Sunday, November 23, 2008

Second-Order Economic Effects

One thing I find so disconcerting in most economic discussions taking place right now is the lack of consideration for the second-order effects of policies. Almost every government policy that effects economic policy has some unintended consequence that creates new economic distortions which then create new problems which are then inevitably treated with a new distortive policy, and so if goes.

To that end, it was nice to see an article that discusses one of those distortions.

The truth is that most economic behavior can be distilled to three principles. One, people generally act in their perceived self-interest. Two, people respond to incentives. Three, people are smart enough to apply some level of game theory to the first two principles. That third principle is the source of second-order effects, and it is remarkable how little the third principle is given consideration when policy makers form policy.

Unfortunately, Keynesian thinking has become so ingrained in the government and the media tha the first reaction to any problem is to assume reflexively that a new government policy is needed to combat that problem. Often this policy comes in the form of some new regulation, but while it might feel emotionally satisfying to think that there is an office full of wizened gurus somewhere able to create perfect policy, it is impossible. Not just improbable, but impossible, as any distortive policy will have distortive second-order effects. Of course, even if it was possible, there's not much evidence the government is run by wizened gurus.

2 comments:

Restless Native said...

I agree largely with the points made in the article. As regulation seems to be on the tip of everyone's tongue (the gentleman who wrote this article seems to think banking is the most hevily regulated industry), I thought it would be interesting to mention the insurance industry regulatory monstrousity. While banking is largely regulated at the federal level, insurance is almost exclusively regulated at the state level. That means your average multi-line property and casualty insurance company (think Chubb) actually has to deal with as many regulators as states/territories in which it wants to do business. The Insuance Commissioner of Washington state can and does regulate Chubb on any policies issued by Chubb in the state of Washington. Moreover, given the patchwork nature of this regulation, most insurers of this type actually maintain several different "insurance companies" with licenses and filings in different states, depending on their needs and the type of regulation mandated in each. So AIG has/had Lexington to deal with non-admitted, or excess and surplus, policies, National Union Insurance to do D&O, Granite State for some other stuff, AIICO in Puerto Rico, etc.

Now, AIG got itself into some serious trouble, but did so in the relatively unregulated market in the UK and by doing CDSs which not even they understood. Also, all these hodgepodge regulators couldn't keep Reliance from failing, or Mission, or effectively Continental and Kemper, and those are just some of the US companies who have gone tits-up in the last 20 years.

You guys tease the insurance dudes about being the "Special Olympics" of finance, and I largely agree, but we have had to deal with 50+ regulators for a long time (certainly a drag on productivity, but apparently not all that much of a drag on creativity , see AIG's CDS unit mentioned above) and have managed to maintain a relatively wide-open market for about 100 years.

While I am certainly not suggesting that the Big Boy financial sectors go to a 50-state model for regulation (quite the contrary, I'd rather insurance regulation be unified at the federal level) I do think the chupacabra fear of sensible regulation spewed by the likes of the WSJ is unfounded.

Don't throw the baby (solid rules and enforcement) out with the bathwater (distortions created by unclear rules and unintended consequences).

Aztec Tomb said...

"Unfortunately, Keynesian thinking has become so ingrained in the government and the media tha the first reaction to any problem is to assume reflexively that a new government policy is needed to combat that problem."

That's a non sequitur. New government spending, maybe. Not new regulations though.