Corruption and Media Concentration: A Panel Data Analysis

My master thesis studies the relationship between media concentration and corruption based on a panel data analysis, with a panel dataset which provides information about 29 countries over a span of 19 years. Based on a cross-section analysis Djankov, McLiesh, Nenova and Shleifer (2003, JLE) which focused on the relationship between corruption and media state-ownership, I enhance their results thanks to panel data fixed effects, to control for more unobservable effects, and several robustness checks. As Djankov et al did, we focus on two specific media markets: television (TV) and daily newspapers. Thanks to new data from the book “Who Owns the World‟s Media?” (Noam, 2016), I broaden the spectrum of their article to focus on the correlations between corruption and all types of media concentration (public, private and the industry). I confine their previous results: a positive correlation is found only for public TV with large shares of the market. In fact, I find a negative correlation between public TV shares and corruption for lower levels of state-ownership, especially in the case of developed countries. Contrary to daily newspapers, this result remains after many robustness checks. I provide evidence that low-levels of state-ownership limit concentration of private media, reducing the risk of media capture. Indeed, competition within the private sector is found to be negatively correlated with corruption. Finally, I find weak evidence for a positive correlation between corruption and media industry concentration in only two cases: when considering all types of media and when considering TV alone.