Over issues of little consequence, all the various flavors of economists can have fierce debates over issues such as the proper role of government. Libertarians might stake out one extreme position, someone will argue the other side, and there are always at least a few who will take the middle ground. But when it's a policy issue of critical importance, such as in the wake of a financial crisis, the differences among us are much less apparent.
Transparency - the availability of essential information about the market - is needed for markets to function optimally. A key point of this argument from Arnold Kling is that financial markets replace transparency with trust by necessity, and this makes financial markets inherently unstable. The problem is that trust is fragile, it can be broken easily, and thus some sort of insurance is needed to overcome lack of trust and keep these markets functioning smoothly. One way to do this is through government provided deposit insurance, and I fully agree that this is a critical element in maintaining trust in financial markets (and as noted below, not all events can be anticipated and this means that some interventions to provide the necessary insurance will necessarily be ad hoc). But the main point I want to emphasize is that government must step in when there is a danger that trust will be significantly eroded by the failure of financial markets:
What Should Government Guarantee?, by Arnold Kling: ...On financial markets, I am in the middle between the Progressive view that government can guarantee everything and the libertarian view that the government should guarantee nothing.
My view is that financial markets are inherently unstable, because financial intermediation inherently replaces transparency with trust. If my bank were perfectly transparent, then I would know everything about its loans, including the underlying risks of the real estate developers, small businesses, and individuals to whom it is lending money. But in that case, I would not need a bank--I could just make those loans myself. So if you assume perfect transparency, you assume away the need for financial intermediation.
In fact, you have to assume the opposite of perfect transparency. You have to assume highly imperfect transparency, with reputation and trust serving as substitutes.
In banking, deposit insurance helps facilitate trust. A private insurance pool might work, but people trust government-provided deposit insurance even more.
With deposit insurance, the consumer loses all motivation for worrying about the bank's risk management. By the same token, the insurer has to worry a lot. In the U.S., the FDIC has been getting better over the years, but you can never get complacent. Any system can be gamed eventually, so it's a challenge for the regulators to stay one step ahead of the banks.
What we see with Bear Stearns, Freddie and Fannie, Lehman, and AIG insurance are institutions that are not FDIC-insured banks where nonetheless a question arises about whether some of their creditors ought to be protected by the government. I think that just about everyone is unhappy that these decisions are being made ad hoc, after the firms got in trouble, rather than having rules set ahead of time. But maybe what the government is doing is actually pretty reasonable. ...
Overall, I think that having some regulated, insured institutions, like the banks, is good. I don't think we can or should try to regulate everyone. Regulators should try to anticipate crises and prevent them. But almost by definition, the crises that do occur will be ones that they did not anticipate, and the responses will have to be somewhat ad hoc.
My grandmother lived through the Great Depression. After she passed away, we found money hidden all over her house - I'm sure there was some we never found, maybe buried in the yard or something. Even with deposit insurance, after the experience of the Great Depression she never trusted banks again, and nothing could convince her otherwise. That lack of trust takes assets that could be used productively to finance investment projects and hides them in the house or yard. And when you spread this behavior over millions and millions of people, the result is lower investment and lower growth.
In addition, when trust evaporates, the withdrawal of assets from the financial system is not limited to households with relatively modest savings worried about their bank deposits. People and businesses with large accumulations of assets do the equivalent of hiding their money in the cookie jar, and this can cause investment markets to dry up very fast. And as my grandmother's case shows in its own small way, once trust is gone it can take a long time to be reestablished, if ever.
The consequences of a big financial crash are not necessarily temporary, it is not simply a case of wiping out that which needs to be creatively destructed and moving on, the damage to trust can be permanent, and if it is, the consequence will be lower investment, lower growth, and fewer jobs than if the trust-busting crash had been prevented. The cost of, say, a quarter percent lower growth for 25 years is large, far larger than the cost of a typical bailout, and the costs do not fall solely on those who made the choices that caused the problems, the costs fall on all of us.
Update: Given the above, and this, it shouldn't be a surprise that I think this is a good idea:
Fed Readies A.I.G. Loan of $85 Billion for an 80% Stake, NYT: In an extraordinary turn, the Federal Reserve was close to a deal Tuesday night to take a nearly 80 percent stake in the troubled giant insurance company, the American International Group, in exchange for an $85 billion loan, according to people briefed on the negotiations.
In return, the Fed will receive warrants, which give it an ownership stake. All of A.I.G.’s assets will be pledged to secure the loan, these people said. ...