## Abstract

Research suggests that immigrants from the English-speaking Caribbean surpass the earnings of U.S.-born blacks approximately one decade after arriving in the United States. Using data from the 1980–2000 U.S. censuses and the 2005–2007 American Community Surveys on U.S.-born black and non-Hispanic white men as well as black immigrant men from all the major sending regions of the world, I evaluate whether selective migration and language heritage of immigrants’ birth countries account for the documented earnings crossover. I validate the earnings pattern of black immigrants documented in previous studies, but I also find that the earnings of most arrival cohorts of immigrants from the English-speaking Caribbean, after residing in the United States for more than 20 years, are projected to converge with or slightly overtake those of U.S.-born black internal migrants. The findings also show three arrival cohorts of black immigrants from English-speaking African countries are projected to surpass the earnings of U.S.-born black internal migrants. No arrival cohort of black immigrants is projected to surpass the earnings of U.S.-born non-Hispanic whites. Birth-region analysis shows that black immigrants from English-speaking countries experience more rapid earnings growth than immigrants from non-English-speaking countries. The arrival-cohort and birth-region variation in earnings documented in this study suggest that selective migration and language heritage of black immigrants’ birth countries are important determinants of their initial earnings and earnings trajectories in the United States.

## Introduction

Between 1980 and 2005, the foreign-born black population in the United States more than doubled, increasing from 816,000 to more than 2.8 million (Kent 2007). Although immigration from the Caribbean accounts for most of this growth, the share of immigrants from Africa also increased substantially during this period. Specifically, African immigration accounted for 20 % of black immigration during the 1980s, 36 % during the 1990s, and 53 % between 2000 and 2005 (Kent 2007). Despite the changing composition of the black immigrant population, few studies have evaluated whether recent waves (post-1990) of black immigrants have earnings trajectories similar to those of earlier black migrants from the English-speaking Caribbean.

Data from the 1970–2000 U.S. censuses show that black immigrants from the English-speaking Caribbean earn more than U.S.-born blacks (“native blacks”). Because of skin-shade similarities between these two groups, this earnings disparity and other labor market differences led researchers such as Thomas Sowell (1978) to question whether patterns of stratification in the United States stemmed from cultural deficiencies in attitudes toward work among native blacks rather than from racism and discrimination (Glazer and Moynihan 1979; Sowell 1978). Although subsequent research has shown that social and demographic characteristics partly explain earnings disparities between the two groups, scholars have documented that immigrants from the English-speaking Caribbean earn more than native blacks as their U.S. tenure increases (Kalmijn 1996; Model 1995, 2008). Few studies, however, have examined whether a similar earnings crossover exists among other subgroups of black immigrants—particularly among African immigrants who share a similar linguistic heritage—or whether any subgroup of black immigrants is able to surpass the earnings of native blacks who are also selected on migration (i.e., U.S.-born black internal migrants) (Butcher 1994; Tolnay 2003). Such an analysis could help disentangle the relative importance of the primary mechanisms—culture, disparate patterns of discrimination, and selective migration—argued to produce labor market differences between native and immigrant blacks (Mason 2010; Model 2008; Sowell 1978; Waters 1999).

The convention in this literature is to compare the labor market outcomes of black immigrants with representative samples of native blacks (Borch and Corra 2010; Dodoo 1999; Dodoo and Takyi 2002). Black immigrants, however, do not comprise random samples of residents of their birth countries, and using representative samples of native blacks as the referent could generate biased estimates of the effect of culture and discrimination in explaining earnings differences between the two groups. To address this problem, Butcher (1994) suggested that U.S.-born blacks who are selected on migration—domestic interstate migrants (“native black movers”)—are a more appropriate comparison group to evaluate nativity earnings differences among blacks. Using data from the 1980 U.S. census, she showed that the two groups have remarkably similar earnings profiles, a finding that Model (2008) reproduced using more recent data on immigrants from the English-speaking Caribbean.

These two studies, however, did not examine whether any subgroups of black immigrants are able to surpass the earnings of native black movers as their tenure of U.S. residence increases. Instead, they evaluated differences between the two groups at single points in time. Prior literature has suggested that immigrants’ initial earnings are considerably lower than the earnings of natives as they adjust to the nuances of the U.S. labor market (Borjas 1995; Chiswick 1978; Kalmijn 1996). Consequently, determining black immigrants’ earnings trajectories is key to understanding the theoretical implications of earnings differences among blacks based on nativity. For example, if native black movers and black immigrants share common unobserved characteristics correlated with selective migration, then the earnings of black immigrants should converge with the earnings of native black movers rather than the earnings of native blacks collectively. Such a finding could potentially reduce the unexplained earnings gap that researchers have suggested stems from cultural differences or from disparate patterns of discrimination.

This study evaluates whether arrival cohorts of black immigrants from the English-speaking Caribbean, Latin America, and Haiti—as well as English- and non-English-speaking African countries—converge with the earnings of four subgroups of U.S.-born men: native blacks (collectively), native black movers, native black nonmovers, and U.S.-born non-Hispanic whites (native whites). Results show that upon arrival in the United States, all cohorts of black immigrants have lower weekly earnings than both native blacks (collectively) and native black movers. Although the rate of earnings growth varies by birthplace, several arrival cohorts of black immigrants from English-speaking countries in Africa and the Caribbean are projected to overtake the earnings of native blacks (collectively) as their tenure of U.S. residence increases. Fewer of these arrival cohorts, however, are projected to converge with or surpass the earnings of native black movers, and no arrival cohort is projected to achieve earnings parity with native whites. Most arrival cohorts of black immigrants from Latin America, Haiti, and non-English-speaking African countries are not projected to surpass the earnings of any subgroup of native blacks. These findings suggest that selective migration and the language heritage of black immigrants’ birth countries are important determinants of their initial earnings and earnings trajectories in the United States.

In the next sections, I discuss the popular theories used to explain variation in earnings among blacks and then describe the data and methodology used in this study. I then present results and discuss implications of this study’s findings for understanding disparate patterns of earnings among blacks in the United States.

## Background

### Theoretical Considerations

The extant literature offers several explanations for labor market differences between native and immigrant blacks. This research, primarily based on the experiences of immigrants from the English-speaking Caribbean, posits that culture, favoritism toward black immigrants by white Americans, or processes associated with selective migration explain labor market differences between the two groups (Butcher 1994; Kalmijn 1996; Model 1995, 2008; Sowell 1978; Waters 1999).

#### Culture

Researchers have argued that cultural differences in work orientation between native and immigrant blacks result from two mechanisms. First, Sowell (1975) argued that the disparate experiences of blacks during slavery and post-slavery in the English-speaking Caribbean and the United States produced contemporary differences in work attitudes, initiative, skills, and rational thinking that favor immigrants from the English-speaking Caribbean. He posited that these differences were primarily generated by unequal opportunities to acquire skills during and after slavery.

The second cultural explanation attributes differences to socialization in a majority black context (the Caribbean) versus a minority black context (the United States). For example, researchers have argued that black immigrants from the English-speaking Caribbean are more confident, more ambitious, and less willing to accept marginalized roles in society—all attributes theorized to result from their socialization in a context where there are greater numbers of black role models. Such a setting, scholars have claimed, creates a norm of success among black immigrants from the English-speaking Caribbean (Vickerman 1999; Waters 1999).

#### Favoritism and British Heritage

Researchers also suggest that because Americans are enamored by characteristics associated with Great Britain (e.g., British accents and mannerisms), black immigrants from the English-speaking Caribbean occupy a more favorable position in the discrimination hierarchy of the United States, leading to better treatment in the U.S. labor market (Arnold 1984). Similarly, Grosfoguel (2003) suggested that the social hierarchy in the United States is partly defined by colonial heritage and race, a framework that places black immigrants from the English-speaking Caribbean (who migrate from countries in which the United States was not a colonial power) above native blacks (defined as America’s colonial subjects) in America’s social hierarchy. This framework also implies that these immigrants could receive preferential treatment in the U.S. labor market. In contrast, Waters (1999) claimed that relative to native blacks in the United States, black immigrants from the English-speaking Caribbean have low expectations of racism and a strong sense that they can overcome racism, which contributes to more favorable interpersonal interactions with whites and over time could generate preferential treatment by white employers.

Although these two frameworks are compelling, it might be knowledge of English itself rather than white Americans’ admiration of British heritage that produces better labor market outcomes for blacks from the English-speaking Caribbean. Many immigrants from countries with a British influence have a readily transferable skill upon arrival in the United States: English language proficiency (Kalmijn 1996; Moore and Amey 2002). For example, Chiswick (1991) showed that the earnings of immigrants who speak and read English well grow substantially faster than the earnings of immigrants with no knowledge of English.

British heritage may also impact immigrants’ earnings in other ways. Most English-speaking countries in Africa and the Caribbean are former British colonies. Consequently, relative to non-English-speaking countries, the degree of social distance, institutions (e.g., education systems), and customs within English-speaking countries are likely more similar to those in the United States. These factors could reduce the labor market adjustment process needed for immigrants from English-speaking countries to succeed in the U.S. labor market.

#### Selective Migration and Immigration Policy

In contrast to the cultural theories or white favoritism, selectivity theorists argue that some subgroups of immigrants might outperform their U.S.-born counterparts because migration is correlated with hard-to-measure characteristics, such as hard work and ambition, which increase immigrants’ productivity and result in greater earnings (Chiswick 1978). Immigrants, however, may not immediately capitalize on these traits. Chiswick (1978:899) noted that recent immigrants arrive in the United States with limited knowledge of the customs, language, and firm-specific training needed to succeed in the U.S. labor market—deficits leading recent immigrants to have lower initial earnings than native-born individuals with similar human capital characteristics. Over time, immigrants improve their earnings by acquiring U.S.-specific skills. How much their earnings grow over time depends on the degree to which immigrants are selected on unobserved traits and the transferability of their pre-migration skills to the U.S. labor market (Chiswick 1978; Duleep and Regets 1999). For example, Chiswick (1978:900–901) noted,

Economic theory suggests that migration in response to economic incentives is generally more profitable for the more able and more highly motivated. This self-selection in migration implies that for the same schooling, age, and other demographic characteristics immigrants to the United States have more innate ability or motivation relevant to the labor market than native-born persons. If so, holding measured variables constant, as earnings rise with time in the United States, the earnings of immigrants may, but would not necessarily, exceed that of native-born persons. The earnings crossover is less likely to occur if the migration is less selected in favor of the more able or more highly motivated. The self-selection may be weaker, for example, if the migration is induced by political pressure in the country of origin, if it is mass migration of an entire community, or if it is induced by the availability of more generous welfare benefits in the place of destination than if it is the more conventional economic migration of workers for higher real earnings. The number of years since migration at which this earnings crossover occurs, if it does occur, is a parameter of considerable interest.

These predictions suggest that economically motivated migrants may have steeper earnings trajectories than political refugees or immigrants motivated by family reunification. Given this, people from the English-speaking Caribbean who emigrate from countries sending few political refugees to the United States may have greater earnings growth than immigrants from Haiti or Africa, the two regions from which most black refugees emigrate.

U.S. immigration policy also impacts the degree of selectivity among arrival cohorts of black immigrants from the same region. For example, both the Diversity Visa Lottery Program and the increase in the number of immigrants admitted on the basis of job skills introduced in the Immigration Act of 1990 provided easier paths for highly educated African immigrants to come to the United States (Lobo 2001; Thomas 2011). Indeed, more than 27 % of diversity visas awarded between 1998 and 2006 went to immigrants from sub-Saharan Africa (Kent 2007). Prior to the introduction of the Diversity Visa Lottery Program, political refugees accounted for a greater proportion of migration streams from Africa, suggesting that the earnings profiles of post-1990 African immigrants may differ from those of earlier migrants from Africa.

### Previous Empirical Studies

Model (2008) argued that the primary distinction between the selectivity argument and the cultural or disparate discrimination arguments is the scope of applicability. That is, the culture and disparate discrimination theories imply that the factors producing favorable labor market outcomes for immigrants from the English-speaking Caribbean (e.g., socialization in a majority black context or British heritage) extend to the entire black population of the region. In contrast, selectivity theorists have argued the unobserved characteristics propelling black immigrants’ labor market outcomes apply exclusively and differentially to individuals who choose to move. Using this argument, Model (2008) showed that neither variation in slave histories within the Caribbean and the United States nor variation in the racial composition of black immigrants’ birth countries consistently explains differences in labor market outcomes among black immigrants. Model also found that relative to earlier arrival cohorts, more recent cohorts from the English-speaking Caribbean have less-favorable initial labor market outcomes, leading her to conclude that the declining labor market advantage of more-recently arrived black immigrants from the English-speaking Caribbean is uniquely consonant with selectivity. That is, “neither culture nor white favoritism anticipates a diminution of West Indian [English-speaking Caribbean] advantage by cohort of arrival” (Model 2008:81).

Butcher (1994) argued that native blacks who are also selected on migration are a better comparison group for evaluating the labor market outcomes of black immigrants. Using data on men from the 1980 census, she compared the earnings of black immigrants, native black movers, and native black nonmovers. Butcher found that the earnings of black immigrants are more similar to the earnings of native black movers than to those of native black nonmovers. She concluded that selective migration—rather than culture—better explains nativity earnings differences among blacks. Model (2008) similarly showed that recently arrived immigrants from the English-speaking Caribbean and native black movers have statistically similar labor market outcomes.

### Current Study

Although previous studies provide valuable insights toward explaining the mechanism producing nativity earnings differences among blacks, several gaps exist within this literature. Research on the earnings profiles of immigrants, in general, has shown that upon arrival in the United States, immigrants have significantly lower earnings than natives; however, their earnings grow with time and may converge with or surpass the earnings of natives (Chiswick 1978). Despite this work, most studies evaluating nativity earnings differences among blacks have either not accounted for the dynamic nature of immigrants’ earnings or have not used the most appropriate reference group: native black movers. Model (2008) attempted to address these issues by comparing the earnings of native black movers and black immigrants who had both moved in the previous five years. However, this analysis implicitly assumed that the adjustment costs of migration are similar for both internal and international migrants. Because immigrants from other countries most likely require more time to adjust to their new labor market than native migrants, evaluating differences between native and immigrant blacks five years after migration biases the result toward native blacks.

In addition, few comparative studies have documented the earnings trajectories of arrival cohorts of black immigrants in the United States from all the major sending regions. If socialization in a majority black country or having a slave history distinct from that of the United States confers an advantage to immigrants from the English-speaking Caribbean, these features of the birth country might also produce a labor market advantage for black immigrants from Haiti, Latin America, and Africa. Similarly, proponents of the white favoritism framework suggest that immigrants’ British heritage is the mechanism producing disparate treatment by white Americans. If this is the case, similar to blacks from the English-speaking Caribbean, immigrants from English-speaking African countries might also overtake the earning of native blacks.

This study builds on the existing literature by answering three questions: (1) Do the earnings of arrival cohorts of black immigrants converge toward the earnings of native black movers or native black nonmovers? (2) Do the earnings trajectories of black immigrants vary by arrival cohort? (3) Does linguistic heritage impact the earning trajectories of black immigrants?

## Data, Measures, and Method

### Data

This article combines data on males aged 25 to 64 from the 5 % Integrated Public Use Micro Series (IPUMS) samples of the 1980, 1990, and 2000 U.S. censuses with data from the IPUMS samples of the 2005–2007 American Community Surveys (ACS) to analyze earnings differences between black immigrant men and U.S.-born black and white men in the United States (Ruggles et al. 2004). Black immigrants are defined as individuals who self-identify as black and who were born outside the United States. The black immigrant population is separated into five origin subgroups: Haiti, the English-speaking Caribbean, Latin America, African countries where English is an official language (immigrants from English-speaking African countries), and African countries where English is not an official language (immigrants from non-English-speaking Africa countries).1

It is important to note the English-speaking/non-English-speaking distinction among African countries contains measurement error because the ability to speak English varies considerably among African countries where English is an official language and is primarily spoken by the well-educated in some countries. This distinction does, however, capture the degree of linguistic distance between countries in Africa and the United States as well as the institutional legacy of British colonial heritage. The African sample is restricted to immigrants from sub-Saharan Africa. To ensure that individuals who arrived as children do not impact the results, the immigrant sample is restricted to men who arrived in the United States after age 18.

For computational efficiency, the U.S.-born non-Hispanic white male sample (native whites) is generated by taking a 1 % sample of each 5 % U.S. census wave and a 2 % sample of each ACS wave. The U.S.-born black population is separated into three subgroups: native blacks (collectively), native blacks who reside in their state of birth (native black nonmovers), and native blacks who reside in a state other than their state of birth (native black movers).

Individuals who reside in institutions or group quarters, were born abroad to American parents, reported having a disability restricting work, or were born in U.S.-outlying areas (including Puerto Rico) are excluded from the sample. The sample is also restricted to men who reported working at least one week in the previous year and who have positive total earnings. After these restrictions, the final analytic sample comprises 649,600 U.S.-born black men, 45,913 foreign-born black men, and 84,806 U.S.-born non-Hispanic white men. Because the factors driving male migration differ significantly from those driving female migration (Model 2008), this article focuses exclusively on the earnings of males.

Although U.S. census data provide the largest samples available to study black immigrants, these data have several limitations. First, this article’s empirical analysis implicitly assumes no systematic return migration of immigrants back to their countries of origin. Thus, the comparisons presented in this study represent differences between Americans and black immigrants who chose to stay in the United States at the time of the survey. Previous research has suggested that return migrants may have poorer labor market outcomes than immigrants who remain in the United States (Borjas 1994). This assumption could introduce some degree of selection bias into the estimates. Second, many immigrants make frequent trips between the United States and their country of birth before they settle in the United States. As a result, some degree of measurement error likely exists in immigrants’ reports of their year of immigration. This issue could bias estimates of earnings growth among immigrants. Third, relative to native blacks, black immigrants are less likely to be incarcerated and are more likely to be employed. The study is limited to the noninstitutionalized population and to individuals with positive earnings, which also introduces a degree of selection bias into comparisons. Fourth, the sample is also restricted to individuals who self-report their race as black. This decision may also introduce some degree of measurement error if some subgroups of darker-skinned migrants do not self-report as black.

Last, to provide estimates for the most recent black immigrant arrival cohorts, data from the U.S. census are combined with data from the ACS. Although the two data sets are comparable, several differences exist between them. First, the ACS is collected throughout the year, whereas the U.S. census is collected at a specific point in time during the year. This might lead to systematic undercounts or overcounts of different populations. Additionally, in some years, the ACS does not identify respondents’ metropolitan area of residence or provide a continuous measure of the number of weeks worked during the previous year. To address these specific issues, the study uses ACS data from only 2005–2007, survey waves in which both variables were collected.

### Measures and Method

#### Dependent Variable

The outcome measure used in this study is the logarithm of weekly earnings. Summing the total wage/salary income with any positive business or farm income earned by respondents generates their total earnings. Total earnings are then divided by the total number of weeks worked by individuals. This measure of earnings allows immigrants to reach wage parity with natives by either increasing the number of hours worked or by achieving a higher wage.2 A logarithmic transformation of weekly earnings is used to account for the nonlinear nature of wages.

#### Labor Market Determinants

Factors such as education, experience, marital status, English proficiency, region of residence, and whether respondents reside in a metropolitan area partly determine the labor market outcomes of U.S.-born workers (Borjas 1986, 1987; Chiswick 1991; Chiswick and Miller 1995; Correll et al. 2007; Korenman and Neumark 1991; Model 2008). In addition, skills acquired in immigrants’ birth countries and time spent in the U.S. labor force help determine the earnings of immigrants (Funkhouser and Trejo 1995). From a modeling perspective, immigrants’ arrival cohort and tenure of U.S. residence influence their earnings.

As shown by Borjas (1985, 1987), tracking cohorts of immigrants across multiple cross-sections of data, when available, allows for the separate identification of the impact of arrival cohort and tenure of U.S. residence on earnings. The estimation equations used to conduct this exercise are as follows:
(1)
(2)

In Eqs. (1) and (2), Y is weekly earnings. X is a vector of standard social and demographic characteristics, including predicted experience, predicted experience squared, education, marital status, region of current residence, metropolitan area status, and English proficiency. A is a vector of dummy variables indicating how long an immigrant has lived in the United States; these variables are set to zero for U.S. native-born individuals. C is a vector of dummy variables identifying immigrants’ arrival cohorts. T is a vector of dummy variables indicating the survey year. In Eq. (1), ϕ1 captures the effect of being a white native. Native blacks (collectively) are the reference group for both native whites and the immigrant arrival cohort variables in the equation. In Eq. (2), captures the impact of migrant status of native blacks. In this equation, native black movers are the reference group for native whites and the immigrant-arrival-cohort variables.3 Last, and υ are random error terms.

To identify the effects of both arrival cohort and duration of U.S. residence, Eqs. (1) and (2) impose the restriction that the period effect on earnings is the same for both immigrants and those born in the United States. Therefore, the period effect is estimated for U.S.-born blacks and whites, and this information is used to identify cohort and duration effects for immigrants (Borjas 1987).

## Results

### Descriptive Results

Table 1 shows summary statistics using only data from the 2000 U.S. census. Column 1 shows that native blacks (collectively) have weekly earnings of $761 USD, which is more than the weekly earnings of black immigrants from Haiti (column 5), Latin America (column 6), and non-English-speaking African countries (column 9). Black immigrants from English-speaking Caribbean (column 4) and African countries (column 8) earn$75 and \$154 more than native blacks, respectively.

The descriptive results in columns 2 and 3 of Table 1 show that native black movers are systematically different from native black nonmovers.4 Relative to nonmovers, movers earn more, have more years of education, and are more likely to be married. In fact, across most measures included in Table 1, native black movers are more similar to foreign-born individuals than to native black nonmovers.

Table 1 also shows dummy variables capturing the year in which immigrants arrived in the United States. These variables show significant time-of-entry variation among black immigrants. For example, column 4 shows that 29 % of black immigrants from the English-speaking Caribbean immigrated to the United States prior to 1980. In contrast, only 17 % of African immigrants (column 7) arrived prior to 1980. Table 2 provides summary statistics for the entire merged data set. Because this table includes data from the 2005–2007 ACS, it contains a dummy variable for black immigrants who migrated between 2000 and 2007. Although the summary statistics in Tables 1 and 2 are different, most of the relative patterns across the subgroups are similar.

### Regression Results for Weekly Earnings

Table 3 presents coefficients from estimating Eq. (1) for weekly earnings, with native blacks (collectively) as the reference group for both the native whites and the dummy variables for immigrant arrival cohort. The coefficients are estimated using ordinary least squares, and robust standard errors are shown in parentheses. Because the weekly earnings variable represents nominal earnings, the survey-year dummy variables in each model capture any inflation and real wage changes occurring over the survey waves.

Table 3 shows native whites earn approximately 24 % (exp(.212) – 1) more than native blacks, the reference group. The coefficients on the arrival-cohort variables measure initial earnings differences (i.e., when immigrants’ tenure of U.S. residence is evaluated between zero and five years) between immigrant arrival cohorts and native blacks. For example, the coefficient on the 1970–1974 arrival cohort in column 1 shows that black immigrants from the English-speaking Caribbean earn 15 % (exp(–.168) – 1) less than native blacks (the reference group) in their first five years after arriving in the United States. Indeed, the arrival-cohort coefficients in Table 3 are uniformly negative, implying that every black immigrant arrival cohort earns less than native blacks within their first five years after arriving in the United States.

The coefficients on the years-since-arrival dummy variables, which are the focus of this study, show how black immigrants’ earnings change as they adjust to the U.S. labor market. Table 3 also shows that these variables are uniformly positive and steadily increase in magnitude as immigrants’ tenure of U.S. residence increases. For example, black immigrants from the English-speaking Caribbean (column 1) who have been in the United States 6–10 years earn 12 % (exp(.115) – 1) more than immigrants from this region who have been in the country 0–5 years, the reference category. Likewise, black immigrants from the English-speaking Caribbean who have resided in the United States 16–20 years earn 21 % (exp(.189) – 1) more than immigrants from this region who have been in the country 0–5 years.

The negative coefficients on the arrival-cohort dummy variables, in conjunction with the positive coefficients on the years-since-arrival dummy variables, suggest that some black immigrant arrival cohorts may overtake or converge with the earnings of native blacks as their tenure of U.S. residence increases. To determine if and when a particular arrival cohort is projected to overtake the earnings of native blacks, the coefficient on a particular arrival-cohort variable must be added to the coefficient on one of the years-since-arrival variables. For example, the coefficient on the 1980–1984 arrival-cohort variable for black immigrants from the English-speaking Caribbean (column 1) shows that this cohort earns 14 % (exp(–.146) – 1) less than native blacks shortly after arriving in the United States. The coefficient on the dummy variable for 6–10 years since arrival in column 1 suggests that the earnings gap for the 1980–1984 immigrant-arrival cohort from the English-speaking Caribbean declines by 12 percentage points after they have resided in the United States between 6 and 10 years. After members of this cohort have lived in the United States for more than 20 years, they are projected to earn approximately 7 % (exp(–.146 + .211) – 1) more than native blacks—an estimate calculated by adding the coefficient on the 1980–1984 dummy variable to the coefficient on the dummy variable for more than 20 years in column 1. Indeed, column 1 shows that every cohort from the English-speaking Caribbean is projected to overtake the earnings of native blacks.

When I compare the coefficients on immigrant arrival cohort to coefficients on the years since arrival for the immigrant subgroups in columns 2–6 of Table 3, only one cohort from Latin America (the 1990–1994 arrival cohort) and four cohorts (post-1984) from English-speaking African countries are projected to overtake the earnings of native blacks. No arrival cohort from Haiti (column 2) or non-English-speaking African countries (column 6) is projected to surpass the earnings of native blacks.

The coefficients on years since arrival in Table 3 also show that black immigrants from English-speaking African (column 5) and Caribbean (column 1) countries experience the most rapid earnings growth as their tenure of U.S. residence increases. For example, relative to black men from English-speaking African countries who have been in the United States 0–5 years (the reference group), those who have been in the country 6–10 years earn 18 % (exp(.167) – 1) more. In contrast, the earnings growth of black men who emigrate from non-English-speaking African countries (column 6) who have been in the country 6–10 years is less than one-half that of black men from English-speaking African countries who have the same U.S. tenure. Indeed, the years-since-arrival variables show that relative to black men from non-English-speaking African countries, those from English-speaking African countries experience greater earnings growth at every tenure of U.S. residence. Similarly, the years-since-arrival dummy variables in columns 1–3 show that relative to black immigrants from Haiti and Latin America, black immigrants from English-speaking Caribbean countries experience greater earnings growth.

Table 4 presents weekly earnings regressions based on Eq. (2) This specification includes a dummy variable for native black nonmovers, where native black movers are the reference group for native whites, native black nonmovers, and the arrival-cohort dummy variables. The coefficients on the native black nonmover variables indicate this group earns approximately 8 % less than native black movers. This result implies that the initial earnings gap between native black movers and any immigrant arrival cohort is greater than the initial earnings gap between native blacks (collectively) and any arrival cohort. Because this specification has no impact on the years-since-arrival variables estimated in Table 3, every black immigrant arrival cohort will require a greater tenure of U.S. residence to converge with or surpass the earnings of native black movers, which reduces the number of arrival cohorts that are able to earn substantially more than native black movers as their tenure of U.S. residence increases.

To illustrate this point, Figs. 1, 2, and 3 show the projected earnings gaps between native blacks (both collectively and the subset of native black movers) and black immigrants from the English-speaking Caribbean, Latin America, and English-speaking African countries. I display these three subgroups because each contains at least one arrival cohort projected to surpass the earnings of native blacks (collectively) in Table 3, based on the dummy variable for the longest duration of U.S. residence (more than 20 years). The y-axis on these figures shows the adjusted weekly earnings gap between subgroups of black immigrants and either native blacks (collectively) or native black movers.

Panel a of Fig. 1 shows that every cohort of immigrants from the English-speaking Caribbean is projected to achieve greater earnings than native blacks (collectively), as discussed previously. In contrast, panel b indicates that no arrival cohort from this region is projected to surpass the earnings of native black movers by more than 2 %. For example, among cohorts from the English-speaking Caribbean, the 1980–1984 arrival cohort earns approximately 7 % more than native blacks after residing in the United States for more than 20 years (Fig. 1, panel a). When the 1980–1984 cohort is compared with native black movers (Fig. 1, panel b), their earnings advantage declines to 1.8 %. Similarly, although panel a of Fig. 2 shows that the 1995–1999 arrival cohort is projected to earn 1.8 % more than native blacks after residing in the United States for more than 20 years, panel b of Fig. 2 indicates that no immigrant cohort from Latin America is projected to surpass the earnings of native-black movers.

Turning to projections for immigrants from English-speaking African countries, panel a of Fig. 3 shows that five of the eight arrival cohorts studied from this region are projected to converge with or surpass the earnings of native blacks. In contrast, panel b of Fig. 3 shows that only the three post-1990 arrival cohorts are projected to surpass the earnings of native black movers.

## Discussion

### Main Findings

Four major findings emerge from this study. First, after 20 years of U.S. residence, the earnings of most arrival cohorts of English-speaking black Caribbean immigrant males are projected to converge with or slightly overtake the earnings of native black movers (see Fig. 1, panel b). Second, post-1990 arrival cohorts of black men from English-speaking African countries are projected to earn more than native black movers after residing in the United States for more than 20 years (see Fig. 3, panel b). Third, relative to black immigrants from non-English-speaking countries, black immigrants from English-speaking countries in Africa and the Caribbean experience more rapid earnings growth (see Tables 3 and 4). Last, no arrival cohort of black immigrant men from non-English-speaking countries (i.e., Haiti, Latin America, and non-English-speaking African countries) is projected to surpass the earnings of native black movers (see Table 4). These findings raise two important questions for the study of earnings differences among blacks in the United States.

#### Why Do Black Immigrants from English-speaking Countries Overtake the Earnings of Native Black Movers?

The earnings convergence pattern of immigrants from the English-speaking Caribbean is consistent with predictions of immigrant selectivity. Chiswick (1978:899) contended that “if the foreign and native born have the same level of innate labor market ability and work motivation, the earnings of the foreign born would approach, and might equal, but would not exceed that of the native born, ceteris paribus.” This argument suggests that immigrants from the English-speaking Caribbean and native black movers are equally selected on unobserved labor market characteristics correlated with earnings. Similarly, the result for immigrants from English-speaking African countries implies that some arrival cohorts from this region are more positively selected on unobserved labor market characteristics than native black movers—a reasonable finding given the relatively high costs of migration (e.g., social, psychological, and economic) for African immigrants.

A number of factors likely explain why only post-1990 arrival cohorts surpass the earnings of native black movers (Konadu-Agyemang et al. 2006; Thomas 2011), but changes in immigration policy is a prominent reason. The Immigration Act of 1990 introduced a Diversity Visa Lottery Program designed to increase the number of immigrants from countries underrepresented in the United States and increased the number of immigrants admitted on the basis of job skills (Kent 2007). This policy change provided new avenues for highly skilled African immigrants to enter the United States (McCabe 2011). Prior to the introduction of the Diversity Visa Lottery Program, political refugees accounted for a greater proportion of migration streams from Africa, which could explain why the earnings profiles of post-1990 immigrants differ from those of earlier migrants from English-speaking African countries.

#### What Explains Earnings Differences among Black Immigrants by Linguistic Heritage?

Because the Diversity Visa Lottery Program also increased immigration from non-English-speaking African countries (McCabe 2011), other factors must be at play to explain the poor earnings outcomes for immigrant cohorts from this region. Some scholars have argued that migration in response to political unrest or pressure is less favorably selected on the more able or more highly motivated (Borjas 1994; Chiswick 1978). Thus, the relatively flat earnings trajectories of immigrants from Haiti and non-English-speaking African countries—the country and region from which most black refugees or asylum seekers emigrate—could explain these differences. Moreover, most English-speaking countries in Africa and the Caribbean are former colonies of Great Britain. Consequently, relative to non-English-speaking countries, the ideological and institutional distance between English-speaking countries and the United States is smaller, potentially having a positive impact on the post-migration earnings trajectories of men from these countries.

Lack of proficiency with the English language may also negatively impact the earnings trajectories of immigrants from non-English-speaking countries, particularly among immigrants from Latin America. To account for this factor, I control for whether respondents speak or read English well. This measure, however, may not completely capture nuanced differences in the ability to speak English, such as difficulties that U.S. natives have communicating with individuals with particular foreign accents. Therefore, immigrants from non-English-speaking countries may also have weaker earnings trajectories because of their unfamiliarity with the idiosyncrasies of English spoken in the United States (Portes and Rumbaut 2007).

### Limitations

One of the study’s central findings—that black immigrants from English-speaking countries are projected to converge with or overtake the earnings of native black movers—assumes that the impact of U.S. tenure is constant across arrival cohorts. For example, the study assumes that more recent cohorts of black immigrant men from English-speaking African countries will experience earnings growth similar to that of earlier arrival cohorts from this region. Duleep and Regets (1999:188) argued, however, that variation in initial skills transferability creates an inverse association between immigrants’ initial earnings and their earnings trajectories such that “immigrants with low initial skill transferability will have greater human-capital investment, and hence greater earnings growth, than immigrants with high initial skill transferability.” The small differences across arrival cohorts from the English-speaking Caribbean imply that the transferability of pre-migration skills does not vary much for this subgroup. However, because of the sizable variation among African immigrants, the earnings growth of more recent immigrant arrival cohorts from this region should be reexamined after they have resided in the United States longer.

## Conclusion

Black immigrants are one of America’s most diverse immigrant subgroups: they speak a variety of languages and migrate from vastly different birth-country contexts. In spite of this diversity, few studies have used the broader black immigrant population to understand labor market differences between black immigrants and native blacks. This study finds that earnings differences between the two groups stem primarily from factors correlated with selective migration and characteristics unique to the language heritage of immigrants’ birth countries.

## Acknowledgments

Support for this research was provided by a grant from the Eunice Kennedy Shriver National Institute of Child Health and Human Development (Grant #5R24HD047879) and the IC2 Institute at the University of Texas at Austin.

## Notes

1

The Latin American subgroup comprises black immigrants from Mexico and Spanish-speaking countries in Central America, South America, and the Caribbean. The linguistic distinction among countries comes from the U.S. Central Intelligence Agency’s 2012 World FactBook (https://www.cia.gov/library/publications/the-world-factbook/index.html).

2

Models were also estimated using hourly earnings rather than weekly earnings. The substantive findings are the same using either measure.

3

Models are also estimated with movers defined as black natives who moved within the last five years and who do not currently reside in their state of birth (results not shown); the substantive results of the article remain the same.

4

A supplemental analysis not included with this manuscript indicates that native whites who moved across states since birth also earn significantly more than native whites who reside in their birth state.

## References

Arnold, F. W. (
1984
).
West Indians and London’s hierarchy of discrimination
.
Ethnic Groups
,
6
,
47
64
.
Borch, C., & Corra, M. K. (
2010
).
Differences in earnings among black and white African immigrants in the United States, 1980–2000: A cross-sectional and temporal analysis
.
Sociological Perspectives
,
53
,
573
592
. 10.1525/sop.2010.53.4.573
Borjas, G. J. (
1985
).
Assimilation, changes in cohort quality, and the earnings of immigrants
.
Journal of Labor Economics
,
3
,
463
489
. 10.1086/298065
Borjas, G. J. (
1986
).
The self-employment experience of immigrants
.
Journal of Human Resources
,
21
,
485
506
. 10.2307/145764
Borjas, G. J. (
1987
).
Self-selection and the earnings of immigrants
.
American Economic Review
,
77
,
531
553
.
Borjas, G. J. (
1994
).
The economics of immigration
.
Journal of Economic Literature
,
32
,
1667
1717
.
Borjas, G. J. (
1995
).
Assimilation and changes in cohort quality revisited: What happened to immigrant earnings in the 1980s?
.
Journal of Labor Economics
,
13
,
201
245
. 10.1086/298373
Butcher, K. F. (
1994
).
Black immigrants in the United States: A comparison with native blacks and other immigrants
.
Industrial and Labor Relations Review
,
47
,
265
284
. 10.2307/2524420
Chiswick, B. R. (
1978
).
The effect of Americanization on the earnings of foreign-born men
.
Journal of Political Economy
,
86
,
897
921
. 10.1086/260717
Chiswick, B. R. (
1991
).
Speaking, reading, and earnings among low-skilled immigrants
.
Journal of Labor Economics
,
9
,
149
170
. 10.1086/298263
Chiswick, B. R., & Miller, P. W. (
1995
).
The endogeneity between language and earnings: International analyses
.
Journal of Labor Economics
,
13
,
246
288
. 10.1086/298374
Corra, M. K., & Kimuna, S. R. (
2009
).
Double jeopardy? Female African and Caribbean immigrants in the United States
.
Journal of Ethnic & Migration Studies
,
35
,
1015
1035
. 10.1080/13691830902957767
Correll, S. J., Benard, S., & Paik, I. (
2007
).
Getting a job: Is there a motherhood penalty?
.
American Journal of Sociology
,
112
,
1297
1339
. 10.1086/511799
Dodoo, FN-A (
1999
).
Black and immigrant labor force participation in America
.
Race and Society
,
2
,
69
82
. 10.1016/S1090-9524(00)00008-5
Dodoo, FN-A, & Takyi, B. K. (
2002
).
Africans in the diaspora: Black-white earnings differences among America’s Africans
.
Ethnic and Racial Studies
,
25
,
913
941
. 10.1080/0141987022000009377
Duleep, H. O., & Regets, M. C. (
1999
).
Immigrants and human-capital investment
.
American Economic Review
,
89
,
186
191
. 10.1257/aer.89.2.186
Funkhouser, E., & Trejo, S. J. (
1995
).
The labor market skills of recent male immigrants: Evidence from the Current Population Survey
.
Industrial and Labor Relations Review
,
48
,
792
811
. 10.2307/2524357
Glazer, N., & Moynihan, D. P. (
1979
).
Beyond the melting pot: The Negroes, Puerto Ricans, Jews, Italians, and Irish of New York City
.
Cambridge, MA
:
MIT Press
.
Grosfoguel, R. (
2003
).
Race and ethnicity or racialized ethnicities?
.
Ethnicities
,
4
,
315
336
. 10.1177/1468796804045237
Kalmijn, M. (
1996
).
The socioeconomic assimilation of Caribbean American blacks
.
Social Forces
,
74
,
911
930
. 10.1093/sf/74.3.911
Kent, M. M. (
2007
).
Immigration and America’s black population
.
Population Bulletin
,
62
(
4
).
Konadu-Agyemang, K., Takyi, B. K., & Arthur, J. A. (
2006
).
The new African diaspora in North America: Trends, community building, and adaption
.
Lanham, MD
:
Lexington Books
.
Korenman, S., & Neumark, D. (
1991
).
Does marriage really make men more productive?
.
Journal of Human Resources
,
26
,
282
307
. 10.2307/145924
Lobo, A. P. (
2001
).
US diversity visas are attracting Africa’s best and brightest
.
Population Today
,
29
(
5
),
1
2
.
Mason, P. L. (
2010
).
Culture and intraracial wage inequality among America’s African diaspora
.
American Economic Review
,
100
,
309
315
. 10.1257/aer.100.2.309
McCabe, K. (
2011
).
African immigrants in the United States
.
AfricaFocus Bulletin
Model, S. (
1995
).
West Indian prosperity: Fact or fiction?
.
Social Problems
,
42
,
535
553
. 10.2307/3097045
Model, S. (
2008
).
West Indian immigrants: A black success story?
.
New York, NY
:
Russell Sage Foundation
.
Moore, A. R., & Amey, F. K. (
2002
).
Earnings differentials among male African immigrants in the United States
.
Equal Opportunities International
,
21
(
8
),
30
50
. 10.1108/02610150210787244
Palmer, R. W. (
1974
).
A decade of West Indian migration to the United States, 1962–1972: An economic analysis
.
Social and Economic Studies
,
23
,
571
587
.
Portes, A., & Rumbaut, R. G. (
2007
).
Immigrant America: A portrait
.
Berkeley
:
University of California Press
.
Ruggles, S., Sobek, M., Alexander, T., Finch, C. A., Goeken, R., Hal, P., & Ronnander, C. (
2004
).
Integrated Public Use Microdata Series
Minneapolis
:
Minnesota Population Center
.
Sowell, T. (
1975
).
Race and economics
.
New York, NY
:
D. McKay Co.
.
Sowell, T. (
1978
).
Three black histories
. In Sowell, T., & Collins, L. D. (Eds.),
Essays and data on American ethnic groups
(pp.
7
64
).
Washington, DC
:
Urban Institute
.
Thomas, K. J. (
2011
).
What explains the increasing trend in African emigration to the US?
.
International Migration Review
,
45
,
3
28
. 10.1111/j.1747-7379.2010.00837.x
Tolnay, S. E. (
2003
).
The African American “Great Migration” and beyond
.
Annual Review of Sociology
,
29
,
209
232
. 10.1146/annurev.soc.29.010202.100009
Vickerman, M. (
1999
).
Crosscurrents: West Indian immigrants and race
.
New York, NY
:
Oxford University Press
.
Waters, M. C. (
1999
).
Black identities: West Indian immigrant dreams and American realities
.
New York, NY
:
Russell Sage Foundation
.