Viewpoint: Lessons Learned From the 1918 Influenza Pandemic

LONDON — Lockdowns have brought most of the world’s advanced economies to a near standstill as policymakers fight the coronavirus pandemic by banning all but essential movement outside the home.

Tens of millions of employees across retail, entertainment, food services, personal services, professional services, education and other heavily impacted sectors have already been put on short time, furloughed or let go.

Small businesses and the self-employed are among the worst affected groups, but the loss of output, income and jobs is spreading all along the supply chain to businesses of all sizes.

Lockdowns have successfully disrupted the transmission of the coronavirus and show some signs of reducing the number of reported new cases and deaths.

But the high economic and social cost is fueling demands for governments to articulate a clear exit strategy that would allow the economy to restart.

Policymakers and their science advisers, as well as businesses and voters, must weigh the relationship between the health benefits and economic costs of lockdowns.

“We cannot let the cure be much worse than the problem itself,” U.S. President Donald Trump wrote on Twitter on March 23.

But some economists have pushed back, insisting there is no trade off between fighting the epidemic and minimizing the economic fallout.

In their view, lockdowns actually reduce both deaths and the economic damage compared with an uncontrolled epidemic by reducing the duration and severity of the crisis.

For proponents, lockdowns are a policy intervention that has zero opportunity cost and actually improves outcomes in all dimensions compared with an uncontrolled epidemic.

The claim lockdowns have both health and economic benefits has been most forcefully made in a paper published by three researchers from the Federal Reserve System and the Massachusetts Institute of Technology.

“Pandemics Depress the Economy, Public Health Interventions Do Not: Evidence from the 1918 Flu”, was published by Sergio Correia, Stephan Luck and Emil Verner on March 26.

“We find that cities that intervened earlier and more aggressively do not perform worse and, if anything, grow faster after the pandemic is over,” they conclude.

“Our findings thus indicate that non-pharmaceutical interventions not only lower mortality; they also mitigate the adverse economic consequences of a pandemic.”

Even without a lockdown, severe epidemics cause lots of economic disruption through deaths, absences from work, and as individuals and households adopt their own voluntary social-distancing measures.

If non-pharmaceutical interventions (NPIs) can reduce the duration and severity of the epidemic, it may be possible to reduce the economic costs as well as cutting the number of deaths.

Correia, Luck and Verner suggest NPIs can indeed improve health and economic outcomes based on the experience of the influenza pandemic in the United States in 1918/19.

“Our analysis yields two main insights. First, we find that areas that were more severely affected by the 1918 Flu Pandemic see a sharp and persistent decline in real economic activity.

“Second, we find that early and extensive NPIs have no adverse effect on local economic outcomes. On the contrary, cities that intervened earlier and more aggressively experience a relative increase in real economic activity after the pandemic.

“Altogether, our findings suggest that pandemics can have substantial economic costs, and NPIs can have economic merits, beyond lowering mortality.”

For their evidence about NPIs, Correia, Luck and Verner rely on an earlier data set of restrictions introduced by 43 cities during the influenza pandemic.

“Nonpharmaceutical Interventions Implemented by US Cities During the 1918-1919 Influenza Pandemic”, was published by Howard Markel, Harvey Lipman and Alexander Navarro, and others, in 2007.

The NPIs cataloged in the earlier paper were mostly school closures; public gathering bans; and legally mandated isolation and quarantine.

But it also considered other measures, including altering work schedules, limited closure or regulations of businesses, transportation restrictions, public risk communications and face mask ordinances.

Correia, Luck and Verner correlated the speed with which these measures were enforced and the length they were maintained with four health and economic outcomes.

They looked at death rates; manufacturing output and employment; bank assets and loan charge offs; and motor vehicle ownership, at city or state-level.

Death rates were taken from the annual “Mortality Statistics of the United States”; manufacturing and vehicle ownership from the annual “Statistical Abstract of the United States” for 1909, 1914, 1919, 1921, and 1923; and bank assets and charge offs from the “Annual Reports of the Comptroller of the Currency”.

One key limitation is the low frequency of many of these data sources, nearly all of which present annual statistics rather than monthly statistics, which makes it harder to identify short-term impacts.

Unfortunately for economic historians, the influenza pandemic occurred right at the start, or just before, the big expansion in federal statistics collection between the First World War and the Second.

The newly created Federal Reserve System, for example, only started publishing monthly estimates for industrial production from 1919 (“Celebrating 100 Years of the Industrial Production Index”, Federal Reserve, January 2019).

There are several reasons to be cautious about relying on the correlations in the Correia, Luck and Verner paper about influenza in 1918 to draw conclusions about the economic impact of coronavirus lockdowns in 2020.

First, the influenza pandemic coincided with the end of the war and the transition from a wartime to a peacetime economy, a point the authors acknowledge, but only briefly and without much further analysis.

Disentangling the economic effects of the pandemic from post-war demobilization has proved notoriously hard for economic historians.

Second, restrictions introduced in 1918, mostly school closures, mass gathering bans, and isolation/quarantine, were much less severe than the travel bans introduced as part of the lockdowns in 2020.

For the most part, the NPIs employed in 1918 were far less disruptive for business activity, and the most severe of those restrictions were often enforced for relatively short periods.

Third, the lack of high-frequency monthly data makes it much more difficult to assess the short-term economic impact of the NPIs in 1918.

In theory, the longer-term impact of NPIs can be tracked in the annual data, but comparisons across the war years (1919 with 1914) or over a period of rapid economic change (1923 with 1919) are fraught with difficulty.

The authors make some effort to control for cross-state differences in economic structure (for example the share of agriculture and manufacturing in the state economy) but this was a period of very rapid change.

Correia, Luck and Verner have made the best of the very limited data that is available on economic dynamics in 1918/1919, and it is an impressive study. But their conclusions rest on very fragile data foundations.

I am not convinced the limited available data can support their relatively bold conclusions and policy recommendations about the economic impact of epidemic controls.

In any event, the local non-pharmaceutical interventions employed in 1918 were very different and more modest than the whole-economy lockdowns being employed in 2020.

Policymakers should be wary before concluding there is no trade off between lockdowns to save lives and the impact on the economy.