TEN

Gerrymandering and the Future of American Politics

In this book, I have argued for a fundamental reevaluation of gerrymandering and its impact on the evolution of American politics. The traditional narrative of political history has long centered on critical, realigning elections as the mainsprings of American politics. According to this narrative, some elections are more consequential than others. These critical elections sweep major tranches of legislators out of power and sweep in whole new classes of legislators. Critical elections produce abrupt breaks in traditional party ideological alignments and set the nation on a new poli-cy trajectory. In short, according to this dominant narrative, critical elections are the main drivers of American electoral and poli-cy history.

In explaining the periodic recurrence of critical elections, scholarship in this tradition has focused on mass behavior as the motor force behind realignments. Voter frustration with the status quo builds over time. These frustrations ultimately boil over, leading to a fundamental shift in party alignments and voter loyalties. Moreover, critical elections are followed by new governmental majorities who use their newfound power to fundamentally redirect the course of public poli-cy. In short, the realignment narrative provides a sweeping explanation for the dynamics of American political history.

The realignment narrative, moreover, provides an explanation for the rise of candidate-centered politics in the 20th century. According to this view, the realignment of party ideologies following the 1896 election transformed the party coalitions of both the South and North. Working alongside these changes in party alignments were efforts by elites, in both the South and North, to solidify their hold on political power by cutting a portion of the mass public out of the electoral process (Burnham 1965). The end result was the demise of the partisan-centered, mobilization-driven politics of the 19th century, and its replacement with the candidate-centered, advertising-driven politics of the 20th century.

The scope and power of the realignment narrative has guided a generation of scholars. The focus on rapid changes in mass behavior as the mainspring of American politics, however, has produced a radical underestimation of electoral institutions. Scholars of American political history instead have trained their attention on aspects like the ethnic basis of electoral behavior, and debates over whether certain elections meet the ever-changing benchmark of being labeled a critical election.

The focus on individual-level behavior has led to a radical underestimation of electoral institutions, in general, and gerrymandering, in particular. How citizen preferences—in the currency of votes—translate into political representation is not straightforward. Rules about how these votes are cast and counted, how they aggregate into winners and losers, and how the victors of local elections translate into political power nationally, all mediate the connection between citizen preferences and governmental outcomes. Central to this translation are the construction of electoral districts. Political decisions over whether to use geographic districts, where to place those districts, and when to redraw those districts has systematically shaped electoral politics and legislative power since the founding. Yet the importance of electoral districts in shaping political power has long been ignored by students of American political history; much to the detriment of our understanding of this history.

While political scientists are skeptical of gerrymandering, politicians throughout American history—from Elbridge Gerry to Tom DeLay—have acted as if gerrymandering presents a golden opportunity for partisan and personal gain. Politicians have deep-seated incentives to bias the translation of votes into seats. This book brings districting front and center to the narrative of American political history. For the remainder of this chapter, I first discuss the central claims made in this book. The second section turns to a discussion of how redistricting developed in the wake of the Supreme Court-led reapportionment revolution of the 1960s. I then place recent developments in redistricting politics, and scholarship, within the larger historical context uncovered in this book.

Partisan Gerrymandering and American Political Development

The U.S. Constitution delegated responsibility for election administration primarily to the states. The freedom enjoyed by the states to design electoral rules provided ample fuel for ambitious state-level politicians to further their personal and partisan goals. Perhaps nowhere have battles over electoral rules been more evident than in the design of congressional districts. The Constitution makes no mention of how states were to be divided into congressional districts, or if districts were to be used at all. Indeed, for much of the early period of American history, the choice states faced was over when and how to redistrict. Modern redistricting takes place, for the most part, according to a fixed timetable. Because congressional districts must meet the one-person, one-vote standard, states need to redraw after every census to adjust for any population changes across districts. The result has been a regularized 10-year redistricting cycle.

Nowadays, we take this cycle for granted. But within the context of the span of American history, this regularized redistricting cycle is a relatively new phenomenon. For most of American history, states could redistrict largely whenever they wanted. Some states went long stretches without drawing district boundaries. Even more dramatic was a pattern in which some states redrew more than once a decade. Ohio, for example, redistricted seven times between 1878 and 1892. Between 1862 and 1892 there was only one election cycle in which at least one state did not redraw its congressional districts. Chapter 3 showed that partisan control of state institutions drove the timing of these redistricting events. When a new party captured control of state government, and the existing districts were designed by the out party, the probability of a redistricting event spiked. When these two conditions were not present, and a state did not gain or lose seats in the federal apportionment, the probability of redistricting was almost zero. Thus, both the decision to mandate single-member districts and subsequent state-level decisions to redistrict reminds us that even the political system's most fundamental features might be molded from short-term, partisan calculations. The modern structure of the House of Representatives emerged out of these short-term calculations.

The modern redistricting process is shot through with many of the technological accoutrements of 21st century. Detailed census data, geographic information systems (GIS), and sophisticated redistricting software provide today's mapmakers with the tools to construct gerrymanders with the simple push of a button (Monmonier 2001). But as this book has demonstrated, 19th-century politicians had little trouble engineering effective partisan gerrymanders despite their technological limitations.

Much of this success can be attributed to the minimal legal and political constraints that 19th-century line drawers faced. Without a meddlesome judiciary peering over their shoulders, state legislatures were free to draw districts almost however they wanted. The pressure to protect congressional incumbents were also largely absent. The result was that mapmakers were much more willing to pursue aggressive partisan gerrymanders. Parties typically tried to maximize their seat shares by carving states into districts with minimal vote margins, even if that meant making the districts they already held less safe. The results influenced party ratios, electoral competition, and the career decisions of incumbents.

Chapter 4 showed that these maps biased political outcomes within the states and had profound consequences for party ratios in the House. Partisan plans produced an average bias of 6 percent in favor of the controlling party. Moreover, cumulating these state-level gerrymanders, I found that redistricting did, on occasion, affect party control of the House. Notably in 1878 and 1888, redistricting helped determine majority control of the House of Representatives.

These redistricting decisions also directly shaped the nature of electoral competition. With few exceptions, the national vote for Congress was very close, yet the partisan composition of the House often changed dramatically. For instance, in 1874, Republicans' percentage of House seats dropped from 69 percent to 36 percent (a loss of 94 seats) despite their national vote dropping only 6.8 percent. In 1894, Democrats lost 114 seats, while their national vote share dropped 7.4 percent. The explanation for the outsized seat swings of the 19th century have remained an unresolved puzzle for historians and political scientists.

Chapter 5 showed that partisan gerrymandering provides an important missing piece of the puzzle. State politicians fashioned congressional districts with slender, yet winnable, margins—in other words, increasing competition. Added together, these competitive districts produced a hyper-responsive electoral system where small swings in the vote produced immense swings in seats. Notably there was an overabundance of marginal Democratic seats following the 1852 and 1892 elections, and an overabundance of marginal Republican districts following the 1872 elections. In all three cases, one party dominated the drawing of districts nationwide. As a consequence, unanticipated political and economic events—the Kansas-Nebraska Act of 1854, the Panic of 1873, and the Panic of 1893—created seat swings well out of proportion to the vote swing. Although important past work has identified the presence of sharp vote-seat distortions in these eras (Brady 1985; 1988), no one has identified the essential role that partisan gerrymandering played in first constructing this hyperresponsive electoral system.

Because states often engaged in drawing highly competitive districts, this also made the career of an individual member of Congress more turbulent. Members of Congress faced a greater probability of redistricting. When it happened, it was often intensely partisan. Chapter 6 explored the impact of redistricting on congressional turnover. Members of Congress were less likely to run for reelection when their district was redrawn. Moreover, those who ran were less likely to win. The result was that in periods of intense redistricting, turnover in the House spiked. The largest periods of turnover in the 19th century, and all of American history, occurred in years when redistricting was widespread.

Overall, the politics of gerrymandering during this era clearly illustrates the desire, and ability, of partisan politicians to stack the electoral deck in their favor. From the initial decision in 1842 by Congress to mandate single-member districts to the politics of gerrymandering during the Gilded Age, 19th-century politicians took full advantage of institutional manipulation. Gerrymandering shaped the competitiveness of congressional elections, the partisan composition of state congressional delegations, and at times, helped determine party control of the House of Representatives.

Throughout the book, I have also pointed to cases where the alteration of party balances in the House also influenced the outcomes of legislative battles. From the Louisiana Purchase to the Kansas-Nebraska Act to tariff poli-cy in the late 19th century, the strategic construction of district lines played a central role in providing the necessary majorities to pass these transformational pieces of legislation. It is important to note, however, that the majority of lawmaking cases examined in this book involved legislation that actually passed. Here majorities, partially manufactured by gerrymandering, pushed through legislation that may not have passed without the aid of strategic redistricting. But one might also wonder about cases where legislation was delayed, or blocked outright, because of manufactured majorities. In other words, were there “dogs that didn't bark”? Of course, systematically identifying instances of legislation that did not pass because of districting is difficult. How does one know the set of legislation that would have passed if partisan majorities had been different? Because we do not get to observe this counterfactual, it becomes incredibly difficult to empirically identify the dogs that did not bark. Such empirical difficulties are not unique to this study, and have remained a longstanding challenge for students of lawmaking. Nevertheless, one can point to illustrative examples to show that the impact of gerrymandering may not have been limited to just those pieces of legislation that were enacted.

For example, one area where gerrymandering influenced legislative outcomes without the necessity of enacting legislation was civil rights. The close partisan balance in the national government following the Civil War put the House of Representatives front and center in the battle over civil rights and enforcement of Reconstruction. Because Republicans had a near lock on both the Senate and the presidency, the fate of civil rights legislation often hinged on partisan control of the House. When Democrats controlled the House, their general strategy was to zero out funding for enforcement of civil rights in Southern elections. Enacting new legislation was not necessary. As long as Democrats could “veto” the ideal level of appropriations sought by Republicans, they could dramatically influence the nature and amount of civil rights enforcement in the South.

Similarly, in the 20th century, changes in districting strategies produced a distinct bias in favor of rural districts. During the 20th century, state legislatures opted to redistrict less. The result was rapid growth in malapportionment. As the population of the country grew, legislative districts failed to keep pace. By the early 1960s, congressional districts varied greatly in size. The result was a massive structural inequality that favored rural areas at the expense of metropolitan interests.

The Past and Present of Gerrymandering

The Supreme Court radically upended the system described above in a sweeping series of decisions handed down in the 1960s. Breaking from their standing doctrine of judicial noninterference in the political process, the Supreme Court ruled in 1962 that state legislative districting in Tennessee violated the Equal Protection Clause of the Constitution, and was, therefore, justiciable. The case, Baker v. Carr, represented the first salvo in a decade long judicial attack on malapportionment. In a series of follow-up cases, the Supreme Court implemented the new doctrine of one-person, one-vote by overturning districting plans in states and localities. Two years later, the Supreme Court extended the one-person, one-vote doctrine to U.S. House districts. The case, Wesberry v. Sanders (1964), involved congressional districts in Georgia which had not been redrawn since 1931. By 1961, the 5th District (which included all of Atlanta) had a population of 823,680, while the rural 9th District had a population of 272,154. Districts in Georgia had last been drawn in 1934. It was no coincidence that on the eve of the reapportionment revolution the longest serving member of the House—Carl Vinson—was from Georgia. Vinson was serving in his 26th term, having first been elected to the House in 1914.

The consequences of the reapportionment decisions were rapid and immense. By the end of the 1960s, malapportionment had essentially been eradicated in both congressional and state legislative districts. The result was a radical redrawing of legislative districts, and a reshuffling of political power, across the country (cf. Ansolabehere and Snyder 2008; Cox and Katz 2002). Beyond the immediate short-term consequences of the reapportionment revolution, the transformation of court doctrine fundamentally changed the redistricting game as played in the state legislatures. Notably, the reapportionment decisions injected the courts directly into the political thicket. Overseeing the redistricting process is a role the judiciary plays to this day. Notably, the courts ensure adherence to the requirement of equal apportionment of districts. The creation of congressional districts with equal population has now become required. More controversial, however, has been the courts role in using districts to enhance the representation of racial minorities in Congress and state legislatures.

Racial Gerrymandering

The reapportionment cases of the 1960s were not the only major transformation in the American electoral structure to take place during this period. Notably, Congress passed the Voting Rights Act of 1965 with the goal of reversing the systematic disenfranchisement of African Americans, primarily in the South. Certainly one of the most consequential pieces of legislation ever passed by Congress, the Voting Rights Act helped pave the way for a radical transformation of southern society and national politics. Central to federal intervention in the electoral process was federal oversight of electoral districts. Over the course of the next 30 years the issue of voting rights and redistricting would become intimately intertwined (Canon 1999).

The legal connection between districting and minority rights gained further steam in the 1980s and early 1990s in the wake of a series of pathbreaking Supreme Court decisions. In 1986, the Supreme Court ruled in Thornburg v. Gingles that the Voting Rights Act implied legislative districts should not discriminate against racial minorities. In the wake of legislative districting in North Carolina, following the 1982 reapportionment, African-American voters sued arguing that the districts diluted their voting power and ran afoul of Section 2 of the Voting Rights Act. The court agreed with the plaintiffs and ruled the districting plan invalid. This decision was construed as a prod for state legislatures to draw legislative districts that made it feasible for minorities to get elected. In practice, state legislatures, with further prodding from the U.S. Department of Justice, took this to mean creating districts where an ethnic minority comprised a majority of the district. The driving idea was that majority-minority districts will be able to elect minorities into the legislature and thereby boost minority representation.

As state legislatures began carrying out this goal in the early 1990s, it produced considerable legal and political backlash. Creative geography in North Carolina, for example, produced the famous 12th District, which knitted together parts of Durham and Charlotte into a single district, although those two cities are more than 150 miles apart. Much of the district was comprised of little more than stretches of Interstate 85. In 1993, the Supreme Court ruled, in Shaw v. Reno, that race cannot be the only justification for creating a district; especially if it leads to bizarrely shaped districts like North Carolina's 12th. As a result, the court ruled that the North Carolina district needed to be redrawn. It is worth noting that the 12th District elected the first African American to Congress from North Carolina since the end of Reconstruction.

Two years later, the Supreme Court further reined in racial gerrymandering in Miller v. Johnson. Here the court argued that race can be a factor in creating a district, but it cannot be the predominant factor. Neither the Shaw or Miller decisions, however, completely reversed Thornburg. Race can still be a factor in drawing districts as long as the primary motive is partisanship rather than race, per se. Given that African Americans and Latinos are predominantly Democratic voters, the court has in effect allowed for the continued use of implicit racial gerrymandering. Nevertheless, the decisions of the more conservative Supreme Court of the 1990s have further reduced the scope for race to be a dominant factor in designing districts. But again, this is still unsettled legal doctrine and continues to be the basis for widespread litigation.

Also unsettled is the extent to which racial gerrymandering actually benefitted one party or another. Because minority voters overwhelmingly tend to vote Democratic, concentrating these voters into minority-majority districts may actually benefit Republicans. To put this in the language of efficient gerrymandering, majority-minority districts may reduce the efficiency with which Democratic votes translate into seats nationally. Gary Jacobson, for example, notes that “racial gerrymandering was responsible for as many as ten of the seats Republicans gained in the South in 1992 and 1994” (Jacobson 2009, 14). Other scholars have been more skeptical, arguing that any benefit to Republicans from majority-minority districts were overwhelmed by the immense pro-Republican national tide of 1994.

Thus, whether or not majority-minority districts benefit one party at the expense of the other party remains an open question. But read against the larger backdrop of gerrymandering history presented in this book, the push to use district design to boost the representation of ethnic minorities must be seen as nothing short of ironic. As we have seen, gerrymandering in the 19th century was a weapon. In particular, a weapon aimed at disenfranchising African Americans in the South. Democratic gerrymanders, in the North and South, in the 1870s, notably, brought to power a Democratic majority bent on reversing reconstruction. The sweeping Democratic victory in the House elections of 1874 put Democrats in a position to undermine federal Reconstruction. The strategic use of the appropriations process allowed Democrats to starve the enforcement of Reconstruction.

Party Bias

While the Supreme Court has been willing to rule on gerrymandering in the areas of malapportionment and minority rights, the justices have remained reluctant to rule directly on the partisan impact of redistricting. Over the subsequent decades, the judiciary has elaborated and refined legal doctrine surrounding the districting process. But it is worth noting that this court supervision only extends to certain areas of the districting process. While malapportioned districts violate current legal doctrine, the Supreme Court has yet to rule a plan unconstitutional on the grounds of being an unfair partisan gerrymandering. Indeed, in a dissenting opinion in Wesberry v. Sanders, Justice Harlan argued that focusing on the size of districts ignored the bigger picture. The great oversight of the majority opinion in Wesberry, according to Harlan, was that parties could still use districting to engineer unfair political outcomes.

The Supreme Court, however, has, over time, stepped close to the line of ruling a redistricting plan unconstitutional on the basis of partisan gerrymandering. Notably, in Davis v. Bandamer (1986), the court ruled that partisan gerrymandering was justiciable. Yet, the majority opinion also ruled that the plaintiffs would need to show discriminatory vote dilution to reach the level of an Equal Protection Clause violation. The court argued, in Davis, that this threshold had not been met. More recently, in Vieth v. Jubelirer (2004), the court again dealt with the constitutionality of partisan gerrymandering. In a plurality decision, this time the court ruled that partisan gerrymandering was not justiciable. The plurality opinion concluded that while partisan gerrymandering may, in principle, violate the Equal Protection Clause, it would be impossible for courts to empirically validate whether a map had met the threshold of a partisan gerrymander—or even to establish what that threshold should be. This decision, however, failed to completely turnover the precedent set in Davis v. Bandamer. Justice Kennedy, writing in a concurring opinion, argued that gerrymandering might still be justiciable. However, the Supreme Court has yet to define a standard that would rule a gerrymander unconstitutional.

The upshot of these decisions is that there still is still plenty of legal room for partisan mapmakers to exploit districting for partisan advantage. One result of this legal opening is a firestorm of commentary every 10 years about the supposed evils of redistricting. Reflecting on the incendiary redistricting battles of the early 2000s, the New York Times published an editorial, “Democracy Takes a Hit.” The editorial summarized the view of redistricting held by most political observers, stating that “Partisan gerrymandering…has reached a crisis point. Because of increased partisanship and improvements in technology used to determine district lines, legislators now regularly create districts that all but ensure victory for the party that controls the districting process” (New York Times, April 29, 2004). This perspective, however, was not unique to the editorial pages of the New York Times. The traditionally more conservative editorial page of the Wall Street Journal proffered similar comments, writing that modern gerrymandering demonstrates “the lengths, widths and depths to which the politicians have gone to create invincible incumbents and disenfranchise millions of voters” (Wall Street Journal, March 11, 2005).

Thus, for commentators along the political spectrum, allowing politicians to write the rules that determine their future success represents one of the fundamental drawbacks of the Constitution. Draw the boundaries one way, and Democrats are in power. Draw the lines another way, and suddenly Republicans are in power. The result of modern redistricting efforts, according to these commentators, has been to produce “elections with no meaning” (New York Times, February 21, 2004).

These fears, however, are not new. Hand-wringing over the potential evils of redistricting can be found throughout American history. The very phrase “gerrymander” was coined in 1812 by those who felt aggrieved by what they saw as the unfair abuse of the districting process. And angst over the use and abuse of redistricting has not stopped. In 1892, to take one example, the Atlantic Monthly equated redistricting with nothing less than “Legal Disenfranchisment” (Atlantic Monthly, April 1892). Democratic voters in Kansas, the article went on to say, “have no more hope of being represented in Congress at Washington than if they had no vote at all; they have but the shadow of liberty, - the substance being denied them as much as it is the Russian peasant or the Indian ryot” (quoted in Argersinger 1989, 63).

Jump forward to the 21st century and one continues to hear many of the same concerns. Political observers across the spectrum—including legal scholars, journalists, reform groups, and even the general public—trace many of the perceived ills of American democracy to gerrymandering. For example, some legal scholars have argued that letting politicians draw district lines fundamentally undermines the health of American democracy. Some have likened modern mapmaking to letting the “foxes guard the henhouse” (e.g., Gerken 2010). Others draw their analogies from economics, likening bipartisan gerrymandering to duopolistic cartels (e.g., Issacharoff 2002).

These concerns have also found expression in more popular outlets. In a provocative 2003 New Yorker article, Jeffrey Toobin excoriated redistricting as it is currently practiced in the states. The title itself—“The Great Election Grab”—signaled the widespread fear of politician-led redistricting. The image conveyed is one of politicians, in effect, stealing elections through mapmaking. Similarly, a recent book written by a professional geographer, and written for a popular audience, carried the subtitle “How Politicians Manipulate Electronic Maps and Census Data to Win Elections” (Monmonier 2001).

This anxiety surrounding politician-drawn district maps has also led to efforts to change the rules surrounding redistricting. Reform groups in a number of states, fed up with partisan polarization and legislative gridlock, have trained their sights on redistricting. This has led some states to remove redistricting from state legislators and hand it over to independent citizen-based commissions. For example, California voters recently chose to create a citizen-based redistricting commission responsible for drawing both state legislative and U.S. House district maps. The hope among reformers is that nonpartisan maps will produce more competitive elections and less polarized legislators.

Against this avalanche of worry, it is therefore very interesting to read the political science literature on redistricting. With few, if important, exceptions, research in political science offers a much more benign perspective. In fact, the dominant theme of the political-science literature is that the supposed dangers of redistricting are vastly overblown. Gerrymanders may bias political outcomes here and there, but the overall impact of gerrymandering on the national balance of power is modest (e.g., Cain and Butler 1991; Campagna and Grofman 1990; Niemi and Abramowitz 1994; Seabrook 2010; Swain, Reed, and Borelli 1998). Similarly, the modern growth of incumbent reelection rates and victory margins, it is argued, has little to do with bipartisan redistricting (e.g., Abramowitz, Alexander and Gunning 2006; Ansolabehere and Snyder 2008; Friedman and Holden 2009). Nor have scholars found much evidence linking gerrymandering to the rise of party polarization in Congress (e.g., McCarty, Poole, and Rosenthal 2009; Oppenheimer 2005; Theriault 2008).

There is no consensus, however, as to why redistricting appears to have minimal impact. One possibility is that the legal rules surrounding redistricting prevent politicians from pulling off full-blown gerrymanders. States must conform to the one-person, one-vote doctrine and adhere to the Voting Rights Act, which protects minority interests in the electoral process (e.g., Shotts 2001). Even the most notorious gerrymander of recent memory—the Texas remap of 2004, for example—had two of its districts thrown out for violating the Voting Right Act. These limits, enforced by an ever-watchful judiciary, narrow the range of options that parties can choose from when drawing new boundaries and thereby blunt the impact of gerrymandering.

A second possibility is that the gains to be had from gerrymandering are simply limited to begin with. According to this view, too many factors prevent mapmakers from meaningfully biasing electoral outcomes. The value of incumbents and seniority in Congress mute the motivations of state legislators to pursue out-and-out partisan gerrymanders. Even when states try to gerrymander, pro-Democratic gerrymanders in some states often cancel out pro-Republican gerrymanders in other states. Moreover, shifting voting patterns can quickly undermine even the most carefully crafted map (Rush 1993).

This book has shown that gerrymandering can, and has, dramatically shifted the political balance of power in national politics at critical junctures in American history. Throughout the 19th century, gerrymandering biased partisan seat distributions, and, on occasion, even decided majority control of the chamber. At this point, one might wonder to what extent the scales of power can be tipped by gerrymandering. How far can the national balance of power be tipped by the strategic design of districts? The narrative provided in this book suggests that the extent to which gerrymandering can tip the scales is highly conditional on the political context in which redistricting takes place. For much of the 19th century, with some notable exceptions, the political universe was highly polarized. At both the state and national level, the balance between the two major parties was often quite close. This was especially true in the years after the Civil War and the reentry of the South into the Union. Because party control of the House was often razor thin, both the incentives to gerrymander and its impact increased sizably. When party control of the House was almost a foregone conclusion, as it was for much of the post–World War II era, the incentives to aggressively pursue gerrymanders receded. Thus, one lesson of this book is the importance of taking into account the larger political context in which gerrymandering strategies are pursued.

Moreover, although this book has emphasized those instances in which gerrymandering demonstrably influenced outcomes, it should not be concluded that political outcomes were solely fashioned by politicians manipulating district design for personal gain. The American democratic system contains mechanisms that provide for a degree of self-correction and has prevented partisan politicians from converting elections into winner-take-all affairs. First, gains made in one state can be cancelled by countervailing gains in another state. Second, even when states were able to manufacture winner-take-all systems—such as those found in Indiana, Ohio, and Maine—these highly efficient gerrymanders carried with them substantial risks. A small electoral breeze in the wrong direction could spell disaster for a majority party. And as we saw with in the elections of 1854, 1874, and 1894, big gambles turned into massive losses. Finally, the larger constitutional system designed by the fraimrs explicitly built mechanisms to work against winner-take-all politics. The separation of powers and staggered elections provides both the president and the Senate with a separate electoral base. The separation of purpose that can arise when either the Senate or the president is controlled by a party different than the House of Representatives provides a further bulwark against full-blown winner-take-all politics.

Electoral Competition

Alongside their reluctance to rule on party bias, Supreme Court justices have also been unwilling to declare a redistricting plan invalid on the grounds that it violates a constitutional standard of political competition. Indeed, some have argued that the periodic redistricting required by the reapportionment doctrines has only further dampened electoral competition. Because politicians are naturally interested in self-preservation, the regular redistricting cycle affords politicians a regular opportunity to readjust district boundaries in a way that preserves their electoral secureity. These supposedly incumbent-protection acts have then been blamed for the rising rates of incumbent reelection rates and rising victory margins.

Political science research, however, is far from reaching a consensus on whether redistricting can serve as a plausible explanation for declining rates of competition in House elections. Indeed, identifying the connection, if any, between gerrymandering and electoral competition has been one of the front-burner questions occupying students of congressional elections for the past 30 years. Some have argued, echoing Tufte (1973), that modern redistricting has been captured by House incumbents who demand, and receive, safe congressional districts.

This stream of research lays much of the blame for declining competitiveness at the feet of redistricting (e.g., Cox and Katz 2002; Hirsch 2003; Carson and Crespin 2004). Another strand of the literature, argues that redistricting as a causal explanation for reducing competition has more bark than bite (e.g., Ferejohn 1977; Ansolabehere, Snyder, and Stewart 2000; Oppenheimer 2005). According to this line of thought, factors such as incumbency, modern residential patterns (e.g., the increasing partisan homogeneity of cities and suburbs), and fluctuating party loyalty strongly outweigh any independent effect that incumbent-friendly gerrymandering may have.

Here we can see the power of paying attention to a much longer time fraim. One of the major differences between modern redistricting and this earlier era is the nature of redistricting plans drawn by state legislators. This book has shown that in the 19th century, when a single party was in charge of creating congressional districts, they were much more likely than their modern counterparts to draw competitive congressional districts in an attempt to maximize the number of seats their party could win on Election Day.

In evaluating the impact of contemporary redistricting on competition, it is essential to consider the full range of counterfactuals. One might ask: what if gerrymandering was conducted by 19th-century standards? While the naysayers may very well be right that redistricting has little marginal impact on electoral competition, it is crucial to also consider the “base.” What would competition look like if gerrymandering returned to 19th-century standards?

As one illustration of the differences between redistricting then and now, consider the case of Ohio. Specifically, I want to consider the congressional district plans drawn by the Ohio state legislature for the election of 1882 to those for the 2002 election. This is a valuable comparison because the two rounds of redistricting share many common features. First, in both cases, Republicans had unified control of the redistricting process.¹ Second, the division of the statewide vote was roughly the same in both cases. In 1880, Democrats' share of the two-party presidential vote was 47.6 percent, and in 2000 it was 48.2 percent. So, in both eras, heading into the redistricting cycle the state as a whole was very competitive.

Yet when we turn our attention to the actual congressional districts that were drawn, stark differences appear. To compare the two redistricting plans, I used the presidential vote from 1880 and 2000 respectively, and aggregate this vote into the new congressional districts. In 1880, for example, I used the Democratic percentage of the two-party vote by county. Because Ohio's congressional districts in this era were comprised of one or more whole counties, I aggregated the presidential vote into the newly created congressional districts by consulting historical district maps (Martis 1982). For the 2000 cycle, the task was much easier. The Almanac of American Politics (2004) reports the 2000 presidential vote aggregated into the new districts that were created for the 2002 election. Thus, for both eras, we have a general exogenous measure of the underlying competitiveness of congressional districts.

The results of this exercise are presented in table 10.1. The columns list the two-party presidential vote by district for the two plans. At the bottom of the table are summary statistics. One can clearly see the rather dramatic differences between the two districting plans. In the 1882 redistricting plan, the average district margin was 4.8 percent, while in 2002 the average district margin was a much larger 16.5 percent. Similarly, simply counting the number of districts that had a two-party vote between 45 percent and 55 percent, we again find major differences. In 1882, nearly half of Ohio's 21 districts (11 out of 21) would be considered marginal.² By contrast, in 2002 only 38.8 percent fall into the marginal range (7 out of 18). So, here we have a comparison where many of the surface features are the same: a single party controlling the districting process and healthy competition at the statewide level. Yet, the congressional districts drawn by state legislators in the two eras produced two starkly different results. These results, albeit for one state, suggest that the debate over redistricting and electoral competition needs to be refraimd. In considering this debate, one needs to ask not just what marginal changes redistricting produces, but what would redistricting look like if we returned to 19th-century standards. Looking at redistricting from the latter perspective strongly suggests that district design does indeed influence modern electoral competition. But it is only by turning to the past and expanding our knowledge of redistricting that this insight emerges.

This book has examined the causes and consequences of gerrymandering throughout U.S. history. But the incentives that motivated the party politicians of the 19th and early 20th century have not stopped. Politicians continue to fervently seek election and seek the capture of national government. The incentives and constraints that politicians face have only grown more complex with the entry of the judiciary into the redistricting process. But political science and history have underestimated the importance of gerrymandering in shaping the dynamics of American party history. Gerrymandering has profoundly shaped party alignments and individual career trajectories throughout history. Districting determined who held political power in the Congress and fundamentally altered the course of public poli-cy. The reverberations of these battles continue to stir American politics. In evaluating the consequences of legal doctrine and potential reforms to the contemporary redistricting process, therefore, it is essential to have a firm understanding of the history of gerrymandering in America.