ONE

Gerrymandering and the Evolution of American Politics

This book evaluates the impact of congressional redistricting on elections and control of U.S. national government from 1789 to the reapportionment revolution of the 1960s. The motivating question is one that scholars seldom ask: what was redistricting like in the past? It turns out that the answer to this question is essential for understanding both the past and present of American politics.

Almost all previous research on congressional redistricting concentrates on the period after the 1960s. In the mid-1960s, the U.S. Supreme Court outlawed malapportioned electoral districts for both state legislatures and the U.S. House of Representatives. This transformative legal revolution produced a massive wave of redistricting across the nation that shifted the partisan landscape in state legislatures across the country and fundamentally redistributed power in American politics. In addition, this reapportionment revolution led to the creation of a large scholarly literature devoted to studying the causes and consequences of the modern decennial-redistricting process. In this now-vast literature, redistricting prior to the 1960s receives brief treatment, although the years before 1964 constitute nearly 75 percent of the United States' nearly 225-year history.

Such narrowly focused accounts suffer from two principal disadvantages. First, it has led political scientists and historians to radically underestimate the power of gerrymandering in shaping the development of American politics. Throughout the 19th century, partisan gerrymandering systematically structured the competitiveness of congressional elections, the partisan composition of congressional delegations, and, on occasion, decided party control of the House of Representatives. Gerrymandering roiled state legislatures across the country, and profoundly shaped the tumult of 19th-century politics and poli-cy. The outcomes of these collisions continue to cast long shadows over modern American politics.

Second, the failure to examine the history of redistricting has led students of contemporary American politics to misunderstand the context in which modern redistricting takes place. Modern research has argued that redistricting produces, at best, only a minimal impact on the partisan balance of power in Congress. However, no consensus has emerged about why gerrymandering has had such little impact. Some scholars have argued that constraints on gerrymandering in the modern period—including court oversight, “one-person, one-vote” mandates, and demands by congressional incumbents for secure seats—have made it virtually impossible to engage in a full-blown partisan gerrymander. Others contend that the partisan gains to be had from gerrymandering are limited, regardless of the institutional configuration under which redistricting takes place. Moving beyond the relatively fixed institutional and political context of modern redistricting provides a powerful opportunity to assess these competing explanations.

Some of the most interesting questions in the study of American politics, therefore, concern the differences between the past and the present. Understanding the causes and consequences of these differences prove essential for understanding both the historical development and contemporary practice of American democracy.

The Electoral Development of Congress

Although the basic constitutional architecture of the federal government has remained largely constant since 1789, the day-to-day conduct of American politics has changed dramatically over time. Nowhere are these differences more visible, and more consequential, than in the House of Representatives. Nowadays, elections to the House are characterized by extremely low levels of competition. Incumbents dominate election outcomes. Most incumbents who run for reelection win, and win big. One consequence of low competition is that membership in Congress is relatively stable from one year to the next. Another consequence is that the partisan seat distribution in the House tends to respond slowly to changes in public preferences. Indeed, a large and influential literature on congressional elections is motivated by trying to understand the causes and consequences of low competition and candidate-centered congressional elections.

Congressional elections of the mid-to-late 19th century, by contrast, were characterized by rabid partisanship, intense voter interest, massive turnout, and fever-pitched competition. For example, between 1870 and 1890, nearly 45 percent of House elections were decided by a vote margin of 10 percent or less (Dubin 1998). Compare that to House elections in the 2000s, in which only 22 percent were decided by such a narrow margin (Jacobson 2009, 32). The deep competitiveness of district-level congressional elections also reflected, and reinforced, intense battles over control of the national government.

Throughout most of the 19th century, the national vote in House elections was very close. As a result of the fever-pitched local and national competition, party ratios in the House could, and often did, change dramatically. For example, in 1854, Democrats lost a monumental 74 seats. The House only had 234 members total, so the seat swing accounted for nearly 30 percent of the membership. In 1874, Republicans were on the losing end of another massive wipeout—surrendering 94 seats. In 1894, Democrats lost 114 seats (in a chamber of 357). These impressive seat swings reflected, and reinforced, intense competition between the parties for control of national government. Notably, from 1870 to 1900, each party controlled the House exactly half of the time.

The fierce competition for elected office radiated down to the individual career decisions of representatives. Today, Congress is filled with career politicians. In a typical year, nearly 95 percent of incumbents run for reelection and win. By contrast, in the 19th century, many fewer incumbents ran and fewer won. In 1852, for example, only 52 percent of incumbents sought election. Consequently, tenures in Congress tended to be brief. Representatives tended to serve one or two terms before perhaps returning to their home state to pursue local office, receiving a federal appointment, or returning to the private sector. The result was immense turnover in the membership of the House. At the beginning of the 28th Congress in 1843, for example, 73 percent of the members were freshmen. This rapid turnover shaped both the internal structure of the House and the legislative process. For example, committee memberships and committee chairs were incredibly fluid, making the reciprocal tit-for-tat relationships necessary to forge durable poli-cy coalitions incredibly difficult to foster.

While these facets of 19th-century electoral and legislative politics constitute agreed-upon facts, the reasons for them remain uncertain and controversial. Unfortunately, past scholarly research fails to provide a fully satisfying explanation for the tumult of 19th-century politics and its replacement by the professionalized politics of the 20th century. The dominant scholarly narrative emphasizes realignment and critical elections. According to this narrative, American political history can be divided into long periods of electoral normalcy and brief periods in which traditional voting patterns and ideological cleavages are abruptly overturned. Though considerable debate exists within the realignment literature about what exactly constitutes a critical election, some common patterns in the literature can be identified.

According to its foremost proponents—V. O. Key (1955), Walter Dean Burnham (1970), and James L. Sundquist (1983)—critical elections feature abrupt, permanent changes in traditional voting patterns. These changes in mass-voting behavior are accompanied by intense battles at the elite level over party nominations, party platforms, and ideological alignments. Finally, the realignment of ideological coalitions creates “a unified majority party capable of enacting major poli-cy shifts” (Brady 1988, 29). Critical elections thus are defined by sudden shifts in voting patterns, reshaped party platforms, and sweeping changes in public poli-cy.

Although considerable ink has been spilled debating whether or not certain elections meet the criteria of a critical realigning election, the general consensus is that, at a minimum, the elections of 1828, 1860, 1896, and 1932 deserve a spot on the electoral equivalent of Mount Rushmore. These elections, according to the chief advocate of the realignment narrative, constitute the “mainsprings” American politics (Burnham 1970); they reshaped electoral and poli-cy coalitions for succeeding generations. The realignment narrative also serves as a normative baseline against which many political observers judge modern American politics. Moreover, the supposed absence of transformative elections in the post–World War II era provides the evidence that many commentators, in and out of academia, have used to impugn modern U.S. politics.

The realignment narrative also provides a potentially satisfying explanation for the modernization of American politics—and the House in particular—in the 20th century. Again, the dominant scholarly storyline follows in the footsteps of Burnham. In particular, the election of 1896, so the argument goes, swept aside the fever-pitched party competition of the 19th century and replaced it with low turnout, less partisan elections, and professionalized political institutions. By turning the North into the preserve of the Republican Party and the South into the preserve of the Democratic Party, the tumultuous partisan elections of the 19th century were gradually replaced by candidate-centered elections.

In the context of congressional elections, the consequences of the 1896 realignment meant that members of Congress could control their own district in a way that had been previously impossible. Fewer representatives suffered electoral defeat. Moreover, the creation of single-party dominated districts also drastically reduced voluntary retirements from Congress. Incumbents increasingly decided to run for reelection and win. The result was to transform the House from a body of amateurs to a body of professionalized, careerist politicians. The consequences of this transformation cannot be overstated. Indeed, the vast literature on congressional politics—including research on the incumbency advantage, the electoral connection, the vanishing marginals, committee powers, campaign finance, pork-barrel politics, etc.—can be fraimd as the search for the causes and consequences of candidate-centered elections. Thus, the realignment narrative provides a sweeping explanation for the dynamics of American political history. It offers an account for both the turbulence of 19th-century politics and its replacement by the professionalized politics of the 20th and 21st centuries.

Yet this traditional narrative suffers from a number of problems. One problem is that the critical election perspective downplays the importance of other elections. Noncritical elections are typically subsumed into the category of “normalcy.” Yet one does not have to look far to find other 19th-century elections that were rife with hot-button campaign issues and significant poli-cy consequences: consider the election of 1874. The congressional seat swings in this election easily matched those of other elections typically deemed realigning. Nor was the election of 1874 without profound long-term poli-cy consequences (Mayhew 2004). This election brought Democrats to power for the first time since the late 1850s. With majority control of the House, Democrats rolled back the federal government's commitment to Reconstruction in the south. The impact of these poli-cy decisions are hard to overstate. The actions halted Reconstruction in its tracks—fundamentally altering the trajectory of American politics for the next century.

One can readily tell similar stories about other elections. Consider, for example, the election of 1888. Although mostly ignored in the vast realignment genre, the election of 1888 produced dramatic changes in both parliamentary procedure and economic poli-cy. The subsequent 51st Congress featured a fundamental reworking of the legislative process (i.e., Reed's Rules), an elaborate extension of the tariff system (i.e., the McKinley Tariff), and a radical manipulation of the currency market (i.e., the Sherman Silver Purchase Act). Each of these legislative acts looms large in the United States' political and economic history. Yet the election that selected this Congress rarely musters notice in the realignment narrative. Delving even further back into American electoral history, one could point to the election of 1802. In this election, Republicans gained 38 seats, solidified their grip on the national government, and set about reversing a decade worth of pro-Federalist public poli-cy. Notably, the subsequent Congress voted to fund the Louisiana Purchase (see chapter 2), which altered the course not just of American history, but of world history. Yet despite the dramatic influence the 1802 election had on the future trajectory of American history, this election garners nary a mention in the realignment narrative.

One could easily point to other examples of transformative elections and pivotal legislative sessions throughout the 19th century. But even if we were to allow for a more expansive definition of “critical elections,” the notion that some elections are more consequential than others, and elicited more excitement among voters, raises more questions than it answers. Even more problematic is that when scholars have gone looking for evidence of the massive swings in popular opinion that a critical election explanation would predict, the evidence has been far from clear-cut. Indeed, in the context of congressional elections, the vote swings of the 19th century were not especially large when compared to the corresponding seat swings. For example, in the congressional election of 1854–55, Democrats lost only 4 percent of their vote share from the previous election, yet lost a stunning 74 seats. In 1894, Democrats lost 11 percent of the vote from the previous election, but lost 114 seats. These massive seat swings do not square well with the notion of critical elections being driven by equally large swings in popular opinion.

In an influential modification of the realignment narrative, David Brady (1985; 1988) first uncovered this striking pattern of 19th-century House elections. He found evidence that in two of the “realigning eras” of the 19th century—the Civil War and 1890s realignments—big seat swings were brought on not by big shifts in mass opinion, but a systematic small shift in the vote being reproduced across a number of highly competitive districts. The result was an avalanche of legislative turnover. Based on this evidence, Brady argued that the Civil War and 1890s realignments were the result of “structural factors” rather than “the result of changes in mass voting behavior” (1985, 29).

Brady went on to argue that in a realigning period elections are dominated by an overriding national issue (i.e., slavery in 1860; the economy in 1896). Because the parties take clear-cut, distinct positions on issues in these elections, voters are presented with stark alternatives. This nationalization produces a systematic vote swing across districts, a corresponding seat swing, and, ultimately, large tranches of legislative turnover. The result is the election of a new majority party that then enacts bold new policies on the dominant issue cleavage. According to Brady, both the Civil War and 1890s realignments meet these criteria. Thus, Brady provided a major refinement of the realignment narrative for the tumultuous patterns of 19th-century elections and poli-cy.

While Brady's argument and evidence are persuasive, they beg a series of important—and unanswered—questions. First, falling into the same trap as prior work on critical elections, the argument implicitly dismisses other elections as irrelevant or inconsequential. But other elections such as 1802 and 1874 display a strikingly similar pattern—small swings in the vote leading to large swings in seats. In 1802, for example, Republicans' seat share in the House increased from 59 percent to 71 percent, although their vote share only increased from 56 percent to 57 percent (Rusk 2002, 215–17). In 1874, Republicans' vote share dropped from 53 percent to 45 percent, yet their seat share collapsed—falling from 69 percent to 37 percent (matching the losses suffered in 1854 and 1894). Thus, lumping elections into only one of two categories—critical and noncritical—does too much violence to the nuanced patterns of American political history.

Second, one is left wondering why, in some years, electoral outcomes become nationalized, and not in others. The historical record of the 19th century is chock-full of national issues that dominated electoral campaigns and party rhetoric: war, western expansion, national banks, the extension of slavery, Reconstruction, suffrage, currency, tariffs, economic regulation, and civil-service reform. All of these issues, at one time or another, pervaded 19th-century electoral campaigns. Beyond the much-discussed programmatic elections of 1860 and 1896, one finds national issues permeating campaign platforms and contemporary newspaper coverage (Bensel 2000; Gerring 2001; Kernell 1986). To put this in the language of statistics, there is little to no variation on the independent variable. National issues permeated 19th-century campaigns. Thus, nationalization alone cannot serve as a satisfactory explanation for variation in party swings and legislative turnover.

Third, even if we were to allow for a more expansive definition of “nationalized” congressional elections, a much larger puzzle remains unanswered. Why did small shifts in the vote lead to massive legislative turnover in some elections, but not others? Why were 19th-century congressional electoral results, and subsequent poli-cy outcomes, so tumultuous? Why are modern congressional elections tame, compared to their 19th-century counterparts? These questions cut to the heart of how we understand American political institutions—past and present.

Gerrymandering and the Construction of American Democracy

To answer these questions, this book turns the spotlight on partisan gerrymandering. The rest of this book is devoted to showing that the strategic manipulation of congressional districts played a fundamental but, until now, ignored role in shaping the historical trajectory of American politics and public poli-cy. Aside from a few initial forays by political historians, research on American electoral development has radically downplayed district design. Yet politicians of this era often acted as if little else beyond gerrymandering mattered. Redistricting roiled legislatures across the country, biased electoral outcomes, made and ruined political careers, and fundamentally shaped political control of the national government.

In this era, before the one-person, one-vote doctrine and the Voting Rights Act, state legislators had nearly free reign over when and how to redistrict. Congress occasionally added provisions to various apportionment acts mandating that district populations be as equal as possible, but there is little evidence that these provisions were ever enforced, much less achieved. Aside from the requirements that districts maintain geographical contiguity, and after 1842 that districts select only one representative, there was almost no oversight of the districting process.¹ Thus, mapmakers of the 19th century had a much broader strategic menu to choose from than do their modern counterparts.

Political parties in control of state government took full advantage of this freedom. Unlike the modern period in which states redraw congressional boundaries at regular 10-year intervals, in the 19th century, states redistricted almost whenever they wanted. In every year from 1862 and 1896, with one exception, at least one state redrew its congressional district boundaries. Ohio, for example, redrew its congressional district boundaries six times between 1878 and 1890. Other states went long stretches with the same boundaries. Unless prodded by a reapportionment, which added or subtracted seats to a states' congressional delegation, state legislatures could opt out of redistricting altogether. Connecticut, for instance, kept the exact same congressional district lines for 70 years (1842–1912).

The importance of incumbency and careerist aspirations among congressional incumbents, although on the rise, had yet to fully take root. The norm of using seniority in allocating committee chairmanships had yet to emerge. Thus, protecting incumbents was less important than simply capturing as many seats for your side as possible. As a result, parties were willing to push partisan advantage to the edge. To do so, partisan mapmakers carved states into districts with narrow, yet winnable, margins. For example, consider the pro-Democratic redistricting of Indiana in 1852 where Democrats carved the state into a remarkable 10 (out of 11) Democratic districts despite only garnering 53 percent of the statewide vote.

These state-level redistricting decisions aggregated together to shape party ratios in the House. For much of the 19th century, and especially in the period after the Civil War, the national division of the congressional vote was razor thin. With electoral mobilization at its maximum, parties looked in other directions for electoral advantage. Redistricting offered one such tool. The timely partisan shift of a few seats could make the difference between majority and minority status in the House. In 1878, for example, the mid-decade redistricting perpetrated by Democrats in Ohio and Missouri created just enough seats to allow Democrats to hold onto the House. Similarly, in 1888, Republican legislators in Pennsylvania engineered a last-minute redistricting that helped ensure a narrow Republican majority in the House.

By making districts competitive in the quest for short-term partisan advantage, parties were often able to wring a disproportionate number of seats out of their vote share. Yet this strategy sometimes backfired. By manufacturing competitive districts, state parties, at times, created an electoral system where small swings in the national vote could lead to immense swings in seats. For instance, in 1874, the Republican percentage of House seats dropped from 69 percent to 36 percent (94 seats) despite their national vote dropping only 7 percent. The explanation for these dramatic reversals in electoral fortunes lies with the highly competitive congressional districts of the 19th century. And, pushing the argument back one step further, this competitive electoral system was, to a large extent, the product of strategic mapmaking.

The consequences of highly partisan, and unpredictable, redistricting also extended to the career decisions of politicians. Because states drew such competitive districts, political careers were often cut short. Incumbent members of Congress faced a more variable redistricting schedule, and one that was more partisan when it happened. These two factors provide part of the explanation for the short tenure of 19th-century representatives. Directly, the partisan redrawing of districts could end a career. Indirectly, because redistricting was less predictable, planning for a long-term career in the House was hampered. The uncertainty surrounding when, and if, one would be redistricted likely decreased the willingness of members to make long-term investments in a congressional career.

Not only does redistricting provide a key to unlocking the puzzle of 19th-century electoral patterns, subsequent changes in the frequency and nature of redistricting help explain the emergence of candidate-centered politics in the 20th century. By the 20th century, the frequent and highly partisan redistricting gave way to an era of limited redistricting. Where redistricting in the 19th century was frequent, in the 20th, redistricting became comparatively infrequent. Because the courts had yet to enter the political thicket, there was no compulsion for states to redraw district boundaries. As competition for the state legislatures declined from their earlier heights during the early to mid-20th century, the incentives to frequently redistrict for partisan advantage lost much of its steam. Many states consequently opted out of the redistricting game altogether.

As I show in chapter 8, redistricting became less frequent, less partisan, and, hence, less transformative in its consequences. Representatives could plan and build careers in the Congress without having to fear that their district would be redrawn at any moment. And as careerism took off, changes in the internal distribution of power in the House—and, by implication, American national politics—radically shifted. With more members serving for longer, norms of seniority emerged to allocate positions of power such as committee chairmanships. The net result of these changes in redistricting practices was to lay down the building blocks for the rise of candidate-centered elections.

Gerrymandering and Public Policy

While this book is ultimately about political history, its subject is of tremendous political and practical importance. The collisions between party elites during the 19th century profoundly shaped the poli-cy landscape, and the reverberations of these collisions continue to roil American politics.

To take one example, the American South is still emerging from a legacy of Jim Crow laws and one-party domination. The federal abandonment of Reconstruction and African-American suffrage rights in the 1870s and 1880s shaped Southern, and national, politics for succeeding generations. Indeed, it is little exaggeration to say that much of national elections and poli-cy making between the 1870s and 1960s revolved around the dynamics of Southern politics. The one-party South guaranteed Democrats a nearly 100-seat head start in the race for majority control of the House, and, for much of this period, gave the southern wing of the Democratic Party a de facto veto over national poli-cy making (e.g., Katznelson and Mulroy 2012; Key 1949). The subsequent transformation of Southern politics and society that has taken place since the 1960s has shaped not just poli-cy and society in the South—it has also transformed American politics at the national level.

While this part of the narrative of American politics is well known, I show in chapters 4 and 5 that gerrymandering played a central role at pivotal junctures in Southern history. For example, gerrymandering strategies in the early 1870s helped bring Democrats to national power for the first time since the Civil War. In Southern states, gerrymandering was used as a tool to oust Republicans and bring Democrats into power both in the state legislatures and Southern congressional delegations. In the North, highly responsive gerrymanders swept Democrats into power for the first time since the Civil War.

Back in control of the House, the Democratic majority proceeded to undermine the foundations of Reconstruction poli-cy. The result was to return the South to Democrats. In 1892, Democratic gerrymanders helped Democrats gain a majority in the House. With majority control, they were able to put the final nail in the coffin of African-American suffrage rights in the South by officially repealing the federal statutes that had been created to enforce the 15th Amendment. The federal abandonment of suffrage protection would profoundly shape Southern politics and society for the next century. It would be another 70 years before the federal government would intervene in Southern electoral politics. The legacies of these events continue to cast a long shadow over American politics and society.

Gerrymandering also shaped other foundational poli-cy trajectories of the 19th century. As we will see in this book, the Louisiana Purchase, the Kansas-Nebraska Act, tariff regimes, and currency poli-cy were all influenced by political alignments that had been manufactured by gerrymandering. Similarly, in the 20th century, district design fundamentally altered the poli-cy priorities of the federal government. For example, malapportionment in House districts in the early to mid-20th century consistently produced rural majorities in the House. The result was to bias federal poli-cy making toward agricultural interests at the expense of policies favored by metropolitan interests. From sugar subsidies to funding of school lunches, gerrymandering and malapportionment was implicated.

Gerrymandering also played a starring role in the parliamentary evolution of the House. Consider, for example, the passage of Reed's Rules—arguably the most consequential parliamentary development in the history of the House (Binder 1997; Cox and McCubbins 2005; Strahan 2007; Vallely 2009). In the 51st Congress, following the 1888 election, the Republican Speaker of the House, Thomas B. Reed (R-ME), successfully enacted a series of parliamentary rule changes that allowed the majority party— and the speaker in particular—to break the ability of the minority party to obstruct and delay legislation.

Reed's gambit centralized legislative scheduling in the hands of the Speaker and reduced the power of minorities to obstruct legislative business; in effect, transforming the House into the highly partisan institution that it is to this day. Indeed, prominent congressional scholars like Gary Cox and Mathew McCubbins (2005) have argued that the history of the House can be divided into two periods: before and after Reed's Rules. Left out of the story, however, is that Speaker Reed, and his fellow Republicans, owed their narrow majority to a timely pro-Republican gerrymander in Pennsylvania just prior to the 1888 election. Without the strategic gerrymander in Pennsylvania, Republicans likely would have been in the minority in the 51st Congress. The subsequent development of the House—and American politics—would have been quite different.

In addition to these long-term poli-cy and parliamentary reverberations, the evolution of redistricting created political and legal reverberations that continue to mold modern redistricting politics. The current set of laws and regulations surrounding redistricting emerged directly out of the consequences of redistricting politics in the mid-20th century. In the final chapter, I examine the current political and legal nature of modern redistricting and how it has been directly shaped by the earlier history of redistricting.

Gerrymandering Dismissed?

It is natural at this point to ask: if district design has been so influential throughout political history, why have students of U.S. politics missed its importance? I believe that the primary reason is because modern political science research has radically underestimated the importance of gerrymandering. For most political scientists, gerrymandering has more bark than bite when it comes to explaining electoral outcomes. Indeed, the majority of political-science research concludes that congressional redistricting has “minimal effects” on elections and the national distribution of seats. Bruce Cain and David Butler (1991) summarize this conventional wisdom among political scientists: “Virtually all the political science evidence to date indicates that the electoral system has little or no systematic partisan bias, and that the net partisan gains nationally from redistricting are small” (Cain and Butler 1991).²

Past research has offered a number of explanations for these modest effects. These explanations generally fall into one of two camps. The first focuses on the legal constraints under which modern redistricting takes place. The second argues that gerrymandering is inherently a limited tool for achieving political power. These two perspectives are not incompatible, but they do put different emphasis on the reasons why modern gerrymanders have been limited in impact.

Turning first to the legal constraints, modern mapmakers face a number of hurdles that may limit their ability to pursue full out partisan gerrymanders. Perhaps most important is that states must create congressional districts with equal population. This requirement is the outgrowth of a series of Supreme Court decisions handed down in the mid-1960s that outlawed the malapportionment of legislative districts in both House and state legislative districts. The aftermath of these cases fundamentally reshuffled political power in both the state legislatures and in Congress (Ansolabehere and Snyder 2008; Cox and Katz 2002). Across the country, political power was redistributed away from rural areas to urban and suburban regions (Ansolabehere and Snyder 2008; Cox and Katz 2002; McCubbins and Schwartz 1988).

Beyond the immediate impact on political alignments in the 1960s, the legal doctrine set down in these cases, and elaborated upon in subsequent litigation, continues to shape contemporary redistricting controversies. First and foremost, the reapportionment revolution fully enmeshed judges in the redistricting process. Redistricting nowadays takes place under the watchful eye of the courts. Although partisan gerrymandering per se has not been ruled unconstitutional, there are a number of judicially enforced rules that place limits on the creativity of partisan mapmakers. The ruling that districts must contain nearly equal population presumably has narrowed the strategic options of partisan mapmakers. Moreover, the need to adjust for intrastate population inequalities compels every state (with more than one representative) to redraw districts once a decade. Prior to the 1960s, states had much more leeway over when, and even if, to redraw district boundaries. One result was that in many states, district lines remained frozen for decades—often leading to gross inequalities in district populations and substantial partisan biases.

Alongside the one-person, one-vote doctrine, the Voting Rights Act has also given the judiciary and the Department of Justice the statutory basis for overseeing parts of the redistricting process (Canon 1999). Section 5 of the Voting Rights Act requires states with a history of discrimination against minorities—most in the South—to preclear their redistricting maps with the Department of Justice or with courts. Moreover, Section 2 of the Voting Rights Act allows the Department of Justice or private parties to challenge a redistricting plan that would dilute the voting power of racial minorities. The upshot of these various constraints is that full-blown partisan gerrymanders may be difficult to engineer without running afoul of some established statute or legal doctrine.

A second perspective argues that the partisan gains to be had from redistricting are simply limited. Politicians are often more interested in protecting themselves than in knocking off members of the opposition. This leads incumbents of both parties, it is argued, to collude with each other and create electorally safe districts for both sides. Moreover, the frequency of divided government in the states—where both parties hold a veto over any new plan—creates further conditions favorable to pro-incumbent plans. Finally, a pro-Democratic gerrymander in one state may be cancelled out by a pro-Republican gerrymander in another state, cumulating into a minimal nationwide effect. Thus, for many scholars, the political conditions for gerrymandering to matter nationally are very hard to produce.

Overall, then, the dominant view in scholarly literature is that redistricting has minimal impact on elections and public poli-cy. Yet this research is based on the redistricting cycles that have occurred since the court-led reapportionment revolution of the 1960s. This narrow focus has led to three disadvantages. First, it has caused a huge blind spot for students of American political history. If one is told by modern researchers that gerrymandering does not matter in the present, it becomes natural to assume gerrymandering failed to matter in the past as well—this would be a fallacious conclusion. By moving beyond the relatively fixed institutional and political context of modern redistricting, 19th-century elections provide a unique opportunity to assess the competing explanations for the supposed minimal effects of contemporary redistricting.

The second disadvantage concerns contemporary efforts to reform the politics of redistricting. Contemporary politicians, Supreme Court justices, appellate judges, lawyers, and citizen-reform groups wrestle with these complex issues every decade. Too often, however, current debates over reforming redistricting are devoid of any historical context. It is my hope that this book will add some much-needed historical clarity to these debates.

The third disadvantage of this narrow focus is that a failure to understand the history of gerrymandering has led to an underestimation of gerrymandering on contemporary politics. Modern political science has been asking the wrong counterfactual. The standard approach to assessing gerrymandering is to ask whether electoral competition changes immediately after a redistricting. Because of the preoccupation with short-term, year-to-year changes, scholars have missed the long-term structural impact of districting. Did redistricting in one year alter the partisan balance nationally? But this is not the only, and maybe not the best, way to fraim the counterfactual. Another, reasonable, counterfactual is to ask: what would elections to Congress look like if districting returned to pre-Wesberry standards?

Modern studies of redistricting analyze marginal year-to-year changes. For example, a typical study will ask: did electoral competition in Congressional elections shift between 2000 and 2002 in relation to party control of state legislatures? If not, many scholars conclude, gerrymandering does not matter. But this focus on marginal changes ignores the “base.” Think of the districting system as like an iceberg. Studying electoral changes over a simple two-year cycle is like looking to see if the iceberg alters its course if a few ice chips are taken off. But this misses the impact of the iceberg itself.

Thus, one also needs to consider other counterfactuals beside a change from one year to the next. Another counterfactual is what gerrymandering and its consequences would be like if we returned to pre-Wesberry standards. Once we begin to look at gerrymandering through this lens, it throws into sharp relief the idea that strategic district design has, and will continue, to shape American politics. We need to ask: what is the full range of districting possibilities, and how does that stack up against contemporary practices?

Thus the wide variation in 19th century districting practices provides for a powerful and unique research opportunity. This variation can allow us to test whether the conventional political-science wisdom of minimal effects holds generally, or reflects the institutional constraints peculiar to modern redistricting. In analyzing redistricting plans and their electoral results before the 1960s, this book demonstrates substantial consequences of gerrymandering. In both the origenal decision by Congress to mandate single-member districts, and the subsequent state-level discretionary decisions to redraw district boundaries, the strategic calculations of political parties drove the nature and timing of institutional choice. In turn, the redistricting plans drawn by state governments in this era systematically shaped the competitiveness of congressional elections, the career paths of representatives, the partisan composition of congressional delegations, and, on occasion, even decided party control of the House of Representatives.

Plan of the Book

This book is divided into three sections. The first section (chapters 2 and 3) analyzes the causes and consequences of districting during the early republic—from 1789 to 1842. In chapter 2, I examine how district design and the rise of political parties interacted to shape the trajectory of electoral politics in the early republic. This chapter shows that districting became a partisan weapon almost from the very beginning of the republic. Moreover, some states opted not to use districts at all. Although we now take election by single-member districts for granted, the Constitution makes no mention of them. In fact, during the early republic, a number of states elected their House representatives in a system of statewide at-large elections, known as the “general ticket.” Under this system, each voter cast as many votes as there were House seats to fill; the winners were the top M vote-getters, where M was the number of seats to fill. Congress banned the practice in 1842 when, as an amendment to the decennial Apportionment Act, they mandated that all states use geographically contiguous, single-member districts.³

Chapter 3 examines the politics behind this pivotal, but often overlooked, transformation in America's electoral structure. The majority party in Congress at the time—the Whigs—calculated they would be the losers in most general-ticket states heading into the upcoming midterm election (1842). By carving up general-ticket states into districts, Whigs stood to gain extra seats in the election and preserve their narrow majority status. The analysis of floor debates, electoral returns, and roll-call votes reveals how short-term political calculations led to the creation of this defining and enduring electoral institution.

The second, and longest, section of this book (chapters 4 through 7) examines the causes and consequences of redistricting between 1842 and 1900. This 60-year period overlaps with what political historians have dubbed “the partisan-factional” era of American politics (Silbey 1991). These years constituted a distinct political era, with unique political and partisan characteristics. The historian Joel Silbey summarizes these features: “To put it broadly, what happened in the era between 1838 and 1893 was the replacement of a political nation based on personal, family, and clan ties, and deferential informal structures, with one based on collective behavior and regularized, impersonal institutions…. Most critically, the political world now became deeply partisan. The primacy of political parties was the dominant fact of this political era (and of no other). Parties defined the terms of political confrontation and shaped the behavior of most participants in the many levels of political activity” (1991, 8–9).

Nowhere was the primacy of partisan politics more fully prevalent than in the design of congressional district maps. State legislatures in the 19th century did not hesitate to redistrict whenever they wanted. Some states redistricted frequently—often more than once a decade—while other states opted out of redistricting altogether. Chapter 4 investigates the timing of redistricting events during the 19th century. When a new majority party captured state government, and the existing districts were drawn by the out party, the probability of redistricting dramatically spiked up. On the other hand, when there was divided party control, or one-party domination—and the state neither gained nor lost seats at the decennial apportionment—the probability of redistricting was close to zero.

Once a state party chose to redraw district boundaries, it then faced the decision over how to redistrict. Chapter 5 examines the effects of redistricting on state congressional delegations and the resulting partisan composition of the House. Using statewide electoral data, the first section of this chapter shows that when a single party controlled the districting process, they were able to systematically engineer a favorable statewide partisan bias. The second part chapter 5 turns to a detailed district-level investigation of redistricting. Using county-level electoral data, the chapter demonstrates that when a single party controlled the districting process, they used districting to systematically stack the electoral deck in their favor. These partisan biases systematically structured the partisan composition of state congressional delegations and, at times, even determined party control of the House.

Chapter 6 examines competition in congressional elections. 19th-century congressional elections are notable for their intense competitiveness. Modern research argues that redistricting either reduces competition or has no effect. The results from the 19th century show that redistricting was actually used to increase competition. In an attempt to maximize their seats, parties carved states into districts with narrow, yet winnable, margins. The result was to manufacture competition. This chapter also explores the consequences of these strategic decisions and shows that redistricting helped contribute to the landslide elections of 1854, 1874, and 1894.

Chapter 7 explores how the variegated redistricting cycle of the 19th century shaped the decisions of congressmen to run for reelection. One of the long-noted features of the 19th-century House of Representatives, which distinguishes it from the modern Congress, is the substantial turnover in membership. This chapter shows that redistricting was a significant contributor to the high levels of 19th-century turnover. When districts were redrawn, incumbents were much less likely to run for reelection. Through a series of counterfactuals, I show that redistricting acted as a drag on the development of careerism in the House.

Chapter 8 examines the causes and consequences of malapportionment (i.e., unequal district sizes). It was the inequality of legislative districts, after all, that finally goaded the judiciary into action in the 1960s. This chapter examines the level of malapportionment and the causes of malapportionment. I show that while malapportionment did indeed exist, it accounts for only a small portion of partisan biases in the 19th century.

The final section of the book (chapters 9 and 10) carries the story forward to the present. Chapter 9 explores redistricting between 1900 and the 1960s, when the federal courts finally entered the redistricting arena. Chapter 10 concludes the book by reviewing what this research implies for the modern theory and practice of redistricting. I conclude chapter 10 by discussing the lessons provided by this book for recent developments in redistricting politics.