The 2020 Census is complete and the results will be released soon. Because of how we count people in prison, there may be inaccurate demographics and a virtual overestimate of people living and accessing services in rural communities. In this two-part series, we look at how rural prisons can skew county-level analyses of rural demographics and health statistics.
Census policy and rural prisons
Thousands of researchers and policy makers rely on census data. U.S. census policy dictates counting individuals at the address where they currently reside, including prisons.
Most prisons built in the past 40 years are in rural communities. In fact, hundreds of rural counties now have a state or federal prison. In 2019, 1.4 million people were held in state or federal prisons, a substantial population of at-risk individuals. And prisons in rural counties may account for anywhere from 5-25% of a county’s population.
A prison population’s demographics and socioeconomic make-up can be markedly different than the general population. As a result, we may wrongly interpret county health statistics, impacting the distribution of state and federal funding for county services.
The demographic data disconnect
Let’s look at one rural county as an example.
Logan County, Colorado has a population of 22,000 in an area of 1,800 square miles. It is home to the state’s largest prison with 2,585 inmates. Incarcerated individuals account for a striking 12% of the population. Yet, the demographics of the incarcerated population are vastly different than the county overall. Compared to the county as a whole, Black, Latino and Native-American individuals are over-represented among incarcerated individuals.
As a result, census data show that Logan county has a 5-10 times higher rate of African Americans than the surrounding counties. The data also show a 6 percent lower proportion of females, and a 2.1% lower proportion of children under age five. Two other rural Colorado counties with prisons (Lincoln and Crowley counties) have a 10-20 times higher rate of African Americans and lower rates of females. Crowley County, Colorado for example, reported that just 25% of the population is female.
Numerators, denominators and diseases
HIV and Hepatitis C (HepC) in jails and prisons are referred to as a bellwether for community infection rates over time. It is crucial to have accurate, reliable, up-to-date HIV and HepC infection rate data in communities. In 2003, Logan County reported 36 cases of HepC. With 20,504 in the county, this was a rate of 176 cases per 100,000 people, which is high for a rural county. However, the HepC numerator and the county denominator both included the prison population. Without prison inmates in the calculation, the rate was much lower. There were just 6 cases and 19,114 residents, a rate of 31 per 100,000 that is consistent with surrounding rural county rates [doc].
Since initial publication of these findings, the Colorado Department of Health and Environment has changed how inmate HepC is reported. Now Colorado groups all inmate cases into one “other” location category. If a person with HepC is incarcerated in a county jail, the case is assigned to the county. However, if that inmate is at a state or federal prison, Colorado assigns the person to an institution type rather than a county. This essentially groups all state and federal prison inmates into one separate group. In 2017, more inmates in Colorado prisons had chronic HepC than people in any individual county.
A different story for HIV
Prevalence of individuals living with HIV in Colorado counties (dark red= highest rate). Source: AIDSVu.org
While HepC in Colorado prisons is no longer reported at a county level, HIV is still assigned to the county where the prison is located. As a result, 2 of the 4 counties with the highest rates of HIV infection are rural counties with large prisons.
Freemont County, for example, has 46,000 residents, 13 prisons, and nearly 9000 inmates. Incarcerated people make up nearly 20% of the county population. Crowley County, Colorado has the highest ratio of inmates (34%) of any county in the U.S., with 1,955 inmates out of a total population of 5,800. Both counties have the highest rates of HIV infection.
Other states are doing this too. In the lower panhandle of Texas, Childress County stands out on the map due to the much higher rate of HIV infection. It is home to a prison with over 1,600 inmates who account for 23% of the population. At least three other rural counties in the state (Coryell, Falls and Anderson counties) stand out as having a much higher rate of HIV than surrounding counties. This is likely due to the presence of a prison in the county.
Health services publications may misinterpret data due to rural prisons
Inclusion of prison populations in the numerator and/or denominator can lead to erroneous interpretation and publication. For example, Van Handel, et al. of the Centers for Disease Control and Prevention conducted a study to identify counties at risk of rapid spread of HIV or HepC. They based their findings on several unique county data points.
First, they used the county rate of confirmed acute HepC infections as a proxy for county-level rates of persons who inject drugs (a common reason for rapid spread). Second, they identified counties with high prevalence that were close to populations with a higher number of people with HIV infection. They ranked counties based on their vulnerability to rapid HIV spread due to presumed intravenous drug use.
Crowley County in Colorado–a county we mentioned above–was in the 95th percentile of most vulnerable counties. While inmates have higher rates of HepC and HIV, they are less likely to be routinely injecting drugs. And they are unlikely to be traveling frequently outside the county (an assumption of the CDC study). The result is a skewed vision of the true risk of the rural county.
Appearing in a CDC report may seem innocuous. But when the report enters the public media, such as it did with the Wall Street Journal, it can hurt rural communities. Rural counties already struggle in many ways, including with COVID-19. Imagine undertaking rural development, community marketing, and teacher recruitment when your county stands out for HIV vulnerability.
Stay tuned for Part 2 soon, with another example and thoughts on implications for redistricting and other issues.
Jack Westfall
Latest posts by Jack Westfall (see all)
Hoon Byun