Richard H. Pildes is the Sudler Family Professor of Constitutional Law at NYU.
As soon as the Court decided to hear Evenwel, a barely suppressed anger emerged in many quarters, on grounds of both process and substance. On process: how dare the Court address this issue, when a 1966 precedent seemingly settled the issue, and no conflict existed in the lower courts, to boot. On substance: how disturbing for the Court to consider any change in the legal status quo, in which states are perfectly free to define the “one person, one vote” baseline (total population or eligible voters) for themselves. But on both process and substance, these complaints and anxieties are misplaced and misguided.
The Court is right to confront this issue. And more importantly, the most likely outcome is that the Court will either re-affirm the status quo or conclude that equal protection requires states to use population, not voters, as the measure of political equality – a possibility almost none of the commentary, thus far, seems to recognize.
Let’s start with the substantive issue. The issue is whether “one person, one vote” is a principle of “representational equality” or one of “electoral equality.” Once the Court fully grapples with the issue, I consider it extremely unlikely a majority will conclude that the constitutional metric must be voters. Four reasons of principle and practicality, at least, lead to this conclusion. First, states have the power to extend even the right to vote itself to non-citizens; in the mid-nineteenth century, for example, non-citizens typically were given the right to vote (outside the Northeast), as Alex Keyssar’s leading history, The Right to Vote: The Contested History of Democracy in the United States, chronicles. States are not required, of course, to extend the vote to non-citizens, but doing so is constitutionally permissible and does not dilute the vote of citizens, if this historical experience provides guidance. If states can constitutionally include non-citizens in the population of eligible voters, it would be incongruous to conclude states lack similar discretion to include them in the population that counts for designing election districts.
Second, the Constitution’s text itself recognizes the validity of basing political representation on persons, rather than only voters. In Article I, Section 2 of the original Constitution, the apportionment among the states of members to Congress was based on the number of “persons”; when the Fourteenth Amendment revised the apportionment provisions to reflect the end of slavery, the same judgment was again made that political representation of the states in the House should be based on “the whole numbers of persons in each State.” Indeed, Congress specifically rejected proposals to base apportionment on eligible voters instead. The legitimacy of basing political representation on population, not voters alone, is embodied in these provisions. These provisions might not require states to equalize population across districts, but they strongly suggest using “persons” as the relevant baseline is constitutionally permissible.
Third, the practice of all states for several decades has been to use persons, not voters (whether voting-age population, citizen voting-age population, or eligible voters) as the redistricting metric. Both at the time of the Founding and the Fourteenth Amendment, and throughout American history, many states have used population as the standard. To the extent these political practices can “liquidate” or settle the Constitution’s meaning, they confirm that population is, at the least, a constitutionally permissible metric. In addition, the fact that states uniformly use population will make the Court realize just how destabilizing it would be to impose a sudden new constitutional rule requiring states to equalize the number of eligible voters across districts, even when doing so creates significant inequalities in the number of people across districts. Fourth, and finally, is the technocratic and practical problem: since the Census no longer asks respondents whether they are citizens, a constitutional requirement that states equalize the number of eligible voters across districts would be difficult for States and courts to administer. Citizenship data would have to come from the ACS rolling-survey data sets; others have pointed out the difficulties with basing once-a-decade redistricting on this data (should we ever have a system of automatic voter registration for all eligible voters when they come of age, this technocratic issue would evaporate).
So for all these reasons, the most interesting question in Evenwel is not actually whether the Constitution requires “electoral equality.” That the Court would reach this conclusion is highly unlikely. Once the Court rejects this conclusion, the more interesting question is whether the Court will remain content with the principle that the Constitution gives the states discretion to choose either “electoral equality” or “representational equality” as the proper interpretation of the Equal Protection Clause. Remarkably, the Court has only focused on this substantive question at all in one case, Burns v. Richardson (1966), decided at the dawn of the reapportionment revolution; Burns concluded states could make either choice. Now that the issue is back before the Court nearly fifty years later, the jurisprudential issue is whether all the developments in redistricting and voting-rights law in those intervening years should lead the Court to conclude that equal protection requires a uniform understanding concerning the correct population measure that must be used. (My co-authored casebook, The Law of Democracy, asks whether “Burns survives the subsequent development of voting rights law.”) If the Court does conclude that a uniform understanding of “equality” is required, the most likely outcome is representational equality – equality of the total number of persons across districts.
The argument for a uniform understanding of “equality” is strong, as a matter of both constitutional principle and pragmatic judicial implementation of the Constitution. In the apportionment cases, the Court has spoken eloquently many times about the importance of political equality in designing districts – but equality of whom, people or voters? If the basic principle is of such constitutional magnitude, there is much force to the conclusion that the Court has an obligation to specify equality of whom, or equality with respect to what value or principle. The choice between electoral equality and representational quality is not a fine-grained technical detail of how to implement the Equal Protection Clause. That choice is a fundamental, categorical one about the essential interpretation and meaning of equal protection in the context of designing our basic democratic institutions. Does the clause require that all persons in a jurisdiction (non-eligible voters as well as voters) have roughly equal political representation? Or does it require that all eligible voters have a roughly equal voting power? Those are fundamentally different-in-kind understandings of equal protection that flow from the Court’s “one person, one vote” jurisprudence – precisely the kind of question, in other contexts, to which the Court would provide the answer.
The reason the Court gave in Burns for leaving this choice instead to state discretion was that the decision of which groups to include in the baseline for districting “involves choices about the nature of representation with which we have been shown no constitutionally founded reason to interfere.” But in the context of the Reapportionment Cases, this explanation is off-key. After all, it was the vehement position of the dissenting Justices in these cases, such as Justices Harlan and Frankfurter, that the Court should not get involved in these issues at all because to get involved was to require the Court to choose among competing theories of political representation.
The Court crossed that Rubicon when it decided that equal protection did not permit representation to be based on geographic units, such as towns and counties, and did require it to be based on equal numbers of sentient beings (people or voters). Having completely redefined the basis of political representation the Constitution requires, the Court’s reticence about not wanting to choose between competing theories of representation when it comes to voters or people rings hollow. Instead, Burns reads like a tentative, interim, and transitional decision in the early stages of working out the meaning of the Reapportionment Cases. Decided only two months after argument, Burns arose with elections imminently pending and dealt with what was only an interim districting plan; in other words, the stakes were low, the need for an immediate decision pressing.
With the much fuller development of the “one person, one vote” doctrine in the fifty years since, it is not obvious the Court will be comfortable with leaving states as much discretion to choose “equality of whom” in districting. And given the intensity of today’s political conflicts over immigration, it is not difficult to imagine those politics coming to further poison redistricting, if states are free to move back and forth between using voters or persons as the measure of district equality. Given how aware the Court is of the extreme partisan polarization of our era, and how that polarization plays out already in districting, the Justices might conclude that strong pragmatic reasons further support adoption of a uniform principle concerning district “equality.”
The courts of appeals, in the three major cases raising this issue, have all explained why representational equality is the better interpretation of the principles underlying the “one person, one vote” doctrine. But all have recognized that the issue is important and the question close. In Evenwel, this issue arose for the first time in the Court’s non-discretionary appellate jurisdiction; the Court was right to take the case, rather than summarily affirm, and to give this issue the attention it deserves. Texas, as the defendant-appellee, will only ask the Court to affirm the status quo and let Texas (and other States) continue to have discretion to choose whether to create district equality between persons or voters. Texas will succeed to at least that extent, I believe. But now that the Court will be forced to confront these issues, the Court might well conclude that it has an obligation to decide whether there is a right answer to the question under the Equal Protection Clause of “equality of whom” and that the better answer is equality of political representation for all persons.
In my view, if population were the baseline, the state could limit “population” to those who are bona fide permanent residents, as long as the state employs a legitimate, non-discriminatory standard for permanent residence. One such example is Kostick v. Nago.