The reason that it's easier to get redistributive policies passed in areas of law like corporate law is that people don't understand corporate law. When you try to pass redistributive policies in areas of policy that people understand---like tax policies---they meet with more resistance.

Do you really think tax law is easier to understand than corporate law? At anything other than the most general level, that is almost certainly wrong (I strongly suspect a survey of random 2Ls would confirm that). At a very general level maybe there's something to it, but even there I wonder, for instance, how many people really get the difference between a flat tax and a graduated income tax.

Brett, I admit that I'm not familiar with the critiques of Kaplow and Shavell that you mentioned. But with respect to executive compensation, it would be mathematically impossible to reach a Pareto-optimal solution to the question "how much should an executive be paid" if you added a global wealth redistribution criteria.

Firm economics already dictate three difficult trade offs in executive compensation--management retention, mgmt/shareowner alignment, and shareholder cost. Adding a fourth consideration completely divorced from firm economics would make the board's position untenable. In analytical terms, there is no objective function that could, even in theory, satisfy all four constraints, unless the government created some sort of formula that would fix the price of executive contracts. I could hardly imagine how distorting that would be to both our economy and politics.

BTW - I'm not sure I read the first comment right. I certainly read it different than Brett. I think he was saying that we could get away with more social engineering via corporate law than tax law because so few people would understand the implications in corporate law. I agree with that, for the same reason that one would go along with that tack for "feasibility" reasons. But just because something is possible doesn't make it desirable.

Brett, I thought these were great arguments. I still haven't seen a good rebuttal -- have you?

You might also consider the possibility that the costs of administering redistributive rules might vary across systems. For example, along the lines of Kaplow's rules/standards analysis, if you can acheive similar amounts and accuracy of redistribution in a system that requires only a few transactions, rather than one that has thousands, you should pick the one with the small number of transactions.

I suspect this guideline would lead us to prefer redistribution outside of tax, but possibly that would be at some cost in accuracy, since much of the reason tax has so many rules is to achieve precision. As Kaplow & Shavell point out, unless we're in a strictly welfarist framework, it's hard to think about how we compare different cost/accuracy of fairness tradeoffs.

So, in short, there's still lots of interesting questions to ask in this area. Your next post, maybe?

BDG: Thanks for the comment. I won't be able to address it in the next post, in part because my visiting time here is up, but I will eventually be addressing this in one or more paper, so much more to say in the future.