Partisan theories of political economy expect that bondholders will panic with the election of a left-wing presidential candidate. The latter seems to be what happened in Brazil in the 2002 presidential elections. However, quantitative analysis of perceptions of sovereign credit risk in Argentine, Brazilian, Mexican, and Venezuelan presidential elections from 1994 until 2007 shows no real evidence of a link between partisanship and perceptions of risk, even if the left-right divide is further broken down into left, center-left, centerright, right. Instead, international and domestic economic fundamentals have a stronger influence on risk evaluations. Qualitative analysis of the individual presidential elections shows the importance of policy uncertainty in explaining why certain electoral periods seemed more critical than others and how bondholders select between multiple equilibria. This research helps shift political analysis away from partisanship and more in the direction of policies and articulation. [ABSTRACT FROM AUTHOR]