
henceforth) was created to plan, coordinate and oversee many of the important activities of
the higher education system within the provisions of the higher education law. This was an
important step toward the creation of a centralized and unified higher education system that
at the same time entailed a compromise in autonomy for individual universities.
The new 1982 constitution also included a provision that allowed non-profit
foundations to establish higher education institutions. This officially marked the beginning
of the private or “foundation” university system in Turkey3. The first private university,
Bilkent, was formed soon after in 1984 and started accepting students in 1986. Since then,
following the enactment of the Foundation University Law4 (Vakıf Üniversitesi
Yönetmeli i) in 1991, which clarified the conditions under which foundation universities
could be formed and managed, 23 new private universities have been created. The newly
established private university system in Turkey has succeeded in attracting talented foreign
and Turkish academicians from abroad by offering competitive wages and state-of-the-art
equipment and facilities. On the other hand, private universities charge tuition fees that are
because they know that getting into university is important to us.” See also Tansel and Bircan
(2002) for an analysis of private tutoring and the demand for education in Turkey.
3
Previous attempts, during the late 1960s and into the 1970s, at forming private universities to meet
the growing demand for higher education were thwarted on the ground that they were
unconstitutional, and the existing for-profit private higher education institutions were absorbed into
the state university system.
4
Law No. 3785 passed in 1992.
55
generally out of the income range of a majority of Turkish families, although they provide
scholarships to exceptional candidates scoring high on the national placement exam.
Enrolments at the private universities are lower than for the state universities partly because
these universities promise a lower student to teacher ratio, but more importantly because
families find the tuition and education costs prohibitive. Thus, while private universities
have partially reversed the academic brain drain to other countries, they have not eased the
demand pressures on the higher education system. Relatively few students are able to take
advantage of the opportunities provided by the private universities in Turkey. Those who
can afford the high tuition fees come from a higher socioeconomic group, and this serves to
aggravate the existing problem of unequal opportunities in education.
The number of state universities has also increased dramatically over the years.
While there were only eight state universities prior to 1970, this number reached 53 by the
year 2000, compared to 21 for the private universities. State universities are free of tuition
by law, although students must still pay a mandatory “contribution fee” at the start of each
term, which is much lower that the tuition in private universities. For this reason, a majority
of students enroll in state university programs. State universities, therefore, carry an
essential part of the responsibility of providing post-secondary education to a broader group
of students. The distance education program offered by Anadolu University since 1982,
consisting of both 2-year technical college and 4-year university programs, has become an
important means for absorbing some of the demand for higher education, accounting for 30
percent of total enrolments (YÖK, 2001). This unique distance education program has been
called the “largest university on Earth” by the World Bank since nearly half a million
students are enrolled in this program from different parts of Turkey as well as from
different countries (MacWilliams, 2000).
Despite the rapid increase in the number of both private and public universities and
the removal of quota restrictions in distance education programs, only a third of all
candidates taking the entrance exam in 2001 could be placed in a higher education
institution, including distance education. A significant number of those who are placed in
higher education programs do not enroll. Many students, for example, who qualify for the
distance education program choose not to enroll and instead wait to take the exam the
following year in order to be placed in a regular university program. Similarly, those who
do not qualify for the more prestigious universities or their desired programs also wait
56
before enrolling. Ministry of Education statistics indicate that only about a third of all
students taking the university placement exam are final year high school students; many
others take the exam several times in order to be placed in their desired program or school.
There are significant disparities in the quality of higher education institutions as
well. The sharp rise in the number of higher education institutions after 1980 has sparked
the quantity-quality debate in higher education. It is claimed that the quantitative expansion
of universities has occurred at the expense of quality, which is measured in part by
indicators such as student-teacher ratios, and the physical resources devoted to teaching and
research ( enses, 1994). The public and private resources devoted to higher education have
not kept up with the expansion in enrolments, institutions and programs, and there appears
to be chronic understaffing in terms of student-teacher ratios, especially for the state
universities (Dündar and Lewis, 1999). Academic staff at state universities also receive
salaries that are far below those of the private universities. Like the wages of other civil
servants in Turkey, the salaries of academicians in state universities are set by legislation
and they have not kept up with inflation. The February 2001 economic crisis has made the
situation worse by more than halving the value of the academic salaries at the state
universities. There is indication that moonlighting and extra teaching activities to
supplement incomes are becoming more prevalent (Cumhuriyet Gazetesi, 2001). Such a
trend will undoubtedly have dire consequences for research-related activities, and
inevitably lead to the loss of some the best researchers to private and overseas universities.
The quality gap, both perceived and real, at the university level also has important
consequences for university graduates entering the labor force in Turkey. The quantitative
expansion of universities, with little regard for quality, has yielded graduates with diplomas
that appear to have little value in the Turkish labor market. For example, the most lucrative
jobs in the labor market are offered to the graduates of a small number of universities with
well-established reputations5. The “signal” value of obtaining a diploma from one of these
institutions, therefore, creates immense competition among high school students for getting
acceptance to the more prestigious universities. It is also interesting to note that almost all
5
A cursory look at the job openings in the classified section of the major Turkish newspapers reveals
that for many top-level firms, there is a strong preference for graduates of established universities,
and in particular, those that produce candidates who are fluent in at least one of the major foreign
languages, with English topping the list. Even when the ads do not specifically mention any
universities by name, many are given in English or German which strongly favours candidates with a
foreign language education background.
57
of the private universities, most of which have been formed after 1995, have adopted
English as the language of instruction in order to attract students, because the job market
strongly favors candidates with fluency in at least one major foreign language.
4.3 A Closer Look at Non-Returning Students
Demand pressures have led to an increase in the number of students who are
studying abroad with their own means (private students) or on foreign scholarships. A
majority of these students are pursuing undergraduate studies. Some are recruited by
prestigious foreign universities, while others choose foreign study after failing to be placed
in a program in a national university. There are also those who do not want to go through
the stress of taking the very competitive nation-wide university placement exam. Another
important reason for wanting to study at a foreign university is the belief that it will provide
better quality education. Section 4.3.1 takes a close look at private students studying
abroad. In response to the pressures on the higher education system in Turkey outlined in
the previous section, the Ministry of Education and the Higher Education Council increased
the numbers of scholarships for post-graduate studies abroad. These scholarships hold the
condition that scholarship recipients return and fill positions in the newly established state
universities. Section 4.3.2 shows that non-returning scholarship recipients have become a
concern for the education authorities.
4.3.1 Private Students
According to Ministry of Education statistics, a total of 21,570 Turkish students
were studying abroad with their own means in mid-2001. Two-thirds of these students
chose universities in Western Europe and North America, while a significant proportion (22
percent) also chose the Turkic republics in Central Asia as study locations. The majority of
private students are pursuing undergraduate studies and nearly 90 percent of them are male.
This gender gap also persists at the postgraduate levels of study, being slightly higher in the
technical fields in comparison to the social sciences. Figure 4.1 below provides the figures
for the number of private students studying abroad in 2000 by program of study and gender.
Part of the explanation for the great number of students in overseas undergraduate
programs can be traced back to the inability of the higher education system in Turkey to
absorb the demand for education at the university level. Demographic factors, including a
58
high population growth rate and a high percentage of the young in the total population,
have led to both an expansion in demand for schooling and an increase in the Turkish labor
force. Labor force participation rates, however, have not kept pace with population growth,
showing instead a decline over the years. This is attributed partially to the “discouraged
worker effect” from a lack of employment generation despite a high growth rate compared
to OECD levels, except during the crisis periods, ( enses, 1994; Tansel 2002b). The return
rate of private students is not known. However, it is expected that non-return will be more
prevalent in the absence of a “moral contract” to break as in the case of national scholarship
recipients.
13,074
14,000
12,000
10,000
8,000
6,000
2,475
816
4,000
2,000
206
481
0
rate
Do cto
1,559
male
female
er's
M ast
uate
rg rad
Unde
Figure 4.1 Private Overseas Students in 2000, by Program of Study and Gender
Source: SIS (2002: 171), Table 105.
4.3.2 Government-Sponsored Students
In addition to private students, there are several thousand government-sponsored
students who are studying abroad, most of them at the postgraduate level as part of the goal
of training academicians to fill positions in state universities. The great majority (90
percent) of the government-sponsored students are studying in the United States and Great
Britain. Law 1416 (Law Regarding Students to be Sent to Foreign Countries), enacted in
59
1929, provided many students with the opportunity to study abroad on a scholarship
provided by the National Education Ministry (Milli E itim Bakanlı ı - MEB). The original
aim of these scholarships was to train civil servants to fill positions in the growing public
sector of the newly formed Turkish Republic. With the expansion of the higher education
system, the emphasis shifted to the creation of a cadre of foreign-educated academicians to
staff the newly-established universities in Turkey and to thus enrich the educational
standards of these universities. The number of government-sponsored students for the
period 1963-1998 is given in Figure 4.2 and includes all levels of study, undergraduate and
graduate. In October 2002, the number of students sponsored by the government was 7206,
a majority of whom were pursuing doctorate level studies (77.2%).
1400
1200
1000
800
600
400
200
19
63
19
65
19
67
19
69
19
71
19
73
19
75
19
77
19
79
19
81
19
83
19
85
19
87
19
89
19
91
19
93
19
95
19
97
0
Social Sciences
Natural Sciences
Figure 4.2 Government-Sponsored Students, 1963-1998
Source: Various issues of the Statistical Yearbook of Turkey.
Note: Includes students sponsored by various Ministries and other Government Institutions.
In 1987, the Higher Education Council (YÖK) also began awarding scholarships to
university graduates for postgraduate studies abroad. The YÖK scholarships share the same
purpose as the Ministry of Education scholarships, which is to supply the Turkish higher
education system with qualified academic staff. These scholarships also provide foreign
study opportunities for students who would otherwise not have been able to finance the
expenses involved in overseas education, provided that they meet at least the minimum
6
Figures are from MEB Sayısal Veriler 2003-2004, available at the MEB website.
60
criteria specified in the terms of these scholarships. The number of research assistants sent
abroad on scholarships awarded by the Higher Education Council (YÖK) is given in Figure
4.3. YÖK awarded the greatest number of scholarships in 1993, but this number declined
sharply after 1993 as a result of a change in policy to award fewer scholarships under more
stringent requirement in order to increase the quality of recipients (YÖK, 2003).
1400
1200
1000
800
600
400
200
0
1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002
total
149
399 187
226 208
159 1282 131
285 162
76
97
80
63
53
74
USA
67
138
66
67
60
745
64
169
85
40
56
50
47
32
53
England
38
208 112
127 110
74
446
52
93
57
17
21
8
9
10
11
55
Figure 4.3 Number of Research Assistants Sent Abroad on YÖK Scholarships
Source: YÖK (2003).
Both the MEB and YÖK scholarships are given in return for compulsory academic
service in the universities of Turkey. This generally means that for every year of study
abroad, the scholarship recipient must spend two years working in a pre-chosen university
in Turkey when they complete their studies. Since most of the scholarships are given for
doctoral level studies, the amount of the academic service amounts to eight years on the
average. Students who fail to comply with the terms of the scholarship must pay back the
value of their scholarship plus interest. Between 1987 and 2002, a total of 3631 research
assistants were sent abroad on YÖK scholarships to pursue graduate level education (see
Table 4.1 below). Nearly 90 percent were sent to the United Stated (49.2 percent) and
England (38.4 percent), with the remaining dispersed over twenty five countries (YÖK,
2003). In 2002, 762 YÖK scholarship recipients were continuing with their studies abroad.
61
Table 4.1 YÖK Scholarship Recipients by Status, 2002
1987-2002
Number
%
3631
100.0
762
21.0
1667
45.9
Returning with a Master’s Degree
375
10.3
Returning without a diploma
351
9.7
Non-returning students
473
13.0
Scholarships awarded
Students:
Continuing their Education Abroad
Returning with a PhD
Source: YÖK (2003).
Despite the good intentions behind these scholarships, there is indication that they
may not be fulfilling their purpose, at least to the extent that they had been envisioned.
According to the 2003 report by YÖK, 473 of the total of research assistants sent abroad to
study since 1987 have not returned to Turkey. While some scholarship recipients have
officially resigned from their position of research assistant, others have been considered as
“resigned” for not complying with the terms set out in the scholarship or ending their
communication with the Higher Education Council. There is indication of high
dissatisfaction among scholarship recipients with regard to the terms of the scholarship, the
bureaucratic processes they have had to face, and the general inflexibility shown for special
or changing circumstances of the recipients. There is also indication of some abuse of the
state scholarships by a number of recipients who view these scholarships primarily as a
stepping stone for taking advantage of overseas opportunities that they otherwise could not
have afforded. Some of these students opt to repay the value of the scholarship after earning
money abroad instead of fulfilling the compulsory academic service requirement. The YÖK
scholarship program has had a 58% success rate7 so far in terms of fulfilling its stated
purpose of producing PhD recipients with foreign degrees who return and take academic
positions in Turkish universities.
7
Those continuing with their overseas studies are excluded from the calculation (1667÷2869×100).
62
4.4 Output of the Higher Education System: Stock of Graduates
Turkey as a middle-income developing country does not experience the same degree
of difficulty in producing a university-educated population as some of the least developed
countries, which lag further behind in terms of the number of teachers, institutions and
educational infrastructure. Figures 4.4 and 4.5 present the number and percentage of
graduates respectively from Turkish universities by major discipline8 for the three decades:
1970s, 1980s and 1990s. Education science had the greatest share of graduates in the 1970s,
followed by the social sciences, engineering and medicine. The 1980s witnessed a
substantial rise in graduates from the social sciences at the expense of education, and
modest increases were seen in the shares of the remaining disciplines with the exception of
the share of law school graduates, which declined. Education’s share continued to decline
in the 1990s, while the share of social sciences increased to more than a third of the total
number of graduates.
Within the social sciences, half of all graduates have graduated from programs in the
economic and administrative sciences. The rise of the private banking sector in the 1990s
created many employment opportunities for social science graduates, which may explain
the striking increase in enrolments and in the number of graduates within the discipline.
The economic crises took many of these opportunities away, however, and left numerous
university graduates unemployed. One of the effects was to increase the number of
applications for graduate level study at Turkish universities and overseas.
The State Planning Organization (SPO) has made supply-demand projections in the
Eighth Five-Year Development Plan for education and health personnel as well as other
occupations (see Table 4.2). In the health sector, the projected shortfall in supply is greatest
for the nursing profession; this is followed by doctors. The low share of graduates from the
health sciences indicates that the projected shortage may become a reality unless measures
are taken to increase the incentive to enroll in health programs. The contention of the
beneficial brain drain models is that the possibility of emigration increases an individual’s
desire to invest in tertiary education. The less developed country benefits since not all of
the university graduates will be admitted as immigrants due to restrictions or quotas
imposed by foreign countries. If individuals do indeed base their education decisions on
8
Classifications are based on UNESCO’s International Standard Classification of Education
(ISCED). Medicine also includes nursing and other health services.
63
500,000
400,000
300,000
200,000
100,000
0
g
s
e
n
s
es
in
ce
rin
t io ence
it i
c
e
n
a
i
n
e
e
a
uc
ci
ed
ci
in
um
M
Ed
lS
lS
ng
a
H
a
E
i
r
c
tu
So
Na
1970-71/1979-80
e
w
ur
La cult
ri
Ag
1980-81/1989-90
Ar
ts
r
he
Ot
1990-91/1999-00
Figure 4.4 Graduates from Universities in Turkey, by Discipline 1970-1999
40.0
35.0
percentage
30.0
25.0
20.0
15.0
10.0
5.0
th
er
O
rts
A
La
w
gr
ic
ul
tu
re
A
So
c
Ed
uc
at
io
ial
n
Sc
ie
nc
En
es
gi
ne
er
in
g
M
ed
ici
ne
H
um
N
an
at
it i
ur
es
al
Sc
ie
nc
es
0.0
1970-71/1979-80
1980-81/1989-90
1990-91/1999-00
Figure 4.5 Graduates from Universities in Turkey, by Discipline 1970-1999 (%)
Source: SIS, various issues of the Statistical Yearbook of Turkey.
Note: The Natural Sciences category includes Mathematics and Computer Science.
64
possibilities at home and abroad, then they must also base their choice of occupation on the
same criteria. This means that the “beneficial brain drain” models have ignored the
potential effect of emigration on career choice when making their prediction that home
countries will benefit from emigration. It is possible that students will choose study
programs that are in high demand in developed countries so as to increase their chance of
being accepted as immigrants. The potential for emigration can therefore alter the incentive
structure for choosing study programs.
Table 4.2 Demand and Supply Projections for Selected Occupations, 2005 (‘000)
Supply
Demand
Projected
Shortfall in
Supply
Primary School Teachers
394.8
413.0
18.2
Secondary School Teachers
210.1
180.0
-30.1
85.0
119.5
37.5
Doctors
89.0
121.7
32.7
Dentists
16.0
28.3
12.3
Pharmacists
21.3
26.2
4.9
Nurses
77.1
212.8
135.7
Occupation
Education
Higher Education – Academic Personnel
Health
Source: SPO (2000) Eighth Five-Year Development Plan.
4.5 Can Turkey Afford to Ignore Skilled Emigration?
One of the views on the brain drain is that it ceases to be a problem when countries
develop. This suggests that a positive development path is a given for developing countries.
Figure 4.6 demonstrates the striking differences in educational attainment among
industrialized and less developed countries. The United States has the highest level of
average educational attainment over the period 1960-1999. This is followed by Japan and
Taiwan, and then by China, Turkey, India and Pakistan. These countries all have significant
numbers of students and academic staff in United States universities (see Open Doors).
While Turkey has made progress in development over time, her position in education
with respect to other countries has marginally improved, if at all. Figure 4.7 below shows
the difference in the average years of schooling between Turkey and the United States,
65
14
United States
12
10
Japan
Taiwan
8
6
China
Turkey, India
4
Pakistan
2
0
1960 1965 1970 1975 1980 1985 1990 1995 1999
United States
Taiwan
China
India
Pakistan
Japan
Turkey
Figure 4.6 Mean Years of Schooling for Selected Countries, 1960-1999
Source: Barro and Lee (2000) education dataset.
10
9
Years of Schooling
8
7
6
5
4
3
2
1
0
1960
1965
1970
1975
United States
1980
1985
Germany, West
1990
1995
1999
Japan
Figure 4.7 Difference in Mean Years of Schooling between Turkey and
Selected Countries, 1960-1999.
Source: Barro and Lee (2000) education dataset.
66
Japan and West Germany. In the thirty years spanning the period 1960-1999, the difference
between Turkey and Japan has remained steady at five years of schooling. The greatest
difference in the average years of schooling is with the United States at around seven years.
For the same period, there appears to be a worsening in Turkey’s position vis-à-vis the
United States. The only improvement appears to be with that of West Germany, and that is
partially a consequence of West Germany’s unification with East Germany and the
resulting influx of less educated individuals.
Table 4.3 School Expectancy* in 2000 for Selected Countries
United Kingdom
Germany
United States
Argentina (1)
Brazil (1)
Turkey
China
All Levels
Primary and
lower secondary
education
Upper
secondary
education
Tertiary
education
18.9
17.2
16.7
16.4
15.7
10.1
10.1
8.9
10.1
9.4
10.6
10.9
7.5
8.5
7.4
3.0
2.6
2.1
2.6
1.7
1.2
2.5
2.0
3.4
2.7
0.9
0.8
0.4
Source: OECD Education at a Glance 2002, Table C1.1 (www.oecd.org).
*
Expected Years of Schooling for a 5-year-old Under Current Conditions.
(1) Reference year is 1999.
Table 4.3 shows the school expectancy for a five-year-old under current conditions.
These figures are for the year 2000 and are taken from the OECD publication Education at
a Glance 2002. A five-year old in Turkey can expect to receive a total 10 years of
schooling, which is below the secondary level. The expected level of schooling at the
primary and lower secondary education levels is only 7.5 years, which is below Turkey’s
current goal of universal primary level education. The figures for China are similar. In the
industrialized countries, a five-year-old can expect to receive over 16 years of schooling.
The United States has the highest expected number of years of schooling at the tertiary
level at 3.4 years, while in Turkey a five-year-old can only expect to receive less than a
year of tertiary education. Given this bleak outlook, the migration of tertiary-educated
individuals carries added significance.
67
Turning our attention to the health sector, it would appear that Turkey has made
some progress here. The numbers of physicians and nurses per population have risen
considerably since the forming of the Republic (see Figures 4.8 and 4.9). While this may
seem impressive, when these figures are compared to those for other OECD countries, a
significant gap remains between Turkey and the OECD average. Figure 4.10 gives the
practicing physicians per 1000 population for Turkey and selected countries for the years
1960 and 1998. Turkey’s ranking among the OECD countries in terms of the number of
physicians has remained the same. It should also be kept in mind that these aggregate
figures mask serious regional differentials within Turkey, where the southeast provinces
face a serious shortage of qualified personnel in health and education. The return of
educated individuals, therefore, does not necessarily imply a quick solution for these
internal disparities.
4.6 Concluding Remarks
The overview of recent labor market conditions and the higher education system in
Turkey, provided in the current chapter, sets the macroeconomic context for the study of
return intentions of overseas professions and students from Turkey. Turkey’s recent
experience with economic crises has created great uncertainty for both the unskilled and
skilled workforce. One of the characteristics of the current spell of economic instability is
the high rate of unemployment among the university educated. It is expected that economic
conditions will be a prominent factor in the return decision of overseas professionals and
students from Turkey. The next three chapters are devoted to the presentation of the
empirical investigation of return intentions based on a survey of over 2000 Turkish students
and professionals residing overseas. Chapter Five gives details of the survey methodology,
including survey design and strategies employed to collect data for the exploratory and
empirical study. Chapter Six provides the preliminary survey results, while Chapter Seven
presents the econometric analysis of the determinants of return intentions.
68
90,000
80,000
70,000
60,000
Physicians
50,000
Inhabitants per
physician
40,000
30,000
20,000
10,000
19
28
19
33
19
38
19
43
19
48
19
53
19
58
19
63
19
68
19
73
19
78
19
83
19
88
19
93
19
98
0
Figure 4.8 Stock of Physicians and Inhabitants per Physician in Turkey, 1928-99
120,000
100,000
80,000
Inhabitants
per Nurse
60,000
Nurses
40,000
20,000
Figure 4.9 Stock of Nurses and Inhabitants per Nurse in Turkey, 1928-99
Source: SIS (1996); SIS (2002).
Note: Totals include practitioners, specialists and assistant physicians.
69
19
98
8
197
19
83
198
8
199
3
3
197
3
19
48
195
3
19
58
19
63
196
8
194
19
28
19
33
193
8
0
ted Tur
Ki key
ng
do
m
Ne Ja
w pa
Ze n
al
an
Po d
l
Au and
str
al
No ia
rw
ay
Un Swe
ite d e
n
d
Ne Sta
t h t es
er
la
nd
Fi s
nl
an
d
Fr
an
c
Au e
str
Po ia
rtu
Hu gal
ng
D ary
en
m
a
Ic rk
e
Sw lan
itz d
er
la
Ge n d
rm
a
Be ny
lg
iu
m
Gr
ee
ce
Ita
ly
Un
i
7.0
6.0
5.0
4.0
3.0
2.0
1.0
0.0
1960
Source: OECD Health Data, 2002 (www.oecd.org).
70
1998
Figure 4.10 Practicing Physicians per 1000 Population for Selected OECD Countries
CHAPTER 5
SURVEY METHODOLOGY
5.1. Introduction
Chapter Three reviewed some of the theoretical explanations for why skilled
migration occurs, including possible reasons for the phenomenon of student non-return. In
Chapter Four, institutional, demographic and political factors were considered in seeking to
understand the reasons behind overseas study and the migration of professionals from
Turkey. A purpose of this thesis is to determine to what extent some of the models of brain
drain set out in Chapter Three hold for the population of Turkish professionals and students
currently working and studying abroad. With this aim, an Internet-based survey was
conducted to collect data on the return intentions of Turkish students and professionals
residing abroad. Data collection took place in the first half of 2002 and over 2000 responses
were received from the targeted groups of Turkish students and professionals residing in
various countries. The purpose of this chapter is to present a brief review of previous
studies relating to the determinants of skilled migration, and to provide a detailed
discussion of the survey methodology used in the current study.
A variety of empirical studies on the brain drain have been carried out, usually
benefiting from data drawn from custom designed surveys. Section 5.2 reviews some of the
empirical work investigating the reasons for skilled migration in various countries and
regions. The sample of studies reviewed display great diversity in terms of targeted
populations (student non-return vs. professional migration), time period, survey strategies
and methodologies, and in terms of their research focus (economic, sociological, or
psychological), which makes comparisons difficult. The vast array of studies and each
study’s particular perspective serve to highlight the complexity of the factors involved in
the decision to migrate. Many studies make use of “push-pull analysis” to delineate the set
of factors important in making the mobility decision.
71
Section 5.3 gives a detailed discussion of the survey methodology used in obtaining
data for the empirical analysis of the Turkish brain drain presented in Chapter Six and
Chapter Seven, including the choice of target populations, data collection procedures,
questionnaire design and survey implementation. The chapter concludes with a discussion
of the limitations of the survey and possible improvements in survey design, as well as a
discussion of other data sources that may be considered in future studies relating to skilled
labor mobility.
5.2. Review of Some Empirical Studies of the ‘Brain Drain’
Chapter Three presented a discussion of some of the economic theories of the brain
drain, including theories based on asymmetric information, increasing returns to advanced
education, and on complementarities between education and training received abroad, all of
which have their basis in human capital theory. Empirical studies of the human capital
theory of migration are relatively few owing to a lack of reliable data at both the aggregate
and micro levels. Data collection procedures for recording migration and its breakdown into
skilled or unskilled categories show great variation across countries.
The existing empirical studies, for the most part, have relied on data obtained from
questionnaire responses or face-to-face interviews to collect information on return
intentions and various factors believed to be important in the decision to return or stay
overseas. Some of these include studies on the Asian engineering brain drain (Niland,
1970), studies on China (Kao and Lee, 1973; Zweig and Changgui, 1995), and on Latin
America (Cortés, 1980). Studies focusing on the Turkish brain drain include O uzkan
(1971, 1975) and Kurtulu (1999). O uzkan’s study is based on a survey conducted in 1969
of 150 respondents holding a doctorate degree and working abroad. The study by Kurtulu
looks at the responses of 90 students studying in the United States in 1991.
While the studies cited above have relied on primary data collection through
questionnaires or interviews, one notable exception is the study by Huang (1988), which is
an empirical analysis of foreign student brain drain to the United States using data on 25
countries, including Turkey. The data used in Huang’s study was compiled mainly from
statistical documents published by various US and international agencies. As an immigrant
country, the United States possesses a comprehensive collection of records on foreign
students, foreign scholars and immigrants by different criteria. In his study, Huang tries to
72
explain why there is so much cross-country variation in the non-return rates of foreign
students studying in the United States. His dependent variable is the number of adjustments
made from an F-1 student visa to immigrant status in each year for the period 1962-1976.
These adjustments represent the non-return rate of foreign students. One of the important
findings of his study is that although income differentials are found to be statistically
significant in the econometric analysis, professional opportunities in the United States as
well as the social and political progress of the home country appear to be no less significant
in determining return rates. These results suggest that narrowing the wage differential alone
will not be sufficient to persuade students to return home.
Kao and Lee (1973) investigate the Chinese brain drain to the United States. Their
sample consists of scholars from mainland China and Taiwan. The variable of interest is the
“propensity to stay in the United States”, which is measured by a preference scale ranging
from 0 to 9. Their study confirms the importance of income, lifestyle preferences, political
freedom and the “lack of fair competition in Taiwan” in the propensity to stay in the United
States. A greater inclination to stay was found among scholars from mainland China
compared to scholars from Taiwan, which was as expected given the differences in political
freedom between mainland China and Taiwan.
An important characteristic of Taiwan is that it is competitive in many respects
with the economies of the more advanced countries. Despite the higher rate of return for
Taiwanese scholars compared to the mainland Chinese in the Kao-Lee study, Taiwan still
loses a segment of its skilled workforce, mainly in the form of student non-return, to the
United States and elsewhere. Chen and Su (1995) have proposed an explanation of why this
is so. Their model of on-the-job training as a cause of brain drain was discussed in detail in
Chapter Three. To reiterate, the argument is that advanced education and on-the-job
training received in the same country are complementary to each other in capital-intensive
occupations, such as science and engineering. Taiwanese students with advanced degrees in
these fields from overseas universities will be more productive and receive higher wages if
they remain in the country in which they received their education. Chen and Su attempt to
test this empirically with data from Japan. They analyze the likelihood that Taiwanese
students will remain in Japan after completing their studies there. However, they fail to find
significant differences in the “stay” inclination for students graduating from the so-called
“capital-intensive” fields of study.
73
Kao and Lee have also looked at differences in return inclinations across
disciplines and obtained similar results. Their expectation was to find a greater propensity
to stay among Chinese scholars in the natural sciences compared to those in the humanities
or social and behavioral sciences. They suggest that to the extent that differences in stay /
return inclinations across disciplines is the result of income differentials in these fields,
including the income variable in the regression analysis controls for this and makes the
field dummy variables statistically insignificant.
Zweig and Changgui (1995) provide a more recent study of the Chinese brain
drain. Their analysis incorporates both bivariate and multivariate techniques (multivariate
logistic regression) to investigate the return intentions of Chinese scholars and students in
the United States. Several important findings emerge from their study, one of which is that
a very high percentage of those interviewed are from a very high socio-economic
background. More than half of interviewees were the children of intellectuals and an
important proportion of them came from the “middle-level cadres”, which suggested
“unequal access to channels out of China.” Another important result was that previous
intentions about returning held significant predictive power over current intentions.
Political instability and economic conditions were equally important in return
considerations, while family considerations also played a prominent role. Zweig and
Channgui found women to be more reticent about returning than men, and they attributed
this to the relative lack of opportunities for personel development in China. A thorough
review of previous studies on the Chinese brain drain can also be found in their study.
Another study (Niland, 1970) investigates the engineering brain drain to the United
States from five Asian countries: India, China, Korea, Japan and Thailand. The focus is on
the determinants of student non-return for graduate students studying in various
engineering fields in the United States. Niland divides the respondents into four mobility
groups in terms of their work plans after completing their studies. The first group consists
of those who plan to return home immediately after completing their studies. The remaining
mobility groups consist of respondents who intend to work in the United States for a certain
period of time: up to eighteen months for the second group, up to five years for the third
groups and longer than five years for the fourth group. Based on this distinction, there
appear to be important differences in the reasons for wanting to work in the United States
across the three mobility groups and among the countries under study. Lifestyle preferences
74
hold the greatest significance for those planning a long period of stay in the United States,
while the savings motive appears to be important for those who intend a medium length of
stay. Niland also points to significant differences in the determinants of brain drain across
countries, and makes the recommendation that policies to curb brain drain should be
tailored to each country.
Hekmati (1973) presents the findings from a survey applied to students from five
developing countries, including Turkey. His study focuses on sociological factors rather
than the economic reasons of migration. A survey conducted in mid-1998 as part of the
South African Migration Project (SAMP), looks at the emigration potential of skilled South
Africans. This study reveals that South Africa is in danger of losing a significant portion of
its skilled population, both black and white, to countries such as the United States, Canada,
UK, Australia and New Zealand. More than two thirds of the sample of 725 South Africans
interviewed in a telephone survey have revealed that they were thinking about leaving
South Africa for better conditions abroad including greater safety, better services, more
favorable tax conditions, and lower cost of living (SAMP, 2000).
O uzkan (1971) has conducted a survey of Turkish professionals working outside
Turkey and does a qualitative analysis of the causes of the migration of Turkish scholars
and high skill workers to the rest of the world. The current study on the return intentions of
Turkish professionals and students residing abroad is based on a survey conducted in the
first half of 2002. This survey serves to update and extend the previous studies of the
Turkish brain drain by O uzkan (1971) and Kurtulu (1999), and uses the push-pull
perspective which is common among mobility studies. Details of the survey methodology
are given in the next section.
75
Table 5.1 A Sample of Previous Brain Drain Studies
Study
Data
Source
Destination Population
Country
under Investigation
Niland
(1970)
mail-out
questionnaire
United
States
O uzkan
(1971)
mail-out
questionnaire
OECD
Countries
Kao and
Lee (1973)
mail-out
questionnaire
United
States
Chen and
Su (1995)
Rotary Club
records
Japan
Zweig and
Changgui
(1995)
face-to-face
interviews
Kurtulu
(1999)
questionnaire
Lucas
(1975)
Huang
(1988)
various
statistical
documents
from the US
and other
sources
Observations Period
in Sample of Study
graduate students in
engineering from India,
China, Japan, Korea and
Thailand
Turkish professionals holding
doctorate degrees and
working abroad
447
individuals
1968
150
individuals
1968
Chinese Scholars and
Scientists from Taiwan and
mainland China possessing
PhD degrees
Taiwanese students who
received scholarships from
the Yoneyama Rotary Club
of Japan during the course of
their studies
372
individuals
1969
776
individuals
19621988
United
States
Chinese students, scholars
and former students in
workforce
273
individuals
1993
United
States
Turkish students studying in
the United States
90
individuals
1991
US Dept. of
United
Labor / INS;
States
United Nations
applications by male
candidates for labor
certificates in the US divided
by the male labor force in the
country of origin
103 countries
(both DC and
LDCs)
1973
United
States
Students switching from F-1
student visa to immigrant
status (Australia, Egypt,
France, Greece, W.Germany,
Hong Kong, Indonesia, India,
Ireland, Iran, Iraq, Israel,
Italy, Jordan, Japan, Korea,
Lebanon, Netherlands,
Philippines, Spain, Sweden,
Switzerland, Turkey, Taiwan,
United Kingdom)
375 (25
countries ×
15years)
19621976
76
5.3. Survey Design and Methodology
Advances in communication technology and the rapid spread of computer use
especially among the young and educated populations around the world during the past
decade have expanded the techniques and strategies available for data collection. While
Internet-based surveys employing web technologies and e-mail communication are
relatively new tools in the study of various behavioral phenomena, these types of surveys
have become increasingly commonplace. The new technology has considerably eased the
process of collecting responses, thus shortening the time frame for implementing previously
time consuming and costly survey studies. The current study is also based on an Internet
survey, designed for the purpose of collecting data on the return intentions of Turkish
students and professionals. The details of the survey methodology, including the selection
of the target populations, the sampling strategy used and questionnaire design followed by a
discussion of survey implementation, are presented in this section.
5.3.1 The Survey Population Defined
In this sub-section, a working definition of brain drain is presented for the purposes
of the survey study and the empirical analysis presented in subsequent chapters. The term
“brain drain” was initially coined by the British Royal Society in the 1950s to refer to
skilled Britons migrating to United States and Canada (Cervantes and Guellec, 2002). This
term is also used today to describe skilled individuals who leave their native lands to seek
better prospects elsewhere, but relates more to the skilled emigration from the developing
countries to the developed countries. More recently, student non-return has become
recognized as an increasingly important form of brain drain. “Student non-return” involves
individuals who go abroad to complete higher level studies and do not return to their home
countries after receiving their degrees.
However, it is often not clear what the term “skilled migration” refers to, as
evidenced by the wide variance in the definitions used in different studies. In many
migration studies, the educational attainment of migrants is taken to be an indicator of
skills, and “brain drain” usually refers to migrants with at least a tertiary (university) level
degree. This is, in general, done as a matter of convenience. Data on other activities that
contribute to the skills or productivity of individuals, such as learning-by-doing and on-thejob training are not available across countries. For the purpose of making international
77
comparisons of skilled immigration to destination countries, therefore, the educational
attainment levels of migrants are used to measure losses in human capital.
Studies that focus exclusively on university educated migrants are also likely to
ignore the effects of a loss in “entrepreneurial capital”. Since entrepreneurs in developing
countries have varying educational backgrounds, those with less than a university education
will be left out of the analysis of skilled migration when this refers only to those with a
university education. This entrepreneurial base, regardless of the educational attainment of
the emigrating entrepreneurs, may be crucial to the economic development of a country.
The South African Migration Project (SAMP, 2000), for example, includes businessmen
and businesswomen in their definition of the skilled population in South Africa, since they
are a crucial element of the South African economy.
In order for the migration of skilled individuals to be considered “brain drain”,
some investment must have been made by the home country in the education or skill
formation of the individuals involved. This investment need not be confined to public funds
since private indigenous funding of education and skills is also an investment made by
domestic agents within the home economy. Thus, no distinction should be made between
public and privately funded educational endeavors overseas. Students studying abroad
using their own means (e.g. family savings) should be treated the same as students who are
sent abroad for further studies on national scholarships in terms of the possible losses
incurred by the domestic economy if they do not return (e.g. the externalities they bring).
In this study, two separate but related populations are targeted. The first group
consists of students at the undergraduate or graduate level currently studying at higher
education institutions outside Turkey. The second group consists of individuals who have
earned at least a bachelor’s degree and are currently working abroad. In the second group, a
significant number have earned their highest degree in the country they are currently
working, and may be viewed as being part of the phenomenon of student non-return. The
remaining have left Turkey to work abroad after completing their highest degree from a
Turkish university. The individuals in this group who intend to settle permanently in
another country form part of the brain drain in the traditional sense. Also, a broad view of
“skilled migration” is adopted in terms of including students who have completed high
school in Turkey and are pursuing undergraduate studies abroad.
78
These two populations – students studying abroad and individuals who have earned
at least a bachelor’s degree – are chosen to constitute the pool of highly skilled individuals
abroad. It was believed to be appropriate to apply a separate survey to these two groups.
12,000
10,983
10,100
9,081 9,377
10,000
7,678
8,000
6,000
8,194
6,716
4,978
5,474
4,000
2,000
19
92
/9
19 3
93
/9
19 4
94
/9
19 5
95
/9
19 6
96
/9
19 7
97
/9
19 8
98
/9
19 9
99
/0
20 0
00
/0
1
0
academic year
Figure 5.1 Turkish Student Enrollments at US Universities
Source: Open Doors, IIE (various years, 2001).
Uncovering the population of all skilled individuals abroad as defined above is not
an easy task, since many host countries do not publish data that distinguishes between
skilled and unskilled emigrants. Since the precise population of skilled emigrants abroad is
unknown, determining an appropriate sample size at the outset for all of the groups
concerned proves not to be feasible. According to informal sources, an estimated 30,000 to
40,000 Turkish students are believed to be studying in various higher education institutions
in Germany, which makes Germany the single most important destination country for
Turkish students. The numbers in other Western European countries are unknown. The
most extensive dataset on foreign students and foreign scholars is found for the United
States in the annual publication of the Institute of International Education (IIE), Open
Doors. Turkish student enrolments at US universities has more than doubled in less than a
decade, reaching nearly 11,000 in the academic year 2000-2001 (see Figure 5.1). In the
1997-1998 academic year, more than a quarter of Turkish students were studying business
79
(27%), followed by engineering (23%), social science (10%), and math/computer science
(7%)1.
5.3.2. Sampling and Distribution Strategies
Construction of the Sampling Frame. Given the difficulties in determining the
actual size and location of the two groups targeted for the survey study, the initial part of
the sampling strategy involved compiling a list of the names and e-mail addresses of
potential participants that would serve as the sampling frame. Undergraduate students from
the Middle East Technical University were employed to help carry out the search for
individuals who fit the definition of the targeted populations. A considerable amount of
time was allotted to the construction of the list of potential survey candidates.
The e-mail addresses and names of Turkish professionals, scientists and students
were collected from various sources. An extensive internet search was undertaken to obtain
the e-mail addresses of Turkish academicians. The EDUCAUSE directory of higher
education2 provides a list of American universities and colleges based on the Carnegie
classification. The web sites of the academic departments of all faculties as well as all
affiliated research centers for various universities and colleges present on this list were
searched. A similar web search for names was also done for Canadian universities, and to a
lesser extent universities in the UK. Time and resource limitations prevented a full search
of the all of the universities listed. The information obtained through the above channels
was supplemented by various other sources.
While contact information for academicians was obtained from a search of staff
directories and department websites, it was somewhat more difficult to gather information
for Turkish workers in the overseas non-academic private sector. Some could be reached
from alumni listings and directories published in the websites of Turkish universities.
Overseas Turkish professionals associations, such as the Society of Turkish American
Architects, Engineers and Scientists, were also very helpful in reaching a portion of the
targeted population.
1
Figures are obtained from Open Doors Profiles Survey, which covers more than half (51%) of
Turkish Students studying in the United States for the 1997-1998 academic year.
2
Found on the web site: http://www.educause.edu/asp/dheo/carnegies.asp
80
As mentioned, another important source of brain drain candidates for the survey
was the alumni pages of Turkish universities where these were available. The BURCIN
(Bo aziçi University–Robert College) Database, for example, provided an additional 133
individuals. Several departments at Bo aziçi University maintain their own alumni lists on
the web: The Department of Computer Engineering provided a partial list of graduates from
the class of 1986 through 19993. An important problem with obtaining e-mail addresses
through these channels was that these pages were often outdated and many of these
addresses turned out to be invalid. However, once the names of former graduates were
reached through the channels mentioned, a search for their e-mail addresses could be made
from the internet search engines. Similarly, the e-mail addresses of students studying
abroad were also collected from the directories of universities and research centers located
in the United States, Canada, England and Australia, and the alumni pages of universities in
Turkey.
The collection of potential participant names and contact information depended to a
great extent on the existence and accessibility of student and personnel directories at
institutions of higher learning and research centers, the existence of accessible and up-todate alumni directories of Turkish universities, and the help of various Turkish associations
abroad. Unfortunately, the reliance on internet search procedures in the construction of a
list of potential participants has inevitably set limitations on who could be reached. For
example, individuals who were not members of any overseas Turkish associations, nor
listed in any directories, and without e-mail address information (especially older
participants) cannot be said to be adequately represented. Another limitation is that the
search for survey participants concentrated on universities and associations in North
America and England; time considerations did not permit expanding the search to other
important destination countries, such as Germany in the case of students and the Middle
East for skilled workers. The construction of a list of candidates, given the limited time
frame for conducting the survey, could not be expected to be exhaustive and uncover each
possible survey candidate.
Sampling and Distrubution Strategies. Since the size and distribution of the
populations are not known with certainty, the probability that a given respondent will be
picked as part of the sample is also unknown. A nonprobability sampling method known as
3
Found on the web site: http://www.cmpe.boun.edu.tr/∼alimoglu/cmpeaddr.html.
81
snowball sampling was chosen as an appropriate strategy to adopt (Rea and Parker, 1997).
Snowball sampling is also called “referral sampling” since it involves asking the initial
group of contacts on the list to assist in reaching other potential participants who are in the
targeted population. This strategy has the advantage of allowing a great number of
respondents to be reached in a relatively short period of time.
An e-mail cover letter was sent to potential respondents discovered through the
search process described above. The cover letter was used to introduce and explain the
purpose of the study, and contained a link to the web address of the survey page. The
potential respondents were invited to participate in the study and to forward the cover email letter to colleagues and friends who they believed would fit the targeted survey
population. The cover e-mail letter is provided in Appendix C, Section C.1.
Turkish student associations in the US, UK and Canada were also contacted in
order to help in the distribution of the initial e-mail message containing a link to the survey
website. The students from the targeted group who were contacted during the initial search
process were asked to distribute the cover email letter to their friends and acquaintances
who met the survey criteria. The distribution of the cover letter began in the middle of
December, 2001 and was ended in summer 2002. The address of the web page containing
the survey form was sent to the e-mail addresses of potential respondents. Turkish student
associations in the US, UK and Canada were also contacted in order to help with the
distribution of the cover e-mail containing a link to the survey website. The data collection
process began in mid-December 2001 and ended in Summer 2002.
Referral sampling is a fast and efficient, but potentially biased, means of reaching
the targeted populations. As mentioned, the “snowball effect” was an important method for
reaching potential participants. As an example, the METU Alumni North America
discussion group consisted of over 600 members at the time of the mail-out. A large
number of responses were obtained from this group within a short period of time.
The data collection procedures and the sampling strategy used suggest the
possibility that non-participants may differ systematically from participants in terms of
their characteristics and in their return intentions. For this reason, the survey results cannot
be used to generalize to the targeted population or universe as defined in Section 5.3.1.
Nevertheless, a good participation rate was reached with the strategy employed. The
82
combination of internet search and “snowball” or referral sampling resulted in a total of
1170 responses from Turkish students studying abroad, and 1282 responses from Turkish
professionals working abroad. After eliminating responses from non-target populations and
incomplete answers4, the number of valid responses totaled 1103 for the student survey, and
1238 for the survey of Turkish professionals. The list collected from the present study can
be used as part of longer term research agenda to study the Turkish brain drain and overall
mobility of skilled individuals from Turkey.
5.3.3 Questionnaire Design
A web-based survey was thought to be an appropriate method for gathering
responses since familiarity with computers and computer-based survey technologies would
be more widespread for the targeted groups of university students and university-educated
professionals. Greater acquaintance with and access to computers may differ from
discipline to discipline in the targeted group. For example, students and workers in
computer-related fields may have an advantage over other fields, such as the humanities, in
participating in a web-based survey. Nevertheless, for the period that the survey was
conducted it is reasonable to presume that the use of these technologies had become quite
widespread over all disciplines. The complete web version of the survey was hosted on the
Middle East Technical University server. The use of an academic domain address possibly
helped to increase the confidence of participants and convince them that of the “legitimacy”
of the survey study.
Figures 5.2a-b, 5.3, 5.4 and 5.5 below provide illustrations of the cover page and
sample pages from the web survey. The cover page provided links to two separate
questionnaires: the Turkish Student Survey Form, in English and Turkish, and the Turkish
Professionals Survey Form, in English and Turkish (Figure 5.2a-b). The respondent could
choose to answer either the English or Turkish version of the appropriate form. All of the
survey questions appeared on a single web page and respondents were asked to scroll down
to reveal more questions as they filled out the form. At the end of the questionnaire,
respondents were asked to send the completed form by clicking on the submit button
4
Non-target populations included respondents from the Turkish Republic of Northern Cyprus and
second-generation citizens of Turkish origin. Incomplete responses were eliminated on the basis of
the extent of incompleteness (e.g. if a majority of the questions were left unanswered or if important
portions of the survey were not filled out).
83
(Figure 5.5). If the respondent provided a valid e-mail address, a courtesy e-mail reply was
automatically sent to his/her e-mail address (Appendix C, Section C.2). Clicking on the
submit button also redirected the respondent to the “thank you” page (Figure 5.6).
The questionnaires were structured as a set of close-ended questions with an
optional open-ended question at the end that respondents could fill in as they liked with
comments about the survey questions or the topic of Turkey’s brain drain in general. The
survey consisted of several broad question groups that included sections on demographic
information, educational background, work-related information (job search and careerrelated intentions for students), as well as a section on return intentions and the related
“push” and “pull” factors that might be important in the decision to stay overseas. The full
set of questions for both surveys is provided in Appendix C (Sections C.3 and C.4). Figures
5.3-5.5, which relate to the student survey, give an idea about the appearance of the web
forms and the division of the survey into groups and sets of related questions. To ease
readability and eye-strain, non-imposing pastel colors were used in the background and to
separate blocks of questions. Blue was used to indicate section headings and alternating
shades of pink were used to separate each question. One respondent indicated, however,
that the lavender patterned background made the survey appear longer than it was and
suggested that solid coloring be used instead.
5.3.4 Survey Implementation: Some Caveats
The following points should be kept in mind when interpreting the survey results.
1) Self-selectivity: There may be self-selection bias since many respondents
volunteered their responses without being prompted. Many responses were obtained
through the “snowball effect”: those solicited for their participation were asked to forward
the e-mail message containing the survey cover letter and instructions to those who they
knew to be eligible for participation in the study. On the positive side, this increased the
number of responses received. This problem was overcome to some extent for students and
academicians at higher education institutions since extensive web surveys were carried out
to find student and academician names and e-mails.
2) While the survey was presented in two languages (English and Turkish), there is
the possibility that the interpretation of the questions may differ based on the choice of
84
language. There were several reasons for including the language option. One was a
technical reason. Limited international character support in older internet browsers would
make the survey difficult to read in Turkish and discourage individuals from responding.
The English language alternative was included to circumvent this problem. Since the study
focused primarily on North America—the United States and Canada—(but allowed for
responses from other countries), it was believed that many respondents would choose to fill
the survey in English, especially if they had been residing for some time outside Turkey in
an English-speaking environment.
3) Some expressed reservations about participating in the survey because
participation entailed disclosing private information, including e-mail addresses. “Network”
effects, similar to those mentioned in Chapter Three, appear to be influential in determining
the participation of those who were contacted. Those who responded positively to the
survey, in terms of thinking that it was important and worthwhile to do, influenced others in
their network to also participate despite their individual reservations. On the other hand,
some groups (networks) collectively chose not to respond after consulting with each other
and deciding, for example, that the risk of transferring private information over the internet
was too great. This was revealed by some of the respondents through their e-mail
communication. A “contagion effect”, therefore, appears to have worked in determining
both participation and non-participation.
5.4 Concluding Remarks
Improvements can be made to the design of the survey instrument in order to
increase the response rate, eliminate mistakes in data processing and improve the content to
ease data analysis and the relevance for policy analysis. The use of a single web page for
the full survey, for example, resulted in several technical difficulties. The first of these was
that the respondent had to scroll down the page to proceed with the survey. Since the
survey was rather long, scrolling also increased the possibility of questions being skipped.
This produced more “non-responses” than would have been the case if, instead of scrolling
down, the respondent could have simply gone on to a new section of the questionnaire by
submitting her answers to the previous section. One disadvantage of this alternative
strategy is that it requires frequent interactions with the server and may exacerbate server
traffic. Another difficulty experienced with the single page option was that during periods
85
of heavy METU server traffic, some respondents complained of not being able to download
the complete web page, which of course meant not being able to submit their answers by
clicking on the submit button located at the bottom of the page.
Another possible improvement that can be made in the survey design is to provide
a means for saving answered questions for future reference so that, if they need to,
respondents can complete survey at a more appropriate time, instead of filling out the
survey all in one shot. Errors in filling out the form may also be reduced by allowing
respondents to save and review their responses before submitting the form. In terms of
alleviating fears about sending private information over the internet, a further improvement
in survey design would be to provide password entry to the survey form. The password
would be unique to each participant and be provided in the cover e-mail. This would not
only prevent respondents from submitting private information over the internet, but allow
the investigators to identify legitimate participants.
As summarized in Chapter Three, many different factors have been provided as
explanations of skilled migration or student non-return, where each explanation addresses a
different aspect of the brain drain phenomenon. Specific questions on on-the-job training
and formal specialized training were asked in the professionals questionnaire in order to test
whether the Chen-Su model had some validity for skilled individuals from Turkey.
However, the sole purpose of the survey was not only to test these theories, but also to
provide an exploratory analysis of the determinants of return intentions. The next chapter
provides a summary of the qualitative characteristics of the respondents as well as
exploratory data analysis of the determinants of return intentions using categorical data
techniques.
86
Figure 5.2a Homepage of the Brain Drain Survey (English Version)
Figure 5.2b Homepage of the Brain Drain Survey (Turkish Version)
87
Figure 5.3 Turkish Brain Drain Student Survey Sample Web Page – Top
88
Figure 5.4 Turkish Brain Drain Student Survey Sample Web Page – Middle
89
Figure 5.5 Turkish Brain Drain Student Survey Sample Web Page – End
90
Figure 5.6 “Thank You!” Page
91
CHAPTER 6
SURVEY RESULTS AND PRELIMINARY DATA ANALYSIS
6.1 Introduction
Chapter Six presents the survey results and provides a preliminary, exploratory
analysis of the data. Given the characteristics of the two survey groups, the sample may not
be truly representative of the total population of overseas Turkish students and
professionals abroad for the period of the survey. However, the volume and diversity of the
responses received have been tremendously important for gaining insight into why Turkish
students studying abroad and Turkish professionals working outside Turkey are not
returning. The results indicate that family considerations play a prominent role in shaping
return intentions for the two groups, while there is some variation in the reasons for going
overseas and in the relative importance given to various push and pull factors.
Section 6.2 provides a summary of the characteristics of respondents and compares
the response patterns of participants in the student and professionals surveys. Sections 6.3
and 6.4 give separate, more detailed analyses of respondents taking part in the student and
professionals surveys respectively, which serves as a guide to interpreting the results of the
empirical investigation on return intentions presented in Chapter Seven. Simple bivariate
analysis is used to identify significant relationships among the background characteristics
of respondents and return intentions. The relationships that are found to be significant
through this analysis form the set of regressors in the empirical model of return intentions.
6.2 Respondent Profiles
In this section, the characteristics of respondents are compared under various
headings. These include: demographic characteristics, such as age, gender and marital
status; educational background; parental education levels; country of residence; stay
92
duration; initial reasons for going abroad; initial and current intentions about returning to
Turkey; family support; general assessments about various aspects of life in current country
of residence versus in Turkey; and the respondents’ evaluations of various push and pull
factors that may affect their decision to return or stay.
6.2.1 Age, Gender and Marital Status
The respondents are predominantly male, although the share of females is greater for
the student survey (38.7 versus 28.2 percent). The student survey comprises a younger
group of individuals, and this may explain the greater number of female respondents.
Traditionally, educational and migration opportunities have been greater for men than for
women. These prospects are slowly changing as reflected in generational differences in the
educational and career opportunities available for women in Turkey. Women currently have
greater options for pursuing overseas studies and overseas careers than they had previously.
The fact that female respondents in the professionals survey are, on average, younger than
their male counterparts also appears to corroborate this (Table A.1, Appendix A).
Nearly three-quarters of student respondents are single compared to only two-fifths
in the professionals survey (Table A.2, Appendix A). This is to be expected given the
younger profile of the student respondents. Of those who are married in the professionals
sample, more than a quarter are married to a foreign spouse suggesting that family
considerations may play a prominent role in their return intentions.
6.2.2 Stay Duration and Country of Residence
Slightly more than half of females (55 percent) in the professionals survey have
stayed in their current country of residence for five years or less, while the same share for
males is only 43 percent (Table A.3, Appendix A). A third of respondents for the total
professionals group have a stay duration of between 6 and 15 years. The sample is therefore
tilted toward those with shorter stays. In the student group, there is no significant difference
in the duration of stay among males and females, the majority having a stay duration of
between 3 and 4 years.
The majority of survey respondents are residing in North America. This is due to the
considerable amount of effort spent in collecting e-mail addresses from the United States
93
and Canada (Table A.4, Appendix A). A greater range of countries is represented in the
professionals survey, including countries in Europe, Asia, Africa and Australia, which may
reflect the possibility that overseas study options are more limited than international work
opportunities. However, the sample is not a true reflection of the actual number of Turkish
students studying at foreign universities. Germany is by far the largest recipient of students
from Turkey with an estimated number ranging between 30,000 and 40,000; Germany is,
therefore, severely under-represented. Recent years have also shown an increase in
enrollments at universities in nearby countries, such as Bulgaria and the Turkic Republics
in Central Asia. Again these countries are not represented given the focus on North
America in the data collection period.
6.2.3 Parental Education Levels and Parental Occupations
Parental educational attainment levels are used as the main indicators of socioeconomic status. Tables 6.1 and 6.2 present the breakdown of parental educational
attainment levels by gender and survey type. Respondents’ parents are, in general, highly
educated; two-fifths of mothers and more than two-thirds of fathers in the two groups hold
a tertiary level degree, which provides confirmation for the existence of unequal
opportunities in education. The average years of schooling for Turkey’s 25 years of age and
older population in 2000 was 5.7 years1, which corresponds to a little above the primary
level of schooling. From this, it appears that existing opportunities in education, both in
Turkey and abroad, are concentrated at higher socioeconomic levels.
The figures in Table 6.1 and Table 6.2 also reveal that the parents of female
respondents tend to be more educated than those of male respondents. While half of all
mothers of female students hold a tertiary level degree, the same is true for only two-fifths
of the mothers of male students. Similarly, while three-quarters of the fathers of female
students have a higher education degree, a little less than two-thirds of the fathers of male
students hold the same. These figures are slightly lower for the professionals group, but
show the same tendency: the parents of female respondents have greater educational
attainments than the parents of male respondents. This result is to be expected, since as
Tansel (2002a) has verified empirically, a stronger relationship exists between a girl’s
education and her parents’ education than for a boy’s and his parents’ in Turkey. In general,
1
Calculated from SIS (2003), Table 3.9, p. 51.
94
sons tend to be encouraged more than daughters to pursue educational opportunities or
goals, but this difference lessens as the socioeconomic position of the family increases.
Thus, it is expected that girls with more educated parents will be given more
encouragement to pursue higher education and for overseas studies.
Table 6.1 Respondents by Father’s Educational Attainment Level (%)
Education Level
Students
Professionals
Total
Male Female
Male Female
Below primary
2.6
2.7
2.6
3.2
0.6
Primary
11.1
14.7
7.6
11.7
7.4
Middle
4.5
3.9
3.3
5.4
5.3
High
13.6
13.0
11.6
15.0
13.9
Tertiary
68.0
65.7
75.0
64.5
72.6
Bachelors
42.7
42.0
48.8
42.4
37.5
Masters
14.0
13.9
14.2
11.9
19.5
...................................
Doctorate
11.3
9.8
12.0
10.2
15.6
Not known
0.1
...
...
0.2
0.3
n (missing excluded)
2265
662
424
840
339
Nonresponses
62
14
3
39
6
2
Test of Independence
(6) = 15.59**
2
(7) = 28.48***
Notes: ***p < 0.001, **p < 0.005, *p < 0.010; Cell percentages sum to 100 across columns.
Table 6.2 Respondents by Mother’s Educational Attainment Level (%)
Education Level
Students
Professionals
Total
Male
Female
Male Female
Below primary
8.3
10.3
3.8
10.6
4.7
Primary
17.3
19.2
13.7
19.2
13.6
Middle
7.3
5.7
6.1
9.6
6.5
High
26.4
24.1
25.4
27.0
30.4
Tertiary
40.6
40.7
51.1
33.5
44.8
Bachelors
30.7
29.9
38.4
27.0
30.4
Masters
5.9
6.0
8.2
4.2
7.4
Doctorate
4.0
4.8
4.5
2.7
4.7
Not known
0.0
...
...
0.1
0.0
n (missing excluded)
2276
668
425
844
339
Nonresponses
51
8
2
35
6
2
Test of Independence
(6) = 26.80***
2
(7) = 28.70***
Notes: ***p < 0.001, **p < 0.005, *p < 0.010; Cell percentages sum to 100 across columns.
The breakdown of parents’ occupations for each of the groups, students and
professionals, also confirms the above findings (Table A.5 and Table A.6, Appendix A).
Half of all fathers are in the “scientific, technical and related” professions, where 16 percent
fall into the “architects, engineers and related professionals” category and 11 percent are
“legal, business or public service professionals”. On the other hand, half of all mothers are
95
homemakers, and a little over a third are in the “scientific, technical and related”
professions. There are fewer engineers and architects and a greater proportion of health
professionals and teachers, at all levels, among the mothers of respondents reflecting
differences in both career opportunities and preferences. In 2000, the share of the scientific,
technical and related workers in the economically active population in Turkey was 7.5
percent for males and 6.9 percent for females. The same figures for the respondents’
parents are well above the average for Turkey: 50.4 percent for fathers and 34.8 percent for
mothers.
6.2.4 Bachelor’s Degree Institutions and Fields of Study
Since many of the students responding to the survey are postgraduate students, a
majority of them hold a bachelor’s degree. In both the students and professionals groups,
about a third of respondents are graduates of Middle East Technical University (METU)2.
This is followed by universities such as Bo aziçi, Bilkent, stanbul Technical, stanbul,
Ankara and Hacettepe Universities (Figures 6.1 and 6.2; see Table A.8, Appendix A for the
full list). These universities count among the more prestigious higher education institutions
in Turkey. The higher share of graduates from universities that have English language
instruction, such as METU, Bo aziçi and Bilkent, is perhaps not surprising since previous
exposure to a foreign language makes the transition to a foreign country easier. Foreign
language instruction starting from high school and sometimes even earlier in Turkey is
considered be an important catalyst in facilitating adaptation to a new environment and thus
non-return. Indeed, more than half of all respondents in the two groups have graduated
from high schools with foreign language instruction (Table A.7, Appendix A). The
remaining respondents are graduates of other universities in Turkey and various universities
abroad, each of which constitutes less than three percent of the share of graduates. It is
important to note that an important share of respondents hold a foreign undergraduate
degree (11.5 percent for professionals and 3.6 percent for students, not including those
currently in an undergraduate program), indicating early exposure to a new environment.
2
The relatively higher share of METU graduates in the total raises the question of whether there may
be a response bias because of the survey’s affiliation with Middle East Technical University.
96
Yildiz Teknik 1.6%
other 11.4%
ODTÜ (METU)
32.0%
Marmara 2.8%
Hacettepe 3.4%
Foreign Univ.
3.5%
Ankara 4.1%
Bo aziçi 17.0%
stanbul 4.3%
stanbul Teknik
6.2%
Bilkent 10.9%
Figure 6.1 Bachelor’s Degree Institution of Turkish Students Abroad (n = 967)
Notes: The total number of bachelor’s degree holders is 993; There are 26 missing responses.
Orta Do u Teknik
33.5%
Bo aziçi
16.9%
Other
6.5%
Foreign
University
11.5%
Ege
1.8%
Marmara
1.9%
Ankara
2.4%
Hacettepe
4.2%
stanbul
4.3%
Bilkent
5.8%
stanbul Teknik
11.2%
Figure 6.2 Bachelor’s Degree Institution of Turkish Workforce Abroad (n = 1223)
Note: There is one missing response.
97
The clear majority in both groups hold an undergraduate degree in engineering and
technical sciences (Detailed undergraduate fields are listed in Appendix A, Tables A.9 and
A.10). This share is significantly higher for male respondents than for females (62 percent
versus 35 for student respondents, and 71 percent versus 44 for professionals). The greater
share of engineering and related sciences majors in the professionals survey may be due to
the greater demand for technical manpower in the United States and worldwide. The
economic and administrative sciences discipline comprises the next highest category of
majors for both groups. The share of females in this category in comparison to the other
categories is significantly higher in both survey groups (Table 6.3).
Table 6.3 Bachelor’s Degree Disciplines by Survey Group and Gender (%)
Students
Bachelor’s Degree Discipline
Total
Male
Female
Architecture and Urban Planning ..........
2.5
1.0
2.6
Economic and Administrative Sciences .
17.6
13.6
25.5
4.9
8.1
Educational Sciences ..............................
3.4
Engineering and Technical Sciences .......
57.9
61.9
34.9
Language and Literature .........................
1.0
1.2
2.1
Math and Natural Sciences .....................
9.2
12.2
14.1
Medical and Health Sciences ..................
3.7
2.5
3.4
Social Sciences .......................................
4.2
2.5
8.6
0.3
0.8
Arts .........................................................
0.4
n (valid responses)
Nonresponses
2206
11
598
7
2
Test of Independence
Professionals
Male
Female
2.4
5.5
13.2
27.0
0.6
3.2
70.9
43.8
0.2
1.7
5.6
7.8
4.9
3.2
2.2
7.3
0.1
0.6
384
4
( 8) = 80.37***
879
0
2
345
0
( 8) = 108.97***
Notes: ***p < 0.001, **p < 0.005, *p < 0.010; Cell percentages sum to 100 across columns; Missing responses
are not reflected in the percentages.
6.2.5 Reasons for Going Overseas
Respondents were asked to choose, from a pre-determined list, a set of factors that
were important for their initial decision to study or work overseas. There were significant
differences among students and professionals in their reasons for the initial decision to go
abroad (Table 6.4). For both groups, foreign education is associated with greater prestige
and opportunities, and is in itself an important motivation. For professionals, the prestige or
opportunities associated with acquiring foreign education ties with the need for change as
the most often marked reason for going abroad (43 percent). Lifestyle and family factors
appear to carry somewhat greater importance for professionals relative to students. Among
students, factors related to the study program, facilities and research opportunities, and the
98
desire to improve language skills appear to be of greater relevance. Surprisingly perhaps—
given the bleak employment outlook in Turkey for the tertiary-educated workforce
following the economic crises—“not being able to find a job in Turkey” was chosen by less
than 10 percent of respondents in each group as their reason for going abroad, although the
proportion of students for which unemployment played an important role in leaving Turkey
is significantly greater than that of professionals. This is probably, in part, a reflection of
the unemployment problem facing recent university graduates in Turkey.
Table 6.4 Reasons for Going Abroad by Survey Type (%)
Professionals Students
(n = 1210) (n = 1102)
Reason
Learn language, improve language skills
Need change, experience new culture
Job requirement in Turkey
Could not find employment in Turkey
No program in specialization in Turkey
Insufficient facilities, equipment for research in Turkey
Prestige and advantages of study abroad
Lifestyle preference
To be with spouse, family
Provide better environment for children
Get away from political environment
Other
18.2
43.4
21.7
5.2
9.8
26.2
43.3
33.5
13.1
17.8
31.2
24.6
26.4
48.4
40.3
7.6
16.8
44.0
70.9
23.9
7.9
7.5
25.8
12.7
2
(1)
22.66
5.76
93.54
5.65
24.42
80.77
178.26
25.90
16.69
53.83
8.44
52.76
***
**
***
**
***
***
***
***
***
***
***
***
Notes: ***p < 0.001, **p < 0.005, *p < 0.010; Cell percentages reflect the number of positive responses for
each item out of the total number of valid/nonmissing responses (n).
Table 6.5 Reasons for Going Abroad by Gender (%)
Females
(n = 770)
Reason
Learn language, improve language skills
Need change, experience new culture
Job requirement in Turkey
Could not find employment in Turkey
No program in specialization in Turkey
Insufficient facilities, equipment for research in Turkey
Prestige and advantages of study abroad
Lifestyle preference
To be with spouse, family
Provide better environment for children
Get away from political environment
Other
19.1
46.9
31.3
6.4
13.9
30.5
58.4
29.1
19.1
8.4
19.9
17.1
Males
(n = 1542)
23.6
45.2
30.2
6.4
12.8
36.8
55.5
28.8
6.4
15.1
33.0
19.8
2
(1)
6.08
0.59
0.28
0.00
0.56
8.86
1.87
0.02
86.72
20.34
43.39
2.33
**
***
***
***
***
Notes: ***p < 0.001, **p < 0.005, *p < 0.010; Cell percentages reflect the number of positive responses for
each item out of the total number of valid/nonmissing responses (n).
99
Table 6.5 presents the gender differences in the initial reasons for going abroad. In
general, males and females have responded similarly to this question. More than half of all
respondents, both male and female, have marked “the prestige and advantages of overseas
study” as an important reason for going abroad. Both groups have also chosen “the need for
change”, “requirement in Turkey” and “lifestyle preference” in nearly equal proportions as
important reasons for going abroad. There are several factors that appear to be significantly
different among males and females. A greater proportion of female participants are
influenced by family constraints: “being with or near their families” was an important
reason for one-fifth of female participants. For male participants, on the other hand,
learning a new language, the lack of facilities and resources for research in Turkey, the
desire to provide a better environment for children, and political considerations were
chosen more often as reasons for going abroad.
Respondents in each survey were also asked to choose the most important reason for
their initial decision to pursue international education or employment opportunities (see
Figure 6.3 and Figure 6.4). Taking advantage of educational opportunities was selected as
the most important reason by many respondents, because many believe that international
study programs offer higher quality education in their chosen field of study compared to
universities in Turkey. Thus, one-sixth of professionals and one-fourth of students chose
“the prestige and advantages associated with study abroad” as the most important reason for
going abroad.
For professionals, this was followed by “other” reasons, the need for change, lifestyle
preference, and the lack of facilities and necessary equipment for carrying out research in
Turkey (Figure 6.3); For students, insufficient facilities, overseas experience being a job
requirement in Turkey, the need for a change, the lack of an academic program in the
respondent’s specialization, and lifestyle preference were the next most popular “most
important reasons” (Figure 6.4).
Some of the participants did not feel that the categories presented to them adequately
represented their reasons for going, and a substantial number of respondents in both survey
groups chose the “other” category (13 percent of professionals and 5 percent of students).
The “other” reasons included: gaining international work experience / global business
vision; being part of an inter-company transfer; being invited by the foreign country
employer; being frustrated with corruption in Turkey and wanting to be part of a more
100
18.0
Prestige and advantages of study abroad
12.6
Other
12.0
Need change, experience new culture
11.7
Lifestyle preference
10.3
Insufficient facilities, equipment for research
7.8
Job requirement in T urkey
7.1
Get away from political environment
T o be with spouse, family
6.5
Provide better environment for children
6.3
3.4
No program in specialization in T urkey
2.4
Could not find employment in T urkey
1.8
Learn language, improve language skills
0
5
10
15
20
Figure 6.3 Most Important Reason for Going Abroad, Professionals (%)
Notes: Respondents were asked to choose the most important factor. There are 28
nonresponses; (n = 1196).
26.6
Prestige and advantages of study abroad
17.7
Insufficient facilities, equipment for research
14.6
Job requirement in T urkey
10.1
Need change, experience new culture
7.5
No program in specialization in T urkey
5.9
Lifestyle preference
5.3
Other
3.6
Provide better environment for children
Could not find employment in T urkey
2.8
Get away from political environment
2.5
Learn language, improve language skills
1.9
T o be with spouse, family
1.6
0
5
10
15
20
25
30
Figure 6.4 Most Important Reason for Going Abroad, Students (%)
Notes: Respondents were asked to choose the most important factor; There are 17
nonresponses; (n = 1086).
101
professional work environment; to postpone / delay / shorten the military service
obligation; to get an “acceptable” doctorate; the belief that little value is placed on science /
technology / knowledge / academics in Turkey; to be able to use the latest technology not
available in Europe; disagreements, etc. with the Higher Education Council in Turkey; to
work with and learn from the best in their chosen field of specialization; more opportunities
for international recognition and mobility, higher quality undergraduate and post-graduate
education; political and social disorder in Turkey prior to 1980; and wanting to be in an
economically stable country. While some of these reasons are similar in spirit to the
categories presented in the survey, they provide somewhat more detailed explanations for
why participants have chosen to go abroad. Below is a sample of some of the explanations
in the participants’ own words:
At the university I worked in Turkey, research opportunities and support were very
insufficient, and the overall atmosphere was negative for scholarly activities.
[I left because of the] lack of organization and planning in Turkey, having to struggle
with daily things, lack of trust in people and institutions, [and] lack of optimism for the
future in Turkey.
It was difficult to get an academic job in Turkey, so I decided to study in the US.
METU [Middle East Technical University] would not let me teach as Assistant
Professor and wanted me to do a second dissertation for Associate.
Bogazici [University] requires a PhD from abroad to employ as an assistant professor.
At the time I wanted to be a professor at Bogazici University and thought that I needed
a PhD from the USA for that.
Working environment in Turkey is simply not professional, and very political.
[I left in order] to stay on the technical track (it’s impossible to work as an engineer
and survive in Turkey).
I had no career prospects in Turkey’s bleak technology sector.
Most of the faculty had left Turkey due to [the] political atmosphere at the time,
leaving no qualified professors in the universities to advance my studies.
[I wanted to use] my existing skills more efficiently, [and be] able to use my creativity.
Some participants also viewed overseas experience as a personal challenge to grow
as individuals in the absence of “a family support structure”, and some as a way to discover
their “professional abilities and limitations, in a high paced, competitive, international
environment.” For respondents of the student survey, the opportunity to receive better
102
quality education and to get away from the stress of preparing for the nationwide university
placement exam (ÖSS) also figure in as important reasons. It is worth noting that many
respondents believe that they will have better employment opportunities in Turkey in terms
of both workplace quality and better positions if they acquire overseas study and work
experience.
Reasons for going abroad by study program: For students, the reasons for going
abroad may differ according to the academic program of study (Table 6.6). Lifestyle factors
and the prestige of study abroad are important for a greater proportion of students in
bachelor’s and master’s degree programs, while, not surprisingly, a significantly greater
proportion of PhD students and postdoctoral fellows have marked the lack of a program in
their specialization, and the lack of resources for research in Turkey as important reasons
for going abroad. For students pursuing a master’s degree, the need for change, fulfilling a
job requirement in Turkey, learning a new language and the inability to find employment
were marked proportionally more as important factors.
Table 6.6 Reasons for Going Abroad by Program of Study, Students (%)
Bachelors
Masters
PhD
Postdoc
(n = 119)
(n = 303)
(n = 625)
(n = 55)
Learn language, improve language skills
26.1
36.3
22.2
20.0
22.02
***
Need change, experience new culture
47.9
56.4
46.7
23.6
22.06
***
Job requirement in Turkey
30.3
47.5
39.0
36.4
12.33
***
Could not find employment in Turkey
5.0
12.5
5.4
10.9
16.61
***
No program in specialization in Turkey
9.2
13.2
19.5
21.8
11.97
***
Insufficient facilities, equipment for research
in Turkey
30.3
23.8
54.4
67.3
99.01
***
Prestige and advantages of study abroad
76.5
74.3
69.3
58.2
8.55
Lifestyle preference
35.3
26.4
21.1
16.4
13.92
***
6.7
12.2
6.2
5.5
10.79
**
Reason
To be with spouse, family
Provide better environment for children
2
(3)
6.7
7.9
7.2
10.9
1.18
Get away from political environment
24.4
26.7
25.0
32.7
1.87
Other
20.2
11.9
11.7
12.7
6.75
**
*
Notes: ***p < 0.001, **p < 0.005, *p < 0.010; There is one missing response; Cell percentages reflect the
number of positive responses for each item out of the total number of valid/nonmissing responses (n).
Table 6.7a below presents the factors chosen as the most important reasons for going
abroad, broken down by academic program. Close to half of those enrolled in a bachelor’s
program abroad indicate that the most important reason for their decision to study in a
103
foreign country is “prestige or better quality education”. This is followed by “lifestyle
preference” and “other factors”. At the master’s level, close to a third of respondents
indicate that “prestige and better quality education” is the most important factor in their
decision to pursue a degree abroad, followed by “requirement in Turkey”, and the “need for
change / learn a new culture”. On the other hand, a good proportion of PhD students and
postdoctoral scholars have chosen the lack of facilities and resources necessary for research
in their field of specialization as the most important reason.
Table 6.7a Top Reasons for Going Abroad by Academic Program,
Students
Program Type
%
bachelors (n = 118)
Prestige and advantages of study abroad
Lifestyle preference
Other
44.9
10.2
8.5
masters (n = 300)
Prestige and advantages of study abroad
Job requirement in Turkey
Need change, experience new culture
29.7
14.7
12.3
doctorate (n = 614)
Insufficient facilities, equipment for research in Turkey
Prestige and advantages of study abroad
Job requirement in Turkey
24.1
23.5
16.3
postdoc (n = 54)
Insufficient facilities, equipment for research in Turkey
Job requirement in Turkey
No program in specialization in Turkey
37.0
14.8
9.3
Notes: 1086 out of 1103 participants responded to this question; n is the
number of valid responses.
The top three reasons for going abroad for professionals are listed in Table 6.7b
according to the highest degree completed. Although there is substantial variation among
the respondents in their reasons for going abroad, the top three reasons nevertheless account
for about half of all respondents in each category. The need for change and lifestyle factors
are given greater importance by bachelor’s and master’s degree holders, while those with
doctorate degrees give importance to research-related factors. These findings indicate that
the initial purpose or factors that are important for deciding to study or work overseas differ
according to level of specialization in higher education and in terms of gender. Female
respondents are more constrained by family considerations, while bachelor’s and master’s
degree holders are motivated to a greater degree by lifestyle preferences.
104
Table 6.7b Top Reasons for Going Abroad by Highest Degree,
Professionals
Highest Degree
%
bachelors (n = 266)
Need change, experience new culture
Lifestyle preference
Other
20.7
13.9
10.9
masters (n = 489)
Prestige and advantages of study abroad
Need change, experience new culture
Lifestyle preference
21.3
13.3
12.9
doctorate (n = 441)
Prestige and advantages of study abroad
Insufficient facilities, equipment for research in Turkey
Other
19.3
18.6
15.2
Notes: 1196 out of 1224 participants responded to this question; n is the
number of valid responses.
6.2.6 Family Support
Two-thirds of respondents have indicated that their families were “very supportive”
in the initial decision to study abroad, while less than 10 percent indicated that they were
“not very supportive” or “not at all supportive” (Table 6.8). This proportion is higher for
the student group than for professionals, possibly reflecting generational decisions in family
support. When asked whether their family would support them in the decision to settle
permanently outside Turkey, less than a third indicated that their family “would definitely
support” them, while one quarter of respondents believed that they “would most likely
support” them. This indicates that more than half of the respondents think that their family
would “definitely” or “most likely” support their decision to settle abroad, while only 20
percent indicate that their family “would not be very supportive” or “would actively
discourage them”. While family support is lower for the decision to settle permanently
outside Turkey compared to that for the decision to study abroad, it is still quite high. This
may be a reflection of the current economic circumstances in Turkey and parents’ desire for
their children to have a “better future”.
105
Table 6.8 Family Support for the Decision to Study or Work Overseas and for Settling
Abroad Permanently (%)
Support for the Decision
Survey Type
Gender
to Study or Work Overseas
Total
Professionals Students
Male Female
Not at all supportive
Not very supportive
Somewhat supportive
Very supportive
n (valid responses)
Nonresponses
2.2
5.8
24.9
67.1
2260
67
3.3
7.3
29.6
59.8
1176
48
2
Test of independence
Support for Settling
Abroad Permanently
Total
Discourage
Not very supportive
Somewhat supportive
Most likely support
Definitely support
Not sure
n (valid responses)
Nonresponses
6.6
12.4
24.7
24.9
28.3
3.2
2247
80
1.0
4.1
19.8
75.1
1084
19
(3) = 65.15***
2
Survey Type
Professionals
Students
6.3
11.1
27.2
23.6
29.5
2.3
1160
64
2
Test of independence
(5) = 18.18**
2.1
5.8
25.5
66.6
1508
47
6.9
13.8
22.0
26.2
27.0
4.1
1087
16
2.4
5.6
23.8
68.2
752
20
(3) = 0.98
Gender
Male Female
6.7
12.7
23.7
24.9
28.3
3.7
1499
56
2
6.3
11.8
26.6
24.9
28.2
2.3
748
24
(5) = 5.31
Notes: ***p < 0.001, **p < 0.005, *p < 0.010; Cell percentages sum to 100 across columns. Missing
responses are not reflected in the percentages.
6.2.7 Initial and Current Return Intentions
Initial return intentions at the outset of the stay in a foreign country may have some
explanatory power for the subsequent decision to migrate or return. The combined results
for the two groups are given in Table 6.9. Half of all respondents have indicated that their
initial intention was to return. Only about one-tenth have indicated they left Turkey without
the intention of returning, while an important proportion (more than one-third) was
undecided about whether to return. There is no significant difference in initial return
intentions between professionals and students, or between males and females.
The categories for current return intentions differ slightly for students compared to
professionals. About a quarter of the respondents taking part in the professionals survey
have indicated that they have definite return intentions, while slightly more than a third are
less certain about returning. Another third indicate that it is unlikely for them to return,
while about 7 percent say they will definitely not return. For students, there is a greater
106
tendency to indicate return intentions and a smaller proportion of student respondents have
strong non-return inclinations compared to professionals.
Table 6.9 Initial Return Intentions (%)
Initial Intentions
Return
Stay
Undecided
Professionals
(n = 1224)
51.6
12.0
36.4
Test of independence
2
Students
(n = 1103)
53.0
9.4
37.5
Male
(n = 1555)
51.6
11.3
37.1
2
(2) = 4.02
Female
(n = 772)
53.5
9.8
36.7
(2) = 1.32
Notes: ***p < 0.001, **p < 0.005, *p < 0.010 ; Cell percentages sum to 100 across columns.
Table 6.10 Initial and Current Return Intentions, Professionals (%)
Current Intentions
%
Return
54
272
416
401
81
4.4
22.2
34.0
32.8
6.6
83.3
74.3
51.7
36.7
27.2
14.8
23.2
43.3
42.9
28.4
1.9
2.6
5.1
20.5
44.4
1224
100.0
631
446
147
Definitely return, plans
Definitely return, no plans
Return probable
Return unlikely
Definitely not return
n
Initial Intentions
Undecided
Stay
n
2
Test of Independence
Measures of ordinal-ordinal association:
(8) = 232.16***
gamma = 0.5776; ASE = 0.032
Kendall’s tau-b = 0.3921; ASE = 0.024
Notes: ***p < 0.001, **p < 0.005, *p < 0.010; Cell percentages sum to 100 across rows; ASE refers to
the asymptotic standard error.
Table 6.11 Initial and Current Return Intentions, Students (%)
Current Intentions
Return without completing studies
Return immediately after studies
Definitely return, but not soon after
Return probable
Return unlikely
Definitely not return
n
Initial Intentions
Return Undecided
Stay
n
%
13
149
389
308
211
33
1.2
13.5
35.3
27.9
19.1
3.0
61.5
82.6
67.6
42.5
27.0
9.1
1103 100.0
585
38.5
15.4
29.1
51.6
52.6
9.1
414
2
Test of Independence
Measures of ordinal-ordinal association:
0.0
2.0
3.3
5.8
20.4
81.8
104
***
(10) = 388.25
gamma = 0.5776; ASE = 0.032
Kendall’s tau-b = 0.3921; ASE = 0.024
Notes: ***p < 0.001, **p < 0.005, *p < 0.010; Cell percentages sum to 100 across rows; ASE refers to
the asymptotic standard error.
107
The relationship between initial and current return intentions is presented for
professionals and students in Tables 6.10 and 6.11 respectively. According to the gamma
and Kendall’s tau-b statistics—two measures of ordinal-ordinal association (Agresti,
1984)—a strong, positive relationship exists between initial and current return intentions.
This is also evident from examining the percentages in the tables: current return intentions
are more likely to be in favor of remaining abroad when initial intentions are also to stay.
6.2.8 Reasons for Returning and the Time Frame of Return
While a majority of respondents (61% of professionals and 88% of students) have
indicated that they intend to return to Turkey, it appears that many of them do not have
short term return plans: half of professionals and 41 percent of students intend to return in
five years or more. About one third have indicated they will return within 2-5 years, while
another third intends to return within 5-10 years. A significant proportion of professionals
(18%) plan to return after 10 years, which is much higher than the proportion for students
(8%). On the other hand, a greater percentage of students have immediate return plans
(11%) compared to professionals (6%). There are no significant differences in the predicted
return dates of male and female respondents.
Table 6.12 Predicted Return Dates for Respondents with Return Intentions (%)
Professionals Students
Females
Males
(n = 699)
(n = 827)
(n = 490) (n = 1036)
within 6 months
6 - 12 months
1 - 2 years
2 - 5 years
5 - 10 years
over 10 years
2.9
2.7
10.6
33.6
32.1
18.2
2
Test of significance
***
**
5.7
5.0
13.1
35.0
33.7
7.6
(5) = 48.04***
4.1
3.9
13.5
36.3
31.6
10.6
2
4.5
4.0
11.2
33.4
33.6
13.3
(5) = 4.78
*
Notes: p < 0.001, p < 0.005, p < 0.010; Cell percentages sum to 100 across columns. n
indicates total valid responses; there is a total of 85 missing responses (43 for the professionals
survey and 32 for the student survey).
Figures 6.5 and 6.6 present the return reasons for professionals and students,
respectively. More than half of respondents marked “missing family” as an important
reason for returning in the professionals survey. Achievement of specific goals was also an
important reason for nearly as many respondents, while achieving career goals, retirement
108
and “other” reasons were the next most often marked options. Three quarters of students
who indicated that they will be returning indicated that reaching specific goals, such as
56.5
M iss family
Achieved specific goals (e.g., work exp.)
55.4
Achieved career goal
42.3
39.7
Retirement
Other
38.3
Achieved savings goal
32.1
22.5
Childrens' education
Job opportunity in Turkey
10.6
Complete military service
8.7
5.9
Lack of safety in current envr.
Expiry of overseas job contract
Complete university service
3.3
1.0
0.0
10.0
20.0
30.0
40.0
50.0
60.0
Figure 6.5 Return Reasons for Turkish Professionals (%), n = 728
Notes: Respondents were asked to mark all valid choices; there are 14 nonresponses.
72.3
Achieved specific goals (e.g., work exp.)
61.6
M iss family
Childrens' education
23.1
Other
16.8
Complete military service
16.3
Complete university service
15.1
Achieved career goal
13.2
Job opportunity in Turkey
12.0
10.6
Achieved savings goal
7.2
Do not feel safe in current envir.
Retirement
0.0
Expiry of overseas job contract
0.0
0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0
Figure 6.6 Return Reasons for Turkish Students (%), n = 847
Notes: Respondents were asked to mark all valid choices. The options “achievement of career goal” and
“achievement of savings goal” were not available in the English version of the student survey. As a
result, the percentages for these categories are valid only for students who answered the Turkish version
of the survey (n = 269).
109
work experience, was an important return reason3. This was followed by “missing their
family while abroad” (62 percent), and the desire to have their children educated in Turkey
(23 percent). Return reasons do not show significant variation between male and female
respondents, although for male respondents “military duty” is another important reason for
returning.
Table 6.13 presents the future overseas stay plans for those who have indicated
they will return to Turkey. In general, students appear to have plans for longer term stays
(e.g., greater than 2 years) as well as being slightly more inclined toward permanent
settlement abroad. There are, however, a significant number of respondents in each group
who plan short term stays of up to three months at most. This category of stay length
constitutes one-quarter of professionals and one-fifth of students. This indicates that
“migration” is not a once and for all decision for a great number of participants. Frequent
visits back and forth or to third countries may provide a more realistic view of the career
trajectories of many highly skilled individuals from Turkey. There are a significant number
of academicians and employees of multinational firms in the professionals survey, which
means that frequent international travel may be expected from this group.
Table 6.13 Future Plans for Overseas Stay, by Survey Type and Gender (%)
Predicted Length of Future Stay
Professionals Students
(n = 631)
(n = 737)
Females
(n = 433)
Males
(n = 935)
Few days to several weeks at most
1-3 months at most
4-6 months at most
7-12 months at most
1-2 years at most
Longer than 2 years, but will definitely return
Permanent settlement
No plans for future overseas travel, etc.
18.7
26.8
12.8
2.7
5.6
23.0
1.1
9.4
18.6
20.5
7.3
5.6
10.7
28.5
2.7
6.1
17.1
23.6
8.8
3.9
7.9
31.0
1.9
6.0
19.4
23.3
10.4
4.4
8.6
23.6
2.0
8.3
Total
Nonresponses
742
111
859
122
517
84
1084
149
2
Test of significance:
***
**
(7) = 46.85***
2
(7) = 10.19
*
Notes: p < 0.001, p < 0.005, p < 0.010; Cell percentages sum to 100 across columns; n is the valid
number of responses.
3
This is a much higher figure than for professionals. This may be due to the fact that respondents to
the English survey were not given two related reasons—reaching career goals and reaching savings
goal—as options in the survey. It is, therefore, difficult to make comparisons between students and
professionals for this particular question.
110
6.2.9 General Assessments of Study, Work, Social and Living Conditions
Respondents were asked to compare various aspects of their life in their current
country of residence with that in Turkey. A third indicated that social life was worse in
their current country of residence than in Turkey, while two-fifths felt it was neither better
nor worse. Less than one fifth of respondents indicated that their social environment was
“better” or “much better” than in Turkey. There is no significant difference in this
evaluation between professionals and students and between males and females.
In general, respondents believe that their standard of living was “better” or “much
better” in their current country of residence (80 percent). However, for students and for
females the share of those who felt that their standard of living was better than in Turkey
was somewhat lower: 70 percent for students compared to 88 percent for professionals, and
74 percent for females compared to 83 percent for males.
Table 6.14 Respondents’ General Assessment of Social Conditions in their Current
Country of Residence versus in Turkey (%)
Survey Type
Gender
Assessment
Total
Professionals
Students
Male
Female
Much worse
Worse
Neither better nor worse
Better
Much better
Do not know
n (excludes missing)
Nonresponses
10.5
33.0
39.3
8.8
7.9
0.5
2317
10
10.3
33.3
40.8
7.4
7.5
0.7
1218
6
2
Test of independence
10.7
32.6
37.7
10.3
8.5
0.4
1099
4
11.1
33.8
38.0
8.7
7.9
0.5
1546
9
2
(5) = 8.78
9.3
31.4
41.9
8.8
8.0
0.5
771
1
(5) = 4.31
Note: ***p < 0.001, **p < 0.005, *p < 0.010
Table 6.15 Respondents’ General Assessment of the Standard of Living in their
Current Country of Residence versus in Turkey (%)
Survey Type
Gender
Assessment
Total
Professionals
Students
Male
Female
Much worse
Worse
Neither better nor worse
Better
Much better
Do not know
n (excludes missing)
Nonresponses
1.2
4.2
14.6
26.1
53.7
0.3
2315
12
Test of independence
0.7
1.5
9.1
26.2
62.2
0.3
1217
7
2
(5) = 138.57***
Note: ***p < 0.001, **p < 0.005, *p < 0.010
111
1.7
7.2
20.7
26.0
44.3
0.2
1098
5
1.1
3.2
12.9
26.5
56.1
0.2
1545
10
2
1.3
6.2
17.9
25.2
49.0
0.4
770
2
(5) = 26.32***
Professionals were also asked to make a general assessment about their work
environment in their current country of residence in relation to that in Turkey (Table 6.16).
Similar to the living conditions assessment, the work conditions assessment is tilted toward
the “better” and “much better” categories. Those in academia and male respondents appear
slightly more satisfied with their work environments than female respondents and
respondents working in other types of organizations.
Table 6.16 Turkish Professionals’ General Assessment of Work Conditions
in their Current Country of Residence versus in Turkey (%)
Gender
Type of Organization
Assessment
Male Female
Academiaa
Other
Much worse
0.5
0.6
0.0
0.7
Worse
1.8
2.1
0.3
2.5
Neither better nor worse
9.0
10.3
7.5
10.1
Better
23.9
27.9
22.0
26.2
Much better
59.2
52.8
65.9
54.0
Do not know
5.6
6.5
4.3
6.5
Freq. (n)
871
341
346
866
Nonresponses
8
4
2
10
2
Test of independence
2
(5) = 4.21
(5) = 20.45***
Note: ***p < 0.001, **p < 0.005, *p < 0.010
a
Academia includes Universities, Research Centers and Medical Schools.
Table 6.17 Turkish Students’ General Assessment of Academic Conditions in their
Current Institution of Study versus in Turkey (%)
Gender
Academic Programa
Assessment
Male Female
Bachelors Masters Doctorate
Much worse
0.3
0.0
0.9
0.0
0.2
Worse
1.9
1.9
0.9
5.9
0.3
Neither better nor worse
10.4
8.2
3.4
17.5
7.1
Better
24.6
27.5
18.6
28.1
27.1
Much better
62.6
62.4
75.4
48.5
65.3
Do not know
0.2
0.0
0.9
0.0
0.0
Freq. (n)
671
425
118
303
620
Nonresponses
5
2
1
0
5
Test of independence
2
2
(5) = 4.09
Postdoc
0.0
0.0
7.3
12.7
80.0
0.0
55
1
(15) = 99.34***
Note: ***p < 0.001, **p < 0.005, *p < 0.010
a
Respondents in associates degree post-bachelors programs are included in bachelors program figures.
Respondents in post-masters programs are included in masters program figures.
Students, on the other hand, were asked to assess academic conditions at their current
institution (Table 6.17). A great majority (87 percent) indicated academic life to be “better”
or “much better” in their current country of study. A breakdown of student responses by
112
academic program indicates that students in master’s degree programs appear to be less
enthusiastic about academic conditions than those in bachelors, doctorate and postdoctorate programs. In future survey studies, more specific questions could be asked about
study and work conditions to pinpoint which aspects of their jobs or academic programs
that respondents are particularly dissatisfied with in Turkey.
6.2.10 Difficulties Abroad and Adjustment Factors
The difficulties faced while studying or working abroad may be interpreted as being
part of the “psychic” costs of moving to a new location. A list of potential difficulties was
presented to respondents, and they were asked to mark the difficulties that were significant
for them. Table 6.18 gives the results by gender and survey type. Four-fifths of
professionals and students marked “missing family members left behind in Turkey” as an
important difficulty. For females, this proportion was significantly higher (87 percent).
High cost of living was the next most often marked category for student respondents,
followed by lack of leisure time, and loneliness or being unable to adjust. For professionals,
lack of leisure time, children growing up in a foreign culture and high cost of living were,
Table 6.18 Difficulties Faced Abroad by Gender and Survey Type (%)
Gender
Survey Type
2
(1)
Difficulties
Female Male 2(1)
Profes. Students
Being away from family
Children growing up in a different
culture
87.2
78.5 25.59
***
80.9
11.4
15.5
7.03
***
21.8
Loneliness, not being able to adjust
Fast-paced life
Little or no leisure time
Unemployment
No jobs in my area of specialty
22.7
17.4
31.1
5.1
2.9
21.9
16.0
27.7
4.0
2.2
0.19
0.65
2.89
1.38
0.84
Discrimination against foreigners
Lower income than in Turkey
Higher Taxes
Crime, lack of personal security
High cost of living
Other
No difficulties experienced
12.1
5.1
11.5
5.1
29.8
16.3
0.7
16.1
2.7
11.6
5.0
28.5
14.1
1.7
6.35
8.55
0.01
0.01
0.43
2.04
4.17
n (valid responses)
766 1515
Notes:
***
**
*
*
***
***
**
0.38
5.5 125.74
***
18.6
12.8
25.7
3.9
2.1
26.2
20.6
32.3
4.9
2.9
19.21
24.7
12.33
1.34
1.48
***
16.2
1.8
15.3
4.1
21.7
13.2
2.0
13.2
5.5
7.4
6.1
36.9
16.6
0.6
3.83
23.18
34.79
4.90
64.76
5.07
9.56
**
1201
1080
p < 0.001, p < 0.005, p < 0.010; There are 46 missing observations.
113
81.9
***
***
***
***
**
***
**
***
in that order, shown to be important difficulties. Significant chi-square statistics in the
gender and survey groups indicate that there are important differences in the response
patterns between males and females and between students and professionals for different
types of difficulties faced while abroad. A greater proportion of male respondents are
concerned about their children growing up in a different culture, as well as discrimination
from foreigners compared to female respondents, while more females indicate that a lower
income level and having less leisure time abroad are significant difficulties.
Table 6.19 gives the most important difficulties for the same groups. The top three
difficulties do not change by gender or survey type, although the proportions with which
they are chosen are slightly different. Being away from or missing family is chosen by
more than half the respondents as the most important difficulty faced while abroad. This is
followed by loneliness / inability to adjust and “other” difficulties.
Table 6.19 Top Difficulties by Gender and Survey Type (%)
Gender
Difficulties
Female
Male
Being away from family
Children growing up in a different culture
Loneliness, not being able to adjust
Fast-paced life
Little or no leisure time
Unemployment
No jobs in my area of specialty
Discrimination against foreigners
Lower income than in Turkey
Higher Taxes
Crime, lack of personal security
High cost of living
Other
No difficulties experienced
n (valid responses)
2
Test of Independence:
65.0
1.3
8.9
1.5
4.2
2.0
1.3
2.6
0.4
0.4
0.0
4.2
7.5
0.7
52.7
5.2
9.7
2.5
5.6
1.3
0.4
4.1
0.5
1.4
0.9
5.4
8.8
1.7
757
1493
***
(10) = 388.25
Survey Type
Profes. Students
2
58.0
6.0
7.8
1.9
4.8
1.3
0.8
4.4
0.3
1.6
0.5
3.1
7.4
2.1
55.5
1.5
11.1
2.4
5.5
1.8
0.6
2.7
0.7
0.5
0.7
7.2
9.5
0.6
1187
1063
(13) = 82.60***
Notes: ***p < 0.001, **p < 0.005, *p < 0.010; There are 77 missing observations; Cell percentages
sum to 100 across columns.
Some of the most cited adjustment problems faced by the respondents were
communication difficulties related to language problems, cultural barriers, and having a
more limited social network compared to that in Turkey (e.g., lack of close social ties;
114
family ties not being as strong), as well as a lack of sense of community and feeling of
“belonging”, and hence alienation. A number of participants emphasized the difficulties of
having to adjust to small town life after living in a cosmopolitan city like stanbul, while
others were dissatisfied with the social and cultural lifestyle in their host countries:
[There is] no real quality of life socially and culturally. [There is] misfit with the
extreme individuality and selfishness of American society in general. Too much work,
too little leisure. Too isolated, systematic and cold.
We are from totally different historical and cultural backgrounds: We don’t laugh, cry,
or enjoy the same things.
It was a step down in my social status. In Turkey, I was a member of the privileged
group. Here I am a typical middle income.
There were two factors in particular that were important in terms of having an impact
(in terms of a significant bivariate association) on the return intentions of both students and
professionals: loneliness and missing family. For professionals, a high chi-square statistic
indicates that high cost of living and “other” factors also significantly affected return
intentions.
Figure 6.7 presents the adjustment factors marked by professionals and students as
being important in overcoming the difficulties of life abroad. A group of participants
indicated they had no adjustment problems; this group is represented in the “other”
category. It is also interesting to note that many of the adjustment factors have a negative
association with foreign high school language instruction. Some respondents explicitly
mention that they experienced little or no difficulty in adjusting to life outside Turkey
because of the foreign-language education they received in Turkey. For students, “time”
and having Turkish friends at current institute of study were the most often marked
adjustment factors, while for professionals, the presence of their spouse and having prior
overseas experience are also important. In Figure 6.8, the top adjustment factors are
depicted for each survey group. “Spouse”, “previous experience” and “time” are the top
three adjustment factors for professionals, while “time”, “having Turkish friends” and
“previous experience” are the top three for students.
115
Time
58.7
42.4
Turkish Friends
34.3
Previous Experience
22.9
Presence of Spouse / Loved One
Turkish Community
Other
Turkish Student Association
4.6
6.5
4.6
0.8
1.1
Turkish Internet Network
Cultural Attache / Embassy
0
10
47.4
56.4
43.4
43.1
17.9
19.0
17.0
16.5
12.0
20
30
40
Professionals
50
60
70
Students
Figure 6.7 Adjustment Factors by Survey Type (%)
Note: The sum of the percentages exceed 100 since more than one factor could be chosen.
Time
28.4
19.1
Turkish Friends
17.0
16.7
Previous Experience
22.3
21.7
14.7
Presence of Spouse / Loved One
25.8
11.7
12.0
Other
3.3
3.1
2.3
0.7
0.4
0.4
0.3
0.2
Turkish Community
Turkish Student Association
Turkish Internet Network
Cultural Attache / Embassy
0
5
10
15
Professionals
Figure 6.8 Top Adjustment Factors by Survey Type (%)
116
20
Students
25
30
Interestingly, a male respondent working in a multinational corporation indicated that
the difficulties faced abroad are just part of the price of having a more comfortable lifestyle:
Personally, I never want to return to Turkey. My main reasons are a better standard of
living, much better job satisfaction and more liberal way of life (especially here in
California). I believe the loneliness I feel (not too many close friends, no family
members) is the price I need to pay to have all the other stuff. In my company, 90% of
the engineers are first generation immigrants (from Russia to Brazil, China to Iran)
and they all told me about the loneliness. And just like them, I will endure the
loneliness to have a much better life.
6.2.11 Evaluation of Various Push and Pull Factors
“Push” factors are those characteristics or circumstances of the home country that
prompt a person to migrate to another country, while “pull” factors are the characteristics of
the receiving country that provide incentives for individuals to settle in the receiving
country. Economic factors or differences in income levels have been cited most often as
reasons for the loss of highly skilled workers in developing countries. Respondents were
asked to rank various “push” and “pull” factors on a five-point scale ranging from least
important “1” to most important “5”4 in terms of their relative significance in the decision
to remain abroad.
Table 6.20 gives the percentage of respondents marking the various push and pull
factors as “important” or “very important” in the professionals and student surveys.
Economic instability is the top push factor for both groups: 76 percent of students and 84
percent of professionals indicate that economic instability is an important reason for not
returning. This is not surprising since unemployment among high school and university
graduates reached nearly 30 percent in the aftermath of the February 2001 economic crisis
according to the State Institute of Statistics Household Survey results. For students,
economic instability is followed by low income levels (73.4 percent), little opportunities for
advancing in career (71.5 percent) and bureaucratic obstacles (71.3 percent). For
professionals, bureaucracy (79.4 percent), unsatisfactory income levels (68.4 percent),
political instability (64.7 percent) and lack of opportunities for advancing in occupation
(61.7 percent) were the next most often marked push factors. Less than a quarter of
respondents in both surveys chose an “unsatisfactory social and cultural life in Turkey” as
important push factors. Many of those who marked the “other” category included
4
It is technically a 6-point scale since items that are “not applicable” are given a score of “0”.
117
Table 6.20 Push and Pull Factors Viewed as Important‡ by Professionals and Students (%)
Professionals Students
(n = 1189) (n = 1095)
Push and Pull Factors
2
(1)
PUSHA:
PUSHB:
PUSHC:
PUSHD:
PUSHE:
PUSHF:
PUSHG:
PUSHH:
PUSHI:
PUSHJ:
PUSHK:
PUSHL:
Low income in occupation
Little opport. for advancement in occupation
Limited job opport. in specialty
No opportunity for advanced training
Away from research centers and advances
Lack of financial resources for business
Less than satisfying social/cultural life
Bureaucracy, inefficiencies
Political pressures, discord
Lack of social security
Economic instability
Other
68.4
61.7
53.0
36.1
39.5
29.1
24.6
79.4
64.7
59.0
83.7
11.9
73.4
71.5
58.6
57.8
58.7
34.4
22.9
71.3
58.3
51.3
76.2
8.1
7.03
24.82
7.36
108.17
84.04
7.48
0.84
20.08
9.91
13.74
20.20
8.72
PULLA:
PULLB:
PULLC:
PULLD:
PULLE:
PULLF:
PULLG:
PULLH:
PULLI:
PULLJ:
PULLK:
PULLL:
Higher salary or wage
Greater advancement oppr. in profession
Better work environment
Greater job availability in specializ.
Greater oppr. to develop specialty
More organized, ordered envir.
More satisfying social/cultural life
Proximity to research and innov. Centers
Spouse’s preference or job
Better educational opport. for children
Need to finish /continue with current project
Other
79.1
76.1
71.3
65.9
69.9
76.4
26.6
42.0
31.0
37.4
15.2
4.8
76.8
82.1
67.8
75.1
82.1
76.6
28.5
60.4
21.4
19.7
30.0
3.7
1.69
12.36
3.26
22.84
45.67
0.01
1.05
76.66
26.43
87.10
71.66
1.82
***
***
***
***
***
***
***
***
***
***
***
***
*
***
***
***
***
***
***
Notes: ***p < 0.001, **p < 0.005, *p < 0.010
‡
Marked as “Very Important” or “Important” by Respondents
corruption (bribery, partisanship, nepotism) and, in the case of male respondents,
compulsory military duty as important push factors.
The top pull factors for both groups complement these results. In the student sample,
the great majority of respondents have marked greater advancement in occupation (82
percent) and greater opportunities for developing their specialty (82 percent) as important
or very important pull factors in their host country. This is followed by higher salaries (76.8
percent), a more organized and ordered environment (76.6 percent), and greater job
availability in specialty (75.1). The emphasis on professional opportunities advancing in or
developing the field of specialization is not surprising given that the majority of students
are post-doctorate and PhD students. The majority of Turkish professionals, on the other
hand, indicate that a higher salary in the host country is an important pull factor (79.1
118
percent). Three-quarters also indicate that a more organized / ordered environment and
greater opportunities for advancement in occupation are very important pull factors.
One of the most common views expressed in the survey by those who have chosen
an academic career is that there is a lack of value given to science and to academics in
Turkey. Many respondents have indicated that, as a result of this, they fear they will find
themselves in an “unproductive environment” when they return. Others have stated that
“there is a point where money is no object” and that they would be willing to work for
lower wages in Turkey provided that they are “valued and respected”. The following
comments illustrate the dilemma faced by respondents contemplating return:
Everyone should realize [the] fact that we stay abroad because of the lack of scientific
advancements and economic instability in Turkey. Like the movie says, “If you build
it, they will come...” If the government/industry/institutions work together and build a
good structure, why should we work for another country? This is a close loop and the
good approximation is the “Chicken-egg” analogy. Which one comes first? Chicken?
or Egg? Should we build the structure first or should we come back without a good
structure? This is the main question! How much money I am making in this country or
how happy I am, these are all nonsense. How can you be happy when you are away
from your family, culture, and people?
I advise many Turkish students who work for their PhD, either with me or in my
institution, or field of work (Experimental Physics). My advice to them is to stay
rather than to return. [...] The research budget of Turkey is negligible compared to
many developed countries. That translates directly to the fact that there cannot be a
sustained, competitive, internationally recognized research programs in Turkish
institutions. Yet, this is precisely why young people spend 5-to-10 years extra after
their Bachelor's degree to get their PhD's. So in a way, returning is tantamount to
negating all of your hard work. Once the importance of original creative work is
understood, and appreciated by the society, and the required resource allocations are
made by the politicians, the situtation will remedy itself over a period of time, like a
decade.
Unfortunately, many respondents contemplating an academic career after completing
their studies abroad are hesitant about working in newly created state universities in
Turkey, even when they have a compulsory service requirement. Many believe the private
or foundation universities offer them better conditions.
Bu yaz Amerika'da doktora ö renimimi bitirdim. Türkiye’de bir üniversiteden burslu
olarak gelmi tim. Masterimi TR’nin bursu ile, doktoramı ABD üniversitesinin bursu
ile bitirdim. Ama TR ile ili kimi kesmedim. Bu yaz Türkiye'de burs aldı ım
üniversitenin rektörü ve bölüm ba kanım ile görü tüm. Amacım onların geri dönmem
konusunda ne dü ündüklerini ö renmek, bize sa lanacak imkanları görmek idi. Hem
rektör, hem de bölüm ba kanı bana gelmemin gereksiz oldugunu, dönmem halinde
bana sa layacakları hiç bir imkan olmadı ını, benim ABD’den onlara daha fazla
faydalı olaca ımı do rudan veya dolaylı olarak söylediler. Hatta bölüm ba kanı ...
119
bilgisayar verilip verilmeyece ini sordu umda, e er masa ve sandalye bulursam
kendimi san lı saymam eklinde cevabı çok ilginçti. Gerçekten de TR’ye dönmek çok
istiyorum. Ama devlet üniversitesine degil. Özel bir üniversiteye. (After finishing my
doctoral studies in the United States, I visited the university where I have a
compulsory service requirement and spoke with the department head and the rector. I
wanted to find out about what they thought about my returning and what kind of
opportunities they could offer me. I was told, both directly and implicitly, that there
was no reason why I should return, there were no opportunities they could offer me
and that I would be more useful to them if I stayed in the United States. When I asked
if they could provide a computer, the department head said I would be lucky if I could
find a chair and table. I really do want to return to Turkey. Not to a state university,
but a private one.)
You need to assess the importance of and contributions made by private universities in
Turkey. My main reason for wanting to return to Turkey is to join one of these
institutions. I have already contributed to Sabanci and Koc University programs.
Facilities provided in Turkish private universites are as good as abroad but they need
to be scrutinized by independent academic groups in order to maintain and enhance
quality of teaching and research.
While many academic participants would be willing to work in state universities with
established reputations, there is no guarantee that those who return will be employed in one
of these institutions.
As I had a firm belief of returning and giving back what was given to me by my
country after my PhD in 1975, I taught at ODTU in 1975-77, and Bogazici, 78-80. I
returned to USA because of political turmoil; moved to Sydney to join my partner in
1989. I am now an academic living abroad; in 1993, I came and presented myself to
ODTU and Bogazici; had I been offered a job, we would have moved back.. I still
maintain very close contact, and participate in training and development [activities].
Other respondents’ comments give more detailed explanations for why many of the
educated are choosing not to return to Turkey. It is usually a combination of factors that
keep professionals and students abroad. There are also generational differences in the
reasons for not returning. Below are some of these explanations as well as suggestions for
remedies.
I think the main factor [in not returning] is, lack of good jobs, lack of opportunities.
People move away and they get treated so much better professionally and they get
used to the salary and the opportunities other countries have to offer that they don't
consider going back. Why would you move back and take a job cut, a pay cut and
make your life more difficult. People move to make things better not worse.
My personal belief is that the most important reason is the business climate; and
mostly the lack of entrepreneurial culture. My school (METU), TUBITAK and others
[have spent] a lot of effort on technoparks, etc but nothing came out of them because
they are isolated efforts.
120
In the early years (1970s) terror in Turkey was the main factor causing us to stay in
[the] USA. Later on, political instability and lack of opportunities in our fields. But,
overall, government policies to encourage growth of private sector, especially in terms
of regulations, taxation, bureucracy, corruption kept us working in USA rather than
returning. Later on, after a year of living in Turkey, 1992-3, we decided to return to
USA since we had two elementary school children and we felt we could not get them
into acceptable middle education schools (özel okullar), and comparably we could find
better quality schools in USA for them.
Please add the mandatory military service as a reason to work abroad. For me, the
main reason [for continuing to live] in the States is the business environment (lack of
professional environment) and corruption.
Due to the fact I will not be able to find a job (a job close to this one) in Turkey, It will
not be easy to [return]. I design, analyze and construct and manage the wireless sites.
Türkiye’ye gitmek istemememdeki bir faktör de Türkiye’deki trafik. Ailemden 2 ki i
(annem ve teyzemin o lunu trafik kazasında kaybettim. 4 ki i ciddi ekilde yaralandı).
Ayrıca, sa lık hizmetlerinin kötü olması (hastanelerin durumu, ambülans sistemi vs),
insana de er verilmemesi, kanunların uygulanmaması, her eyin torpil ve tanıdıklar
vasıtasıyla yürütülmesi kendimi Türkiye’de güvende hissetmememe sebep oluyor.
Türkiye’ye dönsem bile orada bir eyi de i tiremeyece imi ve yeni bir ey
getiremeyece imi dü ünüyorum. (Traffic in Turkey is another factor that makes me
not want to return. I lost my mother and cousin in a traffic accident. In addition, the
poor health services (in terms of hospitals, ambulance system, etc), lack of value given
to human life, lack of law enforcement, and everything being done through nepotism
or other such connections add to my apprehension about being in Turkey. I feel that if
I return I won’t be able to change anything or bring anything new to Turkey).
I believe the most important factors of brainpower not returning to Turkey are: 1)
money and increased likelihood [for promoting] your career abroad 2) economic and
political stability and order abroad. However, the social environment and culture of
foreign countries are very different from that of Turkey, and most people I know
would return immediately if they knew the situation [was] more stable and predictable,
and that they knew they would be financially secure.
I think that the brain drain argument implies two things: First, what I know is not
known in Turkey; second, Turkey would be interested in implementing what I know.
Turkey has professionals who are very capable. However, the majority of Turkish
people and the governments are not listening to them. Under these circumstances,
what would be the contribution of a Turkish professional to Turkey, if she returned to
Turkey? Not much, I think.
I was planning to return to Turkey but ... the crisis in banking delayed my decision
again. Another main reason not to return is the education of my children. Each time
you decide to go back you remember the race they have to enter for their higher
education.
I think this is a great concern to Turkey and that there are no strategic planning to
recover any of the brain drain. While most of us would like to entertain the possibility
[of coming] back, even for lesser opportunities, there is no structure that creates
platforms for capturing the value of brains outside of Turkey. I would even say that
there is some resentment and/or resistance to such attempts.
121
In US you feel like you can really contribute to the society. [For] some reason I hardly
felt this in Turkey, the feeling of doing something really useful and making a
difference.
I think one thing we need to do to prevent the "brain drain" is to give a little hope and
inspiration to young people. With no hope for the future, no trust, and no opportunities
to make a difference or to speak up, stand up for what we believe, life comes pretty
much down to basics: food, shelter, etc. Unfortunately, on that scale, I am far better
rewarded for my efforts here than I would be in Turkey. So I make my decision based
primarily on that "quality of life" criterion. Sad and materialistic maybe, but true.
Anecdotal evidence further indicates that the inability to find satisfying work is a
relevant factor in looking for overseas jobs in the non-academic private sector. Many
university graduates do not work in their field of study, but in unrelated sectors as noted by
one respondent:
There should be a question asking if the person is practicing the profession he/she has
studied. A lot of people, particularly those who have studied liberal arts, do not
practice their professions and do unrelated things to make a living (they may be
practicing their studies as a hobby or 2nd job, etc).
Lack of planning or knowledge when making study or work decisions also appears to
contribute to the drive to go abroad to work or study among young people in Turkey. It is
not difficult to imagine that a considerable number of young people are influenced by their
peers and by societal pressures (e.g., conform to society norms) to do what is acceptable in
terms of career and life choices:
I think making a decision to go abroad is just like choosing a major for your college
degree. You do not know much about what is waiting [for] you, until you get into it.
For the college degree you choose whatever is most popular, or whichever one is the
hardest to get into. And once you are done with your degree, the next definition of
"success" is going abroad to get your Masters degree.... Sometimes in this rush you
forget why you started it all.
I believe that the most important reason people don't return is the fact that they get
caught up in daily activities and never look at the big picture.
I personally feel confusion about returning because I really am not aware of the
opportunities in Turkey in many fields. Resources and professional information and
information for potential future are not very clear and accessible in and about Turkey.
I wish there would be more aggressive and promotional governmental and professional
activities in Turkey to bring people back.
As these responses illustrate, much of Turkey’s brain drain problems may be
attributed to a lack of planning at the individual level through the education and career
choices people make (which is of course a response to the current education system and
labor market conditions) and lack of planning at the national or institutional levels.
122
6.3 A Closer Look at Student Respondents
This section gives a more detailed presentation of the responses of Turkish students
studying overseas in terms of their current program and field of study, sources of financial
support, and the reasons for choosing the current institution of study, as well as future work
destinations and expected work activities when overseas schooling is completed. The
importance of various push and pull factors differ with the level and field of study and
according to the current return intention of the respondent. These are also examined in this
section. Finally, two factors not explicitly included in the survey as possible push factors—
namely, compulsory military service and compulsory academic service—are also discussed
in terms of their impact on return intentions.
6.3.1 Current Program and Field of Study
Nearly two-thirds of the respondents are enrolled in a doctoral degree or postdoctoral
program. The remaining respondents are pursuing masters and undergraduate degrees, with
28 percent and 11 percent shares respectively (Table A.11, Appendix A). The highest
degree planned (or obtained in the case of postdoctoral fellows) by three-quarters of
participants is the doctorate, while nearly one-quarter plan to get a master’s degree
(Table A.12). The most popular field of study among participants is “Engineering and
Technical Sciences”, except for females and those pursuing master’s degrees (Table A.15;
Table A.16).
The high percentage of respondents in the technical fields is likely to be a reflection
of the greater number of graduates produced in these fields by the Turkish higher education
system. Engineering and related sciences is surpassed only by the social sciences, where
business administration is also a popular subject. Traditionally the technical fields hold
great prestige in Turkey and there is a great desire to get accepted into a technical program.
This requires a relatively high score on the nation-wide entrance exam, which is even
higher for the more prestigious universities as a result of the greater demand. There is also a
proportionately higher percentage of postdoctoral students in the “Math and Natural
Sciences” and “Medicine and Health-Related programs”, which is perhaps an indication of
the greater emphasis on basic science at this level of study.
123
Table 6.21 presents return intentions according to the field of study. Return
intentions appear to be the greatest in the social sciences and education fields. This may be
due to the greater number of government- or public sector-sponsored students in these
fields, where there is a compulsory academic service obligation (see Table A.17).
Table 6.21 Fields of Study and Return Intentions‡‡, Students (%)
Likely to
Return
Somewhat
Likely to Unlikely
Return
to Return
(n = 160)
(n = 696)
Engineering and Technical Sciences
Economic and Administrative Sciences
31.9
25.6
47.1
28.9
42.6
26.2
43.9
27.8
Math and Natural Sciences
Social Sciences
9.4
11.3
11.2
5.3
13.5
7.0
11.5
6.6
Educational Sciences
Medical and Health Sciences
12.5
4.4
3.6
1.6
5.7
2.5
5.4
2.2
Architecture and Urban Planning
Language and Literature
3.1
1.9
0.7
0.6
1.2
0.8
1.2
0.8
Arts
0.0
1.0
0.4
0.7
Field
2
Test of Independence
Total
(n = 244) (n = 1100)
(16) = 51.84***
Notes: ***p < 0.001, **p < 0.005, *p < 0.010; Cell percentages sum to 100 across columns; There are
three missing responses.
‡‡
The six categories of the return intentions variable have been collapsed into three as follows:
“Likely to Return” = “Return Immed. without Completing Studies” + “Return Immed. after Completing
Studies”; “Somewhat Likely to Return” = “Return, but not soon after completing studies” + “Probably
Return”; “Unlikely to Return” = “Return Unlikely” + “Definitely Not Return”.
6.3.2 Types of Financial Support
According to Ministry of Education statistics, the majority of Turkish students
studying abroad are private students who are studying with their own means. In our sample,
the great majority of respondents are private students, which reflects the aggregate
distribution. Only about one-fifth are sponsored by public or private organizations in
Turkey. Approximately 17 percent of respondents are government-sponsored students who
hold scholarships that have a compulsory service requirement in Turkey: 11 percent from
the Turkish Ministry of Education (MEB), 5 percent from the Higher Education Council
(YÖK), and less than one percent from the Turkish Academy of Sciences (TÜBA) and the
Scientific and Technical Research Council (TÜB TAK) (Figure 6.9).
124
44.8
T eaching or research assistant salary
33.4
Savings and/or support from family
29.7
Financial support from current university
23.0
Part-time job
11.3
MEB (Ministry of Education) scholarship
7.8
Other (loans, full-time job, etc.)
4.9
YÖK (Higher Education Council) scholarship
Other national scholarship
3.9
International scholarship / support
3.6
T urkish Academy of Sciences scholarship
1.2
Fulbright scholarship
0.7
0
10
20
30
40
50
Figure 6.9 Students Abroad by Type of Financial Support (%)
Notes: The sum of the figures does not add to 100, since respondents could have more than
one relevant source of financial support for their study abroad; (n = 1098); There are five
missing answers.
Table 6.22 Return Intentions and Compulsory Academic Service (%)
Compulsory
Academic Service
No
Yes
Return Intention
(n = 907)
(n = 191)
Return without completing studies
Return immediately after studies
Definitely return, but not soon after
Return probable
Return unlikely
Definitely not return
0.7
8.6
37.5
28.8
21.2
3.3
2
Test of independence
Total
(n = 1098)
3.7
37.2
25.1
23.6
8.9
1.6
1.2
13.6
35.3
27.9
19.0
3.0
(5) = 129.36***
Notes: ***p < 0.001, **p < 0.005, *p < 0.010; Cell percentages sum to 100 across columns.
Many private students later obtain scholarships from the foreign universities they are
attending or from foreign governments. In our sample, many of the private students are
research or teaching assistants at the institutions they are studying. Many private- and
government-sponsored students also receive financial support from their families during the
125
course of their studies. One third of respondents have received financial support from their
families or used previous savings. Many of those without scholarships finance their
education by working at a part-time job, usually within the university. Loans, full-time job
and spouse’s job were also indicated as means for financing overseas studies.
6.3.3 Reasons for Choosing Current Overseas Institution
Various factors have been cited as being important in choosing an overseas study
location. For three-fifths of the respondents the fact that their institution provided the most
relevant program in their field of specialization was important for their choice of institution.
One undergraduate student indicated that she chose to study at an American university
because she was provided greater diversity in terms of the fields of study and curriculum.
The reputation and relevance of the program (61 percent) was followed by the respondent’s
ability to get acceptance (44 percent), better financial support or scholarship opportunities
offered by the university (42 percent), recommendation of the adviser or other professors
(37 percent), and the possibility of greater job opportunities (26 percent). The “other”
category was also marked by 22 percent of the respondents which indicates that the
categories provided did not give the full range of possible reasons for choosing current
institution of study. The two categories “having Turkish contacts at institution” and “being
with or near spouse” was marked as important by 18 and 11 percent of respondents
respectively. This information is summarized in Figure 6.10.
Most relevant program
Acceptance
Best financial support / s cholars hip
Recommended
Job opportunities
Other
Turkish contacts
Spouse
0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0
Figure 6.10 Reasons for Choosing Current Institution of Study
(by % of respondents marking category)
Notes: Respondents were asked to mark all valid choices; n = 1099; missing responses = 4.
126
The respondents were also asked to choose the factor they considered to be the most
important in their decision to study at their current institution (See Figure 6.11). “Provided
most relevant program” is indicated to be the most important factor for nearly one third of
respondents, followed by “best financial support / scholarship” (18 percent) and “able to get
acceptance” (11 percent) which ties with the “other” category. Some of the factors
indicated as important by those who marked the “other” category are “prestige of
institution” (e.g., institution ranked in top 5 percent for field), “recommended by Ministry
of Education”, “lower costs”, “friends are there”, “location”, and “weather”. Private
students base an important part of their decision on cost considerations and family contacts
in the destination location.
30.8
Most relevant program for specialized field
18.1
Best scholarship / financial support
Other
11.7
Able to Get Acceptance
11.5
9.6
Recommended by adviser or other professors
9.0
Provided greater job opportunities
5.2
T o be close to spouse
Having T urkish contacts at the institution
4.1
0
5
10
15
20
25
30
35
Figure 6.11 Most Important Reason for Choosing Current Institution (%)
Notes: Respondents were asked to choose the most important factor; There are 17
nonresponses; (n = 1086).
6.3.4 Work Intentions after Completion of Studies
Students were asked in which country they expected to be working immediately after
completing their studies. The United States was the most popular work location for twothirds of the respondents. Turkey, on the other hand, was chosen by only a quarter of
students as their immediate work destination. The majority of the remaining respondents
127
chose countries in the West as possible work locations (Table A.18, Appendix A). Not
surprisingly, there appears to be a tendency for choosing a location that one is already
familiar with, since this reduces the costs involved in job search and adjusting a new
environment (Table 6.23). The majority of those studying in Canada, for example, indicated
Canada to be their immediate work location, and so on. This is less pronounced for those
residing in Europe, where there is a greater tendency for returning to Turkey. Language
also appears to be a deciding factor in choosing a work destination; respondents who have
experienced German language instruction in high school, for example, also tend to choose
German-speaking countries or regions, such as Germany and Austria.
Table 6.23 Work Destinations and Current Country of Residence
Current Country of Residence
USA
Canada
n
%
n
%
Work Destination
n
%
USA
Canada
Europe
Turkey
Other/Don't Know
652
3
15
222
16
71.8
0.3
1.7
24.4
1.8
3
26
1
6
2
7.9
68.4
2.6
15.8
5.3
11
3
40
28
4
12.8
3.5
46.5
32.6
4.7
Total
908
100.0
38
100.0
86
100.0
Europe
Notes: There are 49 missing responses (n = 1049); 17 individiuals residing in locations outside the
USA, Canada and Europe are not shown in the tabulations.
6.3.5 Types of Organizations and Activities at Work after Completion of Studies
The work intentions of survey respondents are presented in this subsection. Students
were asked the type of organization they planned to work for (or believed they would be
working for) and the type of job activities they expected to be involved in, both
immediately after and five years after completing their studies.
The majority (73 percent) of those who intend to return to Turkey immediately after
completing their studies indicate that they will start work in a university or technical
college, while the percentage of those who plan to work in the private sector is relatively
low (13.8%). A shortage of academicians persists at higher education institutions in
Turkey. In 1995, the number of positions available at these institutions was pretty much
balanced by the supply. In 2000, the number of academicians fell short of demand by
128
19,000. This gap is projected to widen further to 35,000 in 2005 (SPO, 1995, 2000). The
proliferation of higher education institutions in Turkey from the early 1990s onward has
increased the demand for higher education personnel. On the other hand, the environment
created by the economic crises has led to a contraction of private sector jobs, exacerbating
the private sector’s ability to absorb educated individuals. This may explain why
respondents who plan to return to Turkey are headed for careers in academia rather than the
private sector or other public sector jobs (Table 6.24; see Tables A.19 and A.20, Appendix
A for more detailed organizational classifications). Of those who intend to work in a
university in Turkey, the great majority believe they will be working at a public (state)
university. For some, this is because they have an academic service obligation at a state
university in Turkey.
Close to one half (48 percent) of those who indicated that they will be working in the
United States immediately after completing their studies believe they will be working in the
non-educational private sector, while 39 percent believe they will be working in a 4-year
higher education institution. The great majority of those who expect to be working at a
four-year educational institution in the US indicated they will work in a private university.
More than a quarter of respondents indicated that they will work in US-based private firm,
and one-fifth in a multinational corporation. The remaining students expect to be employed
either in a non-profit organization, international organization, or be self-employed.
Table 6.24 Intended Work Destinations and Organizations Immediately after Completing
Studies (%)
Canada
Other /
Not
Known
Total
19.6
23.2
30.4
19.6
3.6
3.6
21.9
15.6
25.0
25.0
6.3
6.3
27.6
27.6
6.9
6.9
10.3
20.7
20.7
27.5
16.9
21.6
5.9
7.5
56
32
29
1042
United
States
Turkey
Europe
University / School – Private
University / School – Public
Multinational Corporation
Other Private Organization
Government / Non-Profit / Int. Org.
Not sure
24.1
14.9
19.9
27.8
5.1
8.3
11.5
61.9
6.5
7.3
7.7
5.0
Total (n)
665
260
Organization Soon After Studies
Notes: There are 61 missing responses; Cell percentages sum to 100 across columns.
129
Table 6.25 presents respondents’ workplace intentions five years after the completion
of their studies. The percentage of respondents who believe they will be working in a state
university falls to about one half.
Table 6.25 Intended Organization Five Years after Completing Studies by Initial Work
Destination (%)
Canada
Other /
Not
Known
Total
17.9
14.3
28.6
28.6
3.6
7.1
27.3
9.1
21.2
27.3
6.1
9.1
34.5
17.2
3.5
17.2
3.5
24.1
24.6
22.5
13.5
19.0
6.7
13.8
56
33
29
1024
Organization 5 Years After Studies
United
States
Turkey
Europe
University / School – Private
University / School – Public
Multinational Corporation
Other Private Organization
Government / Non-Profit / Int. Org.
Not sure
26.2
14.7
15.6
21.3
6.8
15.3
20.5
46.5
4.7
9.8
7.9
10.6
Total (n)
652
254
Notes: There are 79 missing responses; Cell percentages sum to 100 across columns.
The majority of those who will be working in a public university believe that their
main activity would be teaching (48.3 percent), followed by applied research (30 percent),
basic research (14.5 percent), and development (3.4 percent). For respondents who
indicated that they will be working in a private university, the majority believe their main
activity will be applied research (43.2 percent), followed by basic research (27 percent),
teaching (27 percent), and development (2.7 percent). Therefore, we may conclude that
students who expect to be working in a public university, also expect to be involved more
in teaching activities than research, while those who plan to work in a private university
believe their activities will be research-oriented. Furthermore, some of those who intend to
work in a public university initially are intending to move to a private university within five
years.
6.3.6 Push-Pull Factors by Degree Program and by Return Intentions
The push-pull motivations may be different for students at different levels of study.
Table 6.26 gives a breakdown of the push-pull factors by the level of study: bachelors,
masters, doctorate and post doctorate. As expected, at the higher levels of study more
importance is given to opportunities for advanced research and training. Salary
130
considerations, lifestyle preferences and economic instability appear to be important for a
greater proportion of respondents at the bachelors and masters levels of study.
Table 6.26 Program of Study and Push / Pull Factors Viewed as Important‡ by Students (%) (n = 1095)
Push and Pull Factors
bachelors
masters
PhD
(n = 116)
(n = 300) (n = 620)
Postdoc
(n = 55)
2
(3)
PUSHA
Low income in occupation
68.1
73.0
75.0
69.1
3.03
PUSHB
Little opport. For advancement in
occupation
73.3
70.0
71.6
74.6
0.77
PUSHC
Limited job opport. in specialty
55.2
56.3
59.8
65.5
0.45
50.9
49.7
62.3
65.5
17.02
***
39.7
35.7
72.1
72.7
133.82
***
41.4
35.3
33.3
27.3
4.17
25.0
33.3
17.8
20.0
28.26
***
***
PUSHD
PUSHE
PUSHF
PUSHG
No opportunity for advanced
training
Away from research centers and
advances
Lack of financial resources for
business
Less than satisfying
social/cultural life
PUSHH
Bureaucracy, inefficiences
72.4
63.0
75.2
70.9
14.76
PUSHI
Political pressures, discord
59.5
57.3
58.3
60.0
0.25
PUSHJ
Lack of social security
53.5
54.0
50.2
45.5
2.17
PUSHK
Economic instability
78.5
80.7
74.2
69.1
6.53
PUSHL
Other
12.1
8.0
7.1
12.7
4.95
PULLA
Higher salary or wage
84.6
79.7
74.7
69.1
8.80
**
PULLB
Greater advancement oppr. in
profession
82.9
77.3
83.7
89.1
7.63
*
PULLC
Better work environment
70.9
69.3
66.6
65.5
1.39
75.2
72.3
76.5
74.6
1.84
74.4
75.7
86.3
85.5
21.00
73.5
73.3
78.9
74.6
4.32
33.3
37.0
23.6
27.3
19.49
***
51.3
38.7
71.6
70.9
98.40
***
12.0
23.0
22.1
25.5
7.36
21.4
19.3
18.7
29.1
3.69
23.9
21.3
34.8
34.6
20.26
4.3
4.7
2.9
5.5
2.49
PULLD
PULLE
PULLF
PULLG
PULLH
PULLI
PULLJ
PULLK
PULLL
Greater job availability in
specializ.
Greater oppr. to develop
specialty
More organized, ordered envir.
More satisfying social/cultural
life
Proximity to research and innov.
centers
Spouse's preference or job
Better educational opport. for
children
Need to finish or continue with
current project
Other
Note: ***p < 0.001, **p < 0.005, *p < 0.010
‡
Marked as “Very Important” or “Important” by Respondents
131
*
***
*
***
Table 6.27 gives the breakdown of the push-pull factors according to the return
intention of respondents. With few exceptions, a greater proportion of the respondents who
are unlikely to return rate each of the push and pull factors as being “important” or “very
important”. Therefore, it is difficult to determine which factors will be significant in the
empirical analysis of return intentions.
Table 6.28 gives the top five push and pull factors according to each return intention
category (made compact by combining adjacent categories as explained in the notes to the
table). Almost 90 percent of respondents who are unlikely to return have marked economic
instability as an important push reason, compared to 74 percent for those who are
“somewhat likely to return” and 66 percent for those who are “likely to return”. Threequarters of those who are somewhat likely or unlikely to return have marked low salary
levels in Turkey as an important push factor, while only two-thirds of those who are likely
to return have done so. Higher salary in the current country of residence is also an
important pull factor for a majority of respondents at each level of return intention. Greater
opportunities for developing specialty and greater advancement opportunities in profession
are also among the top five pull factors. The pull factors are, in general, chosen as
important by a greater proportion of respondents compared to the push factors. This is to be
expected since respondents are likely to give more weight to their current surroundings
rather than the environment they left behind in Turkey. Similarly, one would expect push
factors to be more prominent in a survey on the brain drain “e.g., intention to go overseas”
conducted within Turkey.
6.3.7 Compulsory Military Service as a Reason for Not Returning
The military service requirement for males in Turkey is generally viewed as a career
interruption. For a considerable number of male respondents, postponing their military
service was an important reason for pursuing study and work opportunities overseas.
Military service in Turkey ranges between 15 to 18 months, and thus represents a
significant break from participating in the labor force. The time spent out of the labor
market signifies a greater economic loss for the university-educated population in Turkey,
since, as corroborated by empirical studies, the economic returns to education are highest at
the tertiary level. The time lapse can also lead to significant skill erosion and lower
132
Table 6.27 Return Intentions‡‡ and Push and Pull Factors Viewed as Important‡ by Students (%)
Likely to
Return
Push and Pull Factors
Somewhat
Likely to Unlikely to
Return
Return
2
(n = 160)
(n = 694)
(n = 241)
PUSHA: Low income in occupation
65.6
74.5
75.5
5.94
PUSHB: Little opport. For advancement in occupation
58.1
74.6
71.4
17.41
***
PUSHC: Limited job opport. in specialty
40.0
62.4
60.2
27.18
***
PUSHD: No opportunity for advanced training
50.0
58.4
61.4
5.37
*
PUSHE:
Away from research centers and advances
51.3
58.2
65.2
7.86
**
PUSHF:
Lack of financial resources for business
20.6
37.5
34.9
16.36
***
PUSHG: Less than satisfying social/cultural life
13.8
21.5
33.2
22.84
***
PUSHH: Bureaucracy, inefficiences
68.8
69.2
79.3
9.51
***
PUSHI:
Political pressures, discord
56.9
54.9
68.9
14.53
***
PUSHJ:
Lack of social security
41.3
49.1
64.3
24.11
***
PUSHK: Economic instability
65.6
73.9
89.6
35.78
***
PUSHL:
6.3
6.1
15.4
21.61
***
PULLA: Higher salary or wage
59.0
77.9
85.8
40.07
***
PULLB:
Greater advancement oppr. in profession
67.7
83.9
86.7
27.75
***
PULLC:
Better work environment
59.0
67.3
75.0
11.48
***
PULLD: Greater job availability in specializ.
55.9
76.7
83.3
41.37
***
PULLE:
Greater oppr. to develop specialty
72.1
83.8
83.8
12.83
***
PULLF:
More organized, ordered envir.
66.5
74.7
88.8
30.39
***
PULLG: More satisfying social/cultural life
16.2
25.3
45.8
50.88
***
PULLH: Proximity to research and innov. centers
54.7
59.2
67.5
7.70
PULLI:
Spouse's preference or job
13.7
20.7
28.8
13.63
***
PULLJ:
Better educational opport. for children
16.8
17.2
28.8
15.99
***
PULLK: Need to finish / continue with current project
31.7
29.4
30.4
0.36
PULLL:
2.5
3.2
5.8
4.29
Other
Other
(2)
**
**
Notes: ***p < 0.001, **p < 0.005, *p < 0.010
‡
Marked as “Very Important” or “Important” by Respondents
‡‡
The six categories of the return intentions variable have been collapsed into 3 as follows:
"Likely to Return" = "Return immed. w/o Completing Studies" + "Return immed. after Completing Studies"
"Somewhat Likely to Return" = "Return, but not soon after completing studies" + "Probably Return"
"Unlikely to Return" = "Return Unlikely" + "Definitely Not Return"
133
Table 6.28 Top Five Push and Pull Factors according to Return
Intentions
PUSH Factors
Likely to Return (n = 160)
PUSHH Bureaucracy, inefficiences
PUSHA Low income in occupation
PUSHK Economic instability
PUSHB Little opport. For advancement in occupation
PUSHI Political pressures, discord
%
68.8
65.6
65.6
58.1
56.9
Somewhat Likely to Return (n = 694)
PUSHB Little opport. for advancement in occupation
PUSHA Low income in occupation
PUSHK Economic instability
PUSHH Bureaucracy, inefficiences
PUSHC Limited job opport. in specialty
%
74.6
74.5
73.9
69.2
62.4
Unlikely to Return (n = 241)
PUSHK Economic instability
PUSHH Bureaucracy, inefficiences
PUSHA Low income in occupation
PUSHB Little opport. for advancement in occupation
PUSHI Political pressures, discord
%
89.6
79.3
75.5
71.4
68.9
PULL Factors
Likely to Return (n = 160)
PULLE Greater oppr. to develop specialty
PULLB Greater advancement oppr. in profession
PULLF More organized, ordered envir.
PULLA Higher salary or wage
PULLC Better work environment
%
72.1
67.7
66.5
59.0
59.0
Somewhat Likely to Return (n = 694)
PULLB Greater advancement oppr. in profession
PULLE Greater oppr. to develop specialty
PULLA Higher salary or wage
PULLD Greater job availability in specializ.
PULLF More organized, ordered envir.
%
83.9
83.8
77.9
76.7
74.7
Unlikely to Return (n = 241)
PULLF More organized, ordered envir.
PULLB Greater advancement oppr. in profession
PULLA Higher salary or wage
PULLE Greater oppr. to develop specialty
PULLD Greater job availability in specializ.
%
88.8
86.7
85.8
83.8
83.3
134
productivity upon resumption of career-related or educational pursuits. The career break
may be even more crucial for those with advanced graduate degrees who are pursuing
careers in academia and in cutting-edge occupations in which skills must be renewed or
upgraded continuously.
In 1980, an important change was made in the military service law. Individuals
working abroad for at least three years were allowed exemption from long term military
service in return for the payment of approximately
5,000. Instead of the 18 months of
regular service, they were required to finish only one month of basic military training.
Several other important changes were made in the military service system in 1992, which
include the shortening of service duration to 15 months and the extension of the short term
military service in return for fees to those living in Turkey. This exemption from long term
service, however, could take place only through legislation during periods when the supply
of new recruits exceeded the military’s demand5. While compulsory military service was
not listed as a “push” factor in the survey questionnaire, many male respondents indicated
that for them and for many of their friends delaying or shortening military service duty
played an important role in the decision to not return. One respondent explained in this
way:
Compulsory military service is perhaps one of the most important reasons why Turks
studying abroad, particularly the male students pursuing a masters degree, delay
returning to Turkey. Almost all of the male students studying abroad plan to work
three years abroad in order to qualify for short-term military service. Some of these
students return to Turkey after three years but others want to continue with their
careers abroad and so make plans for permanent settlement in their country of work.
A 25-year-old master’s student studying in the United States
6.3.8 Views of National Scholarship Recipients
The Ministry of Education (MEB), the Higher Education Council (YÖK) and the
Turkish Academy of Science (TÜB TAK) all award scholarships in return for compulsory
5
Most recently, a law was passed in 1999 allowing those born before 1973 to take advantage of
short-term military service provided they would pay the fee of around 7,500 to 10,250. Those
born before 1960 were allowed to bypass the one month basic military training if they wished. The
demand for short term military duty was huge, but not everyone who wanted to benefit from it did,
either because of the age limit or the high exemption fee. As a result, some of those who have not
completed their military service are waiting for a new law to pass. In the mean time, education and
training abroad allow many to delay their military duty, and after three years of full-time work
abroad they qualify for short term service anyway, though subject to a higher fee.
135
academic service, usually to be served in the state universities of Turkey. As indicated in
Chapter Four, non-returning scholarship recipients have become a concern. While a greater
percentage of students who have an academic service obligation are returning compared to
private students, the significant number of non-returning scholarship recipients point to the
lack of efficiencies within the scholarship system and to a lack of planning in terms of
making the return home more productive for both the recipient and for the development of
the higher education system. One respondent, who returned to Turkey to complete her
compulsory academic service, was dismayed to learn that her university did not have a
program in her specialization. Her requests to transfer to another university that included
her field of study were turned down without explanation, and her attempts to engage in
research projects were mired in bureaucratic obstacles. A different respondent listed the
following deficiencies of the scholarship and higher education system:
1) There are no facilities or the department in the specific university [in Turkey] which
I have been funded through. The [rector] of the university (I think he is really like that)
is thinking of assigning me to the technical college. I do not see any reasons to send
me studying abroad for that need. I bet just an instructor with a BS degree would be
sufficient... 2) YÖK spent almost $90,000 on me, excluding the tuition fees for four
years. So it might have been about $140,000 if I had not received a tuition waver.
However, they do not want to spend any more money for us to establish a lab or to
bring our own software, computers, equipments when we return. I guess for my
particular case, I need to have $10,000-$20,000 (It seems high but I can earn this
money within one year here) to establish my work environment in Turkey in order to
be successful and productive for my country. Otherwise, it is not making sense just to
bring people back immediately after their graduation without technology or the things
they need. 3) I need to spend a few more years here before going back to learn really
what the overall picture is. The Ph.D. is so specialized that I don’t think [it is
sufficient] for a person to continue with his/her career without some other sources. I
believe there should be [more] inputs, supportive information, and environment for us
to be fruitful and productive. These are again not provided in Turkey.
A frustrated participant made the following comment:
Beyin göçünü biz isteyerek yapmıyoruz. Bizi buna itekliyorlar. Biz buraya geldikten
sonra YÖK olsun MEB olsun bizimle hiç iyi yönde ilgilenmiyorlar. Hep kar ımıza bir
sürü zorluk çıkartıp bezdiriyorlar. Aslında hepimizin yüre i gelmek için yanıp
tutu uyor. Ama burada doktorasını deprem üzerine yapan n aat Mühendisi
arkada ımızın Türkiye'ye döndükten sonra Kayseri Milli E itim Müdürlü üne memur
olarak atandı ını ö rendikten sonra içimiz kan a lıyor ve Türkiye’ye dönme
hevesimiz, ate imiz, a kımız zarar görüyor.
To summarize, these examples illustrate that the advanced education and training received
abroad is not being put to the best possible use for both the returnees and the higher
education system. As the anecdotal evidence indicates, scholarship recipients have, by and
136
large, come to share a negative perception about working conditions in the universities
where they have to complete their compulsory academic service, especially the newly
established universities located in less developed regions. These impressions, in turn, have
a negative impact on the decision to return to Turkey for some. Despite the dissatisfactions
outlined above, the current survey results indicate that national scholarship recipients are
more likely to be returning to Turkey immediately after completing their studies: 37.2
percent indicate they will return immediately after completing their studies compared to 8.6
percent for the remainder.
6.4 A Closer Look at Professionals
6.4.1 Highest Degree Held and Field of Highest Degree
A majority of respondents hold a masters degree (41%); this is followed by those
with doctorate (37%) and bachelors degrees (22%). The most common field of study at all
levels of education is the engineering and technical sciences, followed by economic and
administrative sciences (see Table 6.29). These two broad fields account for 84%, 89% and
70% of respondents with bachelors, masters and doctoral degrees, respectively. The
mathematical and natural sciences, and the medical and health sciences also accounts for a
significant proportion—more than one-fifth—of doctorate holders. These patterns,
including the greater emphasis on technical fields, are possibly a reflection of the demand
for skilled foreign workers in the country of residence.
Table 6.29 Highest Degree Held and Field of Highest Degree (%)
Highest Degree
Highest Degree Field
Bachelors Masters Doctorate
Engineering and Technical Sciences
62.2
51.5
52.9
Economic and Administrative Sciences
22.2
37.6
17.3
Architecture and Urban Planning
4.4
2.6
2.0
Math and Natural Sciences
3.6
3.4
11.5
Social Sciences
3.6
2.8
4.2
Educational Sciences
1.5
0.6
0.2
Medical and Health Sciences
1.5
0.4
10.8
Language and Literature
0.7
0.2
1.1
Arts
0.4
0.8
0.0
Total
100.0
100.0
100.0
n
275
497
452
2
Test of Independence
Note: ***p < 0.001, **p < 0.005, *p < 0.010
137
(16) = 152.18***
Where a respondent receives his/her highest degree may also be of significance.
Table 6.30 below gives both the level and country of highest degree of respondents. More
than two-thirds have obtained their highest degrees from a foreign country and this is
generally at the masters or doctoral level. Of those who received their highest degree from
Turkey, more than half hold a bachelors degree, about a third hold a masters degree and
only one in seven hold a doctorate.
Table 6.30 Highest Degree by Level and Country (%)
Country of Highest
Degree
Foreign
Highest Degree
Country
Turkey
Bachelors
Masters
Doctorate
Total
7.3
45.5
47.2
55.9
29.8
14.4
100.0
100.0
841
n
2
Test of independence
383
***
(2) = 369.90
Note: ***p < 0.001, **p < 0.005, *p < 0.010
6.4.2 Stay Duration and Return Intentions
One of the purposes of this thesis is to do an econometric investigation of the
determinants of return intentions to Turkey. Before proceeding with the empirical analysis
of the determinants of return intentions, it may be useful to do a preliminary analysis of the
relationship between some of the variables of interest. A very useful inductive method for
analyzing and interpreting the associations in large datasets comprised of categorical
variables is the technique called correspondence analysis. This methodology allows the
associations between the categories of a set of variables to be described in terms of a small
number of dimensions. It is thus similar to principal components analysis, which is used to
uncover common dimensions among a set of continuous variables. One of the advantages
of correspondence analysis is that it doesn’t require making any restrictive assumptions
about the characteristics of the dataset (see Clausen, 1998 for further details). This
technique is used to examine the relationship between stay duration, initial return intentions
and current return intentions in this section.
138
Simple correspondence analysis (CA) gives a visual depiction of the relative
proximity between the categories of two categorical variables as measured by the chisquare distance. Figure 6.12 illustrates the relationship uncovered by CA between the
responses given by survey participants on their initial and current intentions about returning
to Turkey, and their length of stay in the current country of residence. The boxed categories
represent current return intentions, while the remaining points represent the categories of
the combined ‘stay duration’ and ‘initial intention’ variables. The initial intention variable
has three categories—return, uncertain, and stay—that are indicated by R, U, and S
respectively.
Points-rows and Points-columns (axes F1 and F2: 91 %)
2
1.5
S_30+
U_30+
-- axis F2 (20 %) -->
1
R_less_1y
S_20_25y
0.5
S_1_5y
definitely return, plans
R_1_5y
definitely return, no
plans
R_30+
R_25_30y
S_16_20y
U_25_30y
R_6_10y
U_16_20y
R_20_25y
R_16_20y
U_20_25y S_less_1y
U_6_10y U_11_15y
U_less_1y U_1_5y
S_25_30y
return probable
-1
-1.5
-1
S_6_10y
S_11_15y
R_11_15y
0
-0.5
definitely not return
-0.5
return unlikely
0
0.5
1
1.5
2
-- axis F1 (71 %) -->
Figure 6.12 Correspondence Analysis of Initial and Current Return Intentions and
Stay Duration
Two things are noteworthy: first, initial intentions appear to be positively
associated with current return intentions, and secondly, return intentions also appear to
weaken with the length of stay. For example, survey participants who have stayed for less
than a year in their current country of residence and who have also indicated an initial
intention to return are associated with definite return plans. Return plans weaken for the
139
group with initial return intention when the length of stay increases to between one and five
years, and further still when the duration of stay is longer than five years. The same pattern
holds for those who were initially uncertain about returning; as stay duration increases, the
likelihood of returning declines. Those with an initial intention of not returning (staying) lie
close to the “unlikely to return” and “definitely not return” categories regardless of stay
duration.6
6.4.3 Return Intentions According to Location of Highest Degree
In Figure 6.13, correspondence analysis is used to reveal the response pattern of
three separate groups in terms of their current intentions about returning to Turkey. The
three groups are 1) those who have obtained their highest tertiary-level degree from a
Turkish university, represented by HDTUR; 2) those holding their highest degree from a
foreign institution and whose first full time job after completing their studies is located
outside Turkey, whether in the same city or same country as their studies or in another
country [HDFOR(samecity); HDFOR(samecountry); HDFor(dif_country)]; and 3) those
with a foreign highest degree who initially returned to Turkey to work after completing
their studies and then went abroad to work, represented by HDFOR(Turkey).
The upper-left cluster of Figure 6.13 reveals that those who have obtained their
highest degree from a Turkish university appear to be closely associated with definite return
intentions. The second group, forming the bottom left cluster, represents the phenomenon
of student non-return—those who have remained abroad to work after completing their
studies. The members of this group appear less definite about their return intentions; the coordinates of the points representing this group lie close to the “return probable” and “return
unlikely” points. The third group forming the center-right cluster differs from the other two
in that it comprises those who returned to Turkey to work at a full-time job immediately
after completing their studies at a foreign university and who then decided to go abroad
again to work. The members of this group appear more likely to indicate that they will
definitely not return to Turkey. If intentions translate into reality, it would appear that the
migration of professionals—or brain drain in the traditional sense—as measured by those
6
There is the possibility that the current intentions of respondents may cloud their memory of their
initial intentions about returning. One way of remedying this would be to undertake a longitudinal
study of the same individuals over time and comparing their recent responses to previous responses.
140
whose highest degree is from a Turkish university, is less of a concern than non-returning
students for Turkey’s brain drain problem. Even more troublesome is the third group of
returning students who have experienced working in Turkey after completing their studies;
they appear to be the least likely to return to Turkey.
Points-rows and Points-columns (axes F1 and F2: 98 %)
0.3
0.25
definitely return, plans
0.2
-- axis F2 (29 %) -->
0.15
definitely return, no
plans
HDTUR
0.1
definitely not return
0.05
HDFOR(samecity)
0
HDFOR(Turkey)
HDFOR(dif_country)
-0.05
return unlikely
return probable
-0.1
HDFOR(samecountry)
-0.15
-0.2
-0.1
0
0.1
0.2
0.3
0.4
0.5
-- axis F1 (68 %) -->
Figure 6.13 Correspondence Analysis of Return Intentions: Student Non-return
versus Professional Migration
6.4.4 Return Intentions by Level of Highest Degree
Disaggregating the three groups in Section 6.4.2 further by level of highest degree
(bachelors, masters, or doctorate) also reveals interesting information. Figure 6.14 presents
the correspondence analysis of return intentions for respondents differentiated by their level
and location of highest degree (FOR_bach, FOR_mast, FOR_PHD; HDTUR_bach,
HDTUR_mast and HDTUR_PHD) and whether they initially started work in Turkey or a
foreign country after completing their studies (workTUR, workFOR). Since the level of
highest degree is an indication of the level of specialization achieved by the respondent
through formal study, a pattern of non-return for students with foreign doctorate degrees
141
will provide some confirmation that specialized training in a foreign country has an adverse
impact on return intentions.
Points-rows and Points-columns (axes F1 and F2: 87 %)
0.4
0.3
definitely not return
FOR_PHD;w orkTUR
definitely return, plans
-- axis F2 (22 %) -->
0.2
HDTUR=masters
0.1
definitely return, no
plans
HDTUR=bach
HDTUR=PHD
FOR_mast;w orkFOR
0
FOR_mast;w orkTUR
-0.1
FOR_bach;w orkTUR
return unlikely
return probable
FOR_PHD;w orkFOR
-0.2
-0.3
-0.6
FOR_bach;w orkFOR
-0.5
-0.4
-0.3
-0.2
-0.1
0
0.1
0.2
0.3
0.4
-- axis F1 (65 %) -->
Figure 6.14 CA of Return Intentions and Level of Highest Degree: Student
Non-return versus Professional Migration
Figure 6.14 shows that respondents with a foreign highest degree, regardless of
level, are more disinclined to return than those holding degrees from Turkish universities.
Respondents with foreign doctorate degrees who also have some work experience in
Turkey after completing their studies constitute the group that is least associated with return
intentions. The following comments by the survey participants are insightful:
I come from a family of professors and I lived in a university campus (lojman)
throughout all my life in Turkey. I have seen some cases of failed attempts to return to
Turkey after getting a degree abroad. People come back after 5-10 years and get a
university position, but re-adaptation is not very easy. Your own country becomes
harder to adapt to than US was when you left Turkey years ago. Turkey is easier to
live in if you haven't seen the other side and what’s worse is that the changes Turkey
goes through “culturally” is a lot faster than what you can find here in the US.
142
There was no question [in the survey] about job experience (length of time, etc.) in
Turkey. In my case my first 4 years of employment were in Turkey, as well as one
year of sabbatical. It might have shed some light on informed comparisons on the part
of those who've elected to remain abroad.
6.4.5 Respondents by Occupation and Job Activities
A little over one-fifth of the sample of professionals is working in educational
occupations, almost entirely at the university level. The sample is roughly equally divided
between management, computer & mathematical science, architecture & engineering,
education and the remaining occupations. The first four broad occupation groups thus
account for about 80 percent of the total sample. The remaining fifth is divided mainly
between those in business and finance and those in the life, physical and social sciences
(see Tables A.21 and A.22 for more detailed groupings).
Table 6.31 Broad Occupation Groups and Return Intentions
DRP
(y = 1)
DRNP
(y = 2)
RP
(y = 3)
RU
(y = 4)
DNR
(y = 5)
253
87
255
234
83
263
49
3.2
2.3
4.3
4.7
3.6
5.7
8.2
22.5
29.9
26.3
23.1
25.3
14.5
18.4
35.2
40.2
35.3
35.0
32.5
32.7
14.3
34.0
26.4
27.5
29.9
31.3
38.4
51.0
5.1
1.2
6.7
7.3
7.2
8.8
8.2
1,224
54
272
416
401
81
Occupation
n
Managerial
Business / Finance
Computer & Math
Arch / Engineering
Social & Life Sciences
Education
Other
Total
2
Test of significance:
(7) = 46.85***
p < 0.001, **p < 0.005, *p < 0.010; Cell percentages sum to 100 across each row.
Lower values for y indicate greater return intentions.
Notes:
***
Table 6.32 Occupation Categories Sorted by Return and Non-Return
Intentions
%
%
Occupation
RU/DNR
Occupation
DRP/DRNP
Business / Finance
32.2 Other
59.2
Computer & Math
Social & Life Sciences
Arch / Engineering
Other
Managerial
Education
30.6
28.9
27.8
26.5
25.7
20.2
143
Education
Managerial
Social & Life Sciences
Arch / Engineering
Computer & Math
Business / Finance
47.2
39.1
38.6
37.2
34.1
27.6
Table 6.31 presents the occupation groupings by return intention. A significant chisquare statistic indicates that return intentions differ by occupation classification. However,
much of this variation appears to be between education (academe), where return intentions
are weakest, and the other groups. In Table 6.32, the two strongest (DRP and DRNP) and
weakest (RU and DNR) return intention categories are combined together, and the
occupation groups are sorted according to the two new return intention categories.
Respondents working in education and in “other” occupations are the least likely to return,
while those in business or finance are the least likely to indicate non-return intentions. In
terms of definite return plans, those in the education/academic occupations appear to have
the weakest return intentions: only one-fifth of respondents in education are definitely
planning to return. The proportion of respondents with definite return plans does not appear
to be significantly different from each other in the other occupations: approximately 30
percent have definite return intentions.
Table 6.33 Percentage of Time Spent on Various Job Activities (valid n = 1186)
20406080Topa
Code
Activities
<20% 40% 60% 80% 100% >50% Activ.
Teaching
Applied Research
Basic Research
Development
Computer Related
Administrative Activities,
Supervision
Professional Services
ACTV7
Quality Control, Production
ACTV8
Management
Accounting, Contracts
ACTV9
ACTV10 Marketing, Consumer Services
ACTV11 Other
ACTV1
ACTV2
ACTV3
ACTV4
ACTV5
ACTV6
R&D
Research & Development
(2+3+4)
77.3
67.2
79.1
73.8
64.5
11.1
19.1
12.7
15.4
12.1
8.9
8.6
4.7
7.3
9.5
1.8
2.5
2.5
1.4
4.9
0.9
2.5
1.1
2.3
8.9
6.7
9.1
5.8
6.6
19.4
13.7
17.6
10.0
14.0
26.6
80.8
84.2
11.6
2.8
4.8
3.5
1.1
3.3
1.7
6.2
5.5
11.6
10.5
14.0
95.3
97.0
91.4
95.2
2.5
1.9
4.3
1.3
1.1
0.6
1.9
1.3
0.6
0.3
0.6
0.8
0.5
0.3
1.8
1.5
1.8
0.8
3.7
3.0
3.2
1.7
6.0
4.0
35.2
18.4
20.1
12.4
14.0
35.5
45.6
Notes: R & D activities are applied and basic research and development.
a
Top activity is defined as the activity that respondents spend most of their time on compared to
other activities.
The percentage of time spent on various job activities is presented in Table 6.33.
These job activities are the same as those in the US National Science Foundation’s Survey
of Doctorate Recipients. One-fifth of respondents spend more than half their time on
computer related activities, which is not surprising since a good proportion of participants
144
are in computer related occupations. The relationship between job activities and
occupations is given in Table A.23 in Appendix A. More than a third of respondents spend
the majority of their time in research and development activities. These activities constitute
highly specialized work that may be difficult to find in Turkey. One would, therefore,
expect return intentions to decrease with increases in the R&D content of the overseas job.
However, there is no discernible positive or negative association between the R&D
intensity of job activities and return intentions (Table A.24, Appendix A).
6.4.6 Work Experience and Overseas Training
Previous work experience, in Turkey or abroad, is likely to be an important
determinant of return intentions. The great majority (70 percent) of the survey participants
have held one or more full-time jobs in Turkey (Table A.23, Appendix A). Work
experience in Turkey could have two possible effects on return intentions. Respondents
who have held a full time job in Turkey have firsthand knowledge of the work environment
and work conditions in Turkey and are, therefore, able to make comparisons based on this
information. Those who judge work conditions to be worse in Turkey are more likely to
remain abroad. Having work experience in Turkey may also increase the chance of return
since individuals with previous experience in Turkey can perhaps re-adapt more easily to an
environment they already have knowledge about.
Full-time overseas work experience is also expected to be important in determining
who is more likely to return to Turkey. Many of the respondents (about 30 percent) have
only one to two years of overseas job experience. The sample, in general, is tilted toward
those with fewer years of job experience (Table A.24 and Table A.25, Appendix A). Return
intentions are expected to decrease with the number of years of work experience in the host
country (see Section 3.3.4 in Chapter Three).
Transfer of knowledge and technology may be difficult when the training received
abroad is highly specific to an organization or to an industry which is not developed in the
home country (see Section 3.3.3, Chapter Three). To determine the impact of different
types of work experience (on-the-job training) and formal training, questions were asked on
the type of training received abroad—whether general, specific to industry or specific to the
current organization. The tabulations for on the job training and formal training are given in
Table 6.34 and Table 6.35 respectively.
145
Table 6.34 Type of On the Job Training and Return Intentions (%) (valid n = 1213)
Return Intentions
Definitely return, plans
Definitely return, no plans
Return probable
Return unlikely
Definitely not return
None
(n = 524)
Type of On the Job Training
Industry Organiz.
Specific
Specific
General
(n = 230) (n = 353) (n = 111)
5.2
19.9
32.1
35.3
7.6
2.6
25.7
36.1
30.4
5.2
100.0
100.0
Notes: Cell percentages sum to 100 across columns;
2
Total
(n = 1,213)
4.3
24.4
35.4
30.3
5.7
5.4
19.8
35.1
32.4
7.2
4.4
22.3
34.1
32.7
6.6
100.0
100.0
100.0
(12) = 11.40
Table 6.35 Type of Formal Training and Return Intentions (%) (valid n = 1213)
Return Intentions
Definitely return, plans
Definitely return, no plans
Return probable
Return unlikely
Definitely not return
None
(n = 485)
Type of Formal Training
Industry Organiz.
General
Specific
Specific
(n = 301) (n = 384)
(n = 43)
5.2
19.8
34.6
33.2
7.2
3.7
24.9
31.9
32.9
6.6
100.0
100.0
Notes: Cell percentages sum to 100 across columns;
2
Total
(n = 1,213)
3.7
23.7
35.2
32.3
5.2
7.0
20.9
32.6
27.9
11.6
4.4
22.3
34.1
32.7
6.6
100.0
100.0
100.0
(12) = 8.87
Only 3.5 percent of respondents have received formal training that is specific to the
organization they are working for. This is a somewhat higher (about 10 percent) for
informal on the job training. There does not appear to be a significant relationship between
the type of training and return intentions, as one would expect.
6.4.7 Respondents by Type of Organization
Close to half (46 percent) of respondents are working in multinational corporations,
while 17 percent are working in other private firms. Slightly less than a third are working in
a university (22 percent), research center (3 percent), or in a hospital/medical center (3
percent) (Table A.29, Appendix A). Return intentions are weaker for those working in an
academic environment: 46 percent are either unlikely to return or definitely not considering
returning, compared to 36 percent for the non-academic group (Table 6.36). Many (43
percent) found their current job while already in their current country of residence, while 30
146
Table 6.36 Return Intentions by Whether Respondent is
Working in an Academic or Related Environment
Return Intentions
Academic2
No
Yes
Definitely return, plans
Definitely return, no plans
Return probable
Return unlikely
Definitely not return
4.0
24.5
34.7
30.9
5.8
5.5
16.4
32.2
37.4
8.6
n
876
348
Notes: Columns sum to 100; Academic2 refers to those working in a
university, research center or hospital/medical center; 2(4) = 15.23***
where *** denotes significance at the 1 percent significance level.
41.1
Direct contacts initiated with firm (unsolicited CV)
22.7
22.2
Informal channels (friends / colleagues)
17.4
Faculty or advisor recommendation
Other means
12.4
Ads in professionals journals
10.7
Professional recruiters or headhunters
8.7
13.3
14.8
12.0
12.4
Placement office at university
7.7 6.2
Newspaper ads
7.0 6.5
"Career days" held at Turkish universities
40.8
1.4 1.3
0.3 0.3
Turkish internet network
0
10
20 30
40 50
FFTJ
60 70
80
90
CJ
Figure 6.15 Channels for Finding First Full-Time Job Abroad (FFTJ) and Current
Job (CJ) (%)
Note: The figures do not sum to 100 since more than one channel could be picked.
percent were located in Turkey and close to 30 percent were located in another country
(Table A.30, Appendix A). Figure 6.15 shows the channels respondents have used to find
their current job and their first full-time job abroad. It is clear that in both cases many
respondents have used their own initiative to contact potential employees by sending their
CVs. A greater proportion of respondents (30 percent) who found their full time job while
147
in Turkey or in a third country have made use of informal channels (e.g., friends and
colleagues) compared to those who found their current jobs while in their current country of
residence. This points to the importance of information exchange through informal
channels for taking advantage of work opportunities at a global level.
6.4.8 Positive Contributions to Turkey During Stay
The extent of positive contributions to Turkey during the stay abroad is given in
Figure 6.16. Most respondents believe they contributed by increasing knowledge about
Turkey in the country they are staying. About 40 percent are involved in lobbying activities
on behalf of Turkey. Over one-third believe they have helped increase professional contacts
between their colleagues in their host countries and colleagues in Turkey. Over a third has
also made donations to Turkish organizations (36 percent). Some (mostly those in
academe) have participated in conferences and teaching activities in Turkey, which is a
potential route for knowledge transfer. Those in academe also help Turkish students find
scholarships in their institutions. Some of the respondents have been very active in terms of
increasing contacts and knowledge transfer between their current residence and Turkey, as
the comments of one university professor clearly shows:
I spent six weeks in Turkey in 2000 visiting 8 universities (including METU) and the
TUBITAK research centre, giving 25 lectures on my research programs. Over the past
year I had two visiting scientists from Anadolu University in my lab working on joint
projects. We are looking at organizing a conference next year in Eskisehir. Another
colleague of Turkish origin who is currently in USA has organized two NATO
summer schools in Kemer and I attended both as a presenter. Another colleague
organized a conference in Istanbul in 1996 and is organizing another one in 2001 in
Istanbul again, which I will be attending. I am working towards increasing my
collaborations with colleagues in Turkey and act as a resource for them. I currently
have a PhD student who is a graduate of METU.
On the other hand, others believe the right environment in Turkey must be created
before their knowledge and skills can be put to efficient use:
Risk yatırımı ile u ra ıyorum. Kendi ekonomik gücüm arttı ında ve Türkiye'de
giri imcilik için uygun artlar olu tu unda bu i i ülkemde yapmak isterim. Silikon
Vadisi'nde elde etti im tecrübe ve ili ki a ım sayesinde Türkiye'ye de daha faydalı
olabilirim. Bulu u, fikri olan Türkiye'de ya ayan Türk giri imcilere elimden gelen
yardımı yapmaya da çalı ırım. E er bir Türk Teknoloji ile ilgili
Adamları ve
Giri imcileri isimli veri tabanı olu turursanız ülkeye büyük faydanız olabilir. (I am
involved in risk capital. I would like to do this in Turkey when the right conditions for
entrepreneurship are created and when my own economic situation strengthens. Then I
can be of greater use to Turkey through the experience I have gained and my personal
network in Silicon Valley. I will do everything that I can for Turkish entrepreneurs in
148
Turkey who have new ideas or inventions. I believe that a database for linking Turkish
businessmen and enrepreneurs in and outside Turkey will be very useful.)
I do not believe that we can help Turkey from where we are despite some of your
questions along those lines. Turkey needs to create the environment to attract the talent
abroad. Then again, many people wouldn’t want their positions to be challenged by
“outsiders”.
Increased knowledge about Turkey in general
86.0
Lobbying actitivies on behalf of Turkey
38.9
Increased professional contacts
36.4
Donations to Turkish organizations
36.0
Helped in the transfer of knowledge
29.0
Overseas scholarships for Turkish students
27.4
Increased overseas business contacts with Turkey
19.8
Other positive contribution
12.4
0
20
40
60
80
100
Figure 6.16 Positive Contributions to Turkey During Stay (%) (n = 1099)
Note: The percentages to not add to 100 since more than one item could be picked.
6.5 Concluding Remarks
Overseas work and study opportunities are seen by participants as a means for
investing in themselves and as a way to increase their value in the marketplace at home
(Turkey) and abroad. It also appears that the quality of both the work environment and the
greater amount of career and study opportunities are important factors for going overseas.
For those contemplating an academic career, overseas experience is often a requirement for
tenure positions at some of Turkey’s best universities, and this acts as a significant “push”
factor. There is also a positive association between initial return intentions and current
return intentions, although it is weaker for those who initially intended to return to Turkey.
Return intentions weaken considerably when stay duration increases. Student non-return
compared to professional migration also appears to be more significant: those with foreign
degrees in the professionals survey are less likely to be returning.
149
The significance of the brain drain from Turkey becomes apparent when the
average years of educational attainment of the adult population in Turkey is considered. In
the year 2000, the average years of educational attainment of the adult (25 and older)
female population was 4.6, which is below the primary level. Male educational attainment
levels are somewhat higher than the female levels as a result of the gender gap in education.
The adult male population had an educational attainment level that was below the middle
school level but above the primary level of education.
In comparison, the respondents’ parents have an average educational attainment
level of nearly 11 years which is equivalent to the high school level in Turkey. There is a
considerable difference of 6 years between mothers’ educational attainments and the 25 and
over female population. The average educational attainment level of respondents’ fathers,
on the other hand, is nearly 13 years. The difference between the average educational
attainment level of the respondents’ parents and the adult male population in Turkey is also
six years. If we take into consideration that social mobility is limited, in that the probability
of receiving more education and thus greater earnings is considerably lower for those with
less educated parents, then these are striking figures. This suggests that Turkey is losing a
significant amount of human capital that will be difficult to replace.
150
CHAPTER 7
AN EMPIRICAL INVESTIGATION OF THE RETURN INTENTIONS OF
TURKISH STUDENTS AND TURKISH PROFESSIONALS
7.1 Introduction
In this chapter, the information collected through the Internet survey is used to
determine the empirical importance of various factors on the return intentions of the target
populations: Turkish professionals working abroad and Turkish students studying abroad.
The first sample of respondents consists of individuals with bachelors or higher level
degrees who were employed or who were between jobs during the period of the survey. The
second sample consists of students who were in the process of obtaining a tertiary level
degree from a foreign university or college. Section 7.2 presents a brief discussion of model
selection and estimation methodology. The empirical specification of the model and the
explanatory variables used in the empirical analysis are given in section 7.3. This is
followed by the empirical investigation of the determinants of return intentions of Turkish
professionals and other skilled workers in section 7.4; and by a similar analysis in section
7.5 for Turkish students studying abroad.
7.2 Estimation Procedures and Model Selection
The purpose of the empirical study is to determine the factors that are significant in
explaining the skilled migration from Turkey and the non-return of Turkish students. The
dependent variable is the likelihood of returning to Turkey based on the response to the
question “What are your current intentions about returning to Turkey?”. The following
possibilities were presented to respondents in the Turkish professionals survey:
151
Table 7.1 Dependent Variable: Return Intentions of Turkish Professionals
RESPONSE CATEGORIES
Label
Index
I will definitely return and have made plans to do so.
DRP
1
I will definitely return but have not made concrete plans to do so.
DRNP
2
I will probably return.
RP
3
I don’t think that I will be returning.
RU
4
I will definitely not return.
DNR
5
For the student sample, the choices forming the categories of the dependent
variable “likelihood of returning to Turkey” are slightly different from the ones used for
those working abroad. The table below gives these choices:
Table 7.2 Dependent Variable: Return Intentions of Turkish Students
RESPONSE CATEGORIES
Label
Index
I will return as soon as possible without completing my studies.
R_BS
1
I will return immediately after completing my studies.
R_IAS
2
I will definitely return but not soon after completing my studies.
R_NSAS
3
I will probably return.
RP
4
I don’t think that I will be returning.
RU
5
I will definitely not return.
DNR
6
These choices form a set of ordered categories in which each consecutive category
indicates an increase in intensity in the respondents’ intentions to stay in their current
country of residence. Because of the way the index is constructed, categories with a higher
index value imply a greater intensity in feeling about not returning (staying). In the
econometric analysis, this means that positive coefficients on the independent variables
indicate an increase in the probability of “not returning”. However, the change in intensity
between categories cannot be assumed to be uniform. Given the ordered and non-uniform
nature of these choices, the appropriate model appears to be an ordered response model
(Maddala, 1983). Formally, the observed discrete index is given by
yi = {1, 2, 3, ... , J}
(7.1)
where i indexes the observations and J is the number of categories of the dependent
variable. It is assumed that a continuous, latent variable underlies the discrete, ordered
152
categories. This latent variable is explained by a set of observed characteristics and a
random element as given below:
yi* = ’Xi + ui
(7.2)
where y* is the unobserved “return intention” variable, X is the (k×1) vector of explanatory
is the parameter vector to be estimated and u is the random disturbance term.
variables,
The relationship between the discrete, observed y and unobserved, continuous y* is given as
follows:
1 if y*i ≤ 0 (=
2 if 0 <
yi =
y*i
3 if
2
<
4 if
3
<
≤
y *i
y *i
1)
2
≤
3
≤
4
(7.3)
...
J if
where
1
,
2
,
3
...
J -1
J-1
≤ y*i
are the threshold parameters that link y to y*. Since the threshold
parameters are not known, they are estimated along with the explanatory variable
coefficients. Normalizing
1
to 0 will reduce the number of threshold parameters to be
estimated to three (Liao, 1994).
Whether to use an ordered logit or an ordered probit model depends on the
assumption made about the distribution of the error term u. Since the two models
essentially give similar results, choosing one model over the other appears for the most part
to be a matter of preference. When a very large number of observations are concentrated at
the tails of the distribution, however, the logit specification with an underlying logistic
distribution has been shown to be the appropriate specification. In this study, the ordered
probit specification, which assumes an underlying normal distribution for the error term, is
used. Choosing between a logit and probit model also means making an assumption about
the nature of the latent dependent variable. A logit specification implies a discrete latent
variable, whereas a probit specification implies a continuous latent variable (Pampel, 2000).
Given an ordered probit specification, the probability that an observed response
falls into an arbitrary category j is given below as:
153
Prob( y i = j ) =
where
(
j
) (
− ′x i −
j −1
− ′xi
)
(7.4)
(.) is the cumulative normal distribution. Differentiating this probability with
respect to the explanatory variables gives the marginal effect of each on the probability of
choosing category j. Model estimation is carried out by using maximum likelihood (ML)
estimation techniques since it has been shown that ML gives unbiased and efficient
estimates for nonlinear models.
Figures 7.1 and 7.2 present the observed frequencies of the dependent variable
return intentions for the two samples. These figures show that the distribution of responses
is concentrated in the middle rather than the extreme categories, which justifies the initial
choice of an ordered probit over an ordered logit specification.
Choosing between an ordered probit or logit model also implies making the
assumption that the explanatory variables of the model will have the same impact across
each of the categories of the dependent variable. This is known as the “parallel regression
assumption” (Long and Freese, 2001). It could well be that the coefficients of some or all of
the explanatory variables are significantly different across each categorical choice, in which
case alternative models must be considered, such as the multinomial logit model or
generalized ordered logit / probit models. In the generalized ordered models, a separate
parameter vector is estimated for each of the J categories (e.g.,
1
,
2
, ... ,
J
). The parallel
regression assumption may be tested with an approximate LR test or a Wald test (see Long
and Freese, 2001, p. 151 for details).
After choosing an appropriate estimation method based on the characteristics of the
dependent variable, a suitable model selection procedure must be decided on to determine
the set of regressors to keep in the final estimation model. There are several things to note.
One is that the set of possible factors (variables) presented in the bivariate analysis in
Chapter Six do not have the same number of valid points (cross-sections) because of
missing responses1. Including some of these regressors will come at the cost of reducing
the sample size and thus the precision of the estimated parameters. On the other hand,
excluding key variables will also compromise the fit of the estimated model.
1
Table B.1 in Appendix B provide a quick reference to the associations between the dependent
variables and the set of possible regressors for the professionals survey.
154
Dependent Variable, Return Intentions
DNR
RU
RP
DRNP
DRP
0
100
200
Frequency
300
400
Figure 7.1. Return Intentions of Turkish Professionals, Observed Frequencies
Dependent Variable, Return Intentions
DNR
RU
RP
R_NSAS
R_IAS
R_BS
0
100
200
Frequency
300
Figure 7.2. Return Intentions of Turkish Students, Observed Frequencies
155
400
The analysis in Chapter Six provides an initial criterion for reducing the number of
regressors: variables with a large number of missing responses that are not significantly
associated with the dependent variable(s), based on the chi-square test of independence, are
excluded. The various migration theories, set out in Chapter Three, also serve to provide a
guideline for keeping or excluding variables from the initial model.
After determining the initial set of explanatory variables, which are discussed in
detail in Section 7.3, the next stage in model selection involves adopting an appropriate
strategy for choosing the best possible model—one that fits the data well and is relatively
easy to interpret. The model may be complicated by non-linearities and interactions among
the regressors. Some of these significant interactions were uncovered in Chapter Six. One
approach to take would be to start from a saturated model—a model that incorporates all
possible variables, interactions and higher-order terms—and to use a backward elimination
procedure. At each step, terms that are not statistically significant individually and that also
do not contribute significantly to the fit of the model are eliminated. The elimination
procedure continues until further model reduction involves a significant deterioration in
model fit. The advantage of this approach is that all of the reduced or pared down models
are nested in the previous models so that one could use testing procedures, such as the
likelihood ratio (LR) test, that are suitable for testing nested non-linear models. Otherwise,
measures of fit based on information criteria must be used to compare non-nested models or
models with different sample sizes.
One difficulty of the current study is that the response rates vary considerably
across different sets of questions in the survey study. For example, there is a lower response
rate for questions appearing at the end of the survey than for those appearing at the
beginning. This means that starting from a saturated model with all possible sets of
regressors, even with the initial reduction in the variable set, leads to a significant reduction
in the sample size. Another approach that can be used is that of forward selection where the
explanatory variables are added sequentially to the model. The criteria for adding a variable
is based on whether the new variable significantly improves the fit of the model. With this
strategy, the explanatory variables that have the greatest significant bivariate association
with the dependent variable are used in the initial regression; then, more complicated
models are gradually built up from this preliminary model. The disadvantage of this
approach is that the final model may be sensitive to the initial set of regressors and to the
156
order in which the remaining regressors are added. The ultimate strategy adopted in the
current study is a combination of both approaches, while keeping in mind the hypotheses to
be tested.
The parallel regression assumption underlying the ordered probit model is violated
in both the student and professionals samples. A possibility is to estimate a multinomial
logit model. The drawback of using the multinomial logit model is that it does not preserve
the inherent ordering of the return intention categories and therefore does not incorporate
this information when estimating the coefficients of the explanatory variables. This results
in a loss in the efficiency of the estimators (Long, 1997). While the generalized ordered
logit model provides an alternative model that does preserve the ordering (e.g., it is a
restricted version of the multinomial logit model), it is very sensitive to low frequency
counts (e.g., small cell sizes). Thus, it is often necessary to combine the dependent variable
categories that have low frequencies with adjacent categories in order for the estimation
procedure to work2. However, combining categories may also lead to a loss in information,
especially if the underlying latent variable is multi-leveled or continuous. For example,
while the “definitely not return” category has relatively few observations, it expresses a
much more intense feeling about returning than the “unlikely to return” category, which is
an important distinction within the context of the current study. As a result, we have chosen
to present the results from the ordered probit model. A larger sample size and fewer
explanatory variables would have made the use of generalized models more feasible.
7.3 Empirical Specification of the Model: Explanatory Variables
The models estimated in this study are based on the human capital theory of
migration, which was presented in Chapter Two. Human capital theory predicts that
individuals will migrate when the net present value of benefits from migration is positive.
Wage differentials between the host and source countries provide the main motivation for
moving to a foreign country. This basic assumption is pertinent to both skilled and nonskilled labor migration. However, since the focus is on the return intentions of skilled
individuals who are currently residing outside Turkey, a slightly different set of explanatory
variables may be relevant. These variables represent a combination of economic, social,
2
The gologit command in Stata 7.0 was used to obtain estimates for the generalized ordered logit
model. The number of categories of the dependent variable was reduced to three. These results are
not included.
157
political, psychological and institutional factors. This section provides descriptive details of
some the explanatory variables that are considered in the econometric analysis of return
intentions.
7.3.1 Income Differentials
According to human capital theory, the difference in the expected foreign and
domestic income levels is the key determinant of skilled migration. Since expected income
is the relevant variable, employment opportunities and labor market conditions both at
home and abroad play an important role in the perceptions of economic opportunity held by
skilled individuals. Given the importance of perceptions in making the migration decision,
a set of “subjective” variables are used to determine the significance of economic factors.
These include the respondents’ rankings of various push-pull factors in terms of their
importance in their decision to return or stay. To account for the pecuniary aspects of this
decision, lack of a satisfactory income level in the home country was included among the
push factors and a competitive income level in the current country of residence was
included as a pull factor (pushA and pullA).
The approach of using subjective measures to test the impact of income differences
may be justified by the fact that each migrant may have different perceptions of the income
differential based on incomplete information of all alternative employment opportunities
available to him or her. Not everyone may be equally informed of the prevailing income
differentials, and more importantly, they may not place equal weight or importance to the
same information. Another difficulty in using actual income differences is that it would
require income information for a diverse range of occupations, and comparisons across
countries would also need to take into account cost-of-living differences.
As the analysis of the previous chapter has revealed, the income differential is an
important consideration (marked as “very important” or “important”) for a majority of
respondents. The task of the econometric analysis, however, is to determine the factors that
distinguish between respondents with strong return intentions versus those with weak return
intentions. It is possible that the income differential may fail to be a discerning factor since
it is considered to be important for a good proportion of respondents.
158
7.3.2 Explanatory Variables for Testing Specific Brain Drain Theories
In addition to the assessments made by respondents about their level of income in
their current country of residence versus what they expect when they return to Turkey, the
survey included variables that were designed to test some of the explanations of wage
differentials outlined in Chapter Three. These theories all adopt the human capital
framework but provide different explanations for the existence of income differentials
between the sending and receiving countries. Examining the validity of the first model
based on the asymmetric information hypothesis would require firm-level data on the
recruitment and compensation practices taking place in Turkey and the receiving countries.
This is not possible from information collected in the current study; therefore, the empirical
analysis excludes the evaluation of this particular model.
Miyagiwa’s model of agglomeration economies. The second hypothesis based on
the human capital framework is Miyagiwa’s “increasing returns to scale in advanced
education” hypothesis. The argument was that skilled individuals migrate to more advanced
countries because physical proximity to other skilled individuals concentrated in
institutions and research centers in developed countries has the effect of increasing their
individual productivities, and thus wages. There are several variables that come close to this
idea, although implicitly3. One of these is the importance of proximity to research centers
for respondents as an important reason for not returning. This is given by the variables
pushE (being away from research centers and advances in the home country) and pullH
(proximity to research centers and advances in the host country), both of which are
constructed as dummy variables where “one” indicates that the item scored high on the
Likert scale (received either a score of “five” or a “four”) whereas “zero” indicates the item
was not important to the respondent (received at most a “three”).
Because they are closely associated (e.g.,
2
(1) = 489.9, Pr = 0.000, n = 1176 for
professionals), including both pushE and pullH as separate regressors in the model would
3
This hypothesis may be more readily tested at the aggregate level or separately for different
occupations, given available data. The ratio of the number of skilled individuals (for example, PhD
holders) in a sending country (or within a specific occupation in the sending country) to the number
of skilled individuals in the receiving country (e.g., the United States) could be used as an
explanatory variable in a model explaining human capital flows into the receiving country, with the
sending countries representing the cross-sectional unit in the study. A negative, significant
relationship could then be interpreted as confirming Miyagiwa’s “agglomeration economies”
hypothesis.
159
be redundant. Thus, only pullH is included in the model. Since proximity to research
centers may be more important for respondents in academia or with a higher degree of
specialization, an interaction term, ACADxpullH, is added to the model. In future survey
studies, more detailed questions could be asked about the importance of being in close
proximity to experts in a given field.
Chen and Su’s model of on-the-job training. Another hypothesis to be tested is “onthe-job training” as an explanation for brain drain, especially student non-return, as set out
in the model by Chen and Su. In the Chen-Su model (1995), six disciplines are looked at:
medicine, engineering and sciences, which are labeled the “hard-sciences” or capitaldependent disciplines; and law, business and humanities, which are labeled the “liberalarts” or non-capital-dependent disciplines. The capital-dependent and non-capitaldependent distinction among disciplines is an important one, since it is used to test whether
the theory that on-the-job training after a period of study abroad provides an important
explanation for brain drain in the form of student non-return. It is hypothesized that brain
drain will be more prominent for graduates from capital-dependent disciplines. This is
because in capital-dependent disciplines education and training that take place in the same
country are believed to be complementary and lead to higher productivity than when
training occurs in another country. On-the-job training in the foreign country is therefore
expected to increase the likelihood of not returning to the home country for students who
completed their studies in the foreign country. In the empirical analysis conducted by Chen
and Su, whether a student studied in a capital-dependent discipline as defined above did not
provide an explanation for the Taiwanese brain drain.
In addition to the division of disciplines as capital-dependent or not according to
the Chen-Su definition, specific questions about on-the-job training and formal training in
the workplace were asked in the professionals survey. Becker’s pioneering work on human
capital formalized the notion that workers’ productivities improve with the amount of time
they spend on the job, and with the amount and type of training they receive. With general
training, for example, workers acquire skills that are easily transferable to other firms. The
more specific the training a worker receives, the more difficult it is to transfer the acquired
skills to other firms. Thus, workers with specific training will tend to be less mobile since
mobility will have a higher cost. Two sets of variables are included in the empirical model.
One has to do with the formal training received by respondents, while the other has to do
160
with less formal on-the-job training. These variables are represented by the following set of
dummy variables:
FTr1: No formal training
FTr2: General formal training
FTr3: Formal training specific to industry
FTr4. Formal training specific to current firm
OTJT1: No on-the-job training
OTJT2: General on-the-job training
OTJT3: On-the-job training specific to industry
OTJT4: On-the-job training specific to current firm
Wong’s model of learning-by-doing. Wong’s (1995) model of brain drain based on
learning-by-doing interprets the greater output level in the host country as representing a
cumulative base of experience. Foreign workers choosing to stay in the host country are
able to take advantage of the greater base of experience and increase their productivities
from learning-by-doing. This model can be tested by including the variable “number of
years of overseas work experience” in the model (yrs_wrkd_abrd) or the number of years
of experience in current country of residence (yrs_wrkd_cc) in the professionals survey.
Return intentions are expected to decline as the number of years spent working abroad
increases. If this is the case, Wong’s learning by doing model will receive confirmation.
7.3.3 Other Explanatory Variables
Gender: Although it is expected that there would be differences in the likelihood of
returning between the male and female samples, there is no a priori expectation about the
direction of this difference. The dummy variable for gender takes on the value 1 for
“female” and 0 for “male”. In previous empirical studies, women have been found to be
more reticent about returning to their homelands. In the case of China (Zweig and
Changgui, 1995: 36-7), this is believed to be caused by a lack of career opportunities for
women (e.g., the biases they face in the workplace) and constraints imposed on their
behavior in China, as well as certain convenience factors abroad, aside from greater wage
levels, that offer them many more modern conveniences and a more comfortable lifestyle
161
than they could expect to experience in China. These factors, including less lifestyle
freedom, may also be important for women in Turkey making them less willing to return.
According to one respondent:
There is a very specific reason for why I stayed in the USA initially. I had had all the
intentions of returning at the end of my PhD. When I left Turkey I was 24 and had
been married for three years. Toward the end of my PhD I got a divorce at the age of
26. In 1986, Turkey was not ready to accept the notion of a 26 year old divorced
woman living by herself. My family expected me to live with them. That was not
acceptable to me. Even today I do not feel that I would be as comfortable (or receive
the same amount of respect I get in the USA) living in Turkey as a divorced 42 year
old.
Age: “Age” and “Age squared” are included as explanatory variables in order to
control for cohort effects and possible nonlinearities. Previous empirical research has
established age as an important factor in determining the net present value of migration.
Older workers tend to be less mobile than younger workers since the “psychic costs” of
moving increase with age (Stark and Bloom, 1985). Older workers in the sample of
professionals may therefore be expected to indicate a greater intention of remaining in the
host country. Chen and Su (1995) have suggested, however, that a younger graduate has a
greater likelihood of staying in the foreign country than an older one since the present value
of her income streams in the foreign country is greater (amplifying the wage differential
between home and host country and the relative returns to be earned). Workers approaching
retirement may therefore exhibit stronger intentions for returning than younger workers
who face a longer time frame for working and earning a high salary level in the foreign
country.
Stay Duration: Stay duration is the number of years spent in the current country.
When stay duration increases, the incentive to return is expected to diminish, since
individuals become more accustomed to living abroad. Thus, there may be an “inertial
effect” with an increase in the length of stay. Longer stay duration may also be indicative of
a preference to live abroad, whether existing initially or acquired with time. Since the stay
duration variable also incorporates the effects of age, initial preferences and work
experience (and hence the effect of on-the-job training on the migration decision),
controlling for these variables will reveal the “pure inertial effects” of stay duration.
Another possibility, which appears to be pertinent for Turkey and other developing
countries, is given by one of the survey respondents:
162
Dı arıda 3-5 yıl ya adıktan sonra dönmek çok zorla ıyor. Türkiye'de i ler informal
ili kilerle bulunuyor. Dı arıda olmaktan dolayı informal ili kiler geli medi inden,
dönünce olanakların ne olabilece ini kestirmek zor oluyor. (Returning becomes very
difficult after living abroad for 3-5 years. In general, finding a job in Turkey depends
on informal relations, and being outside Turkey means that you can’t develop these
informal networks. Therefore, it is difficult to imagine what kind of opportunities you
will be facing when you return.)
According to another respondent, re-adapting to Turkey can be as difficult as the
initial adjustment to a foreign culture when stay duration increases:
Yabancı bir ülkede uzun sure kalınca, insan Türkiye’deki alı kanlıklarını unutuyor, dil
de i imini kaçırıyor, kültür de i imini takip edemiyor. Hatta bazen Türkiye’ye gitmek
yabancı bir ülkeye gitmek gibi stresli oluyor. (When a person stays for a long time in a
foreign country, they miss the cultural and language changes that take place in Turkey
and can forget their old habits and living patterns. Going to Turkey can sometimes be
as stressful as going to a foreign country.)
Years of Work Experience: The number of years of work experience is believed to
contribute to the general skills level of the respondents, which is believed to increase
mobility. Goss and Paul (1986), argue that when the number of years of work experience is
not controlled for, the coefficient on the “age” variable will be the sum of two
countervailing factors. If the distinction between work experience in the home country
versus in the foreign country is important for return intentions, then the number of years of
work experience abroad may be the more pertinent variable (Wong, 1995), since this
implies that respondents with greater overseas work experience will have acquired skills
that are related to the capital stock of the host countries (see previous section).
Initial Return Intentions: Respondents were asked about their initial return
intentions prior to going abroad to work or study. The possible responses were “return”,
“undecided” and “stay”. The dummy variables, init_RETURN, init_STAY and
init_UNSURE, were constructed to reflect the initial intention of the respondent. In the
ordered probit analysis, “stay” was chosen as the reference category. It is expected that
respondents who left Turkey with the intention to return will be more likely to express the
same intention at the time of filling out the survey.
Marital Status and Family Support: Family considerations are also expected to
have considerable weight in the mobility decision of individuals. The marital status of
respondents is included as an explanatory variable to account for family constraints. The
effect of this variable on return intentions can work in either direction. Marriage to a
163
foreign spouse is expected to reduce return intentions, while marriage to a Turkish spouse
may either reduce or increase return intentions depending on the spouse’s preferences and
position in the family. The respondents were asked about the attitudes of their families both
in terms of their initial decision to go abroad (fam_sup1) and in terms of settling down
permanently in their current location (fam_sup2). In a family-oriented culture, family
attitudes may be expected to have a significant impact on the return decision of
respondents. Both of the family support variables are ordinal categorical variables4, which
are treated as interval variables in the econometric model whenever appropriate (e.g. this
decision is based on whether the null hypothesis of evenly spaced categories is rejected by a
likelihood ratio test).
Parental Education Levels:
Parents’ educational backgrounds provide information about the socio-economic
background of their children. Socioeconomic background is probably more important in
whether a person is ever able to go overseas for study or work experience. The educational
attainment of parents will determine the educational opportunities available for their
children. Children from higher income, more educated families are more likely to get a
better education (e.g., since their families will be able to afford better quality schools or be
able to spend more time with them on schoolwork), and proceed on to higher level studies.
Those with higher, university-level skills have greater prospects for finding overseas
education and employment opportunities. Since the more educated are more mobile than
the less educated, and because the level of educational attainment increases with the
parents’ education levels (see Tansel, 2002a), it is not surprising to find that the sample of
respondents come from highly educated backgrounds (see Section 6.2.3 in Chapter Six.).
Occupation and Work Activities: A distinction can be made between academic and
non-academic occupations. A dummy variable representing working in academia (or plans
for working in academia in the case of students) was constructed to determine whether
academicians are more or less likely to return than those in other occupations. Respondents
were also asked to give the percentage of time they spend on various job-related activities.
The first three job activities (basic research, applied research and development) are R&D
activities (OECD, 1994: “Frascati Manual 1993”). The other activities considered are
technical support, administrative and various other activities. These activities have been
4
See Questions 21 and 25 in the student and professionals surveys respectively.
164
used as part of the National Science Foundation (NSF) Survey of Doctorate Recipients in
the US (NSF, 1997). The same definitions of job activities are also used in the current
survey study. It is expected for respondents involved in activities related to research and
development have weaker return intentions, since they are doing very specialized work that
may be difficult to duplicate or develop in Turkey.
Previous Overseas Experience: Prior overseas experience (work, study or travel)
before coming to the current country of residence may be an influential factor in adjusting
to or feeling comfortable with the current country of stay. Some of those with previous
overseas experience who returned to Turkey to work for a period of time have also had the
opportunity to compare the work environments and therefore base their return decisions on
this comparison. In addition to prior experience overseas, various adjustment factors were
included in the questionnaire, including having a large Turkish community in the city of
residence (see Section 6.2.10). These factors and difficulties faced while abroad are
included in the model as dummy variables.
Level and Location of Highest Degree Completed: Each consecutive level of
higher education represents an increasing degree of specialization. It is postulated that those
who have received more specialized formal education abroad, based on the degree level, are
less likely to return since their advanced training will be more relevant or attuned to the
needs of the foreign country and thus provide them with higher monetary returns in the
foreign country than in their native country. The level of highest degree is represented by
the following set of dummy variables: bachelors, masters and doctorate.
If the highest degree completed by a respondent is from a Turkish institution of
higher education, then the individual is part of the “classic brain drain” (HD_TUR). On the
other hand, if the highest degree completed is from an educational institutional outside
Turkey, then the respondent is part of the phenomenon of “student non-return” (HD_FOR).
Language Facility / Skill: Language skills may also be an important part of
adjusting to life abroad. The greater the command of a foreign language, the easier it is to
make the transition to a foreign culture. Language acquisition is also related to the age of
the respondent, which suggests that those who go abroad at an earlier age will generally
have better command of the foreign language in question. As mentioned before, foreign
language instruction in the home country should also increase language skills and prepare
165
students for foreign study or work experience. To account for early exposure to a foreign
language, language of instruction in high school for science and social science classes are
included as dummy variables in the model (HSsci_TUR and HSsoc_TUR). The expectation
is that those who have received foreign language instruction in high school will adjust more
easily to a foeign culture (since it will be less foreign to them) and exhibit less intense
return intentions than those who complete their high school education in Turkish language
schools.
Economic Instability and Uncertainty: General economic conditions and
economic stability will determine relative employment opportunities and can lower or
increase an individual’s expected income accordingly. Economic instability and uncertainty
in the home country was included among the Likert scale items as a push factor (pushK).
This variable is expected to have a strong deterring effect on return intentions for the
sample considered since at the time of the survey the Turkish economy was experiencing
the effects of the 2001 economic crisis.
The variables discussed above may be divided into policy and non-policy variables
in order to distinguish between those factors for which “something can be done about”,
such as income differences, and those that form part of the respondent’s lifestyle
preferences and constraints including brain drain due to marriage to a foreign spouse.
7.4. Determinants of the Return Intentions of Turkish Professionals
In the ordered probit model, the independent variable “return intention” is
constructed in a way such that categories that suggest greater intensity in feeling about not
returning (staying) are assigned higher values. As a result, positive coefficients on the
independent variables indicate an increase in the probability of “not returning” while
negative coefficients imply an increase in the probability of “returning”. Table B.2 in
Appendix B provides summary statistics and descriptions of the variables used in the final
model, which was chosen on the basis of goodness-of-fit statistics: mainly the AIC and
McFadden’s adjusted R2. In comparing nested models, the likelihood ratio test was also
used. In general, these three statistics gave very similar results. The final model has 59
regressors, many of which are qualitative or dummy variables. The ordered probit model
results are used in the analysis of the determinants of return intentions, since model
selection (e.g., determining the appropriate explanatory variables) are based on the results
166
from fitting various ordered probit models5. Estimates of the coefficients and the associated
marginal effects are provided in Appendix B for both the ordered probit model and the
alternative multinomial logit model (Tables B.3 and B.4). The effects of various factors on
the “non-return” decision are discussed under separate headings below.
Gender Effects:
There are gender differences in the estimated probabilities of return intentions.
Positive, statistically significant coefficients on the dummy variable, female, indicates that
female respondents have a higher probability of indicating an intention of “non-return” in
the ordered probit results. Table 7.3 summarizes the marginal effects of gender on the
probabilities associated with each outcome. The marginal effects were computed by
holding all other explanatory variables at their means and accounting for gender interaction
effects (e.g., setting femalexpullK to zero for males and to 1x(mean of pullK) for females).
The gender differences in the marginal effects show a clear tendency for females to indicate
that they plan to remain abroad compared to males. The probability of returning to Turkey
being unlikely is 0.10 points higher for female respondents, and the probability of definitely
returning (y = 1 or 2) decreases by 0.07. This may be because educational and migration
opportunities for women are more limited, which makes the migration of females a more
selective process (e.g., as evidenced by the higher socio-economic background of females
in the survey as measured by parental education levels). Another important factor may be
the greater freedom of lifestyle that some of them may enjoy while abroad.
Table 7.3 Marginal Effect of Gender, Professionals
Probabilities:
DRP
y=1
Male
Female
Difference
0.0045
0.0018
-0.0027
DRNP
y=2
0.1785
0.1139
-0.0646
RP
y=3
RU
y=4
DNR
y=5
0.5157
0.4744
-0.0413
0.2937
0.3935
0.0998
0.0076
0.0164
0.0088
Cohort Effects:
The age and agesq variables are statistically significant at the 1% significance level
for the ordered probit model when the stay duration and work experience variables are
5
Since the ordered probit model violates the parallel regression assumption, the results of the
multinomial logit model are given as an alternative. The fit, however, is not as good as the ordered
probit model, and the results are less intuitively appealing. Further studies can explore different
estimation strategies.
167
excluded. A positive sign on the age coefficient indicates a higher intensity in non-return
intentions for older respondents. This may be a reflection of the possibility that older
respondents have spent more time abroad than younger respondents and are more firmly
established in their overseas careers and/or have become more accustomed to the lifestyle
abroad. As such, the “age” variable may be echoing the effects of the “stay duration”
variable. Older individuals also tend to be less mobile than younger individuals, and
therefore may exhibit a greater tendency (“inertia”) to stay in their current place of
residence. A negative sign on agesq means that the tendency for individuals to “not return”
increases with age at a diminishing rate. When stay duration, years of work experience and
possible interaction effects (e.g., AGExSTAYDUR and AGESQxSTAYDUR) are
controlled for, the coefficients become marginally statistically insignificant.
Effects of Stay Duration and Work Experience:
The probability of returning to Turkey is expected to decrease as stay duration
increases, holding everything else constant (including age, work experience, lifestyle
preference). Stay duration may be thought of as reflecting “inertial effects”: returning
becomes more difficult after individuals become accustomed to living conditions abroad.
Increases in the length of stay duration may also speed up the acculturation process and
shift personal lifestyle preferences toward the culture of the host country. Another
important effect of stay duration is that “psychic” or adjustment costs associated with the
initial move to a foreign country diminish as the length of stay increases.
Figures 7.3a and 7.3b show the effects of stay duration on return intentions holding
age constant at 35 years, which is close to the average age for the sample. The marginal
effects for the extreme categories (DRP and DNR) are small and lie close to the origin as
illustrated in Figure 7.3a, although definite return plans show a decrease in probability with
stay duration, while the probability of definitely not returning shows an increase. The
overall trend is an increase in the probability of not returning and a decrease in the
probability of returning as stay duration increases, which is as expected.
The number of years of work experience in the current country of residence is
included as a separate explanatory variable in the model. This measure serves as a proxy for
the amount of learning-by-doing accumulated in the host country. Figure 7.4 presents the
effect of different amounts of work experience on return intentions. The same qualitative
results apply as for the stay duration variable, except that increases in work experience
168
Probability (y|x)
0.5
0.4
0.3
0.2
0.1
0.0
1
2
5
7
10
12
15
DRP(1)
0.0048 0.0045 0.0036 0.0032 0.0025 0.0022 0.0017
DRNP(2)
0.185
RP(3)
0.5176 0.5158 0.5088 0.5030 0.4926 0.4846 0.4711
RU(4)
0.2854 0.2932 0.3170 0.3332 0.3578 0.3743 0.3993
DNR(5)
0.0071 0.0076 0.0092 0.0105 0.0126 0.0143 0.0171
0.1789 0.1614 0.1503 0.1345 0.1246 0.1108
Stay Duration
Figure 7.3a Effect of Stay Duration on Return Intentions (Age = 35 years)
1.0
0.9
Return Unlikely
Cum. Probability
0.8
0.7
0.6
0.5
Return Probable
0.4
0.3
0.2
0.1
Definitely Return, no plans
0.0
1
2
5
7
10
12
15
Stay Duration (Average Respondent, Age = 35)
Figure 7.3b Cumulative Probabilities: Stay Duration & Return Intentions
169
0.6
Probability (y|x)
0.5
0.4
0.3
0.2
0.1
0
Pr(y=DRPx):
1
2
5
7
10
12
15
0.0082 0.0072 0.0046 0.0034 0.0021 0.0015 0.0009
Pr(y=DRNPx): 0.2374 0.2227 0.1815 0.1566 0.1235 0.1043 0.0795
Pr(y=RPx):
0.5236 0.5235 0.5166 0.5065 0.4836 0.4638 0.4285
Pr(y=RUx):
0.2267 0.2418 0.2899 0.3237 0.3762 0.4117 0.4642
Pr(y=DNRx):
0.0041 0.0048 0.0074 0.0097 0.0145 0.0187 0.0269
Years Worked in Current Country
Figure 7.4 Effect of Work Experience in Current Country on Return Intentions
appear to have a stronger negative effect on return intentions than do increases in stay
duration. The probability of not returning (y = 4 or 5) increases by 0.07 for the first five
years of work experience, and then by 0.09 for the second five years, and finally by 0.10 for
the next five years after that. By comparison, the same figures for stay duration are 0.03,
0.04 and 0.05 respectively. The negative impact of foreign work experience on return
intentions provides empirical support for Wong’s learning-by-doing model of brain drain.
Whether a respondent has had any work experience in Turkey also appears to be an
important determinant of current return intentions, in addition to the amount of work
experience obtained in the host country. When a respondent has no full-time job experience
in Turkey (NWexpTUR=1), the probability of not returning (y = 4 or 5) increases by 0.08,
and is slightly higher for females (see Table 7.4 below).
170
Table 7.4 Marginal Effect of Having No Work Experience in Turkey
Probabilities:
DRP
y=1
DRNP
y=2
RP
y=3
RU
y=4
DNR
y=5
Total:
NWexpTUR=0
NWexpTUR=1
Difference 0 1
0.0043
0.0023
-0.0020
0.1751
0.1269
-0.0482
0.5145
0.4865
-0.0280
0.2982
0.3704
0.0722
0.0079
0.0139
0.0060
Males:
NWexpTUR=0
NWexpTUR=1
Difference 0 1
0.0055
0.0029
-0.0026
0.1963
0.1444
-0.0519
0.5204
0.4994
-0.0210
0.2716
0.3421
0.0705
0.0063
0.0112
0.0049
Females:
NWexpTUR=0
NWexpTUR=1
Difference 0 1
0.0023
0.0011
-0.0012
0.1273
0.0888
-0.0385
0.4869
0.4433
-0.0436
0.3697
0.4434
0.0737
0.0138
0.0233
0.0095
The correspondence analysis of the previous chapter (see Section 6.4.4) suggested
the possibility that respondents who returned to Turkey to work after obtaining foreign
degrees are less likely to return a second time. The dummy variable FFTJ_TUR takes on a
value of 1 for respondents completing their highest degree abroad if their first full-time job
(FFTJ) after completing their studies is located in Turkey. Table 7.5 shows the marginal
effects of working in Turkey immediately after completing foreign studies for each return
intentions category. The probability of not returning (y = 4 or 5) increases by 0.18, while
the more positive return intention categories—“definitely return, no plans” (DRNP: y = 2)
and “return probable” (RP: y = 3)—decrease in total by about the same amount. The
probability of choosing the “definitely return, no plans” category decreases by 0.10 for
male respondents compared to a decline of 0.07 for females, and the probability of
“probably returning” (RP) decreases by 0.11 for female respondents versus a decline of
0.07 for males.
These results (e.g., the negative impact of work experience in Turkey for
respondents with foreign degrees and the phenomenon of student non-return) have
important implications for the “brain circulation” hypothesis, which is pervasive in the
current literature on the impact of migratory flows. It appears that respondents who start
their work life abroad after completing their overseas studies are less likely to have strong
return intentions, and respondents with foreign degrees who start their work life in Turkey
171
are less likely to have plans for returning to Turkey again6. Those who make contributions
to Turkey during their stay abroad are also more likely to indicate they will return. This is
included in the model as the dummy variable contr, which takes on a value of 1 when
respondents have contributed either by making donations, taking part in lobbying activities
or by participating in activities such as attending conferences in Turkey. The effect of this
on the likelihood of returning is substantial: the probability of definitely returning increases
by 0.09. This suggests perhaps that those who are already likely to return are also those
contributing the most to Turkey through various activities.
Table 7.5 Marginal Effect of Working in Turkey Immediately after
Completing Overseas Studies
DRP
y=1
Probabilities:
DRNP
y=2
RP
y=3
RU
y=4
DNR
y=5
Total:
FFTJ_TUR=0
FFTJ_TUR=1
Difference 0 1
0.0040
0.0009
-0.0031
0.1693
0.0774
-0.0919
0.5122
0.4249
-0.0873
0.3061
0.4691
0.1630
0.0084
0.0278
0.0194
Males:
FFTJ_TUR=0
FFTJ_TUR=1
Difference 0 1
0.0051
0.0012
-0.0039
0.1900
0.0899
-0.1001
0.5190
0.4449
-0.0741
0.2792
0.4412
0.1620
0.0067
0.0229
0.0162
Females:
FFTJ_TUR=0
FFTJ_TUR=1
Difference 0 1
0.0021
0.0004
-0.0017
0.1225
0.0514
-0.0711
0.4827
0.3678
-0.1149
0.3780
0.5361
0.1581
0.0147
0.0442
0.0295
RP
y=3
RU
y=4
DNR
y=5
0.4699
0.5207
0.0508
0.4013
0.2698
-0.1315
0.0174
0.0062
-0.0112
Table 7.6 Marginal Effect of Contributions to Turkey
Probabilities:
DRP
y=1
contr=0
contr=1
Difference 0
0.0017
0.0056
0.0039
1
DRNP
y=2
0.1097
0.1978
0.0881
6
Toward the end of the survey questionnaire respondents were asked about the frequency of their
visits to Turkey for various purposes, including for educational and work endeavours. Unfortunately,
this part of the survey had a low response rate and could not be used to determine the degree to
which productive brain circulation is occurring on behalf of Turkey.
172
Effect of Initial Intentions:
Two dummy variables, init_UNSURE and init_RETURN, are included in the model
to determine whether differences in the initial intention of the respondent prior to his/her
venture abroad is important in determining his/her current intentions about returning to
Turkey. The reference variable is “stay”. Both the “return” and “undecided” variables are
negative and significant at the 1 percent significant level in the ordered probit model. Table
7.7 shows the marginal effects of initial return intentions with all other variables held at
their mean values.
Table 7.7 Marginal Effects of Initial Return Intentions, Professionals
DRP
y=1
DRNP
y=2
RP
y=3
RU
y=4
DNR
y=5
init_STAY = 1
init_UNSURE = 1
init_RETURN = 1
0.0001
0.0026
0.0078
0.0210
0.1371
0.2317
0.2509
0.4945
0.5237
0.6311
0.3536
0.2324
0.0968
0.0122
0.0044
Change in Probability:
init_STAY
init_UNSURE
init_UNSURE
init_RETURN
init_STAY
init_RETURN
0.0025
0.0052
0.0077
0.1161
0.0946
0.2107
0.2436
0.0292
0.2728
-0.2775
-0.1212
-0.3987
-0.0846
-0.0078
-0.0924
Probabilities:
The probability of definitely returning (y = 1, 2) increases by 0.22 for respondents
with an initial intention to return compared to those with an initial return intention of
staying abroad. The increase in the probability of definitely returning is lower (0.10) when
the comparison group is those who are initially unsure about returning. The probability of
being unlikely to return is quite high (0.63) for those whose initial intention is to stay in the
host country. The probabilities of definitely not returning and of return being unlikely
increases by 0.09 and 0.40 respectively, when respondents have initial “stay” intentions
compared to those with initial return intentions. These figures suggest that the initial or
prior intentions of individuals tend to shape their current intentions about whether to return
to Turkey or not. This tendency, however, appears to be strongest for those with initial
plans to remain abroad. These results may be reflecting the “self-fulfilling” tendency of
prior intentions and expectations: e.g., those who start out more determined from the outset
to make a career or succeed abroad will try harder to make this come true; they may also
tend to try to protect themselves psychologically from setbacks or initial adjustment
problems, and exhibit greater tolerance when they occur.
173
Effect of Family Support and Marriage to Foreign Spouse:
Respondents were asked about the degree of support (encouragement) that they
received from their families (parents, wife, and children) in the initial decision to work or
study abroad and in the decision to settle overseas permanently. Maximum likelihood
testing procedures were performed to determine whether the ordered family support
variables could be treated as interval7. On the basis of the LR test results for the ordered
probit model and the Wald test results for the multinomial logit model8, fam_sup1 and
fam_sup2 were included as interval variables in the models.
Family support for the initial decision (fam_sup1) is negative and significant
(
= 0.01) in the ordered probit model. This means that the probability of returning
increases when there is support for the initial decision to go abroad. In the analysis of the
previous chapter, it is clear that there is strong family support the initial decision to acquire
overseas study or work experience for a majority of respondents. This variable may be
indicative of the strength of ties to family in Turkey, which offers a possible explanation of
the negative sign on the fam_sup1 coefficient and higher probability of return.
The second “family support” variable is a measure of how much encouragement
the respondent believes that she/he would receive from her/his family for the decision to
settle abroad permanently. The interpretation of the positive and statistically significant
coefficient ( = 0.01) in the ordered probit model for the fam_sup2 variable is more clearcut. Respondents with greater family encouragement in the decision to settle abroad
permanently have a greater probability of not returning to Turkey. This outcome appears to
validate the importance of family encouragement in the decision to migrate, especially for
individuals coming from a traditional, family-oriented society such as Turkey. (This could
be compared with other country studies that contain “family” variables).
7
To illustrate: in performing the LR test, the model containing the ordinal variable fam_sup1 is
compared to the model that includes both fam_sup1 and all but two of the categories of fam_sup1. If
the restricted model leads to a loss in information, then the ordinal variable cannot be treated as an
interval variable (see Long and Freese, 2001: 268-9). Wald tests are performed instead for the
multinomial logit model since only the restricted model is required, which considerably speeds up
computation.
8
Test results:
fam_sup1 (ordered probit model): LR 2(2) = 5.16, Prob > 2 = 0.0757;
fam_sup1 (multinomial logit model): Wald 2(8) = 10.80, Prob > 2 = 0.2133
fam_sup2 (ordered probit model): LR 2(4) = 5.48, Prob > 2 = 0.2414;
fam_sup2 (multinomial logit model): Wald 2(16) = 20.84, Prob > 2 = 0.1848.
174
Table 7.8 Marginal Effect of Family Support and Marital Status
DRP
y=1
DRNP
y=2
RP
y=3
RU
y=4
DNR
y=5
Initial family support
fam_sup1
marginal effect
z-value
0.0019
(2.21)**
0.0413
(2.79)***
0.0206
(2.61)***
-0.0593
(-2.82)***
-0.0045
(-2.31)**
Family support for
permanent settlement
fam_sup2
marginal effect
z-value
-0.0016
(-3.11)***
-0.0362
(-5.28)***
-0.0181
(-4.25)***
0.0520
(5.43)***
0.0039
(3.49)**
0.0042
0.0012
-0.0030
0.1733
0.0910
-0.0823
0.5138
0.4465
-0.0673
0.3005
0.4388
0.1383
0.0081
0.0226
0.0145
Probabilities:
Marriage to foreign
spouse:
spousenat2=0
spousenat2=1
Difference 0 1
Another important consideration is marriage to a foreign spouse, which is given by
the dummy variable spousenat2. The sign of the coefficient on spousenat2 in the ordered
probit estimates is negative and statistically significant at the 1 percent significance level,
indicating a lower intention of returning. The marginal effects of the family support
variables and being married to a foreign spouse are presented in Table 7.7. Family support
for permanent settlement and marriage to a foreign spouse decrease the probability of
definitely returning by 0.037 and 0.085 respectively. Initial family support for overseas
study or work, on the other hand, tends to increase definite return intentions by 0.04. As
expected, marriage to a foreign spouse has a very large positive effect (0.14) on the
probability of “being unlikely to return”, which is much larger than the effect of family
support for settlement abroad (0.04).
Effect of Parental Education:
Differences in the social background of respondents, as reflected in the educational
attainment of their parents, were found to be statistically insignificant in determining
current return intentions. In the ordered probit estimates, “high school” was used as the
reference educational attainment category for each parent. No significant relationships were
found when the other categories of educational attainment are used as the reference. As a
result, parental education levels were not included in the final model. Although parental
175
education levels are not important in determining the likelihood of return of respondents, it
is clear that, as shown in Chapter Four, the socioeconomic background of individuals is
important in determining who leaves Turkey for study and work opportunities in another
country.
Effects of the Initial Reasons for Going:
Since initial return intentions appear to be important in determining current return
intentions, the initial reasons for going overseas may also provide important information
about who is planning to return and who is not. Only six of the possible twelve reasons
presented to the respondents are found to have statistical significance. They are the ones
included in the final model. Some of these factors become significant only when their
interactions with certain variables such as age, female and academic are controlled for.
The results from the estimated ordered probit model indicate that respondents are
more likely to return if their initial reason for going was any of the following: having a job
requirement in Turkey (whygo_C), prestige of overseas study (whygo_G), or to join spouse
(whygo_I). The first two are statistically significant at the 10 percent and the last at the 1
percent significance level. A positive, significant ( = 0.10) coefficient for the interaction
term between female and whygo_I (FxWHYGOI)9 and between female and whygo_C
(FxWHYGOC)10 indicates that these results hold for males. Male respondents are more
likely to return if they initially went abroad as a requirement or to be with their spouses.
The result for whygo_G (the prestige of overseas study), on the other hand, is moderated by
age (through a positive and significant coefficient of the term AGExWHYGOG at the 10
percent significance level) and strengthened if the respondent is working in academia
(through a negative and significant coefficient of the term ACADxWHYGOG at the 5
percent significance level).
9
The in-sample bivariate association between return intentions and whygo_C as measured by the
chi-square statistic 2(4) is 1.84 (Pr = 0.76) for females and 8.68 (Pr = 0.07), even though a greater
percentage of female respondents have indicated that their reason for going abroad is to be with their
spouses (23.1 percent versus 8.2 percent).
10
The percentage of females in the sample whose initial reason for going abroad was to fulfil a job
requirement in Turkey is approximately the same as that for males (21.7 percent versus 22.6
percent). Interestingly, the chi-square statistic between return intentions and whygo_C is significant
only for males ( 2(4) = 41.57, Pr = 0.00), and there is a clear tendency (based on an examination of
table percentages) for males who chose whygo_C as their reason for going abroad to have stronger
return inclination than those who did not.
176
As expected, respondents who left Turkey because of lifestyle preferences
(whygo_H) or due to political factors (whygo_K) are not likely to indicate strong return
plans. The coefficients of these variables are positive and statistically significant at the 5
percent and 10 percent significance levels respectively. Respondents who left because they
found facilities and equipment for doing research in Turkey to be inadequate (whygo_F) are
also less likely to be returning (significant at 1 percent). Table 7.9 below presents the
marginal effects of each reason on the probabilities of the return intention categories.
Table 7.9 Marginal Effects of the Initial Reasons for Going
Probabilities:
DRP
y=1
DRNP
y=2
RP
y=3
RU
y=4
DNR
y=5
Job requirement in Turkey
whygo_C=0
whygo_C=1
Difference 0 1
0.0033
0.0044
0.0011
0.1536
0.1760
0.0224
0.5048
0.5148
0.0100
0.3283
0.2970
-0.0313
0.0101
0.0078
-0.0023
Insufficient facilities, etc
whygo_F=0
whygo_F=1
Difference 0 1
0.0036
0.0028
-0.0008
0.1595
0.1412
-0.0183
0.5079
0.4973
-0.0106
0.3197
0.3471
0.0274
0.0094
0.0117
0.0023
Prestige of study abroad
whygo_G=0
whygo_G=1
Difference 0 1
0.0031
0.0036
0.0005
0.1481
0.1600
0.0119
0.5017
0.5082
0.0065
0.3364
0.3189
-0.0175
0.0107
0.0094
-0.0013
Lifestyle Preference
whygo_H=0
whygo_H=1
Difference 0 1
0.0042
0.0024
-0.0018
0.1728
0.1321
-0.0407
0.5136
0.4908
-0.0228
0.3012
0.3617
0.0605
0.0081
0.0130
0.0049
To be with spouse
whygo_I=0
whygo_I=1
Difference 0 1
0.0032
0.0086
0.0054
0.1515
0.2420
0.0905
0.5037
0.5233
0.0196
0.3313
0.2221
-0.1092
0.0103
0.0039
-0.0064
Escape Political Environment
whygo_K=0
whygo_K=1
Difference 0 1
0.0040
0.0026
-0.0014
0.1697
0.1366
-0.0331
0.5124
0.4941
-0.0183
0.3055
0.3544
0.0489
0.0084
0.0123
0.0039
Lifestyle preference has the greatest negative marginal effect on return intentions,
followed by getting away from the political environment and insufficient facilities for
conducting research in Turkey. The probability of not returning (y = 4 or 5) increases by
0.07 for those who have indicated lifestyle preference to be their reason for going abroad,
compared to 0.05 for political reasons and 0.03 for insufficient facilities. Respondents who
177
indicated they went abroad to be with their spouse have the highest return intentions: the
probability of choosing one of the “definitely return” categories increases by 0.096
(0.0054+0.0905), compared to 0.024 for those who went because of a job requirement in
Turkey and 0.017 for those who went abroad to take advantage of study opportunities.
Effect of Work, Social and Standard of Living Assessment:
Respondents were asked to assess in general terms their personal work environment
(e.g., job satisfaction), the social aspects of life (e.g., friendships, social relations) and
standard of living in their current country of residence versus that in Turkey on a 5-point
scale ranging from “much worse” to “much better” (see Section 6.2.9 for details). Work and
standard of living assessments (work_assess and SOL_assess) are skewed toward the
“better” or “much better” categories. These two variables are positively associated with
lifestyle preferences. The distribution of the social assessment variable appears not to be as
slanted toward extreme points, although it is tilted toward the “worse” categories. The
work_assess variable was not statistically significant and was therefore excluded from the
model11. The coefficients of social_assess and SOL_assess12 are positive and statistically
significant at the 5 percent and 1 percent significance levels respectively, indicating a
decrease in return intentions when more positive assessments are made about conditions
abroad compared to Turkey.
The marginal effects are given in Table 7.10. It is clear that positive assessments of
living conditions abroad lead to greater decreases in the probability of indicating return
intentions than do positive assessment about social conditions abroad. Figures 7.5 and 7.6
give the cumulative probabilities associated with each value (1 to 5) that the social_assess
and SOL_assess variables take on. Areas toward the bottom represent more definite plans
and areas at the top represent more definite non-return intentions. These diagrams also
show that standard of living assessments have a greater impact on return intentions.
11
Wald test of significance:
2
(1) = 0.12, Prob >
2
= 0.7321.
12
The likelihood ratio test results for whether the ordinal variables can be treated as interval are as
follows: social_assess: LR 2(4) = 2.95, Prob > 2 = 0.5663;
SOL_assess: LR 2(4) = 11.58, Prob > 2 = 0.0207.
The likelihood ratio test results indicate that social_assess can be used at the interval level, but
treating SOL_assess as an interval variable leads to loss of information. Despite this, both variables
were included as interval variables in order to keep the model simple. This did not lead to a change
in the qualitative results.
178
Table 7.10 Marginal Effects of Social and Standard of Living Assessments
Probabilities:
Social Assessment
social_assess
DRP
y=1
DRNP
y=2
RP
y=3
RU
y=4
DNR
y=5
-0.0011
-0.0237
-0.0118
0.0340
0.0026
(-2.09)**
(-2.42)**
(-2.29)**
(2.42)**
(2.25)**
Standard of Living Assessment
SOL_assess
-0.0014
-0.0304
-0.0152
0.0436
0.0033
(-2.21)**
(-2.78)***
(-2.57)***
(2.79)***
(2.36)**
Notes: Figures in parentheses are z-statistics. The table summarizes information
from Table B.3 in Appendix B.
1.0
0.9
Cum. Probability
0.8
0.7
Pr(y=DNRx):
0.6
Pr(y=RUx):
0.5
Pr(y=RPx):
0.4
Pr(y=DRNPx):
0.3
Pr(y=DRPx):
0.2
0.1
0.0
1
2
much
worse
3
4
5
much
better
Social Assessment
Figure 7.5 Cumulative Probabilities: Social Assessment of Life Abroad
1.0
0.9
Cum. Probability
0.8
0.7
Pr(y=DNRx):
0.6
Pr(y=RUx):
0.5
Pr(y=RPx):
0.4
Pr(y=DRNPx):
0.3
Pr(y=DRPx):
0.2
0.1
0.0
0
much
worse
1
2
3
4
Standard of Living Assessment
5
much
better
Figure 7.6 Cumulative Probabilities: SOL Assessment of Life Abroad
179
Level and Location of Highest Degree:
It is expected that higher levels of formal education received abroad (e.g., PhD
level education), corresponding to a greater degree of country or institution-specific
specialization, will result in a lower tendency for returning to Turkey.
While the highest degree held by the respondent has no significant effect on the
return intentions of respondents, where the highest degree is received is statistically
significant at the 1% significance level. Those who have received their highest degree from
a Turkish university are more likely to indicate they will return than those whose highest
degree is a foreign degree. Therefore, higher education received abroad, regardless of the
level, is important in the decision to return or stay13. This also means that student nonreturn is a potentially more serious problem for Turkey.
Table 7.11 Marginal Effect of Highest Degree being a PhD from a
Turkish University
Probabilities:
HDPHDxTUR=0
HDPHDxTUR=1
Difference 0 1
DRP
y=1
0.0033
0.0126
0.0093
DRNP
y=2
0.1539
0.2859
0.1320
RP
y=3
RU
y=4
DNR
y=5
0.5050
0.5163
0.0113
0.3277
0.1826
-0.1451
0.0100
0.0025
-0.0075
Effect of the Field of Study: Capital Intensive versus Non-Capital Intensive Fields
According to Chen and Su (1995), students in capital-intensive fields (where a
complementary relationship exists between the education received and the physical and
social capital stock of the host country) will be less likely to return than students in non
capital-intensive fields (such as law, sociology and the like). To test this, the highest degree
fields were arranged into three groups: HDnew1 (architecture, economics and
administrative sciences); HDnew2 (education, language, sociology, art) and HDnew3
(engineering, mathematics, science and medicine). The reference category is HDnew2. In
the ordered probit analysis, the coefficients on HDnew1 and HDnew3 are both positive and
statistically significant at the 1 percent significance level, indicating that those in the “hard
sciences” or more capital intensive fields (HDnew3), as defined by Chen and Su, are more
13
The analysis was done with the dummies HD_TUR (highest degree is from Turkey), FHD_BS
(highest degree is a foreign bachelors degree), FHD_MS (highest degree is a foreign master’s
degree) and FHD_PHD (highest degree is a foreign doctoral degree).
180
likely to stay abroad compared to those in education, language, and so on. However, the
least likely to return are those who hold their highest degrees in architecture, economics or
administrative sciences. Economic instability and the crisis environment in Turkey, which
has had important repercussions in the banking and finance sectors, offers an explanation
for this.
Table 7.12 Marginal Effects of Fields of Study, Professionals
DRP
y=1
DRNP
y=2
RP
y=3
RU
y=4
DNR
y=5
HDnew2 = 1 (educ / lang / soc / art)
HDnew1 = 1 (arch / econ / admin)
HDnew3 = 1 (engin / math / science / medic)
0.0063
0.0012
0.0029
0.2100
0.0907
0.1430
0.5225
0.4461
0.4985
0.2557
0.4393
0.3443
0.0054
0.0226
0.0114
Change in Probability:
HDnew2
HDnew1
HDnew2
HDnew3
HDnew1
HDnew3
-0.0051
-0.0034
0.0017
-0.1193
-0.0670
0.0523
-0.0764
-0.0240
0.0524
0.1836
0.0886
-0.0950
0.0172
0.0060
-0.0112
Probabilities:
On-the-Job Training and Formal Training:
One of the main arguments set forth by Chen and Su (1995) to explain the Taiwanese
brain drain to Japan is on-the-job training. Training received on the job abroad after
completing overseas studies is expected to instill skills that are given a higher premium in
the country in which they are received. This wage differential, in turn, is supposed to favor
the host country and keep foreign workers abroad. To test on-the-job training as a cause of
brain drain directly, respondents were asked whether they have received informal on-thejob training at their current overseas jobs. Nearly 60 percent of respondents have received
some on-the-job training, and for 10 percent, this training is specific to the organization and
cannot be easily transferred to other organizations.
The following dummy variables were constructed: OTJT1 (did not receive on-the-job
training), OTJT2 (general), OTJT3 (specific to industry), and OTJT4 (specific to
organization). The signs on these variables were as expected. With “no on-the-job training”
as the reference category, the coefficients of the “general”, “specific to industry” and
“specific to organization” were positive but not statistically significant. This indicates that
on-the-job training does not have explanatory power for differences in return intentions. On
the other hand, formal training specific to the organization (represented by FTr4) is positive
and statistically significant at the 10 percent level indicating that respondents who have
181
gone through formal specialized training are less likely to return. The marginal effects are
given below in Table 7.13. The probability of not returning to Turkey (y = 4 or 5) increases
by 0.14 while the probability of definitely returning (y = 1 or 2) falls by 0.08. Firm-specific
training as a cause of brain drain is limited to a very small proportion of participants in the
sample (3.8 percent).
Table 7.13 Marginal Effect of Organization-Specific Formal Training
Probabilities:
DRP
y=1
FTr4=0
FTr4=1
Difference 0
0.0037
0.0012
-0.0025
1
DRNP
y=2
0.1618
0.0892
-0.0726
RP
y=3
RU
y=4
DNR
y=5
0.5090
0.4439
-0.0651
0.3164
0.4425
0.1261
0.0092
0.0231
0.0139
R&D Activities and Return Intentions:
R&D activities may be grouped into three basic categories: basic research, applied
research, and development (OECD, 1994). Respondents were asked what percentage of
time they devoted to job-related activities that also included R&D. If respondents spent at
least half their time on R&D activities, they were labeled R&D workers and placed in the
R&D category. Again, a dummy variable was used: R&D (1 if R&D worker, 0 otherwise).
About 40 percent of those engaged in research and development activities are
academicians (166/421*100). The R&D dummy variable was not significant at any
conventional significance level. This is not an expected result since R&D activities are
given a greater premium abroad and those engaged in R&D are expected to be less willing
to return. The problem here may be how respondents interpreted the different job
activities14.
Academic vs. Non-Academic Professions:
In the following analysis, “academic” refers to individuals who are teaching and/or
doing research at a 4-year university or at research centers and medical schools affiliated
with a 4-year university. Academicians make up 30 percent of the overseas labor force
sample. A dummy variable, academic2, is used (1 for academic, 0 for non-academic) to
14
The respondents were also asked if they had any patented inventions. A dummy variable ‘patent’
was constructed (1 = ‘has patent’; 0 = ‘does not have patent’) to determine whether return intentions
for individuals with patents differed from those without. The coefficient for this variable was not
statistically significant.
182
determine whether the return intentions of the academicians in the sample differ from the
non-academic labor force. This variable is not found to be statistically significant, although
it is an important modifier or interaction variable in the analysis of push and pull factors.
Table 7.14 Marginal Effect of Working in Academia or a Research
Institution
Probabilities:
DRP
y=1
academic2=0
academic2=1
Difference 0
0.0041
0.0020
-0.0021
1
DRNP
y=2
0.1704
0.1194
-0.0510
RP
y=3
RU
y=4
DNR
y=5
0.5127
0.4798
-0.0329
0.3046
0.3836
0.0790
0.0083
0.0153
0.0070
Effects of Various Push and Pull Factors:
Income or wage differentials are cited as among the most important reasons for the
brain drain. Many elaborate models of the brain drain found in the literature are based on
explaining how this differential occurs. We use a relatively simple test of whether income
differentials are important. To determine whether income differentials are important, we
include a dummy variable that takes on the value 1 when a respondent indicates that a
higher salary or wage is a “very important” or “important” reason for not returning or
postponing returning to Turkey on a 5-point Likert scale. The disadvantage of this construct
is that it is a subjective measure (for further elaboration see Section 7.3.1). The income
variable was found to be statistically significant and therefore excluded from the final
model.
Of the twelve “push” factors presented to participants, only four were found to be
statistically significant: pushC (limited job opportunity in specialty), pushD (no opportunity
for advanced training), pushF (lack of financial resources for business) and pushK
(economic instability and uncertainty). Having limited job opportunities in specialization
carries greater significance for those in academia or research-oriented institutions (given by
dummy variable academic2). While the coefficient of pushC is not statistically significant,
the coefficient of the interaction between pushC with academic2 (ACADxpushC) is positive
and significant at the 5 percent significance level. A significant interaction effect (at the 1
percent significance level) was found between having little or no opportunities for
advanced training (pushD) and the age of participants (AGExpushD). Respondents who
indicated that the lack of financial resources and opportunities for starting a business in
183
Turkey (pushF) was an important push factor for them are more likely to be returning. The
coefficient on pushF is negative and significant at the 10 percent significance level.
Economic instability and uncertainty, on the other hand, appears to have a strong negative
effect on return intentions (statistically significant at 1 percent). The marginal effects on
each of the significant push factors are presented in Table 7.15:
Table 7.15 Marginal Effects of Various Push Factors
Probabilities:
DRP
y=1
DRNP
y=2
RP
y=3
RU
y=4
DNR
y=5
Limited job opportunity in
specialty (academic2=1)
pushC=0
pushC=1
Difference 0 1
0.0018
0.0013
-0.0005
0.1121
0.0933
-0.0188
0.4725
0.4497
-0.0228
0.3968
0.4339
0.0371
0.0168
0.0218
0.0050
No opportunity for advanced
training
pushD=0
pushD=1
Difference 0 1
0.0036
0.0030
-0.0006
0.1613
0.1454
-0.0159
0.5088
0.5000
-0.0088
0.3171
0.3405
0.0234
0.0092
0.0111
0.0019
Lack of financial resources for
starting a business
pushF=0
pushF=1
Difference 0 1
0.0031
0.0046
0.0015
0.1494
0.1812
0.0318
0.5025
0.5165
0.0140
0.3344
0.2903
-0.0441
0.0106
0.0074
-0.0032
Economic Instability
pushK=0
pushK=1
Difference 0 1
0.0086
0.0030
-0.0056
0.2423
0.1462
-0.0961
0.5233
0.5005
-0.0228
0.2219
0.3393
0.1174
0.0039
0.0110
0.0071
It is clear that the greatest negative effect on return intentions is due to economic
instability and uncertainty: the probability of not returning (y = 4 or 5) increases by 0.12 for
those indicating that pushK was a “very important” or “important” push factor (which
accounts for 85 percent of respondents in the sample). For those working in academic or
research-oriented organizations, having no job opportunities in their specialization in
Turkey increases the probability of not returning by 0.04. Having no advanced training
opportunities increases the probability of non-return by 0.03 for the average respondent.
However, this negative impact of pushD on return intentions is greater for older
respondents (see Figure 7.7). On the other hand, the probability of definitely returning
increases by 0.03 for those indicating that the lack of business opportunities in Turkey is an
important push factor. This may be reflecting the fact that the percentage of non-academic
184
respondents who indicated pushF is an important factor is much greater than that of
academics (33 percent versus 22 percent), who have a much higher non-return probability.
0.7
Probability of Non-Return
0.6
pushD=1
0.5
0.4
pushD=0
0.3
0.2
pushD0
1
0.1
0
-0.1
25
30
35
40
45
50
-0.2
Age
RU+DNR (pushD=0)
RU+DNR (pushD=1)
0 to 1
Figure 7.7 Effect of the Interaction between Age and Importance of Advanced
Training Opportunities on the Probability of Not Returning (y = 4 or 5)
The number of significant pull factors is greater compared to the push factors.
Eight of the twelve pull factors presented to participants are found to be statistically
significant. Since respondents in the target group are residing outside Turkey, it is natural
that factors in their immediate environment will have a greater impact on their current
return intentions. Table 7.16 gives the marginal effects of the significant pull factors. The
greatest negative impact on the probability of returning is from family considerations (pullI
and pullJ), but there are gender differences. Spouse’s job or preference appears to play a
greater role in the stay decision of males. Greater opportunities for developing specialty
(pullE), a more satisfying social and cultural life (pullG), proximity to research centers
(pullH) and a more organized, ordered environment (pullF) follow. The other two pull
factors—the need to finish or complete an overseas project (pullK) and other reasons
(pullL) for male respondents—are associated with positive return intentions. For males, the
effect of “other” factors is mainly that of wanting to return to complete military service in
Turkey.
185
Table 7.16 Marginal Effects of Various Pull Factors
Probabilities:
DRP
y=1
DRNP
y=2
RP
y=3
RU
y=4
DNR
y=5
Greater opportunity to
develop specialty
pullE=0
pullE=1
Difference 0 1
0.0061
0.0028
-0.0033
0.2062
0.1414
-0.0648
0.5221
0.4975
-0.0246
0.2600
0.3467
0.0867
0.0057
0.0116
0.0059
More organized, ordered
environment
pullF=0
pullF=1
Difference 0 1
0.0051
0.0031
-0.0020
0.1898
0.1499
-0.0399
0.5189
0.5027
-0.0162
0.2795
0.3337
0.0542
0.0067
0.0105
0.0038
More satisfying social /
cultural life
pullG=0
pullG=1
Difference 0 1
0.0043
0.0019
-0.0024
0.1756
0.1151
-0.0605
0.5146
0.4756
-0.0390
0.2976
0.3913
0.0937
0.0079
0.0162
0.0083
Proximity to research and innovation
centers (academic2=1)
pullH=0
0.0029
pullH=1
0.0012
Difference 0 1
-0.0017
0.1434
0.0904
-0.0530
0.4988
0.4456
-0.0532
0.3436
0.4401
0.0965
0.0113
0.0228
0.0115
Spouse’s preference or job
pullI=0
pullI=1
Difference 0 1
0.0049
0.0016
-0.0033
0.1861
0.1075
-0.0786
0.5179
0.4675
-0.0504
0.2840
0.4055
0.1215
0.0070
0.0179
0.0109
Better educational
opportunities for children
pullJ=0
pullJ=1
Difference 0 1
0.0050
0.0019
-0.0031
0.1877
0.1161
-0.0716
0.5184
0.4767
-0.0417
0.2821
0.3894
0.1073
0.0069
0.0159
0.0090
Need to finish / continue with
current project
pullK=0
pullK=1
Difference 0 1
0.0026
0.0149
0.0123
0.1370
0.3064
0.1694
0.4944
0.5103
0.0159
0.3537
0.1664
-0.1873
0.0123
0.0021
-0.0102
Other pull reason (male=1)
pullL=0
pullL=1
Difference 0 1
0.0042
0.0148
0.0106
0.1731
0.3059
0.1328
0.5137
0.5105
-0.0032
0.3009
0.1668
-0.1341
0.0081
0.0021
-0.0060
186
Effect of Difficulties Faced Abroad and Adjustment Factors
The main difficulty with life abroad that was statistically significant ( = 0.05) in
the empirical analysis is that of missing one’s family in Turkey (difabrdA). The probability
of returning (y = 1 or 2) increases by 0.05 for those who indicate that missing family is one
of the difficulties they have faces while abroad. “Missing family” was an important
difficulty for a great proportion of respondents in the sample (83%). Previous experience
and involvement in a Turkish student association also have a similar, but slightly greater
impact on return intentions. The greater return intentions associated with these adjustment
factors may be due to the fact that respondents who indicate they have had difficulties
abroad also have to adjust compared to those who indicate they had no difficulties and
therefore did not need to adjust.
Table 7.17 Marginal Effects of Difficulties Faced Abroad and Adjustment Factors
Probabilities:
DRP
y=1
DRNP
y=2
RP
y=3
RU
y=4
DNR
y=5
Difficulty: missing family
difabrdA=0
difabrdA=1
Difference 0 1
0.0020
0.0039
0.0019
0.1199
0.1674
0.0475
0.4803
0.5115
0.0312
0.3827
0.3086
-0.0741
0.0152
0.0086
-0.0066
Adjustment factor: previous
experience
adj_A=0
adj_A=1
Difference 0 1
0.0025
0.0055
0.0030
0.1327
0.1967
0.0640
0.4912
0.5204
0.0292
0.3608
0.2711
-0.0897
0.0129
0.0063
-0.0066
Adjustment factor: TurkishStudent Association
adj_C=0
adj_C=1
Difference 0 1
0.0034
0.0070
0.0036
0.1557
0.2197
0.0640
0.5060
0.5234
0.0174
0.3251
0.2451
-0.0800
0.0098
0.0049
-0.0049
Effect of Language of Instruction in High School
The effect of foreign language high school instruction was looked at with the
dummy variable HSsciTUR, which takes on a value of 1 when language instruction for
science courses is Turkish. However, this variable is positively associated with difficulties
faced abroad (difabrdA) and previous experience as an adjustment factor (adj_A), as well as
other factors. As a result it is statistically insignificant in the model. In a model with only
gender, initial intentions and stay duration, HSciTUR becomes statistically significant at the
5 percent level.
187
Effect of Last Impressions
Return intentions may be shaped by the last impression from the latest trip to
Turkey. In this section we consider the effect of the last visit made to Turkey on the return
intentions of participants. A visit to Turkey made after a long period of time abroad may
radically change an individual’s perceptions about conditions in Turkey, either for the
better or for the worse. Whatever the case, these personal observations lead to changes in
the probability of returning. The probability of returning (y = 1 or 2) decreases by about
0.04 for those who were negatively effected by their last trip to Turkey, and increases by
0.22 for those who were left with more positive impressions. From this, it appears that
positive impressions appear to have a greater impact on the probability of returning.
The effect of the September 11, 2001 terrorist attacks in New York is also
considered in this section. The effect, in general, is to increase return intentions (sept11_inc
is negative and statistically significant at the 5 percent significance level. The probability of
returning (y = 1 or 2) increases by 0.07. For a small minority of respondents, Sept.11 had
the opposite effect on return intentions (sept11_dec is not statistically significant and is
therefore excluded from the final model). In one participant’s opinion:
Experiencing first hand both the earthquake in Turkey and the 9/11 attacks in NYC,
my determination of staying in the US has grown even stronger. The organization of
the rescue efforts, the value to human life, the role of gov't and many other aspects that
influence our lives directly are far superior in this country then my home country.
Table 7.18 Marginal Effect of the Last Visit to Turkey and of September 11
Probabilities:
DRP
y=1
DRNP
y=2
RP
y=3
RU
y=4
DNR
y=5
Last visit to Turkey
decreased return intentions:
lastvis1=0
lastvis1=1
Difference 0 1
0.0040
0.0025
-0.0015
0.1687
0.1337
-0.0350
0.5120
0.4920
-0.0200
0.3068
0.3590
0.0522
0.0085
0.0128
0.0043
Last visit to Turkey
increased return intentions:
lastvis3=0
lastvis3=1
Difference 0 1
0.0029
0.0204
0.0175
0.1435
0.3479
0.2044
0.4988
0.4934
-0.0054
0.3434
0.1369
-0.2065
0.0113
0.0014
-0.0099
Sept. 11 increased return
intentions:
sept11_inc=0
sept11_inc=1
Difference 0 1
0.0033
0.0070
0.0037
0.1525
0.2196
0.0671
0.5043
0.5234
0.0191
0.3298
0.2451
-0.0847
0.0102
0.0049
-0.0053
188
7.5 Determinants of the Return Intentions of Turkish Students
The previous section examined the determinants of return intentions of Turkish
nationals who are currently working abroad. In this section, the results of the empirical
investigation for students are presented. The focus is on the return intentions of Turkish
students studying at higher education institutions in different parts of the world. Much of
the analyses presented in the previous section are in agreement with that of students; thus, a
more brief treatment of the results will follow. The same estimation strategies and
methodologies apply for the investigation of the return intentions of Turkish students.
Gender and Age Effects:
Unlike the results for the overseas working population, gender and age do not
appear to be significant in explaining differences in return intentions for the overseas
Turkish student population. The coefficients on the “female”, “age”, and “agesq” variables
are not statistically significant at any of the conventional significance levels. This result
continues to hold when the stay duration variable is excluded.
Table 7.19 Marginal Effect of Gender, Students
R_BS
R_IAS R_NSAS
Probabilities:
RP
RU
DNR
y=1
y=2
y=3
y=4
y=5
y=6
0.0007
0.0005
-0.0002
0.0678
0.0532
-0.0146
0.4523
0.4179
-0.0344
0.3633
0.3867
0.0234
0.1141
0.1392
0.0251
0.0018
0.0026
0.0008
Gender
female=0
female=1
Difference 0
1
Stay Duration:
The stay duration variable is positive and statistically significant at the 1 per cent
significance level. As the length of stay in the host country increases, the tendency to “not
return to Turkey” also increases. This is as expected, since time helps overcome adjustment
problems, if they exist. As time passes, ties to Turkey may weaken while ties to the country
of study may strengthen (e.g. brain drain caused by marrying a national of the host country
may increase, but this is tested with a separate variable). Figure 7.8 gives the marginal
effects of different stay durations for each return intention category.
189
0.60
Prob. (y|x)
0.50
0.40
0.30
0.20
0.10
0.00
1
2
5
7
10
12
15
17
Pr(y=R_BSx):
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
Pr(y=R_IASx):
0.08
0.07
0.04
0.03
0.02
0.01
0.00
0.00
Pr(y=R_NSASx): 0.48
0.46
0.38
0.33
0.25
0.20
0.14
0.11
Pr(y=RPx):
0.34
0.36
0.40
0.42
0.43
0.42
0.39
0.36
Pr(y=RUx):
0.10
0.11
0.17
0.21
0.29
0.34
0.43
0.48
Pr(y=DNRx):
0.00
0.00
0.00
0.01
0.01
0.02
0.03
0.05
Stay Duration
Figure 7.8 Effect of Stay Duration on Return Intentions, Students
Notes: R_BS: return as soon as possible without completing studies; R_IAS: return
immediately after completing studies; R_NSAS: definitely return but not soon after
completing studies; RP: probably return RU: return unlikely; DNR: definitely not
return
Effect of Initial Intentions:
Initial intentions about whether to return to Turkey prior to starting overseas studies
are important in determining current return intentions. A little more than half the of the
students sampled intended to return, while one out of every ten student intended not to
return (stay in current country). The same dummy variables as in the previous section,
INIT_STAY and INIT_UNSURE, are used in the model, the reference variable being the
“intention to return”. The coefficients on both variables are positive and statistically
significant ( = 0.01), which indicates that those who have indicated that they will “stay” in
the current country or are “unsure” about returning are more likely to indicate that their
current intention is to “not return”. The probability of not returning (y = 5, 6) increases by
0.32 when initial intention changes from “stay” to “unsure” and by 0.38 when the change is
190
from “stay” to “return”. These large effects suggest that initial determination becomes an
important factor in shaping current intentions for Turkish students.
Table 7.20 Marginal Effects of Initial Return Intentions, Students
Probabilities:
init_STAY = 1
init_UNSURE = 1
init_RETURN = 1
Change in Probability:
init_STAY
init_UNSURE
init_UNSURE
init_RETURN
init_STAY
init_RETURN
R_BS
R_IAS
R_NSAS
RP
RU
DNR
y=1
y=2
y=3
y=4
y=5
y=6
0.0000
0.0003
0.0039
0.0426
0.1278
0.3860
0.3785
0.4038
0.4495
0.1638
0.0402
0.0036
0.0017
0.1089
0.5132
0.3039
0.0715
0.0007
0.0003
0.0014
0.0017
0.0387
0.0663
0.105
0.2582
0.1272
0.3854
0.0253
-0.0999
-0.0746
-0.2857
-0.0923
-0.3780
-0.0366
-0.0029
-0.0007
Effect of Family Support:
The student sample was also asked the degree that they felt that their families
supported them in the initial decision to study abroad and whether they would support them
in the decision to settle abroad permanently. For the initial decision to study abroad, threequarters of the student sample indicated that their families were very supportive. In general,
this initial support does not have any statistical significance with respect to the current
intention to return. Compared to the initial decision to study abroad, family encouragement
to settle abroad is considerably less, although it is still high (53% of the sample).
Initially, dummy variables for each category were included in the model as
regressors. Since the first three categories “actively discourage”, “not very supportive” and
“not sure” are not statistically different from each other, they are combined into the broader
category FAMSUP2_NS: “not supportive”, which is used as the reference category. The
same is done for the “somewhat supportive” and “most likely supportive” categories since
they are also not statistically different from each other. They are combined into a new
“somewhat supportive” category: FAMSUP2_SS. Only the “definitely not support”
category is not changed (FAMSUP2_DS). The signs on the FAMSUP2_SS and
FAMSUP2_DS dummy variables are positive and statistically significant at the 5 percent
and 1 percent significance level respectively. Greater family encouragement to settle abroad
results in a greater tendency to indicate non-return intentions, and vice versa. The marginal
effects of the family support variables are given below. Compared to respondents whose
families are not supportive (NS), the likelihood of not returning (y = 5 or 6) increases by
191
0.04 for those whose families are somewhat supportive (SS), and by 0.08 for those whose
families are definitely supportive (DS).
Table 7.21 Marginal Effects of the Family Support Variables, Students
Probabilities:
FAMSUP2_NS=1
FAMSUP2_SS=1
FAMSUP2_DS=1
R_BS
R_IAS
R_NSAS
RP
RU
DNR
y=1
y=2
y=3
y=4
y=5
y=6
0.0012
0.0006
0.0003
0.0921
0.0616
0.0411
0.4931
0.4387
0.3808
0.3270
0.3732
0.4061
0.0856
0.1239
0.1679
0.0010
0.0021
0.0038
-0.0006
-0.0003
-0.0009
-0.0305
-0.0205
-0.051
-0.0544
-0.0579
-0.1123
0.0462
0.0329
0.0791
0.0383
0.044
0.0823
0.0011
0.0017
0.0028
Change in Probability:
NS to SS
SS to DS
NS to DS
Effects of Parents’ Education:
Parents’ educational levels were included in the ordered probit model as possible
socioeconomic background indicators for the respondents. A dummy variable was
constructed for each level of education and different levels of education were used as
reference to determine whether any significant differences existed in the return intentions of
students with different family backgrounds. None of the parents’ education level dummies
were statistically significant except for the master’s level for fathers’ educational attainment
( = 0.05). Again, as for the working population sample, there was no a priori reason to
believe that we would find significant effects for these two social background variables. As
shown in the previous chapter, the student sample also comes from highly educated
backgrounds. Three-quarters of female students and two-thirds of male students have
fathers who possess a bachelor’s or higher degree. These are the same percentages as for
the working population sample. Mothers’ educational attainments, on the other hand, are
slightly higher for the student sample (51% vs. 47% for female respondents and 41% vs.
34% for male respondents).
Effect of Academic Conditions:
Students were asked to compare their academic environments in their current country
of study to that in Turkey. The great majority (close to 90 per cent) of students indicated
that academic conditions were either “better” or “much better”. A dummy variable was
constructed for each assessment category, and only the “much worse” category appeared
statistically significant at the 5 per cent significance level with reference to the other
192
categories. However, only two individuals chose the “much worse” category, and when this
category was chosen as the reference, none of the other categories were statistically
significant. This indicates that the academic assessment variables do not have any
explanatory power and may be excluded from the model.
Effect of Social Conditions:
In the previous section, social environment was found to be important in explaining
differences in return intentions for the working population. Hence, it is expected that this
will be true for the student sample as well. A third of respondents have indicated that their
current social environment is “neither better nor worse” than it was in Turkey, and a
significant number (43 per cent) indicate that it is “worse” or “much worse”.
The above categories above were reduced to three (not counting the “don’t know”
category) by combining the “worse” and “much worse” categories, and the “better” and
“much better” categories. With “much worse” as the reference category, both the “neither
better nor worse” and “better” categories are positive and statistically significant at the 1
per cent significance level. When the reference category is “much better”, both the “neither
better nor worse” and “worse” dummy variables are negative and statistically significant, at
the the 5 per cent and 1 per cent significance levels respectively. As before, the social
environment is found to be an important determinant of current return intentions. Those
who are less satisfied with their social conditions abroad are more likely to indicate that
they will return.
Standard of Living Assessment:
Students were also asked to assess their standard of living using the same scale as
above. The distribution of responses is tilted toward the “much better” end of the scale.
Since the coefficients of the “much better” and “better” dummy variables are not
statistically different from each other, they are combined. Similarly, the first four categories
can also be combined into a single category because they are statistically insignificant with
respect to each other. This latter variable is used as the reference. The coefficient of the
“standard of living is better” variable (SOL_B) is positive and statistically significant at the
5 percent significance level. Not surprisingly, once again, students who assess their
standard of living abroad as being better or much better than in Turkey show greater
intention to stay (not return).
193
Table 7.22 Marginal Effects of Social and Standard of Living Assessments, Students
Probabilities:
R_BS
R_IAS
R_NSAS
RP
RU
DNR
y=1
y=2
y=3
y=4
y=5
y=6
0.0004
0.0011
0.0007
0.0458
0.0882
0.0424
0.3964
0.4877
0.0913
0.3986
0.3325
-0.0661
0.1555
0.0893
-0.0662
0.0032
0.0011
-0.0021
0.0009
0.0005
-0.0004
0.0776
0.0557
-0.0219
0.4708
0.4245
-0.0463
0.3482
0.3826
0.0344
0.1011
0.1343
0.0332
0.0014
0.0024
0.0010
Social Assessment: Worse or
Much Worse
soc_W=0
soc_W=1
Difference 0
1
Standard of Living
Assessment: Better or Much
Better
SOL_B=0
SOL_B=1
Difference 0
1
Turkish Student Association Membership:
More than half the students responding to the survey belong to a Turkish student
association or society (TSA) at their institution of study (see the Table below). Membership
in these cultural associations turns out to be an important determinant of return intentions.
The coefficient of the dummy variable for membership (TSA_member) is negative and
statistically significant at the 1 per cent significance level, indicating that students who are
members of TSAs are more likely to have return intentions. This probably reflects a
preference on the part of TSA members to be with fellow nationals compared to nonmembers and is possibly an indication of stronger “cultural ties” to Turkey.
If a student is not a member of a TSA, this is because of personal choice or because
no TSA exists. Not being a member by choice and not being a member because no TSA
exists were not statistically different from each other and were, therefore, used combined as
the reference category.
Table 7.23 Marginal Effect of Turkish Student Association Membership, Students
Probabilities:
Turkish Student Association
membership
TSA_member=0
TSA_member=1
Difference 0
1
R_BS
R_IAS
R_NSAS
RP
RU
DNR
y=1
y=2
y=3
y=4
y=5
y=6
0.0004
0.0008
0.0004
0.0511
0.0709
0.0198
0.4121
0.4583
0.0462
0.3901
0.3586
-0.0315
0.1435
0.1098
-0.0337
0.0020
0.0016
-0.0004
194
Effects of the Field of Study:
In the previous section on the return intentions of Turkish professionals, the Chen
and Su (1995) hypothesis that on-the-job training causes “brain drain” was tested. Chen and
Su used a dummy for capital-dependent disciplines, which they determined to be medicine,
engineering and business. In their econometric analysis, they found that capital dependent
disciplines suffered more from brain drain than non-capital dependent disciplines. The
same dummy variable for capital-dependent disciplines is constructed in our analysis to see
if the same result will hold for the sample of Turkish students currently studying abroad.
This dummy variable turned out to be statistically insignificant15.
Effect of the Initial Reasons for Going:
The initial reasons for pursuing overseas studies also determine who is more likely to
return immediately after completing their studies (Table 7.24). The greatest positive
marginal effect on the probability of returning immediately after finishing studies is when
the main reason why respondents have gone abroad is to be with their spouse or families:
the probability of returning immediately increases by 0.11. When there is compulsory
service or job requirement—such as when higher education institutions in Turkey require
foreign degrees before they grant tenure positions—the probability of returning
immediately increases by 0.03. This is one of the important “push” factors that cause many
who are contemplating academic careers in Turkey to go abroad to get foreign higher level
degrees. While the probability of return increases when respondents have left because of a
job requirement, many do not have immediate return plans. Given that stay duration affects
the probability of returning negatively, many are not expected to return, especially if they
find good positions abroad. According to one participant:
Having gone through graduate programs both at METU and Northeastern, I can easily
say that METU had a much better program. Most of my grad coursework at
Northeastern was at the level of METU undergrad. I suspect this is the case for most
US universities. Given this fact, it's remarkable that METU forces (or at least forced in
1995) its assistants to get degrees in the US. It's no surprise that METU graduates get
the best jobs in the US.
15
A dummy variable for each discipline, in turn, was also used in the model to determine whether
certain fields of study are more prone to brain drain than other. The disciplines are “architecture”,
“economic and administrative sciences”, “engineering and technical sciences”, “education sciences”,
“language and literature”, “math and natural science”, “medicine”, “social sciences”, and “arts”.
None were found to be statistically significant from each other except for econ./admin. and
engin./tech. with education at the 5 per cent significance level.
195
Table 7.24 Marginal Effects of the Reasons for Going Abroad, Students
Probabilities:
Learn / improve language
skills
whygo_A=0
whygo_A=1
Difference 0
1
Job requirement in Turkey
whygo_C=0
whygo_C=1
Difference 0
1
Insufficient facilities for
research
whygo_F=0
whygo_F=1
Difference 0
1
Prestige and advantages of
international study
whygo_G=0
whygo_G=1
Difference 0
1
Lifestyle preference
whygo_H=0
whygo_H=1
Difference 0
1
To be with spouse / family
whygo_I=0
whygo_I=1
Difference 0
1
Get away from political
environment
whygo_K=0
whygo_K=1
Difference 0
1
Reason for choosing current
institution: job
opportunities
DC_E=0
DC_E=1
Difference 0
1
Reason for choosing current
institution: same location as
spouse
DC_F=0
DC_F=1
Difference 0
1
R_BS
R_IAS
R_NSAS
RP
RU
DNR
y=1
y=2
y=3
y=4
y=5
y=6
0.0005
0.0008
0.0003
0.0580
0.0742
0.0162
0.4302
0.4646
0.0344
0.3789
0.3535
-0.0254
0.1300
0.1054
-0.0246
0.0023
0.0015
-0.0008
0.0004
0.0010
0.0006
0.0503
0.0814
0.0311
0.4099
0.4771
0.0672
0.3913
0.3426
-0.0487
0.1452
0.0967
-0.0485
0.0028
0.0013
-0.0015
0.0007
0.0005
-0.0002
0.0662
0.0542
-0.0120
0.4487
0.4204
-0.0283
0.3660
0.3851
0.0191
0.1166
0.1372
0.0206
0.0018
0.0025
0.0007
0.0005
0.0006
0.0001
0.0549
0.0638
0.0089
0.4222
0.4437
0.0215
0.3840
0.3697
-0.0143
0.1359
0.1202
-0.0157
0.0025
0.0019
-0.0006
0.0007
0.0003
-0.0004
0.0684
0.0446
-0.0238
0.4533
0.3927
-0.0606
0.3625
0.4005
0.0380
0.1134
0.1585
0.0451
0.0017
0.0034
0.0017
0.0005
0.0038
0.0033
0.0562
0.1629
0.1067
0.4257
0.5495
0.1238
0.3818
0.2407
-0.1411
0.1333
0.0429
-0.0904
0.0024
0.0003
-0.0021
0.0009
0.0002
-0.0007
0.0767
0.0301
-0.0466
0.4691
0.3376
-0.1315
0.3497
0.4221
0.0724
0.1022
0.2043
0.1021
0.0014
0.0057
0.0043
0.0008
0.0003
-0.0005
0.0715
0.0399
-0.0316
0.4596
0.3767
-0.0829
0.3576
0.4080
0.0504
0.1089
0.1712
0.0623
0.0016
0.0040
0.0024
0.0007
0.0001
-0.0006
0.0676
0.0270
-0.0406
0.4518
0.3227
-0.1291
0.3637
0.4259
0.0622
0.1144
0.2177
0.1033
0.0018
0.0066
0.0048
196
The other reasons for pursuing foreign studies abroad that have a positive effect on
return intentions are when respondents go abroad in order to improve their language skills
or if they want to take advantage of the prestige and opportunities associated with overseas
studies. International diplomas are an important signal to employees in Turkey and those
with foreign degrees are more likely to get accepted or promoted. Foreign degrees,
therefore, increase the employability of individuals in Turkey, which is a factor that has a
positive effect on return intentions. Language skills are also given a premium by Turkish
employers.
When respondents go abroad to get away from the political environment, or due to
lifestyle preferences, or because they find the facilities and equipment in Turkey to do
research insufficient, they are very unlikely to return. The probability of not returning (y =
5 or 6) increases by 0.11 for those who left due to political reasons, by 0.05 for those who
left due to a lifestyle preference, and 0.02 for those who left due to insufficient facilities for
research. If students choose their current institution of study because of the job
opportunities they are given or to be in the same location as their spouse, the probability of
non-return increases by 0.06 and 0.11, respectively. Interestingly, the effect of family
considerations can have quite different effects on the intention of returning.
Effect of Difficulties Faced Abroad and Adjustment Factors:
Just as in the professionals case, the probability of definitely returning increases
when the psychic costs associated with being in a foreign country are high. When
employment prospects abroad are dim, the probability of returning immediately after
completing studies increases by 0.03 (Table 7.25). When respondents indicate that they had
to adjust to their environment (which is implied when they choose certain factors such as
previous experience as important in adjusting), the probability of returning also increases.
While Turkish friends at current institution of study may be important for easing
adjustment, those who indicated that this was an important adjustment factor for them are
more likely to be returning. This may also be an indication of strong ties to Turkish
community and to Turkey for some.
Effects of Compulsory Academic Service and Plans for Academic Career
As expected, students who finance their studies with national scholarships that have
a compulsory academic service requirement are more likely to be returning immediately
after completing their studies. The probability of returning immediately is 0.05 for those
197
without a compulsory academic service requirement, and 0.17 for those who have this
requirement. While the marginal effect between these two groups appears to be large
(0.12), what is worrisome is that the probability of returning immediately is not higher.
Non-returning students are an indication that the scholarships are not as successful as they
can be. Those who are planning an academic career are also more likely to have return
intentions. Despite the difficulties within the higher education system in Turkey,
universities provide greater opportunities for employment compared to other sectors,
especially in the recent economic crisis environment where many university graduates face
the prospect of being unemployed.
Table 7.25 Marginal Effects of Difficulties Abroad and Adjustment Factors, Students
R_BS
R_IAS R_NSAS
RP
RU
DNR
Probabilities:
Adjustment factor: previous
experience
adj_A=0
adj_A=1
Difference 0
1
Adjustment factor: Turkish
friends at institution
adj_F=0
adj_F=1
Difference 0
1
Difficulties faced while
abroad: unemployment
difabrdF=0
difabrdF=1
Difference 0
1
y=1
y=2
y=3
y=4
y=5
y=6
0.0005
0.0009
0.0004
0.0548
0.0772
0.0224
0.4221
0.4701
0.048
0.3841
0.3488
-0.0353
0.1360
0.1015
-0.0345
0.0025
0.0014
-0.0011
0.0005
0.0007
0.0002
0.0536
0.0688
0.0152
0.4188
0.4541
0.0353
0.3861
0.3619
-0.0242
0.1385
0.1128
-0.0257
0.0026
0.0017
-0.0009
0.0006
0.0013
0.0007
0.0606
0.0925
0.0319
0.4363
0.4936
0.0573
0.3749
0.3264
-0.0485
0.1256
0.0852
-0.0404
0.0021
0.0010
-0.0011
Table 7.26 Marginal Effects of Compulsory Academic Service and Plans for an Academic
Career, Students
R_BS
R_IAS R_NSAS
RP
RU
DNR
Probabilities:
Respondent plans to work
in academia
academic_b=0
academic_b=1
Difference 0
1
Respondent has compulsory
academic requirement
compulsory=0
compulsory=1
Difference 0
1
y=1
y=2
y=3
y=4
y=5
y=6
0.0003
0.0007
0.0004
0.0409
0.0694
0.0285
0.3802
0.4554
0.0752
0.4064
0.3609
-0.0455
0.1684
0.1119
-0.0565
0.0038
0.0017
-0.0021
0.0004
0.0039
0.0035
0.0481
0.1658
0.1177
0.4033
0.5505
0.1472
0.3950
0.2377
-0.1573
0.1503
0.0418
-0.1085
0.0030
0.0003
-0.0027
198
Effects of Various Push and Pull Factors:
Two push factors were important in determining return intentions for students:
being away from research centers / recent advances and finding the cultural or social life to
be less than satisfying in Turkey. The negative impact of finding the cultural and social life
in Turkey less satisfying is slightly less for those contemplating academic careers (0.07
compared to 0.10). The marginal impact of being away from research centers and recent
advances on the probability of not returning is 0.04.
The pull factors that significantly affect the return intentions of students are a
higher income level in the host country (pullA), a more ordered and organized life (pullF),
and spouse’s preference or job (pullI). The greatest negative impact on return intentions are
due to family considerations, followed by income levels and a more ordered lifestyle. The
marginal impact on each return intention category is given in Table 7.28. The importance of
salary levels for students contemplating an academic career is confirmed by the following
observation.
From talking with students who decide to stay here rather than go back to Turkey, the
primary reason is financial. Very able PhD graduates who can become excellent
faculty in Turkey, most of the time decide on even a mediocre job here (which will not
satisfy them in the long run) rather than become a faculty member in Turkey with the
current salaries. If Turkey does not improve the living standards of University faculty
... the price paid will be incalculable. Here in US the best go into academia, there it
looks like it is the people who either have money or could not find anything else (most
of the time). The first thing the country should do is to invest in [the] education of the
new generation.
Effect of Last Impressions:
For professionals, the last impression from the latest trip to Turkey has an important
impact on return intentions. The same is true for students. The last visit to Turkey changes
an individual’s perceptions about conditions in Turkey. The probability of returning (y = 1
or 2) decreases by about 0.04 for those who were negatively effected by their last trip to
Turkey, and increases by 0.05 for those who were left with more positive impressions. The
effect of the September 11, 2001 terrorist attacks in New York is given by sept11_inc. The
effect of Sept. 11 is to increase return intentions. The probability of returning (y = 1 or 2)
increases by 0.04 which is less than that of professionals (0.07).
199
Table 7.27 Marginal Effects of Various Push Factors, Students
R_BS
R_IAS R_NSAS
Probabilities:
Push factor: being away
from research centers and
recent advances
pushE=0
pushE=1
Difference 0
1
RP
RU
DNR
y=1
y=2
y=3
y=4
y=5
y=6
0.0009
0.0005
-0.0004
0.0765
0.0528
-0.0237
0.4688
0.4169
-0.0519
0.3499
0.3873
0.0374
0.1024
0.1399
0.0375
0.0014
0.0026
0.0012
0.0004
0.0001
-0.0003
0.0473
0.0222
-0.0251
0.4011
0.2968
-0.1043
0.3962
0.4306
0.0344
0.1519
0.2421
0.0902
0.0031
0.0083
0.0052
0.0010
0.0003
-0.0007
0.0803
0.0408
-0.0395
0.4753
0.3798
-0.0955
0.3442
0.4066
0.0624
0.0979
0.1687
0.0708
0.0013
0.0038
0.0025
Push factor: less than
satisfying cultural / social
life in Turkey
non-academic (academic_b=0)
pushG=0
pushG=1
Difference 0
1
academic (academic_b=1)
pushG=0
pushG=1
Difference 0
1
Table 7.28 Marginal Effects of Various Pull Factors, Students
R_BS
R_IAS R_NSAS
Probabilities:
Pull factor: higher level of
income in host country
pullA=0
pullA=1
Difference 0
1
Pull factor: more organized,
ordered environment
pullF=0
pullF=1
Difference 0
1
Pull factor: spouse’s
preference or job
pullI=0
pullI=1
Difference 0
1
RP
RU
DNR
y=1
y=2
y=3
y=4
y=5
y=6
0.0012
0.0005
-0.0007
0.0920
0.0542
-0.0378
0.4929
0.4206
-0.0723
0.3272
0.385
0.0578
0.0857
0.1371
0.0514
0.0010
0.0025
0.0015
0.0011
0.0005
-0.0006
0.0855
0.0557
-0.0298
0.4835
0.4243
-0.0592
0.3366
0.3827
0.0461
0.0922
0.1344
0.0422
0.0012
0.0024
0.0012
0.0008
0.0002
-0.0006
0.0718
0.0340
-0.0378
0.4601
0.3544
-0.1057
0.3572
0.4167
0.0595
0.1085
0.1898
0.0813
0.0016
0.0049
0.0033
200
Table 7.29 Marginal Effects of the Last Visit to Turkey and Sept. 11, Students
R_BS
R_IAS R_NSAS
RP
RU
DNR
Probabilities:
Last visit to Turkey
decreased return intentions
lastvis1=0
lastvis1=1
Difference 0
1
Last visit to Turkey
increased return intentions
lastvis3=0
lastvis3=1
Difference 0
1
Effect of Sept. 11: increased
return intentions
sept11_inc=0
sept11_inc=1
Difference 0
1
y=1
y=2
y=3
y=4
y=5
y=6
0.0009
0.0003
-0.0006
0.0764
0.0377
-0.0387
0.4687
0.3687
-0.1000
0.3501
0.4113
0.0612
0.1025
0.1778
0.0753
0.0014
0.0043
0.0029
0.0005
0.0017
0.0012
0.0579
0.1100
0.0521
0.4300
0.5143
0.0843
0.3791
0.3025
-0.0766
0.1302
0.0707
-0.0595
0.0023
0.0007
-0.0016
0.0005
0.0014
0.0009
0.0573
0.0973
0.0400
0.4284
0.4999
0.0715
0.3801
0.3197
-0.0604
0.1314
0.0808
-0.0506
0.0023
0.0009
-0.0014
7.6 Concluding Remarks
The impact of various factors on the “probability of not returning” and on the
“probability of returning” are presented in order of importance in Tables 7.30-7.33. In both
the students and professionals groups, the greatest positive impact on the probability of not
returning occurs when the initial return intention is to stay compared to those who initially
intended to return. Family considerations such as marriage to a foreign spouse and family
support for settling abroad are also influential in non-return.
Stay duration, work experience in the host country and specialized training are all
found to have significant negative impacts on the return intentions of Turkish professionals.
In addition, work experience in Turkey after obtaining a PhD abroad increases the
likelihood of not returning. Among the push and pull factors, economic instability has the
greatest deterrent effect on return. Female participants and those in academe are also less
likely to be returning in the professionals group.
The results for Turkish students studying abroad suggest that family considerations,
lifestyle factors, higher salaries and the political environment are prominent in non-return
intentions. On the other hand, the compulsory academic service requirement has a positive
201
effect on return intentions, although many of those who intend to return are not planning to
return immediately after completing their studies.
Table 7.30 Factors that have the Greatest Negative Impact on Return Intentions, Professionals
Marginal
Effect on
Prob(y =4 or 5)
Variable
Initial return intention is to stay versus return
Initial return intention is to stay versus unsure
Highest Degree: Arch/Econ/Admin versus Educ/Lang/Soc/Art
First full time job after getting foreign degree is in Turkey
Married to foreign spouse
Received organization-specific formal training
Initial return intention is unsure versus return
Push Factor: economic instability
Pull Factor: better educational opportunities for children
Gender: female
Pull Factor: proximity to research centers
Pull Factor: more satisfying social / cultural life
Highest Degree: Engineer/Math/Science/Medicine versus Educ/Lang/Soc/Art
Pul Factor: greater opportunity to develop specialty
Academic and research related occupation
No work experience in Turkey
Reason for going: lifestyle preference
Pull Factor: more organized, ordered environment
Last visit decreased return intention
Family support for settling abroad (1 point increase)
Reason for going: get away from political environment
Standard of living assessment of life abroad (1 point increase)
Push Factor: limited job opportunity in specialty
Social assessment of life abroad (1 point increase)
Reason for going: insufficient facilities, equipment for research
Push factor: no opportunity for advanced training
202
0.4911
0.3621
0.2008
0.1824
0.1528
0.1400
0.1290
0.1245
0.1163
0.1086
0.1080
0.1020
0.0946
0.0926
0.0860
0.0782
0.0654
0.0580
0.0565
0.0559
0.0528
0.0469
0.0421
0.0366
0.0297
0.0253
Table 7.31 Factors that have the Greatest Positive Impact on Return Intentions, Professionals
Marginal
Effect on
Prob(y =1 or 2)
Variable
Last visit increased return intention
Initial intention is to return versus to stay
Need to finish / continue with current project
Other pull reason (male=1) (e.g. military service requirement)
Respondent has a PhD from a Turkish university
Initial intention is unsure versus to stay
Initial intention is to return versus unsure
Reason for going: to be with spouse, family
Active in contributions to Turkey during stay abroad
September 11 increased return intentions
Adjustment factor: Turkish Student Association
Adjustment factor: previous experience
Highest degree field is in Engineering/Math/Science/Medicine versus
Arch/Economics/Admin
Difficulty faced while abroad: missing family
Received family support for initial overseas venture
Lack of financial resources for business
Reason for going: job requirement in Turkey
Reason for going: prestige and advantages of study abroad
203
0.2219
0.2184
0.1817
0.1434
0.1413
0.1186
0.0998
0.0959
0.0920
0.0708
0.0676
0.0670
0.0540
0.0494
0.0432
0.0333
0.0235
0.0124
Table 7.32 Factors that have the Greatest Negative Impact on Return Intentions, Students
Marginal
Effect on
Prob(y =5 or 6)
Variable
Initial return intention is to stay versus return
Initial return intention is to stay versus unsure
Reason for choosing current institution of study: job opportunities
Reason for going: get away from political environment
Push factor: less than satisfying cultural or social life in Turkey (non-academic)
Initial return intention is unsure versus return
Pull factor: spouse's preference or job
Family support for settlement abroad: definitely versus not supportive
Last visit decreased return intention
Reason for choosing current institution of study: same location as spouse
Pull factor: higher salaries in host country
Reason for going: lifestyle preference
Family support for settlement abroad: definitely versus somewhat supportive
Pull Factor: more organized, ordered environment
Family support for settlement abroad: somewhat versus not supportive
Push factor: being away from research centers and recent advances
Standard of living assessment: better or much better
Gender: female
Reason for going: insufficient facilities, equipment for research
0.3787
0.3223
0.1081
0.1064
0.0954
0.0952
0.0846
0.0785
0.0782
0.0647
0.0529
0.0468
0.0457
0.0434
0.0394
0.0387
0.0342
0.0259
0.0213
Table 7.33 Factors that Have the Greatest Positive Impact on Return Intentions, Students
Marginal
Effect on
Variable
Prob(y =1,2,3)
Initial intention is to return versus to stay
Initial intention is unsure versus to stay
Compulsory academic service
Reason for going: to be with spouse
Initial intention is to return versus unsure
Last visit increased return intention
Social assessment of life abroad: worse or much worse
September 11 increased return intentions
Respondent has plans for an academic career
Reason for going: job requirement in Turkey
Difficulties abroad: unemployment
Adjustment factor: previous experience
Turkish Student Association member
Reason for going: learn / improve language skills
Adjustment factor: Turkish friends at institution
Reason for going: prestige and advantages of international study
204
0.4921
0.2972
0.2684
0.2338
0.1949
0.1376
0.1344
0.1124
0.1041
0.0989
0.0899
0.0708
0.0664
0.0509
0.0507
0.0305
CHAPTER 8
CONCLUSIONS
This study deals with skilled migration from a developing country perspective. The
first part of the study brings up to date both the theoretical and the policy debate on the
impact of skilled migration on the sending economies. In economic models of migration,
skilled labor mobility is treated in a similar way as physical capital movements. In an open
economy setting, under perfect capital mobility, capital will flow to where it will earn a
higher rate of return. Similarly, skilled migration is believed to be the result of differences
in the rates of return awarded to skills or educational attainment levels in different
countries, as measured by the wage rate. According to neoclassical theory, higher wages
signal excess demand for skilled workers and skilled workers respond by relocating to
where they will earn a higher income. Migration changes the relative quantities of skilled
workers in both the sending and receiving countries and as a result alters their rate of return
so that wage differentials disappear in the long run. This will then eliminate migratory
movements motivated purely for economic reasons. Wage differentials, however, are not
disappearing as predicted by the neoclassical theory of migration. Instead, they appear to be
quite persistent in spite of the large volume of skilled migration from the developing
countries.
The theoretical brain drain models considered in the thesis offer different
perspectives on the reasons for the wage differential between sending and receiving
countries. They all adopt the view that wages are determined by the marginal productivity
of individuals and that wage differentials provide the main motivation for migration. The
Kwok-Leland model of asymmetric information provides an alternative theory of the wage
differentials existing between host and source countries. The argument is that host firms
have an advantage over home firms in terms of their knowledge about the true
productivities of students completing their studies in the host country that enables them to
205
give the appropriate level of income to each student. Home firms, on the other hand, can
offer only the average wage level of the returning students. Another explanation for the
persistence of income differentials is the higher social capital stock (physical and human)
existing in developed countries. Since educated workers complement both physical and
human capital in production, they are more productive and earn higher wages in locations
where physical and human capital are relatively more abundant. Miyagiwa’s model of the
brain drain is based on such “agglomeration economies” in which there is increasing returns
to the accumulation of human capital. Wong’s model of learning-by-doing offers a slightly
different explanation for the wage differential that is based on the greater cumulative base
of “experience” in the host country, which leads to higher productivity levels for those
working and therefore taking part in the production process in the host country. An increase
in work experience through learning on the job in the host country increases the
productivity and salaries of individuals. Chen and Su, on the other hand, propose a slightly
different potential explanation for student non-return based on-the-job training. In their
model, the education received in the host country is complementary to the social capital
stock of the host country. Education received abroad thus increases the productivity of
individuals much more in the host country than in the home country.
The second part of the thesis provided an evaluation of the findings of the survey on
the return intentions of Turkish students and Turkish professionals. The majority of Turkish
students responding to our survey are single, male, studying in the engineering and
technical fields, holding a degree from a university in Turkey with English instruction, and
having parents who are highly educated. The most cited reason for studying abroad is the
perception that a better quality education will be received at the foreign institution of study,
based on the institution’s reputation, ranking of the program or the presence of an academic
thesis supervisor in the case of master’s or doctorate level students. Professionals are, on
average, slightly older than the student respondents and have a longer length of stay. A
significant proportion of them are married and the proportion of female respondents is
lower. A much greater majority have earned degrees in the engineering and technical
sciences.
The most important reason for not returning or delaying return appears to be the
uncertainty created by the February 2001 economic crisis, which has also hit the educated
segment of the population. Many university-educated individuals fear that they will not be
206
able to find employment upon their return to Turkey and therefore choose to stay abroad for
a period of time to acquire work experience. More than three-quarters of respondents in
both surveys cited economic instability and uncertainty as a “very important” push factor.
Thus, the economic crisis combined with existing problems of unemployment or
underemployment in certain fields appears to be a prominent factor in delaying return. The
crisis has also prompted many students to seek either jobs or study opportunities abroad.
The increasing demand for these types of graduates in the United States has made the US a
popular destination for recent graduates, although the job market has tightened in the US.
For professionals lower income levels, which is among the most often cited reasons for
brain drain from developing to developed countries, appears to be less important than other
“push” factors such as bureaucratic obstacles. For students, higher income in the host
countries does not appear to exert as great a “pull” as opportunities for advancement in the
chosen occupation or for further development and training in specialization in terms of the
number of respondents marking this factor as important. This emphasis may be due to the
higher number of doctoral level students answering the survey.
The models estimated in the ordered probit analysis are based on the human capital
theory of migration, which predicts that individuals will migrate when the net present value
of benefits from migration is positive. Wage differentials between the host and source
countries provide the main motivation for moving to a foreign country. According to
human capital theory, the difference in the expected foreign and domestic income levels is
the key determinant of skilled migration. Since expected income is the relevant variable,
employment opportunities and labor market conditions both at home and abroad play an
important role in the perceptions of economic opportunity held by skilled individuals.
While the income differential is an important consideration (marked as “very important” or
“important”) for a majority of respondents in both survey groups, higher salary levels in the
host country are found to be statistically significant in determining the return intentions of
only Turkish students studying abroad.
Family considerations, not surprisingly, have considerable weight in the mobility
decisions of the survey participants. In some cases, remaining abroad is not simply a matter
of earning a higher salary or enjoying better work conditions. Marriage to a foreign spouse
is obviously an important reason for not returning. For others, however, concern over
childrens’ adaptation to the highly competitive education system in Turkey may also
207
dominate the return decision. In both the student and professionals survey groups, family
support for the decision to settle abroad is found to be an important factor determining
return intentions.
Female respondents appear less inclined to be returning to Turkey than male
respondents. In general, the parental education levels of female participants are greater than
that of males indicating that they come from a higher socio-economic background. This
may be indicative of a more selective migration process working in the case of females.
Some female participants have indicated that they enjoy greater freedom in lifestyle choice
abroad than they do in Turkey, which may also be an important factor in the non-return
decision.
Information on past mobility patterns of the respondents reveal important
information about the dynamics of return intentions: For example, respondents with no
previous work experience in Turkey were more likely to indicate return intentions than
those who had some work experience. This suggests that dissatisfaction with the work
environment in Turkey and the ability to compare workplaces and work situations
decreases the likelihood of return. The length of stay abroad is also another important
determinant of return intentions. As expected, return intentions weaken with the length of
stay. Initial return intentions (the intentions at the beginning of the stay abroad) are
positively associated with current return intentions. However, this association is weaker
when the initial intention is to return and the length of stay increases.
In general, respondents appear to be satisfied with economic conditions in their host
countries but indicate that they find social life “lacking”. In spite of this dissatisfaction
with social life, nearly a quarter of all respondents are not considering returning to Turkey.
One third of those who are considering returning to Turkey are planning to do so within 2
to 5 years, and another third are planning to do so within 5 to 10 years. There is a high
probability that delaying return could in time come to mean “no return”. Taking this fact
into consideration, one could surmise that the number of students who will never return to
Turkey could reach significant proportions.
Respondents’ comments have also been important in understanding the various
motivations in the decision to return to Turkey or stay abroad. Compulsory military service
has been given both as a “push factor” in the decision to go abroad and as a reason for non-
208
return. A considerable number of male respondents have indicated “delaying compulsory
military service” as a reason for pursuing an overseas degree. Those who have not
completed their military service regard long-term military service as an “interruption”
causing a “time loss” in education and career. As a result, many go abroad or delay
returning in order to fulfill the requirements of short-term military service. For some this
constitutes the first step toward settling in a foreign country, since it means that they are
starting their professional careers abroad and adapting to life and work conditions in their
country of work. As well, some of those who have entered into working life abroad delay
returning to Turkey because they fear the uncertainty of finding employment. Many
respondents have cited the unfavorable conditions created by the February 2001 economic
crisis as an example.
Some of those who have settled abroad, or who plan to, say that they will continue
with their lives abroad without cutting their ties to Turkey and act as a sort of “cultural
bridge” between their native country and their country of destination. This indicates that
although the return potential for these individuals may not be very high, their value as both
cultural diplomats and mediums for information and technology transfer between Turkey
and their resident countries should make them an important target group for Turkish
policymakers. Turkish academic advisors abroad, for example, help ease the transition to a
foreign university for many students.
In Turkey, the academic brain drain appears to be particularly troubling, since the
number of universities in Turkey has grown rapidly over the last decade in response to the
growing social demand for higher education created by demographic pressures. This has
created the problem of staffing the newly formed universities. While the compulsory
academic service requirement of government-sponsored overseas scholarships was planned
as a way to meet part of this need, non-returning scholarship recipients have become a
major concern. One of the most common views expressed in the survey by governmentsponsored research assistants is the perceived lack of value given to science and to
academics in Turkey. Some respondents have indicated that, as a result of this, they fear
they will find themselves in an “unproductive environment” if they return to Turkey. Others
have stated that “there is a point where money is no object” and that they would be willing
to work for lower wages in Turkey provided that they are “valued and respected”.
209
Have the state investments in higher education, through the national scholarship
program, gone to waste? The number of returning students is not the best measure to assess
this. Even if all of the government-sponsored students were to return, there is indication
that the advanced overseas training they received will not be put to efficient use, especially
in the newly-established state universities that lack facilities, equipment and other
important resources. Several government-sponsored research assistants have expressed the
fear that they will be devoting most of their time in teaching activities at the undergraduate
level with little opportunity to do research and develop their knowledge. The current needs
of the expanding higher education system seem to be favoring a teaching role for the
returning government-sponsored students, and this has led to some disillusionment and lack
of motivation among the scholarship recipients. The Higher Education Council has also
begun to question the value of sending so many students for overseas studies. As a result,
the number of YÖK scholarship recipients has been reduced, and greater emphasis is
currently placed on producing new academicians internally through the graduate programs
of the established universities in Turkey. However, this requires that a greater amount of
resources be devoted to the development of graduate programs. In turn, a greater amount of
public investment in higher education is required if undergraduate programs are not to be
compromised by a shift of teaching staff to graduate level studies.
State universities in Turkey, like many in the developing world, are unable to
compete with the scale of research funding, provision of resources and incomes offered by
universities in leader countries. Public universities are in danger of losing their best
researchers and teaching staff to the private universities in Turkey and to universities
abroad. The recent economic crises, however, have led to serious cutbacks in university
funding that were already inadequate before the crises. It is also well known that university
salaries especially at the state universities are inadequate and lead to moonlighting and
extra teaching activities.
Newly established departments are small in terms of the number of full-time staff
they employ and often have to resort to using research assistants as lecturers in order to
make up for shortages in teaching staff. Cutting edge, innovative research cannot be
expected from these universities until they mature as institutions, and without a research
agenda devoted to specific research problems that is complementary to the needs of
indigenous industry and local conditions. While the return of overseas academicians may
210
create the right atmosphere for positive changes toward institution building and the
development of a national research agenda, there may be valid concern that returning
scholars are simply “importing the host country’s agenda” (Kreimer, 2003).
The recent brain drain from Turkey should not be looked at solely in terms of an
employment problem created by the conditions of the economic crises and ensuing
uncertainties. Turkey must take seriously the need to develop and expand research and
development activities and create opportunities for the transfer of skills and training for
which so much investment has been undertaken. What is promising is that a great number
of survey respondents have indicated their willingness to return even if some progress is
made toward creating the right environment for research and better career development
opportunities.
Further Research:
The current survey research on Turkey’s brain drain involved collecting information
from an Internet-based survey on the return intentions of individuals. This information was
then used to examine various characteristics of the respondents and to determine the
importance of different factors in the decision to return to Turkey or stay in the current
country of residence. The study combined a mixture of inductive and deductive methods in
the analysis of the determinants of return intentions.
One of the limitations of the survey study is that it deals with return intentions rather
than actual behavior. The return intentions of individuals who were studying or working
abroad at the time of the survey may not be realized, no matter how certain respondents
may have been in their plans about returning or staying. Returning also does not guarantee
a permanent settlement in Turkey, since new opportunities and new circumstances can arise
at any time and radically alter previous plans. Many of the theoretical contributions to the
migration literature treat the migration decision as a single, once-and-for-all decision. The
new literature on the brain drain, on the other hand, emphasizes the positive aspects of
migration for developing countries, including return migration and brain circulation. The
dynamics of migration in developing countries suggest however that many who return to
their home countries have difficulties re-adapting and as a result may decide to settle
abroad permanently if they can find the opportunity. This pattern is also found in the
current survey where work experience in Turkey after studying abroad is found to be an
211
important factor contributing to non-return. The reason is that work experience in Turkey
allows individuals to compare work environments and conditions in Turkey and very often
these comparisons have a negative effect on return intentions.
Answers to many questions about mobility can be found through micro level
studies. The present study can be extended by following up on some of the participants and
seeing whether their return intentions have turned to reality and for what reasons. It is also
useful to examine mobility patterns within specific occupations or specialties in order to
obtain a better understanding of the concerns within specific occupation groups. The
database obtained from the survey study can be integrated into a long term study for
studying the career paths and mobility patterns of highly educated individuals from Turkey.
In addition to the questionnaire responses, information on educational and career mobility
may be supplemented from various sources some of which may be available directly from
the Internet, such as curriculum vita data.
212
REFERENCES
Adams, W. (1968) The Brain Drain, New York: The Macmillan Company.
Aghion, Phillipe and Peter Howitt (1998) Endogenous Growth Theory, Cambridge, MA:
The MIT Press.
Agresti, A. (1984) Analysis of Ordinal Categorical Data, New York: John Wiley & Sons.
Akerlof, G. A. (1970) “The Market for ‘Lemons’: Quality Uncertainty and the Market
Mechanism,” Quarterly Journal of Economics, 84, 488-500.
Alesina, A. and D. Rodrik (1994) “Distributive Politics and Economic Growth,” Quarterly
Journal of Economics, 109(2), 465-90.
Arrow, K. J. (1962) “The Economic Implications of Learning by Doing,” Review of
Economic Studies, 29, 155-173.
Baldwin, G. B. (1970) “Brain Drain or Overflow?” Foreign Affairs, 48(2), 358-72.
Banerjee, A. V. (1992) “A Simple Model of Herd Behavior,” Quarterly Journal of
Economics, 107(3), 797-817.
Barro, R. J. and J.-W. Lee (2000) “International Data on Educational Attainment: Updates
and Implications,” Harvard University CID Working Paper No. 42, Data Appendix,
http://www.cid.harvard.edu/ciddata/ciddata.html.
Bauer, T., G. S. Epstein and I. N. Gang (2000) “What are Migration Networks?” IZA
Discussion Paper No. 200, Bonn, Germany: Institute for the Study of Labor.
Bauer, T., G. S. Epstein and I. N. Gang (2002a) “Herd Effects or Migration Networks? The
Location Choice of Mexican Immigrants in the US,” IZA Discussion Paper No. 551,
Bonn, Germany: Institute for the Study of Labor.
Bauer, Thomas, Gil S. Epstein and Ira N. Gang (2002b) “Enclaves, Language and the
Location Choice of Migrants,” IZA Discussion Paper No. 558, Bonn, Germany:
Institute for the Study of Labor.
Becker, G. S. (1993) Human Capital: A Theoretical and Empirical Analysis with Special
Reference to Education, Third Edition, Chicago and London: The University of
Chicago Press.
213
Becker, Gary S., Kevin M. Murphy and Robert Tamura (1990) “Human Capital, Fertility,
and Economic Growth,” Journal of Political Economy, 98(5), S12-S37.
Becker, S. O., A. Ichino, and G. Peri (2003) “How Large is the ‘Brain Drain’ from Italy,”
CESifo Working Paper No. 839, http://papers.ssrn.com/abstract_id=378522
Beine, M., F. Docquier and H. Rapoport (2001) “Brain Drain and Economic Growth:
Theory and Evidence,” Journal of Development Economics, 64, 275-289.
Berry, R. A. and R. Soligo (1969) “Some Welfare Aspects of International Migration,”
Journal of Political Economy, 77, 778-94.
Bhagwati, Jagdish and Koichi Hamada (1974) “The Brain Drain, International Integration
of Markets for Professionals and Unemployment: A Theoretical Analysis,” Journal
of Development Economics, 1, 19-42.
Bhagwati, Jagdish and Carlos Rodriguez (1975) “Welfare-Theoretical Analysis of the Brain
Drain,” Journal of Development Economics, 2, 195-221.
Blau, Francine D., Lawrence M. Kahn, Joan Y. Moriarty and Andre Portela Souza (2002)
“The Role of the Family in Immigrants’ Labor-Market Activity: Evidence from the
United States,” NBER Working Paper No. 9051, Cambridge, MA: National Bureau
of Economic Research (http://www.nber.org/papers/w9051).
Blomqvist, Ake G. (1986) “International Migration of Educated Manpower and Social
Rates of Return to LDCs,” International Economic Review, 27(1), 165-174.
Carrington, William J. and Enrica Detragiache (1998) “How Big is the Brain Drain?,” IMF
Working Paper WP/98/102, New York: International Monetary Fund.
Carrington, William J. and Enrica Detragiache (1999) “How Extensive is the Brain
Drain?”, Finance and Development, 46-49. (Reprinted as Carrington, W. J. and E.
Detragiache (1999) “International Migration and the ‘Brain Drain’”, The Journal of
Social, Political and Economic Studies, 24(2), 163-71.
Cervantes, M. and D. Guellec (2002) “The Brain Drain: Old Myths, New Realities,” OECD
Observer, published on the internet on May 7, 2002.
Chang, P. and Z. Deng (1992) “The Chinese Brain Drain and Policy Options,” Studies in
Comparative International Development, 27(1), 44-60.
Chen, Been-Lon (2003) “An Inverted-U Relationship Between Inequality and Long-Run
Growth,” Economics Letters, 78, 205-212.
Chen, T.-J. and H.-Y. Su (1995) “On-the-Job Training as a Cause of Brain Drain,”
Weltwirtschaftliches Archiv, 131(3), 526-41.
214
Cheng, Lucie and Philip Q. Yang (1998) “Global Interaction, Global Inequality, and
Migration of the Highly Trained to the United States,” International Migration
Review, 32(3), 626-653.
Choi, Hyaeweol (1999) “Shifting Human Resources in South Korean Science and
Technology,” Comparative Education Review, 43(2), 212-32.
Commander, S., M. Kangasniemi and L. A. Winters (2002) “The Brain Drain: Curse or
Boon? A Survey of the Literature,” paper prepared for the CEPR/NBER/SNS
International Seminar on International Trade, Stockholm, 24-25 May, 2002.
Coppel, Jonathan, Jean-Christophe Dumont and Ignazio Visco (2001) “Trends in
Immigration and Economic Consequences,” Economics Department Working Paper
No. 284, Paris: Organisation for Economic Co-operation and Development.
Cortés, C. (ed.) (1980) The Latin American Brain Drain to the United States, New York:
Arno Press.
Cumhuriyet Gazetesi (2000) “Beyin Göçü: Ara tırma Görevlisi Azalıyor,” (Brain Drain:
The Number of Research Assistants is Declining), 14.01.2000.
Cumhuriyet Gazetesi (2001) “Yüksek Sefalet,” (Higher Poverty), 16.07.2001.
Davis, Donald R. And David E. Weinstein (2002) “Technological Superiority and the
Losses from Migration,” NBER Working Paper No. W8971, Cambridge, MA:
National Bureau of Economic Research, http://www.nber.org/papers/w8971.pdf.
Davies, James B. And Ian Wooton (1992) “Income Inequality and International Migration,”
The Economic Journal, 102, 789-802.
Dayıo lu, M. and Z. Kasnako lu (1997) “Kentsel kesimde kadın ve erkeklerin i gücüne
katılımları ve kazanç farklılıkları”, (Labour force participation of men and women
in urban Turkey and earnings inequality between genders), METU Studies in
Development, Vol. 24, No. 3, pp. 329-361.
Desai, Mihir A., Devesh Kapur and John McHale (2001) “The Fiscal Impact of the Brain
Drain: Indian Emigration to the U.S.,” paper prepared for the Third Annual NBERNCAER Conference, 17-18 December.
Drinkwater, Stephen, Paul Levine and Emanuela Lotti (2002) “The Economic Impact of
Migration: A Survey,” paper prepared for the 2nd Workshop of the Fifth Framework
Programme Project ‘European Enlargement: The Impact of East-Wast Migration on
Growth and Employment,’ Vienna, 6-7 December.
Dos Santos, Manon Domingues and Fabien Postel-Vinay (2003) “Migration as a Source of
Growth: The Perspective of a Developing Country,” Journal of Population
Economics, 16, 161-175.
Dündar, H. and D. R. Lewis (1999) “Equity, quality and efficiency effects of reform in
Turkish higher education,” Higher Education Policy, 12, 343-66.
215
Eicher, Theo and Cecilia Garcia-Penalosa (2001) “Inequality and Growth: The Dual Role
of Human Capital in Development,” Journal of Development Economics, 66, 173-97.
Eldridge, Robert Huyck (1966) “Emigration and the Turkish Balance of Payments,” Middle
East Journal, 20, 296-316.
Ellerman, D. (2003) “Policy Research on Migration and Development,” World Bank Policy
Research Working Paper No. 3117, Washington, D.C.: The World Bank, August
2003.
Epstein, G. S. (2002) “Informational Cascades and Decision to Migrate,” IZA Discussion
Paper No. 445 Bonn, Germany: Institute for the Study of Labor.
Ekonomist, “En Büyük Arz-Talep Açı ı Hem irelikte” (The Supply-Demand Gap is
Greatest in Nursing), 30 Temmuz 2000.
Eliasson, G. (2000) “Industrial Policy, Competence Blocs and the Role of Science in
Economic Development,” Journal of Evolutionary Economics, 10, 217-241.
Ermisch, J. and M. Francesconi (2001) “Family Matters: Impacts of Family Background on
Educational Attainments,” Economica, 68, 137-156.
Faist, Thomas (2000) The Volume and Dynamics of International Migration and
Transnational Social Spaces, Oxford and New York: Oxford University Press.
Fan, C. S. and O. Stark (2002) “Addition through Depletion: The Brain Drain as a Catalyst
of Human Capital Formation and Economic Betterment,” Royal Economic Society
Annual Conference 2003, 192, Royal Economic Society.
Findlay, Allan M. And Emma Stewart (2002) “Skilled Labour Migration from Developing
Countries: Annotated Bibliography,” International Migration Papers No. 55,
International Migration Programme, ILO.
Galor, Oded and Joseph Zeira (1993) “Income Distribution and Macroeconomics,” Review
of Economic Studies, 60, 35-52.
Galor, Oded and Daniel Tsiddon (1996) “Income Distribution and Growth: the Kuznets
Hypothesis Revisited,” Economica, 63, S103-S117.
Galor, O. and D. Tsiddon (1997) “Technological Progress, Mobility, and Economic
Growth,” The American Economic Review, 87(3), 363-82.
Godfrey, E. M. (1970) “The Brain Drain from Low-Income Countries,” The Journal of
Development Studies, 6(3), 235-47.
Goss, E. P. and C. Paul (1986) “Age and Work Experience in the Decision to Migrate,” The
Journal of Human Resources, 21(3), 397-405.
216
Görün, F. (1996) “Higher Education and Unemployment in Turkey,” in Tuncer Bulutay
(ed.) Education and the Labour Market in Turkey: Proceedings of a Seminar Held in
Ankara, 19 April 1996, Ankara: State Institute of Statistics.
Greene, William H. (1997) Econometric Analysis, 3rd Edition, London, UK: Prentice-Hall
International.
Greenwood, Michael J. (1983) “The Economics of Mass Migration from Poor to Rich
Countries,” The American Economic Review, Papers and Proceedings, 73(2), 173-7.
Greenwood, Michael J. and John M. McDowell (1991) “Differential Economic
Opportunity, Transferability of Skills, and Immigration to the United States and
Canada,” The Review of Economics and Statistics, 73(4), 612-23.
Grubel, H. B. and A. D. Scott (1966a) “The International Flow of Human Capital,”
American Economic Review, Papers and Proceedings, 56(2), 268-274.
Grubel, Herbert B. and Anthony D. Scott (1966b) “The Immigration of Scientists and
Engineers to the United States, 1949-61,” Journal of Political Economy, 74, 368-378.
Haque, N. U. and S. Kim (1994) “Human Capital Flight: Impact of Migration on Income
and Growth,” IMF Working Paper WP/94/155, New York: International Monetary
Fund.
Haque, N. U. and M. Ali Khan (1997) “Institutional Development: Skill Transference
Through a Reversal of ‘Human Capital Flight’ or Technical Assistance, IMF
Working Paper WP/97/89, New York: International Monetary Fund.
Hamada, K. (1975) “Efficiency, Equality, Income Taxation and the Brain Drain,” Journal
of Development Economics, 2, 281-7.
Hamada, Koichi (1977) “Taxing the Brain Drain: A Global Point of View,” in Jagdish N.
Bhagwati (ed.) The New International Economic Order: The North-South Debate,
Cambridge, MA: The MIT Press, 125-55.
Hamada, K. and J. Bhagwati (1975) “Domestic Distortions, Imperfect Information
and the Brain Drain,” Journal of Development Economics, 2, 265-79.
Harris, John R. And Michael P. Todaro (1970) “Migration, Unemployment and
Development: A Two-Sector Analysis,” American Economic Review, 60, 126-42.
Helmenstein, C. and Y. Yegorov (2000) “The Dynamics of Migration in the Presence of
Chains,” Journal of Dynamics and Control, 24, 307-323.
Hekmati, M. (1973) “Non-Returning Foreign Students: Why Do They Not Return Home?,”
Die Dritte Welt, 2(1), 25-43.
Hoch, Paul K. (1987) “Migration and the Generation of New Scientific Ideas,” Minerva: A
Review of Science, Learning and Policy, 25(3), 209-237.
217
Huang, W.-C. (1988) “An Empirical Analysis of Foreign Student Brain Drain to the United
States,” Economics of Education Review, 7(2), 231-243.
IIE (various years) Open Doors: Report on International Educational Exchange, New
York, NY: Institute of International Education. (http://www.opendoors.org)
IIE (2001) Open Doors 2001: Report on International Educational Exchange, New York,
NY: Institute of International Education.
Iredale, Robyn (1999) “The Need to Import Skilled Personnel: Factors Favouring and
Hindering its International Mobility,” International Migration, 37(1), 89-120.
I ı ıçok, Ö. (2002) “Türkiye’de ya anan son ekonomik krizlerin sosyo-ekonomik sonuçları:
Kriz i sizli i ve beyin göçü”, (The socioeconomic results of the last two economic
crises
in
Turkey:
Crisis
unemployment
and
brain
drain),
www.isguc.org/calhay_beyingocu.php.
Jaffe, A. B. (1989) “Real Effects of Academic Research,” American Economic Review, 79,
957-70.
Johnson, H. G. (1965) “The Economics of the ‘Brain Drain’: The Canadian Case,”
Minerva, Spring, 299-311.
Johnson, H. G. (1967) “Some Economic Aspects of the Brain Drain,” The Pakistan
Development Review, 7(3), 379-441.
Johnson, H. G. (1968) “An ‘Internationalist’ Model,” in W. Adams (ed.) The Brain Drain,
New York: The Macmillan Company.
Johnson, Harry G. (1972) “Labour Mobility and the Brain Drain”, in Gustav Ranis (ed.)
The Gap between Rich and Poor Nations, New York: St. Martin’s Press, 380-91.
Jolly, Richard and Dudley Seers (1972) “The Brain Drain and the Development Process”,
in Gustav Ranis (ed.) The Gap between Rich and Poor Nations, New York: St.
Martin’s Press, 365-79.
Kao, C. and J. W. Lee (1973) “An Empirical Analysis of China’s Brain Drain to the United
States,” Economic Development and Cultural Change, 21(3), 500-13.
Kaya, M. (2002) “Beyin göçü / erozyonu” (Brain drain / erosion), Technology Research
Centre Report, Eski ehir: Osmangazi University, Eski ehir, November 5, 2002.
Katz, E. and O. Stark (1984) “Migration and Asymmetric Information: Comment,”
American Economic Review, 74, 533-534.
Kepenek, Yakup and Nurhan Yentürk (2000) Türkiye Ekonomisi, stanbul: Remzi Kitabevi.
Khadria, Binod (1999) The Migration of Knowledge Workers: Second-Generation Effects
of India’s Brain Drain, London: Sage Publications.
218
Kim, Jinyoung (1998) “Economic Analysis of Foreign Education and Students Abroad,”
Journal of Development Economics, 56, 337-65.
Kwok, V. and H. Leland (1982) “An Economic Model of the Brain Drain,” American
Economic Review, 72(1), 91-100.
Kwok, V. and H. Leland (1984) “Migration and Asymmetric Information: Reply,”
American Economic Review, 74, 535.
Koç, smet and I ıl Onan (2001) “The Impact of Remittances of International Migrants on
the Standard of Living of the Left-Behind Families in Turkey,” paper presented at
IUSSP XXIVth General Population Conference in Salvador de Bahia, Brazil, August
18-24 (http://www.iussp.org/Brazil2001/s20/S26_03_Koc.pdf).
Kunze, A. (2002) “The Timing of Careers and Human Capital Depreciation,” IZA
Discussion Paper No. 509, Bonn, Germany: Institute for the Study of Labor.
Kurtulu , B. (1999) Amerika Birle ik Devletleri’ne Türk Beyin Göçü (Turkish Brain Drain
to the United States), stanbul: Alfa Basım Yayım Da ıtım.
Kuznets, Simon (1955) “Economic Growth and Income Inequality,” American Economic
Review, 45(1), 1-28.
Lambooy, Jan G. (1997) “Knowledge Production, Organisation and Agglomeration
Economies,” GeoJournal, 41(4), 293-300.
Laudel, G. (2003) “Studying the Brain Drain: Can Bibliometric Methods Help?”
Scientometrics, 57(2), 215-37.
Liao, T. F. (1994) Interpreting Probability Models: Logit, Probit, and Other Generalized
Linear Models, Sage University Paper series on Quantitative Applications in the
Social Sciences, 07-101, Thousand Oaks, CA: Sage.
Lien, D.-H. D. (1987a) “Economic Analysis of Brain Drain,” Journal of Development
Economics, 25, 33-43.
Lien, D.-H. D. (1987b) “Asymmetric Information and Multi-Stage Brain Drain,”
Economics Letters, 23(3), 305-309.
Lien, Da-Hsiang Donald (1988) “Appropriate Scientific Research and Brain Drain: A
Simple Model,” Journal of Development Economics, 29, 77-87.
Long, J. S. and J. Freese (2001) Regression Models for Categorical Dependent Variables
Using Stata, College Station, TX: Stata Press.
Lowell, B. L. and A. M. Findlay (2001) “Migration of Highly Skilled Persons from
Developing Countries: Impact and Policy Responses,” draft report prepared for the
International Labour Office (ILO), Geneva, June 1, 2001.
219
Lucas, R. E. Jr. (1975) “The Supply-of-Immigrants Function and Taxation of Immigrants’
Incomes,” Journal of Development Economics, 2, 289-308.
Lucas, R. E. Jr. (1988) “On the Mechanics of Economic Development,” Journal of
Monetary Economics, 22, 3-42.
Lundahl, Mats (1985) “Brain Drain, Illegal Migration and Capital Exports from Less
Developed Countries: A Neoclassical Approach,” Economics Letters, 17(3), 277280.
MacWilliams, B. (2000) “Turkey's old-fashioned distance education draws the largest
student body on earth: Anadolu University uses radio, TV, and (gulp) mailed
material to educate a rural population,” The Chronicle of Higher Education,
September 22.
Maddala, G. S. (1983) Limited-Dependent and Qualitative Variables in Econometrics,
Econometric Society Monographs, Cambridge: Cambridge University Press.
Mahroum, Sami (1999) “Highly Skilled Globetrotters: The International Migration of
Human Capital,” paper presented at OECD Workshop on Science and Technology
Labour Markets, 17 May, (http://www.oecd.org/pdf/M000014000/M00014667.pdf).
Maresová, Jarmila (1999) “Labor Migration to Austria: Czech and Slovak Temporary
Workers in Vienna,” in Paradigms and Contentions, IWM Junior Visiting Fellows
Conferences, Volume 7, Vienna.
Massey, Douglas S., Joaquin Arango, Graeme Hugo, Ali Kouaouci, Adela Pellegrino and J.
Edward Taylor (1993) “Theories of International Migration: A Review and
Appraisal,” Population and Development Review, 19(3), 431-66.
Massey, Douglas S., Joaquin Arango, Graeme Hugo, Ali Kouaouci, Adela Pellegrino and J.
Edward Taylor (1994) “An Evaluation of International Migration Theory: The North
American Case,” Population and Development Review, 20(4), 699-752.
Mattoo, A. and A. Carzaniga (eds.) (2003) Moving People to Deliver Services, Oxford:
Oxford University Press and the World Bank.
Meyer, J.-B. and M. Brown (1999) “Scientific Diasporas: A New Approach to the Brain
Drain,” Management of Social Transformations Discussion Paper No. 41, paper
prepared for the World Conference on Science, UNESCO-ICSU, Budapest, Hungary,
26 June-1 July.
McCormick, Barry and Jackline Wahba (2000) “Overseas Employment and Remittances to
a Dual Economy, The Economic Journal.
McCulloch, Rachel and Janet L. Yellen (1975) “Consequences of a Tax on the Brain Drain
for Unemployment and Inequality in Less Developed Countries,” Journal of
Development Economics, 2, 249-64.
220
Milanovic, Branko (2002) “True World Income Distribution, 1988 and 1993: First
Calculation Based on Household Surveys Alone,” The Economic Journal, 112,
51-92.
Ministry of Foreign Affairs, “Turks Living Abroad”, Ankara: Republic of Turkey Ministry
of Foreign Affairs (http://www.mfa.gov.tr/grupa/ac/acb/default1.htm; accessed
19.11.2000).
Mincer, Jacob (1962) “On-the-Job Training: Costs, Returns, and Some Implications,”
Journal of Political Economy, Supplement, 70(5), 50-79.
Mincer, Jacob (1978) “Family Migration Decisions,” Journal of Political Economy, 86(5),
749-73.
Miyagiwa, K. (1991) “Scale Economies in Education and the Brain Drain Problem,”
International Economic Review, 32(3), 743-59.
Mountford, A. (1997) “Can a Brain Drain be Good for Growth in the Source Economy?”
Journal of Development Economics, 53, 287-303.
Myers, Robert G. (1972) Education and Emigration: Study Abroad and the Migration of
Human Resources, New York: David McKay Company, Inc.
National Science Board (2002b), Science and Engineering Indicators – 2002. Volume 2:
Appendix Tables, Arlington, VA: National Science Foundation.
Naumova, T. V. (1998) “Russia’s ‘Brain Drain’,” Russian Social Science Review, 39(2),
49-56.
Niessen, Jan (2002) “International Mobility in a Globalising World,” paper presented at the
ACP-EU Joint Parliamentary Assembly Workshop on Migration and Development,
Capetown, 20 March.
Niland, J. R. (1970) The Asian Engineering Brain Drain: A Study of International
Relocation into the United States from India, China, Korea, Thailand and Japan, D.
C. Heath and Company, Lexington, Massachusetts.
NSF (1997) Survey of Doctorate Recipients, Arlington, VA: National Science Founcation,
Division of Science Resources Studies.
OECD (1994) Proposed Standard Practice for Surveys of Research and Experimental
Development: “Frascati Manual 1993”, The Measurement of Scientific and
Technological Activities Series, Paris: Organization for Economic Co-operation and
Development.
OECD (1998a) Human Capital Investment: An International Comparison, Paris:
Organization for Economic Co-operation and Development.
OECD (1998b) Trends in International Migration, Paris: Organization for Economic
Co-operation and Development.
221
O uzkan, T. (1971) Yurt Dı ında Çalı an Doktoralı Türkler: Türkiye’den Ba ka Ülkelere
Yüksek Seviyede Eleman Göçü Üzerinde Bir Ara tırma (Turks with Doctorate
Degrees Working Abroad: An Investigation of the Migration of Highly Skilled
Workers from Turkey to Other Countries), Ankara: Orta Do u Teknik Üniversitesi.
O uzkan, T. (1975) “The Turkish Brain Drain: Migration of Tendencies among Doctoral
Level Manpower,” in R. E. Krane (ed.) Manpower Mobility Across Cultural
Boundaries: Social, Economic and Legal Aspects, The Case of Turkey and West
Germany, Leiden, Netherlands: E.J. Brill.
Oni, Bankolé (2000) “Capacity Building and Brain Drain in Nigerian Universities,” The
Regional Conference on Brain Drain and Capacity Building in Africa,
Addis Ababa, 22 - 24 February.
Pampel, F. C. (2000) Logistic Regression: A Primer, Sage University Paper series on
Quantitative Applications in the Social Sciences, 07-132, Thousand Oakes, CA:
Sage.
Partridge, Mark D. (1997) “Is Inequality Harmful for Growth: Comment,” The American
Economic Review, 87(5), 1019-32.
Patinkin, D. (1968) “A ‘Nationalist’ Model,” in W. Adams (ed.) The Brain Drain, New
York: The Macmillan Company.
Persson, Torsten and Guido Tabellini (1994) “Is Inequality Harmful for Growth?” The
American Economic Review,” 84(3), 600-21.
Poot, Jacques (1996) “Information, Communication and Networks in International
Migration Systems,” The Annals of Regional Science, 30, 55-73.
Ranis, Gustav (ed.) (1972) The Gap Between Rich and Poor Nations, New York: St.
Martin’s Press.
Rauch, James E. (1991) “Reconciling the Pattern of Trade with the Pattern of Migration,”
American Economic Review, 81(4), 775-96.
Rea, L. M. and R. A. Parker (1997) Designing and Conducting Survey Research: A
Comprehensive Guide, Second Edition, San Francisco: Jossey-Bass Publishers.
Regets, Mark C. (2001) “Research and Policy Issues in High-Skilled International
Migration: A Perspective with Data from the United States,” IZA Discussion Paper
No. 366, Bonn, Germany: Institute for the Study of Labor.
Riley, John G. (1975) “Competitive Signalling,” Journal of Economic Theory, 10, 174-186.
Rodriguez, Carlos Alfredo (1975) “Brain Drain and Economic Growth: A Dynamic
Model,” Journal of Development Economics, 2, 223-47.
222
Russell, Sharon Stanton (1995) “International Migration: Implications for the World Bank,”
Human Resources Development and Operations Policy Working Paper HROWP 54,
Washington, DC: The World Bank,
(http://www.worldbank.org/html/extdr/hnp/hddflash/workp/wp_00054.html) *
SAMP (South African Migration Project) (2000) Losing our Minds: Skills Migration and
the South African Brain Drain, Cape Town, South Africa: Idasa Publishers.
Schelling, T. C. (1978) Micromotives and Macrobehavior, New York and London: W. W.
Norton & Company.
Schott, Thomas (1998) “Ties between Center and Periphery in the Scientific WorldSystem: Accumulation of Rewards, Dominance and Self-Reliance in the Center,”
Journal of World-Systems Research, 4: 112-44.
Schultz, T. Paul (1998) “Inequality in the Distribution of Personal Income in the World:
How it is Changing and Why,” Journal of Population Economics, 11(3), 307-44.
enses, F. (1994) “Labour market response to structural adjustment and institutional
pressures: The Turkish case,” METU Studies in Development, 21(3), 405-48.
SIS (1996) Statistical Indicators 1923-1996, Publication No. 1883, Ankara: State Institute
of Statistics, Republic of Turkey.
SIS (2002) 2001 Statistical Yearbook of Turkey, Publication No. 2690, Ankara: State
Institute of Statistics, Republic of Turkey.
SIS (2003) 2000 Census of Population: Social and Economic Characteristics of the
Population, Publication No. 2759, Ankara: State Institute of Statistics, Republic of
Turkey.
Sjaastad, L. A. (1962) “The Costs and Returns of Human Migration,” Journal of Political
Economy, 70(5), 80-93.
Solimano, Andrés (2001) “International Migration and the Global Economic Order: An
Overview,” Policy Research Working Paper No. 2720, Washington, DC: The World
Bank, (econ.worldbank.org/files/2728_wps2720.pdf). *
Solow, R.M. (1957) “Technical Change and the Aggregate Production Function,” Review
of Economics and Statistics, 47, 312-320.
SPO (2000) Sekizinci Be Yıllık Kalkınma Planı, 2001-2005 (Eighth Five-Year
Development Plan, 2001-2005), Ankara: State Planning Organization.
Stark, O. and D. E. Bloom (1985) “The New Economics of Labor Migration,” American
Economic Review, Papers and Proceedings, 75(2), 173-178.
Stark, O., C. Helmenstein and A. Prskawetz (1997) “A Brain Gain with a Brain Drain,”
Economics Letters, 55, 227-234.
223
Stark, O., C. Helmenstein and A. Prskawetz (1998) “Human Capital Depletion, Human
Capital Formation, and Migration: A Blessing or a Cure?” Economics Letters, 60,
363-367.
Stark, O. and Y. Wang (1999) “Externalities, Human Capital Formation, and Corrective
Migration Policy,” ZEF Discussion Papers on Development Policy No. 11, Bonn:
Center for Development Research, Bonn University.
Stark, O. and Y. Wang (2002a) “Inducing Human Capital Formation: Migration as a
Substitute for Subsidies,” Journal of Public Economics, 86, 29-46.
Stark, Oded and You Qiang Wang (2002b) “Migration Dynamics,” Economics Letters, 76,
159-64.
Straubhaar, Thomas (1986) “The Determinants of Workers’ Remittances: The Case of
Turkey,” Weltwirtschaftliches Archiv, 122(4), 728-40.
Straubhaar, Thomas (2000) “International Mobility of the Highly Skilled: Brain Gain,
Brain Drain or Brain Exchange,” HWWA Discussion Paper No. 88, Hamburg
Institute of International Economics.
Tansel, A. (1994), “Wage employment, earnings and returns to schooling for men and
women in Turkey”, Economics of Education Review, 13(4), 305-320.
Tansel, A. (1999) “Türkiye ve Seçilmi Ülkelerde E itimin Getirisi”, (Returns to education
in Turkey and selected countries), METU Studies in Development, 26(3-4), 473-491.
Tansel, A. (2002a) “Determinants of School Attainment of Boys and Girls in Turkey:
Individual, Household and Community Factors,” Economics of Education Review,
21(5), 455-470.
Tansel, A. (2002b) “Economic development and female labour force participation in
Turkey: Time series evidence and cross province estimates,” in T. Bulutay (ed.)
Employment of Women, Ankara: State Institute of Statistics, 111-51.
Tansel, Aysıt (2002c) “General versus vocational high schools and labour market outcomes
in Turkey,” in . Sirageldin (Ed.), Human Capital: Population Economics in the
Middle East, Cairo: American University in Cairo and Economic Research Forum,
258-72.
Tansel, A. and F. Bircan (2002) “Demand for schooling and private tutoring in Turkey,”
paper presented at the VI. ERC/METU International Conference on Economics,
September 12-14, Ankara.
Tansel, Aysıt and Nil Demet Güngör (1997) “The Educational Attainment of Turkey’s
Labor Force: A Comparison Across Provinces and Over Time,” METU Studies in
Development, 24(4), 531-47.
Tunalı, nsan (2000) “Rationality of Migration,” International Economic Review, 41(4),
893-920.
224
Tunalı, nsan (2003) “Rationality of Migration,” in smail Sirageldin (ed.) Human Capital:
Population Economics in the Middle East, Cairo: American University in Cairo
Press.
U.S. Congress, House of Representatives (1967) The Brain Drain into the United States of
Scientists, Engineers, and Physicians, A Staff Study for the Research and Technical
Programs Subcommittee of the Committee on Government Operations, 90th
Congress, 1st Session, July 1967.
U.S. Congress, House of Representatives (1968a) The Brain Drain of Scientists, Engineers,
and Physicians from the Developing Countries into the United States, Hearing before
a Subcommittee on Government Operations, 90th Congress, 2nd Session, January 23,
1968.
U.S. Congress, House of Representatives (1968b) Scientific Brain Drain from the
Developing Countries, Twenty-Third Report by the Committee on Government
Operations, House Report No. 1215, Union Calendar No. 474, 90th Congress, 2nd
Session, March 28, 1968.
Valdés, Benigno (1999) Economic Growth: Theory, Empirics and Policy, Cheltenham, UK:
Edward Elgar.
Vidal, J.-P. (1998) “The Effect of Emigration on Human Capital Formation,” Journal of
Population Economics, 11, 589-900.
Vogler, M. and R. Rotte (2000) “The Effects of Development on Migration: Theoretical
Issues and New Empirical Evidence,” Journal of Population Economics, 13, 458508.
Watanabe, S. (1969) “The Brain Drain from Developing to Developed Countries,”
International Labour Review, 99(4), 401-33.
Winters, L. A., T. L. Walmsley, Z. K. Wang, and R. Grynberg (2003) “Liberalising
Temporary Movement of Natural Persons: An Agenda for the Development Round,”
World Economy, 26, 1137-61, http://ssrn.com/abstract=443784.
Wong, K. (1995) “International Labor Migration,” Chapter 14 in International Trade in
Goods and Factor Mobility, Cambridge, MA: The MIT Press.
Wong, K. and C. K. Yip (1999) “Education, Economic Growth, and the Brain Drain,”
Journal of Economic Dynamics and Control, 23, 699-726.
World Bank (2000) “Scientific and Technical Manpower Development in India,” Report
No. 20416-IN, Education Sector Unit, South Asia Region, Washington, DC: The
World Bank.
Yoon, B.-S. L. (1992) “Reverse Brain Drain in South Korea: State-led Model,” Studies in
Comparative International Development, 27(1), 4-26.
225
YÖK (2001) Yüksekö retim Kurumu Faaliyet Raporu (Activities Report of the Higher
Education Council), Ankara: Higher Education Council.
YÖK (2003) Yüksekö retim Kurumu Faaliyet Raporu (Activities Report of the Higher
Education Council), Ankara: Higher Education Council.
Zang, Wei-Bin (1990) “Brain Drain and Economic Cycles with International Migration: A
Case of Minimum Wage in the Unskilled Sector,” Journal of Development
Economics, 32, 191-203.
Zweig, D. and Changgui, C. (1995) China’s Brain Drain to the United States: Views of
Overseas Chinese Students and Scholars in the 1990s, China Research Monograph,
Institute of East Asian Studies, University of California, Berkeley, CA.
Zweimüller, Josef (2000) “Inequality, Redistribution, and Economic Growth,” Empirica,
27: 1-20.
226
APPENDIX A
SUPPLEMENTARY TABLES FOR CHAPTER SIX
Table A.1 Respondents by Age and Gender (%)
Professionals
Male
Female
Total
Age
18-20
21-25
26-30
31-35
36-40
41-45
46-50
50+
(n = 879)
...
4.6
27.5
26.5
13.4
8.8
8.0
11.3
2
(n = 345)
Male
Age
(n = 1224)
...
7.5
39.7
24.9
8.7
8.1
6.4
4.6
...
5.4
31.0
26.1
12.1
8.6
7.5
9.4
18-20
21-25
26-30
31-35
36-40
41-45
46-50
50+
(6) = 33.31***
(n = 676)
2.8
34.8
46.5
13.8
1.8
0.4
...
...
2
Students
Female
Total
(n = 427) (n = 1103)
2.6
33.0
50.1
12.2
1.4
0.7
...
...
2.7
34.1
47.9
13.2
1.6
0.5
...
...
(5) = 2.06
Notes: ***p < 0.001, **p < 0.005, *p < 0.010 for the chi-square test of independence. Cell percentages sum
to 100 across columns.
Table A.2 Marital Status of Respondents
Marital Status
Professionals
n
%
Never Married
414
35.1
777
70.9
72
6.1
22
2.0
692
58.7
297
27.1
422
190
80
...
35.8
16.1
6.8
...
254
25
12
6
23.2
2.3
1.1
0.5
1178
100.0
1096
100.0
Divorced / Separated / Widowed
Married
Spouse’s Nationality = Turkish
Spouse’s Nationality = Foreign
Spouse’s Nationality = Dual Citizen
Spouse’s Nationality = Not Indicated
Total
Students
n
%
Note: There are 46 missing responses in the professionals and 7 missing responses in the student survey.
227
Table A.3 Stay Duration of Respondents by Gender (%)
Stay Duration
Male
Professionals
(n = 879)
< 1 year
1 - 5 years
6 - 10 years
11 - 15 years
15 - 20 years
20 - 25 years
25 - 30 years
> 30 years
10.4
32.7
25.0
11.3
5.2
9.0
4.3
2.2
Students
< 6 months
6 - 12 months
1 - 2 years
3 - 4 years
5 - 6 years
7 years
(n = 676)
Female
Total
(n = 345) (n = 1224)
8.1
46.1
24.1
9.0
3.5
6.4
1.7
1.2
9.7
36.4
24.8
10.6
4.7
8.3
3.6
1.9
(n = 427) (n = 1103)
9.9
12.9
26.6
29.4
13.0
8.1
12.7
11.7
29.0
26.0
15.2
5.4
11.0
12.4
27.6
28.1
13.9
7.1
Note: Cell percentages sum to 100 across columns.
Table A.4 Respondents by Country of Residence
Country
ISO Code
Freq.
%
944
40
39
22
13
9
8
8
6
5
4
3
1
1
85.6
3.6
3.5
2
1.2
0.8
0.7
0.7
0.5
0.5
0.4
0.3
0.1
0.1
1103
100
Students
United States
Canada
United Kingdom
Germany
Japan
France
Australia
Austria
Belgium
Finland
Netherlands
Switzerland
Italy
Spain
USA
CAN
GBR
DEU
JPN
FRA
AUS
AUT
BEL
FIN
NLD
CHE
ITA
ESP
Total
228
Table A.4 continued
Country
ISO Code
Freq.
%
856
75
62
48
34
25
24
23
18
10
9
8
7
5
3
2
2
2
2
1
1
1
1
1
1
1
1
1
69.9
6.1
5.1
3.9
2.8
2
2
1.9
1.5
0.8
0.7
0.7
0.6
0.4
0.3
0.2
0.2
0.2
0.2
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
1224
100
Professionals
United States
Canada
Germany
United Kingdom
Australia
Belgium
Switzerland
Netherlands
France
Austria
United Arab Emirates
Japan
Finland
Saudi Arabia
Italy
Hungary
Kazakhstan
Norway
Sweden
Algeria
China
Ireland
Israel
Malaysia
Mexico
Romania
Singapore
South Africa
USA
CAN
DEU
GBR
AUS
BEL
CHE
NLD
FRA
AUT
ARE
JPN
FIN
SAU
ITA
HUN
KAZ
NOR
SWE
DZA
CHN
IRL
ISR
MYS
MEX
ROM
SGP
ZAF
Total
229
Table A.5. Respondents by Father’s Occupation
Father’s Occupation
Total
Students
Professionals
n
%
n
%
n
%
1144
50.4
548
50.3
596
50.5
Architect, engineer or related professionals
Science and technology professionals
Health professionals
Other health-related workers (e.g., nurses)
Legal, business or public service professionals
Academicians
Teachers – pre, primary or secondary
Teachers – other
Culture, media or sports professionals
356
44
152
4
254
129
123
62
11
15.7
1.9
6.7
0.2
11.2
5.7
5.4
2.7
0.5
189
16
65
1
107
66
72
28
4
17.4
1.5
6.0
0.1
9.8
6.1
6.6
2.6
0.4
167
28
87
3
147
63
51
34
7
14.1
2.4
7.4
0.3
12.4
5.3
4.3
2.9
0.6
Administrators, managers ................................
Clerks, secretaries, other admin. workers ........
Sales or related workers ...................................
Services workers ..............................................
Trades, crafts, arts and related workers ............
Armed forces occupations ...............................
Other ................................................................
267
32
32
68
177
154
386
11.8
1.4
1.4
3.0
7.8
6.8
17.0
149
17
17
28
85
57
183
13.7
1.6
1.6
2.6
7.8
5.2
16.8
118
15
15
40
92
97
203
10.0
1.3
1.3
3.4
7.8
8.2
17.2
Not known .......................................................
10
0.4
5
0.5
5
0.4
2270
66
100.0
1089
14
100.0
1181
52
100.0
Scientific, Technical and Related Professions
Total (valid responses)
Missing Responses
230
Table A.6. Respondents by Mother’s Occupation
Mother’s Occupation
Total
Students
Professionals
n
%
n
%
n
%
Scientific, Technical and Related Professions
784
34.8
402
37.1
382
32.7
Architect, engineer or related professionals
Science and technology professionals
Health professionals
Other health-related workers (e.g., nurses)
Legal, business or public service professionals
Academicians
Teachers – pre, primary or secondary
Teachers – other
Culture, media or sports professionals
81
12
111
27
72
58
299
116
8
3.6
0.5
4.9
1.2
3.2
2.6
13.3
5.1
0.4
53
5
57
15
31
35
153
51
2
4.9
0.5
5.3
1.4
2.9
3.2
14.1
4.7
0.2
28
7
54
12
41
23
146
65
6
2.4
0.6
4.6
1.0
3.5
2.0
12.5
5.6
0.5
Administrators, managers ................................
Clerks, secretaries, other admin. workers ........
Sales or related workers ...................................
Services workers ..............................................
Trades, crafts, arts and related workers ............
Armed forces occupations ...............................
Other ................................................................
87
44
9
20
32
1
168
3.9
2.0
0.4
0.9
1.4
0.0
7.5
41
23
4
9
18
1
86
3.8
2.1
0.4
0.8
1.7
0.1
7.9
46
21
5
11
14
0
82
3.9
1.8
0.4
0.9
1.2
0.0
7.0
Homemaker .....................................................
.......................................................
Not known .......................................................
1106
2
49.1
0.1
499
1
46.0
0.1
607
1
51.9
0.1
Total (valid responses)
Missing Responses
2253
74
100.0
1084
19
100.0
1169
55
100.0
Table A.7 Language of Instruction in High School, Science and Social
Science Courses (%)
Language
of Instruction
Turkish
English
French
German
Italian
Professionals (n = 1224)
Science
Social
Students (n = 1103)
Science
Social
44.6
94.3
43.8
96.2
44.0
4.9
3.8
1.2
47.4
2.7
2.8
0.5
6.2
0.3
0.5
0.3
5.9
0.2
0.0
Note: Cell percentages sum to 100 across columns.
231
0.5
Table A.8 Bachelor’s Degree Institutions of Respondents
Students
Alma Mater
Professionals
n
%
Orta Do u Teknik
Bo aziçi
Bilkent
stanbul Teknik
stanbul
Ankara
Foreign University
Hacettepe
Marmara
Yıldız Teknik
Dokuz Eylül
Koç
Çukurova
Ege
Gazi
Uluda
Anadolu
Akdeniz
Gaziantep
Karadeniz Teknik
Osmangazi
Selçuk
Abant zzet Baysal
Balıkesir
Çanakkale
Galatasaray
nonu
Atatürk
Ba kent
Çankaya
I ık
Kocaeli
Mimar Sinan
Ni de
Polis Akademisi
Sabancı
Sakarya
Samsun 19 Mayıs
Hava Harp
318
169
108
62
43
41
35
34
28
16
15
15
12
11
10
7
6
3
3
3
3
3
2
2
2
2
2
1
1
1
1
1
1
1
1
1
1
1
1
32.9
17.5
11.2
6.4
4.5
4.2
3.6
3.5
2.9
1.7
1.6
1.6
1.2
1.1
1.0
0.7
0.6
0.3
0.3
0.3
0.3
0.3
0.2
0.2
0.2
0.2
0.2
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Total
Not indicated
967
26
100.0
Alma Mater
Orta Do u Teknik
Bo aziçi
Foreign University
stanbul Teknik
Bilkent
stanbul
Hacettepe
Ankara
Marmara
Ege
Yıldız Teknik
Dokuz Eylül
Gazi
Do u Akdeniz
Koç
Mimar Sinan
Anadolu
Atatürk
Çukurova
Uluda
Deniz Harp
Hava Harp
I ık
Karadeniz Teknik
Kocaeli
Abant zzet Baysal
Akdeniz
Dicle
Fırat
ktisadi ve Ticari limler
Kadir Has
Kara Harp
Mersin
Osmangazi
Zonguldak Karaelmas
Total
Not indicated
232
n
%
410
207
141
137
71
53
51
29
23
22
19
10
6
4
4
4
3
3
3
3
2
2
2
2
2
1
1
1
1
1
1
1
1
1
1
33.5
16.9
11.5
11.2
5.8
4.3
4.2
2.4
1.9
1.8
1.6
0.8
0.5
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.2
0.2
0.2
0.2
0.2
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
1223
1
100.0
Table A.9. Detailed Undergraduate Fields of Students with Bachelor’s Degrees
Bachelor’s Degree Fields
n
%
Architecture and City Planning, Total ..................................
16
1.6
Architecture
City and Urban Planning
Economic and Administrative Sciences, Total .....................
Business Administration
Economics
Finance
International Relations
Political Science and Public Administration
International Trade
Educational Sciences, Total ....................................................
Art Education
Curriculum Planning
Educational Sciences
Elementary Education
Foreign Languages Education
Physical Education and Sport
Science and Mathematics Education
Social Sciences Education
Special Education
Music Education
Counselling
Education, field not specified
Engineering and Technical Sciences, Total ..........................
Agricultural Sciences / Agricultural Engineering
Aeronautical / Aerospace Engineering
Biomedical Engineering
Chemical Engineering
Civil Engineering
Computer Science
Computer Engineering
Electric-Electronic Engineering
Engineering Sciences
Environmental Engineering
Food Engineering
Forestry
Geological Engineering
Geomatic Engineering (Geodesy/Photogrammettry)
Geophysics Engineering
Industrial Engineering
Maritime Eng. (Naval Arch., Ship Building, Marine Eng.)
Mechanical Engineering
Metallurgical and Materials Engineering
Mining Engineering
Nuclear Engineering
Petroleum and Natural Gas Engineering
233
7
9
0.7
0.9
179
18.0
44
74
2
31
25
3
4.4
7.5
0.2
3.1
2.5
0.3
60
6.0
2
1
2
3
12
4
25
1
2
1
3
4
0.2
0.1
0.2
0.3
1.2
0.4
2.5
0.1
0.2
0.1
0.3
0.4
504
50.8
13
10
1
38
41
15
34
134
1
18
13
4
2
2
2
50
1
72
20
7
3
9
1.3
1.0
0.1
3.8
4.1
1.5
3.4
13.5
0.1
1.8
1.3
0.4
0.2
0.2
0.2
5.0
0.1
7.3
2.0
0.7
0.3
0.9
Table A.9 continued
Bachelor’s Degree Fields
Physics Engineering
Textiles Engineering
Engineering Management
Mathematical Engineering
Fishery Sciences and Engineering
Engineering, field not specified
Language and Literature, Total ............................................
Turkish Language and Literature
Eastern Languages and Literatures
Western Languages and Literatures
Comparative Literature Studies
Interpretation/Translation
Math and Natural Sciences, Total .........................................
Astronomy and Space Sciences
Biology / Molecular Biology and Genetics
Biochemistry
Chemistry
Mathematics
Physics
Statistics
Medical and Health Sciences, Total ......................................
Child Care and Development
Dentistry
Medicine – Genera l
Nursing
Pharmacology and Pharmaceutical Sciences
Sports Medicine
Veterinary Sciences
Social Sciences, Total ..............................................................
Communication
Geography
History
Law
Philosophy
Public Relations
Psychology
Social Work
Sociology
Tourism and Hotel Management
Social Sciences, field not specified
Arts, Total ................................................................................
Music
Radio, Television and Cinema
Discipline and field not specified ...................................................
TOTAL
234
n
%
2
2
4
1
1
4
0.2
0.2
0.4
0.1
0.1
0.4
15
1.5
1
1
8
1
4
0.1
0.1
0.8
0.1
0.4
127
12.8
1
28
3
24
31
37
3
0.1
2.8
0.3
2.4
3.1
3.7
0.3
28
2.8
1
3
13
1
5
1
4
0.1
0.3
1.3
0.1
0.5
0.1
0.4
48
4.8
1
1
1
5
4
1
14
1
16
2
2
0.1
0.1
0.1
0.5
0.4
0.1
1.4
0.1
1.6
0.2
0.2
5
0.5
3
2
0.3
0.2
11
1.1
993
100.0
Table A.10 Detailed Undergraduate Fields of Overseas Turkish Workforce
Bachelor’s Degree Fields
Architecture and City Planning, Total ..................................
Architecture
City and Regional Planning
Industrial Design
Interior Architecture and Environmental Design
Economic and Administrative Sciences, Total .....................
Accounting
Business Administration
Economics
Econometrics
Finance
International Relations
Labour Economics and Industrial Relations
Political Science and Public Administration
Economic and Administrative Sciences, field not specified
Educational Sciences, Total ....................................................
Art Education
Computer Education and Instructional Technology
Educational Sciences
Foreign Languages Education
Secondary Science and Mathematics Education
Education, field not specified
Engineering and Technical Sciences, Total ..........................
Agricultural Sciences / Agricultural Engineering
Aeronautical and Aerospace Engineering
Biomedical Engineering
Chemical Engineering
Civil Engineering
Communications Technologies
Computer Science
Computer Engineering
Electrical-Electronics Engineering
Engineering Sciences
Environmental Engineering
Food Engineering
Forestry
Geological Engineering
Industrial Engineering
Maritime Engineering
Mechanical Engineering
Metallurgical and Materials Engineering
Mining Engineering
Nuclear Engineering
Petroleum and Natural Gas Engineering
Physics Engineering
235
Freq.
%
40
27
5
5
3
209
3
85
81
3
4
13
1
15
4
16
1
3
1
5
5
1
774
1
18
1
48
61
4
33
88
249
2
13
7
2
10
86
3
92
10
9
2
13
4
3.3
2.2
0.4
0.4
0.3
17.1
0.3
6.9
6.6
0.3
0.3
1.1
0.1
1.2
0.3
1.3
0.1
0.3
0.1
0.4
0.4
0.1
63.2
0.1
1.5
0.1
3.9
5.0
0.3
2.7
7.2
20.3
0.2
1.1
0.6
0.2
0.8
7.0
0.3
7.5
0.8
0.7
0.2
1.1
0.3
Table A.10 continued
Bachelor’s Degree Fields
Textiles Engineering
Automotive Engineering
Management Engineering
Engineering, field not specified
Language and Literature, Total ............................................
Ancient Languages and Cultures
Western Languages and Literatures
Comparative Literature Studies
Math and Natural Sciences, Total .........................................
Biology / Molecular Biology and Genetics
Biochemistry
Chemistry
Mathematics
Physics
Science – General
Statistics
Math and Natural Sciences, field not specified
Medical and Health Sciences, Total ......................................
Medicine – General
Pharmacology and Pharmaceutical Sciences
Veterinary Sciences
Social Sciences, Total ..............................................................
Anthropology
Archeology
Art History
Communication
History
Journalism
Law
Library Sciences
Linguistics
Philosophy
Public Relations
Psychology
Social Work
Sociology
Tourism and Hotel Management
Arts, Total ................................................................................
Fine Arts
Graphic Arts
Ceramic and Glass
TOTAL
236
Freq.
%
2
4
2
10
0.2
0.3
0.2
0.8
8
1
5
2
0.7
0.1
0.4
0.2
76
10
4
14
22
15
2
7
2
6.2
0.8
0.3
1.1
1.8
1.2
0.2
0.6
0.2
54
47
3
4
4.4
3.8
0.3
0.3
44
1
3
1
1
2
2
4
1
2
3
2
11
1
3
7
3.6
0.1
0.3
0.1
0.1
0.2
0.2
0.3
0.1
0.2
0.3
0.2
0.9
0.1
0.3
0.6
3
1
1
1
0.2
0.1
0.1
0.1
1224
100
Table A.11 Current Program of Study by Gender, Students (%)
Male
Female
Total
(n = 676)
(n = 427)
(n = 1103)
Bachelors
11.0
10.5
10.8
Masters
Doctorate
25.6
57.7
30.4
55.0
27.5
56.7
Postdoctorate
5.8
4.0
5.1
Test of independence
2
Program
(3) = 4.26
Notes: ***p < 0.001, **p < 0.005, *p < 0.010; Cell percentages sum to 100
across columns; The bachelor's category includes three students pursuing an
associate’s degree and three students in the post-bachelor's certificate
program; Five students in the post-master’s certificate program are included
in the master’s category.
Table A.12 Highest Degree Planned by Gender, Students (%)
Male
(n = 676)
Female
(n = 427)
Total
(n = 1103)
Doctorate
Masters
Bachelors
Post-Masters Certificate
Post-Bachelors Certificate
Associates
74.9
22.5
2.1
0.4
0.2
0.0
74.0
23.9
0.9
0.5
0.5
0.2
74.5
23.0
1.6
0.5
0.3
0.1
Test of independence
2
Degree
***
Notes: p < 0.001,
across columns.
**
(5) = 4.89
*
p < 0.005, p < 0.010; Cell percentages sum to 100
Table A.13 Living Accommodations by Study Program, Students (%)
Living
Accomodation
Apartment
Room in apartment
Dorm
House
Room in house
Other
Bachelors
(n = 118)
30.5
44.9
11.0
2.5
5.1
5.9
Masters
(n = 300)
47.3
13.0
13.3
2.3
16.3
7.7
Doctorate
(n = 623)
62.0
5.0
10.1
1.6
13.0
8.4
Postdoc
(n = 56)
Total
(n = 1097)
67.9
3.6
16.1
1.8
1.8
8.9
Notes: There are six missing responses; Cell percentages sum to 100 across columns.
237
54.9
11.4
11.4
1.9
12.5
7.9
Table A.14 Living On or Off Campus by Study Program (%)
Program
No
119
303
625
56
51.3
76.9
77.9
82.1
48.7
23.1
22.1
17.9
1103
75.0
25.0
Bachelors
Masters
Doctorate
Postdoc
Total
On Campus?
Yes
n
Note: Cell percentages sum to 100 across rows.
Table A.15 Current Field of Study by Gender, Students (%)
Male
Field
(n = 674)
Engineering and Technical Sciences
Economic and Administrative Sciences
Math and Natural Sciences
Social Sciences
Educational Sciences
Medical and Health Sciences
Architecture and Urban Planning
Language and Literature
Arts
Female
(n = 426)
Total
(n = 1100)
29.3
33.6
11.7
11.3
7.3
2.6
1.9
1.4
0.9
43.9
27.8
11.5
6.6
5.4
2.2
1.2
0.8
0.7
53.1
24.2
11.3
3.6
4.2
1.9
0.7
0.5
0.6
2
Test of independence
(8) = 77.09***
Notes: ***p < 0.001, **p < 0.005, *p < 0.010; Cell percentages sum to 100 across columns.
There are three missing responses.
Table A.16 Students by Current Program and Field of Study (%)
Current Program
Bachelors
Masters
Doctorate
Field
(n = 116)
(n = 303) (n = 625)
Engineering and Technical Sciences
Economic and Administrative Sciences
Math and Natural Sciences
Educational Sciences
Social Sciences
Medical and Health Sciences
Architecture and Urban Planning
Language and Literature
Arts
48.3
31.0
6.0
1.7
7.8
2.6
0.0
0.0
2.6
37.0
44.6
2.3
5.9
7.6
0.0
1.3
1.0
0.3
2
Test of Independence
45.9
21.4
15.8
6.1
5.9
1.8
1.4
1.0
0.6
Postdoc
(n = 56)
50.0
1.8
23.2
1.8
5.4
17.9
0.0
0.0
0.0
(24) = 188.22***
Notes: ***p < 0.001, **p < 0.005, *p < 0.010; Cell percentages sum to 100 across columns. There are
three missing responses.
238
Table A.17 Field of Study and Compulsory Academic Service Requirement (%)
No
Yes
Total
Field
(n = 904) (n = 191) (n = 1100)
Engineering and Technical Sciences
Economic and Administrative Sciences
Math and Natural Sciences
Social Sciences
Educational Sciences
Medical and Health Sciences
Architecture and Urban Planning
Language and Literature
Arts
45.9
30.9
11.2
6.5
1.8
1.9
0.6
0.6
0.8
Test of independence
2
34.6
13.1
12.6
6.8
22.5
3.7
4.2
2.1
0.5
43.9
27.8
11.4
6.6
5.4
2.2
1.2
0.8
0.7
(8) = 173.32***
Notes: ***p < 0.001, **p < 0.005, *p < 0.010; Cell percentages sum to 100 across columns; There are
three missing responses.
Table A.18 Work Destinations after Completion of Studies
Work Destination
n
%
703
667
33
3
67.0
63.6
3.2
0.3
57
14
13
12
6
3
2
1
1
1
1
1
1
1
5.5
1.3
1.2
1.1
0.6
0.3
0.2
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Australia
7
0.7
Japan
1
0.1
263
25.1
18
1.7
1049
100.0
North America
USA
Canada
North America, unspecified
Europe
Europe, unspecified
Germany
Great Britain
France
Belgium
Spain
Austria
Switzerland
Denmark
Finland
Italy
Nederlands
Portugal
Turkey
Do not know
Total
Note: There are 54 missing responses.
239
Table A.19 Intended Organization Immediately after Completing Studies by Work Destination
USA
Turkey
Europe
Organization
n
%
n
%
n
%
University – private
University – public
College / Tech. Inst. - private
College / Tech. Inst. - public
Pre / Primary / Secondary School - private
Pre / Primary / Secondary School - public
Government Department
Government Owned Corporation
Multinational Corporation
Other Private Sector Organization
Self-Employed in Incorp. Business / Practice / Farm
Self-Employed in Non-Inc. Business / Practice / Farm
International Organization
Non-profit Organization
Armed Forces
Not Sure
154
91
5
4
1
4
4
2
132
170
13
2
23
5
0
55
23.2
13.7
0.8
0.6
0.2
0.6
0.6
0.3
19.8
25.6
2.0
0.3
3.5
0.8
0.0
8.3
Total
665 100.0
30
161
0
0
0
0
15
2
17
14
3
2
1
1
1
13
11.5
61.9
0.0
0.0
0.0
0.0
5.8
0.8
6.5
5.4
1.2
0.8
0.4
0.4
0.4
5.0
260 100.0
9
13
1
0
1
0
0
0
17
11
0
0
2
0
0
2
16.1
23.2
1.8
0.0
1.8
0.0
0.0
0.0
30.4
19.6
0.0
0.0
3.6
0.0
0.0
3.6
56 100.0
Table A.19 continued
Canada
n
%
Organization
University – private
University – public
College / Tech. Inst. - private
College / Tech. Inst. - public
Pre / Primary / Secondary School - private
Pre / Primary / Secondary School - public
Government Department
Government Owned Corporation
Multinational Corporation
Other Private Sector Organization
Self-Employed in Incorp. Business / Practice / Farm
Self-Employed in Non-Inc. Business / Practice / Farm
International Organization
Non-profit Organization
Armed Forces
Not Sure
5
4
2
1
0
0
0
0
8
8
0
0
1
1
0
2
Total
32 100.0
Note: There are 61 missing responses.
240
15.6
12.5
6.3
3.1
0.0
0.0
0.0
0.0
25.0
25.0
0.0
0.0
3.1
3.1
0.0
6.3
Other / Not
Known
n
%
8
7
0
0
0
1
0
1
2
2
0
0
2
0
0
6
27.6
24.1
0.0
0.0
0.0
3.4
0.0
3.4
6.9
6.9
0.0
0.0
6.9
0.0
0.0
20.7
Total
n
%
206
276
8
5
2
5
19
5
176
205
16
4
29
7
1
78
19.8
26.5
0.8
0.5
0.2
0.5
1.8
0.5
16.9
19.7
1.5
0.4
2.8
0.7
0.1
7.5
29 100.0 1042 100.0
Table A.20 Intended Organization Five Years after Completing Studies by Work Destination
USA
Organization
n
%
University – private
University – public
College / Tech. Inst. - private
College / Tech. Inst. - public
Pre / Primary / Secondary School - private
Government Department
Government Owned Corporation
Multinational Corporation
Other Private Sector Organization
Self-Employed in Inc. Business / Practice / Farm
Self-Employed in Non-Inc. Business / Practice / Farm
International Organization
Non-profit Organization
Armed Forces
Not Sure
164
95
6
1
1
6
3
102
85
42
12
28
7
0
100
25.2
14.6
0.9
0.2
0.2
0.9
0.5
15.6
13.0
6.4
1.8
4.3
1.1
0.0
15.3
Total
652 100.0
Turkey
n
%
52
118
0
0
0
12
1
12
11
9
5
6
0
1
27
20.5
46.5
0.0
0.0
0.0
4.7
0.4
4.7
4.3
3.5
2.0
2.4
0.0
0.4
10.6
254 100.0
Europe
n
%
10
7
0
1
0
0
0
16
9
3
4
2
0
0
4
17.9
12.5
0.0
1.8
0.0
0.0
0.0
28.6
16.1
5.4
7.1
3.6
0.0
0.0
7.1
56 100.0
Table A.20 continued
Canada
n
%
Organization
University - private
University - public
College / Tech. Inst. - private
College / Tech. Inst. - public
Pre / Primary / Secondary School - private
Government Department
Government Owned Corporation
Multinational Corporation
Other Private Sector Organization
Self-Employed in Inc. Business / Practice / Farm
Self-Employed in Non-Inc. Business / Practice / Farm
International Organization
Non-profit Organization
Armed Forces
Not Sure
9
3
0
0
0
0
0
7
7
1
1
1
1
0
3
Total
33 100.0
Note: There are 79 missing responses.
241
27.3
9.1
0.0
0.0
0.0
0.0
0.0
21.2
21.2
3.0
3.0
3.0
3.0
0.0
9.1
Other / Not
Known
n
%
10
5
0
0
0
0
1
1
4
1
0
0
0
0
7
34.5
17.2
0.0
0.0
0.0
0.0
3.4
3.4
13.8
3.4
0.0
0.0
0.0
0.0
24.1
Total
n
%
245
228
6
2
1
18
5
138
116
56
22
37
8
1
141
23.9
22.3
0.6
0.2
0.1
1.8
0.5
13.5
11.3
5.5
2.1
3.6
0.8
0.1
13.8
29 100.0 1024 100.0
Table A.21 Respondents by Standard Occupation Classification,
Broad Groups
SOC
Occupations
Code
Management
Business and Financial Operations
Computer and Mathematical Science
Architecture and Engineering
Life, Physical and Social Science
Community and Social Services
Legal
Education, Training and Library
Arts, Design, Entertainment, Sports and Media
Healthcare Practitioner and Technical
Healthcare Support
Food Preparation and Service Related
Personal Care and Service
Sales and Related
Office and Administrative Support
Installation, Maintenance and Repair Occupations
Total
242
11
13
15
17
19
21
23
25
27
29
31
35
39
41
43
49
n
%
253
87
255
234
83
1
2
263
9
19
1
2
1
9
4
1
20.7
7.1
20.8
19.1
6.8
0.1
0.2
21.5
0.7
1.6
0.1
0.2
0.1
0.7
0.3
0.1
1224
100.0
Table A.22 Respondents by Detailed Occupation Categories, SOC classification
Occupation
Engineering Teachers, Postsecondary
Computer Software Engineers
General and Operations Managers
Computer Programmers
Engineering Managers
Operations Research Analysts
Business Teachers, Postsecondary
Sales and Marketing Managers
Electronics Engineers, Except Computer
Financial Analysts
Computer Specialists, unclassified
Mechanical Engineers
Economics Teachers, Postsecondary
Management Analysts
Electrical Engineers
Civil Engineers
Computer and Information Systems Managers
Private Sector Executives
Economists
Financial Managers
Computer Hardware Engineers
Chemical Engineers
Industrial Engineers
Medical Scientists, Except Epidemiologists
Architects, Except Landscape and Naval
Health Specialties Teachers, Postsecondary
Aerospace Engineers
Network and Computer Systems Administrators
Computer Science Teachers, Postsecondary
Industrial Production Managers
Computer Systems Analysts
Mathematical Science Teachers, Postsecondary
Physics Teachers, Postsecondary
Physicists
Biological Science Teachers, Postsecondary
Construction Managers
Physicians and Surgeons, All Other
Natural Sciences Managers
Business and Financial Operations Managers, unclassified
Computer and Information Scientists, Research
Materials Engineers
Petroleum Engineers
Chemists
Architecture Teachers, Postsecondary
Medical and Health Services Managers
Personal Financial Advisors
Nuclear Engineers
Market Research Analysts
Political Science Teachers, Postsecondary
Secondary School Teachers, Except Special and Vocational
243
SOC code
25-1032.00
15-1030.00
11-1021.00
15-1021.00
11-9041.00
15-2031.00
25-1011.00
11-2020.00
17-2072.00
13-2051.00
15-1000.00
17-2141.00
25-1063.00
13-1111.00
17-2071.00
17-2051.00
11-3021.00
11-1011.02
19-3011.00
11-3031.00
17-2061.00
17-2041.00
17-2112.00
19-1042.00
17-1011.00
25-1071.00
17-2011.00
15-1071.00
25-1021.00
11-3051.00
15-1051.00
25-1022.00
25-1054.00
19-2012.00
25-1042.00
11-9021.00
29-1069.99
11-9121.00
13-0000.00
15-1011.00
17-2131.00
17-2171.00
19-2031.00
25-1031.00
11-9111.00
13-2052.00
17-2161.00
19-3021.00
25-1065.00
25-2031.00
n
95
84
58
48
47
45
43
40
38
34
34
33
33
30
29
25
22
21
20
19
19
16
16
15
14
14
13
12
11
10
10
10
9
8
8
7
7
6
6
6
6
6
6
6
5
5
5
5
5
5
%
7.8
6.9
4.7
3.9
3.8
3.7
3.5
3.3
3.1
2.8
2.8
2.7
2.7
2.5
2.4
2.0
1.8
1.7
1.6
1.6
1.6
1.3
1.3
1.2
1.1
1.1
1.1
1.0
0.9
0.8
0.8
0.8
0.7
0.7
0.7
0.6
0.6
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.4
0.4
0.4
0.4
0.4
0.4
Table A.22 continued
Occupation
Computer Software Engineers, Applications
Statisticians
Environmental Engineers
Political Scientists
Securities, Commodities, and Financial Services Sales Agents
Administrative Services Managers
Education Administrators, Postsecondary
Computer Software Engineers, Systems Software
Mining and Geological Engineers, Including Mining Safety
Food Scientists and Technologists
Biologists
Clinical, Counseling, and School Psychologists
Social Scientists and Related Workers, All Other
Area, Ethnic, and Cultural Studies Teachers, Postsecondary
Psychology Teachers, Postsecondary
Graphic Designers
Obstetricians and Gynecologists
Pediatricians, General
Sales Engineers
Advertising and Promotions Managers
Treasurers, Controllers, and Chief Financial Officers
Purchasing Managers
Transportation, Storage, and Distribution Managers
Credit Analysts
Computer Support Specialists
Actuaries
Biomedical Engineers
Biochemists and Biophysicists
Materials Scientists
Urban and Regional Planners
Biological Technicians
Lawyers
English Language and Literature Teachers, Postsecondary
Foreign Language and Literature Teachers, Postsecondary
History Teachers, Postsecondary
Philosophy and Religion Teachers, Postsecondary
Anesthesiologists
Psychiatrists
Eligibility Interviewers, Government Programs
Government Service Executives
Financial Managers, Branch or Department
Human Resources Managers
Lodging Managers
Wholesale and Retail Buyers, Except Farm Products
Purchasing Agents, Except Wholesale, Retail, and Farm Products
Compliance Officers, Except Agriculture, Construction, Health
Business Operations Specialists, All Other
Accountants
Auditors
Budget Analysts
Insurance Underwriters
Loan Officers
244
SOC code
15-1031.00
15-2041.00
17-2081.00
19-3094.00
41-3031.00
11-3011.00
11-9033.00
15-1032.00
17-2151.00
19-1012.00
19-1020.01
19-3031.00
19-3099.99
25-1062.00
25-1066.00
27-1024.00
29-1064.00
29-1065.00
41-9031.00
11-2011.00
11-3031.01
11-3061.00
11-3071.00
13-2041.00
15-1041.00
15-2011.00
17-2031.00
19-1021.00
19-2032.00
19-3051.00
19-4021.00
23-1011.00
25-1123.00
25-1124.00
25-1125.00
25-1126.00
29-1061.00
29-1066.00
43-6011.00
11-1011.01
11-3031.02
11-3040.00
11-9081.00
13-1022.00
13-1023.00
13-1041.00
13-1199.99
13-2011.01
13-2011.02
13-2031.00
13-2053.00
13-2072.00
n
4
4
4
4
4
3
3
3
3
3
3
3
3
3
3
3
3
3
3
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
1
1
1
1
1
1
1
1
1
1
1
1
1
%
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.3
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Table A.22 continued
Occupation
Financial Specialists, All Other
Network Systems and Data Communications Analysts
Landscape Architects
Surveyors
Electrical and Electronic Engineering Technicians
Industrial Engineering Technicians
Mechanical Engineering Technicians
Foresters
Environmental Scientists and Specialists, Including Health
Geoscientists, Except Hydrologists and Geographers
Geologists
Life, Physical, and Social Science Technicians, All Other
Social and Human Service Assistants
Chemistry Teachers, Postsecondary
Anthropology and Archeology Teachers, Postsecondary
Education Teachers, Postsecondary
Law Teachers, Postsecondary
Social Work Teachers, Postsecondary
Graduate Teaching Assistants
Archivists, Curators, and Museum Technicians
Museum Technicians and Conservators
Librarians
Library Technicians
Fine Artists, Including Painters, Sculptors, and Illustrators
Commercial and Industrial Designers
Fashion Designers
Exhibit Designers
Music Directors and Composers
News Analysts, Reporters and Correspondents
Internists, General
Surgeons
Psychiatric Aides
Cooks, Institution and Cafeteria
Cooks, Restaurant
Flight Attendants
Sales and Related Occupation, unclassified
Sales Representatives, Mechanical Equipment and Supplies
Customer Service Representatives
Desktop Publishers
Automotive Service Technicians and Mechanics
Total
245
SOC code
13-2099.99
15-1081.00
17-1012.00
17-1022.00
17-3023.00
17-3026.00
17-3027.00
19-1032.00
19-2041.00
19-2042.00
19-2042.01
19-4099.99
21-1093.00
25-1052.00
25-1061.00
25-1081.00
25-1112.00
25-1113.00
25-1191.00
25-4010.00
25-4013.00
25-4021.00
25-4031.00
27-1013.01
27-1021.00
27-1022.00
27-1027.02
27-2041.00
27-3020.00
29-1063.00
29-1067.00
31-1013.00
35-2012.00
35-2014.00
39-6031.00
41-0000.00
41-4011.05
43-4051.00
43-9031.00
49-3023.00
n
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
%
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
1224
100.0
Table A.23 Percentage of Time Spent on R&D Activities by Occupation (valid n = 1186)
RD1
<20%
RD2
20-40%
RD3
40-60%
RD4
60-80%
RD5
80-100%
Total
Managerial
Business / Finance
Computer & Math
Arch / Engineering
Social & Life Sciences
Education
Other
48.0
62.5
49.2
27.0
16.5
8.9
64.4
26.2
15.0
16.8
13.5
11.4
19.0
24.4
16.0
11.3
17.2
11.3
10.1
42.6
6.7
5.3
5.0
5.6
20.4
16.5
21.7
0.0
4.5
6.3
11.2
27.8
45.6
7.8
4.4
244
80
250
230
79
258
45
Total
35.2
18.4
12.4
14.0
1,186
Occupation Group
Notes: Cell percentages sum to 100 across rows;
significance at the 1 percent level.
20.1
2
(24) = 397.26
***
where *** indicates
Table A.24 Return Intentions and R & D Intensity of Job Activities (%) (valid n = 1186)
Return Intentions
Definitely return, plans
Definitely return, no plans
Return probable
Return unlikely
Definitely not return
RD1
<20%
(n = 417)
RD2
20-40%
(n = 218)
RD3
40-60%
(n = 238)
4.6
24.7
35.3
27.8
7.7
5.1
19.7
32.1
36.2
6.9
3.8
16.4
34.9
38.7
6.3
4.1
21.1
30.6
39.5
4.8
4.8
28.3
36.8
25.9
4.2
4.5
22.2
34.2
32.7
6.4
100.0
100.0
100.0
100.0
100.0
100.0
Notes: Cell percentages sum to 100 across columns;
at the 10 percent level.
246
2
RD4
RD5
Total
60-80% 80-100%
(n = 147) (n = 166) (n = 1186)
(16) = 23.95* where * indicates significance
Table A.25 Full-Time Jobs in Turkey (#)
Number of jobs
None
One
Two
Three or more
Total
n
%
384
417
228
195
31.4
34.1
18.6
15.9
1224
100.0
Table A.26 Full-Time Jobs Abroad (#)
Number of jobs
One
Two
Three or more
Total
n
%
520
357
334
42.9
29.5
27.6
1211
100.0
Table A.27 Number of Years Worked Abroad
Years
1_2
3_4
5_6
7_8
9_10
11_12
13_14
15_16
17_18
19_20
21_22
23_24
25_26
27_28
29_30
31or more
Total
n
%
Cum.
344
219
173
104
57
49
36
44
36
34
43
18
18
10
10
18
28.4
18.1
14.3
8.6
4.7
4.0
3.0
3.6
3.0
2.8
3.5
1.5
1.5
0.8
0.8
1.5
28.4
46.4
60.7
69.3
74.0
78.0
81.0
84.6
87.6
90.4
93.9
95.4
96.9
97.7
98.5
100.0
1213
100.0
247
Table A.28 Sector of Current Organization
Sector
Private
Public
Non-profit / other
Total
n
%
520
357
334
42.9
29.5
27.6
1211
100.0
Table A.29 Type of Organization
n
%
368
267
177
162
49
41
39
37
28
27
13
7
6
2
1
30.1
21.8
14.5
13.2
4.0
3.4
3.2
3.0
2.3
2.2
1.1
0.6
0.5
0.2
0.1
1224
100.0
Organization
Multinational Corporation - Headquarters in Current Country
University
Multinational Corporation - Headquarters in Third Country
Other Incorporated Firm
Non-incorporated firm or business
Research Center at a University
Other
Hospital / Medical Center
International Organization (IMF, ILO, World Bank, etc.)
National Government
Multinational Corporation - Headquarters in Turkey
Non-govermental organizaiton
Local Government
Secondary School
College / Tech. Institute
Total
Table A.30 Location Where Current Job was Found
Location
Current country of residence
Turkey
A Third Country
Total
248
n
%
520
357
334
42.9
29.5
27.6
1211
100.0
APPENDIX B
Table B.1 Associations of Explanatory Variables with Return Intentions (y), Professionals
valid
Code
VARIABLE DEFINITIONS
n
chisq
df
Pr
Sig.
gamma
female
initial_int
Respondent is female
Initial return intentions
1224
1224
13.39
232.17
4
8
0.010 ***
0.000 ***
0.1046
0.5018
spouse_nat
spousenat1
spousenat2
spousenat3
Nationality of spouse
Spouse's nationality: Turkish
Spouse's nationality: Foreign
Spouse's nationality: Dual
Citizen
Never married
Divorced/Separated/Widowed
1178
1178
1178
122.70
11.11
86.67
16
4
4
0.000 ***
0.025 **
0.000 ***
-0.1390
0.5384
1178
1178
1178
12.69
40.48
5.11
4
4
4
0.013 **
0.000 ***
0.276 ns
0.1983
-0.2739
1176
26.32
12
0.010 ***
-0.1007
1160
164.11
20
0.000 ***
0.3405
1212
56.22
20
0.000 ***
0.2067
1218
129.83
20
0.000 ***
0.3183
1217
93.82
20
0.000 ***
0.3320
1213
1213
1213
8.87
4.12
2.21
12
4
4
0.714 ns
0.390 ns
0.697 ns
1213
3.00
4
0.558 ns
1213
2.78
4
0.595 ns
1218
1218
1218
11.40
8.08
5.17
12
4
4
0.495 ns
0.089 +
0.271 ns
1218
2.76
4
0.598 ns
1218
0.70
4
0.951 ns
1221
90.26
12
0.000 ***
1221
1221
24.60
3.64
4
4
0.000 ***
0.456 ns
0.1859
1221
75.01
4
0.000 ***
-0.5873
1221
2.57
4
0.632 ns
1224
23.98
8
0.002 ***
spousenat4
spousenat5
fam_sup1
fam_sup2
work_assess
social_assess
SOL_assess
FTr_type
FTr1
FTr2
FTr3
FTr4
OTJT_type
OTJTtype1
OTJTtype2
OTJTtype3
OTJTtype4
lastvis
lastvis1
lastvis2
lastvis3
lastvis4
HD2
Family support for initial
decision to go abroad
Family support for permanent
settlement
Assessment of work
conditions
Assessment of social
conditions
Assessment of standard of
living
Skills transferability of tormal
training (4-point)
Formal training: none
Formal training: general
Formal training: specific to
industry
Formal training: specific to
organization
Skills transferability of onthe-job training (4-point)
On-the-job training: none
On-the-job training: general
On-the-job training: specific
to industry
On-the-job training: specific
to organization
Effect of last visit to Turkey
on returning (4-point)
Last visit effect: Decreased
return intentions
Last visit effect: No effect
Last visit effect: Increased
return intentions
Last visit effect: Not
Applicable
Highest degree held by
respondent (3-point)
249
0.0883
Table B.1 continued
Code
VARIABLE DEFINITIONS
valid
n
chisq
df
bachelors
masters
doctorate
Highest degree: bachelors
Highest degree: masters
Highest degree: doctorate
1224
1224
1224
7.35
7.19
22.20
HD_TUR
Highest degree is from
Turkey
Highest degree: doctorate
from foreign university
Location of first full-time job
for foreign degree holders
Same city and country where
degree is conferred
Same country, different city
Turkey
A different country
Last degree held is not a
foreign degree
Turkish instruction, high
school science courses
Turkish instruction, high
school social science courses
TYPE OF ORGANIZATION
Academic / Medical School /
Research Center
Government / International
Org / NGO / Other
Private Organization
1224
ForHD_PHD
FFTJ_where
FFTJloc1
FFTJloc2
FFTJloc3
FFTJloc4
FFTJloc5
HSsci_TUR
HSsoc_TUR
orgtype
academic2
publicserv
privateorg
JOBACTV1
JOBACTV2
JOBACTV3
JOBACTV4
JOBACTV5
JOBACTV6
JOBACTV7
JOBACTV8
JOBACTV9
JOBACTV10
JOBACTV11
JOBRandD
DOM_ACTV1
DOM_ACTV2
DOM_ACTV3
Teaching
Applied research
Basic research
Development
Computer use, programming,
system development
Administrative, supervisory
activities
Professional services
Quality control, production
management
Accounting, contracts
Marketing, consumer
services, public relations
Other activities not defined
above
Research and Development
(2+3+4)
Teaching
Applied research
Basic research
Sig.
gamma
4
4
4
0.118 ns
0.126 ns
0.000 ***
0.1890
10.66
4
0.031 *
-0.1311
1224
29.70
4
0.000 ***
0.2411
1219
47.33
20
0.001 ***
0.0427
1224
1224
1224
1224
0.98
11.93
26.02
0.34
4
4
4
4
0.913
0.018
0.000
0.987
0.0138
0.2847
1224
10.66
4
0.031 *
-0.1311
1224
12.85
4
0.012 **
0.1116
1224
7.73
4
0.102 ns
1224
15.23
4
0.004 ***
0.1466
1224
1224
4.74
20.88
4
4
0.315 ns
0.000 ***
0.1519
-0.1793
1186
1186
1186
1186
10.98
1.50
2.23
5.08
4
4
4
4
0.027
0.827
0.693
0.279
1186
4.16
4
0.384 ns
1186
1186
4.03
0.84
4
4
0.401 ns
0.933 ns
1186
1186
2.02
1.54
4
4
0.732 ns
0.820 ns
1186
1.50
4
0.827 ns
1186
1.70
4
0.791 ns
1186
1.94
4
0.746 ns
1186
1186
1186
5.47
1.15
2.35
4
4
4
0.242 ns
0.886 ns
0.672 ns
250
Pr
ns
**
***
ns
*
ns
ns
ns
Table B.1 continued
Code
VARIABLE DEFINITIONS
DOM_ACTV4
DOM_ACTV5
Development
Computer use, programming,
system development
Administrative, supervisory
activities
Professional services
Quality control, production
management
Accounting, contracts
Marketing, consumer
services, public relations
Other activities not defined
above
Research and Development
(2+3+4)
Teaching
Applied research
Basic research
Development
Computer use, programming,
system development
Administrative, supervisory
activities
Professional services
Quality control, production
management
Accounting, contracts
Marketing, consumer
services, public relations
Other activities not defined
above
Research and Development
(2+3+4)
Respondent has no full time
work exp. in Turkey
DOM_ACTV6
DOM_ACTV7
DOM_ACTV8
DOM_ACTV9
DOM_ACTV10
DOM_ACTV11
DOM_RandD2
ACTV1
ACTV2
ACTV3
ACTV4
ACTV5
ACTV6
ACTV7
ACTV8
ACTV9
ACTV10
ACTV11
RDintensity
NWexpTUR
contrA
contrB
contrC
contrD
contrE
contrF
contrG
contrH
Overseas scholarships for
Turkish students
Lobbying actitivies on behalf of
Turkey
Increased overseas business
contacts with Turkey
Increased knowledge about
Turkey in general
Donations to Turkish
organizations
Increased professional contacts
betw. Turkey and cc.
Helped in the transfer of
knowledge
Other positive contribution
valid
n
chisq
df
1186
1.83
4
0.768 ns
1186
4.56
4
0.336 ns
1186
1186
2.64
1.18
4
4
0.620 ns
0.882 ns
1186
1186
2.75
2.65
4
4
0.600 ns
0.618 ns
1186
1.44
4
0.837 ns
1186
2.30
4
0.681 ns
1186
2.88
4
0.578 ns
1186
1186
1186
1186
37.07
11.04
14.16
11.64
16
16
16
16
0.002
0.807
0.587
0.768
1186
14.12
16
0.590 ns
1186
1186
10.72
15.29
16
16
0.826 ns
0.504 ns
1186
1186
13.12
16.40
16
16
0.664 ns
0.425 ns
1186
10.67
16
0.829 ns
1186
11.14
16
0.801 ns
1186
23.95
16
0.091 +
0.0037
1224
13.54
4
0.009 ***
0.0682
1099
1.36
4
0.852 ns
0.0209
1099
5.72
4
0.221 ns
-0.0986
1099
5.04
4
0.283 ns
-0.0898
1099
8.64
4
0.071 +
-0.1393
1099
7.62
4
0.106 ns
0.0113
1099
7.71
4
0.103 ns
-0.0574
1099
1099
20.68
4.25
4
4
0.000 ***
0.373 ns
-0.1585
0.0579
251
Pr
Sig.
***
ns
ns
ns
gamma
0.2132
Table B.2 Summary Statistics and Descriptions of the Variables used in the Final Model,
Professionals (n = 1031)
Std
Variable
Variable Descriptions
Mean
Min
Dev.
y
Dependent variable: return intentions
(1=definite return plans; 2=definite return,
no immediate plans; 3=return probable;
4=return unlikely; 5=definitely not return)
0.97
1
5
female
init_UNSURE
init_RETURN
age
agesq
staydur
Gender of respondent (1=female)
0.28
0.45
Initial return intentions: Unsure (1=yes)
0.36
0.48
Initial return intentions: Return (1=yes)
0.53
0.50
Age of respondent in 2001
35.04
8.90
Square of Age
1307.99 722.14
Stay duration in current country of residence
12.78
6.89
(years)
Work experience in current country (years)
6.84
6.88
Married to a foreign spouse (1=yes)
0.15
0.36
Respondent has no work experience in
0.32
0.47
Turkey (1=yes)
Country of work after completing studies
0.09
0.29
abroad is Turkey (1=yes)
Respondent's highest degree is a PhD from a
0.04
0.20
Turkish university (1=yes)
Assessment of social conditions abroad
2.63
1.00
Assessment of standard of living abroad
4.48
0.81
Family support for initial decision to go
3.48
0.75
abroad
Family support for settling abroad
4.39
1.51
Type of organization: Academic / Research
0.27
0.44
Center / Medical School
Job requirement in Turkey
0.22
0.42
Insufficient facilities, equipment for research
0.27
0.44
Prestige and advantages of study abroad
0.46
0.50
Lifestyle preference
0.33
0.47
To be with spouse, family
0.12
0.33
Get away from political environment
0.32
0.47
Limited job opport. in specialty
0.54
0.50
No opportunity for advanced training
0.37
0.48
Lack of financial resources for business
0.30
0.46
Economic instability
0.85
0.35
Greater oppr. to develop specialty
0.71
0.45
More organized, ordered envir.
0.77
0.42
More satisfying social/cultural life
0.26
0.44
Proximity to research and innov. centers
0.42
0.49
Spouse’s preference or job
0.31
0.46
Better educational opport. For children
0.37
0.48
Need to finish /continue with current project
0.16
0.36
0
0
0
22
484
1
1
1
1
72
5184
32
1
0
0
31
1
1
0
1
0
1
0
0
1
5
5
4
1
0
6
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
yrs_wrkd_cc
spousenat2
NWexpTUR
FFTJloc3
HDTURXPHD
social_assess
SOL_assess
fam_sup1
fam_sup2
academic2
whygo_C
whygo_F
whygo_G
whygo_H
whygo_I
whygo_K
pushC
pushD
pushF
pushK
pullE
pullF
pullG
pullH
pullI
pullJ
pullK
252
3.15
Max
Table B.2 continued.
Variable
Variable Description
pullL
Hdnew2
Hdnew3
adj_A
adj_C
difabrdA
contrB2
FTr4
lastvis1
lastvis3
sept11_inc
Other
Field of Highest Degree:
Education/Languages/Social Sciences/Arts
Field of Highest Degree:
Engineering/Math/Science/Medicine
Adjustment factor: previous experience
Adjustment factor: support from TSA
(Turkish Student Association)
Difficulties abroad: being away from family
Contribution to Turkey: Lobbying actitivies on
behalf of Turkey
Formal training received abroad is specific to
organization (1=yes)
Last visit to Turkey decreased return
intentions (1=yes)
Last visit to Turkey increased return
intentions (1=yes)
Effect of September 11, 2001 (1=increased
return intentions)
253
Mean
Std
Dev.
Min
Max
0.05
0.04
0.21
0.20
0
0
1
1
0.66
0.47
0
1
0.43
0.05
0.50
0.21
0
0
1
1
0.83
0.60
0.38
0.49
0
0
1
1
0.04
0.19
0
1
0.28
0.45
0
1
0.09
0.29
0
1
0.10
0.30
0
1
Table B.3a Estimation Results and Marginal Effects for Outcomes y = 1 and y = 2,
Ordered Probit Model, Professionals
y = DRP = 1
y = DRNP = 2
(a) z-value
dy/dx
z-value
dy/dx z-value
female (b)
init_UNSURE (b)
init_RETURN (b)
age
agesq
staydur
yrs_wrkd_cc
AGExSTAYDUR
AGESQxSTAYDUR
spousnat2 (b)
NWexpTUR (b)
FFTJloc3 (b)
HDTURXPHD (b)
social_assess
SOL_assess
fam_sup1
fam_sup2
academic2 (b)
whygo_C (b)
whygo_F (b)
whygo_G (b)
whygo_H (b)
whygo_I (b)
whygo_K (b)
FxWHYGOC (b)
FxWHYGOI (b)
ACADxWHYGOG (b)
AGExWHYGOF
AGExWHYGOG
pushC (b)
pushD (b)
pushF (b)
pushK (b)
pullE (b)
pullF (b)
pullG (b)
pullH (b)
pullI (b)
pullJ (b)
pullK (b)
pullL (b)
femalexpushC (b)
femalexpullI (b)
femalexpullK (b)
femalexpullL (b)
0.355
-0.950
-1.323
0.085
-0.001
0.327
0.051
-0.012
0.000
0.403
0.213
0.475
-0.477
0.101
0.129
-0.176
0.154
0.078
-0.190
1.536
-0.666
0.178
-0.454
0.144
0.347
0.396
-0.465
-0.042
0.021
-0.070
-0.966
-0.132
0.368
0.263
0.164
0.275
-0.215
0.357
0.317
-0.618
-0.460
-0.257
-0.469
0.380
0.813
(2.40)**
(6.65)***
(8.87)***
(1.11)
(0.54)
(3.40)***
(3.23)***
(2.77)***
(2.05)**
(3.43)***
(2.45)**
(3.18)***
(2.31)**
(2.43)**
(2.80)***
(2.82)***
(5.46)***
(0.39)
(1.92)*
(4.22)***
(1.69)*
(2.14)**
(2.95)***
(1.69)*
(1.69)*
(1.73)*
(2.49)**
(4.14)***
(1.74)*
(0.69)
(2.96)***
(1.65)*
(3.38)***
(2.59)***
(1.76)*
(3.05)***
(2.10)**
(3.58)***
(3.67)***
(4.99)***
(2.12)**
(1.61)
(2.73)***
(1.58)
(1.99)**
-0.0031
0.0172
0.0186
-0.0009
0.0000
-0.0034
-0.0005
0.0001
0.0000
-0.0030
-0.0020
-0.0031
0.0093
-0.0011
-0.0014
0.0019
-0.0016
-0.0008
0.0023
-0.0111
0.0085
-0.0017
0.0078
-0.0014
-0.0025
-0.0027
0.0082
0.0004
-0.0002
0.0007
0.0174
0.0015
-0.0056
-0.0033
-0.0020
-0.0025
0.0024
-0.0033
-0.0031
0.0122
0.0087
0.0035
0.0084
-0.0026
-0.0034
254
(-2.24)**
(3.06)***
(3.56)***
(-1.08)
(0.54)
(-2.58)***
(-2.39)**
(2.28)**
(-1.85)*
(-2.94)***
(-2.12)**
(-2.81)***
(1.37)
(-2.09)**
(-2.21)**
(2.21)**
(-3.11)***
(-0.41)
(1.55)
(-2.74)***
(1.25)
(-1.97)**
(1.74)*
(-1.61)
(-2.15)**
(-2.16)**
(1.53)
(2.8)***
(-1.57)
(0.69)
(1.63)
(1.44)
(-2.12)**
(-1.91)*
(-1.47)
(-2.48)**
(1.75)*
(-2.61)***
(-2.66)***
(2.5)**
(1.25)
(1.19)
(1.68)*
(-2.21)**
(-2.99)***
-0.0773
0.2433
0.2930
-0.0199
0.0001
-0.0767
-0.0120
0.0029
0.0000
-0.0824
-0.0482
-0.0918
0.1320
-0.0237
-0.0304
0.0413
-0.0362
-0.0179
0.0466
-0.2538
0.1595
-0.0407
0.1217
-0.0331
-0.0700
-0.0782
0.1253
0.0098
-0.0050
0.0164
0.2466
0.0318
-0.0961
-0.0648
-0.0399
-0.0605
0.0512
-0.0787
-0.0716
0.1694
0.1264
0.0650
0.1267
-0.0750
-0.1244
(-2.57)***
(6.43)***
(9.43)***
(-1.11)
(0.54)
(-3.36)***
(-3.19)***
(2.74)***
(-2.04)**
(-3.95)***
(-2.52)**
(-3.94)***
(2.06)**
(-2.42)**
(-2.78)***
(2.79)***
(-5.28)***
(-0.4)
(1.83)*
(-5.8)***
(1.67)*
(-2.2)**
(2.64)***
(-1.74)*
(-2.04)**
(-2.11)**
(2.24)**
(4.06)***
(-1.74)*
(0.69)
(2.83)***
(1.62)
(-3.08)***
(-2.46)**
(-1.69)*
(-3.25)***
(2.06)**
(-3.8)***
(-3.74)***
(4.44)***
(1.89)*
(1.5)
(2.45)**
(-1.95)*
(-3.65)***
Table B.3a continued
y = DRP = 1
(a)
ACADxpushC (b)
ACADxpullE (b)
ACADxpullH (b)
AGExpushD
HDnew2 (b)
HDnew3 (b)
adj_A (b)
adj_C (b)
difabrdA(b)
contrB2 (b)
FTr4 (b)
lastvis1 (b)
lastvis3 (b)
sept11_inc (b)
0.387
-0.292
0.493
0.030
0.544
0.270
-0.268
-0.248
-0.217
-0.390
0.366
0.154
-0.716
-0.262
z-statistic
dy/dx
(2.24)**
(1.36)
(2.40)**
(3.14)***
(3.03)***
(3.29)***
(3.58)***
(1.51)
(2.21)**
(4.99)***
(1.90)*
(1.87)*
(5.64)***
(2.06)**
-0.0029
0.0039
-0.0036
-0.0003
-0.0031
-0.0033
0.0030
0.0036
0.0019
0.0039
-0.0025
-0.0015
0.0175
0.0037
z-value
(-2.41)**
(1.02)
(-2.36)**
(-2.35)**
(-2.91)***
(-2.3)**
(2.45)**
(1.12)
(2.08)**
(2.97)***
(-2.35)**
(-1.74)*
(2.71)***
(1.39)
y = DRNP = 2
dy/dx
-0.0791
0.0736
-0.0991
-0.0069
-0.0988
-0.0658
0.0640
0.0639
0.0475
0.0882
-0.0726
-0.0350
0.2044
0.0671
z-value
(-2.6)***
(1.28)
(-2.83)***
(-3.11)***
(-4.1)***
(-3.17)***
(3.49)***
(1.39)
(2.37)**
(5.13)***
(-2.31)**
(-1.92)*
(4.95)***
(1.91)*
Notes: * significant at 10%; ** significant at 5%; *** significant at 1%;
(a) Robust z-statistics in parentheses; Observations = 1031; Log-likelihood = -1028.82;
LR chi2(59)= 651.57; Maximum Likelihood R2 = 0.527; McFadden’s Adjusted R2 =
0.228; McKelvey-Zavoina R2 = 0.583; AIC = 2.118; BIC= -4658.626.
(b) dy/dx is for discrete change of dummy variable from 0 to 1.
255
Table B.3b Estimation Results and Marginal Effects for Outcomes y = 3, 4 and 5,
Ordered Probit Model, Professionals
Explanatory
y = RP = 3
y = RU = 4
y = DNR = 5
Variables
dy/dx
z-value
dy/dx
z-value
dy/dx
z-value
female (b)
init_UNSURE (b)
init_RETURN (b)
age
agesq
staydur
yrs_wrkd_cc
AGExSTAYDUR
AGESQxSTAYDUR
spousnat2 (b)
NWexpTUR (b)
FFTJloc3 (b)
HDTURXPHD (b)
social_assess
SOL_assess
fam_sup1
fam_sup2
academic2 (b)
whygo_C (b)
whygo_F (b)
whygo_G (b)
whygo_H (b)
whygo_I (b)
whygo_K (b)
FxWHYGOC (b)
FxWHYGOI (b)
ACADxWHYGOG (b)
AGExWHYGOF
AGExWHYGOG
pushC (b)
pushD (b)
pushF (b)
pushK (b)
pullE (b)
pullF (b)
pullG (b)
pullH (b)
pullI (b)
pullJ (b)
pullK (b)
pullL (b)
femalexpushC (b)
femalexpullI (b)
femalexpullK (b)
femalexpullL (b)
-0.0518
0.0532
0.1480
-0.0099
0.0001
-0.0382
-0.0060
0.0014
0.0000
-0.0674
-0.0279
-0.0874
0.0113
-0.0118
-0.0152
0.0206
-0.0181
-0.0096
0.0181
-0.2912
0.0670
-0.0229
0.0200
-0.0183
-0.0598
-0.0707
0.0189
0.0049
-0.0025
0.0083
0.0556
0.0140
-0.0228
-0.0246
-0.0162
-0.0390
0.0234
-0.0504
-0.0417
0.0159
0.0130
0.0204
0.0184
-0.0679
-0.1877
(-2)**
(3.56)***
(6.04)***
(-1.09)
(0.54)
(-3.01)***
(-2.92)***
(2.54)**
(-1.95)*
(-2.62)***
(-2.14)**
(-2.38)**
(0.68)
(-2.29)**
(-2.57)***
(2.61)***
(-4.25)***
(-0.37)
(2.27)**
(-4.18)***
(1.97)**
(-1.88)*
(2.24)**
(-1.52)
(-1.3)
(-1.33)
(1.83)*
(3.53)***
(-1.66)*
(0.68)
(3.63)***
(1.75)*
(-3.59)***
(-2.97)***
(-2.04)**
(-2.47)**
(2.13)**
(-2.87)***
(-3.09)***
(1.21)
(0.83)
(2.63)***
(1.76)*
(-1.21)
(-1.57)
256
0.1211
-0.2928
-0.4107
0.0286
-0.0002
0.1100
0.0172
-0.0042
0.0000
0.1383
0.0722
0.1630
-0.1451
0.0340
0.0436
-0.0593
0.0520
0.0263
-0.0627
0.4475
-0.2177
0.0604
-0.1416
0.0489
0.1195
0.1363
-0.1443
-0.0140
0.0071
-0.0237
-0.2979
-0.0442
0.1174
0.0867
0.0543
0.0937
-0.0718
0.1215
0.1073
-0.1873
-0.1407
-0.0835
-0.1454
0.1309
0.2632
(2.39)**
(-7.2)***
(-9.71)***
(1.11)
(-0.54)
(3.37)***
(3.2)***
(-2.75)***
(2.04)**
(3.4)***
(2.42)**
(3.24)***
(-2.66)***
(2.42)**
(2.79)***
(-2.82)***
(5.43)***
(0.39)
(-1.96)**
(6.98)***
(-1.77)*
(2.12)**
(-3.25)***
(1.68)*
(1.69)*
(1.75)*
(-2.76)***
(-4.08)***
(1.73)*
(-0.69)
(-3.37)***
(-1.67)*
(3.62)***
(2.66)***
(1.79)*
(3)***
(-2.12)**
(3.54)***
(3.65)***
(-5.7)***
(-2.42)**
(-1.68)*
(-3.05)***
(1.59)
(2.51)**
0.0111
-0.0210
-0.0488
0.0022
0.0000
0.0083
0.0013
-0.0003
0.0000
0.0145
0.0060
0.0194
-0.0075
0.0026
0.0033
-0.0045
0.0039
0.0021
-0.0043
0.1085
-0.0172
0.0049
-0.0080
0.0039
0.0128
0.0153
-0.0080
-0.0011
0.0005
-0.0018
-0.0217
-0.0032
0.0071
0.0060
0.0038
0.0083
-0.0053
0.0109
0.0090
-0.0102
-0.0074
-0.0054
-0.0080
0.0146
0.0523
(1.87)*
(-4.39)***
(-4.7)***
(1.07)
(-0.54)
(2.78)***
(2.82)***
(-2.4)**
(1.88)*
(2.45)**
(2.12)**
(1.97)**
(-3.22)***
(2.25)**
(2.36)**
(-2.31)**
(3.49)***
(0.37)
(-1.91)*
(2.08)**
(-1.47)
(1.85)*
(-3.05)***
(1.53)
(1.2)
(1.17)
(-3)***
(-3.33)***
(1.63)
(-0.67)
(-2.54)**
(-1.66)*
(3.25)***
(2.42)**
(1.79)*
(2.39)**
(-1.91)*
(2.61)***
(2.74)***
(-3.95)***
(-2.85)***
(-1.81)*
(-3.04)***
(1.11)
(1.07)
Table B.3b continued
y = RP = 3
dy/dx
z-value
ACADxpushC (b)
ACADxpullE (b)
ACADxpullH (b)
AGExpushD
Hdnew2 (b)
Hdnew3 (b)
adj_A (b)
adj_C (b)
difabrdA(b)
contrB2 (b)
FTr4 (b)
lastvis1 (b)
lastvis3 (b)
sept11_inc (b)
-0.0650
0.0236
-0.0848
-0.0035
-0.1089
-0.0266
0.0293
0.0174
0.0312
0.0507
-0.0651
-0.0200
-0.0054
0.0191
(-1.75)*
(2.26)**
(-1.87)*
(-2.83)***
(-2.27)**
(-3.27)***
(3.32)***
(3.07)***
(1.83)*
(3.86)***
(-1.45)
(-1.66)*
(-0.27)
(3.38)***
y = RU = 4
dy/dx
0.1330
-0.0950
0.1688
0.0100
0.1857
0.0893
-0.0896
-0.0800
-0.0741
-0.1316
0.1262
0.0522
-0.2065
-0.0847
z-value
(2.23)**
(-1.42)
(2.43)**
(3.11)***
(3.17)***
(3.34)***
(-3.62)***
(-1.59)
(-2.18)**
(-4.93)***
(1.91)*
(1.85)*
(-6.96)***
(-2.18)**
Notes: * significant at 10%; ** significant at 5%; *** significant at 1%;
(b) dy/dx is for discrete change of dummy variable from 0 to 1.
257
y = DNR = 5
dy/dx
0.0140
-0.0062
0.0187
0.0008
0.0252
0.0063
-0.0067
-0.0049
-0.0066
-0.0112
0.0140
0.0043
-0.0100
-0.0053
z-value
(1.64)
(-1.53)
(1.61)
(2.7)***
(1.85)*
(2.89)***
(-2.83)***
(-1.9)*
(-1.77)*
(-3.44)***
(1.32)
(1.65)*
(-4.09)***
(-2.21)**
Table B.4 Marginal Effects for the Multinomial Logit Model, Professionals
y = DRP = 1
y = DRNP = 2
y = RP = 3
dy/dx z-value
dy/dx z-value
dy/dx z-value
female*
init_UNSURE*
init_RETURN*
age
agesq
staydur
yrs_wrkd_cc
AGExSTAYDUR
AGESQxSTAYDUR
spousnat2*
NWexpTUR*
FFTJloc3*
HDTURXPHD*
social_assess
SOL_assess
fam_sup1
fam_sup2
academic2*
whygo_C*
whygo_F*
whygo_G*
whygo_H*
whygo_I*
whygo_K*
FxWHYGOC*
FxWHYGOI*
ACADxWHYGOG*
AGExWHYGOF
AGExWHYGOG
-0.0002
0.0002
0.0005
0.0003
0.0000
-0.0003
-0.0001
0.0000
0.0000
0.0002
-0.0004
-0.0003
0.0003
-0.0001
-0.0001
0.0000
-0.0001
-0.0001
0.0003
-0.0021
0.0158
0.0001
0.0004
-0.0004
-0.0003
-0.0003
0.0000
0.0001
-0.0001
(-0.63)
(0.4)
(0.89)
(0.87)
(-0.93)
(-0.85)
(-0.95)
(0.6)
(-0.12)
(0.65)
(-1)
(-0.99)
(0.41)
(-0.99)
(-0.48)
(0.33)
(-1.02)
(-0.23)
(0.86)
(-0.77)
(0.49)
(0.32)
(0.56)
(-1.09)
(-0.94)
(-0.94)
(0.14)
(0.9)
(-0.92)
-0.0957
0.1810
0.3067
0.0030
-0.0002
-0.0489
-0.0126
0.0016
0.0000
-0.1162
-0.0496
-0.0783
0.1407
-0.0346
-0.0388
0.0549
-0.0499
-0.0420
0.0587
-0.3013
0.1840
-0.0443
0.2214
0.0048
-0.0758
-0.0933
0.1609
0.0133
-0.0070
(-2.05)**
(1.68)*
(3.76)***
(0.1)
(-0.44)
(-1.21)
(-1.92)*
(0.79)
(-0.37)
(-3.37)***
(-1.65)*
(-2.22)**
(1.32)
(-2.31)**
(-2.41)**
(2.43)**
(-4.75)***
(-0.61)
(1.47)
(-4.02)***
(1.16)
(-1.46)
(2.05)**
(0.15)
(-1.65)*
(-1.84)*
(1.55)
(3.25)***
(-1.48)
-0.1179
0.1816
0.1907
-0.0398
0.0005
-0.0405
-0.0092
0.0012
0.0000
-0.1021
-0.0217
-0.0335
0.0412
-0.0062
-0.0149
-0.0217
-0.0078
0.1387
0.0349
-0.2033
0.1185
-0.0202
-0.1797
-0.1011
-0.0502
0.1407
0.0740
0.0019
-0.0044
258
(-1.49)
(1.86)*
(2.23)**
(-0.77)
(0.66)
(-0.67)
(-1)
(0.42)
(-0.21)
(-1.63)
(-0.44)
(-0.4)
(0.4)
(-0.28)
(-0.54)
(-0.66)
(-0.47)
(1.32)
(0.61)
(-1.13)
(0.54)
(-0.45)
(-1.95)*
(-2.15)**
(-0.45)
(1.17)
(0.71)
(0.36)
(-0.62)
y = RU = 4
dy/dx z-value
0.2141
-0.3610
-0.4950
0.0357
-0.0003
0.0884
0.0218
-0.0027
0.0000
0.2174
0.0715
0.1091
-0.1817
0.0410
0.0531
-0.0329
0.0576
-0.0979
-0.0940
0.4975
-0.3143
0.0636
-0.0412
0.0965
0.1257
-0.0804
-0.2345
-0.0153
0.0113
(2.57)***
(-5.07)***
(-6.61)***
(0.65)
(-0.37)
(1.43)
(2.4)**
(-0.92)
(0.42)
(3.4)***
(1.42)
(1.23)
(-2.69)***
(1.7)*
(1.71)*
(-0.94)
(3.12)***
(-0.95)
(-1.64)*
(2.74)***
(-1.54)
(1.4)
(-0.51)
(1.97)**
(1.03)
(-0.82)
(-3.58)***
(-2.69)***
(1.56)
y = DNR = 5
dy/dx z-value
-0.0003
-0.0017
-0.0030
0.0008
0.0000
0.0012
0.0001
-0.0001
0.0000
0.0006
0.0001
0.0030
-0.0006
0.0001
0.0006
-0.0004
0.0002
0.0013
0.0001
0.0091
-0.0040
0.0008
-0.0009
0.0002
0.0006
0.0333
-0.0005
-0.0001
0.0001
(-0.84)
(-2.04)**
(-2)**
(1.92)*
(-1.87)*
(1.94)*
(1.37)
(-1.99)**
(1.5)
(1.04)
(0.46)
(1.29)
(-1.86)*
(0.44)
(1.56)
(-1.23)
(1.09)
(0.82)
(0.18)
(0.63)
(-1.13)
(1.48)
(-1.83)*
(0.63)
(0.39)
(0.69)
(-1.32)
(-1.29)
(1.53)
Table B.4 continued
y = DRP = 1
dy/dx z-value
pushC*
pushD*
pushF*
pushK*
pullE*
pullF*
pullG*
pullH*
pullI*
pullJ*
pullK*
pullL*
femalexpushC*
femalexpullI*
femalexpullK*
femalexpullL*
ACADxpushC*
ACADxpullE*
ACADxpullH*
AGExpushD
HDnew2*
HDnew3*
adj_A*
adj_C*
difabrdA*
contrB2*
FTr4*
lastvis1*
lastvis3*
sept11_inc*
-0.0007
0.0099
0.0002
-0.0012
-0.0001
0.0001
-0.0003
0.0001
-0.0003
-0.0001
0.0012
0.0061
0.0022
0.0010
0.0000
-0.0004
-0.0002
0.0047
-0.0005
0.0000
-0.0001
-0.0004
0.0002
0.0006
0.0000
0.0002
0.0004
-0.0005
0.0016
0.0007
(-0.97)
(0.41)
(0.83)
(-0.95)
(-0.28)
(0.72)
(-1)
(0.46)
(-0.93)
(-0.62)
(0.89)
(0.89)
(0.66)
(0.52)
(-0.13)
(-0.91)
(-0.83)
(0.6)
(-0.85)
(-0.86)
(-0.57)
(-0.95)
(0.74)
(0.69)
(-0.06)
(0.76)
(0.44)
(-1.01)
(0.9)
(1.15)
y = DRNP = 2
dy/dx z-value
0.0348
0.2097
0.0227
-0.1190
-0.0772
-0.0478
-0.0362
0.0388
-0.0809
-0.1027
0.1699
0.1286
0.0553
0.1678
-0.1046
-0.1350
-0.1130
0.1342
-0.1386
-0.0056
-0.0596
-0.0701
0.0473
0.0569
0.0356
0.1072
-0.0571
-0.0607
0.1165
0.0304
(1)
(1.21)
(0.73)
(-2.47)**
(-1.89)*
(-1.38)
(-1.15)
(1.01)
(-2.39)**
(-3.21)***
(2.68)***
(0.97)
(0.78)
(1.63)
(-2.31)**
(-3.03)***
(-2.81)***
(1.23)
(-3.25)***
(-1.27)
(-0.99)
(-2.06)**
(1.66)*
(0.83)
(1.02)
(3.96)***
(-1.14)
(-2.13)**
(1.83)*
(0.75)
y = RP = 3
dy/dx z-value
0.0543
0.1738
0.0483
0.1137
0.0162
-0.0139
-0.0595
-0.0206
-0.0680
-0.0108
0.0271
-0.0301
0.0158
-0.0187
-0.0105
-0.2132
-0.1372
-0.2047
0.0249
-0.0071
-0.1352
0.0185
-0.0138
0.0963
0.0635
0.0695
-0.0190
0.0592
0.1870
0.0902
259
(0.95)
(0.91)
(1.09)
(1.94)*
(0.29)
(-0.27)
(-1.17)
(-0.39)
(-1.18)
(-0.23)
(0.41)
(-0.24)
(0.19)
(-0.19)
(-0.08)
(-1.01)
(-1.44)
(-1.91)*
(0.22)
(-1.23)
(-1.29)
(0.4)
(-0.34)
(1.27)
(1.11)
(1.62)
(-0.17)
(1.28)
(2.81)***
(1.27)
y = RU = 4
dy/dx z-value
-0.0879
-0.3922
-0.0712
0.0059
0.0602
0.0612
0.0947
-0.0173
0.1487
0.1132
-0.1973
-0.1046
-0.0733
-0.1497
0.1103
0.3465
0.2497
0.0663
0.1133
0.0127
0.1933
0.0518
-0.0330
-0.1535
-0.0982
-0.1763
0.0718
0.0021
-0.3011
-0.1206
(-1.48)
(-2.72)***
(-1.64)*
(0.1)
(1.12)
(1.15)
(1.79)*
(-0.33)
(2.42)**
(2.34)**
(-3.5)***
(-1.21)
(-0.97)
(-2.1)**
(0.79)
(1.57)
(2.45)**
(0.52)
(0.98)
(2.22)**
(1.83)*
(1.14)
(-0.81)
(-2.41)**
(-1.66)*
(-3.81)***
(0.59)
(0.04)
(-7.08)***
(-1.65)*
y = DNR = 5
dy/dx z-value
-0.0005
-0.0012
0.0000
0.0006
0.0009
0.0005
0.0013
-0.0011
0.0006
0.0004
-0.0009
-0.0001
0.0000
-0.0005
0.0049
0.0022
0.0007
-0.0006
0.0008
0.0000
0.0017
0.0002
-0.0008
-0.0003
-0.0009
-0.0006
0.0040
0.0000
-0.0039
-0.0007
(-1.19)
(-1.03)
(-0.17)
(1.83)*
(1.73)*
(1.21)
(1.71)*
(-1.71)*
(1.38)
(1.06)
(-1.92)*
(-0.12)
(0.03)
(-1.49)
(0.57)
(0.6)
(0.74)
(-1.25)
(0.7)
(1.25)
(0.83)
(0.61)
(-1.73)*
(-0.58)
(-1.41)
(-1.32)
(1.08)
(-0.15)
(-1.8)*
(-1.87)*
Table B.5 Summary Statistics and Descriptions of the Variables used in the Final Model,
Students (n = 960)
Std
Variable
Variable Descriptions
Mean
Min
Dev.
y
female
age
agesq
init_UNSURE
init_STAY
staydur1
FAMSUP1_S
FAMSUP2_SS
FAMSUP2_DS
soc_W
SOL_B
TSA_member
res_USA
fieldnew1
fieldnew3
div_sep
not_married
spousenat2
whygo_A
whygo_C
whygo_F
whygo_G
whygo_H
whygo_I
whygo_K
DC_E
DC_F
adj_A
adj_F
Dependent variable: return intentions
(1=return without completing studies;
2=return immed. after compl. studies;
3=return probable; 4=return unlikely;
5=definitely not return)
Gender of respondent (1=female)
Age of respondent in 2001
Square of Age
Initial return intentions: Unsure (1=yes)
Initial return intentions: Return (1=yes)
Stay duration in current country of residence
(years)
Family support for initial decision to go
abroad (1= supportive)
Family support for settling abroad
(1=somewhat supportive)
Family support for settling abroad
(1=definitely supportive)
Assessment of social conditions abroad
(1=much worse or worse)
Assessment of standard of living abroad
(1=better or much better)
Turkish Student Association membership
(1=yes)
Current residence is USA (1=yes)
Current field of study: arch / econ / admin
Current field of study: engin / math / science
/ medic
Respondent is divorced or separated
Respondent has never married
Respondent is married to a foreign spouse
Learn language, improve language skills
Job requirement in Turkey
Insufficient facilities, equipment for research
Prestige and advantages of study abroad
Lifestyle preference
To be with spouse, family
Get away from political environment
Chose current institution because of job
opportunities
Chose current institution to be near spouse
Adjustment Factor: previous experience
Adjustment Factor: Turkish friends at
institution of study
261
3.57
Max
1.06
1
6
0.49
0.39
26.96
3.67
740.40 207.08
0.37
0.48
0.09
0.29
0
18
324
0
0
1
44
1936
1
1
2.79
2.31
0
13
0.95
0.21
0
1
0.48
0.50
0
1
0.27
0.44
0
1
0.44
0.50
0
1
0.69
0.46
0
1
0.57
0.86
0.29
0.49
0.35
0.45
0
0
0
1
1
1
0.58
0.02
0.71
0.02
0.25
0.41
0.45
0.72
0.24
0.08
0.25
0.49
0.15
0.45
0.14
0.44
0.49
0.50
0.45
0.43
0.27
0.44
0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
1
1
1
0.26
0.11
0.34
0.44
0.31
0.47
0
0
0
1
1
1
0.57
0.50
0
1
Table B.5 continued.
Variable
Variable Description
difabrdF
Difficulties faced while abroad:
unemployment
Respondent plans to work in academia 5
years after completing studies
Respondent is bound by compulsory
academic service requirement
Push Factor: being away from research
centers and recent advance
Push Factor: less than satisfying cultural and
social life
Pull Factor: a higher level of income in host
country
Pull Factor: better work environment
Pull Factor: greater job availability in
specialization
Pull Factor: more organized, ordered
environment
Pull Factor: proximity to research and
innovation centers
Pull Factor: spouse's preference or job
academic_b
compulsory
pushE
pushG
pullA
pullC
pullD
pullF
pullH
pullI
pullJ
pullK
pullL
lastvis1
lastvis3
sept11_inc
Mean
Pull Factor: better educational opportunities
for children
Pull Factor: need to finish current project
Pull Factor: other factors
Last visit to Turkey decreased return
intentions (1=yes)
Last visit to Turkey increased return
intentions (1=yes)
Effect of September 11, 2001 (1=increased
return intentions)
262
Std
Dev.
Min
Max
0.05
0.21
0
1
0.47
0.50
0
1
0.18
0.38
0
1
0.59
0.49
0
1
0.23
0.42
0
1
0.76
0.68
0.43
0.47
0
0
1
1
0.75
0.43
0
1
0.76
0.42
0
1
0.60
0.49
0
1
0.21
0.41
0
1
0.19
0.30
0.04
0.39
0.46
0.19
0
0
0
1
1
1
0.32
0.47
0
1
0.09
0.29
0
1
0.14
0.34
0
1
Table B.6 Estimation Results and Marginal Effects for each Outcome, Ordered Probit Model,
Students
Explanatory
dy/dx
Variables
y=1
y=2
y=3
y=4
y=5
y=6
(a) z-statistic
female(b)
age
agesq
init_UNSURE(b)
init_STAY(b)
staydur1
FAMSUP2_SS(b)
FAMSUP2_DS(b)
soc_W(b)
SOL_B(b)
TSA_member(b)
div_sep(b)
not_married(b)
spousenat2(b)
whygo_A(b)
whygo_C(b)
whygo_F(b)
whygo_G(b)
whygo_H(b)
whygo_I(b)
whygo_K(b)
ACADxwhygoF(b)
ACADxwhygoG(b)
ACADxwhygoI(b)
ACADxwhygoK(b)
DC_E(b)
DC_F(b)
adj_A(b)
adj_F(b)
difabrdF(b)
academic_b(b)
compulsory(b)
pushE(b)
pushG(b)
pullA(b)
pullC(b)
pullD(b)
pullF(b)
pullI(b)
pullJ(b)
pullK(b)
pullL(b)
0.124
0.036
-0.001
0.495
1.434
0.087
0.216
0.415
-0.339
0.172
-0.167
0.542
0.181
0.545
-0.127
-0.248
0.220
-0.241
0.213
-0.331
0.280
-0.252
0.349
-0.604
0.370
0.290
0.436
-0.178
-0.128
-0.227
-0.430
-0.705
0.191
-0.061
0.279
-0.104
0.092
0.225
0.365
-0.116
-0.087
-0.469
(1.61)
(0.34)
(0.60)
(5.66)***
(8.55)***
(4.26)***
(2.55)**
(3.80)***
(4.49)***
(1.99)**
(2.15)**
(2.44)**
(1.60)
(1.64)
(1.47)
(3.05)***
(2.14)**
(2.12)**
(2.06)**
(1.65)*
(2.42)**
(1.67)*
(2.13)**
(2.67)***
(2.03)**
(3.58)***
(2.82)***
(2.19)**
(1.64)
(1.33)
(2.51)**
(5.75)***
(2.25)**
(0.56)
(3.27)***
(1.26)
(1.02)
(2.50)**
(3.53)***
(1.12)
(0.77)
(1.53)
0.000
0.000
0.000
-0.001
-0.001
0.000
0.000
-0.001
0.001
0.000
0.000
-0.001
0.000
-0.001
0.000
0.001
0.000
0.000
0.000
0.001
0.000
0.001
-0.001
0.003
-0.001
-0.001
-0.001
0.000
0.000
0.001
0.001
0.004
0.000
0.000
-0.001
0.000
0.000
-0.001
-0.001
0.000
0.000
0.002
263
-0.015
-0.004
0.000
-0.055
-0.077
-0.010
-0.026
-0.044
0.042
-0.022
0.020
-0.044
-0.023
-0.044
0.016
0.031
-0.026
0.027
-0.024
0.049
-0.031
0.033
-0.039
0.107
-0.036
-0.032
-0.041
0.022
0.015
0.032
0.053
0.118
-0.024
0.008
-0.038
0.012
-0.011
-0.030
-0.038
0.015
0.011
0.077
-0.034
-0.010
0.000
-0.139
-0.379
-0.024
-0.060
-0.119
0.091
-0.046
0.046
-0.163
-0.048
-0.163
0.034
0.067
-0.061
0.068
-0.061
0.080
-0.080
0.066
-0.099
0.118
-0.109
-0.083
-0.129
0.048
0.035
0.057
0.116
0.147
-0.052
0.017
-0.072
0.029
-0.025
-0.059
-0.106
0.031
0.024
0.102
0.023
0.007
0.000
0.085
-0.005
0.017
0.041
0.068
-0.066
0.034
-0.031
0.062
0.036
0.062
-0.025
-0.049
0.042
-0.043
0.038
-0.072
0.049
-0.052
0.061
-0.140
0.056
0.050
0.062
-0.035
-0.024
-0.048
-0.082
-0.157
0.037
-0.012
0.058
-0.019
0.018
0.046
0.060
-0.023
-0.017
-0.107
0.025
0.007
0.000
0.106
0.404
0.017
0.043
0.092
-0.066
0.033
-0.034
0.138
0.035
0.139
-0.025
-0.048
0.045
-0.051
0.045
-0.056
0.060
-0.047
0.075
-0.086
0.085
0.062
0.103
-0.034
-0.026
-0.040
-0.085
-0.108
0.038
-0.012
0.051
-0.021
0.018
0.042
0.081
-0.022
-0.017
-0.073
0.001
0.000
0.000
0.004
0.057
0.001
0.001
0.004
-0.002
0.001
-0.001
0.008
0.001
0.008
-0.001
-0.002
0.001
-0.002
0.002
-0.001
0.002
-0.001
0.003
-0.002
0.004
0.002
0.005
-0.001
-0.001
-0.001
-0.003
-0.003
0.001
0.000
0.001
-0.001
0.001
0.001
0.003
-0.001
-0.001
-0.002
Table B.6 continued
Explanatory
Variables
ACADxpushG(b)
ACADxpullK(b)
ACADxpullL(b)
lastvis1(b)
lastvis3(b)
sept11_inc(b)
(a)
z-statistic
y=1
y=2
0.403
-0.188
0.864
0.352
-0.350
-0.284
(2.12)**
(1.18)
(1.84)*
(3.99)***
(2.91)***
(2.79)***
-0.001
0.000
-0.001
-0.001
0.001
0.001
-0.038
0.025
-0.055
-0.039
0.052
0.040
dy/dx
y=3
y=4
-0.119
0.049
-0.253
-0.100
0.084
0.072
0.058
-0.039
0.048
0.061
-0.077
-0.060
y=5
y=6
0.095
-0.035
0.240
0.075
-0.059
-0.051
0.004
-0.001
0.020
0.003
-0.002
-0.001
Notes: * significant at 10%; ** significant at 5%; *** significant at 1%;
(a) Robust z-statistics in parentheses; Observations = 960; Log-likelihood = -1073.44; LR
chi2(48)= 583.83; Maximum Likelihood R2 = 0.491; McFadden’s Adjusted R2 = 0.194;
McKelvey-Zavoina R2 = 0.535; AIC = 2.347; BIC= -4081.431.
(b) dy/dx is for discrete change of dummy variable from 0 to 1.
264
APPENDIX C
SURVEY LETTERS AND SURVEY QUESTIONNAIRES
C.1 E-Mail Cover Letter (English and Turkish Versions)
Dear ...,
We are conducting a survey on the Turkish brain drain and we need your help.
The purpose of our study is to identify the reasons why skilled individuals of Turkish
origin, students studying abroad or professionals, do not return or postpone returning to
Turkey. With this study we also hope to shed light on mobility patterns, and interactions or
feedback patterns between those that have gone abroad and their friend, family and
colleagues staying in Turkey.
The survey will take 15, at most 20, minutes of your time. The questions are easy to answer
but need to be responded to carefully. The information you provide will be kept strictly
confidential and will in no way be presented in a way that will identify you.
The survey is not anonymous, but confidential. We ask for your name and email for the
purpose of identifying who we have successfully reached. We will also be sending a
summary of our findings to respondents via email.
We would appreciate it if you would forward this message to any friends or colleagues of
Turkish origin who meet the following criteria:
1) those studying abroad at the university level (associate, bachelors, masters, doctorate,
postdoc)
2) those who are working abroad holding at least a bachelor’s degree.
We are trying to reach as many people as we can who hold the above qualifications.
Our survey web address is http://www.metu.edu.tr/home/survey . To fill out the survey
please go to this address and follow the appropriate links.
We would greatly appreciate your prompt response.
We thank you again for your help.
Yours sincerely,
Aysit Tansel,
Nil Demet Güngör
Prof. Dr. Aysit Tansel / Research Assistant Nil Demet Güngör
Department of Economics
Faculty of Economic and Administrative Sciences
Middle East Technical University
Ankara, Turkey 06531
email: survey@metu.edu.tr
264
[Turkish Version]
De erli ...,
Türkiye’den beyin göçü üzerine bir anket çalı ması yapıyoruz ve sizin yardımınıza
ihtiyacımız var.
Çalı mamızın amacı yurtdı ında okuyan Türk ö rencilerin ve yurtdı ında çalı an nitelikli
Türk i gücünün yurda dönmemelerinin veya dönmeyi ertelemelerinin nedenlerini tespit
etmektir. Çalı mamızla ayrıca vasıflı i gücü hareketlerine ve yurtdı ında bulunanların
Türkiye’deki arkada , aile ve meslekta larıyla etkile imlerine ı ık tutmayı ümit ediyoruz.
Anketimizi doldurmanız en fazla 15-20 dakikanızı alacaktır. Sorular kolay
cevaplandırabilece iniz sorulardır ama dikkatli okunmaları gerekiyor. Verece iniz tüm
bilgiler bizde gizli kalacaktır ve çalı mamızın sonuçları hiçbir ekilde bireylerin tespit
edilmesini mümkün kılacak ekilde sunulmayacaktır.
Anketimizde isim ve eposta alanları da yer almaktadır. Bu bilgileri istememimizin nedeni
kimlere ula abildi imizi anlayabilmemizdir. Ayrıca çalı mamızın bitiminde, ankete
katılanlara bulgularımızın bir özetini eposta ile gönderece iz.
Bu mesajı a a ıdaki kriterlere uyan tanıdıklarınıza gönderebilirseniz çok seviniriz.
1) yurtdı ında okuyan Türk ö renciler (lise üstü teknik kolej, lisans, master, doctora,
doktora sonrası e itim düzeyinde)
2) yurtdı ında çalı an ve en az lisans derecesine sahip olan Türkler
Bu niteliklere sahip olan mümkün oldu u kadar fazla ki iye ula maya çalı ıyoruz.
Web adresimiz http://www.metu.edu.tr/home/survey ’dir. Anket formunu doldurmak için
lütfen bu adrese girip, sayfadaki linkleri takip ediniz.
Yanıtınızı en kısa zamanda yollarsanız çok memnun oluruz.
Yardımınız için tekrar te ekkür ederiz.
Saygılarımızla,
Aysit Tansel,
Nil Demet Güngör
Prof. Dr. Aysit Tansel / Ara tırma Görevlisi Nil Demet Güngör
ktisat Bölümü
ktisadi ve dari Bilimler Fakültesi
Orta Do u Teknik Üniversitesi
Ankara, Turkiye 06531
eposta: survey@metu.edu.tr
265
C.2 Courtesy Reply Message (English and Turkish Versions)
Dear friend,
Thank you for participating in our survey on the determinants of student nonreturn and migration of skilled individuals from Turkey to other countries. We
appreciate your help.
If you know any colleagues or friends of Turkish origin who qualify to take part
in our survey (students at the undergraduate or graduate level studying abroad, and
skilled individuals holding at least a bachelor's degree who are working abroad), we
would appreciate it if you would forward them our survey address.
If you would like to receive a copy of the survey results by email when they
become available, please reply to this message indicating that you request a copy.
Regards,
Prof. Dr. Aysit Tansel, Research Assistant Nil Demet Gungor
METU Department of Economics
survey@metu.edu.tr
http://www.metu.edu.tr/home/survey
Degerli arkadas / meslektas,
Turkiye’den diger ulkelere goc eden egitimli bireylerin, yurda donmeme
kararlariyla ilgili olarak yaptigimiz arastirmaya katildiginiz icin tesekkur ederiz.
Eger arastirmamiza katilabilecek niteliklerdeki Turk arkadas ve meslektaslariniz
varsa (yabanci ulkelerdeki universite mezunu veya universiteyi bitirmekte olan
ogrenciler ve disarda calismakta olan en azindan universite mezunu bireyler)
anketimizin web adresini bu kimselere iletirseniz memnun oluruz.
Calismamizin sonuclarinin bir kopyasini isterseniz, bu mesaja isteginizi belirten
bir yanit gondererek bize bildirebilirsiniz. Degerli katkilarinizdan dolayi tekrar
tesekkur ederiz.
Saygilar,
Prof. Dr. Aysit Tansel, Arastirma Gorevlisi Nil Demet Gungor
ODTU Iktisat Bolumu
survey@metu.edu.tr
http://www.metu.edu.tr/home/survey
266
C.3 English Mail-outVersion of Tertiary-Educated Workforce Abroad Survey
Brain Drain Survey of
Academicians, Professionals and other Workers
1) Please write your name and e-mail address in the boxes provided below. This
information is for our record-keeping only; it will not be used in our study, and it will
not be disclosed in any way to other parties. The information you provide will be used
for research or statistical purposes only.
2) Please read and answer carefully. The survey will take approximately 15-20
minutes. Since not all of the questions will apply, you will be able skip those that are
not relevant to you.
3) Please place an X within the square bracket of the appropriate selection for
multiple choice questions.
Thank you again for taking the time to participate in our study.
YOUR NAME:
_____________________________
YOUR E-MAIL ADDRESS:
_____________________________
GENERAL INFORMATION
1. Personal Information: Please indicate your
a) Gender:
Please mark the appropriate box with an X.
[ ] Male
[ ] Female
b) Birthyear:
_____________
c) Birthplace:
city:
country:
_______________
_______________
2. a) What is your current country of residence?
Please mark the appropriate box with an X.
[ ] Australia
[ ] Canada
267
[
[
[
[
]
]
]
]
England
New Zealand
USA
other, please indicate:
_______________
b) Please indicate your current city and (if applicable) state or province of
residence:
city:
___________
state / province:
___________
3. How long have you been staying in your current COUNTRY of residence?
____________
number of years
4. a) Did you have any study, work, travel or other experience outside Turkey
prior to coming to your current country of residence?
Please mark the appropriate box with an X.
[ ] yes
[ ] no
If you have no prior experience abroad please proceed to question 5.
b) What kind of previous experience did you have abroad?
Please mark all that apply.
[
[
[
[
]
]
]
]
study
work
travel
other, please specify: ______________
c) What is the longest period you have spent outside Turkey not counting
your current stay?
_____________
number of months/years
EDUCATIONAL INFORMATION
5. a) Which high school (lycée) did you graduate from?
Please indicate the name of the high school and its location.
NAME:
____________
LOCATION: ____________
(e.g. Ankara)
268
b) What was the language of instruction of the
i. science courses at your high school (lycée)?
Please mark the appropriate box with an X.
[
[
[
[
[
]
]
]
]
]
Turkish
English
German
French
other, please specify: ______________
ii. social courses (e.g., history, geography, philosophy) at your high
school (lycée)? Please mark the appropriate box with an X.
[
[
[
[
[
]
]
]
]
]
Turkish
English
German
French
other, please specify: ______________
6. a) From which university did you receive your undergraduate (bachelor's
or associate's) degree?
__________________________________________________________________
b) What was your major?
__________________________
c) What year did you graduate? ___________________________
7. a) What is the highest academic degree you hold?
Please mark the appropriate box with an X.
[
[
[
[
[
[
]
]
]
]
]
]
associate’s
bachelor’s
post baccalaureate certificate
master’s
post master’s certificate
doctorate
b) From which university did you receive your highest academic degree?
__________________________________________________________________
c) In which country did you receive your highest academic degree?
Please mark the appropriate box with an X.
[
[
[
[
[
[
]
]
]
]
]
]
Australia
Canada
England
New Zealand
USA
Other, please indicate:
_____________
269
d) Please indicate also the city and (if applicable) state or province
where you received your highest degree.
CITY :
_____________
STATE/PROVINCE : _____________
e) What was your field of study?
Please be specific and indicate any areas of specialization as well.
General field of study: ___________________
Specialization 1: ______________
Specialization 2: ______________
Specialization 3: ______________
f) If you received your last academic degree outside Turkey, where did
you start your first first full time job after completing your studies?
Please mark the appropriate box with an X.
[
[
[
[
[
]
]
]
]
]
same city and country where I received my last degree
same country but different city
Turkey
another country, please specify: _____________
not applicable
8. What is the highest academic title you hold or have held in the past?
Please mark the appropriate boxes with an X.
a) in Turkey:
[
[
[
[
[
[
[
]
]
]
]
]
]
]
none
professor
associate professor
assistant professor
instructor / lecturer
research assistant
teaching assistant
b) in your current country of residence:
[
[
[
[
[
[
[
]
]
]
]
]
]
]
none
professor
associate professor
assistant professor
instructor / lecturer
research assistant
teaching assistant
9. a) Do you hold any professional degrees?
(e.g. in law, medicine, dentistry, nursing, pharmacy, etc.
such as M.D., V.M.D., J.D., L.L.M.)
[ ] yes
[ ] no
270
b) If so, which professional degrees do you hold?
____________________________
WORK-RELATED INFORMATION
10. What is your occupation?
Please be specific. For example, high school science teacher, university professor in
sociology, computer programmer, mid-level manager etc.
____________________________
11. What is your current employment status?
[ ] self-employed
[ ] employee
[ ] unemployed, looking for a job
If you are unemployed or between jobs, please refer to your last job when answering
questions concerning your 'current workplace or institution'.
12. a) How long have you been working outside Turkey?
__________________
number of years
b) How long have you been working in your current country of residence?
__________________
number of years
c) How long have you been working at your current workplace/institution?
__________________
number of years
13. How many different organizations have you worked full time for so far?
in Turkey:
___________
abroad:
___________
14. What sector is the firm / organization you are currently working for in?
Please mark the appropriate box with an X.
[ ] private
[ ] public
[ ] other (e.g., non-profit organization or trust)
15. What type of organization do you work for?
Please mark the appropriate box with an X.
[
[
[
[
]
]
]
]
Multinational firm (headquarters in Turkey)
Multinational firm (headquarters in current country)
Multinational firm (headquarters in third country)
Other incorporated firm
271
[
[
[
[
[
[
[
[
[
[
[
[
[
]
]
]
]
]
]
]
]
]
]
]
]
]
Non-incorporated firm or business
Pre-school, primary or middle school (junior high school)
High school (secondary school)
2-3 year arts college or technical institute
University
Research center at a university
Hospital / medical center
International Organization (IMF, ILO, World Bank, etc.)
Armed forces
Government department, organization
Local government
Non-governmental organization
other, please specify:
______________________
16.a) When was the firm or organization you are working for established?
Please mark the appropriate box with an X.
[
[
[
[
[
[
[
[
[
]
]
]
]
]
]
]
]
]
within the past year (Jan. 1 2001 – Dec. 31, 2001)
within the last 2 years
within the last 5 years
within the last 10 years
10-15 years ago
15-30 years ago
30-50 years ago
more than 50 years ago
don’t know
b) Approximately how many people currently work full time in your
organization (at all levels)? Please mark the appropriate box with an X.
[
[
[
[
[
[
[
[
[
[
]
]
]
]
]
]
]
]
]
]
less than 5
5-11
11-25
26-50
51-100
101-200
201-500
501-1000
more than 1000
don’t know
17. In which country were you residing when you found (or established) your
current job abroad? Please mark the appropriate box with an X.
[ ] in my current country of residence
[ ] in Turkey
[ ] in a third country, please specify:
__________
18. a) Through which channel(s) did you find your current job?
Please mark all that apply.
[ ] Direct contacts initiated with firm / organization
(e.g., sending unsolicited CV)
[ ] Professional recruiters (e.g., "headhunters")
[ ] 'Career Days' held at Turkish universities
[ ] Informal channels (e.g., friends, colleagues)
[ ] Ads in professional journals
272
[
[
[
[
[
]
]
]
]
]
Turkish internet network (e.g., alumni networks)
Newspaper ads
Placement office at university
Faculty or advisors
other, please specify: _____________
b) How did you find your first full time job abroad?
Please mark all that apply.
[
[
[
[
[
[
[
[
[
[
]
]
]
]
]
]
]
]
]
]
Direct contacts initiated with firm / organization (e.g., sending unsolicited CV)
Professional recruiters (e.g., "headhunters")
'Career Days' held at Turkish universities
Informal channels (e.g., friends, colleagues)
Ads in professional journals
Turkish internet network (e.g., alumni networks)
Newspaper ads
Placement office at university
Faculty or advisors
other, please specify: _____________
19. During the past year what percentage of your time on your current job
went to each of the following activities listed below? Please indicate the
percentage in the squared brackets. e.g., teaching [50] %, basic research [40] %,
administrative [10] %, others [0] %.
a) Teaching
b) Applied research activities
(research for the purpose of gaining knowledge to meet a specific need)
c) Basic research activities
(research for the purpose of gaining knowledge for its own sake)
d) Development
(transforming knowledge from research into production)
e) Computer use, programming, system development
f) Administrative activities, supervision
g) Professional services
(medical practice, legal practice, financial consultancy)
h) Quality control, production management
i) Accounting, contracts
j) Marketing, consumer services, public relations
k) Other, please specify below:
___________________________________________________
[ ] %
[ ] %
[ ] %
[ ] %
[ ] %
[ ] %
[ ] %
[
[
[
[
]
]
]
]
%
%
%
%
20. a) Have you received any formal job skills training in your current
organization?
(e.g., classroom, seminar, lecture or workshop training in management, professional
and technical skills, computer, clerical, sales, customer relations, service-related or
production-related)
Please mark the appropriate box with an X.
[ ] yes
[ ] no
If you have not received any formal training, please go to question 21.
273
b) In general, would you say that the skills you acquired from formal
training in your current job are:
[ ] Specific to your current organization (cannot be easily
transferred to other organizations)
[ ] Specific to the industry of your organization (can be easily
transferred between organizations in the same industry
but not between industries)
[ ] Generally transferable to other organizations in other industries
c) For what reasons did you receive formal training at your current
organization?
Please mark all that apply.
[ ] To gain new knowledge or skills related to my profession
that would improve my job performance
[ ] Training was compulsory, mandatory
[ ] Training was required for future advancement
[ ] To stay up-to-date with new regulations, laws, technologies, etc.
[ ] To receive promotion at the end of training
[ ] To receive certification / licence upon completion
[ ] To receive higher pay or bonus upon completion
[ ] other, please specify: _________________________
21. a) Have you received any informal on-the-job training in your current
organization? (e.g., learning from senior colleagues during medical internship,
any other learning on the job.)
[ ] yes
[ ] no
If you have not received any formal training, please go to question 22.
b) In general, would you say that the informal training you received in
your current organization is:
[ ] Specific to your current organization (cannot be easily
transferred to other organizations)
[ ] Specific to the industry of your organization (can be easily
transferred between organizations in the same industry
but not between industries)
[ ] Generally transferable to other organizations in other industries
22. a) How many hours did you typically spend on your current job each
week during the past year (2001)?
___________________
HOURS PER WEEK
b) How many weeks did you spend on your current job during the past
year (2001)?
___________________
274
WEEKS PER YEAR
23. What was your 2001 gross annual wage or salary income in U.S.$ from
your current job?
[
[
[
[
[
[
[
]
]
]
]
]
]
]
under $ 20,000
$ 20,000 - $ 49,999
$ 50,000 - $ 74,999
$ 75,000 - $ 99,999
$ 100,000 - $ 149,999
over $ 150,000
rather not answer
If you received your salary or earnings in a different currency, indicate the
currency type and your gross annual income in local units below.
CURRENCY TYPE (CAN$, DM, £, ¥, etc.) : __________
ANNUAL GROSS INCOME :
____________
QUESTIONS RELATING TO THE DECISIONS TO
LEAVE, STAY AND RETURN
24. a) What were your main reasons for going to the country you are
currently staying? Please mark all that apply.
[ ] A. To learn a new language / improve language skills
[ ] B. In need of change / want to experience a new culture
[ ] C. Education or experience in another country is required
by employers in Turkey
[ ] D. Could not find a job in Turkey
[ ] E. No program in my specialization in Turkey
[ ] F. Insufficient facilities, lack of necessary equipment
to carry out research in Turkey
[ ] G. In order to take advantage of the prestige
and advantages associated with study abroad
[ ] H. Preference for the lifestyle in my current country of residence.
[ ] I. To be with spouse or loved one
[ ] J. To provide a better environment for children
[ ] K. To get away from the political environment in Turkey
[ ] L. other, please specify: ____________________
b) Which of the above was the most important reason?
___________________________
25. a) In general, how supportive was your family (e.g. father, mother,
spouse) in your decision to go abroad to work or study?
Please mark the appropriate box with an X.
[
[
[
[
[
]
]
]
]
]
very supportive
somewhat supportive
not very supportive
not at all supportive
not applicable
275
b) Do you think your family in Turkey would support (or supports) your
decision to settle permanently outside Turkey?
Please mark the appropriate box with an X.
[
[
[
[
[
[
[
]
]
]
]
]
]
]
They would definitely support me.
They would most likely support me.
Some family members would support me, others would not.
They are not likely to be very supportive.
They would actively discourage me.
I am not sure.
not applicable
26. Before you left Turkey, what were your thoughts about returning?
Please mark the appropriate box with an X.
[ ] I thought that I would definitely return.
[ ] I was undecided about returning; I would wait and see.
[ ] I did not think that I would return.
27. a) What are your thoughts about returning to Turkey now?
Please mark the appropriate box with an X.
[
[
[
[
[
]
]
]
]
]
I will definitely return and have made plans to do so.
I will definitely return but have not made concrete plans to do so.
I will probably return.
I don't think that I will be returning.
I will definitely not return.
If you marked one of the last two options ('not return') please question 30.
28. When do you think you will be returning to Turkey?
Please mark the appropriate box with an X.
[
[
[
[
[
[
[
]
]
]
]
]
]
]
within 6 months
6 to 12 months
1 to 2 years
2 to 5 years
5 to 10 years
more than 10 years
not applicable
29. a) What are your main reasons for returning to Turkey?
Please mark all that apply.
[ ] to complete compulsory military service
[ ] to complete university service (e.g., YÖK, TÜBA scholarship recipients)
[ ] I will return when my permitted time for working abroad ends
(e.g. I am a visiting scholar)
[ ] I miss my family in Turkey
[ ] I want my children to continue their education in Turkey
[ ] after achieving specific goals (gaining work experience, completing research
project) I want to apply what I have learned in Turkey
[ ] I will return after reaching my savings goal
[ ] I will return after reaching my career goal
[ ] I received a job offer from a firm or institution in Turkey
276
[ ] I want to spend my retirement in Turkey.
[ ] I don't feel safe in my current environment
[ ] other, please specify:
________________________________________
b) After you return, do you plan to go abroad again?
Please mark the appropriate box with an X.
[
[
[
[
[
[
[
]
]
]
]
]
]
]
No
Yes, for a few days to several weeks at most
Yes, for 1-3 months at most
Yes, for 4-6 months at most
Yes, for 7-12 months at most
Yes, for 1-2 years at most
Yes, could be longer than 2 years but I believe I will
definitely return to Turkey.
[ ] Yes, to settle down permanently
[ ] not applicable
30. In general, how does your life in your current country of residence
compare with your life in Turkey?
a) work environment (e.g. your job satisfaction):
[
[
[
[
[
[
]
]
]
]
]
]
much better
better
neither better or worse
worse
much worse
don’t know
b) social aspects (e.g. friendships, social relations):
[
[
[
[
[
[
]
]
]
]
]
]
much better
better
neither better or worse
worse
much worse
don’t know
c) standard of living:
[
[
[
[
[
[
]
]
]
]
]
]
much better
better
neither better or worse
worse
much worse
don’t know
31. a) What are the main difficulties that you have faced / are facing living in
your current country of residence? Please mark all that apply.
[
[
[
[
[
]
]
]
]
]
A. Being away from family
B. Children growing up in a different culture
C. Loneliness, not being able to adjust
D. Fast-paced life
E. Little or no leisure time
277
[
[
[
[
[
[
[
[
]
]
]
]
]
]
]
]
F. Unemployment
G. No jobs in my area of specialty
H. Discrimination against foreigners
I. Lower income compared to the income I had in Turkey
J. Higher taxes
K. Crime, lack of personal security
L. High cost of living
M. Other, please specify: ___________________________________
b) Which of the above factors do you consider to be the most difficult
for you? __________________________
32. a) Which of the following factors were important in helping you adjust to
life abroad? Please mark all that apply.
[
[
[
[
[
[
[
[
[
]
]
]
]
]
]
]
]
]
A. having previous experience abroad
B. the passage of time
C. support from the Turkish Student Association (TSA) at my institution
D. having spouse or other loved one with me
E. having cultural attaché / embassy support
F. having Turkish friends/colleagues at my university/college/research center
G. existence of a large Turkish community in my city
H. being able to share experiences, ask for advise via Turkish internet network
I. other, please specify:
___________________________________
b) Which has been the most important factor in helping you adjust?
_________________________________________________
278
33. What are the greatest difficulties RELATING TO TURKEY that may cause
you NOT to return? Please indicate how important for you the following
factors are in this decision.
Please answer even if you have indicated that you will definitely return.
REASON
Very
Somewhat Not
Not at all
Important Important Important Important Important
5
4
3
2
1
A. Low income in my occupation
___
___
___
___
___
B. Little opportunity for advancement
in my occupation
___
___
___
___
___
C. Limited job opportunities in my field of
expertise
___
___
___
___
___
D. No opportunity for advanced
training in my field
___
___
___
___
___
E. Being far from important research
centers and as a result from new
advances
___
___
___
___
___
F. Lack of financial resources and
opportunities to start up my business
___
___
___
___
___
G. Less than satisfying social and
cultural life
___
___
___
___
___
H. Bureaucracy, inefficiencies in
organizations
___
___
___
___
___
I. Political pressures, discord
___
___
___
___
___
J. Lack of social security
___
___
___
___
___
K. Economic instability, uncertainty
___
___
___
___
___
L. Other reason, please indicate below:
___
___
___
___
___
______________________________
279
34. Please indicate the relative importance FOR YOU of each of the following
factors relating to your CURRENT COUNTRY OF RESIDENCE in deciding not
to return or postpone returning to Turkey.
Please answer even if you have indicated that you will definitely return.
REASON
Very
Somewhat Not
Not at all
Important Important Important Important Important
5
4
3
2
1
A. Higher salary or wage
___
___
___
___
___
B. Greater opportunity to advance
in profession
___
___
___
___
___
C. Better work environment
(flexible work hours, relaxed setting, etc.)
___
___
___
___
___
D. Greater job availability in my area
of specialization
___
___
___
___
___
E. Greater opportunity for further
development in area of specialty
___
___
___
___
___
F. A more organized and ordered life
in general
___
___
___
___
___
G. More satisfying social and cultural life
___
___
___
___
___
H. Proximity to important research
and innovation centers
___
___
___
___
___
I. Spouse's preference to stay or
spouse's job being in current country
___
___
___
___
___
J. Better educational opportunities for children /
want children to continue their education
___
___
___
___
___
K. Need to finish or continue with current project
___
___
___
___
___
L. Other reason, please specify below:
___
___
___
___
___
______________________________________
280
INTERACTIONS WITH TURKISH AND NON-TURKISH COLLEAGUES
AND MEMBERSHIPS IN ORGANIZATIONS RELATING TO PROFESSION
35. a) Are you a member of any professional, cultural or alumni associations /
societies?
[ ] Yes
[ ] No
If you are not a member of any associations or societies, please go to question 36.
b) If so, how many associations are you a member of?
i. Turkish associations located in Turkey
(e.g., Genç Yönetici ve adamları Derne i in stanbul)
______
ii. Turkish associations located in your current country of residence
______
(e.g., Turkish Canadian Business Council or ITU Alumni Assoc.- Canada
if you are living in Canada):
iii. National or local associations in your current country of residence ______
(e.g., Manitoba Association of Architects, American Dental Association):
iv. International or regional associations ______
(e.g., International Association of Agricultural Economists, European Association
of Archeologists)
c) In the past year (January 1, 2001 - Dec. 31, 2001) did you attend or
participate in any of the activities (meeting, conference, fundraiser,
dinner, etc.) organized by these associations?
_____
i. Turkish associations in Turkey:
ii. Turkish associations in current country:
_____
iii. National / local associations in current country:
iv. International or regional associations:
_____
_____
36. a) Do you have any patented inventions?
[ ] Yes
[ ] No
If you do not have any patented inventions, please go to question 37.
b) If so, how many patented inventions do you have? _______
For how many of these inventions are the patents owned:
i. by you?
_____
ii. by firms or universities in your current country of residence? _____
iii. by firms or universities in Turkey? _____
iv. by firms or universities in another country? _____
281
c) How many patented inventions do you have where you are the sole
_____________
inventor?
Of these, how many are the product of research or experiments you
have undertaken mostly:
i. in Turkey? _____
ii. in your current country of residence? _____
iii. in another country? _____
d) For how many of your patented inventions were you part of a team of
inventors that included colleagues of Turkish origin?
_____________
Of these, how many are the product of research or experiments you
have undertaken mostly:
i. in Turkey? ____
ii. in your current country of residence? ____
iii. in another country? _____
37. a) How many of your studies have been published (in journals, books,
reports, etc.) within the past two years (Jan. 1, 2000 - Dec. 31, 2001)?
i. total: _____
ii. with you as the sole author: _____
iii. written with Turkish colleagues: _____
iv. written with non-Turkish origin colleagues:
_____
How many of these were published in Turkish journals or by Turkish
publishers? ___________________
b) How many ongoing projects or studies are you currently involved in?
i. total: _____
ii. by yourself: _____
iii. with Turkish colleagues residing in Turkey: _____
iv. with Turkish colleagues residing in current country: _____
v. with Turkish colleagues residing in other countries: _____
vi. with non-Turkish origin colleagues: _____
c) What percent (%) of your studies do you believe contributes to:
Please indicate the percentage in the squared brackets.
i. the knowledge stock of Turkey: [ ] %
ii. the knowledge stock of your current country of residence: [ ] %
iii. the universal stock of knowledge: [ ] %
You can write down your thoughts about this question below:
____________________________________________________________
282
38. What type of positive contribution(s) do you think your stay abroad is
making or has made to Turkey? Please mark all that apply.
[
[
[
[
[
[
]
]
]
]
]
]
Helped Turkish students find scholarships abroad
Participated in lobbying activities on behalf of Turkey
Helped increase business contacts with Turkey
Helped increase knowledge about Turkey in general
Made donations to Turkish organizations
Helped increase professional contacts between colleagues in my current
country and colleagues in Turkey
[ ] Helped transfer knowledge gained in my current country of residence to
colleagues in Turkey (e.g., by presenting papers in conferences or teaching
in Turkey)
[ ] other, please specify:
_______________________________________
OTHER INFORMATION
39. Please indicate your marital status:
[
[
[
[
]
]
]
]
married, spouse with me
married, spouse away
never married
divorced / widowed / separated
If you marked either 'never married' or 'divorced / widowed / separated', please go to
question 41.
40. Please indicate your spouse's:
__________
a) Age:
b) Nationality:
[ ] Turkish
[ ] other
[ ] dual citizen (Turkish and other)
c) Education level:
[
[
[
[
[
[
[
]
]
]
]
]
]
]
less than primary
primary school
middle school
high school
bachelor’s or equivalent
master’s or equivalent
doctorate
d) Occupation:
_______________
e) Employment status:
[ ] not employed
[ ] employed full time
[ ] employed part time
283
41. Indicate the number of children living with you as part of your family in
the following age categories.
under 2 years
between 2-5 years
between 6-11 years
between 12-17 years
18 and over
_____
_____
_____
_____
_____
42. Please indicate your
a) mother's education level:
[
[
[
[
[
[
[
[
]
]
]
]
]
]
]
]
less than primary
primary school
middle school
high school
bachelor’s or equivalent
master’s or equivalent
doctorate
don’t know
b) mother’s occupation:
__________________
c) father’s education level:
[
[
[
[
[
[
[
[
]
]
]
]
]
]
]
]
less than primary
primary school
middle school
high school
bachelor’s or equivalent
master’s or equivalent
doctorate
don’t know
d) father’s occupation:
__________________
43. a) How many of your family** are living in Turkey? ______
**e.g., mother, father, sibling, spouse, children, or any other family member
who is close to you.
b) How many of your relatives are living abroad?
______
c) How many of your relatives are living in your current country of
residence?
_______
44. a) How do you maintain contact with family members in Turkey?
Please mark all that apply.
[
[
[
[
[
[
]
]
]
]
]
]
telephone calls
regular mail
email
visits to Turkey
visits by family
other, please specify: ___________________________________________
284
b) Which has been your most frequent means of contact? _________
c) Has your contact with family members in Turkey increased, decreased
or remained the same over time?
[
[
[
[
]
]
]
]
increased
decreased
stayed the same
not applicable
Reason:
___________________________________________________
45. a) Do you currently subscribe to any Turkish publications?
[ ] yes
[ ] no
If you do not currently subscribe to any publications in Turkey, go to question 45c.
b) How many Turkish publications do you currently subscribe to?
i) newspapers _________
ii) journals related to your studies _________
ii) other
____________
, please specify: _________________________
c) How frequently do you keep in touch with news from Turkey?
[
[
[
[
[
[
]
]
]
]
]
]
daily
weekly
monthly
once or twice per year
infrequently
not at all
d) How do you keep current with the news from Turkey?
Please mark all that apply.
[
[
[
[
[
[
]
]
]
]
]
]
looking at Turkish internet sites
through visits from family / friends in Turkey
phone conversations with relatives in Turkey
email messages from family/friends in Turkey
through Turkish embassy or cultural attaché
other, please specify: ____________________________________
46. a) Indicate the number of visits you have made to Turkey where the main
reason for your visit was the following:
If you have not made any trips to Turkey during your current stay abroad please go
on to question 47.
A. vacation / family visits: _____
B. participate in conferences or seminars: _____
C. take part in research activities: _____
D. take part in business activities: _____
E. other: _____
Describe other here: _____________________________________________
285
b) When was your last visit to Turkey? month: ______
year: _______
c) How did your last trip to Turkey affect your views about returning to
Turkey?
[
[
[
[
]
]
]
]
increased my likelihood of returning
decreased my likelihood of returning
did not change my views
not applicable
Reason:
_______________________________________________
47. Have the events of September 11, 2001 - the terrorist attacks in the US –
and the aftermath affected your views about returning to Turkey?
[ ] increased my likelihood of returning
[ ] decreased my likelihood of returning
[ ] did not change my views
48. How did you find the length of this survey?
[ ] too long
[ ] too short
[ ] just right
49. Please write down any comments or questions about any part of this
survey in the text box below. We would greatly appreciate receiving your
input.
___________________________________________________________
___________________________________________________________
___________________________________________________________
___________________________________________________________
___________________________________________________________
Thank you for taking part in our survey!
Prof. Dr. Aysıt Tansel
Research Assistant Nil Demet Güngör
Middle East Technical University
FEAS Department of Economics
survey@metu.edu.tr
286
C.4 English Mail-out Version of Turkish Students Abroad Survey
Turkish Brain Drain Student Survey
1) Please write your name and e-mail address in the boxes provided below. This
information is for our record-keeping only; it will not be used in our study, and it will
not be disclosed in any way to other parties. The information you provide will be used
for research or statistical purposes only.
2) Please read and answer carefully. The survey will take approximately 15-20
minutes. Since not all of the questions will apply, you will be able skip those that are
not relevant to you.
3) Please place an X within the square bracket of the appropriate selection for
multiple choice questions.
Thank you again for taking the time to participate in our study.
YOUR NAME:
_____________________________
YOUR E-MAIL ADDRESS:
_____________________________
GENERAL INFORMATION
1. Personal Information: Please indicate your
a) Gender:
Please mark the appropriate box with an X.
[ ] Male
[ ] Female
b) Birthyear:
_____________
c) Birthplace:
city:
country:
_______________
_______________
2. a) What is your current country of residence?
Please mark the appropriate box with an X.
[ ] Australia
[ ] Canada
[ ] England
287
[ ] New Zealand
[ ] USA
[ ] other, please indicate:
_______________
b) Please indicate your current city and (if applicable) state or province of
residence:
city:
___________
state / province:
___________
3. How long have you been staying in your current COUNTRY of residence?
____________
number of years
EDUCATIONAL INFORMATION
4. a) What is the highest degree you hold?
[
[
[
[
[
[
[
]
]
]
]
]
]
]
high school certificate
associates degree (e.g. 2 year program)
bachelor’s (BA / BS)
post baccalaureate certificate
master’s degree (MA / MS / MBA)
post master’s certificate
doctorate (e.g., Ph.D., Ed.D., D.Sc.)
b) In which country did you receive your highest degree?
[
[
[
[
[
[
[
]
]
]
]
]
]
]
Australia
Canada
England
New Zealand
United States
Turkey
other, please specify: ___________________-
c) What is the highest degree that you plan to receive?
[
[
[
[
[
[
[
]
]
]
]
]
]
]
high school certificate
associates degree (e.g. 2 year program)
bachelor’s (BA / BS)
post baccalaureate certificate
master’s degree (MA / MS / MBA)
post master’s certificate
doctorate (e.g., Ph.D., Ed.D., D.Sc.)
5. a) Which high school (lycée) did you graduate from?
Please indicate the name of the high school and its location.
NAME:
____________
LOCATION: ____________
(e.g. Ankara)
288
b) What was the language of instruction of the
i. science courses at your high school (lycée)?
Please mark the appropriate box with an X.
[
[
[
[
[
]
]
]
]
]
Turkish
English
German
French
other, please specify: ______________
ii. social courses (e.g., history, geography, philosophy) at your high school
(lycée)? Please mark the appropriate box with an X.
[
[
[
[
[
]
]
]
]
]
Turkish
English
German
French
other, please specify: ______________
6. If the highest degree you hold is a 'high school certificate', please go on to
question 7.
a) From which university did you receive your undergraduate (bachelor's or
associate's) degree?
_________________________________________
b) What was your major?
__________________________
c) What year did you graduate?
___________________________
d) In how many years did you complete your undergraduate studies? _______
e) What was your CGPA (cumulative grade point average) at the end of your
undergraduate studies?
Please indicate the scale as well: e.g., 3.2/4.0 or 6.1/10
__________
7. What type of program are you currently enrolled in abroad?
[ ] student exchange program
[ ] visiting student / scholar program (e.g., you are a TÜBA or TÜB TAK
scholarship recipient enrolled in a Turkish university and completing part of
your program requirements abroad)
[ ] intensive language program (as prerequisite for continuing with undergraduate
or graduate studies abroad)
[ ] associate's degree program
[ ] bachelor's degree program
[ ] post baccalaureate certificate program
[ ] master's degree program
[ ] post master's certificate program
[ ] doctoral degree program, course work not yet completed
[ ] doctoral degree program, course work completed
[ ] postdoctoral fellow
[ ] other, please specify: ____________________________-
289
8. If you are an exchange student or a visiting student / scholar, please
answer the following questions. Others please go on to question 9.
a) From which university will you be receiving your degree?
_________________________________________________________________
b) What degree will you be receiving from this university?
[
[
[
[
]
]
]
]
bachelor's degree
master's degree
doctorate degree
other, please specify: _____________________
c) What type of activities are you involved in at the university or research center
you are currently visiting? Please check all that apply.
[
[
[
[
[
[
]
]
]
]
]
]
lab work / experiments
participating in seminars
attending courses
giving lectures
independent research activities
other, please specify: _____________________
d) Do you plan to get a separate degree or certificate from the university /
research center you are visiting?
[ ] yes
[ ] no
e) If so, which degree or certificate do you plan to receive?
[
[
[
[
[
]
]
]
]
]
bachelor's degree
master's degree
doctorate degree
other, please specify:
not applicable
9. a) What is the name of the institution (university / research center ) you are
currently attending abroad? (or will be attending after you complete your
language program)
_____________________________________________
b) What was your field of study?
Please be specific and indicate any areas of specialization as well.
General field of study: ___________________
Specialization 1: ______________
Specialization 2: ______________
Specialization 3: ______________
290
c) If you will be receiving a degree or certificate from this institution,
please answer the following questions. If you will not be receiving any degree
from this institution, please go on to question 10.
i) When did you start the program? Please include any compulsory
language training that formed part of the degree requirement.
MONTH ________
YEAR ________
ii) When do you expect to receive your degree?
MONTH ________
YEAR ________
iii) Were you required to take part in an intensive language training
program prior to being accepted into the degree program?
[
[
[
[
]
]
]
]
Yes
No
I am currently enrolled in a language program.
not applicable
10.a) Which of the following factors played a significant part in your decision
to choose your current university or research center for studying
abroad. Please check all that apply.
[
[
[
[
[
[
[
[
]
]
]
]
]
]
]
]
A. provided the most relevant program for my field of specialization
B. provided the best scholarship or financial support
C. having Turkish contacts at the institution
D. recommended by advisor or other professors
E. greater job opportunities
F. being with or near spouse
G. able to get acceptance
H. other, please specify: _____________________________
b) Which was the most important factor? _______________
11. Which source(s) of financial support do you or (did you) have available to
you for your current studies abroad?
Please check all that apply. (To uncheck click on the box again.)
[
[
[
[
[
[
[
[
]
]
]
]
]
]
]
]
savings or support from family
part-time job (university)
part-time job (private sector)
part-time job (public sector)
teaching or research assistant salary
YÖK (Yüksek Ö renim Kurumu) scholarship
MEB (Milli E itim Bakanlı ı) scholarship
TÜBA or TÜB TAK scholarship
291
[
[
[
[
[
]
]
]
]
]
other national scholarship or support (including private sector)
financial support from current university
Fulbright scholarship
international scholarship or support
other, please specify: _________________________
12. a) Do you intend to go on to the next level of studies immediately after
receiving your degree or certificate? i.e., continue with the master's program
after receiving your bachelor's degree, or go on to do a postdoc after receiving your
Ph.D., etc.
[ ] Yes
[ ] No
[ ] I am not sure
[ ] not applicable
b) If yes, in which city / country are you most likely to continue your
studies?
CITY: _______________ COUNTRY: _______________
13. The following questions are about your living arrangements abroad.
a) Which of the following best describes your living
accommodations abroad?
[
[
[
[
[
[
]
]
]
]
]
]
dormitory
house
room in a house
apartment
room in an apartment
other, please specify:
__________________________
b) Are you living on or off campus?
[ ] on campus
[ ] off campus
14. a) Which of the following factors were important in helping you adjust to
life abroad? Please check all that apply.
[
[
[
[
[
[
[
[
[
]
]
]
]
]
]
]
]
]
A. having previous experience abroad
B. the passage of time
C. support from the Turkish Student Association (TSA) at my institution
D. having spouse or other loved one with me
E. having cultural attaché/embassy support
F. having Turkish friends/colleagues at my university/college/research center
G. existence of a large Turkish community in my city
H. being able to share experiences, ask for advise via Turkish internet network
I. other, please specify: ___________________________________
292
b) Which has been the most important factor in helping you adjust?
________________________________________________
c) Are you a member of the Turkish Students Association (TSA) at your
institution?
[ ] yes
[ ] no
[ ] There is no TSA at my institution
Why or why not? ___________________________________________
15. a) Did you have any study, work, travel or other experience outside
Turkey prior to coming to your current country of residence?
[ ] yes
[ ] no
If you have no prior experience abroad then go on to question 16.
b) What kind of previous experience did you have abroad?
Please select all that apply.
[
[
[
[
]
]
]
]
study
work
travel
other, please specify: ________________________
c) What is the longest period you have spent outside Turkey not
counting your current stay? _____________
JOB SEARCH / WORK RELATED INFORMATION
16. In which country do you think you will be working immediately after
completing your studies?
[
[
[
[
]
]
]
]
Turkey
USA
another country, please specify: _____________________
I do not plan to work
If you do not plan to work, please go to question 20.
293
17. What type of organization will you most likely be working for?
a) SOON after completing your studies:
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
]
]
]
]
]
]
]
]
]
]
]
]
]
]
]
]
University (private)
University (public)
College / technical institute (private)
College / technical institute (public)
Pre, primary, secondary school (private)
Pre, primary, secondary school (public)
Government department
Government owned corporation
Multinational corporation
Other private sector organization
Self-employed in incorp. business, practice, farm
Self-employed in nonincorp. business, practice, farm
International organization
Non-profit organization
Armed forces
not sure
b) 5 YEARS AFTER completing your studies:
[Same choices as above]
18. What type of activities will you most likely be doing at work?
a) SOON after completing your studies:
[
[
[
[
[
[
[
[
[
[
[
[
]
]
]
]
]
]
]
]
]
]
]
]
Teaching
Applied research (gaining knowledge to meet a specific need)
Basic research (gaining knowledge for its own sake)
Development (transforming knowledge from research into production)
Computer use, programming, system development
Administrative activities, supervision
Professional services (medical, legal, financial, etc.)
Performing arts, visual and related arts
Quality control, production management
Accounting, contracts
Marketing, consumer services, public relations
not sure
b) 5 YEARS AFTER completing your graduate studies:
[Same choices as above]
19. a) During your current stay abroad did you apply to any firms /
organizations for jobs in Turkey or other countries?
[ ] Yes
[ ] No
If you did not apply for any jobs, please go to question 19d.
294
b) In which countries are the firms and organizations that you applied to
for jobs located? Please select all that apply.
[
[
[
[
[
[
[
]
]
]
]
]
]
]
Turkey
Australia
Canada
England
New Zealand
United States
other, please specify: __________________________
c) What were your reasons for applying?
[ ] To find a full time job that is directly related to my career or education
[ ] To find a full time job (which may not be related directly to my education)
after I graduate
[ ] To find a part time job to cover my education or other expenses
(e.g., university bookstore, library, shop)
[ ] To make extra money during the summer months
[ ] To gain work experience in my field during the summer months
[ ] other, please specify: __________________
d) During your current stay abroad did you receive any job offers from
firms / organizations in Turkey or other countries?
[ ] Yes
[ ] No
If not, please go on to question 20.
e) Were these job offers directly related to your education /training
abroad?
i) job offers from Turkey:
[
[
[
[
]
]
]
]
most are directly related
most are somewhat related
most are unrelated
not applicable
ii) job offers from other countries:
[
[
[
[
]
]
]
]
most are directly related
most are somewhat related
most are unrelated
not applicable
f) From which channels did you seek jobs or receive job offers?
Please select all that apply.
[ ] Direct contacts initiated with firm / organization (e.g., sending unsolicited CV)
[ ] Professional recruiters (e.g., "headhunters")
[ ] 'Career Days' held at Turkish universities
295
[
[
[
[
[
[
[
]
]
]
]
]
]
]
Informal channels (e.g., friends, colleagues)
Ads in professional journals
Turkish internet network (e.g., alumni networks)
Newspaper ads
Placement office at university
Faculty or advisors
other, please specify: ________________________
QUESTIONS RELATING TO THE DECISIONS TO
LEAVE, STAY AND RETURN
20. a) What were your main reasons for going to the country you are
currently staying? Please mark all that apply.
[ ] A. To learn a new language / improve language skills
[ ] B. In need of change / want to experience a new culture
[ ] C. Education or experience in another country is required
by employers in Turkey
[ ] D. Could not find a job in Turkey
[ ] E. No program in my specialization in Turkey
[ ] F. Insufficient facilities, lack of necessary equipment
to carry out research in Turkey
[ ] G. In order to take advantage of the prestige
and advantages associated with study abroad
[ ] H. Preference for the lifestyle in my current country of residence.
[ ] I. To be with spouse or loved one
[ ] J. To provide a better environment for children
[ ] K. To get away from the political environment in Turkey
[ ] L. other, please specify: ____________________
b) Which of the above was the most important reason?
___________________________
21. a) In general, how supportive was your family (e.g. father, mother,
spouse) in your decision to go abroad to work or study?
Please mark the appropriate box with an X.
[
[
[
[
[
]
]
]
]
]
very supportive
somewhat supportive
not very supportive
not at all supportive
not applicable
b) Do you think your family in Turkey would support (or supports) your
decision to settle permanently outside Turkey?
Please mark the appropriate box with an X.
[
[
[
[
[
[
[
]
]
]
]
]
]
]
They would definitely support me.
They would most likely support me.
Some family members would support me, others would not.
They are not likely to be very supportive.
They would actively discourage me.
I am not sure.
not applicable
296
22. Before you left Turkey, what were your thoughts about returning?
Please mark the appropriate box with an X.
[ ] I thought that I would definitely return.
[ ] I was undecided about returning; I would wait and see.
[ ] I did not think that I would return.
23. What are your thoughts about returning to Turkey now?
Please mark the appropriate box with an X.
[
[
[
[
[
[
]
]
]
]
]
]
I will return as soon as possible without completing my studies.
I will return immediately after completing my studies.
I will definitely return but not soon after completing my studies.
I will probably return.
I don't think that I will be returning.
I will definitely not return.
If you marked one of the last two options ('not return') please go to question 26.
24. When do you think you will be returning to Turkey?
Please mark the appropriate box with an X.
[
[
[
[
[
[
[
]
]
]
]
]
]
]
within 6 months
6 to 12 months
1 to 2 years
2 to 5 years
5 to 10 years
more than 10 years
not applicable
25. a) What are your main reasons for returning to Turkey?
Please mark all that apply.
[ ] to complete compulsory military service
[ ] to complete university service (e.g., YÖK, TÜBA scholarship recipients)
[ ] I will return when my permitted time for working abroad ends
(e.g. I am a visiting scholar)
[ ] I miss my family in Turkey
[ ] I want my children to continue their education in Turkey
[ ] after achieving specific goals (gaining work experience, completing research
project) I want to apply what I have learned in Turkey
[ ] I will return after reaching my savings goal
[ ] I will return after reaching my career goal
[ ] I received a job offer from a firm or institution in Turkey
[ ] I want to spend my retirement in Turkey.
[ ] I don't feel safe in my current environment
[ ] other, please specify:
________________________________________
b) After you return, do you plan to go abroad again?
Please mark the appropriate box with an X.
[ ] No
[ ] Yes, for a few days to several weeks at most
[ ] Yes, for 1-3 months at most
297
[
[
[
[
]
]
]
]
Yes, for 4-6 months at most
Yes, for 7-12 months at most
Yes, for 1-2 years at most
Yes, could be longer than 2 years but I believe I will
definitely return to Turkey.
[ ] Yes, to settle down permanently
[ ] not applicable
26. In general, how does your life in your current country of residence
compare with your life in Turkey?
a) work environment (e.g. your job satisfaction):
[
[
[
[
[
[
]
]
]
]
]
]
much better
better
neither better or worse
worse
much worse
don’t know
b) social aspects (e.g. friendships, social relations):
[
[
[
[
[
[
]
]
]
]
]
]
much better
better
neither better or worse
worse
much worse
don’t know
c) standard of living:
[
[
[
[
[
[
]
]
]
]
]
]
much better
better
neither better or worse
worse
much worse
don’t know
27. a) What are the main difficulties that you have faced / are facing living in
your current country of residence? Please mark all that apply.
[
[
[
[
[
[
[
[
[
[
[
[
[
]
]
]
]
]
]
]
]
]
]
]
]
]
A. Being away from family
B. Children growing up in a different culture
C. Loneliness, not being able to adjust
D. Fast-paced life
E. Little or no leisure time
F. Unemployment
G. No jobs in my area of specialty
H. Discrimination against foreigners
I. Lower income compared to the income I had in Turkey
J. Higher taxes
K. Crime, lack of personal security
L. High cost of living
M. Other, please specify: ___________________________________
b) Which of the above factors do you consider to be the most difficult
for you? __________________________
298
28. What are the greatest difficulties RELATING TO TURKEY that may cause
you NOT to return? Please indicate how important for you the following
factors are in this decision.
Please answer even if you have indicated that you will definitely return.
REASON
Very
Somewhat Not
Not at all
Important Important Important Important Important
5
4
3
2
1
A. Low income in my occupation
___
___
___
___
___
B. Little opportunity for advancement
in my occupation
___
___
___
___
___
C. Limited job opportunities in my field of
expertise
___
___
___
___
___
D. No opportunity for advanced
training in my field
___
___
___
___
___
E. Being far from important research
centers and as a result from new
advances
___
___
___
___
___
F. Lack of financial resources and
opportunities to start up my business
___
___
___
___
___
G. Less than satisfying social and
cultural life
___
___
___
___
___
H. Bureaucracy, inefficiencies in
organizations
___
___
___
___
___
I. Political pressures, discord
___
___
___
___
___
J. Lack of social security
___
___
___
___
___
K. Economic instability, uncertainty
___
___
___
___
___
L. Other reason, please indicate below:
___
___
___
___
___
______________________________
299
29. Please indicate the relative importance FOR YOU of each of the following
factors relating to your CURRENT COUNTRY OF RESIDENCE in deciding not
to return or postpone returning to Turkey.
Please answer even if you have indicated that you will definitely return.
REASON
Very
Somewhat Not
Not at all
Important Important Important Important Important
5
4
3
2
1
A. Higher salary or wage
___
___
___
___
___
B. Greater opportunity to advance
in profession
___
___
___
___
___
C. Better work environment
(flexible work hours, relaxed setting, etc.)
___
___
___
___
___
D. Greater job availability in my area
of specialization
___
___
___
___
___
E. Greater opportunity for further
development in area of specialty
___
___
___
___
___
F. A more organized and ordered life
in general
___
___
___
___
___
G. More satisfying social and cultural life
___
___
___
___
___
H. Proximity to important research
and innovation centers
___
___
___
___
___
I. Spouse's preference to stay or
spouse's job being in current country
___
___
___
___
___
J. Better educational opportunities for children /
want children to continue their education
___
___
___
___
___
K. Need to finish or continue with current project
___
___
___
___
___
L. Other reason, please specify below:
___
___
___
___
___
______________________________________
300
OTHER INFORMATION
30. Please indicate your marital status:
[
[
[
[
]
]
]
]
married, spouse with me
married, spouse away
never married
divorced / widowed / separated
If you marked either 'never married' or 'divorced / widowed / separated', please go to
question 32.
31. Please indicate your spouse's:
__________
a) Age:
b) Nationality:
[ ] Turkish
[ ] other
[ ] dual citizen (Turkish and other)
c) Education level:
[
[
[
[
[
[
[
]
]
]
]
]
]
]
less than primary
primary school
middle school
high school
bachelor’s or equivalent
master’s or equivalent
doctorate
d) Occupation:
_______________
e) Employment status:
[ ] not employed
[ ] employed full time
[ ] employed part time
32. Indicate the number of children living with you as part of your family in
the following age categories.
under 2 years
between 2-5 years
between 6-11 years
between 12-17 years
18 and over
_____
_____
_____
_____
_____
33. Please indicate your
a) mother's education level:
[
[
[
[
[
[
]
]
]
]
]
]
less than primary
primary school
middle school
high school
bachelor’s or equivalent
master’s or equivalent
301
[ ] doctorate
[ ] don’t know
b) mother’s occupation:
__________________
c) father’s education level:
[
[
[
[
[
[
[
[
]
]
]
]
]
]
]
]
less than primary
primary school
middle school
high school
bachelor’s or equivalent
master’s or equivalent
doctorate
don’t know
d) father’s occupation:
__________________
34. a) How many of your family** are living in Turkey? ______
**e.g., mother, father, sibling, spouse, children, or any other family member
who is close to you.
b) How many of your relatives are living abroad?
______
c) How many of your relatives are living in your current country of
residence?
_______
35. a) How do you maintain contact with family members in Turkey?
Please mark all that apply.
[
[
[
[
[
[
]
]
]
]
]
]
telephone calls
regular mail
email
visits to Turkey
visits by family
other, please specify: ___________________________________________
b) Which has been your most frequent means of contact? ____________
c) Has your contact with family members in Turkey increased, decreased
or remained the same over time?
[
[
[
[
]
]
]
]
increased
decreased
stayed the same
not applicable
Reason:
___________________________________________________
36. a) Do you currently subscribe to any Turkish publications?
[ ] yes
[ ] no
If you do not currently subscribe to any publications in Turkey, go to question 45c.
302
b) How many Turkish publications do you currently subscribe to?
i) newspapers _________
ii) journals related to your studies _________
ii) other
____________
, please specify: __________________________
c) How frequently do you keep in touch with news from Turkey?
[
[
[
[
[
[
]
]
]
]
]
]
daily
weekly
monthly
once or twice per year
infrequently
not at all
d) How do you keep current with the news from Turkey?
Please mark all that apply.
[
[
[
[
[
[
]
]
]
]
]
]
looking at Turkish internet sites
through visits from family / friends in Turkey
phone conversations with relatives in Turkey
email messages from family/friends in Turkey
through Turkish embassy or cultural attaché
other, please specify below:
_______________________________________________________
37. a) Indicate the number of visits you have made to Turkey where the main
reason for your visit was the following:
If you have not made any trips to Turkey during your current stay abroad please go
on to question 38.
A. vacation / family visits: _____
B. participate in conferences or seminars: _____
C. take part in research activities: _____
D. take part in business activities: _____
E. other: _____
Describe other here: _____________________________________________
b) When was your last visit to Turkey?
month: ________
year: ____________
c) How did your last trip to Turkey affect your views about returning to
Turkey?
[
[
[
[
]
]
]
]
increased my likelihood of returning
decreased my likelihood of returning
did not change my views
not applicable
Reason:
_______________________________________________
303
38. Have the events of September 11, 2001 - the terrorist attacks in the US –
and the aftermath affected your views about returning to Turkey?
[ ] increased my likelihood of returning
[ ] decreased my likelihood of returning
[ ] did not change my views
39. How did you find the length of this survey?
[ ] too long
[ ] too short
[ ] just right
40. Please write down any comments or questions about any part of this
survey in the text box below. We would greatly appreciate receiving your
input.
___________________________________________________________
___________________________________________________________
___________________________________________________________
___________________________________________________________
___________________________________________________________
Thank you for taking part in our survey!
Prof. Dr. Aysıt Tansel
Research Assistant Nil Demet Güngör
Middle East Technical University
FEAS Department of Economics
survey@metu.edu.tr
304
APPENDIX D
TURKISH SUMMARY
Çalı mada, yüksek e itimli i gücü göçü kalkınmakta olan ülkeler açısından
irdelenmektedir. Geli mekte olan ülkelerden geli mi ülkelere gerçekle en nitelikli i gücü
göçü, geli en ülkeler açısından yüksek maliyetli bir hibe olarak nitelendirilebilir.
Çalı manın ilk bölümünde bu göçün göç veren ülkeler üzerindeki etkisini tartı an yazın ele
alınarak tartı mada ula ılan son noktanın ortaya konulması amaçlanmaktadır. Çalı manın
di er amacı, Türkiye’den yurt dı ına gerçekle en nitelikli insan göçünü belirleyen etmenleri
inceleyerek, bu göçte en etkili olanları belirlemektir. Türkiye’den yurt dı ına nitelikli
i gücü göçü özellikle son dönemlerde pe pe e ya anan ekonomik krizlerden sonra daha da
önem kazanmı tır, çünkü ekonomik krizlerin ardından e itimli gençlerde i sizlik önemli bir
ölçüde artmı tır.
Nitelikli i gücü göçüne iktisadî açıdan bakan modellerde nitelikli i gücü hareketleri
fizikî sermaye hareketleriyle benzer bir ekilde ele alınır. Buna göre, yüksek e itimli
ki ilerin daima kendilerine daha yüksek getiri sa layaca ı bölgelere ve ülkelere do ru
hareket etti i görü ü benimsenir. Neoklâsik kurama göre, geli mi ülkeler ve geli en
ülkeler arasında olu an gelir farkları bu ülkelerdeki yapısal i gücü talep-arz açıklarından
kaynaklanır. Nitelikli i gücüne daha fazla gereksinim duyan geli mi
ülkeler, talep
fazlalarını geli en ülkelerden göç alarak kar ılarlar. Geli mi ülkelerdeki nitelikli i gücü
sayılarının artmasıyla nitelikli i gücünün marjinal verimlili inin ve getirisinin dü mesi
beklenir. Öte yandan, göç veren geli mekte olan ülkelerde ise yüksek e itimli insanların
sayıları azaldıkça kalan nitelikli i gücünün marjinal getirisinde ve verimlili inde artı
beklenir. Neoklâsik yakla ım, gerçekle en bölgesel göç hareketlerinin sonucunda ülkeler
arası gelir farklarının kapanmasını öngörür ve buna göre de göç hareketlerin
kısıtlanmayarak tamamen serbest bırakılmasını önerir. Ancak, geli mi ülkelere nitelikli
i gücü göçünün artan sayılarla gerçekle mesi neoklâsik yakla ımın öngördü ü gibi ülkeler
arasındaki gelir farklarının kapanmasına neden olamamı tır. Bunun aksine, bazı
çalı maların bulgularına göre geli mi ve geli en bölgeler arasındaki gelir uçurumu daha da
derinle mi tir.
305
Beyin göçünü açıklamayı amaçlayan kuramsal çalı maların pek ço u iktisadî
nedenleri ön plâna çıkartarak ülkeler arası gelir farklarını en önemli göç nedeni olarak
göstermektedir. Bu çalı malar beyin göçüne neden olan gelir farklarının nasıl ortaya
çıktı ını incelerken genelde gelirlerin verimlili e göre belirlendi i varsayımını benimserler.
Daha çok yurt dı ına ö renim görmek için gidenlerin geri dönmemelerini açıklamak için
geçerli olan bazı yakla ımlarda, göç alan ve göç veren ülkelerde bulunan i verenler
arasındaki bilgi asimetrilerinin gelir farklarına yol açabilece i vurgulanmaktadır. Kwok ve
Leland’ın (1982) çalı masında, geli mi ülkelerdeki i verenler ülkelerine ö renim görmek
için gelen ö rencilerin verimlilikleri ve kabiliyetleri hakkında ö rencilerin ana yurtlarında
bulunan i verenlere göre daha çok bilgiye sahiptirler. Bu yüzden onlara gerçek
verimliliklerini ya da üretime sa ladıkları katkılarını yansıtan gelirleri verebilecek
durumdadırlar. Göç veren ülkelerdeki i verenler ise yurda dönen ö rencilerin verimlilikleri
hakkında aynı bilgiye sahip olmadıkları için onlara ancak daha önceden dönen ö rencilerin
ortalama verimlili ini yansıtan gelirleri verebilirler. Bu durumda gerçek verimlilikleri
ortalama gelirin altında olan ö renciler dönmeyi tercih ederken, en verimli ve en kabiliyetli
ö renciler de yurt dı ında kalmayı tercih eder. Bu yakla ıma getirilen ele tirilerde asimetrik
bilginin ancak kısa vadede geçerli olabilece i, orta ve uzun vadede ise geri dönen
ö rencilerle ilgili bilgi eksiklerinin tamamen yok olaca ı savunulmaktadır.
Di er yakla ımlarda geli mi ve geli mekte olan ülkeler arasındaki sosyal (be eri
ve fiziksel) sermaye farkları önemlidir. Miyagiwa’nın modeline göre yüksek e itimli
ki ilerin birarada toplanması verimliliklerini ve gelirlerini olumlu ekilde etkiler. Beyin
göçünün nedeni nitelikli çalı anların daha verimli ve daha fazla kazanç sa layabilecekleri
nitelikli i gücü sayısının yüksek oldu u ülkelere yönelmeleridir. Wong’un modelinde ise
yurt dı ında çalı mak e itimli ki ilere yurt dı ındaki toplu i tecrübesinden faydalanma
olana ı tanır ve verimliliklerini artırır. Chen ve Su bu konuya farklı bir yakla ım daha
getirirler. Yurt dı ında ö renim görenlerin gördükleri e itim bulundukları ülkenin sermaye
sto uyla daha çok uyumludur. Bu yüzden yurt dı ında e itim görenler yurt dı ında daha
fazla kazanç elde edebilirler. Bu modeller ikinci bölümdeki ampirik çalı manın teorik
çerçevesini olu turmaktadır.
Ampirik analizde kullanılan veriler 2002 senesinin ilk yarısında gerçekle tirilen
anket uygulamasının sonuçlarına dayanmaktadır. Anketin hedef kitlesi yurt dı ında
ö renimlerini sürdüren lisans, yüksek lisans ve doktora ö rencileri ile üniversite e itimli
306
i gücü olarak belirlenmi tir. Buna göre, bu iki gruba ayrı anket soruları da ıtılmı ve
2000’in üzerinde yanıt toplanmı tır. Anketlerden elde edilen verilerle çe itli gitme/kalma
nedenlerinin önemlerini belirlemek amacıyla sıralı probit analizi yapılmı tır. Bu analizin
sonuçları a a ıda yurt dı ında çalı an üniversite e itimli Türkler ve yurt dı ında okuyan
Türk ö renciler için ayrı ayrı verilmi tir.
Yurt Dı ında Çalı anlar: Sıralı Probit Kestirim Sonuçları
Cinsiyet ve Ya Etkileri:
Yurt dı ında çalı an yüksek nitelikli i güçü için Türkiye’ye geri dönme niyeti
cinsiyete göre farklılık göstermektedir. ‘Kadın’ de i keninin katsayısı pozitif ve %1
oranında anlamlıdır. Buna göre, kadınlar erkeklere göre daha kuvvetli yurt dı ında kalma
niyeti belirtmektedirler.
Modelde ‘ya ’ ve ‘ya kare’ de i kenleri katılımcı ya ının geri dönme niyeti
üzerindeki etkisini gösterir. Genç katılımcıların daha ya lı katılımcılara göre dönme
niyetlerinin daha az olması beklenebilir. Bunun nedenlerinden biri gençlerin önündeki
i gücüne katılım süresinin daha uzun olması, ve buna göre de yurt dı ındaki yüksek
gelirden daha uzun süre faydalanma olana ına sahip olmalarıdır (Chen ve Su, 1995). Geri
dönmeme niyetini, ‘ya ’ de i keni artı yönde, ‘ya kare’ de i keni ise eksi yönde ektiledi i
anla ılmaktadır. Katılımcı ya ı arttıkça, geri dönme niyeti azalan hızda azalmaktadır. Di er
bir deyi le, daha ya lı katılımcılar daha kuvvetli geri dönmeme (yurt dı ında kalma) niyeti
belirtmektedir. Bu olgunun nedeni bulunan yerde uzun süre geçirilince, alı kanlıkların
geli mesi ve yerle mesi dolayısı ile dönü ün güçle mesi olabilir. Bazı katılımcılar
ya larının ilerledi i için geri dönmenin zor olaca ını ifade etmi lerdir. Ya de i kenleri
yurt dı ında kalma ve çalı ma süresiyle ili kili oldu undan, modele bu de i kenler dahil
edildi inde ya de i kenleri istatistiksel olarak anlamlı bulunmamaktadır.
Yurt Dı ına Çıkmadan Önceki Niyetlerin Etkileri:
Türkiye’ye geri dönme niyetinde en belirleyici etkenlerden biri yurt dı ına
çıkmadan önce katılımcıların geri dönme konusundaki tutumlarıdır. Yurt dı ına çıkmadan
önceki dönme/dönmeme e ilimlerini ölçmek amacıyla katılımcılara üç kategori içeren bir
soru yöneltilmi tir: “Türkiye'den ayrılmadan önce, Türkiye'ye geri dönme konusundaki
dü ünceniz neydi?”. Kategoriler, “mutlaka geri dönmeyi dü ünüyordum”, “kararsızdım” ve
“kesinlikle geri dönmeyi dü ünmüyordum” seçeneklerinden olu maktadır. Modelde bu
307
e ilimler, “mutlaka geri dönme” kategorisi baz alınarak, kukla de i kenlerle gösterilmi tir.
Her iki kukla de i kenin katsayısı pozitif ve istatistiksel olarak %1 düzeyinde anlamlı
bulunmu tur. Yurt dı ına çıkmadan önce kesin dönmeme niyeti veya dönme konusunda
belirsizlik gösteren katılımcılar kesin geri döneceklerini belirtenlere göre daha kuvvetli geri
dönmeme e ilimi göstermektedirler. Bu sonucu, geri dönmeme konusunda daha kararlı
olanların yurt dı ına intibak etmek ve yurt dı ında ba arılı olmak için daha fazla azim ve
çaba göstermelerine de ba layabiliriz, ve de “kendi kendini do rulayan kehanet” olarak
nitelendirebiliriz. Bazı katılımcıların açıklamaları da bu tür bir yorumu destekler
niteliktedir.
Aile deste inin etkisi:
Aile deste ininin Türkiye’ye geri dönme niyetindeki rolünü ölçmek için ankette iki
soru sorulmu tur. Birinci soru, katılımcıların yurt dı ına ilk çıkma kararlarında gördükleri
aile deste ini belirlemek amacıyla sorulmu tur. Modelde bu deste in derecesini ifade eden
bir ve be arasında de er alan bir de i ken kullanılmı tır. Bu de i ken aile deste i
gördüklerini belirten katılımcılar için daha yüksek de er, aile deste i görmediklerini
belirtenler için daha dü ük de er almaktadır. Birinci aile deste i de i keninin katsayısı
negatif ve %1 oranında anlamlıdır. Daha fazla aile deste i gördüklerini belirten
katılımcıların Türkiye’ye geri dönme niyetleri daha kuvvetlidir. Bu de i ken, aile ba larını
ve dolayısıyla yurta olan ba ları temsil ediyor olabilir.
kinci soruda, ailelerin katılımcıların yurt dı ına yerle meleri konusundaki
tutumları sorulmu tur. Modelde, bu tutumun etkisini göstermek için bir (hiç desteklemez)
ile altı (çok destekler) arasında de er alan bir de i ken kullanılmı tır. Bu de i ken pozitif
ve %1 düzeyinde anlamlı bulunmu tur. Bu sonuç, yurt dı ına yerle me konusunda aile
deste inin önemini göstermektedir.
Anne-Babaların E itim Düzeyleri:
Anne ve babaların e itim düzeyleri modele sosyoekonomik gösterge olarak dahil
edilmi tir. Daha yüksek e itim düzeyleri i gücü piyasalarında daha çok gelir getirdi inden,
e itim düzeyi daha yüksek olan ailelerin çocukları daha fazla e itim olanaklarına
sahiptirler. Anne ve babaların e itim düzeyleri, Türkiye’de kız ve erkek çocukların okulda
eri imini belirleyen en önemli etkenler arasında gösterilmi tir (Tansel, 1999 ve 2002). Bu
göstergenin Türkiye’ye geri dönme niyetinde ne yönde bir etki gösterece i önceden belli
de ildir. Probit analiz sonuçlarına göre, anne-baba e itim de i kenleri istatistiksel olarak
308
anlamlı de ildir. Bu göstergelerin daha çok yurt dı ına çıkmakta etkili oldu unu
dü ünebiliriz. Bu yüzden ço u katılımcının Türkiye ortalamasına göre daha yüksek e itimli
aileden geldi ini görmek a ırtıcı de ildir.
Yurt Dı ında Çalı ma artlarının Etkisi:
Anketi yanıtlayanlardan, çalı tıkları ülkede, içinde bulundukları çalı ma artlarını
(örne in, çalı tıkları i in verdi i tatmini), Türkiye’deki tecrübelerine kıyasla ‘çok daha
kötü’den ‘çok daha iyi’ arasında de i en altı kategoride de erlendirmeleri istenildi.
Kategoriler kukla de i ken olarak modele eklendi. ‘Ne daha iyi, ne daha kötü’ kategorisi
baz kategori olarak seçildi. Probit kestirim sonuçları, tüm kategoriler için alınan sonuçların
istatistiksel olarak anlamsız oldu unu gösteriyor. Öte yandan, ankete katılanların büyük
ço unlu u çalı tıkları ülkedeki çalı ma artlarının Türkiye’dekine kıyasla ‘daha iyi’ veya
‘çok daha iyi’ oldu una inanıyor.
Yurt Dı ında Sosyal Ya amın Etkisi:
Aynı ekilde, anketi yanıtlayanlardan, çalı tıkları ülkedeki sosyal ya amı (örne in,
arkada lıklar, sosyal etkinlikler) Türkiye’deki sosyal ya ama göre de erlendirmeleri
istenildi. Bu de i kende görülen istatistiksel da ılım, ‘daha kötü’ kategorisine do ru
e ilimlidir. ‘Ne çok daha iyi, ne çok daha kötü’ kategorisi baz alınarak, di er kategoriler
kukla de i ken olarak modele konulmu tur. a ırtmayan bir sonuç ise, yurtdı ındaki sosyal
ya amlarını ‘çok daha iyi’ olarak nitelendirenlerin Türkiye’ye geri dönmeme niyetlerinin
baz kategoriye göre daha yüksek olmasıdır. Bu kategori, pozitif ve 5% oranında istatistiksel
olarak anlamlı çıkmı tır.
Çalı anların Yurt Dı ındaki ‘Ya am Standartları’na dair De erlendirmeleri:
Çalı ma artları ve sosyal ya am için yapılan de erlendirme yurt dı ındaki ya am
standardı için de yapılmı tır. Gene a ırtıcı olmayan bir sonuç ‘çok daha iyi’ kategorisinin
pozitif ve %1 düzeyinde anlamlı çıkmasıdır. Baz olarak alınan ‘ne çok daha iyi, ne çok
daha kötü’ kategorisini seçenlere göre ya am standartlarının çok daha iyi oldu unu
dü ünenlerin geri dönmeme niyetleri daha yüksektir.
Yurt Dı ında yeri Tecrübesinin Etkisi:
Yurt dı ında alınan i yeri e itimi (on-the-job training), bu tecrübeye sahip olan
çalı anların maa larının yükselmesi anlamına gelece inden, yabancı i çilerin kendi
ülkelerine dönmeme olasılı ını kuvvetlendirebilir (Chen ve Su, 1995). Yurt dı ındaki i yeri
309
tecrübesinin etkisini do rudan ölçmek için katılımcılara çalı tıkları son kurumda i yeri
e itimi alıp almadıklarını sorduk.
yerinde e itim görenlere aldıkların e itimin çalı tıkları
i yerine mi özgü, çalı tıkları sektöre mi özgü yoksa genel bir e itim mi oldu unu sorduk.
Anketi yanıtlayanların ço u böyle bir tecrübeye sahip olduklarını belirtmi tir. Öte yandan,
probit analizi, ‘i yeri tecrübesi’ni temsil eden de i kenlerin istatistiksel olarak anlamsız
oldu unu gösteriyor.
Yurt dı ı i tecrübesi yurt dı ında çalı ılan yıl sayısı olarak da gösterilebilir.
Katılımcıların bulundukları ülkede çalı tıkları yıl sayısı, pozitif ve %5 oranında istatistiksel
olarak anlamlı çıkmı tır. Bu da, katılımcının yurt dı ı i tecrübesinin artmasıyla geri
dönmeme niyetinin kuvvetlendi i anlamına gelmektedir. Modele dahil edilen ba ka bir
de i ken ise katılımcının Türkiye’de tam zamanlı bir i te çalı madı ı belirten kukla
de i kendir. Bu de i kenin katsayısı positif ve %1 oranında anlamlıdır. Bu sonuç,
Türkiye’de hiç çalı mayan katılımcıların geri dönmeme niyetlerinin, çalı anlara göre daha
kuvvetli oludu u anlamına gelir.
Akademik ve Di er Meslekler:
Anketi yanıtlayanların yakla ık dörtte birini akademisyenler te kil ediyor.
Akademik alanda çalı anların di er meslek gruplarına göre geri dönme niyetlerindeki fark
bir kukla de i kenle ölçülmü tür. Sonuçlar, bu mesle i seçmenin Türkiye’ye dönüp
dönmeme kararında etkili olmadı ına i aret ediyor; kullanılan de i ken istatistiksel olarak
anlamsız çıkmı tır.
AR-GE Çalı maları:
Ankette, çalı ma saatlerinin en az yarısını ara tırma-geli tirme faaliyetlerine
ayırdıklarını belirten elemanlar ‘AR-GE çalı anı’ olarak nitelendirilmi tir.
Anket
sonuçlarına göre, AR-GE çalı anlarının üçte biri akademisyenlerden olu uyor. Öte yandan,
AR-GE faaliyetlerine yurtdı ında daha fazla prim verildi i dü ünülürse, probit analizinde
AR-GE de i keninin istatistiksel olarak anlamsız çıkması beklenmedik bir sonuç olarak
nitelendirilebilir.
Çekici ve tici Etkenler:
Yurt dı ında kazanılan yüksek maa lar, beyin göçünün en önemli nedenlerinden
biri olarak görülmektedir. Yurtdı ında kazanılan yüksek maa ın, Türkiye’ye dönüp
dönmeme kararında ne denli etkili oldu unu ara tırmak için, ankete bu soru dahil edilmi tir
310
ve anketi yanıtlayanlardan, yurtdı ında kazandıkları nispeten yüksek olan maa ları,
Türkiye’ye dönmeme kararında veya dönmeyi ertelemede bir etken olarak, ‘çok önemli’den
‘önemsiz’e kadar be
kategoride de erlendirmeleri istenmi tir. ‘Önemli’ ya da ‘çok
önemlidir’ yanıtını verenlerin ‘bir’, di erlerinin ‘sıfır’ de erini alan bir kukla de i ken
yaratılmı tır. Tablo 2’de görüldü ü gibi, bu de i ken istatistiksel olarak anlamsız çıkmı tır.
Katılımcılar, ankette verilen di er itici ve çekici faktörleri de aynı
ekilde
de erlendirmi lerdir ve bu faktörler modelde kukla de i kenlerle temsil edilmektedir.
Verilen çekici faktörler arasında istatistiksel olarak anlamlı bulunanlar unlardır: daha
düzenli ve sistemli bir ya am olana ı, daha doyurucu kültürel ya am, çocuklarım için daha
iyi e itim olanaklarının bulunması, e in i inin yurt dı ında olması ya da e in yurt dı ında
ya amayı tercih etmesi, ve yurt dı ındaki çalı ılan projenin devam etmesi /
tamamlanmaması. Belirtilen son iki etken %1 oranında anlamlı çıkmı tır; di erleri %5
oranında anlamlıdır. Bu sonuçlar, Türkiye’ye geri dönmeme kararında ailenin önemi
kanıtlamı tır. ‘Yurt dı ında giri ilen projenin devam etmesi’ etkeninin katsayısı negatif
çıkmı tır. Bu da bu nedeni çok önemli olarak belirten katılımcıların projelerini
bitirdiklerinde geri dönme niyetinde olduklarını gösterebilir. Di er çekici etkenlerin
katsayıları beklenildi i gibi pozitifdir.
statistiksel olarak anlamlı çıkan itici de i kenler unlardır: ihtisas alanında daha
ileri seviyede deneyim kazanma olanaklarının azlı ı (%5), i kurmak için gerekli maddî
destek ve finansmanın bulunmaması (%1), ekonomik istikrarsızlık, belirsizlik (%1), ve
‘di er’ kategorisi (%5). Parantez içindeki yüzdeler istatistiksel anlamlılık düzeyini
vermektedir. ‘
kurmak için olanakların azlı ı’ dı ındaki etkenlerin katsayıları pozitifdir.
Geri dönmeme niyetindeki en önemli itici nedenin ekonomik istikrarsızlık ve belirsizlik
oldu u görülmektedir. ‘Di er’ kategorisinin itici neden olarak önemli oldu unu belirtenler,
i yerinde torpil, toplumsal yozla ma, askerlik mecburiyeti gibi nedenler göstermi lerdir.
Ba ımsız De i kenler ve Kestirim Sonuçları: Ö renciler
Bu bölümde yurt dı ında bulunan Türk ö rencilerin geri dönme niyetlerini
belirleyen etkenler incelenmektedir. Pek çok ba ımsız de i ken yurt dı ında çalı an Türk
i gücü analizinde kullanılan de i kenle aynıdır. Bu yüzden, üçüncü bölümdeki
de i kenlerle ilgili açıklamalar bu bölümde de geçerlidir.
311
Cinsiyet ve Ya Etkileri:
Ö renci grubunda geri dönme niyeti cinsiyet ve ya a göre anlamlı bir farklılık
göstermemektedir. Ö renci katılımcıların ya ları çalı anlara göre daha dü ük varyanslı
oldu undan böyle bir sonuç beklenebilir.
Yurt Dı ına Çıkmadan Önceki Niyetlerin Etkileri:
Yurt dı ında yüksek ö renim görenler için gitmeden önceki dönme niyetlerini
belirleyen de i kenlerin katsayıları pozitif ve istatistiksel olarak %1 düzeyinde anlamlı
bulunmu tur. Yurt dı ında çalı anlar analizindeki gibi, yurt dı ına çıkmadan önce kesin
dönmeme niyeti veya dönme konusunda belirsizlik gösteren katılımcılar kesin geri
döneceklerini belirtenlere göre daha kuvvetli geri dönmeme e ilimi göstermektedirler.
Aile deste inin etkisi:
Birinci aile deste i (yurt dı ına ilk çıkı taki aile deste i) de i kenin katsayısı
negatif ve istatistiksel olarak %5 düzeyinde anlamlı çıkmı tır. Ailelerin katılımcıların yurt
dı ına yerle meleri konusundaki tutumun etkisi beklendi i gibidir. Ailenin deste ini
gösteren iki kukla de i ken, ‘çok destek’ ve ‘biraz destek’, pozitif ve %1 ve %5
düzeyinlerinde anlamlıdır. Ailenin yurt dı ına yerle me konusundaki deste i arttıkça,
katılımcının geri dönmeme niyeti de artmaktadır.
Yurt Dı ında Sosyal Ya amın Etkisi:
Çalı anlar anketinde oldu u gibi, ö rencilerin ö renim gördükleri ülkedeki sosyal
ya amı Türkiye’deki sosyal ya ama göre de erlendirmeleri istenildi. ‘Çok daha kötü’
kategorisi baz alınarak, di er kategoriler kukla de i ken olarak modele konulmu tur. ‘Daha
kötü’ kategorisi dı ındaki de i kenlerin katsayıları pozitif ve istatistiksel olarak %1
oranında anlamlı bulunmu tur. Yurt dı ındaki sosyal ya amlarını ‘ne daha iyi, ne daha
kötü’, ‘daha iyi’ veya ‘çok daha iyi’ olarak nitelendirenlerin Türkiye’ye geri dönmeme
niyetleri baz kategoriye göre daha yüksektir.
Ö rencilerin Yurt Dı ındaki ‘Ya am Standartları’na dair De erlendirmeleri:
Ankette, yurtdı ında okuyan ö rencilerden, dı ardaki ya am standartlarını
de erlendirmeleri istenmi tir. Bu soruya verilen yanıtların istatistiksel da ılımı ‘çok daha
iyi’ kategorisine do ru e ilimlidir. Dı ardaki ya am standartlarını Türkiye’dekine göre
‘daha iyi’ ya da ‘çok daha iyi’ olarak de erlendiren ö rencilerin baz alınan di er
312
kategorilere göre yurtdı ında kalma niyetlerinin daha fazla oldu u anla ılmaktadır. Bu
de i kenin katsayısı pozitif ve %1 de anlamlıdır.
Yurt Dı ında Türk Ö renci Birliklerine Üye Olmanın Etkisi:
Anketi yanıtlayan ö rencilerden yarıdan fazlası yurt dı ında okudukları
üniversitelerdeki Türk ö renci birliklerine üye. Probit analizi sonuçlarına göre, yurt dı ında
Türk ö renci birliklerine üye olmanın beyin göçüne etkisi, negatif ve istatistiksel olarak
yüzde bir oranında anlamlı. Bu sonuç, Türk ö renci birli i üyeleri arasında Türkiye’ye
dönme niyetine sahip olanların daha fazla oldu una i aret ediyor. Türk ö renci birliklerine
üye olmanın, Türkiye’ye hissedilen ‘kültürel ba ların’ belki daha güçlü oldu unun bir
göstergesi olarak dü ünülebilir.
Yurt Dı ında kalma süresinin etkisi:
Regresyon sonuçlarına göre, yurtdı ında kalma süresinin Türk beyin göçüne olan
etkisi pozitif ve istatistiksel anlamda yüzde bir oranında anlamlı. Sonuçlara göre,
yurtdı ında kalma süresi uzadıkça, Türkiye’ye dönmeme e ilimi de kuvvetleniyor. Bu
beklenilen bir sonuç, zira yurtdı ında kalma süresinin uzaması, yurtdı ındaki hayata
intibakı güçlendirdi i gibi (örne in, yurtdı ında bir yabancıyla evlenmek), anavatana olan
ba ların zayıflamasına da yol açabiliyor.
Meslek Alanının Etkisi:
Chen ve Sue (1995), daha önce de de indi imiz çalı malarında, tıp, mühendislik ve
i letme gibi ‘capital dependent’ mesleklerde görülen beyin göçünün di er mesleklere göre
daha yo un oldu unu bulmu tur. Chen ve Sue’nun bu çalı malarında kullandıkları
ekonometrik analiz, Türk beyin göçüne yönelik olarak yürüttü ümüz anketten elde
etti imiz verilere uygulanmı tır, fakat sonuçlarımızda, çalı ılan meslek alanının Türk beyin
göçüne olan etkisi istatistiksel olarak anlamsız çıkmı tır.
Çekici ve tici Etkenler:
Ö rencilere yönelik olan ankette, yurtdı ına beyin göçünde önemli rol oynadı ı
dü ünülen 12 ‘çekici’ ve 12 ‘itici’ etken sıralanmı , ve anketi yanıtlayanlardan bu etkenleri,
kendi aldıkları yurtdı ına çıkma kararında ta ıdıkları öneme göre de erlendirmeleri
istenmi tir. Regresyon sonuçlarına göre, Türkiye ile ilgili sıralan itici faktörlerin ço u
istatistiksel olarak anlamsız çıkmı tır. Modelin tanımına göre anlamlı etkenler unlardır:
‘Uzmanlık alanında i olanaklarının azlı ı’ ve ‘di er’ itici nedenler. ‘Di er’ kategorisini
313
i aretleyenler ‘mecburî askerli i ertelemek’, ‘Türkiye’deki yolsuzluklar’ gibi nedenler ileri
sürmü lerdir (%5’de anlamlı). Türk ö renci beyin göçünü kuvvetlendiren yabancı ülkeye
ba lı çekici etkenler unlardır: Yurt dı ında sistemli ve düzenli bir ortamın olması ve e in
yurtdı ında bulunması istatistiksel olarak %1 oranında anlamlı bulunmu tur; Daha yüksek
maa lar ve yurt dı ında henüz bitirilmemi olan bir proje üzerinde çalı mak istatistiksel
olarak %5 oranında anlamlı bulunmu tur.
Sonuç
Probit analiz sonuçları, son dönemde ya anan ekonomik krizin ve siyasi
belirsizli in yurt dı ında çalı anların Türkiye’ye geri dönme niyetlerinde etkileyici rol
oynadı ını kanıtlamı tır. Yurt dı ında ö renim gören ö renciler için geri dönmeme
niyetlerinde çekici faktörlerin daha a ırlıklı oldu u gözükmektedir. Literatürde, yüksek
nitelikli
i gücünün
yurt
dı ına
göç
etmesinde
ekonomik
nedenlerin
önemi
vurgulanmaktadır. Yurt dı ında kazanılan yüksek maa lar, beyin göçünün en önemli
nedenlerinden biri olarak görülmektedir. Çalı mada beklenenin aksine yurt dı ında
çalı anların Türkiye’ye geri dönmeme kararında yurt dı ındaki yüksek gelirler istatistiksel
olarak anlamlı bulunmamı tır. Ö renci grubunda iste gelir farkları beklenildi i gibi önemli
bulunmu tur. Ö rencilerin yurt dı ında kalma kararındaki en önemli çekici faktörlerden biri
yurt dı ındaki sistemli ve düzemli ya am tarzı olmu tur. Yurt dı ında çalı anların
Türkiye’ye geri dönmeme kararındaki en önemli itici nedenlerden biri ise Türkiye’deki
ekonomik ve siyasî istikrarsızlık olmu tur. Analizde, her iki grup için Türkiye’ye geri
dönme veya yurt dı ında kalma kararında gitmeden önceki dönme niyetleri ve ailenin rolü
önemli çıkmı tır. Geri dönmeme niyetinde ya ve cinsiyet farkları, yurt dı ında çalı an
Türkler için önemli bulunmu tur. Ö renci grubunda geri dönme niyeti cinsiyet ve ya a göre
anlamlı bir farklılık göstermemektedir.
Katılımcıların anne ve babalarının e itim düzeylerine bakıldı ında, ebeveynlerin
genelde yüksek tahsilli oldukları görülmektedir; bu da yurt dı ında e itim görme ve çalı ma
fırsatlarının yüksek gelirli ailelerde toplandı ına i aret etmektedir. Çalı mada ortaya çıkan
“fırsat e itsizli i” sonucu di er benzer çalı maların bulgularını desteklemektedir. Katılımcı
ebeveynlerinin e itim düzeylerinin Türkiye ortalamasının üzerinde olması Türkiye’den
gerçekle en nitelikli insan göçünün önemini göstermektedir. Ailelerin genelde çocuklarının
yurt dı ına gitmelerini te vik edip yurt dı ında kalmalarını (daha dü ük oranda olsa da)
desteklemeleri katılımcıların geri dönmeme kararında etkileyici oldu u anla ılmaktadır.
314
VITA
Nil Demet Güngör was born in stanbul on September 26, 1971. She resided in
Ottawa, Ontario, Canada with her family for 14 years. In 1993, she received her bachelor’s
degree (magna cum laude) in economics from the University of Ottawa, where she also
minored in Public Policy and Public Management. In 1994, upon her return to Turkey, she
was accepted into the master’s program in economics at the Middle East Technical
University (METU). After receiving her degree in 1996, she was accepted into the PhD
program in economics at METU. She worked as research assistant in the economics
department and also as editorial assistant for the journal METU Studies in Development for
the duration of her doctoral studies. Her main areas of interest are labor economics,
economics of education, applied econometrics and economic growth and development.
Publications:
Tansel, A. and N. D. Güngör (2003) “‘Brain Drain’ from Turkey: Survey Evidence of
Student Non-Return,” Career Development International, special issue on Career
Development in the Middle East, 8(2), 52-69.
Tansel, A. and N. D. Güngör (1997) “The Educational Attainment of Turkey’s Labor
Force: A Comparison Across Provinces and Over Time,” METU Studies in
Development, 24(4), 531-47.
Güngör, N. D. (1997) “Education and Economic Growth in Turkey 1980-1990: A Panel
Study,” METU Studies in Development, special issue on Education, 24(2), 185-214.
Awards and Scholarships:
2000-2002 Turkish Academy of Sciences Scholarship for Integrated Doctoral Studies in
Turkey and/or Abroad in the Social Sciences and Humanities;
1996 Young Researcher Award (awarded jointly by the Turkish Statistical Association and
the Association of Statistics Graduates).
315