This paper investigates the impact of airborne pollen on serious and fatal traffic accidents in the UK. Using daily county-level data from UK Met Office reporting sites spanning 2013–2023, I find that higher pollen concentrations cause an increase in severe traffic accidents. The effect seems to be driven by tree pollen in particular. The results are robust to a range of weather and air pollution controls and are further validated using simulated pollen data from Ambee (2016–2022). These findings underscore the role of seasonal health-related factors in shaping road safety outcomes. Additionally, I explore the use of thermal inversion as an instrumental variable for pollen exposure and raise serious concerns about the validity of the exogeneity assumption in this context, with implications for its application in air pollution research.

Die Einführung des ElterngeldPlus hatte unter anderem zum Ziel, die partnerschaftliche Aufteilung von Erwerbs- und Sorgearbeit zu fördern. Obwohl die Reform die gleichzeitige Teilzeitbeschäftigung von Müttern und Vätern im Bezugszeitraum deutlich attraktiver gemacht hat, wird diese Option weiterhin selten gewählt. Wir möchten untersuchen, inwiefern dieser Befund darauf zurückzuführen ist, dass die finanziellen Anreize im ElterngeldPlus nicht richtig verstanden werden. Zu diesem Zweck konzipieren wir ein Surveyexperiment, in dem sich die Teilnehmenden zwischen unterschiedlichen Möglichkeiten, die Erwerbsarbeit innerhalb eines Paares zu teilen, entscheiden müssen. Dabei variieren wir die Elterngeldregelungen, die Art der Informationsbereitstellung und die Einkommen der beiden Partner vor der Geburt. Somit können wir den Effekt der Informationsbereitstellung, der Reform selbst und deren Interaktion kausal identifizieren.

Since intra-household resource distribution is unobserved it is difficult to comparehow women and children fare across countries. To address this we analyze 45 household expenditure surveys from predominantly low- and middle-income countriesi.e. an international sample of around 2.4 million individuals. Using harmonized estimations of intra-household resource sharing we construct globally comparable measures of gender inequality and child poverty in consumption. Our findings reveala widespread imbalance: women receive about one-fifth less than men leading to a 60% higher poverty rate. Children appear to fare even worse though this is partlyexplained by differences in needs and sibling economies of scale. Intra-householdinequalities are more pronounced in poorer countries and among low-income households within countries. Cross-checks with nutritional proxies tend to validate ourresults linking household poverty and intra-household disparities to child undernutrition. Finally we decompose global individual consumption inequality and findthat 13%-32% (across measures) stems from inequality within households.

This paper examines the impact of corporate income tax rates on establishment-level productivity growth using a panel dataset of approximately 50,000 German manufacturing and service establishments from 1994 to 2019. We document a robust nonlinear relationship: higher tax rates enhance productivity growth among establishments near the technological frontier, while hindering convergence among those further behind. This pattern holds across multiple productivity estimation methods and a wide range of alternative specifications and robustness checks. To investigate the underlying mechanisms, we show that higher tax rates induce the exit of low-productivity incumbents, reallocating labor and capital toward more efficient survivors. This reallocation amplifies gains among high-productive firms but simultaneously deepens divergence for lagging establishments. A detailed decomposition reveals labor reallocation as the primary channel driving these effects, with capital reallocation—particularly through equipment investment—playing a secondary role. Overall, our results underscore the importance of local tax policy in shaping productivity dynamics through firm selection and resource reallocation mechanisms.

We study in-group favoritism in the evaluation of grant proposals, focusing on gender and nationality in the European Research Council (ERC) funding process. Using data on over 37,000 proposals and the structure of ERC reviews, we find that individual reviewers tend to give higher scores to applicants who share their nationality or gender. Nationality-based favoritism is particularly strong while gender-based favoritism is smaller but still significant. These patterns persist after accounting for reviewer characteristics and proposal-reviewer topic match. At the level of final panel decisions—where reviewer scores are aggregated and discussed—we find weaker and more mixed evidence of in-group favoritism, with some persistence of nationality bias but little effect for gender.

Advancements in artificial intelligence have enabled the creation of highly realistic audio deepfakes, raising concerns about their potential for deception in scams, romance fraud, and other exploitative schemes. This study examines the extent to which audio deepfakes can be used strategically to generate trust using a two-player trust game with one-sided pre-play communication. I investigate the perceived trustworthiness of authentic and artificial voices and whether trust is exploited by the use of deepfakes. Artificial voices significantly affect trust, with artificial female voices being perceived as more trustworthy than artificial male voices. Moreover, male speakers are more likely to select artificial female voices compared to female speakers selecting artificial male voices. Despite this strategic choice, speakers do not exploit listeners’ trust, as return rates remain consistent across authentic and artificial messages. These results highlight the role of voice-based social identity and deepfakes on trust in digital communication.