In the Natural Experiment
- 462 Downloads
The natural experiment offered by the cluster of political and public health reforms partitions the childhood years of the set of cohorts into pre-reform and post-reform era childhoods. This chapter explores the conditions required for any inference from such a natural experiment to be valid, as well as the limitations and the strengths of the inference. It finds that the differences in the profiles of cohorts brought up in the pre-reform era and the post-reform era were statistically significant, suggesting that political reform may have influenced modern aging. While the results may have some external validity, their internal validity is poor because the childhoods of the cohorts are likely to have differed along several dimensions other than being pre- or post-reform.
KeywordsAging Age-profiles Generations Non-communicable diseases Natural experiment Pre-reform Post-reform Selection bias External validity Internal validity
- Angirst, J. D., & Pischke, J.-S. (2009). Mostly harmless econometrics, an empiricist’s companion. Princeton: Princeton University Press.Google Scholar
- Heckman, J. J. (2008). Econometric causality (IZA Discussion Papers, No. 3425). http://nbn-resolving.de/urn:nbn:de:101:1-2008040906
- Wooldridge, J. M. (2009). Econometric analysis of cross section and panel data. Boston: MIT Press.Google Scholar