Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2023 Mar 1.
Published in final edited form as: Health Place. 2022 Jan 25;74:102742. doi: 10.1016/j.healthplace.2022.102742

WHERE ARE THE LABOR MARKETS?: EXAMINING THE ASSOCIATION BETWEEN STRUCTURAL RACISM IN LABOR MARKETS AND INFANT BIRTH WEIGHT

Tongtan Chantarat a,b,c, Kari M Mentzer a,c, David C Van Riper c, Rachel R Hardeman a,b,c
PMCID: PMC8923951  NIHMSID: NIHMS1774780  PMID: 35091167

Abstract

Racist policies and practices that restrict Black, as compared to white workers, from employment may drive racial inequities in birth outcomes among workers. This study examined the association between structural racism in labor markets, measured at a commuting zone where workers live and commute to work, and low-birthweight birth. We found the deleterious effect of structural racism in labor markets among US-born Southern Black pregnant people of working age, but not among African- or Caribbean-born counterparts in any US region. Our analysis highlights the intersections of structural racism, culture, migration, and history of racial oppression that vary across regions and birth outcomes of Black workers.

Keywords: structural racism, birthweight, health equity, workers, labor markets

INTRODUCTION

Infants born with a birthweight of under 2,500 grams (low birthweight; LBW) face increased risks of adverse health outcomes, such as jaundice, necrotizing enterocolitis, breastfeeding problems, and death before their first birthday (Martin et al., 2017). LBW infants are also more vulnerable to having chronic diseases later in life, incurring excess healthcare expenses and creating economic burdens to their family and the society (Almond, Chay and Lee, 2005). For centuries, an inequity in birthweight between infants born to Black versus white pregnant people in the United States (US) has persisted despite improvements in access to prenatal care and initiatives aimed at reducing the exposure to risk factors of poor pregnancy (e.g., tobacco and substance use, short time between pregnancies) (Martin et al., 2017). LBW inequity also exists among infants born to Black pregnant people of different nativities. For example, in the Minneapolis-Saint Paul area, home to one of the largest Somali communities outside Somalia (United Nations Development Program, 2009), the incidence of LBW was 12.2% and 5.9% for infants born to US-born and African-born Black pregnant people, respectively (Minnesota Compass, 2020). A similar trend is observed in New York City, the largest settlement in the US for Black people from the Caribbean Islands (Hamilton, 2019); the incidence of LBW was 11.9% and 9.9% for infants born to US-born and Caribbean-born Black pregnant people, respectively (Mason et al., 2010).

Black people from all socioeconomic backgrounds weather the exposure to various stressors due to the lived experience of racism that elicits the same response as chronic stress (Geronimus, 1992; Berger and Sarnyai, 2015). Over time, the accumulation of physiological wear and tear (weathering) caused by racism and the embodiment of oppression renders Black people more susceptible to poor health (Geronimus et al., 2006; Hicken et al., 2013; Doamekpor and Dinwiddie, 2015). In a society like the US, where there are over 400 years of history of structuring advantage based on race, genocide, and colonialism, structural racism – the interconnected system of ideologies, policies, and practices in our social and economic institutions that produce racialized outcomes, even in the absence of racist intent (Bailey et al., 2017) – has been linked to several poor pregnancy outcomes, including stress during pregnancy (Mendez, Hogan and Culhane, 2013), severe maternal morbidity among Black pregnant people (Liu et al., 2019), small-for-gestational-age infants (Wallace et al., 2015), and infant mortality (Wallace et al., 2017; Pabayo et al., 2019).

Structural racism, specifically anti-Black racism, influences the social and economic experiences of Black people across the life course. One major milestone in adulthood is entry into the labor force. For most workers, this process repeats itself over ten times throughout their career (Bureau of Labor Statistics, 2019). Employment provides access to economic and social resources (e.g., wages, healthcare coverage, professional networks) and a sense of independence and accomplishment – all of which are keys for maintaining good physical and mental health (Pavalka and Smith, 1999). A system of racist policies and practices that exclude Black, as compared to white workers, from employment likely contributes to racial health inequities in the workforce (Krieger et al., 2008; McCluney et al., 2018; McClure et al., 2020). We characterize this form of racial injustice as structural racism in labor markets (SRL). The scholarship on SRL has been the focus of economists and sociologists for decades (Reskin, 1998; Hamilton, Austin and Darity, Jr, 2011; Cajner et al., 2017; Quillian et al., 2017; Wingfield, 2019; McClure et al., 2020). SRL has been exposed through a variety of methodological approaches, including field experiments, audit studies, and the decomposition of gaps (Pager and Shepherd, 2008). For example, in a meta-analysis of field experiments of hiring discrimination since 1989, Quillian and colleagues (2017) found that job seekers with a “white sounding” name received a job callback at about a 36% higher rate than those with a “Black sounding” name with similar credentials (Quillian et al., 2017). Cajner and colleagues (2017) documented employment trends since the Great Recession and found that Black workers are more likely than white workers to be employed in part-time jobs, even after the US economy has recovered from the Great Recession (Cajner et al., 2017). Black workers are more likely than their white counterparts to work in low-wage, low-skill occupations that are associated with job insecurity, the significant likelihood of job loss, and low job control (Hamilton, Austin and Darity, Jr, 2011; Wingfield, 2019; McClure et al., 2020).

Despite our extensive knowledge of SRL and its impact on workers’ economic conditions, few scholars have operationalized this type of structural racism as a driver of poor health among Black workers, with varied results. Prior studies operationalized and measured SRL by evaluating downstream employment patterns for Black compared to white workers to capture the combined health effect of a system of racist labor market policies and practices. SRL is often operationalized to be area-based, grounded on the scholarship in residential segregation as a fundamental force behind sociopolitical inequity between Black and white people (Groos et al., 2018). For example, in the groundbreaking study by Lukachko and colleagues (2014), SRL was the state-level ratio of employment rates for Black to white workers (Lukachko, Hatzenbuehler and Keyes, 2014). An alternative geographic unit for measuring SRL is the county (Liu et al., 2019; Pabayo et al., 2019). Measuring SRL at different geographic scales likely explains inconsistent findings concerning the health effects of SRL. This problem is known among geographers and environmental scientists as modifiable area unit problems (MAUP) (Wong, 2004). By design, area-based measures implicitly assume that all residents (or workers in the case of SRL) living in the same geographic unit are exposed to the same level of risk. We can minimize the impact of the MAUP in structural racism and health research by selecting the areal unit that is consistent with the way structural racism operates (Riley, 2018). Because SRL directly affects the job search process, the most appropriate geographic area to measure SRL is labor markets where local economies operate. Measuring SRL at the labor market level will also reduce the uncertain geographic context problem, which generally arises when scholars fail to consider how areal exposure measures “exert contextual influence on the individuals being studied and the temporal uncertainty in the timing and duration in which individuals experience these contextual influences” (Kwan, 2012).

But where are the labor markets? Boundaries of labor markets are not rigid like those drawn for political purposes. As a result, using states or counties to represent the boundaries of labor markets can be problematic. Consider one of the largest labor markets in the US: the New York Tri-State area. This labor market consists of the southern part of New York State (Hudson Valley, Long Island, the five boroughs of New York City (NYC)), western New Jersey, and most of Connecticut, and is home to more than nine million jobs (New York State Department of Labor, 2020). If we measure SRL at the state level, there are three possible values of SRL (New York, New Jersey, and Connecticut). Additionally, each state’s SRL value, in theory, would also represent the extent to which racist policies and practices restrict the employment opportunities for all Black workers in the entire state (e.g., workers in Buffalo, NY, are implicitly assumed to experience the same level of SRL as those working in the borough of Manhattan). Yet, if we measure SRL at the county level, we risk misclassifying workers’ exposure to SRL. NYC residents who live in the outer boroughs (Queens, Bronx, Kings, and Richmond counties) may work in Manhattan (New York county), yet we would assume their SRL exposure based on their county of residence. This example highlights one crucial consideration missing from the current measurement of SRL: workers tend to search for jobs across state or county borders. The measure of SRL must account for workers’ travel patterns between home and work.

Economists have proposed several areal units that best represent labor markets (Fowler and Jensen, 2020). One of the units that accounts for the workers’ travel patterns is the commuting zone (CZ). First developed by the US Department of Agriculture in 1980, CZ is considered the most accurate delineation of local economies where people live and commute to work (US Department of Agriculture Economic Research Service, 2020). CZ is a product of hierarchical cluster analysis based on the Census’ journey-to-work data. Counties are grouped based on the commuting patterns of their residents. Each CZ usually has a metropolitan core and consists of counties with a high correlation of average wage rates (Fowler, Rhubart and Jensen, 2016). Based on the 1990 Census, there are 741 CZs compared to 50 states and the Districts of Columbia and 3,143 counties in the US. Since SRL likely operates at the labor market level where workers search for employment, not the states or counties where they live, a CZ-based measure of SRL is more appropriate than other area-based measures.

In this study, we improve our understanding of the relationship between SRL and LBW by using a novel measure of SRL. We developed a CZ-based measure of SRL with data from the American Community Survey and merged it with birth data of infants born to working-age Black pregnant people in all 50 states and the District of Columbia. Using the US birth data rather than data from specific cities, states, or geographic regions improves the generalizability of our findings. We theorized that Black workers who endure the effect of structural racism when searching for a job in a racist labor market and/or working in a racially structured organization experience greater wear and tear (weathering) to their body than those in the low-exposure group. Consequently, those in the high-exposure group are more likely to experience poorer pregnancy outcomes (e.g., LBW) than those in the low-exposure group. We stratified our sample into three nativity groups: US-born Black, African-born Black, and Caribbean-born Black pregnant people, to capture the potential effect heterogeneity by differences in length of time in the US, the reasons for migration, and general health status. Although our main goal was to examine the average effect of SRL on the risk of LBW among US residents overall, given the geographic reach of our data, we also assessed the potential effect variation that may stem from different historical sociopolitical characteristics associated with different US regions, and/or other unaccounted residuals.

METHODS

Data

The birth outcomes data for our analysis were from the restricted-use National Vital Statistics System (NVSS), housed at the National Center for Health Statistics. The NVSS data consolidates birth certificates submitted to the local health departments in all US states and territories. The NVSS birth data include details of the sociodemographic characteristics of the parents, where the parents live, extensive health and perinatal health care use, health behaviors, and birth outcomes. The restricted-use version also provides the county of residence of the parents. This information is the key we used to merge our measure of SRL to the LBW status of infants born to Black pregnant people.

Our analytical sample included singletons born to working age (i.e., 25 to 64 years; 25 is the age where average US workers have completed their terminal educational degree (US Bureau of Census, 2020b)) Black pregnant people (both Hispanic and non-Hispanic) between January 1 and December 31, 2017. Because the employment status of the pregnant people is not routinely collected on the birth certificate or reported to the NVSS, we were unable to ascertain whether the pregnant people in our sample were in the labor force or actively searching for a job during pregnancy. We excluded infants born to the pregnant people who were not born in the US, Africa, or the Caribbean Islands, as well as infants of the pregnant people whose country of birth was missing. A complete list of counties we classified as Africa and the Caribbean Islands is available in the Appendix.

We classified infants as LBW if their birthweight is less than 2,500 grams. Other covariates we measured were the pregnant person’s age/career stage (25–34/early career; 35–44/mid-career; 45 and older/late career), marital status (not married, married, not reported), and educational attainment (less than high school; high school diploma; some college; bachelor’s degree; advanced degree). Despite the availability of information about health conditions (e.g., diabetes, hypertension, and viral infection), risk behaviors (e.g., smoking, alcohol consumption), and healthcare access and utilization (e.g., insurance coverage, prenatal care) in the NVSS birth data, we purposely did not measure them in our study. Because SRL influences these determinants of infant birthweight (Gee and Ford, 2011), these determinants are considered mediators in the relationship between SRL and LBW (Martin et al., 2017). Controlling for these mediators in our analysis would have biased the effect of interest towards the null and may have led to an inaccurate conclusion (Schisterman, Cole and Platt, 2009).

Measuring Structural Racism in Labor Markets

We developed a CZ-based measure of SRL using the data from the American Community Survey (ACS). The ACS is conducted annually by the US Census Bureau (“the Census”), allowing scholars to examine the effect of racist policies and practices in labor markets in real time. One thing to note about the ACS is the data suppression policies created by the Census to protect the confidentiality of participants sampled from areas with less than 65,000 residents (US Bureau of Census, 2016). From the data users’ point of view, this policy creates systematic missingness, particularly for non-metropolitan areas. Since CZ population exceeds the data suppression threshold, missing data for our analysis was minimal. Our study used a time-lag design (Gollob and Reichardt, 1987), that is, we measured SRL at time t to predict LBW at time t+1. This approach allows us to infer temporal causality between SRL and LBW better than if SRL and LBW are from the same time period. To create our SRL measure, we first extracted the 2016 ACS data from the Integrated Public Use Microdata Series (IPUMS) USA database (Ruggles et al., 2020) for all Public Use Microdata Units (PUMAs) in the US. Because CZs and PUMAs do not align perfectly, we applied Autor and Dorn’s (2013) weighting scheme to assign sampled workers from PUMAs to CZs. Details regarding the development of the PUMA-to-CZ weighting scheme and its application are described elsewhere (Autor and Dorn, 2013).

Our SRL measure was based on predicted probabilities of being employed for Black and Non-Hispanic (NH) white workers aged 25–64 years after controlling for worker’s gender, career stage by age (25–34; 35–44; and 45 and older), educational attainment (less than high school; high school diploma; some college; bachelor degree; advanced degree), occupational groups (IPUMS USA, 2020b), and industry of work (IPUMS USA, 2020a). For each CZ, the SRL measure was calculated as:

SRL=mean(Pr(employmentforNHWhites∣gender,careerstage,education,occuaption,industry))mean(Pr(employmentforBlacks∣gender,careerstage,education,occuaption,industry))

In the equation above, the conditional probabilities were estimated by the OLS regression, with being employed and unemployed coded as 1 and 0, respectively. The SRL for 2016 ranged from 0.976 to 2.125 and was normally distributed. In some CZs, there were no sampled Black workers in the labor force aged 25–64 years; hence, SRL was missing for those CZs. This did not mean that there were no Black workers in those CZs in 2016, only that none of them were sampled by the Census to participate in the ACS. Figure 1 shows the levels of SRL, categorized by quartiles, in the US in 2016. The grey color on this map indicates the CZs where SRL could not be estimated (106 out of 741 CZs; 14.3%). Overall, we observed high levels of SRL in the lower Midwest, the upper and central Southern US, and large urban areas, consistent with the trends described in previous studies (Massey and Denton, 1993; Iceland and Sharp, 2013; Bell and Owens-Young, 2020). In our final step, we used the CZ-to-county crosswalk from Autor and Dorn (2013) to assign the CZ-based SRL to counties where the pregnant people in our sample lived.

Figure 1:

Figure 1:

Open in a new tab

Geographic distribution of structural racism in the US labor markets, 2016

Statistical Analysis

We conducted all analyses separately for infants of US-born Black (n=289,537), African-born Black (n=47,969), and Caribbean-born Black pregnant people (n=28,513) and only included those with non-missing correlates. The proportions of those excluded due to missingness were similar for the three groups (1.6% US-born; 2.5% African-born; 1.7% Caribbean-born).

First, we conducted a descriptive analysis using means and proportions to examine the distribution of the key sociodemographic characteristics, SRL levels, and the incidence of LBW across the three nativity groups. SRL was treated as a binary variable: pregnant people who lived in the CZs where SRL was in the fourth quartile (1.070 to 2.125) were considered high SRL exposure compared to those in the first to third quartiles (0.976 to 1.069). For sensitivity, we also classified pregnant people in the CZs where SRL was greater than the 90th percentile (1.100) to the very high SRL exposure group and those in the CZs in the lower deciles to the low exposure group.

Second, we examined the association between SRL and LBW with logistic regression and calculated odds ratios (OR) and confidence intervals (CI). We used the Huber-White method to calculate standard errors (SE) to account for the clustering effect among pregnant people who lived in the same CZ (Huber, 1967; White, 1980). We examined the association between SRL and LBW with bivariate regression as well as fitted multivariate regression to control for sociodemographic factors that may confound the relationship between SRL and LBW.

Lastly, we examined whether the effect of SRL on LBW exists across the US or is observable only among some geographic areas of the country. To do so, we further stratified our analytical sample into four Census regions (Northeast, Midwest, South, and West) (US Bureau of Census, 2020a), then fitted multivariate regression with these data. In all analyses, we set the alpha level to 0.05. This study protocol was reviewed and determined to be non-human research by the University of Minnesota Institutional Review Board. All data management and analyses were conducted in R version 4.0.2.

RESULTS

Table 1 shows the characteristics of Black pregnant people in our study, their exposure to SRL, and birthweight of their infants. Over 99% of the pregnant people in our study gave birth by age 44. A larger fraction of the US-born Black pregnant people gave birth during their early-career age and were unmarried at the time of childbirth compared to African-born and Caribbean-born pregnant people. African-born pregnant people are more likely to have a bachelor’s or advanced degree than the other two (US-born and Caribbean-born) groups, although most pregnant people in our sample have received a high school diploma. Approximately 13.2% of the US-born Black pregnant people lived in the CZs with high SRL (i.e., fourth quartile), compared to 7.0% of African-born and 9.1% of Caribbean-born Black pregnant people. The proportions of those who lived in the CZs with very high SRL (i.e., higher than 90th percentile) were small for all nativity groups. Table 1 also highlights the inequity in birthweight of infants by the Black pregnant people’s nativity. US-born Black pregnant people had the highest incidence of LBW (12.2%), followed by Caribbean-born Black pregnant people (9%) and African-born Black pregnant people (6.2%).

Table 1:

Sample characteristics

US
(n=289,537)		 African
(n=47,969)		 Caribbean
(n=28,513)
Age
 Early career: 25–34 years 81.4% 69.0% 64.7%
 Mid career: 35–44 years 18.4% 30.1% 34.7%
 Late career: 45 years and older 0.2% 0.9% 0.5%
Marital Status
 Not married 63.9% 22.6% 38.4%
 Married 31.2% 73.4% 60.5%
 Not reported 4.9% 4.1% 1.1%
Educational Attainment
 Less than high school 9.5% 15.9% 11.5%
 High school diploma 29.9% 21.4% 29.9%
 Some college 39.1% 23.9% 33.2%
 Bachelor’s degree 13.7% 26.2% 17.3%
 Advance degree 7.8% 12.5% 8.1%
US Census Region
 Midwest 18.6% 24.5% 2.5%
 Northeast 13.8% 21.0% 46.6%
 South 59.5% 41.9% 48.8%
 West 8.1% 12.6% 2.2%
Birth Outcome
 Low birth weight 12.2% 6.2% 9.0%
Structural Racism in Labor Markets
 Forth quartile 13.2% 7.0% 9.1%
 Greater than the 90th percentile 1.3% 0.9% 0.1%
Open in a new tab

Table 2 shows the association between high exposure to SRL and LBW. High SRL was significantly associated with LBW for US-born Black pregnant people after controlling for sociodemographic characteristics of the pregnant person (OR=1.051, CI: 1.018–1.086). SRL appears to increase the risk of LBW among Caribbean-born Black pregnant people, but this association did not reach the significance level of 0.05 (OR=1.069, CI: 0.930–1.228). Interestingly, we observed a slight reduction in the risk of LBW for the African-born Black pregnant people living in the CZs with high SRL in the multivariate model; however, this association was not statistically significant (OR=0.977, CI: 0.842–1.134).

Table 2:

Association between structural racism in labor markets and low-birthweight birth; Forth quartile vs. first to third quartiles

Pregnant Person’s Nativity Bivariate Model Multivariate Model
US
(n=289,537)		 1.091 (1.057, 1.127) 1.051 (1.018, 1.086)
African
(n=47,969)		 0.954 (0.822, 1.107) 0.977 (0.842, 1.134)
Caribbean
(n=28,513)		 1.045 (0.910, 1.201) 1.069 (0.930, 1.228)
Open in a new tab

Note: Multivariate models control for age, marital status, and educational attainment.

Results from our multivariate models were slightly sensitive to the threshold we used to determine the level of SRL. Table 3 shows the associations between SRL and LBW when SRL was classified as very high (i.e., greater than the 90th percentile) or low. We observed slightly larger effect size for US-born Black pregnant people but the confidence interval in this case covers the null value (OR=1.086, CI: 0.989–1.194). The fact that the number of pregnant people living in the CZs with very high SRL is extremely small may explain the widening of the confidence interval and the effect disparity by the threshold used to classify the level of SRL.

Table 3:

Association between structural racism in labor markets and low-birthweight birth; Greater than the 90th percentile vs. 90th percentile or less

Pregnant Person’s Nativity Bivariate Model Multivariate Model*
US
(n=289,537)		 1.138 (1.036, 1.250) 1.086 (0.989, 1.194)
African
(n=47,969)		 0.908 (0.605, 1.361) 0.947 (0.630, 1.424)
Caribbean
(n=28,513)		 1.795 (0.753, 4.279) 1.787 (0.743, 4.297)
Open in a new tab

Note: Multivariate models control for age, marital status, and educational attainment.

Figure 2 shows the effect of SRL on LBW from the multivariate models for US-born (purple), African-born (blue), and Caribbean-born Black pregnant people (green) in each Census region. SRL predicts LBW only for US-born Black pregnant people who lived in the South. Because of the small size of the stratified sample and positivity assumption violations, we were not able to pinpoint whether this effect is consistent for the West South Central states (Arkansas, Louisiana, Oklahoma, and Texas), the East South Central states (Alabama, Kentucky, Mississippi, and Tennessee), and the South Atlantic states (District of Columbia, Delaware, Florida, Georgia, Maryland, North Carolina, South Carolina, Virginia, and West Virginia). In addition, we did not observe a significant association between SRL and LBW for Black pregnant people of any nativity groups in other regions of the US.

Figure 2:

Figure 2:

Open in a new tab

Regional variation in the effect of structural racism in labor markets on the risk of low-birthweight birth.

Note: Purple, blue, and green dots and lines represent the odd ratios (ORs) and confidence intervals (CIs) for the association between SRL and LBW for US-born, African-born, and Caribbean-born Black pregnant people, respectively. ORs and CIs in this figure are from the multivariate models controlling for age, marital status, and educational attainment.

DISCUSSION

Within the past decade, the scholarship on the role of structural racism as a key driver of social determinants of health and as a fundamental cause of racial inequities in birth outcomes have increased (Wallace et al., 2015, 2017; Bailey et al., 2017; Chambers et al., 2018; Liu et al., 2019; Pabayo et al., 2019; Hardeman et al., 2021). Our study contributes to this growing body of research by offering a solution to the modifiable area unit problems in structural racism research. We demonstrated the use of the geographic unit that relates to the underlying mechanism of the domain of structural racism for measurement, minimizing the uncertain geographic context problem. Because SRL mainly interferes with the job search process, we proposed a novel measure that evaluates SRL at the commuting zone where workers search for jobs, not simply the county or state where they live. By linking our novel measure to the national birth data of infants born to US-, African-, and Caribbean-born, working-age Black pregnant people in all regions of the US, we were able to estimate the nativity- and region-specific associations between SRL and the risk of LBW and further highlight the intersections of structural racism, culture, migration status, and history of racial oppression in different parts of the country.

First, we found a significant association between SRL and LBW for US-born Black pregnant people. The magnitude of the association we observed is in line with other birth outcome studies that measured SRL at the state level (Wallace et al., 2015, 2017). Putting aside differences in data sources, analytical model specifications, and birth outcomes studied, residual effect size variation is likely explained by the geographic unit used to define labor markets. While some racist policies that exclude Black workers from employment are implemented at the state level (e.g., credentialing) and assigning SRL for each state make theoretical sense, scholars may still risk misclassifying SRL for workers in multi-state labor markets. The magnitude of the misclassification bias likely depends on the nature of the analytical samples. For studies that rely on data from a single state or a cluster of states in which interstate travel for work is minimal (i.e., a majority of structural racism studies), such misclassification should be relatively small, and the associations between SRL and birth outcomes observed should not be significantly affected by misclassification bias. However, the use of local data limits the generalizability of these findings. Furthermore, these studies usually do not have enough power for subgroup analyses due to the small sample size. Our CZ-based measure allows workers who live and work in adjacent states that are a part of the same commuting zone to have the same SRL. This measure is useful for future studies that may draw their analytical sample from larger geographic areas or nationally representative datasets (e.g., NVSS birth and death certificate data, National Survey of Family Growth, medical record data from multi-state health systems, etc.).

Second, the magnitude of the association between SRL and LBW for US-born Black pregnant people in our study is smaller than reported in the studies that use a county-based SRL measure (Liu et al., 2019; Pabayo et al., 2019). This variation is likely due to the reduced variance of the CZ-based SRL measure. Our CZ-based measure assigned the same level of SRL to a larger pocket of workers than a county-based measure (e.g., now that everyone in the NYC area is exposed to one level rather than five levels) and, thus, decrease in the variance of SRL in our sample. Regardless, we still found a conservative association between SRL and LBW for US-born Black pregnant people in our study, suggesting the deleterious influence of SRL on workers’ health. Another possible reason behind our conservative effect size relative to previous studies may stem from the way we operationalized SRL. We measured SRL as the ratio of predicted probabilities of gaining employment for white to Black workers, controlling for white-Black educational inequity and occupational segregation, rather than the Black-to-white unemployment ratio used in previous studies. By controlling for other forms of racial inequity, our measure likely netted out the joint effect of other forms of structural racism on LBW. Because of the pervasiveness of structural racism, racial inequity in several social and economic domains reinforce one another to cause poor health among Black people (Bailey et al., 2017). Future research should test whether measuring SRL with the Black-to-white unemployment ratio at a CZ level reveals a large effect size.

Third, we found that the association between SRL and LBW for US-born Black pregnant people differs by region. Specifically, the association is significant in the stratified sample of US-born Black pregnant people living in the South only. It is important to note that 59.5% of US-born Black pregnant people who gave birth in 2017 lived in the South, compared to 18.6% in the Midwest, 13.8% in the Northeast, and 8.1% in the West, and the difference in sample size across strata may contribute to the regional discrepancy we found. To rule out the sample size effect, we conducted a post-hoc analysis where we fitted the multivariate regression using the fourth quartile as an SRL cutoff with the exact same number of US-born Black pregnant people from all regions (n=23,392; this was the sample size from the West strata, the smallest strata in our primary analysis). Based on 1,000 randomly sampled datasets, the association between SRL and LBW was significant 97 times for the South, 31 times for the Midwest, and none for the Northeast and West stratum (results not shown). This finding provides supporting evidence that regional variation in the association between SRL and LBW exists.

The weathering of Black bodies results from encounters with multiple forms of racism over the life course and across generations (Geronimus, 1992). The long and deep history of racism in the US can be traced to the South, where the vestiges of the enslavement of Black people persist today (Hannah-Jones, 2021). As a result, Southern Black people may be disproportionately exposed to centuries of racial oppression in employment, housing, education, and voting rights (Geronimus and Thompson, 2019; Williams, Lawrence and Davis, 2019). While Jim Crow laws physically segregated Southern Black and white people in the past, the covert colorblind laws that restrict Southern Black people’s freedom, civil rights, and social and economic mobilities persist (Bonilla-Silva, 2017). For example, the median household income of Black families in the “Deep South” states (i.e., Alabama, Georgia, Louisiana, Mississippi, and South Carolina) from 2015 to 2019 was $7,000 lower than the national median of $41,935 for Black families (Guzman, 2020). Excluding Louisiana, no Deep South states expanded Medicaid after the passage of the Affordable Care Act (Graves et al., 2020; Taylor, Liu and Howell, 2020). Taken together with the decreased access to hospital obstetric services (Hung et al., 2017), increased restrictions to routine reproductive health services (Stevenson et al., 2016), and zero state laws mandating paid family or sick leave for workers in the South (Henry J. Kaiser Family Foundation, 2020), Southern Black people may certainly be more vulnerable to SRL than their counterparts in other regions. Our regional variation findings should not be interpreted as SRL is the issue only in the South nor that it is bad only for Southern Black people. It is rather a call for future analyses using non-local data to uncover other mechanisms that operate in concert with SRL, causing more harm or protecting Black workers, relative to particular areas of the country.

Finally, we found no significant association between SRL and LBW for African-born and Caribbean-born Black pregnant people anywhere in the US. We attribute the lack of association among foreign-born Black pregnant people to a shorter duration of exposure to SRL and, thus, less weathering among this population compared to their US-born counterparts. One of the major adjustments for foreign-born individuals after relocating to the US is understanding US society’s long-standing racialized structures and practices (Benson, 2006; Viruell-Fuentes, Miranda and Abdulrahim, 2012). In a racist society, an individual’s identity is predominantly defined by the membership to a racial group. As immigrants live in the US longer, they are increasingly exposed to racial cues and racist events (Hoggard, Jones and Sellers, 2017). Individuals interpret the meanings of those racial cues in a broader social context and internalize the beliefs regarding their racial identity and racial norms (Hoggard, Jones and Sellers, 2017). Unlike US-born individuals who have been socialized regarding racism and racial oppression from a young age, foreign-born individuals may not embody the stress of living in racialized US society until they have lived in the country for some time (Viruell-Fuentes, 2007). The notion that this socialization process occurs later for immigrants is supported by evidence that recent immigrants are less likely to report discrimination based on their race and that the reporting rate of racial discrimination is similar between those who have lived in the US for an extended duration or have immigrated to the US at a younger age and US-born individuals (Dominguez et al., 2009; Krieger et al., 2011; Brondolo et al., 2015). Additionally, the birth outcomes of children of first-generation African and Caribbean Black immigrants are more like their peers with US-born parents than their own parents’ outcomes (Ramraj, Pulver and Siddiqi, 2015; Andrasfay and Goldman, 2020). This evidence further supports our theory that different racialization durations between US- and foreign-born Black people contributes to in-group heterogeneity.

Selective migration also contributes to the extent to which SRL affects the risk of LBW among foreign-born Black people. The odds ratio for SRL for Caribbean-born Black pregnant people in our multivariate model was greater than one but did not reach a significant level. On the contrary, the odds ratio for African-born Black people was consistently below one. The first wave of Caribbean immigrants, searching for higher wages and better lives, began in the 1910s and 1920s after the completion of the Panama Canal (Hamilton, 2019). Comparatively, a large number of African immigrants entered the US during the 1990s after the passage of the 1980 Refugee Act and the 1986 Immigration Act that provide a path to legal residency for African immigrants, particularly those with high education and skills (Hamilton, 2019). Although these qualifications did not always translate into the same level in the US systems, African people who migrated to the US have higher socioeconomic status and are healthier than Caribbean immigrants on average. These characteristics may explain the varied health impact of structural racism between the two groups. Research on the health impact of structural racism for Black immigrants to date remains limited, despite the fact that the Black community in the US is becoming more diverse. We encourage future research to capitalize on representative datasets that oversample Black immigrants to verify our findings and further characterize the extent to which duration in the US and migration selection contribute to nativity variations observed in our study.

Despite our innovative approach, our study has limitations besides several we discussed earlier. First, we measured SRL for 2016 but infants in our sample could be born at any time in 2017. Because of the cross-sectional time-lag design we used, we ended up making the implicit assumption that the pregnant people’s CZ at the time of delivery was the same as where they resided one year prior. Some pregnant people in our study who moved to the CZ they resided around the time of birth from other CZ were incorrectly assigned the SRL level. However, because CZs generally cover a larger land area, the pregnant people for whom we misclassified their SRL exposure are potentially movers from the non-adjacent state (e.g., moving from Minnesota to New York rather than from Minnesota to Wisconsin) and likely made up of a small portion of our sample. Second, we could not adjust for labor market characteristics (e.g., urbanicity, income inequity) that may vary regionally. The CZ-based estimates of these characteristics were not available when we conducted our analyses. Future analyses to examine how labor market characteristics influence the relationship between SRL and LBW for Black workers are needed. Third, despite our use of a national sample, our findings are still not generalizable to all Black people living in the US. This is particularly true for those who live in areas with a very small number of Black workers (e.g., the state of Oregon, Montana, New Mexico, North and South Dakota, and Utah) where our SRL measure could not be estimated. For the residents of these white-majority states, alternative measurements that do not require population-based data should be explored. Fourth, heterogeneity in the effect of SRL on LBW for Black immigrants (likely, even within nativity groups) reflects varying cultural practices and reasons for migration. Future studies may consider consolidating multiple years of birth certificates or other representative data that oversampled Black immigrants to create a larger, more granular sample for such investigation. Lastly, our SRL measure focuses on inequity in employment opportunities for Black and white workers. We did not focus on other labor market aspects like pay inequities for workers with the same job (Frogner and Schwartz, 2021), racialized occupational exposure (Nazareno et al., 2021), union protections and discriminatory practices by unions (Rosenfeld and Kleykamp, 2012) that may also contribute to racial inequities in birth outcomes. Future studies may operationalize SRL as a multidimensional latent construct to examine the combined effect of various aspects of SRL on LBW and other adverse birth outcomes.

Despite these limitations, our study highlighted the complexity of the relationship between structural racism in labor markets and infant birth weight and offered a solution to the modifiable area unit problem prevalent in this area of research. As the population health scientists continue to move from documenting racial trends to uncovering the racist labor market mechanisms that cause and sustain racial health inequities, this paper provides the next steps for future research, particularly in assessing such effects among Black immigrants.

HIGHLIGHTS.

  • Structural racism is operationalized and measured as an area-based health determinant.

  • Modifiable area unit problems explain effect inconsistencies in the literature.

  • We measured structural racism in labor markets (SRL) across 635 US commuting zones.

  • A commuting-zone measure considers geography where workers conduct a job search.

  • SRL is associated with low-birthweight birth for US-born Southern Black pregnant people of working age.

Source of Funding:

This project was supported by the Minnesota Population Center, which is funded by the Eunice Kennedy Shriver National Institute of Child Health and Human Development of the National Institute of Health (Grant P2C HD041023).

APPENDIX

Lists of countries classified as Africa and the Caribbean Islands

Africa:

Angola, Benin, Botswana, Burkina Faso, Burundi, Cameroon, Cape Verde, Central African Republic, Chad, Comoros, Congo, Cote d’ Ivoire, Dahomey, Djibouti, Equatorial Guinea, Ethiopia, Europa Island, Gabon, The Gambia, Ghana, Glorioso Islands, Guinea, Guinea-Bissau, Kenya, Lesotho, Liberia, Madagascar, Malawi, Mali, Mauritania, Mayotte, Midway Island, Mozambique, Namibia, Niger, Nigeria, Reunion, Rwanda, Sao Tome and Principe, Senegal, Sierra Leone, Somalia, South Africa, Southern Rhodesia, Spanish North Africa, Spanish Sahara, Sudan, Swaziland, Tanzania, Togo, Tromelin Island, Uganda, Western Sahara, Zambia, Zimbabwe

Caribbean Islands:

Anguilla, Antigua and Barbuda, Aruba, The Bahamas, Barbados, British Virgin Islands, Cayman Islands, Cuba, Dominica, Dominican Republic, Grenada, Guadeloupe, Haiti, Jamaica, Martinique, Montserrat, Netherland Antilles, Puerto Rico, Saint Kitts and Nevis, Saint Lucia, Saint Vincent and the Grenadines, Swan Islands, Trinidad and Tobago, Turks and Caicos Islands, US Virgin Islands

Footnotes

Declarations of Interest: None

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

REFERENCES

  1. Almond D, Chay KY and Lee DS (2005) ‘The Cost of Low Birth Weight’, The Quarterly Journal of Economics, 120(3), pp. 1031–1083. Available at: https://about.jstor.org/terms (Accessed: 28 February 2019). [Google Scholar]
  2. Andrasfay T and Goldman N (2020) ‘Intergenerational Change in Birthweight: Effects of Foreign-born Status and Race/ethnicity’, Epidemiology, 31(5), pp. 649–658. doi: 10.1097/EDE.0000000000001217. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Autor DH and Dorn D (2013) ‘The Growth of Low-skill Service Jobs and the Polarization of the US Labor Market’, The American Economic Review, 103(5), pp. 1553–1597. doi: 10.1257/aer.103.5.1553. [DOI] [Google Scholar]
  4. Bailey ZD et al. (2017) ‘Structural Racism and Health Inequities in the USA: Evidence and Interventions’, Lancet, 389, pp. 1453–1463. doi: 10.1016/S0140-6736(17)30569-X. [DOI] [PubMed] [Google Scholar]
  5. Bell CN and Owens-Young JL (2020) ‘Self-Rated Health and Structural Racism Indicated by County-Level Racial Inequalities in Socioeconomic Status: The Role of Urban-Rural Classification’, Journal of Urban Health. doi: 10.1007/s11524-019-00389-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Benson JE (2006) ‘Exploring the Racial Identities of Black Immigrants in the United States’, Sociological Forum, 21(2), pp. 219–247. doi: 10.1007/s11206-006-9013-7. [DOI] [Google Scholar]
  7. Berger M and Sarnyai Z (2015) ‘“More than skin deep”: Stress Neurobiology and Mental Health Consequences of Racial Discrimination’, Stress, 18(1), pp. 1–10. doi: 10.3109/10253890.2014.989204. [DOI] [PubMed] [Google Scholar]
  8. Bonilla-Silva E (2017) Racism without Racists. 5th edn. Rowman & Littlefield Publisher. [Google Scholar]
  9. Brondolo E et al. (2015) ‘Place of Birth Effects on Self-reported Discrimination: Variations by Type of Discrimination’, International Journal of Intercultural Relations. Elsevier Ltd, 49, pp. 212–222. doi: 10.1016/j.ijintrel.2015.10.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Bureau of Labor Statistics (2019) Number of Jobs, Labor Market Experience, and Earning Growth: Results from a National Longitudinal Survey. Washington, DC. Available at: www.bls.gov/nls (Accessed: 17 September 2019). [Google Scholar]
  11. Cajner T et al. (2017) Racial Gaps in Labor Market Outcomes in the Last Four Decades and over the Business Cycle. Washington, DC. doi: 10.17016/FEDS.2017.071. [DOI] [Google Scholar]
  12. Chambers BD et al. (2018) ‘Using Index of Concentration at the Extremes as Indicators of Structural Racism to Evaluate the Association with Preterm Birth and Infant Mortality—California, 2011–2012’, Journal of Urban Health. Journal of Urban Health, pp. 1–12. doi: 10.1007/s11524-018-0272-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Doamekpor LA and Dinwiddie GY (2015) ‘Allostatic Load in Foreign-born and US-born Blacks: Evidence from the 2001–2010 National Health and Nutrition Examination Survey’, American journal of public health. American Public Health Association, 105(3), pp. 591–7. doi: 10.2105/AJPH.2014.302285. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Dominguez TP et al. (2009) ‘Differences in the Self-reported Racism Experiences of US-born and Foreign-born Black Pregnant Women’, Social Science and Medicine. Elsevier Ltd, 69(2), pp. 258–265. doi: 10.1016/j.socscimed.2009.03.022. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Fowler CS and Jensen L (2020) ‘Bridging the Gap between Geographic Concept and the Data We Have: The Case of Labor Markets in the USA’, Environmental and Planning A: Economy and Space, 52(7), pp. 1395–1414. doi: 10.1177/0308518X20906154. [DOI] [Google Scholar]
  16. Fowler CS, Rhubart DC and Jensen L (2016) ‘Reassessing and Revising Commuting Zones for 2010: History, Assessment, and Updates for U.S. “Labor-Sheds” 1990–2010’, Population Research and Policy Review, 35, pp. 263–286. doi: 10.1007/s11113-016-9386-0. [DOI] [Google Scholar]
  17. Frogner BK and Schwartz M (2021) ‘Examining Wage Disparities by Race and Ethnicity of Health Care Workers’, 59(10), pp. 471–478. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Gee GC and Ford CL (2011) ‘Structural Racism and Health Inequities: Old Issues, New Directions’, Du Bois Review, 8(1), pp. 115–132. doi: 10.1017/S1742058X11000130. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Geronimus AT (1992) ‘The Weathering Hypothesis and the Health of African-American Women and Infants: Evidence and Speculations’, Ethnicity & disease, 2(3), pp. 207–21. Available at: http://www.ncbi.nlm.nih.gov/pubmed/1467758 (Accessed: 28 December 2018). [PubMed] [Google Scholar]
  20. Geronimus AT et al. (2006) ‘“Weathering” and Age Patterns of Allostatic Load Scores among Blacks and Whites in the United States’, American Journal of Public Health, 96(5), pp. 826–833. doi: 10.2105/AJPH.2004.060749. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Geronimus AT and Thompson JP (2019) ‘To Denigrate, Ignore, or Disrupt: Racial Inequality in Health and the Impact of a Policy-induced Breakdown of African American Communities’, Du Bois Review, 1(2), pp. 247–279. doi: 10.1017/S1742058X04042031. [DOI] [Google Scholar]
  22. Gollob HF and Reichardt CS (1987) ‘Taking Account of Time Lags in Causal Models’, Child Development, 58(1), pp. 80–92. [PubMed] [Google Scholar]
  23. Graves JA et al. (2020) ‘Medicaid Expansion Slowed Rates of Health Decline for Low-income Adults in Southern States’, Health Affairs, 39(1), pp. 67–76. doi: 10.1377/hlthaff.2019.00929. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Groos M et al. (2018) ‘Measuring Inequity: A Systematic Review of Methods Used to Quantify Structural Racism’, Journal of Health Disparities Research and Practice, 11(2), pp. 190–206. Available at: http://digitalscholarship.unlv.edu/jhdrp/http://digitalscholarship.unlv.edu/jhdrp (Accessed: 18 August 2019). [Google Scholar]
  25. Guzman G (2020) Household Income by Race and Hispanic Origin: 2005–2009 and 2015–2019. Washington, DC. Available at: https://www.census.gov/content/dam/Census/library/publications/2020/acs/acsbr19-07.pdf. [Google Scholar]
  26. Hamilton D, Austin A and Darity WA Jr (2011) Whiter Jobs, Higher Wages. Occupational Segregation and the Lower Wages of Black Men. Washington, DC. Available at: https://www.epi.org/files/page/-/BriefingPaper288.pdf (Accessed: 23 August 2019). [Google Scholar]
  27. Hamilton TG (2019) Immigration and the Remaking of Black America. New York, NY: Russell Sage Foundation. [Google Scholar]
  28. Hannah-Jones N (2021) The 1619 Project: A New Origin Story. One World. [Google Scholar]
  29. Hardeman RR et al. (2021) ‘Association of Residence in High-Police Contact Neighborhoods With Preterm Birth Among Black and White Individuals in Minneapolis’, JAMA Network Open. American Medical Association, 4(12), pp. e2130290–e2130290. doi: 10.1001/JAMANETWORKOPEN.2021.30290. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Henry J Kaiser Family Foundation (2020) Paid Family and Sick Leave in the U.S. Available at: https://www.kff.org/womens-health-policy/fact-sheet/paid-family-leave-and-sick-days-in-the-u-s/ (Accessed: 31 December 2019). [Google Scholar]
  31. Hicken MT et al. (2013) ‘“Every Shut Eye, Ain’t Sleep”: The Role of Racism-related Vigilance in Racial/ethnic Disparities in Sleep Difficulty’, Race Soc Probl, 5(2), pp. 100–112. doi: 10.1007/s12552-013-9095-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Hoggard LS, Jones SCT and Sellers RM (2017) ‘Racial Cues and Racial Identity: Implications for How African Americans Experience and Respond to Racial Discrimination’, Journal of Black Psychology, 43(4), pp. 409–432. doi: 10.1177/0095798416651033. [DOI] [Google Scholar]
  33. Huber PJ (1967) ‘The Behavior of Maximum Likelihood Estimates under Nonstandard Conditions’, in Proceeding of the Fifth Berkeley Symposium on Mathematical Statistics and Probability. Berkeley, CA: University of California Press, pp. 221–233. [Google Scholar]
  34. Hung P et al. (2017) ‘Access to Obstetric Services in Rural Counties Still Declining, with 9 Percent Losing Services, 2004–14’, Health Affairs, 36(9), pp. 1663–1671. doi: 10.1377/hlthaff.2017.0338. [DOI] [PubMed] [Google Scholar]
  35. Iceland J and Sharp G (2013) ‘White Residential Segregation in U.S. Metropolitan Areas: Conceptual Issues, Patterns, and Trends from the US Census, 1980 to 2010’, Population Research and Policy Review, 32(5). doi: 10.1007/s11113-013-9277-6.White. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. IPUMS USA (2020a) 2013–2017 Industry Codes (IND). Available at: https://usa.ipums.org/usa/volii/ind2013.shtml (Accessed: 15 September 2019).
  37. IPUMS USA (2020b) ACS Occupation Codes (OCC) 2000–2017. Available at: https://usa.ipums.org/usa/volii/occ_acs.shtml (Accessed: 15 September 2019).
  38. Krieger N et al. (2008) ‘The Inverse Hazard Law: Blood Pressure, Sexual Harassment, Racial Discrimination, Workplace Abuse and Occupational Exposures in US Low-income Black, White and Latino Workers’, Social Science & Medicine, 67, pp. 1970–1981. doi: 10.1016/j.socscimed.2008.09.039. [DOI] [PubMed] [Google Scholar]
  39. Krieger N et al. (2011) ‘Racial Discrimination, Psychological Distress, and Self-rated Health among US-born and Foreign-born Black Americans’, American Journal of Public Health, 101(9), pp. 1704–1713. doi: 10.2105/AJPH.2011.300168. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Kwan M (2012) ‘The Uncertain Geographic Context Problem’, Annals of the Association of American Geographers, 102(5), pp. 958–968. [Google Scholar]
  41. Liu SY et al. (2019) ‘Structural Racism and Severe Maternal Morbidity in New York State’, Clinical Medicine Insights: Women’s Health, 12, pp. 1–8. doi: . [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Lukachko A, Hatzenbuehler ML and Keyes KM (2014) ‘Structural Racism and Myocardial Infarction in the United States’, Social science & medicine (1982). NIH Public Access, 103, pp. 42–50. doi: 10.1016/j.socscimed.2013.07.021. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Martin JA et al. (2017) National Vital Statistics Reports, National Vital Statistics Reports. Hyattsville, MD. Available at: http://www.cdc.gov/nchs/data_access/Vitalstatsonline.htm (Accessed: 6 September 2019). [Google Scholar]
  44. Mason SM et al. (2010) ‘Ethnic Density and Preterm Birth in African-, Caribbean-, and US-Born non-Hispanic Black populations in New York City’, Am J Epidemiol, 172(7), pp. 800–808. doi: 10.1093/aje/kwq209. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Massey DS and Denton NA (1993) American Apartheid. Cambridge, MA: Harvard University Press. [Google Scholar]
  46. McCluney CL et al. (2018) ‘Structural Racism in the Workplace: Does Perception Matter for Health Inequalities?’, Social Science and Medicine. Elsevier Ltd, 199, pp. 106–114. doi: 10.1016/j.socscimed.2017.05.039. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. McClure ES et al. (2020) ‘Racial Capitalism within Public Health: How Occupational Settings Drive COVID-19 Disparities’, American Journal of Epidemiology. doi: 10.1093/aje/kwaa126/5866668. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Mendez DD, Hogan VK and Culhane JF (2013) ‘Stress during Pregnancy: The Role of Institutional Racism’, Stress and Health, 29(4), pp. 266–274. doi: 10.1002/smi.2462. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Minnesota Compass (2020) Early Childhood - Key Measures. Available at: https://www.mncompass.org/early-childhood/low-birth-weight#1-5872-g (Accessed: 28 February 2019).
  50. Nazareno J et al. (2021) ‘From Imperialism to Inpatient Care: Work Differences of Filipino and White Registered Nurses in the United States and Implications for COVID-19 Through an Intersectional Lens’, Gender, Work and Organization. John Wiley and Sons Inc, 28(4), pp. 1426–1446. doi: 10.1111/GWAO.12657/FORMAT/PDF. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. New York State Department of Labor (2020) Labor Statistics. Available at: https://www.labor.ny.gov/stats/index.shtm (Accessed: 17 September 2019).
  52. Pabayo R et al. (2019) ‘Structural Racism and Odds for Infant Mortality Among Infants Born in the United States 2010’, Journal of Racial and Ethnic Health Disparities, 6, pp. 1095–1106. doi: 10.1007/s40615-019-00612-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Pager D and Shepherd H (2008) ‘The Sociology of Discrimination: Racial Discrimination in Employment, Housing, Credit, and Consumer Markets’, Annual Review of Sociology, 34, pp. 181–209. doi: 10.1146/annurev.soc.33.040406.131740. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Pavalka E and Smith B (1999) ‘The Rhythm of Work : Health Effects of Women ‘ s Work Dynamics *’, Social Forces, 77(3), pp. 1141–1162. [Google Scholar]
  55. Quillian L et al. (2017) ‘Meta-analysis of Field Experiments Shows No Change in Racial Discrimination in Hiring Over Time’, Proceedings of the National Academy of Sciences, 114(41), pp. 10870–10875. doi: 10.1073/pnas.1706255114. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Ramraj C, Pulver A and Siddiqi A (2015) ‘Intergenerational Transmission of the Healthy Immigrant Effect (HIE) Through Birth Weight: A Systematic Review and Meta-analysis’, Social Science and Medicine. Elsevier Ltd, 146, pp. 29–40. doi: 10.1016/j.socscimed.2015.10.023. [DOI] [PubMed] [Google Scholar]
  57. Reskin BF (1998) The Realities of Affirmative Actions in Employment. American Sociological Association. [Google Scholar]
  58. Riley AR (2018) ‘Neighborhood Disadvantage, Residential Segregation, and Beyond-Lessons for Studying Structural Racism and Health’, J Racial and Ethnic Health Disparities, 5, pp. 357–365. doi: 10.1007/s40615-017-0378-5. [DOI] [PubMed] [Google Scholar]
  59. Rosenfeld J and Kleykamp M (2012) ‘Organized Labor and Racial Wage Inequality in the United States’, American Journal of Sociology, 117(5), pp. 1460–1502. doi: 10.1086/663673. [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Ruggles S et al. (2020) ‘IPUMS USA’. Minneapolis, MN. Available at: 10.18128/D010.V9.0. [DOI] [Google Scholar]
  61. Schisterman EF, Cole SR and Platt RW (2009) ‘Overadjustment Bias and Unnecessary Adjustment in Epidemiologic Studies’, Epidemiology, 20(4), pp. 488–495. doi: 10.1097/EDE.0b013e3181a819a1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  62. Stevenson AJ et al. (2016) ‘Effect of Removal of Planned Parenthood from the Texas Women’s Health Program’, New England Journal of Medicine, 374(9), pp. 853–860. doi: 10.1056/nejmsa1511902. [DOI] [PMC free article] [PubMed] [Google Scholar]
  63. Taylor YJ, Liu TL and Howell EA (2020) ‘Insurance Differences in Preventive Care Use and Adverse Birth Outcomes among Pregnant Women in a Medicaid Nonexpansion State: A Retrospective Cohort Study’, Journal of Women’s Health, 29(1), pp. 29–37. doi: 10.1089/jwh.2019.7658. [DOI] [PMC free article] [PubMed] [Google Scholar]
  64. United Nations Development Program (2009) Somalia’s Missing Million: The Somali Diaspora and Its Role in Development.
  65. US Bureau of Census (2016) American Community Survey Data Suppression. Washington, DC. [Google Scholar]
  66. US Bureau of Census (2020a) Census Regions and Divisions of the United States. Available at: https://www2.census.gov/geo/pdfs/maps-data/maps/reference/us_regdiv.pdf (Accessed: 9 December 2019).
  67. US Bureau of Census (2020b) Educational Attainment in the United States: 2019. Available at: https://www.census.gov/data/tables/2019/demo/educational-attainment/cps-detailed-tables.html (Accessed: 12 January 2019).
  68. US Department of Agriculture Economic Research Service (2020) Commuting Zones and Labor Market Areas. Available at: https://www.ers.usda.gov/data-products/commuting-zones-and-labor-market-areas/ (Accessed: 17 September 2019).
  69. Viruell-Fuentes EA (2007) ‘Beyond Acculturation: Immigration, Discrimination, and Health Research among Mexicans in the United States’, Social Science and Medicine, 65(7), pp. 1524–1535. doi: 10.1016/j.socscimed.2007.05.010. [DOI] [PubMed] [Google Scholar]
  70. Viruell-Fuentes EA, Miranda PY and Abdulrahim S (2012) ‘More Than Culture: Structural Racism, Intersectionality Theory, and Immigrant Health’, Social Science and Medicine. Elsevier Ltd, 75(12), pp. 2099–2106. doi: 10.1016/j.socscimed.2011.12.037. [DOI] [PubMed] [Google Scholar]
  71. Wallace M et al. (2017) ‘Separate and Unequal: Structural Racism and Infant Mortality in the US’, Health and Place. Elsevier Ltd, 45(November 2016), pp. 140–144. doi: 10.1016/j.healthplace.2017.03.012. [DOI] [PubMed] [Google Scholar]
  72. Wallace ME et al. (2015) ‘Joint Effects of Structural Racism and Income Inequality on Small-for-Gestational-Age Birth.’, American journal of public health. American Public Health Association, 105(8), pp. 1681–8. doi: 10.2105/AJPH.2015.302613. [DOI] [PMC free article] [PubMed] [Google Scholar]
  73. White H (1980) ‘A Heteroskedasticity-Consistent Covariance Matrix Estimator and a Direct Test for Heteroskedasticity’, Econometrica. JSTOR, 48(4), p. 817. doi: 10.2307/1912934. [DOI] [Google Scholar]
  74. Williams DR, Lawrence JA and Davis BA (2019) ‘Racism and Health: Evidence and Needed Research’, Annu. Rev. Public Health, 40, pp. 105–125. doi: 10.1146/annurev-publhealth. [DOI] [PMC free article] [PubMed] [Google Scholar]
  75. Wingfield AH (2019) Flatlining: Race, Work, and Health Care in the New Economy. 1st edn. University of California Press. [Google Scholar]
  76. Wong DW (2004) ‘The Modifiable Area Unit Problem (MAUP)’, in Janelle D (ed.) WorldMinds: Geographical Perspectives on 100 Problems. Kluwer Academic Publishers, pp. 571–575. [Google Scholar]

RESOURCES