Taking Ideas Seriously in Political Science: The Diffusion of Presidentialism in Latin America after Independence

Today it is rare to find a political scientist who rejects that "ideas matter"-at least in the abstract. But if ideationally inclined theorists have gained seats at the disciplinary table, their approaches still occupy disadvantaged positions. Ideational theories typically must confront nonideational alternatives to achieve salient publication, but nonideational theorists routinely design and publish research without considering ideational alternatives. This is even true on topics where all scholars seem to agree a priori on the importance of ideas. Erratic attention to ideas appears to be justified by widespread views that even if ideas plausibly "matter," they are too intractable to address in concrete research: too difficult to measure empirically, to relate in explanatory ways to action, or to connect to goals of theoretical generalization. This article first highlights major problems with these views in the abstract, and then illustrates them in the example of early Latin American constitutional design. On this terrain, there are good reasons to think that an ideational account connects in more concrete ways to available evidence than leading nonideational hypotheses about constitutional choice. Political science should move toward better balanced debates between plausible explanations, upgrading the rigor of the discipline overall.