Does the Constitution guarantee one person, one vote? Or is it one citizen, one vote? This deceptively simple question is actually profound and the U.S. Supreme Court has agreed to decide it in the term that will begin in October. The answer will define the nature of American democracy for generations to come.
The legal nature of the question can be stated simply. In the 1961 case Reynolds v. Sims, the Supreme Court announced a principle that was then referred to as "one man, one vote." Until then it had been up to state legislatures to allocate congressional districts according to whatever principle they wanted. There was no requirement that districts have roughly equal numbers of residents, which meant that some districts might have many fewer residents and voters than others. The court said this in balance violated equal protection of laws because it diluted the votes of those who lived in relatively overpopulated districts.
When the court in Reynolds announced this brand-new constitutional right, the language of its opinion and of some subsequent opinions strongly hinted that the goal was for equal numbers of voters to vote for equal numbers of representatives. Yet in the 54 years since the Reynolds decision, the court has assiduously avoided explicitly stating whether districts must have the same number of people or the same number of citizens.
So long as most residents of a given state were also citizens, the court's unwillingness to answer this question didn't much matter. True, citizens younger than 18 can't vote, and some states disenfranchise felons, so very disproportionate numbers of children or convicts might've caused imbalance between districts.
But what really makes the question salient today is the question of immigration and the emergence of districts containing large numbers of residents who are not citizens, and are therefore not entitled to vote. If one district has vastly more noncitizens living in it than another, the state legislature making the districting decision must choose: Should it equalize the number of people living in the district when it allocates seats? Or should it equalize the number of citizens who are eligible to vote?
History provides no definitive answer but it does show that analogous questions have plagued elective democracies since the beginning. Consider the generation of the American Revolution, who rose against Britain under the slogan, "no taxation without representation." To the colonists, the fact that they did not elect members of Parliament meant they were excluded from representation.
But the standard British response was that colonists, acknowledged subjects of the Crown, were in fact "virtually" represented in Parliament. Although the colonists did not themselves elect members of Parliament, neither did most other Britons. Women were not allowed to vote but were said to be represented virtually. Parliamentary districts were wildly disproportionate in population. And the overwhelming majority of white Englishmen didn't have the franchise either.
When the American Revolution created a new country with new voting rules, the idea of virtual representation didn't disappear. The franchise in the U.S. was extended to a larger number of white men, but property requirements persisted for some time in some states. Women and blacks continue to be excluded, on the theory that they were virtually represented.
Most notoriously, at the Philadelphia convention that drafted the U.S. Constitution in 1787, Northern and Southern white men argued about how blacks should be counted in the census that would determine the allocation of members of the House of Representatives. They settled on the three-fifths compromise, which counted each black person as three-fifths of a white person.
Ironically, Southern slaveholders wanted blacks to count as full people for allocation purposes, and anti-slavery Northerners wanted blacks not to count all. No one expected enslaved blacks to be allowed the vote but the Southerners wanted to inflate the size of their populations to get more representatives in Congress and therefore more power relative to the Northern states. In essence, the Southerners wanted one person, one vote, while the Northerners wanted one citizen, one vote.
Today the political configuration is different. The friends of the court briefs filed so far in favor of the one citizen, one vote principle come from conservative organizations. Their idea is that, if noncitizens are counted in the census, it gives an unfair advantage to smaller numbers of citizen voters who live in the same districts as the noncitizens. That in turn, they believe, dilutes the votes of people who live in districts full of citizens.
You can expect responses to come from liberal groups, who will argue that all people deserve representation whether they are voting citizens or not. Indeed, when the U.S. Court of Appeals for the Ninth Circuit ruled on this issue 25 years ago, it said that the people as a whole, not just voters, are represented in government, and that nonvoters had a right of access to elected officials just like citizens do.
It's a fascinating twist that this contemporary liberal view is a version of the old British theory of virtual representation, according to which you're represented by elected officials even if you have no right to vote for them. In principle, liberals reject the idea of virtual representation, which was used to justify denying the franchise to women, blacks and whites who didn't hold property.
To be fully consistent, liberals probably should argue that noncitizens who are de facto permanent residents should be allowed to vote. But because almost no liberal today is prepared to argue that noncitizens should have the right to vote in the U.S., virtual representation becomes the implicit theory for how noncitizens can be represented.
The court could go either way on the issue. An originalist approach would allow for virtual representation, which could favor the liberal side. But conservative originalists may not like that result.
No matter what the decision, the impact on American democracy will be far-reaching. The question of who is represented is the question of who we the people really are. And whether that includes noncitizens is as important a matter as can be imagined.Noah Feldman
• 01/27/15: Why judicial elections are idiotic and bad for law