Because of their relatively small size, city and county governments often experience the most distortive impacts of prison gerrymandering. When drawing districts in these communities, a single prison can easily account for an entire city council or county commission district, making it impossible for officials to draw fair districts. So it should come as no surprise that, while state government efforts to address the practice receive most of the attention, the movement to end prison gerrymandering started and has been most successful at the local government level, in mostly small, rural, conservative cities and counties. The 2020 redistricting cycle continued this pattern.
Prison gerrymandering is a problem created because the Census Bureau incorrectly counts incarcerated people as residents of their prison cells rather than their home communities. As a result, when states and local governments use Census data to draw new political boundaries, they inadvertently give residents of districts with prisons greater political clout than all other residents. Progress in addressing this problem has been so swift in recent years that the National Conference of State Legislatures recently called the movement to end prison gerrymandering “the fastest-growing trend in redistricting.”
The distortions caused by prison gerrymandering have real-world implications in these communities. It gives some residents a louder voice in the political process, which means local government often prioritizes the issues they want addressed mdash; whether it be deciding which potholes get filled, where to expand sewer services, or which parks to improve — over the needs of residents who live in all other parts of the town or county.
Hundreds of local governments ended prison gerrymandering on their own, with still new ones joining their ranks this decade
In addition to the hundreds of cities and counties that avoided prison gerrymandering after the 2000 and 2010 Censuses (decades when zero and two states, respectively, tackled the issue), we identified an additional 21 local governments1 scattered across 10 states that addressed the problem for the first time after the 2020 Census.2
Local governments that ended prison gerrymandering for the first time after the 2020 Census
State
Local Government
Arkansas
Izard County
Arkansas
Jackson County
Delaware
City of Willmington
Florida
Columbia County
Florida
Jefferson County
Georgia
Coffee County
Georgia
Emanuel County
Georgia
Long County
Georgia
Monroe County
Georgia
Montgomery County
Georgia
Wayne County
Louisiana
Allen Parish
Louisiana
Catahoula Parish
Oklahoma
Atoka County
Oklahoma
Woodward
Rhode Island
City of Central Falls
North Carolina
Granville County
South Carolina
Williamsburg County
Tennessee
Hardeman County
Tennessee
Johnson County
Tennessee
Morgan County
It is important to note that this does not represent the complete list of communities that addressed prison gerrymandering; it is a list of places we identified that have done so for the first time this decade — and an incomplete list at that. That’s because many of the laws passed in recent years ending prison gerrymandering in state legislative districts also apply to city and county districts. For example, California’s law, which ended the practice, applies to local governments, too, meaning the communities containing the 34 state prisons and 114 local jails also avoided the practice. As a result, our analysis certainly underrepresents the number of local governments that no longer take part in prison gerrymandering.
Local governments got creative to avoid prison gerrymandering
To avoid prison gerrymandering, communities found creative solutions to overcome the Census Bureau’s flawed counting methods. Regardless of which method they employ, they all led to the same result: districts more representative of the community’s actual population.
There are several common ways local governments avoided prison gerrymandering:
Adjusting their redistricting data to remove the out-of-town population counted at the prison;
“Overpopulating” the district with the prison so that it contains the same number of true residents as every other district in the city/county;
Cutting holes in the maps around the prison to exclude the prison population altogether;
Trying to split prison facilities across multiple districts to even out the impact.
In addition, a handful of communities avoid prison gerrymandering by forgoing redistricting altogether, a particularly common practice when a new prison has been built in the area since it last drew new district lines. Communities can maintain equal representation without redistricting if their population outside of the prison has not changed. We’ve discovered a number of places that have chosen not to redistrict because the only significant change in the total Census population is the correctional population.
It is important to remember, though, even when local governments can solve their own prison gerrymandering problems, their jurisdiction ends at their city or county line. While they can ensure that prison isn’t included in their redistricting data, they can’t actually count incarcerated people in their home communities, meaning those people often aren’t included anywhere for redistricting purposes.
Some local governments face unnecessary obstacles
Even as more governments take on the task of ending prison gerrymandering in their local districts, some face arcane state rules that turn these simple reforms into complex mazes.
In Wisconsin, for example, leaders in Juneau County recognized prison gerrymandering was a problem and wanted to fix it, so they sought guidance from the state on how to do that but were told their hands were tied. They learned an outdated and confusing 1981 attorney general’s opinion directed local governments to draw new district lines without excluding incarcerated people. It requires local governments to take part in prison gerrymandering, even though they didn’t want to. Wisconsin’s attorney general can and should issue a new opinion that gives local governments the freedom to finally fix this problem, but until he does, their hands are tied.
Rural municipal and county districts still suffer from the largest impact
Despite this progress, there are still some communities that face huge impacts of prison gerrymandering.
Local governments’ efforts to end prison gerrymandering are sometimes stymied by state governments, such as in Wisconsin. But even when state governments reverse course to support avoiding prison gerrymandering, some counties stay adrift on their previous heading. For example, since 2016, Tennessee has explicitly allowed (Tenn. Code Ann. S 5-1-111) counties to deviate from the Census population to avoid prison gerrymandering. Since then three counties (Hardeman County, Johnson County, Morgan County) that had engaged in prison gerrymandering have stopped doing so. Unfortunately, most counties impacted by prison gerrymandering in the state continue to use unadjusted Census data for redistricting and, as a result, draw some of the most unequal districts in the country.
As a result, Tennessee, along with Wisconsin — which, as we explained above, prohibits local governments from ending prison gerrymandering — account for six out of the eight cities or counties in the country where correctional facilities make up at least half of local government district
Cities or counties where incarcerated people account for more than 50% of the population
County or City
State
Percent of the population in a City Council or County Commissioner district counted at a correctional facility
Lake County
Tennessee
86%
Jackson County
Wisconsin
83%
Wayne County
Tennessee
79%
Juneau County
Wisconsin
79%
Mansfield City
Ohio
66%
Adams County
Wisconsin
64%
Marion City
Ohio
61%
Lauderdale County
Tennessee
56%
Rural residents would benefit from the Census Bureau solving the problem
Despite the progress made in recent years to address the problem, residents of rural communities continue to be disproportionately harmed by prison gerrymandering and stand to uniquely benefit if the Census Bureau were to solve the problem.
There are three reasons for this:
As we explained in this briefing, because of their small size, rural cities and towns that contain prisons often have their local government districts extremely distorted by prison populations.
Rural state governments have been slower to adopt statewide reforms to address prison gerrymandering. As a result, our analysis earlier this year found most of the worst state legislative prison gerrymanders are located in rural areas. This analysis also shows the issue defies many preconceived notions about the partisan impact of the problem, with Democrats controlling six of the most gerrymandered districts, while Republicans control four.
Mass incarceration is no longer just a problem in urban areas. Our analysis of where incarcerated people come from found that rural areas often have the highest incarceration rates in their state. These increasing incarceration rates mean that while some rural communities see their political boundaries dramatically distorted by prison gerrymandering, others lose political clout because of the problem.
If, in 2030, the Census Bureau chose to count incarcerated people in their home communities instead of in prison and jail cells, all three of these problems would be solved for rural communities. This change would save these communities time and money, ensure they didn’t lose political clout to prison gerrymandering, and help communities that contain prisons create local government boundaries that accurately reflect their true population.
Looking toward 2030
Roughly half of all U.S. residents now live in a city, county, or state that has ended prison gerrymandering. Even as more local governments and states address the issue, they can only do so much. Prison gerrymandering is a national problem that crosses jurisdictional boundaries. No city, county, or state can completely fix this issue on its own. Only the Census Bureau can fully solve this problem nationwide. As it begins to plan the 2030 count, it should finally listen to the growing consensus among state and local governments and end prison gerrymandering across the country.
Keep reading for an explanation of why this number undoubtedly underrepresents that number of communities that addressed the problem for the first time after the 2020 census. ↩
Rhode Island’s adjusted data did not account for Wyatt Correctional Facility, located in Central Falls. So the state’s partial data adjustment was not usable for their local redistricting. ↩
2030 Census Advisory Committee's inaugural meeting ended in recommendations that future agenda include discussions of the Census Bureau's approach to counting incarcerated people in the 2030 Census.
On Friday the 2030 Census Advisory Committee held their inaugural meeting and voted to put prison gerrymandering on its agenda. (That’s actually more exciting than it sounds.) The recommendation to discuss the topic in depth was both in reaction to a public comment submitted to the Committee by the Leadership Conference on Civil and Human Rights on behalf of its prison gerrymandering working group, as well as Committee members’ own expertise identifying prison gerrymandering as an issue that needs to be addressed for the 2030 Census.
The topics of discussion for the Committee’s first meeting were largely (and procedurally) limited to on hard-to-count and historically undercounted populations, but committee members Jeri Green (National Urban League’s 2020 Census Black Roundtable) and Ben Williams (National Conference of State Legislatures) requested that the committee address two issues related to prison gerrymandering at a future meeting:
The Census Bureau’s residence criteria for incarcerated people, which is based on an outdated interpretation of the Bureau’s “residence rule”, and means that people are counted as if they were residents of a correctional facility.
The 2030 Census Advisory Committee is charged with “advis[ing] the Census Bureau in conducting an accurate decennial census.” And it is in that role that it recommended that the “a future committee meeting include a discussion on residence criteria for those who are incarcerated. … [And] that there be a future discussion regarding the impact of disclosure avoidance policies on the 2030 Census.” These issues accounted for two out of the three topics recommended for discussion by the Committee.
The 2030 Census Advisory Committee operates in an advisory capacity to the Census Bureau, and as such their focus on prison gerrymandering sends a strong message to the Census Bureau that it must consider how to count incarcerated people at home in the 2030 Census.
Last week, The Leadership Conference on Civil and Human Rights — a coalition of more than 240 national organizations to promote and protect the rights of all persons in the United States — sent a letter to the Census Bureau urging it to end prison gerrymandering. Specifically, the group called on the Bureau to “update the current residence rules to maintain the accuracy of the census, ensure fair representation of incarcerated individuals within their home communities, and to prioritize equity among all residents regardless of their incarceration status.”
Prison gerrymandering is a problem created because the Census Bureau incorrectly counts incarcerated people as residents of their prison cells rather than their home communities. As a result, when states use Census data to draw new state or local districts, they inadvertently give residents of districts with prisons greater political clout than all other state residents.
The letter follows up on a meeting with staff from the Census Bureau’s Population Division, which oversees how the “residence rule” — is applied to people in different living situations at Census time. To end prison gerrymandering, the Bureau needs to update how it applies its residence rule for incarcerated people to ensure an accurate Census. While is a plethoraof reasonsit should do so, the letter focuses on 5 key points:
The residence rules must reflect the real-life circumstances of incarcerated persons and the needs of the states and communities in which they live.
Representation and, therefore, democracy, are distorted by counting incarcerated persons where they are housed on Census Day rather than in the home communities to which most will return.
Incarcerated people are not part of the social and economic fabric of the communities surrounding the facilities where they are serving their sentences.
Criteria used to apply the residence rule to incarcerated people should be applied in the same manner as for other people away from home on Census Day.
Equity requires counting incarcerated people at home.
For details on all 5 points, check out the letter (or PDF).
We explore state redistricting reports to show that, despite facing immense challenges, states were remarkably successful ending prison gerrymandering after the 2020 Census.
During the 2020 redistricting cycle, more than a dozen states took it upon themselves to do what the Census Bureau has refused: end prison gerrymandering. Through their redistricting process, they worked to unwind the flawed way the Bureau counts incarcerated people — as residents of a prison cell — and instead count them where they actually live, creating legislative districts that more accurately reflect actual resident populations.
Prison gerrymandering is a problem created because the Census Bureau counts incarcerated people as residents of a prison cell rather than at their true communities. When states then use Census data during their redistricting process, they artificially inflate the population of prison communities and give them a larger voice in government.
It is an important question with huge implications; 19 states are already poised to adjust their redistricting data to address prison gerrymandering in the 2030 redistricting cycle. These questions can help states evaluate how they can do better during the 2030 redistricting cycle. It can guide states that haven’t yet adopted anti-prison gerrymandering laws but are considering them. Finally, it can illustrate how states are limited in their ability to solve this problem in ways the Census Bureau is not.
To answer these questions, we reviewed redistricting reports and data from 13 states that addressed prison gerrymandering after the 2020 census to understand how many people in state prisons they attempted to count in their home communities and how many they were able to reallocate successfully.
Perfection isn’t possible, so it’s not the goal
Before diving into our analysis, it is important to remember that it is impossible and unrealistic for states to successfully reallocate everyone in their prisons to their home communities.
There are three primary reasons why a state can’t count an incarcerated person at their home address:
The person’s last address is in a different state: States don’t have the power to reallocate people across state lines, so they cannot count people from out of state in their home community.
The state has decided not to reallocate an incarcerated person for a policy-based reason: Unfortunately, some states have implemented arbitrary limitations on their reforms by excluding certain incarcerated people from being counted in their home communities. For example, Rhode Island did the mapping work for all people incarcerated in the state correctional facilities. However, it chose to count incarcerated people with more than two years left on their sentence at the prison location in the redistricting data. While we and other organizations think this is misguided, it is the choice policymakers have made, so redistricting staff could not count them in their home community.
Erroneous or missing data: State redistricting officials often rely on their Department of Corrections (DOC) to get address data for incarcerated people. Generally, that data contains everything they need to implement their reforms. However, sometimes data is missing or erroneous. Apart from typos and other record-keeping errors in home address data, there are a few situations where home addresses are missing altogether. For example, if an incarcerated person indicated they were unhoused before coming to prison, the DOC will likely leave the address field blank or note “homeless.” This may be an appropriate response for prison staff, but it creates problems for redistricting officials.1While there are some steps redistricting officials can take to reduce this number, it often takes time, which is something they don’t have much of.
These limitations make it impossible for a state to reallocate 100% of people in its prisons to their home communities, so when they evaluate their success, they shouldn’t grade themselves against perfection. Instead, they should ask themselves: “Of the people it was possible to reallocate, how many did we reallocate?”
How did states do?
For our analysis, we looked at two key numbers:
The state’s success rate for reallocating people that it was possible to count at home.2
The total percentage of people in state prisons counted at their home address.
Our analysis 3 shows that most states did remarkably well. This is noteworthy because 2020 was a notoriously difficult redistricting cycle, in which Census data was delayed, and many states adopted prison gerrymandering reform laws late in the decade.
Of the 13 states we analyzed, 10 published their methodology. Nine out of 10 were able to reallocate at least 80% of people with addresses that met the states’ criteria, with Nevada being the lone exception. Impressively, California successfully mapped every address that it attempted.
2020 redistricting state reallocation success rates
State
Success rate for reallocating people who’s address met the state’s criteria
Percentage of people in state prisons successfully counted at their home address
California
100.0%
99.7%
Colorado
88.5%
82.1%
Connecticut
96.9%
87.6%
Delaware
85.5%
79.2%
Maryland
84.8%
77.0%
Montana
96.3%
45.3%
Nevada
68.9%
64.1%
New Jersey
95.3%
89.5%
New York
91.9%
91.9%
Pennsylvania
State did not publish methodology
60.6%
Rhode Island
83.1%
41.0%
Virginia
State did not publish methodology
88.2%
Washington
State did not publish methodology
91.4%
States were so successful that even in two of the three states where we could not fully analyze the success rate because they didn’t publish their methodology — Virginia and Washington — we can tell that roughly 90% of their total state prison population was successfully reallocated to their home communities. Importantly, like Rhode Island (which we explained above), Pennsylvania also made a policy choice to exclude certain incarcerated people from these reforms, instantly lowering the portion of its prison population that could be counted at home.
Getting to 100%
While states have done a remarkable job implementing these reforms, every incarcerated person should be counted in their home communities for redistricting purposes. What can government officials do to accomplish this?
Start collecting, assessing, and cleaning data early: Getting good data from the Department of Corrections is critical to successfully implementing reforms. That’s why state redistricting officials should work with that agency now to ensure it is collecting the data they’ll need during the next redistricting cycle. Similarly, states that haven’t yet addressed prison gerrymandering should adopt and implement policies to do so now. This will ensure prisons collect and maintain high-quality address data for people they incarcerate.
Don’t adopt arbitrary limitations on reforms: When adopting anti-prison gerrymandering reforms, states should include all incarcerated people, regardless of the length of their sentence. Even if a person is likely to spend decades behind bars, they’re still not a resident of the legislative district that contains the prison4 — they don’t have ties to that community, they don’t call it home, and if they have a problem that they want lawmakers to fix they’re not likely to call the legislator who represents the prison but rather the elected officials from their home community — so the town containing the prison shouldn’t get an outsized voice in government simply because the facility happens to be located there.
The Census Bureau must change its policies: Even if states do everything in their power to address prison gerrymandering, there will still be some incarcerated people that they can’t count in their home districts, most frequently those who come from different states. Similarly, but not covered in this analysis, if a state contains a federal prison, redistricting officials cannot reallocate those incarcerated people to their home communities. The Census Bureau is the only government entity with the power to address this problem completely and across state lines.
This analysis shows that states have been incredibly successful in implementing reforms to address prison gerrymandering. History suggests that if the Census Bureau fails to resolve this problem, states will do even better during the 2030 redistricting cycle. This should give confidence to places considering similar reforms and motivate them to take action now.
We measured the states’ success in two different ways:
Success rate for reallocating people who met the state’s criteria for reallocation
Total percentage of people in state prisons successfully counted at their home address
It is worth noting that we only provided partial analyses for three states that have ended prison gerrymandering for the following reasons: Pennsylvania and Virginia have not made their methodologies publicly available; Washington has some documentation available, but it does not provide the number or percentage of people successfully reallocated, or the total prison population the state began with.
Our state-specific methodologies are below.
California
The state documentation reports a total of 122,730 people in state prisons across the state (MacDonald presentation, p. 4). Ultimately, the adjusted dataset includes 122,393 people in state prisons who were successfully reallocated home in California (Statewide Database). Based on the 122,393 successfully reallocated people, we can assume that the remaining 337 people had a last known place of residence outside of California or an address that “cannot be determined” (Final Maps Report, p. 20). These people were counted in an “unknown geographical location” and excluded “from the population count for any district, ward, or precinct” (Final Maps Report, p. 20).
State documentation:
Final Maps Report from the California Citizens Redistricting Commission
The Colorado state documentation reports 17,506 people in prisons across the state. This does not include any people in federal prisons.5
1,270 people in state prisons reported a known previous address outside of the state and, therefore, were counted at the facility (p. 3). 6
That leaves a remaining 16,236 people in prison with Colorado addresses. 1,872 of those people in state prisons with in-state addresses had unusable addresses and were subsequently counted at the facility (p. 3). The state defined “unusable” as addresses with “no or incomplete street address information” or people who reported being homeless before incarceration (p. 3).
The remaining 14,364 people in state prisons could be reallocated home in Colorado (p.3).
The Connecticut state documentation states the Department of Corrections reported a total of 11,853 people in state prisons across the state (p. 4). An additional 1,057 people were in federal prisons and subsequently counted at large (p.4).
392 records had incomplete or unusable addresses and were subsequently counted at large (p. 5). The reasons for these 392 unusable addresses were listed as: 301 records did not include a street address, 75 records did not have a street number, four records listed a PO box in the street address field, eight records listed “an apartment complex where the precise address location and [Census] block could not be determined,” and four records were for “temporary locations (e.g., hotel, shelter, or campground) where the precise census block location could not be determined” (p. 5). 687 records had an out-of-state address and were subsequently counted at large (p. 4). 7 10 records were for people serving life-without-parole sentences and subsequently counted at the facility. 49 records were for people under 18 years old and were “excluded” (p.3).
Of the remaining 10,715 records, 333 “could not be geocoded in cleaned dataset” (p. 7). 8 The remaining 10,382 people in state prisons were successfully reallocated home in Connecticut (p.6).
The Delaware state documentation reports that a list of records from the Department of Corrections included 4,748 records (p.3). 350 of these records were for out-of-state addresses and were excluded from the counts of the facility and the population totals (p. 3). There were 4,398 remaining in-state addresses. 637 of these records were “without a good geocoded home address” and were counted at the facility (p. 3). The remaining 3,761 people in state prisons could be reallocated home in Delaware (p.3).
The Maryland state documentation reports a total of 19,802 people in prisons across the state (p. 1). This does not include any people in federal prisons. 1,821 of those people had out-of-state addresses and were not counted. That leaves a remaining 17,981 people with in-state residence. Out of those 17,981 people, 2,793 people had “no address” – meaning they did not have “a valid prior address or they did not provide a previous residential address” (p. 1) – and were subsequently counted at the facility. The remaining 15,242 people in state prisons could be reallocated home in Maryland.
The Montana state documentation reports 2,840 people in state prisons (p. 5). 1,505 were deemed “not appropriate for reallocation,” which included: 1,389 records with no addresses, 56 with out-of-state addresses, 28 with addresses in jails or other similar facilities, 20 indicating “transience,” and 12 with P.O. boxes as addresses (p. 5). The remaining 1,335 people in state prisons had addresses deemed “appropriate for geocoding” (p. 2). 1,286 were successfully geocoded, and 49 were not geocoded, therefore excluded from reallocation (p. 2).
The Nevada state documentation reports 12,214 records of people in Nevada state prisons (Killian presentation, p. 3). 860 of these addresses were for out-of-state residences and were not counted (Killian presentation, p. 9). This leaves 11,354 records with in-state addresses. Ultimately, the adjusted dataset includes 7,826 people in state prisons who were successfully reallocated home (Nevada Reapportionment and Redistricting website). Based on the 7,826 successfully reallocated people, we can assume that the remaining 3,528 people had addresses that could not be reallocated. The available documentation does not explain why some addresses could not be reallocated.
State documentation:
October 2021 presentation from Asher Killian, Haley Proehl, and Matthew Lawton on the interim status of the Nevada reallocation process. (Note: This documentation is from October 2021, and subsequently, further work was done to reallocate people in prisons to their addresses. The numbers in this presentation are not the final reallocation counts that are in the adjusted dataset published at the completion of the reallocation process. Those numbers are linked below.)
The New Jersey state documentation reports a total of 18,109 people in state prisons (p. 3). An additional 4,067 people are in federal prisons and were subsequently counted at large (p. 10). 9
Out of the 18,109 people in state prison, 1,114 people had out-of-state addresses. 10
Of the remaining 16,995 people in state prisons with in-state addresses, 797 had in-state addresses that were unusable (p.5). The state defines “unusable” addresses as: residential addresses that were unknown/homeless, incomplete, PO boxes, rural route addresses, or contained an invalid/nonexistent address or street name.
The remaining 16,198 people in state prisons could be reallocated home in New Jersey (p.5).
The New York state documentation (p.2) reports a total of 46,418 people in prisons across the state. 3,926 of those people are in federal prisons and were subsequently counted at large. That leaves a remaining 42,492 people in New York state prisons. The three-step geocoding and address verification process used in New York included the following:
Utilization of the Google Maps Geocoding API 12 to extract the latitude and longitude coordinates from addresses. At this point, geocoding results were accepted when the program confirmed street address precision. These results were then converted to Census blocks.
Reverse geocoding the information generated by the Google Maps Geocoding API latitude and longitude points to a mapped address, and then these results were manually compared to the original addresses. Discrepancies at this stage were investigated using Google StreetView, Bing Maps, and street name lists.
At this point, any remaining “invalid addresses” (i.e., addresses that were not successfully mapped to an address) were investigated manually using Google StreetView, Bing Maps, and street name lists. The state reports that “many of these initially invalid addresses were effectively geocoded” at this point by visually confirming house and building numbers or rectifying street name misspellings. Google StreetView and Bing Maps had poor coverage or limited visibility for some apartment complexes and trailer parks. In these instances, a third source was used, the Cornell University-based New York Block Browser LUCA Evaluation System (NYBBLES), allowing for rooftop geocoding of many individual structures in mobile home parks and apartment complexes.
After these three steps, 3,465 of those people in state prisons still had “invalid” addresses and were subsequently counted at large. Invalid addresses were loosely defined in the documentation (p. 2-3) as addresses that were not able to be geocoded successfully via New York’s three-step process. 39,027 people in state prisons were able to reallocated home in New York.
The Rhode Island redistricting consultant’s documentation reports 2,618 people in Rhode Island Department of Corrections custody across the state (Nov. 15 presentation, slide 13). 691 addresses had “obstacles to geocoding,” which included: 271 records with no permanent addresses, 154 with out-of-state addresses, 146 serving home confinement,11 44 incarcerated out of state, 23 addresses listed at the facility, 10 addresses with no street, and 4 with P.O. boxes as addresses (Nov. 15 presentation, slide 13). The remaining 1,927 people in custody had addresses that were deemed geocoded (Jan. 12 presentation, slide 3).
While 1,927 addresses were successfully geocoded, the Rhode Island redistricting process only reallocated people who were not yet sentenced or had less than two years until their expected release date. Out of all 1,354 people (Jan. 12 presentation, slide 3) in Rhode Island Department of Corrections custody who met that criteria, only 1,074 people were reallocated to their home address.
See our methodology for a full explanation of how we conducted our analysis. ↩
There are obviously some people who actually do live in the same legislative district where the prison they are incarcerated is located. However, our analysis shows this is the exception, not the rule. ↩
The Colorado legislation requires the state to request this data from the federal Bureau of Prisons, but that federal agency refused to share it. ↩
The Colorado documentation states, “Nonpartisan staff will not reallocate incarcerated persons with a previous known address outside Colorado and will instead leave these persons in the state correctional facility census block they are listed under in the DOC report. Section 2-2-902 (4), C.R.S. requires only an incarcerated person with an in-state address (in-state incarcerated persons) to be counted at their last known address. Per the DOC report, there were 1,270 incarcerated persons in a Colorado correctional facility on April 1, 2020 with a last known address outside Colorado” (p. 3). ↩
There were 688 records with out-of-state addresses, but one was also under 18 years old, so that individual is counted in the total count of people excluded because they were younger than 18 years old. ↩
These 333 records could not be “geocoded in the cleaned dataset,” but there is no clear definition of what that means. ↩
Of note, the Federal BOP did not provide the population count of federal prisons in NJ: “As required under N.J.S.A. 52:4-1.2 and N.J.S.A. 52:4-1.3, the Secretary of State requested a report from the NJ Department of Corrections and from agencies that operate a federal corrections facility in New Jersey. The Department of Corrections provided such report, and the Federal Bureau of Prisons advised that it would not provide the requested data because the state’s request is not for law enforcement purposes and thus, did not meet the criteria for permissible disclosure of the data sought.” To address this, New Jersey relied on Census data for the population of the two federal facilities in the state, which is where the federal population count of 4,067 originates. The Census data included an additional 35 people who were not accounted for in the “group facilities count” and New Jersey counted these 35 individuals at large (p. 8). ↩
As the state documentation explains, “for each individual included in the report received from the NJ Department of Corrections pursuant to N.J.S.A. 52:4-1.2 where the individual’s address was unknown or not located within the State, and for all individuals reported in the census as residing in a federal incarceration facility for whom a report was not provided under N.J.S.A. 52:4- 1.3, the relevant population counts reported in the census have been reallocated to reflect that the individual resided at “an unknown address within the State” on the day the census was completed, and the individual is not included in any applicable population counts reported in the census for the geographic units where the individual’s prison facility was located on the day the census was completed.” (p. 2-3) ↩
The state documentation (p. 2) explains this program as “an interface that gives programmers a method to actively send requests to the Google geocoding server through customized scripts and functions in order to obtain geographic coordinates from valid addresses.” (p. 2) ↩
In calculating the success of the geocoding process, we removed the following addresses from the denominator: out-of-state addresses, which are by definition not mappable to a location within Rhode Island, and those listed as home confinement because people in that situation would have had the opportunity to fill out the Census at home and be counted there directly. ↩
With Governor Tim Walz's signature, the state is the latest to reject the Census Bureau’s flawed and outdated way of counting incarcerated people.
May 20, 2024
On Friday, Minnesota Governor Tim Walz signed HF 4772 — an omnibus elections policy bill — into law, officially ending prison gerrymandering in the state. With this action, Minnesota joins the rapidly growing list of states that have taken action on this issue. The measure requires state and local governments to count incarcerated people at their home addresses when drawing new political districts during their redistricting process.
Prison gerrymandering is a problem created because the Census Bureau incorrectly counts incarcerated people as residents of their prison cells rather than their home communities. As a result, when states use Census data to draw new state or local districts, they inadvertently give residents of districts with prisons greater political clout than all other state residents.
“With this law, yet another state has rejected the Census Bureau’s flawed way of counting incarcerated people to ensure its residents have an equal voice in their government,” said Aleks Kajstura, Legal Director of the Prison Policy Initiative and the head of the organization’s campaign to end prison gerrymandering. “Roughly half the country now lives in a place that has ended prison gerrymandering. With so many places taking action on this issue, it raises the question, ‘Will the Census Bureau listen to the growing consensus on this issue, or will it cling to its outdated and misguided way of counting people in prisons and jails in 2030?'”
The provisions ending prison gerrymandering in the state were initially introduced as standalone measures and were rolled into this omnibus election bill.
Avoiding carveouts in prison gerrymandering reforms
Carveouts that exclude certain incarcerated people from bills to end prison gerrymandering undermine their impact and continue to distort democracy.
During the House debate on the measure, a representative introduced an amendment to limit the impact of the reforms by continuing to count people who have served ten consecutive years in prison or have over a decade remaining on their sentence at the prison. Fortunately, this poison-pill amendment was rejected, but it explains why other states considering similar reforms should also avoid this type of carveout.
Misguided amendments like this ignore the reality that, regardless of an incarcerated person’s sentence length, they are still not a member of the prison community. Additionally, nationally 75% of people in prison serve time in more than one facility, while 12% of people serve time in at least five facilities before returning home. These amendments also assume that a person will serve their full sentence and that sentencing policies or release mechanisms in a given state will remain static over a ten-year period, two things that are far from certain.
States shouldn’t undermine their own progress on this issue. When they end prison gerrymandering, they should ensure those reforms include all people incarcerated in the state’s prisons.
Most states have clauses in their constitutions or statutes that explicitly say that prisons are not a residence — whether someone is incarcerated away from home for a few months or a few decades — yet the Census Bureau continues to count people as if they are. The Census Bureau has the authority to change how it counts incarcerated people and officially end prison gerrymandering at the national level, but inaction has forced state and local officials to pass reforms and shoulder the burden of correcting flawed Census redistricting data to count incarcerated people at home.
In addition to Minnesota, eighteen other states — including “red” states like Montana, “blue” states like New York, and “purple” states like Maine — have recognized the impact of prison gerrymandering on political representation and have taken action to provide more equal representation to their residents. Progress on this issue has been so swift that the staunchly bipartisan National Conference of State Legislatures recently called the movement to end prison gerrymandering “the fastest-growing trend in redistricting.”
With Minnesota adding itself to this rapidly growing trend, there is yet another reason for the Census Bureau to finally change how it counts incarcerated people and end prison gerrymandering nationwide.
Download our one-pager for a quick summary of the evidence that shows incarcerated people return home after their release.
Incarcerated people return to their home communities after release. That’s a fact that is obvious to anyone who lives in a community hard hit by mass incarceration. Yet when states seek to end prison gerrymandering by counting incarcerated people at home for redistricting purposes, some people ask why we should use someone’s home address — concerned that they might not return to that exact place after release or even might stay in the prison town.
Unfortunately, because we can’t predict the future and because prisons and jails generally don’t track where people go after their release, this has been a hard question to answer. However, we collected several unique datasets to fill this gap and provide the best evidence possible of where people go after being released from prison. We found:
Incarcerated people don’t generally stay in the area of the prison after their release.
Incarcerated people almost always go back to the communities they came from — if not the exact address they lived at before their arrest.
Counting incarcerated people in prisons is likely the least accurate place to count them.
In theory, every legislative district within a state is supposed to have the same population to ensure everyone has equal representation from elected officials. However, the Census frustrates this goal by counting nearly 2 million incarcerated people as residents of the places in which they are detained instead of at their home addresses. The resulting Census data leads to “prison gerrymandering” — legislative districts that skew representation in favor of people who live near prisons and other correctional facilities.
To avoid prison gerrymandering, states have taken it upon themselves to count incarcerated people at home in their redistricting data. In practical terms, this means counting people at the pre-incarceration address on file with the Department of Corrections.
People don’t stay near the prison after release
There are two widely-accepted truths about mass incarceration:
With this in mind — along with the fact that more than half a million people enter and leave prison every year — you can conclude that if even a small portion of people stayed near the prison after release, those communities would, over time, begin to look similar to those on the inside, but that’s not the case.
In 2015, we released a report that explored the racial geography of mass incarceration. It showed that in 208 counties with prisons, the portion of the free population that was Black was at least 10 times smaller than the portion of the prison population that was Black. These trends were so strong, in fact, that we found 161 counties where the number of incarcerated Black people was actually bigger — by raw number — than the number of free Black people.
If people in prison stayed in those communities after their release, these dramatic disparities would not exist.
People go home after release
Unfortunately, while prison systems generally keep people’s home addresses on file, they don’t track whether they actually return to that address upon release. However, addresses for people on probation and parole can be a reasonably accurate proxy to fill this gap. That’s because a significant number of incarcerated people are on probation or parole after their release.
Comparing these two sets of addresses, we can get a sense of whether people return to where they came from after incarceration. If the portion of incarcerated people from an area is similar to the portion of people on probation or parole from that area, you can reasonably conclude that people are returning to their home communities after incarceration.
Rhode Island serves as a useful example for this analysis because all of its correctional facilities, including pre-trial facilities, are located in one correctional complex (in the city of Cranston). In most states, people are held pre-trial in jails, usually close to home, but in Rhode Island, they are incarcerated equally far from home, whether they are held for a few days or a few years.
Percentage of people incarcerated or on probation/parole in Rhode Island’s five largest cities
Municipality
Percentage of incarcerated people from the city
Percentage of people on probation or parole from the city
Providence
38.2%
33.0%
Pawtucket
11.9%
11.2%
Woonsocket
8.1%
8.2%
Cranston
7.6%
6.2%
Warwick
4.6%
5.4%
Examining this data for the five largest cities in the state, you see that the numbers align quite closely, reinforcing the fact that incarcerated people almost always return to their home communities after their release.
It is worth noting that Providence, the state’s biggest city, represented a larger portion of the incarcerated population than the probation/parole population. While some may be tempted to say this disproves our point, it is important to remember the context of this time period. The probation/parole data was from December 2020, at the pandemic’s peak, when many urban areas saw their populations dramatically decrease due to health and economic concerns and the desire for more space. Based on the patterns in the other communities, it is reasonable to conclude this data represents pandemic-era fluctuations and should be treated as such.
Importantly, you can also see more incarcerated people come from Cranston than are there while on probation and parole. This is another sign that most people don’t stay near the prison after incarceration unless they are from there originally.
These numbers aren’t a perfect match, which means that the exact address someone ends up at is sometimes different than the one they had on file while incarcerated. However, it clearly shows that people almost always return to their home communities after incarceration.
Prison is the least accurate place for the Census Bureau to count incarcerated people
Even if someone is unconvinced by the data showing that incarcerated people are likely to return to their home communities after their release, there are plenty of reasons — backed up by hard data — that make clear that by counting incarcerated people at prisons, the Census Bureau is choosing the least accurate option.
People are incarcerated further from home than most people choose to live
When someone is incarcerated, they are moved far from their daily surroundings and usually locked up somewhere they never visit on their own — let alone live. They’ll likely never know anyone outside of the prison walls, nor rent a house, visit a business, or attend a community event there. So why would the Census Bureau decide this was their home?
People counted at the prison on Census Day aren’t likely there for long
There is a common misconception that a person in a prison will be in that facility for a long time — this is the mistaken belief that is the foundation of the Census Bureau’s choice to count incarcerated people there. But this simply isn’t true.
The average time served by people in state prison is 2.7 years, with this “average” pulled up by the few people serving very long sentences; the median time served is 15 months. Short stays in prison are common. For example, in Rhode Island, the median length of stay for people serving a sentence in the state’s correctional facilities is only 99 days. This means that most people counted at a state prison on Census Day will spend the vast majority of the next decade outside the prison walls.
The home address someone has on file with the Department of Corrections may not be the exact place they return to after incarceration. However, it is a close approximation of where they reside through and after incarceration. While some of these address records may vary in precision, the one address we know is wrong is the facility address.
States that adjust their redistricting data to count incarcerated people at home use the home addresses contained in their Departments of Corrections records. This practice creates redistricting data that better reflects the populations of communities hardest hit by mass incarceration and communities that contain large prison populations. The Census Bureau should count incarcerated people at home in the 2030 Census using home address data as the states have done.
The Probation and Parole data is from December 2020 and the redistricting data is from April 2020 — these are the closest dates available. While the distribution of people in prison in April 2020 geocoded to their hometowns is not a perfect match to the hometowns of people on probation and parole in December 2020, it is close enough to draw some conclusions. It clearly shows that the towns where incarcerated people come from are the same towns where incarcerated people go after release. ↩
The U.S. Office of Management and Budget (OMB) recently announced changes to how federal agencies will collect race and ethnicity data, the first change to these rules since 1997. We and other civil rights groups welcomed these changes because they will present a more nuanced and accurate depiction of the racial and ethnic diversity of the country. The Census Bureau will be following these new standards for the 2030 Census, and so these changes should also prompt states — both those that have ended prison gerrymandering and those that are considering taking that step — to change how they track race and ethnicity data for people in their prisons.
What are the changes and why do they matter?
The new rules do two primary things to the Census count:
Combine race and ethnicity categories into one question: This better aligns with the way people identify themselves — for example, it gives people the option to choose Latino or Hispanic as a racial category, rather than just ethnicity.
Add Middle Eastern and North African (MENA) as a new racial category: Previously people of Middle Eastern or North African descent were aggregated into the white category against their wishes.
How will this impact prison gerrymandering reforms in states?
These changes present some challenges — and opportunities — for states that have ended, or are considering ending, prison gerrymandering.
Prison gerrymandering is a problem the Census Bureau created by counting incarcerated people as residents of their prison cells rather than in their home communities. This artificially inflates the populations of areas that contain prisons, giving them additional political clout when state and local governments use this Census data to draw new district lines every ten years. In states that have ended prison gerrymandering, state officials adjust their redistricting data to count people in prison at their pre-incarceration address, giving a more accurate picture of the area’s population and more equal representation in government.
To address prison gerrymandering, states take the data provided by the Census Bureau and cross-reference it against data collected by their Department of Corrections to reallocate incarcerated people back to their home communities before they draw new political boundaries. Importantly, when drawing these new maps, they not only consider the number of people in a district, but also who the residents of that area are. This is largely to comply with state and federal Voting Rights Acts and constitutional protections to ensure the ability of Black communities — and other racial and ethnic groups — to elect representatives of their choice.
In an ideal world, states and the Census Bureau would track race and ethnicity data in the same way. It would make it easier and more efficient for states to match Census data to their state data when adjusting redistricting data counting incarcerated people in their home communities. However, as the National Conference of State Legislatures reported last year, that’s not the case. Of the states that ended prison gerrymandering after the 2020 redistricting cycle, only one state’s Department of Corrections used the same race and ethnicity categories as the Census Bureau.
States that currently track race and ethnicity differently than the Census Bureau, will still have to do the work to make their data and the Bureau’s data compatible. However, these states should consider these Census changes as an opportunity to improve their data collection and make it more compatible with the Bureau’s data. Here’s how:
States should reform how they collect race and ethnicity data to conform with the updated Office of Management and Budget standards to make comparing data easier for reallocation purposes;
States should provide opportunities for incarcerated people to self-report race and ethnicity data whenever possible to ensure accuracy in the data they collect. When that is not possible, at the very least the Department of Correction should disclose how that data is collected.
Regardless, these new standards do not stand in the way of ending prison gerrymandering.
Roughly half of all U.S. residents now live in a city, county, or state that has ended prison gerrymandering, with more almost certain to join them before the 2030 redistricting cycle. That raises the important question: Instead of forcing states to keep up with the Bureau’s changing data collection and reporting, why not finally change how it counts incarcerated people in 2030?
Last month, U.S. District Judge Shelly Dick struck down Louisiana’s state legislative districts, noting that they unfairly deprived Black residents of political representation in violation of the Voting Rights Act. She ordered the state to draw new legislative maps to resolve this issue.
Importantly, in her ruling, Judge Dick noted how the over-incarceration of Black people harms their ability to engage with their government, saying:
“The preponderance of the evidence proved that Black voters’ participation in the political process is impaired by historic and continuing socio-economic disparities in education, employment, housing, health and criminal justice.”
While the judge did not specifically cite prison gerrymandering in her ruling, it is an issue that makes the inequalities at the heart of the case even worse. Prison gerrymandering is a problem created because the Census Bureau counts incarcerated people as residents of the wrong place — a prison cell — rather than their actual homes. When state governments then use the Bureau’s data to draw new legislative maps during their redistricting process, they inadvertently give residents who live closest to prisons greater political clout at the expense of everyone else.
When the Louisiana legislature convenes to draw new legislative districts to comply with the Judge’s order, its maps should address the injustice of prison gerrymandering.
Prison gerrymandering diminishes Black political representation in Louisiana
Louisiana is the incarceration capital of the world, locking up not only a greater share of its residents than any other state in the country but also more than any nation on the planet. So it should come as no surprise that prison gerrymandering has stark consequences in the state.
A unique dataset, which we gathered with the help of Louisiana-based Voice of the Experienced and Redistricting Data Hub, gives us a more precise look at how this plays out.1 Most importantly, we can see that incarcerated people come from all over the state. This matters because under the current flawed way the Census Bureau treats incarcerated people, the 26,000+ people in state custody are counted in just a handful of state prisons, dramatically inflating the population of those communities — and, with it, their political clout.
This data shows that prison gerrymandering robs nearly every state legislative district and its residents of some of their political voice. This political clout is then given to those districts that contain prisons.
While nearly every community is harmed by prison gerrymandering, that pain is not felt equally. Louisiana has stark racial disparities in its prisons. Black residents account for 32% of the state’s population, but they account for 66% of the people in state prisons.
Slideshow 1. Racial disparities in Louisiana prisons are stark. Learn more by visiting our Louisiana state profile page.
With such dramatic disparities in incarceration in the state, it is clear that prison gerrymandering is undermining Black political representation in Louisiana.
Prison gerrymandering gives two majority-white districts a particularly loud voice in government
Because of prison gerrymandering, people in Louisiana who happen to live close to state prisons get more political clout at the expense of every other resident of the state.
In Louisiana, two majority-white districts particularly benefit from this dynamic. First is House District 22, which contains a federal prison in Pollock. The other is House District 18, home to the Louisiana State Penitentiary, better known as Angola Prison.
In each of these places, incarcerated people make up more than 11% of the district population. As a result, 89 actual residents of these districts have the same political clout as 100 residents of any district in the state without a prison.
Two majority-white districts in Louisiana benefit the most from prison gerrymandering, at the expense of every other district in the state.
Percentage of the district population that is incarcerated
Percentage of the district population that is white (non-Hispanic)
House District 22
11.73%
62%
House District 18
11.36%
62%
Addressing prison gerrymandering will correct this imbalance and ensure that residents of these districts get the same voice in government — no more and no less — as every other resident of the state.
Louisiana’s embrace of failed, punitive criminal legal system policies shows why prison gerrymandering matters
At its core, the failure of Louisiana to address prison gerrymandering is an issue of political representation. However, this isn’t simply an academic exercise about ensuring everyone has an equal voice. It has real and dramatic consequences. To see this, you have to look no further than the recently ended special session of the Louisiana legislature.
The state’s new governor called the special session under the guise of focusing on crime despite the evidence that crime is at its lowest levels in roughly 60 years. During the session, lawmakers passed a slew of regressive policies that will increase prison populations and make the criminal legal system more punitive, including effectively eliminating parole, trying more young people as adults, and expanding the use of the death penalty.
It is reasonable to conclude that lawmakers who benefit politically from prison gerrymandering and the state’s large prison population have a strong incentive to protect and even grow this advantage by supporting the type of punitive criminal legal system policies passed during the special session. On the other hand, lawmakers who represent districts that have large numbers of residents in state prisons are more likely to seek solutions to the root causes of incarceration, such as poverty and untreated mental health and substance use issues. This conclusion was borne out during the special session when the two lawmakers who represent House Districts 22 and 18 — and benefit most from prison gerrymandering — supported all of the new draconian changes to the criminal legal system.
Addressing prison gerrymandering will help to disrupt this incentive toward bad policy and ensure every resident gets an equal say in government.
Now is the time for Louisiana to end prison gerrymandering
Eighteen states — including “red” states like Montana, “blue” states like New York, and “purple” states like Maine — have recognized the injustice of prison gerrymandering and rejected the flawed Census data. Instead, they took it upon themselves to correct the redistricting data to provide more equal representation to their residents. Progress on this issue has been so swift that the staunchly bipartisan National Conference of State Legislatures recently called the movement to end prison gerrymandering “the fastest-growing trend in redistricting.”
Louisiana should join this rapidly growing list.
We don’t yet know when the legislature will take action in response to the court ruling, and the judge did not set a clear timeframe for when they must act. However, when they do, their maps should address the problem of prison gerrymandering. Failure to act would mean that residents of the state — particularly Black residents — will continue to have their political representation diminished until at least 2031.
Footnotes
This analysis only includes data on the people incarcerated in state prisons in Louisiana. It does not include information on people held at the two federal prisons in the state. ↩
The 2020 redistricting cycle was a turning point in the effort to end prison gerrymandering. Fixing the Census Bureau’s flawed way of counting incarcerated people was once considered a niche issue, but after 2020 is now “the fastest-growing trend in redistricting,” according to the fiercely bipartisan National Conference of State Legislatures.
We previously looked at the experiences of states that have addressed prison gerrymandering after the 2020 Census. We wanted to also see what happened in states that haven’t fixed this problem to see what it tells us about how this issue plays out across the country as we look toward the 2030 count.
To do this, we examined the ten state legislative districts1 where, according to Census data, people held at correctional facilities account for the largest percentage of the district population.
These ten worst prison gerrymanders in the nation show the problem is increasingly harming residents of rural states and defying some of the preconceived notions about partisan impacts.
The 10 Worst Prison Gerrymanders After 2020 Redistricting
State
District
Percent of district’s Census population incarcerated
What party holds the seat?
New Hampshire
House District Merrimack 17
49.5%
Democratic
West Virginia
House District 83
18.4%
Republican
Mississippi
House District 47
17.7%
Democratic
Mississippi
House District 68
16.6%
Democratic
Oklahoma
House District 56
14.0%
Republican
Wyoming
House District 2
13.6%
Republican
Mississippi
House District 26
13.6%
Democratic
Arkansas
House District 65
13.5%
Democratic
West Virginia
House District 45
12.6%
Republican
South Carolina
House District 73
12.0%
Democratic
Rural states bear the burden of the Census Bureau’s error
This analysis shows that prison gerrymandering is a problem increasingly harming residents in some of the most rural states. While some might be tempted to chalk this up to the political dynamics of those states — we explain later in this piece why this seems increasingly unlikely — the truth is this is likely more a reflection of the Census Bureau shifting its responsibilities to over-burdened state governments and the series of trade-offs this forces states to make.
Prison gerrymandering is a problem created by the Census Bureau. Every ten years, it erroneously counts incarcerated people at the location of the facility they happen to be on Census Day rather than in their home communities. When state and local governments use this data for redistricting, they inadvertently give communities that contain prisons greater political clout at the expense of everyone else.
If states want to draw fairer districts by ending prison gerrymandering, they have to take on the task of fixing the flawed Census Bureau data to count incarcerated people at their true homes. This takes time, effort, and money, three rare and valuable commodities during redistricting.
To be clear, fixing prison gerrymandering is not a hugely expensive endeavor.2 But state budgets are a series of trade-offs and compromises. If lawmakers have the chance to spend even a few thousand dollars on direct services to their residents or can use that money to fix the Census Bureau’s mistake that caused prison gerrymandering, which do you think they’re likely to choose?
The experience of Montana, a state that has addressed prison gerrymandering despite its comparably small budget, shows the budgetary tricks or luck that states have to rely on to fix this Census Bureau mistake. During the 2020 redistricting cycle, commissioners in that state were unanimous in their desire to address prison gerrymandering, but it almost did not happen because they didn’t have the money to adjust the Census redistricting data to count incarcerated people at home. Ultimately, they could only implement this change because COVID-related restrictions meant that they could use money initially allocated for travel instead to end prison gerrymandering.
Some might be tempted to ask, “Why should the Census Bureau take on this effort instead of the places that want to change the data?” The simple fact is that the Census Bureau has a duty to provide states with redistricting data fit for use — a duty that the emerging consensus among states shows they are not meeting. State and local governments have been loud and clear, they want the Bureau to fix this problem.
Roughly half of all U.S. residents now live in a city, county, or state that has ended prison gerrymandering, with more almost certain to join them before the 2030 redistricting cycle. The Bureau’s stubborn refusal to change how it counts incarcerated people is a liability for a rapidly growing swath of the country and a disincentive for more places to join their ranks.
Additionally, it is important to remember that apart from any concerns about taking on additional burdens and redistricting time, states can only partially solve the problem. Prison gerrymandering is a national problem that crosses jurisdictional boundaries. No city, county, or state can completely fix this issue on its own. Only the Census Bureau can fully solve this problem nationwide.
Prison gerrymandering is not a partisan issue
Redistricting is notoriously contentious, with politicians working to gain leverage for their party in state houses.
While some have sought to paint the effort to end prison gerrymandering as a partisan tool in those fights, this analysis firmly busts that myth. These data show that both parties are harmed by prison gerrymandering.
Of these ten districts, six are districts controlled by Democrats, while Republicans control four. Notably, the district most distorted by a prison, New Hampshire’s Merrimack 17, is currently held by a Democrat. Prison gerrymandering distorts this district so severely that residents there have twice the political clout of residents of any other district in the state without a prison.
This is the latest piece of evidence that prison gerrymandering reform regularly defies traditional political expectations:
More than 200 rural city and county governments — many in some of the most conservative parts of the country — have recognized how prison gerrymandering harms their residents and have taken it upon themselves to fix Census data to address prison gerrymandering.
This analysis makes clear that neither political party is a guaranteed winner in prison gerrymandering. But, the clear loser in prison gerrymandering is every resident who doesn’t live in a legislative district with a prison.
What does this data tell us about 2030?
The 2030 Census may seem far away, but the Bureau has already begun planning for the count. It will likely decide where to count incarcerated people in the next two to three years.
This new analysis, which shows that the burden to correct flawed Census data increasingly falls on rural states and that prison gerrymandering is an issue that defies traditional partisan assumptions, should weigh heavily on their decision.
As we wait for the Census Bureau to decide, state lawmakers — particularly in these states with the worst prison gerrymanders — should start laying the groundwork to fix this problem should the Bureau fail to act. This means passing legislation now to ensure the state has the data and expertise it needs to successfully implement reform to end prison gerrymandering during the 2030 redistricting cycle.
Footnotes
This analysis is based on districts as they stood after the initial redistricting period after the 2020 Census. Some of these states may have adjusted their districts since then. ↩
Courts in Michigan and Wisconsin have stuck down their legislative maps, providing a rare mid-decade opportunity to immediately end prison gerrymandering.
While most people around the country were focused on the festivities of the holiday season, courts in two states — Michigan and Wisconsin — ordered them to redraw their legislative districts after they were ruled unconstitutional. This move provides each state with a rare opportunity to address prison gerrymandering in the middle of a decade, rather than waiting for the next Census.
Prison gerrymandering is a problem created by the Census Bureau. During its tally every decade, it erroneously counts incarcerated people as residents of prison cells, rather than in their home communities, artificially inflating the population of prison towns. When states then use this data to draw new legislative districts, areas that contain prisons get additional political clout, at the expense of everyone else.
More than a dozen states have taken action to fix the Census Bureau’s error — including “red” states like Montana and “blue” states like California. Momentum on this issue has been so swift, in fact, that the fiercely bipartisan National Conference of State Legislatures recently called it “the fastest-growing trend in redistricting.” Roughly half the country now lives in a place that has addressed the problem.
The court-ordered redistricting in Michigan and Wisconsin gives each state a chance to join this growing movement, but they have to act fast.
Michigan’s chance to address racial disparities
In late December, federal judges ruled that 13 Detroit-area state legislative districts unfairly diluted the political clout of Black residents, and ordered the state’s independent redistricting commission to redraw them for the 2024 elections. Commissioners have until February 2 to produce new maps.
While the court only explicitly ordered that 13 districts be redrawn, changing the lines of these districts will almost certainly have statewide ripple effects.
Because the court ruling involved the dilution of Black representation, it is important to remember the disproportionate impact that prison gerrymandering has on Black residents in the state. Black residents are 13% of the state’s population but 51% of the people in state prisons. As a result of this disparity and the state’s failure to address prison gerrymandering, Black political representation across Michigan is being unfairly diluted.
Addressing prison gerrymandering won’t solve all of the issues with Wisconsin’s current legislative districts, but it will be a meaningful step toward ensuring fairer representation in the state.
Is there enough time for these states to address prison gerrymandering?
Time is tight, but there is an quick — albeit imperfect — solution to address the worst impacts of prison gerrymandering.
Both of these states are unquestionably working on tight timeframes. Some will question whether they have the time or data necessary to fully address prison gerrymandering in these new legislative maps.
Ending prison gerrymandering completely by counting incarcerated people in their home districts is obviously the preferred solution, but mapmakers in these states shouldn’t let the perfect be the enemy of the good. There is a simple — albeit partial — solution that can address the most distortive impacts of prison gerrymandering and can be implemented quickly, even if the states don’t have address data for people in prison.
Here’s how this solution works. When drawing new maps, instead of counting incarcerated people as residents of the district that contains the prison, mapmakers reallocate this population statewide. While this will not fully return political representation to the communities most impacted by incarceration, it will ensure that communities that contain prisons don’t get a dramatically outsized voice in government.
For example, in Wisconsin’s Assembly District 53, roughly 8% of people counted in that district are incarcerated and live outside of that area. That means that 92 residents of that district have the same political sway as 100 residents in any other district that doesn’t have a prison. Ending prison gerrymandering will ensure that residents of this district don’t get a louder voice in government, simply because they live near a prison.
Recognizing that the governor and state legislature would be unlikely to reach a consensus on new maps, the court appointed two experts to review the submitted maps. They can then either alter the submissions or produce their own maps that resolve the court’s concerns. Notably, these experts have helped Virginia, Pennsylvania, and New York — three states that have addressed prison gerrymandering — draw their current legislative districts. Since these two experts already understand this issue and the ways to solve it, the final maps they propose should address prison gerrymandering to the fullest extent possible.
Failure to act means at least eight more years of prison gerrymandering
Redistricting officials in these two states have a lot of work ahead of them, and not much time to do it in. Ending prison gerrymandering in the state should be near the top of their list of priorities.
We’re hopeful that the Census Bureau will fix its prison gerrymandering mistake in the 2030 count. But Michigan and Wisconsin don’t have to wait that long to fix their prison gerrymandering problem. They have an opportunity to end prison gerrymandering while addressing problems with their state legislative maps right now. They should take advantage of it.