Mayhew: Electoral realignments
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.
Mayhew. 2002. Electoral realignments: A critique of an American genre. New Haven: Yale University Press.
Main Point
Mayhew distills the realignment literature into 11 key claims and evaluates whether they are empirically valid. He finds that the performance of the claims ranges from mediocre to poor.
Evaluation of the 11 Realignment Claims
- Some elections are realigning and others are not: The attempt to dichotomize elections between critical and non-critical ones is flawed. Rather than thinking about partisan change in terms of peaks followed by long periods of stasis, we should think about change as a mixture of large, medium, and small effects. In other words, a gradual change model is better than a punctuated equilibrium model.
- Periodicity: The "30-year pattern" seems very dubious. If the critical/non-critical dichotomy is invalid, then so is the periodicity hypothesis.
- Tension and flash points model: The realignment model of a gradual build-up of tension that then triggers a major shift at a critical election is flawed. The realignment model looks too far back in history to explain shifts in party ID, which can often be better explained by proximate events.
- Concern and turnout are higher at realigning elections: Seems modestly accurate for 1860 and 1896, but not 1932.
- The emergence of a new cleavage causes a realignment: Not necessarily; events may do a better job of explaining changes in voter attachments.
- Ideological polarization is associated with realignment: No, the realigning elections (1860, 1896, 1932) don't all seem to be accompanied by ideological polarization, particularly 1932.
- Realigning elections are national; non-realigning ones are local: Empirically, this doesn't seem to be so clear.
- Realignment leads to major policy changes: Seems to be true for 1860s and 1930s, but not the 1890s.
- Realignments are about redistribution: Seems to be true of 1860s and 1930s, but not the 1890s.
- Major contributions to the political system are made about once a generation: No, political parties respond to changing preferences on an on-going basis.
- The 1896 election ushered in an era of business dominance: Probably not. Business was already doing fine before 1896 and appears not to have needed insulation.
In sum
- We should think of realignment as a gradual process, not a punctuated equilibrium / tipping point model.
- Events matter. Short-term variables (events like recessions and wars) may do a better job of explaining partisan change than long-term variables (like ideology change and/or new cleavages).
- Many political issues are valence issues (we all want less crime; the debate is about how to do it). When we factor them in, the realignment paradigm looks less useful, based as it is on divisions over policy positions.
