Federal Census policy breaks Michigan’s democracy — state lawmakers can fix it
Michigan’s redistricting data was once again skewed after the 2020 Census; the state needs to take action to fix the issue for 2030.
by Aleks Kajstura, August 6, 2025
Everyone in Michigan is supposed to have an equal voice in their government’s decisions, but an outdated and misguided Census Bureau policy that counts incarcerated people in the wrong place gives a few residents of the state a megaphone. It is a problem known as prison gerrymandering, and Michigan lawmakers can fix it.
Michigan blindly follows outdated bureaucratic federal policy
Every ten years, when the Census Bureau conducts its official tally of the nation’s population, it incorrectly counts incarcerated people as residents of prison cells rather than in their home communities. This is despite the fact that they usually are not from the prison town, have no family or social ties there, likely won’t stay there for long, and state residence law says they’re not residents there. When state officials then use that incorrect Census data in the legislative redistricting process, they inadvertently inflate the populations of those areas – in violation of constitutional principles of equal representation. This gives residents of state legislative districts that contain correctional facilities a particularly loud voice in government, allowing them to have an outsized influence on debates about minimum wage, gun regulation, election rights, and more, at the expense of nearly every other person in the state.
To ensure equal representation, states across the country have taken steps to fix this problem that the Census Bureau created. But, Michigan is one of the remaining states still suffering from this “prison gerrymandering.” While the 2030 Census count is still years away, Michigan needs to act now to avoid prison gerrymandering the next time it redraws its districts.
Prison gerrymandering significantly distorts Michigan’s state legislative districts
In Michigan, there are three state House districts that powerfully illustrate how prisons distort district populations and give some residents a louder voice in government as a result of prison gerrymandering.
In Districts 78 and 46, for example, correctional facilities make up 7% of the population. That means that just 93 residents in those districts have as much political clout as 100 residents in any normal district. District 93 is close behind with more than 5% of its population being incarcerated.
The three most prison-gerrymandered House of Representatives districts in Michigan:
| District | District Location | Notable facilities | Percent of the district that is incarcerated |
|---|---|---|---|
| 78 | Parts of Ionia, Kent, Eaton, and Barry Counties | Ionia Correctional Facility, Richard A. Handlon Correctional Facility, Bellamy Creek Correctional Facility, Michigan Reformatory | 7.4% |
| 46 | Parts of Jackson and Washtenaw Counties | Charles Egeler Reception and Guidance, Cooper Street Correctional Facility, Parnall Correctional Facility, G. Robert Cotton Correctional Facility | 7.0% |
| 93 | Parts of Saginaw, Gratiot, Montcalm, Ionia, and Clinton Counties | Carson City Correctional Facility | 5.6% |
In fact, one-third of all people incarcerated in Michigan are counted as residents of facilities in these three districts, despite the fact that incarcerated people come from every district in the state.1 That imbalance in representation comes from the state choosing to redistrict based on Census numbers that don’t match the reality of where people live.
The Census counted people incarcerated in those places as if they were residents of the facility location, even though state correctional facilities contain people who are incarcerated far from home, have no ties to the communities where the facilities are located, and are moved regularly between facilities for administrative convenience.2 Simply put, being incarcerated in a specific facility doesn’t make someone a resident of the surrounding district.
Prison gerrymandering disproportionately harms Michigan’s Black residents
Prison gerrymandering reduces the political power of nearly all Michigan residents by allowing a handful of districts with large correctional facilities to claim residents from all over the state. But prison gerrymandering particularly harms Black residents as it enshrines the racial inequities of mass incarceration into the state’s legislative districts.
In Michigan, like across the country, mass incarceration has a disproportionate impact along racial lines. In Michigan, Black residents are incarcerated at higher rates and, therefore, are counted in the wrong place more often than Michigan’s white residents.
The racial impact of prison gerrymandering is so strong that, for example, 88% of Black people counted in State House District 78 were actually behind bars, rather than living in the community.
Over half of the people incarcerated in Michigan’s prisons are Black. Black residents make up just 13% of the state population, but a whopping 51% of people in prisons and 36% of people in jails.
Additionally, 26% of the state’s incarcerated people come from Wayne County, even though the county only accounts for 18% of the state’s population.3
Counting incarcerated people in the wrong place adds up. Just in the three districts highlighted above, over 7,608 Black people were counted in the wrong place. This means that the Census Bureau policies are effectively silencing the voices of a large portion of the state’s Black residents.
Michigan law says a prison cell is not a residence. Census Bureau policy disagrees
Not only does the Census Bureau’s redistricting data cause prison gerrymandering, it also doesn’t comply with Michigan’s law. The state’s constitution explicitly states that being incarcerated doesn’t change a person’s residence:
“An elector shall not be deemed to have gained or lost a residence … while confined in a jail or prison.”
Michigan Compiled Laws §168.11(2).
Additionally, Michigan law also explicitly requires local governments to address prison populations in their own redistricting data. We explore this more deeply in the local government section, below.
Instead of following state law, though, the Census Bureau follows its own “residence rule” to choose where to count incarcerated people – where they “live and sleep most of the time.” But it doesn’t even follow this rule properly when it comes to counting incarcerated people.
The Census Bureau counts incarcerated people at the location of the facility where they happen to be held on Census Day under the mistaken belief that that is where incarcerated people “live and sleep most of the time.” The facts, however, do not support its interpretation of its own definition of residence. It is well-established that in the modern era of mass incarceration, incarcerated people do not “live and sleep most of the time” at the facility where they are held on any given day (including Census Day). Nationally, 75% of people serve time in more than one prison facility, and 12% of people serve time in at least five facilities before returning home.
Representation in local government is also harmed by prison gerrymandering
The impact of prison gerrymandering is even more visible at the small scale: in city and county governments. So much so that Michigan already requires local governments to exclude state prisons from their redistricting data when drawing county board of commissioners districts or city council wards.
Michigan’s statutes (MCL §46.404(g) for counties and §117.27a(5) for cities) specify that people incarcerated in state institutions who do not actually reside in that county or city shouldn’t be counted when drawing local government districts. These statutes protect Michigan residents from wildly distorted political power distribution – for example, in neighboring Wisconsin, there are three counties where prison populations make up more than half of a county commission district, giving the residents of the districts that contain the prisons twice the influence over county affairs compared to other county residents.
An unfortunate oversight, however, is that these laws fail to account for local jails and, even more importantly, the federal facility (FCI Milan).
The experience of Washtenaw County shows a county working to avoid prison gerrymandering under the state law but being prevented by that same statute from entirely solving the problem. Washtenaw has four correctional facilities within its borders: two state facilities (Camp Cassidy and Hurton Valley), a federal facility (FCI Milan), and the Washtenaw County Jail. The county excluded the state facilities, as required by law, but was left with the jail and FCI Milan skewing its redistricting data. There were 231 people counted at the jail, not enough to significantly affect a roughly 40,000-person district. But the 1,573 people counted at FCI Milan accounted for 4% of County Commissioner District 3. Meaning that every 96 people in District 3 have the same voice in county government as 100 people in the other districts.
Michigan has taken important first steps in requiring that local governments avoid prison gerrymandering, but it should make its solution more comprehensive by extending it to apply to federal and local facilities. Or, even better, the state should provide a more efficient and complete solution for its local governments by counting incarcerated people at home and sharing that redistricting data with cities and counties. While it is not fair that the state has to correct for this federal issue, the state is in a better position to take on that burden than each individual city.4
Nationally, state and local governments are addressing the problem, but Michigan is lagging behind
Over the last few decades, more than 200 local governments and a growing number of states have taken action on their own to fix this problem. Nearly half of the US population now lives in a place that corrects redistricting data they receive from the Census to avoid prison gerrymandering.
States that have ended prison gerrymandering on their own include deep “blue” states like California, “purple” states like Maine and Pennsylvania, and deep “red” states like Montana — where prison gerrymandering-reform legislation received wide bipartisan support. But Michigan is falling behind and letting the state’s democracy be skewed by an outdated federal system.
Michigan needs to take action now
Adjusting redistricting data to avoid prison gerrymandering is now a well-tested strategy with a proven track record. In fact, the bipartisan National Conference of State Legislatures called this effort “the fastest-growing trend in redistricting.” Michigan can now confidently pass legislation to count incarcerated people at home for redistricting purposes. Other states have already been successful in these efforts, paving the way for Michigan. And the state would have the benefit of refining its approach based on lessons learned by states that have gone through the process before. And it is easier than ever for states to act; even the Census Bureau is starting to acknowledge the problem and help.
2030 may seem far away, but other states have learned that the earlier that reforms are put in place, the less expensive, easier to produce, and more accurate the final redistricting data becomes. Every state has a different legislative approach to ending prison gerrymandering, but as a practical matter, this model bill, prepared by a coalition of civil rights, voting rights, and criminal justice organizations, is a great place to start. It provides clear guidance on how this data should be collected, by whom, and how it will be used for the redistricting process. And Michigan has already taken its first steps, with Senate Bill 494 in the 2023 regular session. The bill nearly made it across the finish line, but died during the session’s closing day.5
The Census Bureau is unlikely to change its policies about how to count incarcerated people in time for the 2030 Census, meaning that unless Michigan acts quickly, the state will once again be driven into prison-gerrymandering their legislative districts.
Michigan needs to end prison gerrymandering now.
About the Data
Correctional Facility Populations: To calculate the percentage of each district’s population that was in correctional facilities, we used the redistricting data (PL 94-171) from the 2020 Census. Table P1 provides the total population for each Census block and Table P5 provides the number of incarcerated people for each Census Block. Notably, this approach includes people in all kinds of correctional facilities, including state prisons, federal prisons, private prisons, local jails, halfway houses, etc.
Identifying specific facilities: Table P5 provides the population of correctional facilities without distinguishing between state, federal, or private facilities and it is published for each Census block. Census blocks do not necessarily translate directly to facilities, as some facilities are counted in multiple blocks and some blocks contain multiple facilities. To aid redistricting officials and advocates with using this data, the Prison Policy Initiative maintains a Facility Locator Tool that contains annotations of most of the Census blocks in the country that contain correctional facilities. These annotations rely on publicly-available data to identify facility names and types in each of these blocks.
Calculating how many Michigan residents are held by the Bureau of Prisons:
Our calculations on the number of people in federal prisons in each state are based on data provided by the Bureau of Prisons in response to our periodic Freedom of Information Act requests.
How this report quantifies prison gerrymandering compared to other analyses: There are a few ways to calculate the impact of prison gerrymandering, so other researchers may have used slightly different approaches that generate slightly different numbers for the same general problem. For example, some analyses only focus on prisons and exclude jail populations. That choice makes sense when looking at state-level policies and state districts because people in jails are very likely to also live in the legislative district where the jail is located. However, for this analysis, we included jails as well as state correctional facilities for two reasons: because State House districts often split county lines in Michigan, and over 6% of the people held in Michigan jails are held for state and federal authorities. Still other approaches, such as that taken by the Redistricting Data Hub, are based on estimates of incarcerated people’s home addresses. That approach adds an additional level of precision for counting people held in state facilities because it seeks to not only address where these people were counted incorrectly – which accounts for the bulk of prison gerrymandering’s population distortion – but to also estimate where they should have been counted. Unfortunately, this approach isn’t able to reflect where people in federal facilities, most jails, and private facilities are from. And so, for simplicity, this report doesn’t use that approach.
Each of these approaches has its own merits, and none are universally better than others; they all highlight different aspects of how prison gerrymandering skews population numbers, and each has its own use. The complexities inherent in the current patchwork approach to identifying and solving prison gerrymandering point to the need for the Census Bureau to count incarcerated people at home in the first place in order to provide a comprehensive solution to prison gerrymandering.
Correctional facility populations in Michigan House of Representatives Districts, 2020 Census
| State House District | Total Population | Incarcerated Population | Percent of the District that is Incarcerated |
|---|---|---|---|
| 1 | 91,856 | 0 | 0.0% |
| 2 | 89,622 | 18 | 0.0% |
| 3 | 93,531 | 6 | 0.0% |
| 4 | 90,903 | 0 | 0.0% |
| 5 | 92,744 | 9 | 0.0% |
| 6 | 93,629 | 0 | 0.0% |
| 7 | 92,948 | 368 | 0.4% |
| 8 | 92,670 | 0 | 0.0% |
| 9 | 90,818 | 34 | 0.0% |
| 10 | 90,534 | 564 | 0.6% |
| 11 | 91,145 | 0 | 0.0% |
| 12 | 90,630 | 0 | 0.0% |
| 13 | 90,393 | 0 | 0.0% |
| 14 | 90,555 | 150 | 0.2% |
| 15 | 92,301 | 0 | 0.0% |
| 16 | 93,035 | 0 | 0.0% |
| 17 | 90,737 | 0 | 0.0% |
| 18 | 92,169 | 0 | 0.0% |
| 19 | 90,931 | 0 | 0.0% |
| 20 | 93,017 | 0 | 0.0% |
| 21 | 93,876 | 0 | 0.0% |
| 22 | 91,654 | 0 | 0.0% |
| 23 | 90,719 | 18 | 0.0% |
| 24 | 91,480 | 0 | 0.0% |
| 25 | 90,562 | 0 | 0.0% |
| 26 | 91,723 | 0 | 0.0% |
| 27 | 90,457 | 0 | 0.0% |
| 28 | 91,598 | 0 | 0.0% |
| 29 | 92,583 | 17 | 0.0% |
| 30 | 93,460 | 413 | 0.4% |
| 31 | 92,978 | 1,573 | 1.7% |
| 32 | 92,092 | 1,974 | 2.1% |
| 33 | 92,730 | 231 | 0.2% |
| 34 | 92,371 | 2,280 | 2.5% |
| 35 | 93,023 | 1,469 | 1.6% |
| 36 | 89,634 | 84 | 0.1% |
| 37 | 91,456 | 0 | 0.0% |
| 38 | 93,422 | 0 | 0.0% |
| 39 | 90,270 | 0 | 0.0% |
| 40 | 90,211 | 0 | 0.0% |
| 41 | 91,872 | 39 | 0.0% |
| 42 | 91,192 | 0 | 0.0% |
| 43 | 92,518 | 169 | 0.2% |
| 44 | 89,974 | 359 | 0.4% |
| 45 | 90,612 | 0 | 0.0% |
| 46 | 91,041 | 6,371 | 7.0% |
| 47 | 91,302 | 0 | 0.0% |
| 48 | 92,373 | 142 | 0.2% |
| 49 | 93,247 | 331 | 0.4% |
| 50 | 93,139 | 209 | 0.2% |
| 51 | 91,507 | 0 | 0.0% |
| 52 | 91,098 | 0 | 0.0% |
| 53 | 93,056 | 1,012 | 1.1% |
| 54 | 92,949 | 0 | 0.0% |
| 55 | 91,805 | 0 | 0.0% |
| 56 | 90,410 | 0 | 0.0% |
| 57 | 89,693 | 0 | 0.0% |
| 58 | 90,454 | 0 | 0.0% |
| 59 | 89,336 | 0 | 0.0% |
| 60 | 92,742 | 0 | 0.0% |
| 61 | 93,156 | 536 | 0.6% |
| 62 | 90,539 | 0 | 0.0% |
| 63 | 90,638 | 0 | 0.0% |
| 64 | 91,060 | 296 | 0.3% |
| 65 | 92,892 | 1,345 | 1.4% |
| 66 | 93,014 | 0 | 0.0% |
| 67 | 92,816 | 997 | 1.1% |
| 68 | 93,065 | 0 | 0.0% |
| 69 | 91,698 | 0 | 0.0% |
| 70 | 90,738 | 507 | 0.6% |
| 71 | 91,966 | 85 | 0.1% |
| 72 | 92,844 | 0 | 0.0% |
| 73 | 91,543 | 239 | 0.3% |
| 74 | 90,782 | 0 | 0.0% |
| 75 | 93,554 | 57 | 0.1% |
| 76 | 92,354 | 174 | 0.2% |
| 77 | 92,594 | 0 | 0.0% |
| 78 | 92,264 | 6,834 | 7.4% |
| 79 | 90,952 | 0 | 0.0% |
| 80 | 92,350 | 0 | 0.0% |
| 81 | 91,516 | 674 | 0.7% |
| 82 | 91,219 | 145 | 0.2% |
| 83 | 91,341 | 0 | 0.0% |
| 84 | 91,890 | 0 | 0.0% |
| 85 | 90,127 | 0 | 0.0% |
| 86 | 90,575 | 0 | 0.0% |
| 87 | 91,376 | 3,991 | 4.4% |
| 88 | 90,900 | 0 | 0.0% |
| 89 | 93,134 | 0 | 0.0% |
| 90 | 91,549 | 0 | 0.0% |
| 91 | 91,350 | 104 | 0.1% |
| 92 | 92,520 | 3,805 | 4.1% |
| 93 | 89,410 | 4,966 | 5.6% |
| 94 | 90,438 | 38 | 0.0% |
| 95 | 91,439 | 144 | 0.2% |
| 96 | 90,544 | 116 | 0.1% |
| 97 | 93,159 | 1,492 | 1.6% |
| 98 | 92,049 | 149 | 0.2% |
| 99 | 89,375 | 111 | 0.1% |
| 100 | 91,751 | 212 | 0.2% |
| 101 | 92,604 | 1,505 | 1.6% |
| 102 | 91,886 | 1,096 | 1.2% |
| 103 | 93,426 | 135 | 0.1% |
| 104 | 89,466 | 71 | 0.1% |
| 105 | 89,541 | 107 | 0.1% |
| 106 | 90,875 | 90, | 0.1% |
| 107 | 92,701 | 1,378 | 1.5% |
| 108 | 89,366 | 2,739 | 3.1% |
| 109 | 89,410 | 2,657 | 3.0% |
| 110 | 90,788 | 113 | 0.1% |
Table notes
- Total Population
- Total population reported for all blocks in the district (as redistricted in 2022). Block populations reported for the 2020 Census in the PL 94-171 redistricting summary files Table P1.
- Incarcerated Population
- Total incarcerated population reported in all blocks in the district (as redistricted in 2022), based on the incarcerated population in group quarters reported for the 2020 Census in the PL 94-171 redistricting summary files Table P5.
- Percent of the District that is Incarcerated
- This is the number of incarcerated people counted in the district divided by the total population of the district.
Footnotes
-
The Redistricting Data Hub’s report on prison gerrymandering in the state includes maps showing a district-by-district analysis of where incarcerated people would have been counted if the state had counted people at home. And our analysis of their underlying data shows that incarcerated people come from every county in the state. ↩
-
Even worse, some facilities don’t contain Michigan residents at all. FCI Milan, located in District 31, for example, is used by the Federal Bureau of Prisons to hold people from all over the country. And less than 2% of the Bureau of Prisons population comes from Michigan, which means out of the 1,573 people held by the Bureau of Prisons at FCI Milan, only about 30 people are likely Michigan residents. And of those 30 Michigan residents, it’s unlikely that even one person would be an actual resident of District 31. ↩
-
Wayne County, which contains Detroit, is home to nearly 40% of Michigan’s Black residents. And although the state does not keep records of where incarcerated people live, we were able to use sentencing data as an estimate of where incarcerated people come from. We received the sentencing county data courtesy of the Redistricting Data Hub, which obtained it through a Freedom of Information request from the Michigan DOC. ↩
-
In most cases, adjusting redistricting data to avoid prison gerrymandering is quite easy for local governments. Outside of Michigan, local governments are taking matters into their own hands and rejecting the Bureau’s way of counting incarcerated people, even in states that fail to act. Nationwide we identified over 200 local governments that avoided prison gerrymandering after the 2000 and 2010 Censuses (decades when zero and two, respectively, states adjusted their redistricting data to count people at home). This decade we limited the scope of our research but still found an additional 21 local governments scattered across 10 states that started doing so after the 2020 Census despite those states continuing to use unadjusted data for state-level districts. ↩
-
The bill passed the Senate and died in the House in the chaotic final hours of the legislative session. ↩
