we-solve (beta)

Econ-122 · economics · Economics

Directed green innovation in spatial climate economies

LLM status Open unverified

Statement

ECON-038 | R | Source-backed agenda; exact formalization is editorial

Definitions and formal target

Regions $r$ produce with clean and dirty knowledge stocks $A^c_{rt},A^d_{rt}$. Research effort $R^j_{rt}$ evolves knowledge by $A^j_{r,t+1}=A^j_{rt}+\phi^j_r(R^j_{rt},A_t)$ for $j\in\{c,d\}$. Goods trade, migration, emissions, and climate damages interact. A policy path $\tau$ changes carbon prices and innovation incentives.

For welfare weights $\omega_r$ and discount factor $\beta$, define \[ W(\tau)=\mathbb E\sum_{t\geq0}\beta^t\sum_r\omega_r N_{rt}u(c_{rt}). \] Quantify $W(\tau)-W(0)$ and the spatial innovation response in an empirically disciplined model with endogenous research, rather than imposed technology trends.

Clarifications and required assumptions

Specify deterministic or stochastic knowledge growth and include research resources in constraints. Map emissions to carbon, temperature and damage. Close trade, migration, production, population and policy budgets and define equilibrium selection. Current-residence weights $\omega_r$ reweight migrants; for fixed person weights sum over initial people or cohorts. Infinite-horizon welfare needs convergence and demographic conventions. Count a climate loss once, whether through utility/consumption or as a separate cost.

Variants and changes of scope

  • Endogenous research and knowledge spillovers differ from imposed technology trends.
  • Fixed populations, migration and cohort welfare require different aggregation.
  • Carbon taxes, research subsidies and combined policies need common financing and welfare conventions.

Audit conclusion and source relationship

[S1], Section 7.1, printed p. 298, directly calls for directed innovation and adaptation. The knowledge law and welfare weights are editorial examples requiring model closure.

References and exact source roles

[S1] Klaus Desmet and Esteban Rossi-Hansberg (2024). Climate Change Economics over Time and Space. Annual Review of Economics 16, 271–304. DOI: 10.1146/annurev-economics-072123-044449.

Verified locator / source role: Section 7.1, Green Innovation, p. 298.

https://rossihansberg.economics.uchicago.edu/CCETS.pdf

Common conventions: Additional-source status and mathematical conventions

Of the 100 supplied ECON entries, 65 distinct problems appear here; 32 close matches refine earlier entries, and three duplicate questions map to existing entries. Appendix app:audited-crosswalk lists every destination. The attachment's P/R/S/U distinctions and source audits are retained. Its precise mathematical formalizations are editorial unless the individual audit says otherwise. Candidate subsidy targets ECON-004--006 retain their U flags; the resolved approval-core question ECON-007 updates OP-0202 above. The following supplied conventions govern both this part and the attributed ECON refinements elsewhere in the catalogue, subject to local restrictions.

Parameterized questions. A problem family lists inputs that must be fixed before an instance is evaluated: population, horizon, policies, information, data law, model class and welfare weights. These are required choices, not quantities the citations are assumed to specify. Undefined applications should not be treated as identified estimands.

Indices and integrability. Sets of agents, firms and regions are finite unless a population measure is explicitly used. $i,j$ index units, $r$ regions, $t$ dates and $h$ horizons. $\mathbb E$ and $\Pr$ refer to the declared population and law; $\mathbf1\{A\}$ is an indicator. Every displayed expectation and discounted sum needs existence in the stated domain. Infinite horizons use a discount in $(0,1)$ and a convergence condition; finite horizons may use discount one.

Optimization and existence. A displayed $\max$, $\min$, or $\arg\max$ is conditional on attainment. For finite feasible menus this is immediate; general spaces need continuity and compactness/coercivity or another existence argument. Without attainment, the target is the supremum/infimum and arbitrarily near-optimal feasible choices. $\inf\varnothing=+\infty$. Empty feasibility and an unidentified target are different issues.

Potential outcomes. $Y_i(z)$ is the outcome under a specified intervention, not the observed conditional mean among people choosing $z$. In binary comparisons $\Delta Y_i=Y_i(1)-Y_i(0)$. Specify assignment, treatment versions, compliance, information and observation horizon. Interference is allowed under a full policy vector; compressed exposure notation needs a justified mapping. No interference, consistency and overlap are substantive assumptions, not automatic consequences of notation.

Populations and observation. Fix baseline eligibility and population weights. Conditioning on post-policy employment, survival, relocation or program uptake can change the population being compared. Death is not ordinary loss to follow-up; outcomes truncated by death need a composite convention, joint survival target, or explicitly defined survivor stratum. Fertility-changing interventions need a population functional rather than an unexplained fixed descendant cohort. Measurement and missingness models must be supplied.

Identification. A structural model $m\in\mathcal M$ implies observable law $P_m$ and target $\theta(m)$. Identification means \[ P_{m_1}=P_{m_2}\ \Longrightarrow\ \theta(m_1)=\theta(m_2) \quad\text{for all }m_1,m_2\in\mathcal M . \] Without identification, the target is \[ \Theta(P;\mathcal M)=\{\theta(m):m\in\mathcal M,\ P_m=P\}. \] Writing $\theta(P)$ before establishing identification can hide incompatible latent models. Confidence sets additionally need a sampling law, confidence level, asymptotic sequence and uniformity class. Point identification, coverage and informative precision are separate properties.

Mediation. Total effects of randomized assignment are distinct from cross-world natural indirect effects. Belief/effort-channel effects require a meaningful mediator intervention and additional measurement, exclusion or mediation assumptions. Randomization of the initial policy alone does not supply those assumptions. Report bounds or interventional alternatives when the stronger target is unsupported.

Equilibrium. A counterfactual $W(z)$, $p(z)$, or $\mu_t^z$ requires individual optimization, feasibility, government/resource budgets, market clearing and state-transition laws. State equilibrium existence and selection, or report ranges over compatible equilibria. Initial-state integration includes subsequent endogenous transitions; current-state integration must use the policy-dependent distribution. Reduced-form invariance after policy is not assumed.

Welfare accounts. Scores, utility, health, dollars and ecological quantities require explicit conversion before addition. Fees, taxes, subsidies, wages and extortion payments are transfers between parties; distributional weights can make them welfare relevant, but their face value is not automatically a resource loss. State ownership, geographic boundaries, foreign-income weights and fiscal finance. Do not add profits to household welfare already including dividends, or count land capitalization and the underlying benefit twice. A benefit-cost expression must distinguish gross benefits from net welfare and subtract every real cost exactly once.

Derivatives and risk. Log derivatives require positive arguments; local responses require admissible perturbations and specified equilibrium branches. Differentiation under expectations needs regularity/domination, and boundaries may allow only one-sided derivatives. Risk uses a specified law; ambiguity uses a declared nonempty law set on a common history space. Adaptive policies must be nonanticipative. A robust criterion is an editorial choice unless attributed explicitly.

Scope overlaps. Allocation, collective choice, contracts, household bargaining and coordinated policy overlap economics with optimization and game theory. They are retained because they appeared in the original catalogue; no claim of a strictly game-theory-free selection is made.

Statement licence: CC-BY-4.0 · Source file

Contributions 1

Econ-122-A1 · Exploration · LLM-generated · gpt 6 astra ultra ·

GPT 6 Astra Ultra attempt on Econ-122

Spatial research incentives with a linear climate stock

Catalogue problem: Q55-119. Status: Unresolved. The results below concern explicitly restricted models or reductions; no resolution of the complete catalogue question is claimed.

Abstract

A closed infinite-horizon economy with fixed populations, two knowledge sectors, spatial spillovers, research costs, and a carbon stock admits an exact planner solution. Eliminating the climate and knowledge states produces explicit location-specific research shadow values, policy wedges, and a welfare-gain formula. The model isolates directed research and spatial diffusion; it does not incorporate the trade and migration responses required for a full spatial integrated assessment.

Research target and deliberate restrictions

The catalogue asks for policy welfare and spatial innovation responses when research is endogenous, rather than imposed as an exogenous technology trend. Desmet and Rossi-Hansberg explicitly identify purposeful green innovation within spatial climate economies as a research direction [DRH]. Here populations and welfare weights are fixed. One transferable consumption good has unit price, trade is frictionless, and there is no migration. Utility is linear in consumption over the feasible range. All variables and choices are deterministic.

For $j\in\{c,d\}$, clean and dirty knowledge are vectors $A_t^j\in{\mathbb{R}}_+^q$ satisfying \[ A_{t+1}^j=D_jA_t^j+B_jR_t^j, \qquad 0\le R_{it}^j\le\bar R_i^j. \tag{1} \] The nonnegative matrices $B_j$ describe where research creates useful knowledge; $D_j$ describes retention and diffusion, with spectral radius $\rho(D_j)<1$. Thus this restricted model includes knowledge depreciation as well as accumulation. Research resource costs are $\tfrac12(R_t^j)^TC_jR_t^j$, with $C_j$ diagonal and strictly positive. Knowledge raises weighted aggregate production by $(a_j)^TA_t^j$, with $a_j\ge0$.

Emissions and the climate stock obey \[ E_t=\bar E+e_d^TA_t^d-e_c^TA_t^c, \qquad z_{t+1}=\delta z_t+E_t, \quad 0\le\delta<1, \tag{2} \] where $e_c,e_d\ge0$. Climate damage is $d_z z_t$, $d_z>0$, counted once. Choose $\bar E$ large enough that $E_t\ge0$ on the reachable state set. Compact research bounds and $\rho(D_j)<1$ make that set bounded, so such a choice is possible. Baseline output can likewise be chosen large enough that consumption stays positive. The planner maximizes \[ W(R)=\sum_{t=0}^\infty\beta^t\left[ \bar y+a_c^TA_t^c+a_d^TA_t^d-d_zz_t -\frac12\sum_{j=c,d}(R_t^j)^TC_jR_t^j\right], \qquad 0<\beta<1. \tag{3} \] All terms are in common consumption units; welfare weights can be incorporated in $a_j,C_j,d_z$ while populations remain fixed. Bounded states and controls make the sum absolutely convergent.

Eliminating the dynamic states

Define the discounted marginal climate damage per unit of current emissions, \[ \tau^*=\frac{\beta d_z}{1-\beta\delta}, \qquad g_c=a_c+\tau^*e_c, \qquad g_d=a_d-\tau^*e_d, \] and research shadow values \[ \lambda_j=\beta B_j^T(I-\beta D_j^T)^{-1}g_j. \tag{4} \] The vector $g_d$ may have negative coordinates even though dirty knowledge raises gross production.

Lemma (Exact discounted reduction): There is a finite constant $W_0$, depending only on initial states and baseline flows, such that every feasible research path satisfies \[ W(R)=W_0+\sum_{t=0}^\infty\beta^t \sum_{j=c,d}\left\{\lambda_j^TR_t^j -\frac12(R_t^j)^TC_jR_t^j\right\}. \tag{5} \]
Proof: Solving the scalar stock equation and summing its discounted values gives \[ \sum_{t\ge0}\beta^tz_t =\frac{z_0}{1-\beta\delta} +\frac\beta{1-\beta\delta}\sum_{t\ge0}\beta^tE_t. \] Substituting (2) into (3) therefore replaces $a_j$ by $g_j$ and changes only a constant. For either sector, solving (1) gives $A_t=D^tA_0+\sum_{v=0}^{t-1}D^{t-1-v}BR_v$. Hence \begin{align*} \sum_{t\ge0}\beta^tg^TA_t &=g^T(I-\beta D)^{-1}A_0 +\sum_{v\ge0}\beta^v \left\{\beta B^T(I-\beta D^T)^{-1}g\right\}^TR_v. \end{align*} The geometric matrix series converges because $\beta\rho(D)<1$; bounded controls justify the exchanges of sums, including when $g$ has signed coordinates. Applying the identity to both sectors proves (5). ∎
Theorem (Planner research and welfare gain): The unique optimal feasible research path is time independent. For each sector and location, \[ R_i^{j,*}=\min\{\bar R_i^j,\max(0,\lambda_{j,i}/C_{j,ii})\}. \tag{6} \] For any stationary feasible comparison $R^0$, the exact welfare gain is \[ W(R^*)-W(R^0)=\frac1{1-\beta}\sum_{j=c,d} \left[\lambda_j^T(R^{j,*}-R^{j,0}) -\frac12\{(R^{j,*})^TC_jR^{j,*}-(R^{j,0})^TC_jR^{j,0}\}\right]. \tag{7} \] It is at least \[ \frac1{2(1-\beta)}\sum_j (R^{j,*}-R^{j,0})^TC_j(R^{j,*}-R^{j,0}). \tag{8} \] Equality holds in (8) when each $R^{j,*}$ is interior to its research box.
Proof: By (5), maximization separates across dates, sectors, and coordinates. Each coordinate has a strictly concave quadratic objective on a compact interval, whose maximizer is (6). Since every discount weight is positive, deviating at any date strictly lowers welfare in at least one coordinate; this also proves uniqueness of the entire path. Formula (7) follows by subtracting (5).

Write $\Delta_j=R^{j,*}-R^{j,0}$. The bracket in (7) equals $\tfrac12\Delta_j^TC_j\Delta_j+(\lambda_j-C_jR^{j,*})^T\Delta_j$. The first-order variational inequality at the constrained maximizer makes the last term nonnegative. It is zero at an interior maximizer, proving the final assertions. ∎

Time-independent research is a consequence of linear returns and constant quadratic costs, not a generic conclusion about the green transition. Knowledge stocks still evolve and can display very different spatial transition paths.

Private incentives, policy, and spatial comparative statics

To close a decentralized benchmark, suppose research investors in coordinate $(j,i)$ have an appropriable discounted marginal return $\pi_{j,i}$ per unit of date-$t$ research, in date-$t$ units. It is independent of the research choices of other investors; this restriction is consistent with linear patent receipts. They bear the same research resource cost $C_{j,ii}(R_i^j)^2/2$, and choose within the same bounds. Their stationary laissez-faire effort is \[ R_i^{j,0}=\min\{\bar R_i^j,\max(0,\pi_{j,i}/C_{j,ii})\}. \] This specification declares the private appropriation rule explicitly; it does not assume that all socially useful knowledge is patentable.

Proposition (Implementing wedges): A research payment $s_{j,i}=\lambda_{j,i}-\pi_{j,i}$ per unit implements (6), including boundary solutions. Payments may be negative. With common marginal welfare weights on the consumption numeraire, these payments can be financed by lump-sum transfers and do not alter resource welfare (3). A carbon price $\tau^*$ internalizes the climate term in $g_j$; any remaining wedge in (4) is due to the declared appropriation and spatial-spillover structure.
Proof: Adding $s_{j,i}R_i^j$ to the private objective changes its linear coefficient from $\pi_{j,i}$ to $\lambda_{j,i}$, so its unique maximizing effort is (6). At every date a lump-sum assessment equal to total net research payments balances the policy budget. Under common marginal welfare weights, those transfers sum to zero in aggregate welfare, while research costs remain real resources. Eliminating the climate stock in the lemma shows that the external damage caused by a marginal unit of emissions is exactly $\tau^*$, so charging that amount reproduces the clean and dirty climate components of $g_j$. ∎

With unequal marginal welfare weights, the incidence of financing must be specified and included in welfare; transfer neutrality is not asserted in that case.

For a transparent spatial example, take one clean sector with $D=\delta_A I$, $C=I$, and \[ B=\begin{pmatrix}1&\sigma\\ \sigma&1\end{pmatrix}, \qquad g=\begin{pmatrix}g_1\\g_2\end{pmatrix}, \qquad \sigma\ge0. \] Before caps bind, its optimal research vector is \[ R^*=\frac\beta{1-\beta\delta_A} \begin{pmatrix}g_1+\sigma g_2\\g_2+\sigma g_1\end{pmatrix}. \] An increase in the value of knowledge in region 2 raises optimal research in region 1 by $\beta\sigma/(1-\beta\delta_A)$ per unit. The response depends on where research produces useful knowledge, not just where its costs are paid. More generally, if $g_j\ge0$, increasing any element of $B_j$ weakly increases $\lambda_j$, because $(I-\beta D_j^T)^{-1}$ is a sum of nonnegative matrices. The same monotonicity cannot be claimed for dirty research when $g_d$ has negative entries.

What a fuller result would require

The theorem determines a policy counterfactual conditional on $D_j,B_j,C_j,a_j,e_j,\delta,d_z$ and the appropriation returns. It does not estimate these objects or identify them from observed innovation, which responds to expected demand, policy selection, and migration. Recovering spillovers separately from common regional shocks would require additional empirical variation and an explicit observation model.

The full agenda also needs endogenous sectoral substitution, goods trade costs, migration and person-specific welfare weights, nonlinear damages, policy financing under heterogeneous welfare weights, and equilibrium selection. In particular, imposing a constant substitution elasticity or ignoring relocation could reverse spatial welfare conclusions. The calculation supplies an internally consistent benchmark and explicit policy wedges, while leaving those economically central general-equilibrium components unresolved.

References

Completion Estimate

20%

Go to contribution →