In this book, Matthew Levendusky addresses fundamental questions of relevance to scholars of political behavior and political institutions. Given the broad consensus that political elites have become more polarized in the past half century, he focuses on the implications of elite polarization on belief systems in the mass public. He argues that elite polarization leads to partisan sorting among ordinary citizens, but not necessarily mass polarization. for Levendusky's purposes, sorting means that the party affiliation and ideology/issue positions of ordinary citizens have become aligned. In contrast, mass polarization refers to extremity in the distribution of preferences. Elite polarization leads citizens to become sorted into different camps, but not necessarily to become more polarized in their views.

Levendusky provides three key contributions. First, he articulates a theoretical argument regarding how sorting occurs and why. Second, he leverages experimental data to isolate the causal mechanism by which sorting occurs. Third, he elucidates the implications of sorting for both the mass public and political elites.

The author's theory is based on elite-driven cue taking. As elites become polarized, the connections between partisanship and ideology become clearer, citizens become aware of these clearer connections, and they align their partisanship with their ideology. The theory harkens back to central concepts articulated by Phillip Converse (“The Nature of Belief Systems in Mass Publics,” 1964): As elites send more messages regarding “what goes with what” (Converse 1964, p. 212), citizens receive and internalize such messages, leading to greater levels of constraint among partisanship and ideology/issue positions (or increased sorting, to use Levendusky's terms). Levendusky argues that the direction of influence is initially top-down: Elites polarize and the masses sort; once the masses sort, they can subsequently exert some checks upon elite behavior. Yet, what predicts first movement in the dance between elites and citizens; that is, what are the conditions under which elites decide to polarize? Do elites act because they anticipate a readiness and receptiveness among citizens? Are elites responding to movement among key segments of the citizenry? Levendusky argues that elite position taking is exogenous, but establishing this convincingly is a tricky business worthy of some additional attention.

After using content analysis to demonstrate that elite polarization has increased in the past few decades, Levendusky investigates its implications for mass belief systems. In his analysis of cross-sectional surveys from the American National Election Studies (ANES), he shows an association between increased elite polarization and increased clarity among the mass public regarding where the parties stand on important issues of the day, and, more importantly, an association between increased elite polarization and partisan and ideology/issue position alignment.

Levendusky also conducts an initial probe into whether there is heterogeneity in sorting. He shows that sorting is not isolated to particular regions, birth cohorts, or religious affiliations; he argues that it is a generalized, national phenomenon. These analyses are valuable, but more sustained investigation into heterogeneity is warranted. For example, African American voters, the staunch Democratic Party supporters that they are, are curiously conservative when it comes to some of the most important cultural issues at the center of the partisan divide. Controlling for partisanship, African Americans are significantly more conservative on cultural issues such as gay rights, abortion, and gender roles than are their white counterparts, and they are also more ideologically conservative. These conservative issue positions stand in direct opposition to the Democratic Party line and provide a stark counterexample to the general story that the author tells. An extension of this work would inquire into the conditions under which, and the cases for whom, partisan sorting does not occur, and why.

Another point of unexamined heterogeneity also deserves additional attention. Sorting, as Levendusky points out, can occur when ideology/issue positions move partisanship or when partisanship moves ideology/issue positions. His argument again resonates with a concept developed by Converse—that of centrality. Converse (1964) argued that the more central a belief element, the more likely that it would tether the remaining items in the belief system, and the more likely that changes in that particular element would ripple through the remaining items in the belief system. Levendusky leverages survey data to argue that under ordinary conditions, partisanship is a central element in most Americans' belief systems. As such, other belief elements (ideology and issue positions) will move to be consistent with partisanship. He does note that under particular contextual conditions, and for particular groups of individuals, belief elements aside from partisanship may be more central; at those times and for those individuals, partisanship itself may change to accommodate the more central belief element. Here, Levendusky's work is certain to inspire more sustained investigation into heterogeneity into which the mechanism of sorting occurs across context and among citizens.

Levendusky's strongest evidence comes from two experiments that demonstrate the effects of elite position taking on the alignment of partisanship with issue positions. The first, a natural experiment based on the issue of stem cells, identifies the increased role that partisanship plays in opinion on an evolving issue when there is a stark modification in elite position taking. The second, a survey experiment, manipulates the polarization and heterogeneity of party elites. Both experiments establish the author's central point that exogenous changes in elite polarization provoke sorting among the mass public. A natural extension of the work would probe how sticky these linkages are: Once policy opinions become linked with partisanship, can they become unlinked if elites begin to depolarize? Can a once-partisan issue evolve into a bipartisan one? On this possibility, he provides little guidance, but his results should trigger additional theoretical development and empirical inquiries.

Elites polarize, the mass public sorts, and, Levendusky notes, there are important consequences for voters and political elites. Sorted citizens vote with more party loyalty, they are more affectively charged in regard to the parties, and their policy opinions correspond more closely with the party line. Such effects likely ripple through the political system, affecting the conduct of campaigns and potentially hampering the possibilities for political compromise.

These consequences raise important normative questions about the nature of citizen competence and its implications for democratic politics.

What are the normative implications of sorting, especially in light of Levendusky's prediction that sorting may actually increase in the future? On the one hand, Converse's use of an organized and constrained belief system as a normatively desirable criterion for democratic competence suggests that we should take heart from Levendusky's core findings. On the other hand, do sorted citizens fall short in other respects? Are they more motivated (and selective) information processors (Milton Lodge and Charles Taber, “Three Steps Toward a Theory of Motivated Political Reasoning,” in Elements of Reason, ed. A. Lupia, M. D. McCubbins, and S. L. Popkin, 2000)? Are they more biased interpreters of political events—even objective political events (Larry M. Bartels, “Beyond the Running Tally: Partisan Bias in Political Perception,” Political Behavior 24 [2002]: 117–150)? Are they less open-minded in evaluating political candidates (Cindy D. Kam, “Political Campaigns and Open-Minded Thinking,” Journal of Politics 68 [2006]: 931–945)? Are citizens who are bound by ideology and partisanship “dogmatic and intolerant” (Donald R. Kinder and Don Herzog, “Democratic Discussion,” in Reconsidering the Democratic Public, ed. G. E. Marcus and R. L. Hanson, 1993, p. 354)? When large proportions of citizens sort, will we eventually lose the heterogeneity that Bernard R. Berelson, Paul F. Lazarsfeld, and William N. McPhee (Voting: A Study of Opinion Formation in a Presidential Campaign, 1954) saw as so crucial for democratic stability, flexibility, and reconciliation? And, do these implications of sorting suggest an incremental movement toward a more polarized electorate? Although Levendusky argues that sorting and polarization are conceptually distinct phenomena, will the perceptual and motivational biases inherent in sorting eventually lead to mass polarization?

The Partisan Sort offers a compelling theoretical and empirical account of the connection between elite polarization and mass belief systems. Levendusky's lucid book convincingly demonstrates that the evolution of the partisan landscape in America has important implications at both the mass and elite levels. As that evolution continues, it also provokes us to ponder whether increased sorting will bode well or poorly for democratic politics in America.