The Illusory Arbitrariness of Deploring?
By Bryan Caplan
I’ve previously argued that public deploring is exceedingly arbitrary. The “outrages” we hear about on the news – the outrages that “We cannot tolerate!” – are usually no worse than dozens of other problems that we barely acknowledge or discuss. In my words:
When I witness the unbearable arbitrariness of deploring, two unsympathetic types of explanations come to mind.
First, people’s negative emotions depend far more on the vividness of the evil than its badness. A hundred stories about celebrity harassers would upset the world far more than ironclad statistical proof that 10% of celebrities harass. Indeed, it’s likely that one detail-rich story about a celebrity harasser would upset the world more than the best statistical study ever performed.
Second, people’s negative emotions are intensely social. People don’t want to rage alone. They want to get mad with their friends and countrymen. So when a new round of ugly stories pop up, almost no one asks, “Is this really the best target of our collective anger?” Instead, they jump on the bandwagon. Who cares where we’re going, as long as we’re united in negativity?
As examples, I name the crusades against chemical weapons and sexual harassment. I don’t see why chemical weapons are worse than the endless alternative methods of mass murder, and I don’t see why sexual harassment is worse than the endless alternative ways of mistreating employees and co-workers.
Yesterday, however, renowned rationalist Scott Alexander defended the basic rationality of the conventional view. Unlike most people, however, Scott doesn’t try to defend conventional priorities. Instead, he finds a hidden wisdom of effective deterrence lurking beneath the surface. Scott:
I have a different theory: people get upset over the violation of already-settled bright-line norms, because this is the correct action if you want to use limited enforcement resources efficiently.
Imagine a town with ten police officers, who can each solve one crime per day. Left to their own devices, the town’s criminals would commit thirty muggings and thirty burglaries per day (for the purposes of this hypothetical, both crimes are equally bad). They also require different skills; burglars can’t become muggers or vice versa without a lot of retraining. Criminals will commit their crime only if the odds are against them getting caught – but since there are 60 crimes a day and the police can only solve ten, the odds are in their favor.
Now imagine that the police get extra resources for a month, and they use them to crack down on mugging. For a month, every mugging in town gets solved instantly. Muggers realize this is going to happen and give up.
At the end of the month, the police lose their extra resources. But the police chief publicly commits that from now on, he’s going to prioritize solving muggings over solving burglaries, even if the burglaries are equally bad or worse. He’ll put an absurd amount of effort into solving even the smallest mugging; this is the hill he’s going to die on.
Suppose you’re a mugger, deciding whether or not to commit the first new mugging in town. If you’re the first guy to violate the no-mugging taboo, every police officer in town is going to be on your case; you’re nearly certain to get caught. You give up and do honest work. Every other mugger in town faces the same choice and makes the same decision…
The police chief’s public commitment solves mugging without devoting a single officer’s time to the problem, allowing all officers to concentrate on burglaries. A worst-crime-first enforcement regime has 60 crimes per day and solves 10; a mugging-first regime has 30 crimes per day and solves 10.
But this only works if the police chief keeps his commitment. If someone tests the limits and commits a mugging, the police need to crack down with what looks like a disproportionate amount of effort – the more disproportionate, the better. Fail, and muggers realize the commitment was fake, and then you’re back to having 60 crimes a day.
I happily grant that Scott’s story is logically possible. But I see minimal real-world relevance of his thought experiment. Sticking to his example:
- In the real world, muggers can become burglars.
- In the real world, deterring muggers more requires deterring burglars less. Zero-tolerance for muggers means a free hand for burglars.
- In the real world, mugging and burgling are not equally bad. If authorities choose to prioritize one, it will be because of vividness and herding, not actual badness.
To be fair to Scott, he does argue that his model successfully explains the standard approach to both chemical weapons and sexual harassment. But I find neither of his analyses convincing:
This looks to me like what’s happening with chemical weapons. The relevant difference between chemical weapons and conventional weapons is that the international community made a credible commitment to punish chemical weapons use, and so far it’s mostly worked. People with chemical weapons expect to be punished for using them, so they rarely get used. If there are some forms of atrocity that are easier with chemical weapons than with conventional ones – ie a dictator with a limited arms budget can kill more people with a choice between chemical and conventional weapons than they can when restricted to conventional weapons alone – then the taboo against chemical weapons saves lives. And so when a dictator tests the limits by trying a chemical weapon, it’s worth responding to that more forcefully than if they used conventional weapons to commit the same massacre. You’re not just preventing the one attack, you’re also acting to enforce the taboo.
I agree that chemical weapons rarely get used. But has the chemical weapons ban actually reduced total number of war deaths? On the contrary, it probably increased total war deaths because (a) chemical weapons aren’t especially cheap or effective, and (b) the ban was the key rationale for the Iraq War. Furthermore, the mere fact that almost no one reflects on this distinct possibility confirms that the ban derives from vividness and herding, not shrewd deterrence strategy.
The sexual harassment situation seems like the same dynamic. We can’t credibly demand our elites are never jerks to their subordinates – jerkishness is too vague a concept, there’s too much of it around, and it’s just not really an enforceable norm. But we have sort of credibly demanded our elites don’t sexually harass their subordinates, and it seems like we might be getting enough of a coalition together to enforce this in a lot of cases. If we can solidify this into an actual social norm, such that the average elite expects to be punished for sexual harassment, then elites will stop sexually harassing their subordinates and we won’t have to keep calling the whole coalition together all the time to enforce the punishment.
This is a more promising case for Scott: Governments that can’t kill their enemies with gas will simply bomb them, but sexual predators that can’t harass probably won’t switch to asexual emotional abuse. But again, the obvious question to ask is: Has the taboo on sexual harassment actually raised overall job satisfaction? It’s possible, but how many people with strong feelings on the subject have ever perused the numbers? And this is an especially strange position for Scott to take, because he’s blogged extensively on the vagueness of romantic-sexual norms. There’s probably stronger agreement that “Bosses who publicly scream at employees are jerks” than “Bosses who date employees are presumptive sexual harassers.”
All this aside, the fundamental question is not whether selective extreme reactions to chemical weapons or sexual harassment has done some good, but whether it was the best possible use of the moral energy and resolve employed to fight them. As far as I can tell, Scott claims nothing like this. True, that’s a high bar. But almost no normal person even pretends to try to meet it. Instead, they pick their targets in exactly the way my original piece described: vividness and herding. If you lose sleep over Moloch, isn’t this precisely what you should expect humans to do?