Discrimination: less is more

‘Discrimination’ can mean all sorts of things. One of the main ones, and what it will mean in this post, is differential treatment of people from different groups, due to real or imagined differences in average group features.  Discrimination is a problem because the many people who don’t have the supposed average features of a group they are part of are misconstrued as having them, and offered inappropriate treatment and opportunities as a result. For instance a capable and trustworthy middle aged man may miss out on a babysitting job for which he is truly the best candidate because the parents take his demographic information as reason not to trust him with their children.

This means that ‘discrimination’ is really a misnomer; this problem is due to lack of discrimination. In particular lack of discrimination between members of the groups. For instance if everyone could instantly discriminate between women with different levels of engineering ability, generalizations would be useless, assuming engineering ability is really the issue of interest to the discriminators. Generalizations aren’t even offensive when enough discrimination is possible. Telling a 6’5” Asian man that he’s probably short since he’s Asian is an ineffective and confusing insult.  Even if observers can’t discriminate perfectly, more ability to discriminate means less misrepresentation. For instance a test score doesn’t perfectly determine people’s abilities at engineering, but it is much more accurate than judging by their gender. This is assuming the generalizations have some degree of accuracy, if they are arbitrary it doesn’t make much difference whether you use false generalizations of larger groups or smaller ones.

The usual solution suggested for ‘discrimination’  is for everyone to forget about groups and act only on any specific evidence they have about individuals. Implicitly this advice is to expect everyone to have the average characteristics of the whole population except where individual evidence is available. Notice that generalizing over a larger group like this should increase the misrepresentation of people, and thus their inappropriate treatment.  Recall that that was the original problem with discrimination.

If the parents mentioned earlier were undiscriminating they would be much more trusting of middle aged men, but they would also be less trusting of other demographics such as teenage girls. All evidence they had ever got of any group or type of person being untrustworthy would be interpreted only as weaker evidence that people are untrustworthy. This would reduce the expected trustworthiness of their best candidate, so more often they would not find it worth going out in the first place. Now the man still misses out on the position, but so does the competing teenage girl plus the parents don’t get to go out. Broadening group generalizations to the extreme makes ‘discrimination’ worse, which makes sense when we consider that discriminating between people as much as possible (judging them on their own traits) is the best way to avoid ‘discrimination’.

It may be that something else about discrimination bothers you, for instance if you are most concerned with the equality status of competing social groups, then population level generalizations are the way to go. But if you want to stop discrimination because it causes people to be treated as less than they are, then work on making it easier to discriminate between people further, rather than harder to discriminate between them at all. Help people signal their traits cheaply and efficiently distinguish between others. In the absence of perfect discrimination between individuals, the other end of the spectrum is not the next best thing, it’s the extreme of misrepresentation.

10 responses to “Discrimination: less is more

  1. You might mention that you are assuming that people accurately infer things about the average characteristics of groups, and that this is the only kind of “discrimination” you discuss here. People do discuss other problems under the label “discrimination.”

  2. Oh, come on. Obviously complaints about discrimination aren’t concerned with the thing discussed here, but with group grievance and power. Use of accurate predictive tests is no defense against “antidiscrimination” laws in most cases: the laws are a tool to redistribute to groups with average characteristics such that would be less rewarded in the absence of the antidiscrimination efforts.

  3. “Unimpressed”, bullying of the sort you describe is still limited by what fair-minded people think about discrimination. For instance, even though Google is famous for its employee perks, I don’t see anyone forcing them to hire lots of black and Hispanic software engineers.

    Katja, of course you’re right that if you have accurate information on subgroup means, and you ignore it in your estimation procedure, your errors grow. The $1 trillion question (which Robin alludes to) is, how accurate are the typical person’s priors? Are they inaccurate enough that, in practice, the “don’t discriminate” heuristic is actually error-reducing for most people? Even if they aren’t, do they exhibit unpleasant systematic biases that may be worth reducing even at an “inefficient” variance cost? If one of these previous two cases hold, how much should Kant’s categorical imperative constrain the discriminatory activities of statistical sophisticates like us?

  4. Pingback: Tweets that mention Discrimination: less is more | Meteuphoric -- Topsy.com

  5. Pingback: Discrimination « Econstudentlog

  6. This means that ‘discrimination’ is really a misnomer; this problem is due to lack of discrimination.

    To discriminate is to draw distinctions; to choose. I feel complimented if someone says I have discriminating tastes; that suggests that I have the capacity to appreciate distinctions that others cannot perceive.

    Presumably a person who opposes “discrimination” would advocate never making any decisions, or making all decisions at random. Clearly that’s not what people who oppose discrimination are talking about. Rather, the word discrimination has become a shorthand for “wrongful discrimination” or “undue discrimination.” It refers to a practice of drawing distinctions based on irrelevant or wrongful criteria.

    When you say that problems arise from “lack of discrimination,” I do not understand you to refer to a lack of people drawing distinctions on the basis of inappropriate or irrelevant criteria. Rather, I understand you to refer to a lack of people drawing distinctions based on appropriate, relevant criteria.

    As you observe, discrimination seems to be associated with ascribing some characteristic of a group to each member of that group – even though we all know that people vary in their attributes. Why would people do that? Animus is a possible explanation — but not a necessary one. I postulate that much of discrimination can be explained by simple economics: it’s a heuristic that gets you more bang for the buck. Consider:

    1. Gathering information is costly. Life constantly requires me to make decisions on the basis of insufficient information and time. That person walking down the hall toward me might be my significant other or it might be a blood-thirsty rival; any wrong move on my part may impair my ability to pass on my genes to the next generation. Consequently I develop rules of thumb to guide my decisions, squeezing the maximum guidance from the least costly info.

    One cheap source of info is vision; I can derive info about you at a safe distance. And what info can I derive at a distance? Age, sex, race, physical ability, social status, etc. Another even cheaper source is socialization: I can learn about the world based on what other people tell me. And what they’re likely to tell me about is, at least in large part, all the info they’ve picked up cheaply. Consequently I’m likely to focus on gender, etc. , not necessarily because I think it’s important, but because it’s easy to observe.

    Now, if I had been socialized in a culture with a high degree of racial uniformity I might well not focus or race as a relevant variable — not because I’d be too high-minded for that, but because race would no longer serve as a cheap basis upon which to draw distinctions among people. In the US we have biases about blondes and redheads. I’m told the Japanese have no such biases about hair color — perhaps because the burden of discerning between the range of Japanese hair colors renders this form of distinction too burdensome to work as a useful heuristic.

    2. I feel anxious about my inability to guard against unknown threats or exploit unforeseen opportunities. To soothe my anxieties, I imagine that I can detect patterns based on limited data. For example, I might imagine that I can predict lots of things about you based solely on easily-gathered data such as gender. Here I’m focusing on gender because I want to have greater certainty about you, and gender is among the few things I can perceive about you.

    3. Finally, I have attribution bias. That is, I tend to give the benefit of a doubt to myself and people with whom I identify, and make harsher assumptions about everyone else. Bad things that are associated with me and people like me are the result of accident or misunderstanding or innocent mistake, see? Whereas bad things associated with other people are a result of their weakness or laziness or carelessness or malice or nature. Here I focus on your gender as an easily-recognized marker for determining whether you are like me or not, and therefore whether I should extend the benefit of a doubt to you — and vicariously, to myself.

    If you combine my constant need to draw conclusions from inadequate info + my psychological need for greater certainty than the evidence allows + attribution bias, and mix well, you get a good foundation for the standard forms of discrimination we observe.

  7. Also, perhaps a propos to this discussion —

    In US federal jurisprudence, discrimination can be shown by if a party accords disparate treatment to similarly-situated entities OR accords similar treatment to differently-situated entities. In that sense, I understand you to be advocating that latter test: Treat people who are differently situated differently, even if they come from the same social group.

  8. Actually, in Japan, people with blond hair are stereotyped as juvenile delinquents… it has something to do with actual delinquents dying their hair blond.

  9. Pingback: Does SI make everyone look like swimsuit models? | Meteuphoric

  10. This is years later, but it seems I can still respond: people object to discrimination “because it causes people to be treated as less than they are”, but more to the point, people object to discrimination because it causes people to be treated as less than they are in an unbalanced manner which particularly puts the burden on particular people. If black people have a high crime rate and so law-abiding black people get treated as criminals, the same black person is going to get treated as a criminal over and over again. And having the same person suffer over and over again is worse than having more crime victims but having those crime victims spread evenly throughout society.


Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.