NEW YORK, May 5 (WNM/New York Fed) - A staff report by the Federal Reserve Bank of New York has found a correlation between deaths from the influenza pandemic in 1918-1920 and extremist voting in Germany in 1932 and 1933 (https://www.newyorkfed.org/medialibrary/media/research/staff_reports/sr921.pdf).

The researchers merged several historical data sets from Germany to show that influenza mortality in 1918-1920 is correlated with societal changes, as measured by municipal spending and city-level extremist voting, in the subsequent decade.

First, influenza deaths are associated with lower per capita spending, especially on services consumed by the young. Second, influenza deaths are correlated with the share of votes received by extremist parties in 1932 and 1933.

The researchers conjecture that the findings may be the consequence of long-term societal changes brought about by a pandemic:

"We show that the deaths brought about by the influenza pandemic of 1918-1920 profoundly shaped German society going forward. Regional variation in influenza mortality was related to subsequent city spending on various amenities for its population. Cities that saw a greater share of their population die due to influenza spend less, per-capita, going forward. We tie in to literature on the effects of austerity by showing that lower expenditures by local municipalities affect voting for more extremist parties.

However, we also show that influenza deaths themselves had a strong effect on the share of votes won by extremists, specifically the extremist national socialist party. This effect dominates many other effects and is persistent even when we control for the influences of local unemployment, city spending, population changes brought about by the war, and local demographics or when we instrument for influenza mortality. The same patterns were not observable for the votes won by other extremist parties, such as the communists. Our results are striking in part because they are robust to a large battery of alternate specifications despite being based on a relatively small sample.

We argue that our results are likely the consequence of changes in societal preferences following a pandemic. In particular, the fact that the pandemic affected predominantly one part of the demographic spectrum, in this case younger people, may have altered preferences in communities. It may also have spurred resentment of foreigners among the survivors (as has happened in past pandemics), driving voters towards parties whose platform matched such sentiments. Given a number of econometric challenges, we are cautious about the interpretation of our results. Nevertheless, the study offers a novel contribution to the discussion surrounding the long-term effects of pandemics."