A federal district court handed down a memorandum and order for summary judgment in Davidson v. City of Cranston, a case concerning "prison gerrymandering" in Rhode Island. The court concluded that the city improperly drew districts that included all of the incarcerated individuals in a prison into a single district, distorting representation and voting strength in other districts.

One might have concluded that the Supreme Court's recent decision in Evenwel v. Abbott mandated this case come out the other way. There, the Court permitted Texas to use total population in drawing its districts, even if it included non-citizens (i.e., non-voters) in its population base. Had the case come out the other way and some voter- or citizen-based measure been required for drawing districts, prison gerrymandering may well have ended.

As Adam Liptak and others noted, a win for Texas in Evenwel would deeply undermine constitutional arguments against prison gerrymandering. After all, prisoners are people ineligible, just like non-citizens or children. They're drawn into districts with the rest of the total population. Instead, one must come up with a political theory for excluding this set of non-voters from redistricting, but not other sets of non-voters.

In a recently-published article in the Harvard Journal of Law & Public Policy, Perpetuating "One Person, One Vote" Errors, I highlight the deep problems that arise when courts attempt to insert ever more-detailed theories of political representation into the constitutional doctrine of "one person, one vote." Mercifully, I remarked, the decision in Evenwel leaves some discretion to the states (and cities) in redistricting.

But the decision in Davidson takes away some of this discretion. And it does so using bizarre support from the Three-Fifths Clause. It even suggests prisoners are less worthy of representation than slaves.

To start, one has to make a pair of logical inferences to get from the Court's theoretical findings in Evenwel (and its predecessors) to this case. As I explained at the George Washington Law Review Docket, the Court linked the theory of apportionment of congressional representatives to the reason for drawing equal numbers of people in congressional districts, and then linked that finding to the reason for drawing equal numbers of people in other non-federal districts. It opens with this theory of apportionment--how many congressional representatives should each state get? As the Davidson court explained:

In its review of the drafting history of the Constitution and later the Fourteenth Amendment, the Supreme Court repeatedly stresses the prevailing view that women, children, slaves, tax-paying Indians, and non-land-holding men (in some areas) all deserved representation - though none of these groups could vote.

That syncs with the Court's holding in Evenwel:

In other words, the basis of representation in the House was to include all inhabitants - although slaves were counted as only three-fifths of a person - even though States remained free to deny many of those inhabitants the right to participate in the selection of their representatives. Endorsing apportionment based on total population, Alexander Hamilton declared: "There can be no truer principle than this - that every individual of the community at large has an equal right to the protection of the government."

One must appreciate a serious concern here--the Three-Fifths Clause. By including slaves (albeit fractionally) in the apportionment base, the South would receive dramatically more political power in the House than if slaves counted as zero-fifths.

But one should not seriously claim that the Three-Fifths Clause somehow means that slaves were virtually represented by their masters--that voters in the South somehow were protecting slaves through their votes! Instead, as I've pointed out in another piece, one justification commonly raised by the Founders for including slaves in the apportionment base was a notion that it was a proxy for wealth. Population (and slave total) meant roughly the wealth in an area; and more wealthy (or more "productive") areas should have more political power.

Such political theory likely would not garner much support today--indeed, it has been expressly rejected in the "one person, one vote" line of cases. (This is also a major reason Justice Alito wrote separately in Evenwel--he viewed the apportionment question as distinct from the question of drawing districts.) But the only way for "one person, one vote" to work is if one adopts a kind of theory of "virtual representation," where the elected representatives serve the entire population of the region, regardless of whether they are eligible to vote. (Indeed, the Court in Evenwel expressly made this point concerning children.)

The district court's next findings, then, are fairly breathtaking:

The inmates . . . share none of the characteristics of the constituencies described by the Supreme Court. They don't have a stake in the Cranston public school system and they are not receiving constituent services, such as help with public-benefits bureaucracies. They are not making requests of and suggestions to Cranston elected officials (or if they are, they are receiving no response), nor are they receiving "the protection of government," at least not from Cranston elected officials.

Really? First, descriptively, this is false at almost all levels. A quick look at Kramer v. Union Free School District No. 15 offers myriad ways (in the right-to-vote context) someone can "have a stake" in the public school system--merely being "interested in" the school board as a resident may be enough, or the note that "[a]ll members of the community have an interest in the quality and structure of public education," which may have "grave consequences to the entire population." Prisoners may write letters to representatives--even if they are not, they are able to do so. It may well be that these are still pretty weak reasons for including prisoners in a district; but that is a far cry from saying that they are constitutionally forbidden from being included in a district.

Second, it rather brazenly suggests that eighteen-century slaves were better represented, and better deserving of representation, than twenty first-century prisoners. How else is one supposed to construe the phrase that "the inmates . . . share none of the characteristics" of "slaves"? It turns the Three-Fifths Clause into a rather favorable view of representation. Justice Alito's concurring opinion in Evenwel is perhaps wise guidance on the Three-Fifths Clause. But its failure to carry a majority of the Court leads to curious decisions such as this.

It may well be the case that Rhode Island and other states should include incarcerated persons in the districts where they last resided. It makes sense to me, as a political matter. As an administrative matter, it also seems to be sensible to include the easy total of the prisoners in the prison where they reside. Indeed, that's how the Census counted them in 2010! It becomes a fairly significant task, then, to create a new population total for redistricting, one that the Census Bureau does not provide. (Indeed, this was a major claim raised in Evenwel.)

But for a federal court to make the claim that slaves were favorably represented in apportionment while prisoners ought never be included in redistricting is a deeply problematic claim. The "one person, one vote" line of cases cannot sustain such decisions of representation equality or political theory, which are best left to the political process.