< Return To Hearing
Professor Keith Gaddie
May 16, 2006
On the Need for Renewal of Section 5 of the Voting Rights Act
Remarks prepared for presentation to the United States Senate Committee on the Judiciary
My name is Ronald Keith Gaddie. I am a professor of political science at the University of Oklahoma. Your former colleague now the President of the University of Oklahoma, David Boren, with whom I have the privilege of teaching with in the Political Science Department, has asked me to give the members of the committee his best wishes.
Since 2001 I have worked as a litigation consultant and expert witness in voting rights and redistricting cases in several states, including Texas, Oklahoma, Wisconsin, New Mexico, Virginia, South Dakota, and Georgia, for jurisdictions, plaintiffs, Democrats and Republicans. I have authored or coauthored eight books on aspects of American politics. Currently, with my colleague Charles S. Bullock III of the University of Georgia, I am completing an analysis of the progress on minority voter participation and elections in the jurisdictions covered by Section 5 of the Voting Rights Act. This study is supported by the American Enterprise Institute, through the Blum Thernstrom Project on Fair Representation.
The Voting Rights Act has framed American electoral politics for forty years. The Act stands as the enforcement mechanism for one of two "superior" redistricting principles of voting rights, that of racial fairness (the other principle being the one-person, one-vote guarantee). The most proactive tools of the Voting Rights Act are up for renewal. This periodic review and renewal of legislation gives us the chance to ask, "what have we done and how far have we come?"
This statement will highlight trends in minority participation in the seven states originally covered by the Act. I will then frame some topics for discussion as we move toward the renewal of the Act, with some attention paid to the history and prospects for minority voter participation in the states originally covered by Section 5.
The initial concern of the Voting Rights Act was access to the political process. Political scientist V. O. Key, writing over a half-century ago in his classic Southern Politics: In State and Nation, observed that "the South may not be the nation's number one political problem . . . but politics is the South's number one problem." (1949: 3) Participation was necessary to a functioning democracy, for Key, who observed that the problem of participation in South, like every other problem, could be traced to the status of blacks.
What was the status of Southern blacks? Well, depending on where you went in the South variations were in evidence, but southern blacks were generally disfranchised, generally discriminated against, and generally held at a distance from white society -specifically the prosperous part of white society -- by public policy. Key observed that "whites govern and win for themselves the benefits of discriminatory public policy" and further noted that "discrimination in favor of whites tends to increase roughly as Negroes are more completely excluded from the suffrage" (1949: 528). Exclusion from the vote did not cause discriminatory treatment, but it most certainly reinforced the status of Southern blacks. Key observed in a clinical fashion what Martin Luther King, Jr. argued passionately in 1965, "give us the vote and we will change the South." It was only by the exercise of political power through ballots that politicians would change policy in the long run.
We have the opportunity for a frank, informed conversation about the shape of the Voting Rights Act for the future. And I thank the chairman and committee for holding these hearings in order to advance this conversation. What should take place in this conversation?
In 1964, there was one black state legislator in the seven states originally covered by Section 5. The South lumbered under an archaic and outdated political and social culture that clung to the past at the possible cost of the future. There was no viable competition to the Democratic Party, which was locally a contrary adjunct to the national party, opposed to the Democrats in the rest of the nation on most every dimension of social policy politics.
The contemporary South is vibrant, the most populous and fastest-growing region of the nation. Southern children are more likely to attend integrated schools than in the rest of the nation, and an African American is more likely to have black representation in the South than anywhere else in the nation. Education and income differences across the races are matters of degree rather than orders of magnitude. Southern blacks are registered and voting at rates comparable to black voters in the rest of the nation. There is a two-party system in the South which fosters black political empowerment and office-holding. However, that empowerment comes as the party of choice for most African-Americans, the Democratic Party, has been relegated to minority status in the legislature of three of the original Section 5 states and also in the covered states of Texas and Florida.
Race still divides the South, but southern blacks are not helpless in the pursuit of political, social, and economic goals when compared to five decade ago. The context of race relations and the status of minorities in the South are dramatically changed from four decades ago.
The times, they change. But the times do not change at the same pace in every place. The law too must change, so that it might address the prevailing circumstances of the times, with respect for the past. Every renewal of the Voting Rights Act has brought some change, some evolution of the law to address the circumstances of the present with respect for the past.
In my previous testimony to the US Commission on Civil Rights and before the Judiciary Committee of the U.S. House of Representatives, I used as a starting point Table 1, which contains information from Earl and Merle Black's Politics and Society in the South. This table shows the growth of black voters in the South. By 1984, South Carolina and Mississippi ranked at the top of proportion black electorate. Mississippi and Alabama registered the largest proportional gain of size in the black electorate, though Mississippi simultaneously ranked "high" and "low" because the baseline for minority participation was so very low, large proportional gains were inevitable. Georgia and Louisiana conversely rank near the bottom of proportional gain in part because of having the highest rates of black registration of any state originally covered by Section 5. By 1984, the black percentage among registrants tracks closely with the black percentage in the voting age population, evidence of the success of the Voting Rights Act in eliminating obstacles to participation. The states with the largest potential black electorates (Mississippi and South Carolina) had the most-heavily African-American voter registration rolls.
The Black brothers' analysis informs us as to the proportionately largest black electorates in the South. Tables 2 and 3 present Census Bureau estimates of black voter registration and participation since 1980 for the seven states originally covered by Section 5. Black registration lags white registration for most of the time period in the seven covered states analyzed (as it does in nonsouthern states throughout the time series). But, for the last four elections for which there are comparative data, black registration in six of the seven states (all but Virginia) exceeds black registration rates in the nonsouthern states. In three of the states (Georgia, South Carolina, Mississippi), black registration rates exceeded white registration rates within those states for at least two of the last four elections.
Black turnout rates are less consistently above the national average. As indicated in Table 3, three of the original Section 5 states - Alabama, Mississippi, and Louisiana - have black turnout consistently above the national average for black turnout. Every covered state except Virginia reports higher black than white turnout rates at least once since 1990, and Georgia reports higher black than white turnout in three of the last four general elections. Differences in racial registration and participation have become differences of degree rather than of magnitude.
These votes are generally translated into seats. Figure 1 presents time-lines, since 1964, of the percentage of state legislative seats held by black incumbents in the state legislatures of the seven original Section 5 states. Of these states, Alabama has achieved proportionality in the legislature relative to their respective citizen voting age populations, while Georgia, Mississippi, and North Carolina are approaching proportionality (data for this graphic appear in Table 4).
At the congressional level, the 1990s saw significant advancement of descriptive African-American representation. The number of southern, African-American members of Congress tripled. In the states covered by Section 5, the number increased from three in 1991 to a current eleven (one from Virginia, two from North Carolina, one from South Carolina, four from Georgia, one from Alabama, one from Mississippi, and one from Louisiana) -- 18% of all congressmen from these states are African-American, compared to 25% African-American citizen voting age population. If we also add the black congressmen elected from the other two Section 5 southern states - Texas and Florida - we total seventeen black MCs, or 15% of all MCs from nine states that are collectively 18.9% black by citizen voting age population. Black representation in the Section 5 states is not proportional to the black citizen voting age population. But, black descriptive representation is as high as it has ever been in southern legislatures in modern times, and is approaching proportionality to the extent that the geographic placement of black voters and the tendencies of electorates in general elect black candidates who seek legislative office (see Table 5). As is widely recognized, in single-member, plurality political systems like in the US (in contrast with the proportional systems used in most of Europe), the majority group gets a disproportionate share of the legislative seats and the minority groups gets less than its proportional share.
What is Retrogression?
Assume, for example, an existing thirty-district state legislative map had three majority-minority districts, all of which elected candidates of choice of the minority group. The new map eliminates a minority districts and does not create an offsetting one elsewhere. The new map retrogresses against existing minority opportunities. Were a new district plan to eliminate a minority district while creating a new one, the number of minority opportunities for access would be unchanged, and retrogression would not have occurred. The submitting jurisdiction rather than the Department of Justice (DOJ) bears the burden of proof for demonstrating non-retrogression. Another facet of the retrogression standard generally prohibited the reduction in the minority concentration in an existing majority-minority district.
The US Department of Justice may not apply other standards in addition to retrogression when determining whether to preclear new districting plans. The Supreme Court has ruled out standards that go beyond the charge to the agency under Section 5 which only sets the floor of ensuring no loss of political ground by minorities.
Georgia v. Ashcroft altered the retrogression standard. Georgia lowered the black percentage of the voting age population in a number of state legislative districts and redistributed this population to craft more districts that were competitive for Democratic candidates. In majority-white districts with increased numbers of blacks it would be possible for a biracial coalition to elect Democrats. In reviewing the Georgia plan, the Supreme Court held that evidence of non-retrogression can include coalition districts - identified by plaintiffs as districts between 30% and 50% black by population. This offered to jurisdictions two avenues for satisfying the non-retrogression: fewer, safer minority districts, more less-safe minority districts and coalition districts, or some combination. The second option offered in Ashcroft permits reducing the concentrations of minorities in majority-minority districts - which may result in less certainty of minority candidates being elected -- in exchange for a greater number of districts in which minorities may be able to coalesce with white voters to elect candidates preferred by the minority voters. In other words, with less certainty comes greater opportunity to spread influence, assuming one were willing to pull, trade, and haul. Such districts were deemed more permissible if the elected representatives belonging to the minority community supported the creating of access and influence districts in the political process. As stated by Justice O'Conner, writing for the majority:
the retrogression inquiry asks how "voters will probably act in the circumstances in which they live." Post, at 19. The representatives of districts created to ensure continued minority participation in the political process have some knowledge about how "voters will probably act" and whether the proposed change will decrease minority voters' effective exercise of the electoral franchise.
So there are multiple avenues to satisfy Section 5. Does this broadening of solutions also broaden the scope of districts protected from retrogression? To understand the means by which one satisfies nonretrogression, we need to consider the nature of Section 5. Has it become so blurred by recent litigation that the provision is emerging as a vehicle for the pursuit of partisan advantage rather than ensuring minority group access to the political process?
Republican administrations, specifically the first Bush Administration, used the Voting Rights Act as a lever to encourage the creation of majority-minority districts, and to limit the opportunities to create cross-racial coalitions in support of Democrats. White Democrats in turn preferred districts with sizeable (but not majority) minority populations because of the biracial coalitions that could command more seats. In the 1980 and 1990 rounds of redistricting, African-American Democrats preferred districts with black majorities sufficient to elect an African American.
The aggressive use of Section 2 of the Voting Rights Act to create majority-minority districts in the early 1990s resulted in an electoral map that shifted one-third of all southern congressional districts to the GOP in a three-election period. That these newly acquired Republican districts were largely bereft of minority voters and next-door to majority-minority districts is more than coincidence. These districts were urged by the Justice Department as part of a "maximization strategy", using preclearance as a policy lever. Congressional plans which maximized minority seats and had been approved (in some cases demanded) by the Justice Department were overturned by courts in Georgia, Louisiana, North Carolina, Virginia, and Texas. A quote from John Dunne, assistant attorney general for civil rights in the first Bush Administration Justice Department, taken at deposition in Miller v. Johnson, is instructive in acknowledging the political dimension to the use of preclearance:
You know, I can't tell you that I was sort of like a monk hidden away in a monastery with only the most pure of intentions. I am a Republican. I was part of a Republican administration. And to tell you that at no moment during the course of my, the discharge of my responsibilities, was I totally immune or insensitive to political considerations, I don't think would justify anybody's belief. But I can't really tell you much more than that.
The consequence -- concentrating the most loyal Democratic voters into the fewest districts possible -- paid political benefits. The number of congressional districts with between 20 and 40% African-American population southwide - districts especially likely to elect white Democrats -- fell from 50 seats to 22, and within two elections the number of Republicans from southern states nearly doubled.
Another effect was the shape of the new congressional constituencies. Described as bizarre, tortured, irregular, and non-compact, many of the new congressional districts created by states to comply with Justice Department efforts - combined with the pursuit of other political goals such as incumbency protection -- stretched the credibility of the terms "compact" and "contiguous." This "spiral down" effect on compactness resulted from the meeting of the policies of the Department of Justice and the determination of the legislatures in the jurisdictions to protect incumbents.
DOJ refused to enforce any compactness rule asserting that compactness was a state policy and therefore the level of compactness in districts was an issue outside of the scope of the preclearance process. As stated in its preclearance letter for the Texas congressional redistricting scheme [w]hile we are preclearing this plan under Section 5 the extraordinarily convoluted nature of some of the districts compels me to disclaim any implication that our preclearance establishes that the proposed plan is otherwise lawful or constitutional...Our preclearance of the submitted redistricting plan in no way addresses the state's approach to its redistricting obligations other than with regard to section 5.
DOJ's policy that "reductions in the minority percentages in one district might be effectively counterbalanced by increases in others" essentially meant that jurisdictions did not have to be geographically specific when attempting to remedy the dilution of a minority community's voting strength. In jurisdictions such as North Carolina, Georgia, and Texas, mapmakers responded by drawing far less compact minority districts than might have been possible, in order to ameliorate the political effects of drawing the compact majority minority district. This "new" standard of compactness was then used to prompt the crafting of additional majority minority districts, which could not be drawn under the original standard of compactness advocated by the jurisdiction. The result was a downward spiral of demands for crafting minority districts, lowered compactness, and sparring to protect incumbents, which in turn led to the least compact plans, but with more majority minority districts than ever before.
So, we see two political dimensions of the implementation of voting rights creating further legal and political conflict: the effort by southern legislatures to protect incumbents and facilitate (possibly) politically-motivated Justice Department demands to create new minority opportunities, results in the torturous shape of the legislative districts challenged in the Shaw/Reno-styled cases of the 1990s.
Partisan goals and the role of minority voters continue to define redistricting. Most recently, Georgia and Texas offer opposite perspectives on the effort to seize electoral advantage while playing politics that affect or relate to the Voting Rights Act.
In 2001 Georgia Democrats moved to retain control of the state legislature while also expanding their foothold in the state's congressional delegation. This was accomplished through the efficient allocation of black, Democratic voters in a fashion partially opposed by the Justice Department, and which required litigation to establish. This efficient allocation reduced minority majorities particularly in some state Senate districts and was considered retrogressive by the Justice Department. Because the elected representatives of the community of interest approved of the strategy, and because minority choices could prevail in most of the coalition districts, the Supreme Court held that the use of coalition districts as an alternative to less heavily-minority districts was permissible (though not required) to satisfy Section 5.
This change in the definition of retrogression occurred during the recent Texas redistricting. In Texas, plaintiffs challenged the mid-decade congressional redistricting on several dimensions. One claim was that districts lacking a majority of a minority, but electing candidates preferred by minority voters, were protected from change under Section 5. One Plaintiff's expert testified that districts as low as 5% minority population might be protected from change under the Voting Rights Act, unless agreed to by the minority community's leadership. This reasoning was rejected by both the Justice Department, which precleared the new Texas map and the Federal district court in Austin, which explicitly rejected the argument that there is an obligation to create coalition districts under federal law.
The use of Section 2 as incorporated into Section 5 reviews was a powerful lever for concentrating instead of spreading minority populations in creating minority-majority districts and accompanying, largely white districts that presented electoral opportunities for Republicans. From the perspective of the Republican Party, it has been successfully used, given the dramatic realignment of southern congressional delegation in the early 1990s. The redistricting compelled by the Justice Department under Section 5 is not solely responsible, but when combined with the departure of incumbents and wedge issues, the redistricting facilitated the doubling of Southern Republican congressional strength. The interpretation under Ashcroft facilitates the reintroduction of coalition constituencies, with the approval of the representatives of the minority community, or, in other words, allows in theory for the crafting of constituencies of the sort that once contributed to the Democratic southern congressional majority until 1994.
This latest change raises a question that I first articulated in 2003 at the Texas State Senate redistricting hearings, of how one establishes a baseline for evaluating retrogression. My perspective is that of a social scientist charged with conducting analyses to inform those who make and interpret the law, rather than from the perspective of a legal thinker, and should be taken as such. If retrogression is evaluated in the context of an entire map, and constituencies in which a white legislator relies on biracial support are among the districts protected from retrogression, then how are those districts to be treated in subsequent efforts to baseline minority access and evaluate retrogression?
My concern with efforts to use retrogression to maintain coalition districts is that it sets up a circumstance where part of the legislative map becomes immune to political change under redistricting. Why? Because, with the exception of Dade County Cubans, the minority populations covered by the Voting Rights Act are Democratic constituencies. If districts where a cohesive minority electorate is not in a position to control the election of consequence are counted among protected districts, then party bias is introduced. Any district, anywhere, in which minorities, no matter how small a percentage of the electorate, vote for the Democratic candidate, conceivably becomes immune from change. In this instance, Democratic districts are locked in as part of the district format. One party gets a guarantee of protection for its seats, but the other does not.
The use of the Ashcroft retrogression standard also presents an empirical problem for the people who measure the baseline. When coalition districts are part of the baseline for measuring retrogression, a question that arises is "how much representation do minority voters get out of these coalition districts?" Is a district where minority voters are a quarter of the electorate as effective for those voters as one where minority voters control the electoral process? Is it half as effective? A quarter as effective? Do we compute the expected utility of different districts, based on the probability of the minority group to control the respective districts, and compare the expected utility of different maps? While Justice O'Connor's decision does not explicitly advocate an algorithm to ascertain the performance of coalition districts, it certainly invites the use of an algorithm by virtue of its logic.
If minority representation guarantees can be achieved through coalition, another question arises. If minority candidates and candidates of choice can be elected from districts with minority percentages of the voting age population of under a majority, say as low as 44.3%, or even as low as the hypothetical 30% coalition district advanced by plaintiffs and noted by the court in Ashcroft, then is there a need to have Section 5 coverage of the jurisdiction? In order to have an "even chance" at winning a 44.3% voting age population district, and we assume equal turnout with 90% minority voter cohesion, a candidate of choice will need to capture 18.1% of the Anglo vote. To have an even chance at winning a 40% minority-turnout district requires 23.3% of the Anglo vote. And, to have an even chance at 30% minority-turnout and 90% cohesion requires 32.8% of the white, Anglo vote. These thresholds for white crossover voting increase as the rate of minority turnout falls.
If candidates are capable of winning in less-than-majority districts (as Sanford Bishop, Cynthia McKinney, David Scott, and, in the past, Andrew Young have done)-- or can exercise control of seats under circumstances where the minority of white voters coalesced with the cohesive minority vote to create winning coalitions -- is Section 5 coverage still necessary? If the prevailing candidate is not just a candidate of choice, but a candidate of color, is Section 5 coverage still necessary? The circumstances that favor the use of coalition districts - sufficient white cross-over vote and political support from minority elected officials - seem to satisfy the notion and circumstances that Section 5 coverage is no longer necessary.
Virginia offers evidence that local circumstances can change in order to allow jurisdictions to "bail out" from under Section 5. Efforts should be made to explore how the Justice Department can further work with jurisdictions that have made real strides in improving their racial political climate, in order to remove the footprint of federal oversight where it is no longer required. The existing rules for bailing out from Section 5 set high evidentiary standards for jurisdictions to attain. Do those standards impede the removal of the preclearance mechanism in states where recent evidence of progress is overwhelming?
A state in which this question is relevant is Georgia. The fastest-growing of the original Section 5 states offers real evidence of voting rights progress in the last decade. African-American candidates run as well or better than white candidates for statewide office of the same party. The work of Professor Epstein indicates that African-American legislative candidates are capable of winning non-majority black districts on an even basis. There are currently two black Republicans in the Georgia Legislature, from heavily-white Gwinnett County and Middle Georgia Houston County. The state has the most-heavily black congressional delegation in the US House (31% of seats). Georgia's African-American Attorney General Thurburt Baker asserted that:
The State (sic) racial and political experience in recent years is radically different than it was 10 or 20 years ago, and that is exemplified on every level of politics from statewide elections on down. The election history for legislative offices in Georgia - - House, Senate and Congress - - reflect a high level of success by African American candidates [Post-trial brief of the state of Georgia, Georgia v. Ashcroft C.A. No. 01-2111 (EGS) (D.C., DC 2002), p. 2].
The current rules governing bailing out from under Section 5 preclude Georgia's departure, due to recent objections by the Justice Department. And, many local jurisdictions have a history of Section 5 objections. At the highest levels of government, Georgia accomplished more than any other state covered by Section 5. However, given the numerous jurisdictions in Georgia that have not been subject to a Section 5 objection for the last ten years, and the generally high rate of voter participation in the state, many localities may wish to explore this option.
Then there are other jurisdictions where one must wonder why Section 5 coverage continues. Certainly, outside the South, the continued coverage of ten townships in New Hampshire defies logic. These ten townships are all at least 96% Anglo white by population. In the 2000 election these ten townships averaged over 50% voter participation as a share of voting age population. One township - Pinkham's Grant - has no people residing in its borders. No Section 5 objection has ever been registered against a New Hampshire township. Two townships in Michigan are subject to preclearance, and the Department of Justice has issued no objection in Michigan in over three decades of coverage. Yet still Michigan submits its congressional and state legislative plans for preclearance because the Section 4 trigger picked up two townships containing a total of six voting precincts.
In Alaska native voter turnout lags that of whites, Native Americans and Aleuts are represented well in excess of their popular proportions in the legislature, and according to the preclearance analysis submitted for the Alaska legislative districts in 2001, Native and Aleut candidates have a fair chance of winning districts as low as 35% Native/Aleut population. The Alaska state constitution contains provisions for diversity of representation that extend beyond the guarantees of retrogression arising from Section 5.
A careful examination of the need to continue coverage under Section 5 could pay tremendous benefits in terms of defining the appropriate scope of the application of the law. The Lawyers Committee on Civil Rights Under the Law offers an efficient description of the conditions to be met by a jurisdiction seeking to bail out from under Section 5 coverage. The jurisdiction must show that, for the previous ten years: it has not used a test or device that has a discriminatory purpose or effect as a precondition to registering or voting; not have had a US court issue a final judgment against the city or county for voting discrimination; show full compliance with section 5 -- including timely submission of voting changes and no implementation of objectionable changes before final resolution; not had a proposed voting change objected to by the attorney general and no declamatory judgment denied under section 5 by the US District Court for the District of Columbia; no Federal examiners were assigned to the city or county under the Voting Rights Act. In addition, a jurisdiction must show that it has not engaged in other discriminatory practices prohibited by the law; must show that it has taken constructive steps to increase minority access to the political process; and show that there has been an increase in minority political participation.
Three Different Cases from the South
Change is afoot in Georgia, and throughout the South. Circumstance and politics have changed, and both black political empowerment and white acceptance of black politicians is part of that New South. Part of this change is the ability of black politicians to pull, haul, and trade, and the willingness of sufficient white voters to pull the lever for those politicians. Again, as observed by Representative Lewis:
I think many voters, white and black voters, in metro Atlanta and elsewhere in Georgia, have been able to see black candidates get out and campaign and work hard for all voters. And they have seen people deal with issues as, I said before, that transcend race: economic issues, environmental issues, issues of war and peace. . . So there has been a transformation, it's a different state, it's a different political climate, it's a different political environment. It's altogether a different world that we live, really.