Incentives to innovate could be considerably improved by enhancing data-package exclusivity. In the case of incentives to develop new medicines, data-package exclusivity protects for a period of years an innovator’s regulatory submission package to the Food and Drug Administration from being used as a source of information by a company that produces generic products. The period in Europe is 10 years plus an additional year if the innovator has gained approval for more than one indication. The United States grants data exclusivity for a new chemical entity for 5 years; a second indication is entitled to 3 years of exclusivity. Those periods are generally too short to stimulate investment. Thus, innovation incentives in the United States are almost entirely patent-driven.

The current system has been successful in stimulating the creation of new molecules, but the limitations of the patent system sometimes result in denying patients the best that the pharmaceutical industry could offer. The limitations are due largely to the time constraints under which the patent system operates. Patents generally must be filed as quickly as possible after an invention occurs, and the ticking clock creates a tension with other aspects of drug development.^a

The demands for data on a molecule’s safety and efficacy are increasing. The generation of the necessary data requires time and money. It is to patients’ benefit for as much time as appropriate to be devoted to the development of the data, but spending the time lessens the return on the developer’s investment because it encroaches on the patent term. Bringing a new medicine to patients requires a sequence of major breakthroughs, which in the current system must be accomplished well before the life of a patent runs out. Often, the clock does run out, and the innovator must start over with a new molecule simply to get time “back on the clock.” As a result, there is an ever-growing “graveyard” currently comprising more than 10 million compounds. There is no incentive to exhume these compounds in the absence of substantial data-package exclusivity, because patents will be either unavailable or of such narrow coverage that they would be easy to avoid in developing a related drug.

In addition, there is little incentive to pursue new indications for old molecules without appropriate data-package protection. Indeed, when no compound patent covers the product, there is a disincentive to develop new indications. Generic medicines may be approved for a smaller number of indications than those associated with the innovator’s drug. If there is no compound patent and one of the indications is unpatentable, the generic medicine may be approved only for the unpatented indication. The innovator’s entire market could then be eroded because