The American economy in WW2 is a very weird case of a country which apparently had massive economic growth during wartime. Usually, economies collapse with sharp drops of GDP (France, Germany, Austria-Hungary in WW1, USSR and Japan in WW2) or maintain themselves through massive imports of labor (Germany in WW2).
Also, it appears to be a general case that in WW1 GDP collapses were much higher than in WW2. Why is that? In all cases war was financed thourgh massive borrowing which was partly paid by central banks which means central banks were "printing money" to pay for war expenditures which lead to a massive increase in money supply and hence in nominal GDP's.
However, in the US in WW2, prices didn't increase much while nominal GDP boomed. That's because in WW2 governments took direct control of economies to a much greater degree than in WW1, instead of allowing prices to rise to allocate scarce resources, governments began to allocate resources directly (which also means that in WW2 the war effort was perhaps less efficient than in WW1) while freezing prices: people's nominal incomes increased greatly and their "real" incomes also increased because prices were frozen, however, because of rationing, they couldn't actually SPEND their incomes ("real" wages in Germany increased in WW2, for instance, while decreasing 32% in WW1, but in WW1 these 67% remaining could be spent on whatever the workers wanted).
In other words, the main reason why WW1 national accounts show massive decreases in GDP while WW2 national accounts show increases in GDP (massive in the US's case) was because in WW1 markets were less distorted and so GDP better reflected levels of market activity. Another important reason is that in WW1, conscription was much more severe than in WW2, as result the civilian labor forces decreased abruptly while in WW2, countries like US and UK didn't allow for the civilian labor force to decrease significantly. Still this discrepancy is not easy to explain only in reference to real output, I believe the main cause is plain statistical manipulation caused by price freezing.
Well, in WW1, let's look at French GDP and price indexes:
-------- Prices ------- GDP
1913 - 100 ---------- 100
1914 - 105.6 ------- 84.0
1915 - 123.9 ------- 71.6
1916 - 145.4 ------- 80.6
1917 - 176.3 ------- 78.9
1918 - 232.5 ------- 66.2
1919 - 288.5 ------- 71.5
1920 - 405.8 ------- 77.3
Prices more than doubled and real GDP (nominal GDP divided by prices) decreased. After the war is over, however, the economy begins to recover and GDP grows at 8% a year during the first couple of years after the war, a natural consequence of the return to the labor force of the conscripted soldiers. Although inflation continued at increasing rates.
Now, let's look at the US's GDP and price indexes, US's GDP estimate is made by Kendrick (https://en.wikipedia.org/wiki/John_Whitefield_Kendrick):
-------- CPI ---------- GDP
1939 - 99.30 ------- 100.0
1940 - 100.30 ------ 109.7
1941 - 105.31 ------ 128.7
1942 - 116.53 ------ 145.5
1943 - 123.68 ------ 160.6
1944 - 128.68 ------ 172.4
1945 ----------------- 171.3
1946 ----------------- 156.7
1947 ----------------- 153.4
So consumer prices increased 29% but GDP magically jumps 72%, growing at over 12% a year from 1940 to 1944. After the war is over, GDP magically DROPS, even though now the economy is released from having a large fraction of it's labor force being "held hostage" under conscription. Notice that real GDP is nominal GDP divided by the price index: in both France's WW1 case and US's WW2 case
Let's see how civilian (employed) labor force behaved in the process, in millions:
-------------- France ---- US
1st year ---- 17.53 ----- 46.1 (1940)
2nd year --- 14.96 ----- 48.6 (1941)
3rd year --- 13.15 ----- 52.8 (1942)
4th year --- 14.68 ----- 53.9 (1943)
5th year --- 15.64 ----- 53.6 (1944)
6th year --- 15.76 ----- 53.7 (1945)
So while France's labor force declined greatly following mobilization over the course of the war France mobilized 8 million men out of 33.2 million strong population outside of German occupation (the Germans occupied 6.5 million French people in WW1), most of that is shown in the decrease of 4.4 million of the civilian labor force from 1913 to 1915. And then they substituted the conscripted men with women and workers unfit for military duty, who are inexperienced or less physically able and hence less productive: labor productivity decreased 11% from 1913 to 1917. In the US, however, labor productivity appears to have magically increased 35.2% between 1940 and 1944.
Data for specific industries in Germany and France show a decrease in productivity as well: in Germany industrial productivity per worker decreased 22.3% from 1913 to 1918 and GDP decreased by about 25%, similar to France's.
One might think that the US's situation was different given the US's prodigious output of munitions in WW2, however, if you look at France and Germany in WW1, their production of munitions was extremely impressive at 280 million shells for France and 340 million for Germany from 1914 to 1918 (that's more artillery ammunition than produced by US and Germany combined in WW2) and over 101,000 aircraft were produced by the two countries combined in WW1. Productivity in producing munitions certainly exploded in the two countries during the period, but munitions industries only employed 10% of French's civilian labor force in 1918.
So why this massive discrepancy? Well, Simon Kuznets tried to produce more realistic estimates of US GDP for WW2, his results:
1939 - 100
1940 - 109.0
1941 - 121.8
1942 - 126.5
1943 - 132.5
1944 - 135.8
1945 - 139.4
Using this index, after the US entered the war in December 1941 productivity per worker did not appear to increase much: in fact, from 1941 when the war began to 1943, productivity per civilian worker DECREASED 2%, which is more in line with the experiences in WW1 of Germany and France. Although it doesn't fully match it.
And also, due to the massive distortions caused by centralized planning in the US's economy in WW2, even these revised figures are problematic:
From: http://www.independent.org/newsroom/article.asp?id=138Robert Higgs wrote:Finally, one can make an even more unorthodox—which is not to say incorrect—argument for rejecting the conventional wisdom. One can simply argue that outside a more or less competitive equilibrium framework, the use of prices as weights in an aggregation of physical quantities loses its essential theoretical justification. All presumption that price equals marginal cost vanishes, and therefore no meaningful estimate of real national product is possible.
In fact, price was “never a factor” in the allocation of resources for war purposes. The authorities did not permit “the price-cost relationship . . . to determine either the level of output or the distribution of the final product to individual uses.” Clearly, all presumption of equalities between prevailing prices, consumers’ marginal rates of substitution, and producers’ marginal rates of technical substitution vanished. Absent those equalities, at least as approximations, national income accounting loses its moorings; it necessarily becomes more or less arbitrary.
Some economists appreciated the perils at the time. Noting that the government had displaced the price system, Wesley Mitchell observed that comparisons of the war and prewar economies, even comparisons between successive years, had become “highly dubious.” Index number problems lurked around every corner. Much output during the war, especially the weapons, consisted of goods that did not exist before the war. Even for physically comparable goods, price structures and output mixes changed radically. Production of many important consumer goods was outlawed. Surrounding everything were the “obvious uncertainties concerning [price] quotations in a land of price controls and evasions.” Kuznets declared that the “bases of valuation for the war and nonwar sectors of the economy are inherently noncomparable . . . . It is impossible to construct directly a price index of war products that would span both prewar and war years.” Kuznets’s own efforts to overcome these problems never escaped from arbitrariness, as he himself admitted.
This essentially explains in detail in what I was saying before: the magical growth in productivity (specially true in metal working industries as measured by Restuccia and Tooze (2013) where they try to explain why US labor productivity in metal working increased magically by 70% from 1939 to 1944) measured in WW2, but maybe isn't it a consequence of statistical distortion caused by price control/manipulation? Although it's true productivity increased a lot in munitions industries because supervisors workers were learning how to make goods they didn't make in peacetime: that only meant that productivity in those industries was "low" in pre-war times when munitions were produced in very small quantities.
These issues are also present in other countries that practiced price controls and centralized allocation (Germany for instance, although I would think that the US had achieved a higher level of central planning than Germany, because in Germany according to Tooze, munitions production was still driven by the orders made by the Wehrmacht to the private industries who produced them. Although there was central planning in the allocation of labor and in consumer goods, so distortions were there as well).
This implies that my estimates of military expenditures in WW2 are problematic because I converted then in pre-war prices using price indexes that mostly consists of prices frozen by government decree and hence do not reflect real scarcities. For example in WW2, German military outlays from 1940 to 1944 appear to be 361% of pre-war national income while in WW1, military outlays from 1914-18 were only 171% of pre-war national income. This discrepancy can be partly explained by the fact that Germany controlled a lot of occupied countries in WW2 and could tax them to finance gargantuan expenditures but they were also so high because prices were artificially frozen in WW2 so that inflation was not accurately measured which lead to inflation of estimated level of military expenditures.