Blog Post - Views on the Economy and the World

History Warns Us to Avoid a W-shaped Recession

| May 03, 2020

“Those who do not study history are condemned to repeat it.”  And the rest of us are condemned to repeat George Santayana.

Will the Coronavirus Recession of 2020 be V-shaped?  Or U-shaped?  If we fail to heed the lessons of history it is likely to be W-shaped, with incipient recovery followed by successive relapses into sickness and recession.

As has been widely noted, we would have been better prepared to cope with the Covid-19 pandemic in the first place if everyone had paid more attention to the history of epidemics. Be that as it may, the world is now deep into the pandemic and its economic consequences, the most severe such events since the interwar period, 1918-1939.  As decision-makers in every country contemplate their next steps, they would do well to ponder the precedents of that interwar period.

Don’t expect a V-shaped recovery

As recently as March, a pattern of V-shaped bounce-backs in individual economies seemed plausible.  The argument was that once the rates of infection and death peaked and declined, people will eagerly go back to work.  The economic activity might even get an extra boost, as consumers release pent-up demand, and firms rush to fill backorders and to re-stock inventories.  The macroeconomic effects of earthquakes and hurricanes and other natural disasters usually exhibit V-shapes of this sort.   When China was hit by SARS in 2003, for example, there was a clear negative effect on output in the second quarter of that year, estimated at 2 %, but the subsequent recovery was so fast that there was little observable impact on the year’s GDP.

Indeed, China itself reports that industrial production bounced back sharply in March from the February plunge.

But it is clear by now that a V-shape forecast would be too optimistic overall.

For one thing, even in a rosy scenario where the health effects of Covid-19 in individual countries faded quickly and allowed people to go back to work safely, the aggregate global outlook would not be V-shaped.  Different countries are in different time phases, with the epicenter having started China, then moving to other East Asian countries, then Europe, the US, and the developing world.   Even a country that is in a relatively healthy phase (either before or after the worst of the attack) is adversely impacted by the loss of trade with other countries that are in bad phases.

A U-shape, then?

More likely than a V-shape, the best we can hope for is a U-shaped recovery. Certain segments of the economy go back to work, with employees spaced out spatially as well as spaced out over time in shifts, if possible.  Meanwhile, other segments remain shut down for longer. One would think that some of the lowest candidates for early re-opening would be stadiums, theaters, bars, beauty salons, bowling alleys, and tattoo parlors.   In a poll of 3,500 chief executives,  the World Economic Forum reported April 24 that 60% of business leaders expect a “U” shape.

Under the U-shaped scenario, a full economic recovery would have to wait until we achieve a health recovery that is more solid than a mere peaking in the reported rates of infection and death.  Lasting economic recovery requires lasting health recovery.

We need massive, frequent, convenient, free testing for the virus, which is technically perfectly feasible (at least, in high-income countries). It has been delayed in the US, UK, and some other countries by poor government performance. Contact-tracing should complement testing, with which countries such as South Korea have had success.  Effective tests for anti-bodies would also make a big difference if they could reliably identify workers who are immune to contagion. The big game-changer would, of course, be the discovery of a proven vaccine; but we are told that in the best case this is likely to take a year or more.  Another important technological breakthrough, which could come sooner, would be the discovery of an effective treatment.

Spending a year with economic activity severely depressed would confirm the status of the current recession as by far the worst since the 1930s.  But there are worse possible scenarios than the U-shape.

A repetition of historical mistakes would give a W shape

I worry about a prolonged W-shaped recovery, arising out of mistakes by our political leaders.  Two likely policy-mistakes are well-illustrated by events of the inter-war period: 1918 and 1936, respectively.

Under mistake #1, leaders sound the all-clear signal on contagion too soon, resulting in a second wave of infections.  Many politicians in many countries have shown in their words and actions a grave bias towards optimism with respect to this coronavirus.  This over-optimism has already done tremendous harm by delaying action.  It can do a lot more harm if they try to re-open the economy prematurely.

The 1918-19 global flu pandemic is an informative precedent. The sickness first hit the United States in early 1918.  But there was a second wave in September 1918, far worse than the first, and then a third wave which lasted into 1919-20.  (Negative economic impacts were statistically associated with the fatality rate across countries and were estimated at 6% of lost growth in the typical country, in a recent paper by Barro, Ursua, & Weng.)

American cities instituted public health interventions, such as face-mask requirements and the cancellation of public gatherings and school classes, but varied widely in the length of time that they maintained them.  Research in 2007, such as a study from the National Academy of Sciences, found corresponding effects on excess mortality across cities.  The success in reducing the number of deaths “was often very limited because of interventions being introduced too late and lifted too early.”  San Francisco, my original hometown, takes first prize with the most effective interventions, which reduced mortality by an estimated 25 % or more.  (The next most effective interventions were in St. Louis, Milwaukee, and Kansas City.)  But controls were withdrawn too early in all cities.  If the interventions had been kept for longer — assuming that were feasible — the death rate would have been further reduced, by an estimated 95 % altogether in the case of San Francisco.

Instead, by November 1918, with mortality down sharply in the City by the Bay, the public health authorities relented.  The masks came off on November 21. One can imagine the residents going out into the streets to celebrate the Armistice ending World War I.  As a result, another flu wave arrived in December-January and the death toll cumulated to one of the worst in the country.

Thus, the lesson of the first mistake is not to declare victory and abandon public health measures too early.

Mistake #2 is abandoning economic stimulus too soon, resulting in a renewed economic downturn.  Here the best US precedent is 1936-37.  The Roosevelt Treasury, satisfied with the recovery from the Depression that had begun in 1929, prematurely curbed federal spending and raised taxes to restore a balanced budget (which it achieved in 1938).  At the same time, the Fed also tightened monetary policy in 1936.  It doubled banks’ reserve requirements (which Friedman and Schwartz identified as the key mistake in their famous Monetary History of the United States) and began sterilizinggold inflows.  Probably as a result of the premature withdrawal of stimulus, the US went back into a severe recession in 1937-38.

This could happen again. We have a more recent US precedent in the US, revealing something about the politics of how we could see a repeat of premature withdrawal of economic stimulus in 2021.

The response to the Great Recession of 2007-09 was a fiscal stimulus, relatively rapid by historical standards.  I believe this fiscal policy, particularly President Barack Obama’s Recovery Act of February 2009, along with the Fed’s response, explains why the economy’s freefall ended in the first quarter of that year and the recovery began in the second quarter.

The recovery was slow, however, with unemployment still at 9% in 2011, from which it came down only gradually.  A premature end to fiscal stimulus in 2011, the result of a new Congress, helps explain the unsatisfactory speed of the expansion.

Why might one fear a repeat in 2021?  The US entered 2020 with a $1 trillion budget deficit – which was an awfulfiscalpolicy at a cyclicalpeak, by the way.  Fortunately, that did not stop Republicans and Democrats from coming together in March and April to respond to the coronavirus with unprecedented fiscal measures, matched with unprecedented monetary measures from the Fed.  That economic relief and stimulus are helping. But the job is far from finished.

It is easy to see how yet another premature withdrawal of stimulus could come about in 2021, for the same political reasons as 2011:  namely, Congress suddenly discovers the evils of debt when the opposite party holds the presidency.  The scenario is that the White House changes hands in November, but the Senate does not.  It is easy for politicians to remember the virtues of the counter-cyclical policy if their party holds the White House during a recession, but hard to remember them under any other circumstances.

In short, if our political leaders are too quick to declare victory – hastily lifting lock-downs and social distancing or ending fiscal support – we could see a nascent economic recovery cut short by a second downturn.  That is the W scenario, where “W” stands for withdrawal of measures prematurely.  It is all too plausible.


For more information on this publication: Belfer Communications Office
For Academic Citation: Frankel, Jeffrey.History Warns Us to Avoid a W-shaped Recession.” Views on the Economy and the World, May 3, 2020,