Regression discontinuity (RD) designs with multiple running variables arise in a growing number of empirical applications, including geographic boundaries and multi-score assignment rules. Although recent methodological work has extended estimation and inference tools to multivariate settings, far less attention has been devoted to developing global testing methods that formally assess whether a discontinuity exists anywhere along a multivariate treatment boundary. Existing approaches perform well in large samples, but can exhibit severe size distortions in moderate or small samples due to the sparsity of observations near any particular boundary point. This paper introduces a complementary global testing procedure that mitigates the small-sample weaknesses of existing multivariate RD methods by integrating multivariate machine learning estimators with a distance-based aggregation strategy, yielding a test statistic that remains reliable with limited data. Simulations demonstrate that the proposed method maintains near-nominal size and strong power, including in settings where standard multivariate estimators break down. The procedure is applied to an empirical setting to demonstrate its implementation and to illustrate how it can complement existing multivariate RD estimators.

Artificial intelligence weather prediction (AIWP) models now often outperform traditional physics-based models on common metrics while requiring orders-of-magnitude less computing resources and time. Open-access AIWP models thus hold promise as transformational tools for helping low- and middle-income populations make decisions in the face of high-impact weather shocks. Yet, current approaches to evaluating AIWP models focus mainly on aggregated meteorological metrics without considering local stakeholders' needs in decision-oriented, operational frameworks. Here, we introduce such a framework that connects meteorology, AI, and social sciences. As an example, we apply it to the 150-year-old problem of Indian monsoon forecasting, focusing on benefits to rain-fed agriculture, which is highly susceptible to climate change. AIWP models skillfully predict an agriculturally relevant onset index at regional scales weeks in advance when evaluated out-of-sample using deterministic and probabilistic metrics. This framework informed a government-led effort in 2025 to send 38 million Indian farmers AI-based monsoon onset forecasts, which captured an unusual weeks-long pause in monsoon progression. This decision-oriented benchmarking framework provides a key component of a blueprint for harnessing the power of AIWP models to help large vulnerable populations adapt to weather shocks in the face of climate variability and change.

The establishment of the UK National Health Service (NHS) in July 1948 was one of the most consequential health policy interventions of the twentieth century, providing universal and free access to medical care and substantially expanding maternal and infant health services. In this paper, we estimate the causal effect of the NHS introduction on early-life mortality and we test whether survival is selective. We adopt a regression discontinuity design under local randomization, comparing individuals born just before and just after July 1948. Leveraging newly digitized weekly death records, we document a significant decline in stillbirths and infant mortality following the introduction of the NHS, the latter driven primarily by reductions in deaths from congenital conditions and diarrhea. We then use polygenic indexes (PGIs), fixed at conception, to track changes in population composition, showing that cohorts born at or after the NHS introduction exhibit higher PGIs associated with contextually-adverse traits (e.g., depression, COPD, and preterm birth) and lower PGIs associated with contextually-valued traits (e.g., educational attainment, self-rated health, and pregnancy length), with effect sizes as large as 7.5% of a standard deviation. These results based on the UK Biobank data are robust to family-based designs and replicate in the English Longitudinal Study of Ageing and the UK Household Longitudinal Study. Effects are strongest in socioeconomically disadvantaged areas and among males. This novel evidence on the existence and magnitude of selective survival highlights how large-scale public policies can leave a persistent imprint on population composition and generate long-term survival biases.

I introduce and study a nested search problem modeled as a tree structure that generalizes Weitzman (1979) in two ways: (1) search progresses incrementally, reflecting real-life scenarios where agents gradually acquire information about the prizes; and (2) the realization of prizes can be correlated, capturing similarities among them. I derive the optimal policy, which takes the form of an index solution. I apply this result to study monopolistic competition in a market with two stages of product inspection. My application illustrates that regulations on drip pricing lower equilibrium price and raise consumer surplus.

Reliance on generative AI can reduce cultural variance and diversity, especially in creative work. This reduction in variance has already led to problems in model performance, including model collapse and hallucination. In this paper, we examine the long-term consequences of AI use for human cultural evolution and the conditions under which widespread AI use may lead to "cultural collapse", a process in which reliance on AI-generated content reduces human variation and innovation and slows cumulative cultural evolution. Using an agent-based model and evolutionary game theory, we compare two types of AI use: complement and substitute. AI-complement users seek suggestions and guidance while remaining the main producers of the final output, whereas AI-substitute users provide minimal input, and rely on AI to produce most of the output. We then study how these use strategies compete and spread under evolutionary dynamics. We find that AI-substitute users prevail under individual-level selection despite the stronger reduction in cultural variance. By contrast, AI-complement users can benefit their groups by maintaining the variance needed for exploration, and can therefore be favored under cultural group selection when group boundaries are strong. Overall, our findings shed light on the long-term, population-level effects of AI adoption and inform policy and organizational strategies to mitigate these risks.

I study symmetric competitions in which each player chooses an arbitrary distribution over a one-dimensional performance index, subject to a convex cost. I establish existence of a symmetric equilibrium, document various properties it must possess, and provide a characterization via the first-order approach. Manifold applications--to R&D competition, oligopolistic competition with product design, and rank-order contests--follow.

Finite mixture models are widely used in econometric analyses to capture unobserved heterogeneity. This paper shows that maximum likelihood estimation of finite mixtures of parametric densities can suffer from substantial finite-sample bias in all parameters under mild regularity conditions. The bias arises from the influence of outliers in component densities with unbounded or large support and increases with the degree of overlap among mixture components. I show that maximizing the classification-mixture likelihood function, equipped with a consistent classifier, yields parameter estimates that are less biased than those obtained by standard maximum likelihood estimation (MLE). I then derive the asymptotic distribution of the resulting estimator and provide conditions under which oracle efficiency is achieved. Monte Carlo simulations show that conventional mixture MLE exhibits pronounced finite-sample bias, which diminishes as the sample size or the statistical distance between component densities tends to infinity. The simulations further show that the proposed estimation strategy generally outperforms standard MLE in finite samples in terms of both bias and mean squared errors under relatively weak assumptions. An empirical application to latent group panel structures using health administrative data shows that the proposed approach reduces out-of-sample prediction error by approximately 17.6% relative to the best results obtained from standard MLE procedures.

The paper analyzes four classical signal-plus-noise models: the factor model, spiked sample covariance matrices, the sum of a Wigner matrix and a low-rank perturbation, and canonical correlation analysis with low-rank dependencies. The objective is to construct confidence intervals for the signal strength that are uniformly valid across all regimes - strong, weak, and critical signals. We demonstrate that traditional Gaussian approximations fail in the critical regime. Instead, we introduce a universal transitional distribution that enables valid inference across the entire spectrum of signal strengths. The approach is illustrated through applications in macroeconomics and finance.

This paper replicates Chen and Kung's 2019 analysis ($The$ $Quarterly$ $Journal$ $of$ $Economics$ 134(1): 185-226). Inspecting the data reveals that nearly one-third of transactions (388,903 out of 1,208,621) are perfect duplicates of other rows, excluding the transaction number. The analysis on the data sans duplicates replicates their statistically significant princeling effect, robust across various specifications. Further analysis reveals a disagreement between Chen and Kung's text and code: the paper's ''logarithm of area'' is actually area ($\text{m}^2$) divided by one million. This therefore necessitates a reinterpretation of the estimation results, revealing that the princeling effect is extremely large.

This paper develops a peer effect model for count responses under rational expectations. The model accounts for heterogeneity in peer effects across groups based on observed characteristics. Identification is based on the linear model condition that requires the presence of friends of friends who are not direct friends. I show that this identification condition extends to a broad class of nonlinear models. Parameters are estimated using a nested pseudo-likelihood approach. An empirical application to students' extracurricular participation reveals that females are more responsive to peers than males. An easy-to-use R package, CDatanet, is available for implementing the model.

Determining an individual's strategic reasoning capability based solely on choice data is a complex task. This complexity arises because sophisticated players might have non-equilibrium beliefs about others, leading to non-equilibrium actions. In our study, we pair human participants with computer players known to be fully rational. This use of robot players allows us to disentangle limited reasoning capacity from belief formation and social biases. Our results show that, when paired with robots, subjects consistently demonstrate higher levels of rationality and maintain stable rationality levels across different games compared to when paired with humans. This suggests that strategic reasoning might indeed be a consistent trait in individuals. Furthermore, the identified rationality limits could serve as a measure for evaluating an individual's strategic capacity when their beliefs about others are adequately controlled.

We develop a methodology for conducting inference on extreme quantiles of unobserved individual heterogeneity (e.g., heterogeneous coefficients, treatment effects) in panel data and meta-analysis settings. Inference is challenging in such settings: only noisy estimates of heterogeneity are available, and central limit approximations perform poorly in the tails. We derive a necessary and sufficient condition under which noisy estimates are informative about extreme quantiles, along with sufficient rate and moment conditions. Under these conditions, we establish an extreme value theorem and an intermediate order theorem for noisy estimates. These results yield simple optimization-free confidence intervals for extreme quantiles. Simulations show that our confidence intervals have favorable coverage and that the rate conditions matter for the validity of inference. We illustrate the method with an application to firm productivity differences between denser and less dense areas.

This paper studies the long-run economic and institutional consequences of Iran's confrontation with the West, treating the 2006-2007 strategic shift as the onset of a sustained confrontation regime rather than a discrete sanctions episode. Using synthetic control and generalized synthetic control methods, I construct transparent counterfactuals for Iran's post-confrontation trajectory from a donor pool of countries with continuously normalized relations with the West. I find large, persistent losses in real GDP and GDP per capita, accompanied by sharp declines in foreign direct investment, trade integration, and non-oil exports. These economic effects coincide with substantial and durable deterioration in political stability, rule of law, and control of corruption. Magnitude calculations imply cumulative output losses comparable to civil-war settings, despite the absence of internal armed conflict. The results highlight confrontation as a deep and persistent economic and institutional shock, extending the literature beyond short-run sanctions effects to sustained geopolitical isolation.

This paper shows that the mid-20th century was characterised by a considerable reduction in breastfeeding rates, reducing from over 80% in the late 1930s to just over 40% only three decades later. We investigate how maternal breastfeeding during this period has shaped offspring health and human capital outcomes in the UK. We use a within-family design, comparing children who were breastfed to their sibling(s) who were not. Our results show that breastfeeding increases adult height, as well as fluid intelligence, but does not affect educational attainment, nor adult BMI. In further analyses, we examine whether and how this impact varies with individuals' genetic "predisposition" for these outcomes, proxied by the outcome-specific polygenic index. We find that the "height-returns" to breastfeeding are larger among those genetically predisposed to be taller, with no genetic heterogeneity for the other outcomes, though we note that power in the within-family GxE analysis is more limited. Overall, our estimates suggest that breastfeeding plays a non-negligible role in child development.

The governance of common-pool resources-resource systems characterized by high subtractability of yield and difficulty of exclusion-constitutes one of the most persistent and intricate challenges in the fields of economics, ecology, and applied mathematics. This comprehensive review delineates the historical and theoretical evolution of the mathematical frameworks developed to analyze, predict, and manage these systems. We trace the intellectual trajectory from the early, deterministic bioeconomic models of the mid-20th century, which established the fundamental tension between individual profit maximization and collective efficiency, to the contemporary era of complex coupled human-environment system models. Our analysis systematically dissects the formalization of the "Tragedy of the Commons" through the lens of classical cooperative and non-cooperative game theory, examining how the N-person Prisoner's Dilemma and Nash Equilibrium concepts provided the initial, albeit pessimistic, predictive baseline. We subsequently explore the "Ostrom Turn," which necessitated the integration of institutional realism-specifically monitoring, graduated sanctions, and communication-into formal game-theoretic structures. The review further investigates the relaxation of rationality assumptions via evolutionary game theory and behavioral economics, highlighting the destabilizing roles of prospect theory and hyperbolic discounting. Finally, we synthesize recent advances in stochastic differential equations and agent-based computational economics, which capture the critical roles of spatial heterogeneity, noise-induced regime shifts, and early warning signals of collapse. By unifying these diverse mathematical threads, this review elucidates the shifting paradigm from static optimization to dynamic resilience in the management of the commons.

We establish dual attainment for the multimarginal, multi-asset martingale optimal transport (MOT) problem, a fundamental question in the mathematical theory of model-independent pricing and hedging in quantitative finance. Our main result proves the existence of dual optimizers under mild regularity and irreducibility conditions, extending previous duality and attainment results from the classical and two-marginal settings to arbitrary numbers of assets and time periods. This theoretical advance provides a rigorous foundation for robust pricing and hedging of complex, path-dependent financial derivatives. To support our analysis, we present numerical experiments that demonstrate the practical solvability of large-scale discrete MOT problems using the state-of-the-art primal-dual linear programming (PDLP) algorithm. In particular, we solve multi-dimensional (or vectorial) MOT instances arising from the robust pricing of worst-of autocallable options, confirming the accuracy and feasibility of our theoretical results. Our work advances the mathematical understanding of MOT and highlights its relevance for robust financial engineering in high-dimensional and model-uncertain environments.

Structural Equation Modeling (SEM) systematically validated hierarchical pathways among multiple factors by constructing a dual framework integrating latent variable measurement and path analysis, utilizing covariance matrices derived from online questionnaires of Wuliangye consumers in Sichuan Province. Statistical analysis quantified path coefficient significance through maximum likelihood estimation, revealing via factor loadings and goodness-of-fit tests that consumer ethnocentrism directly promotes purchase intention, while simultaneously refuting the null hypothesis regarding perceived behavioral control-thus deconstructing the "trigger-transmission" causal chain among variables. Crucially, SEM findings revealed environmental stimuli as the predominant factor, indirectly influencing purchasing behavior through perceived value, contrary to existing literature asserting equal impacts from consumer ethnocentrism, environmental stimuli, and perceived behavioral control. Statistical evidence further demonstrated higher online purchase frequency for premium Wuliangye liquor, aligning with Generation Z's e-commerce preferences. By implementing stricter website-based participant screening than prior studies, this research optimized the analytical model, yielding data-driven strategic recommendations: strengthening e-commerce platforms, enhancing promotional expertise, leveraging cultural localization, and prioritizing premium product development. These actionable insights significantly advance sales optimization strategies for Wuliangye products in Sichuan's dynamic market.

We provide systematic evidence on the potential for estimating household well-being from mobile phone data. Using data from four countries - Afghanistan, Cote d'Ivoire, Malawi, and Togo - we conduct parallel, standardized machine learning experiments to assess which measures of welfare can be most accurately predicted, which types of phone data are most useful, and how much training data is required. We find that long-term poverty measures such as wealth indices (Pearson's rho = 0.20-0.59) and multidimensional poverty (rho = 0.29-0.57) can be predicted more accurately than consumption (rho = 0.04 - 0.54); transient vulnerability measures like food security and mental health are very difficult to predict. Models using calls and text message behavior are more predictive than those using metadata on mobile internet usage, mobile money transactions, and airtime top-ups. Predictive accuracy improves rapidly through the first 1,000-2,000 training observations, with continued gains beyond 4,500 observations. Model performance depends strongly on sample heterogeneity: nationally-representative samples yield 20-70 percent higher accuracy than urban-only or rural-only samples.

This paper evaluates the causal impact of Generative Artificial Intelligence (GenAI) adoption on productivity and systemic risk in the U.S. banking sector. Using a novel dataset linking SEC 10-Q filings to Federal Reserve regulatory data for 809 financial institutions over 2018--2025, we employ two complementary identification strategies: Dynamic Spatial Durbin Models (DSDM) to capture network spillovers and Synthetic Difference-in-Differences (SDID) for causal inference using the November 2022 ChatGPT release as an exogenous shock. Our findings reveal a striking ``Productivity Paradox'': while DSDM estimates show that AI-adopting banks are high performers ($β> 0$), the causal SDID analysis documents a significant ``Implementation Tax'' -- adopting banks experience a 428-basis-point decline in ROE as they absorb GenAI integration costs. This tax falls disproportionately on smaller institutions, with bottom-quartile banks suffering a 517-basis-point ROE decline compared to 129 basis points for larger banks, suggesting that economies of scale provide significant advantages in AI implementation. Most critically, our DSDM analysis reveals significant positive spillovers ($θ= 0.161$ for ROA, $p < 0.01$; $θ= 0.679$ for ROE, $p < 0.05$), with spillovers among large banks reaching $θ= 3.13$ for ROE, indicating that the U.S. banking system is becoming ``algorithmically coupled.'' This synchronization of AI-driven decision-making creates a new channel for systemic contagion: a technical failure in widely-adopted AI models could trigger correlated shocks across the entire financial network.

We develop an iterative framework for economic measurement that leverages large language models to extract measurement structure directly from survey instruments. The approach maps survey items to a sparse distribution over latent constructs through what we term a soft mapping, aggregates harmonized responses into respondent level sub dimension scores, and disciplines the resulting taxonomy through out of sample incremental validity tests and discriminant validity diagnostics. The framework explicitly integrates iteration into the measurement construction process. Overlap and redundancy diagnostics trigger targeted taxonomy refinement and constrained remapping, ensuring that added measurement flexibility is retained only when it delivers stable out of sample performance gains. Applied to a large scale public employee retirement plan survey, the framework identifies which semantic components contain behavioral signal and clarifies the economic mechanisms, such as beliefs versus constraints, that matter for retirement choices. The methodology provides a portable measurement audit of survey instruments that can guide both empirical analysis and survey design.

The global decline in the labor income share has challenged the classical Kaldor facts; however, the macroeconomic aggregation mechanism -- namely, how aggregate factor shares emerge from firm-level heterogeneity -- remains underexplored. This paper bridges this gap by constructing a theoretical framework that links firm size distribution to aggregate factor shares. We extend Houthakker's aggregation theory and formalize the \textit{weighting effect}: when large firms are systematically more capital-intensive than small firms, a shift in market structure toward larger firms mechanically reduces the aggregate labor share. Using comprehensive firm-level data from Chinese manufacturing (2001--2015), we empirically validate this mechanism. First, we estimate production function parameters and confirm that capital elasticity significantly exceeds labor elasticity, implying a negative relationship between firm size and labor share. Second, we find that the negative effect of firm size on labor share is significant only in industries with high technological heterogeneity. Counterfactual decomposition reveals that the shift in the size distribution toward ``superstar firms'' during 2001--2015 constitutes the primary driver of the labor share decline. Our findings provide a technical micro-foundation for the ``superstar firm'' hypothesis and highlight the distributional consequences of ``winner-take-all'' market structures.

Aligning AI systems with human values remains a fundamental challenge, but does our inability to create perfectly aligned models preclude obtaining the benefits of alignment? We study a strategic setting where a human user interacts with multiple differently misaligned AI agents, none of which are individually well-aligned. Our key insight is that when the users utility lies approximately within the convex hull of the agents utilities, a condition that becomes easier to satisfy as model diversity increases, strategic competition can yield outcomes comparable to interacting with a perfectly aligned model. We model this as a multi-leader Stackelberg game, extending Bayesian persuasion to multi-round conversations between differently informed parties, and prove three results: (1) when perfect alignment would allow the user to learn her Bayes-optimal action, she can also do so in all equilibria under the convex hull condition (2) under weaker assumptions requiring only approximate utility learning, a non-strategic user employing quantal response achieves near-optimal utility in all equilibria and (3) when the user selects the best single AI after an evaluation period, equilibrium guarantees remain near-optimal without further distributional assumptions. We complement the theory with two sets of experiments.

As a key advancement in artificial intelligence, large language models (LLMs) are set to transform transportation systems. While LLMs offer the potential to simulate human travelers in future mixed-autonomy transportation systems, their behavioral fidelity in complex scenarios remains largely unconfirmed by existing research. This study addresses this gap by conducting a comprehensive analysis of the value of travel time (VOT) of three popular LLMs. We employ a full factorial experimental design to systematically examine LLMs' sensitivities to various transportation contexts, including the choice setting, travel purpose, and socio-demographic factors. Our results reveal a high degree of behavioral similarity between LLMs and humans. Some LLMs exhibit an aggregate VOT similar to that of humans, and all tested models demonstrate human-like sensitivity to travel purpose, income, and the time-cost trade-off ratios of the alternatives. Furthermore, the behavioral patterns of LLMs are highly consistent across varied contexts. However, while the behavior of every single model is highly robust, we also find some heterogeneity across models regarding the magnitude and direction of sensitivity to travel contexts and income elasticity. Overall, this study provides a foundational benchmark for the future development of LLMs as proxies for human travelers, demonstrating their robust decision-making capabilities while cautioning that misaligned magnitudes of economic trade-offs between humans and LLMs necessitate rigorous validation and additional conditioning of LLMs before their application.

This work introduces a unified framework for analyzing games in greater depth. In the existing literature, players' strategies are typically assigned scalar values, and equilibrium concepts are used to identify compatible choices. However, this approach neglects the internal structure of players, thereby failing to accurately model observed behaviors. To address this limitation, we propose an abstract definition of a player, consistent with constructions in reinforcement learning. Instead of defining games as external settings, our framework defines them in terms of the players themselves. This offers a language that enables a deeper connection between games and learning. To illustrate the need for this generality, we study a simple two-player game and show that even in basic settings, a sophisticated player may adopt dynamic strategies that cannot be captured by simpler models or compatibility analysis. For a general definition of a player, we discuss natural conditions on its components and define competition through their behavior. In the discrete setting, we consider players whose estimates largely follow the standard framework from the literature. We explore connections to correlated equilibrium and highlight that dynamic programming naturally applies to all estimates. In the mean-field setting, we exploit symmetry to construct explicit examples of equilibria. Finally, we conclude by examining relations to reinforcement learning.

We study the design of one-to-one matching mechanisms that are strategy-proof for both sides and as stable as possible. Motivated by the impossibility result of Roth (1982), we formulate the mechanism design problem as a linear program that minimizes stability violations subject to exact strategy-proofness constraints. We consider both an average-case objective (summing violations over all preference profiles) and a worst-case objective (minimizing the maximum violation across profiles), and we show that imposing anonymity and symmetry when the number of agents in both sides are the same can be done without loss of optimality. Computationally, for small markets our approach yields randomized mechanisms with substantially lower stability violations than randomized sequential dictatorship (RSD); in the $3\times 3$ case the optimum reduces average instability to roughly one third of RSD. For deterministic mechanisms with three students and three schools, we find that any two-sided strategy-proof mechanism has at least two blocking pairs in the worst case and we provide a simple algorithm that attains this bound. Finally, we propose an extension to larger markets and present simulation evidence that, relative to sequential dictatorship (SD), it reduces the number of blocking pairs by about $0.25$ on average.

We develop a theoretical framework to investigate the link between rising scale economies and stagnating productivity. Our model features heterogeneous firms, imperfect competition, and firm selection. We demonstrate that scale economies generated by fixed costs have distinct impacts on aggregate productivity compared to those driven by returns to scale (slope of marginal cost). Using UK data, we estimate long-run increases in both fixed costs and returns to scale. Our model implies that this should increase aggregate productivity through improved firm selection and resource allocation. However, increasing markups can offset the productivity gain. Higher markups cushion low-productivity firms' revenues, allowing them to survive, and constrain firm output, which limits exploitation of scale economies.

In semi-logarithmic regressions, treatment coefficients are often interpreted as approximations of the average treatment effect (ATE) in percentage points. This paper highlights the overlooked bias of this approximation under treatment effect heterogeneity, arising from Jensen's inequality. The issue is particularly relevant for difference-in-differences designs with log-transformed outcomes and staggered treatment adoption, where treatment effects often vary across groups and periods. This paper proposes new estimation and inference methods for an estimand that accounts for heterogeneity across observable subgroups and improves upon conventional measures. The estimand provides a lower bound on the ATE in percentage points, and coincides with it in the absence of within-group heterogeneity. I establish the methods' large-sample properties and study their finite-sample performance through Monte Carlo experiments, which reveal substantial discrepancies between conventional and proposed measures when systematic heterogeneity is large. Two empirical applications further underscore the practical importance of these methods.

We formulate factorial difference-in-differences (FDID), a research design that extends canonical difference-in-differences (DID) to settings in which an event affects all units. In many panel data applications, researchers exploit cross-sectional variation in a baseline factor alongside temporal variation in the event, but the corresponding estimand is often implicit and the justification for applying the DID estimator remains unclear. We frame FDID as a factorial design with two factors, the baseline factor $G$ and the exposure level $Z$, and define effect modification and causal moderation as the associative and causal effects of $G$ on the effect of $Z$, respectively. Under standard DID assumptions of no anticipation and parallel trends, the DID estimator identifies effect modification but not causal moderation. Identifying the latter requires an additional \emph{factorial parallel trends} assumption, that is, mean independence between $G$ and potential outcome trends. We extend the framework to conditionally valid assumptions and regression-based implementations, and further to repeated cross-sectional data and continuous $G$. We demonstrate the framework with an empirical application on the role of social capital in famine relief in China.

Appropriately regulating artificial intelligence is an increasingly urgent and widespread policy challenge. We identify two primary, competing problem. First is a technical deficit: Legislatures and regulatory face significant challenges in rapidly translating conventional command-and-control legal requirements into technical requirements. Second is a democratic deficit: Over-reliance on industry to provide technical standards fails to ensure that the many values-based decisions that must be made to shape AI development and deployment are made by democratically accountable public, not private, actors. We propose a solution: regulatory markets, in which governments require the targets of regulation to purchase regulatory services from a government-licensed private regulator. This approach to AI regulation could overcome the limitations of both command-and-control regulation and excessive delegation to industry. Regulatory markets could enable governments to establish policy priorities for the regulation of AI while relying on market forces and industry R&D efforts to pioneer the technical methods of regulation that best achieve policymakers' stated objectives.