Fall 2004

Special Orders

Two Generations of Redistricting: An Overview

Charles S. Bullock, III

University of Georgia

In 1962 when the Supreme Court plotted a course through the political thicket in Baker v. Carr ¹, it revolutionized representation in the United States. The next four decades have witnessed restrictions on the discretion that legislatures had enjoyed in distributing seats among the public while also eliminating some of the constraints that had checked legislators' tendency to pursue partisan advantage. Population equalization, fair treatment of minorities, and maximization of partisan strength have come to the fore. In the course of redrawing legislative districts, urban areas have benefited while rural communities have lost and minorities have gained additional seats at the expense of whites. At the state level, legislators have increasingly been able to choose their constituents rather than vice-versa.

Equal Population

In the immediate aftermath of the Baker decision, the courts extended the equal population requirement to other jurisdictions (both chambers of the state legislature, Congress and local governments) and gradually narrowed the tolerances allowed in population deviations. Prior to 1962, states' most populous districts might be many times larger than the least populous so that major strides towards eliminating disparities could be taken even if wide variations continued. During the 1960s, it was not uncommon for states to have to draw and then re-draw their districts as courts refined definitions of what constituted "equal population."

Courts relied on two different sections of the federal Constitution to justify requiring equal populations in congressional districts and legislatures. Courts turned to the Equal Protection Clause of the 14th Amendment to require equal populations among state and local legislative districts. For years, many legislators interpreted Brown v. Thompson² as standing for the proposition that districts, other than for Congress, could vary by as much as +/- 5 percent.

Since the Equal Protection Clause does not apply to the federal government, courts had to find another basis for demanding equal populations among a state's congressional districts. The Supreme Court pointed to the provision requiring a representative form of government in Article1 Section 2. This provision came to be interpreted to mean that congressional districts within a state needed to be as nearly equal as possible. With the development of geographic information systems, rapid multiple iterations became possible so that population variations could be brought closer and closer to zero. As early as 1991, some states eliminated population variations among congressional districts reducing the range between the most and least populous district to a single person. In 1992 when a federal court had to draw Florida's congressional districts since the legislature never completed the task, multiple plans were offered but the only ones considered by the court reduced population deviations to a single person. In 2001, a federal court threw out a Republican plan for Pennsylvania in which the maximum population deviation range was only 19 people, when the Democratic challengers offered a plan that reduced that deviation to a single person.³

Georgia Democrats tested the proposition that population variation of +/- 5 percent provided a safe harbor for state legislative plans. Democrats faced a two-fold challenge as they sought to perpetuate 130 years of control. First, the growth areas voted Republican so simply shifting seats from under- to over-populated parts of the state was unacceptable. Second, in each of the three previous elections, Republican candidates for both chambers had outpolled Democrats statewide. To overcome these challenges the plans adopted by the Democrats soaked up as many Republicans as possible while allocating scarce Democrats to achieve maximum effectiveness. The 2001 plans clustered districts at the extremes of what was thought to be an acceptable population variations. Of 56 Senate districts, 36 were +/- 4 percent, 31 were +/- 4.5 percent and, surprisingly 16 were +/- 4.9 percent. In the House, 90 of the 180 districts were +/- 4 percent while 20 were +/- 4.9 percent. The substantially under-populated districts invariably had Democratic majorities while Republicans found themselves in over-populated districts.

In 2004, a three-judge federal court invalidated Georgia's plan.⁴ The court noted that the state failed to offer an acceptable rationale for its failure to come closer to equalizing population. The state acknowledged – any cursory view of the maps would prove this true – that the deviations were not justified on the basis of deference to traditional districting concerns such as to avoid cutting city or county boundaries or to promote compactness or contiguity. The court found unacceptable the state's rationale which was that it sought to minimize the loss of seats by south Georgia and Atlanta at the expense of more rapidly growing north Georgia and suburbs and to help Democratic incumbents. The court went back to decisions from the 1960s to underscore the unconstitutionality of creating districts of unequal population in order to advantage certain geographic areas. The Larios court did not specify the upper limits of acceptable population deviations for state legislatures. However, when the Georgia General Assembly failed to redistrict itself, the judges assumed the responsibility and imposed plans with maximum deviations of +\- one percent.

Equal Treatment of Minorities

Soon after the Supreme Court opened the way for challenges based on equal protection claims, Congress passed legislation that in time was interpreted as requiring fair treatment of minorities in redistricting. While the 1965 Voting Rights Act did not specifically mention redistricting, the courts included districting plans as a type of action subject to review by federal authorities before implementation by the states caught by the trigger mechanism incorporated into Section 4 of the legislation. This legislation sought both to eliminate current practices used to discriminate against blacks who wanted to participate in the political system and to ensure that state and local authorities not erect new barriers. To this end, states that had tests and devices as prerequisites to registration and where either registration or turnout in the 1964 presidential election fell below half of the voting age population were subject to Section 5. This provision required that any changes affecting elections must be approved either by the Attorney General of the United States or the federal district court sitting in the District of Columbia. The initial legislation applied to Alabama, Georgia, Louisiana, Mississippi, South Carolina, Virginia, and about 40 percent of North Carolina. Later amendments expanded coverage to sixteen states.

In 1972, the Department of Justice (DOJ) objected to Georgia's congressional plan because it carefully excluded the homes of Andrew Young and Maynard Jackson, two likely candidates, from the 5th Congressional District. This district which contained the city of Atlanta had come close to electing Young in 1970 before reconfiguration. After having its initial plan rejected, the state legislature drew the district to include the homes of these two African-Americans which enabled Young to join Houston's Barbara Jordan as the first African-Americans elected from the South in the twentieth century.

The courts clarified the responsibilities of submitting authorities in Georgia v. U.S. and held that the burden of proof had to be borne by the submitting authority. Jurisdictions must demonstrate to the satisfaction of the federal reviewer that the districting plan will not discriminate against minorities. It was not up to the Justice Department to prove that a plan was discriminatory.⁵

By 1980, the Justice Department had interpreted Section 5 as prohibiting retrogression. That is, a plan would not be approved if it reduced the minority concentration in any district in which minorities constituted a substantial share of the population. Certainly it would be unacceptable for a jurisdiction to transform a district in which most residents were minorities into a district in which most were white. Even in the absence of retrogression, Justice would not approve a plan that had been drawn with the intention of discriminating against minorities. Population changes had increased the black concentration in Georgia's under-populated 5th district from 44 percent in 1970 to slightly more than half in 1980. The state House proposal for increasing the population in the district to satisfy the equal population requirement maintained the black percentage as it was in the census. The Senate plan increased the black concentration to 69 percent. A conference committee settled on 57 percent black. After the DOJ rejected the plan, the state appealed to the federal district court in the District of Columbia. The three-judge panel in Brisbee v. Smith noted that the chair of the House redistricting committee had said that, "The Justice Department is trying to make us draw nigger districts and I don't want to draw nigger districts."⁶ Since the committee chair had so clearly shown his racism, the court rejected the plan as submitted.

To secure approval of a plan and hold elections, Georgia had to increase the black concentration in the 5th district to 65 percent. Although the Justice Department has denied that it ever had a rule requiring that minority concentrations be pegged at 65 percent, in the 1980s a number of jurisdictions and even some federal judges thought 65 percent to be a necessary threshold.

DOJ reviewers of plans, as well as federal judges, regularly required that districts be well more than 50 percent black. The rationale for demanding an extraordinary majority was that the black population tended to be younger and that blacks were less likely to register and less likely to turn out than whites. Therefore in order to have an electorate that would be majority black on election day, it was necessary to inflate the black population in the district.

Civil rights supporters contended that districts needed to have a black majority among voters to offset the unwillingness of many white voters to support a black candidate, regardless of the African-American's credentials. An extensive analysis of South Carolina elections in the 1960s and 1970s conducted by sociologist James Loewen (1987) demonstrated the reluctance of white voters to support black candidates.⁷

Amendments passed in 1975 extended coverage to Hispanics and other language minorities. The renewal of the Voting Rights Act in 1982 expanded the geographic scope and conditions under which African Americans could seek redress if they felt themselves to be the victims of discrimination. Congress amended Section 2 to provide that challenges to an existing electoral system would not require plaintiffs to demonstrate an intent to discriminate. Evidence that minorities had less opportunity to elect their candidates of choice would suffice. The entire nation and not just 16 states were subject to Section 2.

When the Supreme Court encountered Section 2, it established three preconditions for successful litigation. The first, and most appropriate for redistricting, requires that the minority population be sufficiently large and concentrated that a single-member district could be drawn in which the minority voting age population would constitute a majority. The other two elements require that the minority population be politically cohesive and that a white bloc vote usually defeats minority preferences.

DOJ relied heavily on Section 2 when reviewing post-1990 census redistricting plans. At the very end of the 1980 cycle, a federal district court in Arkansas overturned part of the state legislative plan upon a showing that it would be possible to draw districts more likely to elect African Americans (Jeffers v. Clinton).⁸ Justice incorporated Section 2 standards into the reviews of districting plans from jurisdictions subject to Section 5 and demanded the drawing of additional majority- minority districts when possible. Avoidance of retrogression no longer sufficed to secure approval of a plan.⁹ DOJ rejected the congressional and state legislative plans from Georgia twice while the North Carolina congressional plan was turned down once. A plan for an educational board elected in Louisiana was also rejected sending a cue to the legislature that it would need to increase the number of majority-black congressional districts from one to two. In each instance, DOJ questioned why additional majority black districts had not been incorporated into the plan.

DOJ demands increased the number of majority-black congressional districts in the South from five to seventeen. Each of these new districts elected an African American in 1992. In addition, two more majority-Hispanic districts were drawn in Texas and one was added in Florida. The Florida district and a Texas district in the Rio Grande Valley sent Hispanics to the Congress, while an Anglo won the Houston district.¹⁰

Almost immediately after the implementation of the new plans, a reaction set in. Duke University law professor Robinson Everett challenged the plan imposed on North Carolina. He was especially troubled by the new twelfth district, a narrow band of geography designed to link minority concentrations in most of the cities in the Piedmont, extending more than 150 miles from Gastonia in the west to Durham in the east. The trial court refused to even hear the case and Everett appealed to the Supreme Court. The high court found that Everett had raised a justiciable issue by invoking the Equal Protection Clause of the Fourteenth Amendment. In an opinion authored by Justice Sandra Day O'Connor, the court fretted about the shapes of the districts which seemed to a majority to bespeak apartheid. O'Connor observed that, "we believe that reapportionment is one area in which appearances do matter."¹¹ Immediately cases bubbled up in other states with the Supreme Court elaborating on its concerns in a Georgia filing, Miller v. Johnson, where the court explained that while jurisdictions could and should consider race when drawing plans, they violated the Fourteenth Amendment if race predominated their deliberations.¹² Ultimately courts overturned racial gerrymanders in Georgia, North Carolina, Virginia, Texas, Florida, and Louisiana. Following each of the decisions, the challenged districts became more compact and the minority concentration reduced. In only one instance, however, did it impact the minority legislator. Eliminating Louisiana's 620-mile long 4th district that extended into almost half of the state's parishes, prompted Cleo Fields to drop out of Congress and make an unsuccessful bid to become governor.

The new plan drawn for North Carolina reduced the black concentration in the 12th District from 57 to 47 percent as it became more compact, extending only from Charlotte northward with one arm reaching to Winston-Salem and the right arm to Greensboro. Everett challenged this plan, also. In what became the fourth review by the Supreme Court of a Tarheel State districting plan during the 1990s, the court gave further clarification to its expectations. Although plaintiffs had sought to disentangle the effects of party and race and demonstrate that race had again predominated legislative considerations, the Supreme Court concluded that party was the overriding factor in the new plan and, therefore, the plan was constitutional.¹³ The Court acknowledged that the strong support black voters give Democratic Party nominees makes disentangling the influence of race and party difficult. So long as the legislators spoke mainly in terms of partisan goals, the court was not inclined to probe.

As already discussed, Georgia Democrats tested the limits of the population equality standard while trying to thwart GOP hopes of winning legislative majorities. A second stratagem, adopted as Democrats tried to stretch their minority support into a legislative majority, prompted them to reduce the minority concentrations in a number of legislative districts. In the past, DOJ would likely have rejected these efforts as retrogression. In anticipation of a challenge, the Democratic leadership convinced the Black Caucus that their best interest lay with a plan that maximized Democratic seats even if it failed to maximize black seats which was a dramatic change from the previous decade. To allay fears of black legislators that they risked defeat if black population was dispersed to bolster white Democrats, the state's expert, David Epstein, testified that blacks had a 50-50 chance of electing their preference with black concentrations as low as 44 percent.¹⁴ The plans reduced black concentrations in several districts represented by African-American senators to slightly more than 51 percent of the voting age population. The Department of Justice objected to three of these and the trial court agreed with DOJ that minority political influence had been diluted in these districts. The state appealed to the Supreme Court.

In Georgia v. Ashcroft, the Supreme Court found for the state.¹⁵ This decision seemed to severely restrict the "no retrogression" standard that DOJ had used to assess compliance with Section 5. The Supreme Court held that Section 5 can be satisfied either by having a relatively small number of districts in which minorities are so concentrated as to almost ensure that the minority-preferred candidate will win, or by having a greater number of districts in which minorities are less concentrated so that they have fewer assurances of electing their preference but will, at a minimum, have an impact on the election. The Court's opinion suggested that minority influence districts, or what the Court referred to as coalitional districts, might have minority concentrations as low as 25 percent of the voting age population.¹⁶

In the challenge to the Texas congressional redistricting that was heard by a three-judge federal court (Sessions v. Perry) a few months after Ashcroft, white Democrats sought to strike down the Republican gerrymander by arguing that the districts filled by white Democrats were protected under Ashcroft. The implications of this claim, which the court rejected, were that districts that elected Democrats could not be dismantled if they had enough minority population to be considered an influence district.¹⁷ The consequence would be to protect Democrats against partisan gerrymanders while leaving Republican incumbents vulnerable to such challenges.

Partisan Gerrymandering

In Davis v. Bandemer, the Supreme Court acknowledged the justiciability of claims asserting a partisan gerrymander. The Court, however, did not find for the plaintiff.¹⁸ During the next 18 years, no plaintiff succeeded in a case alleging a partisan gerrymander. Courts were not blind to the existence of partisan gerrymanders but set such a high standard for securing relief that no one could meet it. To overturn a partisan gerrymander, it would be necessary for plaintiffs to demonstrate that they were completely shut out of the political process. But even when the plaintiffs' party received substantially fewer seats than its proportion of the vote, it might still have seats in the state's congressional delegation or control the governorship and thus continue to have some influence in the political system.

The Supreme Court revisited the issue of partisan gerrymandering in 2004. The new millennium had witnessed blatant examples of one party taking advantage of the opposition. Much of what Georgia Democrats did to the GOP has already been discussed. In Pennsylvania and Michigan – and in 2003 in Texas – Republicans took advantage of their majority status to enhance their positions.

Because of slow population growth, Pennsylvania lost two seats from its delegation that had eleven Republicans and ten Democrats. Republicans pushed through a plan that paired Democratic incumbents and had the potential to give the GOP a 13-6 edge. The 2002 election did not yield everything the GOP had hoped for but they did secure a dozen seats even as their gubernatorial candidate managed just 44 percent of the vote. Democrats went to court to challenge the plan which restricted them to 37 percent of the congressional delegation, far less than the 51 percent vote share Al Gore had amassed in 2002. Four members of the Court wanted to reverse Davis since they believe that there were no standards under which one could successfully prove a partisan gerrymander (Vieth v. Jubelirer).¹⁹ While Justice Kennedy agreed that under present standards it would be impossible to prove a partisan gerrymander and therefore voted against the plaintiffs, he joined the dissenters in holding that claims of a partisan gerrymander would remain justiciable holding out the prospect that at some future date an acceptable standard might emerge.

Conclusion

Since the Supreme Court macheted its way into the political thicket, it has largely eliminated the population deviations that had characterized both congressional and legislative plans. Where once the most populous districts in a jurisdiction had many times more people than the least populous, now congressional districts within a state often vary by no more than a single person and in Larios, a federal court narrowed the deviations in a state legislature to a range of two percent.

While the standards applied for equal population have moved in an almost unbroken trend towards smaller variations, court requirements for racial fairness have been less unidirectional. For most of the period, the courts and DOJ were increasingly demanding in terms of promoting the likelihood that minorities would be elected. The high water point came in the early 1990s when DOJ required a number of jurisdictions subject to Section 5 of the Voting Rights Act to adopt "Max Black" plans and draw majority-black districts whenever possible. The courts undid a number of these on the basis that by concentrating so heavily on race, they violated the equal protection clause. In the next decade, African-Americans often agreed to have minority concentrations dispersed if doing so would increase the number of districts likely to elect Democrats. The Supreme Court has embraced that approach.

The third potential area of change was to establish judicial standards limiting partisan gerrymandering. While the Supreme Court acknowledged that partisanship could become too obtrusive, no plaintiff has succeeded in convincing a court that a plan was so heavily tilted towards one party that it precluded effective participation in the political system by the disadvantaged party.

The development of the current set of standards has come as the result of hundreds of suits. If the past is any guide, the deluge of decennial litigation will resume with the next census. Since redistricting is the most political action that legislatures take, we can expect that those who have an advantage will push the envelope and that the disadvantaged will continue to appeal to the courts for relief.

Notes

1. Baker V. Carr, 369 U.S. 186 (1962).

2. Brown v. Thompson, 462 U.S. 835 (1983).

3. Vieth v. Pennsylvania, 195 F. Supp. 2d 672 (M.D. Pa. 2002).

4. Larios v. Cox, 300 F. Supp. 2d 1320 (N.D. Ga. 2004).

5. Georgia v. U.S., 411 U.S. 526 (1973).

6. Busbee v. Smith, 549 F. Supp. 494 (D.D.C. 1982).

7. James W. Loewen, "Racial Bloc Voting in South Carolina," Unpublished (1987).

8. Jeffers v. Clinton, 730 F. Supp. 196 (E.D. Ark. 1989).

9. The Supreme Court ultimately banned DOJ's practice of applying Section 2 standards when conducting Section 5 reviews (Reno v. Bossier Parish School Board, 528 U.S. 320, 1999).

10. Douglas D. Abel and Bruce I. Oppenheimer, "Candidate Emergence in a Majority Hispanic District: The 29th District in Texas," in Thomas A. Kazee, ed., Who Runs for Congress? Ambition, Context and Candidate Emergence. (Washington, DC: Congressional Quarterly Press, 1994).

11. Shaw v. Reno, 509 U.S. 630 (1993).

12. Miller v. Johnson, 515 U.S. 900 (1995).

13. Easley v. Cromartie, 532 U.S. 234 (2001).

14. David Epstein, report filed in Georgia v. Ashcroft, 123 S. Ct. 2498 (2003). Research into racial voting patterns in congressional elections shows whites in the 1990s more likely to support black candidates than Loewen (1987) observed a generation earlier. Approximately a third of white voters support black Democrats running for Congress (See Charles S. Bullock III and Richard E. Dunn, "The Demise of Racial Districting and the Future of Black Representation," Emory Law Journal 48: 1209-1253) while a black Republican may do even better with white voters. See Charles S. Bullock III, "High Turnout in a Low-Turnout Year: Georgia's Second District," in James A. Thurber, ed., The Battle for Congress (Washington, D.C.: Brookings Institution, 2001).

15. Georgia v. Ashcroft, 123 S.Ct. 2498 (2003).

16. Sessions v. Perry, Civil Action No. 2:03-CV-00354 (E.D. Tx., Marshall Division, 2003).

17. Davis v. Bandemer, 478 U.S. 109 (1986).

18. Vieth v. Jubelirer, 2004 W L 894316 (U.S. Pa).

Charles S. Bullock III (Ph.D., Washington University) is Richard B. Russell Professor of Political Science at The University of Georgia, where he teaches classes in southern politics and legislative politics. He is the author of more than a dozen books, including The New Politics of the Old South (Rowman and Littlefield, 1998) and co-author with R. Keith Gaddie of Elections to Open Seats in the U.S. House (Rowman and Littlefield, 2000). Professor Bullock can be reached via email at cbullock@uga.edu.

This page is best viewed at a resolution of 800 x 600 pixels.
Copyright, The Carl Albert Center