How stereotypes hurt people – even those who don’t conform to them


5:32 pm - January 9th 2012

by Chris Dillow    


      Share on Tumblr

What causes discrimination? Conventionally, there are two views on this. One is taste discrimination; some people would rather not hire blacks or women or gays or whatever simply because they don’t like them.

The other is statistical discrimination; employers believe (rightly or not) that a particular group is, on average, less able and so are reluctant to hire from that group.

However, a recent experiment suggests that there’s another form of discrimination that is harder to eliminate – a simple lack of Bayesian reasoning.

Researchers first got people to perform some mental rotation tasks, and then split the performers into two groups. In one group, which they called K, 43% were top performers (who got 13 or more of 24 right) and 57% poor performers.

In the other group, called L, 14% were top performers and 86% poor ones.

A separate group of people were then given a choice. They could take a small sum of money, or they could gamble on an individual being drawn from group K or L where they would win €20 if the performer were top but nothing if they were not.

Obviously, in this gamble one has a 43% chance of winning if you can choose from group K.

But here’s the thing. Subjects were told that the draw from group L was rigged, in such a way that someone would be drawn from group K and if s/he were a top performer, a top performer from L would also be picked. This means that subjects had a 43% chance of winning, whether they drew from group K or L.

However, despite knowing this, subjects preferred to bet on group K. In other words, they paid attention to base rate probabilities (43-57 vs. 14-86) even though they were irrelevant. This is Bayesian conservatism.

This is not taste-based discrimination, because K and L are neutral terms. Nor is it statistical discrimination because subjects’ actual odds of  success were identical in both groups. It’s plain irrational.

This might have nasty implications in labour markets*. It suggests that group labels stick to individuals, even if the particular individual does not have that particular attribute. So, for example, if an employer is loath to hire women for fear they will leave to have children, he might be reluctant to hire even women who could prove they have no such intent; though lesbians earn more than straight women, they earn less than men. Fannie Hurst’s line that a woman has to be twice as good as a man to go half as far might, therefore, be true.

In this sense, stereotypes are dangerous not merely because they can be self-fulfilling, but also because they damage the life-chances even of people who do not conform to them.


* The standard objection here is that hiring decisions have bigger stakes than lab experiments, and so people are more likely to behave rationally. I find this objection implausible, partly because many hiring decisions are taken by people who are using other people‘s money, partly because of the Yerkes-Dodson law, and partly because of casual empiricism; look how many idiots are in good jobs.

    Share on Tumblr   submit to reddit  


About the author
Chris Dillow is a regular contributor and former City economist, now an economics writer. He is also the author of The End of Politics: New Labour and the Folly of Managerialism. Also at: Stumbling and Mumbling
· Other posts by


Story Filed Under: Blog ,Economy

Sorry, the comment form is closed at this time.


Reader comments


There are personal prejudices, reinforced by personal experience (ie working with a woman how then left to have kids). People must see the individual on every occasion, and we need to be aware. I once asked why a man had asked on other men to apply for a job: he said – but I asked al my friend sin the pub. All white males, then…See? my own anecdotal experience upheld my own suspicion.

Personally Ive always believed that prejudice is a natural animal instinct that serves as protection.

The human brain seems incredibly powerful at categorizing and putting things into its appropriate group in comparison with other things, I also think its a powerful instinct to have a mistrust, or fear of things that are different than ourselves.

I’m white and if i walked into a room with 50 white dudes on one side and 50 black dudes on t’other I think I would naturally and subconsiously choose to walk to the white side of the room, unless I saw someone in the black half who i recognised.

I’m pretty sure most people would do act in s similar way, unless they did the opposite for the same reason. Anyone confirm what they would probably do?

I don’t think this is racism, its quite natural (for me anyway, cant speak for anyone else). The trick is using intelligence and reasoning to see this way of thinking as a throw back to when we were hunted by things, and to try to overcome them.

3. Paul Newman

On first meeting anyone will be cognisant of speech, dress, hair and draw a rough sketch subject to future amendment.
In real life those clues are usually good enough to tell us a lot about someone and we are rarely entirely wrong.
Skin colour might be part of that, along with accent and so on and it is really more like “judgement ” than prejudice, although I would accept that there is a special exclusion of black people by some groups ( Brown people for example )
Laws obliging people not to use their judgement are bound to fail and I am yet to see any demonstration that negative associations with some skin colours and ethnic markers are more or less unfair than judgements about accent dress and manners. That may, of course, be very hard on some individuals who will have to try harder. That, is life, I am afraid .

“They could take a small sum of money, or they could gamble on an individual being drawn from group K or L where they would win €20 if the performer were top but nothing if they were not.

Obviously, in this gamble one has a 43% chance of winning if you can choose from group K.

But here’s the thing. Subjects were told that the draw from group L was rigged, in such a way that someone would be drawn from group K and if s/he were a top performer, a top performer from L would also be picked. This means that subjects had a 43% chance of winning, whether they drew from group K or L.”

You haven’t explained this properly. I can think of at least 5 ways this could be interpreted…and even if I happened to choose the correct interpretation, unless I how small the ‘small sum of money’ is I still can’t get an adequate perspective on what you’re trying to say. As it happens, I’ve heard of this before and know the precise explanation and can therefore assure you that your application to labour markets is a very very loose analogy.

dave

#I’m white and if i walked into a room with 50 white dudes on one side and 50 black dudes on t’other I think I would naturally and subconsiously choose to walk to the white side of the room, unless I saw someone in the black half who i recognised.#

Surely if there were such a room, by the time you entered the ‘black dudes’ would have been divided up by the whites who would by then be effective ‘rulers’ of the room and would probably either have them picking cotton or would have locked them up, having framed them for crimes they didn’t commit?
Incidentally, even if the white dudes were PC metrosexual Guardian readers, the same would apply. You know what white dudes are like…they’re all the same…and they’d be ‘loving it’.
Also, don’t for a second claim that what I’ve written is racist…unless you want to make the same claim about a certain D. Abbott. And anyway I’m talking about a room full of 19th century colonialist white dudes.

@1. rentergirl: “I once asked why a man had asked on other men to apply for a job: he said – but I asked al my friend sin the pub.”

A very common way to gain employment is from social connections. Organisations that train job seekers encourage them to use friends and friends of friends to spot opportunities. By probability, if you are white, the friend or friend of a friend will be white.

@2. dave: “I’m white and if i walked into a room with 50 white dudes on one side and 50 black dudes…”

Superficially that is a clever trick. It won’t work all of the time because some of us are contrary; all that it can demonstrate is that, when no obvious direction about sociability is demonstrated, most people walk towards the people most like them; a more meaningful experiment would be to introduce the subject to a random member of the two groups and to see what follows.

@OP, Chris Dillow: “But here’s the thing. Subjects were told that the draw from group L was rigged, in such a way that someone would be drawn from group K and if s/he were a top performer, a top performer from L would also be picked. This means that subjects had a 43% chance of winning, whether they drew from group K or L.”

I think that, by this point, the experiment has become so abstract that the participants have lost the plot. The experiment does not have a real life metaphor. The results are random and inconsequential.

7. Chaise Guevara

@ 2 dave

Very well put. And yes, I suspect I’d do the same thing in the black-guys white-guys scenario if I wasn’t thinking about it. And if I was at a party and didn’t know anyone, I’d probably gravitate towards people who dressed like me, had a similar accent to me, were about my age etc, unless there was anything else distinguishing the people there.

8. Just Visiting

Charlieman is spot on.

> The experiment does not have a real life metaphor. The results are random and inconsequential.

9. Chaise Guevara

@ 6 Charlieman

“I think that, by this point, the experiment has become so abstract that the participants have lost the plot. The experiment does not have a real life metaphor. The results are random and inconsequential.”

It doesn’t seem so to me. It showed people paying attention to surface appearances rather than rationality. If by “random” you meant it was an unremarkable one-off, there are a LOT of experiments that provide people with a definitively correct rational path, only for them to take a different option.

On the other hand, saying it’s therefore a reason for racial and sexual discrimination is a bit of a leap, but the data is at least suggestive of this being the case.

The problem I have with this particular experiment is that it set up a system whereby both choices were equally good, which encourages people to pay attention to irrelevant details: there’s nothing else by which to distinguish the options. It would be more consequential if it was set up with the same chances of winning, but a slightly higher payout if you bet on L and win. It would be interesting to see if people still bet on K under those conditions.

Stereotyping is simply a stage in inductive reasoning, a cognitive necessity for handling the myriad pieces of information with which we are bombarded. Stereotypes enable us to make predictions about the world, and without them we would have to navigate at random, like making a journey without a map. Neither would there be human language, since concepts are stored in stereotypical form.
If my wife told me there was a bird on my lawn, I would expect a reasonably stereotypical bird, that is, has two legs, two wings, feathers, which flies – perhaps a robin, blackbird, starling. If, on looking, I found that the bird was an an emperor penguin I would feel justified in claiming that she had not been completely truthful. My non-stereotyping wife would probably say “Ah, but you shouldn’t stereotype – penguins are birds too you know”.

“So, for example, if an employer is loath to hire women for fear they will leave to have children, he might be reluctant to hire even women who could prove they have no such intent; though lesbians earn more than straight women, they earn less than men.”

This amused me a bit because every single gay woman I know either has kids, or wants them. I do know 3 militantly child-free women, all very committed to their careers and determined never to reproduce – and they’re all straight. It’s not a big deal in an otherwise fascinating article… but it does go to show quite how prevalent prejudices are, doesn’t it?

@11. Olivia: “but it does go to show quite how prevalent prejudices are, doesn’t it?”

No, it represents that man and women choose to procreate. Hopefully, happy families.

And you just showed your own, Charlieman.

My husband and I chose to procreate in a traditional manner, yes, and luckily to date we enjoy a very happy family. But I know several lesbian families with little kids, all of whom are the biological children of a lesbian mother. As I said, all the gay women I know have kids or want them. Therefore, to explicitly state that lesbians won’t take maternity leave, and thus will never be a cost to the employer, is simply not accurate. It’s a stereotype, not the reality – which is what this article is about, no?

To be honest I’m not really sure why you’re arguing the point, given the law has now accepted it and same-sex couples in a civil partnership can register a baby’s birth and their parental status, under their names alone. There are even law firms who specifically cater to gay and lesbian parents. Do you think that happened without gay families needing it to – because they were having babies? Do you genuinely not know that there are a lot of rainbow families out there?

I should probably clarify, given my shocking grammar didn’t, that those biological children are not of the SAME mother. ;)

Not the same biological mother, I hasten to add. ;)

16. So Much For Subtlety

This might have nasty implications in labour markets*. It suggests that group labels stick to individuals, even if the particular individual does not have that particular attribute. So, for example, if an employer is loath to hire women for fear they will leave to have children, he might be reluctant to hire even women who could prove they have no such intent;

Notice the problem here – the labour market has in fact moved the other way. They have moved towards making it harder for an employer and a female employee to have a conversation to establish whether or not a woman will leave to have children. A sensible employer can only assume that a woman will, as he cannot ask her nor can he sack her later. Asking is forbidden by law. Even though everyone would benefit by a little openness.

though lesbians earn more than straight women, they earn less than men.

Thus proving discrimination does not exist. No sexist will employ a lesbian in dungarees if they can employ a straight woman instead. Thus sexism has no role in employment decisions – although admittedly there was a study recently that said Human Resource Departments are more likely to punish pretty female candidates. Presumably because the women who run them see these women as competitors. They may not see lesbians that way.

Fannie Hurst’s line that a woman has to be twice as good as a man to go half as far might, therefore, be true.

Except it is not. Women continue to cluster in low-demanding jobs where they do not have to be particularly good, they do not have to do anything dirty or dangerous, and then they usually leave for maternity related reasons and return half time. There is no pay discrimination when adjusted for maternity – and given women’s on-going preference for nicer jobs, it suggests they are being paid more than they deserve.

“Subjects were told that the draw from group L was rigged, in such a way that someone would be drawn from group K and if s/he were a top performer, a top performer from L would also be picked. This means that subjects had a 43% chance of winning, whether they drew from group K or L.

However, despite knowing this, subjects preferred to bet on group K. This is… plain irrational.”

Hmm. Bet on group K and you just have to trust that 43% of members of group K are top performers; but bet on group L and you have to trust that 43% of members of group K are top performers, AND that you’re being told the truth about the draw from group L being rigged, AND that the rigging mechanism is reliable (so that a top performer from group L really does get picked every time a top performer from group K gets picked, i.e. 43% of the time), AND that you’re not overlooking any ‘trick’. Since there’s no penalty for betting on group K rather than group L, it looks rational enough to me to bet on group K so as to avoid the risks (however minimal) that you’re being misinformed, that something might go wrong with the ‘select-top-from-L-if-select-top-from-K’ process, or that you’re missing something.

“So, for example, if an employer is loath to hire women for fear they will leave to have children, he might be reluctant to hire even women who could prove they have no such intent”

…and the same applies in this real-world example. If there’s no penalty for simply hiring a man, it doesn’t strike me as irrational to avoid the extra risk involved in relying on whatever ‘proof’ you have that a woman won’t leave to have children (even if that risk is vanishingly small).

…just to add: it looks highly probable to me that the difficulty people have with Bayesian reasoning has implications for discrimination. (For instance: someone might read that 75% of people who take time off work to look after children are women, and mistakenly infer that if (s)he hires a woman, there’s around a 75% chance she’ll end up taking time off to look after children.) But I don’t see that the experiment you cite does much to establish this.

(That 75% figure is for illustrative purposes only, by the way – I made it up.)

19. Torquil Macneil

I think G.O.’s objection stands up, faced with a simple option and a complex one, it might be rational to choose the simple one if the outcomes are rationally supposed to b equal. I would be interested to hear Chris’s response to that.

Superficially that is a clever trick. It won’t work all of the time because some of us are contrary; all that it can demonstrate is that, when no obvious direction about sociability is demonstrated, most people walk towards the people most like them

Thats all i meant it to demonstrate. we feel most comfortable around the familiar.

21. Political_Animal

@2 as a misanthrope, I would ignore BOTH groups and head straight to the bar, whereupon I would stand on my own all night!

One thing though, if you were aware of it, it wouldn’t be ‘subconscious’.

I think the question to be asked here is “would you employ a member of the Conservative Party if they were the best suited for the job, or would you let your prejudices make the decision for you?” We all have prejudices, even we think we don’t.

One thing though, if you were aware of it, it wouldn’t be ‘subconscious’.

I can be aware of the result of subconcious thought, without being aware of those thoughts occuring

24. Political_Animal

@23 the fact that you have precognition of a supposed subconscious act, means that it can’t be subconscious!

25. Chaise Guevara

@ 24

“the fact that you have precognition of a supposed subconscious act, means that it can’t be subconscious!”

That’s not true; the question is whether he was aware of it at the time. The point (as I read it) is that, in that situtation, he would walk over to the white guys without thinking about it, but if he sits down and thinks about the situation, he realises it was due to an instinctive desire to be near “people like me”, a desire that at the time of the event was subconscious.

If I put my hand on a hot stove, I pull away before I’ve had time to register that it’s hot. That doesn’t mean that I’m unaware of the reason for pulling my hand away.

@20. Dave: “Thats all i meant it to demonstrate. we feel most comfortable around the familiar.”

Noted, and what I tried to suggest is that a variant of the experiment might be more meaningful. If a respondent feels that s/he is part of an experiment, his/her behaviour will be affected.


Reactions: Twitter, blogs
  1. Liberal Conspiracy

    Hopw stereotypes hurt people – even those who don't conform to them http://t.co/Xk0Q69w6

  2. Koh Miyaoi

    How stereotypes hurt people – even those who don’t conform to them | Liberal Conspiracy http://t.co/Aq6U3svG via @libcon

  3. leftlinks

    Liberal Conspiracy – How stereotypes hurt people – even those who don’t conform to them http://t.co/MfWYcRD1

  4. Patron Press - #P2

    #UK : How stereotypes hurt people – even those who don ’t conform to them http://t.co/COsq3Clu

  5. alex a pagliuca

    #UK : How stereotypes hurt people – even those who don ’t conform to them http://t.co/COsq3Clu





Sorry, the comment form is closed at this time.