FEATURE CONTENTS INTRODUCTION I. THE WEIGHT OF A VOTE II. EOUAL REPRESENTATION III. WHO ARE THE PEOPLE? CONCLUSION
Equality is a powerful idea, and nowhere more so than in the political sphere. It was the power of an idea of political equality that led the Supreme Court to conclude that it had no choice but to enter the "political thicket" (1) and regulate the way district lines of all kinds are drawn across the United States. The Court's sweeping intervention in this sphere has been enduringly popular, in part because of its straightforwardness. Anyone can understand a rule that says each district must contain an equal number of persons.
But equality is rarely so simple. Political equality in particular is a subtle, multilayered idea. In one of the most striking developments of the present redistricting cycle, the equipopulation rule is now under significant fire from litigants who come bearing arguments that are also couched in terms of a conception of political equality. These litigants argue that equality--and therefore, the Constitution--requires districts with equal numbers of eligible voters, not equal numbers of persons.
The difference is enormous. Certain districts, such as a predominantly Hispanic city council district in Irving, Texas, have the same total population as other districts but only half the citizen voting age population (CVAP). (2) A conservative impact litigation firm brought a lawsuit challenging that particular district in 2010, alleging that the difference in CVAP harmed the voters of the other five districts by diluting their votes. (3) The real prize here is much larger than the government of a mid-sized Texas city. Today, line-drawers across the nation rely almost uniformly on total population, an approach that current Supreme Court precedents neither require nor prohibit, (4) but that has become the de facto national policy. (5) A shift from total population to eligible voters or CVAP (6) would shift power markedly at every level, away from cities and neighborhoods with many immigrants and many children and toward the older, whiter, more exclusively native-born areas in which a higher proportion of the total population consists of eligible voters.
Adjudicating this conflict may force courts to confront a deep question they have long avoided. What is one person, one vote really about? What form(s) of political equality does it protect? (7)
Proponents of switching from total population to eligible voters or CVAP have a straightforward answer-and one with an outstanding pedigree. In Reynolds v. Sims, the 1964 case that interpreted the Equal Protection Clause to require the one person, one vote rule, the Court held that "the right of suffrage can be denied by a debasement or dilution of the weight of a citizen's vote just as effectively as by wholly prohibiting the free exercise of the franchise." (8) In Wesberry v. Sanders, (9) which likewise imposed the one person, one vote rule on congressional redistricting on the basis of Article I, Section 2's requirement that the House be elected "by the People of the several States," (10) the Court similarly framed numerical malapportionment as debasing the "weight" (11) or sometimes the "worth" (12) of individual voters' votes. (13) The Court conceptualized the debasement of the weight of a vote as an individual injury, distinct from but analogous to disenfranchisement. This was a clever move. By focusing on individuals, the Court at a pivotal moment sidestepped the objection that it was intervening in an unprecedented way to restructure American democracy. Avoiding the Guaranty Clause, (14) the Court told an individual rights story. (15) Thus, when litigants today claim that one person, one vote protects the weight of individual votes, they are making an argument as old as one person, one vote itself, using language that courts know well and often cite.
The trouble is this: the closer one examines this argument about the weight of individual votes, the thinner and more insubstantial it turns out to be. As I will argue in this short Essay, no coherent account can be reconstructed of a nontrivial, non-tautological individual interest in the "weight" of a vote that one person, one vote protects. The "weight" of an individual vote, as protected by the one person, one vote rule, turns out to he a somewhat mysterious, ephemeral construct. Protecting the "weight" of my vote does not protect my chances of casting the deciding vote in an election. It does not preserve the proportion of the winning coalition, or the proportion of the total votes cast, that my vote constitutes. Fundamentally, it does not protect my chances of being on the winning side and electing my candidate of choice. In short, the "weight" of my vote does not do much work for me, in terms of the various aspects of my individual situation that I might care about as a voter. Nor does it solve these problems to redefine "weight" by reference to eligible voters or CVAP in place of equipopulation. If the "weight" of an individual vote were really all that one person, one vote protected, it is unclear why this jurisprudence would exist.
My claim is that instead, at its heart, one person, one vote is a rule about group power. It is a rule whose purpose and effect are not about the weight of any individual vote but instead concern a different set of questions: which groups of people will be able to elect representatives, and how many representatives will those different groups of people be able to elect? Will a small group of people in an area of rural Tennessee have the same power to elect candidates of their choice as a much larger group of people in Memphis? It is these group questions that led to the creation of the one person, one vote doctrine, and these remain the key questions on which one person, one vote intervenes. The "weight" argument is a makeweight. This conclusion does not resolve the contemporary challenge to the equipopulation rule, but it situates it on dramatically different ground, as I will discuss in the final Section of this Essay.
The argument of this Essay might seem at first blush to be just the opposite of the argument I made in a recent article, Equal Citizenship and the Individual Right to Vote. (16) In that article, I argued that election law scholars and courts have both been too quick to frame all election law questions in structural terms. Vote denial claims, I argued, are different; they are individual rights claims at their core, and should be adjudicated as such. My project here is complementary to--and indeed grew out of--that argument. In this Essay I argue that in one person, one vote cases, in contrast to vote denial cases, the real action is not in the domain of individual rights, but rather in structural questions about the allocation of group political power.
I. THE WEIGHT OF A VOTE
I am just one voter. In what sense could the "weight" of my individual vote be diluted by numerical malapportionment--or protected by the one person, one vote rule? There are a number of possible answers to this question. Yet both courts and scholars tend to devote surprisingly little--if any--attention to specifying what they have in mind. Given the critical role that the "weight" of an individual vote plays in the dominant understanding of one person, one vote, it is worth being more precise. Let us consider six distinct answers.
The most straightforward and venerable answer, although one the Court has not embraced, begins with the idea that the way my vote matters is that it might alter the election outcome. From that perspective, we can conceptualize "weight" as a probability: the weight of my vote is (1) the probability that I might cast the decisive vote. This probability is small. But all other things being equal, it gets smaller as the number of voters gets larger. A significant literature at the intersection of statistics and law analyzes the weight of an individual vote in these terms, calculating what is now known as the Banzhaf index, after the work of John Banzhaf, (17) by assuming that each voter is equally likely in a two-candidate election to vote for one candidate or the other. With that simplifying coin-flip assumption in place, each additional voter in my jurisdiction predictably and measurably decreases the probability that my vote will be decisive. The Court has rejected the Banzhaf index as a measure of vote dilution. (18) But perhaps this way of thinking nonetheless captures a sense in which, as we add more voters, we dilute the probabilistic weight of each vote--and, by the same token, a sense in which the one person, one vote rule protects a vote's weight. (19)
This promising idea runs quickly aground when one begins to think realistically about when it is more or less probable that any individual vote will decide an election. (20) Far more relevant than adding or taking away some voters, the most determinative factor affecting the probabilistic weight of a vote is how often the elections in the particular jurisdiction are close. In a district that nearly always votes 80%-20% in favor of one party, the probability of a close election, let alone an election decided by one vote, is tiny. An individual vote has far more probabilistic weight in a district that is twice or ten times as populous, but also far more competitive, with patterns of results falling in a tight bell curve around the 50%-50% knife-edge. If we could rerun the 2000 presidential election any number of times, each voter would have far greater probabilistic weight in a close swing state like Florida than in safe states of any size--whether more populous California or less populous Alaska--for the simple reason that the outcomes in California and Alaska were not in doubt. (21) If we wished to increase the probabilistic weight of one vote, the best strategy would not be to make a district less populous, but to make it more competitive. (22)
A related but different conception of the "weight" of a vote posits that there is an "efficacious set" of votes--a subset of the votes for the winner that, together, are the ones that added up to victory. (23) We thus might conceptualize "weight" as (2) the probability that my vote is within the efficacious set, and thus not wasted. (24) This approach has the virtue that it does not depend on the possibility of knife-edge outcomes. But it brings us no closer to a solution. Here, too, it is competitiveness rather than district size that will primarily determine the weight of a vote. (25) If we care about equalizing either of these first two conceptions of weight, we ought to forget about one person, one vote and instead focus on equalizing the competitiveness of districts. (26)
We might conceptualize the "weight" of a vote, instead, as (3) one's chances of being on the winning side. This approach has the virtue of capturing something voters care about: voters want their candidate to win. Here, then, the idea is that votes for the winner are in some important sense "weightier" than votes for the loser. Although the Court presumes that elected representatives adequately represent the interests of both the voters who supported them and the voters who opposed them, (27) this presumption is something of a legal fiction. (28) Representatives literally represent both kinds of voters, but they do not represent them in exactly the same way. Some constituents are part of a representative's political constituency, while others are opposition voters--people whom the representative would probably prefer not to represent at all, may not want much to do with, and often would be happy to jettison at redistricting time. (29) Perhaps the real "weight" of a vote is one's chances of being in the former group.
Although conceptually interesting, this way of thinking about weight is wholly unconnected to the one person, one vote standard. There is no logical relationship between the absolute or relative size of a district--measured by total population, CVAP, or any other way--and a given voter's chances of being on the winning side. The simplest method of raising most votes' weight in this sense would be just the inverse of drawing competitive districts: gerrymander the voters into districts packed with like-minded voters. Then, most of the time, most voters will be part of a winning landslide.
A different way of thinking about the weight of one vote might begin with the fact that each voter who is part of the winning coalition makes up some fraction of that coalition's vote total. The "weight" of one vote could mean (4) the size of the fraction of the winning vote total that this one vote represents. Here, the idea is that a voter is somehow just more important, her vote more "weighty" in this fractional sense, if her vote makes up one six-hundredth of the winning coalition, rather than one twelve-hundredth of it. (30)
Except in extreme cases, the one person, one vote rule does not protect a vote's fractional weight either. First, neither equipopulation nor its opponents' proposed equal-CVAP alternative captures the number of people who actually turn out to vote in a given election. (31) Second, even if everyone turned out to vote, this fractional conception of weight is greatly affected by the magnitude of the victory. Holding turnout constant, the fractional weight of a vote for the winner will be considerably smaller in a landslide than in a squeaker.
To be sure, attacking truly massive interdistrict population disparities will almost always make fractional weight less unequal. But that is not a good description of what the modern one person, one vote regime does. As applied today, in either congressional or state contexts, the rule demands a far more precise degree of equipopulation than could possibly be relevant to the goal of equalizing fractional weight. In each cycle, before redistricting commences, districts are typically close enough in size that turnout, competitiveness, and other variables can easily trump size differences in determining fractional weight.
In any case, there is something odd about this fractional conception of the weight of a vote: where does it leave those who play no part in the majority at all, because they are on the losing side? For voters perennially on the losing side-the "filler people" (32)--the weight of their individual votes in this fractional sense appears to be zero, or perhaps negative.
Perhaps all these conceptions of weight have been insufficiently abstract: all of them have retained some idea that election outcomes matter for understanding the weight of votes. One could instead ignore outcomes and define the "weight" of my vote simply as (5) the fraction of the total votes cast that my vote represents. But this is no solution either. A simple way to raise or lower the "weight" of my vote in this sense is to raise or lower turnout. Making a district so lopsided in partisanship that the outcome will be foreordained will tend to lower turnout and thus raise votes' "weight"; making a district highly competitive will tend to raise turnout and thus lower votes' "weight." One could raise votes' "weight" in this sense simply by holding the election on a special day when no others are held, or lower it with a successful public service announcement urging people to vote. In any case, it is not obvious why equal "weight" in this outcome-independent sense would matter to me as an individual voter in any nontrivial way. It does not give me equal power to affect the outcome, an equal chance to be among the majority, or any particular confidence that my ballot made any difference, either by itself or in combination with others' ballots.
One more option, still more abstract, remains. We could define the "weight" of one person's vote as (6) the fraction of the district's total population (or, alternatively, its CVAP) that that person represents, regardless of whether that person, or anyone else, actually votes. This, finally, is a definition of "weight" that one person, one vote clearly protects. But at this point we have abstracted away too much, and are left with nothing more than a tautology. To state that each person must represent the same fraction of the population of their district is to state that each district must contain the same number of persons. (To state that each voting-age citizen must represent the same fraction of her district's CVAP is to state that each district must contain the same CVAP.) It does no conceptual work to rename equipopulation (or equal CVAP) "equal weight." Weight, in this final sense, has nothing to do with votes at all. This last kind of weight is not, in truth, the weight of my vote. Instead it is the weight that the redistricting scheme accords me, as a person (or as a citizen of voting age). (33) Thus we reach a strange and unsatisfying conclusion: one person, one vote protects the weight of an individual vote if and only if "weight" is defined in an entirely circular, tautological way.
The Court's own discussions of the weight of individual votes do not linger long on the meaning of "weight"--for good reason. Nor do they linger long on the topic of the interests of individual voters. When the Court rejected the Banzhaf index in the one person, one vote context in Board of Estimate v. Morris, it held that
a citizen is, without more and without mathematically calculating his power to determine the outcome of an election, shortchanged if he may vote for only one representative when citizens in a neighboring district, of equal population, vote for two; or to put it another way, if he may vote for one representative and the voters in another district half the size also elect one representative. (34)
This formulation purports to be about the "personal" rights of a single citizen, (35) but curiously, in order to describe those rights, it cannot help but speak of citizens in the plural and also of representatives in the plural. The Court is reaching here for a relationship between the "population" or "size" of the district and the number of representatives it can elect. (36) The individual citizen is playing the same rather limited role she plays in a racial vote dilution case: providing a convenient rights-holder with legal standing on whose behalf the Court can implement a group-based conception of equality. (37)
II. EQUAL REPRESENTATION
Justice Powell argued in Davis v. Bandemer that
[t]he concept of "representation" necessarily applies to groups: groups of voters elect representatives, individual voters do not. Gross population disparities violate the mandate of equal representation by denying voters residing in heavily populated districts, as a group, the opportunity to elect the number of representatives to which their voting strength otherwise would entitle them. (38)
The insubstantiality of the weight of an individual vote strongly suggests that Justice Powell was right. Individual voters, to be sure, have vital interests in the election process. Each voter has an interest in being able to cast a ballot and have it counted. Disenfranchisement injures an individual voter by rendering her something less than a full, equal citizen; one need not invoke group interests to understand this injury. (39) But vote dilution claims are different. They are about the interests of groups, because only groups are capable of electing representatives. As Justice Powell suggests, one person, one vote claims in particular are about the interests of numerical groups--that is, collections of people who happen to live within a given geographic area (40)--in securing representation in proportion to their numbers. (41)
Numerical groups are not the kind of groups whose collective interests would ordinarily seem to rise to the level of constitutional significance. In election law, we are used to thinking about the political interests of racial groups, which the Voting Rights Act and Equal Protection Clause protect. We are also used to thinking about the interests of partisan groups (although the stalemate over exactly how to protect partisan group interests in the gerrymandering context has now persisted for a quarter century). (42) Conceptualizing one person, one vote in terms of the interests of numerical groups enables us to see some of the continuities among all three of these forms of vote dilution. (43) But it does not tell us why the Constitution either does or should protect the interests of numerical groups. Partisan groups by definition share a political point of view. The law's solicitude toward racial group interests often turns on whether racial groups are similarly "polarized" in their political views. (44) Numerical groups are different. The layer of constitutional protection for numerical groups operates without reference to the axes of partisanship and polarization.
And yet, other features of our political and constitutional tradition offer some clues as to why our law views numerical groups as constitutionally significant. Looking out across the landscape of advanced democracies, the most striking (although not unique) feature of the American system is our stubborn insistence on single-member districts. (45) Although not constitutionally required, single-member districts are deeply embedded in American law and political practice and are often required by the Voting Rights Act. (46) What is most distinctive about single-member districts, as compared with alternative arrangements such as party lists, is their emphasis on local interests. A system of single-member districts is well suited to a political world of relatively weak parties and divergent local interests, in which a major part of the job of a representative is to bring "pork" projects home and otherwise to advance interests that are distinctly local. (47) In such a political world, party list systems will often fail to protect local interests. Absent one person, one vote, single-member district systems are prone to the same problems. If gross population disparities and rotten boroughs are allowed, there are almost no constraints on the power of political insiders to draw or maintain maps that entrench the power of some local interests at the expense of others.
Some early one person, one vote cases were explicitly litigated in part as "geographic discrimination" claims, in which part of the claim was that urban voters as a group voted differently from rural voters as a group. (48) Even where this was less explicit, it was the identifiable and stark gap between urban and rural interests--even where both urban and rural groups might elect representatives from the same party--that led the Court to see the need for one person, one vote. (49) Courts could easily understand that the early cases involved not just random numerical disparities, but a "rural strangle hold on the legislature." (50) At the same time, no court has ever required a showing of geographic vote polarization in order to sustain the one person, one vote rule. The primacy of the local is, so to speak, baked in: it is a conceptual ingredient in the rule itself.
And perhaps that is what is so clever about the one person, one vote rule. This rule does not require thick, fact-intensive analysis. There is nothing like the Gingles test (51) in numerical vote dilution cases, let alone the difficulties that have plagued partisan gerrymandering claims. We need not debate which local interests may have had their strength diluted, nor do we need to make any predictions about who is likely to be elected from each district or about which groups will dominate which districts. We need not even decide which of the many overlapping numerical groups one might define deserves protection in the first place--which is a good thing, since each of us is part of many such groups at once. Instead, one person, one vote protects local interests indirectly, by ignoring everything but the simplest question: how many? By protecting numerical groups, the one person, one vote rule provides a certain limited degree of protection for many different kinds of groups, as long as those groups are to some degree geographically concentrated. In this way the doctrine is not only admirably administrable, (52) but also in a certain sense pluralistic. Rather than reifying a particular set of numerical groups as the set to protect, one person, one vote indirectly protects them all. (53) On the other hand, one person, one vote offers no protection to scattered and diffuse groups that lack any degree of geographic concentration; its pluralism extends only to numerical groups defined in geographic terms. It works only if, and to the extent that, we have at least some politically relevant things in common with our neighbors.
III. WHO ARE THE PEOPLE?
My argument up to this point does not resolve the pressing question with which we began. Which numerical groups should the one person, one vote rule protect: groups of people, or groups of eligible voters? (54) Once we leave behind the "weight" of individual votes, and flame the question instead in group terms, the best arguments on both sides come more clearly into focus. This debate is not about votes, but about the terms of belonging in our polity. It is about the ways The People includes or excludes the people who lack the vote, such as noncitizen immigrants, children, and felons.
The strongest argument for switching to CVAP or some other eligible voter measure rests on the following bold claim: virtual representation is dead. The modern history of the expansion of the franchise has gradually eaten away at virtual representation, which was once the rule rather than the exception. (55) As our law enfranchised group after group--propertyless men, women, racial minorities, eighteen- to twenty-one-year-olds--we whittled virtual representation down to its core. Today, only children, noncitizens, most felons, some ex-felons, and very few others are virtually represented by the voting-age citizens who happen to live in their communities. (56) Perhaps we have reached a tipping point and are ready to do away with virtual representation entirely. It is rather awkward, today, for courts to defend total population as the basis of representation on the quasi-originalist grounds that "[t]he framers were aware that this apportionment and representation base would include categories of persons who were ineligible to vote--women, children, bound servants, convicts, the insane, and, at a later time, aliens." (57) The Framers went in for quite a bit of virtual representation, including most notoriously in the case of slaves, (58) but the modern history of the expansion of the franchise has gradually undercut the Framers' framework. Perhaps we are ready to take the plunge and declare that the Constitution no longer sanctions virtual representation at all.
On the other hand, a number of strong arguments support the proposition that the numerical groups the Constitution protects are groups of people, not groups of citizens. First, the Constitution requires that congressional reapportionment--the allocation of representatives to states--proceed on the basis of census population, not eligible voters. (50) It would be odd, to say the least, for states to receive their representatives based on total population but then allocate them on an intrastate basis by eligible voters instead. A state like Texas could conceivably qualify for additional representatives entirely because of a growing population of children and/or immigrants in one part of the state yet allocate all of its additional representatives to some other (older, whiter) part of the state. Such a result would make a mockery of the idea of apportionment on the basis of "the whole number of persons." (60) Of course, congressional redistricting might be a special case. It might be possible to interpret the Equal Protection Clause to require a different result for noncongressional redistricting--although if so, the text of the Clause, which protects persons rather than citizens, is not an especially promising starting point. (61)
Second, there is the problem of the political independence of the count. Contemporary advocates of counting voters rather than persons sensibly avoid arguing for the use of actual voters, or registered voters, in part because these figures can be altered by policy decisions by the government that affect, among other things, how easy or difficult it is to register or vote. (62) It is inherently problematic for the basis of political representation to be subject to political manipulation; the beauty of the one person, one vote rule is that it provides at least some minimal degree of constraint on politicians' ability to decide which groups will be represented. But the eligible voter standard suffers from the same lack of independence from politics as registered or actual voters. Under an eligible voter standard, laws restricting voter eligibility, such as felon disenfranchisement laws, alter the basis of apportionment. Even CVAP is subject to manipulation if the government alters naturalization policy. (63) The simplest way out of these problems is to count something that politicians cannot as easily manipulate: the census count of persons. Of course, no count is completely independent of politics; it is a matter of degree. (64) Still, the census count of persons is the least manipulable figure, in that each additional requirement one imposes beyond the basic census count--citizenship, eligibility, registration--adds one more layer of additional variables subject to potential manipulation. (65)
Third, the very political assumptions that make one person, one vote make sense in the first place--the primacy of local interests and pork in a decentralized political world of relatively weak parties--point strongly to a regime of equipopulation. Call it the "equal pork" argument: if what legislators do, in significant part, is fight for resources for their districts, and if all persons (not just all adult citizens) are to enjoy the equal protection of the laws, then it would seem to follow that each legislator ought to be responsible for bringing resources home to roughly the same number of persons. Children--and for that matter resident aliens--need roads, bridges, schools, and Teapot Museums (66) as much as the rest of us do, if not more. (67)
If, in the end, the constitutional rule is that one person, one vote can operate on either basis--total population or CVAP-it is worth noting an interaction with other elements of our voting rights regime. Specifically, if the Constitution permits either approach, then the choices jurisdictions make to use total population or CVAP would be subject to both section 2 and section 5 of the Voting Rights Act. (68) This would not matter in an all-white jurisdiction in which the persons-versus-citizens debate amounted to a political tug-of-war between child-filled suburbs and age-restricted retirement communities. But in our contemporary politics, this debate is primarily about Latino political power. Proponents of CVAP should be careful about pressing jurisdictions to adopt a rule that might be a permissible interpretation of one person, one vote, but might also amount to racial vote dilution.
There is something undeniably elegant about the multilayered protections for political equality that our voting rights regime has come to embody. In the first, foundational layer, our law treats voters as individuals, granting each of us a right against disenfranchisement that cannot be sacrificed to anyone's claims of group-based equality. Second, in the legal regime that is my subject here, the one person, one vote rule looks out at The People and ensures that equally sized numerical groups of us will have equal representation. This second layer of political equality views The People stripped of race, language, party, and--at least under the present de facto national policy--even citizenship and age. All that matters about each group of us is how many of us are in it. This is a bare, thin layer of democratic equality, innocent of political facts that divide us. Third, race and language reenter the picture, as our law protects racial groups and language minority groups against vote dilution and against barriers to political participation. Finally, fourth, partisanship returns to the picture, at least if the possibility of a jurisprudence of partisan gerrymandering were to be realized.
Doctrinally, each of these layers is lexically prior to the ones that follow. One must not violate the individual right to vote in the course of vindicating one person, one vote; nor is it permissible to violate one person, one vote to protect against racial vote dilution; nor may we engage in racial vote dilution to vindicate values of partisan group fairness. Each of these layers protects a recognizable, yet distinct, form of political equality. Our law protects them all. It is not obvious, in terms of the substantive values at stake in these different forms of political equality, why this lexical arrangement would have arisen or whether it is justified. Why not allow for some flexibility in weighing these different forms of equality--for example, why not allow a wider range of departures from strict population equality in the name of preventing racial vote dilution? (69) Indeed, why not place these distinct forms of political equality on an equal footing? The trouble is that requiring several competing forms of equality, with no rule for how to order or balance them, is a way of requiring very little. (70) A lexical ordering shapes these different forms of political equality into a more coherent and administrable body of law. (71) The route our law has taken is to protect each of these forms of equality, but in cases of conflict, to give priority to the forms of equality that treat us more abstractly. The choice of this priority rule seems to reflect, in an inchoate way, an idea that before we are partisans, and even before we are members of racial groups, we have a more fundamental kind of equal standing--as people, without all of that weightier political baggage.
However, one person, one vote does not abstract us all the way. We are not, in this legal regime, isolated, atomistic individuals whose only interest is in having our vote count, or in maximizing our own individual vote's "weight" or power. Despite the Court's oft-repeated rhetoric about the weight of individual votes, the reality of one person, one vote is that we appear before the law in a collective way. This layer of the law views us with an eye toward representation; it views us as groups of human beings sufficiently numerous to merit our own representatives in the halls of power. Before considering partisanship or racial polarization, this layer of the law provides for a thinner, simpler, and arguably more fundamental baseline of group-based fairness in the relationship between The People and our representatives. It is, in a way, a reflection of a long-term democratic triumph that we have reached the moment when litigants can come forward and argue seriously that The People consist only of the eligible voters. As the franchise has grown, virtual representation has narrowed in ways that the Framers of our Constitution could scarcely have imagined. This is a wonderful development by any measure. But courts should take care before holding that only eligible voters count as The People. There is value in having a government that represents all of the people living in its jurisdiction and subject to its laws, a value we ought not to trade away lightly.
(1.) Colegrove v. Green, 328 U.S. 549, 556 (1946) (plurality opinion) ; see Baker v. Carr, 369 U.S. 186, 209-10 (1962).
(2.) See Lepak v. City of Irving, No. 3:10-CV-0277-P, 2011 WL 554155, at *1 (N.D. Tex. Feb. 11, 2011), aff'd No. 11-10194, 2011 WL 6217946 (5th Cir. Dec. 14, 2011).
(3.) Id; see also infra note 59 (discussing Louisiana's challenge to the U.S. Census on related grounds).
(4.) See Burns v. Richardson, 384 U.S. 73, 91 (1966) ("We start with the proposition that the Equal Protection Clause does not require the States to use total population figures ...."); Garza v. Cnty. of L.A., 918 F.2d 763, 773-74 (9th Cir. 1990) (holding that equal total population may be used, notwithstanding unequal numbers of eligible voters); see also Chen v. City of Houston, 532 U.S. 1046, 1047 (2001) (Thomas, J., dissenting from denial of certiorari) ("We have never determined the relevant 'population' that States and localities must equally distribute among their districts.").
(5.) See, e.g., Sanford Levinson, One Person, One Vote: A Mantra in Need of Meaning, 80 N.C.L. REV. 1269, 1281 (2002) ("[T]he United States currently seems to operate under a system in which any given representative should have (roughly) the same number of constituents as any other representative.").
(6.) Proponents of citizen voting age population (CVAP) argue that it can serve as a proxy, albeit an imperfect one, for eligible voters, since the vast majority of those ineligible to vote are either minors or noncitizens. See, e.g., Brief Amicus Curiae of Edward Chen and the Project on Fair Representation in Support of Appellants at 11-17, Perry v. Perez, No. SA-11-CV-788, 2011 WL 5904716 (W.D. Tex. Nov. 25, 2011), vacated, 132 S. Ct. 934 (2012) (arguing, as amicus in a lawsuit challenging interim statewide maps adopted by a federal court in Texas, that one person, one vote ought to protect "eligible voters" and that CVAP would be the best way to do this).
(7.) See Heather K. Gerken, The Costs and Causes of Minimalism in Voting Cases: Baker v. Cart and Its Progeny, 80 N.C.L. REV. 1411, 1414-15 (2002) (discussing the Court's refusal to articulate any larger theory of what the one person, one vote cases are about).
(8.) Reynolds v. Sims, 377 U.S. 533, 555 (1964) (emphasis added).
(9.) 376 U.S. 1 (1964).
(10.) U.S. CONST. art. I, [section] 2, cl. 1.
(11.) Wesberry, 376 U.S. at 7-8.
(12.) Id. at 8.
(13.) See also Colegrove v. Green, 328 U.S. at 570 (Black, J., dissenting) (arguing that "the constitutionally guaranteed right to vote, and the right to have one's vote counted clearly imply the policy that state election systems ... should be designed to give approximately equal weight to each vote cast").
(14.) U.S. CONST. art. IV, [section] 4; see Baker v. Carr, 369 U.S. 186, 209-10 (1962) (holding that challenges to districting plans are justiciable under the Equal Protection Clause, not the Guaranty Clause).
(15.) See Pamela S. Karlan, Politics by Other Means, 85 VA. L. REV. 1697, 1716 (1999) (discussing the Court's "resolutely individualistic" approach). Michael McConnell argues that this choice was a mistake: instead of relying on an individualistic reading of the Equal Protection Clause, the Court should have read the Guaranty Clause--which guarantees each state a republican form of government--to prohibit malapportionment so egregious that it thwarts majoritarian government. Michael W. McConnell, The Redistricting Cases: Original Mistakes and Current Consequences, 24 HARV. J.L. & PUB. POL'Y 103, 106-07, 114 (2000). But at the time of Baker, the Court had reason to proceed the way it did. See Samuel Issacharoff & Richard H. Pildes, Not by "Election Law" Alone, 32 LOY. L.A.L. REV. 1173, 1180 (1999) (arguing that the Court adopted the individual rights approach "to break the restraints of the political question doctrine that had long kept the issues of democratic design outside of the constitutional law arena," but that this approach "locked into place conceptual tools that soon proved insufficient").
(16.) Joseph Fishkin, Equal Citizenship and the Individual Right To Vote, 86 IND. L.J. 1289 (2011).
(17.) John F. Banzhaf III, Weighted Voting Doesn't Work: A Mathematical Analysis, 19 RUTGERS L. REV. 317 (1965). This literature goes back even further, to L. S. Penrose, The Elementary Statistics of Majority Voting, 109 J. ROYAL STAT. SOC. 53 (1946). For a critical survey of this literature, see sources cited infra note 20.
(18.) See infra notes 22 and 34 and accompanying text.
(19.) See Jonathan W. Still, Political Equality and Election Systems, 91 ETHICS 375, 380-82 (1981).
(20.) See Andrew Gelman, Jonathan N. Katz & Joseph Bafumi, Standard Voting Power Indexes Do Not Work: An Empirical Analysis, 34 BRIT. J. POL. SCI. 657 (2004) (explaining why the coin-flip approach that characterizes measures such as the Banzhaf index fails in certain systematic ways to deliver accurate estimates of the probability of a decisive vote); Nathaniel Persily, Thad Kousser & Patrick Egan, The Complicated Impact of One Person, One Vote on Political Competition and Representation, 80 N.C.L. KEY. 1299, 1313 (2002) ("Only if we control for all other political variables does population size become the dispositive factor .... Once the model becomes more complex and realistic, it falls apart ...."); see also Andrew Gelman, Jonathan N. Katz & Francis Tuerlinckx, The Mathematics and Statistics of Voting Power, 17 STAT. SCI. 420, 424-32 (2002) (discussing a variety of empirically-driven models that provide more accurate estimates of the probability of a decisive vote than the Banzhaf coin-flip approach).
(21.) Indeed, this strategic fact was obvious enough to ordinary observers that it led some Gore voters in safe states (of all sizes) to offer to "trade" votes with would-be Nader voters in swing states like Florida. See, e.g., Lynda Gledhill, California Shuts Down Vote-Trader Web Site, S.F. CHRON., Oct. 31, 2000, http://www.sfgate.com/cgi-bin/article.cgi?file=/chronicle/ archive/2000/10/31/MN119607.DTL; see also Jamin Raskin, Nader's Traders, SLATE, Oct. 25, 2000, http://www.slate.com/id/91933 (suggesting a Gore-Nader vote-swapping plan).
(22.) This is part of why the Court originally rejected the Banzhaf index. See Whitcomb v, Chavis, 403 U.S. 124, 145-46 (2971) (explaining, in the context of racial vote dilution, that the Banzhafindex fails to account for factors such as voters' partisan leanings).
(23.) RICHARD TUCK, FREE RIDING 44 (2008) ("If, say, in an election 10,000 votes were cast for Candidate A and 3,999 for Candidate B, the efficacious set must have consisted of 4,000 of the 10,000 votes cast for A, and there is therefore a 2 in 5 chance that our ballot [for A] was part of it.").
(25.) In a squeaker, as the margin of victory approaches zero, the chance of your vote for the winner being within the efficacious set approaches 100%.
(26.) See Persily et al., supra note 20, at 1313 (noting that equal probabilistic weight "would require that each district be equally competitive").
(27.) Davis v. Bandemer, 478 U.S. 109, 132 (1986) ("An individual or a group of individuals who votes for a losing candidate is usually deemed to be adequately represented by the winning candidate and to have as much opportunity to influence that candidate as other voters in the district.").
(28.) It is a normative statement about representation, not an empirical claim: its empirical truth or falsity does not matter. See Frederick Schauer, Legal Fictions Revisited, 9 (Univ. of Va. Sch. of Law, Pub. Law & Legal Theory Research Paper Series No. 2011-29, 2011), available at http://ssrn.com/abstract=1923314 (discussing "afactual" legal fictions).
(29.) Of course, when a representative aims to maximize both her own chances of reelection and the number of seats her party will win, she will not try to jettison all the opposition voters. Instead she will aim to represent just enough of them that they pose no realistic threat to her reelection chances. In this case, far from wanting such voters to be part of her own political coalition, her aim in representing them is to ensure that they do not become part of any winning political coalition, in her district or elsewhere.
(30.) Cf. Still, supra note 19, at 378-80 (discussing an "equal shares" conception of political equality).
(31.) This problem also affects the probabilistic weight of a vote.
(32.) See T. Alexander Aleinikoff & Samuel Issacharoff, Race and Redistricting: Drawing Constitutional Lines After Shaw v. Reno, 92 MICH. L. REV. 588, 601 (1993) (inventing and deploying this evocative term in the racial gerrymandering context).
(33.) One might also understand this last kind of "weight" as having an expressive dimension. By enforcing one person, one vote, the state is endorsing a certain form of political equality--whether of persons or of citizens--that does not depend on who actually casts ballots.
(34.) 489 U.S. 688, 698 (1989).
(36.) But see Garza v. Cnty. of L.A., 918 F.2d 763, 782 (9th Cir. 1990) (Kozinski, J., concurring in part and dissenting in part) (apparently reading this passage from Morris in precisely the
opposite way, as an affirmation that the Court was talking about individuals rather than groups).
(37.) See Heather K. Gerken, Understanding the Right to an Undiluted Vote, 114 HARV. L. REV. 1663, 1666-67 (2001) (arguing that although individuals bring racial vote dilution cases, these and other "aggregate rights" claims turn on group-based conceptions of equality).
(38.) Davis v. Bandemer, 478 U.S. 109, 167 (1986) (Powell, J., concurring in part and dissenting in part). Justice Powell goes on to acknowledge the individual "weight" idea but strongly suggests that, by itself, it does not do the work. "While population disparities do dilute the weight of individual votes," he writes, "their discriminatory effect is felt only when those individual votes are combined." Id.
(39.) See Fishkin, supra note 16, at 1332-36. One need not, but one could: groups also have interests in preventing their members' disenfranchisement.
(40.) Conceptually, one need not define numerical groups in geographic terms. A different kind of numerical group might consist of a widely dispersed collection of people who share some politically relevant thing in common. Protecting such a group's political interests might require voting schemes other than the single-member district and tools other than the one person, one vote rule. However, in the one person, one vote context that is our subject here, the numerical groups that matter are those defined in geographic terms.
(41.) The one person, one vote cases often recur to the principle of "equal representation for equal numbers of people." Wesberry v. Sanders, 376 U.S. 1, 18 (1964). At least in the context of geographically defined districts, that principle means equal representation for numerical groups as I have just defined them. Yet courts in the one person, one vote cases tend to make such arguments alongside, and sometimes even in the same breath as, arguments about the weight of votes that sound in a much more individualistic register. See, e.g., Kirkpatrick v. Preisler, 394 U.S. 526, 530-33 (1969) (repeatedly invoking "the constitutional command to provide equal representation for equal numbers of people," but also stating at one point that this very principle is "designed to prevent debasement of voting power and diminution of access to elected representatives").
(42.) See League of United Latin Am. Citizens v. Perry, 548 U.S. 399 (2006); Vieth v. Jubelirer, 541 U.S. 267 (2004); Davis, 478 U.S. 109.
(43.) For instance, although all three forms of vote dilution involve claims by individual voters, all these claims are resolved by evaluating the interests of the relevant group. See Gerken, supra note 37, at 1681-89.
(44.) See, e.g., Thornburg v. Gingles, 478 U.S. 30, 56 (1986) (explaining that an inquiry into racially polarized voting is necessary "to ascertain whether minority group members constitute a politically cohesive unit and to determine whether whites vote sufficiently as a bloc usually to defeat the minority's preferred candidates").
(45.) See, e.g., Lani Guinier, The Representation of Minority Interests: The Question of Single-Member Districts, 14 CARDOZO L. REV. 1135, 1135 (1993) (discussing, and decrying, "the particular bias toward winner-take-all single-member electoral districts common in the United States").
(46.) See, e.g., a U.S.C. [section] 2c (requiring single-member districts for Congress).
(47.) See BRUCE CAIN, JOHN FEREJOHN & MORRIS FIORINA, THE PERSONAL VOTE: CONSTITUENCY SERVICE AND ELECTORAL INDEPENDENCE (1987) (arguing that legislators, especially in the United States, devote substantial effort to advancing local "allocative" interests--pork--and helping individual constituents). Cain, Ferejohn, and Fiorina argue that single-member districts in which the voters choose among individual candidates for specific offices by name, as in the United States, produce the strongest incentives for constituency service and electoral independence--that is, for candidates and elected officials to side with local interests over party. See id. at 214-24. This effect is compounded by an unusual feature of the American system: U.S. voters separately elect many different officeholders, multiplying their opportunities to split their tickets, choosing different parties' candidates for different offices. See id. at 214.
(48.) Richard Briffault, Bush v. Gore as an Equal Protection Case, 29 FLA. ST. U. L. REV. 325, 362 (2001) (arguing that the reapportionment revolution began with claims of "geographic discrimination", or "the problem of state laws favoring voters in some areas over voters in other areas"); Gerken, supra note 7, at 1422 n.38 ("[T]hese cases were originally litigated as traditional equal protection claims based on the assumption that these two groups had different interests."); see, e.g., Swarm v. Adams, 385 U.S. 440, 446 (1967) ("[B]esides unequally populated legislative districts ... [a]ppellants also claim that the legislative plan discriminates invidiously by underrepresenting the populous urban counties and by overrepresenting the sparsely settled rural counties...."); W.M.C.A., Inc. v. Simon, 370 U.S. 190, 191 (1962.) (per curiam) (remanding a claim of "invidiously discriminatory geographic classification"); Baker v. Carr, 369 U.S. 186, 273 (1962) (Frankfurter, J., dissenting) (quoting the complaint as challenging a "purposeful and systematic plan to discriminate against a geographical class of persons").
(49.) Cf. James A. Gardner, Liberty, Community and the Constitutional Structure of Political Influence: A Reconsideration of the Right To Vote, 145 U. PA. L. REV. 893, 921 (1997) (describing Wesberry and Reynolds as "a kind of Gomillion for urbanites") ; Lani Guinier & Pamela S. Karlan, The Majoritarian Difficulty: One Person, One Vote, in REASON AND PASSION: JUSTICE BRENNAN'S ENDURING INFLUENCE 207, 219 (E. Joshua Rosenkranz & Bernard Schwartz eds., 1997) (noting that because malapportionment "stymied the interests of voters in the burgeoning cities and suburbs" in favor of "extremist politicians from underpopulated and disenfranchised 'Black Belt' regions [of the South who] were at the forefront of massive resistance, much of the Supreme Court's workload was an indirect consequence of malapportionment's hold on state political processes").
(50.) Md. Comm. for Fair Representation v. Tawes, 377 U.S. 656, 674 n.19 (1964). The Court in Tawes invokes this "strangle hold" language almost in passing. As a descriptive matter this observation was so obviously true as to be unremarkable. The "strangle hold" language also underscores a related point: the gross malapportionment that preceded one person, one vote did not just privilege some numerical groups over others, but also locked those group-based inequalities into place, creating a "lockup" of the democratic process. See Samuel Issacharoff & Richard H. Pildes, Politics as Markets: Partisan Lockups of the Democratic Process, 50 STAN. L. REV. 643, 668-70 (1998).
(51.) See Thornburg v. Gingles, 478 U.S. 30, 36-37 (1986) (citing a set of factors probative of a violation under section 2 of the Voting Rights Act).
(52.) Cf. JOHN HART ELY, DEMOCRACY AND DISTRUST: A THEORY OF JUDICIAL REVIEW 121 (1980) ("[A]dministrability is its long suit, and the more troublesome question is what else it has to recommend it.").
(53.) This is the flip side of an important criticism of one person, one vote: that the doctrine thwarts regional government by preventing the formation of regional bodies on which each locality has a distinct and perhaps equal voice, notwithstanding differences of population. See Richard Briffault, Who Rules at Home?: One Person/One Vote and Local Governments, 60 U. CHI. L. REV. 339, 401-19 (1993). For better and worse, one person, one vote treats numerical groups that happen to straddle or crisscross the borders of localities as no more or less entitled to representation than numerical groups whose edges match those local borders.
(54.) Cf. Chad Flanders, How To Think About Voter Fraud (And Why), 41 CREIGHTON L. REV. 93, 115 (2008) (arguing that to decide what counts as vote dilution, "we need a theory of which groups matter").
(55.) The concept of virtual representation has played a number of roles in American political life, from the colonists' supposed representation in the British Parliament to the longstanding virtual representation of women and children by male heads of household. For our purposes here, we can define virtual representation simply as the state of being counted as a constituent yet not enfranchised as a voter.
(56.) In the case of incarcerated felons, the crucial question is how to define the community that is to virtually represent them. The argument from the priority of the local, sketched above, provides some reasons why we might believe felons' own communities, rather than their places of incarceration, might make for a more plausible form of virtual representation. See supra text accompanying notes 45-47. In addition, counting felons in their places of incarceration, rather than in their own communities, gives politicians more power to manipulate the population of districts by building, expanding, or closing prisons in those districts. See infra notes 63-64 and accompanying text. But these are arguments for another day.
(57.) Garza v. Cnty. of L.A., 918 F.2d 763, 774 (9th Cir. 1990).
(58.) See Levinson, supra note 5, at 1285 ("Interestingly enough, the [Garza] court did not include in its originalist litany perhaps the best example for its point--the Three-Fifths Clause.").
(59.) U.S. CONST. amend. XIV, [section] 2 ("Representatives shall be apportioned among the several States according to their respective numbers, counting the whole number of persons in each State, excluding Indians not taxed."). This appears to be the only operative part of Section 2 of the Fourteenth Amendment, the rest of the Section having never been enforced.
The "whole number of persons in each State" language seems an awfully clear example of the "Constitution of Settlement"--one of those parts of the Constitution that lays down a rule that structures our politics rather than offering an invitation to conversation and interpretation. See SANFORD LEVINSON, FRAMED: AMERICA'S FIFTY-ONE CONSTITUTIONS AND THE CRISIS OF GOVERNANCE 17-27 (2012) (contrasting the "Constitution of Settlement" with the "Constitution of Conversation," the open-textured clauses that generate legal conversation and argument). However, this language did not stop the State of Louisiana from filing a remarkable complaint in the Supreme Court of the United States in November 2011, see Complaint, Louisiana v. Bryson, No. 220140 ORG (U.S. Nov. 14, 2011), available at http://sblog.s3.amazonaws.com/wp-content/uploads/2011/11/ Louisiana_v_Bryson.pdf, attempting to unsettle the "whole number of persons" rule and exclude illegal immigrants from the census count. Louisiana argued that both the Equal Protection Clause (citing Sims) and Article I, Section 2 (citing Wesberry) protect "the weight of a citizen's vote," and that therefore, it amounts to vote dilution to apportion congressional districts based on total population, including illegal immigrants, since districts then contain different numbers of "voters." Id. at 31-35. The Court ultimately denied Louisiana's motion to file this complaint. Louisiana v. Bryson, _ S. Ct._ (March 19, 2012). The argument of this Essay provides some reasons to doubt whether disparities of this kind do any real constitutional work at all--let alone so much work that courts should unsettle a provision as clear as "the whole number of persons in each State."
(60.) U.S. CONST. amend. XIV, [section] 2; see Kalson v. Paterson, 542 F.3d 281, 289 n.16 (2d Cir. 2008) (Calabresi, J.) ("In light of the fact that Article I apportions seats between states based on total population and not on the basis of the number of voters, it would seem to follow that apportionment within a state must also be based on total population."). Conversely, if within-state apportionment were to be based on CVAP, it would seem only fair to apportion representatives among states on the same basis and to redistribute to other states the extra representatives that states like Texas receive on account of persons who would no longer be counted as part of the basis of apportionment. However, squaring this result with constitutional text presents serious difficulties. See supra note 59.
(61.) U.S. CONST. amend. XIV, [section] 1; see Yick Wo v. Hopkins, 118 U.S. 356, 369 (1886) (aliens are persons entitled to full protection of the Equal Protection Clause); Garza, 918 F.2d at 775 (aliens' equal protection rights include political rights, even if aliens cannot vote or hold certain offices).
(62.) See Burns v. Richardson, 384 U.S. 73, 92 (1966) ("Use of a registered voter or actual voter basis.... [would be] susceptible to improper influences by which those in political power might be able to perpetuate underrepresentation....").
(63.) For example, speeding up or slowing clown the processing of applications for naturalization in advance of the census could have large effects. See Adam B. Cox & Eric A. Posner, The Rights of Migrants: An Optimal Contract Framework, 84 N.Y.U. L. REV. 1403, 1450 n.149 (2009). More generally, immigration law changes that affect the mix of legal statuses of newcomers, or of those already here, can have substantial effects--albeit after a significant delay--on the future citizen population. See id. at 1447-53.
(64.) In certain cases, such as the large military base at issue in Burns v. Richardson, 384 U.S. 73, a government decision such as the choice to deploy or transfer troops may have large direct effects on census counts. More generally, immigration policy affects total population figures, and myriad government policies have some impact on where people live or on the number of children born. Debates about undercounts and sampling affect the census itself, as do debates about counting Americans overseas. See Nathaniel Persily, The Law of the Census: How To Count, What To Count, Whom To Count, and Where To Count Them, 32 CARDOZO L. REV. 755, 783-86 (2011); see also supra note 59 (discussing Louisiana's challenge to the 2020 census).
(65.) Related to this problem of political manipulation is the problem of accuracy. As the Court has noted, "the census data provide the only reliable--albeit less than perfect--indication of the districts' 'real' relative population levels." Karcher v. Daggett, 462 U.S. 725, 738 (1983). Mixing census data with other data in order to estimate variables that are not part of the census--especially citizenship--raises significant questions about data quality and overall accuracy. See Persily, supra note 64, at 774-82.
(66.) See CITIZENS AGAINST GOV'T WASTE, 2006 CONGRESSIONAL PIG BOOK SUMMARY 47, http://www.cagw.org/assets/pig-book-files/2006/2006pigbooksummary.pdf.
(67.) In addition, children, aliens, and others sometimes seek assistance from their representatives. Constituency service is truly an individual interest. It is not an interest in the "weight" of a vote, but it is an interest that one person, one vote helps equalize, assuming that the number of constituents affects the quality of service. Thus, to this "equal pork" argument one might add a related "equal constituency service" argument. See generally CAIN, FEREJOHN & FIORINA, supra note 47, at 50-76 (arguing that legislators, especially in the United States, devote significant energies to helping individual constituents with their dealings with government).
(68.) Proponents of using CVAP in one person, one vote cases also make a different point: that racial vote dilution cases under the Voting Rights Act themselves commonly make use of CVAP, and so one person, one vote cases should as well. See, e.g., Brief Amicus Curiae of Edward Chen and the Project on Fair Representation in Support of Appellants at 15-16, Perry v. Perez, No. SA-11-CV-788, 2011 WL 5904716 (W.D. Tex. Nov. 25, 2011), vacated, 132 S. Ct. 934 (2012) ("[U]sing CVAP would harmonize the one-person, one-vote inquiry with the test for evaluating [racial] vote dilution claims under Section 2 of the VRA." (internal citations omitted)).
CVAP is not the only possible starting point for racial vote dilution analysis and it may not be the best. See Persily, supra note 64, at 781-82 (arguing that VAP is a better starting point than CVAP because of the questionable accuracy of available citizenship data). Nonetheless, racial vote dilution cases under the VRA commonly use CVAP in several contexts. These include (a) determining which districts offer minority groups an opportunity to elect their candidates of choice, and similarly (b) the threshold inquiry under section 2 into whether any such district(s) can be drawn. In both of those contexts, CVAP is one step in a process of predicting the actual voters who are expected to turn out in a district and pick a winner, which we must predict in order to know which group, under polarized conditions, would have the opportunity to elect the candidates of its choice. One person, one vote is different: it involves no such inquiries into which groups will win or lose, who will be in the majority and who in the minority, in any given district. Instead of imagining an electorate divided and polarized, one person, one vote works from the premise that each geographically defined group has some interests in common, and that those interests will be represented regardless of who votes. Thus, the use of CVAP in racial vote dilution cases in these outcome-predicting contexts is no guide to the proper basis of representation for one person, one vote.
More problematic and interesting is the use of CVAP in yet a third Voting Rights Act context: judging "proportionality," or the relationship between a minority group's share of the districts and its share of the relevant population. See Johnson v. De Grandy, 512 U.S. 997, 1013-16 (1994) (introducing proportionality, and basing a finding of proportionality on a baseline of voting age population). Using CVAP to establish a baseline for a proportionality comparison seems in at least some tension with using total population as the basis of representation in one person, one vote cases. The reasons for this difference, and whether it is justified for these two layers of the law to frame the problem of equal representation differently, are topics for another day.
(69.) See Grant M. Hayden, Resolving the Dilemma of Minority Representation, 92 CAL. L. REV. 1589, 1627 (2004) ("[L]egislatures and, if necessary, courts should be able to manipulate both the shape and the relative population of the district in order to strengthen minority political participation.").
(70.) Cf. Georgia v. Ashcroft, 539 U.S. 461, 495-97 (2003) (Souter, J., dissenting) (noting that if section 5 protects majority-minority, coalition, and "influence" districts, with no rule for balancing these competing objectives, it creates no administrable constraint on jurisdictions' choices).
(71.) Cf. JOHN RAWLS A THEORY OF JUSTICE 36-40 (rev. ed. 1999) (explaining, analogously, how a lexical ordering of moral principles allows a theory to avoid "intuitionistic" balancing).
AUTHOR. Assistant Professor of Law, University of Texas School of Law. I would like to thank Mitch Berman, Steve Bickerstaff, Josh Chafetz, Adam Cox, Justin Driver, Chris Elmendorf, Jim Fishkin, Chad Flanders, Willy Forbath, Cary Franklin, Jim Gardner, Heather Gerken, Adam Gordon, Bernie Grofman, Rick Hasen, Sandy Levinson, Stefanie Lindquist, Nate Persily, Rick Pildes, and Nick Stephanopoulos for helpful comments, and Patrick Yarborough for excellent research assistance.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||equal representation for equal numbers of persons|
|Publication:||Yale Law Journal|
|Date:||May 1, 2012|
|Previous Article:||Districting for a low-information electorate.|
|Next Article:||Recognizing character: a new perspective on character evidence.|