Research

Contemporaneous waves of democratization and decentralization in Latin America have divided residents of large metropolitan areas into numerous politico-administrative districts in both federal and unitary countries. While there is much debate about the effects of decentralization on representation in cities, two questions—one descriptive and one conceptual—have yet to be answered: how fragmented are cities in Latin America? And what exactly is being fragmented when new districts are created, or old ones are divided? This paper introduces the Fragmented Latin American Urban Democracy (FLAUD) dataset, which measures the degree and nature of urban fragmentation in over 1,000 cities in the region. Coalescing data on and coding all major administrative and political districts that intersect each metropolitan area, FLAUD uses micro-level spatial data on global population density to generate novel estimates of urban administrative fragmentation, weighted by the relative proportion of residents in each urban district. Using these estimates as a benchmark, FLAUD proposes two derivative dimensions of urban fragmentation central to the study of democracy and governance in cities. Electoral fragmentation characterizes whether the constituents of each district used in legislative and subnational elections are predominantly urban or rural, facilitating the study of potential coalitional dynamics that animate urban politics and the representation of urban voters in different cities and across countries. Sectoral fragmentation, by contrast, categorizes the core policy competencies allocated to representatives in each tier of subnational government to identify which policy sectors are affected by fragmentation across and within tiers. The descriptive statistics from the resulting dataset alone help to identify cities in which coherent urban policy may be impeded by excessive or insufficient metropolitan decentralization, opening several new channels for future research. To demonstrate potential applications of FLAUD, the paper estimates the relationship between urban fragmentation and both the spatial distribution of amenity access within cities and reported satisfaction with local governance and service provision by social class.

Cities have emerged as critical venues for democratic contestation and representation as countries rapidly urbanize and political authority and resources are increasingly channeled to elected subnational governments. Recent research finds that cities’ institutional frameworks influence their capacity to address persistent inequality among urban residents, which is spatialized in the form of class-based segregation. While fragmented metropolitan areas, in particular, are believed to facilitate effective representation of constituents at the local level, few studies have explored how electoral dynamics operative at the scale of the city moderate the quality of local democracy enjoyed by different social classes within cities over time. This paper considers how class-based partisan attachments formed in higher-level elections lead parties biased in favor of relatively rich or poor voters to dominate lower-level elections in class-segregated urban districts. Such dynamics systematically alter the nature and size of mayors’ winning coalitions among divided urban electorates. Drawing on electoral and demographic data from Buenos Aires, Santiago, Mexico City, and Lima, I estimate whether class-based preferences appear to inform voting behavior in city-wide elections over the last two decades and whether these trends, in turn, influence the electoral dynamics in the various urban districts that fragment each city. I generate city-level estimates of the degree to which voters from different social classes are exposed to local electoral contestation and a district-level typology of the mandates with which mayors are elected, a function of their support base, political ties, and margin of victory. Apart from highlighting significant variation in the functioning of urban democracy across major Latin American cities, the analysis illustrates why urban fragmentation and class-based voting are no guarantee of effective representation among segregated urban residents. Indeed, their co-occurrence can inadvertently weaken the incentives among elected officials to govern accountably and equitably.

The tendency for parties to perform similarly in higher- and lower-level elections—i.e., high party localization—can generate nested advantages for candidates affiliated with dominant up-ballot parties in local districts. While voters may use party labels as a heuristic to guide their vote choice in local elections because they hold genuine partisan preferences, party localization may also materialize where voters lack alternative information about local candidates. Under such conditions, candidates in lower-level elections affiliated with dominant parties may ride party coattails to victory even if voters know little about them; large, nested advantages may even crowd out the election of stronger candidates about whom voters hold apparent preferences. This paper turns to municipal elections in Lima, Peru as an instructive case where robust party localization exists between provincial and district elections even though parties are fragile, personalistic, and transient. Across two decades of municipal elections, I find that provincial voting patterns, which are dynamically class-based and candidate-centric, constitute strong baseline predictions for party performance in Lima’s 42 autonomous districts. Voters deviate from this baseline when they have independent, reliable information on the quality of mayoral candidates in the form of past political experience. Yet district candidates affiliated with especially dominant provincial parties often ride their coattails to victory, even if they dramatically underperform their provincial counterparts. Popular candidates that overperform their provincial coattails are less likely to get elected at the district level when their competitors enter the race with large, often unjustifiable, electoral advantages associated with voters casting straight-ticket ballots.