Looking through the Shades

Abstract

The purpose of this study is to determine whether a labor market penalty exists for members of immigrant groups as a result of being phenotypically different from white Americans. Specifically, the authors examine the link between skin shade, perhaps the most noticeable phenotypical characteristic, and wages for immigrants from five regions: (1) Europe and Central Asia; (2) China, East Asia, South Asia, and the Pacific; (3) Latin America and the Caribbean; (4) Sub-Saharan Africa; and (5) the Middle East and North Africa. Using data from the New Immigrant Survey, a nationally representative multi-cohort longitudinal study of new legal immigrants to the United States, the authors find a skin shade penalty in wages for darker immigrants. However, disaggregating by region of origin shows that this finding is driven exclusively by the experience of immigrants from Latin America; the wage penalty for skin tone is substantial for self-reported nonblack Latin American immigrants. The effects of colorism are much less pronounced or nonexistent among other national-origin populations. Furthermore, although a skin shade penalty is not discernible among African immigrants, findings show that African immigrants experience a racial wage penalty.

Keywords

skin shade discrimination earnings immigrants United States

Since the Immigration Act of 1965, there has been substantial growth in the presence of nonwhite immigrants to the United States. Although the foreign-born population in the United States stood at slightly less than 10 million in 1970, it currently stands at roughly 40 million, nearly 13 percent of the total U.S. population (Grieco and Trevelyan 2010). These immigration patterns have contributed to dramatic changes in the demographic composition of the United States. For example, Latin American immigrants alone, who currently constitute more than half of the nation’s foreign-born population, will be the source of over 60 percent of the nation’s population growth over the next 40 years and are projected to constitute nearly one-third of the total U.S. population by 2050 (U.S. Census Bureau 2008).

Similarly, since Congress’s repeal of the Chinese Exclusion Act in 1943 and the relaxation of immigration laws under the McCarran Act of 1952, there has been a dramatic increase in immigration from Asian countries. The population of Asian origin within the United States has tripled over the past three decades (Xie and Goyette 2004). Immigrants from Caribbean countries accounted for 10 percent of the foreign-born population in 2000 (Gelatt and Dixon 2006), and African immigrants constitute 3 percent of the foreign-born population, corresponding to over 1 million people (Grieco 2004). It is estimated that almost one in four youths aged 18 years and younger in the United States is foreign born or has foreign-born parents, and this proportion continues to grow (Portes and Rumbaut 2014).

The rise in the foreign-born population has occurred simultaneously with shifts in the structure of the U.S. labor force since the mid-1970s (Danziger and Gottschalk 1995; Morris and Western 1999). In this context, the economic and educational position of immigrant groups within the United States is varied and drastically unequal. On average, Asian immigrants have achieved comparatively high degrees of economic success in the United States, but many immigrants from Latin America, the Caribbean, and Africa have been incorporated into lower status occupations. Specifically, 51 percent of foreign-born Asians aged 25 years and older hold bachelor’s degrees or higher, compared with only 12 percent, 38 percent, and 20 percent of their counterparts born in Latin America, Africa, and the Caribbean, respectively (Kent 2007; U.S. Census Bureau 2006). Similarly, the median household income of Asians is over $68,000, which is nearly $30,000 more than the median income of immigrants from Latin America, Africa, and the Caribbean. Only one tenth of foreign-born Asians live below the poverty line, compared with one fifth of the population for the other aforementioned groups (Kent 2007; U.S. Census Bureau 2006).

The magnitude of socioeconomic differences between immigrant groups and the substantial increases in their respective populations make it essential for researchers and policy makers to understand the pathways taken by groups entering the United States from different regions of the world. Previous studies have shown that assimilation and integration patterns can be conditional on a range of structural factors that result in segmented patterns of immigrant incorporation (Portes and Zhou 1993; Zhou 1997). Given the history of differential treatment of ethnic minorities within the United States, one factor that warrants greater attention is discrimination. Although the Immigration Act of 1965 dismantled the restrictive barriers of earlier immigration laws and reduced preference for individuals from European countries, members of nonwhite immigrant groups still may experience discriminatory penalties in U.S labor markets.

The purpose of this study is to determine whether a labor market penalty extends to some members of immigrant groups as a result of being phenotypically different from the white American norm, the racial majority in the United States. Specifically, we examine the link between skin shade, perhaps the most noticeable phenotypical characteristic, and wages for immigrants from five regions: (1) Europe and Central Asia; (2) China, East Asia, South Asia, and the Pacific; (3) Latin America and the Caribbean; (4) Sub-Saharan Africa; and (5) the Middle East and North Africa. Previous research has established a wage penalty associated with darker skin tones among immigrant workers in the United States. For example, using data from the New Immigrant Survey (NIS), the only nationally representative data set on new immigrants to the United States that also contains measures of skin shade, Hersch (2008) found that a lighter skin tone can result in approximately 17 percent more in earnings for immigrants, even after controlling for educational attainment and other productivity-linked demographic characteristics.

However, Hersch (2008) and most previous studies using the NIS grouped immigrants from all countries and racial groups into a single sample or focused exclusively on immigrants from particular regions (e.g., Frank, Redstone Akresh, and Lu 2010), making it unclear whether the skin shade effect identified in previous studies is invariant across all racial and national-origin groups. We address this issue in this study by conducting an analysis of the link between skin tone and wages for immigrants disaggregated by region and by race.

In the remainder of this article, we explain why it is reasonable, given previous research, to expect that immigrants with darker skin tone will experience a wage penalty relative to those with lighter skin tone, followed by a description of our analytic plan and results. We conclude by discussing the implications of our findings for understanding the differential patterns of assimilation and integration for immigrants from different regions of the world.

Labor Market Penalties and Minorities within the United States

The black experience in the United States is deeply rooted in socioeconomic disadvantage stemming from government sanctioned discriminatory treatment, dating back to the era of slavery (before 1865) through the era of Jim Crow (1876–1965). Historically, much of the discrimination in the United States has been directed against black Americans as the primary targets of stigmatization. Thus, to the extent that immigrants experience discrimination, one might anticipate that their exposure to discrimination also will fall along racial lines.

Most previous studies on discriminatory labor market penalties have examined the labor market experiences of black Americans. In general, research shows that blacks experience differential treatment relative to their white counterparts (Darity, Jason, and Guilkey 2001). For example, although both whites and blacks older than 25 experience increased earnings with each additional level of schooling, whites earn more than blacks at each level of schooling. Moreover, research suggests that blacks with some college or associate’s degrees have higher unemployment rates than whites who never finished high school (Harris 2011).

Studies based on the National Longitudinal Survey of Youth have shown that although no black-white wage gap exists at labor force entry, a racial wage gap develops over time as blacks reap relatively smaller gains from job mobility (Oettinger 1996; Tomaskovic-Devey, Thomas, and Johnson 2005). The black wage disadvantage persists net of education, experience, hours worked, occupation, authority, region, and city size (Smith 1997). Furthermore, the Council on Contemporary Families has shown that although this income gap declined over the last quarter of the twentieth century, blacks in supervisory positions earned roughly 80 percent of the income earned by whites near the end of the century (McBride Murry and Liu 2014).

Many of these patterns transcend gender. Despite the gains made by black women between 1940 and 1980, they continue to earn less than their white counterparts (Anderson and Shapiro 1996). Furthermore, Anderson and Shapiro (1996) showed that black women are less likely to hold high-wage occupations than their white counterparts, even if they have parallel levels of education, work experience, job tenure, and type of occupation. In fact, they must have higher levels of education, work experience, and job tenure relative to their white female counterparts to obtain high-wage occupations. Hamilton, Austin, and Darity (2011) produced similar findings.

The black labor market penalty is robust. It also exists for nonwage labor market outcomes, including promotions, processes that determine promotions, and layoffs. For example, using data from the National Organizational Study, which contains interviews with the employers of respondents from the General Social Survey, Baldi and McBrier (1997) found that relative to whites, black workers with comparable education, experience, and training, and at similar types of firms, are only half as likely to receive promotions. On the basis of data from the Multi-City Survey of Urban Inequality, Elliott and Smith (2004) found that both black men and women are less likely to receive promotions at the higher end of the occupational hierarchy than white men. These differences exist even if the comparisons are made among individuals with similar years of education, total work experience, prior job-specific experience, and employer tenure (and family considerations for women).

Racial differences also exist in processes that determine promotions in the labor force. Smith (2005) found that before receiving promotions, relative to white men, black men must work longer periods of time after leaving school, and black women must have more prior job-specific experience and log more time on the job, all else equal. Furthermore, previous research provides evidence that blacks’ human capital credentials receive more intense scrutiny than whites’ credentials when competing for promotions (Baldi and McBrier 1997), particularly when vying for managerial (Wilson, Sakura-Lemessy, and West 1999) and supervisory positions (Smith 2001).

With respect to job layoffs, Wilson and McBrier (2005) found that the percentage of layoffs for blacks is nearly twice that for whites (31 percent and 16 percent, respectively). Blacks’ greater layoff rates exist in both the private and public sectors. They also found that blacks are more vulnerable to being laid off than whites; increases in tenure with employer, college and postcollege education, and union membership are associated with decreases in the layoff likelihood for whites but not for blacks. Thus, blacks face a route toward downward occupational movement that is structured less by traditional stratification-based causal factors and experience downward mobility more rapidly than whites (Wilson and Roscigno 2010). Along the same line, reentry into the labor market comes with additional costs for blacks, as blacks experience longer periods of job search than white job seekers with similar levels of human capital such as education and cognitive skill (Tomaskovic-Devey et al. 2005).

Some scholars have suggested that the racial wage gap stems from blacks’ deficiencies in human capital, skills in particular, or the labor market characteristics of blacks rather than discrimination (Johnson and Neal 1998; Neal and Johnson 1996; O’Neill 1990). However, a long line of research that directly tests for discrimination in the labor market through field experiments finds evidence of a black labor market penalty (Bertrand and Mullainathan 2004; Pager 2007; Zhao, Ondrich, and Yinger 2006). These studies typically match job applicants on every category (e.g., résumé or credentials) other than the factor on which discrimination is being tested (e.g., sex, race).

Furthermore, Coleman (2003) found that even if employers’ evaluations of workers’ skills on the basis of actual on-the-job performance ratings relative to other workers are controlled, blacks earn 19 percent less than whites. Less than half (42 percent) of this disadvantage can be attributed to blacks’ lower levels of human capital (i.e., education, tenure on the job, and age). He also found that when white and black men have the same employer’s competitive performance rating and similar levels of human capital, racial wage differences actually slightly increase. Thus, human capital and employers’ job-specific skill ratings explain less than half of the black male wage disadvantage.

Research also suggests that certain immigrant groups, particularly Mexican immigrants with less European features, experience significant levels of discrimination, which blocks their ability to fully assimilate into American society (Alba 2005; Portes and Rumbaut 2014). Using data on four generations of Mexican-origin individuals in the United States, Telles and Ortiz (2008) also found that among fourth-generation Mexican children, 66 percent reported being stereotyped by others and 46 percent reported experiencing discrimination. Moreover, Pager, Western, and Bonikowski (2009) used data from a field experiment conducted in the low-wage sector of the New York City labor market and found that relative to whites with similar resumes, both black and Latino applicants were significantly less likely to receive callbacks or job offers.

Skin Shade and the Labor Market

Perhaps a useful proxy for race is skin shade, which on average differs substantially between blacks and nonblacks. However, skin tone also varies within racial groups, which enables researchers to investigate whether darker skin shade exercises an independent adverse effect on wages. Research on skin shade and social outcomes has focused on five main issues: employment (Goldsmith, Hamilton, and Darity 2006, 2007; Diette et al. 2015), educational attainment (Keith and Herring 1991; Monk 2014; Seltzer and Smith 1991), political beliefs (Hochschild and Weaver 2007), treatment by the criminal justice system, marital outcomes (Eberhardt et al. 2006; Gyimah-Brempong and Price 2006; Hamilton, Goldsmith, and Darity 2008), and income (Monk 2014).

With regard to wages, research has established that darker skin shade is associated with lower wages for black men. Goldsmith et al. (2006, 2007) found that although wages for light-skinned black men are lower than (but close to) those of whites, wages substantially decline as skin tone darkens, with medium- and dark-skinned blacks earning substantially less (by at least 10 percent) than whites. These findings hold even if the comparison is made between people with similar occupations, levels of education, retrospective high school performance level, labor market experience, health status, and self-esteem. Their studies even accounted for respondents’ socioeconomic background and neighborhood quality at age 16 as well as current workplace features, such as union status, full-time work status, firm size, having a supervisory role, and degree of contact with customers. Because skin shade is separate from culture and unrelated to intelligence among U.S. blacks (Hill 2002), these differences cannot be attributed to cultural or genetic factors.

Similar to the research on labor market discrimination within the United States, the majority of research on skin shade has concentrated on black Americans. However, skin tone also varies within groups, making it possible to estimate whether labor market penalties exist for darker complexioned people from the same immigrant group. Given the findings of previous research that labor market discrimination exists along racial lines within the United States and on the general penalty associated with darker skin shade, it is reasonable to expect that immigrants with darker skin shade experience a penalty in the labor market (Frank, Akresh, and Lu 2010; Hersch 2002, 2008, 2011).

For example, researchers have found that darker skinned Mexican Americans with more Native American characteristics earned significantly less and achieved fewer years of schooling than lighter skinned Mexican Americans with more European characteristics (Murguia and Telles 1996; Telles and Murguia 1988). Hersch (2008, 2011) tested whether lighter skin shade is statistically associated with greater earnings among recent immigrants to the United States using data from the NIS. She found that even after controlling for education, English-language proficiency, occupation in source country, family background, ethnicity, race, and country of birth, lighter skinned immigrants earned significantly more than their darker skinned counterparts. No previous study, however, has comparatively examined variation in earnings by skin shade for immigrants by region of origin or by race, which leaves open the question of whether the association between skin shade and earnings varies among immigrants from particular regions of the world. Below we address this gap in the literature.

Data and Analytic Plan

Data for this study are from the NIS, a nationally representative multi-cohort longitudinal study of new legal immigrants to the United States. The data were collected from May to November in 2003 and contain responses for 8,573 legal immigrants to the United States. who received permanent resident status. Because disaggregating by particular nation of birth yields some subsamples that lack sufficient size to provide reliable estimates, we sorted immigrants into the following regions of birth: (1) Europe and Central Asia; (2) China, East Asia, South Asia, and the Pacific; (3) Latin America and the Caribbean; (4) Sub-Saharan Africa; and (5) the Middle East and North Africa. Analysis was not performed on immigrants from Canada, because this region only contained 23 observations.

Nearly half of the NIS sample immigrated from Latin America and the Caribbean (45.9 percent); this is almost twice as many observations as the next largest subsample, China, East Asia, South Asia, and the Pacific. In contrast, the Middle Eastern and North African region accounts for only 3.7 percent of the NIS sample, corresponding to 66 observations.

The first panel of Figure 1 illustrates the distribution of region of birth across the total NIS sample. Because a vast majority of the sample hails from Latin America and the Caribbean, Figure 1 also shows the racial composition of this subgroup. About 70 percent of the respondents classify themselves as white, and fewer than 10 percent report their race as black. Note that the Latin American and Caribbean subsample is composed predominantly of immigrants from Latin America (roughly 90 percent) rather than from the Caribbean (roughly 10 percent).

Figure 1.

Breakdown of New Immigrant Survey sample.

There are several noteworthy points regarding the analytical approach used in this study. First, with regard to the measurement of our primary variables of interest, wages were measured as the natural logarithm of the hourly wage before taxes and deductions. For those who reported that they were salaried workers, yearly salary was divided by the number of weeks worked per year, with that figure divided by the number of hours worked per week. Second, data on skin shade were based on an assessment and report by the interviewer using an 11-point scale ranging from 0 (no pigment/albino) to 10 (darkest skin shade possible). The skin shade scale is displayed in Figure 2.

Figure 2.

Skin shade scale.

Third, in addition to estimating the association between skin shade and wages for the full sample, we repeated this analysis for four sets of subsamples. The five aforementioned regions of birth represent the first group of subsamples. The analysis was conducted separately for each region. Next, we stratified the sample by race and conducted the analysis separately for blacks, whites, and Asians. These three self-classified racial groups constitute 89 percent of the total observations in the larger sample; other races reported are not large enough to yield reliable results. Because the Latin American and Caribbean region accounts for a large proportion of the total sample, we then report findings by race after excluding this region. Finally, we also report findings by race for the Latin American and Caribbean subsample alone.

Fourth, we collapsed the racial categories used for the Latin American population into black and nonblack (insufficient sample size prevented analysis of Asian respondents in the Latin American subsample). Self-reported race is rather problematic among Latin American survey respondents (Hattam 2005; Lopez 2005; and Prewitt 2005). Prior research has shown that certain exogenous factors, such as tenure in the United States and socioeconomic status, may affect the race reported by Latino immigrants (Tafoya 2005). Furthermore, there is often a preference for whiteness (or nonblackness) among respondents, regardless of phenotype (Darity, Dietrich, and Hamilton 2005; Golash-Boza and Darity 2008). Latino immigrants are more likely than other immigrants to decline to respond to questions about race or to select “other” options when available (Hitlin, Brown, and Elder 2007). Therefore, we consider race among Latinos as blacks and nonblacks.

Finally, each subsample analysis included three separate regression specifications.¹ The first model is the most parsimonious, controlling for general human capital, socioeconomic background, and region of residence within the United States. We refer to this model as the baseline model. The second model adds a control for race/ethnicity when the analysis is conducted for each region of origin subsample. When the analysis is conducted for subsamples stratified by race (both with and without the Latin America and Caribbean subsample), we also control for Hispanic versus non-Hispanic and region of origin. The final model includes factors that capture immigrants’ current employment characteristics. Although the impact of skin shade on earnings may vary by gender within region of birth and by race, because of sample size issues, we did not conduct these analyses.

Table 1 presents the measures used in this study. The measures are grouped according to the models in which they are entered into the analysis.

Table 1.

Means, standard deviations, and descriptions for variables used in the analysis.

Variable Name	Description	Metric	Mean (SD)
Outcomes
Ln(wages)	Natural logarithm of hourly wage before taxes and deductions	0–4.6	2.400 (.640)
Actual wages	Hourly wage before taxes and deductions	0–99.4	11.030 (1.900)
Predictors for model 1
Skin shade	Based on interviewer report of skin shade	0 = lightest, 10 = darkest	4.110 (2.240)
Physical characteristics
Height below U.S. average	Average height (based on CDC data; Ogden et al. 2004) < 69 inches for men and < 64 inches for women	0–29.5	4.640 (2.830)
Height above U.S. average	Average height (based on CDC data; Ogden et al. 2004) > 69 inches for men and 64 inches for women	0–26.4	.300 (1.450)
BMI	Body mass index: 703 × (weight in pounds/height in inches squared)	9.27–100.2	26.600 (6.280)
Male	Respondent is male	0 = no, 1 = yes	.590
Age	Year of interview minus year of birth	18–69	36 (9.810)
Age²/100	(Year of interview minus year of birth)²/100	3.2–47.6	13.920 (7.730)
Human capital (profession before immigrating to U.S.)
Professional	Worked in a professional/managerial position	0 = no, 1 = yes	.200
Health	Worked in a health profession	0 = no, 1 = yes	.040
Service	Worked in a service profession	0 = no, 1 = yes	.050
Sales	Worked in a sales profession	0 = no, 1 = yes	.080
Production	Worked in a production profession	0 = no, 1 = yes	.160
U.S. experience	Interview date minus date first worked in the United States	0–42	5.310 (6.190)
(U.S. experience)²/100	(Interview date minus date first worked in the United States) ²/100	0–17.6	.660 (1.340)
Education in United States	Years of education completed within the United States	0–18	1.080 (2.590)
Non-U.S. education	Total years of education minus years of education in the United States	0–25	12.010 (4.920)
English proficient	Respondent speaks English well or very well	0 = no, 1 = yes	.510 (.500)
New arrival	Did not live in United States before acquiring green card	0 = adjustee, 1 = new arrival	.660
Spouse from United States	Respondent’s spouse is a U.S. citizen	0 = no, 1 = yes	.180
Memo	Interview performed after memo to code 0 skin shade only for albinos (NIS noticed overuse of zeroes for skin shade)	0 = no, 1 = yes	.330
Socioeconomic background
Father’s education	Father’s educational attainment	0–30	6.940 (6.610)
Income far below	Family income at age 16 far below average	0 = no, 1 = yes	.030
Income below	Family income at age 16 below average	0 = no, 1 = yes	.150
Income above	Family income at age 16 above average	0 = no, 1 = yes	.180
Income far above	Family income at age 16 far above average	0 = no, 1 = yes	.100
Region of residence in United States
Northeast	Living in the Northeast	0 = no, 1 = yes	.310
Midwest	Living in the Midwest	0 = no, 1 = yes	.110
West	Living in the West	0 = no, 1 = yes	.350
South	Living in the South	0 = no, 1 = yes	.230
Additional controls for model 2: race/ethnicity
Hispanic	Respondent is Hispanic	0 = no, 1 = yes	.410
AIAN	Respondent is American Indian/Alaskan Native.	0 = no, 1 = yes	.020
Asian	Respondent is Asian	0 = no, 1 = yes	.230
Black/African American	Respondent is black or African American	0 = no, 1 = yes	.110
NHPI	Respondent is Native Hawaiian/Pacific Islander	0 = no, 1 = yes	.010
White	Respondent is white	0 = no, 1 = yes	.550
Multiracial	Respondent is multiracial	0 = no, 1 = yes	.010
No race reported	Race not reported	0 = no, 1 = yes	.070
Additional controls for model 3: employment characteristics
Employment visa	Respondent holds an employment visa	0 = no, 1 = yes	.200
Diversity visa	Respondent holds a diversity visa	0 = no, 1 = yes	.140
Tenure	Start date of job minus interview date	0–37 years	2.410 (3.790)
Tenure²/100	(Start date of job minus interview date)²/100	0–13.6	.200 (.640)
Professional	Currently works in a professional/managerial position	0 = no, 1 = yes	.190
Health	Currently works in a health profession	0 = no, 1 = yes	.090
Service	Currently works in a service profession	0 = no, 1 = yes	.230
Sales	Currently works in a sales profession	0 = no, 1 = yes	.130
Production	Currently works in a production profession	0 = no, 1 = yes	.300
Government job	Respondent is employed by a governmental entity	0 = no, 1 = yes	.040
Union	Respondent’s job is covered by a union contract	0 = no, 1 = yes	.130
Outdoor job	Profession has high probability of outdoor work	0 = no, 1 = yes	.170
Paid hourly	Respondent reports wage by the hour	0 = no, 1 = yes	.780
Full-time	Respondent works 35 hours or more per week	0 = no, 1 = yes	.950
Self-employed	Respondent is self-employed	0 = no, 1 = yes	.000

Results

The first set of results is displayed in Table 2, which contains the findings of the baseline model for the total NIS sample and each subsample (five by region, three by race, three by race excluding the Latin American and Caribbean region, and two by race including only the Latin American region). Recall that the baseline model accounts for general human capital, socioeconomic background, and region of residence within the United States. The first model in Table 2 shows that darker skin shade is associated with lower wages for the full NIS sample (b = –.024). However, the next five models show that when the analysis is conducted by region of origin, the wage penalty associated with darker skin shade is experienced only by immigrants from Sub-Saharan Africa (b = –.077). The next three models show that the skin shade penalty in wages exists among white immigrants (b = –.031), and the penalty for white immigrants remains after excluding Latin America and the Caribbean (b = –.035). The final two models suggest a significant wage penalty for nonblack immigrants from Latin America and the Caribbean (b = –.017).²

Table 2.

Unstandardized coefficients from regressions of wages on skin shade and model 1 predictors listed in Table 1.

Baseline Model^a
		Region of Birth					Self-reported Race			Excluding Latin America and Caribbean			Including Only Latin America and Caribbean^b
Individual Variable	Total NIS Sample	Latin America and Caribbean	Europe and Central Asia	China and East/South Asia	Sub-Saharan Africa	Middle East and North Africa	Black	White	Asian	Black	White	Asian	Black	Nonblack
Constant	.897***(.189)	1.378***(.207)	.971**(.469)	.179(.485)	1.813***(.607)	.620(.694)	.729(.462)	.852***(.220)	.423(.469)	.947*(.553)	.615(.422)	.477(.473)	.344(1.018)	1.349***(.200)
Skin shade	–.024***(.006)	–.012(.008)	–.012(.022)	–.019(.015)	–.077***(.016)	–.010(.026)	.002(.016)	–.031***(.010)	–.020(.015)	–.007(.017)	–.035*(.019)	–.019(.016)	–.001(.042)	–.017**(.008)
R ²	.284	.240	.346	.525	.613	.753	.410	.317	.536	.613	.378	.534	.049	.263
n	1,799	825	327	425	133	66	198	988	408	130	408	405	68	754

Note: Numbers in parentheses are standard errors.

All models control for the variables listed under “Predictors for model 1” in Table 1, which are intended to capture factors that might be related to wages, such as physical characteristics, human capital, socioeconomic background, and region of residence within the United States.

Because of insufficient sample size, Asian respondents are excluded from the Latin American and Caribbean subsample.

p < .10, **p < .05, and ***p < .01 (one-tailed tests).

Table 3 contains the results for the second model specification, which adds race and Hispanic versus non-Hispanic to the baseline model when the analysis is conducted for the region of origin subsamples and Hispanic versus non-Hispanic and region of origin when the analysis is conducted for subsamples stratified by race. The first model shows that the skin shade penalty in wages remains significant for the full NIS sample after including race and ethnicity in the model. However, the next model shows that this result is driven primarily by the Latin American and Caribbean subsample; there is no evidence of a skin shade penalty in wages for any of the other region of birth subsamples examined.

Table 3.

Unstandardized coefficients from regressions of wages on skin shade and predictors for both models 1 and 2 listed in Table 1.

		Baseline Model + Additional Controls (Race/Ethnicity)^a					Baseline Model + Additional Controls (Hispanic and Region of Birth)^b						Baseline Model + Additional Controls (Hispanic)^c
		Region of Birth					Self-reported Race			Excluding Latin America and Caribbean			Including Only Latin America & Caribbean^d
Individual Variable	Total NIS Sample	Latin America and Caribbean	Europe and Central Asia	China and East/South Asia	Sub-Saharan Africa	Middle East and North Africa	Black	White	Asian	Black	White	Asian	Black	Nonblack
Constant	1.027***(.226)	1.357***(.255)	.717(.765)	–.093(.585)	1.479**(.603)	–.396(.788)	.922(.604)	1.554***(.247)	.221(.585)	1.944***(.590)	1.419**(.480)	.542(.657)	173(1.077)	1.400***(.244)
Skin shade	–.020**(.009)	–.014*(.008)	–.007(.023)	–.020(.015)	–.015(.018)	–.004(.028)	–.006(.017)	–.016*(.009)	–.020(.016)	–.011(.016)	–.029(.019)	–.019(.016)	–.012(.043)	–.019**(.009)
R ²	.297	.245	.350	.529	.705	.787	.439	.478	.538	.665	.410	.535	.511	.270
n	1,799	825	327	425	133	66	198	988	408	130	408	405	68	754

Note: Numbers in parentheses are standard errors.

All models control for the variables listed under “Predictors for model 1” and “Additional controls for model 2: race/ethnicity” in Table 1, which are intended to capture factors that might be related to wages, such as physical characteristics, human capital, socioeconomic background, and region of residence within the United States. The models for region of birth also control for race/ethnicity.

In addition to the controls included in the baseline model, the models for self-reported race and excluding Latin America and the Caribbean also control for Hispanic versus non-Hispanic and region of birth.

In addition to the controls included in the baseline model, the models for self-reported race including only Latin American and Caribbean region of birth also control for Hispanic versus non-Hispanic ethnicity.

Because of insufficient sample size, Asian respondents are excluded from the Latin American and Caribbean subsample.

p < .10, **p < .05, and ***p < .01 (one-tailed tests).

The next three models show that the skin shade penalty also remains significant for the white population. The inclusion of a control for Hispanic versus non-Hispanic ethnicity and region of birth reduces the magnitude of the penalty by about half (b = –.016). In contrast, the skin shade penalty for white respondents excluding those from Latin America and the Caribbean becomes insignificant. Finally, the wage penalty remains for the nonblack subsample within the Latin American and Caribbean region. The findings in Table 3 remain unchanged even after including current employment characteristics into the analysis, which we show in Table 4.

Table 4.

Unstandardized coefficients from regressions of wages on skin shade and predictors for models 1, 2, and 3 listed in Table 1.

		Baseline + (Race/Ethnicity and Employment Characteristics)^a					Baseline +(Hispanic, Region of Birth, and Employment Characteristics)^b						Baseline + (Hispanic and Employment Characteristics)^c
		Region of Birth					Self-reported Race			Excluding Latin America and Caribbean			Including Only Latin America and Caribbean^d
Individual Variable	Total NIS Sample	Latin America and Caribbean	Europe and CentralAsia	China and East/South Asia	Sub-Saharan Africa	Middle East and North Africa	Black	White	Asian	Black	White	Asian	Black	Nonblack
Constant	1.569***(.261)	1.658***(.259)	2.224**(.731)	0.708(.581)	0.802(.636)	–.328(.982)	1.400**(.651)	1.554***(.247)	0.726(.555)	.856**(.576)	2.103***(.466)	1.032*(.614)	1.07(1.539)	1.595***(.249)
Skin shade	–.017**(.008)	–.016**(.008)	.002(.021)	–.017(.014)	–.009(.017)	–.024(.030)	.004(.016)	–.016*(.009)	–.017(.014)	–.005(.015)	–.023(.018)	–.017(.014)	.021(.054)	–.020**(.008)
R ²	.430	.350	.534	.613	.786	.908	.543	.478	.644	.856	.553	.642	.620	.370
n	1,799	825	327	425	133	66	198	988	408	130	408	405	68	754

Note: Numbers in parentheses are standard errors.

In addition to the controls included in the baseline model, the models for self-reported race and including only observations from Latin America and the Caribbean also control for Hispanic versus non-Hispanic and employment characteristics.

Because of insufficient sample size, Asian respondents are excluded from the Latin American and Caribbean subsample.

p < .10, **p < .05, and ***p < .01 (one-tailed tests).

Summary and Discussion

The purpose of this study was to determine whether a skin shade penalty exists within the labor market among immigrants within the United States. Specifically, we examine whether darker skin tone is associated with lower wages among immigrants using the NIS. The analysis was conducted separately for immigrants from five regions. Several findings are worthy of note.

First, the results for the analysis on the full immigrant sample are similar to those of Hersch (2008). We find that the difference in wages between two immigrants at the opposite ends of the skin shade scale is as large as 24 % an immigrant with the darkest skin shade on the scale earns roughly 76 cents for every dollar earned by an immigrant with the lightest skin shade on the NIS scale. This is the case after accounting for human capital, socioeconomic background, and region of residence within the United States. This difference declines by only 17 percent after accounting for self-reported race and occupational characteristics.

Second, the findings of this study suggest that when immigrants are disaggregated by region of birth and race, skin shade penalties remain significant only for those who were born in Latin America and the Caribbean and, intriguingly, for white immigrants. We offer an answer for this last finding in the discussion that follows.

Thus, wage penalties observed among darker immigrants in the NIS are driven by the large percentage of immigrants from Latin America and the Caribbean. This is consistent with findings produced by Frank et al. (2010), who also concluded that dark-skinned Latinos faced skin shade discrimination with respect to earnings. Additionally, the negative skin shade coefficient is quite pronounced among Latin American and Caribbean immigrants who describe themselves as being white.

We find that the skin shade coefficients are not significant for black or Asian immigrants when the Latin American and Caribbean observations are excluded from the analysis and are significant for white immigrants only when Hispanic versus non-Hispanic ethnicity and region of birth are not taken into account. Thus, it seems likely that colorism is experienced predominantly by Latin American and Caribbean immigrants who do not report their race as black and that the effects of colorism are much less pronounced or potentially nonexistent among other racial and national-origin populations. Furthermore, we find a significant penalty for nonblack Latin American immigrants with the darkest skin shade, who earn roughly 20 percent less than their lightest counterparts.

This conclusion is supported by the finding that the skin shade penalty in wages observed for the white immigrant subsample is explained by Hispanic status in the second model specification. As noted above, self-reported race is problematic among Latin American survey respondents. Self-reported race may reflect a preference for a white identity among respondents, regardless of their phenotypic characteristics (Darity et al. 2005). The racial and skin shade heterogeneity of the Latin American population could cause dissonance between the reported race given by respondents from that region and how they are perceived racially by others, especially potential American employers. This conclusion is consistent with our finding that the skin shade penalty in wages becomes insignificant for the white subsample when Hispanic versus non-Hispanic status is added to the baseline specification; immigrants from this region could identify as white racially while having phenotypical features often associated with nonwhites.

This is illustrated in Figure 3, which shows the distribution of skin shade for white immigrants from both outside (top) and inside (bottom) the Latin American and Caribbean region. Taken together, these panels suggest that there is greater discordance between self-reported race and physical appearance among whites from this region than those from the other regions. The mean skin shade for white immigrants from regions other than Latin America and the Caribbean is 2.38, with a modal value of 3. In contrast, white immigrants from Latin America and the Caribbean have substantially darker skin, with a mean of 4.09 and a modal value of 5 on the NIS scale. These differences suggest that those reporting their race as white from Latin America and the Caribbean are significantly darker than those reporting their race as white from countries in other regions. Furthermore, 92 percent of the immigrants in the total sample who report their race as white while having skin shades coded at greater than 5 were from Latin America or the Caribbean. Given the patterns highlighted in Figure 3, it is plausible that there is discordance between Latin American and Caribbean immigrants’ patterns of racial self-classification and potential employers’ perceptions of their races.

Figure 3.

Skin shade for self-reported white respondents: (a) white respondents excluding Latin America and the Caribbean and (b) white respondents from Latin America and the Caribbean.

The only other group for which a skin shade penalty is observed is the Sub-Saharan African subsample. However, the estimated skin shade effect is explained by race; whites from this region earn significantly more ($12.99 per hour) than their nonwhite counterparts ($10.08 per hour). Therefore, although immigrants from Sub-Saharan Africa do not exhibit a penalty for skin shade, they do experience a race penalty. It is important to consider, however, the possibility that the band of skin tone variation among Sub-Saharan Africans is so slight that the inclusion of the race variable eliminates evidence of colorism.

To evaluate this possibility, we analyzed the distribution of skin shades in the black Sub-Saharan African population (see Figure 4). The mean skin shade is 7.66, with a modal value of 8 and a variance of 3.99. Furthermore, only 7 of 119 black Sub-Saharan African respondents were coded as having skin shades lighter than 5. It is possible that the small number of observations with lighter skin shades in the black Sub-Saharan African population contributed to the insignificant skin shade effect when race is controlled.

Figure 4.

Skin shade distribution among black Sub-Saharan African immigrants.

Finally, we do not find a connection between skin shade and wages for immigrants from (1) Europe and Central Asia; (2) China, East Asia, South Asia, and the Pacific; and (3) the Middle East and North Africa.

Conclusion

In sum, it appears that the finding of a skin shade penalty in wages for darker skinned immigrants found in studies using data from NIS (e.g., Hersch 2008) is driven by the large percentage of Latin American immigrants included in the original sample. The skin shade penalty is substantial for Latin American immigrants who report their race as nonblack or as white. The effects of colorism, however, are much less pronounced or nonexistent among other racial and national-origin populations. Furthermore, although a skin shade penalty might not exist among African immigrants, we do find evidence consistent with wage discrimination by race. Below we briefly discuss both the methodological implications for studies on discrimination against immigrants and the social implications of our results.

Methodological Implications

Our analysis demonstrates that using self-reported race in a statistical test for discrimination against immigrants can be problematic. It seems that immigrant populations from different regions self-report race differently from most native-born Americans, which can compromise the accuracy of results and potentially mask underlying effects within subpopulations. For example, Latinos who report their race as white often are dark complexioned, reinforcing the difficulty of relying on self-reported race as the basis for gauging the degree of discrimination they might encounter. Furthermore, the presence of a significant skin shade wage penalty in the NIS for analyses that do not disaggregate by region of origin likely will yield findings that suggest a skin shade penalty for all immigrants to the United States because of the heavy representation of respondents from Latin America and the Caribbean.

To take this into account, we suggest that studies of immigrants based on data that include self-reported race disaggregate by region to ensure that findings for regional subpopulations are consistent with findings for an entire sample of immigrants before conclusions are drawn. Although our findings have important implications for the study of race, they might also apply to other social variables that may suffer from inconsistency in data reports across different types of respondents. As such, disaggregating by region of origin also could be valuable in a variety of analyses.

Our finding that a skin shade penalty is not present among African immigrants presents an interesting question for future research. Given that a number of studies have observed a skin shade penalty for African Americans, the lack of significant results among African immigrants is perplexing and warrants further analysis. Indeed, a lack of sufficient skin shade variation in the African subsample could have contributed to a lack of significance of skin shade for this group in this study. Moreover, the relatively small sample size of immigrants from Africa might also contribute to the insignificant association between skin shade and earnings for this group and perhaps for some of the other region of birth subgroups. Thus, future studies should explore other methods to study the dynamics of skin shade discrimination for the African population, particularly as larger samples of African immigrants become available as the population grows.

Social Implications

The presence of a skin shade penalty for nonblack Latin American and Caribbean immigrants demonstrates that pervasive inequality of opportunity persists within the United States. This is a real and serious concern given that Latin Americans are the largest and fastest growing immigrant population in the United States (Ortman and Guaneri 2009). Although studies such as this compel readers to acknowledge the existence of these disparities, future research should focus on the reasons behind this unequal treatment to foster policy solutions.

There are a few ways to understand this issue. In particular, one must consider the physical and ethnic phenotypic characteristics that are more likely to be associated with light-skinned Latin Americans than with their dark-skinned counterparts. In particular, light-skinned Latin Americans may appear to have a greater number of European characteristics, allowing them to be more readily accepted as “Americans” by U.S. employers. Therefore, they may face fewer employment disadvantages associated with being an immigrant. Furthermore, several studies show that skin shade discrimination exists in Latin America (Alvarez 1993; Cruz-Janzen 2001). It is possible, then, that Latin American immigrants are more likely to be employed by other Latin American immigrants who may discriminate on the basis of biases imported from their home countries. However, this explanation might not explain the observed phenomenon entirely, because skin shade discrimination is also present in Asian countries but is not faced by Asian immigrants in the United States (Karan 2008; Perry 2005).

Although not examined in the present study, the lack of a statistically significant correlation between skin shade and earnings among Asian immigrants might stem from the relatively high proportion of Asian immigrants who are self-employed or who work for family-owned businesses (Portes and Yiu 2013), for whom the intensity of skin shade discrimination might be less severe. Future work in this area, therefore, should examine whether self-employment is correlated with darker skin tone, as darker skinned immigrants may gravitate toward self-employment or entrepreneurship to escape discrimination in the wage and salary sector of the economy.

The findings also suggest darker skin tone among nonblack immigrants from Latin America reduces the likelihood of achieving middle-class status, which increases the risk for downward assimilation among the children of these immigrants. Portes and Rumbaut (2014) suggested that children from working-class parents, particularly those who reside in weak coethnic communities, are at greater risk for experiencing downward assimilation—having low educational achievement and stagnating into subordinate menial jobs—than children of parents with high levels of human capital who have been able to achieve middle-class status in the first generation.

Moreover, although earnings are the only outcome examined in this study, the results have implications for stratification more generally. In addition to any direct effect of skin shade on measures such as health and occupational mobility (Williams et al. 1997), the findings suggest that skin shade may have an indirect effect on the health of nonblack immigrants from Latin America through an income effect whereby darker skinned immigrants are less able to purchase health-promoting goods and services or purchase homes in less stressful neighborhoods.

Similarly, the findings suggest that darker skin shade may negatively affect the types of employment prospects for immigrants. Portes and Rumbaut (2014) noted,

Employers as a whole may be indifferent toward a particular group, or they may have a positive or negative view of it. Positive or negative typification of a specific minority can take, in turn, different forms. For example, widespread discrimination may hold that certain groups are able only to preform low-wage menial labor (“Mexican work” or in an earlier time, “coolie labor”), or it can hold that they are simply too incompetent to be employable at all. In the first instance discrimination contributes to confinement of the group to the low-wage segment of the labor market; in the second it contributes to it exclusion and hence unemployment. (p. 141)

If employers positively correlate skin shade with negative typification, then darker skinned immigrants from Latin America and the Caribbean may have less favorable employment prospects. Although the models in Table 4 directly control for profession, we cannot rule out the possibility that darker skinned immigrants have varying opportunities for upward mobility within a given profession, which could have a negative impact on earnings.

Although we did not find convincing evidence that colorism affects all new immigrants to the United States, it is important to point out that this does not mean that skin shade is inconsequential. On the contrary, the findings from this (and other aforementioned studies) suggest that skin shade remains an extremely important determinant of life outcomes for a large proportion of both immigrants and nonimmigrants. For example, the lightest immigrants in this sample have an average hourly wage that is substantially greater than the average wage of the darkest immigrants in the sample. Figure 5 demonstrates a steady decrease in hourly wage as skin shade becomes darker among immigrants in the NIS.

Figure 5.

Total sample average wage by skin shade.

What is clear is that discrimination toward immigrants is prevalent in the U.S. labor market. Despite the notion that the United States generally considers itself to be the land of equal opportunity for all, it is evident that serious impediments to success exist for a large and growing portion of the immigrant population.

Footnotes

Notes

Author Biographies

Alexis Rosenblum earned a Bachelor of Arts in Public Policy Studies from Duke University in 2010.

William A. Darity Jr. is the Samuel DuBois Cook Professor of Public Policy, African and African American Studies, and Economics at Duke University. He currently serves as director of the Samuel DuBois Cook Center on Social Equity and previously served as chair of the Department of African and African American Studies, both at Duke. While at the University of North Carolina at Chapel Hill, he served as co-director of the Institute of African American Research, director of the Moore Undergraduate Research Apprenticeship Program and as director of the Undergraduate Honors Program and director of Graduate Studies in the Department of Economics. Darity’s research focuses on inequality by race, class and ethnicity, stratification economics, schooling and the racial achievement gap, North–South theories of trade and development, skin shade and labor market outcomes, the economics of reparations, the Atlantic slave trade and the Industrial Revolution, the history of economics, and the social psychological effects of exposure to unemployment. He has published more than 225 articles and written or edited 11 books. He was a fellow at the Center for the Advanced Study in the Behavioral Sciences at Stanford University during 2011–2012, and he will be a visiting scholar at the Russell Sage Foundation during 2015–2016.

Angel Harris is a Professor of Sociology and African and African American Studies at Duke University. He is also Director of the program for Research on Education and Development of Youth (REDY). His research interests include social inequality, policy, and education. His current work focuses on the social psychological determinants of the racial achievement gap.

Tod G. Hamilton is currently an Assistant Professor in the Department of Sociology and a Faculty Associate of the Office of Population Research at Princeton University. His specializations are in the areas of immigration, health, and race, with an emphasis on understanding patterns of stratification and health between immigrant and native populations in the United States. His recent research as been published in Demography, Social Science and Medicine, and Social Science Quarterly.

References

Alba

2005. “Bright vs. Blurred Boundaries: Second-generation Assimilation and Exclusion in France, Germany, and the United States.” Ethnic and Racial Studies 28(1):20–49.

Alvarez

Julia

. 1993. “Black Behind the Ears.” Essence 23(10):2.

Anderson

Deborah

Shapiro

David

. 1996. “Racial Differences in Access to High-paying Jobs and the Wage Gap between Black and White Women.” Industrial and Labor Relations Review 49(2):273–86.

Baldi

Stephane

McBrier

Debra Branch

. 1997. “Do the Determinants of Promotion Differ for Blacks and Whites?” Work and Occupations 24(4):478–97.

Bertrand

Marianne

Mullainathan

Sendhil

. 2004. “Are Emily and Brendan More Employable than Lakisha and Jamal? A Field Experiment on Labor Market Discrimination.” American Economic Review 94(4):991–1013.

Coleman

Major G.

2003. “Job Skill and Black Male Wage Discrimination.” Social Science Quarterly 84(4):892–905.

Cruz-Janzen

Marta I.

2001. “Latinegras: Desired Women—Undesirable Mothers, Daughters, Sisters, and Wives.” Frontiers 22(3):168–83.

Danziger

Sheldon

Gottschalk

Peter

. 1995. America Unequal. Cambridge, MA: Harvard University Press.

Darity

Jr. Jason

Guilkey

D. K.

2001. “Persistent Advantage or Disadvantage? Evidence in Support of the Intergenerational Drag Hypothesis.” American Journal of Economics and Sociology 60(2):435–70.

10.

Darity

William A.

Jr. Dietrich

Jason

Hamilton

Darrick

. 2005. “Bleach in the Rainbow: Latin Ethnicity and Preference for Whiteness.” Transforming Anthropology 13(1):103–109.

11.

Diette

T. M.

Goldsmith

A. H.

Hamilton

Darity

2015. “Skin Shade Stratification and the Psychological Cost of Unemployment: Is There a Gradient for Black Females?” Review of Black Political Economy 42(1–2):155–77.

12.

Dodoo

F.N.-A.

1991. “Earnings Differences among Blacks in America.” Social Science Research 20(2):93–108.

13.

Dodoo

F.N.-A.

Takyi

B. K.

2002. “Africans in the Diaspora: Black-white Earnings Differences among America’s Africans.” Ethnic and Racial Studies 25(6):913–41.

14.

Eberhardt

J. L.

Davies

P. G.

Purdie-Vaughns

V. J.

Johnson

S. L.

2006. “Looking Deathworthy: Perceived Stereotypicality of Black Defendants Predicts Capital-sentencing Outcomes.” Psychological Science 17(5):383–86.

15.

Elliott

James R.

Smith

Ryan A.

2004. “Race, Gender, and Workplace Power.” American Sociological Review 69(3):365–86.

16.

Frank

Reanne

Akresh

Ilana Redstone

. 2010. “Latino Immigrants and the U.S. Racial Order: How and Where Do They Fit In?” American Sociological Review 75(3):378–401.

17.

Gelatt

Julia

Dixon

David

. 2006. “Detailed Characteristics of the Caribbean Population in the United States.” Migration Policy Institute (http://www.migrationinformation.org/usfocus/display.cfm?ID=408).

18.

Golash-Boza

Tanya

Darity

William

Jr . 2008. “Latino Racial Choices: The Effects of Skin Colour and Discrimination on Latinos’ and Latinas’ Racial Self-identifications.” Ethnic and Racial Studies 28(1):1–36.

19.

Goldsmith

Arthur H.

Hamilton

Darrick

Darity

William A.

Jr . 2006. “Shades of Discrimination: Skin Tone and Wages.” American Economic Review 96(2):242–45.

20.

Goldsmith

Arthur H.

Hamilton

Darrick

Darity

William A.

Jr . 2007. “From Dark to Light: Skin Color and Wages among African-Americans.” Journal of Human Resources 42(4):701–38.

21.

Grieco

Elizabeth M.

2004. “The African Foreign-born Population of the US.” Migration Policy Institute (http://www.migrationinformation.org/feature/display.cfm?ID=250).

22.

Grieco

Elizabeth M.

Trevelyan

Edward N.

2010. “Place of Birth of the Foreign-born Population 2009: American Community Survey Briefs.” Washington, DC: U.S. Department of Commerce, Economics and Statistics Administration, U.S. Census Bureau.

23.

Gyimah-Brempong

Price

G. N.

2006. “Crime and Punishment: And Skin Hue Too?” American Economic Review 96(2):246–50.

24.

Hamilton

Austin

Darity

Jr.

2011. “Whiter Jobs, Higher Wages: Occupational Segregation and the Lower Wages of Black Men.” EPI Briefing Paper #288. Economic Policy Institute. Retrieved August 7, 2015 (http://s1.epi.org/files/page/-/BriefingPaper288.pdf).

25.

Hamilton

Darrick

Goldsmith

Arthur H.

Darity

William A.

Jr.

2008. “Measuring the Wage Costs of Limited English: Issues with Using Interviewer Versus Self-reports in Determining Latino Wages.” Hispanic Journal of Behavioral Sciences 30(3):257–79.

26.

Harris

Angel L.

2011. Kids Don’t Want to Fail: Oppositional Culture and the Black-white Achievement Gap. Cambridge, MA: Harvard University Press.

27.

Hattam

Victoria

. 2005. “Ethnicity & the Boundaries of Race: Rereading Directive 15.” Daedalus 134(1):61–70.

28.

Hersch

2002. “Skin Color, Immigrant Wages, and Discrimination.” Pp. 77–90 in Racism in the 21st Century: An Empirical Analysis of Skin Color, edited by Hall

Ronald E.

New York: Springer.

29.

Hersch

2008. “Profiling the New Immigrant Worker: The Effects of Skin Color and Height.” Journal of Labor Economics 26(2):345–86.

30.

Hersch

2011. “The Persistence of Skin Color Discrimination for Immigrants.” Social Science Research 40(5):1337–49.

31.

Hill

Mark E.

2002. “Skin Color and Intelligence in African Americans: A Reanalysis of Lynn’s Data.” Population and Environment 24(2):209–14.

32.

Hitlin

Scott

Brown

J. S.

Elder

Glen H.

2007. “Measuring Latinos: Racial vs. Ethnic Classification and Self-understandings.” Social Forces 86(2):587–611.

33.

Hochschild

Jennifer L.

Weaver

Vesla

. 2007. “The Skin Color Paradox and the American Racial Order.” Social Forces 86(2):643–70.

34.

Johnson

William R.

Neal

Derek

. 1998. “Basic Skills and the Black-white Earnings Gap.” Pp. 480–97 in The Black-white Test Score Gap, edited by Jencks

Phillips

Washington, DC: Brookings Institution.

35.

Karan

Kavita

. 2008. “Obsessions with Fair Skin: Color Discourses in Indian Advertising.” Advertising & Society Review 9(2).

36.

Keith

Verna M.

Herring

Cedric

. 1991. “Skin Tone and Stratification in the Black Community.” American Journal of Sociology 27(3):760–78.

37.

Kent

Mary Mederios

. 2007. “Immigration and America’s Black Population.” Population Bulletin 62:1–20.

38.

Lopez

Ian Haney

. 2005. “Race on the 2010 Census: Hispanics & the Shrinking White Majority.” Daedalus 134(1):42–53.

39.

Massey

Douglas S.

Martin

Jennifer A.

2003. “NIS Skin Color Scale.” Retrieved August 7, 2015 (http://nis.princeton.edu/downloads/NIS-Skin-Color-Scale.pdf).

40.

McBride Murry

Velma

Liu

. (2014). “CCF Civil Rights Symposium: Are African Americans Living the Dream 50 Years after Passage of the Civil Rights Act?” Retrieved August 7, 2015 (https://contemporaryfamilies.org/are-african-americans-living-the-dream/).

41.

Monk

E. P.

2014. “Skin Tone Stratification among Black Americans, 2001–2003.” Social Forces 92(4):1313–37.

42.

Morris

Martina

Western

Bruce

. 1999. “Inequality in Earnings at the Close of the Twentieth Century.” Annual Review of Sociology 25:623–57.

43.

Murguia

Edward

Telles

Edward

. 1996. “Phenotype and Schooling among Mexican Americans.” Sociology of Education 69(4):276–89.

44.

Neal

Derek A.

Johnson

William R.

1996. “The Role of Premarket Factors in Black-white Wage Differences.” Journal of Political Economy 104(5):869–95.

45.

Oettinger

Gerald S.

1996. “Statistical Discrimination and the Early Career Evolution of the Black-white Wage Gap.” Journal of Labor Economics 14(1):52–78.

46.

Ogden

Cynthia L.

Fryar

Cheryl D.

Carrolls

Margaret D.

Flegal

Katherine M.

2004. “Mean Body Weight, Height, and Body Mass Index, United States 1960–2002.” Washington, DC: National Center for Health Statistics, U.S. Department of Health and Human Services.

47.

O’Neill

June

. 1990. “The Role of Human Capital in Earnings Differences between Black and White Men.” Journal of Economic Perspectives 4(1):25–45.

48.

Ortman

Jennifer M.

Guaneri

Christine E.

2009. “United States Population Projections 2000 to 2050.” 2009 National Population Projections (Supplemental), U.S. Census Bureau (http://www.census.gov/population).

49.

Pager

Devah

. 2007. Marked: Race, Crime, and Finding Work in an Era of Mass Incarceration. Chicago: University of Chicago Press.

50.

Pager

Western

Bonikowski

2009. “Discrimination in a Low-wage Labor Market: A Field Experiment.” American Sociological Review 74(5):777–99.

51.

Perry

Alex

. 2005. “Could You Please Make Me a Shade Lighter?” Time South Pacific No. 48.

52.

Portes

Alejandro

Rumbaut

Ruben G.

2014. Immigrant America: A Portrait. Berkeley: University of California Press.

53.

Portes

Alejandro

Yiu

Jessia

. 2013. “Entrepreneurship, Transnationalism, and Development.” Migration Studies 1(1):75–95.

54.

Portes

Alejandro

Zhou

Min

. 1993. “A New Second Generation: Segmented Assimilation and Its Variants.” Annals of the American Academy of Political and Social Science 530(1):74–96.

55.

Prewitt

Kenneth

. 2005. “Racial Classification in America: Where Do We Go from Here?” Daedalus 134(1):1–17.

56.

Seltzer

Richard

Smith

Robert C.

1991. “Color Differences in the Afro-American Community and the Differences They Make.” Journal of Black Studies 21(3):279–86.

57.

Seltzer

Richard

Smith

Robert C.

2005. “Do the Determinants of Promotion Differ for White Men Versus Women and Minorities?: An Exploration of Intersectionalism Through Sponsored and Contest Mobility Processes.” American Behavioral Scientist 48:1157–81.

58.

Smith

Ryan A

. 1997. “Race, Job Authority, and Income: A Cross-Temporal Study of Changes in the Socioeconomic Status of Black and White Men, 1972–1994.” Social Problems 44:701–19.

59.

Smith

Ryan A

. 2001. “Particularism in Control Over Monetary Resources at Work: An Analysis of Racioethnic Differences in the Authority Outcomes of Black, White, and Latino Men.” Work and Occupations 28(4):447–68.

60.

Smith

Ryan A

. 2005. “Do the Determinants of Promotion Differ for White Men Versus Women and Minorities? An Exploration of Intersectionalism Through Sponsored and Contest Mobility Processes.” American Behavioral Scientist 48(9):1157–181.

61.

Tafoya

Sonya M.

2005. “Shades of Belonging: Latinos and Racial Identity.” Harvard Journal of Hispanic Policy 17(1):58–78.

62.

Telles

E. E.

Murguia

1988. “Phenotypic Discrimination and Income Differences among Mexican Americans.” MASRC Working Paper #13. Tucson: University of Arizona.

63.

Telles

E. E.

Ortiz

2008. Generations of Exclusion. New York: Russell Sage.

64.

Tomaskovic-Devey

Donald

Thomas

Melvin

Johnson

Kecia

. 2005. “Race and the Accumulation of Human Capital across the Career: A Theoretical Model and Fixed-effects Application.” American Journal of Sociology 111(1):58–89.

65.

U.S. Census Bureau. 2006. “Foreign-born Population—Selected Characteristics by Region of Origin” (http://www.census.gov/population/foreign/).

66.

U.S. Census Bureau. 2008. “Table 4. Projections of the Population by Sex, Race, and Hispanic Origin for the United States: 2010 to 2050 (NP2008-T4).” Washington, DC: U.S. Department of Commerce.

67.

Williams

D. R.

Jackson

J. S.

Anderson

N. B.

1997. “Racial Differences in Physical and Mental Health Socio-economic Status, Stress and Discrimination.” Journal of Health Psychology 2(3):335–51.

68.

Wilson

George

McBrier

Debra Branch

. 2005. “Race and Loss of Privilege: African American/White Differences in the Determinants of Job Layoffs from Upper-tier Occupations.” Sociological Forum 20(2):301–21.

69.

Wilson

George

Roscigno

Vincent J.

2010. “Race and Downward Mobility from Privileged Occupations: African American/White Dynamics across the Early Work-career.” Social Science Research 39(1):67–77.

70.

Wilson

George

Sakura-Lemessy

Ian

West

Jonathan P.

1999. “Reaching the Top: Racial Differences in Mobility Paths to Upper-tier Occupations.” Work and Occupations 26(2):165–86.

71.

Xie

Goyette

Kimberly

. 2004. The American People: A Demographic Portrait of Asian Americans. New York: Russell Sage.

72.

Zhao

Ondrich

Jan

Yinger

John

. 2006. “Why Do Real Estate Brokers Continue to Discriminate? Evidence from the 2000 Housing Discrimination Study.” Journal of Urban Economics 59(3):394–419.

73.

Zhou

Min

. 1997. “Segmented Assimilation: Issues, Controversies, and Recent Research on the New Second Generation.” International Migration Review 31(4):1975–2008.