The distortion of Census data caused by counting people in prisons is “Too Big To Ignore”

by Peter Wagner, April 12, 2004 Link

Rose Heyer and I will release a new report tomorrow: Too big to ignore: How counting people in prisons distorted Census 2000. The report is the first national look at the impact of prisoner enumeration policies on the county population size, growth, race, ethnicity, gender and income. The Census counts prisoners as if they were residents not of their homes but of the prison town. The report concludes that due to modern uses of Census data and high incarceration rates, the impact is now too big to ignore. It’s time to update the Census Bureau’s method of counting the incarcerated and start counting prisoners at their home addresses.

Findings include:

• Twenty one counties in the United States have at least 21% of their population in correctional facilities.
• One out of every 50 counties reported as growing during the 1990s actual saw a decline in the actual population. New prison cells made these counties appear to be growing.
• Census data report rapidly growing Black populations in rural white counties. This population consists not of willing migrants but primarily of prisoners moved to the county for temporary incarceration.
• Many counties report concentrations of Latino adults without many children. These too are frequently the results of a prison in the community.
• Many of the counties where unmarried men outnumber unmarried women are counties with large male prisons. This ratio is useful for government planning, but due to the impact of prisons is difficult to use below the state level.
• Inclusion of prisoners in per-capita income figures complicates efforts to study and address rural and urban poverty.

How Census Bureau counts prisoners undercounts Michigan’s cities

by Peter Wagner, April 5, 2004 Link

The way the Census Bureau counts state prisoners — not at home but as if they were residents of the prison town — reduces the population of Michigan’s largest cities and makes the prison hosting counties look larger than they really are. Census Bureau figures influence a county’s political clout because these numbers are used to draw legislative district boundaries and distribute some state and federal funds.

Although only 20% of the state’s population, Wayne County (which contains Detroit) is the home of origin for almost 30% of the state’s prisoners. There is a large state correctional facility in the county, but Wayne County had almost 14,000 residents in state prison during Census 2000. As a result, the Census Bureau counted 9,974 Wayne County residents as residents of other counties because they were incarcerated outside the county. Kent, Genesee, Oakland, Ingham and Berrien Counties all lose population to this counting method. (See first map at right.)

The counties with the prisons get credited with this extra population, even though the prisoners have no ties or contact with that community. Chippewa, Iona and Jackson Counties gain more than 4,000 residents this way. Some of the counties with large prison populations are otherwise very small. In Chippewa and Luce Counties, 12 to 13% of the county’s population is not a resident or there by choice: They are incarcerated in a state prison. (See second map at right)

The Michigan Constitutions of 1850 and 1908 recognized that incarceration did not change a residence, and this principle is now preserved in an election law statute. Like most states, the Michigan definition of residence turns on where the person is voluntarily. Incarceration is most definitely an involuntary movement, and as such the Census Bureau should update its “usual residence rule” to count prisoners where they truly reside: at home.