Social Marginalization

Abstract

Social marginalization, which MeSH classifies under social behavior, is the social process by which a person or group is pushed to the edge of society and denied full participation in the rights, resources, and relationships available to its members. This article traces that idea from Park's marginal man through the nursing and public-health literature, and distinguishes marginalization as a structural process from the subjective states it produces. It develops the psychological core — the need to belong and the pain of exclusion — then the mechanisms that convert difference into disadvantage: Link and Phelan's stigma model, Crenshaw's intersectionality, Meyer's minority-stress model, and Marmot's social gradient in health. Three interactive demonstrations let readers assemble stigma from its components, see how marginalized categories compound at their intersection, and trace the social gradient across a status hierarchy.

Keywords: social marginalization, stigma, intersectionality, minority stress, social gradient

Every society sorts its members, and some are sorted outward — to the margin, where participation is partial, resources are thin, and belonging is conditional. Social marginalization is the name for that process and for the position it produces. It is not a feeling, though it produces characteristic feelings; it is not a single act, though it is built from countless acts. It is a durable social arrangement in which whole categories of people are held at the edge of the institutions that distribute work, health, safety, and standing.

What makes marginalization a topic for cognitive and behavioural science, rather than for sociology alone, is the machinery that runs between the structural fact of being held at the edge and its consequences for the person. Exclusion is registered by a brain built to belong; stigma operates through stereotypes and the categories we think with; the chronic stress of occupying a devalued position reshapes attention, emotion, and physiology. These are the mechanisms this article develops, after first fixing the definition that the surrounding vocabulary routinely blurs.

What Social Marginalization Is

Social marginalization is the social process whereby a person or group is denied full participation in, or access to, the rights, resources, and opportunities generally available to the members of a society — a pushing to the edge, or margin, of social life. The definition is structural: it names a relation between a group and the institutions that distribute goods and standing, not a trait of the marginalized person. A useful early synthesis in the nursing literature defined marginalization through its properties — peripheralization from the centre of power, a sense of being outside, differential treatment, and vulnerability — and argued that the concept is valuable precisely because it keeps attention on the process rather than on supposed deficits of the excluded (Hall et al., 1994).

The idea has a clear point of origin in social science. Park, writing in 1928, described the marginal man — the person who lives at the meeting point of two cultures and belongs fully to neither, carrying the instabilities of both — and in doing so founded the study of marginality as a social-psychological condition rather than a mere matter of geography or poverty (Park, 1928). Park's figure is now dated in its particulars, but the move it made endures: it located the problem in the person's position between social worlds, in the relation, not in the person.

Three nearby constructs must be held apart from marginalization itself. Stigma is the mark of devaluation attached to an attribute; it is one of the chief engines of marginalization but is analytically distinct from the process it drives. Social isolation is the objective thinness of a person's social ties, and ostracism is the act of being ignored and excluded; both are often consequences of marginalization but can occur without it, and marginalization can occur without either. Keeping the process (marginalization), its chief cause (stigma), and its common sequelae (isolation, exclusion, distress) separate is the first discipline the topic demands, because collapsing them is how interventions end up aimed at the wrong level — at the feelings of the excluded rather than the arrangements that exclude them.

Exclusion and the Need to Belong

The reason marginalization wounds is that humans are built to belong. Baumeister and Leary's need-to-belong hypothesis holds that the drive to form and maintain lasting, positive relationships is a fundamental human motivation — as basic as hunger — and that its frustration produces a broad range of ill effects on emotion, cognition, and health (Baumeister & Leary, 1995). If belonging is a need rather than a preference, then being held at the margin of the group is not a disappointment but a deprivation, and the system should respond to it as it responds to other deprivations: with distress and a drive to correct it.

The response is, in part, literally painful. Eisenberger, Lieberman, and Williams scanned people as they were excluded from a simple ball-tossing game and found that social exclusion activated the dorsal anterior cingulate cortex — a region central to the distressing component of physical pain — with activity tracking self-reported distress (Eisenberger et al., 2003). The finding gave concrete form to the intuition buried in ordinary language, that rejection hurts, and suggested that the machinery of social pain was built atop the older machinery of physical pain.

Even the mildest exclusion bites. Williams's research programme on ostracism shows that being ignored and excluded — even briefly, even by strangers, even by a computer the participant knows is scripted — reliably threatens four fundamental needs: belonging, self-esteem, control, and meaningful existence (Williams, 2007). The speed and indiscriminateness of the effect imply a crude but sensitive detection system, one that sounds the alarm at the first sign of exclusion and sorts out whether the alarm was warranted only afterward. Wesselmann and Williams extend this to the group level, arguing that inclusion and exclusion in groups range along a continuum from full social life to a kind of social death, with marginalization sitting in the painful middle region where a person is neither fully admitted nor cleanly expelled (Wesselmann & Williams, 2017). That middle region — partial, conditional, revocable membership — is the lived texture of marginalization, and it is more corrosive than clean exclusion precisely because it never resolves.

Stigma as the Engine of Marginalization

If marginalization is the process, stigma is its principal motor. The sociological study of stigma begins with Goffman, who defined it as an attribute that is deeply discrediting and that reduces its bearer from a whole and usual person to a tainted, discounted one, shifting the analytic focus from the attribute itself to the spoiled social relationship it creates (Goffman, 1963). The most influential modern account builds directly on that relational move: Link and Phelan reconceptualized stigma as the co-occurrence of five components within a relation of power (Link & Phelan, 2001). First, people distinguish and label human differences that matter socially. Second, dominant cultural beliefs link the labelled persons to undesirable characteristics — stereotyping. Third, labelled persons are placed in distinct categories so as to accomplish a separation of us from them. Fourth, labelled persons experience status loss and discrimination that lead to unequal outcomes. And critically, the whole sequence depends on access to social, economic, and political power — without it, the same labelling and stereotyping do not produce the systematic exclusion that stigma names.

The power condition is what makes the model more than a list. A marginalized group may hold stereotypes about the dominant group, may label and separate in return, but lacks the power to turn those cognitions into status loss and discrimination at scale. Stigma, on this account, is not symmetric; it is a function of who can make their categories stick to other people's lives. This is why the model travels so well from mental illness — Link and Phelan's home case — to race, disability, sexuality, and poverty: in each, the same five components operate, and in each the outcome depends on the power differential between the categorizers and the categorized.

Intersectionality: Marginalization at the Crossroads

Marginalization rarely runs along a single axis. Crenshaw's concept of intersectionality, developed from the legal treatment of Black women's experiences, shows that a person situated at the convergence of several marginalized categories faces forms of exclusion that an analysis attending to one axis at a time cannot see (Crenshaw, 1991). A doctrine that recognizes race discrimination and sex discrimination as separate wrongs may offer no remedy to someone harmed specifically as a Black woman — harmed in a way that is neither the race discrimination experienced by Black men nor the sex discrimination experienced by white women, but a distinct configuration produced at the intersection.

The theoretical payload is that marginalized identities are not additive. Treating a person's disadvantages as a sum of independent categories — as though one could tally the race penalty and the gender penalty and add them — misses the qualitatively new positions created where categories meet, and systematically renders the most marginalized the least visible in both data and policy. The Worked Example below makes the arithmetic of this concrete, showing how even a generous multiplicative model of compounding disadvantage understates what a single-axis analysis reports, and why intersectional positions demand their own measurement rather than inference from the margins.

The Social Gradient and the Health Toll

Marginalization is not merely unpleasant; it is a determinant of health, and the evidence is among the most robust in social epidemiology. Marmot's work on the social determinants of health established the social gradient: health outcomes worsen in a stepwise fashion as one descends the social hierarchy, so that the effect is not a cliff at the bottom separating the deprived from everyone else but a slope running across the whole range, each grade less healthy than the one above it (Marmot, 2005). The gradient implies that it is relative position — status, control, participation — and not only absolute material want that drives health, which is exactly what a theory of marginalization would predict. Link and Phelan give this its sharpest theoretical form in their account of social conditions as fundamental causes of disease: a marginalized position persistently reproduces health disadvantage across whatever the proximate risk factors of the day happen to be, because the group's lack of access to money, knowledge, power, and beneficial social connections lets it be harmed by each new mechanism in turn (Link & Phelan, 1995). That is why eliminating one specific pathway rarely closes the gap — the fundamental cause simply expresses itself through the next. Two decades on, Marmot's review of health equity in England found the gradient intact and, on several measures, steepening, with improvements in life expectancy stalling for the most marginalized (Marmot, 2020).

The mechanism connecting marginalized status to ill health is chronic stress, and the most developed model is Meyer's minority stress theory. Meyer argued that members of stigmatized minority groups face an excess burden of stress — from experienced discrimination, from the expectation of rejection, from the concealment of identity, and from internalized stigma — and that this excess, layered on ordinary life stress, explains the elevated rates of mental disorder observed in marginalized populations (Meyer, 2003). The model matters because it locates the cause in the social environment rather than in the group: the excess disorder is the signature of marginalization, not of the marginalized identity itself.

The population evidence bears this out. Pascoe and Smart Richman's meta-analysis found that perceived discrimination has a significant negative effect on both mental and physical health, operating through heightened stress responses and through unhealthy behaviours adopted to cope (Pascoe & Smart Richman, 2009). Schmitt and colleagues, pooling 328 effect sizes, confirmed that perceived discrimination is reliably associated with poorer psychological well-being across disadvantaged groups, and — relevant to the rejection-identification idea — that identification with the marginalized group can buffer, though not erase, the harm (Schmitt et al., 2014). Williams, Lawrence, and Davis, reviewing the racism-and-health literature, document how marginalization operates at structural, cultural, and individual levels to produce racial inequities in health, and argue that these pathways are now well enough understood to be acted upon (Williams et al., 2019). And the downstream endpoint is mortality itself: the thinned social relationships that marginalization produces predict earlier death at a magnitude rivalling classic clinical risk factors (Holt-Lunstad et al., 2015).

Discussion

The study of marginalization has achieved something unusual: it has connected a macro-level social process to a micro-level neural and physiological response, by way of a well-specified set of psychological mechanisms. Exclusion is detected by a brain that treats belonging as a need; stigma converts difference into disadvantage through labelling, stereotyping, separation, status loss, and discrimination; and the chronic stress of occupying a marginalized position leaves a measurable mark on mental and physical health that tracks a gradient across the whole social hierarchy. Few topics in social science can point to a causal chain this complete, running from the structure of a society down to the expression of disease.

The conceptual spine of the field is the insistence that marginalization is a property of a relation, not of a person. This is not a rhetorical nicety; it dictates where interventions should aim. If the excess disorder in a marginalized group is the signature of its social environment, as the minority-stress and social-gradient models hold, then treating the distress of marginalized individuals without altering the arrangements that marginalize them addresses a symptom and leaves the cause. The power condition in Link and Phelan's model makes the same point from another direction: because stigma depends on the power to make one's categories stick, the lever that matters most is the distribution of power, not the content of anyone's beliefs.

The limitations are real. Much of the evidence is observational, so reverse causation — ill health eroding social position rather than the other way round — cannot be fully excluded, though the longitudinal gradient studies and the experimental exclusion work together make a strong case against it being the whole story. Measurement is harder than for isolation or loneliness: marginalization is a process observed through its proxies, and perceived discrimination, the most-used proxy, is a self-report that conflates the objective exposure with the person's appraisal of it. And the intersectional critique cuts against the field's own methods, warning that the most marginalized positions are precisely the ones the standard single-axis designs are worst at seeing.

Current Directions

Three strands define the current research front. The first is the move from documenting marginalization's health toll to specifying its pathways with enough precision to intervene. Williams and colleagues' synthesis of racism and health is the exemplar: it organizes the evidence into structural, cultural, and individual levels and identifies the research still needed to convert understanding into effective policy, marking a shift from whether to how and what to do (Williams et al., 2019). The systems approach now urged for social connection more broadly — matching the response to the specific deficit rather than treating connection as a single quantity — applies with equal force to marginalization, where structural, relational, and cognitive levers are distinct (Holt-Lunstad, 2018).

The second is the hardening of the social-gradient evidence into a policy instrument. Marmot's ten-years-on review turned the gradient from a descriptive finding into an accountability measure, tracking whether a society's commitments to health equity are met and naming the stall in progress for the most marginalized as a political choice rather than an inevitability (Marmot, 2020). The third is the integration of the group-exclusion literature with the broader social-death framing, which situates marginalization on a continuum of inclusion and relocates it from a binary of in-or-out to a graded, revocable membership whose very uncertainty is part of its harm (Wesselmann & Williams, 2017). Across all three, the open questions are the same: how to measure positions that single-axis designs miss, how much of the health gradient the known stress pathways actually carry, and which levers — structural, relational, cognitive — move the outcomes that matter.

Common Misconceptions

Marginalization is just another word for social isolation.

They are different. Marginalization is a structural process that pushes a group to the edge of social participation; social isolation is the objective thinness of an individual's social ties. Marginalization often causes isolation, but a person can be isolated without being marginalized, and marginalized without being isolated (Hall et al., 1994).

Stigma is simply holding a negative belief about a group.

A belief is not enough. In Link and Phelan's model, stigma requires labeling, stereotyping, separation, status loss, and discrimination to co-occur within a relation of power; absent the power to turn categories into unequal outcomes, the same stereotype does not produce the systematic exclusion that stigma names (Link & Phelan, 2001).

Multiple disadvantages simply add up.

Intersectionality shows they do not. A person at the convergence of several marginalized categories occupies a distinct position that is not the sum of its parts, and single-axis analysis that tallies each category separately renders that position invisible rather than capturing it (Crenshaw, 1991).

Poor health at the margin is just a matter of material poverty.

The social gradient runs across the whole hierarchy, not only at the bottom, so relative position and the control and participation that come with it — not absolute deprivation alone — drive health. Each step down the ladder is less healthy than the step above, even well clear of poverty (Marmot, 2005).

Demo 1 — The Five Components of Stigma, and Power

Link and Phelan (2001) recast stigma as the co-occurrence of five components — labeling, stereotyping, separation, status loss and discrimination — within a power relation. Toggle each component and the power differential; marginalization emerges only when all five align and the categorizers hold power.

2 of 5 components active. Link and Phelan argue stigma is their convergence, not any one alone — turn on the rest (and power) to see marginalization emerge.

Demo 2 — Intersectionality: Compounding and Invisibility

Crenshaw (1991) showed that a person at the convergence of several marginalized categories faces exclusion a single-axis analysis cannot see — it reports only the one largest category and treats the rest as separate problems. Toggle the axes and watch the gap open between what a single-axis lens records and the compounded lived position.

A single-axis analysis records a burden of 17.0%, attending to the largest category alone. The lived position stacks all 2 to 28.6%, leaving 11.6% of the burden unseen — the person at the crossroads falls through the gap between single-axis frames.

Demo 3 — The Social Gradient in Health

Marmot's Whitehall studies found that health follows a gradient across the whole social hierarchy, not a cliff edge at the bottom: each step down the ladder is less healthy than the step above, even far clear of poverty. Set a position on the hierarchy and read off the stylised relative risk — the line never flattens into a safe zone.

At the Middle rung, relative health risk sits at about 1.50× that of the top. The gradient runs unbroken down every rung — marginalization harms health by position in the hierarchy, not only by absolute deprivation at the very bottom.

Worked Example

Intersectionality's central claim is that marginalized categories do not simply add. The point can be made precise with the same odds ratio (OR) arithmetic used throughout social epidemiology — and, instructively, even the generous assumption that disadvantages combine multiplicatively still exposes how badly single-axis analysis understates the burden at an intersection.

Suppose some adverse outcome — say, being turned away from housing — has a baseline probability of p = 0.10 for an unmarginalized applicant. The odds are

odds = p ÷ (1 − p) = 0.10 ÷ 0.90 ≈ 0.111

Let marginalization on the basis of gender multiply the odds by an OR of 1.4, and marginalization on the basis of race multiply them by an OR of 1.6. A single-axis analysis examines one at a time. Looking only at gender,

p_gender = (0.111 × 1.4) ÷ (1 + 0.111 × 1.4) ≈ 0.135, or 13.5%

Looking only at race,

p_race = (0.111 × 1.6) ÷ (1 + 0.111 × 1.6) ≈ 0.151, or 15.1%

But a person at the intersection carries both. Under the multiplicative model the joint odds ratio is 1.4 × 1.6 = 2.24, so

odds' = 0.111 × 2.24 ≈ 0.249  →  p_joint = 0.249 ÷ 1.249 ≈ 0.199, or 19.9%

The baseline risk of 10% becomes 19.9% at the intersection — a 9.9-percentage-point increase, far more than either single axis reports (13.5% or 15.1%). A policy that attends to gender discrimination or race discrimination, one axis at a time, sees at most a 5-point problem; the person living at the crossroads faces a 10-point one. And this is the conservative reading: intersectionality's deeper claim is that the joint position can be qualitatively distinct — a harm that is neither the gender penalty nor the race penalty but a third thing produced where they meet — so that even the multiplicative 19.9% can understate the lived reality (Crenshaw, 1991). The third interactive demonstration lets the reader set the axis-specific odds ratios and the baseline and read off both the single-axis and the joint risks directly.

Key Researchers

Nyla R. Branscombe

(University of Kansas). Developed the rejection-identification model and meta-analyzed the well-being costs of perceived discrimination, clarifying when group identification buffers the harm of marginalization. Faculty page · Google Scholar

Kimberlé W. Crenshaw

(UCLA School of Law and Columbia Law School). Introduced intersectionality, showing that people at the convergence of multiple marginalized categories experience forms of exclusion that single-axis analysis renders invisible. Faculty page · Google Scholar · Wikipedia

Bruce G. Link

(University of California, Riverside). Reconceptualized stigma as the co-occurrence of labeling, stereotyping, separation, status loss, and discrimination within a power relation, giving marginalization a testable mechanism. ORCID · Faculty page · Google Scholar · Wikipedia

Michael Marmot

(University College London). Demonstrated the social gradient in health through the Whitehall studies, establishing that position in the social hierarchy — not just absolute deprivation — drives health inequality. Faculty page · Google Scholar · Wikipedia

Ilan H. Meyer

(Williams Institute, UCLA School of Law). Built minority stress theory, showing how the chronic stress of stigma, prejudice, and discrimination experienced by marginalized minorities produces excess mental-health burden. ORCID · Faculty page · Google Scholar · Wikipedia

Robert E. Park

(University of Chicago; 1864–1944). Coined the marginal man in 1928 to describe the person living at the intersection of two cultures and belonging fully to neither, founding the social-psychological study of marginality. Wikipedia · Britannica

Jo C. Phelan

(Columbia University). Co-developed the modern conceptualization of stigma and fundamental-cause theory, explaining why marginalized groups face persistent health disadvantage across changing proximate causes. Wikipedia · Wikidata

Glossary

Discrimination.

Unequal treatment and outcomes imposed on a labelled group; the fifth component of stigma in Link and Phelan's model and the point at which cognition becomes disadvantage.

Intersectionality.

Crenshaw's concept that the convergence of several marginalized categories produces distinct forms of exclusion not captured by analysing each category on its own.

Labeling.

The first component of stigma in Link and Phelan's model: selecting a human difference and marking it as socially significant, the step that turns ordinary variation into a basis for exclusion.

Marginal man.

Park's term for the person living at the meeting point of two cultures and belonging fully to neither; the founding image of marginality in social science.

Minority stress.

Meyer's model of the excess, chronic stress borne by stigmatized minorities — from discrimination, expected rejection, concealment, and internalized stigma — that explains their elevated rates of mental disorder.

Need to belong.

Baumeister and Leary's hypothesis that the drive for lasting, positive relationships is a fundamental human motivation whose frustration damages emotion, cognition, and health.

Odds ratio.

The factor by which the odds of an outcome are multiplied in one group relative to another; distinct from, and not equal to, a risk ratio.

Ostracism.

Being deliberately ignored and excluded by one or more others; a common act of marginalization that threatens belonging, self-esteem, control, and meaningful existence.

Perceived discrimination.

A person's report of being treated unfairly on the basis of group membership; the most-used proxy for exposure to marginalization in health research.

Rejection-identification model.

Branscombe's account of how identifying with a marginalized group can partly buffer the well-being costs of the discrimination that group faces.

Social death.

Wesselmann and Williams's framing of the far pole of a continuum of group membership, where a person is wholly excluded; marginalization occupies the painful middle of that continuum.

Social gradient.

Marmot's finding that health worsens stepwise down the social hierarchy, so relative position — not only absolute deprivation — determines health.

Social marginalization.

The social process by which a person or group is denied full participation in the rights, resources, and opportunities available to members of a society.

Stigma.

The co-occurrence of labeling, stereotyping, separation, status loss, and discrimination within a relation of power; the principal engine of marginalization.

Frequently Asked Questions

What is social marginalization?

It is the social process by which a person or group is pushed to the edge of society and denied full participation in the rights, resources, and relationships generally available to its members. It is a structural arrangement, a relation between a group and a society's institutions, rather than a personal trait or a mere feeling, though it reliably produces distinctive feelings and health effects.

How is marginalization different from stigma?

Stigma is the mark of devaluation attached to an attribute, and it is the principal cause of marginalization, but the two are distinct. In the dominant model, stigma is the co-occurrence of labeling, stereotyping, separation, status loss, and discrimination within a relation of power; marginalization is the exclusion from full social participation that this stigma produces.

Why does being marginalized affect physical health?

Occupying a marginalized position is a source of chronic stress, and chronic stress damages the body over time. The evidence shows a social gradient, in which health worsens stepwise down the social hierarchy, and meta-analyses link perceived discrimination to poorer mental and physical health, operating through heightened stress responses and the behaviours people adopt to cope.

What does intersectionality add to the idea of marginalization?

It shows that marginalized categories are not additive. A person at the convergence of several categories (Crenshaw's founding example was Black women) faces a distinct configuration of exclusion that is not the sum of the separate penalties, and that single-axis analysis, attending to one category at a time, systematically fails to see.

What is the social gradient in health?

It is Marmot's finding that health outcomes worsen in a stepwise fashion all the way down the social hierarchy, not only among the poorest. Because each grade is less healthy than the one above it, the gradient implies that relative position (the status, control, and participation that come with rank) drives health, not absolute material deprivation alone.

What is minority stress?

Minority stress, in Meyer's model, is the excess burden of stress borne by members of stigmatized minority groups, arising from experienced discrimination, the expectation of rejection, the concealment of identity, and internalized stigma, that is added to ordinary life stress and explains the elevated rates of mental disorder observed in marginalized populations.

Can identifying with a marginalized group help?

To a degree. The rejection-identification model and the meta-analytic evidence indicate that identifying strongly with one's marginalized group can buffer some of the well-being cost of the discrimination that group faces, by supplying belonging, meaning, and collective support. It softens the harm rather than removing it, and does not address the marginalization itself.

Where should interventions aim?

Because marginalization is a property of a social relation rather than of the marginalized person, the models imply that treating only the distress of marginalized individuals addresses a symptom. The power condition in the stigma model and the environmental locus of minority stress both point to the arrangements that marginalize, namely the distribution of power, resources, and participation, as the level at which the cause actually sits.

References

Baumeister, R. F., & Leary, M. R. (1995). The need to belong: Desire for interpersonal attachments as a fundamental human motivation. Psychological Bulletin, 117(3), 497–529. https://doi.org/10.1037/0033-2909.117.3.497

Crenshaw, K. (1991). Mapping the margins: Intersectionality, identity politics, and violence against women of color. Stanford Law Review, 43(6), 1241–1299. https://doi.org/10.2307/1229039

Eisenberger, N. I., Lieberman, M. D., & Williams, K. D. (2003). Does rejection hurt? An fMRI study of social exclusion. Science, 302(5643), 290–292. https://doi.org/10.1126/science.1089134

Goffman, E. (1963). Stigma: Notes on the management of spoiled identity. Prentice-Hall.

Hall, J. M., Stevens, P. E., & Meleis, A. I. (1994). Marginalization: A guiding concept for valuing diversity in nursing knowledge development. Advances in Nursing Science, 16(4), 23–41. https://doi.org/10.1097/00012272-199406000-00005

Holt-Lunstad, J., Smith, T. B., Baker, M., Harris, T., & Stephenson, D. (2015). Loneliness and social isolation as risk factors for mortality: A meta-analytic review. Perspectives on Psychological Science, 10(2), 227–237. https://doi.org/10.1177/1745691614568352

Holt-Lunstad, J. (2018). Why social relationships are important for physical health: A systems approach to understanding and modifying risk and protection. Annual Review of Psychology, 69, 437–458. https://doi.org/10.1146/annurev-psych-122216-011902

Link, B. G., & Phelan, J. (1995). Social conditions as fundamental causes of disease. Journal of Health and Social Behavior, 35(Extra Issue), 80–94. https://doi.org/10.2307/2626958

Link, B. G., & Phelan, J. C. (2001). Conceptualizing stigma. Annual Review of Sociology, 27, 363–385. https://doi.org/10.1146/annurev.soc.27.1.363

Marmot, M. (2005). Social determinants of health inequalities. The Lancet, 365(9464), 1099–1104. https://doi.org/10.1016/S0140-6736(05)71146-6

Marmot, M. (2020). Health equity in England: The Marmot review 10 years on. BMJ, 368, m693. https://doi.org/10.1136/bmj.m693

Meyer, I. H. (2003). Prejudice, social stress, and mental health in lesbian, gay, and bisexual populations: Conceptual issues and research evidence. Psychological Bulletin, 129(5), 674–697. https://doi.org/10.1037/0033-2909.129.5.674

Park, R. E. (1928). Human migration and the marginal man. American Journal of Sociology, 33(6), 881–893. https://doi.org/10.1086/214592

Pascoe, E. A., & Smart Richman, L. (2009). Perceived discrimination and health: A meta-analytic review. Psychological Bulletin, 135(4), 531–554. https://doi.org/10.1037/a0016059

Schmitt, M. T., Branscombe, N. R., Postmes, T., & Garcia, A. (2014). The consequences of perceived discrimination for psychological well-being: A meta-analytic review. Psychological Bulletin, 140(4), 921–948. https://doi.org/10.1037/a0035754

Wesselmann, E. D., & Williams, K. D. (2017). Social life and social death: Inclusion, ostracism, and rejection in groups. Group Processes & Intergroup Relations, 20(5), 693–706. https://doi.org/10.1177/1368430217708861

Williams, K. D. (2007). Ostracism. Annual Review of Psychology, 58, 425–452. https://doi.org/10.1146/annurev.psych.58.110405.085641

Williams, D. R., Lawrence, J. A., & Davis, B. A. (2019). Racism and health: Evidence and needed research. Annual Review of Public Health, 40, 105–125. https://doi.org/10.1146/annurev-publhealth-040218-043750