Sudler Family Professor of Law
New York University School of Law
May 16, 2006
Testimony of Professor Richard H. Pildes
Sudler Family Professor of Constitutional Law, NYU School of Law
Visiting Professor of Law, Harvard Law School
On The Continuing Need for Section 5 Pre-Clearance
Senate Judiciary Committee
May 16, 2006
I consider it an enormous honor and responsibility to be called to testify on the Voting Rights
Act ("VRA"), a statute I have called a "sacred symbol of American democracy." I also consider it
a painful moment, because I have been asked to testify on concerns I have raised in my scholarship
about adapting Section 5 to the circumstances of the present, circumstances that differ dramatically
not just from 1965, but from 1982 as well, when Congress last re-visited the Act. These concerns
have put me at odds with some members of the civil rights community, including my close friend
and co-author, Professor Karlan, who is also testifying today.
I will limit myself to two particular concerns. Both address tensions between the
fundamental philosophy of Section 5 and the nature of voting rights problems today. First, the bill
proposes to overrule the Supreme Court's recent decision in the redistricting case, Georgia v.
Ashcroft.1 I consider this a mistake, one that will harm the long-term interests of minority voters,
frustrate the formation of interracial political coalitions in the South, and be damaging to American
Let me remind you of the context of that case, which also illustrates powerfully and
concretely the changes between 1982 and today. At the time of the 2000 round of redistricting,
about 20% of Georgia's state legislators were black, and with their virtually unanimous support, a
coalition of white and black Democratic legislators sought to unpack slightly three "safe minority
districts" that the Act had been thought to require in the 1990s.2 With the rise of robust two-party
competition in the South - a dramatic contrast from 1982, whose ramifications for the VRA should
not be underestimated - this coalition of white and black Democrats agreed to maximize the
possibility that Democrats would maintain control of one of Georgia's central political institutions,
the state senate. Black legislators agreed to put a few of their seats modestly more at risk for the
purpose of having a Democratic senate in which black legislators, through control of committee
chairs and the like, would be more effective in serving their constituents' interests. Not a single
Republican voted in favor of this redistricting plan. Yet the Department of Justice (DOJ) and a
lower court divided 2-1 believed that Section 5 precluded a black-white coalition of legislators from
slightly unpacking three districts, and putting their incumbents marginally more at risk, for the
purpose of building an effective, winning political coalition.
Had the Supreme Court not rejected this view, this rigid understanding of the VRA would
have completely inverted the Act's policies. Here were black and white legislators, willing to make
their seats more dependent upon interracial voting coalitions. Yet the Act would have imposed on
them more racially homogenous constituencies. Here was a large contingent of black legislators
who, having entered the halls of legislative power, determined that they and their constituents would
have more effective power as part of a Democratic senate. Yet the Act would have required them
to become the minority in all state representative institutions, for the sake of a marginal potential
gain, at best, in formal black representation in the senate. Here was Congressman John Lewis, his
life risked in the Selma march to help get the VRA enacted, his seat not at stake, testifying after
nearly twenty years in Congress that "giving real power to black voters comes from the kind of
redistricting efforts the State of Georgia has made" and that the South has "come a great distance"
since a generation ago. And here were black legislators, not demanding safer sinecures for
themselves, as officeholders typically do, but taking risks, cutting deals, and exercising political
agency to forge a winning coalition. Yet the Act would have denied these political actors the
autonomy to make the hard choices at issue, even with partisan control of state government at stake.
The Court's decision permitting this deal instead recognizes room in the statewide
redistricting context for modest flexibility in Section 5, given the changes between today's
circumstances and those in the 1970s and early 1980s. Indeed, Georgia's plan involved a modest
amount of flexibility in circumstances about as compelling as I can envision. If Congress overturns
Georgia v. Ashcroft, it will make even this limited amount of flexibility illegal.
Some who advocate overturning Georgia agree that the Court's decision was right on those
facts, but worry that the decision has introduced a vagueness - a more anxious word for flexibility
- of unknown and potentially worrisome scope. To those worries, however, I would say two things.
First, the Court's decision dates to only 2003. We simply do not know how DOJ and the courts will
apply the principles and standards of the case. As far as I am aware, there is not a single court
decision that has relied on Georgia to preclear a plan, nor has DOJ withdrawn any pre-Georgia
objection in light of the Georgia decision. I would not rush to overrule a decision that is right on
the facts and whose future application is unknown. At this stage, we have only abstract speculation
to invoke. But such speculation is just as possible in both directions. Congress adopts the standard
- no "diminishment in ability to elect" is legal, no matter what the context, reasons, or process
involved - then The anti-Georgia standard proposed in the bill might, for example, lead DOJ and
the courts rigidly to lock the covered states into not reducing minority populations in districts to any
extent at all if doing so , even when coalitions of black and white legislators believe black voters
will be more effectively represented by some tradeoffs that enable more powerful white-black
coalitions. Indeed, Rep. Robert Scott of Virginia expressed exactly this concern in the House
3 Voting Rights Act: The Continuing Need for Section 5, Hearing Before the Subcommittee on the Constitution of
the Committee on the Judiciary, House of Representatives, 109th Cong., First Session, October 25, 2005, Serial No.
109-75, at 92-93:
Rep. Scott: There are a lot of areas where you may, for political reasons of effective participation
in the Government and the City Council, whatever, may want to reduce the percentage from a 70,
say, to a 55, in order to create a more accommodating council, and unless you count the influence
districts, you're stuck. If you go from 70 to 55 but create a good council where you might actually
be able to take over, you don't want to foreclose that as a possibility, ever; and if you don't
consider the totality of the circumstances, how do you do that?
hearings if Georgia v. Ashcroft were overruled.3
Second, redistricting plans in covered jurisdictions remain subject to Section 2. A
redistricting plan that involves impermissible vote dilution will be illegal, in covered as well as noncovered
states. To the extent there is concern about the flexibility Georgia provides for Section 5
review of redistricting, Section 2 remains a safety net. There is no shortage of litigation challenging
districting plans in the non-covered states. Nor is it clear why a districting plan that does not involve
illegal vote dilution under Section 2 should be impermissible in a covered jurisdiction. It is
premature to know whether the courts and DOJ, in applying Georgia, will make sound case-by-case
decisions about the contexts in which Section 5 permits some degree of flexibility, and if so, how
much. At this stage, some degree of experimentation with how best to apply Section 5 to statewide
redistricting, overseen by DOJ and the federal courts, is desirable. In this context, I would also
encourage a shorter renewal term for Section 5. A shorter term would enable Congress and others
to judge how courts and the DOJ actually apply Georgia in specific contexts. Locking in any system
for 25 years, in the midst of so much demographic change and other uncertainties, is problematic,
particularly with respect to issues like redistricting.
I said I had two major concerns with the proposed bill. The rejection of Georgia is the first.
The second is a fundamental constitutional and policy concern regarding whether the evidence in
the record is sufficient to justify re-authorizing Section 5 in its current form, as the bill proposes to
do. I am not aware that this particular concern has been addressed in any detail in the hearings here
or in the House. Yet this evidentiary concern affects both sound policymaking and, perhaps, the
constitutionality of a renewed Section 5.
The assumption thus far seems to be that it is sufficient to identify continuing problems in
the covered jurisdictions, such as racially polarized voting, in complete isolation from consideration
of whether similar problems exist in non-covered sites. But Section 5 is a unique law precisely
because it singles out particular areas for a form of federal receivership not imposed on other areas.
The relevant evidence, it seems, should therefore be tied, to some extent, to the Act's pattern of
selective geographic targeting. It is one thing to base uniform national law on evidence from less
than all states. It is another to base a geographically selective national law on a record that does not
account for why some areas are covered, others not. The Supreme Court has been receptive to the
4The Court upheld the constitutionality of the Family and Medical Leave Act of 1993 on this ground. Nevada Dept.
of Human Resources v. Hibbs, 538 U.S. 731 (2003). Similarly, the Court upheld the nationwide ban on literacy
tests, which Congress added to the VRA in 1970, on this ground. Oregon v. Mitchell, 400 U.S. 112 (1970). The
Court has recognized a national interest as such in nationally uniform legislation. See, e.g., id., 400 U.S. at 283
(Stewart, J., concurring in part and dissenting in part, joined by Burger, C.J., and Blackmun, J.) ("Nationwide
application reduces the danger that federal intervention will be perceived as unreasonable discrimination against
particular States or particular regions of the country."). The Court has not recognized a national interest as such in
national laws that are geographically selective, nor is it evident what such a national interest would be, absent
problems arising differentially in different areas of the country.
5See, e.g., City of Boerne v. Flores, 521 U.S. 507 (1997).
6See LAWYER'S COMM. FOR CIVIL RIGHTS UNDER LAW, NAT'L COMM'N ON THE VOTING RIGHTS ACT, PROTECTING
MINORITY VOTERS: THE VOTING RIGHTS ACT AT WORK, 1982-2005 (Feb. 2006) ("LAWYER'S COMM. REPORT").
7Id. at 95 and n.308. These cases come from 16 states, including Colorado, Maryland, Massachusetts, Montana,
Ohio, and Tennessee.
8Id. at 95-96.
9Id. at 90.
former.4 It is not clear how receptive the Court would be to the latter.5
Yet little evidence in the record examines whether systematic differences exist between the
currently covered and non-covered jurisdictions. Indeed, the evidence that does exist suggests the
opposite: the problems identified, such as racially polarized voting, are similar in many places
throughout the country where sizable minority populations exist. Given that Supreme Court
doctrine now requires that congressional remedies in this area be "congruent and proportional" to
identified violations, the absence of such evidence raises concerns about whether the Court will be
able to find that the proposed bill meets these requirements - particularly the requirement that
coverage be "congruent" to the violations. This concern also raises larger questions about whether
the underlying philosophy of Section 5 continues to provide the model for national legislation
protecting the right to vote that is best suited to the problems of today, and about what an alternative
model of federal voting rights legislation might look like instead.
Let me first be concrete about the evidentiary concern. I believe I have read all the major
studies referenced in the House and Senate hearings to date. I will rely here only on studies offered
in support of renewal. The most comprehensive of these is, perhaps, the report of The National
Commission on the Voting Rights Act.6 Of the three chapters devoted to marshaling the evidence
in support of renewal, one relies on judicial findings of continuing racially polarized voting. Yet
the report itself notes that these findings are similar in court cases throughout the country. Of the
23 cases involving statewide redistricting plans since 1982 that have found racially polarized voting,
half came from covered jurisdictions - and half from non-covered ones.7 The report quotes judicial
language in cases from South Carolina and Louisiana - but also virtually identical language from
Maryland, Massachusetts, and Florida.8 Similarly, an experienced voting rights lawyer testified at
one hearing that "there are politically significant statistical levels of racial polarization between
Anglos and Latinos, as between whites and blacks, in almost every locale which I have
experienced."9 This testimony is offered to justify renewal of Section 5, yet he was testifying about
cases in Texas and Maryland - the former covered, the latter not.
10 Ellen Katz and the Voting Rights Initiative, Documenting Discrimination in Voting Under Section 2 of the Voting
Rights Act, Voting Rights Initiative Database (2005), www.votingreport.org. ["Michigan Study"].
11For example, the Michigan study identifies 209 lawsuits that ended in a liability determination under Section 2; of
these, 53.1% came from non-covered jurisdictions. Id. at 8. Similarly, there were 117 published decisions involving
successful Section 2 lawsuits since 1982. Of these, 67 were in covered jurisdictions, 50 in non-covered ones. Id.
The study identifies 24 reported Section 2 cases since 1982 in which courts found intentionally discriminatory voting
conduct; of these, 13 cases were in non-covered jurisdictions, 11 in covered ones. Id. at 21.
12 Of the 322 reported cases since 1982 in which a Section 2 claim was resolved, the "great majority" challenged
vote dilution, including 108 of 117 of the cases that were successful. See LAWYER'S COMM. REPORT, supra note 6,
at 82-83 (data compiled in part from findings reported there).
13 Michigan Study, supra note 10, Report Addition: List of Locations Nationwide Where Courts Found Section 2
Was Violated (Feb. 24, 2006), http://sitemaker.umich.edu/votingrights/files/violationlocations.pdfat. This is not to
say that there are not major differences between these states of relevance to the VRA; I certainly believe there are. It
is only to say that, on the facts of this study, judicial findings of Section 2 violations do not distinguish between
these particular states. The Michigan Study reports on only published judicial decisions. I am not aware of a study
that provides similar comparative information of covered and non-covered jurisdictions for unpublished judicial
decisions or settlements of Section 2 claims.
Another often-cited study examines all published cases since 1982 in which courts have
found violations of the nationwide ban in Section 2.10 Yet, once again, these violations are not
overwhelmingly or systematically concentrated in Section 5 areas; this report itself documents that
these violations arise in many places with significant minority populations.11 There is a deep reason
these geographic similarities emerge in contemporary studies, a reason tied to the nature of votingrights
problems today. The type of problem central to the VRA today is different than in the past.
Earlier, the primary issue was exclusion of minority voters from the polls. While some problems
of intimidation and harassment unfortunately remain, today the vast majority of VRA cases and
violations instead involve vote dilution.12 And to the extent vote dilution is the issue of this era, that
issue arises in many (perhaps most) areas with significant minority populations. It is not
concentrated in any one discrete part of the country. Since 1990, for example, there are as many
judicial findings of Section 2 violations in Pennsylvania as in South Carolina - and more in New
Congress could adapt Section 5 in either of two directions to reflect this reality. It could
expand Section 5 to reach all areas with findings of vote dilution or racially polarized voting. Or,
Congress could more narrowly target Section 5 to areas that, in addition to such findings, evidence
additional voting-rights problems that justify singling out those areas. I have no doubt there are
some areas in which voting-rights problems are uniquely concentrated, particularly if a renewed
Section 5 focuses on the county level. But simply re-adopting a triggering formula that is based on
a state's use of a literacy test in 1964, and voter registration and turnout below 50% in 1972, or
similar trigger criteria, leaves a renewed Section 5 vulnerable to constitutional challenge.
These evidentiary issues are not just legalistic concerns. They point to more profound
questions about what the fundamental basis ought to be for national policy going forward on voting
rights. The most significant legislative initiatives to bring national consistency and uniformity to
American elections - since the short-lived post-Civil War era of Reconstruction - have been the
1965 Voting Rights Act (VRA) and its amendments, the 1993 National Voter Registration Act
14 See Daniel P. Tokaji, Early Returns on Election Reform: Discretion, Disenfranchisement, and the Help America
Vote Act, 73 GEO .WASH. L. REV. 1206, 1220-39 (2005).
(NVRA), and, in response to the 2000 election, the 2002 Help America Vote Act (HAVA). Though
not widely appreciated, however, these statutes embody radically different philosophies regarding
when and why national oversight of elections is needed.
The original VRA, enacted in 1965, has never protected the right to vote as such. Instead,
it protects voting rights in two more selective and narrowly targeted ways: it targets selected regions,
through Section 5, and it targets racially-discriminatory barriers to voting, rather than unjustified
barriers of any sort to voting. By contrast, HAVA and the NVRA are not selectively targeted in
either sense: these more recent statutes provide uniform, national protection for the right to register
and cast a legal vote as such. These statutes protect all citizens in all parts of the country. The
profound question for policymaking today is whether the model of the VRA, or the alternative model
represented by HAVA and the NVRA, should provide the blueprint for the future of national
legislation to protect the right to vote. Although I do not expect that question to be a major focus
of these hearings, I believe it essential to recognize this issue and to put the more specific questions
concerning Section 5 in that broader perspective.
In singling out certain areas for unique federal oversight, Section 5 of the VRA rests on the
philosophy that national policy can identify, in advance, areas of the country in which voting rights
problems (that is, minority voting rights problems) are considerably more likely to arise
systematically. In addition, Section 5 locates the threat in changes to existing voting rules and
practices; it is only these changes, rather than the status quo baseline practices, that the federal
government must, in certain areas, preclear. In earlier periods, these narrow targeting features were
exceedingly easy to apply: the Act was aimed centrally at the states of the Old Confederacy, which
had systematically denied black citizens (and poor whites) the vote for decades, in part through
changing voting rules and practices to frustrate federal oversight. Even in 1982, blacks were still
virtually invisible in elective offices; the South remained, for state and local elections, the virtual
one-party monopoly it had been throughout the 20th century; and the focus of the 1982 process was
eliminating at-large and multi-member election structures that effectively excluded blacks from
elective office. Constitutional doctrine, recognizing these facts, was also more hospitable to the
geographic singling out Section 5 entails.
But consider the kinds of voting issues we face today. As noted above, not only is it that,
with the end of formal political exclusion, these issues are not as obviously unique or confined to
any particular region. In addition, it is not easy for national legislation to predict in advance of
actual elections where such problems are likely to emerge in coming years. A statutory model based
on identifying in advance where those problems are likely to arise is increasingly difficult to adapt.
Indeed, for similar reasons, such a model will underenforce minority voting rights. In the 2004
presidential election, for example, the most significant voting rights issues arose in the battleground
state of Ohio.14 Yet in 1982, when Congress last legislated, there would have been no way to
anticipate that Ohio would be the place where the major voting rights controversies of the 2004
15The five covered Florida counties are Collier, Hardee, Hendry, Hillsborough, and Monroe. See Section 5 Covered
Jurisdictions, http://www.usdoj.gov/crt/voting/sec_5/covered.htm (last visited April 26, 2006). The 2000 election
disputes centered on Palm Beach, Miami-Dade, Broward, Volusia, and Nassau counties, none covered by Section 5.
See Bush v. Gore, 531 U.S. 98, 101 (2000).
16 For details of these election disputes, see SAMUEL ISSACHAROFF, PAMELA S. KARLAN, AND RICHARD H. PILDES,
THE LAW OF DEMOCRACY: LEGAL STRUCTURE OF THE POLITICAL PROCESS 199-205 (rev. 2d ed. Supp. 2005).
17 For excellent general discussion of these issues, see Daniel P. Tokaji, The New Vote Denial: Where Election
Reform Meets the Voting Rights Act, 57 S.C. L. REV. 689 (2006).
presidential election would emerge. Similarly, in the 2000 presidential election, controversy
centered on Florida, most of which, like Ohio, Section 5 does not reach. To some extent, the VRA
did anticipate Florida as a potential problem area, based on its past history, for Section 5 does reach
five counties in Florida.15 Nonetheless Section 5 was of no relevance during the 2000 post-election
legal disputes: none of the covered counties included those that spawned the major conflicts in 2000.
The same structure of the problem arises if we look more generally at the 2004 election cycle: the
most intense post-election litigation over state and local elections took place over the governorships
of Washington and Puerto Rico and the mayor's office in San Diego16 - none of which the VRA's
geographic-targeting approach reaches.
Ohio, Florida, Washington, Puerto Rico, and San Diego do, however, share one feature - all
had exceptionally competitive elections and small margins of victory. This reveals part of the
difficulty with trying to tailor modern voting-rights protection to specific areas picked out by federal
law in advance: the incentive to manipulate voting rights will be greatest today where elections are
extremely competitive and close. Similarly, complaints and perceptions of large-scale deprivations
of voting rights, including minority voting rights, are most likely to emerge in such elections. But
there is little way to base national regulation on ex ante predictions regarding where close contests
for electoral votes, the Senate, the House, or state and local offices are likely to arise over the next
generation. In contrast, when the geographic targeting approach of Section 5 was adopted, distinct
areas existed that systematically, election after election, denied minority voting rights, whether or
not elections were competitive.
Second, as partially noted above, the nature of voting-rights issues today also is less
geographically concentrated in a distinct way than in the past. Consider the kinds of problems that
have received the greatest attention in recent years. These include concerns about voting
technology; lack of clear standards for what counts as a valid vote; ballot-design confusions;
corrections to the provisional balloting system established in HAVA; long lines at polling places;
partisan administration of election laws; sheer incompetence in election administration at the
precinct level; burdensome voter-registration requirements, such as the need to re-register upon
moving; and felon-disfranchisement laws.17 These problems arise in many different parts of the
country, sometimes only in some elections. It is difficult to conclude that they systematically and
uniquely arise in particular areas that federal law can accurately pre-identify.
The emerging controversy over whether voter identification requirements should be
tightened up and if so, what forms of identification should be required, illustrates this point. The
18 See Common Cause/Georgia v. Billups, 406 F. Supp. 2d 1326 (N.D. Ga. 2005).
19 A federal district court recently upheld Indiana's new law. See Ind. Democratic Party v. Rokita, No.1:05-cv-
00634-SEB-VVS (S.D. Ind. filed Apr. 14, 2006).
20 CARTER-BAKER COMMISSION, BUILDING CONFIDENCE IN U.S. ELECTIONS: REPORT OF THE COMMISSION ON
FEDERAL ELECTION REFORM 21 (2005).
most visible of these laws so far was enacted in Georgia. The federal district court enjoined that law
as a violation of the fundamental right to vote under the Constitution (after the DOJ had precleared
the law through the Section 5 process).18 To some, that might confirm the need for continuing the
geographical targeting of Section 5. Georgia was one of the states initially designed to be put under
Section 5's federal receivership regime; it remains a covered state today. But voter ID requirements
are being adopted and considered elsewhere, including many non-covered jurisdictions. Indiana,
for example, recently adopted such a law.19 And the bipartisan, Carter-Baker Commission
recommended a national voting ID requirement (over the dissent of some members).20 To be sure,
debates now taking place over voter ID requirements in several state legislatures have an
overwhelming partisan dimension. But there is not an obvious geographic dimension to the issue,
particularly not one that easily correlates with other voting-rights issues to suggest that certain states
or areas are systematically infringing on voting rights.
Third, recall that Section 5 selectively targets only changes in voting rules and practices.
Yet here too, today's problems differ from those that generated this statutory focus on voting
changes. Laws that disfranchise ex-felons and felons, for example, are among the most significant
barriers today to African-American suffrage, in terms of the number of otherwise eligible voters
affected. But these laws are typically not recent enactments, nor do they reflect recent changes in
state law. These laws spawn novel issues today precisely because they were enacted long ago, in
eras of much lower incarceration rates. They have remained unchanged even as their effect,
including their racially-disparate effect, has mushroomed in tandem with convictions. Yet because
these are not recent "changes in state law," they are completely beyond the reach of Section 5. Nor
could such laws be brought within the scope of Section 5 through modest amendments. For recall
that the entire approach of Section 5 is premised on the assumption that federal oversight should be
targeted most aggressively on changes in voting practices and rules. Similarly, when it comes to
problems with voting technology, such as pre-scored punch-card ballots, or partisan election
administration and incompetence, the problem is not recent changes in law or practice. Indeed, the
problem is the opposite: it is preservation of the status quo - the failure to update old voting
technology, the failure to create non-partisan election administration structures, the failure to train
election officials properly - that is the problem. Far from suspicious, change is precisely what
policy should aspire to in these areas.
For these three reasons at least, the narrow targeting model of Section 5 - its effort to single
out particular areas and changes in voting rules - is less well suited to the voting rights problems
of today than was the original Section 5 to its day. Section 5 is narrowly targeted in another way
as well: it singles out minority voting rights for federal protection, as opposed to voting rights per
se. The most general question the VRA renewal process should ask is whether this is the right or
21See United States v. Reese, 92 U.S. 214 (1876); United States v. Cruikshank, 92 U.S. 542 (1876).
22531 U.S. 98 (2000).
23See, e.g., Kramer v. Union Free School District No. 15, 395 U.S. 621 (1969) (local elections); Harper v. Virginia
Bd. of Elections, 383 U.S. 663 (1966) (state elections).
exclusive model - I call it the antidiscrimination model - for the future. Put another way, the broad
question is whether the model of HAVA and the NVRA, which redefine the problem from an
antidiscrimination model to protection of voting rights as such, should represent the future of voting
There are reasons to believe that such a shift would enhance the protection of voting rights,
including minority voting rights. First, it is important to bear in mind that Congress had historically
limited national voting-rights protections to the context of race (and later, certain other minority
groups) not only because the problems were most severe in this domain, but partly because
constitutional understandings and doctrine were thought to limit Congress' power over voting issues
to the prevention of racially discriminatory voting practices.21 But since 1965, it has become clearer
that Congress has constitutional power to directly protect the right to vote itself. Whatever else the
Supreme Court's decision in Bush v. Gore22 does, it further confirms this point: the Constitution
protects the right to vote from being arbitrarily infringed, for any reason at all, whether or not race
is involved. The Court now recognizes the right to vote as a fundamental constitutional right in all
general elections, whether national, state, or local.23 Simply because earlier Congresses believed
themselves constitutionally limited to protecting voting rights in the context of racial discrimination
is not a reason to remain locked into that model today. Indeed, the Supreme Court today might well
be more accepting of (and deferential to) congressional power to protect the right to vote as such
than of congressional power to single out regions of the country or voting practices that
disadvantage minorities, without a discriminatory animus, for unique voting-rights protection.
Second, in the context of modern politics, it is often difficult to attempt to separate racial
considerations from partisan ones when voting rights are at stake. In the voter ID controversies of
the moment, for example, Republican legislators are the initiators of efforts to adopt ID
requirements; Democratic legislators typically resist. Some charge that these requirements are
adopted for racial reasons. But to those who believe these requirements unjustified, are they being
adopted for partisan or racial reasons? And should it matter, assuming we could answer the
question? The same is true of national legislation that might, for example, ban the intimidation of
voters. Would it be better (for minority voters, as well as others) for such legislation to target
intimidation "based on race" or simply illegal intimidation per se?
The VRA model of selectively focusing on racially-discriminatory voting practices requires
courts to determine whether race or partisan politics is the cause (or the predominant cause or,
perhaps, a cause) for the adoption of certain voting practices. But such an inquiry is often
intractable, for courts or other actors. As long as black voters remain overwhelmingly Democratic,
race and partisanship will remain intertwined, perhaps inextricably so. National legislation based
on separating the two elements will always, therefore, be problematic, at the least. This problem can
lead to underprotection of minority voting rights themselves. The more difficult it is for courts to
separate racial from partisan or other considerations, the greater the risk that courts will reject
voting-rights challenges on the ground that partisan considerations, not racial ones, account for the
practice at issue. Put another way, the voting rights of all citizens, including minorities, are most
threatened today by partisan attempts to manipulate election regulation for self-interested reasons.
Isolating one dimension of that threat for national legislation addresses only part of the problem and
might make it more difficult effectively to address even that part. Perhaps paradoxically, the more
general the form of voting-rights protection, the more minority voting rights will be effectively
Thus, there is reason to believe that both the selective targeting features of the VRA model
of voting rights - the singling out of certain places and of certain subcategories of voting rights for
federal oversight and protection - represent the past of voting rights, but not the future. This is not
to say that racially-discriminatory voting practices are no longer a problem. But racially
discriminatory voting practices are a subset of a more sweeping set of challenges to full and fair
political participation in American democracy. The most effective way of providing legal protection
for voting rights, including minority voting rights, might increasingly be less through an antidiscrimination
vision than through a vision focused directly on the substantive right to vote itself.
Perhaps national policy will need to reflect both visions: uniform national voting-rights
protections as well as protections selectively targeted both geographically and on certain groups of
voters. But at the least, voting-rights policy should not remain so embedded within the model of the
past as to preclude looking beyond that model to consider whether different visions, such as those
reflected in the NVRA and HAVA, ought to become more dominant as we move forward.
To sum up, I have two major concerns with the specific bill proposed. First, a rush to
overrule Georgia v. Ashcroft at this stage is, in my view, a mistake. Second, I am concerned that
Supreme Court doctrine will not permit Congress simply to re-authorize Section 5, with exactly the
same scope of coverage as in 1982 (itself based partly on a formula from 1965), unless evidence
adequately shows that problems of race, ethnicity, and voting rights today differ significantly
between those jurisdictions targeted and those regulated only by Section 2 of the Act.
More generally, debate over renewal of Section 5 need not remain locked within the models
of the past. I append to my testimony a forthcoming article that elaborates on that point. I would
suggest that much of the work the Voting Rights Act so powerfully began might today best be taken
up by building on the models of the Help America Vote Act and the National Voter Registration Act
and basing national legislation on protection of the right to vote as such.