Objective/context:

The objective of this article is to analyze the results of the anti-corruption referendum in Colombia in 2018. Colombia is a country with a significant corruption problem. More than 99% of the voters who came to the polls voted in favor of the proposals. However, the anti-corruption referendum nonetheless failed because not enough citizens were mobilized to participate. The article addresses the reasons why turnout was very low.

Methodology:

I examine the results at the municipal level. I present an original dataset of 1,101 Colombian municipalities. I use ordinary least squares (OLS) regression models to test theories based on the literature on referendums, corruption, and transparency. I also analyze voter turnout in the 2018 presidential election in order to compare it with participation in the referendum.

Conclusions:

I find that the more transparent a municipality, the higher the percentage of the municipal electorate that voted for proposals in the anti-corruption referendum. Moreover, I find that in municipalities where support for Sergio Fajardo in the presidential election was higher and support for Iván Duque was lower, support for the referendum proposals was higher. Also, turnout was lower in municipalities with higher poverty rates and higher homicide rates.

Originality:

This article contributes to the current global debate on direct democracy. As the anti-corruption referendum was held only recently, a proper analysis has not yet been carried out. Moreover, because of the nature of the referendum questions, the topic is closely connected with research on corruption. Therefore, this research represents a unique opportunity to examine corruption and direct democracy at one and the same time.