Friday, September 26, 2008

Regulation: Too Much or Too Little

One comment frequently made about the present financial mess is that it is the fault of deregulation. As I argued recently, it ought rather to be seen as the fault of regulation—more precisely, of government interventions in the housing market designed to make it possible for more people to borrow money in order to buy houses.

There is, however, a germ of truth to the claim. While I am not an expert, I gather that one source of the present difficulties was a loosening of the requirements imposed on Fanny Mae for loans during the Clinton administration. That made it possible to offer loans to less well qualified borrowers than before, a policy which increased the amount of business Fanny Mae did, satisfied political demands to expand home ownership, and helped lead to the present mess.

This raises a general point worth making. The ideal arrangement in my view, for housing and many other things, would be an entirely free market with the government playing no role. But once the government does intervene, less regulation is not necessarily better than more. If, as in the current case and the earlier S&L case, government intervention makes the government ultimately liable for losses by the regulated firms, less regulation may mean more opportunities for firms to gamble on the basis of "heads we win, tails you lose," with "you" being the taxpayers. Once the government is liable for losses, it may be prudent for the government to make rules designed to limit risk.


Anonymous said...

A deeper question to explore is why the managers of these firms took bad risks. One possibility is that they were ex ante good bets that happened to go sour. A more disturbing possibility is that management knew they were making bad bets, but their compensation was tied to financial performance in a way that made this profitable for them.

A CEO can boost the financial performance of a firm in two basic ways, one easy and one hard. The hard way is to be more skillful than other managers. But not everyone can be above average. The easy way is to take more risk. Taking risk is not an activity that should be rewarded, since it doesn't require skill. But it may be very difficult for observers to distinguish between luck and skill, and thus luck may be inappropriately rewarded.

Patri Friedman said...

Agreed. I'd say this is an instance of a genuine slippery slope. You have 2 equilibrium points, one stable, one unstable. The unstable one is laissez-faire. It is unstable because when you move off it into partial regulation, you tend to get failures and bailouts and socialization or increased regulation to prevent future bailouts.

Also, besides the deregulation you mention, 5 banks get SEC exemptions a few years ago to use 40:1 leverage instead of the usual 10:1. Those being Bear Sterns, Lehman, Merill, Morgan Stanley, and Goldman (I think). Notice that only 2 of these 5 are still functioning as independent entities, due to what happened when they were highly leveraged and the market moved against them.

If they'd had the capital ratios of normal banks, they might not have gone bankrupt and needed shotgun weddings, they might not have defaulted on their bonds (which is part of why AIG needed the bailout - because it insured bank bonds against default), etc.

Mike Huben said...

No matter what government does or doesn't do, the existence of capital creates incentive for fraud. And nothing will eliminate all fraud except perhaps extinction. (Therein lies a modest proposal, whereby government CAN solve the problem.)

So there will be messes no matter what government does or doesn't do. Since government has done something, it is easy to argue that the mess is "caused" by the government action. Even when the mess is multicausal. Even if there was a tradeoff of some sort, and the benefit is much greater than the mess. Or you can argue that government has suppressed the emergence of the utopian one true problem-free solution to everything. Both of those arguments strike me as silly.

But then, I don't agree with David.
Once Again, It Wasn't Fannie and Freddie

Anonymous said...

Another argument for regulation has to do with the need for bailouts, right? If there are some companies which are simply too big to be allowed to fail, then there's a strong reason to impose strict regulations on those companies, because they're gambling with my money, not just their own.

Anonymous said...

The government has a difficult time when it tries to give away money. The purpose of the GSEs was to redistribute money to certain homeowners. This could have been done by some kind of refundable tax credit. That would have been a low-risk way to do it. Instead, the government adopted the high-risk approach of guaranteeing mortgages. Now, home prices are falling and homeowners are walking away from their mortgages in droves. The government's on the hook. A lot of the money is owed to foreigners, e.g., the Chinese. So, clearly, this attempt to redistribute income has gone horribly awry.

Regarding the difficult mission of giving away money, previously we had the problem of rent-seeking, wherein it is worth spending up to $1 to collect $1 that the government is offering. We can now add this new example of the failure of government to hand out money. They took a very convoluted and indirect approach (guaranteeing mortgages). So, the pitfall is that the government uses a strange mechanism to hand out the money. The strange mechanism then blows up in an expected way, causing enormous losses. The question arises as to why the government adopted the convoluted and indirect approach. It's probably related to the fact that that government likes to conceal its true motives.

"Card Check" unionization may be coming, if Obama wins. Collective bargaining is another freako, defective way to attempt to redistribute income. No matter how you look at it, it's counterproductive. So if we get widespread unionization in America through "Card Check", we'll have another cool example of the government's failure in its attempts to redistribute income. This, too, represents some kind of attempt to conceal the fact that we're trying to redistribute income. It's another convoluted and indirect approach.

In these cases, we start out with counterproductive laws. Maybe the antidote is additional laws that somehow fix up the problems caused by the original bad laws. But it seems more likely that the real solution is to avoid making these crazy laws in the first place.

John T. Kennedy said...

" may be prudent for the government to make rules designed to limit risk."

What's the incentive for government officials to produce prudent rules? Acting on the idea that regulation will improve things aren't they a cinch to enact even more damaging regulation?

Leonard said...

JTK: there's no incentive except the indirect one, of voters controlling politicians controlling the bureaucracy, etc. It's so attenuated as to be meaningless most of the time.

There is always the desire of well-meaning govt employees to actually do their job as seems right to them. Craftsmen, as it were, of government, who don't care that they'll never see a reward for crafting a good regulation. This strikes me as the main reason why any good regulation happens. And I don't think it is trivial.

It seems to me that neither of these mechanisms is that strong. But a private corporation doesn't do much better. Its stockholders do, at least, have some incentive to push for good regulation. But it's pretty attenuated via board, CEO, and corporate apparat. And of course some insiders may be craftsmen of governance. My guess is there will be less of it, though, given more scrambling for a buck, although there will occasionally be effective control of the stockholders so it will even out.

I think the main mechanism causing the superiority of the free market is bankruptcy, seen as a mindless filtering mechanism. Over time, all the poorly governed corps go bankrupt, even though nobody understands exactly why they failed. The assets are then placed into someone else's management, and over time most assets end up well regulated. But this is not by human design, just human action.