Standard analysis is that military spending for WW II provided the significant Keynesian stimulus that ended the depression. I have nothing new to add to that debate, but I recently looked at the data and found it somewhat surprising.

The surprise was not in what happened to military spending as the military swung from 0.5 million personnel in 1940 to 11.4 million in 1944 and back to 1.6 million in 1947. The total population grew from 132.1 million in 1940 to 144.1 million in 1947.

Rather the surprise is what happened to civilian real GDP ( real GDP less real military spending). Real civilian GDP fell -18.4% in 1942 and -14.8% in 1943. The peak to trough drop of -26.9% was almost as severe as the -30.4% drop from 1929 to 1933. Of course this is tempered somewhat by a -0.5% and -2.4% drop in the civilian population in 1942 and 1943, respectively. No wonder they imposed rationing of many goods, they just were not available. But civilian real GDP rebounded 0.4%, 17.4% and 58.5% in 1944, 1945 and 1946, respectively.
Total real GDP fell -1.1% in 1945 and -11.0% in 1946. Obviously, this recession like the major recession of 1922 was driven by the drop in military spending. This rebound took the civilian economy back to the 5.6% trend growth line that prevailed from 1933 to 1950. I’m not going to make any observations about what this means for economic analysis– that can be done in the comments — but the data is very interesting. The ability of the civilian economy to rebound so strongly clearly reflected the high savings — partially forced — during the war.
Is that just the opposite of what we have now?