The Science Behind Racism: A Psychological Approach

The reactions of many across the world to the recent atrocities in America have been varied. However, one common theme has been a greater desire for education, in the knowledge that understanding is the opposite of bigotry.  The #BlackLivesMatter movement has sought to expose the extent of racism; it is a problem that is manifested not only in extreme crimes perpetrated by a problematic minority, but also in the daily experience of people of colour, through “micro-aggressions” and implicit bias. This article seeks to offer a brief explanation of the psychological answer to the question: what drives the majority of otherwise fundamentally “good” white people — people who, if asked, would likely state an abhorrence of racism — to exhibit such bias? In short, why are “good people” racist? Can we, by learning why we might have certain prejudice, use that knowledge to deal with implicit biases?

To address this problem, we shall first consider evidence for the instinct of humans to form competing groups. By analysing how this can be applied to racism, we will then examine theories, which suggest how such biases can be overcome. 

The evolutionary background

As humans, one of the most basic instincts we have is towards group formation. This desire to form groups can also be seen in our closest evolutionary relation, other great apes (hominidae), and is pervasive in human culture. Indeed, for many, one of the greatest hardships of the Coronavirus Pandemic is being without friends and family, a reflection of how crucial contact with other humans is to our well-being, a need that is derived from its importance in human evolution: For example, group living increases safety and security, and thus also survival chances and reproduction. Evolutionary theorists such as Dunbar et al. (2005) have proposed that genes supporting group living might have been selected for during evolution.

A necessary consequence of groups being beneficial is that there will be many of them, all competing for the same, limited resources. For our ancestors this would have been water sources, land and perhaps the tastiest berries. For us, it might be jobs or career and educational opportunities. It is not hard to see, therefore, why individuals in a group turn inwards, favouring members of their own group (the ‘in-group’) at the expense of the competing groups (‘out-groups’). This idea has been evidenced by a number of studies.

Robber’s Cave Summer Camp Studies

In the 1960s, Sherif et al. (1961) conducted the “Robber’s Cave Summer Camp Studies”. In these experiments, young (12 year old) middle class boys, who did not know one another beforehand were sent together on summer camps.  They had been screened to ensure they did not have psychological disorders such as psychopathy, which could have meant that the actions of the individuals were not truly representative of the general population and thus bias the results.

The boys were randomly divided into two teams and engaged in team bonding activities, to consolidate grouping. The two groups played a series of zero-sum games, where one team’s win necessarily meant another’s loss, such as tug-of-war. As predicted, this led to genuine conflict and dislike between the groups, manifested in explicit negative attitudes and behaviours towards the out-group. At one point, they even organised and met for a fight between the groups, only prevented by the arrival of the experimenters.

The researchers then tried a variety of techniques to get the boys from opposing groups to like each other despite group differences. The only successful method consisted of the introduction of “superordinate goals”, i.e. goals that could only be reached by bothgroups working together.

One such superordinate goal was to obtain lunch from a truck stuck in the mud. The truck could only be retrieved with both groups working together to pull it out. Sherif et al. found that while one superordinate goal was insufficient to reduce hostility permanently, a series of such goals eventually resulted in the groups becoming friends. One of the groups even shared $5 they had won to buy lunch for all of the boys.

This study is pivotal; the boys were all from the same area, same background, had similar levels of sporting ability and were all at the higher IQ range of the spectrum, as assessed by their teachers, and yet the experimenters were able to generate real dislike and rivalry along totally arbitrary lines by simply introducing competition. This demonstrates just how embedded in our nature group rivalry is, and also further supports the evolutionary argument for its causes. Rivalry occurred mostly because of limited resources and disappeared when resources could only be gained by the two groups working together.

The Minimal Group Paradigm study

However, findings from the “Minimal Group Paradigm” by Tajfel et al. (1971) suggests that group rivalry may occur along arbitrary lines, in the absence of real resource competition. Here, the presence of competition for resources is not a necessary precondition for group rivalry.

In this experiment, groups of adults were asked which abstract painting they preferred out of two similar examples. They were then told that they had been separated into groups based on whether they had an overall preference for paintings by Klee or Kandinsky. In reality, however, the two groups were assigned entirely at random, and each person only knew that they were a member of one group, without ever meeting or seeing the other members of their assigned group.

All participants were then asked to distribute money between the groups. While, perhaps unsurprisingly, there was a general preference to favour the in-group, the key finding from this study was that, given the choice, participants always preferred a scenario in which their own group gained more money than the out-group. Participants even decided to renounce additional gain for themselves if the alternative meant equal winnings for the out-group. For example, in a choice between giving their own group £9 and the out-group £8, or their own group £7 and the out-group £2, they would choose the latter. They therefore chose to maximise differencebetween groups rather than maximisenet gain. It is important to note that these groups were entirely arbitrary, they meant nothing for the participants as soon as the experiment was over, and the participants never met the other members of the group. Yet stillthere was a statistically significant move towards derogating the out-group, even if that came at some expense to the in-group.

These two experiments taken together show how deeply entrenched group attachments are. Even though evolutionary science and the summer camp study suggest that these intrinsic biases are derived from the benefits that group membership confers in an environment with conflict and limited resources, they exist beyond that, even in situations where the group lines are meaningless and confer no advantage. This might be due to the evolutionary high ground of genes that promote an instinct for group formation under all circumstances, simply because groups generally confer evolutionary advantage. That instinct would therefore be preserved in our nature today; the context does not necessarily matter.

Practical relevance

What do these findings mean practically? The evidence presented so far goes some way to explaining racism; we have a strong desire to form groups, those groups are mutually competitive, and there does not necessarily need to be a real reason for the groups to be formed other than for their own sake; the lines can be utterly arbitrary. Considered in the context of racism, this starts to explain why different races or nationalities experience conflict with one another; race is, after all, a very simple dividing line. For a species so prone to group formation, it is perhaps unsurprising to also find division based on characteristics so obvious and unavoidable.

Grouping as an identity trait

To further explain how natural predisposition towards group formation can explain why fundamentally good people can be racist, we need to consider the extent to which group membership is important to them. That is, we do not only feel an urge towards group formation, but we also consider group affiliation as part of our identity. For example, if you were asked to describe yourself, you might say, ‘I go to X University and I support Y football team’.

Whilst of course group membership is not the only factor defining your identity, it is an important feature. This was empirically demonstrated by Volz et al. (2009), who evaluated the extent to which individual self-concept was aligned with group self-concept. For this, they used functional magnetic resonance imaging (fMRI), a technique that measures brain activity by detecting changes in blood flow, and examined activity in the medial prefrontal cortex (mPFC). The mPFC is a brain region associated with many functions, including the processing of information relevant to one’s self concept (i.e. what you think of you).

Volz et al.’s study used Tajfel et al.’s arbitrary group allocation paradigm and scanned participants’ brain activity while they were engaged in reward distribution tasks. These tasks were designed to measure strategies of fairness and intergroup discrimination in the allocation of rewards to in-group and out-group members. The participants who showed morein-group bias also had greateractivation of the mPFC. This may suggest that the more group membership is considered part of an individual’s identity, the more likely they are to show in-group bias and consequently derogate the out-group.

This conclusion has clear practical implications. If the extent to which group membership defines self-concept is directly linked to out-group degradation, then helping people to recognise this is vital. For example, the more that individuals identify as “white British” and value that attribute as important to who they are, the more likely they are to hold implicit or explicit biases against those who are not “white British”.

The “Football Fans” experiment

The idea of saliency of group membership (i.e. the extent to which a person’s attention is focused onto a specific group identity)  and the strength of identification with that group being a key causal factor in derogation of the out-group is further highlighted by a significant study by Levine et al.’s (2005) ‘football fans’ experiment.

The subjects in this study were Manchester United F.C. fans, who have significant rivalry with the supporters of Liverpool F.C. They had to answer questions about their support for Manchester United, in order to make their support for their football team salient in their mind, and were then told they had to walk to a nearby building to finish the second part of the study. As the participants walked to the building, another experimenter, acting the part of a passer-by on a jog, ran past before slipping and falling, grabbing his ankle and shouting in pain. The key experimental manipulation was the colour of the shirt that the “jogger” was wearing: one group of fans saw him wearing a red Manchester United shirt; another group saw him with a red Liverpool shirt, and the final (control) group saw him wearing a plain red shirt. The dependent variable was what proportion of the participants would stop and help, and how this varied based on his shirt’s colour.

The proportions did indeed vary; when the “jogger” was wearing a Manchester United shirt, circa 90%of individuals stopped to help, compared tocirca 30%for the Liverpool shirt and control shirt conditions.

Even more interestingly, in a second experiment, the researches recreated the conditions of the original study but with a crucial change – instead of giving questionnaires emphasising team alliance before meeting the “jogger”, they were given questionnaires emphasising their overall love of “the game”, i.e. football. All other conditions were kept the same. This time the results were starkly different; circa 80%helped the Manchester fan, circa 70%helped the Liverpool fan, while only 20% helped the control confederate.

“the war on climate change”, or more topically, “the fight against the virus” are superordinate goals which […] may help to reduce animosity between different countries and groups of people

This is a crucial finding: arbitrary differences (such as which football team one supports and whether this group identity is further reinforced), can significantly influence one’s behaviour towards another person .[1] The fact that the level of help offered to Liverpool fans increased dramatically when the group lines were redrawn points to a key way in which we can help combat racism, namely encouraging ourselves to think of people of differing races as similar to us because of certain characteristics they share, and making those shared characteristics more salient in our minds than those of race. It is obvious that on a global level, encouraging people to think of others as members of “humankind”, and what unites us as humans, is a necessary attitude in order to eradicate divisions between people, especially if that conceptualisation is against something else. For example, “the war on climate change”, or more topically, “the fight against the virus” are superordinate goals which require the type of group efforts highlighted in the “Summer Camp Studies” by Sherif et al.; tackling such problems may thus help to reduce animosity between different countries and groups of people.


A further important point to note about racial divisions is distinctiveness; that is, people in the out-group are seen as more homogenous, similar between themselves and less distinctive than the in-group. This was shown by Hamilton and Gifford (1976) through the illusory correlations effect, which describes a perception of links between variables even when there is no actual relationship between them.

In this study participants were given a series of slides depicting individual behaviours from people who were either supposedly in group A (the majoritygroup), or group B (the minoritygroup). The ratio of good to bad behaviours was kept the same for both groups. Behaviours used included ‘helped an old lady across a road’, or ‘cheated in an exam’, i.e. minor positive or negative acts. The participants were then presented with the slides again, but this time the name of the group the behaviour belonged to, i.e. group A or B was removed, and the participant had to remember which group the behaviour had come from. It was found that the participants ascribed a greater ratio of negative behaviours to group B, the minority group.

Interestingly, this is not simply a case of minority groups being naturally

viewed as worse. When the study was repeated by presenting for both groups a greater ratio of bad to good behaviour, where the amount of bad behaviour significantly outweighed the good, the participants associated the good behaviour more with the minority group.

“Humans, as “cognitive misers”, are thought to be more inclined to search for features that make a group distinctive, in order to quickly and easily categorise them, and not think much more about it.”

The authors explained this finding in terms of the higher accessibility in memory of the most distinctive combination. That is, in the condition where there was greater good than bad behaviour, the bad behaviour was distinctive, as it was the exception to the norm. Similarly, the minority group was distinctive as it was the exception to the majority norm. Conversely, in the condition where the majority of the behaviour was bad, the behaviour that was distinctive was the good behaviour. The authors postulated that people connected the two distinctive features together, causing the bias observed.

While there remains some debate over the mechanism, the observed results are useful to examine a propensity to view a perceived distinctive feature of the out-group as representative of the whole. This study and others following it have thus been used as evidence to support the argument that the out-group is seen as less distinctive. Humans, as “cognitive misers”, are thought to be more inclined to search for features that make a group distinctive, in order to quickly and easily categorise them, and not think much more about it. That said, follow-up studies have shown that motivational factors, deriving from group membership, can overcome this bias.

This bias is in many ways sensible from an evolutionary perspective: why waste valuable cognitive resources evaluating a group you don’t have much contact with? However, it is immensely dangerous from the standpoint of the fight against racism.

It is not the purpose of this article to address the systemic factors that contribute to racism. Yet these are the likely immediate cause for the disproportionate representation of the Black, Asian and Minority Ethnic (BAME) demographic in the lower socioeconomic groups. There is evidence to suggest the presence of institutional racism in the police force, causing people in the BAME demographic to be more likely to be arrested for a given crime than those of other demographics (Eastwood et al. 2013).

With this higher number of arrests being common knowledge, people of non-BAME demographics are more likely to consider criminal activity to be representative of a typical black person. This is likely due to the distinctiveness effect discussed previously, and may also be augmented by a rule of thumb in psychology called the “availability heuristic”. This states that people ascribe greater importance to things they can easily call to mind.  

“We are already predisposed to fail to properly distinguish between individual members of racial out-groups.”

This is further complicated by the findings of Kelly et al. (2005, 2007). This developmental psychology study discovered that from as young as 9 months old, children who are of race A and brought up around race A are significantly more able to differentiate between similar looking faces of race A than similar looking faces of face B. That is to say, two separate but similar faces of race A would be perceived as different, whereas two separate but similar faces of race B would be perceived as the same. This might simply be again an evolutionary effect: there is little point for early humans in wasting valuable cognitive resources learning to differentiate between groups one rarely sees. Yet this has significant practical effects in modern society.

What we have, therefore, is a natural psychological bias to see the actions of some of the out-group, which are highlighted as distinctive to us, as representative of the whole out-group. In a society that regularly exposes us to these negative aspects through systemic racism, and, in addition, because of an evolved innate bias, we are already predisposed to fail to properly distinguish between individual members of racial out-groups, even beforewe consider our psychological disposition towards viewing the out-group as homogenous.

Therefore, the problem is clear. Humans are predisposed to form groups and to prioritise the group which they are part of, at the expense of any out-groups. Despite the fact that these groups can, and do, form along totally arbitrary lines, the dislike between them is real and keenly felt. In addition, due to other psychological factors, we are predisposed to consider the out-group as more homogenous, thereby allowing negative stereotypes to easily end up being applied to them, especially as these stereotypes are propagated by systemic racism. It is all too easy to see this playing out in real-life, such as when the violent actions of small extremist fringe groups of larger organisations or religions are used as the justification for targeting the whole group.

Solutions to the problem

How do we solve this? There is obviously no single simple solution, and this article most certainly does not claim to have one. The explanation this article wishes to give, however, is that we are all, through no fault of our own, predisposed to have racist biases, and these relate to a variety of evolutionary and psychological factors. The problem of racism (and other forms of discrimination) therefore comes from a failure to challenge such predisposed biases.

“The first step to challenging such biases is an awareness of their existence.”

The crucial response this article seeks to promote is that of the reader  seeking to challenge such instincts about members of an out-group. There is substantial evidence affirming that, as would be logically expected, individuals who are more motivated to challenge racial biases become significantly less racist than their peers, and the first step to challenging such biases is an awareness of their existence.

While these biases might be drawn from evolution, we are highly intelligent creatures who are not simply governed by our instinctive feelings; we have the power to recognise prejudice in ourselves and challenge it, and, over time, change ourselves. This ability, that we are not purely governed by our instincts but have the ability to reason morally, is what likely sets us apart from our evolutionary ancestors, and why there is no excuse for racism.

Psychological interventions to decrease racism

All the evidence about means to decrease racism suggests that exposure is an excellent place to start. Ceasing to see others as members of an artificial out-group is important, as this way the idea of reformulating groups makes salient other attributes. For example, if one had a shared interest (e.g. both playing tennis), with a member of an out-group, this shared feature could be made salient, i.e. should be seen as more representative of someone’s identity than the person’s affiliation with, for example, the BAME demographic.

Additionally, it is argued that in order to reduce prejudice, one must combat the predisposition to seeing members of an out-group as representative of that group. In everyday life, to meet people of the BAME demographic and consider them to be “like you” but at the same time thinking of them as different to a “typical” member of their ethnical group completely fails to reduce overall prejudice.

Models of reducing prejudice support a theory of “dual categorisation” (Gaertner & Dovido 2000, Hornsey and Hogg 2000), where bias between groups was found to be maximally reduced when previously opposing groups considered themselves as part of a superordinate identity anda subgroup within this. This is thought to be due to the fact that humans need distinctive group formation, as discussed above, and having only a superordinate identity, i.e. being a member of humankind, does not convey that necessary distinctiveness.

However, it has been found that intergroup bias can be successfully reduced if we think of ourselves as belonging to the superordinate identity and of a sub-group within that, such as belonging to a university and a college within it. This can be easily analogised within the context of Oxford – at Summer VIIIs, one might ardently support one’s own college and feel dislike towards other competing colleges. However, when watching the Boat Race, one would support Oxford, and the dislike towards other Oxford Colleges would have disappeared; rivalry is now centred on Cambridge. Nevertheless, when they introduce the rowers for each boat, if one saw a member of one’s own college, one might be particularly keen for that rower to do well. 

In a racial context, this can be useful, as it means that members of BAME demographics do not have to give up their culture to “fit in”, and members of other demographics are able to recognise that including BAME people does not mean they have to change their cultural preferences.

Therefore, while we may all be predisposed to a background level of unconscious racial bias, we also have the power to do something about it. Challenging one’s thoughts and recognising when one has assumed something because it is easy to do so, is the first step – an implicit bias test is another excellent way to start. Project Implicit from Harvard University provides a good version – it can be found at this link:

Without purporting to offer any solutions further than increased education, this article has merely touched on several of the background psychological considerations which factor into the attitudes of members of perceived in-groups and out-groups.

For further reading, material in this area is extensive, and chapter 14 of Hewstone et al. (2015), An Introduction to Social Psychology is an excellent place to start.

[1]As an aside, it is important to consider when reading this study that the confederate is not in life threatening danger – it is likely, although there is no evidence for this, that had they been, they would have been helped in higher proportions across conditions. However, this is the point of this article; it is not the aim to address here why some police officers are brutally racist, or why there are too many instances of violent assault on minority groups. There are far more factors that feed into this type of behaviour; such a discussion is beyond the scope of this article. What this article does seek to address is why ‘average’ people, although this term is of course in and of itself loaded, behave with discrimination and exhibit covert racism, and this study goes some way toward answering this.

Caroline Wyatt is a second year Experimental Psychology student at Magdalen College. She is happy to address any outstanding questions you may have after reading this article – correspondence should be sent to [email protected] 

References & further reading

Dunbar, R. I. M., Barrett, L., & Lycett, J. (2005). Evolutionary Psychology: A beginner’s guide. Oxford: Oneworld

Eastwood, N., Shiner, M., & Bear, D., (2013) The Numbers in Black and White: Ethnic Disparities in the Policing and Prosecution of Drug Offenses in England and Wales. Retrieved from

Gaertner, S. L., & Dovidio, J. F. (2000). Reducing intergroup bias: The common ingroup identity model. Philadelphia, PA: Psychology press

Kelly, D. J., Quinn, P. C., Slater, A. M., Lee, K., Gibson, A., Smith, M., Ge, L., & Pascalis, O. (2005). Three-month-olds, but not newborns, prefer own-race faces. Developmental science, 8(6), F31–F36.

Kelly, D. J., Quinn, P. C., Slater, A. M., Lee, K., Ge, L., & Pascalis, O. (2007). The other-race effect develops during infancy: evidence of perceptual narrowing. Psychological science, 18(12), 1084–1089

Hamilton, D. K., & Gifford, R. K. (1976) Illusory correlation in interpersonal perception: A cognitive basis of stereotypic judgements. Journal of Experimental Social Psychology, 12, 392-407

Hornsey, M. J., & Hogg, M. A. (2000) Subgroup relations: A comparison of mutual intergroup differentiation and common ingroup identity models of prejudice reduction . Personality and Social Psychology Bulletin, 25 242-256

Levine, M., Prosser, A., Evans, D., & Reicher, S. (2005). Identity and emergency intervention: How social group membership and inclusiveness of group boundaries shape helping behaviours. Personality and Social Psychology Bulletin, 31, 443-453.

Sherif, M., Harvey, O. J., White, B. J., Hood, W. R., & Sherif. W. (1961). Intergroup cooperation and competition: The Robbers Cave experiment. Norman, OK: University of Oklahoma

Tajfel, H., Billig, M. G., Bundy, R. P., & Flament, C. (1971). Social categorisation and intergroup behaviour. European Journal of Social Psychology, 1, 149-178

Volz, K. G., Kessler, T., & Von Cramon, D. Y. (2009). In-group as part of the self: In-group favouritism is mediated by medial prefrontal cortex activation. Social Neuroscience, 4, 244-260

Title image by Markus Spiske on Unsplash.