CHAPTER 2 investigates the impact of the staggered expansion of early child care in Germany on the spread of influenza. Recent data from contact tracing show that educational institutions such as universities, schools, and child care facilities are among the most common locations where airborne diseases are spread (Robert Koch Institute 2012 . Simultaneously, the number of children attending formal child care has increased substantially in many OECD countries over the last two decades (OECD

more »

... ). If child care facilities indeed propel the spread of infectious diseases, the expansion of early child care imposes unintended costs on society. Work in medical literature shows that child care attendance is positively related to infection rates (e.g Ball et al. 2000; Ball et al. 2002; Côté et al. 2010 ). Yet empirical evidence on a causal impact of child care on the spread of infectious diseases is scarce. This chapter attempts to close that gap by providing causal evidence on the transmission of influenza as a consequence of the expansion of early child care -care offered to children under the age of 3 -in Germany. To identify these effects, I leverage the staggered roll-out of early child care provision in Germany between 2005 and 2016. I further exploit detailed, high-frequency data on the incidence of influenza that allow me to uncover age-specific disease transmission patterns over time and space. The analysis proceeds in three steps: First, I provide reduced-form evidence on an economically and statistically significant effect of child care on infection rates using a difference-in-differences strategy. Second, I extend a semi-parametric model of disease diffusion -the Susceptible-Infected-Resistant Model (SIR) -that builds on epidemiological work. The model allows the estimation of age-specific transmission rates and the identification of policyrelevant margins of heterogeneity. Due to the dynamic structure of the SIR model, a non-classical measurement error in the infection rates may result in biased estimates (Adda 2016). To address this endogeneity concern, I create a novel instrument based on lagged temperature. The results of the SIR model document that transmission rates between children aged 0 to 2 and children aged 3 to 6 significantly increase in response to a rise in early child care coverage rates. I further find profound differences between urban and rural areas: While the expansion in child care significantly increases transmission in urban counties, the data do not provide evidence for a significant effect in rural areas. Finally, I evaluate the effect of two counterfactual policy interventions that aim at limiting the spread of infectious diseases in child care facilities: mandatory vaccination before entry into child care and the closure of child care during local outbreaks. I find that both policies significantly reduce infection rates. While the policies have positive spill-over effects on adults, they mostly benefit children. Evaluating the economic costs and benefits of mandatory vaccination policies shows that the policy intervention would be cost-effective reducing net annual costs by about 20% relative to the status quo. CHAPTER 3, which is joint work with Pavel Obraztcov, Gregory Veramendi, and Joachim Winter, studies the role of public information about unexpected local outbreaks of Covid-19 in mitigating the spread of the virus. Covid-19 is an overdispersed pathogen, where a small fraction of individuals is responsible for a large fraction of the transmission. Epidemiological studies have shown * This chapter is based on joint work with