How a Great Power Falls Apart
Decline Is Invisible From the Inside
Fifteen years ago, the United States entered into its first trade deal with pharmaceutical rules exceeding those required by the World Trade Organization (WTO). A debate over the potential consequences of such deals for medicine has raged ever since. Activists have argued these U.S. trade deals raise the prices of lifesaving drugs and limit access to cheap generic medicines for millions of people outside of the United States, effectively acting as a high tariff on medicines and killing patients. Industry supporters insist that the provisions of trade deal allow for greater rewards for innovative drugs and spur research and development, and they argue that weakening the pharmaceutical rules in U.S. trade deals would chill global investment, slow the development of breakthrough treatments for suffering patients, and force drug companies out of business.
The only way to evaluate the large claims about the pharmaceutical terms in U.S. trade agreements, such as those in the proposed Trans-Pacific Partnership (TPP), is to examine the effects of previous deals. Until this spring, however, only two small studies on that subject had been completed. In March, I published a study in Foreign Affairs (“A Dose of the TPP’s Medicine,” March 23, 2016), presenting more than a decade’s worth of pharmaceutical sales data from the IMS Institute for Healthcare Informatics on 15 of the 17 countries with recent U.S. trade deals with drug terms that go beyond WTO requirements. The study also presented data for four comparator nations of similar income but with no U.S. trade deals or no exclusivity requirements. The results suggested that the terms of U.S. trade deals have not led to large increases in drug prices or shifts away from lower-cost generic drugs in the countries that have agreed to them, although more substantial effects might appear over longer periods and for particular classes of medicines. (An interactive feature on the website of the Council on Foreign Relations (CFR) outlines the study's data, sources, and methodology.)
Amy Kapczynski, Bhaven Sampat, and Kenneth Shadlen ("The TPP and Drug Prices," October 27, 2016) do not question the study’s data or results themselves, but contest their relevance to the debate over the effects of trade deals on medicines, particularly in the case of the TPP. Yet their response overlooks most of the findings and data in my study, misrepresents the remainder, and commits basic factual errors. More importantly, these scholars offer no empirical evidence on the effects of U.S. trade deals on other countries’ access to medicines and their prices. In neglecting to do so, Kapczynski, Sampat, and Shadlen have penned a response that perfectly represents why the trade and medicines debate has persisted so long and advanced so little.
A REASONABLE RATIO
First, Kapczynski, Sampat, and Shadlen argue that examining changes in the ratio of drug spending to overall health-care expenditures, as my study does, is an ineffective way to assess the impact of trade deals on drug prices. Yet that ratio is useful for determining whether growth in drug expenditures is outpacing increases in other health-care costs, which is why the Organization for Economic Cooperation and Development uses the metric as one of its three main indicators on drug spending. It is true that, when there are spikes in overall health expenditures, as there have been in the United States in recent years, the ratio can mask increases in drug spending. That is why I also used numerous other measures to assess changes in drug spending and prices that might be attributed to the U.S. trade deals in the study. The three scholars ignore these metrics: the amount of drugs consumed per country; the share of each country’s pharmaceutical market devoted to off-patent originator drugs, by volume and value; the share of each country’s market devoted to branded products, also by volume and value; and annual growth in per capita drug spending.
Examining the data sheds light on some of the claims that activists and lobbyists have made regarding the effects of U.S. trade deals. The volume of pharmaceuticals consumed has grown in the countries with U.S. trade deals in my study, which suggests that governments and patients have not been rationing medicine to compensate for higher prices, as many feared. The data also suggest no discernable trend away from cheap generic drugs toward on-patent or branded medicines, as some observers expected. The growth in per capita pharmaceutical spending in countries with U.S. trade deals has been comparable to nations of similar income without U.S. trade deals or data exclusivity rules. Nor is there any obvious correlation, over the period I studied, between these results and the length of time that the trade deals have been in force or the stringency of their pharmaceutical provisions. These results are also consistent with the testimony of then-Peruvian Ambassador to the United States Luis Miguel Castilla, who reported earlier this year that prescription drug prices in his country rose 2.8 percent and that drug consumption expanded by more than a third between 2009, when Peru’s trade deal with the United States entered into force, and 2014.
Looking closely at these other metrics might have helped Kapczynski, Sampat, and Shadlen avoid some of the factual errors in their response. First, they claim that there are “no more than ten years of data for any country” in my study and “just one year of data” for two of those countries. Second, Kapczynski, Sampat, and Shadlen write that pharmaceutical patenting is “new” in several lower-income countries in my study. For these reasons, these scholars write that my study could not identify the effects of U.S. trade deals on drug prices, since those deals extend patent protections “only after a decade or more” of a drug entering the market, especially in countries with new patent systems and few drugs that have reached the end of their patent terms.
In fact, the IMS data my study used cover not ten but 11 years, from 2004 to 2014, inclusive; the data on off-patent originator drug prices also includes 2015. The data extend as long as 11 to 13 years after several U.S. trade deals in the study entered into force and two years after the passage of the more recent U.S. trade deals the study assesses. (A table detailing those agreements and their relevant dates and terms is in the interactive on CFR's website.) The only metric in the study limited to ten years of data is the ratio of drug spending to overall health-care expenditures; that is because, at the time of the study, the World Bank had not yet released data on aggregate health expenditures for 2014.
Further, none of the countries I examined adopted pharmaceutical product patents later than 2000. Most, including many of the developing nations, adopted drug product patents earlier, in the mid-1990s. As a result, the data my study considered covers periods that extend no less than 14 and as many as 28 years after the countries in the study adopted pharmaceutical product patents. Recent studies have found the average length of time a patented drug spends on the U.S. market before it faces generic competition to be 11 to 12 years. That time may be shorter in developing countries, where fewer secondary patents are filed to delay generic competition. So although Kapczynski, Sampat, and Shadlen raise a fair point about the relative newness of pharmaceutical product patents in some countries, they overstate its implications for my study.
The dispute over the effects of U.S. trade deals on pharmaceutical prices goes beyond the extension of patent terms for drugs. It involves questions about the patent eligibility, patent linkage, and data exclusivity terms in those deals, which many argue keep generics off the market and raise the prices of newly registered, off-patent medicines. There is no reason those effects should occur only after a decade. The trade deals for all 15 countries in my study included linkage and exclusivity provisions, but the current data do not show discernable shifts away from generics in volume or value in these nations, even relative to countries without U.S. trade deals or data exclusivity rules.
For the four countries in my study for which IMS had data on off-patent, originator medicines, there has been no upward trend in the prices of drugs launched in the three years after U.S. trade deals entered into force. Kapczynski, Sampat, and Shadlen point out that the sample sizes for two of these countries were fairly small, since relatively few off-patent originator drugs were launched in the study period. Yet this just illustrates the point that pharmaceutical provisions in U.S. trade deals do not apply retroactively and the number of affected drugs will be small and build only slowly, as my study notes. In Australia, for example, originator drugs launched since the country’s free trade deal with the United States came into effect in 2005 still only represent 13 percent of that market, and only a subset of those originator drugs will gain patents or data exclusivity due to the expanded terms in U.S. trade deals. The scholars also note that two of these four countries had already adopted a version of data exclusivity before their U.S. trade deals, but this is also true of nearly every country in the proposed TPP.
MATTERS OF FACT
The final argument that Kapczynski, Sampat, and Shadlen make is that the effects of previous U.S. trade deals might not reveal much about what to expect from the TPP. This is the case, they argue, because the TPP includes an investor-state dispute settlement (ISDS) mechanism that could let companies sue governments for changing drug patent and pricing laws, as well as restraints on the procurement and formulary practices and patent standards that governments have used to lower drug costs.
Yet none of these provisions are unique to the TPP, and its terms are more favorable to public health than those of previous trade deals. Earlier U.S. agreements with Australia and South Korea also include provisions on transparency and due process in pharmaceutical reimbursement. Kapczynski, Sampat, and Shadlen argue that the TPP’s version of these provisions could undermine government procurement and formulary practices, but they miss that the TPP annex that they reference excludes from its scope “government procurement of pharmaceutical products and medical devices,” including “formulary development and management.” The TPP also states that its pharmaceutical reimbursement provisions may not be used to “to modify a Party’s system of health care” or “determine health expenditure priorities” and are not subject to dispute settlement under the TPP.
Most U.S. trade deals have also incorporated ISDS, including the agreements for thirteen of the countries in my study. All of the countries in the TPP have entered into previous agreements that provide for ISDS—ranging from Brunei, which is a party to 13 active deals, to Malaysia, which has 59. (The United States already has 54 active trade and investment agreements that include ISDS, some of which are nearly 30 years old.) The TPP’s version of ISDS acknowledges the legitimacy of environmental, health, and safety regulations, including pharmaceutical pricing and reimbursement; carves out tobacco control measures; and incorporates new limits on ISDS claims. If a firm were to sue a country that is party to the TPP, it would likely be under the auspices of an older agreement less friendly to public health, such as the 1999 agreement between Australia and Hong Kong under which Philip Morris brought a recent ISDS claim against the Australian government.
Finally, most of the countries that have signed on to the TPP have already entered into U.S. trade deals and have variants of the patentability terms required by the TPP. In the areas where TPP mandates a significant change, the less wealthy countries that would be parties to the agreement would get more time to adopt it. The TPP would provide Vietnam, for example, up to 18 years to adopt data-exclusivity provisions for new biologic drugs. Most TPP countries, including middle-income states such as Mexico and Vietnam, already regulate drug prices. For all these reasons, there are grounds to expect that the TPP’s effects on drug prices are likely to be similar to the effects of previous U.S. trade deals.
Momentum may be gathering for changes in U.S. trade policy and domestic prescription drug pricing, but where that is occurring, it is being driven by empirical evidence on displaced U.S. manufacturing workers, currency manipulation, and abusive drug price hikes. My 19-country study brings forward the current empirical evidence on the debate over U.S. trade deals and medicines. It should be updated as more data become available and extended to the settings and classes of medicines where conflicts are most likely to arise; the results should be weighed in the context of the broader benefits and costs of trade deals. Fifteen years of arguments and large claims without empirical evidence have failed to move the debate over U.S. trade deals and access to medicines forward. It is disappointing to see Kapczynski, Sampat, and Shadlen respond to this study only to do more of the same.