Electoral Proximity and the Political Involvement of Bureaucrats: A Natural Experiment in Argentina, 1904

Abstract

In this paper, I use a slightly modified version of the Becker–Stigler model of corrupt behavior to explain bureaucratic political involvement. Since bureaucrats prefer higher rewards and not to support losing candidates, we expect them to become politically involved near elections – when rewards are expected to be higher, and information more abundant. Taking advantage of a natural experiment, I employ differences-in-means and differences-in-differences techniques to estimate the effect of electoral proximity on the political involvement of justices of the peace in the city of Buenos Aires in 1904. I find a large, positive, and highly local effect of electoral proximity on their political involvement, with no appreciable impact in the months before or after elections. (author's abstract)