The US prison boom of the past 30 years - which has nearly doubled the number of state prisons to more than 1,000 and increased the nation's prison population from 218,000 to 1.3 million - has had widely recognized economic, political, and social effects.
But one important political effect of the forced relocation of millions of inmates has been largely overlooked: The dilution of the urban black vote to the benefit of rural white communities.
A new Urban Institute report shows that inmates tend to come from regions that are demographically distinct from those in which their prisons are located. And because the Census Bureau counts prison inmates as residents of the legislative districts in which they're incarcerated, the relocation of inmates - who are not allowed to vote in 48 states - skews both the distribution of government funds and the apportionment of legislative representation.
This is a particularly grievous injustice in an era in which presidential and legislative races are won by razor-thin margins.
The distortion in representation caused by enumeration of prisoners tends to favor rural residents, whites, and Republicans, at the expense of urban residents, blacks, and Democrats.
In most states, prisons are disproportionately located in rural areas, yet most inmates come from urban areas. For example, in four state house districts in Connecticut, inmates make up more than 10 percent of the population. Each of these districts has low population and is disproportionately white - except, of course, for the inmate populations. And because of the presence of inmates, these districts effectively get about 10 percent more representation than they ought to.
Similarly, in Florida, statistics show a correlation between gains in representation from the presence of prisons and partisan affiliation of voters. Counties that gain population by importing prison inmates - and thus increase their congressional and state legislative representation - tend to vote Republican. Meanwhile, counties that lose population and legislative representation because a share of their population is imprisoned in other counties tend to vote Democratic.
Besides affecting representation inside a state, prison location has the potential to affect congressional apportionment between states. Wisconsin, for example, contracts to send as many as 10,000 inmates to prisons in other states. Wisconsin politicians were so concerned that the state would lose a congressional seat after the 2000 Census because of this practice that Republican Rep. Mark Green actually introduced legislation to count those inmates as citizens of Wisconsin. (In the end, the legislation didn't pass, and the seat was lost - though because of a larger margin of population loss than the inmates accounted for.)
This is only the most recent in a long line of techniques for diluting the voting power of African-Americans. And it's uncomfortably reminiscent of the infamous "three-gifths compromise." Under that clause of the Constitution, which determined how representation in the lower house of Congress would be allocated, slaves were counted as three-fifths of a person - although, of course, they weren't allowed to vote. The result was that the presence of disenfranchised blacks in the South increased the representation of white Southerners in the House. Likewise, under the current method of counting prison inmates, the presence of disenfranchised blacks in rural prisons increases the representation of white, rural, Republican voters both in the House and in state legislatures.
Not only does counting inmates in the districts in which they're incarcerated dilute the African-American vote and therefore violate the Voting Rights Act, it creates unconstitutional inequalities in the number of constituents per district, a violation of the Fourteenth Amendment.
Since the "one man, one vote" Supreme Court rulings in the '60s requiring that legislative boundaries in each state contain equal populations, the right of each citizen to equal representation in the House and state legislature has been a cornerstone of American democracy. The postcensus redistricting is designed to fulfill this critical goal of ensuring that each legislative district has as close to an equal number of constituents as possible.
The method of counting inmates where they're imprisoned, and not in their home districts undermines this doctrine. Even in the eyes of the legislators whose districts benefit from the presence of inmates, those inmates are not their constituents. In a survey I conducted of 40 Indiana state legislators, all - Democrat and Republican, with or without a prison in their district - said that inmates who are incarcerated in other districts, but who have family members in their district or lived there prior to incarceration, are more truly their constituents than inmates from other areas who happen to be incarcerated in their district.
Thus this is further evidence that counting inmates in the districts in which they're incarcerated creates an imbalance in the number of true constituents per district.
The Census Bureau should count prison inmates at their "homes of record." If the Census Bureau and Congress refuse to act, then the courts must enforce the Voting Rights Act and the Fourteenth Amendment and require a change in enumeration methods.
In an era in which America has the largest prison population in the world in both relative and absolute terms, the topography of that population should not be allowed to arbitrarily reshape the political landscape.
• Taren Stinebrickner-Kauffman is a president's research fellow at Duke University.