MI_6_preliminares.p65 NOTA CRÍTICA Does Marriage after Migration Lead to Better Labor-Market Outcomes? Lorenzo Blanco University of Texas at El Paso Do women who marry after mi- gration have a greater chance of finding a job than do women who marry before migration? Do wo- men who marry after migration earn more, or have greater returns on their education, than do wo- men who marry before migration? Do their husbands earn higher wages than those of women who marr y before migration? These questions address a major question in immigration policy: Should that policy be based on family re- unification? Some researchers be- lieve that women who marry after migration are more independent and more successful in the job market. This belief may lead policymakers to curtail policies designed to reunite immigrant families. Harriet Duleep and Seth Sand- ers (1993:684) find that women who marry after migration are more likely to participate in the labor market than are women who marry before migration. However, they found this to be true for women from Japan, China, and the Phil- ippines but not true for women from Korea, India, Europe, and Canada. Migrant women have received substantial attention in the litera- ture. Edward Funkhouser and Stephen Trejo (1998) looked at employment rates for migrant women and find that labor-mar- ket success is a function of years in the United States. Leo Chávez and his co-authors (1997) looked at labor-market outcomes as a func- tion of legal status. Douglas T. Gurak and Mary Kritz (1996) compared labor-force participation for Dominican women in the Do- minican Republic and in New York as a function of number of children and marital status, and they found that among the women in New York, single mothers were less likely to work because of ac- cess to welfare programs provided by the state government. Adding different dimensions to this research could enrich it. This essay contributes by measuring the degree of labor-market success for migrant women based on pre- and 174 MIGRACIONES INTERNACIONALES post-migration characteristics of the women and their husbands. The null hypothesis is that women who marry after migration have gone to the United States in search of a better life for themselves and will be better suited for the U.S. labor market in the sense that they may represent a nonrandom, posi- tively selected sample of the mi- grant population. Another mechanism seems to be at work, however, and this moti- vates the alternative hypothesis. Women who marry after migration may have been brought to the United States by their parents at young age rather than coming here of their own volition, making them “tied migrants,” just as are married women who follow their husbands. If a woman who marries after mi- gration came as a child, then it is likely that her success in the U.S. labor market is due to greater lev- els of assimilation. Characteristics of Married Migrant Women This analysis uses a sample of 1,640 married women from Latin America, Asia, and Eastern Europe, taken from the June 1995 Current Population Survey (CPS)1 (see Table 1 for descriptive statistics for key socioeconomic variables). The 1 The June 1995 Current Population Survey contains key variables such as year of ar- rival and the year of birth for each child. This makes it possible to separate fertility, edu- cation, and other socioeconomic variables in pre-migration and post-migration compo- Table 1. Characteristics of immigrant women based on timing of marriage. Married before Migration Married after Migration (Number of obs. = 777) (Number of obs. = 863) Standard Standard Variable Mean Deviation Mean Deviation WIVES Age 41.98 10.98 35.57 10.58 Age at arrival 31.34 9.31 18.29 8.01 Years in U.S. 10.64 8.89 17.28 11.68 Live Births 2.67 1.91 1.87 1.44 Pre-migration births 2.48 2.04 0.29 0.92 Post-migration births 0.19 0.67 1.58 1.44 Education (years) 10.26 4.76 11.45 4.05 Pre-migration education 10.23 4.73 8.33 5.36 Post-migration education 0.03 0.26 3.12 4.99 Wages 4.97 4.47 4.99 4.18 Number of obs. 5 7 3 6 8 7 HUSBANDS Age 45.14 13.20 38.25 12.35 Age at arrival 33.50 13.27 25.55 12.77 Years in U.S. 11.65 10.27 12.70 11.94 Education (years) 10.82 5.19 11.54 4.65 Pre-migration education 10.47 5.45 10.19 5.37 Post-migration education 0.33 1.59 1.35 3.27 Wage 12.13 7.89 12.72 7.95 Number of obs. 5 8 2 6 9 7 NOTA CRÍTICA 175 sample reveals that migrant wo- men who marry after migration have been in the United States longer than women who marry before migration. Women who ar- rive as teenagers or children will be, ceteris paribus, better accli- mated to the United States and its labor institutions. Longer duration of residence in the United States brings with it greater exposure to English (and thus, greater fluency) and to the U.S. educational sys- tem (and thus, educational capi- tal). Women who arrive in the United States before reaching a marriageable age will marry after migration, if they marry at all. What may seem to be “indepen- dence,” that is, having the initia- tive to go to the United States on their own, without following a spouse or with their natal family, may in fact be a simple matter of having more time to assimilate to the United States. The analysis of educational background by pre- and post-mi- gration marriages shows some interesting characteristics. The women in the sample who marry after migration have more edu- c a t i o n , o n a v e r a g e , t h a n d o women who marry before migra- tion. The difference, however, is not very large, only about one year. Overall, however, the level o f e d u c a t i o n f o r a l l m i g r a n t women is low from a labor-mar- ket perspective. Wages for women who marry af- ter migration are similar to those of women who marry before mi- gration. This result is inconsistent with the theory that women who marry after migration find higher paying jobs. Theoretically, single women who migrate do so in search of work that is more com- patible with their skills and prepa- ration (even if factors might also motivate the migration). Hence, once married, they would already be installed in jobs that they had sought out and for which they were prepared, and consequently, bet- ter compensated. However, in re- ality, it appears that they obtain jobs that are just as incompatible as those chosen by women who were married before migration. For women who marry after mi- gration, the husband generally does not have more education than the wife, and the difference in educa- tional level is only about one year. The notion of “marriage markets” implies that a woman “shops around” for a “better” spouse (usually con- strued as one with a higher income). In the case of migrant women, it implies that she is seeking a “better” spouse than one she could find back home. The descriptive data on hus- bands’ earnings do not bear this out: In the sample, of the husbands of women who married before migra- tion and those who married after migration are nearly identical, ap- proximately $12.00 an hour. nents. Other years of the CPS do not contain this type of information. In the analysis, edu- cation and fertility are divided into pre-migration and post-migration components based on the year of the immigrant’s arrival in the United States. 176 MIGRACIONES INTERNACIONALES Labor-market Outcomes and Timing of Marriage In this statistical analysis, the depen- dent variable is the probability that a woman reports a positive number of work hours for the week before the survey. The migrant women are divided into two groups: women who marry before migration (777 observations) and women who marry after migration (863 observations). Table 2 lists the magnitude and sta- tistical significance of all the esti- mated coefficients used in the PROBIT procedures. For all women in the survey, a stay in the United States of five years or less reduces the probability of em- ployment.2 Regardless of when a migrant woman marries, during the initial years in the United States, she goes through a difficult adjustment period (Baker and Benjamin, 1997; Ford 1990). However, this adjust- ment is more taxing for women who Table 2. Probability of being employed. Married before Migration Married after Migration (Number of obs. = 777) (Number of obs. = 863) Standard Standard Variable Mean Deviation Mean Deviation Constant 0.247 0.242 -0.075 0.306 Age at arrival 0.005 0.006 0.029*** 0.011 0 <= years of U.S. < 3 -0.599*** 0.158 -0.764*** 0.177 3 <= years of U.S. < 5 -0.625*** 0.184 -0.712*** 0.198 5 <= years of U.S. < 7 -0.197 0.188 -0.284 0.199 0 <= husband’s years in U.S. < 3 0.205 0.145 0.058 0.111 3 <= husband’s years in U.S. < 5 0.511** 0.213 0.346 0.256 5 <= husband’s years in U.S. < 7 0.38 0.205 0.140 0.217 Husband’s union -0.301 0.284 -0.262 0.248 Pre-migration education <= 6 yrs. -0.612*** 0.123 -0.523*** 0.136 6 yrs. < pre-migration education <= 9 yrs.-0.414*** 0.145 -0.181 0.138 9 yrs. < pre-migration education <= 12 yrs.-0.103 0.197 -0.457*** 0.168 Post-migration education (years) 0.306** 0.183 0.061*** 0.017 Pre-migration births -0.074*** 0.028 -0.171 0.056 Post-migration births -0.165** 0.075 -0.121*** 0.036 Resides in Northeastern U.S. -0.013 0.178 -0.077 0.165 Resides in Southern U.S. 0.140 0.191 -0.058 0.174 Resides in Western U.S. 0.005 0.171 0.007 0.153 Mean dependent variable 0.452 0.512 Log likelihood -498.075 -560.092 McFadden R-squared 0.069 0.063 Notes: The dependent variable is whether or not the woman was working the week before the survey. Huber-White standard errors. Omitted region is the Midwestern United States. 2 The wife’s duration of stay in the United States is included in the PROBIT in the form of three dummy variables: 0-3 years in the United States, 3-5 years, and 5-7 years. A re- cent immigrant is defined as a person who has been in the United States for no more than three years whereas someone who has been in the United States for more than seven years is no longer a recent immigrant. The choice in intervals (0-3, 3-5, 5-7 years) is arbitrary but different time intervals do not produce qualitatively different results. NOTA CRÍTICA 177 marry after than it is for women who marry before migration. The coeffi- cients for years in the United States for women who marry after migra- tion are consistently greater in mag- nitude than for women who marry before migration. This is not consis- tent with the theory that women who marry after migration can weather the storms of their first years in the United States better than women who marry before migration. The idea of “countervailing effects” (Duleep and Sanders, 1993:687) suggests that a migrant woman will feel two opposing pressures. The husband’s difficulties in finding work in the United States pressure the women to work (Baker and Ben- jamin, 1997). At the same time, her own adjustment difficulties pull her away from work. If countervailing effects exist in the CPS sample, the amount of time the husband has spent in the United States should have a positive effect on the woman’s ability to find paid work. For women who married before migration, the longer her husband has been in the United States, the more likely it will be that the wife will be employed. The estimated coefficients that measure this effect for the husband are positive. Only one, however, is statistically signifi- cant. Being married to a husband who has been in the United States between three and five years exerts a positive and significant influence on the wife’s probability of getting a job. Taken together, the three husband duration variables are not jointly significant. For women who marry after mi- gration, none of the variables for the husband’s duration of residence has statistically significant coeffi- cients. The direction of those co- efficients, however, is consistent with the notion that countervailing effects influence the probability of a woman’s entrance into the labor market. Countervailing effects are present for all of the women (see Table 2). However, they are stron- ger among women who marry be- fore migration than for women who marry after migration. In terms of these effects, the labor-force par- ticipation of a woman who mar- ries after migration is more inde- pendent from her husband’s duration of stay in the United States. Fertility, whether pre- or post-mi- gration, has the expected negative effect on the probability of being employed. Post-migration fertility, however, is more significant than pre- migration fertility. One possible ex- planation may be because post-mi- gration because children born after the migration are younger and hence need more extensive childcare. Only post-migration births affect labor- force participation for women who marry after migration. This is also an expected result: most of the women who marry after migration arrive to the United States at a very young age, and most of their repro- ductive years are spent in their new country. Pre-migration education plays an interesting role in the labor-market outcomes of women who marry be- 178 MIGRACIONES INTERNACIONALES fore migration.3 Pre-migration edu- cation has a negative effect on the chances of a woman getting a job. However, the more education the woman has before migration, the lower the magnitude of that nega- tive effect. The estimated coefficient for six years or less of pre-migration education is –0.612, and between six to nine years -0.414. Both of these estimated coefficients are statistically significant at the 1% level. The esti- mated coefficient for nine to twelve years of pre-migration education is -0.103, but it is not statistically dif- ferent from zero. Post-migration edu- cation is statistically significant, and its estimated coefficient is positive. The progression from negative to positive returns is clear. Women who marry after migration also show an interesting reaction to pre- and post-migration education. Pre-migration education has a nega- tive influence on the probability that these women will find work. The estimated coefficient for six years of education is –0.523; it then goes down for six to nine years of educa- tion (–0.181), and then back up to –0.457 for nine to twelve years of education. The estimated coefficient on post-migration education is posi- tive and significant as expected. Whether the women marry before or after migration, one can see the progression from the negative influ- ence of pre-migration education to the positive influence of post-migra- tion education on the ability to find work. It is important to remember, however, that these women have a low level of overall education: the average is well below a high-school- level education (10.26 years for women who marry before migration and 11.45 years for women who marry after). It would appear that it is not just the level of education but the content of that education that matters. Conclusions Women who marry after migration have a slightly higher chance of get- ting a job than women who marry before migration. (The mean of the dependent variable in the PROBIT re- gressions is 0.452 for women who marry before migration and 0.512 for women who marry after migra- tion.) Both groups exhibit positive returns to pre-migration education. The estimated coefficient for pre-mi- gration education is negative at all levels of pre-migration education. However, the higher the level of pre- migration education, the lower the magnitude of that negative coeffi- cient. Both groups of women marry men who are similar to each other in terms of education and earnings. Women who marry after migration are not very different from women who marry before migration. Furthermore, women may not base their choice of spouse solely 3 Pre-migration education is divided into three dummy variables: 6 years or less, be- tween 6 and 9 years, and between 9 and 12 years, reflecting three levels of education (pri- mary, secondary, and high school). Post-migration education is measured simply by the number of years because its level is low. NOTA CRÍTICA 179 on economic criteria. Laura Hill (see this issue) emphasizes that some women from Mexico and Central America migrate to escape restrictive social environments and social stigma. Given those factors and their potential for expanded earnings power in the United States, these women may define a “better” marriage market as one in which they can more easily locate a partner whose attitude is less “traditional” in comparison to what could be found in their send- ing communities. This research concludes that in terms of labor-market performance and the characteristics of the hus- band they marry, there are no sub- stantial differences between women who marry before and those who marry after migration. A policy designed to exclude women who marry before migration may not necessarily have a significant effect on labor-force participation of im- migrant women. Although non- economic factors may be harder to measure, U.S. immigration policy should take them into account. References Bergstrom, Theodore C., “A Survey of Theories of the Family”, in M.R. Rosenzweig and Oded Stark (eds.), Handbook of Population and Family Economics, vol. 1A, Amsterdam, North-Holland, 1997, pp. 21-79. ———, and Mark Bagnoli, “Courtship as a Waiting Game”, Journal of Political Economy,101(1), 1993, pp. 185-202. Chávez, Leo R., F. Allan Hubbell, Shiraz I. Mishra, and R. Burciaga Valdez, “Undocumented Latina Immigrants in Orange County, Cali- fornia: A Comparative Analysis”, International Migration Review, 31(1), 1997, pp. 88-107. Duleep, Harriet Orcutt, and Seth Sanders, “The Decision to Work by Married Immigrant Women”, Industrial and Labor Relations Review, 46, 1993, pp. 677-90. Funkhouser, Edward, and Stephen J. Trejo, “Labor Market Outcomes of Female Immigrants in the United States”, in National Research Council (ed.), The Immigration Debate, Washington, D.C., National Academy Press, 1998, pp. 239-88. Gurak, Douglas T., and Mary M. Kritz, “Social Context, Household Com- position and Employment among Migrant and Nonmigrant Domini- can Women”, International Migration Review, 30, 1996, pp. 399-422. Hill, Laura E.,“Managing Risk through Migration: Marriage, Divorce, and Childbearing Among Mexican and Central American Migrants”, Ph.D. dissertation, University of California at Berkeley, 1998.