UC Berkeley

This dissertation argues contemporary party sorting --- the phenomenon by which liberals and conservatives sorted between parties on effectively ever major policy issue in the latter part of the 20th century --- is shaped by the 1960s racial realignment. I contend the parties more easily divided on issues like abortion and gun control in the 1980s-1990s, because the parties in the electorate had already divided on civil rights in the 1960s. Left-right attitudes on civil rights have long overlapped left-right divides on essentially every other major policy issue among ordinary voters. These trends date to the earliest public opinion polls and persist even among those who know little about politics. When the parties divided on race in the 1960s, pre-existing ties between civil rights and other policy views encouraged the parties to take positions on newly salient issues, such as abortion or gun control, which reinforced this racial divide.

This worked through two complimentary mechanisms. First, once the parties begin to divide on race, conservative voters on abortion or guns begin entering the Republican coalition. This creates an incentive for Republican candidates to stake conservative positions to win the party's nomination. Second, by taking conservative positions on race correlated policies such as abortion or guns, Republican candidates reinforced their appeal to conservative Democrats in the general election.

Chapter 3 uses historical public opinion dating to the 1930s to present an exhaustive analysis of issue linkages between race and policy views. I find that (1) voters that express more conservative racial views have long expressed more conservative attitudes on essentially every other major policy issue; (2) voters expressed this package of issue attitudes before the parties established positions on many now salient policies including abortion, gun control, environmentalism, women's rights and gay rights; (3) voters who do not know where the parties stand on issues still package these attitudes together.

Chapter 4 explores the role of interest groups, politicians and media figures in constructing the contemporary alignment of party and ideology. Using the development of abortion's partisan divide as a case study, I argue that because racial conservatives entered the Republican coalition before abortion became politically activated, issue overlap among ordinary voters incentivized Republicans to oppose abortion rights once the issue gained salience. Likewise, because pro-abortion voters generally supported civil rights, once the GOP adopted a Southern strategy, this predisposed pro-choice groups to align with the Democratic party. A core argument is that pre-existing public opinion enabled activist leaders to embed the anti-abortion movement in a web of conservative causes that had become newly prominent in the Republican party. This is despite many leaders of the pro-life movement's desire to align with other liberal causes inside the Democratic party.

Chapter 5 analyzes this trend in state and sub-state level races in Southern general election contests which feature candidates that had sorted on civil rights with election contests where candidates overlap on civil rights. I find consistent evidence that attitudes towards race and civil rights propelled conservative Democrats from the Democratic party once candidates in sub-national elections had divided by racial views. In doing so, these racially conservative voters brought conservative views on other policy issues with them, too.

Chapter 6 then explores how voters who are liberal on race, but conservative on other policy dimensions, reconcile these differences. I find that these cross-pressured voters consistently update their non-racial policy views to match their pre-existing issue attitudes, but this does not happen in reverse. Finally, chapter 7 discusses the implications of this dissertation for contemporary politics in the United States and abroad.