Does the control of corruption cause progress in democracy and if so, through which channels? A threshold non-linear analysis

This article focuses on three main limitations of the literature regarding the impact of corruption on democracy. These are the distinctions between causality and correlation, channels underlying the impact and the shape of the relationship. The study uses recent developments in panel data causality and non-linear econometrics, breaks democracy down into different components and analyses the form of the relationship. We find that the relationship between the variables is often non-linear. Control of Corruption leads to a higher democracy; the components of democracy are also affected by corruption. These findings add to the literature in three respects: (i) corruption has a causal effect on democracy and, hence, single equation estimation poses problems, (ii) assuming linearity of the relationships between control of corruption and democracy is also possibly problematic and (iii) the channels of transmission of the effects of corruption on democracy are diverse.