By Kim Bellware

Prison gerrymandering is distorting democracy in states across the Midwest and nationwide, leaving incarcerated people with inequitable representation—or none at all

A small green road sign along Route 16, in southwestern Illinois, greets motorists rolling into the town of Hillsboro by announcing the size of the local population: 6,300. Brian Sullivan, who is wrapping up his third and final year as mayor of Hillsboro, is proud of the town, but he also knows there’s a level of Potemkinism in that road sign. The nearby Graham Correctional Center accounts for more than one-quarter of the official population. “We’re really a town of 4,500,” Sullivan told me when I called to ask about it. “They count the prison, which I wish they wouldn’t.”

The “count” Sullivan is referring to is the U.S. Census Bureau’s counting formula, which is unchanged since the first decennial census in the 1700s. Under a frustratingly vague census rule, “usual residence” is defined as the place where someone usually “lives and sleeps,” meaning incarcerated people have always been counted as residents of the town in which they’re imprisoned, rather than the last place they actually lived before they were sentenced.

The problem, at least to policy wonks, was largely theoretical until the prison boom of the 1970s. The growth of the U.S. prison population raised the stakes on the prisoner counting formula and emerged as a threat to voter enfranchisement and fair elections. Reform advocates have dubbed the practice “prison gerrymandering,” because counting these ineligible voters away from their homes dilutes and distorts voting power; it’s like old-school political gerrymandering, but more obscure and (usually) less cutthoat. And it’s still pernicious.

According to Illinois Department of Corrections data, roughly sixty percent of the state’s prison population is sentenced from Cook County, which includes the city of Chicago. Of the twenty-eight adult prisons across the state, none are in Cook County. As a result, prisoners from Cook County—overwhelmingly Black and Latino men—are sent to less populous and typically whiter counties downstate, where their numbers can bolster the voting power of those communities.

Outside of the criminal justice sphere, the term “prison gerrymandering” tends to draw a blank stare.

Paris Knox, who’s from Chicago, spent thirteen years in the Logan Correctional Center, where she was counted as a resident of Logan County, an area of central Illinois that is majority white and votes reliably Republican. Meanwhile, Knox’s home county of Cook is both the blackest and most Democratic county in the entire state and, notably, a historic seat of African-American political power in the U.S.

Knox’s situation is the definition of prison gerrymandering: thousands of people, disproportionately people of color, are imprisoned and shifted disproportionately from urban areas to rural ones where the local population is more homogeneously white and conservative. “We felt isolated because there was nobody to represent us,” she recalled, when I asked her what it was like to experience this trend from the inside. “There was no one to represent that population: ‘Here’s what the prisoners want.’”

To the extent anyone but the aforementioned policy wonks pays attention to the census counting formula, it’s historically been with respect to college students, not prisoners. If you live in a college town, you might think about how the ranks of students can swell a college town’s population and distort data points like local poverty rates, or median ages. But outside of the criminal justice sphere, the term “prison gerrymandering” tends to draw a blank stare.

I’ve been reading, researching or reporting on criminal justice issues for nearly ten years, and, until relatively recently, the issue flew below even my radar. I knew that in rural areas where prisons are an economic tentpole, the arrangement creates a conflict of interest when it comes to legislating prison reforms. But I had failed to consider what was at stake when entire classes of people are removed from their home communities to be counted toward the voting power of counties whose interests are different than their own—or, sometimes, disappear from the count completely.


Like so many public policy issues, prison-based gerrymandering has no chief architect or dramatic origin story, but its contours amplify the way power is typically dispersed to certain groups—and how it’s withheld from others.

Aleks Kajstura works on prison-based gerrymandering at the Prison Policy Initiative (PPI), a prison reform think tank where she serves as legal director. I called her recently to unpack the timeline. “Prisoners emerged as a focus of the census counting problem only in the past two decades,” she explained. “Previously, there just simply weren’t enough people incarcerated, even as late as the 1970s, to make a difference.”

But then the U.S. prison population doubled in the ‘80s, and continued to rise sharply through the ‘90s. Researchers attribute the U.S. prison population spike to a combination of factors, including the criminalization of activities related to immigration, drugs, and homelessness, and harsher sentencing laws for both violent and non-violent crimes, which proliferated in the ‘70s (particularly in states like New York, where the “Rockefeller laws” inspired similar legislation nationwide). These activities coincided with the so-called “war on drugs,” which kicked off in earnest when President Richard Nixon declared, in 1971, that drugs were “America’s public enemy number one.” In subsequent decades, state-based efforts to decriminalize drug possession were abandoned while the Drug Enforcement Administration expanded.

“Where people are incarcerated matters, because if everybody was evenly scattered across the state, no matter how many people are incarcerated, you’d still have an equal distribution.” – Aleks Kajstura

At the same time, the trend of “tough on crime” policies was growing. By the ‘80s, both Democrats and Republicans were learning that answering crime with punishment was a good way to avoid getting walloped at the polls. The Democrat-backed Anti-Drug Abuse Act of 1986, then-presidential candidate George H.W. Bush’s “Willie Horton” campaign ad, and the “superpredator” scare of the ‘90s look like disasters now, but they got people elected—and similar projects still do.

As incarceration became mass, Black citizens were disproportionately punished due to a toxic but systemic mix of racial bias that played out everywhere from juries to the way drug law was enforced. Even now, despite similar rates of drug selling and usage between Black and white Americans, Black Americans are 2.7 times more likely to be arrested for drug-related offenses, according to 2016 data from the Brookings Institute’s Hamilton Project; on the state level, Black Americans continue to be roughly 6.5 times more likely than whites to be incarcerated for drug-related crimes. (There is some debate among researchers about whether drug policies alone contributed to mass incarceration; some argue a more definitive factor was prosecutorial discretion—who states ultimately prosecuted, and how aggressively).

Of course, when a country starts locking up more people than it has in its history (or in the history of any other country, for that matter), capacity becomes an issue. The surging prison population led to a construction boom in the mid-1980s. Private prison companies, a capital-driven alternative to state-run facilities, began an aggressive expansion, especially in the south and west. What both public and private prisons had in common, though, is that they were overwhelmingly built in rural areas. Many small towns, for their part, hoped a prison could spur economic development, provide new jobs to replace those lost from the declining manufacturing industry, and stem the flow of out-migration. Between 1980 and 1991, 213 new adult prisons were built in rural areas, housing more than half of the national prison population.

“Where people are incarcerated matters,” Kajstura explained, “because if everybody was evenly scattered across the state, no matter how many people are incarcerated, you’d still have an equal distribution; in terms of district equality, it would all be a wash. But it’s the coupling of prison clusters along with the strong incarceration rates of the ’70s through the ‘90s.”

Given the harm mass incarceration has caused to poor people and communities of color, it’s easy to overlook another legacy of the system: the twin forces of mass incarceration and rural prison growth led to a large-scale transfer of specific populations from diverse urban centers to whiter rural areas, skewing not only demographics, but political power.


The Graham Correctional Center sits on a sprawling, 117-acre campus surrounded mostly by cornfield. It’s in Hillsboro, Illinois but just barely. On a map, the core of Hillsboro is roughly shaped like a rhombus tipped sideways; Graham looks like a vestigial tail, connected to the larger body only by a narrow stretch that juts out and away from town and follows no particular topographical landmarks, or even county road. But it doesn’t matter how precariously attached the prison is to the rest of the town; come census time, every person incarcerated in Graham will boost the population of Hillsboro, and of Montgomery County, for which it serves as the seat.

In states like Illinois, municipalities do have some discretion in how they choose to count their population when it comes time to draw new district lines. Since the last census, at least eleven of the thirty-one Illinois counties with adult state prisons, including Montgomery, have opted to exclude prisoners from the overall population count to avoid the county’s small size being easily distorted by the extra “residents” in the prison.

Mike Plunkett has served on the Montgomery County Board for the past eighteen years. Plunkett’s district is in Hillsboro, and he says that during his tenure the board has always deducted the prison population from the census totals for mapmaking purposes. “We’re a county of thirty thousand. Each district would be about forty-five hundred people, and the population has to be substantially similar in each district…The prison has about two thousand people. That district—among the seven in the county—would have only half of the voters in it even eligible.”

In other words, the voters in Plunkett’s district would effectively have twice the voting power of neighboring districts. And the smaller a district gets, the bigger the problem grows: “Now Hillsboro, where the prison is, the mayor and the councilmen are all elected at-large,” Plunkett explained. “But if Hillsborough were aldermanic, you’d have a district where the prison is that would have two thousand residents who couldn’t vote; you’d have no one who could even run for office.”

One of the more dramatic examples of Plunkett’s scenario happened in Iowa before the 2010 census. In the city council race in Anamosa, Iowa, population fifty-five hundred, a candidate named Danny Young ran for a seat in a ward that included the Anamosa State Penitentiary, which,  according to the New York Times, held a population of about thirteen hundred at the time, leaving only fifty-eight non-prisoners in the voting district. Young won his seat with only two votes cast: one from his wife, and one from his neighbor. (Young didn’t bother to vote for himself.) Thanks to prison gerrymandering—and a good deal of local voter apathy—his victory set a record for a win by the slimmest margin.

Reform efforts look less successful when viewed as the presence of justice for incarcerated populations, rather than simply the absence of exploitation.

Anamosa ultimately solved its problem by sidestepping the issue of prisoner population counts altogether, changing to at-large elections, rather than district-based. Of course, not every prison town’s officials do the honorable thing; while at least ten counties in Illinois have excluded prisoners from their Census counts, at least five other countries like Will, Rock Island, Lee and Clinton have declined to do so. A cruel irony of prison gerrymandering is that the largely minority population is shifted to rural areas, where it lacks the benefit of real representation (destination districts are overwhelmingly white), but can nonetheless enable the careers of politicians who oppose prison and criminal justice reform measures.

Dale Volker, a long-time conservative Republican state senator from upstate New York, somewhat famously acknowledged that his sparsely-populated district “had more cows than people.” Volker’s district did have a large prison, however, and he expressed relief, in a 2002 interview with the Newhouse News Service, that many of his constituents were prisoners ineligible to vote. If they could, he reasoned, “they’d never vote for me.” Volker used his seat in the state legislature—power drawn from a significant number of Black and brown prisoners from the New York City area—to oppose shortening prison sentences and census changes, ensuring a consistent flow of “constituents” for his district.

Nationwide, states are still grappling with voting rights for incarcerated people. Most every state disenfranchises people serving a felony conviction, except for Maine and Vermont, where prisoners never lose the right to vote and do so absentee while in prison. Even after an individual has served their felony sentence, voting accessibility is still stratified: In fourteen states, including Illinois, voter rights are automatically restored upon completing a sentence; in the remainder, the formerly incarcerated lose the right to vote until they’re off parole or probation—or they never regain it, short of more formal intervention, like a direct gubernatorial pardon.

Since the Dale Volker era, The Prison Policy Initiative reports that seven states, including Michigan and New York, now “encourage or even require” governments to exclude prison data from redistricting—meaningful strikes by most measures. Yet the reform efforts look less successful when viewed as the presence of justice for incarcerated populations, rather than simply the absence of exploitation.

Even when counties like Montgomery don’t count prisoners in their population totals, those individuals still have a hometown in which they’re not being counted, resulting in potentially hundreds of thousands of individuals—effectively, the entirely of the U.S. prison population save for two states—losing out on political representation altogether. Out of fifty states, only New York and Maryland have implemented laws mandating prisoners be counted at home. Advocacy organizations have had marginal success in encouraging districts to leave prisoners out of their counts; getting government agencies to restore the representation as if it had never left? That appears to be a tougher sell.


Unlike other mass incarceration-related issues, such as racial bias or political corruption, prison-based gerrymandering is a modern issue. Aleks Kajstura estimates that the 1990 census was the first to be significantly impacted by the nation’s surging prison population. The influx of incarcerated people skewed rural resident counts, but it wasn’t thoroughly documented for another decade, when the 2000 census came around. The energy to reform is a cycle all its own: it peaks around a census, and wanes in the intervening years. Municipalities typically redraw their district boundaries in the odd year after the census, once all the data is in hand, and decide whether or not to include prison totals in the two or so years before.

Recently, some municipalities, like Hillsboro, have been using their own discretion to whittle prison totals out of the federal census data, but advocates like Josina Morita are still pursuing a legislative fix that would address the patchwork nature of these population counts. In 2013, she promoted HB 62, a bill that would have uniformly changed how all state prisoners are counted in local redistricting. “[Prison-based gerrymandering] is not an explicit policy,” Morita wrote in her testimony in support of the bill. “It is a practice that occurs due to a lack of and misuse of population data, but has far reaching negative impacts on the right to fair representation for all Illinoisans.”

There was political will to fix the issue, just not enough of it: the bill ultimately failed to persuade key downstate lawmakers, just like its predecessor, the Prisoner Census Adjustment Act, the year before. “It would’ve required participation of state department of corrections—part of our bill required prisons to collect people’s address—and the Department of Corrections wasn’t thrilled about it because it was a lot of work,” Morita told me. A similar bill, called the No Representation Without Population Act, was filed in the Illinois legislature in 2015 and again in 2016, and both seem to be languishing in the same place, without a date to resume discussions.

States who have been early to make legislative changes have made it easier for others who want to follow.

As the 2020 Census approaches, progress is happening, albeit unevenly. The U.S. Department of Commerce, which runs the census, declined to change the criteria of “usual residence” last spring, snuffing out hope for federally-mandated reform in time for the next decade. (Using cost estimates from the state of Maryland, PPI estimates counting prisoners at their home address would cost a little more than $4 million for the entire U.S. prison population; roughly 1.3 million people.) But the Bureau did improve the way it publishes data, ultimately making it easier for states to identify prison populations among total census numbers, and adjust the totals on their own.

In 1950, the Census Bureau changed how it counted college students—though it still couldn’t make everyone happy—and redefined the same kind of criteria that would be needed to stymie prison-based gerrymandering. Recently, the Bureau has become more aware of the latter issue, as demonstrated by its willingness to make its data more useful for local governments who wish to exclude their prison populations from redistricting totals. “It’s hard to say how fast or slow change is moving because we’re looking at a time frame where change can only really happen at the federal level every ten years. So this is really their second attempt,” Kajstura said.

Meanwhile, the states who have been early to make legislative changes to end prison-based gerrymandering have made it easier for others who want to follow; there’s model legislation already in place. Advocates like Kajstra also see forward momentum at the state level, as state governments are increasingly attempting and passing bills. “New Jersey, as a legislature, passed a bill ending prison gerrymandering last session, which was then vetoed by then-Governor Christie, who’s now of course out of office. So there’s hope that once that bill gets passed again, the new governor will sign it rather than veto it,” Kajstura told me.

Four more states, including Illinois, have comprehensive legislation under consideration that would exempt prisoners from redistricting counts and track a prisoner’s home address with the secretary of state, increasing the chances they’ll be counted at home, even while in prison. Maryland and California have already passed such legislation, similar to New York and Maryland’s solutions, though these don’t take effect until 2020. And PPI reports at least two hundred local governments have also passed ordinances to ensure prison populations don’t skew redistricting at the county level and below.


Groups like PPI and other advocates have a clear goal when it comes to prison-based gerrymandering: ending it. The vision is to have incarcerated individuals counted in the districts where they lived prior to sentencing (and to be able to cast votes on their own behalf in these districts). This would mean fairer elections, more representative democracy, and decreased exploitation of an already-vulnerable group of people. But even with such a well-defined goal, what constitutes the “finish line” for such a problem is constantly shifting. And, even if you acknowledge that prison-based gerrymandering is a problem, its scope is so intractable, its effects so obscured, that it can be hard to articulate the urgency.

Scot Reynolds, a Michigan attorney who worked on Proposal 2, which recently put an end to other forms of political gerrymandering in the state, seemed to understand this struggle. “In the broader context, when you look at voter suppression [like] eliminating voter booths, implementing rules to prohibit voter turnout—which, in my personal and professional opinion, targets minority districts—I think when you put the prison issue in that context, it becomes just another brick in the wall,” Reynolds said. “These things all have a cumulative effect. If it’s a stand-alone issue—maybe not. But in the context of this broader campaign against voter suppression, I think it plays a role.”

That a problem like prison gerrymandering feels so unwieldy to begin with is a grim reflection on the state of U.S. democracy.

Ending prison-based gerrymandering is critical for Reynolds and others because it’s part and parcel of ending voter suppression, which is itself a crucial step toward better and more representative democracy. Reynolds’s observation illustrates precisely why it’s been such a challenge to mobilize lawmakers and officials and voters around ending prison-based gerrymandering—because which aspect of ending voter suppression is s really urgent when everything about ending voter suppression is urgent?

It’s easier to prioritize solutions when there’s a broader agreement on values: democratic participation, civic duty, fairness—issues that galvanize other “bricks,” as Reynolds would call them, in the wall of voting rights. Kajstura told me that the people who are most interested in how voter equity affects prisoners are the people who are already engaged in civil rights work, especially for the incarcerated. “Caring about voting rights of prisoners, is, I think, not very high on a lot of folks’ agenda,” she admitted.

Ultimately, the speed of change appears to come down to will, which itself is a kind of commentary on how sacred American attitudes are around voting. Even Paris Knox, who experienced prison-based gerrymandering firsthand, and wants to see accurate representation and voting rights for everyone, said it was sometimes hard to keep the issue at the top of mind. “Sometimes that’d be the last thing you think about, being locked up, because there’s so much other stuff going on. Being away from your family, waiting on letters from your family…there’s so much on the front burner, you’re not even talking about voting,” she said.

That a problem like prison gerrymandering feels so unwieldy to begin with is a grim reflection on the state of U.S. democracy. The fix itself isn’t complex, and, in a sense, solving the larger problem of voter suppression isn’t either. What’s harder is admitting that when it comes time to pull down the bricks as they form barriers to our democracy, we continue to look elsewhere, even as the wall forms. ■



This project was supported by Belt members, and by Rise Local, a project of New America.

Kim Bellware‘s reporting covers prisons, mass incarceration, the death penalty, politics, and culture. Her work has appeared in The New York Times, Rolling Stone, The Atlantic, Vice News, Chicago Magazine, and others. Previously, she served as a Chicago-based reporter for The Huffington Post.

Cover illustration by David Wilson.

Belt Magazine is a 501(c)(3) nonprofit organization. To support more independent writing and journalism made by and for the Rust Belt and greater Midwest, make a donation to Belt Magazine, or become a member starting at $5 per month.