David Henderson looks at the export statistics and determines they were not a major factor. And of course, trade only grew slowly in the post-war period until later in the century. I have a few scattered thoughts, not driving at anything in particular:
1. I'm curious how the Marshall Plan is classified in the NIPA tables. It was a lot of loans so I'm not sure if that's done differently or if it's in government spending or what. Presumably it's not in the export statistics because I think the annual flow in 1948 was bigger than that. You see a lot of early Keynesians - I think Leon Keyserling, for example - talking about the Marshall Plan as the quintessential Keynesian policy. It rebuilds Europe and makes use of American industrial capacity: Economic Consequences of the Peace meets The General Theory.
2. Contemporary trade and expected future trade are not the same thing, of course. We just liberated Europe and established Bretton Woods, which everybody had high hopes for. The real surge was in private investment, and even though we weren't exporting much at the time the surge in private investment was a response to expected future demand. That is mostly consumption and investment demand, of coruse - but also expected future exports.
3. I used his new post as an opportunity to review his Mercatus Center paper, and I still have questions about the discussion of saving rates. We are in a growth period which means we are in a virtuous cycle analogous to the paradox of thrift. It seems reasonable that plans to release pent up demand have such a positive effet on income that income booms and people end up saving more than they expected to over a particular time horizon. It's the same as the paradox of thrift where everyone plans on saving, income contracts, and they end up saving less than they wanted to. I don't know if this is determinate at all but it seems like if we're discussing macro trends it's worth thinking beyond the headline savings rate.
The story is clearly private investment, though. I am still befuddled about how anyone thinks this is some kind of death knell for Keynesianism. Isn't the behavior of private investment the centerpiece of the whole Keynesian story? This seems like a clearer case than 1920-21. With 1920-21 I have to say "ya, this doesn't look like a downturn that Keynes spent much time talking about so it's not use testing a theory with a circumstance that it doesn't apply to - it doesn't contradict Keynesianism per se because I don't think Keynesians would make the claims about government spending that you expect they'd make". In short, 1920-21 does not look like a Keynesian bust. But to me, the late 1940s look very Keynesian to me (which is, of course, very different from saying that Keynesianism is the only thing that has a chance of explaining it).