Blocked Paths, Unequal Trajectories: Examining the Career Outcomes of Health Professions Graduates by Race, Gender, and Place of Education

Abstract

Despite the need for more - and more diverse - health professionals to meet the U.S.’ growing healthcare needs, qualitative research suggests that many health professions school graduates (HPGs) are tracked away from pursuing a health professions career. To what extent do HPGs’ career trajectories vary across race, gender, and place of education (i.e., whether one received their health professions education in the U.S. or abroad)? Drawing on the National Survey of College Graduates, I identify that among 5,966 HPGs (medical, dental, veterinary, optometry, and podiatry school graduates), foreign-educated HPGs are less likely than U.S.-educated HPGs to become health professionals in the U.S. However, not all foreign-educated HPGs experience equal degrees of disadvantage; racially minoritized foreign-educated HPGs - and most strikingly, Asian foreign-educated men and women HPGs - demonstrate the lowest likelihoods of matriculating to a health professions career. Moreover, structural barriers, including initial U.S. visa status, significantly disadvantage foreign-educated HPGs on their path towards becoming practicing health professionals. Together, results suggest that the stunted career trajectories of foreign-educated HPGs might further stall the diversification of the health professions and limit the potential of immigrant HPGs to draw on their unique human capital post-migration. Findings both improve understanding of the contributors to the U.S. health professional shortage and advance sociological knowledge around how the health professions serve as key sites in the reproduction of inequality across axes of race, gender, and place of education.

Introduction

As the health professions workforce globalizes (Bell and Walkover 2021; Frenk et al. 2010; Clark, Stewart, and Clark 2006), increasing numbers of foreign-educated health professions school graduates (HPGs) are migrating to high-income countries – including the United States (U.S.) - in the hopes of pursuing a health professions career. Yet despite the U.S.’ need for more (and more diverse) health professionals, research finds that foreign-educated HPGs experience significant barriers to becoming health professionals post-migration, particularly in occupations including medicine (Franklyn 2023) dentistry (Kellesarian 2018; Pannu et al. 2013) and veterinary medicine (Witter 2005).

The health professions are highly regulated, involve extensive and costly licensing procedures, and as a result, are highly competitive to enter (Saks 2015). Regulation in the health professions serves two key objectives; it exists both to protect the public through setting standards of competency and quality for health professionals (Greiner and Knebel 2003), and, as sociological work has shown, it also serves to enforce social closure to protect, “role boundaries and the benefits associated with professional status by monitoring and limiting entry into occupational groups” (King et al 2018, 5). As a result of this closure, foreign-educated HPGs face significant barriers toward entering the health professions. Closure toward foreign-educated HPGs is often justified by the fact that a foreign health professions education is not inherently comparable to a U.S. education (Michalski et al. 2017). However, in examining the career outcomes of physicians specifically, Jenkins (2020) argues that this closure may also be the result of status separation, a tacit process aimed at restricting the opportunities of foreign-educated HPGs in order to provide more career opportunities to the U.S. educated. Consequently, qualitative research suggests that the barriers foreign-educated HPGs face toward entering the U.S. health professions are often driven by explicitly discriminatory hiring practices that are employed to maintain the upper hand for U.S.-educated health professions graduates in a competitive market. Moreover, despite the fact that foreign-educated HPGs are often stereotyped as less qualified than U.S.-educated HPGs, quantitative research has presented mixed evidence to support such claims (Tsugawa 2017; Mick and Comfort 1997) and finds instead that foreign-educated HPGs play an important role in contributing to the U.S. healthcare safety net (e.g., Edelstein 2010).

More research on the integration of foreign-educated HPGs into U.S. health care is warranted for three reasons. First, there is little quantitative, longitudinal evidence showing whether and to what extent a U.S. (versus foreign) health professions education shapes the occupational trajectories of HPGs over the professional life course. Second, existing literature on foreign-educated health professionals has seldom examined how their career trajectories are shaped not only by place of education (i.e., whether one’s education was obtained in the U.S. or abroad) but also by race and gender. Finally, whereas a substantial body of scholarship around health professions education has focused on the experiences of foreign-educated physicians (e.g., Franklyn 2023; Schut 2022; Bell 2023; Jenkins 2020), more work is needed to understand the trajectories of foreign-educated health professionals across fields – i.e., physicians as well as other similar-status professionals, including dentists, veterinarians, optometrists, and podiatrists.

In this article, I leverage longitudinal data from the 2003–2013 National Survey of College Graduates to examine the likelihood that health professions graduates (HPGs) (i.e., individuals with graduate degrees in medicine, dentistry, podiatry, optometry, and veterinary medicine) across axes of race, gender, and place of education match into careers that are concordant with their degree of education. In doing so, I produce some of the first quantitative estimates demonstrating the occupational trajectories of HPGs, complementing existing qualitative evidence showing the pervasiveness of inequality across various social axes in the health professions.

Findings indicate that the disadvantage faced by foreign-educated HPGs toward pursuing a U.S. health professions career is magnified for those with minoritized race and gender statuses. Foreign-educated Asian and underrepresented minority (URM) men HPGs, as well as foreign-educated Asian women HPGs, are least likely to work as health professionals in the U.S., with Asian foreign-educated HPGs in particular experiencing the lowest chances of obtaining a health professions career post-migration and across the course of their working lives. I argue these findings bolster understanding of how the demographic composition of the health professions has remained relatively homogenous over time, despite growing emphasis on diversification in in fields such as medicine (e.g., Nguemeni Tiako et al. 2022; Deville et al. 2015) and dentistry (Salsberg et al. 2021). Moreover, findings showing that many HPGs do not matriculate to a career in the health professions has implications for our understanding of the factors contributing to a growing U.S. health professions shortage. HPGs who do not or cannot pursue a health professions career forfeit valuable human capital, leading to “brain waste” (Boesch and Nunn 2024). Critically, an ever-diversifying U.S. population necessitates an increasingly diverse healthcare workforce (Snyder, Frogner, and Skillman 2018). Consequently, findings from these analyses both inform a growing body of medical sociology research around inequalities in health professions education (see Jenkins et al. 2021 for a review) and can inform the development of health policies aimed at ameliorating the barriers foreign-educated HPGs face toward pursuing a health professions career.

Background

Throughout much of the 20^th century, the health professions were largely white, male-dominated fields that practiced explicit exclusion of individuals possessing minoritized racial and gender statuses (e.g., Trotter 2022; Beagan 2001). Yet even as the U.S. population diversifies, medicine, dentistry, veterinary medicine, and other similar-status health professions face significant challenges toward achieving greater racial and gender equity (Snyder, Frogner, and Skillman 2018). Figures 1a and 1b show the distribution of race and gender across five health professions (medicine; dentistry; veterinary medicine; optometry; and podiatry) using 2009–2013 American Community Survey (ACS) data. In terms of race, dentistry, optometry, medicine, podiatry, and veterinary medicine were majority White between 2009–2013 (between 69% for physicians and 92% for veterinarians). Whereas among dentists, optometrists, and physicians, Asians were the second largest racial group (between 13% for optometrists and 20% for physicians), for podiatrists and veterinarians, underrepresented minorities (those identifying as Black, Hispanic/Latino, Native American, or Other) were the second largest group (11% for podiatrists and 5% for veterinarians).

Figure1a and 1b. Percent of Individuals Employed in Occupations Represented in the NSCG Health Professions Category by Race (1a) and Sex (1b).

Note: Descriptive statistics are population weighted. Data are drawn from the 2009–2013 American Community 5-year estimates. Sample is restricted to employed individuals, aged 25–70. URM includes all individuals who identified their race as Black, Hispanic/Latino, Indigenous, or Other.

In terms of gender, Figure 1b shows the distribution of men and women across the health professions. Most of the professions skewed male (74% for dentists; 60% for optometrists; 64% among physicians; and 74% among podiatrists). The exception was veterinary medicine, which prior work has identified as a “feminized” profession (Irvine and Vermilya 2010); ACS data show that 53% of veterinarians were women between 2009–2013.

Although data from the ACS are unable to speak to the percent of foreign-educated individuals represented in each of the aforementioned health professions, and while available data on the number of foreign-educated veterinarians, optometrists, and podiatrists are limited, existing estimates suggest around 25% of physicians in the U.S. are foreign-educated (Andrews et al. 2024), as are 17% of dentists (Mertz et al. 2016). Foreign-educated health professionals can contribute substantially to racial and gender diversity within the health professions, but many face significant barriers toward entering these occupations, which have long practiced exclusion across racial and gender lines. What factors might prevent foreign-educated HPGs from working as health professionals once in the U.S., and how might the challenges these HPGs encounter be magnified among those with minoritized racial and/or gender statuses? In the following sections, I highlight how a host of factors – including immigration status, cultural signaling, and racism and sexism within the health professions - might impact HPGs’ successful integration into a U.S. health professions career.

First, immigration status can serve as a significant barrier towards foreign-educated HPGs’ labor market integration. The two most commonly held visas among foreign-educated HPGs are employer-sponsored H-1B and J-1 visas. Immigrants with H-1B visas must regularly re-apply for legal status until permanent residency is obtained – a process which can take decades for some immigrant groups. These visa holders must also remain employed relatively consistently and without gaps. Because H-1B visas are costly for employers to sponsor – in terms of both financial and bureaucratic costs - visa holders may be seen as burdensome to potential employers (Banerjee 2006). Moreover, those holding a J-1 “exchange visitor” visa experience a separate set of challenges and vulnerabilities. Namely, although J-1 visa holders are typically required to return to their home country for two years after a certain period of stay in the U.S., foreign-educated physicians can be granted a visa “waiver” and remain in the U.S. if they agree to work in underserved areas of the country (Braga, Khanna, and Turner 2024). While this pathway may result in underserved populations gaining access to health care, it also marginalizes foreign-educated physicians through concentrating them in safety-net positions, consequently resulting in the “best” and most desirable positions being reserved for U.S. physicians (Schut 2022). Moreover, this waiver is typically only available to physicians (Gunselman 2009), and therefore other groups of HPGs may not be eligible for this waiver and must return home after their visa expires.

Immigration policies are often portrayed as “color-blind,” but their impacts are both racialized and gendered (Douglas, Sáenz, and Murga 2015). Namely, Asian immigrants comprise a large percentage of high-skilled, H-1B visa holders, and their precarity reflects the continued treatment of Asians as, “inexpensive, short-term, and ideal flexible labor” that the U.S. has constructed as, “not only non-white, but also as non-citizen, property-less, and employer-dependent” (Banerjee 2006:428). The precarity of H-1B workers further extends to their spouses, who may be granted H-4 dependent spousal visas. Uniquely, H-4 visa holders are barred from obtaining employment for several years after migration, until they become eligible for work authorization. This visa has consequently been called the “career destroyer,” as it tracks holders out of the labor force for a number of years, leading to brain waste among a group of highly qualified individuals. Moreover, because over 90% of H-4 holders are women (Wilson 2018), this visa contributes to gender inequities in labor market participation and brain waste among highly-educated immigrant women who are tracked out of the labor force post-migration (Mallapragada 2017).

Second, powerful U.S. health professions organizations (e.g., the American Medical Association, American Dental Association) work together with federal and state governments to produce regulations around which HPGs will be able to pursue U.S. licensure without needing to re-complete graduate medical education. These organizations also determine how easy (or how difficult) it will be for foreign-educated HPGs to obtain licensure in the U.S. Whereas European medical graduates are more likely to have the lowest requirements to obtain licensure in the U.S., those from African, Latin American, and many Asian medical schools are required to undergo a longer and involved licensure process (Bell 2023; Schut 2022). Thus, just as being a member of a socially majoritized group (i.e., “white” and “male”) has been found to credential individuals by increasing their likelihood of being hired, promoted, and integrated into organizations (Ray 2019), so too might where an individual was educated credential them, facilitating their entry and integration into organizations, including health care.

Indeed, existing sociological scholarship demonstrates that the location of an individual’s degree-granting institution powerfully dictates their U.S. labor market outcomes. Zeng and Xie (2004) introduced place of education as a key determinant of immigrant labor market outcomes, showing that for Asian immigrants, only those who obtained their highest degree from a foreign institution faced a wage penalty in the U.S. labor market, net of nativity. Work also finds that foreign-educated immigrants are less financially renumerated in the labor market than their U.S.-educated counterparts (e.g., Sakamoto and Kim 2010) and that foreign-educated individuals are less likely than their U.S.-educated counterparts to work in jobs that match their level of education and skill (Li and Lu 2021; Damelang, Ebensperger, and Stumpf 2020).

Signaling theory (Spence 1973) may help explain why place of education determines labor market outcomes. The theory argues that because skills may be difficult to directly observe, employers use educational credentials as a proxy for skill-level and productivity. Consequently, a foreign education may signal to employers that a worker is less capable than a U.S.-educated worker, regardless of the worker’s actual level of skill and productivity (Lancee and Bol 2017). More specifically, a U.S. education may signal to potential employers that a worker possesses certain qualities (e.g., cultural capital, English language ability, etc.) that are believed to make them “better” workers and more aligned with the values of the hiring organization. Consequently, those educated in countries that are more culturally similar to the U.S. (e.g., European countries) may experience advantages in the hiring process because employers and organizations hold pre-conceived notions about perceived cultural (and racial) similarities that make these individuals better or more qualified workers (He, Gerber, and Xie 2023).

In the health professions, evidence of U.S. education serving as a credential is seen in the degree to which U.S. residency program administrators have been found to routinely filter through applications in search of where an applicant attended medical school; if educated outside of the U.S., applicants may be eliminated from further review solely on the basis of having a foreign degree (Desbiens and Vidallet 2010). Foreign-educated HPGs are more likely to be rejected by residency training programs, fellowships, and career positions than their U.S.-educated counterparts, net of having the same licensing exam (e.g., USMLE) score (e.g., Kureshi et al. 2019). Moreover, specifically along axes of foreign education and country of education, the process of meeting licensure requirements is more laborious and time-intensive for foreign-educated HPGs from countries in the Global South (ECFMG 2023; Oikeleme and Healy 2013).

Finally, racism and sexism continue to be widespread in the health professions. Multi-leveled discrimination processes occurring at the structural, organizational, and interpersonal levels have not only been found to prevent the entry of diverse HPGs into the health professions but may also increase the likelihood that minoritized health professionals experience burnout resulting from exposure to discrimination from co-workers and patients (Wingfield and Chavez 2020; Mertz et al. 2016). Foreign-educated HPGs who possess racially minoritized statuses may face compounding disadvantages in entering into and remaining in the health professions. Specifically, a foreign education – and particularly a foreign education from a non-European or Anglophone country - might evoke both nativist and racist stereotypes through signaling that an HPG is not well-matched culturally and/or educationally to the U.S. healthcare system. For example, Asian health professionals face numerous stereotypes, including being perceived as “forever foreigners” who are less fluent in English and culturally dissimilar from American patients (Louie-Poon et al. 2023; Min and Jang 2015). As a result, foreign-educated Asian HPGs might encounter both racial stereotypes and discrimination related their foreign education status which is widely viewed as less valuable and less credible than a U.S. education.

Moreover, foreign-educated women – in addition to facing potential challenges due to dependent visa status - may also face discrimination if they are perceived to be less motivated or career-driven than men, given widespread stereotypes about the relative amount of dedication men versus women show toward furthering their careers (Gerstel and Clawson 2014; Irvine and Vermilya 2010). Additionally, as immigrant women more often take on the burden of shouldering family and domestic responsibilities, immigrant women health professionals have been found to delay the process of accreditation, retraining, and relicensing in order to support the careers of their male partners (Bourgeault et al 2016).

Together, a host of factors – including citizenship and immigration status, occupational closure, and discrimination within the health professions – have been identified in largely separate literatures as powerful forces determining the likelihood that HPGs successfully integrate into a U.S. health professions career. Few studies, however, have examined how these factors may combine to shape the outcomes of HPGs, and particularly, how they might disadvantage the outcomes of foreign-educated HPGs. Thus, expanding the literature on the reproduction of inequality within the health professions, the analyses that follow describe the career trajectories of HPGs by race, sex, and importantly, place of education. Moreover, I estimate to what degree HPGs’ career trajectories are shaped by sociodemographic, family, and immigration-related factors, thus speaking to the barriers HPGs might encounter on their pathway to a career in the professions. Specifically, the following research questions guide the scope of the study:

1. To what extent are the career trajectories of health professions graduates stratified by place of education, race, and gender?

2. What characteristics (e.g., sociodemographic, family, or immigration-related) shape HPGs’ likelihoods of matching into health professions careers over the course of their professional lives?

Data and Methods

Analyses draw on longitudinal data from the 2003–2013 National Survey of College Graduates (NSCG). The NSCG is a biennial survey of college graduates residing in the U.S., sponsored by the National Science Foundation and conducted by the U.S. Census Bureau. It collects detailed information (including sociodemographic, family, education, and career histories) on individuals living in the U.S. during the survey “reference week” who are under the age of 76 and have a bachelor’s degree or higher. Because of its focus on college graduates, the NSCG is useful for examining the career trajectories of high-skilled individuals (Hersch and Meyers 2018; Tong 2010). Moreover, unlike other panel datasets (i.e., NLSY and SIPP), the NSCG directly asks respondents about their place of education (in the U.S. or abroad), making it well-suited to capture the role of place of education in shaping occupational trajectories.

The NSCG is a repeated cross-sectional survey with a rotating panel element; in each year of data collection, the survey first draws respondents from a sampling pool of recent American Community Surveys and decennial censuses (depending on whichever was most recently collected relative to the NSCG survey year). A random sample of respondents are followed longitudinally for a maximum of four waves (approximately 8–10 years) and are then “rotated out” of the sample. The weighted sample response rate was 73% in the 2003 NSCG and 74% in the 2013 NSCG (National Science Foundation 2017).

I restrict analyses to non-retired NSCG respondents who received a graduate-level “health professions” degree from a U.S. or foreign university (i.e., “HPGs). This group includes those possessing a degree in allopathic medicine (i.e., an MD, “doctor of medicine” degree), osteopathic medicine (i.e., a DO, “doctor of osteopathy” degree), or a degree in podiatry, optometry, dentistry, or veterinary medicine. Most HPGs in the NSCG had 3 or 4 waves of data collection. The resulting analytical sample is 5,966 respondents with 12,685 observations over the course of a 10-year period (2003–2013).

A limitation of the NSCG is that it is not possible to disaggregate HPGs into specific degree groups, or to disaggregate health professionals into specific occupations. However, despite there being differences in the day-to-day practices of those employed as dentists, physicians, podiatrists, optometrists, and veterinarians, individuals in these occupations require similar, high levels of post-graduate training and experience similar pathways toward licensure. Figure 2 demonstrates national estimates of the number of individuals employed in each of the five health professions explored in this study during the study period (2003–2013), drawn from the U.S. Bureau of Labor Statistics. Among the five health professions represented in this study, estimates show that nationally, physicians make up the majority of health professionals (between 67% in 2003 to 76% in 2013), followed by dentists. Veterinarians, optometrists, and podiatrists comprise a significantly smaller proportion of the overall health professional workforce. Thus, as the distribution of occupations within the NSCG likely follows that of the national distribution, we can expect that findings in this analysis are largely driven by physicians and dentists, and to a lesser degree, podiatrists, optometrists, and veterinarians.

Figure 2: Number of Individuals Employed as Dentists, Optometrists, Physicians & Surgeons, Podiatrists, and Veterinarians in the United States, 2003–2013.

Note. Author’s own calculations from 2003–2013 BLS yearly estimates of the number of individuals employed in each of the five health professions occupations represented in the NSCG.

Variables

I explore variation in whether HPGs were employed as health professionals at the time of survey (1 = employed as a health professional) for each wave of data collection. This outcome is derived from the occupation variable, asked of respondents in all NSCG waves. Respondents were asked what their main job activity was in the week prior to survey response. Individuals who reported that their primary job in the reference week was “diagnosing/treating practitioner” were coded as working as a health professional.

I explore temporal variation in the likelihood that HPGs report working (practicing) as health professionals across gender (women and men), race (white; Asian; and URM), and place of health professions education (U.S.-educated and foreign-educated). The NSCG includes three racial categories: White, Asian, and “Other.” The “Other” category may thus include individuals who identify as Black, Latinx, Indigenous, and/or bi/multiracial. I refer to those categorized as “Other” by the NSCG as underrepresented minorities, or, “URM.”

Place of education is drawn from the NSCG survey question that asks respondents to identify whether they obtained a specific educational degree in the U.S. or abroad. Because the literature on health professions education has identified that place of education is a salient factor contributing to the likelihood of HPGs practicing as health professionals, I stratify respondents by place of education instead of nativity. Furthermore, over 90% of foreign-educated HPGs in the NSCG data were also foreign-born. Sensitivity analyses indicated that place of birth (as opposed to place of education) was not a statistically significant axis of stratification determining the likelihood of practicing as a health professional.

The key covariate in this analysis is a continuous time measure of the number of years since a respondent graduated from a graduate health professions program. This measure is interacted with respondents’ gender, race, and place of education to investigate the career trajectories of HPGs over time. Additional covariates include a continuous measure of respondents’ birth cohort; a time-varying categorical measure of respondents’ marital status (married; widowed/separated/divorced; and never married); a time-varying categorical measure of whether respondents had children under the age of 18 at the time of survey (0 = respondent has no children, 1 = respondent has children); and respondents’ first U.S. citizenship status (U.S.-born/U.S. citizen; permanent U.S. resident; temporary resident/work visa holder; training visa holder¹; dependent visa holder; other type of visa holder). These sociodemographic, family, and immigration factors contextualize how group-level compositional differences might be associated with varying likelihoods of matriculation to a career as a health professional. Importantly, views of family responsibilities are highly gendered in the health professions (Gerstel and Clawson 2014), which often penalize the careers of women who have children; therefore, having children may negatively impact the likelihood that women HPGs matriculate to careers as health professionals, but may benefit men’s careers. Supplemental respondent characteristics are included in Appendix A, Table 1. These variables further contextualize HPGs’ sociodemographic, geographic, and labor market characteristics.

Table 1:

Sociodemographic, family, and immigration characteristics of HPGs in the 2003–2013 National Survey of College Graduates

Characteristics	Gender		Total (%)
	Men (%)	Women (%)
Mean birth year	1957	1964	1959
SD	11.8	11.2	12.0
Race
White	64.7	53.2	60.7
Asian	20.1	27.5	22.7
URM	15.1	19.3	16.6
Place of education
Foreign-educated	25.2	27.9	26.2
Place of education and race
U.S.-educated white	56.0	45.2	52.3
U.S.-educated Asian	8.7	13.0	10.2
U.S.-educated URM	10.0	13.9	11.4
Foreign-educated white	8.7	8.0	8.5
Foreign-educated Asian	11.4	14.5	12.5
Foreign-educated URM	5.1	5.4	5.2
Mean number of years since health professions school
graduation	22.7	16.6	20.4
SD	12.0	10.6	11.9
Marital status
Never married	7.4	14.9	10.0
Separated, divorced, or widowed	5.6	10.4	7.2
Married	87.1	74.7	82.8
Children	55.3	56.9	55.9
First U.S. citizenship status
U.S.-born/U.S. citizen	66.0	60.2	64.0
Permanent U.S. resident	11.7	14.6	12.7
Temporary resident/work visa	3.2	3.2	3.2
Temporary student visa	13.9	11.7	13.1
Temporary dependent visa	2.2	8.2	4.3
Other type of visa	3.0	2.2	2.7
N	3,787	2,179	5,966
Observations	8,273	4,412	12,685
%	63.5	36.5	100.0

Analytical Strategy

I employ linear growth curve modeling (GCM) to evaluate the career trajectories of HPGs. GCM, also known as random coefficient, hierarchical, and mixed effects modeling, was chosen for several reasons. First, GCM accounts for correlation in error terms resulting from the same individuals being measured over time in panel data. Second, GCM does not require a balanced panel for analysis. Third, these models provide precise estimates of individual growth over time and have greater power to detect predictors of individual differences in growth. Fourth, GCM allows both for time-varying and time-invariant covariates to be included in models. Finally, although GCM typically requires at least three repeated measures per individual, these models have been successfully used to analyze instances when some individuals have only one or two observations, but the majority of respondents have three or more (Curran, Obeidat, and Losardo 2010; Raudenbush and Bryk 2002).

I use gender-stratified GCM with random intercepts and slopes to estimate the likelihood that HPGs practice as health professionals over time. Several model iterations were tested, and likelihood-ratio tests identified the models with best goodness of fit. The final GCM chosen for this analysis include random time effects, which allow a respondents’ intercept and slope to vary randomly according to a normal distribution. In each GCM, the key time variable (years since graduating from health professions school) is interacted with social categories of HPGs (i.e., years*race*place of education). Models including the interaction between years since graduating and social categories separately (i.e., years*race; years*gender; and years*place of education) are shown in Appendix A, Figures A1–A6.

The basic equation for the GCM is:

$$\begin{array}{ll}\text{Level 1:}& Y_{it}={\beta }_{0i}+{\beta }_{1i}\mathrm{*}t+{\beta }_m{z}_{it}^{\prime }+e_{it}\end{array}$$

$$\begin{array}{ll}\text{Level 2:}& {\beta }_{0i}={\gamma }_{00}+{\gamma }_{0k}{C}_{1i}+{\gamma }_m{x}_i^{\prime }+u_{0i}\\ & {\beta }_{1i}={\gamma }_{10}+{\gamma }_{1k}{C}_{1i}+u_{1i}\\ & {\beta }_m={\gamma }_{m0}\end{array}$$

where in Level 1, $Y_{it}$ represents the likelihood that HPGs work as health professionals. ${\beta }_{0i}$ represents the respondent-specific random intercept (i.e., the baseline likelihood that respondents work as health professionals at graduation), ${\beta }_{1i}$ represents the respondent-specific random slope for time $t$ (i.e., the person-specific rate of change over time), $z_{it}^{\prime }$ represents a vector of time-variant sociodemographic, immigration, and family covariates, and $e_{it}$ represents the unexplained random variation in the likelihood of working as a health professional for respondent $i$ at time $t$. In Level 2, ${\mathrm{\gamma }}_{00}$ is the fixed portion of the baseline likelihood of working as a health professional, ${\mathrm{\gamma }}_{0k}$ reflects the variation of an HPG’s likelihood of working as a health professional across the interaction of race and place of education, ${\mathrm{\gamma }}_m$ indicates the effect of a vector of time-invariant sociodemographic, family, and immigration covariates, $x_i^{\prime }$, and $u_{0i}$ represents the respondent-specific random component.

In the second equation of Level 2, ${\mathrm{\gamma }}_{10}$ represents the fixed part of the growth rate of the likelihood of working as a health professional over time $t$, and ${\mathrm{\gamma }}_{1k}$ represents the varied likelihood that HPGs across race and place of education work as health professionals over time. The residual, $u_{1i}$ represents the random component in the rate of temporal change in the likelihood of working as a health professional that is not explained by the social characteristics of HPGs. In the final equation of Level 2, ${\mathrm{\gamma }}_{m0}$ represents the vector of time-variant covariates.

Results

Descriptive Results

Table 1 presents characteristics among HPGs in the 2003–2013 NSCG. Most HPGs identified their race as white (61%), followed by Asian (23%), and URM (i.e., Black, Latinx, Indigenous, or bi/multiracial) (17%). Women were more likely to identify as Asian (28% versus 20%) or URM (19% versus 15%). Around a quarter of HPGs obtained their education abroad, and women were more likely than men to be foreign educated (28% versus 25%). Fifty-two percent of HPGs were U.S.-educated and white, 10% were U.S.-educated and Asian, and 11% were U.S.-educated and URM. Thirteen percent of HPGs were foreign-educated and Asian, 9% were foreign-educated and white, and 5% were foreign-educated and URM. Women were more likely to report non-white racial identities; fifteen percent of women were foreign-educated and Asian compared to 11% of men. Similarly, 14% of women were U.S educated and URM compared to 10% of men.

Most HPGs were married (83%), although women were less likely to be married than men (75% versus 87%, respectively) and more likely to be never married (15% versus 7%, respectively). Despite being less likely to be married at the time of survey response, women were slightly more likely than men to have minor children (57% versus 55%).

Sixty-four percent of HPGs reported they were U.S.-born or U.S. citizens (64%). Thirteen percent had held either temporary training visas or were permanent U.S. residents as their first U.S. citizenship status, respectively. Four percent had held a temporary dependent visa (such as an H-4), 3% had held a temporary resident/work visa, and 3% had held some other type of work visa. Strikingly, women were four times as likely to have had a dependent visa upon arrival in the U.S. compared to men (8% versus 2%, respectively).

Inequalities in HPGs’ Career Trajectories

To what extent does the likelihood that HPGs work as health professionals vary by gender, race, and place of education? Appendix A, Table 2 shows predicted probabilities from the figures presented in the main text.

Table 2:

Fully-adjusted linear growth curve models estimating the likelihood of practicing as a health professional

Characteristics	Men			Women
	β		SE	β		SE
Place of education and race (ref = U.S.-educated white)
U.S.-educated Asian	−0.130	***	(0.034)	−0.042		(0.038)
U.S.-educated URM	−0.062	†	(0.036)	0.013		(0.038)
Foreign-educated white	−0.191	***	(0.049)	−0.241	**	(0.069)
Foreign-educated Asian	−0.409	***	(0.046)	−0.451	***	(0.054)
Foreign-educated URM	−0.408	***	(0.062)	−0.243	**	(0.094)
Place of education and race x Years since graduation (ref = U.S.-educated white)
U.S.-educated Asian x Years since graduation	0.006	**	(0.002)	0.000		(0.003)
U.S.-educated URM x Years since graduation	0.003	†	(0.002)	−0.001		(0.002)
Foreign-educated white x Years since graduation	0.005	**	(0.002)	0.006	†	(0.003)
Foreign-educated Asian x Years since graduation	0.013	***	(0.002)	0.010	***	(0.002)
Foreign-educated URM x Years since graduation	0.010	***	(0.002)	0.005		(0.004)
Years since graduation	−0.003	*	(0.001)	0.002		(0.002)
Birth cohort	0.001		(0.001)	0.005	***	(0.001)
Marital status (ref = Married)
Never married	0.004		(0.022)	0.012		(0.024)
Divorced, widowed, or separated	−0.038		(0.023)	0.001		(0.027)
Children (ref = has no children)	0.032	**	(0.010)	0.019		(0.015)
First U.S. citizenship status (ref = U.S.-born/U.S. citizen)
Permanent U.S. resident	−0.036		(0.023)	0.016		(0.029)
Temporary resident/work visa	−0.095	**	(0.038)	−0.089		(0.056)
Temporary training visa	−0.171	***	(0.023)	−0.151	***	(0.034)
Temporary dependent visa	−0.114	**	(0.040)	−0.171	***	(0.036)
Other type of visa	−0.078	*	(0.039)	−0.152	*	(0.060)
Intercept	−1.638		1.658	−9.498	***	2.283
Observations		8273			4412
N		3787			2179
Likelihood-ratio test		1996			1085
Random-effects parameters
Years since graduation variance	0.000	***	0.000	0.000	***	(0.000)
Constant variance	0.954	***	0.010	0.056	***	(0.007)
Covariance (years since graduate, constant)	−0.001	***	0.000	0.001	***	(0.000)
Residual variance	0.061	***	0.001	0.075	***	(0.003)

Source: 2003–2013 NSCG

Note.

^***p<0.001,

^**p<0.01,

^*p<0.05,

^†p<0.1

Figures 3 and 4 show the career trajectories of men and women HPGs by place of education and race. Several key findings emerge. First, inequalities in career trajectories by place of education are larger than inequalities across race. For men and women, U.S.-educated HPGs have a higher likelihood of working as health professionals both upon graduation and over time, compared to foreign-educated HPGs. Figure 3 shows that the likelihood of practicing as a health professional varies between 74% for U.S.-educated Asian men to 87% for U.S.-educated white men upon graduation. This likelihood increases to 80% for U.S.-educated Asian and 82% for U.S.-educated white men at 20 years post-graduation. Among foreign-educated men, however, this range is shifted downward: 46% for foreign-educated Asian and URM men, and 68% for foreign-educated white men upon graduation. At 20 years post-graduation, this likelihood increases to 60% for foreign-educated URM men and 74% for foreign-educated white men.

Figure 3: Likelihood of men HPGs practicing as health professionals over time, by race and place of education.

Figure 4: Likelihood of women HPGs practicing as health professionals over time, by race and place of education.

For women (Figure 4) the likelihood of practicing as health professionals ranges from between 68% for U.S.-educated Asian women and 74% for U.S.-educated URM women upon graduation, to between 73% for U.S.-educated Asian women and 77% for U.S.-educated white women at 20 years post-graduation. For foreign-educated women, the likelihood ranges from between 27% for foreign-educated Asian women and 48% for foreign-educated white and URM women upon graduation, to between 52% for foreign-educated Asian women and 64% for foreign-educated white women 20 years post-graduation.

Secondly, Asian, and to some degree, URM, HPGs generally report lower likelihoods of working as health professionals compared to white HPGs within place of education groups. Among men, U.S.-educated white and URM graduates show no statistically significant difference in the likelihood of working as a health professional upon graduation, and their likelihood of working as a health professional remains relatively stable and high over time. U.S.-educated Asian men, however, experience a lower likelihood of practicing as a health professional upon graduation (74%) compared to white men, but experience steep increases in the likelihood of working as a health professional over time; by 20 years post-graduation, the racial gap in working as a health professional is nearly closed between U.S.-educated Asian (80%) and white men (82%).

Whereas foreign-educated white men have a 68% chance of practicing as a health professional upon graduation, the corresponding likelihood for foreign-educated Asian and URM HPGs is 46% for each. Over time, however, there is some evidence among foreign-educated graduates that whiteness serves as a credential to entering the health professions; notably, at 20 years post-graduation, foreign-educated white men have a 74% chance of practicing as a health professional, almost converging to the likelihoods of all racialized groups of U.S.-educated men. Although both foreign-educated Asian and URM men experience increases in the likelihood of practicing as a health professional over time, they are still significantly less likely than all other groups of men to practice as health professionals at 20 years post-graduation (67% for foreign-educated Asian and 60% for foreign-educated URM men).

For women, there is no statistically significant difference in the likelihood of practicing as a health professional among U.S.-educated white and URM HPGs upon graduation. However, U.S.-educated Asian women have a significantly lower likelihood of practicing as health professionals upon graduation (68%). Over time, U.S.-educated Asian women experience the steepest gains in working as health professionals; at 20 years post-graduation, U.S.-educated Asian women have a 73% chance of practicing as a health professional. However, this likelihood is still lower than that of U.S.-educated white (77%) and U.S.-educated URM (75%) women.

Compared to U.S.-educated women, foreign-educated women experience greater disadvantages in working as health professionals. Upon graduating, foreign-educated white and URM women have similar likelihoods of matriculating to a health professions career (48% for each). Both groups experience increases in the likelihood of practicing as health professionals over time, although the slope of increase is steeper for foreign-educated white (64%) compared to foreign-educated URM women (61%). Strikingly, foreign-educated Asian women experience the greatest disadvantage in becoming health professionals compared to all other groups of women; foreign-educated Asian women have only a 27% chance of practicing as a health professional upon graduating. At 20 years post-graduation, foreign-educated Asian women have only a 52% chance of practicing as a health professional.

Factors Shaping HPGs’ Career Trajectories

How do factors such as sociodemographic, family, and immigration characteristics contribute to the career trajectories of HPGs? Table 2 presents coefficients for gender-stratified GCM estimating the likelihood of practicing as a health professional among HPGs by their place of education and race. These analyses assess how different sets of covariates contribute to the likelihood that men and women HPGs practice as health professionals.

Table 2 shows that although marital status appears to have no statistically significant impact on the likelihood of working as a health professional for either men or women, having children increases the likelihood of practicing as a health professional for men (coeff = 0.032, p < 0.01). Comparatively, having children has no statistically significant impact on the likelihood that women work as health professionals. Immigration factors (i.e., initial U.S. citizenship status) appear to shape the career trajectories of both men and women. For men, compared to HPGs born in the U.S./those with U.S. citizenship, HPGs who arrived in the U.S. on any type of temporary visa are less likely to work as health professionals; those who arrived in the U.S. with a training visa are least likely to work as health professionals (coeff = −0.171, p < 0.001). For women, having arrived on a training visa also reduces the likelihood of working as a health professional (coeff = −0.151, p < 0.001), and having arrived on a temporary dependent visa is associated with a substantially lower likelihood of working as a health professional (coeff = −0.171, p < 0.001).

Discussion

This article explored the occupational trajectories of HPGs, investigating their likelihoods of matriculating to a health professions career across axes of race, gender, and place of education. Analyses show that HPGs’ career trajectories are both racialized and gendered, in addition to being shaped by place of education. Foreign-educated Asian HPGs often experience the greatest career disadvantages, corroborating qualitative evidence that “Asianness,” combined with “foreignness,” serve as significant barriers toward entering a U.S. health professions career (Bhatt 2013). Despite the “model minority” stereotype around Asian academic success (Xu and Lee 2013), this article’s findings suggest that such educational success may not inherently translate into career success. Analyses show that Asian foreign-educated men and women HPGs have the lowest likelihoods of matching to a health professions career relative to other groups of HPGs. Given that Asians are perceived to be “overrepresented” in STEM and health care (Min and Jang 2015), stereotypes around Asian academic success may inform the hiring process, and residency directors may be more likely to filter Asians out of their application pool, particularly if they are perceived to be less competitive applicants. For foreign-educated Asian HPGs in particular, this discrimination may also intersect with stereotypes that a foreign education makes individuals less well prepared for a U.S. health professions career, or that foreign graduates are too culturally dissimilar from the “average” American patient to make for qualified health professionals (Jenkins 2020).

Findings that HPGs who once held dependent visas reported lower likelihoods of obtaining a health professions career may indicate that foreign-educated HPGs are significantly disadvantaged by immigration barriers that make it difficult for them to obtain careers in the U.S. For women specifically, this disadvantage may result from the disproportionate likelihood of having an H-4 dependent visa as one’s first U.S. citizenship status. Indeed, GCM estimates suggest that HPGs who held temporary visas, including training and dependent visas, had lower likelihoods of pursuing a health professions career. Given the persistence of educational homogamy (Schwartz and Mare 2005), foreign-educated women HPGs may be spouses of HPGs or other high-skilled immigrants, although the current analyses are unable to capture this directly. Consequently, if an immigrant couple must prioritize one spouses’ labor market opportunities, the other spouse may become a secondary, “dependent” migrant. This may generate “cumulative disadvantage” (Purkayastha 2003) and may explain why foreign-educated Asian women appear to experience the lowest likelihoods of working as health professionals over time. Furthermore, given the expensive and time-consuming process to obtain licensure in the U.S., HPGs who initially held an H-4 visa may be particularly unlikely to go through this licensure process even after they are eligible to adjust their work status.

Determinations about foreign educational quality may appear “color-blind” on the surface, through narratives that all foreign health professions schools are subjected to equal scrutiny, empirically, there are fewer schools in Latin America and Africa that are recognized by the U.S. health professions licensing boards and educational commissions on foreign health professions graduates (e.g., the ECFMG for medicine) (Schut 2022). This means fewer URM HPGs, who largely migrate from Latin America and Africa, are eligible to pursue U.S. licensure. Moreover, even HPGs from “recognized” Latin American and African health professions schools may be required to complete more complex, time-consuming pathways toward licensure in the U.S. compared to other groups of foreign HPGs (e.g., European HPGs) (ECFMG 2024).

Despite the significant disadvantage faced by foreign-educated URM graduates in pursuing a U.S. health professions career, analyses show that U.S.-educated URM HPGs experience relatively high likelihoods of working as health professionals. This finding suggests that foreign education is a larger barrier toward entering the health professions than race. Why might this be? Research shows there is significant selection among Black and Latinx Americans into professional degrees. Structural racism impacts Black and Latinx individuals early on in the educational pipeline, reducing the likelihood that these groups benefit from primary, secondary, and undergraduate education that prepares them for graduate or professional degree programs in the first place (Wingfield and Chavez 2020).

However, for URM individuals who successfully matriculate to and complete health professions educations, their path to a health professions career may be relatively streamlined. That is, findings from this article suggest that U.S.-educated URM HPGs experience some degree of advantage, once passing the hurdle of matriculating to and graduating from health professions programs. However, despite their advantage in obtaining (and maintaining) a health professions career, URM HPGs are potentially more likely to experience meso-level and micro-level discrimination; in medicine, for example, URM physicians are disproportionately tracked into less competitive and lower paying medical specialties and are more likely to experience racism from fellow health care workers and patients (Filut, Alvarez, and Carnes 2020; Wingfield and Chavez 2020).

Several analytical limitations warrant future research attention. First, the category “URM” lumps several different racialized groups together (e.g., Black, Latinx, Indigenous, etc.), masking the diverse experiences of those who are represented by these categories. This limitation prevents a more nuanced, intersectional analysis of HPGs’ career trajectories. Second, analyses presented in this article are descriptive in nature, and therefore do not imply causality. Third, analyses are unable to determine the precise mechanisms driving inequalities in matriculating to a health professions career or in occupational attainment. In lieu of qualitative information that would provide deeper insight into factors driving these inequalities, covariate analyses attempt to further disentangle the factors driving inequalities in matriculation to a health professions career.

Moreover, two key variables - English ability and educational prestige – are not captured by the NSCG, yet these factors have potentially relevant implications for HPGs’ career trajectories. English language ability is a critical factor determining HPGs’ ability to work in the health professions in the U.S, however this variable is not directly captured in the NSCG survey. Moreover, lacking information about the prestige of a respondents’ specific educational institution may bias results, as it is possible that some respondents are not able to pursue a U.S. health professions career in the first place because their institution is not accredited by a U.S. health professions licensing entity. Still, despite this limitation, it is unclear that any racialized or gender group of foreign-educated respondents examined in this article would be more or less likely to have received an education from a non-accredited foreign university; exploring stratification by specific country of education, however, would be more vulnerable to the omission of this variable, as certain countries are perhaps more likely to have accredited universities than others.

Conclusion

The U.S. health care system faces numerous challenges in the 21^st century, including population aging (Barrett and Michael 2022) and widening geographic and racial-ethnic inequities in health and mortality (Rhubart and Santos 2023; Elo et al. 2019). Moreover, health professional shortages across multiple fields (Sabato et al. 2023; Couch et al. 2023; Kirch and Petelle 2017) continue to drive concerns among clinicians and policymakers across the U.S. Increasing the proportion of foreign-educated, non-white, and women health professionals may help ameliorate population health inequalities, as a diverse health professions workforce may also increase rapport between the health professions and minoritized individuals, thus working toward addressing centuries of mistrust that has resulted from abuses of the health care system perpetuated onto marginalized communities. Such efforts can work to fundamentally improve the delivery of equitable health care to an every-diversifying U.S. population.

Highlights.

• Place of education shapes health professions graduates’ (HPGs’) career outcomes.

• Foreign-educated HPGs face barriers toward becoming U.S. health professionals.

• Career disadvantage is magnified for those with minoritized race and gender statuses.

• Initial visa status is a key factor shaping foreign-educated HPGs’ career outcomes.

• The U.S. health professions continue to face challenges toward diversifying.