The Luxturna case — the first in vivo gene therapy approved by the US Food and Drug Administration for an inherited disorder — has become emblematic in pricing debates. A careful and detailed study by Knowledge Ecology International (KEI) shows that governments and charities financed most of the preclinical and early clinical work, while Spark Therapeutics entered late, consolidated exclusive rights, and launched at $850,000 per bilateral treatment.Reference Love 1 A natural reaction, reading the KEI account, is to conclude that those responsible for the key scientific breakthroughs were rewarded far less than downstream actors who assumed commercialization and regulatory risk.
The reaction is natural but not necessarily wise. The Financial Times once broke down the cost of a $3 cup of coffee: only 12 cents reflected the beans. The farmer received barely more than a penny — far less than was paid to the cashier.Reference Bruce-Lockhart and Terazono 2 Yet every participant in the worldwide division of labor that produced the coffee chose to participate and benefited from the process. A price is not a moral judgment, but a signal wrapped in an incentive.Reference Cowen and Tabarrok 3 Our intuition about which stages of a complex production process deserve the most reward does not necessarily match the economic reality of which stages must be most incentivized. Forcing complex processes to match simple intuitions often leads to unintended and undesirable consequences.
Transfers vs. Social Costs
The observed gap between Luxturna’s relatively modest cash outlays and its high price is not itself evidence of monopoly or social waste. In the presence of health insurance, the high price is primarily a transfer from patients and insurers to innovators and investors. The real social costs are the resources used in research, development, manufacturing, and delivery. Lowering the price changes the distribution of gains and losses but does not necessarily reduce social costs. If we reduce the price to patients, we should do so in a way that does not attenuate incentives to innovate. Indeed, public and private health insurance can reduce the deadweight loss from high prices while maintaining incentives to innovate.Reference Lakdawalla and Sood 4 Where insurance is incomplete or absent, high prices can exclude patients and create substantial deadweight loss; the relevant policy problem is to close those coverage gaps without destroying the expected returns that finance innovation.
Reducing prices without replacing lost expected returns is likely to drive investment below the social optimum. Given the large social returns to medical innovation,Reference Philipson and Jena 5 overcorrecting on transfers risks a sizable dynamic-efficiency loss.
Why “Price–Cost” is the Wrong Test in Pharmaceuticals
The standard monopoly screen — price relative to marginal or average cost — is informative in predictable, low-risk industries such as bricks or steel but is misleading when applied product by product in hit-driven industries such as movies, oil and gas exploration, and pharmaceuticals. In any industry where most products fail, product level excess returns are a necessary feature, not an anomaly. If nine projects fail after consuming substantial resources, the tenth must earn far more than its own direct cost for the portfolio to break even. These are excess returns relative to the cost of the hit, not necessarily excess risk-adjusted returns at the portfolio level. Movies exhibit the same pattern: a handful of blockbusters finance many flops, and the revenue of a hit is many times its production cost.
Like movies, pharmaceuticals are hit-driven industries: most projects fail, and a thin right tail funds the portfolio in expectation. The relevant test is therefore not the revenue of a hit relative to its own direct costs, but whether risk-adjusted expected returns across the entire portfolio are sufficient to generate socially desirable research and development (R&D). This conclusion does not depend on accepting pharmaceutical companies’ reported R&D costs. Even perfect accounting for Luxturna’s direct expenditures would not reveal the costs of failed projects by other firms or the ex ante return on the market portfolio. Thus, the key question is not whether Luxturna made excess profits but whether we currently over- or underinvest in pharmaceutical innovation. The evidence discussed below suggests that pharmaceutical innovation is underprovided relative to the social optimum. We need more hits to drive more investment. Before turning to the macro question, however, there are further specifics which make the Luxturna case of special interest.
Cures vs. Chronic Care: Pricing the Time Profile
Most drugs monetize value through repeat use (statins, biologics, insulin). Luxturna is a one-time intervention and as such, the $850,000 price invites sticker shock. But amortized over 40 years at a 5 percent discount rate, it is equivalent to roughly $50,000 per year — comparable to many biologics and far below ultrarare enzyme replacement therapies. On this basis, the “sticker shock” is a financing problem more than a value problem: cures compress decades of value into one payment. If we want firms to prioritize curative R&D, with its high technical risk and uncertain durability, over profitable maintenance drugs, we must either tolerate high upfront prices or replace them with alternative reward mechanisms.
In addition to sticker shock, a high upfront price may incentivize insurers to delay treatment in the hope of passing the large cost to the patient’s next insurer. Amortization, discussed below, may help with both problems.
The Drug Targeting Dilemma
Luxturna also illustrates the dilemma of drug targeting. Many drugs work on average and yet benefit only a minority of patients. Drugs are more valuable when we can target them on the patients who are most likely to benefit or least likely to be harmed. Genetic/biomarker targeting, for example, can improve welfare by avoiding adverse reactions and non-response. The dilemma is that targeting shrinks the market (as measured by the number of patients).
Suppose that, without a diagnostic, a drug is prescribed to 100 patients, each of whom has a 20 percent chance of benefiting. The drug can still pass a cost-benefit test for all 100 because the expected benefit exceeds the cost. Revenue is p × 100. Now suppose a diagnostic identifies in advance the 20 patients who will benefit. This is a significant improvement in health care, but it shrinks the treated market by 80 percent. Holding the required total reward fixed, the price per treated patient must rise by a factor of five: 5p × 20 = p × 100. This is the core economic logic of precision medicine and helps explain why targeted therapies often carry high per patient prices. Luxturna serves a tiny RPE65 mutation population, so a mass-market price model cannot finance investment. The same logic appears in other targeted medicines, such as human epidermal growth factor receptor 2 (HER2)-positive Herceptin, and in therapies such as Sovaldi that rapidly shrink their own future market.
If expected rewards do not rise as targeting shrinks the treated population, firms will underinvest in drugs and diagnostics that reduce volume. We want firms to design targeted drugs, but we must reward them at least as well as less effective drugs that can be sold to larger populations. This is not merely theoretical: the responsiveness of pharmaceutical innovation to market size is well established. Using demographic shifts across therapeutic categories, Acemoglu and Linn estimate that a 1 percent increase in potential market size raises the entry of new nongeneric drugs by approximately 5 percent.Reference Acemoglu and Linn 6 Finkelstein finds that policies increasing demand for particular vaccines produced a roughly 2.5-fold increase in new clinical trials for the affected diseases.Reference Finkelstein 7 Similarly, Blume-Kohout and Sood find that Medicare Part D significantly increased preclinical and clinical development in drug classes with greater exposure to the Medicare market.Reference Blume-Kohout and Sood 8 Together, these results show that expected revenues materially affect both investment in drug development and the arrival of new products.
Evidence Points to Underinvestment
Several independent lines of evidence point to underinvestment in pharmaceutical R&D. First, innovators capture only a fraction of the social value they create. Philipson and Jena estimate that producers captured only about 5 percent of the nearly $1.4 trillion in social surplus generated by HIV/AIDS therapies.Reference Philipson and Jena 9 Consistent with this disease-specific result, Nordhaus estimates that innovators captured only about 2.2 percent of the total surplus generated by innovation in the US nonfarm business sector from 1948 to 2001.Reference Nordhaus 10 More broadly, Jones and Summers conclude that social returns to innovation substantially exceed private returns.Reference Jones and Summers 11 The wedge arises because firms cannot capture all of the benefits of their products such as consumer surplus, knowledge spillovers, or benefits from positive externalities such as reduced disease transmission and increased productivity.
Second, the market size evidence shows that private rewards affect the amount and direction of research. Small patient populations and policies that reduce expected revenues reduce pharmaceutical innovation. Insurance can therefore have a dynamic benefit as well as a static one: by expanding effective demand, it can support innovation while reducing access losses.Reference Lakdawalla and Sood 12
Third, specific institutional and financial distortions push private research away from socially valuable projects. Budish, Roin, and Williams show that private cancer research is distorted away from long horizon projects, including cancer prevention and early stage treatment, because long commercialization lags shorten the effective patent reward; they estimate that the life years at stake are large.Reference Budish, Roin and Williams 13 Krieger, Li, and Papanikolaou find that risk aversion and financing constraints lead even large pharmaceutical firms to underinvest in radical molecular novelty.Reference Krieger, Li and Papanikolaou 14 These studies identify concrete mechanisms — not merely a generic theoretical possibility — through which private pharmaceutical investment falls below the social optimum.
These distortions matter because the potential social gains are enormous. Evidence from several distinct designs indicates that pharmaceuticals have produced substantial improvements in health. Duggan and Evans find that the introduction and rapid diffusion of new antiretroviral therapies sharply reduced HIV mortality.Reference Duggan and Evans 15 Buxbaum et al. attribute 35 percent of the increase in US life expectancy from 1990 to 2015 to pharmaceuticals.Reference Buxbaum 16 Cross-country evidence also links the availability of new cancer treatments to substantial reductions in cancer mortality, consistent with Lichtenberg’s broader disease level estimates.Reference Dubois and Kyle 17 Overall, Cutler et al. find that increases in medical spending have provided reasonable value and Murphy and Topel estimate that the value of annual gains to health are on the same order as the gains to wealth, i.e. on the order of trillions per year.Reference Cutler, Rosen and Vijan 18
As one example, Murphy and Topel calculate that a 1 percent reduction in cancer mortality would be worth nearly $500 billion, and a cure roughly $50 trillion. Those figures are in 2004 dollars aggregated over the year 2000 population; updating for prices, population, and real income growth puts the same calculations at something on the order of $1.5 trillion and $150 trillion today.Reference Murphy and Topel 19
Policy: Preserving Expected Returns While Lowering Deadweight Loss
My aim is not to defend any particular list price or institutional status quo but to highlight the tradeoffs. The challenge is to design an expected-return structure that sustains socially optimal R&D while reducing deadweight loss and barriers to adoption. Achieving this may require new institutional arrangements. Targeted prizes, patent buyouts, advance market commitments, and Operation Warp Speed-style subsidies, for example, have been proposed and, in some cases, successfully used to accelerate pharmaceutical development and access when social returns exceed private returns.Reference Castillo 20 Replacing monopoly-supported returns with more efficient pull incentives can improve both access and innovation. The danger is not lower prices per se, but lowering expected rewards without replacing them.
For the two specific problems highlighted here — the time profile of cures and the volume effects of drug targeting — payment design offers additional tools. One option is to amortize onetime payments into annuities. Instead of an $850,000 payment at year 0, an insurer could pay roughly $110,000 annually for ten years at a 5 percent discount rate, with payments conditional on continued efficacy. Portability would require a clearinghouse, reinsurance arrangement, or new forms of time-consistent insurance so that remaining obligations follow the patient when coverage changes or lapses.Reference Dabbous 21 This transforms a cure from a large upfront sunk cost into a flow of payments, easing sticker shock and reducing incentives for coverage. The innovator can still receive the agreed risk-adjusted net present value while payers smooth costs and patients gain access sooner.
Another mechanism is to bundle diagnostics with drugs through a payer level contract. Recall the earlier example: without a diagnostic, all 100 patients are treated; with a diagnostic, only the 20 responders receive the drug. A two-part contract could charge the insurer a modest fixed amount for each eligible or tested patient and a lower marginal price for each identified responder. The diagnostic itself should be broadly covered with little or no patient cost sharing so that the payment arrangement does not discourage testing. In essence, the contract combines the diagnostic with insurance: payments collected across the full at-risk population finance treatment for the smaller group that tests positive. This preserves the broader risk pool and the social value of targeting while allowing the firm to maintain the portfolio level reward needed to support innovation.Reference Tabarrok 22
Taken together, these strategies address different risks in financing innovation. Annuities mitigate the time problem of front-loaded costs for long-lived benefits, while diagnostic insurance contracts mitigate the volume problem created by targeting therapies to ever-smaller responder groups. Both mechanisms smooth payments across a broader base, reduce adoption frictions, and preserve expected returns consistent with continued R&D investment.
Beyond innovative payment models, a complementary strategy is to lower the denominator in the R&D cost-benefit equation. Reducing the fixed costs of development directly lowers the price needed to ensure a positive expected return. Thus, reducing clinical trial costs and regulatory burdens is a central lever for improving the economics of targeted and curative therapies.Reference Chertman and Teslo 23 Traditional trial designs were built for mass-market drugs, requiring large patient populations, long follow-up periods, and rigid protocols. For rare diseases and highly targeted interventions, these requirements can raise fixed costs to prohibitive levels.Reference Huber 24 Several reforms are possible. Biomarker-enriched and adaptive trial designs can reduce sample sizes by focusing on likely responders and allowing interim analyses to modify or halt trials early. Platform trials and shared control arms can further economize on scarce patient populations. Greater reliance on real-world evidence and registries can supplement long-term randomized follow-up, particularly when assessing the durability of onetime therapies.Reference Madden 25 Finally, reducing duplicative international regulatory processes and harmonizing approval standards can eliminate costly redundancies. Each step lowers the breakeven revenue firms must recoup, easing access barriers while preserving incentives to innovate.
Revisiting the Public Funding Question
Finally, we can return to the KEI study of Luxturna.Reference Love 26 KEI correctly draws attention to the foundational investments made by public and charitable funders. Such funding de-risks early research, which is especially difficult to incentivize and appropriate. It lowers the private cost of a particular project and may justify contractual claims on later returns. But it does not follow that downstream commercialization requires no reward. Late stage trials, manufacturing scale-up, regulatory approval, distribution, and reimbursement require substantial capital and specialized capabilities, and many apparently promising projects still fail before reaching patients. The relevant incentive question is therefore not whether Spark’s identifiable expenditures were small relative to Luxturna’s revenue, but whether expected rewards across the broader portfolio are sufficient to induce firms to build on publicly funded discoveries. The profitability of Luxturna is a market-wide signal that such commercialization can pay. Nothing in this argument establishes that $850,000 was the uniquely efficient price; it establishes that comparing that price with the direct cost of this successful product cannot by itself show that the reward was excessive. Foundational research is only foundational if someone builds upon it.
Conclusion
Luxturna highlights the core economics of precision medicine. In a hit-driven industry, product level excess returns are necessary because a small number of successes must finance many failures. Cures compress decades of value into one payment, while targeting raises per-patient value but shrinks markets. If policy focuses narrowly on direct project costs and squeezes prices without replacing expected returns, dynamic efficiency suffers. The right metric is not the winner’s price relative to its direct costs, but whether portfolio-wide expected returns are high enough to sustain socially optimal investment given smaller markets and high technical risk. The policy playbook is clear: use insurance to reduce access losses; risk-sharing payment designs such as amortization and diagnostic insurance contracts to solve time and volume problems; regulatory reform to reduce fixed costs; and, where appropriate, delinked rewards such as prizes and patent buyouts. That is how we preserve innovation incentives while expanding access.
Disclosures
The author has nothing to disclose.
Open access
