Decentralizing the State: Elections, Parties, and Local Power in the Andes (review)

Kathleen O'Neill develops an electoral theory to explain why and when presidents decentralize political and fiscal powers to subnational governments. At first glance, the decisions of national executives to transfer power to subnational officials appear puzzling. Why would national politicians agree to give power away? O'Neill suggests that presidents are more likely to propose decentralizing measures when electoral support for their parties is weak at the national level but strong at the subnational level and the overall electoral support for their parties has been stable over time. Under these electoral conditions, the seemingly paradoxical decisions of presidents to transfer power downward can actually be viewed as rational strategies allowing their political parties to hold on to power.

The proposed electoral theory is intuitive and simple, and therefore appealing. It is based on a well-established assumption in political science: that politicians want to remain in office. In O'Neill's model, politicians think more broadly and less selfishly as they actually aim to maximize their political parties' chances to retain some degree of power. The author proposes a formal model (chapter 2), which includes four basic elements: the probabilities of winning national and subnational elections in the current and the following term, the distribution of power among national and subnational politicians, the decrease in power that decentralization would produce for the incumbent party in the current term, and how heavily the incumbent party discounts the chance of winning future elections. Whenever the expected utility of holding centralized power in two consecutive (the current and the coming) terms is lower than the utility of decentralizing power in the current term and holding decentralized offices in the coming one, the president will choose to decentralize. With the electoral data provided by the author, I tested the model in two decentralizing administrations, those of Belisario Betancur in Colombia and Carlos Andrés Pérez in Venezuela, and found that the formal model predicted the correct results.

O'Neill also uses a statistical logit model (chapter 3) to examine the relationship between presidents' decisions to decentralize and their electoral support. The three independent variables of interest are the percentage of the vote won by the president's party (at the national level) in the most recent presidential election, the percentage of contests won by the president's party at local levels, and the absolute difference between the percentage of votes for the president's party in the most recent and the previous presidential elections (used as a proxy for the discount rate of winning future elections). The author runs the statistical model in 26 presidential administrations. These include two administrations in Bolivia (1993 and 1997); seven in Colombia (1974 to 1998); four in Ecuador (1979 to 1995); four in Peru (1980 to 1995); and nine in Venezuela (1958 to 1998). The statistical model explains 54 percent of the variance found in the dependent variable, which is quite remarkable, given that it only considers the impact of three independent variables.

Not only does the book propose a simple formal model that translates into an effective statistical test, it also analyzes qualitatively the instances of decentralization in the five Andean countries, arguably the least studied countries of Latin America in political science. This, in itself, is an important contribution. The country-case chapters weave together electoral data, secondary resources, and interviews with politicians and public officials.

The in-depth analysis of country cases is uneven. The chapters on Colombia and Bolivia (4 and 5, respectively) are the most sophisticated and nuanced; it is interesting that these are the two countries where the author conducted fieldwork. Following these, a comparative chapter (6) tells the stories of electoral decisions to decentralize (and sometimes to recentralize) in Ecuador, Peru, and Venezuela. The country analyses in this chapter focus almost exclusively on the evolution of electoral results (which at times are hard to follow, due to...