By Adam Ozimek
I want to follow Karl’s comments about bailouts with some scattered thoughts of my own. If I read him correctly, I think he is arguing that bailouts with risk taking are a second best outcome, and that we can’t have both risk taking and no bailouts unless we want financial panics as well.
One response to his argument that bailouts are needed because we want firms to take risks is to point out that risk taking is important in all industries, and so we should have bailouts in manufacturing and retail industries to encourage more risk taking?
The difference is that it’s the natural state of financial markets, the argument goes, that risk taking will induce occasional panics. This is not true of other markets because consumers don’t panic in other markets, and the nature of fractional reserve banking makes it sensitive to panics.
One response to the notion that panics are the “natural state” if banking is an argument I’ve occasionally made, which is that regulations that separate banking from other types of commerce prevent experimentation with other models for finance that may not be subject to panics and may not require bailouts. I’m not optimistic that we will experiment with new business models here, but I think there is potential.
There are those who blame the Federal Deposit Insurance Corp. (FDIC) and the implicit guarantee of central bankers as lenders-of-last resort for the panics that occur in banking. I am not convinced by this. I believe that banking as it exists in most countries today and as it has existed in most forms in the past is subject to panics. I am not convinced, however, that this is true of all forms of the banking that could exist were it allowed to evolve. Of course one could argue that the costs of experimentation may be too costly to make it worth walking down that path.
A final point that I think is underemphasized today is that we should be concerned about profitable bailouts. No matter how necessary bailouts are at some points in time we want them to be costly both for those being bailed out and for those doing the bailing. Long in the future when the economy is back at full employment, the argument that “we bailed out banks before and made money on it” or “we bailed out G.M. before and made money on it” will be more salient than today’s popular anti-bailout sentiments are. It is an easy debate point that I am afraid will stand the test of time whereas the popular backlash will fade into memory.
When future politicians are deciding whether to bail out or not, it is best if everyone understands they will be voted out of office if they vote for a bailout. They already know that they will be voted out if they fail to bail out when it is necessary to prevent a panic, so the choice in a real panic will then be to end their career by preventing a panic or end it by failing to prevent a panic. I believe the desire to not end up the latter scenario will be enough to overcome coordination problems and freeloader issues.
In contrast, if you can profit on a bailout and create the illusion that you prevented a panic that never was, then I believe we will have too many bailouts. At the very least today’s bailout profitability provides an ex ante justification for one round of future bailouts, even if those bailouts are unprofitable. This is why it’s important that bailouts always come with a political cost, and having them be unprofitable is a good one.
You could argue bailouts will inevitably be followed by populist backlash that will cost politicians, but I think if we had a combination of profitable bailouts and low unemployment now the bailout would be quite popular. It is not bailouts per se that are unpopular, but bailouts that failed to prevent high unemployment. Bailouts in fake panics will always prevent high unemployment, and will thus be popular.
And in any case, even if ex post backlash is inevitable, today’s profitability will be more salient in the future than today’s backlash, which will provide future unnecessary bailouts with ex ante justification.