There was an interesting article in last week’s Economist about the use of prizes to promote innovation. It was supportive of the idea in general, but it seemed to gloss over the economic  arguments.  I think it is a shame that the Economist did not take the opportunity to explain the economics of rewarding innovation, and in particular to explain in economic terms why our current arrangements do not do a good job of creating incentives for innovation that benefits developing countries.

You can think of patents as a kind of prize.  When you invent a new product, the government gives you the right to operate a temporary monopoly. This enables you to charge more than the marginal cost, and the premium is your “prize”. This arrangement has the huge advantage that it links your reward to the amount people are willing to pay for your invention, so it encourages innovations that people actually value.

This kind of prize as a reward for innovation may be fine for a new kind of vacuum cleaner, or for Lady Gaga’s latest album. But it has two big disadvantages which are especially relevant for people who live in developing countries.

First, the use of patents prevents some people from benefiting from the new technology if they are unable to pay the higher price.  If a company develops a drug for heart disease, or a more efficient form of solar panel, the patent will enable them to charge much more than marginal cost for their product. That’s how the inventor gets paid. But the result is that millions of people will not be able to afford that product – though they might be able to afford it at marginal cost. The temporary monopoly results in fewer people benefiting from new technologies than ought to benefit, in the sense that those people would be willing and able to pay the marginal cost.  This is potentially a big welfare cost to society as a whole. It means, for example, that people may die of heart disease because they can’t afford the high price of the drugs, even though they could buy the drug if it were sold at marginal cost; or they can’t use new fertilizers or seed technologies, even though the benefits to them of doing so exceed the cost.

Second, if we reward inventors by granting them temporary monopolies, we only create incentives to develop products for which there are likely to be enough consumers wealthy enough to pay a monopoly price.   Nobody will invent a vaccine against malaria, or a cassava plant that resists mosaic virus, based on the possible rewards they will get from charging high prices to its consumers.  So the patent system is a prize for people who invent cures for baldness, but not a prize for people who invent ways to prevent the spread of malaria.

For these reasons, other incentives, such as prizes, Advance Market Commitments, and similar mechanisms, may be effective either as alternatives or complements to the patent prize of a temporary monopoly, especially for technologies that would have benefits in developing countries.

The Economist quotes Tachi Yamada, the president of Global Health at the Gates Foundation, as suggesting that Advance Market Commitments or prizes may not work well for drugs that require a long time to develop:

Tachi Yamada of the Gates Foundation is a big believer in giving incentive prizes, but gives warning that it can take 15 years or more to bring a new drug to market, and that even AMC’s carrot of $1.5 billion for new vaccines may not be a big enough incentive. No prize could match the $20 billion or so a new blockbuster drug can earn in its lifetime. So, in some cases, says Dr Yamada, “market success is the real prize.”

This seems to reflect the suggestion that is sometimes made that Advance Market Commitments may not be appropriate for for early stage drugs, but the economics of this argument is faulty.

It is clearly true that the reward for bringing to market an early stage medicine, such as an AIDS or malaria vaccine, would need to be higher, both because of the greater uncertainty and risk of failure, and because the rewards are further in the future.  So an AMC for an early stage product would probably need to be larger than for a late stage product that just needs some tweaking for use in developing countries and some investment in bigger production facilities.  But let’s not overstate this.  The median total market size for new chemical entities that pharmaceutical companies actually bring to market is about $3-$4 billion.  Most medicines are not $20 billion blockbusters.  So $3-$4 billion is roughly the market size that the private sector considers sufficient reward to develop new medicines.   We don’t need to match the blockbusters.  An AMC of $4 billion might well be enough to incentivize the development of a malaria vaccine: and let’s not forget that if it turns out not to be enough, it won’t have cost the funders anything.

Furthermore, just as the firms discount the prize by the risk of failure, the funders should similarly discount the cost.  If there is a 25% chance that no vaccine will be developed (because the technology is uncertain) then firms will discount the “prize” – that is, the value of the committed market – when they make their investment decisions.  But in this case, the expected cost to the funders of a $4 billion pledge is $3 billion, and this is what they should include in their value for money calculation.  That means that even though the nominal amount that has to be promised for an early stage product needs to be higher for a given impact on R&D, to take account of the probability of failure, the expected cost to funders is not higher.

The same point can be put another way.  A high probability of failure makes all investment in R&D less attractive, but it does not make AMCs relatively less attractive than other forms of funding.  When the probability of failure is high, the expected return from each dollar spent encouraging innovation is lower. This is true if that dollar is spent up-front in the form of research grants of the kinds normally given by aid agencies and foundations (since the higher probability of failure reduces the expected benefits of the grant), or in the form of a prize or promised market (since the higher probability of failure reduces the expected benefit to firms, and so reduces the incentive for them to invest in R&D).  The effect is the same either way. Higher probability of failure is clearly bad, but it does not make AMCs relatively less efficient as a way to pay for research for early stage products.

Whether an AMC for an early stage product is good value for money depends ultimately on the value of the product.  If donors were to spend $4 billion buying a malaria vaccine for use in developing countries, it would be a hugely good investment, saving millions of lives a year at a fraction of the price of many other interventions. It would result in huge savings on trying to prevent malaria in other ways, or treat to treat malaria; and the resulting reduction in the burden of malaria would have huge economic benefits for developing countries. Given that there is no question that donors would want to spend at least $4 billion paying for a malaria vaccine to be used across the developing world, it is inefficient for them not to say so right away, and thereby create incentives for private sector investment in accelerating its development.  The risk of poor value for money in aid spending comes not from making the commitment, but from failing to do so.

When Dr Yamada says that “market success is the real prize”, he seems to be missing the point that market success is not a good way of rewarding innovation for developing countries.   If we rely on market success, in the form of a temporary monopoly, to reward innovation then we will exclude half the world’s population from being able to access technologies developed with rich markets in mind, such as drugs against cancer and heart disease, clean energy, new agricultural technologies, or new software.  And “market success” creates no incentive to develop technologies which primarily benefit the world’s poor such as a vaccine against malaria or a variety of cassava that resists the mosaic virus, because inventors know that the people in poor countries cannot afford the monopoly prices that would enable inventors to recover their costs.