Getting Serious about the Flu

Friday, October 9, 2009

Here we are, still struggling to find a way out of the worst financial crisis since the 1930s, when along comes the possibility of a global influenza epidemic. Although certainly the first concern about the new strain of H1N1 virus involves health, we also have to worry that a full-blown flu pandemic could intensify the world’s economic problems.

Our ongoing study of economic disasters in 36 countries since 1870 suggests that this concern is well founded. In our sample, we isolated 158 depressions: defined as declines in a country’s real per capita gross domestic product (GDP) by at least 10 percent. The most prominent features of those depressions were wars and financial crises. But the fourthworst global macroeconomic event since 1870 seems to have been the great influenza epidemic of 1918–20, accounting for 13 of the depression events. In contrast, World War II is associated with 25, World War I with 23, and the Great Depression of the early 1930s with 21.

The great influenza epidemic, popularly known as the Spanish flu, began in the spring of 1918 and lasted until 1920, going through three or four waves during that time. The spread of the disease was propelled by international travel, much of which involved World War I troop move ments in 1918. Estimates of worldwide flu deaths cover a wide spectrum but typically range around 50 million.

We have thus far compiled estimates of excess deaths from the flu in 1918–20 for 32 of our 36 countries. The median excess mortality rate was 0.7 per 1,000 people, ranging from 0.1 for Argentina to 4.4 for India and South Africa. (The mean rate was 1.1 per 1,000.) Spain, forever associated with the flu, had a mortality rate of 1.2 per thousand, well above the median. The United States, at 0.65, was close to the median. Those 675,000 American deaths, when applied to today’s U.S. population, would translate into 2 million fatalities.

The great influenza epidemic of 1918–20 appears to have been the fourth worst global macroeconomic event since 1870.

As many commentators have observed, the 1918–20 flu was unusual in that it struck down a large number of working-age people, especially those aged twenty to forty. This suggests that the epidemic had even more serious economic implications than the raw numbers imply.

The troughs in macroeconomic activity that we associate with the great influenza epidemic were typically in 1920 or 1921. Not all of our 36 countries showed economic declines in this period. But the average fall in real per capita GDP from the previous peak in 1918 (or sometimes 1919 or 1920) was 6.6 percent. (For the 24 countries with data, the average decrease in real consumer spending per person was similar to that for real per capita GDP.) Notable declines in GDP among the 13 depression cases appeared in Canada and South Africa (24 percent) and Italy (22 percent). For the United States in 1918–21, the falls in per capita GDP (12 percent) and consumer spending (14 percent) meant that this contraction was second in size (since 1870) only to the Great Depression.

Ironically, given the name Spanish flu, Spain performed better than average. Per capita GDP rose during the epidemic, though real consumer spending per person fell in 1920–21.

The years between 1918 and 1920 or 1921 also featured sharp declines in stock prices in many countries. We have data on real rates of return on stocks in these years for 18 of the 36 countries in our sample. Among those 18, 11 endured stock market crashes, defined as a cumulative real rate of return of minus 25 percent or worse. The weakest markets for this period include Italy at minus 69 percent; Denmark, minus 57 percent; Switzerland, minus 54 percent; Japan, minus 52 percent; and France and Spain, both minus 46 percent. Germany saw a rate of return of minus 78 percent, but this reflected in part the reparations payments imposed by the Versailles Treaty at the end of World War I. The U.S. stock market performed better than average, but the real return in 1920 was still a dismal minus 22 percent.

The potential for a flu pandemic today surely adds to the odds of a depression, particularly because, as in the 1918–20 epidemic, the current strain seems to have a disproportionate impact on people of prime working age.

It seems highly unlikely that the 2009 influenza epidemic will be anywhere near as bad as the 1918–20 pandemic in either illnesses or mortality. For more recent comparisons, one can begin with the flu pandemics of 1957–58 and 1968–70, which were serious in terms of deaths but not comparable to 1918–20. In those more recent cases, stock prices fell temporarily in some countries until the limited scope of the pandemics became clear. In 1976, a swine flu scare prompted a fall in stock prices in some countries and a massive U.S. immunization campaign. The disease was soon found to be mild, however, and the overall effects on financial markets were minor.

The most recent disease scare prior to H1N1 was the outbreak of severe acute respiratory syndrome (SARS) in late 2002. That pandemic caused a widespread fall in stock prices until the limited nature of the threat became evident in the spring of 2003. The economic impacts of the four post–World War II flu events are difficult to pin down but are surely much smaller in magnitude that those from the 1918–20 pandemic.

For most countries—Mexico is a likely exception—the swine flu epidemic of 2009 may turn out not to have greater macroeconomic consequences than the four other post–World War II flu crises. Because of the global financial crisis that began last year, however, we already face substantial depression risks, arguably 20 to 30 percent in the United States. The potential for a flu pandemic surely adds to the odds of a depression, particularly because, as in the 1918–20 epidemic, the current strain appears to have a disproportionate impact on people of prime working age.

As Max Weber, a founding father of sociology, wrote in Science as a Vocation: “Our age is characterized by rationalization and intellectualization, and above all, by the disenchantment of the world.” Thus, the modern thinker relies on “technology and calculation”—even going so far as to use historical data to calculate probabilities of flu pandemics, financial crises, and depressions. Weber clearly would have supported the kind of quantitative analysis we apply to pandemics. Sadly, Weber’s own work was cut short by the great influenza epidemic. He died in Munich in 1920, aged fifty-six.