Many, Many Ounces of Prevention

Friday, October 9, 2009

Every century or so, a major flu pandemic (an epidemic with a global impact) strikes. The last one, the great pandemic of 1918–20, infected many hundreds of millions of people and killed 50 million to 100 million men and women around the world. The Asian flu of 1957 is estimated to have killed 2 million people, and the pandemic of 1968 killed more than 1 million. Various false alarms also have occurred, such as the swine flu outbreak in 1976 in the United States, where more than 40 million people received flu vaccinations, thirty people died from the vaccinations, and few died from the flu itself. This year’s swine flu cases raised fears of the next “big one” before the flu settled down to an apparently manageable, though widespread, outbreak. What would be the economic cost of a global pandemic today?

As of midsummer, about 95,000 people worldwide had confirmed cases of swine flu, with 425 deaths reported. Even as the flu season shifted to the Southern Hemisphere and the illness seemed to be no more dangerous than the common flu, health officials advised caution because flu pandemics, including the great pandemic, often go through phases in which the first spread is rather moderate and the next waves are much more devastating. Whatever the course of this particular outbreak, health officials are confident that before long a major pandemic will strike that could wreak devastation throughout the world.

Note that flu pandemics involve a huge externality because infected individuals have limited incentives to take into consideration the likelihood of infecting others when deciding how much contact to have with other people. This externality justifies a significant public health involvement in trying to control the spread of flu during a pandemic.

Consider the cost of a modern flu pandemic as devastating as the great pandemic. Fifty million deaths in 1918–20 constituted about 2.8 percent of the world population at that time. Because population has grown twofold since then, a flu pandemic today that killed 2.8 percent of all people would take a staggering 150 million lives. That number can be converted into an equally staggering monetary value through findings on what people are willing to pay to avoid fatal health and other risks—what economists call the statistical value of life. It is estimated that this statistical value of life for a typical young adult in the United States is about $5 million.

Flu pandemics involve a huge “externality” because infected individuals have limited incentives to consider the likelihood of infecting others.

To calculate the aggregate cost of another such great pandemic, we assume that the comparable statistical values of life in other countries equal $5 million times the ratio of the per capita incomes to the U.S. per capita income. For example, the statistical value of life for a typical young person in a country with half the per capita income of the United States would be $2.5 million. Then if we assume that the same percent of the population would die from such a pandemic in all countries, the total cost of a pandemic equal in severity to the great pandemic would be more than $100 trillion. This huge amount dwarfs the effects of such a pandemic on world GDP, the economic impact that is usually calculated.

A Lancet study published in December 2006 estimated that a modern pandemic of equal virulence to the Spanish flu that caused the great pandemic would kill not 150 million people but about 60 million. It also claims that these deaths would be very much concentrated in poorer countries. Using the authors’ calculations to adjust my estimate of what people of the world would be willing to pay to avoid such a pandemic would reduce the estimated total cost to about $20 trillion.

Important developments in the world health care system might reduce even further the number of deaths from a virulent flu outbreak today. It’s true that the explosion in world population since 1919, the growth of cities at the expense of the countryside, and the development of air travel have led to much greater movement across national boundaries, which implies that the spread of flu has become a lot easier. But it’s also a lot easier to contain the spread and severity of a flu pandemic. Public health officials can isolate and identify the genetic composition of flu strains much more quickly than during the great pandemic. Officials of different countries are also in much greater contact with each other and can collaborate to partly quarantine victims of future pandemics.

Perhaps the most important lifesaving developments in recent decades are vaccines and antiviral drugs such as Tamiflu. Vaccines might be produced quickly enough to inoculate huge numbers against new flu strains, even highly virulent ones. When taken early enough, the antivirals can greatly moderate the course of an illness and speed recovery. The United States and the European Union apparently have large enough stocks of antivirals to treat about 16 percent of their populations—the U.S. supply covers about 50 million people; Japan has even larger drug supplies relative to its population. The poorer countries of Africa and elsewhere are the least prepared to fight a major pandemic.

Of course, new flu strains may emerge that cannot be treated by the known antivirals (the first cases of Tamiflu-resistant swine flu appeared in the middle of last summer). And bioterrorists may be able to produce and spread highly deadly viruses of all kinds. At the same time, drug companies are better prepared than even a few years ago to ramp up production of old drugs and to develop additional drugs to fight new flu strains and other viruses not treatable by present drugs.

Most people are willing to pay a lot to gain protection against deadly flu viruses.

I have indicated elsewhere that the vast majority of people are willing to pay a lot to gain protection against deadly flu viruses, which is why it would be desirable to greatly increase the stockpile of drugs and vaccines even if the probability of another pandemic were low and its nature unknown. For example, the expected worldwide cost in terms of willingness to pay to avoid the risk of another great pandemic that had a 1-in- 100 probability of occurring during the next twenty years would be approximately one one-hundredth times $20 trillion, or about $200 billion. This cost would justify sizable increases in world spending on antiviral drug and flu vaccines.