Did non-pharmaceutical interventions actually help against the Spanish flu?

From three economics Ph.D students at Harvard, namely Andrew Lilley, Matthew Lilley, and Gianluca Rinaldi:

Using data from 43 US cities, Correia, Luck, and Verner (2020) find that the 1918 Flu pandemic had strong negative effects on economic growth, but that Non Pharmaceutical Interventions (NPIs) mitigated these adverse economic effects. Their starting point is a striking positive correlation between 1914-1919 economic growth and the extent of NPIs adopted at the city level. We collect additional data which shows that those results are driven by population growth between 1910 to 1917, before the pandemic. We also extend their difference in differences analysis to earlier periods, and find that once we account for pre-existing differential trends, the estimated effect of NPIs on economic growth are a noisy zero; we can neither rule out substantial positive nor negative effects of NPIs on employment growth.

I am very willing to publish a response from the original authors on this one.


The Spanish flu was the primary example of medical systems being entangled with a government response that has insignificant impact on the wellbeing of butchers and accountants, at least those in big cities with sewer systems that are permeated by the strain of approximate suffrage.

Such tasty word salads are even better than whatever it is TC is willing to publish.

.. made my day

A flea and a fly in a flue
where imprisoned, so what could they do?

"Let us flee!" said the fly.
"Let us fly!", said the flea.

So they flew through a flaw in the flue.

Just because you don't understand it, that doesn't make it word salad.

'in big cities with sewer systems that are permeated by the strain of approximate suffrage'

Oh-well indeed.

'Apparent superior growth of NPI implementing cities following 1918 Spanish Flu is simply on artefact of pre-existing population growth trends'. Easy.

Only a word salad for those that only subsist on word candy.

"Noisy zero" shall be my new favorite ad hominem insult.

Healthy, young adults were most at risk from the Spanish flu, not seniors and those with existing health conditions as is the case with the coronavirus; thus, it's not surprising that the economics of the two pandemics would differ, as would the effect of non-pharmaceutical interventions.

Wikipedia entries for 1918-1919:

Boston Red Sox- Chicago Cubs The 1918 Series was played under several metaphorical dark clouds. The Series was held early in September because of the World War I "Work or Fight" order that forced the premature end of the regular season on September 1....

Total attendance for six game series: 128,483

1919 Chicago White Sox-Cincinnati Reds Total attendance for eight games of series: 236,936

1918 college football: In an early September meeting between college and War Department officials in Plattsburg, Missouri it became clear that the training regimen envisioned for the soldiers could be incompatible with participation in intercollegiate athletics. On September 13, 1918 newspapers around the country reported that the War Department had asked colleges to reexamine their football schedules.

The influenza outbreak was colloquially called Spanish flu. Most flu outbreaks disproportionately kill juvenile, elderly, or already weakened patients, but the 1918 pandemic predominantly killed previously healthy young adults. To maintain morale, wartime censors minimized early reports of illness and mortality in Germany, the United Kingdom, France, and the United States. One college bowl game was played, Pitt defeating Georgia Tech 32-0.

Sir Barton won the 1919 Kentucky Derby and went on to win the Triple Crown that year.

Texas State Fair: The fair was canceled in 1918 due to the US Army taking control of Fair Park to establish an aviation boot camp known as Camp Dick.

Would economists become better writers if they assembled sentences from words rather than chunks of prose? Thus, instead of the clumsy "had strong negative effects on" they might write that it 'reduced' economic growth, or that it 'slowed' it; if they wanted to hint vaguely at the effect being sizeable they might include "substantially" or even "bigly".

If they were to read their drafts with the idea of the mot juste in mind, rather than trying to mimic the constipated efforts of their elders and worsers, they might end up with more readable work.

Mind you, I suppose that writing readable work might not be their intention.

Econometrics is a very lousy method. Noboby outside the field take it seriously.

Its unclear if the 1918 lockdowns had any benefit whatsoever, given that their medical system wasn't particularly useful to flu victims and thus "flattening the curve" wouldn't have been too important. The disease most likely ended via herd immunity and thus lockdowns failed to prevent a single death from occurring.

Usually this is refuted by claiming that the 1918 virus mutated into a benign form by 1919, but this is impossible to prove.

I expect the better growth before and after reflects better governance. A study that needs to be done is how governance processes differed between cities. By governance I mean both the quality of leadership and quality of implementation of government (not its size).

At a brief look, it seems the original paper is fatally flawed. I sympathize. It's important to get papers out quickly, and that means more of them will be duds. I give credit to the original authors for trying, and the correction paper's authors for finding the mistake. If the original paper is indeed wrong, though, the authors have a duty to tell us so right away, so we don't have to spend time figuring it out.

Comments for this post are closed