Study title

The dataset of Countries at Risk of Electoral Violence (CREV) provides detailed dyadic information on electoral violence in 101 countries between1995 and 2013. For an election to be deemed “at risk” of electoral violence, two criteria have to be met. The country in which the election has taken place must not have been a fully consolidated democracy (defined as having a Polity IV (Marshall, Gurr and Jaggers 2016) score of 10) throughout the entire time period covered by the data, and it must have sufficient media coverage (defined as an average of at least 365 reported events per year in the ICEWS dataset (see below for details)). The dataset of Countries at Risk of Electoral Violence follows the National Elections across Democracy and Autocracy (NELDA) election classification (Hyde and Marinov 2012; 2014). Elections in CREV are for national-level legislative and executive contests only, local and regional elections are excluded, as are referendums and constituent assembly elections. Electoral violence is measured in a ten-month window around each election. We code violence beginning six months before the election, three months after the election, and the month of the election. We provide two versions of the dataset. One is a time series cross-sectional (TSCS) dataset in which the unit of observation is the election, and where events of electoral violence are summed during the ten-month window. The other is a time series cross-sectional (TSCS) dataset in which the unit of observation is the electoral cycle month, and counts of violent events are specific to a given month during an electoral cycle. Elections are a means of adjudicating political differences through peaceful, fair, democratic mechanisms. When elections are beset by violence, these aims are compromised and political crises often result. Despite the undisputed importance of understanding electoral violence, there has been only a limited body of systematic comparative research on this topic. If...