Abstract
Previous research (Carey and Gohdes, 2021) concludes most journalist killings by the state occur in democracies. This paper investigates the link between regime type and journalist killings and argues that democracies are not more hostile to journalists than autocracies. In fact, liberal democratic regimes are the safest of regimes for journalists, i.e. those with by far the least number of journalist killings by the state between 2002 and 2016. Three arguments explain these different results: first, democracies are no homogeneous whole. Over time the definition of “democracy” in public and academic discourse has evolved. The liberal component (ensuring civic rights, protection of minorities, division of power, and constraints on the executive) has become a fundamental building block of “democracy”. Electoral democracies lag behind on this liberal dimension as the journalist killings exemplify. Second, a reporting bias of journalist killings skews the distribution of cases towards the democratic side. Across different measures of democracy most journalists are killed in the middle of the regime spectrum, i.e. in electoral autocracies and electoral democracies. Data transparency and reliability is higher for democracies than autocracies - journalist killings in autocracies remain unobservable to a larger degree than in democracies. Third, focussing on killings that do not occur in conflict settings disregards the connection between democracy and peace. It further skews the distribution of journalist killings against democracy since conflict settings disproportionately occur in autocracies. Failure to account for these biases as well as perpetuating a non-differentiated image of democracy as a theoretical “whole” is dangerous at a time where autocrats such as Victor Orban advertise the idea of illiberal democracy – an oxymoron.