Source: Journal of Politics in Latin America | Published: 2026-06-17
Category: 정권·선거 변동 | Keywords: authoritarian, democracy, election, electoral, mexico, party, politics
The persistence of authoritarian structures within formally democratic systems represents one of the most pressing puzzles in comparative politics. Across Latin America, transitions from military rule or single-party dominance during the late twentieth century were broadly celebrated as democratic breakthroughs, yet the decades that followed have revealed deep tensions between procedural electoral competition and the substantive accountability that democracy promises. Mexico's case is particularly instructive in this regard. The country's protracted transition away from Partido Revolucionario Institucional (PRI) hegemony — culminating in the historic 2000 presidential election — was heralded as a watershed moment for the region. Yet questions about what actually drives electoral outcomes in post-transition Mexico, and whether democratic competition has produced the accountability relationships theorized in the literature, remain sharply contested. A new study published in the Journal of Politics in Latin America addresses precisely these questions at the subnational level, examining gubernatorial elections and arguing that party machine infrastructure, rather than government performance, remains the decisive determinant of electoral success even after the formal transition to multiparty competition.
The paper's central contention is that democratic competition, when combined with the political decentralization that characterized Mexico's post-transition period, paradoxically reinforced rather than dismantled the organizational advantages inherited from the authoritarian era. This is a counterintuitive claim that cuts against the dominant theoretical expectation. Standard retrospective voting models hold that competitive elections create incentives for incumbents to deliver public goods, improve governance outcomes, and cultivate voter loyalty through demonstrated competence. The logic is straightforward: if voters can credibly punish underperformers by voting them out, governors who want to survive politically must govern well. The research challenges this mechanism at the subnational level in Mexico, suggesting that governors who controlled extensive party organizational networks — the clientelistic infrastructure of machine politics developed under PRI rule — outperformed their rivals regardless of measurable policy or economic outcomes under their watch. In other words, the machinery of authoritarian mobilization proved more electorally potent than any record of administrative achievement. This finding has profound implications not only for Mexico scholarship but for how we understand democratic consolidation in societies where single-party regimes left deep organizational imprints on political life.
The theoretical significance of this argument becomes clearer when situated within the broader literature on authoritarian legacies and subnational politics. Scholars including Edward Gibson, Guillermo O'Donnell, and more recently, work in the comparative subnational authoritarianism tradition, have documented how national-level transitions can coexist with entrenched subnational enclaves of machine-style rule. Mexico's federal structure created enormous variation across states in the pace and depth of democratization. Some states moved rapidly toward competitive multiparty systems with genuine accountability pressures; others remained effectively single-party domains well into the 2000s and 2010s. The paper under review contributes to this literature by systematically examining gubernatorial elections across this varied landscape, using incumbency patterns to trace the weight of machine organization against performance-based explanations. The decentralization dimension is particularly important: as fiscal resources and administrative authority were transferred downward to state governments during the 1990s and 2000s, governors accumulated significant patronage capacity. Rather than enabling programmatic governance, this resource access often fed the organizational machines that incumbents — or their chosen successors — could deploy at election time. The political economy of decentralization, in this reading, was captured by pre-existing organizational networks rather than transforming them.
For the international development and ODA research communities, these findings carry direct policy significance. A substantial portion of multilateral and bilateral development assistance to Mexico and comparable middle-income democracies in Latin America has been premised on assumptions about accountability linkages between citizen preferences and government behavior. Programs promoting good governance, anti-corruption measures, and public financial management reform at the subnational level often assume that competitive elections will reward reformist governors and punish those who capture rents or neglect service delivery. If party machine infrastructure systematically overwhelms performance signals in gubernatorial contests, then electoral competition alone provides insufficient incentive for the governance improvements that ODA frameworks typically target. This suggests the need for development partners to invest more seriously in demand-side accountability mechanisms — civil society organizations, investigative journalism, citizen auditing initiatives — that can create reputational and legal pressures on incumbents independent of the electoral channel. It also raises questions about the sequencing of decentralization reforms: devolving fiscal authority to subnational actors before robust accountability institutions are in place may inadvertently strengthen rather than discipline machine politics.
Looking forward, the Mexican case analyzed in this research carries significant lessons for the trajectory of democratic governance in the region and beyond. Mexico's 2018 election and the rise of Morena under Andrés Manuel López Obrador represented a dramatic disruption to the traditional PRI-PAN-PRD competitive framework, and the party's rapid consolidation at both federal and subnational levels raises new questions about whether a new form of machine politics is being constructed atop the ruins of the old one, or whether Morena's organizational model differs meaningfully from its predecessors. The 2024 transition to Claudia Sheinbaum and continued Morena dominance at the gubernatorial level will provide important test cases for the arguments advanced in this paper. More broadly, for researchers working on democratic backsliding and competitive authoritarianism globally, the Mexican subnational experience illustrates a pathway that does not require outright electoral manipulation or coercion: where organizational capacity is sufficiently asymmetric between incumbents and challengers, machine dominance can be reproduced through formally free elections. Understanding how this dynamic can be interrupted — through civil society development, institutional reform, or exogenous economic shocks that disrupt patronage networks — remains one of the most consequential research agendas in contemporary comparative politics. The paper contributes a rigorous empirical foundation for that ongoing inquiry, and practitioners working at the intersection of development assistance and democratic governance would do well to take its findings seriously.