This article reports evidence of misspecification of the measurement model for the index of emancipative values, a value construct used as a key explanatory variable in many important contributions to political science. It shows that the scale on which the index is measured is noninvariant across cultural zones and countries in the World Values Survey. In addition, it demonstrates that the current index composition mixes different value dimensions and their actual associations with various political outcomes, in particular the index of effective democracy. However, an analysis using a novel approximate Bayesian approach shows that at least one specific subdimension of emancipative values, known as pro-choice values, truly exists and may be validly measured and compared cross-nationally. The article also contributes to the recent discussion on whether emancipative values are a reflective or a formative construct by providing thought experiments and empirical evidence supporting the former interpretation.
Every dictator dislikes free media. Yet, many non-democratic countries have partially free or almost free media. In this paper, we develop a theory of media freedom in dictatorships and provide systematic statistical evidence in support of this theory. In our model, free media allow a dictator to provide incentives to bureaucrats and therefore to improve the quality of government. The importance of this benefit varies with the natural-resource endowment. In resource-rich countries, bureaucratic incentives are less important for the dictator; hence, media freedom is less likely to emerge. Using panel data, we show that controlling for country fixed effects, media are less free in oil-rich economies, with the effect especially pronounced in non-democratic regimes. These results are robust to model specification and the inclusion of various controls, including economic development, democracy, country size, size of government, and others.