Economics

We use 70 million policies linked to mortgages and property-level disaster risk to show that credit scores impact homeowners insurance premiums as much as disaster risk. Homeowners with low credit pay 24% more for identical coverage than high–credit score homeowners. Leveraging a natural experiment in Washington State, we find that banning the use of credit information considerably weakens the relationship between credit score and pricing. We discuss the role of credit information in pricing and show that, although insurance is often overlooked in discussions of home affordability, a low credit score increases premiums roughly as much as it raises mortgage rates.

Financial incentives are widely used to encourage beneficial behaviors, but their effectiveness may be limited by inattention and imperfect memory. We study this in a randomized trial of HIV medication adherence in Mozambique. Financial incentives alone increase adherence by 10.6 percentage points, while pairing incentives with reminders increases adherence by 24.3 percentage points. We develop a model in which inattention to daily adherence and imperfect memory of payment eligibility reduce incentive effectiveness and show that reminders mitigate both frictions. Detailed medication refill data support the model’s predictions. The results suggest combining incentives with reminders can sub..

Prepaid consumption is a common feature of modern consumer markets and is often presented as a mutually beneficial arrangement: consumers receive upfront discounts, and firms secure future sales. We analyze a large-scale Pay Now, Buy Later (PNBL) program in which consumers prepay for restaurant credit with bonuses, and spend the balance later. Using detailed transaction data from over 4 million consumers, we document widespread balance breakage: approximately 40% of prepaid value is never used. Because many consumers underutilize their balances, merchants recover significantly more than the bonus cost. The median firm earns roughly $5.5 in breakage profit for every $1 of bonus credit issued...

As Large Language Models (LLMs) are increasingly tasked with autonomous decision making, understanding their behavior in strategic settings is crucial. We investigate the choices of various LLMs in the Ultimatum Game, a setting where human behavior notably deviates from theoretical rationality. We conduct experiments varying the stake size and the nature of the opponent (Human vs. AI) across both Proposer and Responder roles. Three key results emerge. First, LLM behavior is heterogeneous but predictable when conditioning on stake size and player types. Second, while some models approximate the rational benchmark and others mimic human social preferences, a distinct “altruistic” mode emer..

Incomplete contract theory, supported by anecdotal evidence, suggests that when a firm is acquired, workers may be adversely affected in non-contractible aspects of their work experience. This paper empirically investigates this prediction by combining M\&A events from the Refinitiv database and web-scraped Glassdoor review data. We find that: (a) Controlling for pre-trends, mergers lead to lower satisfaction, especially on non-contractible dimensions of the employee experience (about 6% of a standard deviation); (b) The effect is stronger in the target firm than in the acquiring firm; (c) Text analysis of employee comments indicates that the decline in satisfaction is primarily associated w..

We estimate the market value of reproductive rights as capitalized into U.S. housing markets. We do so using a synthetic difference-in-differences design to evaluate the effects of total abortion bans following the 2022 Dobbs decision, and drawing on housing market indices from Zillow and vacancy rate data from the U.S. Census Bureau's Housing Vacancy Survey. The results indicate that total abortion bans reduced rents by an average of 2.2% from July 2022 through June 2025, with the effect reaching 4.0% in the most recent year. Over the same horizon, bans increased rental vacancy rates by an average of 1.1 percentage points, with the effect reaching 1.8 percentage points in the most recent ye..

Sectoral shifts require costly labor reallocation for workers, fueling concerns about how quickly they occur. We study how the pace of such sectoral shifts affects workers at risk of displacement. We develop a life-cycle model with skill heterogeneity and job ladders where labor demand gradually rises in one sector and declines in another. The model reveals three novel insights. First, workers’ lifetime earnings are strongly non-linear and even nonmonotonic in the horizon over which the transition unfolds. Second, more and more workers benefit on the extensive margin as the transition accelerates, but the tail of losses becomes thicker on the intensive margin. Third, labor market frictions..

We show that, to a first-order approximation, estimates of fluctuations in Shiller’s fundamental price relative to observed price depend primarily on forecasts of long-horizon expected returns. Researchers using different measures of cash flow and valuation may reach different conclusions about the extent to which values fluctuate excessively relative to fundamentals, but that is only because return forecasts based on different cash-flow-to-value measures will be different. Using U.S. equity data, we demonstrate that the amount of persistence in expected returns, rather than the amount of short-run return predictability, is the key determinant of implied excess volatility. Disagreements ab..

The Bureau of Labor Statistics Current Employment Statistics (CES) is a high-quality, closely watched indicator of labor market health and the business cycle. Recent revisions to CES data have garnered significant attention. I provide a primer on CES methodology and analyze recent survey response and data revision patterns. CES revisions actually tend to be small, but they have increased in recent years. I discuss potential ideas for improving the CES including expanding the data sample with internal or external data, adjusting the estimates with benchmark data more often, and enhancing estimation of establishment birth and death.

We find that randomly assigned peers play a sizable and unique role in shaping political economy. Closely seated, and exogenously assigned, US Senate peers have a significant impact on Congressional voting, shifting votes by 11.9 percentage points (t=7.34). Physical distance is the largest and most consistent of any characteristic outside of party or state in impacting voting behavior. The distance effect is concentrated in the closest peers, existing for up to 19.6 feet on the Senate floor, then dissipating. Close peers additionally increase the probability of aisle-crossing (voting with the opposite party), with the aisle-crossing impact being roughly eight times larger on the final votes ..

We develop a conceptual framework of political repression. To motivate why states repress, we introduce the notion of the political project. The proposed framework is used to interpret two known stylized facts: 1) repression is higher in autocracies; and 2) repression has declined since the 1990s. We discuss under-researched aspects of political repression such as migration restrictions, the targeting of repression, and backlash. We conclude by bringing attention to the two main challenges for research on repression: 1) conceptualizing whether a given episode of repression is successful, and 2) the inherent selection problem in empirical measures of repression.

As workforces in high-income countries age and shrink, immigrants increasingly fill entry-level, low-skilled jobs. We examine what happens when this labor supply is abruptly reduced, exploiting South Korea’s sudden suspension of its low-skilled guest worker program following the 2020 COVID-19 border closure. Using policy-driven variation in firms’ pre-pandemic reliance on immigrant labor, we show that the collapse in inflows led to a significant increase in firm exit. Among surviving firms, greater pre-pandemic dependence on immigrant workers resulted in production disruptions and operational delays. Firms did not respond by expanding domestic hiring to replace missing guest workers. Ins..

The United States relies primarily on private health insurance markets, yet these markets are highly concentrated and becoming more so over time. We document concentration across commercial, Medicare Advantage, and Medicaid markets. We then examine how asymmetric information—particularly adverse selection—interacts with market power to shape premiums, plan design, and consumer welfare. Empirical evidence confirms that insurer consolidation raises premiums. We discuss how antitrust enforcement, risk adjustment, regulation, and informational interventions shape competition and consumer welfare in these markets.

Human-driven degradation of ecosystems threatens both global biodiversity and the livelihoods of communities that depend on natural systems for production and income. Yet empirical evidence on the economic returns to restoring nature remains scarce. We examine whether nature-enhancing interventions can generate measurable improvements in household wealth, in addition to environmental benefits. We study a large-scale agroecological intervention implemented by Trees for the Future (TREES) across farms in sub-Saharan Africa, combining detailed household surveys with high-resolution satellite imagery. We show that quasi-exogenous increases in natural capital lead to substantial gains in househol..

This paper reassesses the role of financial frictions in capital misallocation through a model disciplined by both firm-level borrowing costs and the average revenue product of capital (ARPK). Using Chinese manufacturing data, we document substantial dispersion in ARPK, alongside a strong positive relationship between ARPK and the borrowing costs firms face---patterns absent in U.S. data. We develop a heterogeneous-firm model with endogenous firm-specific borrowing costs and additional capital distortions modeled as exogenous wedges. In this model, eliminating financial frictions raises total factor productivity (TFP) by 25 percentage points. In contrast, without other capital distortions, r..

We study how teachers' rule violations in grading affect students' ethical behavior. Using administrative data from high-stakes exams, combining teacher-assigned internal scores with externally graded national exam scores, we track teacher grading violations and subsequent student cheating. We explore three potential mechanisms: imitation (learning that rules can be broken), positive reciprocity (responding favorably to favorable treatment), and negative reciprocity (retaliating against unfavorable treatment). Exploiting within-student variation in exposure to different teachers, we find students are significantly more likely to cheat when teachers break the rules to their detriment (systema..

We combine comprehensive data on the rulemaking activities of the U.S. federal government with individual-level personnel and voter registration records to study delegation and principal-agent frictions in the development of new regulations. We present three main results. First, even important pieces of new regulation are frequently delegated to career bureaucrats who are politically misaligned with the president. Second, rules that are overseen by misaligned regulators take systematically longer to complete, are more verbose, generate more negative feedback from the public, and are more likely to be challenged in court. Third, in assigning regulators to rules, agency leaders often face a sh..

Almost a quarter-billion calls are placed to 911 each year in the United States. A large share of them involve social problems, not crimes or emergencies---yet police are dispatched in response. This review traces how the 911 emergency system's institutional design shapes demand for police, who is excluded from or ill served by this system, and what alternatives exist, including nonemergency lines (with police response), government hotlines (211, 311, 988), civilian crisis teams, and community-based resources. Among the universe of municipal police departments with at least 100 sworn officers in 2020, covering 107 million US residents, police have absorbed broad social service functions, wit..

This paper examines how adopting an automated review monitoring system (ARMS) allows restaurants to translate consumer feedback into operational quality improvements. ARMS provides automated alerts for negative reviews and facilitates back-end work ticket management, reducing the costs of identifying, prioritizing, and addressing issues revealed by customer feedback. Using restaurant-level data on ARMS adoption and consumer reviews, we find that adoption increases average star ratings, lowers the share of negative reviews, and raises the positivity of review text. These improvements are not driven by strategic manipulation; they are greater for restaurants with lower initial ratings and in d..

Can protectionism revive domestic production, slow automation, and help routine workers? We address this question in a dynamic open-economy model with heterogeneous firms, endogenous entry, and task-based production in which routine tasks can be performed by workers or robots. Import tariffs reallocate demand toward domestic goods, reshape markups and entry incentives, and generate fiscal revenues rebated to households. As a result, tariffs raise GDP and consumption measured at market prices and temporarily slow automation, even though intermediate output at factor prices and trade volumes decline. The gains are unevenly distributed: routine workers benefit robustly through transfers and red..

We examine two approaches to improving urban school systems: changing who gets to go to existing schools (reallocation) and restructuring school portfolios through closures and reconstitution (resource augmentation). Using data from New York City high schools, we estimate models of school effects allowing for both vertical school quality differences and horizontal student-specific match effects. While sophisticated reallocation policies that optimize student-school matches can generate modest educational gains, they are constrained by limited seats at highly effective schools. Simple resource-augmentation policies targeting replacement of low-performing schools achieve comparable improvement..

Useful social science theories predict behavior across settings. However, applying a theory to make predictions in new settings is challenging: rarely can it be done without ad hoc modifications to account for setting-specific factors. We argue that AI agents put in simulations of those novel settings offer an alternative for applying theory, requiring minimal or no modifications. We present an approach for building such "general" agents that use theory-grounded natural language instructions, existing empirical data, and knowledge acquired by the underlying AI during training. To demonstrate the approach in settings where no data from that data-generating process exists--as is often the case..

The use of federal anti-fraud laws to address unnecessary medical care is controversial. Targeted providers frequently argue that, in their judgment, the treatment in question was appropriate. We examine the effects of anti-fraud litigation against hospitals for over-admitting patients from the emergency department, using 100% Medicare claims for 2005-2019 and a design based on the staggered rollout of these lawsuits. We find that anti-fraud lawsuits reduced admission rates by 3.6 percentage points without increasing mortality rates. We estimate five-year savings to Medicare of $1.3 billion. Our results suggest that anti-fraud enforcement can be successful in reducing costly, unnecessary car..

This paper shows that the pace of technology creation is a key driver of the skill premium. It develops a model in which skilled workers have a comparative advantage in learning new technologies. As technologies age, they become standardized and accessible to other workers. The skill premium is determined by the interplay between the pace of technology creation and standardization. A rapid pace of technology creation leads to a sustained increase in the skill premium. We calibrate the model using novel text-based data on new technologies and their changing demand for skills as they age. These data show that new technologies are initially skill intensive but become less so as they age. The da..