Today, more than 2 million people, or nearly one out of every 100 adults, is sitting in a jail or prison in the United States -- an incarceration rate unprecedented in U.S. history.
Marie Gottschalk is associate professor of political science at the University of Pennsylvania and the author of "The Prison and the Gallows: The Politics of Mass Incarceration in America" (email@example.com).
The total number of prisoners is not in dispute. But how to tabulate them is emerging as perhaps the most controversial issue for the 2010 census.
The U.S. Census Bureau counts prisoners as residents of the towns and counties where they are incarcerated, even though most inmates have no ties to these communities and almost always return to their home neighborhoods upon release.
This has enormous and unsettling political consequences, especially for Pennsylvania. The state banishes many of its urban offenders to prisons in rural areas and is home to a vast archipelago of federal penitentiaries that incarcerate thousands of people from out of state.
In Pennsylvania and 47 other states, imprisoned felons are barred from voting. Yet these disenfranchised prisoners are included in the population tallies used to draw legislative and congressional districts.
This practice dilutes the votes of urban areas such as Philadelphia and Pittsburgh. Nearly 40 percent of the 45,0000 inmates in Pennsylvania's state prisons come from Philadelphia. For census and redistricting purposes, these urban citizens "reside" in counties far from their homes, often in rural districts that are Republican strongholds.
The evidence of political inequities in redistricting due to how the Census Bureau counts prisoners is "compelling," according to a recent National Research Council report.
A provocative analysis by the Prison Policy Initiative estimates that if prisoners held in upstate New York were counted in their home neighborhoods, at least four state Senate seats -- all Republican -- would be in jeopardy after redistricting. Last year, a federal appeals court suggested that counting tens of thousands of black and Latino prisoners from New York City as upstate residents may be a violation of the federal Voting Rights Act.
Nearly 200 counties nationwide, including three in Pennsylvania, have at least 5 percent of their "residents" in prison. Union County, home to several federal penitentiaries, has a whopping 13 percent of its residents behind bars. Those prisoners also distort demographic data. Union County is 97 percent white, but only 90 percent if you count the 5,000 prisoners.
About a dozen counties in Pennsylvania reportedly doubled or tripled their African-American population between 1990 and 2000. But Pennsylvania is not undergoing a mass migration of African Americans. In many instances the sharp increase was due to the opening of a new prison.
Such data make rural prison communities appear, on paper, more ethnically diverse and less affluent. Prisoners' negligible income artificially depresses the per-capita income of prison towns, which ironically then become more eligible for social programs for the poor.
Last year, in a report to Congress, the Census Bureau recoiled from counting prisoners in their home communities. The bureau cited prohibitive costs as well as security concerns, because inmates might have to be individually interviewed by census personnel. But these objections are unconvincing. In many ways, prisoners are cheaper, easier and safer to count than other populations, such as migrant workers or the homeless.
The 2010 census is rapidly approaching, and it may be too late to force a stubborn Census Bureau to change its ways for this go-around. But states can do some things on their own. Pennsylvania legislators could require the state to collect the home addresses of inmates and to adjust the census data before redistricting accordingly. A second-best solution is for the state to subtract its prisoners from the census data before redistricting. At least a dozen counties in upstate New York already do this before drawing their county legislative districts.
State legislators in Pennsylvania and elsewhere also should demand that the Census Bureau conduct a major test in the 2010 census of alternative ways to tally inmates -- as was recommended by the National Research Council.
How prisoners are counted is a critical civil rights issue. Currently 2.5 percent of African Americans nationwide are incarcerated, and 9 percent of African-American men in their 20s are in jail or prison. This issue is particularly salient for Pennsylvania. The state's black-white disparity in incarceration is among the highest in the nation, and its incarceration rate for Hispanics is tops in the country, according to a new report by The Sentencing Project.
The stance of the Census Bureau is reminiscent of the ignoble compromise of the constitutional convention when the founders agreed to count each disenfranchised slave as three-fifths of a white person. This decision allowed the slave-holding South to maintain its dominance in national politics for decades.