Overcoming Bias

“The Tudor landowning justice of the peace (J.P.) was the greatest of of paternalists, rivaled only by the Tudor judges and privy councilors who who controlled the J.P.s. … They wanted to regulate the prices of bread, beer, and wool, the games one played, the amount one drank, the nature of one’s apprenticeship, and the clothes one wore. They arrested drunkards, fined those who did not attend church, and penalized the adulterous. …  a paternal state … only the 20th century has come to eclipse it” (more)

I spent most of the day Tuesday reading papers on paternalism, which was the topic of my job talk paper long ago, and one that I’ve thought a lot about over the years. Alas, almost all writings on the topic seek to argue for or against paternalism, rather than trying to explain it. Now if it were typically efficient, that would in fact be a reasonable explanation. And there are many papers that reasonably argue for the plausible efficiency of mild paternalistic “nudges”, weakly enforced.

But in actual fact we see a huge amount of quite strong paternalism, vigorously enforced. People are greatly discouraged from suicide, and prevented from selling themselves into slavery. Professional licensing limits who can do what, and sex laws limit who can do what with who. Censorship limits what you can read or see. Regulations limit the availability and uses of land, buildings, cars, planes, power plants, food, drugs, and much much more. To prevent “exploitation”, many prices are regulated, purchase is required of schools, doctors, and more. Finally, contract law greatly limited the kinds and levels of penalties that contracts can impose, and the kinds of contracts to which you can agree. And by far the most common rationale offered for all of this is that you are being protected from hurting yourself, not from hurting others. 

This is another one of those subjects where everyone thinks they know the answer, but they all know different answers, almost none of which actually hold up under scrutiny. The most commonly offered explanation is that regulators know more than those they regulate. But then why can’t regulators just tell what they know, such as via very visible certification? If the info for certification is underproduced, why not subsidize it. If it is too easy to forget to check certification, why not offer “would have banned“ stores, where customers must pass a test showing they understand it only sells stuff is otherwise have been banned by regulations. 

Of course it is plausible that some parties extract big selfish gains from these rules, and we do see many examples, such as professionals whose wages are increased via the supply cuts caused by professional licensing. But we need to explain why most everyone else goes along – most actual paternalism is in fact very popular among most people. So for that we’ll need benefits that are much more widely distributed. (In the usual “Bootleggers and Baptists” story, we need to explain the Baptists.) 

The closest I can find to an efficiency explanation is the idea that people make random but correlated mistakes, at which times they are too proud to listen to advice, and at other times they can’t accept that this might later happen to them. Temporary mistakes are easy to fix by requiring modest waiting periods, and temporary individual mistakes can be fixed by requiring groups of associates to choose something together. (Or equivalently, close associates who can veto individual choices.)

But the hypothesis here is that every once in a while a whole group of associates will all go kinda crazy, a “childish” kind of craziness which may last for quite a while. In this rare but correlated childish-crazy mode, this hypothesis says people tend to be especially unwilling to listen to advice, perhaps out of pride. Maybe they see themselves in a status contest with authorities, and are eager to show independence or defiance. Furthermore, people somehow just can’t accept that this problem might happen later to them, and so aren’t inclined to voluntarily choose to commit ahead of time to some more local paternalistic process which would protect them later.

That’s the best I can come up with, and yes this could in fact explain some paternalism. However, I just can’t see it as sufficient to explain the actual typical huge levels of paternalism that we see. So I must look elsewhere. A year ago, I favored this story: 

Thus another possible explanation for min-quality regulations is that, by officially declaring common lower class choices to be bad choices, regulators support upper class claims to be better people. And by forcing everyone to visibly accept this declaration via their not visibly defying the bans, everyone appears to support this claim that elite choices are better choices. … Why would so many non-elites support these policies as voters? Plausibly because they aspire to elite status, and by publicly displaying their agreement with elite attitudes, they affirm that they are themselves good candidates for higher status. (More)

Prestige is a key human process, and a key element is that we all seek to copy the behavior of the prestigious, and to associate with them. So a strong eagerness to push everyone to do what elites do, and what they say that one aught to do, seems completely to be expected. 

Even so, this explanation has still seemed somewhat insufficient to me. There is so so much paternalism! So in this post, let me add one more factor that I think complements the above stories, but also adds substantially to them. 

The key idea is that there are many “rulesy” people in the world. (Think of Sheldon from Big Bang Theory and Dwight from The Office.) These people specialize in learning of and even creating rules, so that they can then find and reveal violations of these rules around them. This allows them to beat on their rivals, and also to raise their own status. It obviously raises their dominance via the power they wield, but they prefer to be instead seen as prestigious, enforcing rules whose purpose is more clearly altruistic. And what could be more altruistic than keeping people from hurting themselves? 

So many people who are especially good at noticing and applying rules, good at finding potential violations, good at framing situations as rule violations, and willing to at least gossip about violators, are eager for a supply of apparently-paternalism-motived rules they can enforce. So they take suggestions by elites regarding what is good behavior and work to turn them into rules they can enforce. They push to turn norms into laws, and to make norms out of the weak behavior patterns of elites, or from their patterns of praise and criticism. 

Now think of the incentives of observers. A declares that B has violated a rule, and audience C has a choice to support A or B in this situation. The rule might be obsolete, A may be stretching its meaning to fit this case, or declaring a new rule from related prior cases. Even so, if B is associated with C, it may seem like corruption for C to support B. If the rule is justified as protecting some folks, then by supporting B you seem to not care about those protected folks. And maybe folks will suspect C of wanting to violate this rule themselves, or of already having violated it. Most of these considerations seem to lean toward supporting A in their case against B.

For example, maybe at first some elites sometimes wear hats. Then they and others start to praise hat-wearers. Then more folks start to wear hats, and get proud of how they are good hat people. Good candidates for promotion to elite they are. Then hat fans start to insinuate that people who don’t wear hats are not the best sort of people in various ways, and are only hurting themselves. They say that word needs to get out about the advantages of hats. And those irresponsible people arguing against hats really need to be dealt with – everyone should be told that their arguments mostly don’t meet the highest possible standards of scientific rigor. (Though neither do most pro-hat arguments.)

It becomes a matter of pride to teach your children to wear hats. And to have hats taught in school. And to include the lack of hats in lists of problems that problem people have. Hat fans start to push the orgs of which they are part to promote hats, sometimes even requiring hats at org functions. Finally it is suggested that wouldn’t it be simpler and more efficient to just have the government require hats. Then foreigners who visit us won’t think we are such backward non-hat people. And its really for their own good, as we all know.

At every step along this path, people can gain by pushing for stricter and stronger hat norms and rules. They are good people, pushing a good thing, which just happens to let them dump harder on rivals. Which is plausibly why we tend to end up with just too many overly restrictive rules. Rules rise with the ratchet of crises that can be blamed on problems said to be fixed by adding new rules. And between the crises, we rarely take away or weaken our rules. 

This sort of tendency to create excess rules can help to explain why many organizations seems to be afflicted by excess “legalism”, including government.

And I’m not sure exactly how, but I suspect that this process is mutually supportive of processes that push for a lot of discretion in rule enforcement: 

To the extent that there are rules, there seems to be a preference for authorizing some people to have discretion to make arbitrary choices, regarding which they are not held strongly to account. … Most people mainly favor discretion … to project to associates an image of being the sort of person who is confidently supports the elites who have discretion, and who expects in general to benefit from their discretion. … The sort of people who are eager to have a fair neutral objective decision-making process tend to be losers who don’t expect to be able to work the informal system of favors well. (More)