The Prison Town Advantage

To say that Danny Young did not win his seat on the Anamosa, Iowa, City Council by a landslide is an understatement of extreme proportions. Young won that seat in 2006 with just two write-in votes, one of them cast by his wife. The definition of the reluctant politician, Young didn’t even vote for himself. There were no other candidates for the seat.

Admittedly, no candidate in Anamosa is going to draw big numbers at the polls, given the city’s tiny size. The community covers 2.2 square miles, and has about 5,700 residents. At the time of Young’s election, voters elected representatives from four wards, each of which included approximately 1,400 residents, as well as two at-large councilors.

But in Young’s Ward 2, fewer than 100 of those residents were eligible to run for the Council seat, or even to vote. That’s because the ward is dominated by Anamosa State Penitentiary, a maximum-security prison where about 1,300 men are incarcerated. And in Iowa, as in 47 other U.S. states (Maine and Vermont are the exceptions), incarcerated felons are not allowed to vote.

They are, however, counted by the U.S. Census Bureau for data used to draw congressional, state and municipal legislative districts. That practice can lead to dramatic power imbalances between communities that have prisons and those that do not—as seen in Anamosa, where the fewer than 100 non-prisoners in Young’s ward have as much representational clout in city affairs as the 1,400 residents in each of the city’s other wards.

While Anamosa presents a particularly dramatic example of the problem, this imbalance exists in communities around the country. The Prison Policy Initiative, an Easthampton-based nonprofit, has released numerous reports in recent years examining the problem in states around the country; this month, PPI is releasing a report, “Importing Constituents: Prisoners and Political Clout in Massachusetts,” that looks at the effects here. The report, co-authored by PPI Executive Director Peter Wagner and colleagues Elena Lavarreda and Rose Heyer, finds that five of the state’s legislative districts would not even exist in their current configurations if their population counts did not include prison inmates.

This apparently unintended data-gathering quirk, Wagner said, has profoundly detrimental consequences for the distribution of political power—consequences that extend further than one might expect.

Counting disenfranchised prisoners to draw up legislative districts “makes no sense,” Wagner said, “and is actually offensive to our notion of democracy.”

It also bears, in the words of Boston-based voting rights attorney Brenda Wright, an “uncomfortable resemblance” to the “three-fifths” compromise between Southern and Northern states written into the U.S. Constitution in 1787. That provision declared that a slave would count as three-fifths of a person for the purposes of apportioning congressional districts.

“The slave states benefited in terms of political power, based on a population that couldn’t vote,” said Wright, who directs the Democracy Program for Demos, a public policy and advocacy organization. More than 220 years later, legislators with prisons in their districts are likewise benefiting from a population that’s also denied the vote—while other districts lose.


Peter Wagner began studying prison-based gerrymandering while a law student at Western New England College. His first project looked at neighboring New York State, where the effects are especially dramatic. There, Wagner noted in a 2002 report, 91 percent of prison cells are located in the upstate region, whose economy depends heavily on the prison industry. But only 24 percent of prisoners actually come from upstate New York; the majority—66 percent—comes from New York City.

As a result, Wagner said in a recent interview, “the whole center of gravity shifts.” For state legislators who have prisons in their districts, the facilities are a boon: the prison population swells local numbers enough to justify the creation of a legislative seat, while the prison creates jobs and spurs related economic activity in a part of the state that sorely needs both. According to PPI, seven legislative districts in upstate New York would not have the minimum population required for a district were it not for their prisoners.

But not everyone wins under this scenario. While upstate legislators may have prisoners in their districts, because those prisoners cannot vote, there’s no incentive for the legislators to support policies that could positively affect the urban districts where the majority of prisoners come from. Meanwhile, because the prisoners are not counted in their hometowns, those communities’ populations, for the purposes of creating legislative districts, drop.

“Prisoners and their families have negative political clout,” Wagner said.

And it’s not just prisoners (and the family and neighbors that remain in their hometowns) who feel the effects of this imbalance, Wagner noted. Residents who live in districts without prisons have, in essence, less political influence than those in districts that do have prisons.

“These … districts get an enhanced say, which hurts every other district in general, and hurts the district where prisoners come from even more,” Wagner said.

Meanwhile, prisoners—despite the fact that they contribute to a prison-district legislator’s political power—have no political influence over “their” representative. “The way things should work is, if a legislator doesn’t represent some of his or her constituents, there’s a check in place—the overlooked residents can vote that person out,” Wagner said. “But when some of those constituents can’t vote, that natural check and balance doesn’t work.”

As Danny Young, the Anamosa councilman, put it in a 2008 New York Times article: “Do I consider [the prisoners within my ward] my constituents? They don’t vote, so, I guess, not really.”


The Census Bureau’s policy of counting prisoners where they’re incarcerated is not new, PPI notes.

But the effects of that policy have become more significant in recent years, as the U.S. prison population has swelled, thanks, in large part, to the trend toward mandatory-minimum sentencing and other “tough on crime” legal reforms. According to the federal Department of Justice, in 1998, there were slightly fewer than 1.3 million people in state and federal prisons in the U.S. A decade later, that number had risen to 1.5 million. (During the same time period, the number of people, nationally, in local jails rose from about 500,000 to 800,000. In Massachusetts and other states, people who are behind bars on misdemeanor convictions or while awaiting trial are eligible to vote.)

There are about 11,000 people in Massachusetts state prisons, according to a 2007 report by the Mass. Department of Corrections. The state’s one federal prison, Fort Devens, has about 1,300 prisoners, according to the Federal Bureau of Prisons.

Until 2000, prisoners in Massachusetts had the right to vote (those who chose to did so by absentee ballot). That year, 60 percent of voters—following the lead of then-Acting Gov. Paul Cellucci—approved a ballot question to amend the state constitution to deny voting rights to prisoners locked up on felony convictions. (The Cellucci administration had previously shut down a political action committee formed by a group of state prisoners, successfully arguing in Superior Court that allowing political activity in prisons presented a security threat.)

The result—in Massachusetts and in the 47 other states where incarcerated felons can’t vote—is that lawmakers derive political power from “constituents” who are legally denied a voice in the political process. By law, each legislative district in Massachusetts should include 39,682 people, “Importing Constituents” notes. The law does allow some deviation from that figure, to ensure that other goals can be met, such as keeping communities with shared interests together in one district. (For instance, the lines might be drawn to keep members of a racial minority in the same district.)

In Massachusetts, that built-in wiggle room allows a district to deviate from the 39,683, in either direction, by 1,984 people—meaning a district may actually have a maximum of 41,667 residents or a minimum of 37,699.

But five state legislative districts in Massachusetts meet that minimum number only because they contain prisons, the PPI report points out. They include the 37th Middlesex district of Democratic Rep. Jennifer Benson, which would not meet the minimum had the 2000 Census not counted the 3,013 prisoners at Fort Devens and three state prisons (one of which, MCI Lancaster, was closed after the district was drawn). Similarly, the 9th Norfolk district of Republican Rep. Richard Ross only meets that threshold because of the 2,596 people who were in its four state prisons at the time it was drawn.

In addition, the 3rd Suffolk district of Democrat Aaron Michlewitz (previously represented by former House Speaker Sal DiMasi, indicted earlier this year on federal corruption charges) counts more than 1,500 Suffolk County House of Corrections inmates in its population total. Without those inmates, the PPI report found, the district would in fact be more than 8 percent smaller than the state’s average district. And the 14th Worcester district, represented by Democrat James O’Day, only meets the minimum because of prisoners at the Worcester County House of Correction.

Locally, the 7th Hampden district of Rep. Tom Petrolati, a Democrat and speaker pro tempore of the House, contains 38,144 people. But 1,660 of those people were counted at the Hampden County House of Corrections in Ludlow; without them, the population would fall to 36,484—again, below the legally required minimum. “The actual population of this district is more than 8 percent smaller than the average district in the state, giving every group of 92 residents in Ludlow and some of the surrounding areas as much political power as 100 residents elsewhere in the state,” the PPI researchers wrote.

A number of inmates at the Ludlow jail—those awaiting trial, and those there for misdemeanors—do, in fact, have the right to vote. But for the most part, “they are credited to the wrong district,” Wagner said, with the exception of those who also happen to be residents of the 7th Hampden district. The rest must vote by absentee ballot in the district where they previously lived.

“The folks in the Hampden County House of Correction [who have the right to vote] are being represented and are voting in other districts, but their presence in the data used to draw the districts enhances the weight of a vote cast by the actual residents of Rep. Petrolati’s district,” Wagner went on. “That ends up turning the concepts of ‘One Person, One Vote,’ and of basing districts on common communities of interest, on their heads.”

Four of the five legislators whose districts benefit from the Census practice—Benson, Ross, Michlewitz and O’Day—were not in the Legislature when the districts were last redrawn, in 2001. (Petrolati was, and, in fact, served as chairman of the House’s redistricting committee, under then-Speaker Tom Finneran. In 2005, Finneran was indicted on federal charges of perjury and obstruction of justice, for allegedly lying about intervening in the redistricting plan to protect certain incumbents. In 2007, he pleaded guilty to one count of obstruction of justice, and received 18 months probation and a $25,000 fine. Petrolati, a Finneran lieutenant, was questioned by investigators but not charged. He did not respond to an interview request from the Advocate.)

Benson, the Middlesex legislator, told the Advocate she didn’t know about the Census’ policy for counting prisoners when she ran for the office, “and was surprised to discover it. …

“I agree that counting prisoners as residents of the towns in which they are incarcerated is counterproductive and that our democracy is based on one man, one vote. [T]herefore equal representation is essential to upholding this belief,” Benson said.

While Massachusetts does not present the extreme cases seen next door in New York, “this small and seemingly benign thing actually affects how our democracy runs,” said report co-author Elena Lavarreda. The imbalance of power created by the census policy hurts all Massachusetts districts without prisons, but it especially hurts the urban areas where prisoners disproportionately come from, Lavarreda added.

For instance, while Boston accounts for 9.1 percent of Massachusetts’ total population, according to 2008 figures from the Mass. Department of Corrections, 15 percent of new court commitments to state prisons reported home addresses in that city. Springfield, meanwhile, accounts for 2.3 percent of the state’s total population, but 9 percent of its state prison population. Holyoke accounts for just 0.6 percent of the state population, but 2 percent of its state prisoners.

“[H]eavily minority urban districts would be entitled to additional representation if prisoners were counted as residents of their home communities for purposes of redistricting,” the PPI researchers wrote.

The U.S. Census Bureau will conduct its next decennial population count in 2010. And it will continue its practice of counting prisoners in their prisons, not their hometowns—this despite the advocacy work of PPI, as well as a 2006 report commissioned by the Census Bureau from the National Academies’ National Research Council. That report—titled “Once, Only Once, and in the Right Place: Residence Rules in the Decennial Census”— described the bureau’s guidelines for determining residency for prisoners and certain other populations as too complicated, and urged the agency to improve that methodology in advance of next year’s count.

Those changes, however, have not happened. Rather, the Census Bureau has been entangled in other matters, including a stand-off with Congressional Republicans who stalled the confirmation of President Obama’s pick to lead the agency, Robert Groves; problems with the effectiveness of new hand-held computers used by canvassers; and the public relations nightmare caused by the agency’s (now-severed) relationship with the controversial community organizing group ACORN.

“The Census is a big bureaucracy, and big bureaucracies are hard to change,” said Lavarreda, adding that she believes many individuals within the Bureau do see the problem created by the current policy. With the next census just around the corner, political leaders are increasingly talking about the importance of getting an accurate population count (including in western Mass., where, it’s feared, population shifts could result in the loss of a Congressional seat). The Census Bureau’s “miscount” of prisoners only adds to those concerns about the count’s accuracy, Lavarreda said.

“The Census Bureau is averse to change,” added Wright, the Demos attorney. “It’s an institution that does very long-term planning for the work it does”; to make any changes to that system “takes a lot of momentum, a lot of impetus.”

Last week, the Advocate contacted the Census Bureau’s Public Information Office seeking a response to the PPI report. At deadline, the office had yet to respond to that interview request.”

By failing to address the prison issue in time for the 2010 count, the Census squandered an important opportunity, PPI contends. But in the absence of reform on the federal level, state and local governments can still address the problem, the organization points out.

While the U.S. Constitution mandates the decennial census, that document only requires that the numbers be used to draw Congressional districts. Over time, states, for the sake of convenience, have come to use the federal census numbers for drawing their own legislative districts, but they are not required to; Massachusetts, for instance, conducted its own census every 10 years from 1855 to 1975. If states conducted their own censuses, PPI suggests, they could ensure that prisoners are counted at their pre-incarceration addresses, and then use those results to draw legislative districts.

“The irony is the legislators who benefit from this [existing policy] is a very small list,” Wagner noted. “Everyone else loses in some way—some more than others.” Still, he said, there appears to be little momentum in the Legislature for addressing the issue, perhaps because some legislators aren’t even aware of its implications.

Fixes can also take place at the local level. In Anamosa, Iowa, citizens addressed the imbalance of power caused by the local state prison through a referendum that changed how city councilors are elected. Starting with next month’s election, instead of electing councilors from each of four wards—including Danny Young’s, where more than 90 percent of his “constituents” are disenfranchised prisoners—all members will be elected at-large.

Closer to home, the Worcester County city of Gardner opted not to count inmates at North Central Correctional Institution at Gardner when redrawing its City Council districts in 2001 to avoid granting too much political influence to the part of the city where the facility is located.

While PPI applauds efforts like those in Anamosa and Gardner, ultimately, Wagner said, “The ideal place to change this is at the Census Bureau.”

The Bureau, he went on, applies a “usual residence rule” to determine where to count a particular individual. In the words of the Census Bureau: “Usual residence has been defined as the place where the person lives and sleeps most of the time”—meaning, for instance, that a person who is on vacation the day of the census would nonetheless be counted at his or her home.

A person’s “usual residence,” according to the Census Bureau, “is not necessarily the same as the person’s voting residence or legal residence.” College students who do not live at home, for instance, are counted at their college housing, even if their parents’ home is their legal residence.

Like college students counted at their dorms, not their parents’ houses, prisoners also “live and sleep most of the time” at their prison, not their previous home. But, PPI points out, there’s a key difference. Unlike college students, people in prison are not there voluntarily. “Students are welcome and encouraged to purchase local goods and services and to rent apartments in town,” the group noted in a related report. “They can register to vote in the community and may decide to stay after graduation. … For the duration of their time at the college, the college is the place they willingly live: that’s the very definition of residence.”

The Census’ decision to count prisoners within prison districts, Wright said, is “based on a fiction that prisoners who can’t vote and are not a permanent part of the community should be treated as though they are”—at least for the purposes of drawing legislative districts. Interestingly, she noted, in other legal matters, such as marriage and divorce laws, prisoners are considered residents of the communities where they lived before they were locked up. Indeed, before incarcerated felons were denied the vote in 2000, courts had ruled against prisoners who wanted to vote in the towns where their prisons were located, saying instead that they must vote in their home communities, by absentee ballot.

The Prison Policy Initiative suggests some ways the Census Bureau could change how it counts prisoners: it could allow prisoners to respond with their pre-incarceration addresses, or use prison records to determine those addresses. The Bureau has made similar adaptations for other populations in the past, PPI notes.

The U.S. Census Bureau, Wagner noted, needs individual states to rely on its figures for their redistricting efforts; if states started conducting their own counts, it would, no doubt, hurt the federal bureau’s funding. Like any business, the Census should view the states as clients, and the states should exercise the client’s right to demand a good product—in this case, fair and accurate population figures. “Part of the incentive to fix this,” Wagner said, “is the aggrieved party is every single person who does not live immediately adjacent to a prison.”

Author: Maureen Turner

Share This Post On

1 Comment

  1. because the prisoners are not counted in their hometowns, those
    communities’ populations, for the purposes of creating legislative
    districts, drop.

    Post a Reply

Submit a Comment

Your email address will not be published. Required fields are marked *

Sign up for our daily newsletter!

You don't want to be left out, do you?

Sign up!

You have Successfully Subscribed!