If NGDP falls 8% below trend, then nominal wages must also fall 8% below trend to maintain full employment, regardless of what is happening to real wages.
So I think his story of the past year really is a sticky-wage story. The Fed tightened (he is not backing down from that view), so that nominal GDP growth went from 5 percent to -3 percent. Meanwhile, nominal wage growth did not drop nearly as far, and therefore unemployment shot up. If we cannot explain the rise in unemployment by showing that real wages increased much faster than productivity, that does not trouble Sumner, because he does not believe the price indexes that are the denominator in real wages. (Note: aggregate wage measures in the numerator are indexes, also. Why do we believe them?)
Here is where I think our views differ. In my view, it is not possible to have a sudden nominal shock. Instead, only long-term changes in monetary policy can affect the trend rate of inflation. So I would explain the sudden slowdown in nominal GDP growth as a reflecting primarily a slowdown in real GDP growth, due to a real shock.
That view puts me in a minority. Most people believe that there was something like a nominal shock last year. However, Sumner is in the minority in attributing the nominal shock to monetary policy. Even though over the last thirty years, the textbooks have described nominal shocks as monetary.
I think the general view is that we had a nominal shock that consisted of the financial crisis. This drove down expectations for inflation, and the Fed could not reduce real interest rates sufficiently in that environment. Therefore, we needed quantitative easing and/or a fiscal stimulus.
Sumner's challenge with me is to get me to believe in nominal shocks again (I used to, back when I was in graduate school). He is welcome to keep trying, but he probably should not get too worked up over it, because (a) what I personally think is not that important in the grand scheme of things, (b) we cannot seem to agree on an empirical way to settle the issue, and (c) I seem to be pretty stubborn.
On the other hand, his challenge with the rest of the profession (I suspect including James Tobin, if Tobin were alive) is to convince them that the nominal shock was tightening of the money supply by the Fed. Maybe folks are just as stubborn and just as unable to settle the issue empirically, but there are a lot more people who have the view that there was some non-Fed source of the nominal shock, and so a lot more minds are at stake.
In order for the price level to have fallen enough to keep real expenditure equal to its long run trend, substantial deflation was necessary-- a 2 percent annual rate of deflation in the third quarter of 2008. For the fourth quarter of 2008, the needed annual deflation rate was 8.4 percent. During the first quarter of 2009, the needed deflation rate would have been 5.5 percent. The needed deflation rate in the second quarter 2009 was 2.4 percent. And finally, deflation would need to be approximately 1 percent in the third quarter of 2009. The problem was not too much deflation, it was rather too little deflation!
Emphasis in the original. I guess the idea is that there was a downward shock to monetary growth, and lower inflation would have insulated real GDP from that shock. But if you believe that deflation would have kept real GDP on track, then I do not see how you can explain the drop in real GDP on the basis of locked-in debt contracts and wage contracts. If you believe in locked-in debt contracts and wage contracts, then deflation should have made real GDP worse, not better.
I still need help understanding this concept of a nominal shock.