Keynes and Sticky Prices: Time to Think Outside the Box

Several recent excellent posts have appeared on Keynesian economics and sticky wages and prices. David Glasner points out that
...the sticky-wages explanation for unemployment was exactly the “classical” explanation that Keynes was railing against in the General Theory.
and quoting David again
it’s really quite astonishing — and amusing — to observe that, in the current upside-down world of modern macroeconomics, what differentiates New Classical from New Keynesian macroeconomists is that macroecoomists of the New Classical variety, dismissing wage stickiness as non-existent or empirically unimportant, assume that cyclical fluctuations in employment result from high rates of intertemporal substitution by labor in response to fluctuations in labor productivity, while macroeconomists of the New Keynesian variety argue that it is nominal-wage stickiness that prevents the steep cuts in nominal wages required to maintain employment in the face of exogenous shocks in aggregate demand or supply. 

Paul Krugman takes off from David's post and argues that
...even if you don’t think wage flexibility would help in our current situation (and like Keynes, I think it wouldn’t), Keynesians still need a sticky-wage story to make the facts consistent with involuntary unemployment. For if wages were flexible, an excess supply of labor should be reflected in ever-falling wages. 
Simon Wren-Lewis  takes up the torch.  He argues that
“the evidence that prices are not flexible is ... overwhelming”.
Lets look at some facts. In contrast to Simon's assertion; the evidence from the Great Depression is that wages and prices are remarkably flexible. During the first six years of the Great Depression, nominal wages and nominal prices fell by thirty percent.  Here is a graph of the normalized CPI and a normalized wage index that I constructed using aggregate data on compensation to employees (the details are in my book Expectations Employment and Prices and you can download the data here)
Of course the fact that the CPI and the wage moved in lock step means that the real wage did not fall and that, I would guess, is the fact that Paul and Simon would point to.  But that is very different from there being "overwhelming evidence" in favor of sticky prices.

The new-Keynesians have tried to discipline their models by looking at micro data on the frequency of price changes.  Here again; the NK model falls short. Klenow and Malin find that prices in the micro data are simply not rigid enough to explain the aggregate data.  Keynesians often cite Truman Bewley's 1999 study as evidence in favor of downwardly rigid nominal wages but a piece by Tomas Hirst (posted over at Pieria by Marco Nappolini) draws our attention to recent work by Elsby Shin and Solon which casts serious doubt on the relevance of Bewley's  finding. ESS find, that in the US, the wage data a surprisingly high frequency of nominal wage reductions.
and for the UK, the authors of previous British studies of nominal wage change, [the authors]  are struck by the apparent flexibility of British wages.
So lets get back to what's really important. High and persistent unemployment is a problem.  But it has nothing to do with inflexible wages or sticky prices. Both Classical AND new-Keynesian models are broken. It's time to think outside the box!