Working Paper: NBER ID: w25884
Authors: Karen Clay; Peter Juul Egedes; Casper Worm Hansen; Peter Sandholt Jensen; Avery Calkins
Abstract: This paper studies the immediate and long-run mortality effects of the first community-based health intervention in the world – the Framingham Health and Tuberculosis Demonstration, 1917-1923. The official evaluation committee and the historical narrative suggest that the demonstration was highly successful in controlling tuberculosis and reducing mortality. Using newly digitized annual cause-of-death data for municipalities in Massachusetts, 1901-1934, and different empirical strategies, we find little evidence to support this positive assessment. In fact, we find that the demonstration did not reduce tuberculosis mortality, all-age mortality, nor infant mortality. These findings contribute to the ongoing debate on whether public-health interventions mattered for the decline in (tuberculosis) mortality prior to modern medicine. At a more fundamental level, our study questions this particular type of community-based setup with non-random treatment assignment as a method of evaluating policy interventions.
Keywords: tuberculosis; public health; mortality; synthetic control; difference-in-differences
JEL Codes: I15; I18; N32
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Framingham community health and tuberculosis demonstration (I19) | tuberculosis mortality (J17) |
| Framingham community health and tuberculosis demonstration (I19) | all-age mortality (J11) |
| Framingham community health and tuberculosis demonstration (I19) | infant mortality (J13) |
| demonstration period (1917-1923) (B14) | tuberculosis mortality in Framingham (I12) |
| demonstration period (1917-1923) (B14) | tuberculosis mortality in control communities (I14) |
| synthetic control method (C53) | significant reduction in tuberculosis mortality attributable to the demonstration (I14) |
| Framingham community health and tuberculosis demonstration (I19) | case detection (C52) |