Alex Tabarrok - Prizes, Prices, and Public Goods

Tabarrok opens with the premise behind the grand innovation prize: in the middle of the pandemic, speed matters more than anything, and the response has been behind the virus at every step. The aim is to speed up the incentives for producing a vaccine, a diagnostic or a therapeutic. Companies already have some reason to be quick, he says, but not as much as society would want.

His example is a vaccine manufacturer. Most vaccines fail because they are hard and complex to produce, so a manufacturer will not build or expand a factory to get doses flowing until the vaccine has been shown to be safe and effective and is about to be approved. If the goal is to make firms move earlier, the incentives have to be bigger.

He lists several tools. One is a large prize, for instance a billion dollars for the first vaccine that meets stated criteria on effectiveness (he gives 60% or 70% as examples) and safety. Another is to pay directly for manufacturing costs so that a firm builds the factory. A third is an advance market commitment or advance purchase, or paying for capacity that is built at risk. Each has pluses and minuses, but they share one logic: firms lack as strong an incentive to ramp up quickly as society would like, so they need "some extra juice." The host adds that the incentive is weaker still because a vaccine that works can only be sold to a person once.

The host asks how a person with a good vaccine idea but no money for manufacturing or testing could ever claim a billion-dollar prize. Tabarrok says this is a tricky trade-off. On one side, government is poor at picking winners and losers, as industrial policy shows, and nobody knows which vaccine type will succeed: he names live attenuated, killed virus, mRNA and DNA vaccines, and says the answer might be a therapeutic or some other innovation. A prize opens the field to "the crazy ones" who would not get through NIH committees, which he links to the longitude prize being won by a clockmaker rather than Newton.

On the other side, some firms may not have the capital, and for them the case for funding up front is stronger. He says there is no easy solution in a pandemic and that one should go "all guns blazing," using almost every available tool. He is working on the problem with Michael Kremer, who is known for the advance market commitment for the pneumococcus vaccine. Tabarrok says that commitment, built on a promise of a fixed price, probably saved 700,000 lives and reached millions of children. For COVID, however, they are pushing toward paying a large share of the manufacturer's costs up front, for building a factory or repurposing capacity, because in this context that turns out to be cheaper. Right now, he says, they lean toward paying for manufacturing capacity.

The host suggests venture capitalists could finance researchers in exchange for a share of the prize. Tabarrok agrees that this is what happens when markets work well, but notes that it takes a long time to set up. There are information problems at every stage, and venture capitalists are not especially good at picking winners either. They do have skin in the game and an advantage over government, but to learn enough about vaccines to tell scams from real possibilities takes time.

now is the time to throw money at the problem

Tabarrok calls himself a conservative free-market economist and says this is not something he usually says. His reasoning is that the cost of the disaster is already running in the trillions, so it pays to advance on all fronts, even though money will be lost and some funded vaccines will fail. He points to the AstraZeneca vaccine from Oxford, which looked great, had roughly two billion doses pre-ordered, and had a clinical trial shut down the previous week because of a negative effect; it was unclear whether the vaccine or something else caused it. He says he hopes it is a false alarm, but that such setbacks can happen at any stage. The lesson he draws is to hold a diverse portfolio: some candidates supported through manufacturing capacity, some through a guaranteed price, and some through fast grants like those Tyler Cowen has been giving out, taking many shots on goal.

Asked whether prizes are only a pandemic tool or are useful wherever the problem or solution space is large, Tabarrok says the latter. He still sees a pandemic-specific advantage: a prize is easy to explain to Congress and can be announced before the money is raised or a prize committee is chosen. This matters to him because he finds Congress lazy, slow and complacent: it took ages to set up testing funding, and then did not spend the money.

He also notes that, as he recalls it from Obama-era law, every agency may set up a prize of up to $50 million, and these can be accumulated. Two, three or four agencies could therefore create a $100 million or $200 million prize by getting together. They have not done so, but could. The second argument is that prizes are especially good when experts have failed. Tabarrok stresses that he is not anti-expert and that experts are usually the best people to ask, but when they have run up against a wall, a prize can bring in outside thinkers.

The host asks why grants dominate research funding. Tabarrok says he does not really know. Prizes were far more common in the 19th century, tailed off in the 20th and have seen a modest resurgence. He offers a tentative explanation: giving out money is a powerful and pleasant job, and a prize with strict criteria that pays only on success gives less power to the people who run it. He calls it a puzzle and says he is not sure.