COUNTING THE INCARCERATED.

The flipside of post-release disenfranchisement is the manner in which the incarcerated are counted as residents of the districts where they are imprisoned. This bizarre arrangement, reminiscent of the three-fifths compromise, means that electoral districts, often in rural areas, are drawn in part on numbers skewed by non-voting incarcerated people, often from cities.This is a post-racial way of saying that, in a state like New York, Republican districts draw some of their "constituents" from urban minorities who aren't actually there by choice.

It's easy to see how this arrangement skews the incentives for elected officials who represent such districts--it's now in their interest to support draconian criminal penalties because the very existence of the districts they represent may depend on them--not to mention the economies of the towns they represent. So it's not just undemocratic--its self-perpetuating. For obvious reasons, it's also hard to change.

New York University Professor Anthony Thompson proposes a compromise solution for the upcoming census in the New York Times today:

The politics are complicated. Municipal leaders — including the mayor of New York City — support counting inmates in their last known address before incarceration. Rural officials support keeping the residence rule as it is. Criminal justice experts think it’s best to count inmates as residents of the communities where they are likely to return after their incarceration. (This, after all, is where the re-entry programs need to be.)

The Obama administration would do well to find a middle path. Commerce Secretary Gary Locke and Robert Groves, director of the Census Bureau, should propose an administrative change to the residence rule: Inmates returning to their home communities before the next census period — those serving a sentence of 10 years or less — should be counted in their home communities. Those serving more than 10 years should be counted where they are in custody. (The residence rules for other large transient groups, like college students, wouldn’t be affected by this change.)