Some Thoughts on Segmented Assimilation

by Reihan Salam

Recently, Sean Trende of RealClearPolitics, author of The Lost Majority, shared a 2012 research report from the Pew Research Center on Hispanic identity. His intention was to rebut the notion that Americans ought to be concerned about the assimilation of recent immigrants, and in particular recent Latino immigrants, by pointing to linguistic assimilation rates. The report distinguishes between “Spanish-dominant” and “English-dominant” persons. The former are those who speak and read Spanish very well, yet who rate their English language proficiency lower. Those who identify themselves as “bilingual” consider themselves proficient in both languages. By the second generation, 8 percent of U.S. Latinos are Spanish dominant, 53 percent are bilingual, and 40 percent are English dominant. For the third and higher generations, only 1 percent are Spanish dominant, 29 percent are bilingual, and 69 percent are English dominant. In Trende’s view, as I understand it, these linguistic trends suggest that the U.S. melting pot is working as well now as it has in the past. Though Trende isn’t advancing a policy argument by making this point about linguistic assimilation, the notion that assimilation is proceeding apace for recent Latino immigrants is central to the case for an increase in legal immigration levels. 

First, it is worth pointing out that the linguistic assimilation trend is not necessarily as impressive as Trende seems to think. “Third and higher generations” is a category that encompasses the grandchildren of immigrants, yet it also includes the great-grandchildren of immigrants and the great-great-grandchildren of immigrants. If there are people who are seriously concerned that there are large numbers of third-, fourth-, and fifth-generation U.S. Latinos who are Spanish dominant, it is reasonable and appropriate to dismiss them as unserious. But the salient question for the immigration policy debate is whether it is reasonable and appropriate to expect that new immigrants have a reasonably high baseline level of English language proficiency. Among first generation U.S. Latinos, 38 percent are bilingual and 24 percent are English dominant, leaving 38 percent who are Spanish dominant. Spanish-dominant parents are, for obvious reasons, more likely to have Spanish-dominant children; and 8 percent of second-generation U.S. Latinos are Spanish dominant. If the U.S. created a strong bias in favor of immigrants who already have a high level of English language proficiency, the share of first-generation Latinos who are Spanish dominant would decrease over time, as would the number of Spanish-dominant second-generation Latinos; the share of bilingual U.S. Latinos would presumably increase to make up the difference. While it may well be true that earlier waves of immigrants experienced a similar linguistic assimilation trajectory, it is also true that in a knowledge-intensive, service-oriented economy, English language proficiency might actually be more important than in earlier eras in which semi-skilled labor was a reliable path to upward mobility and economic self-reliance.

The Migration Policy Institute has reported that the number of limited English proficiency (LEP) individuals in the U.S. increased by 80 percent from 1990 to 2010. And in 2012, Chhandasi Pandya of MPI offered useful context for understanding this population:

There were 23 million immigrants in the US civilian labor force in 2010, accounting for one in six workers ages 16 and older — a figure that has roughly doubled since 1990. Within the immigrant labor force, nearly half speak English less than “very well” (as defined by the US Census Bureau) and are classified as “Limited English Proficient,” a language barrier that affects their employability and wage-earning potential. Though the WIA system is the core source of federal support for workforce training, English as a Second Language (ESL) instruction, and other adult education programs, the current design has created barriers for LEP workers seeking workforce services. In particular, misalignment in goals, program services, and performance accountability mechanisms across WIA-funded programs has made it difficult for these workers to obtain the language and work training services they need to advance in the workforce. At the same time, the LEP population continues to grow. In 2000, about 17.8 million US adults, or 8.5 percent of the population, had limited English proficiency. Today, there are roughly 22.5 million LEP adults, accounting for 10 percent of the US adult population and 9 percent of the labor force. [Emphasis added]

These numbers are not so large that we can’t expect that the children and grandchildren and great-grandchildren of today’s LEP individuals will never become English dominant. But given that access to the U.S. labor market is an attractive prospect for workers living in middle-income and low-income Spanish-dominant societies, it also seems reasonable to suspect that if U.S. immigration policy shifted towards requiring that new arrivals achieve a high degree of English language proficiency before settling in the U.S., many aspiring immigrants would do just that. Far from a reflection of cultural prejudice, such a requirement would make it more likely that new arrivals would be employable and strong wage-earning potential, thus facilitating absolute and relative upward mobility for the immigrants in question and their children. We could think of a strong English language proficiency requirement as a “nudge” that would help ensure that immigrants are in the best possible position to flourish in U.S. society. Moreover, there is good reason to believe that because labor-intensive services are more expensive in affluent countries than in less-affluent countries, the strategy of requiring that aspiring immigrants achieve English language proficiency before they arrive in the U.S. will prove more cost-effective than the alternative, in which U.S. charities and public sector institutions take on the task of investing in the human capital of adult immigrants.

Second, the fact that linguistic assimilation is achieved as families go from the first to the second to the third generation doesn’t necessarily tell us much about other aspects of the assimilation process. In a 2011 article in Social Fores (“Dreams Fulfilled, Dreams Shattered: Determinants of Segmented Assimilation in the Second Generation“), William Haller, Alejandro Portes, Scott M. Lynch summarized the work on assimilation broadly understood. They begin by observing that immigrants and the children of immigrants now represent roughly one-fourth of young Americans, and by differentiating between the social and economic trajectories of high human capital immigrants and manual labor immigrants:

The concept of “mode of incorporation” was coined to highlight key aspects of these contexts of reception pertaining, respectively, to the attitudes of the authorities and the public at large, plus the character of the pre-existing ethnic community. By and large, the mode of incorporation of high human capital immigrants is positive: it is defined by legal status and a receptive, or at least neutral, stance by the native-born population (Portes and Rumbaut 2006; Zhou 2001). When a co-ethnic community exists, it is generally affluent, being formed by earlier migrants with comparable levels of education. By contrast, manual labor immigrants commonly arrive illegally and, by reason of this status, a relatively low level of education and predominantly non-white physical features, are subjected to a negative reception by the authorities and the host population (Suarez-Orozco 1987; Rumbaut 2005). Also, pre-existing co-ethnic communities are frequently weak and lacking in resources needed to counteract a negative official reception (Lopez and Stanton Salazar 2001; Menjivar 2008).

Early on, Haller et al. engage with the work of Joel Perlmann and Roger Waldinger, two sociologists who have argued that despite the fact that contemporary immigration is divided between high human capital immigrants and manual labor immigrants experiencing different modes of incorporation, the assimilation process will unfold as it has in the past with, for example, Italian immigrants and their descendants in the early 1900s. They observe that Perlmann’s data do not bolster his case, and that he repeatedly has to acknowledge that Mexican-American children face formidable barriers in competing for educational and upward mobility that Italian immigrants did not. Mexican Americans lag far behind their native white counterparts in academic achievement, and at the same time the returns to education are much higher than they were a century ago. Comparing the current era to the early 1900s thus seems extremely ill-advised. (As of 2000, for example, the high school dropout rate for second-generation Mexican-American males was 33 percent — considerably higher than the 16 percent rate for black males and the 9 percent rate for native white males.)

The authors then elaborate on the segmented assimilation hypothesis. The idea is that differences in human capital (skilled vs. less-skilled), legal status (authorized vs. unauthorized), and family structure (families that remain together vs. families that do not) translate into different patterns of assimilation. A number of factors might stymie successful adaptation, including racial prejudice; changes in the labor market that have made the relative distance between highly-paid professional occupations and low-paid manual jobs quite substantial; and the emergence of the drug trade as a viable option for those who would prefer not to finish school or work in a conventional job. Second-generation Americans who benefit from high human capital parents, stable families, and a positive mode of incorporation tend to succeed. Some who are raised by parents with low incomes and limited education also succeed, thanks to strong families and supportive co-ethnic communities. Others, however, find it difficult to overcome the problems tied to low parental human capital, a negative mode of incorporation, weak communities, and unstable families.

After drawing on data from the Children of Immigrants Longitudinal Study (CILS), the authors find strong evidence for the segmented assimilation hypothesis. They find that negative modes of incorporation (e.g., having parents who arrive in the country as unauthorized immigrants) are strongly associated with a number of negative outcomes, including arrests and incarceration rates, teenage childbearing among girls, and early school abandonment. Among other things, they highlight the fact that the share of second-generation Mexican- and Caribbean-origin men in prison almost matches that of African Americans while rates of adolescent child-bearing and school abandonment are higher than those seen in domestic minorities. Because they are skeptical of cultural theories, in part because Cubans fare so much better than Mexicans, despite being culturally close, the authors maintain that distinct modes of incorporation are the most plausible explanation for the divergence in life outcomes across immigrant groups. This leads them to the following policy conclusions:

From a policy standpoint, the implications of segmented assimilation are clear. A sizeable proportion of legal immigrants with high levels of human capital is poised to follow a smooth adaptation path, with the majority of their offspring achieving high levels of education and moving solidly into the middle class. At the other extreme, there is the mass of poor and unskilled immigrants coming to fill the labor needs of the American economy. These immigrants face the challenges posed by the poor areas where they settle with few individual resources or external assistance.

Downward assimilation may be regarded as the consequence of the clash between the benefits of unauthorized labor, that are privatized, and its costs, that are socialized. If manual workers are needed in agriculture, construction and other sectors of the economy, they should be brought legally and encouraged to return home after a period of time (Massey 1998; Portes 2007). However, those who settle and create families on this side of the border should be vigorously assisted, to minimize the realities documented by this research. Additional results show that a proactive stance by teachers and counselors and external voluntary support programs can make a significant difference helping children of immigrants overcome the handicaps of a negative mode of incorporation (Fernández-Kelly 2008; Konczal and Haller 2008). Active external intervention in support of these families and their aspirations is needed, lest the country’s hunger for cheap labor devolves over time in the emergence of a new underclass. The proven effectiveness of external voluntary programs targeting these youths offers a blueprint and a ray of hope in an otherwise troubling landscape. [Emphasis added]

Some readers will assume that it is the difference in legal status that matters most, and that the work of Haller et al. ought to lead us to the conclusion that the real problem with our current approach to immigration is that there is such a thing as unauthorized immigration at all. But it seems that it is the interaction of several different factors, like parental human capital, legal status, and family structure, that helps determine the success of the immigrant adaptation process. And if this is the case, we ought to think carefully about a comprehensive immigration reform effort that will not only regularize the status of the current unauthorized immigrant population, but that will also increase the size of the less-skilled influx, which is to say the influx of families in which parents have low levels of education and income and children will have many serious disadvantages to overcome as a result. Rather, we should recognize that regularization and increasing less-skilled immigration might actually work at cross-purposes. Because the current population of unauthorized immigrants and their children will have serious disadvantages relative to other people living and working in the U.S., we need to embrace policies that will help this population adapt successfully to American life. A good first step would be to greatly reduce less-skilled immigration, as new less-skilled immigrants tend to compete with less-skilled immigrants who already reside in the country, and who are raising children who are going to be a central part of the American future.

(For more on immigration, I recommend the latest missives from Ross Douthat and Ramesh Ponnuru.)

The Agenda

NRO’s domestic-policy blog, by Reihan Salam.