"Elbridge Gerry's Salamander" Abstract
The Supreme Court’s reapportionment decisions, beginning with Baker v. Carr in 1962, were immediately hailed as a jurisprudential revolution. They reversed decades of decisions holding that legislative redistricting, fraught though it was with malapportionment and gerrymandering, was not justiciable. They opened the door to a long chain of subsequent litigation, which has continued into the 1990s with important decisions regarding racial gerrymandering. For these and other reasons, the reapportionment decisions now occupy a standard niche in textbooks on the Court.
The Court’s decisions did not simply rewrite case law, however. They also sparked a massive wave of extraordinary redistricting in the mid-1960s. Both state legislative and congressional districts were redrawn more comprehensively—by far—than at any previous time in our nation’s history.
In the aftermath of the Court’s decisions and the consequent redistricting, scholars looked carefully for political consequences. Yet they found few. Neither party seemed to benefit nationwide, as their gains in some states were offset by losses elsewhere. Incumbents did not seem to benefit, as their margins of victory increased even where redistricting did not take place. Policy did not seem to shift toward urban interests in the dramatic way widely anticipated.
These conclusions were surprising, not just because of the magnitude of the judicial shift in doctrine, or the depth and breadth of 1960s redistricting action, but also because of two statistical regularities later described in the scholarly literature. First, work on how congressional votes translated into congressional seats outside the south found a consistent pro-Republican bias prior to the 1960s. By one estimate, the Democrats could expect to win only 44.6% of the nonsouthern seats, when the aggregate vote division was a 50-50 partisan split—indicating a 5.4% pro-Republican bias. This bias abruptly disappeared in the mid-1960s. Second, scholars found that the so-called incumbency advantage—a vote premium putatively derived from the resources of office—jumped up dramatically in 1966, the first election at which a substantial number of districts had been redrawn under court order. The size and abruptness of these two statistical changes, and their coincidence with widespread court-ordered redistricting, seems more than coincidence. Yet the literature has scrutinized these factors and found no causal link between them and redistricting.
*
This book reexamines the electoral consequences of the reapportionment revolution. The bulk of the previous literature has focused on the primary substantive consequence of the Court’s decisions—the eradication of malapportionment in U.S. legislative districts at both the state and federal level. Our work focuses on the primary procedural consequences of the Court’s decisions—the redefinition of the reversionary outcome to the redistricting processes in the fifty states; and the increased regularity and frequency with which redistrictings were undertaken. We use these procedural shifts to explain sea-changes in two struggles central to congressional elections: that between Democrats and Republicans, on the one hand, and that between incumbents and challengers, on the other.
As regards the partisan struggle between Democrats and Republicans, our argument starts by noting that the Supreme Court’s decisions fundamentally altered the reversionary outcome to the redistricting processes in the states. That is, it altered what would happen at law should the state government fail to enact a new congressional districting statute. Once one properly controls for the nature of the reversion when analyzing the impact of redistricting, several consequences of the Court’s reapportionment decisions come into focus.
First, these decisions made the courts strategic players in all subsequent redistricting actions. Even where no suit had (yet) been brought, each party might worry that the other would bring suit.
Second, the wave of court-mandated redistricting peaked just after LBJ’s landslide victory in 1964 had weakened Republican control of nonsouthern state governments; and during a period when the federal judiciary was heavily Democratic. Thus, redistricting in the 1960s was conducted by state governments that were less often under unified Republican control than had historically been the case; and under the supervision of courts that were largely Democratic. This combination produced a substantial net partisan advantage for the Democrats, evidenced not only in the abrupt disappearance of pro-Republican bias outside the south but also in the detailed patterns of how vote shares changed when district lines were redrawn.
Our explanation of the partisan consequences of the reapportionment revolution is quite different from those offered in the literature, both in the line of argument pursued (no one has stressed reversionary outcomes and the strategic role of the courts in the previous literature) and in the conclusion reached.
As regards the struggle between incumbents and challengers, we argue that the key to understanding the dramatic growth in the apparent advantage of incumbents is to recognize that they are strategic agents, deciding whether to seek reelection or not based on their forecast vote shares. We show that much of the incumbency advantage as previously measured reflects incumbents’ prudence—getting out when the getting is good—rather than their superior campaigning ability or resources.
We then explore how redistricting affected incumbents’ prudential exits. One line of argument concerns anticipations of redistricting. After the Supreme Court’s reapportionment decisions, politicians soon realized that redistrictings, rare and not always forseeable beforehand, would now be unavoidable and regular. This recognition facilitated better coordination between incumbents and strong challengers, so that incumbents more often got out in the face of a particularly formidable challenge—thus increasing the statistical association between running an incumbent and the incumbent party’s vote share.
Another line of argument notes that the eradication of pro-Republican bias in the translation of congressional votes into seats resulted in an abrupt decline in the Republicans’ probability of attaining a majority of seats in the House of Representatives. This intensification of the Republicans’ minority status exacerbated a syndrome of woes for the party, resulting in significantly larger estimated incumbency advantages for the Republicans than for the Democrats. This last finding is much at odds with previous theories of why the incumbency advantage arose but follows naturally from our emphasis on strategic entry and exit.
Chapter by chapter descriptions
The book is divided into four parts: an introduction; two main substantive parts, one dealing with competition between Democrats and Republicans, one with competition between incumbents and challengers; and a conclusion. Chapters 1 and 2 set the stage by describing the Court’s decisions, the condition of congressional districts before and after redistricting, and the reasons offered in the literature as to why redistricting should have been relatively inconsequential.
Chapter 3 begins Part II by presenting the first half of a general model of the redistricting process(es) in the American states. Chapter 4 uses this model to estimate bias and responsiveness in postwar congressional elections outside the south. Our results clearly demonstrate the importance of the reversionary outcome, even before the Court’s reapportionment decisions. Chapter 5 extends the model to include the courts as strategic actors, as is appropriate for redistricting actions after Baker v. Carr. Our empirical results show that the partisanship of the judges supervising redistricting cases in the 1960s was at least as important as which party controlled state government in affecting the character of the plan adopted. Chapter 6 investigates the district-level implications of our theory and finds them to hold. In particular, we show that the vote shares of Republican incumbents whose districts were redrawn increased in both mean and variance, while their probabilities of victory decreased; but the vote shares of Democratic incumbents decreased in both mean and variance, while their probabilities of victory increased.
In chapter 7, we set the stage for the analyses of Part III of the book in two ways. First, we review classic evidence that congressional incumbents abruptly began winning by larger vote margins in the 1960s, review previous attempts to explain this and the related increase in the "incumbency advantage," and sketch our own explanation. Second, we show that the Republican incumbents’ margins increased more than did Democratic incumbents’ margins; and that the Republican incumbency advantage increased more than did the Democratic incumbency advantage (when measured, as is typical, in vote shares). We thus add to the list of explananda that a complete model of postwar congressional elections must address: Although there is no systematic difference between the two parties’ incumbency advantages before the reapportionment revolution, afterwards the Republicans tended to "benefit" more from incumbency.
In chapters 8 and 9, we seek to explain growth in the overall incumbency advantage, leaving the differences between the parties to chapter 10. Chapter 8 explores "selection bias" in previous estimators of the incumbency advantage, deriving from incumbents’ ability to forecast votes and enter or exit in light of those forecasts. Chapter 9 demonstrates that entry by strong challengers and voluntary exit by incumbents have followed the redistricting cycle quite regularly since the mid-1960s, arguing that this partly explains the increasing selection bias found in Chapter 8. Chapter 10 explains why selection bias is more severe in estimating the Democrats’ incumbency advantage, thereby partly explaining the difference in party experiences uncovered in Chapter 7.
In chapter 11 (Part IV), we review the various consequences of the reapportionment revolution. Both in understanding the battle between Republicans and Democrats and that between incumbents and challengers, we stress political expectations. Anticipation of what courts would and would not allow had been irrelevant and abruptly became essential; this, along with the wave of redistricting action, put a sudden end to pro-Republican bias. Anticipation of the next redistricting had been relatively infrequent and difficult and abruptly became regular and easy; this suddenly increased the extent to which conventional measures of the incumbency advantage overestimated its size. Finally, anticipations of the Democrats’ probability of securing a majority in the House (either at the next election or over a somewhat longer time horizon) changed after the eradication of pro-Republican bias; this increased several differences between the parties in recruitment, career paths, and campaigning.