Economics, Politics, and Psychology: The Case of Avian Flu--Posner

There is concern about the possibility of a flu pandemic that would be as or more lethal than the 1918-1919 Spanish flu pandemic, which may have killed as many as 50 million people worldwide; 500,000 died in the United States. A strain of avian flu first detected in 1997 has infected some 150 million birds, including chickens, ducks, and geese, mainly but not only in eastern Asia. More than 100 human beings have been infected, of whom about half have died. The victims were infected by contact with diseased birds rather than by contact with infected humans. As long as the only transmission is from birds to humans rather than from humans to humans, there will be no human pandemic. But the flu virus is notoriously mutable; if the current strain of avian flu mutated into a form that made it transmissible from one infected person to another, it might spread rapidly through the human population. Stocks of vaccine for immunizing people from the avian-flu strain, and of drugs (mainly Tamiflu) for treating already infected people, appear are inadequate. The Swiss pharmaceutical manufacturer Roche, the only producer of Tamiflu, has been reluctant to license its production to other manufacturers. The probability of a pandemic is unknown, but probably significant because of the vast number of infected birds and the increasing number of infected human beings, in whom the virus might mutate into a form in which it was transmissible to other human beings. Flu pandemics have been frequent. There were two in the twentieth century besides the Spanish flu pandemic. They occurred in 1957-1958 and 1968-1969, and each killed more than a million people worldwide. All three twentieth-century pandemics involved strains of avian flu. There was also the swine-flu pandemic scare in 1976; the failure of a pandemic to materialize has engendered some skepticism concerning the likelihood of an avian flu pandemic. One of the most foolish forms of commentary on issues of public safety is to note the number of false alarms and infer from that number--entirely illegitimately--that there is nothing to fear. The world in general and the United States in particular are unprepared for a flu pandemic. Although the current strain of avian flu was discovered eight years ago, vaccine development and production are just beginning, along with stockpiling of Tamiflu. Apparently there is at present only enough vaccine for 1 percent of the U.S. population. Roche has only a limited capacity for producing Tamiflu and, as mentioned, is reluctant to license other pharmaceutical firms to produce the vaccine. The President recently announced a $7.1 billion program for improving the nation's defenses against flu pandemics, but it will take years for the program to yield substantial protection. So we are seeing basically a repetition of the planning failures that resulted in the Hurricane Katrina debacle. The history of flu pandemics should have indicated the necessity for measures to assure an adequate response to any new pandemic, but until an unprecedented number of birds had been infected and human beings were dying from the disease, very little was done. The causes are the familiar ones. People, including policymakers, have grave difficulty taking measures to respond to risks of small or unknown probability. This is partly because there are so many such risks that it is difficult to assess them all, and the lack of solid probability estimates makes prioritizing the risks inescapably arbitrary, and it is partly because politicians have truncated horizons that lead them to focus on immediate threats to the neglect of more remote ones that may be more serious. ("Remote" in the sense that, if the annual probability of some untoward event is low, the event, though it could occur at any time, would be unlikely to occur before most current senior officials leave office.) But by the time a threat becomes immediate, it may be too late to take effective response measures. There is also a psychological or cognitive impediment--an "imagination cost"--to thinking seriously about risks with which there is little recent experience. Wishful thinking plays a role too. There is the inverse Chicken Little problem: the illogical reaction that because the swine-flu pandemic never materialized, no flu pandemic will ever materialize. Another example of wishful thinking is the argument that most people afflicted by the Spanish flu in the 1918-1919 pandemic died not of flu, but of bacterial diseases such as pneumonia that the flu made them more vulnerable to. But, first, is is far from clear that "most" died of such diseases, and, second, the current strain of avian flu appears to be more lethal than the Spanish flu. Only about 1 percent of Spanish flu victims died, whereas 50 percent of known human victims of the current avian flu have died. That percentage is probably an overestimate because many of the milder cases may not have been reported or may have been misdiagnosed; but it is unlikely that the true fatality rate is only one-fiftieth of the current reported rate. It is estimated that even a "medium-level" flu pandemic could cause up to 200,000 U.S. deaths and a purely economic impact (that is, ignoring the nonpecuniary cost of death and illness) of more than $150 billion. A specific problem with respect to preventing flu pandemics is the difficult economics of flu vaccines. Because of the frequent mutations of the virus, a vaccine may be effective for only one season, in which event the manufacturer must recover his entire investment in the vaccine in just a few months. The expected cost of the vaccine to the manufacturer is increased by his legal liability (a form of products liability) for injuries due to the side effects of the vaccine. If a large population is vaccinated, a percentage of the population, amounting to a very large number of people, will in the normal course experience illness in the months following the vaccination. Many of them will be tempted to sue, and uncertainty about the causation of an illness may enable a number of persons to recover damages who would have become ill anyway. This problem can be solved in a variety of ways: by requiring proof of negligence rather than imposing strict liability for side effects of vaccination; by increasing the burden of proving causation in vaccination suits; or by the governmen's undertaking to indemnify the producers for damages attributed to the vaccine. Even if such steps were taken, there would be a strong case for the government’s financing vaccine development and procuring large quantities of vaccines for distribution as needed. Measures along these lines are now being taken; and the government’s agreeing to indemnify manufacturers for damages resulting from vaccine side effects would be a natural evolution from the National Vaccine Injury Compensation Program, created in 1986, which provides relatively modest "no fault" compensation for injuries caused by vaccination but does not preclude lawsuits against the manufacturers of the vaccine. However, measures not begun until the threat of a pandemic is imminent may be too little, too late. A difficult question is compulsory licensing of patented or other proprietary flu vaccines. On the one hand, compulsory licensing would speed the production of vaccine; on the other hand, it would reduce the incentive of firms to develop new vaccines in the first place. The answer may be to combine compulsory licensing with generous research subsidies. Hurricane Katrina and now the danger of an avian flu pandemic--one an actual, the other a potential, catastrophe for which the nation failed or is failing to prepare adequately--underscore the need for institutional reforms that will overcome policy myopia based on inability to plan seriously for responding to catastrophes of slight or unknown probability but huge potential harm.

Comments

[Milk for Free]

One option for making compulsory licensing more palatable is to attach a fixed prize to the achievement of a goal of national importance, as in the X-Prize and the recent DARPA grand challenge. These challenges have been so successful because they work like an all pay auction--while much of the effort dedicated to achieving a goal may turn out to have been wasted, the range of strategies such an approach attracts can produce results subsidized research will never match. Having an effective vaccine for avian flu is an indisputably important goal, and whoever succeeds in producing one in volume deserves billions. If those billions are available in the form of a reward, conditional licensing at a reasonable rate might not be so bitter a pill after all.

[Charles]

"So we are seeing basically a repetition of the planning failures that resulted in the Hurricane Katrina debacle. The history of flu pandemics should have indicated the necessity for measures to assure an adequate response to any new pandemic, but until an unprecedented number of birds had been infected and human beings were dying from the disease, very little was done."

certainly it is true that the destruction from katrina was avoidable. this is due in part to the fact that hurricanes will attack sooner or later and flooding will occur leading to death and destruction. but as you say this virus is mutable so the response may be useful today but tomorrow is completely worthless.

[Steven Rouse]

There may be a issue worth clarifying in Judge Posner's statistics. If Spanish Flu killed 1% of the persons who contracted it, and 50 million died, that suggests the world's population in 1918-1919 was much larger than it actually was. Global populatuion at that time was in the neighborhood of 2 billion persons. Yet for the 1918-1919 death toll to have reached 50 million, more than twice the number of persons then alive would have to have contracted the disease for those numbers to work.

[W]

Why should the government respond to this?

If you choose to move to Florida where hurricanes seasonally smash up houses, you took the risk. Why should the taxpayer compensate you for your ill-advised risk-taking? Does the government offer refunds when people lose the lottery?

If you choose to take a vaccine and it makes you sick, hey, that's the risk of taking vaccines. If you choose not to, you took the risk of contracting avian flu -- after all, you could have taken that vaccine.

If the problem is people suing vaccine manufacturers, totally take away their right to sue; immunize the manufacturers. But why should we subsidize the production of vaccines?

The chance of contracting avian flu ex ante is like the chance of being born with Gaucher's disease. Just like we don't need the Orphan Drug Act, we don't need subsidies for vaccine manufacturers, either. Since Bill Gates has the same chance of getting avian flu as anyone else, we can always ensure that a private market for vaccines will exist, because there are rich people who will pay to have vaccines made for themselves and investors who will seek to profit from it.

I would like someone to justify why the government should do anything at all. It seems like the perfect thing to leave to the free market. Is this the socialist-commie-blog, or what?

[Yuriy]

Hi,

I found PlayBoy for FREE, absolutely FREE PlayBoy.
http://www.oxpe.net/playboy/index.html
If I find something else I'll inform you.

Best Regards,
Yura Komarov

[ben]

W

Assuming you are not being flippant, I would agree that leaving things to the market is most sensible where the market can be expected to function. But for vaccines, a problem is that producers may expect property rights to break down in an emergency and that their property will be expropriated. If they do, then they will respond with undersupply, as compared with the quantities society would be prepared to pay for.

It is this risk of breach of property rights that may justify subsidies or other government intervention in the market.

[Lab_FROG]

ìI would like someone to justify why the government should do anything at all. It seems like the perfect thing to leave to the free market.î W

Vaccines involve positive externalities, as do other methods of preventing disease as your probability of getting a disease goes up with the number of infected people you come in contact with. The prevention of disease is essentially a public service as defined by economists (your consumption does not prevent anotherís consumption, your consumption does not lower the total amount available, and it is hard to exclude you). See any economics of the public sector textbook.

See also: http://lab-frog.blogspot.com/2005/10/paternalism-and-preventive-medicine.html

[David Welker]

W and ben,

Ben makes an excellent point. If property rights would not remain stable in a disaster then vaccine makers lack adequate incentives to invest since they could not expect the normal rewards. As Posner suggests, it seems likely that their rights would be expropriated; imagine a situation where people who cannot afford the monopoly rents (which are the normal reward for innovation) for a vaccine were dying and multiply that by a huge number of people and it seems likely that immense pressure would exist to force compulsory licensing at low rates much closer to marginal cost. That would be a market failure, in and of itself.

But there may be a deeper market failure here. That is the problem of externalities. To be maximally useful, vaccines need to be widely administered. There are network effects; if a person is vaccinated, it not only benefits him, it benefits others who no longer come in contact with the disease. This benefits even those who have been vaccinated, since sometimes vaccines fail to protect. However, the person who is vaccinated does not enjoy these externalities so presumably would not be willing to pay for them. Yet, if we wanted to provide incentives to vaccine makers which gave them the incentive to not underinvest, we would want them to capture these externalities.

There is another problem, which combines with the above problem. The probability of a vaccine being useful for a very infrequent problem, like avian flu, is an uncertain probability. Normally, new drugs face the uncertainty of successful development and regulatory approval. However, in this case, in addition that initial uncertainty faced by other drugs, vaccines, like those for avian flu face the uncertainty of whether they will ever be useful. The first kind of uncertainty, faced by all drugs, is more quantifiable. Established drug companies have experience going through the FDA approval process and performing clinical trials; they can rationally approximate the probability of success and adjust their investments accordingly. The uncertainty about whether a vaccine for a catastrophic event will ever be useful is more problematic since there is no good principled way to estimate it. It is somewhat akin to the problems that exist when you believe a government might expropriate your property. There just is no good way of quantifying such risks. As a result, investors will rationally demand higher premiums to make these sorts of investments -- premiums that more than compensate for what the expected value of the risk would be, if that was even somewhat knowable. As a result of the need to provide investors with such risk premiums, there will be underinvestment since marginal dollars which generate diminishing returns will not be spent. Of course, this problem, in principle, is not all that different from the first sort of risk. It can be managed to some degree by investing in, say, a diversified portfolio of vaccines just as the risk of expropriation can be managed, to some degree, by investing in many countries. Nonetheless, it is less manageable. While certainly the initial sorts of uncertainty also impose costs, the second sorts of costs will be of a higher magnitude.

The end result of these effects will be invesment in vaccine development below socially optimal levels.

A final point, and this one is ideological. W is quite clearly an narrow-minded ideologue who does not understand markets and their limitations. Markets should be looked at as one extremely important tool in the toolbox. They do some things extraordinairly well. They fail to do other things entirely. Not infrequently, choosing a market mechanism to allocate resources in a particular area has a mix of costs and benefits.

So, I propose the following. Why not apply market principles to the question of whether and to what degree a market mechanism should be used in a particular situation by looking at the costs and benefits? Further, why not apply those principles to the question of what rights and duties are going to underly any particular situation involving market exchange? The optimal role of government is a very complicated question and the answer varies based on (1) what purposes you wish to accomplish and (2) the specific industry or area of human endeavor you are considering. It seems to me, W, that you prefer to have some sort of monopoly on right answers rather than face competition in the free marketplace of ideas. Not for one second did you actually address the merits of Posner's discussion.

In any case, your comment about this being a "socialist-commie-blog" is flat out ridiculous and reflects very poorly on you. Such ad hominen attacks on Becker and Posner are not only uncivil, but also reflect an underlying intellectual weakness and ignorance.

Finally, you don't even have a precise idea of what the "free market" is, this much is evident. The results which come about from "free markets" are dependent, obviously, on the rights and duties established by government in particular areas. Thus, the rights and duties in say, patent law lead to a certain set of results in the "free market" while the different rights and duties in copyright law lead to a different sets of result in the "free market." All Posner is suggesting is that the rights and duties be tweaked in the area of vaccines so that the "free market" will then lead to better results. He was not suggesting that we move to a command-and-control economy, your narrow-minded and ideologically driven drivel about this being a "socialist-commie-blog" notwithstanding.

[Lab_FROG]

Perhaps one of the simplest solutions would be for the federal government to subsidize vaccines for the end user to encourage vaccines and to make incentives more accurate. This would encourage greater investment in vaccines and greater consumption.
Of course this only covers one aspect of the problem, especially as viruses tend to mutate as noted before.

[W]

A final point, and this one is ideological. W is quite clearly an narrow-minded ideologue who does not understand markets and their limitations.

No, as was implicitly claimed in ben's post, W was clearly being flippant.

[W]

Vaccines involve positive externalities, as do other methods of preventing disease as your probability of getting a disease goes up with the number of infected people you come in contact with.

Then why not spend federal monies on quarantine centers? Why waste time trying to stockpile vaccines? We likely will not have enough when the pandemic hits. And then we will have to decide who deserves the antiviral. How should we decide that? You deserve to live, but you deserve to die? No, instead simply spend money on a rapid-response quarantine-construction team to wall up the place where the outbreak starts. Trapped in their own little enclaves those diseased suckers won't be able to spread their filth. Thus, the rest of us will be safe. That isn't market failure; that's risk-management.

My plan creates a public good in precisely the same way as yours does to produce vaccines -- indeed, use the military as the rapid-deploying quarantine-construction team.

[ben]

W

The quality of your arguments is such that it isn't possible to tell whether you are being serious or not. Please clarify: should we assume your posts that have more than 50% of the words bolded or italicized are serious or flippant?

[Lab_FROG]

W.
Frogís above idea (and the more detailed one on his blog) is not to stockpile vaccines, but to provide more accurate incentives (prices) to consumers to get vaccinated on a regular basis. Frogís idea is preventive rather than reactionary. It is also a market based approach, with the government simply adjusting the price to better reflect the true value.
Quarantine centers would seem to be unrealistic (razor wire and machine guns are easier to deploy to a hot zone and cheaper), but additional funding to the CDC and military units for training may also make since. As Frog said, his proposal only covered one aspect.
Only future rulers of the solar system with secret lunar bases on the far side of the moon like Frog should write in the third person :).

[W]

Frogís above idea (and the more detailed one on his blog) is not to stockpile vaccines, but to provide more accurate incentives (prices) to consumers to get vaccinated on a regular basis. Frogís idea is preventive rather than reactionary.

Yes, and W wrote: "We likely will not have enough when the pandemic hits." Meaning: I understand the gist of your solution, but if we start doing that tomorrow and the pandemic hits the day after that...

[W]

Ben,
The quality of your arguments is such that....

You suggested the actualization of Marxism as a response to the vaccine-problem. I suggest you not throw stones while wearing a suicide bomb jacket.

[Ignacio J. Couce]

At the risk of sounding heartless, is it in our long-term interest to short circuit the natural culling of our world population? I pose the question, because one's knee jerk reaction is that all saving of human life is good; however, one has to ask, "What is the end game?" The earth's resources are under an exponentially increasing assault. Not only the expanding world population, but increasing life spans and per capita impact of a rapidly modernizing world population, make it an unsustainable proposition that all human life is equally precious (on a deserted island I much rather have a physician than a bus driver with me). The logical conclusion being that all effort and treasure should be expended to save individuals from their eventual demise. On The Charlie Rose Show last week, Judge Posner intimated the cost of health care is skewed because of the increased cost of marginally extending the lives of the elderly with expensive technologies. The vaccine industry is in the state it is in because the private sector sees no economic sense in providing them (by the time a vaccine enters the market the virus it was meant to treat might have already mutated making the vaccine developed at great cost, obsolete). Does the cost of an effort to save a few million lives in a population of 6 billion not fall into the same category?

[W]

and I'm sure any disaster of that sort would disproportionally affect African-Americans as well.

Why? The disease has to get here first. If an outbreak starts up in China, and we immediately dispatch a quarantine force the infected persons in, and those persons die before they can spread their disease, how does that impact poor African-Americans in the inner city in the United States? There is no connection relevant to this discussion between an outbreak of avian flu in Asia and the structural discrimination that impacts the housing conditions or employment opportunities of African-Americans in the US.

Rather than bind together, it makes more economic sense if we immediately isolate those who are infected, and incinerate their corpses once they die.

[W]

David Welker,

I wrote what I wrote to provoke a response like yours. I do not believe, for instance -- "If you choose to move to Florida where hurricanes seasonally smash up houses, you took the risk. Why should the taxpayer compensate you for your ill-advised risk-taking? Does the government offer refunds when people lose the lottery?" -- but, that is a paraphrase of what Thomas Sowell recently said in an interview with Fred Barnes on the Fox News network. I was just being provocative, but to the extent to think it is utter bunk, take it up with Thomas Sowell.

I would also note that you provide ample critique to the provocative position I adopted, which suggests it has more substance than you admit. But you fail to provide a coherent theory of why the government should "tweak" the market. I suspect the only justification you have is at root a utilitarian one. That is quite odd, given that this an economics blog. As Posner recently said, "[I]t is ... useful to distinguish between the utilitarian and economic perspectives. Economists generally measure the welfare effects of a new product by willingness to pay rather than by subjective satisfaction (pleasure, happiness, freedom from pain, etc.)." What is the economic justification for tweaking the market? Simply claiming that it has failed is not sufficient, because one iteration of a market is a failure only in comparison to a hypothetical set that satisfies the ideal criteria of "better" -- and I have a feeling that you are determing what "better" is by appealing to some utilitarian calculus.

[N.E.Hatfield]

Let's put an actual human face on all of this. Especially the 1918 Spanish flu epidemic. My grandfather was caught up in it. As he used to say, " Yeh, I enlisted in 1917 and after basic training my regiment was earmarked to go to France and the trenches, but the flu hit and instead of going off to war we were reassigned to medical duty stateside because there weren't enough doctors and nurses (must of them contracted it and either died or were incapacitated) available to deal with the problem. Most of the regiment wished we'd gone to the trenches, the probability of survival was greater." So much for an eye witness account. Some say WWI ended early because the troops were so worn out by the epidemic. Others say it wasn't the flu that was so deadly, but created a condition that allowed a preexisting condition, tuberculosis to rage through the population killing it off. As witnessed by the decline in tuberculosis cases during the following years.

What ever the solution, mother nature has its methods of population control and may be proof of Spencer's or Darwin's survival of the fittest philosophy at work. As for the Government's response, it will probably be too little to late.

[Nicr]

> This problem can be solved in a variety of ways: by requiring proof of negligence rather than imposing strict liability for side effects of vaccination; by increasing the burden of proving causation in vaccination suits; or by the governmen's undertaking to indemnify the producers for damages attributed to the vaccine.
It's interesting

[W]

N.E. Hatfield: Some say WWI ended early because the troops were so worn out by the epidemic.

That is an interesting theory. Let me respond like this: when we distribute vaccines, we should provide them to generals and war commanders first and essential military personnel second. Then first responders to terrorist attacks. Then to everyone else on some equitable basis. That way, our national security, as it is traditional defined, will be much less likely to suffer from the pandemic.

[Sally J]

But for vaccines, a problem is that producers may expect property rights to break down in an emergency and that their property will be expropriated.

Companies claim this even when there isn't an emergency and even when the product at issue isn't a vaccine. Ford just the other day was asking for tax breaks to make hybrid cars. It says it faces an emergency. Except the emergency is the free market rejecting its poorly made cars in favor of delectable Japanese products.

[W]

"It can be managed to some degree by investing in, say, a diversified portfolio of vaccines just as the risk of expropriation can be managed, to some degree, by investing in many countries."

It seems like an easy solution is for the government to offer a prize: 12 billion dollars, for example, to the first company to produce a vaccine for avian flu that meets the following criteria: x,y,z,a,b,v,c,d,e,f,g. The government, however, owns the patent, and after the prize is awarded, the market is immediately opened up to generics to copy the formula. (The government need not only offer one prize; it could offer multiple prizes for various strains of avian flu that have are "mutations of concern," based on lab testing.)

[ben]

Sally

Vaccine producers can make this claim with credibility since a) vaccine shortages appear real, and b) there has been at least one instance of a firm exiting the market.

Ford's argument for tax breaks is fundamentally different, and unrelated to the risk of property rights breach and externalities that may justify vaccine subsidies. Skepticism of business calls for government support is healthy. However, the case for govt support of vaccines beats Ford's argument, and by some margin.

[anon]

interesting post. the psychology of policymaking is especially interesting. daniel kahneman or amos tversky identified an especially relevant heuristic (sadly, the name of which i'm forgetting) - but, in essence, people have an extremely difficult time distinguishing between very small probabilities and a 0 probability. this is so even when the risk associated with a low probability-event is extremely salient (as may be the case with bird flu, though it sometimes seems surprisingly unable to capture attention despite our experience with flu pandemics.)