We propose a theory of political parties in which interest groups and activists are the key actors, and coalitions of groups develop common agendas and screen candidates for party nominations based on loyalty to their agendas. This theoretical stance contrasts with currently dominant theories, which view parties as controlled by election-minded politicians. The difference is normatively important because parties dominated by interest groups and activists are less responsive to voter preferences, even to the point of taking advantage of lapses in voter attention to politics. Our view is consistent with evidence from the formation of national parties in the 1790s, party position change on civil rights and abortion, patterns of polarization in Congress, policy design and nominations for state legislatures, Congress, and the presidency.

What is a party? This article presents the argument that rmal party apparatus is only one part of an extended network of interest groups, media, other advocacy organizations and candidates. The authors have measured a portion of this network in the United States systematically by tracking lists of names transferred between political organizations. Two distinct and polarized networks are revealed, which correspond to a more liberal Democratic group and a more conservative Republican group. Formal party organizations, like the Democratic National Committee and the Republican National Committee, tend to receive information within their respective networks, which suggests that other groups serve to funnel information towards the formal party.

We investigate the relationship between controversial roll call votes and support for Democratic incumbents in the 2010 midterm elections. Consistent with previous analyses, we find that supporters of health care reform paid a significant price at the polls. We go beyond these analyses by identifying a mechanism for this apparent effect: constituents perceived incumbents who supported health care reform as more ideologically distant (in this case, more liberal), which in turn was associated with lower support for those incumbents. Our analyses show that this perceived ideological difference mediates most of the apparent impact of support for health care reform on both individual-level vote choice and aggregate-level vote share. We conclude by simulating counterfactuals that suggest health care reform may have cost Democrats their House majority.

What are the primary factions within the Democratic and Republican parties, and to what extent do rival factions cooperate? We address these questions using a unique data set of information sharing between party organizations, media outlets, 527s, and interest groups. Using social network methods, we identify two major information-sharing clusters, or expanded party networks; these networks correspond to a liberal/Democratic grouping and a conservative/Republican grouping. We further identify factions within each party network, but we find a high degree of cooperation between party factions. That is, our data suggest that beneath the intraparty disagreements we observe in primary elections and policy debates there is a subterranean pattern of organizational cooperation.

This article builds on Matthews and Stimson's (1975) study of legislative cuetaking, analyzing the extent to which legislators sitting next to each other influence each others' voting behavior. Data come from three decades of roll call votes in the California Assembly, a chamber in which each member is paired with a deskmate. By comparing deskmate pairs with nondeskmate pairs, I find that legislators vote identically to their deskmates on a sizeable subset of roll calls. This deskmate effect appears to remain strong even as a rival influence-legislative partisanshipincreases in strength. A typical state or federal legislator will face hundreds or even thousands of roll call votes during a legislative session, mostly concerning topics on which he/she possesses little or no expertise. As Matthews and Stimson (1975, p. 25

Many theoretical and empirical accounts of representation argue that primary elections are a polarizing influence. Likewise, many reformers advocate opening party nominations to nonmembers as a way of increasing the number of moderate elected officials. Data and measurement constraints, however, have limited the range of empirical tests of this argument. We marry a unique new data set of state legislator ideal points to a detailed accounting of primary systems in the United States to gauge the effect of primary systems on polarization. We find that the openness of a primary election has little, if any, effect on the extremism of the politicians it produces.

No abstract

Why are American politicians "single-minded seekers of reelection" in some decades and fierce ideological warriors in others? This article argues that the key to understanding the behavior of members inside a legislative chamber is to follow the actions of key figures outside the chamber. These outsiders-activists, interest groups, and party bosses-use their control over party nominations, conditioned on institutional rules, to ensure ideological behavior among officeholders. To understand how vital these outsiders are to legislative partisanship, this article takes advantage of a particular natural experiment: the state of California's experience with cross-filing (1914-59), under which institutional rules prevented outsiders from influencing party nominations. Under cross-filing, legislative partisanship collapsed, demonstrating that incumbents tend to prefer nonpartisanship or fake partisanship to actual ideological combat. Partisanship quickly returned once these outsiders could again dominate nominations. Several other historical examples reveal extralegislative actors exerting considerably greater influence over members' voting behavior than intralegislative party institutions did. These results suggest that candidates and legislators are the agents of activists and others who coordinate at the community level to control party nominations.