Title: Medical Technology and Life Expectancy: Evidence from the Antitoxin Treatment of Diphtheria

Abstract: In this paper, we explore the impact of the first effective medical treatment for an infectious disease---diphtheria antitoxin---on the historical health transition. In 1895, the Massachusetts State Board of Health began providing free supplies of the antitoxin for medical use throughout the state. This policy has later been recognized as a significant event in the public-health history of Massachusetts. We use cross-municipality variation in pre-antitoxin diphtheria mortality rates and the availability of free antitoxin since 1895 to create an instrumental variable for local adoption rates, as measured by the number of antitoxin bottles per capita. By analyzing approximately 1.6 million death certificates from 1880 to 1914, we find that a hypothetical 10-year delay in the development of antitoxin would have reduced life expectancy at birth by one year, primarily due to reductions in child mortality. Our results suggest that medicine played a significant role in the increase of life expectancy in the early 20th century. Finally, we provide evidence suggesting that antitoxin treatment during the first 9 years increased school attendance but did not affect adult labor-market outcomes.

Title: The Fragility of Sparsity" (with Ulrich Müller and Sebastian Roelsgaard) There is no paper yet, but the abstract is:

We argue, using three empirical applications, that linear regression estimators which rely on the assumption of sparsity are fragile in two ways. First, we document that different choices of the regressor matrix that don't impact long regression estimates, such as the choice of baseline category with categorical controls, move the post-double-selection estimates by one standard error or more. Second, we develop two tests of the sparsity assumption based on comparing sparsity-based estimators with long regression. Both tests tend to reject the sparsity assumption in all three applications. Unless the number of regressors is comparable to or exceeds the sample size, long regression yields more robust results at little efficiency cost.