Disclaimer. Don't rely on these old notes in lieu of reading the literature, but they can jog your memory. As a grad student long ago, my peers and I collaborated to write and exchange summaries of political science research. I posted them to a wiki-style website. "Wikisum" is now dead but archived here. I cannot vouch for these notes' accuracy, nor can I say who wrote them.
Abramowitz and Saunders. 1998. Ideological realignment in the American electorate. Journal of Politics 60:634-652.
According to the existing literature, partisanship is stable and largely inherited from your parents; realignments occasionally happen, but only after some sort of crisis or perhaps the mobilization of a new group of voters. If you believe this literature, though, you'll have a hard time explaining the Republican realignment in the 1980s and 1990s that culminated in the 1994 Republican electoral victories.
The Reagan years pushed the parties toward increased polarization. In turn, increased polarization has made it easier for citizens to differentiate between the two parties (Table 4). The consequences are dramatic: The Democratic party lost many conservative supporters to the Republicans (Table 6). Whites, males, and southerners (and especially white male southerners) abandoned the Democrats en masse (Table 1).
Back before this new polarization, people just stayed with their parents' party; after all, the parties didn't seem very different. But polarization has led people to abandon their parents' party and join a party based on their ideology (Table 2, Table 3).
The authors are confident that respondents' ideology explains the shift in party ID--and not the reverse. Since they have panel data from 1992-1994, they can test the causal direction by correlating 1992 ideology against 1994 party ID. As always, the authors find supporting evidence; 1992 ideology explains 1994 party identification far better than 1992 party identification predicts 1994 ideology (Figure 2).
To summarize: Citizens' ideology seems to explain party identification better in 1994 than it did in 1978, and parental partisanship seems to explain party identification worse (Table 7).
The authors use NES panel data, primarily from the 1976-1978 panel and the 1992-1994 panel.
Have later years supported or refuted their arguments? Apparently the 1996-1998 NES data suggest a slight reversal in these trends. See Bartels (2000).
Research by the same authors
Research on similar subjects