Anthony Downs’s (An Economic Theory of Democracy, 1957) seminal work on electoral politics not only advanced an analytical technique that remains central to the study of elections today but also offered findings that have become tantamount to axioms. In particular, a (somewhat superficial) reading of the Downsian model of two-party competition conjectures that optimal behavior involves convergence to the median voter. While this finding effectively characterized the electoral politics of Downs’s time, it leaves something to be desired in the modern era of highly polarized parties. This tension has sparked myriad efforts by scholars to square the empirical reality of polarization with Downsian convergence. Justin Buchler makes an important contribution to this conversation with his excellent book, Incremental Polarization, in which he argues that an understanding of why legislators adopt extreme policy positions yet win reelection requires an integrated model of electoral and legislative politics.
Buchler contends that a persistent theoretical divide between models of elections and legislative behavior is responsible for the disconnect between theory and reality. Spatial models of elections often treat candidate ideal points as exogenously derived, as opposed to being the product of iterative, binary decisions. Spatial models of legislative decisions frequently overlook the numerous electoral demands that members face in casting roll call votes.
Therefore, Buchler develops a unified theory that integrates these often disparate models, while attempting to avoid making assumptions that compromise foundational spatial principles in a way that extant efforts to reconcile spatial models with the empirical realities of polarization have done. He does so by making the legislative process, by which members incrementally adopt their ideal point, the initial stage of the unified game, challenging the more conventional perspective that the winning candidate’s ideal point (at stage one) is implemented upon assuming office (at a later stage). Buchler’s game unfolds as follows: 1) The majority party caucus elects its leader on the basis of the leader’s disciplinary platform; 2) members engage in a sequence of roll call votes subject to the leader’s disciplinary platform; and 3) voters elect members on the basis of the ideal points revealed by roll call voting.
Buchler, quite ingeniously, demonstrates that roll call voting can lead to conditions under which voters prefer candidates with more distant unrevealed ideal points. A “proximity paradox” emerges when extreme ideal points are concealed by a legislative agenda that includes items that “move policy far from the status quo and divide the ideological factions cleanly around the central ideological break” (pp. 44–45). While this does not necessarily explain why a more extreme revealed ideal point would defeat a more moderate one, it at least gives us context for understanding why pure proximity voting might break down. Furthermore, the core spatial principle that moderation yields electoral benefits is preserved by this line of argumentation. This alone strikes us as an important, stand-alone contribution to the extant literature.
Given that an incentive to moderate remains despite circumstances leading voters to periodically prefer the roll call positions adopted by the more distant candidate, one might ask why electorally risk-averse incumbents would not simply choose to adopt centrist positions. Buchler argues that the answer hinges on the fact that members possess privately held preferences that diverge, in some cases meaningfully, from the median voter. This is a controversial argument. While we know of famous cases of members voting in electorally damaging ways, these cases are remarkable because they are few and far between. Moreover, in these isolated circumstances, we cannot be certain why these members ultimately voted against their constituents. Personal preferences are merely one possible explanation among many. Moreover, some scholars have argued that members are electorally successful because they hold preferences that are naturally aligned with their constituents (e.g., John R. Lott and Stephen G. Bronars, “Time Series Evidence on Shirking in the US House of Representatives,” Public Choice, 76(1-2), 1993). Nevertheless, it is somewhat unconventional to make this distinction, and it proves to be a critical assumption to the model.
It is also worth noting here that challengers generally do not have ideal points that emerge from recorded votes, and therefore their ideal points are indeed exogenous to the legislative process. As Ferejohn (John Ferejohn, “Incumbent Performance and Electoral Control,” Public Choice, 50(1-3), 1986) pointed out, since challengers are unconstrained by a legislative record, why would incumbents ever win? Buchler argues that voters simply impute a challenger’s behavior on the basis of her or his party. This assumption implies that partisanship effectively trumps ideology in the spatial model, obscuring potentially meaningful variation in preferences within a caucus. To that end, this assumption falls prey to the same criticism that Buchler raises about the assumptions made by voting behavior theorists to explain violations of spatial voting.
Parties are central to Buchler’s thesis, as they resolve the collective-action problem facing reelection-seeking members who want to achieve extreme policy goals that will cost them votes. Parties step in to solve this dilemma by threatening punishments to enforce sincere voting among members. Buchler argues that caucuses empower party leaders to pursue more aggressive punishments when its members have homogenous sincere preferences, represent diverse constituent preferences, and possess intense policy preferences. This argument is reminiscent of John Aldrich and David Rohde’s Conditional Party Government Thesis (David W. Rohde, Parties and Leaders in the Postreform House, 1991; John H. Aldrich and David W. Rohde, The Logic of Conditional Party Government: Revisiting the Electoral Connection, 2000), but with a distinction made between sincere and electorally induced preferences. This party pressure leads members to adopt incrementally more extreme positions than would occur in the absence of parties. This insight, in conjunction with the proximity paradox, shows how polarization and electoral success can coexist.
With regard to the role of parties, Buchler’s model raises three questions for us. First, how do party leaders acquire information regarding members’ sincere policy preferences? It would seem that members can avoid punishment by obfuscating their privately held preferences. Without a crystal ball, party leaders would be unable to coerce electorally suboptimal behavior from their members.
Second, why would party leaders care about enforcing sincere preferences as opposed to encouraging members to vote cohesively? There is no guarantee that the sincere preferences held by a member are collectively advantageous. Buchler’s model assumes that members personally prefer their caucus’s position to those of their constituents. Should divergence between personal and electorally induced preferences exist, how are we to know the direction of that divergence? After all, the question of why members’ personal preferences have polarized in recent decades remains a significant and unanswered one. Party rewards/penalties could also generate the electorally suboptimal behavior, even in the absence of a divergence in preferences between a member and his or her constituents.
Finally, what do members gain for hiring party leaders who jeopardize their electoral fortunes? Buchler somewhat trivializes the central motivation of the Cartel Theory—that majority party status motivates party leaders to act in ways that maximize caucus members’ electoral prospects. Buchler’s conception of party influence in some respects is quite the opposite, serving instead “to counteract electoral pressure” (p. 123). In this way, party leaders undermine the goal of maintaining majority party status, which itself yields considerable policy returns. Buchler contends that parties structure the agenda to avoid votes that put members and their constituents on opposing sides. However, in order for the majority to reap nontrivial policy gains, this would require majorities to target only those status quos located well within the minority party space (which we cannot measure). Thus, this model leaves us with some uncertainty over when and how often the majority party will compromise members’ electoral well-being and for what gain. Furthermore, since members reap policy returns to sincere voting in this model, there seem to be alternative options for members to balance electoral and policy constraints without party. In fact, Buchler alludes to this possibility in addressing the extension of his model to the Senate.
As this book relates to ours, we are largely in agreement with the conclusions reached by Buchler in Appendix B. We believe that the spatial model is limited when considering the behavior of voters. Some people make decisions on the basis of spatial proximity and others do not. In the aggregate, though, spatial modeling offers valuable insights about election outcomes. In the absence of a massive shift in partisanship from one election to the next, the rational voting public (i.e., those capable of rational updating) rewards moderation and punishes extremity. However, we would add that members are presented with a somewhat more uncertain context than that proposed by Buchler’s alternative account, in that members have less certainty over which subset of voters processes information in an intuitive versus a reflective way. Therefore, gaming roll call votes to maximize support among the voting public may not be quite so straightforward.
In sum, this is an impressive book that contributes not only to spatial analyses of legislative behavior and elections but also to our understanding of elite polarization more generally. It notably challenges scholars to think more seriously about incorporating electoral demands into our models of legislative decision making, and vice versa. While this might seem an obvious recommendation, it is unfortunately underappreciated in the extant literature. Buchler’s writing is both sophisticated and accessible, making Incremental Polarization a pleasure to read. We commend Buchler for his outstanding scholarship, and believe that this book will receive considerable attention for years to come.
