In response to Tyler Cowen, if the alternative hypothesis to his null that Old Keynesian models have failed is New Keynesian models, and he has rejected the Old in favor of the New, then I don't have many problems with his overall conclusion (which is not to say I agree with every detail of his argument). I thought his alternative hypothesis was broader than just the New Keynesian model, i.e. that he was arguing against Keynesian models of all varieties, but he says "I very much prefer New Keynesianism over Old." So if he is really saying the data support the New Keynesian model, I don't have much to disagree with. (See here for a post highlighting the difference between Old and New Keynesian IS-LM models. I posted this when people tried to claim I support Old Keynesian models as a way of discrediting what I have to say, and I've posted lots of New Keynesian work on fiscal multipliers as well. But people like Williamson, who Tyler points to authoritatively for reasons that escape me, still make the false charge that I promote old-fashioned Keynesian ideas.)
A few notes:
People seem to forget that the federal fiscal policy efforts were almost entirely offset by declines in spending and/or tax increases at the state and local level. Given that, it's not clear why we should expect to see a big effect in the data on output and employment. Fiscal policy at the federal level simply stopped things from getting even worse that they already were -- the bottom would have been much worse without it. Thus, when natural recovery finally began to take hold, it did so from a higher base than without the federal effort (and perhaps started sooner). Think of it this way -- fiscal stimulus allowed us to hold ground we would have lost otherwise -- again things would have been much worse without it -- until the natural recovery process was ready to begin.
I don't see how the fact that the economy is presently recovering at a rate where we will get to full employment by 2019 (or a few years earlier with very optimistic projections) says much about the effectiveness of fiscal policy. It kept things from getting worse, then it ran out, and now we are still looking at a relatively slow recovery by historical standards. What we want to know, but won't find out due to opposition in Congress, is if the recovery would be even faster from this point forward with additional fiscal policy efforts. Nobody ever said the economy wouldn't recover without stimulus, it's the rate of recovery that is at issue. Past efforts have kept GDP and employment from declining even more and made it easier for the natural recovery process to take hold, and additional fiscal policy timed correctly could have helped even more.
On the "timed correctly" point, people also seem to forget about policy lags. The same people who were arguing that infrastructure spending would take too long, that by the time it took hold the economy would already be recovering and it wouldn't be needed, now criticize policy as though it happens instantaneously. It doesn't. How much of the recovery is being driven by the lagged effects of our fiscal policy efforts? That will need to be teased out of the data -- a difficult task since monetary policy easing was going on at the same time and those effects have to be separated from fiscal policy and other factors that affect output and employment. For example, a firm that sees extra spending as a result of tax cuts may do a bit better than otherwise, and then decide to invest in an expansion of the business. It takes time to realize things are a bit better, plan the expansion, and then build it. The expansion is properly attributed to fiscal policy efforts, but this is very hard to see in the data (note that many of the tax measures are still in place, and that spending can also generate these types of effects). Tyler says "Frankly, it is a bit of an embarrassment for many commentators that the (admittedly weak) recovery is coming right after the end of the fiscal stimulus," but I don't see why that necessarily proves the case. Again, past policy efforts allowed us to take off from a higher base, and likely sooner than otherwise, and policy lags (plus the continuation of many tax breaks) imply that fiscal policy could still be active. I think the main effect of fiscal policy was to stop things from getting worse, I am not saying that fiscal policy is still necessarily present to a significant extent, only that we can't rule out that it is still helping without doing the economtric analysis.
Finally, in passing, liquidity traps exist, at least in theory, in both Old and New versions of the model (e.g. see the discussion in Carl Walsh's text on monetary economics). I disagree on with Tyler's point on the liquidity trap -- I think the evidence suggests we did enter a liquidity trap and that it is still a problem -- but in any case the failure to find a liquidity trap does not distinguish one model from the other (though to be fair, it's possible to construct versions of both models where a liqudity trap does not exist, but this is easier in the New Keyneisan model than in thye Old).
Update: Paul Krugman (this is part of a longer post):
The big winners, apart from the American public?: real business cycle theory.
Oh well. I guess we’ve always been at war with Eastasia.