Does corruption kill? Evidence from half a century infant mortality data.

We investigate the relationship between corruption and infant mortality in Turkey. Our study contributes to the literature in a couple of ways in terms of data and estimation method. First, we construct a novel index of corruption in Turkey based on the stories covered in a major newspaper between the years 1960 and 2010. Second, because we now annual data on corruption covering a period long enough, we investigate the integration properties of the data and estimate the cointegrating relationship between corruption and infant mortality using Fully Modified OLS (FMOLS) and Canonical Cointegrating Regression (CCR). We find that corruption increases infant mortality in the long-run. Our results are robust to different specifications and estimation methods.