What is racism?

At the moment, I am writing an empirical study into racism in Queensland, which I will report on at a later date. It made me reflect on the basic question of what racism actually is. Let me give you seven possible scenarios to help us reflect on what we think racism is, whilst I also tell you what the mainstream economic response to those scenarios is.

You might of course try to sidestep the whole issue of racism by saying there is no such thing as race. I am going to hide behind our constitutions on that one: our constitutions condemn discrimination on the basis of race, implicitly presuming there is such a thing as ‘race’. If you need something concrete, simply think of it in terms of skin colour (Black versus White) or Caucasian versus African, but I do acknowledge that ‘race’ is not easy to define, if at all. So dont pester me in the comment boxes with the usual throw-away line that ‘genetic variation is larger within groups than between them’.

The seven scenarios:

  1. People of different races have unequal outcomes, with some groups living longer and wealthier and happier than others. Is this racism? In mainstream economics, the answer is ‘no’ for the outcomes may be due to own effort, due to the fact that people live in different places, etc. Within mainstream economics there is no ‘duty’ amongst the fortunate to make sure the unfortunate have the same outcome as they have. So differential outcomes are not racism.
  2. People of different races are treated differently by public servants and the general population within a country, as particular groups are employed at different wage rates, receive differential amounts of education and health care, are treated differently by the police, etc. Is this racism? In mainstream economics, the answer is again ‘not necessarily’ for it matters in economics what the underlying reasons are. If particular groups live in more remote locations and are themselves averse to receiving particular public services, it is not deemed racism that they get different amounts of it. Similarly, if different treatment by the police reflects higher criminality by some groups, this again is seen as ok. Employers paying people differently because of differences in productivity are again not being racist in mainstream economics.
  3. Individuals of different races are tainted by whatever average behaviour is associated with their group, leading individual members to have to accept lower wages if their group on average is less productive, leading individuals to be treated with less respect and more caution because their group as a whole displays more violent and disruptive behaviour. Is this racism? Again, the answer within economics is ‘no’. This is merely ‘statistical discrimination’ and is seen as the logical consequence of the costs of gathering information about an individual. There is no sense in which it is deemed imperative that employers, civil servants, and the general population makes more effort to find out what the actual characteristics of the individual are: the costs of information gathering are deemed sufficient excuse to treat people as if they are typical proponents of their whole groups or at least to ask them to compensate for the negative signal of their group affiliations even if they cannot choose those signals (such as skin colour).
  4. Individuals of different races are treated less favourably by individuals in the majority group even if they are known to be just as productive and safe simply because the individuals in the majority group find it easier to make friends with someone from their own culture. As a result, the other races have to accept lower wages, fewer clients, less favourable treatment in public services, etc., simply as compensation for the fact that the majority group finds it more tiresome to interact socially with them. Is this racism? Mainstream economics finds this one difficult because it essentially ties two exchanges into one: the market exchanges of labour and consumption goods then get bundled with social exchanges of friendliness and bonding experiences. The discrimination on the social side would be seen by economists as ‘ok’ because it is essentially interpreted as lower social productivity on the side of the other races. But to have that social productivity spillover to all other market transactions is seen with unease as there is a sense to which they ‘should be’ separate. Nevertheless I think that for most economists this type of discrimination would not be seen as racism, but rather a peculiar form of lower productivity of the other races leading again to justifiable discrimination.
  5. In streightforward cases of charity wherein there is no social interaction between individuals, particular races get less favourable treatment than ingroup members, such as anonymous donations being actively withheld from ‘outgroups’. Is this racism? Within economics, it certainly would be seen as a form of ‘ingroup preference’ and thus also a ‘taste against other races’. But would it be enough to qualify as racism?
  6. Individuals of the majority ingroup go out of their way to harm and belittle people of other races, such as when they commit pogroms or cruise around neighbourhoods to beat up people from particular races. Is this racism? Though the vast majority of economists and others would undoubtedly give an immediate ‘yes, of course’, the issue is still not a given. If these activities are due to some kind of revenge motive when, say, the official institutions have not enforced the laws, then I think you will find some economists who dont think this is racism.
  7. Individuals treat individuals of another race deliberately badly simply out of own enjoyment of seeing the discomfort of other races, with no other reason for this enjoyment than a domination motive. Is this racism? Yes, this one would count for there is a displayed ‘taste for discrimination’ without any sense in which mere material motivations or some productivity-related characteristic of the other race justifies the actions taken.

Now, if I reflect on these cases more carefully, I get very uneasy. You see, where actually is the line between someone who wishes to dominate another race out of some personal fantasy (case 7) and someone who does not take up the burden of ensuring equal outcomes for people in other countries (case 1)? Both are actually more similar than they seem at first glance: in both cases are the feelings of the others irrelevant and are we ‘merely’ seeing the outcome of preferences that favour the own ingroup over the outgroup. All that really changes between case 1 and 7 is the proximity of the outgroup and the number of steps between choices and ultimate outcomes. If you like, all that changes is the visibility of cause and result.

Does racism then truly boil down to just that: how visible it is that we care more about ourselves and our ingroup than that we care about others? Are the dividing lines mere sophistry? Or are in fact all such emotive labels as ‘racism’ mere lines on a continuum and arbitrarily drawn depending on the circumstances in a society? What do you think?

Author: paulfrijters

Professor of Wellbeing and Economics at the London School of Economics, Centre for Economic Performance

12 thoughts on “What is racism?”

  1. How familiar are you with the anti-racist movement/literature? That uneasiness you have about the lack of difference between cases 1 and 7 is suggestive of a major point in anti-racist thinking – the offender’s intent is fairly meaningless to the victim.

    In a lot of these examples, there is also a real sense of institutional racism as a result of de facto and de jure power imbalances in the past and present. The potential racist in these examples is often responding to the incentives of the environment provided by those institutions, and they are in turn perpetuating the power imbalances that enforce those institutions. It’s no wonder that the basic definition of racism used in anti-racist literature is “racial prejudice plus power”.


    1. Chris,

      as you might guess, I am much more familiar with the economic literature on this (which is vast) than other literatures, though I have done my fair share on the wider literature (my first paper on discrimination is 20 years old now).

      “Power plus racial prejudice” is not a clear definition to me though: can you tell me which of the 7 examples above involves racial prejudice and which do not? Is ANY preference of the ingroup over the outgroup a racial prejudice (in which case nearly all humans everywhere would be racists)? Can you only be a racist if you have power?


      1. There is room for interpretation of prejudice, but the focus of that definition is generally more on the power element. Essentially, racial prejudice is acknowledged as existing pretty much universally, but power is what makes it truly damaging and racist. So, all humans everywhere can be racially prejudiced, but it doesn’t lead to racist social structures and inequalities without a power imbalance.

        Obviously you’d get a much better response from academics in critical race theory or longtime anti-racist activists. I bring up this other literature mostly because it’s obviously the framework that most anti-racist activists work with, and I’m quite interested in how the economic literature might intersect with it and where it differs.


      2. Hi Chris,

        I think it fair to say that the power element is only implicitly there in economics in that two crucial, but in this context somewhat arbitrary, distinctions are made. One is between maximising production (to which end almost nothing counts as racist) and maximising immaterial consumption (to which end some bits count as racist). The other is between the behaviour of the low-power individual and the other is the high-power law and other ‘official’ institutions, where discriminatory rules on the side of institutions are much quicker branded as racist than on the side of individuals.

        The power angle is not as clear-cut as it sees though: every choice presupposes some power (free will). So is a country racist if it does not take in people from other countries (if so, nearly all countries are racist and so are their ‘social structures and institutions’ supporting such a choice). Is an individual who is rich a racist if he does not use his wealth (power) to achieve equal outcomes simply because he cares more about his own family and friends (an ingroup who might be within some particular race?). The kind of definition you talk about seems to me to make a racist out of every individual and every institution on the planet, i.e. not a very useful definition unless the object is to be able to point an accusing finger at anyone one wants.


  2. James Heckman spoke of Southern ways after growing up in the south for two years in the 1950s and his return in 1963 and in 1970 at http://www.minneapolisfed.org/publications_papers/pub_display.cfm?id=3278

    His parents when they arrived received a delegation of neighbours to explain Southern ways

    There was organised segregation in 1963 when he visted again.

    His 1963 visit with a college roommate from Nigeria was monitored by the local sheriff.

    In Birmingham, he stayed at the black YMCA. People there were frightened to death because he was breaking the local Jim Crow laws. Shops closed in New Orleans to avoid serving them.

    In 1970, Heckman re-visited New Orleans as an academic, going back to the same places.

    They were completely integrated, totally changed. This rapid social change fascinated him.

    The Civil Rights Act of 1964 broke the control of segregationists over their political and legal institutions.

    The racial segregation collapsed because it could no longer rely on Jim Crow laws and the private violence and boycotts through the white citizens councils which police turned a blind eye too when they were not actively involved.

    Paul Rubin wrote a paper in the late 1970s in public choice that modelled racism as a tie-in. easy to hand out limited resources and rent-seek based on an immutable marker. barzel explained the racial basis of slavery on policing costs. it is easier to spot run-aways if the slavery is racial.


  3. Hey Paul, you might want to check out a book by George Frederickson called “Racism: A short history”. It’s a quick primer on how social scientists think about racism. There isn’t much reference to the economics literature or how to measure racism but it instead tackles the normative and epistomological issues that economists are uncomfortable with. The ultimate conclusion is that racism is the systematic exclusion of a certain group. He obviously elaborates on this and illustrates with detailed case studies. There is no magic statistical model to seperate out all the ‘unobservables’ from ‘racism’ and there never will be.The arguments in the book are convincing and the historical perspective is very valuable.


    1. the definetion of systematic exclusion of a certain group is too narrow,

      there are plenty of racism that is still out there despite failing in the systematic exclusion of a certain group. these racists live on in their own little worlds of hate.


    2. Exclusion from equal status with the story of the dominant group would indeed be close to how I see racism. The single racist is then someone violating the norms of the dominant group. Such a definition does not fit within current economics though which has an individual-taste take on it.
      The difficulty with the inclusion idea is that there are so many groups. Is a group that lives separate and wants no interaction racist?


  4. Statistical discrimination (mentioned at point 3) is not efficient compared with full information (do you think the economy would be efficient if absolutely everyone got paid the same (the average) regardless of productivity?), so establishing mechanisms to improve information will improve efficiency. So economics is not ‘okay’ with it – not that I think that is a meaningful concept. Even ‘mainstream’ economics is a discipline, not a person or system of ethics. Just because something is efficient doesn’t mean economics is ‘okay’ with it.
    More generally, the problem with your economic approach is that racism is very much about people’s preferences, and standard economics treats preferences as fixed and inalienable. ‘Mainstream’ economics is therefore not relevant to assessments of what constitutes racism – it simply does not address such questions. Indeed, it is incapable of thinking about what are good and bad preferences.
    You seem to be locked into a mindset that says ‘it is only racism if it has bad outcomes’ (as measured by an economist). Why should this be a precondition for racism? Granted, public policy concern about racism stems from a view that it is bad, but this is not the logical starting point for defining it. My working definition: thinking adversely about a person or group (disliking, thinking inferior, etc.) simply because of their ‘race’ (or traits closely associated with that ‘race’ – although I appreciate this introduces problems in definition).
    A separate question, which I agree is of interest, is ascertaining what the outcomes of racism are, and what outcomes are bad. However, I think the basic premise of most people in this area (which I shall not try to substantiate) is that racist preferences/attitudes/beliefs are bad, and the goal is to change these ‘irrational’ preferences (preferences which could actually be rational, by the way).


  5. I found Schelling’s integration model fascinating here – just the idea that even quite tolerant groups will tend to self segregate, which can lead to adjacent areas with incredibly different cultures, and incomes.

    Say if you’re a white business owner, opening up a business in a predominantly white area, will tend to mean that your employees are largely white. Could it be that somebody coming into that business who wasn’t white would feel uncomfortable? Are they less likely to want to apply to that business, making it even whiter? Is it then racist not to try to change the culture? Or do you accept that your business is excluding a percentage of the possible employees and looks like the Aryan nation and move on…

    Dunno. Interesting thorny, difficult topic. Good luck.


  6. ” Or are in fact all such emotive labels as ‘racism’ mere lines on a continuum and arbitrarily drawn depending on the circumstances in a society?”

    Yes. Of course racism is a social construct that is subject to change depending on contemporary social conditions.

    To me all 7 are basically racism, or at least evidence of racism in the context of my understanding of social norms in this regard.

    The discipline of economics is primarily about utility, and when aggregated, welfare. Surely what you have explained to be non-issues for economists, like unequal outcomes, is a sign that welfare could be increased through redistribution or some reduction of racism at some level. Of course there is ‘no duty’ for the well-off to give to the needy – at least no duty from some all-powerful external party. The duty arises from within the group to which the well-off belong, because the group makes decisions to create institutions and obligations to improve their outcomes as a whole.

    Now, you say that the proximity of the out-group is important. For sure. You essentially want to know the general opinion about when ‘out-group-ism’ is unfavourable, since racism is just one particular example of this bias.

    My view is that the idea of nations, states, countries is currently very the dominant idea about groups at a large scale. Thus, the current social norm is that we can bias towards out country-folk, and it is perfectly acceptable and expected that we do. However, within countries, I would say we are not expected to have major group divisions, particularly along racial lines. You could justify this idea with notions such as civil stability, crime rates with many local groups etc. But ultimately it is just a norm.

    I doubt we will see the world come together as a single strongly identified group. No one ever answer the question ‘Where are you from?’ with ‘Earth’. Because it is self-evident because there is no out-group for Earth.

    Which means, since there is no out-group to push us Earthlings together, we will play a continual game of group formations and dominance strategies, basically forever.

    Back to my point, we have decided that out-groupism along nation-group lines is acceptable (and other groups, like sporting teams etc), but not racial lines, unless the two coincide.


Comments are closed.