Papers

Firms follow changing economic opportunities, but rivalry changes their response. We develop ESCAPE, a rational-share game of distribution-valued positioning with heterogeneous capability costs, establish a unique equilibrium, and derive an exact reallocation restriction separating opportunity and crowding contributions. Competition need not push firms apart: rational firms can enter increasingly crowded regions, because the full reallocation gain can be positive even when its crowding component is negative. Rivalry instead changes how differently firms respond to the same opportunity shift. In the benchmark many-firm economy, the relative-response component accounts for 90.5% of the weighted squared composition-response contrast between strategic and independent firms; it is numerically zero under identical capabilities. The contrast remains distinguishable under corpus-scale measurement. Corporate news distributions document persistent but changing peer structure: optimized peer reconstructions retain 86.5% of their aggregate validation advantage one year later even as absolute distance to both frozen reconstructions increases. Peer structure is persistent without being a fixed destination, so current similarity can remain informative without fixing firms' future relative positions.

Classification systems shape how searchers find relevant objects. We study how revisions to classification architecture affect retrieval by changing the routes through which objects enter consideration. We theorize that creating a cross-cutting category can reduce classificatory distance and translation costs, whereas adding routes to an established system can increase route-selection and updating costs; category cleanup can restore contrast. Empirically, we exploit the 2013 creation and 2020 revision of Y02/Y04 green-technology codes in the Cooperative Patent Classification. We match treated patents to unchanged green patents within technology-cohort cells and estimate examiner- and applicant-specific effects on first-citation timing and citation volume. Category creation is associated with a 25% higher examiner first-citation hazard. In the mature system, additions are associated with 13-15% lower examiner hazards, while cleanup is associated with a 14% higher hazard and 12% higher citation volume. Effects are generally stronger for examiners, whose search task is more tightly coupled to patent classification. Timing and volume need not move together: creation and green-domain entry shift timing without detectable changes in volume, while cleanup increases both speed and recorded reach. Classification infrastructure may therefore change when relevant objects are found even when overall recorded attention is unchanged, and greater classificatory detail need not improve search efficiency.

This paper investigates spatial productivity convergence across Chinese provinces during structural transformation, examining how a shift in auonomous demand composition reshaped regional productivity dynamics. Using Panel Structural Vector Autoregressive modelling and province level data from 2001 to 2021, we analyse eastern, central, ad western regions across two sub-periods through the augmented Kaldor-Verdoorn law. A post-2011 spatial reversal emerges: inland regions exhibit stronger Verdoorn coefficients, while fiscal multipliers converge across coastal and inland areas, validating China's redistributive fiscal policies in narrowing regional productivity gaps. The Kaldor-Verdoorn coefficient remains robust throughout, while export effects on productivity collapse and government expenditure effects strenghten.

Before a flood, governments decide where to invest in protection; after it, how much of the loss to compensate. The European Union informs the first decision with a composite index that ranks every subnational unit by vulnerability. This paper tests the ranking by comparing Italian provinces struck by the same flood between 2005 and 2024. The published index does not order provinces by the losses they suffer, with or without controls for hazard, exposure and income ($\rho$ between -0.06 and 0.09). The failure lies in what it weights. As published, the index mostly measures economic development, while losses follow water: they arrive in recurring groups of provinces that share a river basin, and more than half of the ranking error lies between those groups. In a pre-registered test that rebuilds it and eight rivals at every forecast origin from 2010 to 2020, the index comes last on both questions an allocator asks. On where damage will land, its top ten provinces capture 16 per cent of later damage, less than a random draw; Italy's public flood hazard map, blind to losses, captures 44. On who loses more from the same water, its rank correlation with later losses is 0.03; its own components, reweighted on past losses, reach 0.42. Neither repair needs new data. The misranked loss is paid locally. Of EUR 12.3 billion recorded over 2013-2022, 85 per cent was not covered by national transfers, and prevention funding, scored the same way, does no better than chance. Because correlated losses cluster within borders, national pools in Europe need 39 per cent more reserve than cross-border pools of the same size. The index fails the same way across 1,142 European units, where the repairs are weaker.

Applied researchers increasingly convert images into features with pretrained encoders, then use those features in a downstream prediction model. The encoder is often treated as an implementation detail. We show that it can instead be a consequential source of model uncertainty. We call this uncertainty representation risk: plausible pretrained encoders map the same images into different feature spaces and can yield sharply different out-of-sample conclusions from predictive performance. We compare ten modern and legacy frozen encoders across applications involving house prices, racehorse performance, breast-cancer histology, chest radiographs, continuous facial age, and rice disease. With common dimension control, heads, and group-safe splits, validation selects SigLIP 2 for houses, raising test $R^2$ from 0.396 for ResNet50 to 0.629, and DINOv2 for horses, raising $R^2$ from 0.029 to 0.105. No encoder is best in every task. Candidate procedures are constructed using training data and compared on a separate validation partition. The selected procedure reaches 0.658 for houses and 0.979 accuracy for pneumonia. Fixed-split gains are small for horses and rice, while repeated partitions reveal instability in horse feature union. Continuous age selects SigLIP 2 at 4.786 years MAE. The principal representation gaps persist with neural heads, similarly sized DINOv2 and ViT models, and limited adaptation. These results support a simple workflow: benchmark plausible representations, select on locked validation data, combine only when separate validation evidence justifies the additional cost, and report paired and split-level uncertainty. We implement this workflow in LOOKAGAIN-ML, the software package used to conduct the analyses in this paper.

Refugee integration is a central policy challenge for host countries, and where governments initially place refugees shapes their integration trajectories. Yet placement officers often have limited information about where each case is most likely to succeed. Algorithmic refugee matching uses administrative data, machine learning, and constrained optimization to recommend employment-optimized placements in real time as cases arrive, with human placement officers retaining final authority. Between January 2020 and June 2023, the Swiss State Secretariat for Migration randomly assigned about 2,000 refugee cases to receive a canton recommendation either algorithmically optimized for employment or drawn to approximate existing procedures, with placement officers and refugees blinded to assignment. The two arms used identical but separate canton and origin-group quotas, so gains reflect better refugee-canton matching rather than reallocation toward stronger labor markets. The trial began just before the COVID-19 pandemic shifted labor-market conditions. For the pre-registered primary outcome – the share of months employed during the first three years – the pooled intention-to-treat (ITT) estimate across the 2020-2023 placement cohorts was +2.2 percentage points (about 10% of the 22.3% control mean; 95% CI [+0.05, +4.33]), rising to +3.9 pp (about 17%; [+1.11, +6.68]) for the post-COVID 2022-2023 cohorts. Effects grew over time: at 36 months, the pooled ITT on the employment rate was +5.2 pp (about 11%; 95% CI [+1.10, +9.25]) – comparable to the gains from hundreds of hours of intensive language training. Overall, the results provide rare field evidence that AI-based decision support can improve high-stakes public-sector allocation, offering a scalable, low-cost way to raise refugee employment.

Existing studies consistently find that home batteries and electric vehicles optimized against real-time electricity prices create new load peaks in local distribution grids, because all assets respond to the same price signal in sync. However, all but one of the 18 studies reviewed here treat wholesale prices as exogenous. This paper argues that treating prices as non-responsive biases the result: in reality, charging in low-price hours raises the wholesale price, which dampens the incentive to charge. I test this by simulating households with rooftop solar and home batteries under a spot-based retail tariff, calibrated to German data for 2025 with a model of equilibrium wholesale prices. With exogenous prices, my results confirm the finding from the previous literature: synchronous charging pushes the coincident import peak of the local grid to 83% above its no-battery level. With endogenous prices, the same fleet leaves the peak 6% below the no-battery benchmark. In other words, treating prices as endogenous reduces the coincident peak at high battery penetration nearly by half. This finding is robust across alternative price functions and exogenous price paths, different measures of the coincident peak, 25 scenario variations and three historical years. I conclude that new local peaks remain possible, particularly in grids where battery deployment runs ahead of the national average, but the risk and magnitude are considerably smaller than the existing literature suggests.

This paper develops a theory of accommodation traps in workplace hierarchies. Workers with weak outside options face a higher relational cost of appearing unavailable, resistant, or difficult, and so adopt accommodative communication (extra deference, softened boundaries, visible flexibility) to preserve the employment relationship. This short-run strategy is also an informative signal: a friction-minimizing supervisor rationally infers that an accommodating worker is less likely to resist, and assigns that worker the invisible, low-promotability work. The signal harms the worker who sends it. Unlike agents in ratchet-effect models, who conceal information that would invite heavier demands, dependent workers reveal their low refusal cost, because the relational value of accommodating rises with dependence. The resulting inequality requires no productivity difference and no taste for discrimination, and the supervisor's beliefs are correct, so correcting beliefs cannot remove it. When visible work builds outside options and invisible work erodes them, the allocation becomes self-reinforcing, and even workers who start out identical are locked into unequal roles with probability one. The equilibrium is privately rational for every party yet socially inefficient whenever diminishing returns to outside options outweigh the supervisor's friction cost. Attaching recognition to invisible work, rotating it by rule, or limiting managerial discretion each weakens or breaks the link from communication to assignment that sustains the trap.

A small share of online creators earn fortunes while most earn almost nothing. What separates them: skill, luck, or copying what works? This paper measures what YouTube vloggers actually produce: 3,728 thumbnails, 385 hours of speech (2.9 million words), and video titles and tags, linked to complete view histories of 56,419 videos. Its central sample follows channels from their first upload, including the many that never took off, which most creator statistics miss. The data show that what goes into a video does not predict its success: 36 measures of images, speech and text explain at most 1.1% of the differences in views within a channel, even though they can tell what kind of video it is. Practice does not help either: forty more videos without a hit add nothing. Nothing accumulates except a hit. A channel's first hit raises the views of everything it publishes afterwards by 164% on average, a result that survives all 17 robustness checks. Creators respond by copying their hit, both its look and its topic, yet only copying the topic pays; copying the look adds nothing measurable. A simple model explains why. A hit is a lasting prize, so entering can make sense despite long odds, and because the platform's statistics describe a hit without separating its causes, copying everything about it is the sensible response. Platforms could redirect creators' effort toward what pays by showing whether a hit's topic or its look drove its success.

Online comment sections let a small number of vocal individuals reach far beyond their own networks. We conduct a large-scale field experiment on Facebook that randomizes the presence and stance of comments beneath posts for a racial justice organization, reaching around one million U.S. users. Opposing comments increase reactions, comments, and link clicks by 15-43 percent relative to no comments, whereas supportive comments have little effect. A complementary survey experiment shows that similar opposing comments make attitudes less progressive and reduce donations to the organization. Through a common feature of online platforms, the vocal few can exert outsized influence.

We study gaming and adaptive incentive design in AI adoption. A large medical-device company required roughly 5,000 employees to submit at least 200 queries per month to an internal AI assistant. First-time use increased significantly after the mandate, but usage patterns suggested gaming: query counts bunched at the threshold, and 31 percent of queries were repeated or off-task. The firm subsequently revised incentive design and lowered the target to 100. Using staggered implementation across branches, we estimate that the adjustment reduced query volume by 30 percent, with repeated or off-task queries accounting for about 90 percent of the decline. The estimated change in non-repeated, work-related queries was small and statistically insignificant. Among sales employees, monthly sales increased by 7 percent. The findings identify incentive adaptation as a consequential margin of technology adoption: recalibrating usage requirements can reduce gaming and improve employee performance.

Contemporary fiscal architectures rely overwhelmingly on human labor income and payroll withholdings to finance social insurance and public infrastructure. The rapid diffusion of autonomous software agents, generative foundation models, and embodied robotic systems decouples output growth from human work hours, eroding traditional tax bases while exacerbating capital-income concentration. Prevailing policy proposals, ranging from uniform robot levies to unconditional cash transfers, fail to resolve three foundational challenges: attributing economic value across mixed human-machine workflows, establishing legal fiscal liability without creating fictional machine personhood, and building an equitable distribution mechanism that converts technological surplus into durable public wealth. This paper presents CivicDividendOS, an end-to-end computational fiscal framework for mixed human-AIrobot economies. The architecture introduces an Autonomous Economic Activity Passport (AEAP) that anchors machine activity to accountable corporate beneficiaries, paired with a Factor-Origin Ledger (FOL) that calculates causal marginal contributions across labor, AI services, robotics, data, and conventional capital. Drawing on an endogenous Substitution-Augmentation Classifier, the system levies a dynamic Social Automation Contribution (SAC) on automated value added, offering explicit offset credits for worker upskilling, augmentation, and task innovation. Accrued revenues are recycled through a Human Earned-Income Shield (HEIS) to reduce payroll wedges and a Public Automation Wealth Fund (PAWF) whose structural yields endow citizen dividends, transition insurance, and human capabilities. We formalize the mathematical architecture, illustrate its operation in an urban logistics setting, and outline an evaluation protocol via a macroeconomic policy digital twin.

Many countries require factories to recycle almost all their wastewater and return none to rivers, a standard called zero liquid discharge (ZLD); in India, courts have imposed it on textile and leather clusters. Yet when enforcement tightens, small firms close while larger polluters often endure. This paper asks why, comparing Tirupur, where closures have already occurred, with Kanpur's Jajmau tanneries, where they are ongoing. Small firms depend on shared treatment plants they do not control. As these plants become underused, costs rise, performance declines, and firms with in-house treatment are better able to withstand enforcement. Because compliance is judged by equipment ownership rather than measured discharge, plant failure is attributed to member firms. This final step is the best evidenced, while the earlier links remain suggestive. The paper recommends judging compliance by measured discharge and determining whether the plant or the firm has failed before closure, so enforcement targets pollution rather than the firms least able to absorb it.

Many economic decisions rely on observations that predate the decision time. This paper introduces information latency, the elapsed time between a decision and the most recent observation of an evolving payoff-relevant state, as a distinct source of imperfect information. Unlike classical sources of information friction, latency generates irreducible uncertainty even with exact observations and rational expectations. For sufficiently regular continuous-time Markov processes, conditional variance increases locally with latency; under the Ornstein-Uhlenbeck benchmark, it increases at a decreasing rate toward a finite bound. Following a rare-state observation, however, uncertainty can peak at an intermediate latency before declining. Neither the average latency nor the average update frequency fully characterizes information quality: the regularity of information arrival matters independently of the mean interval. Under a CARA-normal benchmark, latency-induced uncertainty generates a risk premium. With heterogeneous latency, the lowest-risk-adjusted-cost intermediary wins the business, while competition drives the price toward the second-lowest risk-adjusted cost. In a stylized contract-renewal setting, positive latency can be second-best optimal by limiting costly reclassification when direct contractual commitment is unavailable. The framework applies to credit, insurance, financial markets, and operational decisions.

Organizations may perceive disorder in their environment as a threat which can impede the attainment of desired goals. In such a situation, organizations may become single minded, obsessed, and rely on brute force to attempt to bend the external environment to their will. These efforts may be enabled by perceived power imbalances between the organization and other actors in the environment, but they may also be stymied by the complexity of other organizations in the institutional field. This paper examines such an example in Soviet Kazakhstan from 1929 to 1933, when the All Union Communist Party pursued a policy of sedentarization of the nomadic people. Forcibly migrating nomads to collective farms under a brutal policy of repression, the Soviet bureaucracy at all levels saw no policy as off-limits to achieving its goals but continued to run up against resistance from the complex organizational forms inherent in Kazakh nomadism. Using archival documents, testimonies from Kazakhs, and arrest records from this time period, I show just exactly how the Soviet apparatus pulled the devil from the machine and how, ultimately, the machines were doomed to failure. However, the human cost of forced migration continues to linger.

Numerical inflation targets anchor beliefs. Across euro-area and US professional forecasts, inflation swaps and options, and realized inflation, uncertainty about inflation is compressed at the announced number and kinks exactly there. This paper identifies a cost of the same design that, to our knowledge, has not been shown before, and that appears in second moments only. Within the workhorse New Keynesian model, tolerating part of the inflation a supply shock produces is optimal, yet the optimal tolerated share is not identified: optimal look-through and an unwarranted drift of the effective target are observationally equivalent in the inflation history, so a central bank cannot demonstrate that a warranted deviation is warranted, ex post as much as in real time. Agents holding finite, heterogeneous patience then generate predictive variance that is flat below the target and rises linearly with the expected overshoot above it – an observational-equivalence bill, zero at the announced number and accumulating with the point-years inflation spends above it. The first-order benefit of the number stands; what the cost changes is how inflation uncertainty must be measured. The distance from target explains 70% of the variation in professional forecast variance, and the bill lies within that component; Normalized Uncertainty – to our knowledge the first such correction – removes it. The purge changes inference substantially: on French loan-level data, raw dispersion is unrelated to corporate loan rates, while one standard deviation of the purged measure is associated with rates 89 basis points higher, and cross-country growth and time-series results shift similarly. Conventional measures of inflation uncertainty partly record inflation's distance from its anchor rather than uncertainty about the outlook.

LLMs increasingly act as purchasing agents, which makes the LLM, not the user, the one choosing among the options that satisfy a request; its preferences quietly fix what gets bought and what it costs. Hotel booking is a clean instance: a high-volume choice settled on a few comparable attributes, where the pick reveals those preferences. We introduce PriceBench, a diagnostic benchmark that recovers an LLM's price, quality, and brand preferences from its booking choices with a logit choice model, applied to 28 LLMs from 8 providers on 3,600 hotel tasks from 179 real New York City properties. We find that capability is associated with how consistently an LLM chooses, not with what it chooses: more capable LLMs hold stronger, more consistent preferences, while weaker ones either lock onto one position, exploitable by whoever controls listing order, or choose almost indifferently. What those preferences favor varies sharply across providers and even within one family: price sensitivity spans more than an order of magnitude, and the price/quality trade-off moves mean booked nightly price from \$247 to \$393 on identical tasks. What an agent buys must therefore be measured per LLM, not inferred, and we release the tasks, code, and all 28 response sets.

The return migration of skilled workers provides origin countries many important benefits. The likelihood of return migration is thought to depend on relative economic opportunities, but quantitative evidence of its relative importance remains limited. Here we measure return migration rates using the public profiles of software developers on GitHub, geocoded across ten waves between 2012 and 2026, in a panel containing more than 270,000 international movers. Within five years of departure, 8.8% of movers have returned. Origin-country rates range from under 1% to 16%. Hazard models with corridor fixed effects show that migrants whose home countries experience greater catch-up growth return at significantly higher rates. Immutable features of country pairs such as cultural or geographic distance do not explain the estimate. We forecast future return rates using IMF growth projections, predicting, for example, that India's return rate will rise 12% by 2035 and China's 7%. The difference is driven mostly by India's faster projected growth at home. As developing economies close income gaps with the destinations their emigrants have settled in, they recover more of the workers they lost.

A production network can remain largely unidentified even when the economic decision it supports is identified. We characterize sufficient measurements for exposure-based decisions and compute sharp maximum regret over networks consistent with released totals. Using earlier and later vintages of Japan's interregional input-output accounts, we select measurements from the 1995 table and evaluate the frozen design against the 2005 benchmark. At roughly half the statistics required for full disclosure, the resulting monitoring set loses only 0.07 percentage points of average exposure relative to the benchmark optimum, yet its sharp maximum regret across compatible networks is 4.83 points. In U.S. coal deliveries surrounding a 2005 Wyoming rail disruption, additional shipment measurements identify the optimal set of plants to monitor for inventory risk, even though four monitored plants' exposures to the affected coal supply remain unidentified. The results distinguish good benchmark performance from a decision guarantee and show that decisions can be identified before individual exposures. They suggest evaluating network data by the economic decisions they support.

In the last two decades, last-mile delivery (LMD) firms have seen immense growth fueled by the success of e-commerce, leading to faster and cheaper deliveries. Operating on thin margins, LMD firms strive for successful first-time deliveries to avoid the financial and reputational costs of reattempts. Delivery Agents (DAs) are integral to LMD efficiency, influencing customer experience, delivery success, and productivity. However, most LMD performance enhancement research focuses on process, technology, and incentives, which presume workers will conform to procedures and monitoring tools will function flawlessly. Nevertheless, in practice, DAs deviate from expected behaviors, i.e., indulge in misconduct, negatively affecting delivery efficiency, often resulting in returned parcels. One of the major misconducts is fake remarked deliveries, wherein DAs intentionally do not deliver the parcels and provide a fake reason for it. For instance, even without reaching a delivery address, a DA remarks 'customer unavailable' and records a delivery failure. In this study, we collaborated with a leading Indian LMD firm and, using instrumental variable regression, find that such misconduct leads to a spillover productivity loss. This effect reduces the next day's successful deliveries by 1.60% and first-time-right deliveries by 1.86%. We discuss misconduct's correlation with factors such as task complexity and offer novel insights into how opportunistic circumstances can influence worker behavior.

Misinformation is widely reported to propagate faster on engagement-based platforms, yet prior work largely focused on empirical analysis, without identifying a specific algorithmic mechanism that results in this phenomenon. Thanks to the open-sourcing of X's recommendation algorithms, we conduct what is, to our knowledge, the first component-level study of the recommendation algorithm deployed by a social media platform, which examines how each of its components affects misinformation propagation. Specifically, we identify the engagement fungibility mechanism in the algorithm, where the final recommendation score is constructed as a weighted sum of all predicted user activities. As a result, a tweet can be repeatedly recommended simply because it is predicted to draw many instant reactions (e.g., likes and retweets), even when it is not expected to draw thoughtful responses (e.g., replies and quotes). Since misinformation typically draws a larger share of its engagement from instant reactions, this mechanism enables it to receive more recommendation exposure and to propagate faster. To empirically validate this mechanism, we re-implement X's recommendation algorithm on the USC X 2024 election corpus, and build a calibrated simulation study to analyze the impact of different scoring rules. We find that re-tuning the metric weights has little or even a negative impact on reducing the credibility exposure gap, while those scoring rules that set a precondition of thoughtful engagement for amplification would be able to alleviate the gap significantly, across 46 robustness checks. Our diagnosis, therefore, yields a simple and deployable fix, a reflective-threshold gate that withholds amplification until a tweet is predicted to draw thoughtful engagement, which we find to reallocate exposure away from low-credibility content at no cost to mainstream exposure and with no loss of engagement.

Generative AI makes a paradox newly visible: learning from a common source can improve what each firm does today while narrowing the variety available for tomorrow's discoveries. The tension extends beyond AI. Fields turn recurring features of successful organizations into portable practices, such as just-in-time production, freemium business models, or mixture-of-experts architectures. These collective abstractions are partial templates others can adapt, not designs or universal prescriptions. We study how they shape discovery in a model of firms searching interdependent landscapes. A collective repertoire extracts practices from patterns shared by leading configurations; firms vary in how they evaluate those practices locally. The model reveals a generativity chain: population diversity supplies heterogeneous experience, collective abstraction turns it into reusable options, and situated judgment matches options to local needs. Collective learning is not inherently homogenizing. Copying leaders consumes diversity quickly, whereas repertoires paired with local judgment preserve generative capacity. Environmental disruption exposes the consequences: it devalues accumulated knowledge but reopens local search, helping homogeneous, exhausted industries while harming diverse industries whose repertoire remains useful, even though diverse industries still perform better overall. Recurring practices may also outlive their usefulness, making popularity a poor test of validity after change. These findings matter for organizations using generative AI, best-practice communities, and research fields converging on dominant strands. Generativity requires governing the diversity feeding knowledge and the judgment applied to it. Learning systems should be judged not only by the answers they spread today, but by whether they preserve the differences from which tomorrow's answers can be made.

Generative artificial intelligence illustrates a broader organizational puzzle: young, resource-constrained firms can build on knowledge produced across a wider field to address problems that exceed their internal experience. We theorize one mechanism as social imitation: a distributed and recursive process through which communities observe practices and outcomes across organizations, distill recurrent elements into portable strategies, circulate them, and revise the collective repertoire as adoption produces new evidence. This perspective endogenizes the object of imitation. Organizations do not merely copy practices; communities collectively produce what becomes available to copy. We distinguish the source of a candidate (directly observed organizations or a collective) from its mode of adoption (blind or informed), and examine four forms of imitation in an agent-based model of organizational search. Collective sourcing alone does not improve performance. Its value depends on situated evaluation: assessing whether a candidate fits the adopter's configuration. As interdependence increases, informed social imitation improves population performance and the best solution discovered. As problems grow, it becomes more likely to outperform informed imitation from one or several directly observed peers because collective distillation sustains a broader candidate repertoire. Most gains arise from evaluating a small candidate set, and population performance can improve when only a minority of organizations possesses evaluative capability. Social imitation thus reveals a division of cognitive labor: communities expand the strategies organizations can consider, while organizations create value by determining which strategies fit. It explains how firms can mobilize knowledge beyond their boundaries, and why widespread access to collective knowledge does not produce equal benefits.

Appointing more women to corporate boards is widely expected to also raise firms' environmental, social, and governance (ESG) performance. We provide the first meta-analysis of this relationship, drawing on 533 estimates from 106 studies that measure ESG performance with Bloomberg or LSEG ratings. The average reported effect of a one-percentage-point increase in board gender diversity is about 0.28 ESG points, but much of it does not survive scrutiny. Correcting for publication bias with a battery of linear and non-linear methods lowers the effect to between roughly 0.08 and 0.17 points. A best-practice estimate that also imposes sound study design puts it near 0.12 for most of the world, markedly higher for the Middle East, and near -0.11 for the Southeast Asian markets that dominate the Asian evidence. The differences that remain across studies are systematic, driven mainly by geography and by the choice of estimation method rather than by the ESG-rating provider or the controls a study includes. Board gender diversity may be well worth pursuing on its own merits, but the evidence that it reliably raises ESG scores is weaker than the published record suggests.

Consumers increasingly delegate purchasing decisions to Large Language Models (LLMs) acting as surrogate consumers. Using "Tool-Lab," an adaptation of information-board process tracing that places product attributes behind costly tool calls, we examine how marketing pricing cues (i.e., just-below pricing and promotional framing) influence AI shopping agents. Across eight commercially deployed LLMs from three providers, we trace pre-choice information acquisition. Under zero cost, pricing cues rarely mislead. Imposing acquisition costs under a vague goal prompt leads LLMs to omit diagnostic attributes required to compute unit price and choose suboptimal choices resembling human heuristics. Relative to a specific goal prompt that mainly preserves diagnostic search and choice optimality, a vague goal prompt under constraints creates a search-mediated vulnerability. This research demonstrates that marketing heuristics in delegated AI shopping are governed by storefront information architecture, not necessarily immutable LLM flaws.

This paper estimates the causal effect of drug consumption room (DCR) openings on drug-related hospitalizations in Switzerland. I exploit the staggered introduction of DCRs across Swiss cities between 1998 and 2022, using individual-level hospital records and a difference-in-differences estimator. DCR openings reduce drug-related hospitalization rates by roughly 35% within 5 km of a facility. The reduction is most pronounced for fatal outcomes, which fall by 35% while narcotics poisoning hospitalizations rise by a comparable amount, consistent with DCRs converting potentially fatal overdoses into non-fatal hospitalizations. Mental and behavioral disorder hospitalizations decline by 32% while estimates for Hepatitis and HIV are imprecise. The findings support the role of DCRs as an effective harm reduction strategy.

A Central Bank Digital Currency (CBDC) is central-bank money in digital form, held by the public. Leading designs have two limitations: conversion from bank deposits into CBDC can accelerate deposit flight, requiring safeguards, and the CBDC stays outside credit creation and monetary-policy operations. Here we present a CBDC architecture based on grassroots currencies that overcomes these limitations. The architecture has three components: (1) Money: sovereign grassroots coins, which are digital debts of one unit of fiat currency issued by the central bank, constituting a direct CBDC; (2) Credit and Liquidity: non-sovereign grassroots coins, which are digital debts of one unit of the same fiat currency, redeemable at par, that can be issued by any person, natural or legal, thus adding credit; and (3) Interest: grassroots bonds, sovereign and non-sovereign, adding maturity and thus interest, standard banking instruments, and the central bank's instruments of monetary policy. The central bank can therefore lend, absorb liquidity, set its rates and buy and sell securities in the coins and bonds the public holds, choosing the counterparties and terms of its credit operations, and without converting bank deposits into newly issued central bank money on demand. We prove that the arbitrage-free price of any non-sovereign grassroots coin whose issuer redeems it on demand is one unit of the fiat currency. The central bank can choose to deal with any counterparty, not just banks, and we argue that the central bank's interest rates on lending and bonds bound from above and below the corresponding interest rates of its counterparties. Sovereign and non-sovereign grassroots coins and bonds have been implemented and tested on a small scale.

Researchers increasingly use generative AI, particularly large language models (LLMs), to automate tasks across the research pipeline. We study the reliability of these tools at the reading, classification, and synthesis of large bodies of academic literature. We frame LLM-assisted literature reviews as a measurement problem, treating models as measurement systems and tracing how their errors affect downstream conclusions. As a test case, we use three different implementations of ChatGPT to identify and extract metadata from economics papers that use rainfall as an instrumental variable. We benchmark each implementation against a subset of human-labeled evaluation data, and then deploy those implementations to extract metadata from the full corpus. The LLMs perform well on binary classification, but performance deteriorates as tasks demand greater contextual interpretation. More importantly, how much researchers can rely on model outputs depends not only on the complexity of the reading task but also on the type of claims the data is asked to support. The same amount of measurement error substantially affects paper-level claims while having little effect on broader claims about the literature. Measurement error in LLM-generated data is thus most consequential at precisely the level of detail that constitutes an LLM's principal value added over human reviewers. We conclude that standard model performance metrics are informative about the quality of generated data but do not by themselves establish the credibility of downstream inference. Researchers must also evaluate whether substantive claims are robust to the measurement system used to generate the underlying data.

Trust can attract fraud when it delays verification. We develop a two-market signaling model in which generative AI lowers fabrication, verification, and targeting costs. When fabrication becomes profitable before verification, claim credibility first falls and later recovers. Across markets, higher prior quality can delay verification, creating an interval in which only the lower-quality market checks. If targeting becomes profitable in this interval, deceptive sellers enter the higher-quality but less vigilant market, and their entry can initially reverse its reliability advantage. The inflow also triggers verification and deters further entry. We call this self-limiting mechanism trust arbitrage. In the age of generative AI, trust can thus create an endogenous but temporary protection gap that redirects deception across markets.

Rule-based pricing tools are widespread in digital commerce, yet we know little about how their design shapes market outcomes. In a controlled market experiment, participants use dashboards to build pricing algorithms competing in a sequential Bertrand game over multiple periods. We vary design features commonly found in commercial repricing tools: warnings about price wars, pre-configured strategies, and advice from a large language model. Most treatment variations raise market prices with effects driven by an increase in starting prices and more cooperative algorithm designs. The results matter for competition policy, platform regulation and current discussions on regulating algorithm design tools.