Culture, Labor Supply, and Fertility Across Immigrant Generations in the United States

Abstract

Recent immigration to the United States and other developed nations has increasingly been from countries that have relatively traditional gender norms. This study uses data from the Current Population Survey for 2000–14 to investigate how source-country gender norms influence the labor supply and fertility of married women across immigrant generations in the US. It finds that immigrants’ and descendants’ labor supply and fertility are associated with the female-to-male labor force participation ratio and total fertility rate in the source country; importantly, the association declines across successive generations. Husbands’ source-country characteristics are also associated with the labor supply and fertility of immigrant women. These findings indicate evolution and assimilation of traditional gender norms in the long run.

KEYWORDS:

• Gender norms
• immigration
• assimilation
• labor supply
• fertility

JEL Codes:

• J13
• J16
• J22

SUPPLEMENTAL DATA

Supplemental data for this article can be accessed here https://doi.org/10.1080/13545701.2019.1633013.

Notes

1 We also conducted a number of analyses with alternate specifications measuring source-country characteristics when the second- and 2.5-generation women were born and when they were age 10. The results were similar to those reported.

2 Using 1994–2003 Current Population Survey (CPS) data, Blau and Kahn (2007) investigated the labor supply of Mexican first- and second-generation immigrant women and found that although immigrant Mexican women had lower levels of labor supply than native non-Hispanic Whites, the gap was much narrowed in the second generation.

3 We are grateful to Francine Blau for providing us with data from UNSD on the economic activity rates for men and women for 1950–2011.

4 To form a consistent list of source countries, we combined some countries that split or combined between 1980 and 2000 (for example, the former USSR countries, East and West Germany, former Czechoslovakia, and former Yugoslavia). Some countries were combined because data on source-country characteristics were available only in a combined form. The source-country characteristics were typically available at five-year intervals for the period 1950–2010. The fertility rate in the source country was grouped as follows: <1950; 1950–9; 1960–4; 1965–9; 1970–4; 1975–9; 1980–4, 1985–9; 1990–4; 1995–9; 2000–4; 2005–14. The economic activity rate was grouped as follows: <1950; 1950–9; 1960–4; 1965–9; 1970–4; 1975–9; 1980–3, 1984–7; 1988–93; 1994–7; 1998–2001; 2002–6; 2007–14. The GDP per capita was available annually, which we averaged for each of the following five-year categories: <1950; 1950–5; 1956–60; 1961–5; 1966–70; 1971–5; 1976–80, 1981–5; 1986–90; 1991–5; 1996–2000; 2001–5; 2006–10. Language characteristics are time invariant in our data.

5 The complete table of correlation coefficients is available from the authors upon request.

6 The unmatched respondents work for pay more hours than matched respondents on average, but the difference is small (1.3 hours/week) and disappears after controlling for educational attainment. There is no difference between the two groups in number of births. Unmatched respondents are more likely to have arrived in earlier cohorts (1950–65) than matched respondents, which is not surprising because source-country characteristics for that period are not available for several countries. Additionally, unmatched respondents are more educated than matched respondents, and less likely to be first-generation immigrants (than 1.5-generation immigrants) than matched respondents. There are negligible differences in the likelihood of being second or 2.5 generation between matched and unmatched respondents

7 In an alternate specification, we also control for country of origin fixed effects.

8 Clustering the standard errors on country or region of origin yields similar results.

9 P-value for F-test = 0.6187.

10 In additional analyses, we estimated Models 2–3 in Table 3 without the female/male activity rate. The estimated coefficient on fertility rate is modest, sometimes negative, and statistically insignificant.

11 The Supplemental Online Appendix is available on the publisher's website.

Additional information

Notes on contributors

Felix M. Muchomba

Felix M. Muchomba is Assistant Professor in the School of Social Work at Rutgers, the State University of New Jersey. His research examines the effect of policies and institutions on the health and well-being of girls and women.

Nan Jiang

Nan Jiang is Assistant Professor in the Department of Social Work, Faculty of Arts and Social Sciences, at the National University of Singapore. Her research focuses on policies affecting health in later phases of the adult lifespan and intergenerational supports.

Neeraj Kaushal

Neeraj Kaushal is Professor of Social Policy and Chair of the Doctoral Program at Columbia School of Social Work. She is a Research Associate at the National Bureau of Economic Research. She is also Research Fellow at IZA – the Institute for the Study of Labor, Bonn, Germany. She is a labor and health economist and her research focus is on how policies and events affect the well-being of low-income families, with a special emphasis on immigrants.