Prejudice is a feeling or attitude held by an individual toward members of another group, while racism is a broader phenomenon that includes not just personal attitudes but also the institutional policies, cultural norms, and structural arrangements that produce racial inequality. A person can hold prejudiced views without those views being embedded in systems of power, and a system can produce racially unequal outcomes even when the individuals operating it harbor no conscious prejudice at all. The distinction matters because each concept points to a different problem and demands a different kind of solution.
What Prejudice Actually Looks Like in the Mind
Prejudice, at its core, is an attitude. It can be a conscious belief (“I think group X is inferior”) or it can operate below awareness as a gut-level reaction a person would not endorse if asked directly. Research on implicit and explicit prejudice has shown that these two forms are psychologically distinct. In one study, participants completed multiple measures of prejudice, and the results separated into different factors: explicit prejudice (what people say they believe) and two types of implicit prejudice, one involving automatic activation of stereotypes and another involving their automatic application in judgment.1Review of General Psychology. Implicit and Explicit Components of Prejudice In other words, you can simultaneously deny holding prejudice while your brain still fires off biased associations without your permission.
This disconnect between what people consciously endorse and how they spontaneously react is one reason prejudice is so stubborn. The gap between spontaneous affective responses and self-reported evaluations reflects conceptually different forms of prejudice, including both overt and subtle variants.2Social Psychology. Changing Implicit and Explicit Prejudice Someone who sincerely believes in equality can still carry implicit biases shaped by the culture they grew up in. That does not make the person “secretly racist,” but it does mean prejudice is not simply a matter of bad intentions.
Research on group dynamics adds another layer. Studies testing social identity and social dominance theories found that for virtually all groups, stronger identification with one’s own group was linked to warmer feelings toward that group. But people who scored higher on social dominance orientation, a preference for hierarchical relationships between groups, also showed more negative feelings toward lower-status groups.3Political Psychology. Social Dominance and Social Identity in the United States and Israel: Ingroup Favoritism or Outgroup Derogation? Prejudice, then, is not just about liking your own group. It can also manifest as active hostility toward groups seen as occupying a lower position in the social hierarchy.
When Prejudice Begins
If you think prejudice is purely learned through hateful conversations around the dinner table, the picture is more complicated. The cognitive machinery for sorting people by appearance starts remarkably early. By three months of age, infants can perceptually tell faces of different races apart; by six months, they start categorizing faces by race. Between ages six and eight, children can deliberately sort others into racial groups.4PubMed Central. Children’s Racial Categorization in Context
None of this means babies are prejudiced. What it means is that the human brain is wired to notice and classify visual differences very early, and racial features are among the differences it latches onto. Prejudice develops when those classifications get loaded with evaluative content: when a child absorbs, from media, peers, family, or the structure of the world around them, that certain groups are “better” or “worse.” The raw ability to see difference is universal; the meaning attached to difference is culturally constructed. This distinction matters because it shows that prejudice is not inevitable in the way that categorization is. The brain will notice race, but what it does with that information depends heavily on context and socialization.
What the Brain Does With Perceived Threat
A widely cited finding in neuroscience is that the amygdala, a brain region involved in processing emotionally significant information, responds more strongly when white participants view Black faces. This has often been interpreted as evidence that the brain is hardwired for racial bias. But a closer look at the research suggests a different reading. A review of neuroimaging studies argued that the pattern of amygdala activity is better explained by culturally learned threat associations, specifically negative associations between Black males and danger, rather than by some innate racial or outgroup bias.5PubMed Central. A review of neuroimaging studies of race-related prejudice: does amygdala response reflect threat?
This reinterpretation is meaningful. If the brain response were innate outgroup hostility, it would be hard to change. But if it reflects learned cultural associations about threat, those associations can, at least in principle, be unlearned or overridden. The brain’s reaction tells us more about what a person has been taught to fear than about some fixed biological tendency.
There is also evidence that experiencing racism changes the brain. A study of Black women exposed to trauma found that, even after accounting for PTSD symptoms, greater experiences of racial discrimination were linked to heightened neural responses in regions involved in visual attention and emotion regulation when processing threat-related cues.6JAMA Psychiatry. Association of Racial Discrimination With Neural Response to Threat in Black Women in the US Exposed to Trauma In other words, racism does not just hurt in the moment. It can reshape how the brain processes threat going forward, a physiological toll that individual prejudice alone does not fully account for.
How Racism Operates Beyond Individual Attitudes
This is where the concepts of racism and prejudice diverge most sharply. Prejudice lives in a person’s head. Racism, in the way scholars and public health researchers use the term, refers to systems that produce racial inequality whether or not any single person in the system holds prejudiced views. From American colonial history through the present, explicit practices and policies reinforced disadvantage across every domain of life, from slavery through to racially segregated housing, which created racial isolation and imposed disproportionate costs on Black Americans’ opportunities, networks, education, wealth, health, and legal treatment. These institutional and societal systems build individual bias into racialized interactions, resulting in what researchers call systemic racism.7PubMed Central. Systemic racism: individuals and interactions, institutions and society
Consider a concrete example that many people find surprising: you do not need a prejudiced person anywhere in the chain for a system to produce racially biased outcomes. A scoping review of clinical machine learning models found that roughly two-thirds of the studies examined concluded racial bias was present in the algorithms. One algorithm already in clinical use for predicting future health care costs was found to discriminate against Black patients compared to white patients, potentially contributing to disparities in health care delivery. Other studies found that algorithms predicted similar risk scores for Black and white patients even though the Black patients were actually less healthy.8PubMed Central. Evaluation and Mitigation of Racial Bias in Clinical Machine Learning Models: Scoping Review No one programmed those algorithms to be racist. The bias came from training data that reflected decades of unequal access to care, unequal documentation, and unequal treatment. The system inherited historical racism, and it reproduced racial inequality automatically.
Criminal justice offers another clear case. Risk assessment tools used in sentencing and bail decisions typically rely on arrest data, but arrest is a biased proxy for actual offending. Because policing patterns are uneven across racial lines, using arrest records as input can cause algorithms to overestimate risk for Black defendants, even though the tool itself applies the same formula to everyone.9Annual Review of Criminology. Algorithmic Bias in Criminal Risk Assessment: The Consequences of Racial Differences in Arrest as a Measure of Crime The person running the software might genuinely believe they are being objective. The output is still racially skewed.
Real-World Consequences of the Distinction
If racism were just the sum of all the prejudiced people in a society, you could solve it by changing enough individual minds. The evidence suggests that this framing misses much of the problem. Research on health inequities has found that people who report experiencing racism have higher rates of illness. But the researcher who documented this pattern also argued that locating racism only at the individual level understates the real driver: the health of social groups is most strongly affected by structural phenomena, not just individual experiences of prejudice.10PubMed Central. STRUCTURAL RACISM AND HEALTH INEQUITIES: Old Issues, New Directions Segregated neighborhoods mean fewer grocery stores, longer commutes, worse-funded schools, and less access to quality medical care. These conditions damage health over a lifetime in ways that no amount of interpersonal goodwill between individuals can fix.
Hiring discrimination provides a measurable test case for how structural racism persists. A large-scale analysis compiled data from 90 field experiments across six Western countries, covering more than 170,000 job applications, to examine trends in racial and ethnic hiring discrimination.11PubMed Central. Trends in racial and ethnic discrimination in hiring in six Western countries These experiments use matched applications where the only difference is the name signaling race or ethnicity, so they strip away individual qualifications and isolate the effect of racial categorization itself. The persistence of discrimination across countries and over decades reinforces the point: even as explicit prejudice declines in surveys, structural inequality in who gets hired remains stubbornly durable.
Why Majority and Minority Groups See It Differently
One of the thorniest aspects of conversations about racism versus prejudice is that people on different sides of the power divide genuinely perceive the issue differently, and research suggests this is not just a disagreement about values. In experiments where majority-group members (Israeli Jews) and minority-group members (Israeli Arabs) were asked to imagine themselves in the other group’s situation, majority members found outgroup-related scenarios less imaginable, less relevant, and less plausible to themselves compared to minority members doing the same exercise.12PubMed Central. If I were you Minority and majority members evaluate relevancy and subjective experience differently while putting themselves in the other’s shoes The researchers suggested that an individual’s social power and position in the hierarchy can shape their basic capacity to imagine what life is like for others.
This finding helps explain a common impasse in public discussions. When someone in a majority group says “I don’t see racism,” they may be describing their genuine cognitive experience rather than being disingenuous. Their social position makes certain realities harder to imagine as personally relevant. Meanwhile, people in minority groups, who navigate those realities daily, can find it baffling that anyone could miss what seems so obvious. The asymmetry is not a character flaw on either side; it is a predictable cognitive consequence of unequal social positions.
Intersectionality Complicates Both Concepts
Neither prejudice nor racism operates along a single axis. A person’s experience of discrimination is shaped by the intersection of their race with their gender, age, class, and other identities, and those intersections produce outcomes that cannot be predicted by adding up each category separately. Research on how people attribute their experiences of discrimination found significant interactions between these factors: the combination of race and age influenced whether people attributed discrimination to race, while the combination of gender and socioeconomic status influenced whether they attributed it to age, and so on.13PubMed Central. The Intersections of Race, Gender, Age, and Socioeconomic Status: Implications for Reporting Discrimination and Attributions to Discrimination
At the structural level, the same principle holds. A study of eighth-grade students found that even if socioeconomic inequalities were fully addressed, systems of oppression related to race and gender would still produce inequalities in academic outcomes. Racism and sexism had direct effects on achievement independent of socioeconomic disadvantage.14PLoS ONE. Understanding the Influence of Race/Ethnicity, Gender, and Class on Inequalities in Academic and Non-Academic Outcomes among Eighth-Grade Students This undercuts a common argument that racial disparities are “really” about class. Class matters, but race and gender exert independent effects that persist even when you account for economic differences.
Intersectional approaches to stereotyping and prejudice study the interactional and emergent properties of multiple identities, recognizing that the experience of a Black woman, for instance, is not simply “Black experience plus woman experience” but something qualitatively distinct.15Social and Personality Psychology Compass. Stereotyping, prejudice, and discrimination at the intersection of race and gender: An intersectional theory primer For the racism-versus-prejudice question, this means that both concepts need to be understood as operating along multiple dimensions simultaneously.
Why “Just Educate People” Falls Short
If you have been following the logic so far, you can probably anticipate the problem with the most popular intervention strategy: diversity training. The idea that teaching people about their unconscious biases will reduce racism has become enormously popular in workplaces. But a critical analysis concluded that unconscious bias training is a diversity intervention based on unproven suppositions and is unlikely to help eliminate racism in the workplace. Knowing about your biases does not automatically change your behavior. And even when individuals are motivated to act differently, structural constraints can prevent pro-diversity actions.16Work, Employment and Society. Pointless Diversity Training: Unconscious Bias, New Racism and Agency
This critique connects directly to the racism-versus-prejudice distinction. Unconscious bias training treats the problem as prejudice: individual minds that need correcting. But if the structures themselves produce unequal outcomes regardless of what individuals think or feel, then fixing individual attitudes without changing structures is like bailing water without patching the hull. The critique does not mean that addressing individual prejudice is useless, just that it is insufficient on its own.
Intergroup contact, by contrast, targets both individual attitudes and the conditions under which groups interact. The foundational theory in this area, developed by Gordon Allport, identified four conditions for contact between groups to reduce prejudice: equal status between groups within the contact situation, common goals, intergroup cooperation, and support from authorities or institutions.17PubMed. Intergroup contact theory Notice that these conditions are partly structural. You cannot have “equal status within the situation” if the broader system ensures one group always enters the room with fewer resources or less authority. Effective prejudice reduction requires structural change to create the conditions under which contact actually works.
Colorism and the Global Picture
Conversations about racism and prejudice often default to an American or Western European frame, where the categories are organized around discrete racial groups. But in much of the world, the dynamics of skin-color-based discrimination operate somewhat differently. In Asia, for example, color and colorism have deep historical roots that evolved separately from the relatively modern Western concepts of race and racism.18Annual Review of Sociology. Skin Color and Colorism: Global Research, Concepts, and Measurement Lighter skin has been associated with higher status in many Asian, Latin American, and African societies for centuries, often predating European colonialism, though colonialism reinforced and intensified these hierarchies.
Colorism is worth understanding separately from racism because it can operate within racial groups, not just between them. A lighter-skinned Black person in the United States, for instance, may face less hiring discrimination than a darker-skinned Black person, even though both are categorized as Black under the American racial schema. Similarly, caste-based prejudice in South Asia and ethnic hierarchies in parts of Africa do not map neatly onto the Western concept of “race,” but they produce structurally similar outcomes: inherited disadvantage, unequal access to resources, and cultural narratives that justify the hierarchy. The prejudice-versus-racism distinction applies in these contexts too, but the specific categories and histories are different enough that importing American frameworks wholesale can obscure more than it reveals.
When the Terms Themselves Become Contested
Part of the confusion in everyday discussions stems from the fact that people are often using different definitions of “racism” without realizing it. In casual conversation, many people use “racism” to mean personal hostility based on race, which is essentially what researchers would call prejudice or bigotry. In academic and policy contexts, “racism” increasingly refers to the structural and systemic phenomenon, where the emphasis is on outcomes, patterns, and institutional arrangements rather than individual feelings. Neither usage is wrong in any absolute sense, but the mismatch creates enormous miscommunication.
When someone says “I’m not racist,” they usually mean “I don’t harbor hatred toward people of other races.” When someone responds “but you benefit from a racist system,” they are using the structural definition. Both statements can be simultaneously true, but the conversation collapses because each person hears the other’s words through their own definition. Recognizing this definitional gap does not resolve the underlying disagreements about policy, responsibility, or justice, but it does clear away a layer of needless confusion that often prevents those substantive conversations from happening at all.
The structural definition also explains why some scholars argue that, in societies where one racial group holds dominant institutional power, members of subordinate groups can hold prejudice but cannot practice “racism” in the systemic sense. This claim infuriates people who use the individual definition, because under that definition anyone can obviously be racist toward anyone else. Both positions follow logically from their starting assumptions. The argument is not really about facts but about which definition better captures the most important aspect of the problem. People who focus on individual morality tend to prefer the broader definition; people who focus on societal outcomes tend to prefer the structural one.
What is not in dispute, across either definition, is that the consequences of racial prejudice and structural racism are measurable, persistent, and consequential for the people who bear them. Hiring experiments keep showing discrimination. Health data keep showing disparities. Algorithms keep reproducing the biases baked into their training data. Whether you call the root cause “racism” or “prejudice backed by systemic power” matters for how you think about solutions, but the harms are real under any label.