I’m frankly stunned that Krugman would approvingly quote the following passage from Keynes:

[I]f a particular producer or a particular country cuts wages, then, so
long as others do not follow suit, that producer or that country is able
to get more of what trade is going. But if wages are cut all round, the
purchasing power of the community as a whole is reduced by the same
amount as the reduction of costs; and, again, no one is further forward.

Suppose I cut my cleaning lady’s pay by $10 per week.  What happens to the “purchasing power of the community as a whole”?  Her purchasing power clearly falls by $10.  But my purchasing power just as clearly rises by $10.  The net effect, assuming no change in behavior, is -$10+$10=$0. 

And that’s just nominal purchasing power.  When the price of cleaning services goes down, and all other prices stay the same, real purchasing power of the community is actually slightly higher.

Krugman’s next paragraph gratifyingly acknowledges this basic principle for creditor-debtor contracts:

And Fisher pointed out in 1933 that a general fall in wages and prices
actually makes things worse, by making debtors poorer in real terms;
true, creditors are made richer, but because debtors are more likely to
cut spending than creditors are to increase it, the overall effect is to
deepen the depression.

But the same goes for labor contracts, too.  In both cases, the sole Keynesian mechanism at work is differences in the marginal propensity to spend.  All talk about “reductions in communal purchasing power” is a red herring.

Consider: By Keynes and Krugman’s reasoning, mandating a 50% pay raise throughout the economy would be a fool-proof way to raise communal purchasing power by 50%.  It’s not.  Employers foot the bill, and their purchasing power falls just as much as workers’ rises.  Assuming, of course, that they don’t hire fewer workers or declare bankruptcy.  The interaction between income distribution and marginal propensity to consume could outweigh these effects, but I see no reason to think they would.

There is plenty of evidence that nominal wages are downwardly rigid.  There is plenty of evidence that government is far from the sole cause of this rigidity.  Basic labor economics says that if Aggregate Demand falls and nominal wages are rigid, you’ll get unemployment.  Why can’t Krugman stick to these simple truths, instead of maligning wage flexibility as the road to ruin?