Abstract We study the circumstances under which public pressure affects judging. We show that crowd pressure biases decisions in favor of the crowd for “subjective decisions” with respect to which the judge has more discretion but not for “objective decisions.” The bias is strengthened after a judge's error against the crowd and when errors are costlier to the crowd. We use data about referees' decisions and errors from the Bundesliga. We exploit three regimes where, due to the introduction of Video Assistance Refereeing (VAR) and COVID-19, both crowd pressure and the likelihood of errors vary.

Abstract How do consumer credit markets affect the allocation of workers to firms, output, and labour productivity? We address this question in two steps. First, we use new micro-data to estimate empirical elasticities of job search patterns to credit. Second, we estimate our novel theory of sorting under risk aversion to match these elasticities, and then we conduct aggregate counterfactuals. Empirically, we show that an increase in credit limits worth 10% of prior annual earnings allows individuals to take 0.33 weeks longer to find a job. Conditional on finding a job, they earn 1.85% more and work at higher paying firms. We also find that young and high-utilization individuals are more responsive to credit. Theoretically, we integrate risk aversion and borrowing into a model with worker and firm heterogeneity. We estimate the model to match our new empirical elasticities, and we then measure how the credit expansion from 1964 to 2004 affected sorting and output. Sorting improves as credit expands since constrained workers—in particular constrained, young, high human capital workers—find more capital-intensive jobs.