557 P rocesses of direct democracy are an increasingly prominent part of state and local politics (Matsusaka 2005). Direct democracy processes, however, are con- troversial in ways that the more familiar institutions of rep- resentative democracy are not (see for example Broder 2000, Schrag 2004, Smith and Tolbert 2004, Smith 1998). Some of these controversies focus on the outcomes pro- duced by ballot proposition elections (Gerber, Hajnal, and Louch 2002), others focus on the (in) ability of citizens to cope with the demands process (Magleby 1984). In this arti- cle we examine another line of controversy, the role of spe- cial interests in the politics of ballot propositions (Smith 1998; Gerber 1998). Broadly speaking, there are two expec- tations of the role of interest groups and direct democratic institutions. On the one hand ballot proposition process simply allows the continuation of politics as usual but in a different arena. Two broad sets of well-heeled and well defined groups—one left wing one right wing—will join battle. Group politics in ballot proposition politics may thus be characterized by relatively stable coalitions arrayed along a single dimension. A second expectation is that ballot propositions will pro- mote an alternative form of politics, one that is much more grass roots and participatory. Under this view, direct democ- racy is expected to foster more pluralist and fluid coalitional structure because it allows the unbundling of politics from the aggregated packages offered by parties into single issues. In this way, cross-cutting cleavages and cross-party coali- tions may be more readily expressed. If this view is correct at least one of the dimensions may be one where we see out- sider groups (grass roots, citizen inspired) pitted against the insider groups (political parties, unions, or business associ- ations) as the process is used as the Progressives intended. In this article we look to see which of those views of group politics and direct democracy is the more accurate by exam- ining patterns of campaign contributions in ballot proposi- tion elections in California over the period from 2000 to 2004 using social network analysis. Similarities in the patterns of participation by interest groups across multiple ballot propo- sitions are used to identify coalitions in an “interest group space.” Similarities in the propositions, defined by co-partici- pation of interest groups are used to define the structure of “policy space.” Dimensionality and clustering of both interest groups and issues allow us to speak to the coalitional struc- ture of direct democracy politics as a whole, rather than just on a proposition by proposition basis. The article begins with a discussion of the role of groups in direct democracy and outlines two rival pictures of group involvement in the process which we then subject to empiri- cal examination, in three steps. In the first step we present descriptive data on contributions to California ballot proposi- tions between 2000 and 2004. As we show, the level and dis- tribution of contributions to ballot propositions are markedly different from those seen in candidate elections. We then move to examine the ideological stability of contributions. Our second empirical step is, therefore, to examine the structure of the policy space and the interest group spaces independently. We assess the extent to which underlying common dimensions or clusters of issues can be detected on the basis of sharing common major donors. We also assess the extent to which interest groups form stable and coher- ent coalitions as opposed to a series of temporary alliances by examining similarity in the profiles of the contributions they make to campaigns. Our third step is to examine the policy and interest group spaces simultaneously to provide descriptions of the joint structure of issues and interests. A simple picture of this space might be one in which there are stable coalitions of interest groups that regularly take opposing positions across a wide range of ballot propositions. Such a picture would strongly support the “direct democracy as normal politics” view. By contrast, a highly complex (multi-dimensional and Just How Pluralist Is Direct Democracy? The Structure of Interest Group Participation in Ballot Proposition Elections SHAUN BOWLER AND ROBERT HANNEMAN, UNIVERSITY OF CALIFORNIA, RIVERSIDE In this article we look at the pattern of interest group engagement in direct democracy elections. Using a social network analysis of campaign contributions we examine whether or not patterns of group conflict conform to a stable split along a single (left/right) dimension. While we find that it is the case that there are stable com- ponents to group participation in direct democracy elections and that group alliances do lie along a left/right dimension we also find this characterization to be a simplification of actual patterns of engagement. Not only is there evidence of single-issue engagement, there is also evidence of a second—insider/outsider—dimension to direct democracy politics. NOTE: We would like to thank the reviewers of the article for their very careful reading and very helpful suggestions in the course of revis- ing this study. Political Research Quarterly, Vol. 59, No. 4 (December 2006): pp. 557-568