Response to Sunstein

According to Cass, we are now allies. I am as willing to argue with my allies as with my opponents, perhaps more willing; allies are more likely to have enough in common with me to be willing to be listen to my arguments.

Hence this post.

Way back then, I thought that Posner and Epstein (and Becker and Stigler too):

(1) had a stylized, fairy tale account of government failure;

(2) were insufficiently empirical;

(3) were far too allergic to redistribution;

(4) underplayed the endogeneity of human preferences;

(5) underplayed deliberative democracy and the idea of the public interest; and

(6) had an unrealistic and a priori account of human rationality.

…

(By the way, I still think (1), (2), (3), (4), (5), and (6).

(Cass Sunstein, On Classical Liberalism)

(2) Insufficiently Empirical

Cass does not mention Coase, also a colleague of ours at the University of Chicago Law School and arguably a classical liberal, who would have agreed. He thought economists should spend more time looking at real world economic institutions in order to better understand them. As evidence for Coase’s position, I learned quite a lot relevant to the economic analysis of law in the process of researching a book on real world legal systems, past and present.

The criticism is a legitimate one but Cass, unlike Coase, is not in a good position to make it. One of the ideas he is known for is “libertarian paternalism,” structuring the choices presented to people in a way that sets the default at what the choice architect believes they should choose while leaving them free to choose something else if they wish.

I have some real-world evidence of how it works, detailed in a previous post.

Oberlin’s “optional” fees were what Cass had argued for. The default was to pay money for what they considered a good cause. You were free to not pay it — provided that you somehow discovered that option. Why, after all, should they spend time and effort making it easier for you to avoid doing something you ought to do?

Given the assumed situation, a choice architect who knows what choice you should make, it is what one should expect — and could have observed, long before Cass’s invention of “Libertarian Paternalism.”

(1) Government Failure

Our account of government failure, that it is a result of the same causes as market failure in other markets, situations where an actor does not bear the net cost of his actions hence where his interest is not our interest, is stylized, depends on the usual economic assumption of rationality — although if it is a fairy tale it is the kind with a witch and an oven; she too does not bear the net cost of her actions. But our stylized account turns out to explain quite a lot, not only the existence of government failure but much of its pattern. Does Cass have an account that explains the same features of the real world, behavior of actual governments, and more?

(6) Rationality

I too am an admirer of Kahneman’s Fast and Slow; one year it was our Christmas book.¹ What Kahneman did, at least as I interpret it, was not to refute the rationality assumption but expand it, include attention in the budget constraint, take account of the cost of processing information. The value of that expansion is that it explains more behavior. The cost is that it makes theories more complicated, reduces their predictive power, provides a new way of explaining away behavior that does not fit your predictions. This is not a new issue — economists routinely use more or less simplified versions of rationality according to what complications they expect to be relevant to the questions they are asking.

The test is how well the theory fits the data, in particular the data not available when the theory was being constructed. I have not yet seen any areas of economics that have been, by that test, significantly improved by the use of behavioral economics, Kahneman’s legacy, but perhaps Cass can point me at some. For the moment, while I concede that Kahneman identified a problem with the usual version of the rationality assumption, I do not see any way of improving my economics by incorporating his ideas.

One area where behavioral economics might prove useful is macro. The versions I am familiar with depend on people making the same mistake over and over; one of the things Kahneman explains is the pattern of mistakes made by system one.

Perhaps someone is doing it; macro is not my field.

(3) Redistribution

There are two arguments against government redistribution. One, which I do not expect Cass to find convincing, is a moral argument, that individuals are entitled to what they have honestly earned, obtained by voluntary transactions.

The other is the economic argument for the advantages of secure property rights. People arguing for redistribution usually think of it as arguing for some particular pattern, taking some amount from the rich to give to the poor. But what they are really proposing is a legal and political system in which one way of obtaining wealth is getting the government to take it from someone else and give it to you. Neither of us has any good reason to expect the pattern of transfers under such a system to fit our particular preferences but we both have reasons to expect that, under such a system, people well spend substantial efforts trying to arrange to be transferred to rather than from, the dead weight cost of rent seeking.

A further cost is particularly salient at the moment. The same economic arguments that imply that a nation gains by free trade also imply that it gains by free migration, open borders. Those arguments do not apply to a welfare state, since migrants can obtain an income not by voluntary trade but by transfer.

That is one example of a broader point. Under institutions of secure property rights and voluntary transfers, individuals gain by the existence of other individuals as potential trading partners. In a transfer society that is no longer true — individuals may correctly believe that it is in their interest to keep out people, potential free riders or rivals for power.

It might even be in their interest, given the power, to push them out.

(5) Deliberative Democracy

Deliberative democracy is a model of governance where legitimate law-making stems from public deliberation rather than just voting. It emphasizes informed, reasoned, and inclusive discussions among citizens to address complex issues, fostering better, more trusted policy outcomes. (AI Overview)

Neither public choice theory nor the observation of deliberation online — in thirty years online I have observed a few forums where discussions were informed, reasoned and inclusive, but they are vastly outnumbered by the other kind — provides support for this approach. Citizens arguing about policy have little incentive to seek truth and no feedback, or only feedback with a lag of years or decades, to tell them if they have found it. They have, in contrast, a strong incentive to take positions that the people who matter to them will approve of and immediate feedback to tell them if they have done so.²

(4) The Endogeneity of Human Preferences

I will have to get back to this one after Cass tells me what he believes its implications are for economics or public policy. At first glance it seems to argue for the stability of oppressive regimes able to mold the preferences of their population, perhaps also for control of the schools as the back door to power.

The Wrong Question

What was wrong with the policy arguments of economists circa 1960 was that they were answering the wrong question. Their implicit assumption was government as a benevolent actor, the question what it could do to make the world better. The alternative, the unregulated market, individuals acting in their own self-interest, produced outcomes some of which could be improved by a government provided with the necessary powers to intervene, so the government should be given those powers and use them.

The right question was not what government should do but what, given those powers, it would do. The right model for government was not a benevolent philosopher king but a political market, political actors — voters, politicians, lobbyists, government employees —acting each in his own self-interest. That model, public choice theory, produced a less optimistic account of the consequence of government power to intervene in the economy.

I am trying to deduce arguments from a list of criticisms, may be getting some of them wrong, but I think at least half are revivals of that intellectual error.

Sometimes in new dress.

---

My web page, with the full text of multiple books and articles and much else

Past posts, sorted by topic

A search bar for past posts and much of my other writing

1. The default present for people we didn’t have anything else for.

2. An earlier post with some empirical evidence.