Editor’s note: This post examines two cases the Supreme Court will review at its next Term: Evenwel v. Abbott, and Harris v. Arizona Independent Redistricting Commission. This post will be followed this week by a symposium on Evenwel.
In a rare meeting with reporters in early July 1968, Chief Justice Earl Warren answered readily when asked which of the Court’s decisions in his momentous years on the Supreme Court were most important. The series of rulings, he said, that created the constitutional idea of “one person, one vote.” That is a basic theory of democratic representation: no individual’s vote may be weighted more heavily than any other’s. Election districts, for Congress and state legislatures, must be drawn as close to equal in population as can practically be done.
The clearest expression of that idea came in the 1964 decision in Reynolds v. Sims. Now, a half-century later, the Court has assigned itself the task of deciding — remarkably, for the first time — how to get to that goal. What, it will answer in one case, population measure should be used: total people in a district, total citizens, total citizens of voting age, total numbers of registered voters? And, in a second case, the Court will answer whether it violates the equality principle if districts are drawn in a way that favors one party’s candidates.
Both cases involve cross-currents of political theory, and both have the potential to directly change election outcomes and the election fortunes of the two major political parties. There is considerable complexity here, so let’s try to make this simple.
The “overriding objective” in pursuing the “one person, one vote” mandate, the Court said in 1964, “must be substantial equality of population among the various districts so that the vote of a citizen is approximately equal in weight to that of any other citizen in the state.” It would say later that absolute mathematical equality is not required, and that some departure from equality is permitted to serve other legitimate state policies or interests.
Still, the starting point of the exercise is population — so far, population without a constitutionally binding definition. If practice were controlling, because this is how it’s done in most states, the starting point would be a state’s total population divided by the number of election districts, for state legislatures or for a state’s delegation to the U.S. House of Representatives. The usual measure of whether the equality principle has been denied is to compare the numbers gap (technically, the “maximum deviation”) between the most-populated district and the least-populated district.
For state legislative districts, the Court has said that the numbers gap should be no more than ten percent, although that is not a hard-and-fast maximum. Of course, with sophisticated computers working on highly detailed census data, it is quite easy to create maps that have very small numbers gaps between the largest and smallest districts. That expertise, of course, makes it possible — if those responsible for drawing the map are so inclined — to achieve near equality while at the same time making it easier for incumbents to get reelected, or for one party to do better than the other on election day.
In a 1966 decision, in a Hawaii state legislative redistricting case, Burns v. Richardson, however, the Supreme Court found no constitutional fault with a state drawing new districts as measured by the number of registered voters, provided that this would not lead to a gap that is significantly different from the result if total population per district had been used instead. What remains most important, the Court indicated, was that the result satisfies “one person, one vote.”
In 2001, the Court refused to hear a case, Chen v. City of Houston, in which voters challenged a citywide redistricting plan that started with total population. The result, the challengers said, was that some districts were unequal in their voter numbers, compared to other districts, with the result that their own votes were diluted in strength. Only Justice Clarence Thomas filed a dissent, saying that the Court would eventually need to take on a case to fill the gap on the population metric to be used.
The question of the starting point, of course, gets quickly into the democratic theory of who is supposed to be — or entitled to be — represented in elected chambers.
If total population is the measure, many people who can’t vote at all are counted: children under the age of eighteen, prison inmates and others convicted of crime, non-citizens including those living legally in the United States. Are they entitled to have their interests taken into account by an elected official whom they had no role in choosing? Does counting them give them some clout?
If a state is a “red” state or a “blue” state, with heavier registration for voters in one major party or the other, is counting by registered voter totals true to “one person, one vote”? What about a state where one party has among its followers a lot of poor people who tend not to vote or even register: what measure is better for them? Among voters, what is a fairer measure: those who register, or those who are eligible to vote but some of whom don’t register?
If a state is dominated by those in urban areas who vote more dependably, will the Democrats benefit most from one measure instead of another? Or how about a state dominated by voters in suburbia who are more likely to vote: will Republicans do better with one measure or another?
The Court in reviewing the two new state legislative redistricting cases does not necessarily have to answer all of those questions. In fact, it has the option of simply concluding, in the end, that the choice should be left to the states themselves to pick a population metric, and then have that tested to see whether near equality does result. But if the Court is inclined to make one measure constitutionally binding, it would need a representation theory to back that up. The Constitution’s text, of course, is not much help: it just insists on equality, period. And the constitutional doctrine of “one person, one vote” is not self-defining, so the Court has to make it functional.
Those two cases will come up for hearings at the Court’s next Term, and will be decided by next summer. Probably the more important of the two, in terms of constitutional meaning, is the case over the maps drawn up by the Texas state legislature in 2013, for filling the thirty-one seats in the state senate. Its starting point was total population, divided by thirty-one. It came close to equality: the largest-to-smallest numbers gap was 8.04 percent, definitely within the ten percent the Court has allowed.
But those maps were challenged by two voters, Sue Evenwel and Edward Pfenninger, who regularly exercise their right to vote. They interpret “one person, one vote” to require equality of voters, so they argue that the Constitution requires voting-age population to be the starting point. Each of them lives in a district where the voting-age population is considerably larger than in some other districts, so they argue that their votes are diluted, comparatively. In other words, it takes more of them to decide an election in their district, so their votes are less weighty. If there is an “ideal” district in terms of numbers, Evenwel says, her district is thirty-one percent larger, and Pfenninger says that his is forty-nine percent larger.
They sued in a three-judge federal district court, but lost. The judges ruled that the choice of the population starting point is one for the legislature to make. The starting point, that court said, goes directly to “the nature of representation,” and that should be a choice made by the elected representatives of the people.
The other case the Court will take up has a curious twist to it. It involves a decision by the voters of Arizona in 2000 to take away from its legislature the power to redistrict, and give it to an independent commission, with the specific aim of taking the process out of partisan maneuvering. And yet the case as it reached the Court is based upon an accusation that the redistricting commission itself yielded to partisan preferences, in a way that contradicts “one person, one vote.”
Another curious facet of that case is that the challengers are not asking the Court to rule that “partisan gerrymandering” is itself unconstitutional, only that it cannot be used when the result is unequal districts. (The Court has never ruled against partisan gerrymandering, as such, concluding that it has no idea how to judge when there has been too much partisanship.)
By the way, this case involves the same Arizona redistricting commission that the Court examined last Term, ruling that it did not violate the Constitution or federal law for the voters to assign the task of redistricting congressional seats to a body other than the legislature. The new case is focused on the commission’s separate power to redistrict the state legislature — a power that no one questions was a valid choice for Arizona voters to make. It is the result of the commission’s work in 2012, following the 2010 census, that produced the challenge.
A group of voters sued in a three-judge district court, claiming that the maps were unconstitutional because the commission packed more non-minority voters into Republican-dominated districts, making them larger, and put minority voters into smaller, normally Democratic districts, with the result of voter dilution in the GOP districts and a violation of “one person, one vote.” Although the district court found that partisan considerations had played some role, it ruled that those efforts were not responsible for the numbers gaps in the resulting districts.
The main motivation of the legislature, that court found, was to enhance the prospect that the maps would get official approval from the U.S. Department of Justice, under the federal Voting Rights Act. The intent, the court found, was to enhance minority voting influence across the state, in a bid for federal government acceptance at a time when that acceptance was necessary for a l state like Arizona; it is no longer necessary because the Supreme Court has since put an end to the need for some states to get federal approval.
When the Supreme Court accepted the Arizona case for review, it said it would confront two questions: whether a desire for partisan advantage justified the packing plan that distorted voters’ influence between districts, in violation of “one person, one vote” principles, and whether a desire to get Justice Department endorsement justified the creation of unequally sized districts in violation of those same principles. The commission tried to persuade the Justices to simply uphold the district court’s rejection of the challenge, arguing that any partisan influence was minimal, and that the need to satisfy the Justice Department was a valid justification for the maps that resulted. The Court took on the case, anyway.
One of the complications of this case is that there is a clear disagreement between the two sides on just what was done, and why, raising the prospect that the case may not turn out to be a particularly good test case on the larger constitutional issues that the voters’ appeal insisted are at stake.
Incidentally, the Arizona case does not involve the issue of the population metric to be used under the “one person, one vote” doctrine. The commission began with total population.
It is likely that, as the two cases develop further, the Justice Department will decide to take a part in them — especially to offer its views on what population starting point map-drafters should use.