Thursday, July 19, 2018

Safety vs. Status: a Thought Experiment

Scott Alexander has a couple of outstanding posts on one of his favorite themes: the application of the fact-value distinction to matters of public controversy. Scott's general thesis is that many people assert with insufficient warrant that what separates the adversaries in these matters is a difference in fundamental values beyond the reach of reason rather than a difference in their knowledge of various facts. They further assert that since reason can't persuade anyone away from fundamental values, debate is pointless if not counter-productive, and that the only proper course of action is to apply as much force as possible against their opponents: restrict, harass, and shame them by legal and/or extra-legal means at every opportunity. It is this chain of reasoning that gives us the Social Justice Warrior.

Scott illustrates this with a Socratic dialogue between Sophisticus, who argues for fundamental value differences, and Simplicio, who argues that discerning the difference is actually pretty hard in practice, and in any case doesn't always yield meaningful policy differences. (Please read the entire article. Scott's mind operates at a higher level than mine usually does, so my summary of the discussion may be inadequate.) Sophisticus and Simplicio illustrate their differences with two examples: the moral vs information bearing use of the word "lazy", and the appropriate approach to the use of punishment in reducing crime.

My post is somewhat orthogonal to Scott's point in that I want to challenge a point on which Sophisticus and Simplicio appear to agree. Excerpting is hard (again, read the whole thing), but this exchange seems to capture their agreement:

Sophisticus: Oh, I know what you’re going to say. You’re going to say this is just a factual difference between me and the pro-punishment faction. They believe, as a matter of fact, that bad conditions discourage crime extra effectively, since some criminals who would be willing to take the risk of a nice Scandinavian-style prison would be scared off by a dark overheated cage. And I agree this is a possible axis on which people can differ, and that if you proved to me that this was true I could be persuaded to reconsider my views. But I have talked to people who have literally said the words “I don’t care how much it discourages crime or not, I want criminals to suffer.”

Simplicio: Okay. I agree that’s good evidence for your view.

Sophisticus: You…do? Really? I won one of these? REALLY?!

Simplicio: I guess.

Sophisticus: So you admit sometimes there are fundamental value differences?

Simplicio: Sometimes, yeah, I guess. But I want to be really careful with this. Humans are adaptation-executors, not fitness-maximizers. Only one person in a thousand could give the principled consequentialist defense of criminal justice that you’re giving here. The game theory necessary to understand the defense is only a few decades or centuries old, depending on how exactly you define it – but even chimpanzees need to discourage defectors. Since evolution couldn’t cram the whole principled consequentialist defense into a chimpanzee brain, it just gave us the urge to punish.


Basically, Sophisticus is arguing that while it is appropriate that punishments should be carefully calibrated to deter crime, it is not appropriate to use them to satisfy our base instinct for revenge. Simplicio counters that the base instinct may in fact deliver better calibration than an explicit calculation, and even when it doesn't, it's easier to remember. But he appear to stipulate that, yes, the rational point is to reduce crime.

Let me propose a thought experiment.

Suppose I give you a choice between two societies in which you may live. In Society One, you have, in any given year, a 1% chance of being a victim of a crime. (For those of you good at math, you can calculate from this that you have only about a 50-50 chance of experiencing crime in your lifetime.) For the purposes of our thought experiment, let's limit the class of crime to the non-crippling: e.g., you could be assaulted but not murdered or paralyzed; you could be raped but not mutilated or impregnated; your house could be burglarized but not burnt, and in any case you have insurance that caps your out-of-pocket loss to the deductible.

However, let's assume that in Society One, when and if you DID become a victim of a crime, you knew: that law enforcement would make no effective effort to protect you or punish those who had injured you; that you yourself were forbidden from using any violence to either prevent the crime or punish your attackers; that you would subsequently see the criminals walking around the neighborhood with impugnity, suffering no legal, physical or social punishment; that when you spoke of being a crime victim, everyone who's opinion mattered would shrug, or tell you to "get over it", or recommend that you "lie back and enjoy it".

Now let's consider Society Two. Here your annual personal risk of crime is double what it is in Society One: 2%. (Those of you good at math will quickly calculate that you will probably experience crime at some point in your life.) In contrast to Society One, however, when and if you become a victim, you know: that law enforcement will do everything it can to catch the criminals; that personanges of recognized high status -- clergy, politicians, etc. -- will visit you, shake your hand, offer their deepest sympathy and pledge their assistance; and that if you personally use violence to resist or punish your assailant, that use will be celebrated.

In which society would you choose to live?

You might observe that I'm basically describing a low social status person and high social status person in each society, and reply that status is a reliable predictor of having access to nice things: food, love, sex, and yes, physical safety. And this is certainly true in the real world. I'm offering you a hypothetical where you are asked to explicity trade status (specifically as a crime victim) for increased physical safety. Or vice-versa.

You might still reply as a fitness-maximizer: that in a world where community support, reliable law enforcement, and rights to self-defense did NOT reduce crime and our individual risk to it, then those things would be irrational and should be traded away for lower crime. Or you might say that in such a world, the value we place on those things would be de-selected for, evolutionarily speaking. But again, that's not exactly what I asked. I'm asking you, personally, given full knowledge of the scenario above but also given your own personal impulses, to make a trade-off.

This thought experiment is somewhat analogous to sex. Everyone (that I know of) agrees that the reason (if "reason" be the right word) that humans enjoy sex is that sex is the way the race got propagated. Exactly zero of us (that I know of) foreswear non-reproductive sex for that reason. Likewise, I may know that the sense of satisfaction I get from having my status reaffirmed by my community specifically by seeing them seek punishment (or "revenge", or "justice") against those who have unlawfully injured me is an evolved response that reduces my overall risk of victimization, thereby increasing my "adaptive fitness" or whatever, but I don't really care, and neither do you, probably. Even if you think you do, you probably don't have much personal experience with crime.

That said, I will allow that different people will make the trade-off at different points. In my experiment, I minimized both the severity and frequency of crime, while dramatizing the loss of status. We could change these ratios such that far fewer of us would pay the price in victimization that the affirmation cost us.

So let me offer a real-world example, one that involves some controversy. I will assert the following fact:

At no point in history, from our earliest settlements to the present, did the material circumstances of American blacks not exceed those of their counterparts in sub-Saharan Africa by any objective measure.


I concede this is a bold statement. Even a right-wing, racial-realist like me knows that the black American experience involved some pretty tough times: slavery obviously, but also lynching, poverty, crime, you-name-it. My argument is that this must be measured against the black African experience, which in my estimate is a non-stop story of pestilence, disease, famine, starvation, tribal massacres, and getting stomped into yogurt by elephants. Now, an economic historian might be able to compare some particular moment of African flourishing, say 19th century Ethiopia, with the condition of Mississippi slaves and sharecroppers during that time and find the latter wanting. Maybe this is true. But I will stand by the generality that the actual physical circumstances of American blacks on average has been much better than that of sub-Saharan African blacks. (I exempt from this consideration the Atlantic passage itself, which by all accounts was uniquely horrible.)

And yet . . . very few American blacks, and very very few blacks whose thoughts on the matter get broadcast on teevee, are especially appreciative of this fact. Some of this is a function of incentives: there is no money, power, or status derived from being appreciative of anything, not in our present society, not for minorities. Part of it is that we none of us compare ourselves to distant peoples, we only compare ourselves to those around us: no African peasant thinks much about Americans in our SUVs, he only knows that he has one goat and his neighbor has two and that isn't fair.

But part of this lack of appreciation is that material circumstances, beyond a certain threshold, are not the total of our sense of well-being. There is something especially galling about the low status of slavery and Jim Crow, independent of its material effects. So it matters little that you ate better on a Virginia plantation than in the African bush; in the bush, you could hunt as you willed, while in Virginia you did as your taskmaster told. It matters little that, say, slave-rape was a minority taste; you knew that it might happen, and if it did, you could not resist, and no one would resist on your behalf. It matters little that far more blacks were murdered by fellow blacks in New York City during the Dinkins administration than during the entire history of lynching; what matters is that if a white person made an accusation against you, there would be no appeals to law or process.

If any of this sounds familiar, it's because I specifically modeled my description of Society One, above, on an account I read of life in the South before the Civil Rights era. Which brings my point full-circle. Few of us, and exactly zero calling themselves "progressives", expect comparisons to life in Africa to inspire much in the way of gratitude from the descendants of slaves. So we shouldn't really expect appeals to "fitness maximization" to substitute for punishment for its own sake, not to the satisfaction of the vast majority of people, including me.

No comments: