GlaxoSmithKline prices its AIDS drugs above marginal cost. If GSK didn’t have a monopoly, competition would push prices down, more people could afford to buy Combivir, and total surplus would increase (i.e., deadweight loss would decline). So isn’t the solution to the monopoly problem obvious? Open up the industry to competition by refusing to enforce the firm’s patent or force GlaxoSmithKline to lower its price.

In fact, many countries pursue one or the other of these policies. India, for example, has traditionally not offered strong patent protection, and Canada controls pharmaceutical prices. India’s and Canada’s policies have successfully kept pharmaceutical prices low in those countries. Many people argue that the United States should also control pharmaceutical prices. Unfortunately, the story is not so simple. We need to revisit our question, what’s wrong with monopoly?

In the United States, researching, developing, and successfully testing the average new drug cost nearly $1 billion.8 Firms must be compensated for these expenses if people expect them to invest in the discovery process. But if competition pushes the price of a pill down to the marginal cost, nothing will be left over for the cost of invention. And he who has no hope of reaping will not sow.

Patents are one way of rewarding research and development. Look again at Figure 13.3, which shows the green rectangle of monopoly profit. It’s precisely the expectation (and hope) of enjoying that monopoly profit that encourages firms to research and develop new drugs.

244

If pharmaceutical patents are not enforced, the number of new drugs will decrease. India is poor and Canada is small, so neither contributes much to the global profit of pharmaceutical firms. But if the United States were to limit pharmaceutical patents significantly or to control pharmaceutical prices, the number of new drugs would decrease significantly.9 But new drugs save lives. As noted in the introduction, antiretrovirals like Combivir were the major cause of the 50% decrease in AIDS deaths in the United States in the mid-1990s. We should be careful that in pushing prices closer to marginal cost, we do not lose the new drug entirely.

In evaluating pharmaceutical patents, you should keep in mind that patents don’t last forever. A patent lasts for at most 20 years, and by the time a new drug is FDA-approved, its effective life is typically only 12–14 years. Once the drug goes off patent, generic equivalents appear quickly and the deadweight loss is eliminated as price falls.

Pharmaceuticals are not the only goods with high development costs and low marginal costs. Information goods of all kinds often have the same cost structure. Video games like Halo, Madden NFL, and The Sims have typical development costs of $7 million to $10 million; Grand Theft Auto IV cost more than $100 million to develop. Once the code has been written, however, the marginal cost of distributing on the Internet is close to zero. Prices, typically $40–$60, are therefore well above marginal costs. Since prices exceed marginal costs, there is a deadweight loss, which in theory could be reduced by a price control. Reducing prices, however, would reduce the incentive to research and develop new games. What would you rather have: Pong at $2, or, for $50 a game, a constant stream of new and better games?

Video games may seem trivial, but the trade-off between lower prices today at the expense of fewer new ideas in the future is a central one in modern economies. In fact, modern theories of economic growth emphasize that monopoly—when it increases innovation—may increase economic growth.

Nobel prize–winning economic historian Douglass North argues that economic growth was slow and sporadic until laws, including patent laws, were created to protect innovation:

[T]hroughout man’s past he has continually developed new techniques, but the pace has been slow and intermittent. The primary reason has been that the incentives for developing new techniques have occurred only sporadically. Typically, innovations could be copied at no cost by others and without any reward to the inventor or innovator. The failure to develop systematic property rights in innovation up until fairly modern times was a major source of the slow pace of technological change.10

Is there a way to eliminate the deadweight loss without reducing the incentive to innovate? Economist Michael Kremer has offered one speculative idea.11 Take a look again at Figure 13.3. The green profit rectangle is the value of the patent to the patent owner, $800 million. Suppose that the government were to offer to buy the rights to the patent at, say, $850 million? The monopolist would be eager to sell at this price. What would the government do with the patent? Rip it up! If the government ripped up the patent, competitors would enter the field, drive the price down to the marginal cost of production, and eliminate the deadweight loss. In other words, Combivir would fall from $12.50 a pill to 50 cents a pill, and more of the world’s poor could afford to be treated for AIDS.

245

The great virtue of Kremer’s proposal is that it reduces the price of new drugs without reducing the incentive to develop more new drugs. Indeed, by offering more than the potential profit, the government could even increase the incentive to innovate! As usual, however, there is no such thing as a free lunch. To buy the patent, the government must raise taxes, and we know from Chapter 6 that taxes, just like monopolies, create deadweight losses. Also determining the right price to buy the patent is not easy and some people worry that corruption could be a problem.

Kremer’s idea has never been tried on a widespread basis, but despite these problems, economists are becoming increasingly interested in patent buyouts and the closely related idea of prizes as a way to encourage innovation without creating too much deadweight loss.