Despite what you may see in the news, the fastest-growing trend in redistricting isn’t the adoption of redistricting commissions or the use of public, easy-to-use mapping software.
Instead, it’s the way prison inmates are treated by states when redrawing district lines.
Most states, following the U.S. Census Bureau’s “usual residence” rule, count people where they sleep, meaning inmates are counted as residents of the districts where they are incarcerated, not of their home communities. Because inmates are not voters in the districts where prisons are located, some argue that means they are not represented anywhere.
How does this matter for redistricting? A large prison in a small district such as a municipal ward or even a small legislative district can create what some see as a representational imbalance.
Policies to address that are known to some as “ending prison gerrymandering”; others call it, more neutrally, “inmate data reallocation.”
Here’s a hypothetical example of an imbalance, from the new NCSL report Inmate Data Reallocation in the 2020 Redistricting Cycle:
Imagine, for example, a city with 10,000 residents divided into five wards of 2,000 people each. Each ward elects one member to the city council. The city has no children or noncitizens living in it but is home to a medium-sized state prison with 1,000 inmates, making 1 in 10 city residents an incarcerated person who cannot vote. None of the inmates in the prison are from the city. The prison is located entirely within the 1st Ward, with all 1,000 inmates counted as residents for redistricting purposes. Because there are only 1,000 eligible voters in the 1st Ward and 2,000 eligible voters in each of the other four wards, the relative power of a 1st Ward voter is twice as strong in determining municipal electoral outcomes. An inmate data reallocation policy would seek to “move” all 1,000 inmates out of the city by changing their location in the redistricting data set to wherever they lived before incarceration. This would leave the city with 9,000 residents who would be divided into five wards of 1,800 people each, equalizing the power of the city’s voters.
To prevent such a scenario from occurring, legislators began introducing bills two decades ago to change, or “reallocate,” inmate data for redistricting purposes. In 2000, no states did this. In the 2010 cycle, two states (Maryland and New York) adopted the policy, assigning inmates to the places where they lived before incarceration. By 2020, 11 more states had adopted the policy—California, Colorado, Connecticut, Delaware, Montana, Nevada, New Jersey, Pennsylvania, Rhode Island, Virginia and Washington. Illinois will join them in 2030.
Thoughtful people can and do disagree on whether inmate data reallocation is theoretically appropriate or practically significant. For those who think reallocation is the right thing to do, a host of questions follow on how, exactly, to implement the plan. For instance, can redistricting staff get data from the state’s criminal justice system on where inmates lived before conviction? What’s the process for cleaning those addresses? Is it possible to geocode all addresses? What’s to be done when the race categories used by prisons differ from those used by the Census Bureau? Will this added step delay the redistricting process?
According to the Inmate Data Reallocation report, staff in the states that reallocated data in 2020 wished they had more guidance, statutorily or otherwise, on how to address the many issues related to the policy. The report doesn’t give answers, but it lays out the administrative issues any state will face if it embarks on reallocation and digs deep into states’ experiences with the new policies and their recommendations for the future.
With most of the 13 states implementing reallocation for the first time, staff had no historical analogs to rely on, requiring them to troubleshoot problems and coordinate with their states’ counterparts in the criminal justice system. To capture that institutional knowledge while still fresh, NCSL interviewed staff in all of the 13 states to learn what worked, what didn’t and what could be improved in the future. Their recommendations are the core of the new report and will no doubt provide valuable to lawmakers as they consider legislation for 2030 and beyond.
Ben Williams is a program principal in NCSL’s Elections and Redistricting Program.