Knowledge that Transforms

Online experiments have surged in popularity, but how credible are their published results? We examine the universe of hypotheses tested on Amazon Mechanical Turk and reported in leading journals between 2010 and 2021. We find evidence of widespread p-hacking, publication bias, and under-powered studies. The extent of the problems varies substantially across fields, with marketing particularly badly afflicted and little evidence of distortion in economics. We explore correlates of increased credibility and compare the online results to laboratory results from the same outlets, finding similar distortions in both, with bunching just above the significance threshold more pronounced on MTurk.

We study the real and allocative consequences of relaxing gender-specific credit constraints by analyzing a blended finance program designed to expand bank lending to female entrepreneurs in Turkey. Merging credit registry, tax, and matched employer-employee data, we find participating banks raised their credit share to women by over 18%, spanning repeat, poached, and first-time borrowers. Beneficiary firms invest more, grow employment, sales, and profits, diversify their networks, and exit less. Treated banks also expand lending to male entrepreneurs, though the increase is smaller and disproportionately directed toward higher-productivity firms, suggesting reallocation rather than crowding out. District-level effects are absent, reecting the program’s modest scale.

We examine the preferences of low-income households in Kenya over the structure of unconditional cash transfers. Most preferred lumpy transfers, and some preferred deferred receipt—in contrast to the typical structures of safety-net programs, but consistent with evidence on the financial challenges of poverty. Turning to consequences, receiving transfers later in the year raised income 1.5 years later—but willingness to defer receipt was sensitive to small changes in cash flow around the time of decision-making. Taken together, these results illustrate how adapting cash transfer design to the decision-making environment of those in poverty could improve financial choices and outcomes.

How can short episodes of localized hostility shape regional economic trajectories that even amplify over time? I exploit economic differences across the idiosyncratic and short-lived line of contact between the Red Army and the Western Allies in South Austria at the end of WWII. The 74-day hostile presence of the Red Army shaped the age- and occupation-structure via refugee flows, causing amplifying demographic effects over time and backwardness in sector development and tax revenues. My findings also point to the isolated role of the Red Army’s hostile actions after WWII in understanding the European economic East-West divide.

Information is central to consumer credit, but little is known about its equilibrium effects. We study a 2017 natural experiment in which non-bankruptcy public records were purged from U.S. credit reports. The large information-removal differences across states allow us to identify equilibrium impacts. Available credit increased for consumers who lost negative information directly. Consistent with our model, available credit declined indirectly for consumers who did not lose the negative information because they could no longer distinguish themselves. The direct and indirect effects nearly canceled, so the net equilibrium effect after a year was nearly zero but with substantial redistribution.

We assess the empirical evidence for the hypothesis that women more than men respond to changes in treatment. First, we examine whether the results of over two hundred experimental economics papers support the female sensitivity hypothesis. Second, using data from two studies (DellaVigna and Pope, 2022; Exley et al., 2025), we conduct over two hundred pairwise tests of the hypothesis. Both analyses show that gender is not predictive of responsiveness to treatment. We further examine how the hypothesis has been disseminated in the literature and find that it is referenced predominantly by papers that support it.

We examine parental response to a fetal shock using data from 150 Demographic and Health Surveys in 50 developing countries. Focusing on the impact of Ramadan fasting, a common shock experienced by more than a billion Muslims worldwide, we find that children exposed to Ramadan fasting during pregnancy have lower rates of full vaccination for DPT and polio compared to their non-exposed siblings. Further, parents tend to compensate for the fetal shock during booms and reinforce during economic downturns. Our estimations suggest that this has resulted in approximately half a million excess infant deaths over the past three decades.

Using online job postings data, we explore whether auditors adjust their human capital in response to clients' cyber breach incidents. We find that audit offices increase their demand for cybersecurity specialists after their clients' cyber breaches, implying that cyber breaches change not only client risk but also their auditor's production function. Consistent with expectations, we document that this impact is larger when breaches involve their own clients (rather than those of peer offices within the same audit firm) and when breaches directly involve financial information. Our evidence holds for both cyber‐focused and non‐cyber‐focused offices, highlighting the broad impact of cybersecurity risks on audit practices. Additionally, we find that audit offices with relatively high investment in cybersecurity human capital are better able to preserve their market share among clients in high‐cyber‐risk industries after a client's cyber breach.

Managers can boost reported earnings to meet capital market expectations by pursuing aggressive cost-minimization strategies that violate labor laws, such as pressuring frontline employees to forgo safety protocols or work off the clock. We investigate whether labor law violations, which benefit the firm at the expense of employees, fall disproportionately on nonwhite frontline employees. Across similar locations (e.g., stores or plants) within the same firm, we predict and find that locations with a higher share of nonwhite frontline employees are more likely to experience a labor law violation. This racial inequality in labor violations is especially pronounced in settings more prone to racial discrimination and during periods of acute capital market pressure to meet earnings benchmarks, providing insight into the mechanisms underlying the observed disparity. Our evidence suggests that capital market pressure is associated not only with labor violations but also with racial inequality in exposure to those violations. Data Availability: Data are available from the sources cited in the text.

Organizations increasingly operate in uncertain environments, requiring employees to revise task strategies to remain effective. We examine how pay for performance (PFP) influences adaptive performance following unanticipated change and argue that process accountability moderates this effect. Although PFP increases effort, it hinders adaptation by reinforcing persistence with an existing task strategy. Process accountability channels PFP-induced motivation into deeper strategic deliberation, enabling adaptation. We test our predictions in an experiment manipulating compensation (PFP versus flat wage) and process accountability (absent, reflection-only, and present), with participants in the reflection-only condition privately reflecting on their decision-making to isolate the social-evaluative component. Results show that PFP, relative to a flat wage, decreases adaptive performance when process accountability is absent, has no significant effect under reflection-only, and increases it when process accountability is present. Additional analyses suggest that strategic deliberation partially explains this effect. The findings underscore pairing PFP with process accountability to foster adaptation. Data Availability: Data are available upon request.