As Downs famously argued, voters typically lack comprehensive information about the realm of politics: such information is costly and the incentives to obtain it are relatively low (Downs, 1957). Moreover, instead of being exposed to the same amount of limited political information from similar sources, citizens tend to acquire it from an equally limited number of different intermediaries. Finally, the information that flows from those intermediaries to individuals is not necessarily neutral. Instead, it may be biased in ways that end up favoring particular candidates, parties, or positions (Zaller, 1992; Huckfeldt &amp; Sprague 1995; Beck et al., 2002). <br /><br />It is hardly surprising that all this has been seen as potentially consequential for electoral behavior. In fact, the notion that voters should be seen as embedded in concrete ?informational environments? ? understood as relatively stable filters for political communication that structure all kinds of information they acquire (Huckfeldt et al. 1995) ? was one of the main foci of some the very earliest efforts at systematic survey-based electoral research, such as those conducted by the Columbia University?s Bureau of Applied Social Research in the 1940s (Lazarsfeld, Berelson, and Gaudet 1944; Berelson, Lazarsfeld, and Mcphee 1954). However, both Lazarsfeld?s own findings about the mass media?s ?minimal effects? and the prevalence of political predispositions as causes of political behavior and the rise of alternative approaches to the study of electoral choices ? generally focusing on the overpowering role of social, ideological or partisan predispositions on the vote ? have led to general skepticism about the relevance of both interpersonal and impersonal communication for the study of voting choices (Ansolabehere, Behr &amp; Iyengar, 1993; Zuckerman 2005b). <br /><br />The Comparative National Elections Project (CNEP) has a made crucial contribution to reverse this trend, in two main ways. First, it has allowed research on the influence of interpersonal communication on the vote to move from the use of sub-national samples of the United States (Huckfeldt 1983, 1984; Huckfeldt and Sprague 1987, 1995) to the use of nationally representative samples of the electorate in an increasing number of democratic political systems (Burbank, 1997; Beck, 2002; Liu, Ikeda, and Wilson, 1998; Zuckerman, Kotler-Berkowitz, and Swaine, 1998; Levine 2005). Second, it has allowed scholars to go beyond the focus in interpersonal discussants as sources of information and center their attention on the interplay between the effects on the vote exerted by networks of social interaction and those exerted by a number of other informational intermediaries located between individuals and the more remote world of politics, such the organizations to which one belongs, the mass media to which one is regularly exposed, or even the messages conveyed by parties themselves when targeting voters through direct mobilization efforts (Beck, 1991; Dalton, Beck, and Huckfeldt, 1998; Beck et al. 2002; Elder and Greene, 2003; Beck and Curtice, forthcoming)