Thoma concludes with the important point that even if a government intervention loses the taxpayer money on the toxic assets themselves, it may still in a broader sense be a good strategy, once we account in our risk function for the high probability of very severe economic disruption absent any intervention.
Suppose we agree, on these grounds, that some intervention is better than no intervention, even if it loses the taxpayer money in the narrow sense of a book loss on the toxic assets. The question then shifts to which of the plans on offer is best (or least bad)?
In terms of the short-term book loss to the taxpayer, I doubt there will be a huge difference between the plans. As a Deutsche Bank economist memorably put it: "Ultimately, the taxpayer will pay one way or another". This is surely true. I don't want to trivialize such differences as there may be - this may still add up to a huge amount, which could literally have been spent saving lives - but only to emphasize that this is unlikely to be large relative to the costs of inaction for the broader economy.
Why then should we care about precisely how taxpayers end up picking up the tab? To my mind it is crucial here to recognize two things about the present crisis. First, the very fact that so many of the assets held by banks are toxic in the first place has much to do with the behaviour of agents within the financial sector and the regulatory environment and business and cultural norms within which they operate. Second, the entire game is iterated: whatever we do now will provide the backdrop for further rounds of play.
Going back to the beginning of the Thoma analogy:
Imagine a car lot that has 100 cars on it. However, some of these cars have problems. Half of them will have engine troubles that total the cars - the engines blow up and the cars are then worthless - and this will happen just after purchase. The other half are perfectly fine.
Now imagine something further (and I realize I'm stretching the analogy a little): what if the engine troubles that some of the cars will experience were caused by employees in the dealership swapping out good engine parts for bad ones? Or if all the while they had been currying favour with the local authorities and pushing for a relaxation of the town's laws prohibiting such activities? Or even that they were using op-ed pieces in the local newspaper to argue that this new way of selling cars was fundamentally better, and that the old-fashioned car dealer in the next town who suggested otherwise was an old fogey? And that the local dealer whose firm is now in such trouble that it is threatening the very social and economic fabric of the town continues to live in the biggest mansion and drive the fanciest car in town (and opines in the local paper that everything's fine)?
If this were the broader context within which the townspeople had to decide on a rescue plan for the dealership, I suspect many would lean towards some version of the "Saab" plan (i.e. Swedish plan or variant thereof) so as to mitigate the chance of another batch of exploding cars appearing on the streets of the town in the near future.
The key point I want to emphasize is this. Once we recognize the role of incentives and regulatory capture in the crisis, and the fact that once the immediate problems are dealt with the game will start again, this gives us another element to include in the objective function through which we evaluate competing policy proposals. What do such proposals do to incentives going forward? What do they say to players preparing to take part in the next round of the game? The magnitude of the economic fallout from inaction would almost certainly dwarf book losses taxpayers suffer on toxic assets. So it's vital we take action. But the magnitude of future losses from the wrong action, might dwarf even the current crisis. So it's also important we get intervention right.