Measures of Racism, Sexism, Heterosexism, and Gender Binarism for Health Equity Research: From Structural Injustice to Embodied Harm—An Ecosocial Analysis

Racism. Sexism. Heterosexism. Gender binarism. Together, they comprise intimately harmful, distinct, and entangled societal systems of self-serving domination and privilege that structure the embodiment of health inequities. Guided by the ecosocial theory of disease distribution, I synthesize key features of the specified “isms” and provide a measurement schema, informed by research from both the Global North and the Global South. Metrics discussed include (a) structural, including explicit rules and laws, nonexplicit rules and laws, and area-based or institutional nonrule measures; and (b) individual-level (exposures and internalized) measures, including explicit self-report, implicit, and experimental. Recommendations include (a) expanding the use of structural measures to extend beyond the current primary emphasis on psychosocial individual-level measures; (b) analyzing exposure in relation to both life course and historical generation; (c) developing measures of anti-isms; and (d) developing terrestrially grounded measures that can reveal links between the structural drivers of unjust isms and their toll on environmental degradation, climate change, and health inequities. 37 A nn u. R ev . P ub lic H ea lth 2 02 0. 41 :3 762 . D ow nl oa de d fr om w w w .a nn ua lr ev ie w s. or g A cc es s pr ov id ed b y 19 6. 40 .2 4. 24 3 on 1 1/ 16 /2 0. S ee c op yr ig ht f or a pp ro ve d us e. PU41CH03_Krieger ARjats.cls March 17, 2020 8:42


INTRODUCTION
Structural: explicit and nonexplicit unjust "rules of the game" (laws, policies, and rules), as well as area-based or institutional legacies and indicators of injustice Interpersonal: discriminatory practices and beliefs directed at persons in the targeted social group by persons in the privileged social group Internalized: self-subordinating practices and beliefs by persons in the targeted social groups Links to political economy, political sociology, and political ecology How does the privileged group benefit economically, and how is the targeted group harmed economically, by the unjust ism? How does the privileged group benefit socially from, and how is the targeted group socially harmed by, the unjust ism? How is the biophysical ecological context of the privileged group protected, and how is that of the targeted group harmed, by the unjust ism?

Contrasting views of anti-isms
Rejection of biological essentialism and claims of biologically based superiority or inferiority Rejection of definitions of social normality or deviance based on biological essentialism Respect for human rights: recognition of "the inherent dignity and of the equal and inalienable rights of all members of the human family" (186) and that the abrogation of dignity and denial of these rights both constitutes injustice and can produce health inequities Recognition (to varying degrees, as informed by insights from political sociology, political economy, and political ecology) that unjust societal "rules of the game" are established by people in the privileged group with private vested interests in preserving, if not expanding, their power, privilege, property, and resources, primarily at the expense of those in the targeted group(s) a. Forming nouns with the sense 'belief in the superiority of one-over another'; as racism, sexism, speciesism, etc. b. Forming nouns with the sense 'discrimination or prejudice against on the basis of-'; as ageism, bodyism, heightism, faceism, lookism, sizeism, weightism, etc.
These sets of isms likewise provide guidance for action, in this case, unjust practices variously implemented by states, nonstate institutions, and individuals within their societal contexts.

Tenets Relevant to Analyzing Isms and Health Inequities: Ecosocial Theory of Disease Distribution
The second point is that much as isms involving injustice are about beliefs, they are fundamentally rooted in real-world material conflicts over power and property (13,67,74)-as tied to the workings of diverse political economy isms, e.g., capitalism (including neoliberalism), socialism,

Structural
Explicit "rules of the game": laws, policies, and rules

Racism
Measures of one US state-level 1960s civil rights measures in relation to infant mortality (2) National measures of US states' Jim Crow status (legal racial discrimination) and its mid-1960s abolition in relation to concurrent and subsequent infant death rates (101), premature mortality (102), and breast cancer estrogen receptor status (106,107) US federal policies requiring racial discrimination in neighborhood development and home ownership loans ["redlining"; 1930s to mid-1960s (170)] in relation to current rates of within-city urban violence (79), self-rated health (130), cancer stage at diagnosis Nonexplicit "rules of the games": laws, policies, and rules

Racism
US state-level measures of gerrymandering (partisan manipulation of voting district boundaries without using explicitly racial criteria) and proximity of populations of color to US Superfund sites (high-hazard toxic waste sites) (88) Algorithms for dementia status in relation to differential accuracy by race/ethnicity in a US nationally representative data set (60) Sexism Local measures of publicly available toilets (including urinals) by gender across public sites (airports, train stations, shopping centers) in four cities in South Africa (179) National measures of work-hour thresholds (no explicit mention of gender) and gender inequities in mental health in Australia (48) Heterosexism and gender binarism c National study of noninclusion of questions about sexual orientation and gender identity in health surveillance data in the United States (159) (Continued) Area-based or institutional legacies and indicators of injustice Note: The predominant area-based metrics for unjust isms used in health equity research pertain to city-level measures of residential racial segregation (99,192,196); new metrics for spatial polarization enable the use of more granular multilevel measures of racial, economic, and racialized economic segregation (108,112,115). The examples provided below are in addition to measures of residential segregation.

Racism
Local measures of inequitable spatial distributions, by race/ethnicity, of US neighborhood hazards, e.g., warehouses (in the Los Angeles combined statistical area) (202) Indices of US county-and state-level measures of black versus white differences in diverse social and economic indicators in relation to birth outcomes (39, 189), fatal police shootings (134), and myocardial infarction (125) Area-based US Internet data in relation to infant health outcomes, e.g., using Internet data on hate speech (36) or implicit bias surveys (137, 152)

Sexism
Indices of national measures of the gender ratio (women:men) for diverse social and political indicators in relation to femicide rates in 61 countries (156) Measures of gender inequality in (a) political representation in US state legislatures, (b) earnings, labor force participation, and poverty rates, along with data on (c) percent of state population composed of religious conservatives and (d) percent of women who live in a county without an abortion provider, in relation to chronic health conditions, self-rated health, and physical functioning (77) National measures of women's representation in a country's parliament and mortality rates across 155 countries (126) Heterosexism and gender binarism c National measure of voting results for a plebiscite on same-sex marriage in relation to mental health among the LGB population in Australia (163) US community data on explicit and implicit HIV prejudice in relation to the psychological and physical well-being of people with HIV, in 42 communities in New England (137) Individual: exposure to ism Explicit self-report Racism Note: Individual self-reports of exposure are the most widely used measures in research on racial discrimination and health (behavioral, mental, and physical) (99,121,157,196). The reviews and examples listed below are chosen to illustrate the current use of such measures worldwide.
Reviews of national and community-based population self-report data on exposure to racial discrimination (as target or witness) in relation to diverse health outcomes among diverse US populations of color and racialized populations in other countries (12,89,99,157,158,196). National and community-based population self-report data in studies on racial, ethnic, and caste discrimination in relation to health status and health behaviors, respectively, in Brazil (160), Europe (46), Israel (45), India (84), and New Zealand (72)

Sexism
Note: Individual self-report data are the most widely used measures in worldwide research on sexism and health, albeit primarily in relation to gender-based violence, which sometimes includes sexual harassment (86,133,188). The examples listed below are chosen to illustrate current health equity studies, which employ less commonly used measures of gender discrimination and sexism and which also focus explicitly on sexual harassment.
www.annualreviews.org • Measuring Unjust "Isms" and Health Equity Selected examples of measures used in health equity research b Self-report data on gender discrimination and women's drug use in the United States (32) Self-report data on sexist ideology and men's control over women's reproductive health decisions in the United States (164) Self-report data on US women's experiences of sexism and sexual risk behaviors (41) Self-report data on sexual harassment and measured physical health in the United States (103,183) and self-rated health and self-reported psychological health in the United States (28,71) and globally (40) Self-report data on gender equality in couples in relation to self-rated health in Sweden (178) Heterosexism and gender binarism c Note: Individual self-report data are the most widely used measures in worldwide research on anti-LGBT discrimination and health (behavioral, mental, and physical) (141,142). The examples listed below are chosen to illustrate the current use of such measures.
Self-report data on hate crime victimization among LGBT persons in the United States and mental health (29) Self-report data on bisexual-specific stressors in relation to psychological distress and suicidality in the United States (131) Self-report data on acquisition of gender-affirming identity documents for US transgender persons and smoking status (175) Implicit Note: Only four published health equity studies appear to have used the IAT to assess people's sense of themselves as targets versus perpetrators of racial discrimination (33,100,113,114), one of which also tested an IAT for being a target versus perpetrator of gender discrimination (33); these measures were associated with smoking status and measured blood pressure (100,114) Experimental Racism Experimental manipulation, in US studies, of exposure to racial discrimination versus nonracial stressors (e.g., using visual vignettes, written materials, or confederates), analyzed in relation to measured blood pressure, cardiovascular reactivity, and cortisol levels (6, 76, 78, 119, 171) Sexism Experimental manipulation, in US studies, of exposure to sexism (e.g., using confederates and written materials), analyzed in relation to cortisol levels and cardiovascular reactivity (51, 185)

Heterosexism and gender binarism c
No experimental health equity studies identified Individual: internalization Explicit

Racism
Self-report data on internalized negative racial group attitudes in relation to cardiovascular disease among African American men (37) Self-report data on internalized racism and mental health among African Americans, US-born Caribbean blacks, and foreign-born Caribbean blacks in the United States (143)

Sexism
No health equity studies identified Heterosexism and gender binarism c Self-report data on internalized heterosexism or homophobia in relation to mental and physical health in the United States (132,166,180) (Continued)  Table 3. c In this table, the categories of "heterosexism" and "gender binarism" are considered together because studies on sexual and gender minorities (including lesbian, gay, bisexual, queer, transgender, and gender fluid persons) often include persons across this range of social categories, even as some studies may focus exclusively on transgender persons (regardless of sexual orientation) or solely on persons who identify as belonging to a sexual minority (regardless of gender identity).
communism, feudalism, communitarism, etc. At issue are societal structures linking political governance, economic rules, and rules governing who is allowed which rights to live, produce, and reproduce (or not) where, in which community and family structures, and with which rights to use, own, and extract resources from which land, water, forests, and other species (13,21,57,66,67,117,162). The ecosocial theory of disease distribution-which I first proposed in 1994 and have elaborated since-accordingly conceptualizes health inequities in relation to power, levels, life course, historical generation, biology, and ecosystems (91,93,(95)(96)(97) (Figure 1). Its four core conjoined constructs pertain to (a) embodiment; (b) pathways of embodiment; (c) cumulative interplay of exposure, susceptibility, and resistance across the life course and across levels; and (d) accountability and agency. As the ecosocial construct of embodiment has clarified from the start (91,95,97), we jointly live our myriad legally and socially ascribed categories and identities, which notably can vary, among individuals, over time and by context (25,86,122,151,190). We are not a member of a particular racial/ethnic group one day, have a particular gender identity on another, and on still another day have a particular sexual orientation (whether identity, behavior, or desire)-and the same holds for our social class, immigration status, and where we live and work (91,92,95,97,109). Instead, we live embodied-and our bodies each and every day biologically integrate each and every type of unjust, and also beneficial, exposure encountered, at each and every level (96,97). The construct of intersectionality, a term originating in the legal and social sciences (67,74), similarly conveys the socially structured entanglement of multiple types of injustice, albeit without a focus on the pathways and processes of literal biological embodiment (91)(92)(93)(94)(95)(96)(97)(98)(99).

Analyzing Isms Requires Engaging with Biology and Challenging Biological Essentialism
The third point is that, in accord with ecosocial theory (91,97), a structural systems-oriented understanding of isms necessarily challenges two faulty premises of biological essentialism: (a) the spurious construct of biological race, which wrongly presumes humanity can be cleaved into www.annualreviews.org • Measuring Unjust "Isms" and Health Equity distinct biological "races" that intrinsically differ in health status and other traits (169,203), including their human needs for rights and resources to thrive (49); and (b) the erroneous beliefs that gender and sexual orientation are fixed, binary, and biologically determined by sex-linked biology (54, 61,86,94,161,190). Also to contest are (a) conventional legal definitions that restrict "discrimination" to mean unfair treatment based on allegedly innate immutable and discrete characteristics, e.g., "race," "sex," or "sexual orientation" (3,174) and (b) the conventional-and contradictorycausal inference suppositions that "race," "sex," and "sexual orientation" are either "unmodifiable" characteristics (and thus cannot be "causes") or else, unto themselves, are intrinsically the "causes" of health disparities (105). Rather, the very idea that these are fixed attributes that determine who people are-and at the same time are themselves causally sufficient to explain group differences in health (without reference to the causal perpetrators, both systemic and individual)-is part of the belief system of the very isms at issue. Notably, other forms of discrimination recognized and prohibited by law include discrimination on the basis of religious beliefs and political views (3), which, while potentially influenced by familial context, are surely neither innate nor immutable.

Producing Knowledge for Action and Accountability, Not to Prove Injustice
The final point is that, to reiterate what I stated 20 years ago, the reason to study how injustice harms health is not to prove that injustice is wrong, since it is, by definition (92). Instead, the

CORE CONSTRUCTS OF ECOSOCIAL THEORY
Embodiment refers to how we literally incorporate, biologically, the material (biophysical) and social world in which we live. The concept of pathways of embodiment refers to how processes of embodiment are shaped simultaneously by histories of societal arrangements of power and property and by constraints and possibilities of our evolved biology in ecological context, involving gene expression and not just gene frequency. Cumulative interplay of exposure, susceptibility, and resistance across the life course and across levels refers to the importance of timing and accumulation of, and responses to, embodied exposures, taking into account both individuals' life course social and biological development and the historical generation into which they have been born, as well as the levels at which both the exposures-and susceptibility and resistance to these exposures-are occurring (i.e., global, national, regional, area or group, household, individual). The concept of accountability and agency refers to both (a) the institutions and persons responsible for generating or perpetuating health inequities and (b) the public health researchers for the theories used to explain or ignore these injustices. Together, these conjoined constructs underscore why measures of isms must take into account (a) structural, legal, and policy contexts, as tied to changes in political and economic power (at different levels) to uphold or challenge the isms, and (b) timeframes of both exposures and their impacts, in relation to both the biophysical and social worlds and the relevant etiologic periods within and across generations.

PATHWAYS OF EMBODIMENT
Pathways of embodiment especially relevant to exposure to unjust isms involve adverse exposures to (95-99) social and economic deprivation; exogenous hazards (e.g., toxic substances, pathogens, and hazardous conditions); social trauma (e.g., discrimination and other forms of mental, physical, and sexual trauma); targeted marketing of harmful commodities (e.g., tobacco, alcohol, other licit and illicit drugs); inadequate or degrading medical care; and degradation of ecosystems, including degradation linked to the alienation of Indigenous populations from their lands. These pathways, and multiple isms, occur-and are embodied-concurrently and interactively.
reasons are to deepen understanding of how injustice shapes population health, for whose benefit at whose expense; to contest narratives that naturalize inequities; and to generate evidence for accountability. After all, if people have created unjust societal systems and structures, so too can people challenge these systems and structures and advocate instead for human rights, health equity, and ecological sustainability (13,21,67,74,97,155,162,181). Table 2 provides a schema for delineating different types of measures of the specified isms-and anti-isms-relevant for health equity research, with attention to issues of study design. Including anti-isms matters because, from the causal standpoint of inference to the best explanation (105,124), if adverse isms are posited to drive health inequities, then anti-isms should aid health equity. Table 3 in turns provides selected examples of health equity studies that have used such measures, at different levels, from both the Global South and the Global North; Supplemental Table 3 provides the technical details of the metrics these studies employed.

Structural Measures: Isms and Anti-Isms
The term structural-as applied to isms-has, in the past decade, begun to surface in a small but growing number of health equity studies (8,14,58,77,99,196) (Table 3; Supplemental Table 1). "Structural" is meant to convey the "totality of ways" in which societies foster unjust isms through "mutually reinforcing systems" that can variously involve "housing, education, employment, earnings, benefits, credit, media, health care, and criminal justice" (8, p. 1453) and also marriage law, family law, identity documents, and property ownership (13,21,67). These laws, rules, and practices in turn "reinforce discriminatory beliefs, values, and distribution of resources" (8, p. 1453). Structural injustice thus exists above and beyond-while also shaping-unjust interpersonal interactions and individual beliefs aligning with the isms at issue ( Table 1).
When it comes to the measurement of structural isms, however, more precision is required. Clarifying the different types and meanings of "structure" is crucial for accurate identification of etiology, accountability, and agency, and thus avenues for action.
Explicit laws and rules. One class of structural measures pertain to explicitly stated "rules of the game" (13, p. 11) (Tables 2, 3; Supplemental Table 3). These rules can be articulated in explicit laws, policies, or algorithms, for which either government or nongovernmental institutions bear responsibility, thereby pointing to potential legal liability, claims for remedies or reparations, and alternative legislation (13,181,187). Well-known example include, in the United States, legally sanctioned white supremacy, black enslavement, and subsequent legal racial discrimination (i.e., Jim Crow laws) (1, 4, 56, 151); legal and genocidal "racial hygiene" in Nazi Germany (informed by U.S. Jim Crow laws) (56,151); legal apartheid in South Africa (which also drew on U.S. Jim Crow laws) (56); the long-standing legal disenfranchisement of women in most countries around the world [women were first legally granted the right to vote, nationally, in New Zealand in 1893, and, most recently, in 2015, in Saudi Arabia, albeit only at the local level (184)]; national laws in 72 countries (as of 2017) that prohibit same-sex relationships (34); and a new US federal policy, under legal challenge, that prohibits transgender persons from enlisting in the military (123).
Two potential limitations involving these types of structural measures concern inadequate measure of exposure and also correlated exposures. First, health equity studies using these types of metrics typically focus on the presence of such rules; less attention, if any, is given to assessing whether or how these rules are (selectively) enforced, such that important variations in actual exposures may be overlooked (13,66,181). Second, health equity studies typically focus on rules for solely one type of ism, and while this approach may sometimes be warranted, often the ideological orientation of state policies means that rules for related adverse isms travel together (13, 35); a potential corrective could be inclusion, as covariates, of indices that capture multiple related policies (13,35).
Nonexplicit laws and rules. Other times, the laws or rules deliberately do not name the targeted groups, e.g., because it is illegal to do so, but nevertheless discriminate adversely, by design or effect (Tables 2, 3; Supplemental Table 3). Two classic examples from the United States, designed to target African Americans without the relevant laws naming this group explicitly, include (a) the War on Drugs (1,75) and its links to mass incarceration, both of which have been a focus of health equity research (140,195), and (b) voter suppression, whether by (now outlawed) poll taxes and literacy tests or by (still legal) partisan gerrymandering, restrictive voter ID laws, and limits on times and places to vote (4,151), which by contrast has received scant attention in health equity research (88). A newer form of nonexplicit rule-based discrimination is discrimination-by-algorithm, whereby algorithms applied to biased data yield biased predictions, as has occurred with machine learning, big data, and institutional decisions regarding employment, criminal sentencing, mortgage approval, and medical care (63,148,168). In both cases, the antiism measures needed to counter these types of injustice require explicit use of ism-conscious rules to protect the targeted groups (1,4,70,148,168,187).

Area-based or institutional nonrule measures.
Other measures appearing in the health equity research literature that are also referred to as structural do not directly assess actual rules (whether explicit or implicit), but instead comprise area-based or institutional measures regarding legacies and indicators of inferred rule-based structural injustice (Tables 2, 3; Supplemental  Table 3). Most commonly, such measures employ population-based data and quantify the difference (relative or absolute) between the targeted group and the privileged group in relation to one or more specified social outcomes, e.g., group differences in income, education, employment, home ownership, residential segregation, political representation, policing, mass incarceration, access to transportation, access to reproductive health services, proximity to toxic waste sites or polluting industries or facilities, etc. (Tables 2, 3; Supplemental Table 3). The measures used, to date, in health equity research have been highly heterogeneous, rendering comparison of results across studies difficult. They also have thus far been limited by reliance on data that are publicly available and which can be stratified by the relevant social groups. Also potentially of interest, but not yet used in quantitative population-based health equity research, are area-based metrics for exposure to public monuments that support unjust isms (e.g., in the United States, monuments supporting the Confederacy) and also, as anti-isms, the removal of these monuments as well as their replacement by public art that exposes the unjust isms (23,82,136,172).
Additional area-based measures newly being used in health equity research seek to measure social norms regarding isms by employing survey or Internet data ( Tables 2, 3; Supplemental  Table 3). Caution is required, however, given (a) well-known biases (especially underreporting) in self-report survey data on adverse discriminatory attitudes (10,92,99,145,154,157) and (b) fastrising fraudulent postings online, especially on hot-button isms, to foment political divisions (30,148). An alternative approach, not subject to these biases but apparently not yet used in populationbased health equity research, involves audit studies, i.e., investigations that send identical resumes-or identically dressed people of similar body build-to potential employers or landlords; in such studies, the applicants differ solely in their racial/ethnic or gendered appearance or couple status (e.g., single, same-sex couple, heterosexual couple) (118,145,154,167). Another newly employed approach in health equity research is to use area-based implicit bias data (137,152); however, an important caveat concerns how selection bias can affect who voluntarily seeks out and takes such Web-based tests (e.g., mainly highly educated and affluent persons) (18,22,201). Together, these types of area-based measures provide evidence solely about what is termed statistical discrimination (i.e., statistical evidence of group differences inferred to arise due to unjust processes but absent explicit evidence of motivation), and while such evidence may not meet legal standards that require proof of intent (3,174), it can still usefully inform and galvanize action for equity.

Historical contingency and choice of structural measures and comparison groups.
Which types of structural measures can be used depend on societal and historical context. A key consideration is whether the type of injustice at issue is legally under contest.
Consider, first, the case of adverse racial residential segregation in the United States. Its contours were shaped by US federal and state laws and policies, translated into official maps, that mandated neighborhood residential racial segregation and disinvestment both in and outside Jim Crow states from the 1930s to the mid-1960s, after which such legal racial discrimination was nationally outlawed (55, 135,170). Given the post-1965 lack of variation by place or time in laws permitting racial discrimination in housing, health equity studies have relied primarily www.annualreviews.org • Measuring Unjust "Isms" and Health Equity on area-based indicators that are inferred to reflect the impact of historical legal adverse racial residential segregation and also current illegal racial discrimination. Such indicators have variously pertained to neighborhood composition (e.g., racial/ethnic, income) and conditions (e.g., physical environment, housing stock), intergenerational transfers of wealth and impoverishment as tied to home equity, and political power as tied to voting districts and voting facilities (4,8,58,99,135,157,192,196). Only with the recent unearthing and digitization of the original 1930s maps (55, 170) has it become possible for health equity researchers to start to investigate the long-term impacts of the original legal discrimination on both neighborhood trajectories and health inequities (79,130,135 By contrast, present-day legal and policy conflicts over, say, same-sex marriage and civil unions (19,34), and over the rights of transgender and third-gender persons (e.g., travesties in South America, hijras in India, and muxes in Mexico) (19,34,47,190,193), render it feasible to analyze the contemporaneous health impacts of the explicit rules at issue. The same holds, in relation to sexism, for current conflicts over laws and policies regarding sexual harassment (81,86,149,204), gender-based violence (24,86,133,188), and sexual and reproductive rights and health, e.g., regarding access to contraception, abortion, or inclusive health care for sexual or gender minorities, or policies regarding pregnancy and parental leave (13,24,53,59,62,67,68,86,144,181).
Regardless of the type of structural measure employed, careful thought needs to be given to both the comparison groups (101,173) and the etiologic period (14,97). Also at issue is the extent to which people are legally able to move in order to benefit from or avoid policy changes (13,135). Contrasts can include (a) comparisons of the health of the targeted groups before and after the law or policy goes into effect in their polity; (b) within-polity changes in the magnitude of health inequities comparing the targeted and privileged group, before or after the policy change; and (c) cross-sectional comparisons of either the health status of the targeted group or the magnitude of the health inequities between the targeted group and the privileged group across polities that differ in the relevant law or policy. Short-term (or even cross-sectional) temporal comparisons may be apt for outcomes with a short etiologic period (e.g., psychological distress); if longer etiologic periods are involved, time lags must be considered (13,97).
An especially important consideration concerns the long-term life course and intergenerational realities of embodied harm. What may seem "long ago," from a legal or policy standpoint, can be but an instant in terms of biological generations (97,104,107). Jim Crow laws, for example, were decreed illegal in the United States in 1964 (1,4,151). However, anyone born in the United States who, in 2019, is age 55 or older was born when Jim Crow was the law of the land, and any child of theirs is the child of a person who potentially experienced this all-encompassing form of structural racism (2,83,101). The handful of health equity studies on Jim Crow and its abolition demonstrate its relevance to rates of infant mortality, premature mortality, and current risk of more lethal types of breast cancer (2,70,83,101,102,106,107). It is this understanding of embodied history that health equity researchers can bring to their choice of measures of isms.

Individual: Exposure to Isms and Anti-Isms
Despite the importance of structural conditions shaping people's exposures, the most commonly used measures in health equity research on isms and health nevertheless are individual-level explicit self-reports of exposure to discrimination (whether as a target or a witness) ( Tables 2, 3;  Supplemental Tables 1-3). Other individual-level metrics, used less frequently, include implicit measures of exposure to discrimination and measures of internalized oppression. Explicit self-report. As discussed at length in other recent reviews (12,89,98,99,121,141,142,157,158,196), in the case of racism, heterosexism, and gender binarism, most studies rely on self-report explicit measures. These measures typically assess if the exposure has occurred; in which context (e.g., at work, in a public setting, at school, from the police or in the courts, online, getting medical care, etc.); with what intensity, chronicity, and duration of exposure (including age at onset); and also how the exposed person responds (e.g., from resistance to internalization) ( Tables 2, 3; Supplemental Tables 1-3). By contrast, for sexism, measures more typically pertain to phenomena that can be conceptualized as actualizing beliefs and practices emblematic of sexism, e.g., sexual harassment (especially at work, but also in public settings) (31,103,150,204), gender-based violence, and control over sexual and reproductive rights and health (24,77,86,99,133,139,161,188) (Tables 2, 3; Supplemental Tables 1-3).
As with any self-report measures, caveats abound ( Table 2). First, self-reported measures can express only what people are willing and able to say (92,99), as shaped by fear, social desirability, trauma, internalized oppression, and lack of knowledge (10,33,98,145,154,157). Moreover, members of the privileged group may claim reverse discrimination when confronted by potential diminishment of prior privilege (99,144,165,191,200); health equity studies that include members of the privileged groups could-but rarely do-employ relevant measures of resentment [e.g., racial resentment (191) or hostile sexism (164)] to aid interpretation of results. Self-reported experiences of discrimination, moreover, are typically analyzed solely as a psychosocial stressor (12,89,92,98,99,121,141,142,157,158,196), even as some exposures (e.g., unfair policing) pose physical, not just psychosocial, risks (1,52,75,98).
Pointing further to limitations of a solely psychosocial focus, consider the case of wage discrimination, which can harm people's health by causing material (economic) harm (92,99,155,174). Notably, in the United States, the workplace is among the most commonly self-reported sites for racial, gender, and anti-LGBT (lesbian, gay, bisexual, transgender) discrimination (99,147). It is difficult, however, for employees to know if wage discrimination is occurring. Why? The answer is not simply social norms. Rather, it is currently legal in only 15 of the 50 US states, plus the District of Columbia, for employees to share information about their wages (198). Not surprisingly, several recent high-profile wage discrimination suits in the United States have been started by plaintiffs who accidentally discovered the pay stubs of fellow workers who were paid higher wages for the same job (120,199). Also not surprisingly, research has found that states with legally protected transparency about wages have the lowest within-job-title gender wage inequities (85). Thus, where rules govern what individuals cannot know, it is likely that structural measures will outperform self-report data as a measure of exposure.
Additional concerns about self-report measures, discussed at length in other literature (11,27,99,121,157,196), pertain to whether questions should refer explicitly to discrimination [as per the Experiences of Discrimination measure (110,111)] or to unfair treatment [as per the Everyday Discrimination Scale (196,197)], or should solely name the behaviors at issue. Even in research on sexual harassment and gender-based violence, for which the consensus is to focus on behaviors (with no explicit mention of abuse or sexism), controversies abound over which behaviors, enacted by whom, against whom, should count, and how best to counter fears and internalized blame that contribute to underreporting (133,188,204). Another problematic issue is that the discrimination and health literature contains numerous studies that employ solely one-or two-item instruments, despite well-known psychometric flaws affecting reliability and validity of instruments with so few items (12,44,99,141,142).
Despite these challenges, studies from around the world have found that self-reported experiences of diverse types of discrimination and hate crimes rooted in unjust isms are robustly associated with psychological distress and are more complexly associated with adverse health behaviors, biomarkers, and disease outcomes (99,128,157,158,193,196). They are, accordingly, an important complement to structural measures, providing insights into personal experiences that lie on the pathways of embodiment connecting structural drivers to embodied harm and population health inequities. Implicit measurement. Another option is afforded by the implicit association test (IAT), a robustly validated and widely accepted cognitive time-reaction test well-described in the psychological literature, which was designed to address the well-known problems affecting the use of self-report data to measure socially sensitive topics (10,44,64,146). Our preliminary studies (33,100,113,114) of the new IATs we developed to assess someone's sense of being a target of discrimination found that these IATs were not associated with social desirability (as commonly is the case for self-report data), they did not display the well-known person-group discrimination discrepancy that affects self-report measures (which previously had been hypothesized to be an artifact of social desirability), and they were also associated with smoking and with elevated blood pressure (33,114). Work on these measures remains in its infancy.
Experimental. Less commonly, psychosocial research on health impacts of isms has also used experimental techniques, exposing participants either to vignettes or to controlled situations designed to correspond to experiences shaped by these isms, to assess their impact on psychological state, physiological arousal, or diverse biomarkers (44, 99,128,145,158). Despite well-known questions regarding the generalizability of such results to impacts of real-world psychosocial stressors (44, 145,154), such studies can potentially yield insights into sets of psychological and biological mechanisms that may be involved in behaviorally and biologically translating unjust isms into embodied harm (44, 99,116,128).

Individual: Internalized Isms and Anti-Isms
Finally, health equity studies addressing racism, heterosexism, and gender binarism, but rarely sexism, have employed measures of what may be termed internalized isms, e.g., internalized oppression or stigma (Tables 2, 3; Supplemental Table 3). At issue is the extent to which persons in the targeted group consciously or unconsciously accept the privileged group's view of them as inferior, deviant, or shameful, thus potentially increasing the targeted group's risk of harm to health via psychosocial pathways, both behavioral and biological (99,128,157,193,196). Measures of anti-isms ( Table 1) conversely can assess, at the individual level, the extent of pride (and perhaps defiance) that persons in the targeted group feel about their group and their analysis of who discriminates against them and why (99,157); also germane, at the population level, are public demonstrations about, and the appearance of, public slogans regarding, say, LGBT pride or "black is beautiful" (74,127,176,187,190). As in the case of the self-report measures of exposure to discrimination, measures of internalized oppression and pride typically rely on explicit self-report instruments (with all the same caveats about self-report data) (44, 99,157) or, less commonly, implicit measures (10, 38).

CHALLENGES AHEAD: KNOWLEDGE FOR ACTION TO END HEALTH INEQUITIES DUE TO UNJUST ISMS
As should be evident, the field of health equity research is ripe to improve measures of unjust isms, and anti-isms, to generate more powerful knowledge for understanding and for action to advance health equity. As part of moving the field beyond the preponderance of work using individual-level psychosocial measures, I would recommend the following actions: Expand the scope of validated structural measures by type and timeframe to encompass historical injustices whose embodied health harms manifest within and across generations (97,101,102,104) and to integrate exposure (as our bodies do every day) across diverse types of isms (67,74,92,95,99,162), in addition to clarifying whether the structural measures are rule based or indicators of inferred rules. Expand complementary use of anti-ism measures, spanning from structural to internalized, drawing on new work for health justice (21,69,155,162). Expand the scope of comparative health equity research (13,15), taking advantage of crossnational (or subnational) and cross-temporal variations in exposures to unjust isms and corrective anti-isms. Develop terrestrially grounded measures that can reveal links between the structural drivers of unjust isms and their toll on environmental degradation, the climate crisis, and health inequities (16,21,57,87,117,162,182).
I encourage you, the reader, to join in the collective efforts to take on these challenges, as one small part of the multifaceted, multisectoral work needed to advance health equity on a threatened planet.

DISCLOSURE STATEMENT
The author is not aware of any affiliations, memberships, funding, or financial holdings that might be perceived as affecting the objectivity of this review.