When does information meaningfully change behavior? I study the onset of the Germ Theory of Disease (GTOD). Although core principles were formalized by the 1870s and widely accepted by the 1890s, infectious disease mortality in the U.S. remained high well into the 20th century. I exploit the interaction of carefully documented temporal variation in the diffusion of critical GTOD implications and cross-city differences in baseline newspaper circulation and household hygienic infrastructure. Consistent with contemporary accounts, I show that modern hygiene norms did not appreciably diffuse through U.S. science journals and the popular press until the 1910s. Deviations from mortality trends across treatment intensity thresholds post-diffusion highlight infrastructure as a binding constraint. Cities with low baseline access experienced persistently higher post-diffusion mortality than better-equipped cities despite strikingly similar pre-diffusion health dynamics and attitudes. These findings imply that information in isolation may be insufficient to change health behaviors detectably and its potential depends on the extent to which perceived returns shift for individuals with the capacity to act.

The economics literature shows that targeted cash transfers are generally beneficial for disadvantaged children. However, precisely how much these transfers compensate for their disadvantage remains unclear. We examine this by leveraging quasi-random variation in both mining accidents and access to workers’ compensation for families in the early-20th-century United States. Matching detailed individual-level data on coal mining accidents with linked Census data allows us to study how workers' compensation payments may have narrowed income and education gaps between children whose father suffered a serious accident and similar children whose father was unaffected.  

The prominent decline in urban mortality rates during the late 19th and early 20th centuries in the United States is primarily attributed to reductions in deaths caused by infectious and food- and waterborne diseases. While existing literature has explored the role of public health initiatives in explaining this decline from the latter of these types, this paper focuses on the contribution of innovations in and provision of hygienic technologies and their role in eradication of the former. Specifically, we investigate the health impact of the rapid adoption and utilization of bathhouses in urban working-class neighborhoods following the technological advancement of what we recognize today as the modern-day shower. By exploiting differential timing of bathhouse openings beginning around 1890 and utilizing city-level mortality data for U.S. urban areas, we estimate the mortality effects of these facilities introduced prior to the widespread adoption of modern household sanitation infrastructure and critical implications implied by the germ theory of disease. This paper highlights the role that hygienic infrastructure and behavior may have played in improving public health during that era, shedding light on mechanisms policy makers can utilize when designing interventions aimed at encouraging behavioral take-up today.