Prison Gerrymandering Distorts Representation

Prison gerrymandering — the practice of counting incarcerated people at their detention facilities rather than their homes during redistricting — distorts political representation, inflating the clout of active voters living in districts with prisons at the expense of the communities from which incarcerated people come.¹ This report assesses the level of distortion in state lower-house maps due to prison gerrymandering and illustrates how power could shift if all people were instead counted at their homes.

Prison gerrymandering is an artifact of the U.S. Census Bureau’s practice of counting all people incarcerated on Census Day at their detention facility; this count is used to create district plans that remain in place for 10 years, even if a person is being held only temporarily. There are legitimate reasons for tracking where people are incarcerated, but whether political power ought to be assigned to those numbers is a separate matter. Most people in state prisons serve short sentences, return home upon release, and while confined receive constituent services from their home representatives, highlighting the importance of counting incarcerated people at their homes for purposes of equitable representation.²

To analyze the potential distortion of prison gerrymandering, this study considers thousands of redistricting simulations across 11 states, comparing results using traditional census data with the results when incarcerated people are reallocated to their homes. It finds that counting prison populations in their home districts has the potential to result in a cumulative 14 additional Black-majority districts across 8 of the 11 states in the study, with an increase of 6 Black-majority districts in Georgia alone.³ Some states could also see additional Latino-majority districts, although that outcome is less certain. Because these data account only for state prisons — not local jails or federal prisons — the estimates are necessarily conservative.

Prisons are disproportionately located in white and rural areas, despite incarcerated people overwhelmingly coming from urban communities of color, which endure the highest levels of policing.⁴ These communities lose the most political voice to communities with prisons. We know that rural districts lean conservative and urban areas lean liberal, so when urban areas pick up people from incarcerated populations being reallocated back to their homes, we expect that to be reflected in the candidates who are elected. To that end, we assess the effect on Democratic districts (recognizing that the impact would be converse for Republican districts).⁵ This change in population does not necessarily translate to additional Democratic seats, however; instead the simulations create urban districts that have more concentrated Democratic populations.

Thirteen states ended prison gerrymandering for state legislative districts for the 2020 redistricting cycle, and three more are poised to do so in time for the 2030 cycle.⁶

Counting people where they are imprisoned instead of at their homes reinforces a perception that prison is their rightful location. Undoing prison gerrymandering in the remaining states is a crucial step toward ensuring equitable political representation.