David Canon’s U.S. Senate testimony on the Voting Rights Act
Senate Judiciary Committee Subcommittee on the Constitution, Civil Rights, and Property Rights Voting Rights Act: Policy Perspectives and Views from the Field
Testimony of Professor David T. Canon Department of Political Science, University of Wisconsin, Madison
June 21, 2006
1. Statement of Qualifications
I am a professor of political science at the University of Wisconsin, Madison. My research interests are in race and representation, political careers, congressional reform, partisan realignments, and the historical analysis of Congress. My major research on the question of racial representation was published in Race, Redistricting, and Representation: The Unintended Consequences of Black-Majority Districts [University of Chicago Press, 1999]. This book was the winner of the American Political Science Association’s Richard F. Fenno Prize for the best book published on legislative politics in 1999. I am author of approximately 25 scholarly articles and chapters, three scholarly books, seven edited volumes, and a major reference work on congressional committees.
I have testified as an expert in two voting rights cases in federal court and as an expert consultant in two other federal voting rights cases. In one of those cases, I was asked by the attorneys for the United States to analyze the Georgia State Senate redistricting plan under the new “totality of circumstances” test to analyze factors affecting whether the plan is retrogressive under § 5 of the Voting Rights Act (in the remand of Georgia v. Ashcroft). As outlined in the majority opinion in Georgia v. Ashcroft, this analysis must balance the loss of “ability-to-elect” districts that provide minority voters with equal opportunities to elect their candidates of choice against the potential gain of substantive representation in so-called “coalitional” and “influence districts.” The majority opinion also urged the lower court to examine other factors such as substantive representation of black interests that may flow from African American legislators’ leadership positions in the institution. My report conducted that analysis, but the case did not go to trial so I was not deposed and did not testify.
My testimony will examine several important issues that are relevant to why the Voting Rights Act should be renewed. I will focus my comments on issues that I have directly addressed in my research: the importance of the Voting Rights Act (VRA) in providing for the representation of racial and ethnic interests in the U.S. Congress, the importance of Section 5, and ability-to-elect and influence districts in the context of Georgia v. Ashcroft.
2. The Voting Rights Act and Racial Representation in Congress: The Necessity of Ability-to-Elect Districts
The 1982 Voting Rights Act Amendments and subsequent interpretation by the Supreme Court in the 1980s (especially Thornburg v. Gingles, 478 US 30, 1986) required that minorities be provided with equal opportunities able to “elect representatives of their choice” when their numbers and configuration permitted. The 1982 Amendments do not guarantee the right to achieve proportional representation by ensuring that the percentage of minority representatives will be equal to the percentage that minority voters comprise in the jurisdiction. Instead, Section 2 protects the right of minority voters to have equal opportunities to elect their candidates of choice, regardless of whether that candidate of choice is a minority or non-minority. In other words, it is the minority voters’ choice, not the race or ethnicity of the candidate, that matters. As a practical matter, however, in most cases minority voters will prefer to elect candidates who are also minorities. In that sense, the Voting Rights Act provides equal opportunities for minorities to achieve descriptive representation. This is the same context within which the Supreme Court used the term in Georgia v. Ashcroft. Therefore, when I refer to “descriptive representation” in my testimony, I am referring to the equal opportunity of minority voters to elect their chosen candidates, which in most cases will result in descriptive representation.
As a result of the 1982 Amendments, in 1992, 15 new U.S. House districts were specifically drawn to help African Americans to elect their chosen candidates to Congress and ten districts were drawn to help Latinos elect their chosen candidates. Section 5 of the VRA played a key role in producing this outcome, as the Department of Justice required several covered states to create minority majority districts, based on its interpretation of what was required by the 1982 VRA Amendments and the Gingles decision. Some of these ability-to-elect districts were challenged by the landmark Shaw v. Reno (509 U.S. 630, 1993). However, subsequent decisions made it clear that race could continue to be a factor in congressional redistricting as long as it did not predominate over other traditional districting principles, especially when racial and partisan motivations were intertwined (Easley v. Cromartie, 532 U.S. 234, 2001).
Those who argue that certain provisions of the Voting Rights Act should not be extended, and opponents of the Act more generally, point to the isolated success that minority candidates, such as Douglas Wilder, Barack Obama, J.C. Watts, Julia Carson, and Emmanuel Cleaver, have had in winning elections in which a majority of the electorate is white as evidence that special protection for ability-to-elect districts through the pre-clearance process is no longer needed.
There is little support for the optimistic view that blacks will win many House seats in white majority districts. The numbers are stark: of the 28,410 U.S. House elections from the adoption of the 15th Amendment in 1870 through 2004, only 563 (2%) produced black winners (107 individuals). A more meaningful statistic for the purposes of this discussion is the proportion of blacks who have been elected in black-majority and white-majority districts since the passage of the Voting Rights Act in 1965 (before the Voting Rights Act, the distinction between white-majority and black-majority districts was fairly meaningless because a large proportion of blacks were disenfranchised). In the 8,047 House elections in white-majority districts between 1966 and 2004 (including special elections), only 49 (0.61%) were won by blacks.
This number is even more striking when one considers the unusual circumstances surrounding nearly all of those elections. First, 11 of the 49 victories are accounted for by Ron Dellums (D-CA). He represented what The Almanac of American Politics describes as “the most self-consciously radical district in the nation” [Barone, Ujifusa, and Matthews 1975, 66; the 11 elections were from 1970-1990. The district has not been white-majority since 1992]. The combination of the urban ghetto of Oakland and the radical white voters from Berkeley make this district an extreme outlier.
Six elections are accounted for by Alan Wheat, who was elected in 1982 by winning the Democratic nomination with only 31% of the vote. As the only black candidate in the field of eight, he was able to rely on his base of black votes (he had been a state legislator in a black majority district before running for the House). After wining the primary over a field that split the white vote, he was able to win the general election in the heavily Democratic district. Subsequently, Rep. Wheat was able to appeal to white voters after gaining the power of incumbency and was easily reelected to the seat in a district that was only 25% black, until he lost a bid for the U.S. Senate in 1994.
Katie Hall (D-IN) also won election under fortuitous circumstances, but she was not able to maintain her seat in a district that was only 22% black. When Adam Benjamin (D-IN) died of a heart attack in September, 1982, the congressional district party chair was mandated by state law to name the party’s nominee (the primary had already passed). Richard Hatcher, the black mayor of Gary, was the district chair and named Hall for the slot. Party leaders were outraged that Benjamin’s widow was not named, but Hall narrowly won the general election against a Republican whom the Almanac described as “pathetically weak” [Barone and Ujifusa 1985, 452; the Republican nominee only spent $10,526 in losing the election]. In the next election, Hall managed only 33% of the vote in the Democratic primary, a strikingly low figure for an incumbent, and one that closely matches the 30% minority population of the district.
Andrew Young (D-GA) and Harold Ford (D-TN) account for seven elections and will be discussed below. Gary Franks (R-CT) and J.C. Watts (R-OK), who account for five of the cases, won as Republicans and are not considered by black leaders as sympathetic to black interests. In fact, Franks waged a vicious battle simply to be included in the Congressional Black Caucus.
The 1996 elections raised hopes that blacks would start winning more often in white-majority districts. Five black incumbents, all in the South, lost significant numbers of black voters when their black-majority districts were ruled unconstitutional, but were still able to win. Three of those, Cynthia McKinney (D-GA), Sanford Bishop (D-GA), and Corrine Brown (D-FL), won in white-majority districts. Even more significantly, Julia Carson (D-IN) won an open seat race in a district that was 69% white and had been held by Andrew Jacobs (D-IN) for 20 years. These results led to headlines such as “Is the South Becoming Color Blind?” [Fletcher 1996, 13; Sack 1996, 1]. However, McKinney and other supporters of majority-black districts argued that black incumbents were able to win through the power of incumbency [McKinney 1996, 26]. Since 1996, only two additional non-incumbent African Americans were elected to the House in white-majority districts (both in 2004): Gwen Moore (D-WI) in a district that had a bare majority of white voters (50.4%) and Emmanuel Cleaver (D-MO), the two-term mayor of Kansas City, won in a district that is 66% white.
Therefore, only three of the 49 elections (Carson, Moore, and Cleaver) provide much hope for black victories in white-majority districts.[1] On the other hand, 302 of the 353 elections (85.6%) between 1966 and 2004 in black-majority House districts have produced black representatives, including all of them since 2002. (Districts that are neither black-majority, nor white-majority are not included in this analysis).
The success of black politicians noted above in white-majority House districts, the elections of Douglas Wilder as governor of Virginia, Carol Moseley Braun and Barack Obama (D-IL) to the Senate, and Norm Rice as mayor of Seattle, indicate that blacks can win with white-majority electorates. However, as David Lublin points out, “their victories attract attention precisely because of their exceptional nature. Empirical evidence indicates that racial composition of the electorate overwhelms all other factors in determining the race of a district’s representative” [Lublin 1995, 112-13]. The numbers bear repeating: only 49 of 8,047 elections in white-majority U.S. House districts have provided black winners since 1966, and most of those were in unusually liberal districts or with some other idiosyncratic context that prevents generalizing to other districts. While the Voting Rights Act and its amendments, in my opinion, only provide an equal opportunity for black voters to elect candidates of choice rather than guaranteeing that outcome, 49 of 8,047 elections is not much of an equal opportunity.
Therefore, ability-to-elect districts are typically comprised of a majority of the relevant minority voters. This certainly is not always required, and the specific level of minority population required for a “performing” district should be examined by map-drawers and the courts every ten years on a case-by-case basis, depending on the levels of racially polarized voting. In districts in which there is substantial crossover voting in biracial or multi-racial blocs, it is possible for ability-to-elect districts to have less than 50% minority voters. The “packing” of minority voters (into districts comprised of at least 60-65% minority voters) promoted by some voting rights advocates and politicians in the early 1990s appears to be unnecessary.
3. The Necessity of Renewing the Pre-Clearance Provision of Section 5[2]
Given the concentration of black voters in the South, the legacy of legal discrimination, and the centrality of debate over renewal of the pre-clearance provisions of Section 5, it is important to separate the South in this analysis (southern districts and states comprise most of the areas covered by Section 5).[3] Following Reconstruction when federal troops withdrew and the Republican party left the South [Valelly 1995], blacks were almost completely disenfranchised through the imposition of residency requirements, poll taxes, literacy tests, the “grandfather clause,” physical intimidation, other forms of disqualification, and later the white primary [Davidson 1993, Davidson and Grofman, 1994]. Because of these practices no African Americans were elected to Congress from seven southern states originally covered by Section 5 of the Voting Rights Act between 1897 and 1973.[4]
Before the 1990 reapportionment, the South held a majority of the nation’s districts that had between ten and thirty percent black voters, and a large majority of the so-called “black influence” districts (30-50%). However, no districts in the South were black majority in the 1970s and only three were black majority in the 1980s. Given the patterns of racial bloc voting, no blacks were elected from majority white districts in the South between 1980 and 1994. Two, Andrew Young (D-GA) in 1972 and Harold Ford (D-TN) in 1974, were elected in districts that were 44% black and 48% black, respectively, but both districts became black-majority after the 1980 redistricting. Thus, the creation of new black-majority districts in the South in 1992 gave blacks a realistic opportunity to elect black politicians for the first time since Reconstruction.[5]
This is a central argument in favor of extending the pre-clearance provision of Section 5 of the Voting Rights Act. It was nearly impossible for blacks to be elected to Congress from the South before the pre-clearance process required the creation of black-majority districts in the South that provided black voters with equal opportunities to elect their chosen candidates: blacks were not elected in majority-white districts in the South and state legislatures did not draw black-majority districts until they were compelled to by the law and the pre-clearance process. Recall, we are not talking about ancient history here, but 1990-1992.
Critics of the pre-clearance provision also point to the extremely low rate of rejection by the Department of Justice. Some have interpreted this as evidence that pre-clearance is no longer needed because objectionable plans are claimed to be relatively rare. However, this ignores an important mechanism that helps generate this low rate of rejection: because of the pre-clearance process, covered states are less likely to submit electoral arrangements and institutions that violate the Voting Rights Act. There is no doubt that the deterrent effect is real as documented by a recent study by Professor Luis Fraga of the impact of more information requests by the Justice Department on discriminatory voting changes.[6] While the analogy is imperfect, nobody advocates pulling all traffic cops off the streets because voluntary compliance with traffic laws is relatively high. If police officers no longer monitored speed limits or ticketed drivers for running stop signs and traffic lights, it is fairly clear that violations of the law would increase. Similarly, if pre-clearance was abandoned, it is very likely that more local governments and even some states would be more likely to implement laws that harmed minority voting rights. Critics of Section 5 respond by saying the harmed voters would still be able to sue under Section 2 of the Voting Rights Act. Pre-clearance is a more effective tool than relying on litigation to enforce the law because many potential plaintiffs would not have the resources necessary to initiate law suits and discriminatory voting changes would be allowed to go into effect, probably for several years while the litigation was pending.
Second, pointing to the relatively low rate of rejection also ignores the fact that the voting changes that were rejected between 1996 and 2005 represent the full range of tactics that have been used in the South to abridge voting rights including moving and reducing the number of polling places, changing from district-based to at-large elections, annexations and redistricting that dilute minority voting power (including three state-wide redistricting plans: Georgia, South Carolina, and Texas), and an administrative plan for implementation of the National Voter Registration Act of 1993. These are not trivial violations that can be ignored, but important changes in electoral practices that would have hurt minority voting rights if the pre-clearance process had not been in place (see http://www.usdoj.gov/crt/voting/sec_5/obj_activ.htm for a complete list of Section 5 rejections).
Third, there are additional reasons why the number of objections have declined. The Supreme Court’s decision in Reno v. Bossier Parish II (520 US 471, 1997) prevents the Department of Justice from objecting to intentionally discriminatory voting changes merely because they do not have a retrogressive effect. Bossier II has had a tremendous negative impact on the ability of the Department of Justice to object to discriminatory voting changes.[7] Finally the recent decline in Section 5 objections is consistent with the pattern of fewer submissions being made in mid-decade after the substantial volume of submissions are made earlier in the decade during the decennial redistricting cycle.
4. Influence and Ability-to-Elect Districts in the Context of Georgia v. Ashcroft
Congress is also considering restoring the standard for “retrogression” that was in place before Georgia v. Ashscroft (539 US 461, 2003). The proposed legislation would clarify that the purpose of Section 5 of the Voting Rights Act is to protect the ability of minority citizens to elect their preferred candidates of choice (rather than allowing ability-to-elect districts to be traded off for influence districts).[8]
I support this clarification of Section 5. Two types of objections have been made to the new “totality of circumstances” test of Georgia v. Ashcroft: 1) that the new test is vague and unworkable, 2) that allowing influence districts to be traded off for ability-to-elect districts would erode the gains in the opportunities to elect candidates of choice that have been made in the Congress in the past forty years. I find the second point to be the most compelling reason to restore the legal standard that preceded Georgia v. Ashcroft.
Test is unworkable: The majority opinion in Georgia v. Ashcroft outlined a new Section 5 “totality of circumstances” analysis. Specifically, to determine whether a plan is retrogressive under Section 5 of the Voting Rights Act the state must demonstrate that any loss of equal opportunities to elect candidates of choice (often resulting in descriptive representation) that may flow from “unpacking” safe black majority districts, must be offset by gains in substantive representation that may come from the creation of greater number of influence and coalitional districts. While the Court did not provide detailed guidance on how to assess the potential tradeoff between what it termed “descriptive” and “substantive” representation, it asserts that states are permitted a choice in maintaining this balance as long as the overall level of representation of black voters was not diminished under the plan in question. The majority opinion says, “Section 5 leaves room for States to use these types of influence and coalitional districts. Indeed, the States’ choice ultimately may rest on a political choice of whether substantive or descriptive representation is preferable. The State may choose, consistent with Section 5, that it is better to risk having fewer minority representatives in order to achieve greater overall representation of a minority group by increasing the number of representatives sympathetic to the interests of minority voters. [internal citations omitted] . . . In assessing the comparative weight of these influence districts, it is important to consider the likelihood that candidates elected without decisive minority support would be willing to take the minority’s interests into account.” [p.18.]
While the general principles of this new totality of circumstances analysis for retrogression are clear, the specific application of the principles is not. The dissent in Georgia v. Ashcroft lays out the challenges posed by this broadened analysis:
Indeed, to see the trouble ahead, one need only ask how on the Court’s new understanding, state legislators or federal preclearance reviewers under §5 are supposed to identify or measure the degree of influence necessary to avoid the retrogression the Court nominally retains as the §5 touchstone. Is the test purely ad hominem, looking merely to the apparent sentiments of incumbents who might run in the new districts? Would it be enough for a State to show that an incumbent had previously promised to consider minority interests before voting on legislative measures? Whatever one looks to, however, how does one put a value on influence that falls short of decisive influence through coalition? Nondecisive influence is worth less than majority-minority control, but how much less? Would two influence districts offset the loss of one majority-minority district? Would it take three? Or four? The Court gives no guidance for measuring influence that falls short of the voting strength of a coalition member, let alone a majority of minority voters. Nor do I see how the Court could possibly give any such guidance. The Court’s “influence” is simply not functional in the political and judicial worlds.
I share some of these concerns about the complexity of this task. Substantive representation requires that legislators are aware of the preferences of their constituents and take concrete legislative action to address those concerns.[9] Political scientists have studied a broad range of actions that legislators take on behalf of their constituents and there is no universal agreement of which actions provide the best measures. David Mayhew examined the “electoral connection” that drives the behavior of members of Congress. In their effort to stay in office, representatives will appeal to their constituents through “advertising, credit claiming, and position taking” [Mayhew 1974]. Morris Fiorina pointed to the importance of constituency service, especially dealing with bureaucratic red tape, as another dimension of representation [1989].[10] Richard Fenno discussed the range of “home styles” that legislators will cultivate to represent their constituents in a variety of ways [1978; 2003; the more recent book examines how several African American members of Congress represent their constituents in their districts and in Washington D.C. both through their “home styles” and a range of legislative activities]. Richard Hall analyzed participation on committees and on the floor to gauge what members of Congress do to represent their constituents’ interests beyond simple roll call voting [Hall 1996]. My book on race and representation in Congress examined bill sponsorship and cosponsorship, speeches on the floor, roll call voting, committee assignments, leadership positions, and a range of activities in the district to determine how legislators represent racial interests [Canon 1999a].
To return to the questions raised in the dissent: the vast literature on representation makes it clear that the “apparent sentiments of incumbents who might run in the new districts” or an incumbent who “had previously promised to consider minority interests before voting on legislative measures” would not meet the standard of substantive representation outlined by the Supreme Court. Substantive representation means taking concrete action (engaging in committee work to write legislation, building coalitions, sponsoring and cosponsoring legislation, and engaging in constituency service) and staking out specific positions (on roll call votes and in speeches on the floor) in response to constituents’ needs and preferences rather than merely exhibiting an “apparent sentiment” or “previous promise.”
While I believe that it is possible to provide the type of evidence that would be required by this new totality of circumstances test, depending upon what measure of “influence” is used, one aspect of the test raises serious “real world” question of application. Given the large variation in the responsiveness of politicians who are elected in influence districts, one cannot know until well after the fact whether a given representative will be “sympathetic to the interests of minority voters,” as required by the new test. For example, if the Georgia v. Ashcroft misinterpretation of Section 5 is not corrected in the VRA renewal, a state may decide that minority interests would be better served by having three influence district instead of one ability-to-elect district and two districts with less than 25% African American voting-age-population. Under the new retrogression standard, this tradeoff could be upheld as non-retrogressive if a subjective determination is made that minority interests would be better served by this arrangement. However, this claim, of course, is simply a prediction of how the newly elected politicians are likely to behave based on previous patterns of behavior in similar districts. It is quite possible that the representatives elected in the new influence districts would be completely non-responsive to minority interests. At that point, minority voters probably would have no recourse: it is extremely unlikely that the federal courts would be willing to redraw the district lines in the middle of a reapportionment cycle based on the evidence of non-responsiveness and the affected voters would no longer have the ability to elect candidates of their choice given that they would comprise a relatively small proportion of the district. This reason alone is enough to restore the old standard of retrogression. Under the new test it is extremely likely that a plan may be accepted as non-retrogressive, but then prove to be harmful to the interests of minority voters in subsequent years.
Finally, if the “totality of circumstances” test for retrogression is allowed to stand, this would ensure that courts would have to make the political judgment of how much “influence” is enough. It is extraordinarily undemocratic for the least representative branch, the judiciary, to make such fundamental political decisions that will directly affect the ability of minority voters to participate in the political process. Furthermore, the courts are not particularly well suited to make such inherently political decisions. That is precisely why federal courts have been so reluctant to do so, even in cases with compelling evidence of non-responsiveness.[11]
Returning to the point of the non-responsiveness of white representatives elected in influence districts, rather than simply relying on hypothetical examples, it would be useful to examine evidence from the state of Georgia that I collected to be used in the remand of Georgia v. Ashcroft. To determine whether state senators sponsored legislation to represent the interests of their black constituents, I coded all 1,509 bills that were sponsored in the Georgia senate between January 1, 1999, and January, 13, 2004, according to whether the bill’s primary impact was on African American constituents, or what I refer to as having “racial” content in one of four possible categories: non-racial, part-racial, racial (and in support of minority interests), and a few bills that were racial but adverse to the interests of African American constituents.
“Racial issues” include those that explicitly refer to race or policies that specifically deal with issues concerning race and civil rights such as discrimination, affirmative action, and racial profiling. Examples from the 2003 legislative session would include SB 115 on Minority Business Enterprise programs and the various votes on the state flag (the underlying bill for the flag votes was actually a House Bill, HB 380, but there were 20 roll calls in the Senate on the issue). “Part racial issues” are not explicitly racial but are implicitly racial, often involving a subtext of race. These are issues that have a disproportionate impact on the black community and are of central concern to black constituents. Crime, welfare, education programs for the disadvantaged, predatory lending, and issues concerning economic development in Atlanta are some of the general topics. Specific examples from the 2003 session include SB 157 on payday lending, SB 309 on street gangs and graffiti, and 348 on tire scrap and disposal (which disproportionately affects minority communities). A vast majority of issues in the state senate are “non-racial” in their content: highways, taxes, issues concerning municipal elections and judgeships, hunting, veterans issues, and most health and education issues (except those that are targeted for the disadvantaged). Recent specific examples include SB 317 and 328 on deer hunting, SB228 on special license plates for pro sports teams’ foundations, SB 459 the “War on Terrorism Act of 2002,” and my favorite, SB 461 on “hunting marsh hens from boats powered by electric motors.” I cross-checked my list with the subset of roll calls identified in the Georgia Legislative Review, a publication produced by the Clark Atlanta University that identifies issues of great interest to minority and poor constituents in Georgia in each legislative session.
Most of the bills were easy to code from the title of the bill and the abstract. About 5% of the bills required reading the actual text of the bill and in a few cases doing additional research to determine the racial focus of the bill (from newspaper accounts or other sources such as the Georgia Legislative Review). The data were gathered from the Georgia State Senate’s web page. The web page lists the full text of all bills and the sponsor and first four co-sponsors of the bill (more recent sessions list five cosponsors; to ensure comparability of the data across legislative sessions I only analyze the first five names listed even if six are provided). While the process of determining the level of racial representation in influence and ability-to-elect districts was not especially difficult in this instance, it was very time-consuming and somewhat tedious! In some cases, it can be expensive and extremely difficult to prove lack of responsiveness.[12] That is why in the context of Section 2 cases, it is just one factor to be considered, and is not determinative of a claim.[13]
My research revealed that there is a huge gap between the proportion of bills sponsored and cosponsored by whites and blacks in terms of their racial content. African American Democrats average about 40% of their bills with some racial content, Republicans average are 3%, while white Democrats have a much broader range from a little over 5% to about 19% in the different sessions (for an average of about 12%). The number of white Democrats needed to equal the legislative output of an African American senator ranges from two (in the 2003-2004 session) to about six in the 1999-2000 session. The similar ratios for Republicans are in eight to nine range. Neither Democrats nor Republicans become increasingly responsive to black interests as the percentage of BVAP in their district increases. This lack of responsiveness is precisely the type of evidence that addresses the totality of circumstances analysis of retrogression. Based on this evidence white senators in districts with at least 25% BVAP are not adequately responsive to black interests to compensate for the loss of representation of an ability-to-elect district. Many of the state senators in influence districts sponsored little or no legislation with racial content.
A similar pattern is evident in an analysis of resolutions that were submitted in the Georgia State Senate. To identify the resolutions that were in recognition of the black community I did a full-text search of all senate resolutions from January, 1999-January, 2004 (2,247 resolutions total). Some examples of these resolutions from the 2003 session include SR 114 – recognizing African American Business Enterprise Day, SR 201 – honoring the Alpha Kappa Alpha Sorority, and SR 99 which tribute to Horace King, a “master covered bridge builder” and former slave. These resolutions tend to have African American sponsors and co-sponsors, with a few white Democrats on many of the resolutions. There is a huge range in the number of Senate resolutions sponsored by each senator that have a racial focus; for some African American members, nearly a third of their sponsored resolutions have a racial focus. At the other extreme 18 Republicans and 12 white Democrats did not have any resolutions with racial content. Jack Hill (D-4 in 2001-2002), one of the party switchers that gave control of the state senate to Republicans, had 173 resolutions, none of which had any racial content, despite representing a district that was over one-quarter black. While most resolutions are of purely symbolic value, sometimes symbolic politics can have important political meaning, as with the heated debate in Georgia over the state flag. Symbolic resolutions of this nature are one significant indicator of a representative’s sympathy and responsiveness to the minority community. By this measure, many of the white politicians in influence districts provided virtually no representation of minority interests. Clearly the threat of non-responsive representatives in influence districts is not just hypothetical.[14]
My research on the Georgia state senate and my larger research project on racial representation in the U.S. House of Representatives [Canon, 1999] both conclude that the interests of African American voters are best served by being represented by African American politicians. This conclusion is in contrast to the views of Carol Swain, who argues that African American interests are better served through maximizing the number of Democrats in office (which means trading off ability-to-elect districts for influence districts). Swain says, “When representation is defined more broadly than shared race, then there is evidence to suggest that political party or, more specifically, whether a Democrat is in office, is as important as the race of the representative. In fact, there is evidence to suggest that this is more important. Given the way minorities define their policy preferences, their substantive interests are best served by the election of more Democrats” [Swain 1997, 321]. My evidence shows that this is not the case when a employing a broad range of measures of representation. This argument will be elaborated below by examining the links between descriptive and substantive representation.
“Descriptive” versus “Substantive Representation:” The second set of questions raised in the dissent, concerning the balance between more influence districts and fewer ability-to-elect districts, also raises serious concerns about the new retrogression test. While the representation literature provides the basis for developing quantifiable estimates of the representation of minority voters’ interests in influence districts, coalitional, and ability-to-elect districts, there is no obvious way to determine how many influence districts are necessary to balance the loss of an ability-to-elect district.
This task is difficult because of the challenges posed by measuring the representational benefits that flow from what the Court calls “descriptive representation.” The benefits of descriptive representation are widely accepted by most people who study this topic, but difficult to precisely measure. To apply this problem to the analytical task presented in the new retrogression test, consider the following: if an analysis of the more easily measured aspects of substantive representation shows that two white Democrats in influence districts provide the same level of substantive representation as one African American legislator in an ability-to-elect district, one could conclude that trading one ability-to-elect district for two influence districts is non-retrogressive only if descriptive representation is seen as having no additional value (either from additional difficult to measure aspects of substantive representation or from the intrinsic value of representation itself). But nearly everyone who has examined this issue agrees that descriptive representation has some intrinsic value and some tangible but difficult to measure aspects.
Even strong critics of black majority districts such as Abigail Thernstrom argue, Whether on a city council, on a county commission, or in the state legislature, blacks inhibit the expression of prejudice, act as spokesmen for black interests, dispense patronage, and often facilitate the discussion of topics (such as black crime) that whites are reluctant to raise. That is, governing bodies function differently when they are racially mixed, particularly where blacks are new to politics and where racially insensitive language and discrimination in the provision of services are long-established political habits [1987, 239].
The problem is how to define the degree of these additional benefits. The benefits of descriptive representation mentioned here by Thernstrom clearly have substantive impact: changing the terms of debate, bringing up issues that would otherwise not be discussed, forcing others in the room to be more inclusive and tolerant are real, tangible effects but they are very difficult to assess with the typical measures of legislative behavior.
Thus the value of descriptive representation that often results from ability-to-elect districts can be broken into three parts: the purely symbolic benefits (having positive role models), substantive benefits that are difficult to measure such as those mentioned above, and substantive benefits that can be measured (roll call votes, leadership positions, committee positions, and sponsored and cosponsored legislation). Of these three components, only the latter can be assessed in any systematic fashion.
The standard literature on racial representation attempts to measure substantive representation by simply focusing on roll call voting (usually summary indices of roll call voting such as LCCR or ADA scores). However, this is clearly an inadequate method for measuring substantive representation. Consider the behavior of two white members, both of whom represent a 35% black district. Both members are relatively moderate on the standard ADA, LCCR-type of measures and occasionally sponsor legislation and make speeches that would be of interest to black constituents. The first member has an all-white staff, locates her district offices in the mostly-white suburbs, and does not mention any legislative activities that would be of primary interest to black constituents in her newsletters. The second member has a racially diverse staff both in Washington and the district, locates one district office in the inner-city and another in the suburbs, and proudly trumpets in his most recent newsletter his legislative activity on “redlining” in the insurance and financial sectors. The standard roll call analysis of congressional behavior would not uncover any difference between these two members. However, the second member is clearly making more of an effort to maintain a biracial coalition and reach out to African American constituents than the first member. Therefore, more comprehensive analysis of congressional behavior is necessary to uncover the nature of racial representation in any given district.
The intrinsic value of descriptive representation is ultimately a value judgement, but given the obvious substantive benefits, it is unwise to sacrifice the tangible gains that have been made in descriptive representation in Congress resulting from ability-to-elect districts for the uncertain gains that may come from having more influence districts. The unequal opportunities minority voters experienced in most states prior to the 1990 round of redistricting – including no African American members of Congress from seven southern states between 1897 and 1973 – proves that point.
I also want to address one critique that has been made by opponents of overturning this part of Georgia v. Ashcroft. Some have argued that by protecting ability-to-elect districts under Section 5, it will create a “one-way rachet” that would prevent the percentage of minority voters in minority-majority districts from ever being shifted to surrounding districts. This means, the argument goes, that districts will become increasingly segregated by race. However, there is nothing in the proposed legislation that would require that result. Ability-to-elect districts are defined by circumstances that prevail in a given area. Areas that do not have racially polarized voting and have substantial cross-over voting and biracial coalitions could move to lower percentages of minority population and still have performing districts. For example, if an area consistently elects black politicians with substantial white support, it could move from ability-to-elect districts that have 55% black voting age population to having 45% black VAP. Such a move would not be retrogressive under the proposed legislation because they would still be ability-to-elect districts. Minority majority districts are not cemented into place under the proposed legislation, but rather it guarantees that the opportunity to elect candidates of choice cannot be taken away from voters who have enjoyed that opportunity in the past.
5. Conclusion
The Voting Rights Act has been one of the most important pieces of civil rights legislation in U.S. history. The VRA should be renewed and strengthened. In my testimony I have argued the following:
— Ability-to-elect districts are necessary to elect significant numbers of minority legislators. — The protection of ability-to-elect districts does not imply that districts are permanently set at a given level of minority voters; the percentage of minority voters can be reduced if they still have an opportunity to elect candidates of their choice. — The pre-clearance provision of Section 5 should be renewed because the number of minority politicians would almost certainly be reduced without this provision. — The lower rate of objections of Section 5 submissions by the Department of Justice should not be seen as evidence that the provision is no longer needed because of the deterrent effect of Section 5 and other reasons that I have described above. — Creating influence districts may not enhance the representation of minority interests because many politicians in influence districts are not responsive to the interests of their minority constituents. — Non-responsive influence districts mean that minority voters would have no recourse to address the retrogression of their voting rights. — Descriptive representation that often results from ability-to-elect districts has inherent value that cannot be measured by the “totality of circumstances” approach outlined in Georgia v. Ashcroft.
Therefore, I recommend that the Senate pass S. 2703 without amendment.
Bibliography
Achen, Christopher H. 1975. “Measuring Representation.” American Journal of Political Science 22: 475-510.
Amer, Mildred L. “Black Member of the United States Congress: 1870-2004,” Congressional Research Service, CRS Report for Congress, July 28, 2004, #RL 30378.
Canon, David T. 1999a. Race, Redistricting, and Representation: The Unintended Consequences of Black-Majority Districts. Chicago: University of Chicago Press.
Canon, David T. 1999b. “Electoral Systems and the Representation of Minority Interests in Legislatures.” Legislative Studies Quarterly 24:3 (August, 1999): 331-85.
Canon, David T., and Charles Stewart III. 2002. “Parties and Hierarchies in Senate Committees, 1789–1946,” in U.S. Senate Exceptionalism. Edited by Bruce I. Oppenheimer. Columbus, OH: Ohio State University Press, 157-81.
Davidson, Chandler. 1993. “The Voting Rights Act: A Brief History.” In Controversies in Minority Voting: The Voting Rights Act in Perspective. Bernard Grofman and Chandler Davidson eds. Washington, D.C.: The Brookings Institution.
Davidson, Chandler and Bernard Grofman, eds. 1994. Quiet Revolution in the South: The Impact of the Voting Rights Act, 1965-1990 (Princeton, N.J.: Princeton University Press).
Fenno, Richard F. 1973. Congressmen in Committees. Boston: Little, Brown and Co.
Fenno, Richard F. 1978. Home Style: House Members in Their Districts. Boston: Little, Brown.
Fenno, Richard F. 2003. Going Home: Black Representatives and Their Constituents (Chicago: University of Chicago Press).
Fiorina, Morris P. 1989. Congress: Keystone of the Washington Establishment. New Haven: Yale University Press, 2nd ed.
Fraga, Luis Ricardo. “More Information Requests and the Deterrent Effect of Section 5 of the Voting Rights Act,” unpublished paper, delivered at the symposium Protecting Democracy: Using Research to Inform the Voting Rights Reauthorization Debate, The Chief Justice Earl Warren Institute on Race, Ethnicity and Diversity, University of California, Berkeley, School of Law and the Institute of Governmental Studies, University of California, Berkeley, Washington, DC, February 9, 2006.
Hall, Richard L. 1996. Participation in Congress. New Haven, CT: Yale University Press.
King, Gary. 1989. Unifying Political Methodology: The Likelihood Theory of Statistical Inference. Cambridge: Cambridge University Press.
Kessler, Daniel and Keith Krehbiel. 1996. “Dynamics of Cosponsorship,” American Political Science Review 90: 555-566.
Lublin, David Ian. 1997. The Paradox of Representation: Racial Gerrymander and Minority Interests in Congress. Princeton, N.J.: Princeton University Press.
McClain, Paula D., and Garcia, John A. 1993. “Expanding Disciplinary Boundaries: Black, Latino, and Racial Minority Groups in Political Science.” In Political Science: The State of the Discipline II. Edited by Ada W. Finifter. Washington, D.C.: American Political Science Association, 247-79.
McCrary, Peyton. “How the Voting Rights Act Works: Implementation of a Civil Rights Policy: 1965-2005, South Carolina Law Review 57:4 (Summer, 2006): 785-825.
McKee, Seth C. “The Impact of Congressional Redistricting in the 1990s on Minority Representation, Party Competition, and Legislative Responsiveness.” Paper presented at the Annual Meeting of the Southern Political Science Association Meeting, New Orleans, January, 2004.
Matthews, Donald R. 1984. Legislative Recruitment and Legislative Careers.” Legislative Studies Quarterly, 9:4, 547-585.
Mayhew, David R. 1974. Congress: The Electoral Connection. New Haven: Yale University Press.
Miller, Warren E., and Stokes, Donald E. 1963. “Constituency Influence in Congress. American Political Science Review 57: 45-56.
Rohde, David W. 1991. Parties and Leaders in the Postreform House. Chicago: University of Chicago Press.
Swain, Carol M. 1993. Black Faces, Black Interests: The Representation of African Americans in Congress. Cambridge, Mass.: Harvard University Press.
Swain, Carol M. 1997. “Not ‘Wrongful’ by any Means: The Court’s Decisions in the Redistricting Cases.” Houston Law Review 34:315-21.
Thernstrom, Abigail. 1987. Whose Votes Count? Affirmative Action and Minority Voting Rights. Cambridge, MA: Harvard University Press.
Valelly, Richard M. 1995. “National Parties and Racial Disenfranchisment.” In Classifying by Race. Paul Peterson, ed. Princeton, N.J.: Princeton University Press, 188-216.
Whitby, Kenny J. 1997. The Color of Representation: Congressional Behavior and Black Interests. Ann Arbor: University of Michigan Press.
Wilson, Woodrow. 1885. Congressional Government. Cleveland: World Publishing.
[1]Two additional districts, the 1st in Missouri in the 1980s (William Clay) and the 7th in Illinois (George and Cardiss Collins) in the 1970s were barely majority-white if voting-age-population rather than total population is used.
[2]My view that Section 5 of the VRA should be renewed appears to be in direct contrast to Professor Carol Swain’s position. Her web site notes, “With the 40-year-old Voting Rights Act’s key provisions scheduled to expire in 2007, Swain recently told journalist Jonathan Tilove of the Newhouse News Services that ‘Section 5 ought to be allowed to gracefully expire in 2007.’” (http://www.carolmswain.net/news.html, the link “Professor Carol Swain Comments on Voting Rights Act in Newhouse News Article”). The on-line version of the article is Jonathan Tilove, “Voting Rights Act, at 40, Faces Reauthorization Amid Topsy-Turvy Politics,” August 5, 2005, http://www.newhousenews.com/archive/tilove080405.html.
[3]For purposes of my testimony, I will not be discussing areas covered for minority language groups under Section 4(f)(4) of the Voting Rights Act. Instead, I will focus my testimony on African Americans in the South. I would like to note, however, that in my experience, much of my analysis of African Americans is equally applicable to Latinos.
[4]The years in which no African Americans were elected to Congress for the seven covered states are: Alabama (1877-1993), Georgia (1876-1987), Louisiana (1877-1991), Mississippi (1883-1987), South Carolina (1897-1993), Texas (first African American elected in 1973), Virginia (1891-1993; 11 political subdivisions have subsequently “bailed out” from Section 5 coverage, so the entire state of Virginia is no longer covered). Other southern states that are partially covered by Section 5 have a similar record (Florida had a gap from 1876-1993 and North Carolina from 1901-1993. Data are from Amer, 2004.
[5]Additional evidence of the difficulty of electing black in majority-white districts in the South is provided by state legislative races where only one percent of all white-majority districts elected black state legislators in the 1980s. On the other hand, 77% of black majority districts elected blacks to the lower house and 62% of black majority districts sent blacks representatives to the upper house in the 1980s [Handley and Grofman 1994, 345].
[6]Luis Ricardo Fraga, “More Information Requests and the Deterrent Effect of Section 5 of the Voting Rights Act,” unpublished paper, delivered at the symposium Protecting Democracy: Using Research to Inform the Voting Rights Reauthorization Debate, The Chief Justice Earl Warren Institute on Race, Ethnicity and Diversity, University of California, Berkeley, School of Law (Boalt Hall), and the Institute of Governmental Studies, University of California, Berkeley, Washington, DC, February 9, 2006.
[7]Peyton McCrary, “How the Voting Rights Act Works: Implementation of a Civil Rights Policy: 1965-2005, South Carolina Law Review 57:4 (Summer, 2006): 785-825.
[8]The new retrogression analysis mentions three types of districts: ability-to-elect, coalitional, and influence. Ability-to-elect districts are those in which minority candidates of choice are usually elected. Coalitional districts provide a lesser opportunity to elect minority legislators and may or may not elect them depending on the levels of racially polarized voting and turnout (most crucially, the level of white crossover voting for minority preferred candidates). Influence districts are those in which minority candidates do not win, but minority voters are said to be able to play a significant role in electing candidates who will be sympathetic to their interests. The precise lines of division among them depends on the relative levels of turnout and racially polarized voting between white and minority voters.
[9]Miller and Stokes, 1963; Achen, 1975; a review essay by Donald Matthews [1984] on this subject that was written twenty years ago had twenty pages of references and research on the topic has proliferated since then. Several review essays that focus explicitly on racial representation include hundreds of references on the topic [see McClain and Garcia, 1993, Canon 1999b, McKee 2004].
[10]Fiorina also discusses the negative consequences of this behavior: the desire to provide constituency service (because it is a relatively costless way to win votes) means that legislators will tolerate or even encourage inefficiencies in the bureaucracy to provide more demand for their services.
[11]For example, in Presley v. Etowah County Commission, 502 U.S. 491 (1992), the Supreme Court rejected a claim that the powers of elected officials were changed to minimize the power of a newly elected black official because there was no “workable standard” that the court could use to draw “lines between those governmental decisions that involve voting and those that do not.” Other courts have also been reluctant to allow responsiveness claims for similar reasons. See Smith v. Winter, 717 F.2d 191, 198 (5th Cir. 1983); Kardules v. City of Columbus, 95 F.3d 1335, 1358 (6th Cir. 1996) (Batchelder, J., concurring).
[12]See Lee County Branch of NAACP v. City of Opelika, 748 F.2d 1473, 1480 n.11 (11th Cir. 1984) (noting the plaintiffs estimated that “80 percent of their time spent in developing and trying this case originally was devoted to the issue of responsiveness”).
[13] See generally S. Rep. No. 417, 97th Cong., 2d Sess. 29 (1982), reprinted in 1982 U.S.C.C.A.N. at 207 (one of the factors that may be considered in assessing a vote dilution claim is “whether there is a significant lack of responsiveness on the part of elected officials to the particularized needs of the members of the minority group”); Gingles, 478 U.S. at 36-37 (citing the Senate Judiciary Committee Majority Report factors for violation of § 2). The Senate Report also provides that “[u]nresponsiveness is not an essential part of a plaintiff’s case” because of the difficulty of proving it. S. Rep. No. 417, at 29 n.116, reprinted in 1982 U.S.C.C.A.N. at 207 n.116.
[14]I should note that this comprehensive analysis of bill sponsorship and cosponsorship does not capture the full range of legislative behavior that would be necessary to fully describe representation. My book on race and representation [Canon 1999] also examined roll call voting, the racial composition of the member’s staff, speeches on the floor of the House, legislative leadership positions, committee assignments, constituency newsletters, district office location, and coverage of the member’s activities in the local press. This research took nearly two years to complete, and clearly is beyond the scope of what would be possible to conduct for any given court case.