Chapter 12 Key Takeaways: Partisanship, Polarization, and Sorting
Core Concepts at a Glance
Partisan Sorting (Not the Same as Polarization) - Sorting: the process by which liberals have become Democrats and conservatives have become Republicans — ideological alignment with party identity - Polarization: the movement of policy opinions toward the extremes — a distribution shift, not just a realignment - You can have sorting without polarization: if the distribution of opinion is stable but its alignment with party changes, that is sorting - Both have occurred in America, but sorting has been more dramatic and more consequential for electoral politics
The Three Types of Polarization - Ideological polarization: actual movement of policy opinions toward extremes (mixed evidence among voters; clear evidence among elites) - Affective polarization: growing emotional hostility toward the out-party (strong evidence; growing faster than ideological polarization) - Elite polarization: dramatic ideological sorting and extremism among elected officials, donors, and activists (strongest evidence; drives the others)
Negative Partisanship - Many voters are motivated more by hostility to the out-party than enthusiasm for their own - "Independent leaners" often behave like partisans — they vote consistently for one party primarily to prevent the other from winning - The genuine persuadable electorate is much smaller than the apparent "independent" category suggests
Nationalization of Elections - Down-ballot races (Senate, House, state) increasingly track presidential vote share rather than local candidate quality or local issues - Driven by: sorting of elected officials, collapse of local news, campaign finance nationalization, party cue-taking - Consequence: incumbency advantage has diminished; split-ticket voting has declined; local political entrepreneurs face structural headwinds
The Big Sort: Geographic Clustering - Americans are living in increasingly partisan communities, partly through self-selection into different types of places - The red-and-blue map overstates Republican geographic dominance by weighting area rather than population - Geographic sorting reduces cross-partisan contact, reinforces affective polarization, and creates distorted representation in winner-take-all electoral systems
Polarization's Effects on Polling - Partisan differential nonresponse: the party whose voters are more enthusiastic overrepresents itself in poll samples - Social desirability bias: some voters may underreport preference for candidates perceived as socially unacceptable - Herding: pollsters adjust toward consensus to avoid being outliers, compressing apparent uncertainty - No perfect correction for any of these; transparency about methodology is essential
Critical Distinctions to Remember
| Concept A | vs. | Concept B |
|---|---|---|
| Sorting | vs. | Polarization |
| Ideological polarization | vs. | Affective polarization |
| Elite polarization | vs. | Mass polarization |
| Negative partisanship | vs. | Positive partisanship |
| Nationalization of elections | vs. | Federalization of policy |
| Big Sort (behavioral) | vs. | Compositional change (demographic) |
| Differential nonresponse | vs. | Social desirability bias |
The Analyst's Toolkit for a Polarized Environment
When working with polls in a polarized context, ask:
- Is there reason to suspect differential nonresponse? Check the partisan composition of your sample against historical baselines.
- What is the mode-partisan interaction? Social desirability effects are stronger with live interviewers than online. Compare phone vs. online partisan breakdowns.
- Are other polls showing similar results, or is there herding? Look for outlier polls that were not published — these can signal where the real uncertainty lies.
- Is the persuadable universe as large as the "independent" category implies? Disaggregate independents by vote history to identify true persuadables vs. negative partisans.
- How nationalized is this race? Check the historical correlation between presidential and down-ballot vote share in this geography to estimate how much local factors can actually move the needle.
Themes Activated in This Chapter
- Theme 3 — Prediction vs. Explanation: Nationalization and sorting change the predictive power of party ID; understanding sorting explains why.
- Theme 5 — Gap Between Map and Territory: The red-and-blue map is the most visually prominent illustration of this gap in all of political analytics.