Leopoldo Fergusson

Research in Progress

“Population and Social Conflict,” with Daron Acemoglu and Simon Johnson

This paper examines the effect of population growth on violent social conflict. Exploiting the international epidemiological transition that began in the 1940s, we construct an instrument for changes in population (Acemoglu and Johnson, 2007) and find that countries with higher (exogenous) increases in population experimented larger increases in social conflict. A simple falsification test indicates that changes in conflict from 1900 to 1940 are uncorrelated with our instrument, lending support to our identification strategy. Our results are robust to using a variety of standard measures for conflict and are not driven by differential trends between countries with different baseline characteristics often emphasized in the literature on civil wars. Using a simple theoretical framework, we interpret these findings as an indication that a larger population, without a corresponding increase in resources and technology, generates greater competition for resources and makes conflict and civil wars more likely in environments where the institutions cannot handle the higher level of disputes.

“The Rich-poor Divide, Within-groups Inequality and Armed Conflict,” with Juan Fernando Vargas

Economic inequality is often considered a prime cause of violent conflict. However, empirical research and theoretical arguments suggest that there is no reason to expect an unambiguous impact of wealth inequality on conflict. In this paper we use a simple theoretical framework to examine the way in which two different dimensions of inequality influence violent conflict. The first dimension is the "rich-poor divide," or the overall level of inequality between a large majority of relatively poor and homogenous individuals, and a smaller elite of wealthier individuals. The second dimension of inequality is the wealth dispersion within the rich individuals in society. In terms of the poor-rich divide, we identify two possible regimes with opposing predictions in terms of the inequality and conflict relation: a "dispossession regime," where the relationship is negative, and a "grievance regime," where instead inequality increases conflict. Using cadastral records from Colombia, we provide prima facie evidence in favor of the "grievance regime." However, wealth inequality within the relatively rich is correlated with less conflict. We interpret this last result as an indication that state capacity is low and the extent of property rights protection thus depends upon the capacity of groups in society to either organize effectively their own protection or lobby the state for protection. In this context, this negative correlation is consistent with an Olsonian view of collective action: greater concentration of wealth improves collective effort by the rich, and deters the poor from initiating conflict.

“Words versus Bullets: Media and Democracy with Coercion,” with Juan Fernando Vargas

A growing strand of the political economy literature has stressed the crucial role of the media in facilitating the flow of candidates’ information to voters. In general, this research has shown that well-informed voters make better choices. This paper examines whether this result breaks down in weakly institutionalized environments. We construct a unique individual-level high-frequency dataset on the press coverage of over 500 politicians across 14 years, as well as time-varying data on press penetration at the municipal level, to examine the limits of media exposure of corrupt politicians in disciplining voters in Colombia. In particular, we study a nation-wide scandal involving a large number of incumbent legislators proved to have colluded with illegal paramilitaries. To identify the causal effect of media exposure, we rely on differences in the timing of press coverage and compare ‘parapoliticians’ exposed shortly before and after the election. We conduct additional robustness checks to verify that results are not driven by potential endogeneity in the timing of media exposure.