The economic returns from U.S. higher education to foreign students

The economic returns from U.S. higher education to foreign students

67 THE ECONOMIC RETURNS FROM U.S. HIGHER TO FOREIGN STUDENTS Najmul 1. Ho&n EDUCATION and Paul B. Downing’ Introduction The position of the Un...

1009KB Sizes 0 Downloads 44 Views

67 THE ECONOMIC

RETURNS FROM U.S. HIGHER TO FOREIGN STUDENTS

Najmul

1.

Ho&n

EDUCATION

and Paul B. Downing’

Introduction

The position of the United States as the major producer of advanced learning in the western world has for many years influenced foreign students to pursue their higher education in the U.S. Over 260,000 foreign students presently attend U.S. colleges and universities. Economists have done little research on foreign students’ decisions to study abroad; and, more specifically, have not made estimates of the private returns to foreign students of U.S. higher education. Vinod Agarwal, using a human capital theory approach, attempted to test the foreign students’ demand for U.S. higher education.’ But, because of the lack of available data on future expected benefits of U.S. education, he was not able to develop a full capital theory model. Instead, he concentrated on the cost of education and its effect on the flow of foreign students. Work by

1 Najmul Hossain is an Assistant Professor of Economics siana University. Paul B. Downing is a Professor of Economics, Sciences Progrsm at Florida State University.

tation,

at Southeastern LouiDirector of Policy

‘Vinod Aganvai, ” Foreign Students’ Demand for U.S. Higher University of California at Santa Barbara, 1977.

Education,”

dir

Acknowledgements: We wish to thank Bernard La Berge of Virginia Polytechnic Institute and State University; the foreign student advisors at the University of Delaware, the University of Tennessee, and the Florida State University for their help in collection of data; and Roger Faith, Nick Tideman. Art Denzau, and Salim Rashid who provided useful comments.

68

Myer, Glaser, and others explored the sociopolitical factors affecting the decision of foreign students to study and/or migrate abroad.3 They do not estimate the economic returns from education. The objective of this paper is to present estimates of expected present value of pecuniary returns from U.S. education to foreign students in various fields. Following seminal works by T. W. Schultz and Gary Becker, the human capital theory has been widely used to estimate returns from higher education in the U.S. for American students4 Weisbrod, Wilkinson, and Honoch in the 1960’s; Camoy and Marenbach in the 1970’s; and Witmer, and Rumberger in the 1980’s; have used the human capital model to compute the rates of return from formal education.’ However, it was Richard Freeman who extensively explored the various relationships between U.S. college education and the labor

3Robert Myers, Edmcation aud Rmi~mrion SW& Abroad and Humea Resoumes (New York: Penguin Press, 1971); Williun Gluer, - hdgtdcm and Retma (New York: Pergamon Praa, 1978).

Gary

4T. W. S. Becker,

Schultz. ?fmaua

rba Rdi~miion of ‘Ibe Bmim Ihuh

borrtrncnt k ifsman Cupi& (New York: Free Press, Ca@tol @Jew York: Columbii Univc~~ity Press, 1964).

1971);

‘Burton Wchbrod, “Educaioa and Invesuncnt io Human Capital. JounwI of Pditkd Economy (Supplement) Vol. 70 (October 19611, pp. 425436; BNCC Wilkim aon, “Present Value of Lifetime Jhrniogs for Different Occupations,” Journal of Pditicd Ecommy, Vol. 74 (December 1966). pp. 556572; Giora Hmoch, ‘An Econamic Analysis of Earnings and Schooling” Jouml of Human Resowcar, Vol. 2 (1%7), pp. 310329; Martin Carnoy and Dieter Marenbach, “The Returns to Schooling in tbe U.S., 1939-1969,” Journal of Humun Reronnm, Vol. 10 (1975), pp. 312-31; Richard Preunm, “Ovcrinvesunent in College Train@?” Joamul of Huma Reroums, Vol. 10, No. 3 (Summer 1975), pp. 287-311; D. R. Witmcr, “Has the Golden Age of Amcrhn H@ba Education Cane to aa Abrupt End?” Jmmnal of Humm Rewurces, Vol. 15 (Winter 1980). pp. 113-119; R. W. Rumbcrger, ‘The Economic Decline of College Graduates: Pact or Pallacy,” Joumd of Hunnan Resosmxs, Vol. 15 0hter 19801, pp. 99-l 12.

69

market.6 These studies show that the American students’ returns from college education range from 5 to 20 percent. At the international level, country-specific rates of return from education have been computed for various developing countries. Carnoy on Mexico, Psacharopoulos’ study on Greece, and Blaug’s work on Thailand are some of the notable ones7 The novel feature of our study is that we obtained data on the returns foreign students expect from study at U.S. universities. We have derived this data from an extensive survey of foreign born students currently enrolled in several universities. To the best of our knowledge, ours is the first economic survey on the returns to foreign students from studying in the U.S. Our empirical findings on th expected returns from U.S. education will aid in explaining the large stock of foreign students and their choices of fields of study. This paper is presented in three sections. In the first we explain the general methodology used to develop our present value (PV) estimates, outline the specific methods of calculating costs and expected earnings, and explore the limitations of this methodology. In the second section we present our estimates of the PV of the U.S. education, according to field of study. Section III summarizes our findings. An appendix discussing the methodology that was used in calculating the present values of U.S. higher education is also included.

Harvard

‘Richard University

Freeman, Market Press, 1971).

for

College-Trained

Manpow

(Cambridge,

7Martin Cnrnoy, “The Costs of Returns to Schooling in Mexico: dissertation, University of Chicago, 1964; George Psacharopoulous, Analysis of Labor-Skills Requirements in Greece,” dissertation, University 1968; Mark Blaug. Tbe Rate of Rehtm to Investment in Education Report to the National Planning Committee on the Third Educational National Educational Council, 1971.

MA:

A Case Study,” “An Economic of Chicago. in Thailand. A Plan, Bangkok:

70 II.

Data Catbering and Methodology

Our empirical estimates of the private rates of return from U.S. education is based on a cross-sectional survey of foreign students attending (during the academic year 198e81) four U.S. universities: Virginia Polytechnic Institute, the University of Delaware, the University of Tennessee, and the Florida State University. Approximately 1,700 questionnaires were distributed to students enrolled in these institutions. The distribution of students according to their field of study is given in Table 1. One third of the students surveyed are majoring in engineering, while almost twenty-five percent are in basic social sciences. The students were asked to provide actual direct expenditures on schooling for each year since their arrival in the U.S., and also their expected future outlays. In order to calculate expected future returns, we asked the respondents to estimate their future salaries, net of taxes, under four alternative life paths: (1) stay in the U.S. permanently after completion of their studies, (2) return to their native country immediately upon graduation, (3) work in the U.S. for five years after graduation, and then return to the native country, (4) not come to the U.S. for education, but rather work in the native country with a bachelor’s degree from a native college. Each foreign student was asked to estimate future net salaries during the first year of full employment, at fifteen years thereafter, and at the age of 5 5. Questionnaires with only partial answers to the questions pertaining to expected salary were excluded from the present value analysis. Of the 456 respondents, 227 provided adequate information for computing their expected present value. In order to calculate PV, information on schooling costs is needed. If future out-of-pocket costs of education were not estimated by survey respondents, extrapolations were made, based on the assumption that the

71

TABLE1 Dirtrlbutlon According

Requeucy

of Foreign Students to Field of Study

Surreyed

Pstioual our

suntey

Frsouency

l

Baiaaal Studat wuutiou,

Dirtributioa k

th 1901).

fra

UmStod p.9.

States

Doqlaa (Mew

Distribution* 1979-1980

Percent

P*rceat

153

33.3

25.0

24

5.2

3.9

ho

0.7

4.8

68

14.9

0.3

I12

24.6

39.3

56

12.7

10.3

II. .Boym (ad.),

York

Institute

Profiles:

of

Iarrrmtiomal

Thm Poreice

72

direct cost of schooling increasesby $300 annually.8 The rationale for not proxying future salaries is that expectations vary widely because of the heterogeneity of foreign students. On the other hand, the direct costs for the students surveyed were relatively uniform. This data was then used to calculate the present value of the stream of costs and earnings each student expects under each of the four scenarios.The formula for this calculation, presented in an appendix, is an adaptation of the standard present value formula. Earnings expectations have been ascertained for three periods so as to develop a lifetime earnings profile for each scenario. We have not explicitly stated an interest rate for our calculations; instead, we estimate present values at various interest rates. The primary reason for this is that meaningful data on interest rates in various countries are difficult to obtain. By comparing expected returns from education in the native country to returns from U.S. education at the same interest rate, we can determine the maximum interest rate which justifies investment in U.S. education. As we shall see, this rate varies substantially between undergraduatesand graduatesand among fields. There are advantagesto using the student’s expected earnings rather than the actual earnings experience. First, the decision to invest in human capital, like any other investment decision, must be based upon future expectations. Therefore, student responses should better reflect their economic reasoning than would future realizations. Second, as Freeman has shown, expectations are

8The

reason

for making

such an assumption

is explained

in the Appendix.

73

rather close to future realized earnings.’ Finally, data on realized earnings in the native countries of survey respondents is not readily available. The study has certain limitations. In calculating the present values, we assumethat the probability of employment is equal to one under the four career options that the foreign students encounter. In reality, it is not so. Many foreign students seek a U.S. degree to improve their chances of becoming employed, which implicitly effects their present value estimates. Many foreign students seek U.S. education not for the purpose of producing returns in the U.S. labor market or in the native country, but rather in a third country. To keep our analysiswithin manageable limits, the present value estimates of employment in a third country have been dropped from consideration. III. Private Rate of Return From U.S. Education According to our present calculations, of the several career options open to foreign students, staying in the U.S. permanently produces a higher expected present value than returning home either immediately after graduation or after staying in the U.S. for five years. It does not, however, dominate the option of not coming to the U.S. The relative values of the two options depend upon the discount rate. We can obtain an estimate of each student’s expected gain from coming to the U.S. by subtracting the expected present value of staying in the native country and working from the expected return of coming to the U.S. for higher education. These calculations are presented in Table 2. Positive numbers for a career choice imply that the investment in U.S. education is profitable as compared to the alternative of native education and work.

9Freanan,

pp. 202-226.

home

Return

U.S.

home

in

Return

Stay

Undergraduate

home

Return

U.S.

home

in

working

after

working

immediately

permanently

Students

after

immediately

permanently

Student0

Pattern

Return

Stay

Creduate

Career

in

after

for

U.S.

for

graduation

U.S.

five

five

Value

Sraduetion

Preeent

in

after

Net

yeare

yeare

26

26

28

145

202

179

(N)

to of

Sample

of Coming (Thoueande

TABLE

Size

169.1

131.4

IW.2

27.6

-13.2

145.9

5x

the U.S. for U.S. Dollers)

2

63.7

38.3

71.1

-6.9

-31.9

47.5

10%

UiSher

IS.5

5.1

21.8

-25.4

-51.4

2.2

Discount 15%.

Bducation

-4.1

-14.0

-2.4

-36.5

-58.4

-22.4

Rate 20% 251

-16.6

-25.2

-15.6

-43.6

-61.9

-36.6

30%

-26.1

-33.1

-25.5

-50.4

-65.6

-47.5

75

The returns from graduate education in the U.S. are high (16%) only if the foreign students intends to stay in the U.S. permanently. The option of remaining in one’s native country and working with a bachelor’s degreeis a better investment under other career patterns and at interest rates above 16% for permanent residencein the U.S. Recall that the results reported here are the expectations of students currently enrolled in U.S. universities. There are two plausible explanations for this result. One is that students do not plan to return to their native country after completing their graduate education. This may be true but it is inconsistent with the survey responsesof these students. Most (86%) planned to return to their native country immediately or after five years of work in the U.S. The other plausible explanation is that students expect graduate education to enhance their chances of obtaining a job at home. A large portion of the students in our survey are from far easterncountries. A large portion of them stated that graduate education would increasetheir chances of employment. lo Recall that we have assumedthat the probability of employment is one in all career choices including the option of not coming to the U.S. If, in fact, the probability of employment is higher for successfulgraduate students, we have overstated the present value of remaining at home relative to coming to the U.S. for graduate education. The returns from U.S. higher education are much higher for undergraduate students than for graduate students. Under any scenario, the return for undergraduatesis between 15% and 20%. This is on the high end of the range of values estimated for Americans pursuing higher education.’ 1 It is interesting to note

10 This

information

was obtained

during

personal

interviews

“Carnoy and Marcobach, ‘The Returns to Schooling in the U.S., 1939-1969.” Jwmal of Human Resources, Vol. 10 (1975), pp. 312-331; Raymond R. Sesnowitz, “Returns to Invesanent in Higher Education,” Journal of Ifumun Resoucs. Vol. 10 (1975), p. 148, and Richard Frcanan, “Overinvesunent in College Training?” Jour& of Humm Resources, Vol. 19 (19751, pp. 287-311 found private returns fran higher education for Americans to be around 10 to 12 percent.

76

that most of the undergraduates in our survey are from OPEC countries. It may be that the boom conditions of their native countries have elevated their expectations. Distribution

of Students

by Field of Study

The distribution of foreign students in the U.S. in various fields is not uniform. The supply of engineering majors is the largest, about 25 percent of the total foreign student body, followed by businessand management students. Table 1 provides information about the distribution of foreign students by field of study for our survey and for the nation. Becauseof the limitations in the size of our sample in each field, respondents have been classified into the following subgroups:

(1) Engineering

Civil, Electrical, Chemical, and other types

(2)

Biological Sciences Biology, Botany, Zoology, Nutrition, related fields

Ecology, and

(3) Applied Sciences Mathematics, Statistics, Systems Analysis, and Computing (4) Physical Sciences Biochemistry, Chemistry, Geography, Geology, Physics, Earth and Physical Sciences (5) Basic Social Sciences Economics, History, International Relations, Political Science, Sociology, Psychology, Linguistics, and other humanities (6)

Applied Social Sciences Agricultural Economics, Industrial Relations, Public Administration, Business Administration, and other applied sciences.

77 A discussion of foreign students’ expected present values according to field of study is useful in explaining such phenomena as the wide difference between the number of engineeringmajors and the number of applied sciencemajors. Engineering Among the six categories of students, engineering majors have the highest expected present value for U.S. higher education. Tables 3-6 contain the present value results, under various discount rates, for all six fields. The higher return for engineering holds true whether an engineeringgraduatedecidesto work in the U.S. permanently following school, return to his native country immediately, or to work in the U.S. for five years before returning to his homeland. The present value estimatesat 10% discount rate under the three career choices are respectively: $197,900, $104,400, and $138,900. In the event that a foreign engineering student decides to work with a bachelor’s degreefrom a school in his homeland, the expected net presentvalue is $119,400. Thus, as with our general estimates above, the net return is negative if a student plans to return home immediately. In the event the engineering major decides not to seek U.S. education, his present value ranks below that of the social scientists and the biological scientists (Table 6). In short, the returns from U.S. higher education are the highest for engineers. Therefore, it is not at all surprising that we find so many engineering majors from various countries. There are two main causesfor these high returns, and the resulting large number of foreign students majoring in engineering. First, salariesof engineersboth in the U.S. and abroad are equal to or higher than salaries in most other fields. Second, the high salariesof engineersin the U.S. make it more costly (in the form of foreign earnings) for American engineersto pursue graduate education. However, the supply of student slots is relatively inelastic due to a large fixed capital stock and a high proportion of

78 tenured faculty. In order to fill the student slots, U.S. schools must lower admission standards, reduce tuition charges, and/or recruit students from overseassources.’ 2 Since lowering admission standards may result in a lower quality of input, and institutions of higher education often face constraints on minimum quality, the option of recruiting students from overseasis chosen instead. As a result, large numbers of foreign students in the engineering fields are offered assistantshipsand fellowships which lower their costs and increase their expected returns. Foreign students make up almost thirty percent of the graduate students in the engineering departments at Virginia Polytechnic Institute and State University, one of the institutions surveyed. The availability of financial assistanceis a major impetus for seeking U.S. higher education for many of these students.’ 3 BiologIeaI Seleneea. The returns from U.S. higher education are lowest for biological scientists (Tables 3-6). If these students plan to work in the U.S. permanently, the present value at a 10% discount rate is $139,800. Biological scientists rank last, at discount rates of 2096 or lower, if they decide to return home immediately upon graduation or after working in the U.S. for five years. An individual majoring in one of the fields in the biological scienceswould have higher returns if he never came to the U.S. but rather worked in the native country with a bachelor’s degree.Only if the student decides to stay in the U.S. permanently does the decision to seek U.S. education seem a rational choice, on pecuniary grounds. However, at discount rates above 1096,the choice of U.S. education appearsto yield lower returns than the alternative of getting

%iiod 13

This

AguwJ, infom~atio~~

pp. 1-2. WM obtained

during

personal

iotaviews

(364,

(230.

(332.

(288,

(261,

(303.

20

18

26

41

24

Siologicnl Science Point eetlmate 95% confidence bend

band

Applied Sclencee Point estimate 95% confidence

PhyeiCal Sciencea Point estimate 95% confidence band

Ikaic Social Sciences Point eatlmate 95% confidence band

Applied Social Science0 Point eetimete 95% confidence band

348.3 392)

294.8 327)

327.5 366)

389.3 446)

268.8 307)

392.0 419)

5% 197.9 212)

194.3 222)

(148,

(129. 177.1 205)

146.6 163)

. 160.3 (138. 182)

(165.

139.8 (113,'166)

(183.

10%

77.8

83.0 102)

111.9 1m

78.4 100)

'(74.

97.2 119)

(66. 8%

(63.

(93.

(56,

110.8 120)

Value 15%

(loo,

Preeent

42.8

65.4 73)

42.0

68.2 83)

(33.

(29,

54.4 75)

40.1 50)

(20, 63)

(53.

(24. 60)

(57.

($ Thouaanda) 20%

Veluee of U.S. Higher Educetlon to ForeQn Student8 Stay in the U.S. Pewnently. by Fields of Study

band

Field of Study Engineering Point eetimate 95% confidence

Sample Size (srl 87

Preeent If tiy

TABLE 3

17.4 40)

42.2 55)

22.2 39)

38.5 45)

28.7 (7. 49)

18.1 (7. 28)

(-5,

(28,

(5.

(31,

25%

t-6, 28)

10.9

(-7.2i:)

38)

25.1

(-24.G)

(11.

(-8.7;49)

(16. 29)

22.9

30%

of

Study

Social

Social

Social

Point estimate 95X confidence

Applied

Point estimates 95% confidence

Basic

band

band

band

Sciencee

band

Science8

Science0

Point eetimete 95% confidence

Physical

Point eatimote 95% confidence

Applied

band

band

They

Sciencee

Sciences

Point estimete 95% confidence

Biological

Point eetimmte 95% confidence

Engineering

Field

if

return

0

28

37

31

13

15

98

Sire

Value

(151,

(136.

(136,

(91.

(109.

(1845,

the Netive

Suple

Preeant to

4

213.2 275)

197.0 238)

187.5 238)

143.5 1%)

168.6 227)

98.1

95.6 !17)

(645. 13f-v

(73.

65.5 92)

104.4 121)

(38.

(67,

10%

Present

53.4

15%

Value

Porei@l Graduation,

(25,

(33.

49.2 73)

46.7 600)

37.9 50)

29.5 45)

(25.

(13.

311.5 60)

(16,

(6% 63

U.S. UiSber Mucetion to Country Imedietely Upoa

219.3 254)

5x

of

TABLE

t.1.

(lo,

(1,

(-4,

(1,

(20,

Vield

19.7 38)

29)

19.9

13.1 14)

10.4 21)

14.3 30)

26.0 33)

20X

($ Thouunde)

by

studente

11.11

25%

Study

t-15.

12)

i$

2.5 20)

-1.4

c-s.‘i:j

t-15.

,-10:*x,

(-14,

(6, 10

of

-8.6 0)

(-27,

(-16,

-8.6

-8.0

10)

.5)

t-27, 6)

-10.8

t-17.

t-23, 6)

-8.8

,-lf%

30%

Field

Study

They

Sciencea

Social

Social

Point estimates 95% confidence

Applied

Point eetimnte 94% confidence

Basic

band

band

band

Sciences

band

Sclencee

Sciences

Point estimate 95% confidence

Physical

Point eetimete 95% confidence

Applied

band

band

Work

Sciencea

Point eetimate 95% confidence

Siological

Point estimate 95% confidence

8nginearing

of

if

in

the

28

34

33

20

15

230.2

(206,

324)

265.2

4)

244.8 316)

187.8 227)

199.5 238)

(l-51,

(173.

(147.

(160.

265.2 303)

5%

Sample Sire 0 72

(227,

Value of U.S. for Five Yeara

Present U.S.

(111.

(93.

(92.

(85,

138.2 165)

118.3 142)

125.1 157)

104.8 121)

loo.3 123)

Present

Educatio ileturnlng

5

138.9 157)

10%

(77,

(120,

HiSher before

TABLE

Value

Forei8n Native

69.6

63.6 76)

54.1 71)

79.5 95)

64.9 a0)

(63.

(49,

(52, 86)

(51,

(37.

19.5 90)

15%

to

(68.

to

(35,

47.7 59)

35.1 47)

40.2 51)

39.8 50)

29.4 43)

b7.2 55)

(23.

(29,

(29.

(15,

(39,

20%

(9 Thoueenda)

Studente Country, by

of

(17.

(5,

(14,

(13.

28.3 39)

16.2 27)

22.6 30)

34)

24.2

(2. 27)

lb.7

21.5 34)

25%

(2%

Field

Study

lb.2 21)

(4,

(3.

15.2 25)

11.2 18)

(2. 23)

13.1

(-6,5i:)

(7,

30%

co Y

of

Study

Sciencee

Sciences

Social

Social

Point estimate 95% confidence

Applied

Point eetlmate 95% confidence

Basic

band

bend

band

band

band

Sclencea

band

Sciences

Sciences

Point estimate 95% confidence

Physical

Point eetimates 95% confidence

Applied

Point estimate 95% confidence

Biological

Point

estimate 95% confidence

Engineering

Field

Preeent

0

28

2s

24

18

15

61

Selple She

Valoe

of

(138.

(166.

(149,

(75,

(131.

(159.

19s. 1 258)

233.1 300)

201.9 254)

135.8 195)

207.6 284)

199.1 238)

5%

Working

(91.

(110.

(93.

(45,

03.

(94.

6

129.0 166)

157.7 205)

124.6 155)

87.1 129)

136.7 190)

119.4 143)

10%

in the Native by Field of

TABLE

IS%

Value

Without

(66.

(82.

(66,

(31,

(57.

9h:,

122.5 162)

m.5 110)

65.6 100)

103.6 149)

l

U.S.

Degree.

(55.

(67,

104)

79.9

104.8 141)

69.3 86)

(52.

54.6 03

87.3 132)

65.2 WI

(23.

(42,

(49.

20%

($ Thoruende)

(66.%:,

Present

Couotty Study

(47.

(59.

(44,

(19,

94)

70.0

132)

95.8

58.1 72)

16.7 71)

(41.

(36,

(16.

86)

63.9

51.1 63)

45.6 75)

49.0

(32. 67)

(36:i:)

56.3

55.6 69)

30%

(36. 70

(41,

25%

83

a job in the native country with a bachelor’s degree.The low pecuniary returns from U.S. education largely explain why only 3.9% of the foreign student body are studying in the biological sciences. Applied Sciences. Approximately five percent of the foreign students studying in the U.S. are in this field. According to the expected present value estimates, the returns from U.S. education are very high for graduates in these fields only if they decide to stay in the U.S. permanently. They are second only to engineering majors in the present value estimates of working in the U.S. permanently (Table 3), but rank lowest under the other three career pattern choices. The high rates of returns from working in the U.S. can be attributed to the complementarity of human capital and the capital extensive technology prevalent in these fields in the U.S. More specifically, the U.S. is far more advanced in computer technology than are most of the countries from which our sample is drawn; thus the demand for specialists in related fields is extremely high. On the other hand, students who train as statisticians and computer scientists may find that their productivity, and hence their wage rate, falls drastically in their native countries becauseof the dearth of capital goods. Physical Sciencea. Physical and earth scientists, such as chemists, geologists, and physicists, have the highest net present values at discount rates below 15% if they seek U.S. education and decide to become part of the U.S. labor force permanently. From the investment point of view, it does not pay to come to the U.S. for education and then return home immediately at the end of college life. The individual would be better off financially, over the lifetime, if he had instead worked in the native country during the period spent in school in the U.S. for additional degrees.

84

Basic Social Sciences. Because of the limitations in the number of observations, this category is comprised of foreign students from a wide variety of fields - from economics to psychology to literature. This group should have the least incentive to seek education abroad compared to the other five groups. The alternative choice of not coming to the U.S. and working in the native country yields returns even higher than that for engineers (Table 6). For example, at a 15% discount rate, the net present value estimate for social scientists is $122,500, while for the engineers it is $84,300. Yet social scientists represent almost 40% of the foreign students in the nation. An explanation for this apparent contradiction is that foreign students who intend to terminate their studies upon receipt of a bachelor’s degree have very high returns, while graduate students have low returns (Table 2). The average(Tables 3-6) is low becauseof the inclusion of both graduate and undergraduate students. Most of the undergraduatesin our sample were in the basic or applied social sciencesor in engineering. Applied Social Sciences. The businessand management fields which comprise a major fraction of this category has had a substantial increase in the number of foreign students. The demand for applied social scientists, such as accountants and MBA’s has been on the rise not only within the U.S. labor market but abroad aswell. The increase in the demand for applied social scientists has induced a larger number of students to major in these subjects. A degree from a country like the U.S. gives a student a distinct advantage over local graduates in the job market. According to the survey respondents, MBA and other social science degrees from a U.S. university are very marketable in the developing countries. Our present value estimates (Table 5) show that a U.S.educated foreigner with five years of experience in the U.S. labor market is ranked with the engineersas having the highest expected

85

present value. However, if the individual decides to work in the native country with a local bachelor’s degree, he is ranked fourth out of the six categories in terms of present value estimates. IV.

Summary

and Conclusion

Our expected present value estimation procedure included only the pecuniary costs and benefits associatedwith education. The differential pecuniary returns according to field of study largely explain why there has been a large growth in the number of foreigners in the engineering fields. The pecuniary returns from U.S. education for students in other disciplines are higher than the alternative of not seeking U.S. education if students seek employment in the U.S. permanently or work for a few years in the U.S. before returning to their native country. Foreign students majoring in applied scienceshave high returns from U.S. education if they work in the U.S. permanently, and low returns if they seek employment in their native country. For the social scientists the gap is much smaller. The lack of complementary capital goods can largely explain why computer scientists and statisticians expect large differences in pecuniary returns between working in the U.S. and in their native country. The study has certain limitations, some of which can possibly be overcome if a similar type of study is attempted in the future. First, in calculating the present values, we assumethat the probability of employment is equal to one under the four career options. We have uncovered evidence to suggestthat the probability of employment is substantially below one and that a U.S. degree increasesthe probability of employment. A close look at the employment situation of skilled professionals in various counuies with and without a U.S. degreeis suggested. A comparison of respondents’ expected future salarieswith

86 the actual salaries offered in the job market would allow an assessmentof how realistic individuals are in their expectations. Data on actual salaries according to field of study in various countries are difficult to procure; nevertheless, resources employed in such an endeavor may provide useful information. Lastly, an economic analysis of the returns from working in a country other than the U.S. or the native country will add additional insight to the reasons foreign students seek U.S. education.

87

APPENDIX

Methodology for of U.S. Bigher

Calculating Education

returns

capital

The calculation follouillg

of

froa

the Present Value !PV) to the Foreign Students investment

is exprtsrtd

in

the

equation.

(1) g1 (l+r14

RI-

+$U+r)-a+a where

PV = present

+ ffi!L +....+En U+r)

value

of

past,

at tine

Ep - earnings

at time of survey

a - aumber of years of

retirement

= - rate

of

discount

(2)

can

w

also

be

(3)

value

future

expected

earnings

to U.S. aa a student

(interest) ae:

Et

t--a

The present

(l+r)"

t

arrival

written

;

I

period

since

n - year

and

present,

Et - earnings

The equation

+ . . . .

z2 (l+r) -*+l

+

(l+rlt

of cost of schooling,

c-c1

+

(l+r) -a +...

C,

c2 (i+r)-a+l +

P

incurred

+

. . .

over

the

years

+c9 (l+r)-a+a

C n+rjt

is:

or: (4)

n C-i

Ct

where

-

m -

The present pecuniary (5)

Ct

t--a

U+r)’

cost

of schooling

at

year

of

school

finishing

value benefits

Pvt.

(PV) and II

z t-a

of education

time

of

at

period

time

1.

I

is

therefore

(ignoring

the

non-

costs):

-(l+t)

111

% t

Zt-a

Ct (l+r)

t

In the survey analysis of net present value, salariesare used as a proxy for Et. Individuals were askedto provide expected future salaries (net of taxes) during the first year of full employment, fifteen years after entering the labor market, at the age of 55. They were asked to provide U.S. dollar valuesfor the above three categories if they stayed in the U.S. permanently. In addition, respondents were asked what their expected net salary (in native currency) would be if they returned to the native country immediately upon finishing his studies, and after working five years in the U.S.

89

The purpose of not asking the native expected salaries in dollar values is to avoid the potential discrepancies of exchange rate assumptions and/or low cost of living adjustments by individuals. The native salaries have been converted into dollar terms after adjusting them to the real purchasing power of each currency relative to the dollar. The conversion was done basedon studies by Kravis et al * 4 and Isenman.’ 5 The reason for limiting the lifetime earning profile at age 55 is two-fold: (1) life expectancy and retirement age varies from country to country, and reliable data is not available, (2) earnings past the age of 55 will have little impact on the value of net present value, becauseof the discounting of the future income to the present. Each individual is asked about his out-of-pocket cost of schooling for each year, from the time of his arrival in the U.S. until graduation. In the event he did not predict the direct expensesfor future years of college, some assumptionswere made. First, if the student is already receiving financial assistancefrom a source other than family funds, it is assumedthat the assistance will continue for the remainder of the student life. Second, direct cost of schooling is assumed to be going up by $300 annually. The rationale for such an extrapolation is basedon the observation that tuition, and room and board, increase by approximately 5 percent per year.

14 1. B. Kravis, A. Heston, R. Summers, United Nation’s International Comparison Projet.& Phase II; Intemationd Comparisons for Real Product and -baring Pow, (Baltimore: John Hopkins University Press, 1978). Also, Kravis, Heston, Summers, “Real GDP Per Capita for More than One Hundred Countries,” Economic Jour&” Vol. 88 (June 1978), pp. 215242. 15 Paul Isenmsn, “InterCounay and Unresolved Questions,” World D.C., September 1979.

Comparisions (ppp) Incomes: Rank Staff Working Paper No.

Revised Estimates 358, Washington,

90

In certain cases,the individual has savingsafter meeting all the direct costs of schooling from the fellowship/assistantship received. Under such circumstances, the direct cost of schooling is negative. Since financial assistance is generally adjusted for inflation and years of study (seniority) in graduate school, it is realistic to assume that future direct expenditure (savings) on schooling will grow at a steady 5 percent rate. For undergraduate students, the direct cost of U.S. education is assumed to be zero in dollar values. This is based on the assumption that the next best alternative to U.S. schooling is attending school in the home country. The opportunity cost of U.S. education for the graduate student is assumed to be the earnings foregone in one’s native country with a bachelor’s degree. The questionnaire provides information about the earnings function. The contention that the indirect cost of additional years of schooling is earnings foregone in the home country biases the computation of costs downward. It stems from the fact that we are overlooking the possibility of earnings from working in the U.S. Two factors led to overlooking this possibility: (1) the questionnaire does not ask graduate students of their expected earnings in the U.S. labor market with a bachelor’s degree, (2) the possibility of employment in the U.S. with a bachelor’s degree is very slim, especially during the early years of schooling, because of legal restrictions. Over time, the student increases his chances of getting employed. Nevertheless, the legal limitations are still very stringent. In our calculation of PV,, the data for direct and indirect pecuniary costs of schooling for each year are provided by the student and/or extrapolated for future years, according to the methodology discussed above. Information about expected future salaries is provided directly for three time periods. Graphically, let us assume A, B, and C are the three points associated

91

with next salary at the end of the first year of employment, fifteen years thereafter, and at the age of 55, respectively. The hypothetical points are shown in the time-earning profile of Figure 1. The cost of schooling is expressedas negative earnings, and hence it is assumed that net earningsare negative (Ct values) until period m, the year of graduation. A linear growth of salary is assumed between points A and B, and between B and C. Via this method, Et for (ng-3) periods are extrapolated. Based on the above assumptions and methodology, the following four casesof PV calculations are performed:

(1) The individual decides to stay in the U.S. permanently. (2) Returns home immediately on completion of studies. (3) Works five years in the U.S., after graduation, then returns to his native country. The annual salary level of the five years of study in the U.S. is assumedto be the same as under the assumption that the individual decides to work in the U.S. permanently. This biases the expected salaries upward since employers will be less willing to pay the same salary knowing that the employee may leave within a few years.The limitations in the data source, and the fact that an individual may actually succeedin not revealing his future careerplans led us to overlook this potential bias. (4) The individual had not pursued U.S. education (applicable only to the graduate students), and has been working in the home country, since the year of arrival to the U.S. In calculating PV, values, it is assumed that the future expected salaries in the ,abovefour alternatives are associatedwith

92 employment probability of one. In reality, such may not be the case, especially in some developing countries, where there exists an excess supply of skilled manpower, leading to high degreeof unemployment or underemployment. As such, the PV, valuesfor case (4) and to a lesser extent for the other casesmay be biased upward for some individuals. One of the weaknessesof the PV, analysis is the inability to compute individual’s rate of time preference, expressedas the rate of interest/discount in our equation(s). The community’s rate of time preference can be grossly represented by the going rate of interest in the market. However, since (1) the analysis is of disaggregateddata and (2) there doesn’t exist free borrowing and lending in the human capital market; usage of the prevailing market rate of interest may not be valid. In the analysis, this problem is sidestepped of computing PV, for various rates of interest (r). The value of r is held constant over the years, while calculating each PV, figure.