With some of the goods and services we consume, supplying the first unit costs vastly more than the rest. Building a subway line costs billions. The additional cost of making it carry more passengers is minuscule by comparison: pennies per ride for making the trains longer or more frequent. What should users be asked to pay in such cases?
There’s a free-market answer. Whoever builds the subway sets the price. He will build the subway only if he expects to profit from his investment. By setting the price above marginal cost, he earns a mark-up on each ride. He will want to adjust this price so as to maximize his surplus (mark-up multiplied by sales volume) in each time period. His investment will make a profit insofar as it is exceeded by the discounted present value of all future surpluses.
In a city with high inequality, especially, the free-market model leads to a high price (say $15 per ride). This is a pity because providing an additional ride costs much less ($1 per ride, say). It seems that our entrepreneur could make even more profit by selling additional rides at $6 or even $2. But this won’t work because the cheap rides would be snapped up also by people who would otherwise pay full fare. To avoid this problem, special offers must be carefully targeted to those unlikely to ride at full price — students and the elderly, for instance. The bulk of the pity remains: the world is losing many rides priced between $1 and $15 that would be desirable for both the entrepreneur and the rider. This forgone value is what economists call a deadweight loss.
Deadweight losses indicate potential win-win opportunities. The residents of the city can get together and pay the entrepreneur to lower his price. Suppose that, at $15 per ride, he would get 40,000 riders a day, whereas at $1.50 per ride he would get 400,000. His daily surplus in the first, profit-maximizing scenario would be $560,000 (40,000 x $14) and in the second scenario $200,000 (400,000 x $0.50). Our entrepreneur would happily lower the price to $1.50 if we paid him $400,000 per day to do so.
Should we? Absolutely! Those who are now paying full fare will save $540,000 (40,000 x $13.50) daily. Extra passengers will enjoy 360,000 daily rides, each valued by its purchaser at somewhere between $1.50 and $15 — that’s easily worth another half a million each day. All residents — even those who never take the subway — will benefit from reduced traffic, easier parking, and lower pollution. It’s a no-brainer, already implemented the world over. By finding a way to cover much of the fixed up-front investment through the tax system, we are saving much larger amounts from reduced fares and reduced deadweight losses.
Using tax moneys in this way is not merely efficient (citizens collectively gain much more than they lose), but also fair, at least approximately. If tax revenues are used, then affluent residents contribute more. But they also benefit more, financially, as they would either have spent a lot on $15 fares or else have been prepared to pay similar amounts for easier parking and less congested traffic. Poor residents save less. If they would typically avoid the subway at prices above $2.50 per ride, the $1.50 rides they now enjoy are worth only $1 to them. So it’s not unfair that they contribute less than the affluent to the city’s payment for the price reduction. This is rough fairness, to be sure. But when citizens collectively contribute $400,000 daily for benefits that are worth three times as much to them, then rough fairness suffices to ensure that nearly everyone comes out ahead.
The subway solution is desperately needed in the much more important market for new medicines, which still runs on the unmodified free-market model. The cost of pharmaceutical innovation is high — a billion dollars or more per approved new product, on average — while the extra cost of manufacturing additional supply is minuscule. Firms protect their investment in innovation with 20-year product patents that — pursuant to the 1994 TRIPS Agreement — innovators can take out around the world. Patents enable them to veto the manufacture and sale of products embodying their inventions and, thus protected from competition, to charge extremely large mark-ups on their products.
The profit-maximizing mark-up on patented drugs tends to be exorbitant for two reasons. Massive economic inequality, both globally and within most countries, ensures that by pricing a drug to be affordable to the poor, a firm would lose far more in its mark-up than it would gain in sales volume. The poorest two-thirds, after all, have only 6 percent of global household income and 4 percent of global private wealth. Prices are propelled yet higher by the fact that affluent people (or their insurers or governments) are prepared to pay richly for medicines; so price increases, even at very high levels, cause only relatively small declines in the quantity sold.
Like millions around the world, Mary is suffering from a life-threatening disease. She knows that there is a cure that generic firms are willing and able to manufacture cheaply and to sell at an affordable price. But because it is illegal to do this, the cure is actually available only from the patentee, at a price Mary cannot afford. She wonders why the transaction she so desperately needs is illegal. Who would be harmed if generic firms could supply medicines to poor people at competitively low mark-ups? At the high price it is charging, the patentee sells nothing to Mary. What purpose is served by her death?
The core problem is that, if poor people could buy new medicines at low prices, many affluent people would seek to do likewise. Pharmaceutical companies would then be unable to recoup their R&D investments and would therefore stop developing new medicines. The exclusion of poor patients thus enables the world’s more affluent people to pay through mark-ups for the creation of new medicines — which, once the relevant patents expire, will also benefit future poor people.
This reasoning is sound. But it cannot justify the exclusion of the poor — and the resulting suffering and deaths — unless that is the only way of making pharmaceutical innovation sustainable. The subway solution shows that the choice so often presented to us — accept high prices or forgo innovation — is a false dichotomy: a third, viable approach is to use public funds to pay innovating firms enough so they agree to lower prices. The real cost of such payments is much smaller than it appears because governments would also spend much less on patented pharmaceuticals they buy through government insurance schemes (Medicare) or foreign aid programs. And taxpayers would save on health insurance and out-of-pocket expenses.
Negotiating and paying for price reductions on all patented pharmaceuticals would be extremely cumbersome and expensive. It would also be wasteful in many cases — for instance, where products are similar so that, once the price of one is lowered, reducing the prices of the others becomes much less useful. Which price reductions should we be willing to pay for from public funds and how much should we be willing to pay for each? To answer this question, we need some reasonable way of estimating a new medicine’s social value.
In analogy to the subway case, we might estimate this social value in monetary terms. Those who would have the medicine anyway save the difference between the new low price ($2, say) and the high, profit-maximizing price; anyone else buying the medicine for $2 saves the difference between $2 and the most she would have been willing to pay. But such an estimate of social value is implausible, even if willingness to pay could be measured. Consider a very poor patient, like Mary, who buys the cure for her life-threatening disease for $2. She was ready to pay $3, but that was all she had. Surely, enabling her to buy her survival adds more than a mere dollar to the social value of the price reduction.
A more reasonable measure of the social value of medicines is in terms of contribution to human health: to longer and healthier lives. A measure of this sort — the quality-adjusted life year (QALY) — has been in use for a generation. One extra year of healthy life is worth one QALY, and so are two extra years of life in a poor (0.5) health state and likewise five years of life in an improved (+0.2) state of health. Using such a measure, we might offer to pay for each innovation in proportion to its health impact. If the innovator accepts, its innovation is fully paid for and it is not permitted to mark up the price at all. If the innovator refuses, it can continue to charge high prices while public funds are used to reduce the price of some other innovation instead.
You have now seen the central idea of the Health Impact Fund (HIF), a proposed mechanism for improving the way we pay for new medicines. But some important details are still missing.
To make health impact rewards comparable to the earnings an innovator would otherwise derive from patent-protected mark-ups, the former might be based on the health impact during a product’s first ten years on the market. Patents last twenty years from the time of filing but typically have only about 10 years left when the product receives marketing approval. Payments from the HIF should roughly match this reward period. So the HIF offers to pay any innovator in proportion to the QALYs its medicine saves during its first decade in exchange for the innovator agreeing never to charge a mark-up on this product.
How much should we be prepared to pay per QALY? The HIF proposal draws on an elegant solution first proposed by Michael Abramowicz. In order to make public funds go as far as possible, let competition among innovators determine the reward rate. The HIF would fund fixed annual reward pools to be divided among registered medicines according to their health impact in the relevant year. The reward rate (dollar per QALY) is then nicely self-adjusting. If innovators find the rate richly lucrative, it will soon come down through additional product registrations. This assures taxpayers that they are not paying windfall profits to pharmaceutical innovators. Conversely, if innovators find the reward rate unattractive, it will rise as new registrations fall off (and older products reach the end of their reward period). This assures registrants that the reward rate will not sink to an unreasonably low level.
How can the HIF ensure that registered products are sold at the lowest feasible cost? Many pharmaceutical innovators already outsource their manufacturing, often to developing countries. The HIF can require registrants to solicit tenders for the manufacture of their registered products, to accept the cheapest qualifying tenders, and then to price products no higher than the cost of manufacture and delivery. By exploiting economies of scale, this system could make HIF-registered medicines even cheaper than off-patent drugs, which are often produced in small quantities by many competing firms. Manufacturers involved in the HIF system would produce much larger quantities, lowering average costs. The HIF combines the price-lowering element of vigorous competition with cost savings from scale.
Pharmaceutical companies are notorious for their marketing efforts, which often merely shift prescriptions around among similar products. The HIF would reward additional sales only insofar as they result in health gains. It would thereby motivate companies to focus their marketing toward patients who can benefit the most, to subsidize use of the medicine by very poor patients, to safeguard the medicine’s efficacy all the way to the end-user, and to ensure that patients and medical personnel understand the optimal protocol for using the product and are encouraged to adhere to it. This shift in how medicines are promoted would create additional social value: all patients are much better off if pharmaceutical firms focus on the therapeutic benefits they can achieve with their products rather than on the mark-ups they can collect from sales.
The size of the HIF’s annual reward pools is adjustable — the larger they are, the more medicines would be registered. A small pool size, however, is problematic in two ways. Health impact measurement is subject to economies of scale: assessing two products would cost much more than one-tenth the cost of assessing twenty. Small pools also lead to a volatile reward rate, making rewards less predictable and innovators more reluctant to register. A plausible minimum is $6 billion per annum, enough to maintain 20–30 products at any given time with 2–3 entering and exiting each year. In this scenario, the typical product might earn some $250 million annually, minus up to 10 percent in assessment costs, for a total of $2.3 billion over the decade.
HIF registration would be especially attractive for medicines that can bring large therapeutic benefits to many patients but offer unattractive returns outside the HIF. Medicines against diseases that disproportionately affect poor people are the most important example. Under current incentives, many serious health problems are underrepresented in pharmaceutical research because a large majority of those who suffer from them cannot pay high prices. By turning these “neglected diseases” into lucrative profit opportunities, the HIF would compensate for yet another important defect in the existing regime.
If the HIF is to stimulate important research that would not otherwise occur, then reward payments must be secure far into the future. Innovators may take ten or more years to develop a new medicine to the point of marketing approval, and another ten years to exhaust their reward payments. Only governments can credibly guarantee a substantial cash flow many years into the future — and, if they create the HIF at all, they will want to make such a commitment lest their promised funds be discounted by skeptical innovator firms. If all countries participated, a contribution of 0.01 percent of gross national income would suffice to reach $6 billion. Realistically, a contribution of 0.03 percent may be needed. At this rate, just the US plus China would suffice — or the European Union plus India and Brazil. Ideally, of course, many countries of all sizes and levels of development would join the partnership.
The HIF embodies the attractive idea that the health and survival of all human beings matters equally. In fact there are enormous disparities in the health-relevant resources flowing to rich and poor people. Consuming only $1 of every $10,000 of global income, the HIF would not end this disparity. But it would greatly reduce the disease burdens now plaguing poor populations — in three main ways. The HIF would
- foster the development of new high-impact medicines against diseases that are now neglected because innovators cannot recover their research and development costs from sales to the poor;
- promote access to new medicines by limiting the price of any registered product to the lowest feasible cost of production and distribution;
- motivate registrants to ensure that their products are widely available, perhaps at even lower prices, and that they are competently prescribed and optimally used.
Some have argued that international aid is wasteful and may even undermine the countries it is meant to help. The HIF is not, however, a form of aid. It creates a small market where innovators can compete on health impact with products priced at competitive levels, thereby increasing the range of choices for people in both rich and poor countries. To be sure, countries funding the HIF would contribute according to income. But this is not unfair insofar as the richer populations bear most of the cost of innovation now, by buying patented medicines at high mark-ups. By agreeing to pay for some important pharmaceutical innovations differently, the world’s affluent would allow the poor to share access at cost.
Is it possible, at reasonable cost, credibly to assess the therapeutic benefits of a new medicine in diverse populations around the world? Experts in health economics and epidemiology — some convened at a collaborative workshop at the National Institute for Health and Clinical Excellence (NICE) in London, others consulted at the Seattle Institute for Health Metrics and Evaluation — have made valuable contributions to showing how this can be done. But full reassurance of governments and innovators requires conducting “pilots” of the HIF concept. Each pilot would consist of a contractual arrangement in which a firm is rewarded explicitly on the basis of assessed health impact for one product in a single jurisdiction. Depending on the scale of the jurisdiction and the prevalence of the target disease, a pilot could be run at a relatively small expense.
In a suitable pilot, a firm would agree to reduce the price of a product in one jurisdiction, which could be a city, province, country, or region. In exchange, it would receive rewards based on the product’s assessed health impact. The incentives should be designed so that, if the firm appropriately responds to them (making cost-effective efforts to enhance its product’s health impact), its profits would increase. For example, the firm’s sales could be rewarded differentially: it would receive no reward for patients switched from an equally effective drug, small rewards for patients switched from a less effective drug (e.g. one with greater toxicity and therefore typically lower compliance), and large rewards for patients who had previously had no treatment at all. The scheme of rewards would be agreed with the firm in advance.
A pilot would demonstrate the feasibility of reliable health impact assessment and show the effect on behavior of rewarding a firm according to health impact rather than through mark-ups. A pilot would also provide practical evidence on the best methods for assessing health impact and opportunities to learn how to write contracts governing rewards based on health impact. Each pilot will have the desirable property that its costs are straightforwardly related to its health impact. Several promising pilot possibilities emerged from a workshop held in May 2011 at the Rockefeller Foundation’s conference center in Bellagio with experts from Canada, China, Colombia, India, Mexico, South Africa, the UK, the US, and Vietnam.
The current international system for encouraging pharmaceutical innovation is highly inefficient because its rewards are only very tenuously related to health gains. This system is unsustainable — humanity is already spending nearly one trillion dollars annually on pharmaceuticals, and these expenditures are rising faster than global income. Even the wealthiest countries cannot afford skyrocketing health care costs forever. The HIF is a concrete proposal that would not overturn the current system, but would address important inefficiencies so as to realize large net gains. Were it to work as expected, the medicines it supports would bring enormous health gains, especially in the world’s impoverished areas, even while its net costs would be negligible or even negative. Funding the HIF, taxpayers around the world would save through reduced expenses on public health facilities, foreign aid, insurance premiums and private drug purchases. We would save on costly hospitalizations averted by timely pharmacological interventions. And we would benefit from the diffuse economic effects of a massive reduction in the global burden of disease, which would also drive up the tax revenues of governments. Last and foremost, we would have taken an important step toward global justice by reducing the artificial exclusion of poor people from the fruits of pharmaceutical R&D.
Can there really be a reform that benefits all? Yes, if the predecessor regime is sufficiently irrational. This is true of the current scheme of incentives. Pharmaceutical companies make huge profits from developing medicines that barely enhance our pharmaceutical arsenal and from promoting sales that bring no meaningful health gains. Billions of human beings are prevented from buying generic versions of new medicines even though no one profits from this exclusion. The diseases concentrated among the poor are systematically neglected. At least the last two irrationalities are also grave injustices. Given the desperate need for affordable new medicines in many parts of the world — the US very much included — the HIF deserves serious examination. It would cost only $6 billion annually as against the $560 billion now spent worldwide on patented medicines; and a pilot would cost only a few million dollars while realizing savings and health gains for patients. You can help by critically examining the full HIF proposal (at healthimpactfund.org) and by supporting the HIF team’s efforts to launch the first pilots.