Black Immigration, Occupational Niches, and Earnings Disparities Between U.S.-Born and Foreign-Born Blacks in the United States
Using data from the 2000 U.S. census and the 2010 to 2014 waves of the American Community Survey, we examine the importance of occupational niches in explaining earnings disparities between U.S.-born blacks and black immigrants in the United States. Our results show that, relative to U.S.-born blacks, most black immigrant subgroups have similar or greater representation in occupational niches. Employment in a niche occupation has a small but positive association with earnings, and the returns to niche employment are greater for black immigrants, particularly black immigrant women. Niche employment does not, however, explain earnings disparities between U.S.-born and immigrant blacks.
niches, occupation, earnings, blacks, immigrants
Every census of the U.S. population since 1970 has shown that, even after adjusting for a standard set of labor market characteristics, black immigrants from the English-speaking Caribbean (West Indies) have higher labor force participation rates, higher employment rates, and higher earnings than U.S.-born blacks. A number of studies have also shown that West Indian immigrants have become the primary source of labor in sectors of the economy once dominated by U.S.-born blacks (Waldinger 1996; Waters 1999). Although researchers have examined whether selective migration, differential patterns of discrimination, or differences in cultural practices explain disparities between the two groups (Hamilton 2014; Ifatunji 2016; Model 2008; Sowell 1975, 1978, 1981, 1983; Vickerman 1998; Waters 1999), few have [End Page 60] comparatively examined variation in the types of employment typically held by black immigrants and U.S.-born blacks or analyzed whether employment in ethnic niche occupations helps explain labor market disparities between the two groups.
A common feature of immigrant economic incorporation is the proclivity of new immigrants to gravitate toward a narrow set of economic activities (Eckstein and Peri, this issue; Waldinger 1996). The organization of economic activity can take different forms, such as ethnic economies or ethnic enclaves (Adelman, Huishien, and Tolnay 2006; Fong and Shen 2011; Light and Gold 2000; Portes and Jensen 1989; Portes and Manning 2005; Portes and Shafer 2007). One of the most common and basic forms of economic clustering, however, is ethnic niches—the overrepresentation of members of a particular ethnic group in a given set of occupations or industries.
Roger Waldinger (1996) shows that public-sector employment has been a niche for U.S.-born blacks since the 1990s. Conversely, Suzanne Model and Gene Fisher (2001) show that West Indian immigrants are overrepresented in a number of private-sector industries. Prior research has suggested that immigrant occupational niches provide employment opportunities for black immigrants, particularly less-skilled immigrants upon arrival in the United States, but considerable debate remains regarding whether niche-sector employment leads to greater earnings.
Most of the studies on ethnic niches among black immigrants have focused almost exclusively on the experiences of pre-2000 black immigrants from the English-speaking Caribbean residing in New York City (Model 1997a; Waldinger 1994). Between 2000 and 2013, the number of black immigrants in the United States increased by 56 percent: Caribbean immigration increased by 33 percent during this period, while African immigration increased by a remarkable 137 percent (Anderson 2015). Therefore, understanding whether niche employment provides pathways to economic integration for both African and Caribbean immigrants could provide valuable insights into the types of economic incorporation experienced by newer waves of black immigrants, as well as shed light on how niche employment shapes disparities among blacks by nativity in the United States.
This study has three goals. First, given the absence of research documenting the degree to which different black immigrant subgroups are employed in ethnic niches, we examine variation in niche employment for the major black immigrant subgroups: immigrants from the Dominican Republic, Haiti, Jamaica, Trinidad and Tobago, Guyana, Ghana, Nigeria, and Ethiopia. Second, we evaluate the degree to which niche employment explains variation in labor market outcomes between U.S.-born blacks and immigrant blacks. Finally, we look at whether the earnings return to employment in ethnic niches varies between U.S.-born and immigrant blacks as well as among black immigrants by birth country.
Our descriptive results show that, relative to U.S.-born blacks, a similar or greater proportion of individuals from every black immigrant subgroup is employed in an ethnic niche. Immigrant groups with the largest share of individuals employed in niche occupations hail from Ghana and Ethiopia, and the subgroups with the smallest share hail from Guyana and Jamaica. Our regression results show that employment in niche occupations has a small but statistically significant positive association with weekly earnings. However, ethnic niches do not account for earnings disparities between U.S.-born and immigrant black men and women. Results also show that, relative to U.S.-born blacks, the returns to niche employment are greater for black immigrants, particularly immigrant women.
Researchers have proposed three primary explanations for labor market differences between native and immigrant blacks: cultural differences in attitudes toward work, disparate patterns of discrimination, and selective migration (Foner 1985; Hamilton 2014; Ifatunji 2016; Kasinitz 1992; Model 2008; Sowell 1975, 1978, 1981, 1983; Vickerman 1998; Waters 1999).
The Creation of an Ethic Niche
Mary Waters (1999) suggests that a combination of selective migration, network hiring, [End Page 61] and discrimination by white employers produces a labor market advantage for black West Indian immigrants in New York. She argues that because black immigrants are immigrants, they are positively selected on a number of unobserved factors that lead to more favorable outcomes in the United States. Moreover, she contends, immigrants' point of reference regarding the value of a particular type of employment is their country of origin and the opportunities available in their home countries. As a result, at the food services firm that she studied, black immigrants placed greater value on the menial jobs available to them. According to Waters, the result was that West Indian immigrants stayed longer and worked harder at jobs that many Americans, black or white, would have considered dead-end employment. In addition, the firm relied on the social networks of existing immigrant employees to fill vacancies. These factors, combined with discrimination, make it virtually impossible for U.S.-born blacks to find employment in the firm.
The process that Waters (1999) describes is the creation of an ethnic niche. Waldinger (1996) outlines this general phenomenon for other ethnic groups in New York. Because of unusual demand for labor in particular occupations, the resources specific to particular ethnic groups, and differential discriminatory practices across industries and occupations, different ethnic groups gravitate toward different occupations or industries that offer avenues for upward mobility. Waldinger also shows that niches are dynamic and follow a process of ethnic succession whereby members of one ethnic group replace another ethnic group as the dominant workforce in the industry. Two primary factors drive the evolution of niches: better employment opportunities for a particular ethnic group outside of niche sectors, and, perhaps more importantly, immigration.
One of the most visible patterns of ethnic succession is the formation of different immigrant niches in segments of the economy that were once U.S.-born black niches (Waldinger 1996; Rosenfeld and Tienda 1999). Michael Rosenfeld and Marta Tienda (1999) show that Mexican immigrants are now dominant in segments of the economy once held by U.S.-born blacks. Waldinger (1996) also shows that West Indians and Dominicans have replaced U.S.-born blacks as the dominant workforce in many segments of the economy.
Both Waldinger (1996) and Rosenfeld and Tienda (1999) also highlight that while many U.S.-born black niches have been transformed into immigrant niches, U.S.-born blacks, particularly those who are moderately and highly skilled, have become the dominant workforce in many public-sector occupations. Waldinger notes that one of the key factors in the creation of modern U.S.-born black niches is stringent hiring practices that make it more difficult to discriminate. Consequently, U.S.-born black niches tend to form in large private-sector industries and in the public sector. Firms in large industries, being highly visible, are forced to engage in more transparent hiring practices, and the public sector's bureaucratic hiring protocols, which tend to reduce discrimination, help U.S.-born blacks gain employment.
Earnings and Immigrant Occupational Niches
Prior studies show that occupational niches lead to employment opportunities for both new immigrants and U.S.-born blacks, particularly less-skilled new immigrants (Waldinger 2001). However, the evidence regarding the earnings benefits of employment in niche occupations is mixed.
Research has shown that the benefits of niche employment vary by industry and sector of the labor market (Model 1997a; Wilson 2003). For example, Jennifer Lee (2013) finds that, for Asian immigrants, the relationship between occupational niches and earnings depends on whether the niche is in a low-tech or high-tech industry. Steven J. Gold (this issue) shows that Israeli immigrants also tend to gravitate toward high-paying jobs in the technology sector. There is robust evidence that African Americans, because of lower levels of discrimination in public-sector niches, fare particularly well in these niches (Logan and Alba 1999; Rosenfeld and Tienda 1999; Waldinger [End Page 62] 1994). Using data from 1990, Waldinger (1996) also finds that African Americans have higher wages in public-sector niches than in the dominant economy. In contrast, he shows West Indian and Dominican immigrants have lower wages when employed in niche occupations than when employed in non-niche sectors.
The economic benefits of employment in niche sectors also vary by gender. When comparing how nonwhite groups fare in the dominant labor markets of New York and London, Model (1997b) finds that ethnic niches are more beneficial for West Indian female immigrants in New York than for West Indian male immigrants.
The Current Study
Most of the extant literature on black immigrant ethnic niches focuses on the employment experiences of immigrants from the Caribbean. Less is known about the impact of niche-sector employment on earnings disparities between black immigrants from the Caribbean and those from Africa.
There are several reasons to expect black immigrants from sub-Saharan Africa to have different patterns of employment than black Caribbean immigrants. The first relates to education. Among individuals ages twenty-five and older, 30 percent of the entire U.S. population has a college degree or an advanced degree, and 20 percent of U.S.-born blacks have a college degree. This figure is 20 percent for immigrants from the Caribbean and a remarkable 35 percent for immigrants from Africa (Anderson 2015). Although niches can form in high-skilled occupations, much of the existing research suggests that niche employment benefits less-skilled immigrants in occupations where it is easy to bypass bureaucratic hiring procedures and little formal training is needed to start employment (Waldinger 1996). Consequently, it remains an open question whether the patterns of labor market concentration for newer and more-educated waves of black immigrants from Africa will be similar to those for black immigrants from the Caribbean, and whether the benefits to niche occupations will vary across the groups.
Second, the gender composition of immigration varies considerably for immigrants from the Caribbean and those from sub-Saharan Africa. Women represent more than 50 percent of the stock of immigrants from the West Indies. Moreover, relative to West Indian men, West Indian women have a longer tenure of U.S. residence and are also often the family member who initiated the migration decision (Foner 2009; Model 2008), a factor that affects both the likelihood and type of niche employment (Kasinitz and Vickerman 2001). By contrast, men account for a greater proportion of flows from sub-Saharan Africa and also tend to be the primary migrant. These factors, in combination with the gendered nature of occupations in the United States, make it likely that men and women from the two regions would cluster in different types of occupations.
Finally, many of the seminal papers documenting patterns of ethnic niching among blacks were written using data from the 1990s. Given the changes in the U.S. economy since that period, during which many occupations requiring lower levels of skill have been eliminated, it is unclear whether the economic returns to niche employment have remained stable since the 1990s or how the returns to ethnic niching vary across a more diverse black population. Such variation could help explain differences in labor market outcomes between U.S.-born blacks and black immigrants.
We address these gaps in the literature by examining three questions: (1) Is there variation in the degree to which U.S.-born blacks and foreign-born men and women are employed in occupational niches? (2) Does niche-sector employment explain wage disparities between black immigrants and U.S.-born blacks? (3) Do the earnings returns to niche-sector employment vary among blacks?
DATA, MEASURES, AND METHODS
This study uses data on black men and women between the ages of twenty-five and sixty-four taken from the 5 percent Integrated Public Use [End Page 63] Microdata Series (IPUMS) of the 2000 U.S. census and the 2010 to 2014 waves of the American Community Survey (ACS) to evaluate labor market differences between native and immigrant blacks (Ruggles et al. 2015). In this study, blacks are defined as individuals who self-identify as black and who do not reside in institutions or group quarters. To avoid including individuals in the immigrant sample who are more similar to natives than to immigrants, we exclude people who were born abroad to American parents and those born in Puerto Rico. Because one of the primary goals of this study is to examine the impact of occupational niches on earnings, we exclude individuals for whom we do not have occupation data and individuals who are in school, out of the labor force, or in the military.
Blacks are separated into two categories: U.S.-born blacks and immigrant blacks. Immigrants are defined as individuals who were born outside of the United States. We divide the black immigrant population into nine primary source countries: the Dominican Republic, Haiti, Jamaica, Trinidad and Tobago, Guyana, Ghana, Nigeria, and Ethiopia.
Ethnic niches are created and maintained based on the strength of coethnic social networks (Waldinger 1996). Consequently, the influence of ethnic niches on labor market opportunities must be studied in areas with relatively large numbers of coethnics. In this study, we examine occupational niches in the fifteen metropolitan statistical areas (MSAs) that had the largest black populations in 2014 (Anderson 2015): New York–Newark–Jersey City, NY-NJ-PA; Miami–Fort Lauderdale–West Palm Beach, FL; Washington-Arlington-Alexandria, DC-VA-MD-WV; Boston-Cambridge-Newton, MANH; Atlanta–Sandy Springs–Roswell, GA; Philadelphia-Camden-Wilmington, PA-NJ-DE-MD; Orlando-Kissimmee-Sanford, FL; Los Angeles–Long Beach–Anaheim, CA; Houston–The Woodlands–Sugar Land, TX; Minneapolis–St. Paul–Bloomington, MN-WI; Chicago-Naperville-Elgin, IL-IN-WI; Dallas–Fort Worth–Arlington, TX; Baltimore-Columbia-Towson, MD; Seattle-Tacoma-Bellevue, WA; and Columbus, OH.
Measures and Methods
Weekly earnings is the outcome of interest. This variable is generated by summing a respondent's wage or salary income with any positive business or farm income, divided by the reported number of weeks worked in the previous year.
Education and work experience are standard predictors of earnings (Borjas 1986, 1987; Model 2008). To account for these factors, each equation includes years of education and predicted work experience (age-education-6). To capture the nonlinear effect of work experience on earnings, work experience squared is also included in each model. Research suggests that labor market outcomes vary by marital status (Correll, Benard, and Paik 2007; Korenman and Neumark 1991). To account for this factor, each model includes a variable indicating whether an individual is married with the spouse present. In addition, because some immigrants do not speak English or speak English poorly, each equation includes an indicator variable that equals 1 if an individual reports not speaking English or not speaking English well, and 0 otherwise (Chiswick 1991; Chiswick and Miller 1995). To account for the geographic clustering of different immigrant groups, the regression models also include an indicator for the exact metropolitan area in which a respondent resides. Finally, to capture differences in labor market conditions over time, the survey year of each observation is included in each regression.
Immigrants have less favorable labor market outcomes and are more likely to be employed in niche occupations when they first arrive in the United States because they are unfamiliar with the U.S. labor market. As they adapt to the host labor market, however, immigrants improve their labor market outcomes and are more likely to work outside of niche sectors of the economy (Borjas 1985; Waldinger, Bean and Bell-Rose 1999). We account for this factor in our regression models by including a set of variables that control for tenure of U.S. [End Page 64] residence. Labor market outcomes vary considerably by gender in the United States (Model 2008). Moreover, while some occupational niches are gender-neutral, others are extremely segmented by gender. Consequently, we analyze labor market outcomes separately for male and female immigrants.
To evaluate the importance of occupational niching for blacks, we create an odds ratio where the numerator represents the odds of working in a particular occupation for a particular group, and the denominator represents the odds of working in the same occupation for all other persons in the labor force (Rosenfeld and Tienda 1999, 100).
Using this measure, an odds ratio of 1 indicates that an immigrant group is proportionally represented in a particular occupation relative to other ethnic groups (Rosenfeld and Tienda 1999). An odds ratio greater than 1 means that a group is overrepresented, and an odds ratio of less than 1 means that a group is underrepresented in a particular occupation. While the analytical threshold for defining an occupational niche is arbitrary, consistent with prior studies, we define a group as having a niche in an occupation if the odds ratio is greater than 1.5 (Lee 2013; Wilson 2003).1 The odds ratio is calculated separately for each group, in each of the fifteen metropolitan areas, and in each survey year.
We construct the occupational niche variable using the "OCC1990" variable contained in the IPUMS samples of the 2000 census and the 2010–2014 five-year file of the American Community Survey. The OCC1990 variable is a modified version of the 1990 Census Bureau occupational classification scheme. The original 1990 occupation scheme contained 514 categories. OCC1990 combines a number of occupational categories to maximize the variable's consistency over time. The resulting OCC1990 classification scheme contains 389 categories. To avoid small cell sizes, we further collapse these 389 categories down to 79 categories. Given that occupational niches vary across time and place, we allow niches to vary across both metropolitan areas and survey years. For example, the construction of the niche variable allows black Jamaican immigrants to have a different set of niches in Washington, D.C., and in New York. It is also possible for the exact set of niches to vary in 2000 and 2010 for black Jamaican immigrants. Moreover, an occupation is not considered a niche unless at least 200 individuals are employed in it in a given metropolitan area and year (Wilson 2003). We treat the 2010 to 2014 time period as a single cross-section of data to allow for sufficient sample sizes across time periods.
The Empirical Model
The following equations show the fully specified empirical models used in the study. Models are estimated using ordinary least squares (OLS) regression.
In equation 1, Y is weekly earnings. X is a vector of standard social and demographic characteristics that include predicted experience, predicted experience squared, education, marital status, an indicator for each of the fifteen metropolitan areas, English proficiency, and citizenship status. C is a vector of dummy variables identifying immigrants' country of birth. The reference group for these variables is U.S.-born blacks. A is a vector of dummy variables indicating how long an immigrant has lived in the United States; these variables are set to 0 for U.S. native-born individuals, and the reference group is immigrants who have resided in the United States for more than fifteen years. T is a vector of dummy variables indicating the survey year of each observation. Equation 1 [End Page 65] serves as the base model for the study and establishes the magnitude of the wage gap between black immigrants and native blacks.
Equations 2 to 4 attempt to isolate the impact on earnings of employment in a particular occupation from the impact of employment in an occupation that is an occupational niche. Immigrants' social networks might help new immigrants find employment in an occupation where their group clusters. After employment is secured, niche employment might not offer any additional benefits. Under this scenario, after controlling for occupation, the niche variable might not have a significant association with earnings. Alternatively, if employment in an occupation with a relatively large number of coethnics leads to greater earnings than employment in the same occupation with a limited number of coethnics, then, after controlling for occupation, niching would have a statistically significant positive association with earnings.
To examine these possibilities, equation 2 augments equation 1 by including O, a vector of dummy variables that identify the occupation of each respondent. Consequently, wage differences among blacks in equation 2 are based on wage variation between immigrants and native blacks employed in the same occupation. Equation 3 augments equation 1 by including an indicator for whether an individual is employed in a niche occupation. Finally, equation 4, the full model, includes both O and an indicator for whether an individual is employed in a niche.2 Equation 4 examines whether ethnic niche employment is statistically associated with earnings after adjusting for differences in occupation. This niche variable is estimated using variation among individuals with the same occupation.
Given that the impact of niching on earnings varies across origin countries, we also estimate regression models based on equation 4 that include a set of interactions between country of birth and the niche indicator.
Table 1 shows descriptive statistics for native and immigrant black men and women, with the immigrant sample stratified by birth country. Tables 2 and 3 show descriptive statistics for individuals employed in niche occupations and for those employed outside of niche occupations, respectively.
Across the subgroups identified, table 1 shows considerable variation in niche employment. U.S.-born blacks and black immigrants from Guyana had the smallest proportion of individuals employed in occupational niches (36 percent). Among Caribbean immigrants, those from the Dominican Republic (55 percent) and Haiti (55 percent) had the largest percentages of individuals employed in niche occupations, and those from Guyana (36 percent) had the smallest share. Among the three countries from sub-Saharan Africa, Nigeria (52 percent) had the smallest share of individuals employed in niche sectors, ten percentage points less than Ghana and thirteen percentage points less than Ethiopia.
In addition, table 1 shows considerable differences in weekly earnings across the subgroups. For example, black immigrants from Nigeria ($1,062) had higher weekly earnings than U.S.-born blacks ($859) as well as all the other black immigrant subgroups. In contrast, immigrants from the Dominican Republic ($652) and Haiti ($685) earned the least.
The demographic and socioeconomic characteristics of blacks in the sample also vary substantially. Looking at educational attainment, one of the strongest predictors of earnings, table 1 shows that, on average, U.S.-born blacks had 13.27 years of education. Among all blacks, black immigrants from the Dominican Republic had the lowest (11.97) mean years of education, and immigrants from Nigeria had the highest (15.11) mean years of education.
Table 1 also indicates that immigration patterns differ significantly between immigrants from the Caribbean and those born in Africa. [End Page 66]
[End Page 67] A larger share of immigrants from African countries were more recent arrivals. For example, column 4 shows that 8 percent of Jamaican immigrants came to the United States within the last five years. In contrast, column 8 shows that 17 percent of Nigerian immigrants arrived during this period. Similarly, 48 percent of immigrants from Jamaica had resided in the United States for more than twenty years, compared to only 30 percent of Nigerian immigrants.
Tables 2 and 3 show descriptive statistics for individuals who were and were not employed in niche occupations, respectively. Together, these tables highlight two interesting data patterns. First, with the exception of Nigerians, across the subgroups, individuals employed in occupational niches earned considerably less than those employed outside of occupational niches. Second, individuals employed in niche occupations had lower mean years of education than those employed outside of niches, suggesting that niching is more common among less-educated blacks.
This section presents OLS regression models of log weekly earnings. Column 1 of table 4 shows estimates from our base model (equation 1) for the entire male sample. It establishes our baseline estimates before adjusting for occupational niching or occupational composition. The results from this model show that, with the exception of black immigrants from Jamaica, Trinidad and Tobago, and Guyana, the adjusted earnings of all other black immigrant subgroups were lower than or similar to those of U.S.-born blacks, the reference group. Relative to U.S.-born black men, men from Ethiopia had the largest earnings deficits. Column 1 of table 4 also shows how male immigrants' earnings changed the longer they live in the United States. Immigrants who have resided in the United States for more than fifteen years are the reference group for the years of U.S. residence variables. The coefficients on years of U.S. residence variables are universally negative, and less negative for the categories showing longer tenures of U.S. residence. These results suggest that immigrants' earnings increase the longer they live in the United States, a finding consistent with prior studies (Hamilton 2014).
Column 2 of table 4 presents estimates from equation 2, which includes a set of dummy variables for a person's current occupation. Among male immigrants from the Caribbean, controlling for current occupation has no impact on earnings disparities between U.S.-born blacks and black immigrant men from Haiti. Relative to the results in column 1, however, accounting for occupation differences reduces the estimate on the Jamaica coefficient by 42 percent, from 0.07 to 0.04. Moreover, while the Trinidad and Tobago and Guyana coefficients are both statistically and substantially significant in column 1, they lose their significance after accounting for respondents' occupations. This change suggests that differences in earnings between U.S.-born blacks and black immigrants from Trinidad and Tobago and Guyana are driven largely by differences in the distribution of occupations held by members of each group.
Among immigrants from Africa, columns 1 and 2 show that controlling for occupation has no substantive impact on the Ghana or Ethiopia coefficient. By contrast, adjusting for occupation changes the significance and increases the adjusted earnings deficit for immigrants from Nigeria, changing the coefficient from –0.04 to –0.08. This result suggests that, relative to similarly skilled U.S.-born black men, black men from Nigeria tend to hold jobs in better paying occupations.
Column 3 of table 4 presents estimates based on equation 3 for men, which includes a variable that captures whether the respondent is employed in an occupational niche. Importantly, this model does not include controls for occupation. Consequently, comparing the results from columns 1 and 2 to those in column 3 allows us to examine the degree to which adjusting for occupational niches differs from adjusting for occupation. In other words, we can see whether controlling for niche employment, characterized by the overrepresentation of own-group members in a particular occupation, and simply accounting for occupation has a different impact on earnings disparities among blacks.
Consistent with the descriptive results, column 3 shows that men employed in niche occupations [End Page 68]
[End Page 69]
[End Page 70]
[End Page 71] earned less than men who are not employed in non-niche occupations (–0.12). Relative to the results in column 1, our baseline model, the inclusion of the occupational niche variable into the regression model has either no effect or a very modest impact on most of the birth country coefficients. Two notable exceptions are the results for men from Nigeria and Ethiopia. Relative to the results from column 1, the coefficients on Nigeria and Ethiopia increase in magnitude (become less negative). This finding suggests that men from these two countries tend to find employment in low-earnings niche occupations.
Because column 3 does not include occupation, the niche variable is based on variation across occupations. Consequently, the niche variable captures both the impact of employment in a particular occupation and the impact of employment in an occupation in which own-group members are overrepresented (a niche). To determine whether the results in column 3 are driven by variation in the distribution of occupations held by respondents, column 4 of table 4 uses a regression model that takes into account both occupation and niching. After controlling for respondent occupation, the occupational niche variable remains statistically significant, but the sign on the coefficient changes from negative to positive. This result suggests that among individuals with the same occupation, those for whom that occupation is a niche earn more.
Comparing the results in column 2 to those in column 4 provides insights into whether niche-sector employment is a primary driver of earnings disparities among blacks. Note that the coefficients on the birth country variables in columns 2 and 4 of table 4 are nearly identical. Thus, after accounting for occupation, niche employment does not explain earnings disparities between U.S.-born blacks and most subgroups of black immigrants. Together, these results suggest that niching is more likely to occur in low-wage occupations. When U.S.-born blacks and black immigrants with the same occupation are compared, however, niche employment, on average, is modestly associated with greater earnings.
To determine whether the association between niche employment and earnings varies across subgroups, in column 5 we interact the niche indicator with each birth country indicator. In these models, the main effect for occupational niching represents the relationship between niching and wages for U.S.-born blacks, while the coefficients on the interaction terms represent the additional impact of niching on wages for a particular foreign-born group. The coefficient on the occupational niche variable in column 5 is positive and statistically significant, suggesting that U.S.-born black men employed in niche sectors earn approximately 2 percent more than comparably skilled U.S.-born black men employed outside of niche sectors. With the exception of the interaction term for Ethiopian immigrants, all the interaction terms are statistically insignificant, suggesting that there are no additional returns to niche employment for most black immigrant men relative to U.S.-born men. Ethiopian men employed in niche occupations earn approximately 16 percent less than U.S.-born black men employed in niche occupations.
Table 5 shows the results of these models for the sample of women. Column 1 of table 5 shows that women from Jamaica, Trinidad and Tobago, Guyana, and Ghana had significantly greater earnings than U.S.-born black women. In contrast, relative to U.S.-born black women, black immigrant women from the Dominican Republic and Ethiopia had lower adjusted earnings and women from Haiti and Nigeria had similar adjusted earnings. Column 2 shows how earnings disparities among black women change after adjusting for variation in current occupation. The inclusion of occupational controls has no substantive impact on the Jamaica, Trinidad and Tobago, and Guyana coefficients, suggesting that the differences in the distribution of occupations do not explain disparities between these groups and U.S.-born black women. In contrast, the coefficient on the Dominican Republic, Ethiopia, Haiti, and Ghana variables increases in magnitude after accounting for occupation. This result suggests that relative to U.S.-born black women, these immigrant women tend to gravitate toward lower-paying occupations.
Similar to the results for men, column 3 shows that black women employed in niche [End Page 72]
[End Page 73] occupations earned less (–0.10) than women employed outside of niche occupations. Column 3 also indicates that accounting for niche employment either has no impact or modestly increases the magnitude of the immigrant birth country variables relative to column 1.
The model used in column 4 controls for both occupation and niche employment for women. Similar to the results for men, after including occupation controls, the occupational niche variable remains statistically significant, but the sign on the coefficient changes from negative to positive, suggesting that women employed in niche occupations earn slightly more (0.02) than similarly skilled women employed outside of niche occupations. Again similar to the results for men, the fact that results in column 4 of table 5 are virtually identical to those found in column 2 implies that differences in occupations between immigrant and U.S.-born black women account for more of the variation in weekly earnings between the groups than differences in niching patterns.
Finally, column 5 looks for differences in the relationship between niching and wages across groups. In contrast to the results for U.S.-born men, niching is not significantly correlated with wages for U.S.-born black women. Also in contrast to the results for men, niching is significantly and positively correlated with greater wages for five of the eight immigrant women subgroups. This finding suggests that the niche result in column 4 is primarily driven by immigrant women.
Four key findings emerge from this study. First, consistent with prior research, after adjusting for a standard set of labor market characteristics, only black immigrant men from the English-speaking Caribbean (Jamaica, Trinidad and Tobago, and Guyana) had greater earnings than U.S.-born black men. Black immigrants from the other countries studied had similar or lower adjusted earnings. In contrast, every subgroup of black immigrant women except Dominicans and Ethiopians had earnings similar to or greater than those of U.S.-born black women. Second, after controlling for occupational choice, employment in an occupational niche has a small, positive association with earnings (approximately 2 percent for men and 3 percent for women). Models interacting niche employment with country of birth show that the returns to niche employment are similar for U.S.-born black men and most country subgroups of black immigrant men. In contrast, relative to U.S.-born black women, most subgroups of black immigrant women have greater earnings returns to niche employment. Finally, although we document variation in both niche employment and returns to niche employment for earnings, the results suggest that niching does not account for earnings disparities among blacks in the United States.
These findings raise two important questions about the role of niche-sector employment in understanding earnings disparities among blacks in the United States.
1. Why does niche-sector employment seem to play such a limited role in labor market disparities among blacks in the United States? Waldinger (1996) shows that the clustering of coethnics in particular occupations or industries is largely a network-based phenomenon. The social networks found within particular ethnic communities facilitate the flow of information about job opportunities to other coethnics seeking employment. Although our results suggest that this phenomenon may lead to the clustering of different immigrant groups in particular occupations, they also suggest, after accounting for choice of occupation, that the overrepresentation of coethnics within an occupation does not account for earnings disparities between blacks in the United States.
These results are largely consistent with prior research. Model (1997a) compares the economic attainment of ethnic group members within their ethnic economy industries with that of "outsiders" in the same industries. She finds few differences in economic attainment. Together, these findings suggest that occupation choice itself is more consequential than employment in an occupation with a large number of coethnics for understanding earnings disparities among blacks by nativity.
2. What explains variation in the returns to niche employment among blacks? Among men, returns to niche employment for earnings are statistically similar for both U.S.-born and [End Page 74] most black immigrant subgroups. This finding suggests that black men in general tend to form niches in sectors where the presence of coethnics does not facilitate upward mobility. Similar to U.S.-born black men, black immigrant men tend to niche in occupations such as motor vehicle operators, cleaning services workers, and security guards. The presence of niches in these occupations may provide access to jobs, but because they tend to be low-paying jobs that offer few opportunities for advancement, the role that coethnics can play in aiding upward mobility is limited.
In contrast to the results for men, subgroups of black immigrant women from both the Caribbean and sub-Saharan Africa had greater returns to niche employment than U.S.-born black women. This suggests that immigrant women tend to form niches in occupations where the presence of coethnics is beneficial. Like immigrant men, black immigrant women tend to form niches in low-paying occupations with few opportunities for advancement (such as private household work), but a number of niche occupations for immigrant women offer considerable pathways for upward mobility. For example, health care occupations, such as nurses and nurse's aides, are two of the largest niches for women from both sub-Saharan Africa and the Caribbean (authors' tabulations). Both of these occupations offer coethnics pathways to aid each other to upward mobility. Additionally, many of the primary niches that employ U.S.-born black women are in the public sector (Kasinitz and Vickerman 2001; Waldinger 1996). Although these occupations offer opportunities for upward mobility, they often require advanced education, and their more bureaucratic workplaces, with stringent hiring and promotion protocols, may mitigate the influence of coethnic ties (Kasinitz and Vickerman 2001).
Our results also show that the returns to niche employment vary among black immigrant women, among whom Haitian women have the largest returns to niche employment followed by women from Ghana and Nigeria, then by women from Jamaica and Trinidad and Tobago. This variation is likely to be driven by differences in education and U.S. tenure across the groups, which could lead to the formation of different types of niches as well as region-specific demand for female labor.
During the 1990s and 2000s, waves of black immigrants drastically changed the composition of the black population in the United States. Although flows of black immigrants from the Caribbean to the United States date back to the early 1900s (Model 2008), fewer than 60,000 black individuals migrated to the United States from Africa prior to 1990. In contrast, 323,000 black Africans migrated during the 1990s, and another 353,000 between 2000 and 2005 (Kent 2007).
Immigrant occupational niches, an important aspect of the assimilation process for many immigrants, are often associated with favorable labor market outcomes (Lee 2013; Waldinger 2001). Despite the growing demographic diversity of the black immigrant population, few studies have examined the degree to which occupational niche employment explains variation in earnings between U.S.-born blacks and black immigrants or among black immigrants. Our findings show that while occupational niche employment does not explain variation in earnings between the two groups, niche employment remains an important aspect of the economic incorporation of black immigrants, particularly immigrant women.
Tod G. Hamilton is assistant professor of sociology and faculty associate of the Office of Population Research at Princeton University.
Janeria A. Easley is a postdoctoral fellow in the Department of Sociology at the University of Pennsylvania.
Angela R. Dixon is a graduate student in the Department of Sociology and the Office of Population Research at Princeton University.
1. We also estimate models in which occupational niches are defined by an odds ratio of 2. This change does not affect any of the substantive findings reported in this article.
2. In supplemental analysis, tables 4 and 5 were reestimated using data for each survey year. The substantive results from these analyses are the same as those presented here. For simplicity, we use the same arguments to represent the coefficients in equations 1 to 4. However, we are not assuming that the coefficient values or error terms are constant across the empirical models.