Cultivating the Next Generation: Three Studies of College Access among the Children of Immigrants

Currently there is great debate on immigration reform and, given migration trends, policymakers may be interested in finding ways to integrate these new Americans and facilitate their social mobility. Building on research on immigrant assimilation, cultural, social, and human capital, and college choice, this dissertation explores the postsecondary pathway among the children of immigrants. Ensuring smooth transitions to college may be difficult for the children of immigrants since their families are likely unfamiliar with the U.S. system of higher education, which is horizontally stratified and requires a number of steps before admission. Immigrant children, therefore, may have different knowledge about the pipeline to college and react to interventions and policies in varied ways. Additionally, such heterogeneity may depend on the resources that they are able to draw upon within their communities or on policies that limit or facilitate their access to higher education. Testing tenets of assimilation theory, I explore differences by immigrant generation and co-ethnic community human capital as well as the role of state policy contexts and legal status in moderating educational outcomes for immigrant youth.

In the first paper, I explore inequalities in college knowledge using data from the High School Longitudinal Study of 2009. Specifically, I examine whether 11th graders can accurately estimate tuition and fees at two- and four-year institutions and the expected earnings of various levels of educational attainment. Using logistic and ordinary least squares regression analyses, I examine differences in the ability to estimate and the estimation accuracy of college costs and the returns to education among first, second, and third-plus generation immigrants. Although first and second generation immigrants appear to have less information on tuition and earnings, most of these differences are rendered insignificant with control variables. Differences persist, however, with public four-year tuition and fees. First and second generation gaps in these cost estimates appear to concentrated among students in the top two quintiles of the socioeconomic and achievement distributions and among students with four-year college plans. There is some evidence that college knowledge and enrollment are correlated, although the relationship appears to be indirect and possibly operates through information-seeking behaviors.

The second paper builds on the first by testing for heterogeneous treatment effects of social interactions in high school using the Education Longitudinal Study of 2002. If there are differences in college knowledge, it is possible that discussing college with teachers, counselors, and college representatives has disproportionate effects for the children of immigrants. Using propensity score methods to model selection into the treatment condition, I estimate multinomial logistic regression models to determine the effects of social interactions on college choice (none, 2-year, 4-year). With treatment condition interactions, I test for differences by immigrant generation and co-ethnic human capital; the latter is an index constructed using ethnicity- and location-specific data from the U.S. Decennial Census and measures the resources embedded in an ethnic enclave that may affect information and norms. While meeting with teachers, counselors, and college representatives increase college enrollment, there is no evidence of treatment effect heterogeneity, challenging components of classical and segmented assimilation theory. More importantly, if college access among the children of immigrants is a policy priority, interventions not targeted toward certain groups will not reduce inequality.

The final paper examines whether state policy contexts affect college choice and preparation among unauthorized immigrants. Since 2001, 27 states have passed policies that affect college access for undocumented students. Some states have banned them from receiving in-state resident tuition (ISRT), while others allow it or allow it and provide financial aid. First, using the Current Population Survey October Supplement, I estimate the effects of these policy regimes on college choice using a heterogeneous treatment effect difference-in-differences (DiD) research design that compares Mexican foreign-born non-citizens (which include the undocumented) to Mexican citizens. Second, using data from California and Texas, I construct a school-by-race panel dataset to determine if accommodating ISRT policies affect whether high school students meet state curriculum guidelines and earn an advanced diploma. I estimate treatment effects using OLS and a DiD research design comparing Hispanic, white, and black students. Results show that statewide bans reduce college enrollment among likely-unauthorized youth while ISRT policies that offer state financial aid increase enrollment; there is no measureable impact of ISRT without additional aid, probably because unauthorized immigrant families are quite poor. In contrast, I find positive effects of ISRT, with and without additional aid, on college preparation. ISRT policies may improve intermediate steps in the pathway to college, but not enhance enrollment for particularly poor populations.