The journals are full of great studies but can we believe the statistics? Revisiting the Mass Privatisation – Mortality Debate

摘要: Cross-national statistical analyses based on country-level panel data are increasingly popular in social epidemiology. To provide reliable results the societal determinants of health, analysts must give very careful consideration to conceptual and methodological issues: aggregate (historical) typically compatible with multiple alternative stories data-generating process. Studies this field which fail relate their empirical approach true underlying process likely produce misleading if, for example, they misspecify models by failing explore properties longitudinal aspect or ignoring endogeneity issues. We illustrate importance extra need care reference a recent debate whether discussing role rapid mass privatisation can explain post-communist mortality fluctuations. demonstrate that finding was "crucial determinant" male fluctuations world is rejected once better given way generated.