"Once, only once, and in the right place" goes the hopeful mantra of the hard-working and underappreciated Census Bureau. But the current rules for counting incarcerated persons are at odds with the last part of this goal. Next week, the census will count 1.6 million people in the wrong spot, distorting democratic representation in many states and localities.
The bureau tallies incarcerated persons as residents of the prison town rather than as residents of their home communities, where they will typically return within 34 months. This policy is as old as the census, but extraordinary growth in the prison population over the last few decades -- coupled with modern uses of the data to apportion political power at all levels of government -- now spells big problems for representative democracy.
In 1970, the incarcerated population in state and federal prisons was just under 200,000. Today, it has ballooned to 1.6 million -- almost the populations of North Dakota, Wyoming, and Vermont combined. Over the same period, prison construction has become a growth industry in many rural locations far removed from the homes of incarcerated persons, who are disproportionately African American and Hispanic. Jurisdictions with prisons become eligible for greater political representation based on populations that have no real connection to the larger district, while the largely urban home communities of incarcerated persons -- which remain their legal residence under the laws of most states -- are shortchanged in terms of political power and representation.
In Illinois, 60 percent of incarcerated persons are from Cook County, yet 99 percent of them are counted as residents elsewhere in the state. In New York, several upstate Senate districts would not meet minimum population requirements without the "constituents" imported into prison cells there. The strongest advocates for harsh sentencing laws in New York have included some of the state senators elected from these very districts.
But treating prisons as residences harms rural communities as well urban interests. In 2008, a resident of Anamosa, Iowa, was elected to the City Council with exactly two votes. Not a two-vote margin -- two votes total. He wasn't a candidate, but his wife and a neighbor wrote him in.
The problem in Anamosa wasn't voter apathy -- it was a lack of voters. Because the census counted more than 1,300 people in the Anamosa State Penitentiary as residents of a 1,400-person election ward, the ward really contained only 58 residents. The city's three other wards each had 25 times as many real residents as the ward with the prison, giving the few residents who lived near the prison 25 times as much influence in government as anyone else in the city.
Although the Constitution requires districts to be based on population, it does not require blind acceptance of the census counts when drawing electoral boundaries. Local governments around the country are protecting their democratic processes by rejecting the bureau's prison count. More than a hundred rural counties ignore the prison population when determining the population counts for drawing their local legislative districts or designing their weighted voting systems. One New York county board of supervisors explained its decision to do so by observing that state and federal prisoners "live in a separate environment, do not participate in the life of Essex County, and do not affect the social and economic character of the towns."
After the 2000 census, only rural counties took action to avoid prison-based gerrymandering, but now state governments are trying to catch up. In Maryland, legislation to adjust census data to count prisoners at home for redistricting purposes passed initial votes in both the House and Senate just last week. Seven other states are considering similar legislation.
Of course, the Census Bureau should be the one to fix the problem. The bureau has a well-honed data-collection operation, and it is uniquely suited to collect the data in a way that will meet the needs of state legislators and our democratic system. Ideally, the bureau should count incarcerated people where they legally reside: at home.
While it is too late for the bureau to change where people in prison are counted in the 2010 census, the bureau recently has taken steps that should assist states and localities in making their own population adjustments. In February, the bureau announced its intent to accelerate publication of data to identify the incarcerated population within census blocks that contain prisons, which would make it easier for legislatures to identify and remove the prison populations at districting time.
The bureau's method of counting other groups, such as military personnel, has evolved when the country's changing demographics demanded it. Our country's demographics have changed in such a way that the large size of the prison population combines with the bureau's outdated methodology to create a problem for representative democracy that is too big to ignore. The 2010 census should be the last in which 1.6 million incarcerated persons are counted at the wrong location.