The study tackles a long-standing puzzle in public administration. Decades of scholarship have established that managerial quality matters for government performance, and that chief executives with greater public sector experience tend to preside over more effective organizations. Yet the evidence is inconsistent: some studies find strong effects of leaders’ backgrounds, others find none. The authors argue that much of this inconsistency stems from a neglect of context. Most research on the management-performance relationship has been conducted within a single country, treating the surrounding political environment as background noise rather than as a variable that can strengthen or weaken the link between a leader’s skills and measurable outcomes.

To build their theory, the researchers draw on the classic distinction, developed by the political scientist Arend Lijphart, between majoritarian and consociational models of democracy. In majoritarian systems, two main political blocs compete under a winner-takes-all logic, and the losing side serves as a watchful opposition expecting to alternate in power. This adversarial arrangement, the authors contend, creates powerful incentives for executives to perform well, because citizens can more easily identify and reward or punish the party responsible for policy outcomes. Party leaders, in turn, reward politicians whose performance enhances the party’s reputation, promoting their careers within disciplined, hierarchical organizations.

Consociational or power-sharing systems work very differently. Broad coalitions bring most significant parties into government, blurring the line between administration and opposition. Party leaders have incentives to bargain and collaborate with rivals rather than defeat them, and accountability becomes diffuse: when everyone shares credit, no one can be clearly blamed. The authors point to additional mechanisms that erode performance incentives in such systems, including frequent party switching, coalition-driven rotation of government posts, and the dilution of clear policy platforms through compromise. In these circumstances, a politician’s accumulated experience may simply not matter very much, because the electoral arena is not perceived as a zero-sum contest in which superior management translates into political advantage.

Testing this argument required a comparison of two countries that are alike in many respects but differ sharply in their patterns of political competition. Mexico and Colombia both have presidential systems with separation of powers, both underwent decentralization in the 1980s and 1990s, and both assign their intermediate-level governments, 32 Mexican states and 32 Colombian departments, major responsibility for education and health services. Yet their party systems diverged dramatically. Until 2018, Mexico operated a disciplined three-party system in which the Institutional Revolutionary Party, the conservative National Action Party, and the Party of the Democratic Revolution competed adversarially, with rare party switching and strong party loyalty at the state level.

Colombia, by contrast, carries a deep tradition of power sharing. Between 1958 and 1974, the Liberal and Conservative parties governed together under the National Front pact, alternating in the presidency and splitting cabinet posts evenly, an arrangement designed to end political violence. Even after fully competitive elections resumed, winning politicians continued offering co-governing roles to the losing side. The 1991 constitution opened the field to new parties, more than twenty of which now hold Senate seats, but coalition-building norms persisted. The authors illustrate the pattern with the department of Cundinamarca, where a defeated 2007 gubernatorial candidate supported the victor’s government and then won the 2011 election at the head of a coalition that included the very parties that had beaten him four years earlier.

Quantitative evidence reinforces this contrast. The authors calculated the effective number of parties in subnational elections between 1998 and 2015. Gubernatorial races looked similar in both countries, averaging roughly 2.4 effective candidacies. But assembly elections told a different story: Mexico averaged 2.64 effective parties, close to its gubernatorial figure and reflecting high party discipline, while Colombia averaged 5.43, more than double its gubernatorial figure. Colombian governors, in other words, must assemble sprawling, informal coalitions to govern, while Mexican governors could rely on cohesive partisan majorities.

The empirical analysis used two-way fixed-effects regression models on panel data covering the 32 Mexican states from 1995 to 2010 and the 32 Colombian departments from 2001 to 2015, spanning multiple gubernatorial administrations in each country. Performance was measured with two indicators managed primarily at the intermediate level: high school enrollment rates as an output of education policy, and infant mortality rates, deaths in the first year of life per 1,000 live births, as an outcome of health policy. The researchers hand-collected biographical data on every governor, classifying years of public sector experience at the local, state or departmental, and national levels. Mexican governors averaged 14.2 years of public experience, concentrated in national and state posts, while Colombian governors averaged 9.6 years, distributed more evenly across levels. The models controlled for partisan alignment with national and subnational legislatures, party affiliation, margin of victory, postgraduate education, revenue per capita, rural population density, and violence, proxied by homicide rates in Mexico and internally displaced populations in Colombia.

The results split cleanly along the adversarialism divide. In Mexican states, governors’ prior experience in local and state government significantly reduced infant mortality, with each additional year of local or state experience lowering infant deaths, while each additional year of federal-level experience was associated with worse outcomes. Partisan alignment with the health minister also improved health performance. In Colombian departments, by contrast, none of the experience measures showed a significant relationship with either infant mortality or high school enrollment, and a power analysis confirmed that this null result was not an artifact of sample size. Joint models pooling both countries confirmed the pattern: the effect of state-level experience on infant mortality differed significantly between the two systems, negative in Mexico but not in Colombia. Education enrollment showed weaker effects overall, suggesting that the experience-performance link is strongest for outcomes, like infant survival, that depend heavily on executive coordination of health services.

The authors are careful to note that their findings do not constitute a normative endorsement of adversarial politics over power sharing. Adversarial systems may extract more value from experienced executives, but they also expose governments to stronger opposition and a more turbulent external environment. The practical implications cut both ways. In adversarial systems, candidate recruitment should prioritize managerial and technical skills that translate directly into performance, while power-sharing systems may benefit more from executives skilled in coalition building and networking. And where opposition parties are co-opted rather than competitive, accountability may need to come from civil society and the media instead. More broadly, the study delivers a methodological message to the field: theories of public management built in one country, however elegant, cannot be assumed to travel. What works in Mexico may not work in Colombia, and the difference is not culture or capacity but the structure of political competition itself, a variable that shapes whether a leader’s hard-won experience ever gets the chance to matter.

Subject of Research: The moderating effect of political competition on the relationship between executives' experience and subnational government performance in Mexico and Colombia

Article Title: Adversarialism and power-sharing: the effect of political competition on the management-performance relationship

Article References: Bello-Gomez, R. A., Avellaneda, C. N., & Olvera, J. G. (2024). Adversarialism and power-sharing: the effect of political competition on the management-performance relationship. Global Public Policy and Governance, 4(4), 389-419. https://doi.org/10.1007/s43508-024-00104-6

Image Credits: AI Generated

DOI: 10.1007/s43508-024-00104-6

Keywords: political competition, government performance, public management, adversarialism, power-sharing, governors, Mexico, Colombia, subnational governments, infant mortality, education enrollment, comparative public administration