Linda Bilmes presented an interesting paper (not online) at the AEAs looking at the fiscal stimulus in light of Katrina, Iraq and the Big Dig.  Here are some key grafs:

Bilmes is the co-author with Joseph Stiglitz of The Three Trillion Dollar War.

Paul Krugman has some questions:

Financial
institutions that want to “get bad assets off their balance sheets” can
do that any time they like, by writing those assets down to zero – or
by selling them at whatever price they can. If we create a new
institution to take over those assets, the $700 billion question is, at what price? And I still haven’t seen anything that explains how the price will be determined.

I suspect, though I’m not certain, that policymakers are once more
coming around to the view that mortgage-backed securities are being
systematically underpriced. But do we really know this? And how are we
going to ensure that this doesn’t end up being a huge giveaway to
financial firms?

Here is more detail on various plans.  I see it so:  If the assets are undervalued by the market, buying them up is an OK deal.  Presumably the price would be determined by a reverse auction, with hard-to-track asset heterogeneity introducing some arbitrariness into the resulting prices.  If these assets are not undervalued by the market, and indeed they really are worth so little, our government wishes to find a not-fully-transparent way to give financial firms greater value, also known as "huge giveaway."

Right now it seems to boil down to the original TARP idea or nationalization, take your pick.  You are more likely to favor nationalization if you think that governments can run things well, if you feel there is justice in government having "upside" on the deal, and if you are keen to spend the TARP money on other programs instead.

Here is one source:

Here is another:

One Calculated Risk commentator states:

Addendum: Do note that is the gross flow of cash from the government, not the net financial cost or the net social cost.

That's Paul Volcker's idea, namely that the way to prevent "too big to fail" is to prevent "too big."  Details are sketchy but Matt Yglesias offers some comments. 

Is there a precedent for this kind of plan working?  We have antitrust law but antitrust law doesn't actually prevent big firms from becoming big; we've had General Motors, Microsoft, Google, and others, all very large relative to their sectors.

You could imagine an absolute cap on the size of bank assets, so that above the size limit would-be loans and deposits are sent to a rival institution elsewhere by mandate.  One implication is that banks won't have to treat their customers very well, since in a growing economy (we'll get one again, sooner or later) a lot of banks will be at the cap and will be turning away extra business.  If different banks were perfect substitutes for each other, you wouldn't be seeing large banks in the first place.

A second question is whether ten little banks are safer than one larger, 10x bank.  For sure they are if the problem is one bonehead manager at the bigger bank, but what if it's systemic asset price risk?  The smaller banks could well be less safe.  In a financial crisis, would you rather be a larger country or a smaller country?

I believe the plan would require very tight restrictions on off-balance sheet activity.  Something like this may be coming anyway, and its rationale is understandable, but it is easier said than done.  We would be requiring regulators to estimate the net "size" (it's debatable whether that word even applies; what is the "size" of a naked put?) of a bank financial position when banks themselves haven't been very good at doing that.  A simple approach would be to ban all bank trading in derivatives but I believe that would increase bank risk more than decrease it, at least at this point.

By the way the 1927 McFadden Act banned interstate branch banking, in part to keep banks small, and economic historians usually consider that policy to have been a disaster which contributed to the severity of the Great Depression.

Here is a summary of where some of the debate on bank policy is at right now.

Is there a way out?

Buiter's argument — which applies just as much to the US as it does to the UK – is that the only way to get banks lending again is to nationalize them.  Otherwise, they'll simply hoard liquidity in a desperate attempt to avoid being nationalized.

That's Felix Salmon.  

It is a common shibboleth that saved funds mean a decline in aggregate demand but this doesn't have to be true.  Savings often fund investment, which boosts aggregate demand and creates jobs.

Admittedly, savings don't fund investment when the banking system is malfunctioning.  Or it may take so long to translate savings into investment that incomes are falling in the meantime and S and I follow them on the way down (Keynes's scenario).  Still, you shouldn't assume that savings translate into a collapse of aggregate demand.

Michael Mandel adds:

I believe that Obama’s $300 billion tax cut is essential to
‘recapitalize’ the American consumer, just like the banks are being
recapitalized.

With that as background, consider the tax cuts in Obama's stimulus plan.  If the money is spent, you get a boost to aggregate demand.  That was the goal.  If the money is not spent, it is a wash.  The government borrows for people (at a low rate) and people save it.  Since savings has gone up, the borrowing is sustainable and it doesn't even have to crowd out additional government spending, if that is what you want. 

Furthermore these people would have done some borrowing anyway, so their ability to implicitly borrow at a lower interest rate creates a small, positive wealth effect.  The savings also means you have supplied those people with some form of implicit insurance, and at very low risk of moral hazard.

I wouldn't expect a whole lot of recovery from these scenarios, but there's nothing problematic about having some tax cuts in the stimulus package.  If you're looking for another opinion, here is Joseph Stiglitz.