Economic models generally start with the individual person. Any departure from that (i.e. starting with a firm or household, assuming homogeneous population) requires justification, alongside consideration of what happens when that assumption is dropped. This methodology is called methodological individualism.
Today I will be discussing the opposite methodology, collectivism. This form of analysis starts with the whole population, and the individuals are ignored. Unfortunately this is a very poor form of argument which often runs into fallacy of division, and generally lacks a clear tie to reality.
Collectivist arguments are fairly common outside of economics. Arguments like class warfare, support for small/family businesses over large corporations, women vs men, are all collectivist ideas. The basic idea is that belonging to one of these groups makes you part of some shared utility function or decision-making method. Things are said to harm/benefit “women” in general, and therefore any particular woman should feel as if they are being harmed or benefited merely by being a member of the group, regardless of how she was actually effected. Marx argues that employers engage in a systematic underpayment of workers, ignoring the fact that individual employers are actually in competition with each other in the hiring market.
Less popular collectivist viewpoints include various forms of racism. White supremacists have often said that Jews are taking over the media or banks or whatever. It used to be a common fear that jazz music and marijuana were tools used by black men to seduce white women, which of course would undermine the white race and leave the US open to communist infiltration. “White privilege” asserts that people can be categorized as “oppressor” and “oppressed” simply by their race, and not their individual actions.
The fallacy of division is clear from these examples, but there are also other non-sequiturs present. Many of these arguments run an “us vs them” mentality, and pin dissenters as enemies rather than debaters. A closely related issue is the “victim” mentality, where one group is supposedly the victim of another’s actions, as if both groups were a single being. Dissenters are disregarded and placed alongside the oppressors. “The rich are thieves” and “gays are predators” are simple arguments that fit these descriptions.
But are group mentalities always wrong? Many of the arguments above are obviously poor, but one may learn something about an individual from them being a member of a group. What exactly is wrong with racism? The problem lies in statistical error. In principle, if all you know about someone is that they belong in a certain group, then its possible that the stereotype is useful.
This can easily go wrong, however. One way is if the stereotype is false. But even if it is true, one common error is to give the stereotype more weight than evidence that one has about the individual in question. In practice, people interact with individuals, not entire groups, so in virtually all cases we will have more information about the individual than just their membership in a group.
Different races have different average IQ scores. The average IQ score for asians is higher than whites. This may be a way to justify racial discrimination in employment. The problem here is that the employer knows a lot more about the individual candidates, so much so that the average racial IQs can be ignored.
Say you are trying to hire a financial analyst. One of the job requirements is a degree in finance. Say you have two job possible candidates, and the only thing you know about both is their race, one white and the other asian. Who do you pick? You choose at random, 50/50. The reason is simple: you are already filtering by IQ when you require a finance degree. The averages no longer matter in this case, as you are looking towards the right side of the distribution for both parties. The direct evidence for each individual completely overrides the group averages.
Why aren’t 8 year olds allowed to drive? Is it a collectivist argument to say that children as a whole are not mature enough to drive?
Yes. But this is a case where the stereotype is useful information. Having a courtroom examine every individual child constantly to see if they are yet mature enough to drive is costly. Instead, a bright-line rule is set, where all children above that line are considered mature enough to drive. The line is arbitrary, and there is little chance that a kid’s optimum “ready to drive” time will occur the same moment he turns 16. Some will reach that level of maturity at a younger age, others older. The line is set at 16 because it is perceived for most to be around that time.
The age one is allowed to run for president is 35. This bright-line rule doesn’t follow the same justification as the last, as the candidate who is running is already under examination. There is no efficiency to be had by cutting out younger candidates, only loss. Same with the Supreme Court ban of capital punishment for minors. Since the defendant is already being examined by the courtroom, there is nothing to be gained by banning certain punishment. This fallacy runs clearly in the majority opinion, as both dissenters repeatedly noted.
In short, group discrimination is usually bad, but there are cases where it is logically justifiable. The justification usually boils down to costs of finding information about individuals.
Sources: Consultation with professor, Law’s Order by David Friedman, Roper v. Simmons US Supreme Court case
PS: The standard length for patent protection is 20 years from the earliest filing date. This is a similar bright-line rule to the age restrictions on driving. Patents require a fairly complex analysis to come to this conclusion, so I will have to cover them another day. If you want to know more about them now, then check out Law’s Order.
Updated 2020-06-08, removed some unnecessary stuff