From the New York Times (click here for full article):
For decades, predominantly rural and Republican districts have had extra clout in state and local legislative bodies because their large inmate populations were counted as local residents in apportioning representation. Now, the Census Bureau has agreed to give states a tool that could dilute the political power of those districts.
In May 2011, in time for Congressional and legislative reapportionment, the bureau will identify exactly where group quarters like prisons are and how many people occupy them. States would then have the option of counting them in the local population or not.
“This removes a technical problem,” said Peter Wagner, executive director of the Prison Policy Initiative, an advocacy group that favors alternatives to prison sentences and urges that inmates be counted in their hometowns. “The census is going to say where the prisons are and how many people are in them, which will enable states the practical choice of counting them in the wrong place or not counting them at all.”
Mr. Wagner and other groups had originally asked the bureau to determine the home addresses of inmates and to count them at those addresses, but the 2010 census process was too far along for that change to be considered.
A number of states — including Florida, Illinois, Maryland, New York and Wisconsin — are weighing legislation requiring that prisoners be counted at their last known address — for purposes of reapportionment, a change that would likely favor larger and mostly Democratic cities.
In New York, the change could prove pivotal because of the see-saw fight for control of the State Senate and the fact that the state faces the loss of at least one Congressional seat after the 2010 census.
Update: A reader provided us with a link to a document that shows the Census Bureau’s 2006 position on this issue in the from a comprehensive report. The report concludes the following:
“Counting prisoners at a ‘permanent home of record’ address, rather than at their place of
incarceration, would result in increased cost both to the decennial census program and to
the Federal, State, and local correctional facilities that would be required to participate in
data collection efforts. Our study raises concerns that this change would result in
decreased accuracy for a possibly large proportion of millions of individuals confined on
Census day. The completeness of the census count would be compromised for prisoners
that cannot provide a valid address, and we have no method of determining how many
individuals would fall into that category. Further, a fundamental shift for the enumeration
of correctional facilities would likely have a negative impact on other Group Quarters