Innovation Health Economics

Why this matters in health economics

Almost every gain in health economics that a director celebrates — a vaccine that empties a ward, a diagnostic that catches disease early, a drug that turns a fatal condition into a managed one — began as a risky, loss-making bet by someone who did not know it would work. The economics of that bet, not just the economics of the finished product, determine which diseases attract research and which are left untouched. A condition common among people who cannot pay attracts little private investment however great the suffering; a marginal improvement for a wealthy market attracts a great deal. If you care about where health improves and where it stalls, you have to care about the incentives that steer research long before any product reaches a formulary.

The stakes are public money at both ends. Governments fund much of the basic science that private firms later build on, and then public payers buy the resulting products at prices set high partly to reward the original risk. A system that pays twice — once to create the knowledge and again to access it — should at least understand what it is buying and why. Getting the incentive wrong is expensive in a way that never shows up as a line item: the drug that was never developed, the tool adopted a decade too late, the money poured into a marginal "me-too" product because that is where the reward happened to point.

This chapter owns the economics of research and development, the mechanisms that pay for it — patents, prizes, advance market commitments — the way new technology spreads once it exists, and the emerging discipline of assessing value early, before the evidence is complete. It deliberately does not re-litigate how a launched medicine is priced and reimbursed, which belongs to Chapter 2.4 — Pharmacoeconomics, nor the global-access and intellectual-property-treaty questions owned by Chapter 4.2 — Global Health and Trade. Its subject is the stage before those: why innovation under-supplies itself, and how societies can pay to fix that without paying more than the fix is worth.

Core concepts

Research and development (R&D) is the production of new knowledge and its translation into usable products and processes. It runs on a spectrum from basic research — curiosity-driven science with no specific application in view — through applied research to development, testing, and scale-up. The economic problem is that the two ends behave very differently. Basic science is slow, uncertain, and generates knowledge that anyone can use once it exists; development is expensive, closer to market, and easier to keep to oneself.

The root difficulty is that knowledge is close to a public good in the economic sense: it is non-rival (my using an idea does not stop you using it) and hard to exclude others from (once a molecule's structure is known, copying is cheap). This is the innovation face of the market failures catalogued in Chapter 1.3 — Market Failure. A firm that cannot capture the value of what it discovers will under-invest, because much of the benefit — the positive externality — spills to imitators, patients, and future researchers. Left to an unaided market, socially valuable research is systematically under-produced. Every incentive mechanism in this chapter is a response to that single fact.

The economist's frame for the whole process is creative destruction, Joseph Schumpeter's term for the way innovation continually replaces established products and methods with better ones — and displaces the incumbents who depended on the old way. Health care experiences this unevenly: some innovations substitute for expensive care and save money, while many add cost by doing more, better, for more people. "Innovative" and "cost-saving" are not synonyms, and confusing them is a recurring error.

Incentives to innovate divide into two families. Push mechanisms pay for inputs before you know whether the research will succeed — grants, subsidies, tax credits, publicly funded laboratories. They lower the cost and risk of trying, and suit early, uncertain science, but they pay whether or not anything useful results. Pull mechanisms pay for outputs, rewarding success after the fact — the promise of a market. The classic pull instrument is the patent: a time-limited monopoly that lets the inventor charge above cost and recoup the investment. Patents solve the appropriability problem but create a new one — during the patent term, price sits above marginal cost, so some people who would benefit go without. This is the central trade-off of innovation policy: the dynamic efficiency of rewarding invention against the static efficiency of cheap access to what already exists.

Because patents purchase incentive at the price of access, economists have long studied alternatives. An inducement prize contest pays a fixed reward for achieving a defined goal and, in its pure form, places the resulting knowledge in the public domain — decoupling the reward from the price of the product. An advance market commitment (AMC) is a pull mechanism tailored to health: a binding promise by funders to subsidize the purchase of a product that does not yet exist, at a pre-agreed price and quality, if a firm develops it. The idea was developed prominently by the economist Michael Kremer and colleagues to draw private R&D towards diseases of poorer countries, where the expected market is otherwise too small and uncertain to justify investment. The first large-scale test targeted a pneumococcal conjugate vaccine suited to strains prevalent in low-income countries, funded by donors and delivered through Gavi, the Vaccine Alliance. Related instruments include the orphan drug designations that many jurisdictions use to reward research into rare diseases, and priority-review vouchers that reward developers with faster regulatory review of a future product.

Creating a technology is only half the story; the other half is diffusion of innovations — how, and how fast, a proven advance actually reaches the patients who could benefit. Adoption typically follows an S-shaped curve, from a few early adopters through a majority to laggards, and in health care the curve is often slow and uneven: good evidence does not spread itself. The economic cost of slow diffusion — years of avoidable illness while a proven tool sits unused — can rival the cost of the failed innovations that get more attention.

Finally, because health systems increasingly want to shape innovation rather than merely react to finished products, they use early or forward-looking health technology assessment (HTA): applying the valuation tools of Chapter 2.1 — Economic Evaluation to technologies still in development. Two ideas do most of the work. Headroom analysis asks, before much is spent, what a new technology would have to achieve and cost to be worth adopting at all — a cheap way to kill hopeless projects early. And the value of information framework treats evidence itself as something with a price and a payoff: it estimates how much a decision would improve if uncertainty were reduced, telling a funder whether it is worth paying for another trial or better to decide now. The related logic of options — keeping a costly decision open until uncertainty resolves — underpins arrangements such as coverage with evidence development, where a payer funds a promising technology conditionally while the evidence matures.

A growing share of innovation is not a single product but a pair. Precision medicine — also called stratified or personalized medicine — uses biological information, often genomic, to target a therapy at the subgroup of patients likely to benefit rather than treating an average patient. This typically couples a treatment with a companion diagnostic (a test that identifies who should receive the drug), and the two are co-dependent technologies: neither delivers its value alone, and the cost-effectiveness of the drug depends entirely on the accuracy of the test that selects for it. Appraising them separately — running the drug through the standard product pathway of Chapter 2.4 — Pharmacoeconomics while treating the test as a minor line item — misstates the value of both, because the health gain belongs to the pair, not to either half.

Stratification also changes the economics of the market itself. Slicing a disease into biomarker-defined subgroups turns one large market into many small ones, each with orphan-like dynamics that echo the orphan-drug problem discussed above (and developed in Chapter 2.4 — Pharmacoeconomics): thinner trial evidence because each subgroup is small and slow to recruit, weaker statistical power, and higher per-patient prices justified by the tiny eligible population. The same logic that makes targeted therapy attractive — concentrating benefit on those who gain most and sparing the rest — simultaneously erodes the evidence base and pushes prices up, so precision can buy better-aimed treatment at the cost of shakier proof and orphan-like pricing.

Genomic information has an unusual economic character that cuts the other way. A sequence generated once is a durable, reusable asset: it can be re-analysed as knowledge grows, reused across unrelated conditions, and inform many decisions over a lifetime from a single up-front cost. That makes genomics partly a global public good — non-rival and valuable well beyond the patient and system that paid to create it, so its full worth is systematically understated by any appraisal that counts only the immediate decision (the cross-border and equity dimensions of shared genomic knowledge belong to Chapter 4.2 — Global Health and Trade). This durability compounds the core valuation difficulty of any diagnostic: a test produces no health directly; its value is indirect, flowing entirely through the treatment decisions it changes. Drug-style appraisal, which values a product by the health it delivers, handles that poorly — the value-of-information framing above, and the evaluation methods of Chapter 2.1 — Economic Evaluation, are the more natural fit, because a test is in essence a purchase of information whose payoff is a better downstream choice.

Best practices

1. Decide whether you are buying inputs or outputs before choosing an instrument. Push funding (grants, tax credits, public labs) lowers the cost of trying and suits early, high-uncertainty science where no one can yet specify the target; pull funding (patents, prizes, advance market commitments) rewards results and suits problems where the goal is definable but the market is missing or too risky. Most real programmes blend the two, but blending by accident wastes money — fund the science you cannot specify, and reward the outcome you can.

2. Name the market failure the incentive is meant to correct. "We should support innovation" is not a rationale. Is the problem that knowledge spills to imitators so no one invests (an appropriability failure fixed by patents or exclusivity)? That the paying market is too small however great the need (a demand failure fixed by an advance market commitment or a prize)? Or that a proven product is not being adopted (a diffusion failure fixed by procurement and implementation, not more research)? The right tool follows from the diagnosis.

3. Treat the patent bargain as a trade-off, not a gift. A patent buys the incentive to invent by tolerating above-cost prices and blocked access for a period. That bargain is worth striking when the invention would not otherwise appear, and a bad deal when it rewards a marginal change or blocks a product people urgently need. Match the strength and length of exclusivity to the size of the innovation, and remember that access questions during the patent term are real costs, not complaints — the pricing and reimbursement mechanics belong to Chapter 2.4 — Pharmacoeconomics.

4. Prefer pull mechanisms that reward the outcome you actually want, defined in advance. An advance market commitment or prize only works if the target — the disease, the efficacy, the price, the quality, the volume — is specified tightly enough that a firm can aim at it and a funder can verify a hit. Vague targets reward the wrong thing; targets set too high attract no entrants; targets set too low overpay for what would have happened anyway. The discipline of writing the specification is most of the work.

5. Use advance market commitments where need is high but the market signal is absent. When a product would save lives in populations that cannot pay — vaccines for diseases concentrated in low-income countries are the archetype — a credible, binding promise to subsidize purchase at a set price can manufacture the missing demand and pull private R&D towards it. The Kremer-inspired pneumococcal commitment showed the mechanism can accelerate supply of an adapted product; it also showed that design details — who guarantees the money, how price steps down, how quality is assured — decide whether it works (the global-access dimension is developed in Chapter 4.2 — Global Health and Trade).

6. Do early, forward-looking HTA to kill hopeless projects cheaply and steer promising ones. Applying economic evaluation while a technology is still in development — headroom analysis to find the maximum price and minimum benefit that could ever make it worthwhile — costs little and prevents large sums being spent developing things that could never be cost-effective. Done well, early HTA also tells a developer which evidence a payer will demand, so the pivotal trial answers the reimbursement question rather than only the regulatory one.

7. Price the evidence, not just the product. The value-of-information framework asks how much better a decision would be if a specific uncertainty were resolved, and compares that to the cost of resolving it. Sometimes the answer is to commission another study; often it is that further research would cost more than the improved decision is worth, and you should decide now. Either way, treating evidence as a purchasable asset stops both reflexive "we need more data" and reckless adoption on thin proof.

8. Keep expensive decisions open when uncertainty is high and resolving. For a costly, hard-to-reverse technology whose value is genuinely unclear, a conditional arrangement — coverage with evidence development, a time-limited managed access agreement, a staged roll-out — preserves the option to expand or exit as evidence matures. This is real-options thinking: you pay a little to avoid committing fully to a bet you cannot yet judge, provided the deal has real exit conditions and a date, not an open-ended subsidy.

9. Budget for diffusion as deliberately as for discovery. A proven innovation that no one adopts delivers no health, and adoption is not automatic: clinicians, procurement, training, pathways, and payment all have to move. Identify the early adopters, remove the practical barriers the majority face, and fund implementation as a distinct activity — the slow spread of proven advances is one of the largest, quietest losses in every health system.

10. Do not assume innovation saves money. Many valuable innovations raise total spending by doing more for more people, even when each unit is cost-effective. Distinguish substituting innovations that displace costlier care from additive ones that expand it, and be honest with your finance director about which you are adopting. Calling a cost-increasing advance a "saving" because it is efficient per patient is a common and expensive confusion.

11. Shape incentives for the diseases the market neglects, and count the whole cost. Where private returns are low but social returns are high — rare diseases, conditions of poverty, antimicrobials whose value lies in being held in reserve — targeted instruments (orphan designations, prizes, subscription-style "pull" payments) can redirect research. But each carries its own failure: exclusivity that is gamed, prizes that misjudge the target, subsidies captured by firms that would have acted anyway. Weigh the mechanism's cost and gaming risk against the research it actually induces, not the research it is named after.

12. Appraise a therapy and its companion diagnostic as one co-dependent decision. When a treatment is licensed with, or in practice tied to, a test that selects who receives it, the drug and the companion diagnostic form a single co-dependent technology whose value cannot be split cleanly between them. Evaluate the linked pair together — the test's accuracy, the treatment's effect in the test-positive group, and the consequences for the patients the test rules in or out — as one question, not two sequential appraisals. The test's own price is usually trivial beside the therapy it governs, yet a diagnostic that is slightly more or less accurate can swing the pair's cost-effectiveness dramatically, so build the appraisal around the pair from the outset and remember that genomic information generated for one decision may hold reusable value beyond it.

Questions to discuss with your team

1. Which health problems that matter to our population are currently starved of research because the market that would pay for a solution is too small or too poor? This question forces attention from the innovations arriving to the ones that never will. Look for conditions where suffering is high but the paying market is thin — rare diseases, conditions concentrated among people who cannot pay, or products like new antibiotics whose social value lies in restraint rather than volume. The honest discussion names two or three such gaps in your own setting and asks whether any lever available to you — a prize, a guaranteed purchase, a partnership with a funder or a low- and middle-income-country programme — could plausibly change the calculus. Distinguish problems that are genuinely un-researched from those where good science exists but nothing has been adopted, because the remedies differ entirely. An answer that ends with a real candidate for a pull mechanism, and a frank admission of what it would cost, is worth more than a survey of what is already in the pipeline.

2. When a promising but unproven technology arrives, how do we decide between adopting now, refusing, and paying to learn more? This is the value-of-information question made concrete, and most organizations answer it by instinct rather than analysis. The tension is real: adopt too early and you may pour money into something that does not work or bankrupts the budget; wait for certainty and patients lose years of benefit that never comes back. The middle path — conditional coverage while evidence matures — is powerful but only if the arrangement has genuine exit conditions and a date, not an open-ended subsidy that becomes permanent by inertia. An honest answer sets out who in your organization is allowed to say "we will fund this only while it is being studied," how the additional evidence would actually be gathered, and what result would make you stop. It should also confront the political reality that withdrawing a technology patients have started to receive is far harder than never starting.

3. Are our innovation incentives rewarding the health gains we want, or simply the products that are easiest to profit from? This question interrogates the direction of the incentive, not its size. Patents, exclusivity periods, and reimbursement rules reward whatever they happen to point at, and left unexamined they tend to favour marginal improvements for wealthy markets over larger gains for neglected ones, and additive technologies over the unglamorous substitutions that save money. Ask whether the things your system rewards most richly are the things that improve health most, and where the two have come apart. The honest answer usually reveals at least one place where the reward and the health gain are misaligned — a "me-too" product crowding a formulary, a slow-diffusing proven advance no one is paid to spread, a neglected condition no mechanism touches. What matters is naming one such gap and one lever you could pull, rather than concluding that incentives are simply "the market's job."

4. Where in our system is a proven advance failing to reach the patients who would benefit, and who is actually paid to fix that? This shifts attention from inventing technology to spreading it, the half of innovation economics that attracts the least money and the most avoidable harm. Diffusion is not automatic: a genuinely effective tool can sit unused for years while clinicians, procurement, training, pathways, and payment all fail to move together, and the cost of that delay — avoidable illness while a working answer exists — can rival the cost of the failures that get the headlines. The honest discussion identifies one or two proven advances in your own setting that are diffusing slowly, and asks whether the obstacle is evidence, habit, money, or the plain fact that implementation is nobody's job. It should separate a genuine knowledge gap from an uptake gap, because commissioning more research when the failure is adoption wastes money and time. A good answer names who owns diffusion as a distinct, funded activity — with early adopters identified and time-to-uptake measured — rather than assuming good evidence will spread itself.

5. When a therapy comes tied to a test that selects who receives it, do we appraise the pair as one decision or two? Precision medicine increasingly couples a treatment with a companion diagnostic, and the two are co-dependent: neither delivers its value alone, and the cost-effectiveness of the drug depends entirely on how accurately the test picks the patients who benefit. The tension is that our appraisal machinery is usually built to value products one at a time — running the drug through the standard pathway while treating the test as a minor line item — which misstates the value of both, because the health gain belongs to the pair. The honest answer confronts several awkward facts at once: a slightly more or less accurate test can swing the pair's cost-effectiveness dramatically; slicing a disease into biomarker-defined subgroups creates thin, orphan-like evidence and higher per-patient prices; and a genomic result generated once may hold reusable value well beyond the immediate decision. Ask whether your current process could even represent a co-dependent technology, or whether it would quietly appraise the halves in isolation. A workshop-ready answer points to one linked drug-and-diagnostic decision you face and how you would evaluate the two together (the pricing mechanics belong to Chapter 2.4 — Pharmacoeconomics).

6. Are we funding the same piece of research twice — subsidizing the inputs and then paying a premium for the output — and would we even notice? Push mechanisms (grants, tax credits, public laboratories) lower the cost of trying; pull mechanisms (patents, exclusivity, guaranteed purchase) reward the result. Most real programmes blend the two, but blending by accident means paying for the same risk from both ends: funding a firm's science and then rewarding its product as though the firm had borne the whole gamble. The honest discussion asks, for a concrete technology your organization supports, which stage each pound is actually buying, and whether the public purse is paying to create knowledge and again to access it without anyone having decided that on purpose. It should also confront the related self-deception of calling a cost-increasing advance a "saving" because it is efficient per patient, when it in fact expands spending by doing more for more people. A good answer distinguishes deliberate blending from accidental double-payment, and gives the finance function an honest account of both what is being funded and what it will cost.

In practice: a health economics example

The fictional high-income Kingdom of Norlant, a single-payer tax-funded system, faces a familiar frontier problem: resistance is eroding its last-line antibiotics, and no manufacturer will develop new ones. The economics are stark and are a textbook commons failure. A genuinely novel antibiotic would be held in reserve, used as little as possible to preserve its effectiveness — so the more valuable it is to society, the less it sells, and the worse the business case. Norlant's usual purchasing, which pays per pack dispensed, actively punishes the product it most needs. The Ministry of Health convenes an innovation-economics team to design a fix.

The team is clear that this is a demand failure, not a science failure, so a pull mechanism is called for rather than another research grant. They design a "subscription" payment that decouples what the manufacturer earns from how many packs are used: the Kingdom will pay a fixed annual sum for guaranteed access to a qualifying novel antibiotic, calculated from the estimated value the drug provides to the health system — its worth as insurance against a resistant outbreak — rather than from volume. This mirrors the logic of an advance market commitment turned towards a domestic reserve problem, and it borrows from real subscription-style pilots that high-income payers have trialled. The reward now points at availability and stewardship, the two things the Kingdom actually wants.

Before committing, the team runs a forward-looking HTA. Headroom analysis asks the blunt question: what is the most Norlant could rationally pay per year for a reserve antibiotic before the subscription costs more than the resistant infections it would avert? They build the estimate from the burden of resistant infection, the probability and cost of outbreaks, and the value of keeping a working drug in reserve, and it gives them a ceiling to negotiate against rather than a number a manufacturer hands them. A value-of-information step then asks whether to sign now or fund more real-world data first; because delay means continued exposure to untreatable infection, the analysis favours acting, but with a review built in.

The design is deliberately conditional. The subscription runs for a fixed term with defined stewardship obligations — the drug must be conserved, not pushed — and an evidence-development requirement so that resistance patterns and real value are tracked and the payment can be re-based at review. The team is candid about the residual risks they cannot design away: the value estimate is uncertain, a fixed payment could overpay if resistance evolves slowly or underpay if it accelerates, and a small-market scheme may not on its own be enough to pull global R&D — it works best as one of several countries doing the same. The lesson the Ministry takes is the chapter's core one: the market under-supplies exactly the innovation society needs most, the fix is to manufacture the missing demand deliberately, and the fix must be sized by evidence and kept open to revision. The routine pricing of medicines already on the market is a separate discipline, handled in Chapter 2.4 — Pharmacoeconomics.

Four sector lenses

Startup

For a health-innovation start-up, these mechanisms are the difference between a fundable idea and an abandoned one. Founders live on the push–pull boundary: grants, tax credits, and public research funding keep the lights on before revenue, while the eventual reward — patent, exclusivity, or a defined purchase commitment — is what investors are really betting on. The single most valuable early discipline is forward-looking HTA: understanding, before the pivotal trial is designed, what a payer would have to see to buy the product, so that scarce capital proves the reimbursement case and not merely the regulatory one. A start-up that optimizes for approval and neglects the evidence a payer needs can win a licence and still have nothing anyone will fund.

Small business

A small but established innovator — a specialist diagnostics supplier, a single clinic, or a medical-device maker with a steady product line — meets innovation economics mainly as a diffusion and adoption problem rather than a fundraising one. Its concern is less how to finance a moonshot than how to get a proven improvement taken up: navigating procurement, generating the modest real-world evidence a cautious payer wants, and being adopted before a larger competitor crowds the space. With limited capital and no appetite for a decade-long bet, such a firm favours incremental, substituting innovations that pay back quickly and slot into existing pathways, and it is acutely exposed when a patent it depends on expires or a reimbursement rule shifts. Its most useful discipline is to prove, cheaply and credibly, that its product either displaces a costlier alternative or measurably improves an outcome a purchaser already values.

Enterprise

A large manufacturer or provider organization experiences innovation economics as portfolio management. The firm allocates R&D across projects by expected risk-adjusted return, which is precisely why conditions with weak or poor markets get neglected — the incentive points elsewhere, not because anyone is callous but because that is what the numbers say. Enterprises are also the main users and shapers of the diffusion curve: a large provider decides how fast a proven advance spreads through its pathways, and a large manufacturer's pricing and evidence strategy determines who can adopt it. At this scale the recurring temptation is to defend returns on existing products — extending exclusivity, favouring incremental "me-too" work — rather than to fund the riskier, larger advances, and payers increasingly watch for the difference.

Government

Government is the actor that can correct innovation's market failure at its source, and it sits on both sides of the ledger: it funds much of the basic science and then pays for the products that science enables. Its toolkit is the widest — research funding, patent and exclusivity law, prizes, advance market commitments, orphan-disease incentives, and the HTA machinery that decides what to buy — and so is its responsibility to direct innovation towards social value rather than merely private return. A government can convene the demand that no single firm or country can, whether by guaranteeing a market for a neglected-disease vaccine or by pooling purchasing across nations. Its discipline must match its reach: every incentive it creates can be gamed or captured, and a mechanism that pays for research that would have happened anyway is public money spent for nothing.

Common failure modes

• Confusing innovation with saving. Treating every cost-effective advance as a budget reduction, when many valuable innovations add cost by doing more for more people. Fix: separate substituting innovations from additive ones and state the budget impact honestly.

• Funding more research when the failure is diffusion. Commissioning yet another study while a proven advance sits unadopted for want of training, procurement, or payment. Fix: diagnose whether the gap is knowledge or uptake, and fund implementation as a distinct activity.

• Vague pull targets. Launching a prize or advance market commitment without specifying the disease, efficacy, price, quality, and volume tightly enough to aim at or verify. Fix: invest in the specification; a loosely defined reward buys the wrong thing.

• Adopting on thin evidence with no exit. Funding a promising but unproven technology "temporarily" and letting it become permanent by inertia. Fix: build genuine review dates and stopping conditions into every conditional arrangement.

• Rewarding the marginal over the meaningful. Incentive structures that favour incremental "me-too" products and exclusivity extensions over larger, riskier gains and neglected diseases. Fix: examine what your rewards actually point at, and use targeted instruments where the market misdirects.

• Ignoring the access cost of exclusivity. Treating the high price and blocked access during a patent term as someone else's problem. Fix: count static access losses as a real cost of the innovation bargain, and match exclusivity to the size of the advance.

• Push and pull working against each other. Subsidizing a firm's inputs and then also paying a premium for its output, rewarding the same research twice. Fix: decide deliberately which stage you are funding and avoid paying for the same risk from both ends.

• Appraising a companion diagnostic in isolation from its drug. Evaluating a selection test on its own cost and accuracy, or the therapy as if every patient were eligible, when the two are co-dependent and only make sense together. Fix: appraise the drug and diagnostic as a single co-dependent technology, valuing the test through the treatment decisions it changes.

Maturity model

Checklist

• For the problem in front of you, name the specific failure — appropriability, missing demand, or slow diffusion — before choosing an incentive.
• Decide deliberately whether you are funding inputs (push) or rewarding outputs (pull), and avoid paying for the same research twice.
• For any prize or advance market commitment, specify the disease, efficacy, price, quality, and volume tightly enough to aim at and verify.
• Run a headroom analysis on a candidate technology early, to find the maximum price and minimum benefit that could ever make it worthwhile.
• Apply value-of-information reasoning before commissioning more evidence — is the improved decision worth the study's cost?
• For any conditional adoption, build in a genuine review date and stopping conditions, not an open-ended subsidy.
• State whether an innovation substitutes for or adds to existing costs, and give the finance function an honest budget impact.
• Fund diffusion and implementation of proven advances as a distinct activity, and measure time-to-uptake.
• Identify at least one market-neglected need in your population and one lever that could redirect research towards it.
• Count the access cost of any exclusivity you rely on, and match its strength to the size of the advance.

Key sources

• Michael Kremer & Rachel Glennerster, Strong Medicine: Creating Incentives for Pharmaceutical Research on Neglected Diseases — the foundational case for advance market commitments.
• Center for Global Development, Making Markets for Vaccines (advance market commitment working group report) — the design blueprint behind the pneumococcal AMC.
• Gavi, the Vaccine Alliance — pneumococcal advance market commitment — programme documentation on the first large-scale AMC.
• Everett Rogers, Diffusion of Innovations — the standard framework for how technologies spread through a population.
• ISPOR good-practice guidance on health technology assessment and early / forward-looking value assessment.
• National HTA bodies as exemplars of managed access and coverage-with-evidence-development schemes — e.g. NICE (England), CDA-AMC (Canada), PBAC (Australia), IQWiG (Germany).
• Australian Government Department of Health — guidelines for preparing assessment reports for co-dependent and hybrid technologies (PBAC/MSAC) — a leading framework for appraising a drug and its companion diagnostic together.
• EGAPP (Evaluation of Genomic Applications in Practice and Prevention), US Centers for Disease Control and Prevention — an evidence framework for assessing genomic tests.
• Economics Network, Health Economics for Teachers — https://economicsnetwork.ac.uk/health/teachers

References

1. Research and development — Wikipedia — https://en.wikipedia.org/wiki/Research_and_development
2. Public good (economics) — Wikipedia — https://en.wikipedia.org/wiki/Public_good_(economics)
3. Creative destruction — Wikipedia — https://en.wikipedia.org/wiki/Creative_destruction
4. Patent — Wikipedia — https://en.wikipedia.org/wiki/Patent
5. Inducement prize contest — Wikipedia — https://en.wikipedia.org/wiki/Inducement_prize_contest
6. Advance market commitments — Wikipedia — https://en.wikipedia.org/wiki/Advance_market_commitments
7. Michael Kremer — Wikipedia — https://en.wikipedia.org/wiki/Michael_Kremer
8. Pneumococcal conjugate vaccine — Wikipedia — https://en.wikipedia.org/wiki/Pneumococcal_conjugate_vaccine
9. Orphan drug — Wikipedia — https://en.wikipedia.org/wiki/Orphan_drug
10. Diffusion of innovations — Wikipedia — https://en.wikipedia.org/wiki/Diffusion_of_innovations
11. Health technology assessment — Wikipedia — https://en.wikipedia.org/wiki/Health_technology_assessment
12. Value of information — Wikipedia — https://en.wikipedia.org/wiki/Value_of_information
13. Michael Kremer & Rachel Glennerster, Strong Medicine: Creating Incentives for Pharmaceutical Research on Neglected Diseases — Princeton University Press, 2004.
14. Everett M. Rogers, Diffusion of Innovations — Free Press.
15. Economics Network — Health Economics for Teachers — https://economicsnetwork.ac.uk/health/teachers
16. Guidelines for preparing assessment reports for co-dependent and hybrid technologies — Australian Government Department of Health (PBAC/MSAC).
17. Evaluation of Genomic Applications in Practice and Prevention (EGAPP) — US Centers for Disease Control and Prevention.