We study how generative AI affects student learning in a randomized experiment. In proctored, in-person sessions, undergraduates learn about an unfamiliar topic and write an analytical essay with or without access to off-the-shelf generative AI, then complete unaided assessments immediately and one week later. We measure learning with knowledge tests (factual and conceptual understanding) and open-ended essays (higher-order skills). AI access raises immediate test scores by 0.27 standard deviations. These gains persist one week later. Essay quality, by contrast, changes little while students have AI access but improves in style and relevance one week later, when students write unaided. These delayed gains are larger among augmentation users-who use AI to explain concepts rather than generate text-whereas automation users' short-run quality gains vanish once AI is removed. We find evidence for two mechanisms behind the learning gains: students shift time away from drafting text and toward reading and searching for information, and they report greater learning enjoyment.

The adoption of artificial intelligence (AI) by large enterprises is an important potential source of aggregate productivity improvement and labor market impact. We study AI adoption of S&P 500 firms over the period 2016 to 2025, estimating adoption at the enterprise level. While generative AI tools are useful for personal and professional applications, our focus is on the deep integration of AI in the business processes of large enterprises which are bellwethers for firm adoption more broadly. We develop a novel measure to assess deep AI adoption (and distinguish it from AI hype) that is based on SEC 10-K filings, where laws and regulations ``prohibit companies from making materially false or misleading statements." In 2025, 11% of S&P 500 enterprises had AI deeply integrated into their business processes, and a further 10% were using AI in the production of goods and delivery of services. AI adoption has more than quadrupled from 5% in 2022 with slowly accelerating adoption among non-technology firms but very aggressive adoption in the technology sector which accounts for two-thirds of deeply integrated enterprise adoption. Firm profitability shows a "J-curve" as firms move from no adoption to deep adoption, but we observe no differences in capex or productivity. Among technology firms, but not others, AI adoption is higher for firms with more employees and higher values of Tobin's q.

We study how interoperability reshapes competitive price discrimination when consumers are embedded in a social network. Two differentiated platforms set personalized prices; consumers benefit from neighbors' consumption of the same platform and, under interoperability, of the rival. Equilibrium prices obtain in closed form for arbitrary networks and contain a network-position term, proportional to Katz-Bonacich centrality, whose sign is determined by whether interoperability exceeds product substitutability. Below this threshold, platforms contest central consumers and grant centrality discounts; above it, central consumers become gateways to a shared cross-platform network and pay premia; at the threshold, prices are independent of network position. Interoperability softens price competition, can make platforms favor denser consumer networks, and reverses which side of the market gains from price discrimination.

Ellsberg's famous paradox challenged Savage's subjective expected utility theory (EUT) -- which reduces uncertainty to risk -- by suggesting an aversion toward ambiguity. We provide a revealed preference test of the full set of axioms underpinning subjective EUT under uncertainty and compare it to an analogous test of objective EUT under risk. We find that individual choices are as consistent with utility maximization and expected utility maximization under uncertainty as they are under risk. Nevertheless, there is greater empirical scope for non-EUT models under uncertainty than under risk, and the absolute and relative consistency of EUT and non-EUT models vary considerably across subjects.

The coal phase-out's regional economic impact is a key challenge of the energy transition, as employment and fiscal dependence in coal regions face structural adjustment without automatic market solutions. Analyzing European Union NUTS 2 regions from 2000-2022 with fixed effects and clustered errors, coal regions show a consistent 1.1 percentage points unemployment premium and grow faster in gross domestic product per capita at 0.2 percentage points annually, indicating a hollowing-out process where population exit raises per-capita output while employment conditions worsen. Spatial analysis shows strong geographic clustering, supporting coordinated local and sectoral targeted transition policies. South Korea's rapid phase-out, with Chungnam as a major coal-power region, underscores the need for proactive national support to enable concrete regional action before plants shut down.

We use the assisted own goal hypothesis as a lens into media measurement. A demand-generating (upper-funnel) advertising platform such as a short-video social network can cause an incremental purchase, yet see that purchase booked on -- and credited to -- a downstream trusted marketplace, because consumers who discover a product on the platform complete the transaction elsewhere, for example because of distrust of the generating platform as a psychological mechanism. Under attribution-based return-on-ad-spend (ROAS) measurement, the diverted conversions are invisible to the originating platform. Marketing-mix models (MMMs) do not know which channel to credit with the outcome, and channel-by-week aggregation denies the audience-level granularity that budget decisions require. We develop an incrementality-based measurement model with two ingredients: ambient audience-level randomization -- each activated audience carries its own intent-to-treat (ITT) experiment -- and an individual-level extension of Predicted Incrementality by Experimentation (PIE), which learns a mapping from individual features to experiment-identified incremental outcomes. Because ITT contrasts are computed on channel-complete outcomes, the estimator is unbiased and the own goal disappears

Financial and economic history is strewn with bubbles and crashes, booms and busts, crises and upheavals of all sorts. Understanding the origin of these events is arguably one of the most important problems in economic theory: are economies intrinsically unstable, and can one ``stabilize unstable economies''? In this review I argue, from a physicist's vantage point, that the concept of equilibrium -- so central to mainstream economic thinking -- is likely to be the exception rather than the rule in large, complex, interacting systems. Drawing on a series of stylized ``toy'' models, I show how excess volatility, endogenous crises and crashes, inflation swells and persistent inequalities can all emerge naturally from genuinely out-of-equilibrium dynamics, without invoking large exogenous shocks. Three generic mechanisms recur throughout: trapping in a multiplicity of history-dependent equilibria; the impossibility of dynamically reaching equilibrium, leading to oscillations and chaos; and the spontaneous evolution towards fragile, marginally stable states -- the self-organized criticality paradigm. I stress that these are phenomenological scenarios rather than calibrated theories: there is, at this stage, no ``smoking gun''. But the burden of proof, I contend, should be on the equilibrium camp.

As agentic AI systems are increasingly applied to cyber-physical environments, their evaluation requires assessment of both task performance and trustworthiness. In decentralized energy markets, autonomous agents may improve market utility, but may also exploit invalid physical data, create artificial liquidity, and produce unstable governance decisions. Therefore, we propose SolarChain-Eval, a physics-constrained benchmark for evaluating trustworthy economic agents. It formulates market governance as a Gymnasium-compatible Markov Decision Process, where agents make hourly decisions. SolarChain-Eval evaluates each policy across multiple dimensions, including market utility, physical safety, slippage, action smoothness, spatial fairness, and auditability. To support agentic evaluation, SolarChain-Eval incorporates an LLM-based Planner/Auditor layer. The Planner defines episode-level action bounds and audit rules, while the Auditor reviews and revises high-risk actions. All interventions are recorded through structured logs, including trigger signals, proposed actions, revised actions, and audit rationales. Experiments with static, random, myopic, RL, and RL+LLM policies reveal a clear utility-safety trade-off. RL agents improve market utility but can still produce unsafe behavior. When the physics penalty is removed, reward-maximizing agents exploit invalid generation and increase artificial liquidity. The LLM Planner/Auditor improves auditability and mitigates selected risks, but it cannot fully compensate for a misspecified reward function. These results indicate that trustworthy agentic AI evaluation requires both physical constraints and transparent intervention traces. We release data and code as open access on GitHub for replicability.

Decentralized Autonomous Organizations (DAOs) use token-weighted voting to allocate resources, set protocol rules, and legitimate collective decisions. Yet, support in DAO voting is strikingly concentrated. What happens inside the ballot that produces this concentration? We study DAOs' governance at the proposal-choice level, linking each choice's voting-power share to three observable features: whether it expresses an approval-oriented stance, where it appears in the choice list, and whether it is selected by the proposal author. We find that (i) author-selected choices show the strongest and most robust association with voting-power share, with a 58.8% increase relative to non-author choices; (ii) approval-oriented choices retain a positive but slightly less consistent advantage (27.1%); and (iii) first-listed choices also attract systematically higher shares, consistent with position and order effects (7.7%). Results are robust across several specifications, which include subtracting an author's own voting power from computations. We use bias descriptively, to denote systematic associations rather than proven causal distortion. The results shift attention from proposal outcomes alone to the interface and social signals through which choices are presented. In DAO governance, ordering, author signals, and vote visibility should be treated as institutional design choices, not neutral implementation details.

This paper proposes the certainty-equivalent first-order learning (CEFOL) algorithm, a deep learning algorithm for solving discrete-time dynamic programming problems with recursive utility. Dynamic programming with recursive utility is challenging because nonlinear certainty equivalent appears in the Bellman equation and the first-order optimality conditions but is difficult to evaluate. By introducing a separate neural network to represent the certainty equivalent, CEFOL enables the exploitation of the Bellman and model-specific first-order optimality conditions. In addition to certainty equivalent, CEFOL also uses neural networks to learn the value functions, policy functions, and Lagrange multipliers by using model-specific first-order conditions to construct residuals for minimization. By using first-order and KKT residuals to learn the policy, CEFOL directly accommodates general equality and inequality constraints on the controls, including occasionally binding constraints, without requiring penalty functions or problem-specific reformulations. We apply the algorithm to risk-sensitive and Epstein--Zin consumption-saving problems, a small-noise robust-control problem, and a DSGE model with recursive preferences and stochastic volatility. Across these applications, out-of-sample Bellman diagnostics and model-specific optimality residuals, including Euler or first-order residuals where applicable, are generally of order 1.0e-4 to 1.0e-3 over the relevant state regions, with larger values mainly near binding constraints, and the learned value and policy functions closely match VFI benchmarks when available. The CEFOL algorithm also works for dynamic programming problems with expected utility, as expected utility is a special case of recursive utility.

For analytic convenience, existing statistical frameworks either assume random or fixed regressors. However, it is a little awkward that they do not cover the practical case of estimating the average treatment effect in experiments with randomized treatments and non-randomized, fixed pretreatment covariates. We unify the literature by providing the theory for regressions with mixed regressors that contain both random and fixed components. Importantly, our theory allows for misspecification of the regression functions. We first establish general results for estimating equations with both random and fixed components and then use it to analyze misspecified linear regression, with applications to completely randomized experiments. We focus on the causal interpretation of the regression coefficients and standard errors even when the models are wrong. We start with the theory for independent data and then extend the discussion to clustered data.

Large-scale traffic assignment requires equilibrium models that are both behaviorally plausible and computationally tractable. This paper develops a perturbed utility Markovian equilibrium (PUME) framework that preserves the scalability of link-based Markovian traffic equilibrium models and extends their applicability to settings with boundary choice probabilities, undiscounted network loading, and general link interactions. As the behavioral basis of PUME, we first develop the perturbed utility Markovian choice model (PUMCM) in which the Bellman optimality operator is defined through a convex surplus function whose gradient directly yields the optimal policy. The model generalizes existing additive random utility (ARUM) Markovian choice models and admits both interior and boundary choice probabilities. Accordingly, unattractive links can receive zero flow without imposing ex ante choice-set restrictions as in existing ARUM models. We establish conditions under which the corresponding Markov decision problem is well posed and yields a proper demand mapping. We then formulate the equilibrium as a variational inequality (VI) problem on the dual cost space and establish its existence and uniqueness. Particularly, the VI formulation of PUME accommodates non-separable and asymmetric cost structures and thus offers a more flexible modeling framework than existing Markovian traffic equilibrium (MTE) models. For computation, we develop a modified policy iteration method for network loading and a safeguarded accelerated meta-algorithm for computing equilibrium. Both algorithms are proven to be globally convergent and have demonstrated satisfactory numerical performances. Experiments on benchmark and synthetic networks further show that the proposed framework is highly scalable and robust towards a wide variety of demand-supply settings.

This paper explores the connection between two distinct notions of irrationality: the extent to which the consumer fails to maximize his utility function and the extent to which the consumer can be turned into a money pump. We show that the amount of money which can be pumped by an arbitrageur is equal to both (i) the minimum amount of money which the consumer overpays to attain his utility targets (minimum over all utility functions) and (ii) the minimum amount of wasted quasilinear utility (minimum again over all utility functions). Under the assumption that the consumer's true utility function belongs to the aforementioned argmins, we present a method for recovering this true utility function even when choices are non-optimal. Further, we show that this recovered utility function can be interpreted as belonging to a utility maximizing consumer who systematically mis-perceives prices. These results and many more are proved by exploiting a novel connection between revealed preference analysis and optimal transport.

This paper develops a network-based methodology for measuring economic integration in Africa. Conventional indicators such as trade openness and intra-regional trade shares capture the volume of trade but not countries' structural roles within the continental trade system. Using bilateral intra-African trade data for 2000-2019, we construct annual directed trade networks and measure countries' direct connectivity, recursive influence, brokerage role, and core embeddedness using weighted degree, PageRank, betweenness and random-walk betweenness, clustering, and k-core network indicators. We further benchmark observed network positions against a maximum-entropy null model to identify countries whose centrality exceeds that expected from their trade intensity alone. The results show a gradual densification of intra-African trade and the persistence of a core-periphery structure led by a small group of highly connected economies. Dynamic panel estimates indicate that economic development, infrastructure, institutional quality, human capital, regional trade agreements, and FDI are associated with stronger structural integration, whereas trade costs and overlapping regional memberships weaken several dimensions of network position. The findings suggest that African integration depends not only on expanding trade volumes but also on improving countries' positions within continental trade.

We study consumption stimulus using digital coupons, which provide time-limited subsidies contingent on minimum spending. Analyzing a large-scale program in China, we find that the program generates large and heterogeneous short-term effects. Consumption responses vary across both consumers and locations, reflecting both demand-side and supply-side factors. This heterogeneity shapes the incidence of the program: high-response consumers tend to patronize larger businesses, leading to a regressive allocation of stimulus benefits. Through counterfactual analysis, we show that targeting rules can reshape both the size and distribution of stimulus effects. Targeting the most responsive consumers can more than double the aggregate stimulus, while a hybrid design that combines targeted distribution with direct support to small businesses improves both efficiency and equity.

We study optimal auction design when the direction of bidders' deviations is restricted. We show that the optimal revenue when bidders can only underbid their true values cannot exceed the optimal revenue when bidders may freely underbid or overbid. Thus, unidirectional incentive compatibility is sufficient for full incentive compatibility for revenue maximization. We prove this equivalence through linear programming duality in a discrete model, which makes it possible to analyze the feasibility of allocation rules in multi-agent environments.

We extend the Fokker-Planck framework of Froseth (2026, arXiv:2603.05283) to populations of investors with heterogeneous, persistent return-generating ability. When the drift coefficient in the Langevin equation for log-wealth varies across investors, the proportional wealth tax remains a uniform drift shift but ceases to be neutral in the economic sense: its real incidence differs across ability types, and the stationary wealth distribution changes shape. We derive the extended Fokker-Planck equation on the joint space of log-wealth and ability, characterise the conditions under which the drift-shift symmetry breaks, and identify the consequences for asset prices and portfolio allocations. The analysis connects the neutrality results of Froseth (2026, arXiv:2603.05264) and the Fokker-Planck dynamics of Froseth (2026, arXiv:2603.05283) to the heterogeneous-returns literature, notably the "use-it-or-lose-it" mechanism of Guvenen, Kambourov, Kuruscu, Ocampo-Diaz and Chen (2023).

We derive the asymptotic risk function of regularized empirical risk minimization (ERM) estimators tuned by $n$-fold cross-validation (CV). The out-of-sample prediction loss of such estimators converges in distribution to the squared-error loss (risk function) of shrinkage estimators in the normal means model, tuned by Stein's unbiased risk estimate (SURE). This risk function provides a more fine-grained picture of predictive performance than uniform bounds on worst-case regret, which are common in learning theory: it quantifies how risk varies with the true parameter. As key intermediate steps, we show that (i) $n$-fold CV converges uniformly to SURE, and (ii) while SURE typically has multiple local minima, its global minimum is generically well separated. Well-separation ensures that uniform convergence of CV to SURE translates into convergence of the tuning parameter chosen by CV to that chosen by SURE.