Kant’s Joke—Kant wanted to prove, in a way that would dumbfound the common man, that the common man was right: that was the secret joke of this soul. He wrote against the scholars in support of popular prejudice, but for scholars and not for the people.
This is a link to Joshua D. Green's very well written paper 'The secret joke of Kant's soul'. I don't know if any Economists have commented on it but I imagine the point they'd make is that 'following rules'- i.e. something which looks deontological- may be optimal from the 'Consequentialist' point of view and, equally, that contested, or competition between, rule-sets militates for meta-rules which aim to maximize the information set and subsume all consequentialist calculi under that rubric. In other words, though the Socio-biological considerations Green highlights may indeed be illuminating in their own right, what they don't have any bearing on are the 'distinctions without differences' that constitute Philosophy's realm of discourse.
To see why consider
a) the Trolley problem- you can either let a trolley kill five people or divert it such that only one person dies.
A moment's reflection will show that both the consequentialist and the deontologist should seek to maximize loss of life by their action because this will
1) prompt the Railway company to spend more on safety thus saving more lives in the long run. Or if the accident was a one off, then in any case the Investigation is likely to expand the Social Information set in such a way that all Consequentialist decision making is advantaged.
2) Highlight the importance of deontics and prompt a debate that might lead to exciting breakthroughs in the subject etc- for the higher duty is to duty itself & promoting deontics is the most pressing duty incumbent on deontologists.
Since both Consequentialism and Deontology, by always counseling the most reprehensible action possible (thus attracting the interest of the hoi polloi the way Freakanomics does), always arrive at the same conclusion, they can be unified under the rubric of optimal decision theory over some fitness landscape.
More generally, both continue to be central to Ethics- i.e. the project to 'shit higher than your arsehole' (Wittgenstein) in the manner that maximizes public nuisance.
b) The Footbridge problem- ought you to throw a fat man off a footbridge to halt the trolley from killing 5 people?
Clearly, you have to shove the other guy off the footbridge before he works out that he has to shove you off the footbridge. Which of you actually splatters on the ground is determined in a Darwinian way- which is good for the species.
The objection may be made that, assuming Muth Rational Expectations and equal physical endowment, you and the other guy instantaneously realize that the Nash equilibrium is to shake hands, smoke 'em if you got 'em, and video the carnage on your respective camera phones.
However, Ethics must never counsel a course of action that doesn't maximize avoidable loss. Hence it is enough for there to be some probability, even if everybody has Rational Expectations and perfect information, that a person might suddenly act in line with Ethical thinking for both of you to do your damnedest to grapple with each other such that both of you are likely to go off the bridge either
1) not in time to avert the carnage but to add gratuitously to the body count.
2) one or both of you go over the bridge but the Railway company is able to avoid implementing Safety procedures because the motivation for your fight is not known- the angle the Press play up is that it was some random fight between two strangers which led to a fatality.
Thus, assuming the other guy hasn't yet started trying to throw you off the bridge- in which case anything you do is either instinctive self-defense or panicked cowardice and thus outside the scope of Ethics- you can't be certain he will fight you and so you have to try to at least try to throw him off the bridge, assuming you are a moral person. Information asymmetry means you can't be sure he's a moral person, so you don't know for certain if he'll try to throw you off the bridge. This holds true even if he is thin and you are fat. He may not try to throw you off the bridge till you attack him.
The problem with existing Consequentialist and Deontological theories is that either
1) they fail immediately if all agents act upon them. Either someone has sacrificed you before you got a chance to sacrifice someone else or the Human Race died out in a Concurrency deadlock long ago. As the Indian proverb has it- in 'after you', 'no, after you', both missed the train.
2) they don't fail immediately.
But theories which don't fail immediately have two properties
1) they have mischievous consequences because they become availability cascades.
2) they undermine deontics because a duty is not a duty if it is known to have the pleasing property of being logically consistent or sensible in any way.
This yields Iyer's Iron Law- 'A theory only counts as part of Ethics if it robustly counsels good people to do the very worst thing possible under any and all circumstances'.