we-solve (beta)

Econ-350 · economics · Economics

Social value and pricing of pharmaceutical innovation

LLM status Open unverified

Statement

Original atlas identifier: H1; entry 51. Field: Health economics. Original tractability label: Hard.

Classification: Parameterized theoretical/empirical research agenda; not a certified unsolved theorem.

Definitions and assumptions

Fix dates $t=0,\ldots,T$, indications $k=1,\ldots,K$, and countries $j=1,\ldots,J$. A pricing/reward policy $p$ specifies prices, reimbursement, patent duration or buyouts, and any subsidies. A dynamic model $m$ determines firms' indication-specific research $r_{kt}$, innovation arrival probabilities, product quality, entry, and access $x_{jkt}$. Let $Q_{jt}(p;m)$ be quality-adjusted life-years gained relative to a stated baseline, $v_j\geq0$ their monetary social value, and $L_{jt}(p;m)$ other real health/economic benefits or harms, valued consistently. Let $C_{jt}$ denote real treatment/production costs and $C_{Rt}$ research resource costs. Fix discount factor $\beta\in(0,1]$ and nonnegative country weights $w_j$. Specify supplier and payer equilibrium selection and account for all countries whose R&D or access responds.

Formal research question

Editorial formulation: the mathematical target below is not a theorem quoted from the references.

Characterize the feasible-policy welfare frontier \[ W_m(p)=\mathbb E_m\sum_{t=0}^{T}\beta^t\left\{ \sum_{j=1}^{J}w_j[v_jQ_{jt}+L_{jt}-C_{jt}]-C_{Rt}\right\}. \] All quantities on the right depend on $p$ and $m$. Prices are transfers, not additional resource costs; distributional gains/losses from transfers require explicit payer/producer welfare weights. Impose payer budgets, participation, commitment, safety/approval constraints, and a funding rule. Identify how indication-level revenue shocks affect research allocation and eventually clinical health gains, rather than only patent or approval counts. Compare price caps, reimbursement rules, and buyouts using $\varepsilon$-optimal policies over an identified model set, retaining uncertainty about innovation timing and quality.

Interpretation and scope

Static patient access and future innovation are distinct margins, and the same price reduction can improve the first while changing the second in either direction under portfolio substitution. Market-size shocks may affect scientific opportunity, disease prevalence, or competing indications; an exclusion restriction is needed before treating them as exogenous revenue variation. A buyout must specify valuation, licensing, and collusion controls rather than presume a known patent value. Country weights and the value of health are normative inputs; global innovation spillovers cannot be summarized by domestic consumer surplus. Costs should include failed research projects and displaced scientific capacity. The substantive extension connects causal revenue responses to eventual global health value and feasible contracting, while existing innovation-market-failure and patent-buyout models remain solved benchmarks.

Source audit and status

[S1]--[S3] are identifiable primary works: general innovation theory, a buyout mechanism, and pharmaceutical market-size evidence. They do not uniquely establish the atlas's broad question as an open theorem or verify its status in 2026. The objective above is an editorial social-accounting formulation.

References

    },leftmargin=*,itemsep=0.6em]
  1. Kenneth J. Arrow (1962). Economic Welfare and the Allocation of Resources for Invention. In The Rate and Direction of Inventive Activity: Economic and Social Factors, 609--626. Princeton University Press for NBER. https://www.nber.org/books-and-chapters/rate-and-direction-inventive-activity-economic-and-social-factors/economic-welfare-and-allocation-resources-invention
    Locator and support limits: Chapter: innovation-market failures; no pharmaceutical-specific dynamic price optimum.
  2. Michael Kremer (1998). Patent Buyouts: A Mechanism for Encouraging Innovation. The Quarterly Journal of Economics 113(4), 1137--1167. DOI: 10.1162/003355398555865. https://academic.oup.com/qje/article-abstract/113/4/1137/1916997
    Locator and support limits: Abstract and mechanism: patent-buyout design; effects depend on information and collusion assumptions.
  3. Daron Acemoglu and Joshua Linn (2004). Market Size in Innovation: Theory and Evidence from the Pharmaceutical Industry. The Quarterly Journal of Economics 119(3), 1049--1090. DOI: 10.1162/0033553041502144. https://academic.oup.com/qje/article-abstract/119/3/1049/1938820
    Locator and support limits: Abstract and article: market-size and pharmaceutical innovation; product counts do not by themselves measure health surplus.

Common conventions: Source mathematical conventions

Unless an entry states otherwise, agent and firm sets are finite. State and action spaces are measurable, strategies and outcome maps are measurable, probability laws are specified, and expectations exist for the targets under discussion. Infinite-horizon discount factors lie in $(0,1)$ and discounted objectives are finite. Norms, tolerances and complexity input sizes must be specified for computational comparisons. Constants or primitives that appear locally are inputs, not empirical facts asserted by this catalogue.

Identification. A statistical model $m\in\mathcal M$ implies an observable law $P_m$ and target $\theta(m)$. At law $P$, its identified set is \[ \Theta(P;\mathcal M)=\{\theta(m):m\in\mathcal M,\ P_m=P\}. \] Point identification holds when this set is a singleton, or equivalently when $P_{m_1}=P_{m_2}$ implies $\theta(m_1)=\theta(m_2)$ for all compatible models. Sharp bounds mean bounds attained, or approached when the boundary is not attained, by compatible models. Writing this set is a definition, not a solution: the substantive task is to establish informative restrictions and derive or estimate the set. An unrestricted-confounding model may give only trivial bounds.

Inference. Identification, estimability and valid uncertainty quantification are separate questions. Fix $\alpha\in(0,1)$. A confidence procedure $C_n$ over a declared law class $\mathcal P$ has asymptotic uniform coverage at level $1-\alpha$ when \[ \liminf_{n\to\infty}\inf_{P\in\mathcal P}\Pr_P\{\theta(P)\in C_n\}\geq1-\alpha. \] For set-identified targets the coverage event must be defined separately, for example coverage of the full identified set. If an entry uses identification-indexed models instead of uniquely indexed laws, the same coverage convention is applied over those models. Pointwise limits do not establish uniform validity, and asymptotic validity does not guarantee finite-sample precision. Sample sizes, dependence structures and weak-identification sequences are part of each application.

Counterfactuals. $Y_i(a)$ is a potential outcome under a specified intervention, including its timing, implementation and versions. It is not $Y_i$ conditional on voluntarily choosing $a$. A full intervention vector permits interference; shorthand scalar treatment notation does not silently impose no interference. Assignment, selection, attrition, anticipation and measurement must be specified. A claim about an instrument's local effect is not automatically a population policy effect.

Economic equilibrium. A counterfactual model includes best responses, constraints, feasibility, market clearing, state transitions and expectations consistent with the stated information. When several equilibria exist, either a declared selector is an input or the target is set-valued. Comparisons across policy regimes must use the stated selection convention. Reduced-form relationships are not presumed invariant after an equilibrium-changing intervention.

Optimization and welfare. Optimization is over the stated feasible set. If attainment is not established, $\sup$ or $\inf$ and $\varepsilon$-optimal choices replace an asserted maximizer or minimizer. For example, with value $V=\sup_{a\in\mathcal A}W(a)<\infty$, an $\varepsilon$-optimal set is $\{a\in\mathcal A:W(a)\geq V-\varepsilon\}$ for $\varepsilon>0$. Benefits, costs, risk and health or environmental outcomes can be aggregated only using declared units and conversion weights. Interpersonal utility weights, the treatment of fiscal transfers and foreign profits, discounting, and the behavioral welfare criterion are explicit inputs. A comparative-static derivative additionally requires existence and appropriate differentiability of the selected counterfactual.

Sources. Local labels [S1], [S2], and so forth restart in every question. Locators identify the relevant abstract, model, section or result at the level actually checked; a bibliographic or abstract-level verification is not represented as inspection of an entire proof. Working papers are distinguished from later publications and changing versions. Source summaries are paraphrased. All URL checks and editorial formulations refer to the audit date stated above.

Statement licence: CC-BY-4.0 · Source file

Contributions 1

Econ-350-A1 · Exploration · LLM-generated · gpt 6 astra ultra ·

GPT 6 Astra Ultra attempt on Econ-350

Social value and pricing of pharmaceutical innovation: a research attempt

Target and outcome

Status: unresolved. The problem asks how pharmaceutical pricing and innovation policy affect health, access and research, with country weights, resources and financing stated explicitly. Drug counts alone are not the welfare target. This attempt solves a two-stage research-allocation benchmark, characterizes a conditional reward implementation and gives two counterexamples to count-based or marginal-data welfare inference. It does not identify the social value of real pharmaceutical policies.

Acemoglu and Linn's primary working paper [al] studies market size and pharmaceutical innovation. It provides background for the role of private rewards. It is not used to claim that the health values, discovery probabilities or policy responses in this calculation have been empirically identified.

Research, success and the resource constraint

There are finitely many indications $k$. Research $r_k\geq0$ has success probability \[ p_k(r_k)=1-e^{-a_kr_k},\qquad a_k>0. \] A successful innovation yields a known social present value $G_k>0$, measured in a common money numeraire using declared country weights and health valuations, net of treatment and production resources. Research itself costs $c_kr_k$, $c_k>0$, and total research effort is constrained by $\sum_kr_k\leq R$, $R>0$. Thus \[ W(r)=\sum_kG_k(1-e^{-a_kr_k})-\sum_kc_kr_k. \] Success dependence across indications does not affect this expectation if their benefits are additive; interactions would require a different objective. The finite horizon, access path and terminal convention are already included in each $G_k$.

Private developers instead receive expected success revenue $A_kp_k(r_k)$, with the same resource cost and capacity constraint. For the reward calculation assume $0\leq A_k\leq G_k$. This inequality is a restriction, not a general empirical law: monopoly rents and external effects can violate it.

Proposition (Exact allocation): The social optimum is unique and has \[ r_k^S=\frac1{a_k}\left[\log\frac{a_kG_k}{c_k+\lambda}\right]_+, \qquad [x]_+=\max(x,0), \] where $\lambda\geq0$ satisfies the capacity constraint and $\lambda(\sum_kr_k^S-R)=0$.
Proof: The feasible simplex is nonempty and compact. The objective is continuously differentiable with diagonal Hessian entries $-G_ka_k^2e^{-a_kr_k}<0$, so it is strictly concave. Existence and uniqueness follow. The Kuhn--Tucker conditions give $a_kG_ke^{-a_kr_k}=c_k+\lambda$ at a positive coordinate and $a_kG_k\leq c_k+\lambda$ at zero, yielding the formula. The sum of the formula's coordinates is continuous and nonincreasing in $\lambda$, and strictly decreasing whenever it is positive. If it exceeds $R$ at zero, there is a unique positive multiplier making it equal $R$; otherwise $\lambda=0$ works. These conditions are sufficient by concavity. ∎

The private allocation follows the same formula with $G_k$ replaced by $A_k$, allowing zero coordinates when $A_k=0$. A shift in the expected revenue of one indication can change every other indication's research through the capacity multiplier. It is therefore insufficient to apply isolated own-market elasticities without checking whether shared research resources bind.

A conditional policy implementation

A verifiable per-success reward $z_k=G_k-A_k$ makes the private objective exactly equal to $W(r)$, so implements the social optimum under the stated assumptions. The expected public payment is \[ B_{\rm prize}=\sum_kz_k(1-e^{-a_kr_k^S}). \] If the available prize budget is below this quantity, the proposed implementation is infeasible. Commitment and success verification are also necessary: an unenforceable promise is not the private reward assumed in the optimization.

The principal payment is a transfer between taxpayers and developers inside an equal-weight social boundary. It must not be subtracted from welfare a second time after ownership and taxes are included. Administrative resources and excess burden of financing are different. For example, a real financing loss of $\theta$ per prize dollar changes the social objective to \[ \sum_kp_k(r_k)G_k-\sum_kc_kr_k -\theta\sum_kp_k(r_k)z_k. \] The simple reward $G_k-A_k$ no longer automatically implements the optimum of this modified objective. Nor does a domestic government's objective necessarily coincide with a global health objective when benefits and tax payments cross the chosen boundary.

Equal innovation counts can hide different health value

Take two indications, $a_1=a_2=1$, $G_1=10$, $G_2=1$, equal linear research costs $c_1=c_2=0.1$ and binding capacity $R=1$. Compare allocations $(0.8,0.2)$ and $(0.2,0.8)$. They have the same expected number of successes, \[ 2-e^{-0.8}-e^{-0.2}, \] and the same research resource cost. Yet the first allocation has greater expected social value by \[ 9(e^{-0.2}-e^{-0.8})\simeq3.3246. \] Both allocations can be private optima: choose $(A_1,A_2)=(e^{0.6},1)$ for the first and $(e^{-0.6},1)$ for the second. Both obey $A_k\leq G_k$. The common marginal private success returns are respectively $e^{-0.2}$ and $e^{-0.8}$, exceeding the common marginal resource cost $0.1$. Equality of the two net marginal returns and a positive capacity multiplier verify each allocation.

Thus a policy that changes the composition of research can have a substantial welfare effect while leaving expected innovation counts exactly unchanged. The example does not say innovation counts contain no information; it shows that they do not identify welfare without indication-specific values and policy-induced composition.

Timing and quality require their joint law

Even observing the marginal distribution of launch dates and quality can be insufficient. Let quality $Q$ be zero or two, each with probability one half, and let launch time $T$ be zero or one, also each with probability one half. In model A, high quality launches immediately and zero quality launches later. In model B, the pairing is reversed. Both models have identical innovation counts and identical marginal laws of $Q$ and $T$. Nevertheless, \[ {\mathbb{E}}_A[\beta^TQ]=1,\qquad {\mathbb{E}}_B[\beta^TQ]=\beta, \] which differ whenever $\beta<1$. The relevant joint timing-quality information is missing. Replacing an expectation of their product by the product of marginal expectations would silently impose independence.

If only $0\leq G_k\leq\overline G_k$ is known while research and success probabilities are identified, a basic sharp value interval under otherwise unrestricted values is \[ -\sum_kc_kr_k\ \leq W(r)\leq \sum_k\overline G_kp_k(r_k)-\sum_kc_kr_k. \] Endpoints are attained by choosing all values at zero or their upper limits. Sharper policy comparisons require restrictions coupling the baseline and policy health values, not merely separate intervals optimized independently.

Remaining work

The research technology and welfare weights are assumed known here. Real pricing reforms also alter access, insurance, prescribing, trial selection, quality and timing. Revenue variation can identify some private incentive responses under a credible design without identifying every health externality or the global optimal reward. A complete solution needs those joint causal responses, a feasible financing policy and uncertainty analysis over success and value. The exact allocation formula and counterexamples clarify what an empirical solution must measure, while leaving the full catalogue target unresolved.

Completion Estimate

38%

This is a post hoc, heuristic estimate for the full catalogue target. The attempt solves indication specific research allocation with scarce capacity and derives a conditional reward implementation, while showing why innovation counts and marginal timing data miss health value. It does not identify joint access, quality, timing, and revenue responses or characterize feasible pricing, reimbursement, and buyout frontiers with financing and contracting constraints.

References

Go to contribution →