Skip to main content

Wage distributions in origin societies and occupational choices of immigrant generations in the USA

Abstract

This paper studies occupational selection among generations of immigrants in the USA and links their choices to the occupational wage distribution in their country of origin. The empirical results suggest that individuals are more likely to take up an occupation in the USA that was more lucrative in the origin country, conditional on individual demographic characteristics, parental human capital, and ethnic networks. However, the importance of the origin wage declines with the length of time that immigrants spend in the USA and over generations. Information frictions may be an explanation.

This is a preview of subscription content, access via your institution.

Fig. 1
Fig. 2

Notes

  1. Second generation refers to individuals who were born in the USA but have at least one foreign-born parent.

  2. According to statistics from the Current Population Survey, a nontrivial fraction of the native-born Americans in the sample may be second-generation immigrants.

  3. I conduct several sensitivity tests regarding the period to which the origin wage refers and provide the details of these tests in online appendix (https://drive.google.com/file/d/1gR-rPtQhcIdxpgRuHmAqu6obsXiAbnwz/view?usp=sharing).

  4. Variables that do not vary across occupations may also enter the regression by interacting them with a set of occupation dummies. I use the interaction of person-specific characteristics and occupational attributes for computational advantage.

  5. Because the American Community Surveys (ACS) does not document respondents’ parental characteristics, I estimate these characteristics using a group-mean method (Card et al. 2000; Blau et al. 2013). The next section reveals more details about this method.

  6. The last two time-invariant traits are dropped from the regression when country fixed effects are introduced.

  7. Second-generation immigrants are not identifiable in the ACS data. Instead, I examine the 1.5-generation immigrants who migrated to the USA prior to the age of five. These people may obtain all their education in the USA but may be tightly connected to the origin society. Therefore, the 1.5 generation should have a close resemblance to the second generation.

  8. Because both the ACS and the Current Population Survey (CPS) survey nationally representative samples, and because second-generation immigrants can be distinguished in the CPS data, I compare the statistics from the ACS and the CPS of the same years. Were all second-generation individuals to identify their ancestries, statistics suggest that approximately 20% of the native-born American sample are second-generation.

  9. I do not trim the sample of 1.5 generation as this group is much smaller. Appendix Table 11 reports the list of origin countries in the sample and the number of observations in each generation by origin (pre- and post-trimming). Alternatively, I form a 25% random sample of the first generation and a 5% random sample of the native-born Americans, regardless of countries of origin. These two samples have a comparable size to the samples analyzed in the paper. Repeating the regressions in Tables 5 and 6 on the random samples yields very similar results. The results are available from the author upon request.

  10. The parents of the 1.5 and higher generations are more likely to be present in the USA and be observed in the earlier censuses. It is not clear whether the parents of the first-generation immigrants migrated to the USA or not. Therefore, I do not estimate parental education, income, or the number of siblings for the first generation. Nevertheless, the parental ethnic networks estimated for the 1.5 generation may also reflect the ethnic networks of the first generation in the USA.

  11. Appendix Section A.1 offers more information about the group-mean method.

  12. In the ACS sample, about 4% are out of the labor force, and about 3% are unemployed.

  13. Occupation categorization is described in more detail in the Appendix Section A.2.

  14. I measure the years of experience for individuals engaging in various occupations because the census occupational classification includes executive and managerial jobs as a separate category where promotion to these positions largely depends on performance and experience.

  15. The census provides the occupational income score, which assigns each occupation a value representing the median total income of all individuals with that particular occupation. Because I focus on the highly educated individuals, I modify the score to be the median total income score among all college graduates with each occupation.

  16. Work experience is calculated as years of experience = age − years of education − six.

  17. The training cost of each occupation is assigned by the median educational attainment of individuals with a college degree who work in that occupation. Certain occupations, such as physicians and professors, require additional training beyond college.

  18. Individuals who were born between 1956 and 1960 are matched to 1980 occupational attributes; those born between 1961 and 1970 are matched to 1990 attributes; those born between 1971 and 1980 are matched to 2000 attributes; those born after 1980 are matched to 2010 attributes.

  19. Explicitly, I link birth cohort 1956–1960 to origin wages (averaged by period) in 1983–1985, cohort 1961–1965 to origin wages in 1986–1990, cohort 1966–1970 to origin wages in 1991–1995, cohort 1971–1975 to origin wages in 1996–2000, cohort 1976–1980 to origin wages in 2001–2005, and cohort 1981–1985 to origin wages in 2006–2008, respectively.

  20. More details about the data sources are provided in the Appendix Section A.3.

  21. The parental characteristics are not in the control set for the first-generation immigrants as the group means derived from the population in the USA are potentially poor proxies.

  22. In the sample, about 45% of first-generation immigrants have naturalized, and about 87% of 1.5 generation have naturalized.

  23. In response to the two abovementioned issues, I replicate the regressions in columns 2–3 Table 6 (1) excluding individuals who migrated before the age of 18, as there may be substantial heterogeneity in education and motivation for migration among this group and (2) restricting the sample to the first-generation immigrants who have naturalized. Appendix Table 12 presents the results. The estimates are comparable to those in Table 6 and may thus suggest that neither differential education path nor visa policies drive the positive link between origin wages and first generation’s occupational choices in the USA.

  24. More details about these tests are available in the online appendix.

  25. The regressions in columns 2–3 Table 6 control for the years in the USA by including the interactions between the six occupational attributes and the standardized number of years in the USA. For a more straightforward interpretation, I use the actual number of years in the USA in this table. So the estimated main effects and interaction effects of the origin wage are slightly different.

  26. In a similar vein, I investigate if the effect of the origin wage varies by age at migration. I find that the link between origin wage and US occupational choice is stronger if individuals migrated at a later age. The results are available in the online appendix.

  27. Mentioned occupational categories include the following: (1) Architects/engineers, (2) doctors/lawyers, (3) health assessors/teachers, (4) sales clerks, (5) protective service, (6) household/other service, (7) mechanics/system operators, (8) construction trades, and (9) machine operators/drivers. More detailed information about the survey is provided in Appendix Section A.4.

  28. As before, I dismiss the origins with fewer than 10 observations. A total of 2215 individuals from 12 countries of origin are included accordingly, both males and females. Appendix Table 11 displays the included origin countries. Despite the much smaller set of origin countries, these countries represent the ones with a large number of observations in the ACS. Moreover, members of the KWW sample were born between 1970 and 1983 and may resemble the fourth and fifth cohorts in the ACS sample. But I cannot tell whether they were foreign- or native-born.

  29. I use the average standardized occupational wages from 1983 to 1995 in the regressions in Table 9.

  30. I conduct two sensitivity tests to rule out the possibility that the origin occupational wages capture additional relevant information that is not incorporated in the average US wage, such as regional wage differences. First, I replace the national averages of US wages using the regional averages in the analysis just described. Four US regions are considered as follows: Northeast, North Central, South, and West. Second, I restrict attention to occupation pairs with unambiguous wage ranking across states in the USA and exclude the answers to questions 1 and 3, accordingly. These tests provide reassuring results which are available in the online appendix.

  31. The Bureau of Labor Statistics (BLS) provides information about education and training requirements for narrowly defined occupations. Based on the projections data of BLS on occupational entry-level education, over three-quarters of professional jobs require a bachelor’s degree or above, while less than 2% of non-professional jobs require a bachelor’s degree. Source: https://www.bls.gov/emp/tables/education-and-training-by-occupation.htm

  32. Since the full sample is much larger now, I randomly select around 3000 individuals from each ethnic origin among the first generation and the native-born Americans for the origins with the number of observations that well exceeds 3000 given computational capability.

  33. Appendix Table 13 displays the results.

  34. Four censuses are used as follows: 1960 1% sample, 1970 1% metro sample, 1980 5% sample, and 1990 5% sample.

  35. The estimated group mean level is obtained by regressing the variable of interest on age, female, a year-of-survey dummy, and a full set of origin dummies. The estimated group mean level for a certain origin is the predicted value for a 40-year-old male from that nation surveyed in 1980. Besides, the number of children is estimated from only the group of individuals who had children born to them.

  36. The occupational category that a narrowly defined job falls in is reported in the brackets.

  37. I construct the adjusted share of individuals from an origin country choosing a certain option as follows: for each option in every question, I generate a binary indicator of the choice of a respondent and then regress the indicator on a set of origin dummy variables controlling for gender, age, and the year of survey. Accordingly, the coefficient on an origin dummy represents the propensity of 16-year-old males of the origin country selecting that option in 1994.

References

  • Abramitzky R, Boustan LP, Eriksson K (2014) A nation of immigrants: assimilation and economic outcomes in the age of mass migration. J Polit Econ 122(3):467–506

    Article  Google Scholar 

  • Abramitzky R, Boustan L (2017) Immigration in american economic history. J Econ Lit 55(4):1311–45

    Article  Google Scholar 

  • Abramitzky R, Boustan LP, Eriksson K (2020) Do immigrants assimilate more slowly today than in the past?. American Economic Review-Insights 2(1):125–41

    Article  Google Scholar 

  • Arcidiacono P (2004) Ability sorting and the returns to college major. J Econ 121(1):343–375

    Article  Google Scholar 

  • Betts JR, Lofstrom M (2000) The educational attainment of immigrants: trends and implications. In: Borjas G J (ed) Issues in the economics of immigration. University of Chicago Press, pp 51–116

  • Bisin A, Verdier T (2000) Beyond the melting pot: cultural transmission, marriage, and the evolution of ethnic and religious traits. Quart J Econ 115(3):955–988

    Article  Google Scholar 

  • Blau FD, Kahn LM, Papps KL (2011) Gender, source country characteristics, and labor market assimilation among immigrants. Rev Econ Stat 93(1):43–58

    Article  Google Scholar 

  • Blau F, Kahn L, Liu A, Papps K (2013) The transmission of women’s fertility, human capital, and work orientation across immigrant generations. J Popul Econ 26(2):405–435

    Article  Google Scholar 

  • Borjas G (1987) Self-selection and the earnings of immigrants. Am Econ Rev 77(4):531–553

    Google Scholar 

  • Borjas GJ (1992) Ethnic capital and intergenerational mobility. Quart J Econ 107(1):123–150

    Article  Google Scholar 

  • Borjas GJ (1995a) Assimilation and changes in cohort quality revisited: what happened to immigrant earnings in the 1980s?. J Labor Econ 13 (2):201–245

    Article  Google Scholar 

  • Borjas GJ (1995b) Ethnicity, neighborhoods, and human-capital externalities. Amer Econ Rev 85(3):365–390

    Google Scholar 

  • Boskin MJ (1974) A conditional logit model of occupational choice. J Polit Econ 82(2):389–398

    Article  Google Scholar 

  • Card D, DiNardo J, Estes E (2000) The more things change: immigrants and the children of immigrants in the 1940s, the 1970s, and the 1990s. In: Borjas G J (ed) Issues in the economics of immigration. University of Chicago Press, pp 227–270

  • Card D (2001) Immigrant inflows, native outflows, and the local labor market impacts of higher immigration. J Labor Econ 19(1):22–64

    Article  Google Scholar 

  • Chiswick BR (1978) The effect of americanization on the earnings of foreign-born men. J Polit Econ 86(5):897–921

    Article  Google Scholar 

  • Chiswick BR (1983) An analysis of the earnings and employment of asian-american men. J Labor Econ 1(2):197–214

    Article  Google Scholar 

  • Chiswick BR, Taengnoi S (2007) Occupational choice of high skilled immigrants in the united states. Int Migr 45(5):3–34

    Article  Google Scholar 

  • Constant AF, Zimmermann KF (2003) Occupational choice across generations. Appl Econ Q 49(4):299–317

    Google Scholar 

  • Dolton PJ, Makepeace GH, van der Klaauw WH (1989) Occupational choice and earnings determination: the role of sample selection and non-pecuniary factors. Oxf Econ Pap 41(3):573–594

    Article  Google Scholar 

  • Dustmann C, Fadlon I, Weiss Y (2011) Return migration, human capital accumulation and the brain drain. J Dev Econ 95(1):58–67

    Article  Google Scholar 

  • Dustmann C, Preston I (February 2012) Comment: Estimating the effect of immigration on wages. J Eur Econ Assoc 10(1):216–223

    Article  Google Scholar 

  • Dustmann C, Schonberg U, Stuhler J (2016) The impact of immigration: why do studies reach such different results?. J Econ Perspect 30(4):31–56

    Article  Google Scholar 

  • Edin P-A, Fredriksson P, Aslund O (2003) Ethnic enclaves and the economic success of immigrants -evidence from a natural experiment. Quart J Econ 118(1):329–357

    Article  Google Scholar 

  • Eriksson K (2019) Ethnic enclaves and immigrant outcomes: Norwegian immigrants during the Age of Mass Migration. Eur Rev Econ Hist 24(3):427–446

    Article  Google Scholar 

  • Fernandez R, Fogli A (2009) Culture: an empirical investigation of beliefs, work, and fertility. Amer Econ J Macroecon 1(1):146–77

    Article  Google Scholar 

  • Giuliano P (2007) Living arrangements in western europe: does cultural origin matter?. J Eur Econ Assoc 5(5):927–952

    Article  Google Scholar 

  • Hanson GH, Liu C (2017) High-skilled immigration and the comparative advantage of foreign-born workers across us occupations. In: Hanson G H, Kerr W R, Turner S (eds) High-skilled migration to the United States and its economic consequences. University of Chicago Press, pp 7–40

  • Keith K, McWilliams A (1999) The returns to mobility and job search by gender. Ind Labor Relat Rev 52(3):460–477

    Article  Google Scholar 

  • Lafortune J, Tessada J (2012) Smooth(er) landing? The dynamic role of networks in the location and occupational choice of immigrants, Working Paper, Pontificia Universidad Catolica de Chile

  • Lewer JJ, den Berg HV (2008) A gravity model of immigration. Econ Lett 99(1):164–167

    Article  Google Scholar 

  • Munshi K (2003) Networks in the modern economy: Mexican migrants in the u. s. labor market. Quart J Econ 118(2):549–599

    Article  Google Scholar 

  • Peri G, Sparber C (2011) Highly educated immigrants and native occupational choice. Ind Relat J Econ Soc 50(3):385–411

    Google Scholar 

  • Propper C, Reenen JV (2010) Can pay regulation kill? Panel data evidence on the effect of labor markets on hospital performance. J Polit Econ 118 (2):222–273

    Article  Google Scholar 

  • Rooth D-O, Saarela J (2007) Selection in migration and return migration: evidence from micro data. Econ Lett 94(1):90–95

    Article  Google Scholar 

  • Ryoo J, Rosen S (2004) The engineering labor market. J Polit Econ 112(S1):S110–S140

    Article  Google Scholar 

  • Schaeffer PV (1995) The work effort and the consumption of immigrants as a function of their assimilation. Int Econ Rev 36(3):625–642

    Article  Google Scholar 

  • Solon G (1992) Intergenerational income mobility in the United States. Amer Econ Rev 82(3):393–408

    Google Scholar 

  • Spitzer Y, Zimran A (2018) Migrant self-selection: anthropometric evidence from the mass migration of italians to the united states, 1907-1925. J Dev Econ 134:226–247

    Article  Google Scholar 

  • Warman C (2007) Ethnic enclaves and immigrant earnings growth. Can J Econ 40(2):401–422

    Article  Google Scholar 

  • Wobmann L (2003) Schooling resources, educational institutions and student performance: the international evidence. Oxf Bull Econ Stat 65 (2):117–170

    Article  Google Scholar 

  • Zhan C (2015) Money v.s. prestige: cultural attitudes and occupational choices. Labour Econ 32:44–56

    Article  Google Scholar 

Download references

Acknowledgments

I thank Julie Cullen, Catalina Amuedo-Dorantes, Ernest Boffy-Ramirez, Oana Tocoian, Madeline Zavodny, the editor, Klaus F. Zimmermann, and four anonymous referees for helpful comments. Any mistakes are my own.

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Crystal Zhan.

Ethics declarations

Conflict of interest

The author declares that she has no conflict of interest.

Additional information

Responsible editor: Klaus F. Zimmermann

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Data availability

Additional results and copies of the computer programs used to generate the results presented in the paper are available from the author.

Appendix

Appendix

A.1 Group mean levels of parental characteristics

As the American Community Surveys (ACS) does not collect information on the parents of respondents, I use a grouping estimation method (Card et al. 2000; Blau et al. 2013) to estimate the socioeconomic backgrounds of the 1.5 generation and the native-born Americans in the sample. The group means are derived from the population surveyed in the earlier censuses.Footnote 34 I use factors that may affect children’s human capital accumulation, such as educational attainment, family income, and the number of children.

The data matching is as follows. First, groups of “parents” are identified for each individual. For 1.5-generation immigrants, their “parents” would be first-generation immigrants from the same country of origin who were aged 20–40 in an immigrant’s year of birth and who migrated to the USA no more than five years after the birth year. For second- or higher-generation immigrants, their “parents” would be those, either foreign-born or native-born, of the same ethnic origin who were aged 20–40 and lived in the USA in an immigrant’s year of birth. To avoid the potential problems of incomplete education and differential mortality, I examine only individuals aged 20–60 as of the survey year. Then I calculate the mean levels of parental and family characteristics adjusted for age and gender for each group of “parents.”Footnote 35 The summary statistics of parental characteristics that match to my sample are reported in Table 1.

Admittedly, there is slippage in this grouping estimation method. The average parental socioeconomic characteristics of the highly educated sample are likely to be higher than the adjusted group mean levels obtained through the described approach due to positive human capital transmission (Solon 1992). Because of the substantial difference in educational attainment and income across racial and ethnic boundaries, the parental socioeconomic characteristics may be underestimated by different degrees for different ethnic groups. Nevertheless, the group mean levels capture the discrepancy in the human capital across ethnicities, which is an essential input in the formation of one’s human capital (Borjas 1995b) and is an indicator of the ethnic human capital that helps individuals find employment (Fernandez and Fogli 2009).

A.2 Occupation categorization

The 383 three-digit coded occupations in the 1990 Census Bureau occupational classification scheme are categorized into 23 categories as follows: (1) general executives and managers such as legislators, chief executives, public administrators, and mail superintendent; (2) management-related occupations such as accountants, underwriters, and personnel specialists; (3) architects and engineers; (4) scientists and professors, including mathematicians, social scientists, and natural scientists; (5) doctors and lawyers; (6) health assessors, teachers, and librarians such as registered nurses, therapists, and secondary school teachers; (7) social workers such as recreation workers, clergy, and religious workers; (8) writers, artists, entertainers, and athletes; (9) engineering and science technicians such as electrical technicians, cartographers, and airplane pilots; (10) health and legal technicians such as practical nurses, dental hygienists, and paralegals; (11) sales representatives such as insurance agents, advertising agents, and sales engineers; (12) sales clerks such as cashiers, retail sales clerks, and street vendors; (13) office clerks and health service workers such as secretaries, interviewers, and dental assistants; (14) administrative support workers such as office supervisors, computer operators, and expediting clerks; (15) protective service workers such as firefighters, police, and sheriffs; (16) household and other service workers such as housekeepers, cooks, and janitors; (17) farming, forestry, and fishing occupations; (18) mechanics and system operators such as automobile mechanics, aircraft mechanics, and power plant operators; (19) repairers and precision workers such as office machine repairer and miners; (20) construction trades and craftsmen such as concrete and cement workers, engravers, and bakers; (21) heavy machinery operators such as ship crews, locomotive operators, and crane operators; (22) small machine operators and drivers such as printing machine operators, sawyers and bus drivers; and (23) laborers such as construction laborers and stevedores.

A.3 Other origin characteristics

The origin characteristics data are from various sources. The information for democracy status, state of war, and share of labor force in agriculture is acquired from the Wejnert’s Nations, Development, and Democracy Dataset from ICPSR. The per capita GDP data are from GDP and Per Capita GDP by Angus Maddison. The real GDP per capita is adjusted for purchasing power parity and expressed in 1990 International Geary-Khamis dollars. The Gini coefficient data are obtained from World Bank Open Data. The educational attainment data are from the Barro-Lee dataset. Distance to the USA is calculated as the number of air kilometers between the home country’s largest city and the nearest US gateway (Los Angeles, Miami, or New York) using www.timeanddate.com. The information about nations’ official languages is from en.wikipedia.org/wiki/List_of_official_languages.

A.4 Knowledge of the World of Work questions

The Knowledge of World of Work (KWW) survey, a part of the National Longitudinal Survey of Youth 1979 Children and Young Adults (NLSY79 Child/YA), is conducted on individuals aged 14 to 24. In the years 1994, 1996, and 1998, the NLSY79 Child/YA asked the young adults a series of questions concerning commonly held jobs. I focus on a set of eight questions regarding occupational earnings. Each question asks respondents to pick one out of two occupations that they think offers a higher wage. The respondents are also allowed to choose “Don’t know” or “Refuse to answer.” For example, one such question is “Who do you think earns more in a year? A person who is (A) a high school teacher or (B) a janitor.”

Specifically, the questions compare: (1) automobile mechanic [mechanics/system operators]Footnote 36 and electrician [construction trades]; (2) medical doctor [doctors/lawyers] and lawyer [doctors/lawyers]; (3) aeronautical engineer [architects/engineers] and medical doctor [doctors/lawyers]; (4) grocery store clerk [sales clerks] and truck driver [machine operators/drivers]; (5) unskilled laborer in mill [laborers] and unskilled laborer in factory [laborers]; (6) lawyer [doctors/lawyers] and high school teacher [health assessors/teachers]; (7) high school teacher [health assessors/teachers] and janitor [household/other service]; and (8) police officer [protective service] and janitor [household/other service]. Since the second and fifth questions contrast a pair of occupations that are in the same category, I exclude these two questions from the analysis.

Appendix Table 12 presents the summary statistics of the share of respondents of each ethnic origin selecting a certain occupation mentioned in the KWW questions.Footnote 37 Noticeably, in some pairs of occupations, such as those in the latter four questions, one occupation is predominantly selected by individuals of all ancestries. Yet variation exists across origins.

A.5 Tables

Table 11 List of origins and number of observations
Table 12 Subgroups of first-generation immigrants
Table 13 Origin wage distribution and occupational choices by education: conditional logit

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Zhan, C. Wage distributions in origin societies and occupational choices of immigrant generations in the USA. J Popul Econ 35, 89–133 (2022). https://doi.org/10.1007/s00148-020-00811-4

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s00148-020-00811-4

Keywords

  • Immigrants
  • Occupational choices
  • Origin wage
  • Intergenerational transmission
  • Assimilation
  • Human capital
  • Information

JEL Classification

  • J15
  • J24
  • J31