It depends what you mean by nationalize

Not even the most generous observer would say that the Treasury and the Fed have done a good job of explaining the thinking behind their financial stability plan. I think this note by Douglas Elliott does a much better job than they have. If you want to understand the logic of the administration's position, have a look.

On the underlying analysis, Elliott is with Geithner and Bernanke most of the way, which lately puts him in the minority of commentators. The main difference is that he is willing to say that outright nationalization might be necessary for "one or two" large banks, whereas Geithner and Bernanke persist in giving the impression that they will avoid this if at all possible. Most economists who follow this subject now seem to be arguing for outright nationalization of many banks--or, as they put it, "the banks" (choose your own number). People often seem to be talking past each other. Elliott persuades me that some of the discussion of the Fed/Treasury plan is muddled by terminology.

This week Bernanke said nationalization is "when the government seizes the bank, zeroes out the shareholders, and begins to manage and run [the] bank, and we don't plan anything like that... I think that this debate over nationalization kind of misses the point." ("Bernanke Pushes Back Against Nationalization", said the headline in the WSJ.) But he does not exclude bigger public stakes (with taxpayer upside) and strong control rights: indeed, that is what he is proposing. So, as Elliott says, a lot depends on exactly what you mean by nationalization.

Elliott also explains the "franchise value" and "legal complexity" arguments for avoiding full nationalization unless it is really necessary; Bernanke mentioned those points in testimony this week but did not elaborate. The paper carefully goes through the pros and cons of full nationalization. I've been leaning towards that solution but the paper did make me pause. This is how it wraps up.

It] may prove necessary as a last resort for one or two of the larger banks, but should only be undertaken when, and if, it is clearly necessary. More widespread nationalization is unlikely to be needed unless the economy performs substantially worse than most economists expect. Although we all crave certainty, it would be better to wait until we knew that this pessimistic case was likely before nationalizing more widely, given the serious social and financial costs of that extreme step.

Beyond the very small number of full nationalizations that may prove necessary, it may well make sense for the government to take substantial stakes now in additional large banks in a form that some may view as a partial nationalization. The Administration's plan to perform a rigorous, uniform "stress test"on the nation's largest banks is a good one...

[If] it becomes clear through the stress test that a bank is already insolvent or is at high risk of becoming insolvent, then it would be better to go directly to the step of full nationalization...

Many observers would say that most of the banking system is "already insolvent" or "at high risk of becoming insolvent". The Fed and the Treasury evidently think otherwise. ("If a bank does become insolvent then the FDI, of course, will intervene," Bernanke said on Tuesday, "but we're not close to that. All the banks are above their regulatory capital.")

Perhaps Citi and Bank of America--to name two large banks--are or will be insolvent on any plausible scenario; for the rest of the industry it still hinges on the timing and strength of the recovery. Hence the Fed/Treasury stress test. Is the "adverse" case in that exercise--an output decline of 3.3% this year and 10% unemployment in 2010--adverse enough to tell us what we need to know? I wonder. It's not difficult to imagine a worse case than that.

Anyway, take a look at Elliott's article. Even if you don't agree with the conclusion, I think you'll find it helps to clarify the issues.