Votar en medio de la violencia: comportamiento electoral en México

Abstract

Our understanding of political behaviour has been mainly based on research focused on environments characterised by peace and democracy. However, the analysis of such behaviour in contexts where violence and insecurity are a constant in the daily lives of many citizens has been addressed in a limited way. Taking into account the gaps detected in the literature, the present doctoral thesis examined the effects that violence and perceptions of security have on citizens' political behaviour, limiting the analysis to electoral participation, voting orientation, presidential approval and satisfaction with democracy. Mexico was selected as a case study, mainly for the period 2018-2021, as it is a country with high levels of violence and insecurity, which allowed us to test the hypotheses put forward in each of the aforementioned manifestations of political behaviour. However, considering that one of the most recurrent expressions of violence in Mexico comes from the activity of criminal groups, and given that this violence affects the municipal level to a greater extent, an analysis was carried out to determine whether such violence had an impact on municipal elections during the 2006-2018 period in Tierra Caliente, a region known for its persistent history of violence. For this purpose, a mainly quantitative methodology was used, which made it possible to observe how various indicators of violence and insecurity affect political behaviour. Two types of analysis were carried out on electoral participation. On the one hand, an aggregate analysis of the incidence of criminal violence on electoral participation in municipal elections in the Tierra Caliente region was carried out using a database of our own. This aggregate analysis was later complemented with a qualitative approach for a better understanding of the results. On the other hand, using the Latin American Public Opinion Project (LAPOP) 2019 as a data source, an individual-level analysis was carried out on the incidence of violence and perceptions of security on voter turnout in presidential elections. The remaining analyses in this research focused on examining the effects of violence and perceptions of security on voting orientation in presidential elections, presidential approval, and satisfaction with democracy. In these cases, individual survey data were used, with the LAPOP being the main source. Specifically, data from the 2019 wave were used in the seventh and ninth chapters, while data from the 2019 and 2021 waves were used in the eighth chapter. These analyses have yielded several findings. First, criminal violence, as measured by firearm homicides per 1,000 inhabitants, does not show statistically significant effects on voter turnout in municipal elections. Second, knowledge about murders in the neighbourhood is negatively related to voter turnout in presidential elections. Third, the perception of insecurity as the main national problem and direct victimisation negatively affect support for the current president's candidate or party in future presidential elections. Fourth, as far as presidential approval is concerned, the violence and security perception indicators used do not show any impact. Finally, victimisation, the feeling of insecurity in the neighbourhood and knowledge about cases of enforced disappearances have significant effects on citizens' satisfaction with democracy.