Corruption Kills: A Panel Data Analysis of OECD Countries

Sammanfattning: This paper studies the effect of corruption on four selected health outcomes, by using cross- country panel data, containing the 36 member states of OECD. The time period stretches from 1995 to 2017, extending upon previous studies by including data from more recent years. In order to estimate the effects of corruption, the paper uses Ordinary Least Squares, fixed effects and two-way fixed effects methods. In addition, we employ a Two-Stage Least Squares estimation method to check the robustness of the findings. To measure corruption, the main index used is the reasonably new Bayesian Corruption Index (BCI). Moreover, the paper investigates if the choice of corruption index has any effect on the results by using two additional measures of corruption; Corruption Perception Index (CPI) and Control of Corruption Index (CCI). The paper concludes that (1) increased levels of corruption tend to increase infant mortality rates and child mortality rates while corruption has no significant effect on neonatal mortality rates and life expectancy in OECD countries, (2) using BCI as corruption index generates results similar to previous research and (3) using three different corruption indices generate different results, which proves that the choice of index does matter. These findings confirm that corruption indeed kills and that a reduction of the level of corruption would improve public health globally.