GUIDO CATAIFE

PUBLICATIONS

"An Electoral Model with Activists Applied to the Elections of 1989 and 1995 in Argentina." Co-authored with Norman Schofield. Mathematical Social Sciences 53 (2007).

Abstract: The mean voter theorem suggests that all parties should rationally converge to the electoral center. Typically this leads to an outcome which is unattractive to the rich. This paper develops a general stochastic model of elections in which the electoral response is affected by the valence (or quality) of the candidates. Contributions made by policy-motivated activists can influence valence, leading to the failure of the mean voter theorem. The model is then applied to the presidential elections in 1989 and 1995 in Argentina, to suggest why a candidate, who won in 1989 with a populist platform, was able to win in 1995 with quite different policies that favored the upper middle class.

WORKING PAPERS

"The Pronouncements of Paranoid Politicians." (Job Market Paper)

Abstract: Previous results in political competition have shown that, if candidates are expected utility maximizers and voters are risk averse, not announcing policy is never an equilibrium strategy. This paper models the strategic encounter of two office-motivated candidates who may announce policy or opt for no-announcement, according to the following premises. In the case of no announcement, the voters rank the candidates according to prior beliefs. In the case of announcement, the candidates cannot avoid a degree of noise in the voters’ interpretation of their announcements. I find that this simple deviation from the standard Downsian setting suffices to obtain no-announcements in equilibrium, even when voters are risk averse and candidates are maximizers of expected utility.

 

Also, I extend the model to study the equilibrium when candidates are ambiguity averse. An ambiguity averse candidate is interpreted as being concerned about an ongoing negative campaign against him. This negative campaign would consist in inducing the voters to adopt some interpretation of the candidate’s announcement unfavorable to his electoral performance. I show that under ambiguity aversion the candidates opt not to announce position under less stringent conditions than under expected utility.

Finally, I use actual data on U.S. Senate elections to test a distinguishing empirical implication of the model. The passed/failed result of the National Political Awareness Test, conducted on each candidate by Project Vote Smart, is used as a measure of whether or not the candidates announced positions during their campaigns. The price of TV spot ads across media markets is used to construct a proxy for the magnitude of noise in the voters’ interpretation of the announcements by the candidates. I find that the coefficient relating the probability of announcing platform to the magnitude of noise in the voters' interpretation of the announcements is statistically significant and has the sign predicted by the theoretical model.

"Electoral Oscillations in Argentina: Suggestions on Modeling the Political Economy of Hegemon and Periphery." Co-authored with Norman Schofield. Submitted.

 

Abstract: It has often been remarked that Latin American polities seem to swing in a random fashion between pro-market and anti-market democracies. What rationale explains a country's swing from left to right and back again? Is there some mechanism that causes swings in one country to infect its neighbors? We discuss the case of Argentina in 1990s and offer a model that provides an endogenous reason for these oscillations.

The electoral swings are not driven by an exogenous change in preferences, but by a change in the distribution of perceptions that the electorate has of the quality of candidates of left and right. Since these perceptions are influenced by the actions of activists who respond to the interests from the hegemon, electoral equilibria in Latin America are dependent on the electoral choices made in those hegemonic powers playing the game of empire.

WORK IN PROGRESS

"Media Prices and Campaign Spending in the U.S. Senate."

 

Abstract: A recent econometric study on U.S. House races has shown that increasing media prices are not the reason that campaign spending has augmented steadily in the last three decades.

 

In this paper, we show that this result does not apply for the U.S. Senate races. To make the results comparable, we construct a measure of the cost of political advertising equivalent to the one used in the mentioned study, although appropriate for Senate districts. We regress money spent on U.S. Senate campaigns against this measure of the cost of political advertising and find strong evidence that the coefficient relating these two variables is both economically and statistically significant. We discuss the reasons that the results for the House do not apply to the Senate.

 

OTHER PUBLICATIONS

"From Vegetal Selection to Biotech: the Economics of Germoplasm." THEOMAI 6 (2002). (Forthcoming translations from Spanish into English and German.)