WASHINGTON (AP) -- Prisoners will soon be bigger players in those high-stakes redistricting fights, even if unwittingly, thanks to a change in federal policy governing how they're to be counted in the 2010 census.
Prison populations have historically been included in national headcounts, but now Census officials will make data on inmate populations available to states earlier than in the past.
This change will allow states to decide whether to count inmates for purposes of redistricting. If a state makes that choice, it would have to decide where inmates should be considered residents - in rural towns, where prisons are often built, or in cities, where many prisoners come from.
Until now, the bureau provided breakdowns on group quarters, like prisons, only after states had finished their redistricting. That resulted in districts with prisons getting extra representation in their legislatures, despite laws in some states that say a prison cell is not a residence.
The jockeying is all part of a decennial rite - counting the population. The federal government relies on the census not only to learn about Americans and their lives but also to parcel out federal dollars. As required by the Constitution, the census also is used to determine the number of U.S. House seats representing each state.
The census officially began last month in rural Alaska, and most of the U.S. will receive forms by mail the week of March 15.
In this week's policy change on prisoner counts, Census officials said they would release data on prison populations to states when they redraw legislative boundaries next year. This gives states more leeway in tallying their prisoners - a move that could reshape the political map.
Census director Robert Groves made the decision after weeks of discussion with Rep. William Lacy Clay, D-Mo., and with public interest and black groups. They called it an important first step toward shifting federal resources and representation back to urban communities, where they believe the aid is needed the most.
"For too long, communities with large prisons have received greater representation in government on the backs of people who have no voting rights in the prison community," said Brenda Wright, director of the Democracy Program at Demos, a research and advocacy organization. "The Census Bureau's new data will greatly assist states and localities in correcting this injustice."
The impact could be strongly felt in states such as New York, Illinois, Iowa, Wisconsin, California, Texas and Maryland, where prisons are found in more sparsely populated areas. In New York, for instance, most of the 60,000 inmates live in prisons in rural upstate communities, even though half the inmate population committed crimes in New York City.
In Anamosa, Iowa, which boasts a population of roughly 5,700, some 95 percent of the residents are prisoners, none of whom can vote, according to the Prison Policy Initiative.
Still, analysts say the Census Bureau's move could prove politically messy, with the devil in the details. They note that the agency will not release the prison data until May 2011, more than two months after states are given their initial population data by district, so some legislatures may opt not to wait for the additional information or only make cursory use of it.
Also, while the prison data will have breakdowns on where inmates are located, it will not include information on the prisoners' original hometowns. Thus, states will have to gather that information on their own if they choose to count them in different locations.
"This is going to be a big enough deal where states will have to make some decisions," said Kimball Brace, president of Election Data Services, a Virginia-based firm that crunches political numbers. "We may see an impact ultimately where one political party decides to go one way and draws districts accordingly, the other party goes another way, and we end up with a court case to sort it out."
The population count, held every 10 years, is used to apportion U.S. House and state legislative and county seats as well as distribute more than $400 billion in federal aid.
New Yorker Chevelle Johnson, 44, who said he was formerly incarcerated, returned to his community of Bedford-Stuyvesant, in Brooklyn, upon being released from prison in 2007.
"A lot of us come home and we can't even vote," he said. "We need political power in our communities so that when we do come home, we come home to something ... to things that will help us not get reincarcerated."
While the 2010 data will not include hometown information, advocacy groups say they are continuing their push for prisoners to be counted as residents of the communities they came from for the next decennial census in 2020.
"Because incarcerated persons in the United States are disproportionately African Americans and other people of color, the current count of prisoners at their place of incarceration, rather than at their pre-arrest residence, severely weakens the voting strength of entire communities," said John Payton, director-counsel of NAACP Legal Defense and Educational Fund.
Associated Press writer Cristian Salazar in New York City contributed to this report.