Last week, the U.S. Census Bureau issued a proposed rule for the 2020 Census that would continue counting incarcerated persons at the site of their prison rather than at their pre-incarceration homes.
The proposed rule is unfortunate. The current practice ignores the fact that the vast majority of prisoners return to their pre-incarceration communities after their release, and have little or no connection to the communities they are incarcerated in. In New York, for example, 91 percent of prisoners are housed at facilities located in Upstate New York, even though 66 percent come from and ultimately return to New York City. Counting inmates at the site of their prisons distorts the allocation of legislative seats and also leads to a large-scale voting power imbalance in communities with prisons — meaning the citizens in those districts have greater representation and more political power at the expense of the large non-voting population of prisoners in their district.
Instead of continuing the current practice, the Census Bureau could have proposed a rule that would use state and federal Department of Corrections information to count incarcerated individuals in the communities they lived in previously. This step would have given proper representation to communities with large prison populations and ended the distortions caused by prison gerrymandering. From a policy perspective, this rule would have been in line with other proposed changes the Census announced, including counting juveniles in residential treatment facilities as well as deployed military personnel in their pre-stay or pre-deployment homes. In both cases, the Bureau acknowledges the fact that counting either population where they are on census day would lead to inaccuracies, as these situations are temporary, with people returning to their home communities afterwards. This same rationale holds true for incarcerated individuals — who by and large are not permanently incarcerated — and it is difficult to see why a different principle should apply to prisoners. The change would also have been in line with the current practice in most states.
In declining to change its existing practice, the Bureau rejected the urging of 156 groups and individuals, including the Brennan Center, who told the Bureau that the current rule had significant distortive effects. By contrast, only six interested groups advocated the continued counting of incarcerated persons at their prison addresses. Interested parties now have 30 days to provide comments on the proposed rule.
One bright spot in the proposed rule was the Bureau’s announcement that it would continue to provide data that will allow states to voluntarily reallocate their prison population to their homes. Several states have passed laws ending prison gerrymandering by reallocating incarcerated individuals from where they are on Census day to their pre-incarceration homes. Both New York and Maryland, for example, have implemented laws ending the misallocation of incarcerated individuals for drawing districts, and Delaware and California will end prison gerrymandering starting after the 2020 Census. Others have followed suit. However, such voluntary steps are not possible everywhere because some state constitutions require the use of Census counts for redistricting, meaning that in some states even if the legislature wanted to reapportion the prison population to their homes it would be prevented by the Bureau’s refusal to change the residency rule.
The proposed rule is not yet final. The Census Bureau should reconsider its decision to continue its current practice and review and adopt a rule counting incarcerated persons as being residents of their pre-incarceration communities. Doing so would make a more accurate Census as well as help ensure that the nation’s communities are fairly represented in the upcoming round of redistricting that will take place after the 2020 Census.