In modern democracies, political representation is fundamentally an election-based connection between policy makers and citizens (cf. Pitkin 1972). It can be conceived of as purely procedural representation concerned with vote-seat correspondence or as substantive representation where the starting point is citizens' preferences rather than their votes (cf. Powell 2004). Ideological congruence is thus desirable since policy makers should and do take account of what the citizens want, leading to the model of responsible party government (Adams 2001). Such programmatic linking, Kitschelt et al. (1999) show, creates cycles of responsiveness and accountability. <p>The vast majority of studies concerned with substantive representation concentrate on the connection between election rules and ideological congruence. In so doing, they build on theories of strategic party competition and government formation in majoritarian and proportional systems (Cox 1997; McDonald and Budge, 2005; Powell 2000). When voters' preferences are single-peaked, Black's Median Voter Theorem shows that the median voter is the decisive element, representing the majority outcome (Black 1958). Any public policy located close to the median voter's preference will be superior in the sense of producing a collectively desirable outcome and thus higher social utility. Better ideological correspondence occurs when the position of the government is closer to that of the median citizen. Assuming a one-dimensional issue space, a commonly found result is that proportional systems provide a better fit between voters, median legislators, and governments (Huber &amp; Powell 1994; Klingemann et al. 2006; McDonald and Budge 2005; Powell 2000). However, considering different policy issues, Holmberg (2000) finds in the case of the four Nordic European countries higher correspondence between voters and party representatives on issues that are both salient and politicized.&nbsp; </p><p>All these studies concentrate on European parliamentary systems with political parties being the major vehicles of representation. Much less work has been done in the context of Latin American countries where studies tend to be rather recent, a majority of them confined to the analysis of single cases (Hagopian 1998; Levitsky 2001; Mainwaring 1999). One major intent to analyze this topic from a larger cross-national perspective relies on elite surveys to analyze the connection between party elites and voters in nine Latin American countries (Luna and Zechmeister 2005). The authors find a positive relation between representation and party system institutionalization, the presence of leftist parties and higher levels of socioeconomic development. However, studies taking into account the representative role of the president are rare to nonexistent. One exception for the Latin American case are Stokes (1999, 2001) accounts of policy switches in the region during the 1980s until the mid 1990s, when administrations carried out neoliberal &lsquo;reforms by surprise' despite having run on anti-neoliberal election platforms. However, she concludes that mandate switching is rather atypical. Only recently Samuels and Shugart (2006a; 2006b) provided an explanation of how the structure of presidentialism will affect political parties and how presidents and their parties are related to each other. Nevertheless, by modeling the president purely as agent of his own party, the role of the president as independent actor is neglected. </p><p>Wiesehomeier and Benoit (2009) argue and persuasively show that the neglect of the president as a self-contained actor and the inherent assumptions of the alignment of presidents and their parties are not plausible. Actors have institutional and strategic incentives to position themselves individually, thus ideological coherence between presidents and their parties cannot be taken for granted as in separation of power regimes presidents may have reasons to not be limited to the policy platforms held by their own parties, reasons that my comprise political factors and institutional arrangements. With the help of an original dataset from expert surveys conducted by the authors to obtain reliable estimates on the positions and importance attached to different policy dimensions by both 18 presidents and 146 political parties, in a common policy space, in 18 Latin American countries, the authors are able to explore and to explain positioning of the president relative to his own party. Their results show that presidents tend to position themselves independently of their parties more in bicameral and proportional systems with large districts magnitudes, when they enjoy considerable constitutional powers, and the more he and his party diverge in the importance attached to a policy dimension. Furthermore, this latter effect is more pronounced in non-concurrent elections. These results clearly demonstrate that studies concerned with policy competition and the analysis of policy outcomes as a function of preferences and policy positioning have to be aware that the policy issue space in presidential democracies is inhabited by both, parties and presidents. </p><p>Yet, having established that a president indeed has the potential for autonomous positioning can only be the starting point for considering substantive representation. Understanding the mechanisms of issue congruence in presidential democracies by accounting for the president will also avoid flawed conclusions about expected policy outcomes or observed patterns of executive-legislative relations. A key goal of this project is therefore to fill the void in current research concerning representation in presidential systems and to shed light on the representative role of the president and issue congruence achieved in the legislative arena versus the presidency. </p>