As the 2020 census approaches, various controversies have emphasized the importance of census methodology on political representation. One impactful methodology is referred to as “prison gerrymandering.” The census counts incarcerated individuals as residents of the district in which they are imprisoned, as opposed to the district in which they permanently reside. The vast majority of states then use these numbers to apportion state legislative districts. In doing so, states tend to enhance the voting power of those in prison districts, as the residency count is artificially inflated by the inclusion of the prison population. The opposite is true for the communities where incarcerated individuals lived before prison, which lose political clout. This method of counting has led to considerable concern over the constitutionality of “prison gerrymandering” and its distorting effects on voting power and representation.
Recently, in NAACP v. Merrill, the Second Circuit ruled that plaintiffs could proceed with their challenge to prison gerrymandering in Connecticut’s state redistricting plan, rejecting the state’s arguments that Eleventh Amendment immunity precluded suit. This ruling represents a novel win for opponents of prison gerrymandering for two reasons. First, questions of state immunity have not previously been addressed directly because the key constitutional litigation has occurred at the city or county level. Second, the ability to challenge statewide policies meets the harm of prison gerrymandering where it exists: as a statewide dynamic wherein the political power of largely white populations is enhanced at the expense of urban areas densely populated by people of color. That said, given the mixed results of constitutional challenges to local plans, success on the merits is far from guaranteed as this case moves forward.
In 2011, Connecticut adopted a redistricting plan to be used for both the 2018 and 2020 elections. The state’s plan counts incarcerated individuals in the locality in which they are imprisoned, as opposed to the one in which they resided pre-incarceration, despite state law that says one does not lose residency in one’s hometown when “absent therefrom in any institution maintained by the state.”
In 2018, the NAACP and several Connecticut residents adversely affected by the districting plan brought suit in the District of Connecticut, arguing that the redistricting plan violates the “one person, one vote” principle embodied in the Fourteenth Amendment. Plaintiffs sought declaratory relief as well as an injunction preventing use of the plan in the upcoming 2020 elections. The state filed a motion to dismiss, which the district court rejected. The state appealed.
The Second Circuit affirmed in part and remanded the case. Under the Eleventh Amendment, states are largely immune from suits brought against them. But the Second Circuit held that this case fell within the Ex Parte Young exception. Per this exception, if plaintiffs can claim an ongoing violation of federal law, and this claim is neither insubstantial nor frivolous, they can sue a state as long as the relief requested is prospective. The second requirement of the Ex parte Young exception—that only prospective relief can be sought—was not disputed here.
Defendants made three arguments that the claim was insubstantial: First, states are presumptively allowed to have population deviations of up to ten percent between the largest and smallest voting districts, and the plaintiffs had failed to make the requisite prima facie showing of discriminatory intent needed to overcome that presumption. The court rejected this on the ground that such prima facie showing is not required at the pleading stage. Second, defendants argued that redistricting is a state political determination that federal courts may not intrude upon. The court rejected this as well, holding that legislative maps can be challenged in federal court over concerns about racial discrimination. Third, the defendants cited to a First Circuit case, holding that a similar case presented “insubstantial” claims. Here, the court simply noted that the First Circuit’s decision might be persuasive in the Second Circuit, but was certainly not binding. The Second Circuit concluded that federal courts had subject matter jurisdiction over the case.
The panel then rejected the defendants’ argument that the case presented a nonjusticiable political question, rejecting the analogy to Rucho v. Common Cause, in which the Supreme Court held that partisan gerrymandering claims are nonjusticiable.
Finally, the panel observed that the district court judge had erred in ruling on the motion to dismiss for failure to state a claim, although it acknowledged that it did not have jurisdiction to correct the error and was merely providing guidance. Thus, the Circuit affirmed in part and remanded the decision to the district court to in turn refer the case to a three-judge court.
This case appears to the be the first time a federal court of appeals has affirmed the ability to sue a state for its use of prison gerrymandering, indicating that this particular form of gerrymandering presents a substantial enough question regarding a violation of federal law to survive scrutiny at this initial phase. Prior cases involving prison gerrymandering had involved local governments, which do not enjoy Eleventh Amendment immunity. Thus, the issue did not come up, for example, when plaintiffs sued Jefferson County, Florida for its use of prison gerrymandering in redrawing county districts.
The ability to sue at the state level is especially important because, as others have pointed out, the harms of prison gerrymandering are clear only the state level. Cases challenging county or city-level districting plans focus on the inflated political relevance of voters in districts with prisons, as their district’s numbers are artificially inflated with nonvoting incarcerated people. But a local focus obscures the symmetric harms of prison gerrymandering: voters of one district gain an outsized voice in the political realm while the voices of voters elsewhere are diminished. This dynamic most often takes the form of amplifying of the power of residents of rural, white districts where prisons sit, while diminishing the resources of residents of the urban districts where incarcerated individuals, who are largely people of color, tend to be from.
This case demonstrates the necessity of zooming out to the state level to convey the harms of prison gerrymandering. The complaint describes how many of Connecticut’s correctional facilities are in adjacent towns near the Massachusetts border. Yet many of those incarcerated are from some of Connecticut’s biggest cities, like Bridgeport or New Haven, which are located in southern Connecticut. Thus, without being able to make out a statewide claim, the plaintiffs couldn’t have challenged this element of vote dilution.
That the Second Circuit declined to foreclose this claim represents one step forward for those looking to comprehensively combat the effects of prison gerrymandering. That said, as this case moves to arguments on the merits, challenges to prison gerrymandering at the local and county levels will become more relevant, as both have advanced similar substantive arguments as are being made here, with mixed success. The Second Circuit referenced this precedent when it conceded that a First Circuit case holding prison gerrymandering constitutional, though not dispositive at this stage, may be persuasive authority later on. Thus it is far from a guarantee that Connecticut’s 2020 election cycle will be rid of prison gerrymandering’s distortive effects.