What states can do to end prison-based gerrymandering

The problem, the solutions and what can be done today.

by Peter Wagner, October 5, 2010

We’re often asked by state officials what they can do about prison-based gerrymandering in their states and if we can send them some introductory information. We usually send out something like the information below. Additionally, for state-specific information on the problems caused by prison-based gerrymandering, see our report Fixing prison-based gerrymandering after the 2010 Census: A 50 state guide.

The Census count and prison-based gerrymandering:
The problem, the solutions and what can be done today

The Supreme Court’s “one person, one vote” cases require state and municipal districts to be redrawn each decade so that each district contains the same number of people and each resident, therefore, has the same access to government. A long-standing flaw in the decennial census, however, systematically counts more than 2 million people in the wrong place and undermines the “one person, one vote” principle. Although people in prison can’t vote and remain residents of their home addresses, the Census Bureau counts people in prison as residents of their prison cells instead. Using this flawed data to draw districts grants the people who live near large prisons extra influence at the expense of voters everywhere else.

The problem at the state level:

• Crediting prisoners from all over the state to the predominately rural districts that contain large prisons enhances the weight of a vote in those districts, diluting all other votes in the state.
• Prisoners are disproportionately Black or Latino, and outside of the Deep South, most prisons are built in disproportionately white areas. Using Black and Latino prisoners to pad the populations of white legislative districts dilutes minority voting strength state-wide.

The problem at the county and municipal level:

Rural county and municipal districts are smaller than state legislative districts, so prison-based gerrymandering can create an even larger problem for people who live near but not immediately adjacent to a large prison. Some examples:

• A true “rotten borough”: In Anamosa Iowa, a man won a city council seat with two write-in votes, neither of which he cast. There were no candidates because 96% of the district was incarcerated in a large prison. The handful of voters in the district had 25 times as much influence on the city council as residents elsewhere in the city.
• “Majority-minority” in name only: In Somerset County, MD, a county commission district that was deliberately drawn as a majority-minority district in order to settle a Voting Rights Act lawsuit has been unable to elect an African-American because a prison population, which cannot vote, was included in the population. The actual African-American resident population in the district is too small to elect an African-American candidate. It would have been possible to draw an effective majority-minority district had the prison population not skewed the data.

Solutions

• Ask the U.S. Census Bureau to count incarcerated people at home in the next Census, and work to improve state correctional data systems to facilitate that change.
• Use state correctional data to adjust the federal Census for redistricting purposes so that districts are based on data that counts incarcerated people at their home addresses.
• As an interim solution, refuse to pad the populations of legislative districts that contain prisons. Adjust the census data used for redistricting so that all incarcerated people are considered to be residents of unknown locations, and are not included in any individual district.

What states are doing:

• New York, Maryland and Delaware have recently passed legislation to ensure that districts are drawn based on data that counts incarcerated people at home. There is a long history of states making adjustments to the census in other contexts as well.
• The legislative or executive branches in several states (Virginia, Colorado, New Jersey, Mississippi) require or encourage local governments to modify the census and refuse to use prison populations as padding. Additionally, more than 100 rural counties and municipalities around the country make these adjustments on their own.

Cost

Costs are minimal at most because existing infrastructure generally exists at Departments of Corrections and in the redistricting and planning agencies. Furthermore, the Census Bureau has already changed its data publication schedule to make it easier to identify prison populations in its redistricting data. Finally, the most efficient and cost-effective solution has no cost at all to the states: ask the Census Bureau to count prisoners at home.

Conclusion:

The principle of “one person, one vote” requires that legislative districts be drawn so that each contains the same number of people counted in the correct location. The Prison Policy Initiative would be happy to help states determine the most appropriate next steps considering the state constitution, redistricting timelines and available state corrections data.