we-solve (beta)

Econ-254 · economics · Economics

Distribution of trade gains

LLM status Open unverified

Statement

Original subject group: Trade, globalization and geoeconomics

Audit classification: Background; proposed extension. The citation mixed the 2021 NBER WP with a 2024 journal URL. Both are real but distinct versions and are now separately listed. The exact household EV distribution problem is editorial, not verified as a canonical open question.

Definitions and precise research target

Fix households $i$, industries, locations and a trade-cost intervention $\tau$ relative to baseline $0$. Let $U_i(\tau)$ be equilibrium lifetime utility under wages, employment opportunities, consumer prices and wealth returns induced by the intervention. Define equivalent variation $EV_i(\tau)$ as the lump-sum amount added to baseline wealth that equates baseline attainable utility to $U_i(\tau)$, using the declared lifetime budget and preference model. The task is to identify the distribution \[F_{\tau}(v)=\Pr(EV_i(\tau)\leq v)\] for a stated household population, and decompose changes by earnings, expenditures and asset-ownership channels under explicitly ordered counterfactuals. Include migration, occupational switching and general-equilibrium price responses when the horizon permits them. Specify household preference heterogeneity and the assumptions translating measured income and price changes into welfare. Establish the exogeneity of the trade shock or derive an identified set when trade policy responds to domestic conditions. Report uncertainty and dependence on channel ordering, since interactions generally prevent a unique additive decomposition. An aggregate gains estimate or a local wage effect does not identify the full joint welfare distribution. Separate transitional losses from permanent effects and disclose exclusions caused by missing household wealth data.

Additional formal definitions and scope (editorial). Let $v_i(p^0,w_i)$ be baseline lifetime indirect utility at the entire baseline price path $p^0$, with initial wealth $w_i$. Equivalent variation solves $v_i(p^0,w_i+EV_i)=U_i(\tau)$; assume strict monotonicity in wealth and that the target is in its range. Positive $EV_i$ denotes a gain. Compensating variation instead changes wealth at the postintervention prices and is generally unequal. Fix a probability measure $\mu$ over baseline households, including the treatment of births, deaths and migrants, so $F_\tau(v)=\mu\{i:EV_i\leq v\}$. Identify the joint law of household preferences, incomes, spending and assets before constructing this distribution: marginal income or price distributions alone do not determine it. Counterfactual channel changes must retain budget feasibility and use an explicit order.

For the identification part, declare an admissible class $\Theta$ of structural models, including preferences, technologies, shock laws, measurement/sampling rules and any equilibrium selector. If $O$ denotes the complete observable history and $T(\theta)$ the target defined above, the identified set is $\mathcal I_T=\{T(\theta):\theta\in\Theta,\ \mathcal L_\theta(O)=\mathcal L_{\rm obs}(O)\}$; point identification means this set is a singleton. A $do(a)$ comparison replaces stated policy/technology equations while fixing the declared exogenous law and re-solving the remaining equations. These definitions do not assert that an identification theorem holds without further restrictions.

Variants and assumption changes

  1. Static goods-price and wage effects (source-motivated). Fix households and occupations and recover equivalent variation from observed preferences and joint income/expenditure changes.
  2. Lifetime adjustment and wealth (editorial). Allow moving, occupational investment and asset returns; identify the household welfare distribution and ordered channel contributions.

References and source audit

  1. NBER verifies Rodrik, 2021 WP 29507 and DOI; it is distinct from the 2024 journal publication originally linked. Locator: 2021 NBER abstract; author-hosted version, compensation discussion printed pp.3--4. Support: Trade-distribution and compensation background. NBER indexed metadata and author-hosted text inspected. URL: https://www.nber.org/papers/w29507.
  2. Publisher verifies Rodrik, 2024, Oxford Open Economics 3(Supplement 1):i1076--i1082, DOI 10.1093/ooec/odad048; not the 2021 NBER DOI. Locator: Compensation discussion; author-hosted version printed pp.3--4. Support: Compensation background; no exact optimal-program open theorem. Publisher indexed text and author-hosted version inspected; repeated publisher fetch failed. URL: https://academic.oup.com/ooec/article/3/Supplement_1/i1076/7708093.
  1. Dani Rodrik. 2021. A Primer on Trade and Inequality. 2021 NBER abstract; author-hosted version, compensation discussion printed pp.3--4. https://www.nber.org/papers/w29507. DOI: https://doi.org/10.3386/w29507.
  2. Dani Rodrik. 2024. A primer on trade and inequality. Compensation discussion; author-hosted version printed pp.3--4. https://academic.oup.com/ooec/article/3/Supplement_1/i1076/7708093. DOI: https://doi.org/10.1093/ooec/odad048.

Common conventions: Source mathematical conventions

These conventions apply to each entry unless explicitly replaced.

Models and identification. A primitive model $m$ specifies all probability laws, feasible choices, information, equilibrium and measurement rules used by its target. The admissible class $\mathcal M$ is fixed independently of outcomes. If $P_O(m)$ is its observable probability law and $\tau(m)$ the target, its identified set at observable law $P$ is \[ \mathcal I_\tau(P)=\{\tau(m):m\in\mathcal M,\ P_O(m)=P\}. \] Point identification means that this set is a singleton. An empty set signals incompatibility of data and assumptions. Coordinate bounds need not describe the joint set. Bounds are sharp only if every retained target is attainable or arbitrarily approachable by compatible models. Finite bounds require explicit restrictions on unobserved quantities; unrestricted confounding, hidden wealth, extrapolation or arbitrary equilibrium selection can yield uninformative sets.

Causal interventions. A potential outcome $Y_i(p)$ is the outcome under a complete policy regime $p$, including timing, financing, information and market spillovers. The target population and units are fixed before treatment. With interference, use the complete assignment vector or a justified exposure mapping; individual no-interference notation alone cannot identify an economy-wide intervention. A difference of mean potential outcomes does not require the cross-world joint distribution. Conditioning on post-treatment migration, survival, enrollment or employment changes the estimand and needs separate justification.

Statistical statements. An estimator, confidence set, forecast or test specifies a sampling law, model class and loss/coverage criterion. Uniform $(1-\alpha)$-coverage means $\inf_{m\in\mathcal M}\Pr_m\{\tau(m)\in C_n\}\geq1-\alpha$, or an explicitly asymptotic version. The whole parameter space trivially has coverage, so informativeness requires a stated width, risk or power objective. A forecasting improvement is relative to a named benchmark, information set and evaluation population.

Equilibrium and optimization. An equilibrium comprises feasible plans, optimality, beliefs where needed, budget constraints and all market/resource clearing. The primitives or a stated benchmark define each correspondence $\mathcal E(p;m)$. Nonemptiness is assumed or proved, never inferred just from compact policy sets. A selected-equilibrium target uses a declared selection rule; a robust target quantifies over all permitted equilibria. Use suprema or infima unless attainment is established. An $\varepsilon$-optimal policy is feasible and within $\varepsilon$ of the finite optimum value. Report an empty feasible set separately.

Welfare and accounting. Normative utility, interpersonal normalization, social weights, numeraire, discounting and the use of public receipts are primitives. Behavioral utility need not coincide with the welfare criterion. Household welfare includes ownership income and rents once; adding profits again to compensating variation generally double-counts them. A monetary equivalent is defined by an explicit transfer and price convention, with monotonicity and range conditions. Average effects, mechanism contributions, transfers and welfare losses are different targets. Infinite sums require convergence and dynamic feasible plans require terminal or no-Ponzi conditions.

Source versions. A working-paper date, revision date and journal publication year are distinguished. A DOI for a journal version is not automatically the DOI of an earlier manuscript. Locators refer to the stated version and printed pages where specified. A metadata-only check does not verify a claim about a full-text conclusion. Every entry's source subsection narrows the provenance claim accordingly.

Statement licence: CC-BY-4.0 · Source file

Contributions 1

Econ-254-A1 · Exploration · LLM-generated · gpt 6 astra ultra ·

GPT 6 Astra Ultra attempt on Econ-254

Distributional trade gains: welfare coupling and ordered channels

Target and scope

The task is the distribution of household equivalent variation, not only average income changes. The distributional and compensation questions are motivated by [rodrik]. We derive an exact static welfare map, solve the finite-support unknown-coupling problem, and quantify how channel ordering changes an attribution. Migration and dynamic occupational adjustment remain outside the benchmark.

An exact welfare map

A household has Cobb--Douglas preferences $U(c)=\prod_{j=1}^J c_j^{\alpha_j}$, with $\alpha_j>0$ and $\sum_j\alpha_j=1$. Its expenditure function is $e(p,u)=uP(p)$ where $P(p)=\prod_j(p_j/\alpha_j)^{\alpha_j}$. Baseline nominal resources are $m_0>0$, reform resources are $m_1>0$, and all prices are positive. Equivalent variation uses baseline prices: \[ EV=m_1\,{P(p^0)\over P(p^1)}-m_0. \tag{1} \] To verify (1), first-order conditions give expenditure shares $\alpha_j$, so indirect utility is $m/P(p)$. Equating baseline indirect utility at resources $m_0+EV$ to the reform value gives the formula. It applies household by household, including heterogeneous declared shares; it does not infer those shares or the joint resource-price distribution from marginals.

A sharp finite-support coupling calculation

In a quasilinear fixed-price subcase, write $EV=X+Z$, where $X$ is the earnings change and $Z$ the asset-income change. Suppose only their marginal distributions are identified: $\Pr(X=x_i)=a_i$ for $i=1,\ldots,m$ and $\Pr(Z=z_j)=b_j$ for $j=1,\ldots,n$, with nonnegative weights summing to one. No restriction on their joint household assignment is imposed. Baseline wealth can be chosen large enough that all these bounded changes are feasible. Let $\Pi$ be the transportation polytope \[ \pi_{ij}\ge0,\qquad\sum_j\pi_{ij}=a_i,\qquad\sum_i\pi_{ij}=b_j. \]
Theorem (Sharp CDF bounds): At every fixed welfare threshold $v$, the sharp identified interval for $F(v)=\Pr(EV\le v)$ is \[ \left[\min_{\pi\in\Pi}\sum_{ij}\pi_{ij}\mathbf1\{x_i+z_j\le v\},\quad \max_{\pi\in\Pi}\sum_{ij}\pi_{ij}\mathbf1\{x_i+z_j\le v\}\right]. \tag{2} \] Both endpoints are attained. Every intermediate value is attained by a compatible household joint distribution. An endpoint optimizer can be chosen with at most $m+n-1$ positive cells when all marginal weights are positive.
Proof: Every compatible joint distribution determines exactly one matrix in $\Pi$, and every matrix in $\Pi$ defines such a distribution on the finite product support. The displayed objective is precisely its CDF value. The polytope is nonempty because $a_ib_j$ is feasible, and is closed and bounded, so its linear extrema exist. Convex mixtures of minimizing and maximizing matrices retain the marginals and interpolate objective values, proving sharpness. An optimum can be chosen at a vertex. If the bipartite graph of positive cells of a vertex contained a cycle, alternately adding and subtracting a small common amount around that cycle would yield two distinct feasible matrices averaging to the original, a contradiction. The graph is therefore a forest on $m+n$ vertices and has at most $m+n-1$ edges. ∎

These are pointwise CDF bounds. Different thresholds may need different extremal couplings; choosing each endpoint independently need not yield a jointly attainable whole CDF. For a distributional target such as expected shortfall, optimize its own functional over the same common coupling.

Example (Equal means, different household losses): Let both $X$ and $Z$ equal $-1$ and $1$ with probability $1/2$ each. The coupling $Z=X$ gives $EV\in\{-2,2\}$ equally and $F(0)=1/2$. The coupling $Z=-X$ gives $EV=0$ surely and $F(0)=1$. Both models have identical marginal channels and average welfare change zero, but their fractions with strictly negative welfare are $1/2$ and zero. The CDF interval in (2) at zero is exactly $[1/2,1]$: a $2\times2$ feasible table is determined by its upper-left entry $t\in[0,1/2]$, and $F(0)=1-t$.

An explicit channel attribution with interactions

For a household in (1), write $m_1=m_0+e+a$ and $P(p^0)/P(p^1)=1+\delta>0$. Then \[ EV=e+a+m_0\delta+e\delta+a\delta. \tag{3} \] Define channel hybrids by activating a subset of the earnings, asset and price changes while setting the other changes to zero. Assume all eight hybrid resource levels are positive. These are valuation hybrids, not asserted market-clearing equilibria.
Proposition (Shapley attribution for the specified hybrids): Averaging marginal contributions over the six channel orders gives \[ \phi_e=e(1+\delta/2),\qquad \phi_a=a(1+\delta/2),\qquad \phi_p=\delta\{m_0+(e+a)/2\}. \] They sum to (3). The earnings contribution in any single order is $e$ if prices change later and $e(1+\delta)$ if prices change earlier; its order sensitivity has magnitude $|e\delta|$.
Proof: Each additive term belongs to its own channel in every order. For the interaction $e\delta$, whichever of earnings and prices occurs second receives the entire interaction. Each comes second in exactly half the permutations, so the average splits it equally. The same argument applies to $a\delta$, while $m_0\delta$ belongs to prices. This proves the formulas, efficiency, and the two possible single-order earnings contributions. ∎

Remaining gap

The finite programme solves a sharp distributional subproblem conditional on identified marginal causal changes and unrestricted coupling. In actual trade data, the policy shock's exogeneity, household preference shares, wealth coverage and general-equilibrium price responses require their own identification arguments. Neither marginal wage effects nor aggregate gains identify the joint welfare distribution. Lifetime adjustment and feasible counterfactual channel construction remain unresolved in this attempt.

Completion Estimate

40%

This is a post hoc, heuristic estimate for the full catalogue target. The attempt derives a static household equivalent-variation map, sharp finite-support coupling bounds and order-sensitive channel attribution. Identifying the full lifetime welfare distribution still requires causal trade variation, preferences, joint household wealth and earnings, equilibrium prices and endogenous migration or occupational adjustment.

References

Go to contribution →