Significant political stakes ride on how the census counts prisoners

Should the census count inmates as residents of the prisons where they're held — often hundreds of miles from home? Or should they be tallied as citizens of the cities or counties they came from?

An agreement just reached between the U.S. Census Bureau and Rep. William Clay Jr., D-Mo., the chairman of the House subcommittee that oversees census issues, may signal a historic shift in how the bureau reports prisoners to state and local governments. The accord creates at least a chance for prisoners' overwhelmingly urban home areas to get a better break on legislative representation.

Counting prisoners where they're incarcerated didn't matter a lot when America had modest numbers of inmates, usually held in institutions near their homes.

But all this has changed in the past three decades as America's prisoner counts have soared from about 500,000 in 1980 to 2.3 million today. The combination of tough "law-and-order" politics and the development of a vast "prison industrial complex" has led to confinement of predominantly city-based convicts in hundreds of new prisons in small-town areas.

Since the 1970s with their economies and populations declining, rural interests have campaigned hard for the prisons as a source of jobs for local residents. State legislators representing those areas back up the towns' efforts and are likely to support indiscriminate "lock-'em-up" strategies. Increasingly powerful prison-guard unions take the same side and back pro-incarceration legislators in campaigns.

But far-from-home incarceration is bad penology. Inmates who are held at outstate (sometimes even out-of-state) locations receive fewer family visits. More isolated, they're destined to have a harder time readjusting upon release. The practice feeds a negative spiral leading to more recidivism and demands for still more prisons.

But there's another impact: Because the census historically counts inmates where they're imprisoned, their numbers actually swell population counts — and legislative representation — for rural areas. The losers in political clout are then the very urban areas most of the prisoners come from.

The Prison Policy Initiative, an advocacy group that documents the impacts of mass incarceration, found 21 counties across the nation where at least 20 percent of the population were prisoners from other counties. The reformers' cause célèbre is Jones County in eastern Iowa, where a backhoe operator was elected to the Anamosa City Council with just two write-in votes, one from his wife. Why? Because 95 percent of his ward's population consists of 1,300 inmates in Iowa's largest penitentiary — and none of them can vote.

Seven state Senate districts in upstate New York, the Policy Initiative has calculated, would not have met minimum population requirements in the past decade if they had not had significant prison populations. California, Texas, Illinois, Iowa, Wisconsin and Maryland are among other states in which the census count of inmates held in lightly populated small-town areas can impact the political balance.

The distortion makes it less likely that a whole range of smart penal-reform measures — electronic monitoring in place of imprisonment, community-based re-entry programs and halfway houses, basic education and post-release employment programs — will receive adequate state support and funding.

The agreement Rep. Clay has negotiated with the Census Bureau doesn't reach the full solution that reformers would like — to count all inmates as residents of their home cities and counties, not their prison addresses.

But the bureau has agreed to make an early release, in 2011, of its counts of "group quarters" such as prisons. States will then have time to decide how, for purposes of the legislative reapportionment they'll be starting next year, their prisoners should be counted: where the prisons are, or where the prisoners last resided before incarceration. Or they can choose not to count them for reapportionment purposes at all.

How the issue plays out over the long run will influence not just state politics but big decisions in federal grant policies too. The Census Bureau itself last year identified about $436 billion worth of grant and direct assistance money that's "allocated based on Census Bureau data," with the largest items being Medicaid ($203 billion), unemployment insurance ($36 billion), and highway funding ($34 billion). Nutrition, school and college aid, school lunches and Head Start are also impacted.

The reality is that prisons aren't the only issue shortchanging urban areas. As the National Research Council confirmed in a 2009 study, there's a long history of "differential net undercount" of blacks and Hispanics that "has led to their receiving less than their share of federal funds and political representation."

Correcting the imbalance, compounded by many low-income residents' fears of government and/or immigration laws, is no easy task for census takers. But there should be no easier place to start than counting prisoners at their home addresses, not their prison cells.

Neal Peirce's column appears regularly on editorial pages of The Times. His e-mail address is nrp@citistates.com