Occupational Enclaves and the Wage Growth of Latino Immigrants. [Word count: 8,830] Sergio Chavez Ted Mouw Jacqueline Hagan

Transcription

1 Occupational Enclaves and the Wage Growth of Latino Immigrants [Word count: 8,830] Sergio Chavez Ted Mouw Jacqueline Hagan University of North Carolina, Chapel Hill

2 Occupational Enclaves and the Wage Growth of Latino Immigrants ABSTRACT Does the concentration of recent Latino immigrants into occupational enclaves occupations with large numbers of limited English speakers restrict their wage growth? On the one hand, it is possible that immigrants who are concentrated in jobs with co-ethnics may have less need to learn English and/or less on-the-job exposure to it, which may isolate them socially and linguistically and limit their subsequent economic mobility. On the other hand, enclave employment can be seen as a stepping stone for upwardly mobile immigrants who can find work while they improve their English. Using longitudinal data from the 1996, 2001, and 2004 panels of the Survey of Income and Program Participation (SIPP), we test for the effect of occupational level English proficiency on wage growth. We supplement this data with in-depth interviews and observations from immigrants employed in the construction industry. The results indicate that although the proportion of limited English speakers in the respondent s occupation is associated with lower wages for Latino immigrants in the cross section, it is not associated with lower levels of wage growth. These findings demonstrate that occupational enclaves do not trap immigrant workers at least on average but instead can provide a path for immigrants to familiarize themselves with the U.S. labor market.

3 Introduction A central question in the public and academic debate on immigration focuses on the economic assimilation of recent immigrants. While conventional models of assimilation treat the low wages of recent immigrants as the first step on a ladder of upward mobility, proponents of the segmented assimilation perspective argue that reduced opportunities for less educated workers in a postindustrial economy combined with phenotype discrimination may result in the downward assimilation of less educated, darker skinned immigrants (Bean, Leach, and Lowell 2004; Portes and Rumbaut 2006). As the largest group of post-1965 immigrants, with relatively low education levels and the possibility of social and labor market discrimination, Latino immigrants present an important test case for these contrasting perspectives on contemporary immigration in the United States. Empirical evidence on the economic mobility of Latino immigrants paints a mixed picture. Using repeated cross-sections of Census data, researchers have documented that Latino immigrants with the exception of Cubans earn lower wages, on average, than their native counterparts throughout their working lives (Borjas 1982, 1985, 1995, Trejo 1997). Borjas and Katz (2005) show that the largest group of Latinos, Mexicans, lags the farthest behind the nativeborn in terms of wages and education and argue that this disadvantage is transmitted across generations. Lubotsky (2007) uses longitudinal data on earnings and finds that Latino immigrants have lower rates of wage growth than other immigrant groups, and that wage convergence with native-born workers stalls after 10 years in the United States (Lubotsky 2007). In contrast, Smith (2003, 2006) argues that there is considerable evidence of intergenerational educational and earnings gains among Latino immigrants and that concerns about a lack of assimilation are unwarranted. Similarly, Bean, Leach, and Lowell (2004) and Hall and Farkas 1

4 (2008) argue that there is considerably more upward occupational and wage mobility among recent immigrants than one might expect given the expectations of racial stratification and segmented labor market theories. Overall, while there is widespread agreement that a substantial portion of the wage gap between Latino immigrants and native-born workers is due to differences in education and English language ability at the time of immigration (e.g., see Catanzarite and Aguilera 2002), there is considerable disagreement as to the explanation of the remaining wage gap, the degree to which it persists over time, and what this portends for the future. One explanation for the residual wage gap and the apparent lack of wage convergence for Latino immigrants focuses on occupational segregation. Catanzarite (2004) argues that Latino immigrants are crowded into brown collar occupations that have been typecast as immigrant jobs where they receive low wages and have limited prospects for upward mobility. Catanzarite and Aguilera (2002) find that working in jobsites with co-workers who are Latino is associated with lower wages for Latino workers, concluding that working with co-ethnics erodes pay by the equivalent of 8.0 years of education for men or 7.4 for women. Kmec (2003:54) argues that individuals with mostly white co-workers have an unmistakable advantage over those with mostly black or Latino co-workers. Chiswick and Miller (2002) argue that working alongside co-ethnics who speak a minority language has a feedback effect that slows the rate of economic assimilation for immigrant workers. Without discounting the potential role that discrimination and job labeling might play in reducing the wages of workers in brown collar occupations, we argue in this paper that the existing literature is misleading because it relies on cross-sectional data and is not based on a realistic model of immigrant assimilation, where low-skilled immigrants arrive with poor English 2

5 language skills and limited U.S. labor market experience. In contrast, we borrow from the literature on ethnic enclaves to propose the concept of occupational enclaves for Latino immigrants occupations with substantial numbers of Spanish speakers where recent immigrants can find employment while they accumulate job skills, U.S. experience, and English language ability. Rather than hurting wages and opportunities, we argue that occupational enclaves can actually promote the subsequent wage growth of immigrant workers. There are, however, two key differences with the conventional definition of an ethnic enclave. First, there is no presumption that the owner of the hiring firm is a co-ethnic, although in many cases this may be true. Second, enclave occupations are stepping stones for subsequent mobility, and the potentially benefit is only visible in longitudinal data that can trace the upward mobility of successful immigrant workers. In this paper, we use longitudinal data from the 1996, 2001, and 2004 panels of the Survey of Income and Program Participation (SIPP) to test for the effect of working in linguistic enclave occupations on wage growth for Latino immigrants. We merge our SIPP data with state level data on English language ability and Spanish language use within 3-digit Census occupations from the 2005 American Community Study. This allows us to define occupational enclaves based on geographically specific information on the occupational distribution of recent immigrants. We compare cross-sectional and longitudinal estimates of the effect of working in enclave occupations on wages to test whether working in occupations with a substantial proportion of Spanish speakers affects the economic mobility of Latino immigrants. 3

6 Literature Review: Occupational Segregation and Ethnic Enclaves As discussed above, a number of recent papers have made a link between the relatively high levels of occupational concentration among Latino workers, particularly recent immigrants, and the Latino/white pay gap. Catanzarite (2000), for example, uses Public Use Micro Sample (PUMS) data from Los Angeles for 1980 and 1990 to show that Latinos are highly segregated from white workers, and that their earnings are lower than whites even after controlling for education, potential labor market experience, English ability, and family composition. She uses the term brown-collar occupations to refer to occupations with a high proportion of Latino workers, and argues that the labeling of jobs as immigrant jobs simultaneously makes them less desirable, lowers their status, and reduces wages. In a follow-up study, Catanzarite (2003) uses 1990 PUMS data from 18 metropolitan areas to estimate a multilevel model of the effect of immigrant density on wages, and finds that working in an occupation with recent Latino immigrants reduces the wages of all workers, with more pronounced negative effects for blacks and earlier Latino immigrants. In a related study, Catanzarite (2002) uses 1980 and 1990 Census data from Los Angeles to test the relationship between earnings and Latino concentration at the occupational level using a cross-lagged regression model. She finds a negative relationship between occupational earnings in 1980 and the change in the percent of male immigrant Latino workers in the occupation between 1980 and 1990, but this effect disappears after she controls for the occupation s rate of employment growth and the education and experience of native born male incumbents. This finding suggests that a decline in the earnings of brown collar occupations might be an artifact of a general erosion of pay for less-educated workers of all race and ethnic groups during the 1980s. In contrast to the results in Catanzarite (2002), Howell and Mueller 4

7 (2000) use PUMS data from the New York metropolitan area to estimate models of the change in wages on changes in the immigrant occupational share from , and find no effect of changes in immigrant occupation density on changes in the wages of recent immigrants or Latino workers. In addition to effects at the occupational level, it is possible that the negative effect of immigrant concentration on wages is more pronounced at the job or firm level. To test this, Catanzarite and Aguilera (2002) use data from the 1992 Legalized Population Survey, which includes a categorical variable asking respondents to identify the largest race/ethnic group among their coworkers. They found that, on average, legalized Latino males tend to earn 13% less when work at predominately Latino jobsites, even after controlling for occupational characteristics and an extensive set of human capital variables. Kmec (2003) uses data on the race and ethnic composition of jobs from the employer survey of the Multi-City Study of Urban Inequality, and finds that working in jobs with predominately Latino or black co-workers reduces the wages of all workers by 18 and 15 percent less per hour, respectively, compared to jobs in which whites are the majority (Kmec 2003). In perhaps the most comprehensive test of the correlation between Latino concentration and wages at the firm level, Hellerstein and Neumark (2002) use a large data set of matched employer-employee records constructed from the Decennial Census of They document substantial firm level segregation of Latino workers by ethnicity and the degree of English proficiency. Their regression results for log wages show that the share of co-workers who are Latino reduces wages by log points for Latino workers and.037 log points for white workers. 5

8 Although the occupational segregation literature reviewed above argues that working with co-ethnics has a negative impact on the wages of Latino immigrants, the existing evidence is primarily based upon analysis of cross-sectional data or aggregate-level analysis at the occupational level. While it may be true, as this literature suggests, that the crowding of recent immigrants into brown-collar occupations reduces wages in those occupations and represents a structural constraint on race and ethnic equality, the overall portrayal of the economic assimilation of Latino immigrants is too static. Most importantly, the lack of longitudinal data doesn t permit an accurate assessment of the degree to which some immigrant workers move out of these highly segregated occupations over time. If substantial numbers of successful immigrants do experience upward mobility, then any cross-sectional estimate of the effect of working in a brown collar occupation on wages will be biased by the negative selectivity of the remaining workers. In the next section, we turn to the ethnic enclave hypothesis to develop an alternative hypothesis based upon a consideration of the challenges facing recent immigrants, who often arrive in the United States with little English ability and no experience in the U.S. labor market. Ethnic Enclaves: Stepping Stone or Dead End? The ethnic enclave debate of the 1980s helps to situate our reinterpretation of the role of occupational concentration on immigrant wage growth. The central issue in the ethnic enclave literature is whether the sorting of immigrants into segregated neighborhoods and workplaces promotes or impedes their economic and social assimilation (see Sanders and Nee 1987; Wilson and Portes 1980; Zhou and Logan 1989). In general, this literature begins by noting that when immigrants initially arrive in the U.S., they are often relegated to the secondary employment 6

9 sector, which is characterized by low-wages and job insecurity (Piore 1979). Portes and Bach (1985) and Wilson and Portes (1980) argue that immigrants may benefit from avoiding the secondary sector of the labor market and working instead in ethnic enclaves composed of immigrant owned firms. Wilson and Portes (1980) claimed that Cubans employed in an enclave economy earned wages and received a rate of return to human capital comparable to those employed in the primary sector. In contrast, however, Sanders and Nee (1987) argue that these results are misleading and that the main beneficiaries of enclave employment are the employers who benefit by hiring cheap immigrant labor. Sanders and Nee claim that the enclave hypothesis is counterintuitive in that it assumes that despite the social isolation of the enclave, there is no cost to segregation. (Sanders and Nee 1987: 746). Recent research on the ethnic enclave hypothesis has produced a pessimistic and an optimistic view. On the optimistic side, Evans (2004) uses cross-sectional census data from Australia and finds that the proportion of co-ethnics who are entrepreneurs is positively correlated with the occupational status of immigrant workers, while an interaction term between entrepreneurial density and the respondent s English language ability indicated that the benefits of working in enclaves decline as English language skills increase. Edin et al. (2003) uses initial government placement of refugees in Sweden to attempt a quasi-experimental test of the enclave hypothesis, arguing that the initial placement was independent of unobserved factors that otherwise would have influenced the location decision. They find that the earnings of the lesseducated refugees were 13% higher when the size of the ethnic enclave size was increased by a standard deviation. Damm (2006) uses a similar approach based upon a government relocation program in Denmark and finds that a standard deviation increase in the size of the ethnic enclave results in a 4 percentage point increase in employment and a 21 percentage point increase in 7

10 earnings. She interprets these results by arguing that ethnic enclaves provide access to ethnic networks that transmit information about their host country which they would otherwise be excluded from. Finally, Hagan (1998) and Waldinger and Lichter (2003) provide cases studies of the dynamics of ethnic enclaves and argue that in a mature ethnic economy, co-ethnic networks of immigrant workers may come to dominate certain firm or occupational niches, from recruiting and hiring networks to managing work schedules and supervising promotion, resulting in a form of social closure. In these ethnic based economies English is not a requirement for landing a job or acquiring new skills to augment wages. The pessimistic side of the enclave debate focuses on the possibility that working alongside co-ethnics impedes the social assimilation needed for subsequent wage growth. One line of inquiry focuses on linguistic assimilation. Studies of the determinants of wages for Latino and immigrant workers demonstrate that English proficiency is a critical component of the wages of immigrant workers (Cobb-Clark and Kossoudji 2000; Dustmann and Fabbri 2003; McManus 1985; McManus et al. 1983). Building on this, Chiswick and Miller (2002) use PUMS data from the 1990 Census to test a version of the enclave hypothesis based on linguistic concentration. Using a measure of the proportion of speakers of the immigrant s language group at the state level, they find that linguistic concentration is associated with both lower levels of English proficiency and lower earnings for immigrant workers. In a similar article, Warman (2007) uses synthetic cohorts constructed from Canadian Census data from and finds that living in an ethnic enclave is negatively associated with wage growth for immigrants. Overall, the findings based on the quasi-experimental studies of Edin (2003) and Damm (2006) discussed above would seem to be preferable to the results in Chiswick and Miller (2002) and Warman (2007), as the latter studies are based on cross-sectional or synthetic cohort data and 8

11 hence are vulnerable to problems based self-selection based on English ability, as discussed in our theoretical model below. On the other hand, the relatively small size of the ethnic enclaves in the European data that Edin (2003) and Damm (2006) use might point to a qualitatively different process among more larger, more concentrated ethnic enclaves in the U.S or Canada. Overall, the debate on the ethnic enclave hypothesis highlights the complexity of the immigrant economic assimilation process. A crucial component of this hypothesis is the idea that an ethnic enclave may provide an alternative to employment in the secondary sector, and, as such, shelter recent immigrants from direct competition with native workers. Particularly with respect to language, working alongside co-ethnics may provide an entrée into the U.S. labor market for recent immigrants with poor English language skills. As Chiswick and Miller (2002:10) note working within a linguistic enclave is a mechanism for sheltering oneself from or mitigating the adverse labor market consequences of limited destination language proficiency. Three other studies provide important antecedents to our research. First, Chiswick and Miller (2007) use Census PUMS data matched to occupational-level information on English language requirements to test a variant of the enclave hypothesis, and they find that there is a strong positive correlation between occupational English requirements and wages in the crosssection. As a result, immigrants are assumed to be disadvantaged if they do not posses the language of the workplace because they must rely on co-ethnics in the labor process. Second, Hall and Farkas (2008) use data from the 1996 and 2001 panels of the Survey of Income and Program Participation (SIPP) to estimate growth-curve models of wage growth among immigrants and native workers. They find that while the initial wage level is considerably less for immigrants compared to native workers, the estimates of wage growth are statistically 9

12 indistinguishable among natives and different immigrant groups. Finally, Lubotsky (2007) matches earnings data from Social Security records to demographic information from the SIPP and the Current Population Survey (CPS) to study wage growth among immigrant workers. In contrast to Hall and Farkas (2008), Lubotsky s longer panels, augmented with Social Security data, suggest that earnings growth for Latino workers lags behind that of other immigrant groups. In the following section, we combine the theoretical perspective of the ethnic enclave literature with an empirical focus of the brown collar occupations literature to depict a simple model of occupational sorting among immigrant workers. We argue that although immigrants may initially sort into enclaves, this concentration is not necessarily bad during an initial period of adjustment if they eventually are able to develop the skills needed to succeed in the larger labor market. Then, in the analysis section, we propose to combine the empirical approaches of Chiwsick and Miller (2007), Lubotsky (2007) and Hall and Farkas (2008) in order to test a dynamic model of brown collar occupations and wage growth with multiple panels of SIPP data. A Theoretical Model of Occupational Enclaves For recent immigrants, a lack of English fluency and limited knowledge about opportunities represent major constraints in the labor market. Given these constraints, occupational enclaves occupations where there are a significant number of Spanish speakers provide employment for immigrant Latino workers with insufficient English ability, where either the demand for English fluency is minimal or where the presence of large numbers of co-ethnics eases the language difficulties. The key question is whether the sheltering effect of 10

13 working with other Spanish speaking workers reduces wage growth by slowing the process of linguistic and social assimilation. Equations 1 and 2 present this sorting argument more formally. In Equation 1, we depict log wages for immigrant i in occupation j at time t as a function of the level of occupational English proficiency: (1) ln wijt = β1occ-english j + β2englishi + β3x i + αi + εit Where occ english is the proportion of workers with limited English in occupation j, English i j is the worker s English language proficiency, X is a set of other observed individual level control variables, ε it is an error term, and αi represents fixed unobserved factors that affect wages. Note that we will refer to our key independent variable the proportion of workers with limited English as occupational English for the sake of brevity, even though it refers to the proportion of workers in the occupation who lack fluent English language skills. In our theoretical model, we hypothesize that α i represents traits such as ambition and skills that are not measured on typical surveys or adequately proxied by educational credentials, but observed by employers and rewarded in the labor market. In Equation 1, we expect that working in an occupation with a high proportion of workers with limited English is associated with lower wages ( β 1 < 0 ) and that, everything else being equal, workers with better English ability have higher wages ( β 2 > 0). In Equation 2, we present a simple model of occupational sorting based on English language proficiency and unobserved productivity: (2) occ-english i = η1english i + ηα 2 i + ν it 11

14 The benefit of an enclave occupation with lower English language requirements is that it provides employment for immigrants who are not fluent in English, hence we would expect a negative value for η 1. The coefficient η2 depicts the effect of unobserved factors that affect wages such as ambition on sorting into enclave occupations. If occupations with lower English requirements tend to be lower skilled occupations in general, or if more skilled (or ambitious) immigrants learn English more rapidly, then we would expect a negative relationship between occupational English and the unobserved individual-level skills that affect wages, i.e., η 2 < 0. Referring back to Equation 1, we can develop an intuition about how skill-based occupation sorting in Equation 2 will affect our coefficients in Equation 2. A negative correlation between the unobserved factor α i and occupational English (as hypothesized in Equation 2) will tend to result in a downward bias on the coefficient on occupational English in Equation 1, as immigrants with less ambition or lower unobserved skills stay longer in occupations with low English requirements. If this kind of negative sorting is taking place, then regression estimates of β 1 will overstate the negative effect of working in an enclave occupation. If we are worried about the possibility that cross sectional data may overstate the effect of occupational English on wages because of sorting, an alternative approach is to use longitudinal data to model wage growth rather than wage levels. If occupational enclaves restrict economic assimilation by delaying English language acquisition or other skills necessary for upward mobility, then this should result in a negative effect of enclave occupations on subsequent wage growth. This is depicted in Equation 3: Δ ln w ijt (3) = α + φ1occ-english j1+ φ2zi + εit Δtime 12

15 Where the dependent variable is the change in wages over time, occ english i 1 is the level of occupational English in the first wave of data, and Zi represents a set of relevant control variables. If working in a enclave occupation constrains wage growth, then we would expect that φ 1 < 0. In contrast, the sorting argument claims that although occupational enclaves are associated with lower wages in the cross section because of the sorting of workers with poor English into those occupations they do not affect the subsequent wage growth of immigrants workers, hence φ 1 = 0. In other words, the test is quite simple: do immigrants who work in occupational enclaves have lower rates of subsequent wage growth than other immigrants? Data We draw on several complimentary data sources to understand the role of occupational language use on the wage mobility of immigrant workers. First, we use data from the 2005 and 2006 American Community Survey (ACS) and the 1996, 2001, and 2004 panels of the Survey of Income and Program Participation (SIPP). The 2005 and 2006 ACS are 1% samples of the U.S. population and provide a broad overview of immigrant employment by detailed occupation. The benefit of the ACS is that it provides a large number of cases, which allows us to maximize the number of immigrant workers in enclave occupations. In contrast, the SIPP is a much smaller data set of about 60,000 households per panel. The key advantage of the SIPP is that it is a longitudinal study (each panel of data is followed for 3-4 years), which allows us to examine the role of occupational enclaves on wage growth. All three of the SIPP panels provide us with quarterly data on wages and occupation for all respondents who are currently working. The 2004 SIPP panel is the first SIPP panel to include data on language proficiency, so it is the best suited 13

16 for our analysis as it allows us to disentangle the effect of individual language ability from occupation-level effects. We supplement the individual level data from the ACS and the SIPP with aggregate data on the English ability of workers in each occupation at the state level. The variable occupational English measures the proportion of workers in an occupation who report speaking English not very well or not at all. We construct this variable using the 2005 and 2006 ACS by aggregating the individual data at the state level. The state-level variation by occupation is important because it allows us to take regional variation in occupational composition into account. There are geographic differences in the degree to which certain occupations function as occupational enclaves; for example, the proportion of carpenters who speak Spanish may be higher in Texas and California than in South Dakota. We would like to go to a more detailed level of geography, as the ACS would allow us to go the level of a Public Use Micro Area (PUMA, about the size of a county), but the smallest level of geography in the SIPP is the state. We add the variable occupational English to our individual ACS and SIPP data by merging it at the state and occupational level. Quantitative Results [Table 1 about here] American Community Survey Table 1 shows the relationship between time since immigration and self-reported English language ability. Among Latino immigrants who arrived in the last 5 years, about thirty-five percent do not speak English and another thirty-four percent do not speak well. English language proficiency increases steadily as time in the U.S. increases. By the time an immigrant 14

17 has been in the U.S. for longer than 21 years, only about seven percent claim to speak English not very well. The level of those who either speak only English or who speak it very well also climbs from fifteen percent to fifty percent. [Table 2 about here] Table 2 depicts our measure of occupational enclaves, occupational English, which, as discussed above, is the proportion of workers in the occupation, by state, who report either speaking English not well or not at all. This variable is highly correlated with the proportion of Spanish speakers and Latino immigrants in the occupation (above.9 for both) hence there is little empirical difference between either of these variables as our measure of occupational enclaves. Table 2 shows the level of occupational English by the respondent s time since immigration. While recent Latino immigrants work in occupations with, on average, 18.9% limited English speakers, this number falls to 12.7% for immigrants who have been in the U.S. for more than 20 years. Although recent Latino immigrants may work with a large percentage of limited English speakers, the cross-sectional data suggests that immigrants move out of occupational enclaves over time as their experience in U.S. labor markets increases. [Table 3 about here] Table 3 lists the top paying occupations with at least 10% limited English speakers. While Chiswick and Miller (2007) show that, in general, there is a strong negative correlation between occupational English requirements and wages, it is important to note that there are a number of occupations that pay relatively good wages despite having a high proportion of limited English speakers. Inspection of Table 3 reveals that there are several construction related occupations on this list as well as other occupations that require manual skills or involve difficult or dangerous working conditions. The skills and/or danger involved in doing each of these jobs 15

18 may keep wages up for those workers who are able to do the work. Table 3 also shows the proportion of workers in each of these occupations who report speaking Spanish at home. Some occupations on this list have very high proportions of Spanish speaking workers: 50.1% of plasterers and 35.4% of cement masons report speaking Spanish. A worker in these occupations would have a high probability of being in an environment where Spanish would be understood, thereby reducing the necessity of English language fluency. [Table 4 about here] Table 4 presents OLS models of the effect of occupational English on log wages for Latino immigrants in the ACS. In model 1, we estimate a bivariate regression between occupational English and wages and find that a 10 percentage point increase in the proportion of limited English speakers in the respondent s occupation would decrease wages by about 9% (-.921*.1). In Model 2 we add controls for time since immigration and self-reported English ability, and find that the magnitude of the coefficient on occupational English drops by about 30% to In Model 3, we add a variable for construction related occupations and interact this with the occupational English variable. The coefficient on the interaction term (.185) is positive, indicating that the negative effect of occupational English is substantially smaller in construction occupations, which makes sense given our earlier discussion of Table 3 where we found that a number of construction related occupations pay relatively well despite the presence of other coethnics who do not speak English. Thus, although the overall effect of occupational English on wages in construction occupations is still negative ( ), the positive interaction term indicates that the effect of working in an occupation with coworkers who are not fluent in English is not as pronounced in construction occupations. 16

19 Finally, Model 4 adds interaction terms with years since immigration. The results of this model are important in light of our earlier discussion about occupational sorting based on English language ability. The excluded category for years in the U.S. is 0-5, so the coefficient on occupational English in this model (-.281) is the estimated effect for this group. In contrast, the negative interaction term between occupational English and time since immigration (-0.474) indicates that the effect of working in an enclave economy is more pronounced for immigrants with more experience in the U.S. For example, the estimated effect of occupational English in Model 4 for immigrants who have been in the U.S. for more than 20 years is ( plus the interaction term, ). The results in Model 4 are consistent with a sorting model. We hypothesize that there are two types of sorting going on. First, immigrants with lower levels of English proficiency sort into occupations with lower English requirements. Jobs that don t require English proficiency tend to pay less, on average, because they either involve fewer skills or less complex tasks than jobs that involve fluent interaction and communication in English or because of the crowding of non-fluent immigrants into these jobs. In addition to sorting based on language ability, we hypothesize that a second type of sorting occurs over time. As discussed above with respect to Equations 1 and 2, immigrants with higher levels of unobserved skills and motivation may start out in occupational enclaves, but quickly move out as they gain an understanding of U.S. labor markets and the kinds of opportunities available in different occupations. In contrast, immigrants with lower levels of unobserved skills may continue to work in these occupations. As a result of the upward occupational mobility of successful immigrants, the average level of unobserved skills and ambition falls over time in enclave occupations. If we base our interpretation of these results on 17

20 an occupational sorting argument, then a comparison of the greater effect of occupational English on long-term immigrants in Model 4 suggests that close to half of the observed relationship between occupational English and wages observed in Model 2 may be spurious: i.e., the effect of occupational English for recent immigrants in Model 4 (-0.281) is substantially smaller than the overall effect reported in Model 3 (-0.460). The results for the interaction terms in Model 4 should make us cautious about interpreting the negative cross-sectional correlation between wages and occupational English as the causal effect of working in an occupational enclave. Nonetheless, the problem with the cross-sectional data from the ACS is it only tells us what occupation immigrants are currently working in, not what occupations they worked in prior to their current job. SIPP data In order to differentiate between enclave occupations as stepping stones or traps, we turn to longitudinal data from the SIPP. As discussed above, we use data from three different SIPP panels. The 2004 data is preferred because there is data on English language proficiency, which is absent from the 1996 and 2001 panels. However, at the time of writing, only 4 of the 9 waves of data are available for analysis. The 1996 SIPP panel has 12 waves of data running from , and the 2001 SIPP panel has 9 waves of data from [Table 5 about here] Table 5 provides a basic descriptive overview of wages and wage growth for Latino workers in the 2004 SIPP data. The rows of the table correspond to quartiles of the proportion of limited English speakers in the respondent s occupation, as constructed from the ACS data described in the data section above. The second column shows the average proportion 18

21 limited English, ranging from a low of.006 for the first quartile to a high of.350 for the fourth quartile. The third column shows the average wage of the workers in each quartile. The results here are consistent with what we learned with the ACS data in Tables 1-4. Workers in occupations in the first category earned an average of $17.04, versus $9.91 in the bottom quartile of occupational English proficiency. While average wages indicate the strong negative relationship between occupational English and earnings, it is not clear that the effect is causal. As discussed above, a negative correlation between wages and occupational English may reflect sorting rather than a causal effect of occupational characteristics on wages. Workers limited English proficiency may lead some of them to sort temporarily into enclave occupations before transitioning into the mainstream labor market. The fourth column of Table 5 shows wage growth between for Latino workers based on the occupational English of their job in the first wave. If working in an enclave occupation hurts wage growth, then we should see lower levels of wages for workers in the fourth quartile, where the proportion of workers speaking poor English is 35%, compared to the other categories. Table 5 shows that this is not the case; the level of wage growth is actually higher among workers in the fourth quartile compared to the top two quartiles: workers in the fourth quartile see their wages go up by.055 log points, compared to.022 for the first quartile and.034 for the second quartile. This suggests that while working in an enclave occupation is associated with lower wages in the cross section, it does not negatively affect wage growth. To provide a more formal test of the effect of enclave occupations on wage growth, we turn to growth curve models using SIPP data in Table 6. The growth curve model is estimated using the command xtmixed in Stata by treating time as a random coefficient and including 19

22 interaction terms between time and selected covariates (see Rabe-Hesketh and Skrondal 2005 for a more complete discussion of growth curve models). The benefit of the growth curve model is that it takes advantage of the longitudinal SIPP data to model both wage levels and wage growth. Hall and Farkas (2008) provide an example of using a growth curve model to study immigrant earnings trajectories. A basic depiction of the growth curve model we estimate is as follows: First, in Equation 4 we are modeling log wages of individual i at time t with a random intercept, β 0, and slope, β 1 where ε it is a standard error term. ln wijt = β0 + β1 (time it ) + εit (4) In Equation 5, we model the intercept as a function of sets of observed covariates X and Z, along μ i with a person specific random effect,. β = α + α X + α Z + u it 2 it i (5) Finally, in Equation 6, we model the effect of time on wages with a constant, a subset of our observed covariates, Z, and a person specific random effect: β1 = δ0 + δ1 (6) Z it + φi In all of the models in Table 6, the variable for occupational English measures the proportion of limited English speakers in the respondent s occupation in the first wave of data. As a result, we test for the effect of working in an enclave occupation on subsequent wage growth. The results for each model in Table 6 are presented in two panels. The Levels panel presents coefficients for wage levels (the model for the intercept terms in Equation 5), while the Slopes panel presents coefficient for the individual slope of wage growth over time (the model for wage growth depicted in Equation 6). The slope coefficients measure the effect of time and the interaction effects of selected independent variables with time 20

23 Model 1 presents results for the 2004 SIPP panel. The slope coefficient for occupational English (.0362, s.e..066) indicates the impact of this variable on wage growth. While occupational English is negatively correlated with the level of wages, the effect on wage growth is not statistically significant at the.05 level. This result is consistent with the sorting explanation of occupational enclaves; immigrant workers work in enclave occupations because they offer employment opportunities while they are learning English, and there is no negative effect on wage growth over time. Models 2-4 of Table 6 test this hypothesis with the combined 1996 and 2001 SIPP panels. In Models 3 and 4 we find that the effect of occupational English on wage growth is positive but not statistically significant at the.05 level. In Model 4, the effect of occupational English on wage levels is smaller for recent immigrants (i.e., the interaction term between recent immigrant and occupational English is.171), consistent with an interpretation of the results from the ACS data in Table 4 based on sorting. Overall, the results presented in Tables 4, 5, and 6 point to an important divergence in results. An analysis of the effect of occupational enclaves based on cross sectional data with the ACS in Table 4 suggests that working in occupations with a large number of poor-english speakers reduces wages, even after controlling for a large number of individual level variables. In contrast, the longitudinal data from the SIPP indicates that working in an occupational enclave does not constrain wage growth in Tables 5 and 6. Case Study Evidence of Latino Construction Workers To complement our quantitative findings, we now turn to a case study discussion based on a year of field work ( ) in the construction and buildings trade in the Raleigh- 21

24 Durham-Chapel Hill area of North Carolina, an industry and location in which immigrant workers are increasingly concentrated in several occupations where they work alongside coethnics. Complete results from this study are reported elsewhere (Hagan and Lowe 2008), and the goal here is to supplement our statistical results with detailed qualitative data. The study interviewed roughly 50 immigrant workers, their supervisors (encargados), and their employers at job sites, public places, and their homes. Most of the immigrant workers were undocumented and recent newcomers, having migrated from Mexico or Guatemala to North Carolina since None spoke English well. Their supervisors were also Latino but had established work and residential histories in the United States, and many were bilingual and legal residents of the United States. The ethnicity of the employers/owners was more varied: most were white, although a number were either Latino or Black. The field data examines the social mechanisms through which newcomer immigrants experience social mobility in occupational enclaves as it unfolds in the construction industry of the American South. The construction and buildings trade sector of the U.S. economy is a major source of jobs for Latino immigrants. In 2006, Latinos filled two out of every three new construction jobs, and the vast majority of these jobs were filled by newcomers from Mexico and Central America (Pew Latino Center 2007). In North Carolina, the construction industry is the largest employer of Latinos immigrants; at least 40% of the state s construction force is Latino, with estimates reaching as high as 70% in urban areas. The state s Latino construction workers are concentrated in drywall, carpentry, roofing, masonry, concrete and paving, and landscaping. In these areas newcomer immigrants are usually hired as ayudantes (helpers or construction workers) where they earn between $8 and $14, depending upon demonstration of skills, legal status, time on the job, and cooperation with coworkers. At the time of the study, entry level 22

25 workers were earning $8.00 dollars which was significantly higher than the state s $6.55 minimum wage hour. In the study sample, almost all the workers interviewed were recruited to their jobs through social networks. Some workers migrated from Mexico or Guatemala to join family and friends in North Carolina. Others left jobs in Texas and California to take advantage of the booming construction industry and higher wages in the southeast. Most interviewed had for some time enjoyed steady work in a firm or with a work team when they were interviewed. Many agreed that although construction work was grueling and exhausting, it was the best job for a newcomer who lacked English skills and formal training because of the high wages and opportunities to learn new skills. Interestingly, although most immigrant workers we interviewed argued that knowledge of English was very important for economic advancement, most experienced considerable wage growth despite not knowing English. These workers depended on co-ethnics for initial entry into the construction sector and it was after they gained work-related skills that they began to see changes in wage growth. Manuel s narrative reflects the experiences and expectations of many newcomer immigrants we interviewed in construction. Manuel arrived from Guatemala in Back home, he worked as a bank teller. In North Carolina, he found an entry-level job as an ayudante at a large commercial firm that builds schools and other public works throughout the southeast. He was recruited by his cousin, Ricardo, one of several bilingual encargados at the firm (first or second line supervisors). Manuel works in a team with six other Latinos (four Mexicans and two Guatemalans) and under the supervision of Ricardo. Some in the team are more experienced than others, so on-the-hand training is a constant feature of the work process. None of Manuel s co-workers speak more than few sentences in English. 23

26 When Manuel was first hired, his main job was cleaning up the job site, directing traffic, and fetching things for other workers and encargados. As an informal apprentice to his cousin, he learned basic carpentry, drywall installation, and masonry. When Manuel was hired in 2006, he earned $8 an hour; two years later, in 2008, he earned $12 an hour. Because most workers do not speak English, Ricardo s boss relies on him to train his team and reward skills learned with wage increases. Manuel depends on the support of his cousin to receive instructions on what needs to get done at the jobsite. Manuel believes that he can earn up to $17 an hour without English skills, knowledge of blueprints, and operation of heavy machinery. His goal is to follow Ricardo s footsteps and become a maestro (skilled craftsperson who has mastered multiple skills) to surpass his cousin s $28 an hour wage. Eventually, he hopes to start his own construction business. To achieve these goals, Manuel plans to enroll in a local community college that provides English classes and Spanish language classes in carpentry and blueprint reading. He recognizes that learning English is the most important factor to move beyond entry level jobs in the construction sector. In particular, English will allow him to communicate with his white employer and market his own skills without having to depend on an intermediary. Francisco, 35 years of age, is a recently arrived migrant from Guanajuato, Mexico. At the time of the interview, he had only been residing in the U.S. for six months. While in Mexico, he worked as an ayudante in construction before becoming a maestro. Since arriving in the U.S., he has worked as a day laborer and recently had landed a job with a Latino sub-contractor. He recognized that by working for a sub-contractor, he was not remunerated for the construction skills that he brought from Mexico. Still, he recognized that he needed to do his time and learn how to navigate the labor market before he could earn higher wages. Francisco understands that his inability to speak English prevents him from marketing his skills to mainstream construction 24

27 firms. As a result, he was forced to work for a Latino subcontractor that marketed his skills but did not pay him the appropriate wage given his extensive training in Mexican construction. He recognizes this as a temporary roadblock and uses his current job to earn wages until the ideal job comes along. More importantly, he understands that working alongside co-ethnics allows him to acquire U.S. based construction skills. Though he currently only earns $8.00 dollars, he believes that the skills he brought from Mexico will eventually allow him to double his hourly wages. Working alongside co-ethnic bosses and employees has allowed Manuel, Francisco, and other newcomer immigrants the opportunity to acquire on-the-job skills and learn about the construction industry and the U.S. labor market more generally. In the short run, both men sacrificed wages with the understanding that their current situation was only temporary. With the skills, knowledge, and information that they acquire through working with co-ethnics on construction sites, both workers hope to continue to climb the occupational ladder within the construction and building trades. The basic pattern of wage mobility described in these narratives is consistent with a sorting model of occupational enclaves: although the initial occupations of Manuel and Francisco did not pay high wages and involved work among Spanish speaking co-ethnics, they were, at least in these two cases, stepping stones to subsequent upward mobility. Discussion and Conclusion This study uses cross-sectional data from the 2005 and 2006 American Community Survey and longitudinal data from the 1996, 2001, and 2004 panels of the Survey of Income and Program Participation to analyze the effect of working in occupations with large numbers of limited English speakers on the wages and wage growth of Latino workers. In order to measure 25

28 English proficiency at the occupational level, we aggregated data from the ACS to the occupational and state level and then merged this back on to our individual level ACS and SIPP data. Our findings point to a crucial distinction between wage levels and wage growth: although the proportion of workers in the respondent s occupation with limited English was negatively associated with wages in the cross-section, it had no effect on wage growth, based on our analysis of longitudinal data from the SIPP. These results have important implications for understanding the process of economic assimilation for Latino immigrants. While recent studies on Latino occupational segregation have argued that the crowding of immigrant Latinos workers into brown collar occupations and segregated jobs reduces their wages (Catanzarite 2000; Catanzarite and Aguilera 2002; Kmec 2003), our results indicate that this literature paints an overly pessimistic picture of the effect of working with co-ethnics because it relies upon cross-sectional data and, as a result, misses upwardly mobile workers who move on to other occupations over time. Instead, we offer an alternative explanation based on the ethnic enclave hypothesis, which stresses the sheltering effect of working in an ethnically based economy for immigrant workers (e.g., Wilson and Portes 1980, Evans 2004, Bailey and Waldinger 1991). We recast brown collar occupations as occupational enclaves occupations with substantial numbers of Spanish speakers and/or workers with limited English ability and argue that in many cases these occupations provide immigrants with employment opportunities while they adjust to new labor market conditions, learn English, and acquire U.S. based human capital. In contrast to the ethnic enclave hypothesis, however, we maintain that the benefit of these occupational enclaves is temporary; for upwardly mobile immigrants, they are stepping stones to better jobs, a means to an end rather than an end themselves. For immigrant workers who don t move on, however, the 26