(Bloomberg View) -- Prisoners can be counted in population totals for determining a voting district, even though they can’t cast ballots in the place where they’re being held. That's what an appeals court relying on a U.S. Supreme Court decision from last term has said -- even though that case involved noncitizens who are fully members of the community, not inmates who don’t contribute to the city or use local services. Wednesday’s decision casts some doubt on the theory of virtual representation that the justices used, and raises deep issues about the connection between voting and being represented.
The case came out of Cranston, Rhode Island, where the Adult Correctional Institutions, the state’s prison, held 3,433 people, as of the 2010 U.S. Census. Cranston has six wards, each with roughly 13,500 residents. In 2012, the city adopted a new districting plan that counted all the prisoners as part of Ward Six.
But almost none of the prisoners can vote in the ward; 37 percent of the inmates have felony convictions, which in Rhode Island means they can’t vote while serving their terms. The rest are allowed by state law to vote as absentees in their last residences before going to prison. An expert estimated that only six or seven of the 3,433 inmates could vote in Ward Six.
Residents of Cranston’s other wards sued, asserting that by counting the inmates their votes were diluted in violation of the equal protection of the laws. In effect, they said, the roughly 10,000 Ward Six voters who aren’t in prison have more power than residents of the other five wards, because their votes count for more.
A federal district court judge agreed with the challengers and struck down the districting plan. The judge understood that he had to consider the Supreme Court ruling in Evenwel v. Abbott, which was decided in April. But he said the precedent didn’t determine the outcome of the Cranston case.
In the Evenwel case, the court held that it was permissible for states to use overall population as determined by the census to set voting districts. The opinion, by Justice Ruth Bader Ginsburg, rejected the claim that this method was unconstitutional where a district has large numbers of noncitizens who aren’t eligible to vote.
The district judge pointed out, with some plausibility, that the Supreme Court case involved the representation of people who are completely integrated into the political community, even if they aren’t qualified to vote. Noncitizens pay taxes, contribute to the economy and use local government services, including the schools. In these ways, the judge pointed out, noncitizens are similar to other groups who can’t vote but are included in the total population.
Today, that’s mostly kids. In the past, the category included men who didn’t hold land, women, slaves and some Native Americans. All these were considered to be “virtually” represented by elected officials.
The district court reasoned that inmates were different. They are literally segregated from the community where they are held. They are there not by choice but by assignment. They neither use local services nor contribute to the community. When freed, they are very unlikely to stay around. And, perhaps just as important, they may have the right to vote somewhere else.
The U.S. Court of Appeals for the 1st Circuit reversed the district court and approved the city’s redistricting plan. Its opinion first noted that there was no alleged racial or other unlawful discrimination associated with the redistricting, and that the population deviation was within the 10 percent that the Supreme Court usually considers minimal. It also said that the Evenwel case stands for the idea that courts should give wide latitude to local government districting.
In its punchline, the court said that the Supreme Court’s acceptance of census data in the Evenwel case, although not dispositive, was a strong reason to allow the Cranston plan, which also used census data. The court admitted that Ginsburg had noted in the decision that some states don’t count inmates from out of state. But it reasoned that this didn’t mean a state wasn’t allowed to count inmates.
As a doctrinal matter, the 1st Circuit is likely right. But the weird situation of inmates shows how delicate the theory of virtual representation used in the Evenwel case really is.
Why, exactly, should nonvoters be allowed to be counted in districting? If the reason is that they are de-facto citizens because of their integration into the community, then prisoners really shouldn’t count, as the district court thought.
If, in contrast, the reason is that you should be counted if you are breathing in a district, then the 1st Circuit is correct. If that is so, then representation is just a mechanical relation between presence and numbers. Maybe it’s an inherent right that has no connection to whether you take part in community life. That would (partly) explain why a slave was counted as three-fifths of a person in the Constitution despite being excluded from political citizenship.
If that’s what representation is, the Supreme Court should have said so. That it hasn’t suggests that the problem of who counts in districting, and why, hasn’t been adequately resolved.
This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.