Skip to main content

Related projects


Some Unintended Fallout from Defense Policy: Measuring the Effect of Atmospheric Nuclear Testing on American Mortality Patterns

To better understand the health and social costs associated with radioactive pollution and nuclear weapons development, I study a historical period of atmospheric nuclear testing in the 1950s. Using records measuring annual county level fallout patterns for the continental U.S., I analyze how radioactive fallout affects public health in vital statistics records. I find that atmospheric nuclear testing performed in Nevada contributed to substantial and prolonged increases in overall mortality and cancer mortality. These increases in mortality occur over a broader geographic region than previous research would suggest.

Keith Meyers


Paralyzed by Panic: Measuring the Effect of School Closures during the 1916 Polio Pandemic on Educational Attainment

We leverage the 1916 polio pandemic in the United States as a natural experiment to test whether short-term school closures result in reduced educational attainment as an adult. With over 23,000 cases of polio diagnosed in 1916, officials implemented quarantines and closed schools. Since the pandemic occurred during the start of the 1916 school year, children of working age may have elected not to return to school. Using state-level polio morbidity as a proxy for schooling disruptions, we find that children ages 14-17 during the pandemic had less educational attainment in 1940 compared to their slightly older peers. 

Joint work of Keith Meyers and Melissa Thomasson

Keith Meyers


Cold and Mortality in a Cross-Sectional and Lifecycle Perspective: Evidence from Competing Models

This paper models the relationship between cold and mortality using two competing modeling techniques and two unique datasets from Sweden. The first dataset is the Swedish mortality statistic at the county-month level by cause of death. Linear fixed effects models simulate the dominant modeling approach in the economics literature and serve as a benchmark. The second dataset comprises entire life spells of 30,150 Swedes who were born between 1930 and 1935. Along with daily temperatures over 84 years from 1930 to 2013, we then model the impact of exposure to extreme temperatures on cause-specific mortality over humans’ lifecycle through competing risk models. In contrast to linear models, these models consider that temperatures can affect health through more than just one disease channel. Thus, the models allow time-varying temperature indicators to have competing effects on death due to heart diseases, respiratory diseases, and cancer. Although competing risk models have rarely been applied to this context, we show that they are a complementary modeling approach to study extreme temperatures and human health. Both modeling approaches find that extreme cold significantly increases the risk of dying from heart attacks. We compare both approaches and discuss possible biological mechanisms.

Joint work of Martin Karlsson, Emil N, Sørensen, Nicolas R. Ziebarth and Christian Møller Dahl.

Emil N, Sørensen
Christian Møller Dahl

Last Updated 28.07.2021