Each time a recession nears, some call for governments to step in to “stimulate” the economy. Not through monetary means (printing money), but through fiscal ones. If the government initiates more projects, places more orders, and hires more workers, the reasoning goes, it stimulates consumption and thus facilitates the recovery.
The imagery that is frequently used to make that case is that of the wartime spending in the United States during World War II. When it entered the war in 1941, the American government spent large sums to equip soldiers to fight overseas. According to this political imagery, this meant that factories were running full time. That farmers had a strong and steady demand for foodstuffs. That there was no unemployment. Saved from the physical destruction observed in European countries like France, Germany, Poland and Britain, the American economy was stimulated out of the shadow of the Great Depression.
Essentially, the war was a boon to the economy in the United States.
This imagery, however, is incorrect. Those who use it are being fooled by the data’s flaws and limitations. In no way can wartime spending be used to justify fiscal stimulus.
The first is that the price indices needed to adjust income for inflation were plagued by the problems that wartime price controls created. Once adjusted price deflators were used, more than two thirds of the wartime gains commonly reported in the data were eliminated.
The second is that many assumptions needed to estimate economic output in the form of Gross Domestic Product (GDP) vanish or are weakened in wartime. One must account for, for example, the depreciation of capital goods, which means selecting a depreciation rate. Qualitative and quantitative evidence at the firm level suggest that businesses used capital more intensively during the war, and thus that capital depreciated faster — something that is not taken into account. Corrections for the rising depreciation rate during the war only lower the estimated growth rates.
The third is the most important. In wartime, the government’s mandates drafts, seizures, taxes) make the prices used to weigh the quantities produced somewhat meaningless. They do not reflect prices that clear markets for consumers and producers, they reflect prices that bureaucrats and some producers agree to. As a result, these economic historians suggest removing government spending on the military from GDP in order to better estimate the wellbeing of American consumers and workers. Once this is done and added to the previous improvements, there was no wartime boon.
In fact, if these adjustments are extended to 1949 (after the war’s end), one finds that the recovery started once the war ended. In contrast, the unadjusted data suggest that the war’s end brought about a depression, something that is telling, as no economist today is willing to talk about the “Great Depression of 1946”.
One could reply that the American data are hopelessly flawed and that these rebuttals are in no way conclusive. Moreover, it could also be argued that it failed to account for the fact that the war started two years before the American government joined it. During that interceding period, foreign demand for military equipment could have stimulated the economy.
These rebuttals fall flat, as I argue in an article co-authored with Casey Pender that is forthcoming in Social Science Quarterly and which uses data from Canada instead of the United States. This confers two strong advantages. First, Canadian GDP numbers can be argued to be comparatively better than American data in large part due to the early census of manufacturing that was made in Canada, where population censuses asked questions about income far earlier than the United States census (and these questions can be used to assess the plausibility of GDP numbers). Second, Canada joined the war in 1939, rather than 1941. Its involvement was also far more intensive as a small, open economy that was strongly tied to Britain.
We made the same adjustments as Higgs, Vedder, Gallaway, and Field did to Canadian GDP numbers to see if the Canadian experience mirrors that of the United States starting two years earlier. We find that, yes, it did. Adjusting the GDP deflators and removing wartime expenditures suggest that the living standards of Canadians fell during the war. By 1943, Canadians were 10 percent poorer than at the start of the war. From 1945 to 1947, however, they saw their income per capita increase by 50 percent as the economy expanded rapidly.
Proponents of fiscal stimulus may think that large government outlays can pull an economy out of a recession. This is a debatable theoretical proposition. If they want to use wartime spending as an empirical illustration that makes their case, however, they need to be aware that doing so relies on a bad understanding of economic facts.