The World Trade Organization requires members to comply with minimum standards for intellectual property protection. These standards, set out in the ‘TRIPS’ Agreement,¹ were accepted as part of a multi-faceted trade deal, but they did not enjoy broad support in much of the developing world. The rules required for the protection of pharmaceuticals, in particular, have been a source of bitter dispute ever since they were first introduced during the Uruguay round of trade talks in the 1980s. The industry has resolutely insisted on the importance of patents to support sustained research, but developing countries and their advocates fear that patents will raise the price of drugs out of reach of the poor. The recent breakdown of international negotiations over the export of generic pharmaceuticals into countries without domestic manufacturing capacity (the ‘paragraph six’ debate) is only the most recent manifestation of the tensions created by patents on drugs.²

Although the fight over TRIPS has proved extremely difficult to settle, it does not stem from fundamentally conflicting goals. The pharmaceutical industry wants above all to protect its major markets in the developed countries; the desire to see effective protection evolve in poorer countries as they grow into more attractive markets takes a distant second place. Making profit in the poorest countries is not an industry objective. The development community and NGO (non-government organization) activists, on the other hand, are primarily concerned with keeping prices down in the poorest countries, so that drugs remain accessible. In substance these are not opposing objectives.

Of course, as the long debate makes clear, the solution is by no means obvious. The difficulty lies in part in the very confrontational and public way in which the parties involved have addressed disagreements over patents and prices since the advent of TRIPS—an exceedingly counterproductive process. Each side feels that it has made significant concessions in response to the demands of the other, only to be rebuffed, deceived, or further criticized. Whether or not these feelings are reasonable, the deep distrust that now exists on both sides is a major obstacle to constructive engagement.

The lack of attractive options on the table has also limited progress. A viable solution needs several basic features. The first is clarity of language. Ample experience shows that without clarity, future fights over interpretation are inevitable. Second, the proposed system must look fair and be broadly acceptable to most people on both sides. Finally, it must have limitations and structure. Certainty about the rules is clearly crucial to firms considering long-term research investments but it is also important to generics companies considering competitive entry. In the next section I describe a possible solution that satisfies the criteria above.

The Foreign Filing License (FFL) Solution
The basic structure of protection created with this proposal allows generic competition in poorer countries and offers broader IP protection as a country’s market potential increases. The structure is illustrated in Figure 1. Countries are listed in increasing order of annual income per person on the vertical axis. Disease classes are listed along the bottom. These are sorted starting on the left with diseases for which pharmaceutical sales are relatively concentrated in developing countries (for example, malaria drugs), and moving right, to diseases that are prevalent everywhere but have almost all of their pharmaceutical market in the developed countries (for example, cancer drugs). Taking each disease class in turn, the policy would allow generic competition in a group of poor countries that together represent no more than, say, 2 percent of the global sales in that class. The number of countries included for each disease class would thus depend directly on the size and location of the worldwide markets.

Figure 1: Along the horizontal axis are disease classes, starting with the classes where pharmaceutical sales are most concentrated in developing countries. On the vertical axis are countries listed in order of GDP per capita measured in constant US dollars. The white region indicates the area where generic competition would be permitted under the policy. The shaded region indicates the countries and diseases for which patent protection would be available to all inventors. Above the maximum income level (here $5,000) the policy has no effect.

The resulting structure of protection would be as shown in the figure. For countries with incomes below the dotted line, there would effectively be no patent protection and thus no potential for patents to limit generic producers from entering the market.³ As a country’s income increased, patent protection would widen, beginning first with new products treating diseases of specific importance in developing countries. The increasing breadth of protection at higher levels of income is shown as the gray area in the figure. For countries above the highest cutoff (in the figure at $5,000) full protection is available on all pharmaceutical products.

Although the system’s effects occur in developing countries, it does not require those countries to do anything at all. In fact, their obligations under TRIPS would stay just as they are now. The policy is implemented through patent law in developed countries and is achieved as follows (described first for the U.S.). An inventor in the U.S. currently must obtain permission to file for patents overseas. The essence of the policy is simply to require that the patentee sign a declaration in order to obtain this permission. The declaration states that the permission being sought will not be used to prevent the sale of drugs in the countries, and for the diseases, shown as the white area of Figure 1. If the patent holder later starts an infringement suit to prevent a competitor from selling a product in one of the proscribed markets, the firm would have falsified its declaration and in return would lose the ability to enforce the corresponding U.S. patent in respect of the product at issue in the infringement suit. Since the developed country market will almost invariably be vastly more valuable than the developing country market, the policy gives inventors a compelling reason to refrain from exercising their patent rights in the markets indicated in white.

Discussion
The design of the proposed system has industry refrain from enforcing patent rights in markets that representat most, say, 2 percent of sales in any disease class. Adjustments in coverage are made automatically the declaration is updated year by year to reflect changes in markets. For example, India would be effectively allowed to produce generics initially, but it would be within a system where the country would grow into standard patent protection as its market potential increased. This extension in protection would not require India itself to make any changes or invest resources. All of the institutions required to implement and operate the system are in the rich countries and already in place.

While protecting firms’ important markets, the system allows generic competitors to enter the poorer markets without any need to go through the procedural process of compulsory licensing. At the same time it gives as much encouragement as a patent system can offer to do research on products specific to developing countries. Furthermore, the system gives as much support as possible to the building of research capability in the developing world. Because the policy applies only to inventions in the developed countries, inventors in developing countries would continue to enjoy full patent protection on all products. This solution gives local inventors a chance to develop an appreciation of the patent system, and it gives governments an incentive to build the enforcement capability necessary for a patent system to really work. Although the effect of the policy would clearly differ across countries, it would do so in a gradual and rational way, based on relevant objective criteria. It thus avoids the objectionable process of breaking countries into designated “poor” groups. Because the policy does not in any way alter the TRIPS Agreement, existing flexibilities in TRIPS remain to address public health concerns in countries falling outside of the system’s scope.

In sum, both the existing obligations and the flexibilities of TRIPS remain in place. Companies simply forego exercising their rights in the poorest countries, thus removing the major area of disagreement.

Is such an arrangement actually feasible? Yes, albeit with some effort. First, the primary actors need to decide that getting in place a sound and acceptable patent system is a better strategy than “controlling the situation” as more inevitable crises arise. There remain divergent views within and between organizations on this point. There is a role here for those not directly involved to encourage something better than crisis control even if crisis control is the easier route in the short run (which it usually is). Next, these same primary actors need to convince themselves that the “solution” described here does not contain any dangerous features. This requires some real thought, because the mechanism is unusual. With the main actors interested, the next step is to canvass a larger group to ensure that the system is sufficiently “broadly acceptable.” If it is, then the governments of countries with research-based pharmaceutical sectors need to coordinate on implementation. This group includes at least the U.S., Canada, the European Union, and Japan. Each country would revise its patent code to add a foreign filing license provision with the requisite details, if it is not already there.