by Aleks Kajstura, November 6, 2009

The U.S. Census is once again gearing up to count incarcerated people in the wrong place. This census data is soon going to be used to redraw districts at all levels of government throughout the country.

In “What the census will get wrong,” Mary Sanchez, of the Kansas City Star, writes, emphasizing the inconsistency between the Census Bureau’s “patriotic pitches to comply,” and the Census’ method of counting people in prison. According to Sanchez, we’re told that “[e]very breathing soul must be tallied during the massive federal endeavor, the national headcount taken every decade. The census is central to the functioning of our democracy….”

Sanchez correctly notes, however, that the Census reassigns prison populations. The Census data allows legislative districts to pad their numbers using disenfranchised constituents pulled from remote cities. The communities that have high incarceration rates lose their currently incarcerated residents in this count. Sanchez writes:

Criminals forfeit a lot when they get locked up. They lose the right to vote, in all but two states. They lose daily interaction with loved ones and the chance to engage in meaningful work.

The communities of origin for those incarcerated should not be similarly punished.


by Aleks Kajstura, October 16, 2009

report cover

In Pennsylvania, there are eight state house districts that owe their existence to large prison populations in the 2000 Census. With the 2010 Census poised to count an even larger portion of the state behind bars, the prison count will be even more important.

The Census counts prisoners where they are incarcerated rather than where they legally reside. The Census figures are then used by states to draw district lines, apportioning voting power among their citizens. Regardless that the prisoners are not in any way connected to the community, they are counted in the local population making up the district. According to a recent article in Philadelphia’s The Legal Intelligencer, for example, State Rep. Bob Mensch, R-Montgomery, who represents District 147 with a population of 3,404 prisoners, has never gotten a single call from a prisoner.

Because the prisoners cannot vote but are credited to the prison district, their presence adds weight to every vote cast in the district. Such a system cannot be tolerated in a democracy. Indeed the Supreme Court barred practices such as this when it held, in Reynolds v. Sims, that: “The weight of a citizen’s vote cannot be made to depend on where he lives.” The Supreme Court ruled that the 14th Amendment’s equal protection clause required that districts be drawn to be substantially equal in population.

According to our recent research, the vote distortion created by the Census count is not randomly distributed throughout the state. For example, while only 12% percent of Pennsylvania’s population is in Philadelphia, 40% of Pennsylvania’s prisoners are from the city. And virtually all — 99% — are incarcerated outside of the city.

That temporary placement however, does not change where Philadelphia’s incarcerated citizens reside under state law. Like most states, Pennsylvania law says that a prison cell is not a residence. But the growing prison population and an outdated method of counting that population combine to threaten how political decisions are made. Until the Census changes where it counts people in prison, the state of Pennsylvania needs to change how it uses Census data.


Report finds that crediting prisoners to the prison towns leads to unequal distributions of political power within Massachusetts.

October 8, 2009

The 2010 Census is rapidly approaching, but an old error threatens the count, charges a new report by the non-profit Prison Policy Initiative. The report explains that the U.S. Census counts state and federal prisoners as residents of the prison location, and that creates big problems for democracy in Massachusetts. Crediting prisoners to the prison towns leads to unequal distributions of political power within the state.

Under Massachusetts law prisoners — who can’t vote — are legal residents of their pre-incarceration homes. The Census Bureau counts these people in the wrong spot, and that will create a big problem when the state next updates its legislative districts after the 2010 Census. Legislative districts must be updated each decade to ensure that each district contains the same number of residents.

The report states that five Massachusetts House districts meet federal minimum population requirements only because they include incarcerated people as local residents. In these five districts, the presence of the prisons in the Census data enables every group of 95 residents near the prisons to claim as much political power in the State House as each group of 100 residents elsewhere. By using Census Bureau counts of prison populations to pad out legislative districts with prisons, Massachusetts is inflating the votes of residents who live near prisons in violation of the Supreme Court’s “one person one vote” rule.

“How the Census counts people in prison is a little understood or noticed problem,” said report author Elena Lavarreda, “but it’s important that the public know how the Census is diluting their votes.”

The report focuses on the harm to democracy caused by the 2000 Census, but warns that the Census Bureau intends to repeat the mistake of 2000. According to executive director of the Prison Policy Initiative, Peter Wagner, “The next Census is in 6 months, so unless the state acts fast, democracy will have to wait until the 2020 Census.”

The report, “Importing Constituents: Prisoners and Political Clout in Massachusetts”, is available at http://www.prisonersofthecensus.org/ma/


by Peter Wagner, September 30, 2009

In a Tribune-Star (Terre Haute, Indiana) column, Mark Bennett argued that the Census Bureau’s decision to credit the population of the federal prison to the city of Terre Haute was a boon to the region. But while celebrating one regional benefit, he missed a bigger problem for democracy at home, as I point out in my letter to the editor:

The right way to count prisoners

Tribune-Star columnist Mark Bennett writes that the Census Bureau’s practice of counting federal prisoners as if they were residents of Terre Haute (“Multiple factors lead to Terre Haute posting ‘pretty impressive’ population growth“, July 18, 2009, Page D1) misses one thing: City council districts are also based on population.

Each decade, by Supreme Court precedent, districts must be redrawn so that each district contains the same number of residents. In this way, each resident gets the same access to government regardless of where she lives. Unfortunately, the Census count is not the same as the number of residents and this problem is about to get worse in Terre Haute.

When the Terre Haute City Council last updated the districts after the last Census, they unintentionally padded the First District with 1,764 prisoners, granting the 7,778 residents of that district as much say over the future of the city as 10,000 residents in each other district. If uncorrected after the 2010 Census, the expanded prison will be 30 percent of the district, creating an even larger vote dilution problem.

Unfortunately, the Census Bureau won’t be changing where it counts prisoners in next year’s Census. But there is an interim solution: ignore the prison populations when drawing council districts. Give every resident the same say over city council regardless of whether their district happens to contain a prison.

In my research, I’ve found more than 100 places like Terre Haute where the local government rejected the Census Bureau’s prison counts and choose instead to base their democracy on their resident population. The Census may count the prisoners as residents of Terre Haute, but the city isn’t bound by that count for its local districts.

Peter Wagner
Executive Director
Prison Policy Initiative
Easthampton, Mass.
July 25, 2009

It is sometimes tempting to portray the controversy over where incarcerated people should be counted as a “tug-of-war between cities and rural communities”, but that runs the very real risk of obscuring how the current Census policy hurts rural communities right now.

The impact of prison-based gerrymandering on state legislative districts is real and significant, but the impact on local legislatures, such as county boards and city councils, is even more pronounced. Because the district sizes tend to be relatively small in local government, a single prison can have a significant effect. It is for that reason that many of the people working the hardest for Census reform are rural people who live near, but not immediately adjacent to, large prisons.


Report finds 7 Okla. House districts meet federal minimum population requirements only because prison inmates are used to pad the districts' populations.

September 25, 2009

The federal Census counts prisoners as part of the local population, and that creates big problems for state and local government, charges a new report by the non-profit Prison Policy Initiative.

“Governments rely on the Census to count the population so they can update legislative districts,” said Prison Policy Initiative Executive Director and report co-author Peter Wagner. The Supreme Court’s “One Person One Vote” rule requires that legislative districts each contain the same number of people, so that each person has the same access to government.

Where prisoners are counted is a particularly acute problem in Oklahoma. “Oklahoma incarcerates a greater portion of its population than 46 other states, so crediting prisoners to the wrong communities has staggering effects on democracy,” said report co-author Elena Lavarreda.[1]

The report finds that 7 House districts meet federal minimum population requirements only because prison inmates are included in the count. By using Census Bureau counts of prison populations to pad out legislative districts, Oklahoma is inflating the votes of residents who live near prisons in violation of the Supreme Court’s “one person one vote” rule.

“The problem is even larger in some rural county board districts,” said Wagner. The report, Importing Constituents: Prisoners and Political Clout in Oklahoma, finds rural county districts that are as much as 46% prisoners. “This allows the real residents of a district with a prison to unfairly dominate their county government.”

The report identifies 16 counties where prison populations have a large impact on democracy in county government, and applauds Greer County for rejecting the flawed Census counts and drawing districts without regard to the prison miscount.

The report calls on Oklahoma to lobby the Census Bureau to change how prisoners are counted in the future and to develop state solutions to protect the restricting process after the 2010 Census.

The report, “Importing Constituents: Prisoners and Political Clout in Oklahoma”, is available at http://www.prisonersofthecensus.org/ok/

[1] Oklahoma Sentencing Commission, Bureau of Justice Statistics, Prison Inmates at Midyear 2007, Spreadsheet table pim07at05.csv, June 2008. Oklahoma has 670 people serving a state prison sentence for every 100,000 people in the state. The 3 states with higher incarceration rates are Louisiana (857), Mississippi (723) and Texas (682).


by Peter Wagner, September 14, 2009

The 23rd decennial Census will again be counting incarcerated people in the wrong place. But if a proposed constitutional amendment in Wisconsin passes, the state’s days of using prison counts to distort districts (and influence elections) will be over.

Currently, one state legislative district is 10% prisoners, giving the residents of that district disproportionate say over state affairs. And because county board and rural city alderman districts are so much smaller, the inclusion of a single state prison in these districts can allow a handful of residents to dominate the county board or city council.

Ideally, the Census Bureau would change where it counts people in prison. But the pace of change has so far been slow, and with time running short before the next Census in April, the number of options are limited. As the Milwaukee Journal-Sentinel editorialized in 2006:

Congress has asked the Census Bureau to report on how it could change the way it counts inmates, who now number about 1.5 million nationally. The National Academy of Sciences is also doing a study.

Yes, inmates really live in prison, but in no sense are they part of the community in which they are imprisoned. The census must count every U.S. resident, but it needn’t count them this way.

Neither Congress nor the Census Bureau really needs a report here. The bureau should change how it counts inmates, and if it doesn’t, Congress should mandate it.

Unfortunately, the Census Bureau responded with an “obtuse and evasive report” and Congress failed to follow up. The National Academies instructed the Bureau to conduct research on the best way to change how incarcerated people are counted, but there too, the Bureau failed to move forward. By now, the Bureau has squandered too much of the critical planning time to make the change before the next Census in 2010.

Now, the burden to eliminate prison-based gerrymandering is shifting to the states. In Wisconsin, Rep. Frederick Kessler has introduced a constitutional amendment that would require the state and local governments to ignore the prison populations when drawing districts. This approach would not credit incarcerated people back to their homes in Milwaukee and other urban areas, but it would end the practice of crediting them to the rural districts where they count for as much as a tenth of a state legislative district.

Other states are exploring options that would identify where incarcerated people are from and then adjust the federal Census counts to use the home addresses. The Wisconsin amendment [PDF] takes a simpler approach that will address the majority of the problem.

At 10am tomorrow (September 15th), I’ll be in 300 Northeast at the Wisconsin State Capital to testify in support of the Wisconsin Census Correction amendment which would direct state and local governments to omit the Census Bureau’s incarcerated population when drawing legislative districts.


by Peter Wagner, September 10, 2009

Eli Lehrer writes in the conservative National Review Online that the Census Bureau should change where it counts incarcerated people:

“Prisoners aren’t, in any real sense, residents of the places where they are incarcerated…. Their victims and families very likely live in the places where they came from. In short, the costs that prisoners impose on society fall on the municipalities where they lived and will probably return after release.”

See: Eli Lehrer, How Should the Census Count Prisoners? National Review Online.


by Peter Wagner, September 9, 2009

The residents of Maine’s Regional School Unit 13 have launched a petition drive to end prison-based gerrymandering in their school district. The school district’s voting system is based on Census Bureau estimates for 2006 that credited the town of Thomaston with the population of the Maine State Prison that had closed 4 years prior. Not only does Maine state law say that a prison is not a residence, the prison counted in the Census did not even exist at the time of the count.

Unfortunately, the school board has refused to change the voting system, and to date the Commissioner of Education has ignored requests [PDF] to intervene.

By Maine law, voters can use a petition to require the Commissioner to rule on whether a voting system violates the constitutional principles of One Person One Vote.

The petition is only open to the residents of Cushing, Owls Head, Rockland, South Thomaston, St. George, and Thomaston Maine, but this page of question and answers explains the campaign.

Sept 10 update: St. George Select Board Chairman William Reinhardt tells the Knox County Herald Gazette that he supports changing the RSU 13 apportionment to give his town a fairer voice on the school board.


by Peter Wagner, July 23, 2009

Amaris Elliott-Engel blogs about my Tuesday talk at the National Conference of State Legislatures summit on the The Legal Intelligencer Blog: Change is Predicted in Census Prisoner Count.

The 2010 census will be the last to count inmates at their place of incarceration instead of their home communities, an advocate for changing where the U.S. census counts the incarcerated predicted at the National Conference of State Legislatures Tuesday.

But while it is likely too late in the current U.S. Census process to change where prisoners are counted, every state can prevent “prison-based gerrymandering” by omitting prisoners for purposes of legislative districting at the state level, said Peter Wagner of Prison Policy Initiative, an advocacy group based in Northampton, Mass.

Wagner said as long as state officials start now there is enough time for states to take out prisoners from the population data used by elected officials in drawing up legislative districts.

State legislators from Vermont to Wisconsin to Texas all spoke during the conference session in support of adjusting where prisoners are counted.

My talk focused on technical and legal issues involved in state-level changes in how people in prison are counted for redistricting purposes. Also see Amaris Elliott-Engel’s June feature article in The Legal Intelligencer about our Pennsylvania report: Report: Census Prisoner Count Dilutes Urban Political Clout


by Peter Wagner, July 17, 2009

On Tuesday, the delegates to the NAACP’s 100th annual convention approved a resolution calling for an end to prison-based gerrymandering:

THEREFORE, BE IT RESOLVED, that the NAACP, on principle, decries the enumeration of prisoners as local residents as violation of our nation’s fundamental one person one vote ethos of representational democracy, harkening back to the disgraceful three fifths era of constitutionally sanctioned slavery; and

BE IT FURTHER RESOLVED, that the NAACP calls on the U.S. Department of Commerce Bureau of the Census to enumerate prisoners within census blocks where domiciled at their time of arrest; and

BE IT FINALLY RESOLVED, that NAACP units call upon their Congressional representatives to effect such a permanent change to the Census Bureau enumeration procedures.

Continue reading →



Stay Informed


Get the latest updates:



Tweet this page Donate