Introduction to Special Section: Elections in Latin America from a Multi-Level Perspective

In multi-level systems with several arenas of political competition, national electoral results or political outcomes are unlikely to have a homogeneous distribution throughout the territory. Therefore, a nationalised perspective on politics has to date not led to a deep understanding of political dynamics (Došek and Freidenberg, 2013; Snyder, 2001). This is particularly true in Latin America, which was the first region in the world systematically to carry out decentralisation measures, devolving political authority to subnational actors and creating new arenas of competition (Falleti, 2010). Not only have federal countries implemented a wide variety of decentralisation reforms, but unitary countries have become less unitary thanks to the cumulative effects of decentralising processes (Eaton, 2018).

Located at the intersection of debates on agency and structure in politics, this Special Section examines different aspects of elections in Latin America from a multi-level perspective. We aim to highlight the territorial diverseness of multi-level elections as well as the heterogeneous distribution of their political outcomes using disaggregated data. Latin America offers an interesting setting for the study of elections from a multi-level perspective thanks to its outstanding cross-country diversity in territorial power and decentralisation processes. Given that the territorial heterogeneity of national and subnational elections is context-specific, a closer look at these countries is the most appropriate empirical strategy.

This Special Section makes two contributions. First and foremost, we call attention to the fact that the sole use of aggregate data to draw conclusions about electoral participation, party strength and the functioning of intergovernmental relations seems no longer sufficient to grasp political phenomena (Došek and Freidenberg, 2013). Accordingly, the contributions of this Special Section show that the disaggregation of electoral data by constituency is a more accurate way to assess political variation across and within countries, as well as to trace change of cross-territorial political dynamics over time. Second, we show that multi-level political dynamics help to explain national political outcomes regardless of the country's territorial structure (whether federal or unitary).

The six articles included here address three different themes in five Latin American countries. The first focuses on how multi-level institutional settings impact electoral and political outcomes. Méndez de Hoyos and Arellano, as well as Jaramillo, explore the relationship between electoral institutions and turnout rates in Mexico and party fragmentation in Peru, respectively. Analysing the effect of electoral integrity on electoral participation in violent contexts in subnational elections in Mexico, Méndez de Hoyos and Arellano suggest that the effect of violence on electoral participation depends on the level of electoral integrity. They find that when strong electoral institutions become a source of confidence in the way in which elections are organised and conducted, the negative impact of violence on turnout lessens, or almost disappears. Jaramillo's article focuses on how the legislative electoral system and electoral competition contributed to a rapid and constant division in the Peruvian Congress in recent years. It argues that permissive electoral rules regarding party affiliation and the dynamics of subnational party competition diminish party discipline and further increase the incentives for parliamentary groups' fragmentation in a context of already weakly institutionalised political parties.

The second theme focuses on the relationship between geographical divisions and political cleavages. The articles by Forero-Mesa and Rodríguez-Raga on Colombia, and Santos and Moreira on Brazil discuss how geographical divisions inside countries account for the establishment of electoral strongholds and newcomers in politics. The first article analyses how Colombian senators use Twitter to target subnational constituencies. The authors find that legislators' Twitter behaviour not only mirrors their patterns of electoral support but also depends on intra-party competition. Santos and Moreira, for their part, show that the proliferation of small right-wing parties in Brazil is connected to regional economic interests in agribusiness as well as to the ‘catch all’ nature of the traditional right and centre-right parties. Thus, the increasing party fragmentation in Brazil is a consequence of structural changes in the economy, particularly in agriculture.

The third theme concerns the influence of subnational political dynamics on national political outcomes. The articles by Maia et al. as well as by Cherny tackle this topic in Brazil and Argentina, respectively. Maia et al. argue that, despite changes in party systems at the national level around President Bolsonaro's election in 2018 or incongruence between national and subnational levels in Brazil, the subnational level works as a reservoir of established parties, and in fact no real change in party systems is identified when electoral volatility is analysed. In his article, Cherny verifies the legislative influence of governors through quasi-experimental evidence from senatorial elections in Argentina. He shows that legislators who conclude their terms within the administration of their respective governors vote more cohesively with the rest of the legislators in their provincial party delegation, thus pointing to the existence of shared loyalty.

This Special Section is particularly relevant for the community of scholars who are interested in understanding Latin American elections in their complexity. It addresses electoral trajectories and political outcomes from a multi-level perspective. By analysing the contexts in the electoral districts where elections take place, we aim to gain a better understanding of territorial heterogeneity therein and of how national events are reflected and/or explained by subnational political dynamics. This will help us to better grasp political phenomena that are usually blurred by the lens of methodological nationalism.