Skip Navigation

Count This As a Big Step Forward

Now that the Census will deliver prison-count data in time for redistricting plans, learn how to use that data effectively to finally reduce the effects of prison-based gerrymandering.

  • Justin Levitt
February 12, 2010

Yesterday, word spread that the Census Bureau would make a powerful new tool available to the states, counties, and cities needing to redraw political districts after the 2010 census. Along with the standard datafile telling us how many people are assigned to any geographical slice of the country, the Bureau will tell us, at the same time, how many are incarcerated.   

This is a big step toward correcting a problem that the Brennan Center’s been working on for most of the decade. Right now, people in prison show up in the census data at the blocks where they are incarcerated, rather than at the addresses they came from in their home communities. The two are usually far from each other. And with the nation’s rising incarceration rate, they lead to a systematic distortion of the population picture.

When districts are based on this data, they build the distortion into the distribution of local democracy. Districts are constructed on the backs of “ghost voters,” packing in prisoners who count toward the district size but who, with few exceptions, are not permitted to vote, and who have no connection at all to the other residents of the district or its welfare. This artificially inflates the political power of voters in prison districts, skewing the incentives of politicians there — and it artificially deflates the power of voters everywhere else. Peter Wagner has done an enormous amount of homework to show exactly how severe the distortion is, complete with a calculator to drive the math home.

For example, after the last census, 1300 of the 1400 people allotted to Ward 2 of the Anamosa, Iowa, city council were in prison.  This left political power completely lopsided: the few others in ward 2 had far more leverage than any of their neighbors in town.  Indeed, in districts so distorted, we’d hardly recognize what passes for democracy. In 2006, the city council seat for Anamosa’s Ward 2 was won with two write-in votes — one cast by the winner’s wife.  

The right way to handle prison populations is to count people at their last known address before incarceration, which is where virtually all prisoners return when they are released. There are already bills afloat in at least five states in this legislative session, on top of a federal bill, to accomplish just that (here’s a complete legislative list). And it’s already required by Mississippi law, at least for local districts. But the clock has run out on administrative capacity to do the same nationwide, for each and every facility, before redistricting starts.

So we turn to the next best thing: even where it’s not possible to get prisoners’ addresses right, it’s possible to take on half of the skew by ensuring that they’re not inflating the locations that are wrong. Instead, incarcerated populations would be addressed as part of the jurisdiction generally, without a more specific geographic tie, just like servicemembers and federal government personnel who are overseas on Census Day. These people are assigned to the state whence they came in order to apportion members of Congress, but not assigned to a specific address within that state that would affect redistricting.

Removing this half of the skew would not completely fix the prison distortion, by reconnecting the incarcerated population to the communities to which they most tangibly belong. But it would take a significant step toward making districts more fair. And this step would benefit all of the remaining voters in the jurisdiction, whose political power is now artificially diluted.

Some jurisdictions already do this. About 1/3 of the counties with prisons in New York State, for example, don’t tie prison populations to local districts. California counties with large prison populations have asked for permission to make the change; for Colorado counties, it’s required. Until yesterday, however, any new jurisdiction that wanted to follow these leaders had to go through the legwork of collecting the necessary data.

The Census Bureau’s new dataset eliminates that cost entirely. All you need to fix half of the skew is to use the Bureau’s incarceration numbers to reallocate prison populations from lopsided districts to the jurisdiction as a whole. And now that the Bureau will deliver incarceration counts at the same time as the other redistricting data, those that want to draw fairer districts, anywhere in the country, will find it much easier to do so.