on Disability Demographics and Statistics
A Review of Disability Data
for the Institutional Population
David C. Stapleton
Cornell University Institute for Policy Research
For additional information about this Research Brief contact:
Cornell University Institute for Policy Research
(202) 223-7670 ext. 105
Research Brief is being distributed by the Rehabilitation Research and
Disability Demographics and
center is funded to
Institute on Disability and Rehabilitation Research (No. H133B031111). The contents of this
paper do not necessarily represent the policy of the Department of Education, and you should not
assume endorsement by the Federal Government (Edgar, 75.620 (b)).
The Co-Principal Investigators are:
M. Bruyere --
Director, Employment and Disability Institute,
Richard V. Burkhauser -- Sarah Gibson Blanding Professor, Department of Policy Analysis
Andrew J. Houtenville -- Senior Research Associate, Employment and Disability Institute,
C. Stapleton --
Director, Cornell University Institute for Policy Research,
Disability status for the non-institutional population in
According to the Bureau of the Census, institutional group quarters (GQs) include correctional institutions, nursing homes, and a lengthy list of other institutions, many of which exclusively house people with disabilities. The institutional population is a very small share of the entire population, but increased from 1.3 percent of the population in 1990 (3.3 million people) to 1.4 percent in 2000 (4.0 million people). The increase was not uniform across institutional types, however. Nursing home residents, the largest institutional population in 1990, decreased from 0.71 percent of the total population to 0.61 percent, while the incarcerated population increased from 0.45 percent to 0.70 percent, surpassing the nursing home population in size. In 2000, about 98 percent of incarcerated persons were age 18-64, while about 90 percent of nursing home residents were age 65 or over.
The percentage of the institutional population that is of working age (ages 18-64) increased from 0.98 percent in 1990 (1.5 million people) to 1.3 percent in 2000 (2.3 million people), and working-age people accounted for a much larger proportion of the institutional population in 2000 (56 percent) than in 1990 (45 percent). In 2000, a large majority of the institutional working-age population (86 percent) resided in correctional institutions, and the remaining 14 percent were approximately evenly split between nursing homes and other institutions.
When disability is defined as having self-care, mental, physical, or sensory disabilities – 34 million (or 12 percent) of American people had disabilities, including 31 million (or 11 percent) of the household population, 2 million (or 54 percent) of the institutional GQs population, and 0.8 million (or 22 percent) of the non-institutional GQs population. Thus, disability prevalence for the GQs population, especially for the institutional GQs, is much higher than for the household population. Even so, the vast majority of people with disabilities live in household units; 6.4 percent live in institutional GQs and 2.3 percent in non-institutional GQs.
For working-age people with disabilities, the share of males living in institutional GQs is much larger than the share of females (7.7 percent versus 1.7 percent), mostly reflecting the fact that over nine out of ten inmates in correctional institutions are male. Working-age people with disabilities residing in institutional GQs are also disproportionately African American (38.6 percent of those ages 18-49 and 22.4 percent of those ages 50-64). Unfortunately the Census 2000 Public Use Microdata Samples (PUMS) files do not allow us to generate statistics by type of institution.
Based on data from three Department of Justice (DOJ) surveys, about 37 percent of jail inmates, 31 percent of state prison inmates, and 23 percent of Federal prison inmates report a disability of some sort (Exhibit I). Mental and learning disabilities are particularly prevalent in the jail and state prison populations. The prevalence of disability in the incarcerated population, almost all of which is of working age, appears to be two to three times as high as in the household working-age population, although exact comparisons based on published data are problematic due to different definitions of disability and differences in demographics.
[Chart 1] This chart shows the estimated disability prevalence for the incarcerated population in 1996/7, by types of disability and correction institution. The chart contains four pairs of columns. The estimated prevalence rates in 1996/7 are: any disability: 37 percent for jail inmates, 31 percent for state prison inmates, and 23 percent for federal prison inmate; physical disability: 10 percent for jail inmates, 12 percent for state prison inmates, and 11 percent for federal prison inmate; mental disability: 10 percent for jail inmates, 10 percent for state prison inmates, and 5 percent for federal prison inmate; learning disability: 9 percent for jail inmates, 10 percent for state prison inmates, and 5 percent for federal prison inmate. [End Chart 1]
Source: She and Stapleton (2006).
Essentially all nursing home residents have some type of disability. Based on data from the 1999 National Nursing Home Survey (NNHS), the two most common activity limitations are for bathing (94 percent) and dressing (85 percent), and about three quarters of the nursing home residents received help with three or more such activities of daily living (ADLs). The prevalence of various disabilities is generally lower among those under age 65, with one exception: mental disorders other than Alzheimer’s disease – 21 percent for those under 65 versus 17 percent for those 65 and over.
The incarcerated population more than quadrupled from 1980 to 2003, from a half million to more than two million. Although this growth partly reflects underlying population growth, the main reason for growth is increased incarceration rates. From 1989-91 to 1996-97, the incarceration rate for the working-age population grew by 35 percent (Exhibit II). The change in the incarceration rate is greatest for those age 25 to 44, although relative change is greater for those age 35 to 54. Further, the change in the incarceration rate is much greater for males than for females, although the relative change is somewhat larger for females. These statistics, along with the relatively high prevalence of disability in the incarcerated population, suggest that the growth in incarceration could have had a substantial impact on prevalence of disability in the household population, and possibly on the characteristics of the household population with disabilities, especially for working-age males.
[Chart 2] This chart shows the estimated incarceration rate by age and sex for 1989/91 and 1996/7. The chart contains eight pairs of columns. The estimated incarceration rates are: for all ages: 472 per 100,000 in 1989/91 and 638 per 100,000 in 1996/7; for ages 18-24: 1,113 per 100,000 in 1989/91 and 1,474 per 100,000 in 1996/7; for ages 25-34: 1,262 per 100,000 in 1989/91 and 1,690 per 100,000 in 1996/7; for ages 35-44: 669 per 100,000 in 1989/91 and 1,110 per 100,000 in 1996/7; for ages 45-54: 297 per 100,000 in 1989/91 and 476 per 100,000 in 1996/7; for ages 55 or over: 66 per 100,000 in 1989/91 and 87 per 100,000 in 1996/7; for males: 926 per 100,000 in 1989/91 and 1,242 per 100,000 in 1996/7; for females: 66 per 100,000 in 1989/91 and 97 per 100,000 in 1996/7. [End Chart 2]
Source: She and Stapleton (2006).
The nursing home population has been growing slowly in absolute terms, and, as previously indicated, has been declining relative to the size of the entire population. From 1977 to 1999, the nursing home residence rate (the number of residents per 100,000 population) declined for all age groups (Exhibit III). For those under 65, the decline is small, but would be much more substantial if measured relative to the size of the population with disabilities. The decline was especially large for those age 85+, almost 6 percentage points. We do not know the extent to which this decline represents a decline in disability among those 85 and older versus a change in the living arrangements of people with disabilities (e.g., the rise of assisted living).
[Chart 3] This chart shows the estimated nursing home residence rate by age for 1977 and 1999. The chart contains four pairs of columns. The estimated nursing home residence rates are: for ages under 65: 87 per 100,000 in 1977 and 66 per 100,000 in 1999; for ages 65 to 74: 1,473 per 100,000 in 1977 and 1,080 per 100,000 in 1999; for ages 75-84: 6,595 per 100,000 in 1977 and 4,297 per 100,000 in 1999; for ages 85 or over: 24,219 per 100,000 in 1977 and 18,250 per 100,000 in 1999. [End Chart 3]
Source: She and Stapleton (2006).
In the past, the four major household surveys – ACS, CPS, NHIS, and SIPP – have all excluded the institutional population in their sampling frames, and varied in their coverage of persons living in non-institutional GQs. Further, public use files often fail to identify type of living quarters. Survey information is also very limited for the homeless population and the military population. The homeless population is either not covered at all or covered to an unknown extent in major national surveys. Only the March Supplement of the CPS claims to routinely include the military population, but we found no studies providing CPS-based estimates for the military population. Incomplete coverage for the non-institutional GQs population is more problematic for the working-age population than for the elderly population.
The NNHS provides extensive information on nursing home residents, and the three DOJ surveys provide information about the incarcerated population. We found no surveys covering the institutional population not living in correctional institutions or nursing homes, except the Census long-form survey. This component now represents 9 percent of the institutional population, and 7 percent of the working-age institutional population. Further, it is apparent that those with disabilities in this population represent a much larger, although unknown, share of the population of people with disabilities.
In the past, the one survey to collect data on the entire population, the Census, has been conducted only once per decade – in contrast to the annual collection of data on the household population via four major government surveys. The surveys of specific institutional populations are conducted less frequently and less regularly. The DOJ surveys are fielded roughly every five years. The NNHS was fielded biennially from 1995 to 1999, but no information has been available since 1999.
The disability information included in the surveys varies. Currently, the Census and three major non-institutional surveys (ACS, NHIS, and SIPP) all contain information on sensory disabilities, functional limitations, mental disabilities, limitations in ADLs, limitations in instrumental activities of daily living (IADLs), and work disabilities, although definitions vary. The CPS only contains information about work disabilities. The NNHS does not ask respondents about work disability, but this might reflect an implicit assumption that all respondents either have work disabilities, or that almost all are too old for work to be considered a relevant topic. The surveys on inmates do not ask questions on ADLs and IADLs, but do include questions on learning disabilities that are absent in most household surveys. Conceptual definitions of disability in these surveys are substantially in agreement, but there are substantial operational differences in the collection of information for each conceptual definition.
It is apparent from the data available that the size and composition of the institutional population has changed very substantially in the last 25 years. Those changes have been especially important for the population with disabilities, because of the relatively large number of persons with disabilities who live in institutions. Growth in the incarceration rate is especially important for the working-age population, and raises many disability policy issues, such as: how public policies affect the participation of people with disabilities in criminal activity; the provision of equal access to justice; appropriate treatment of, and supports for, prisoners with disabilities; and support for the transition from prison into the household population. The decline in the nursing home residence rate, especially for the elderly, is also very substantial, and raises a host of questions about the reason for the decline and the implications for public policy.
Disability information on the entire population is scarce. That situation is changing; if the Bureau of the Census follows its current plan, the ACS will continuously and consistently provide annual data for the population living in most GQs, including the major institutional GQs, from 2006 forward. This information will be particularly valuable for disability research and statistics, especially for the working-age population, given the large gaps in currently available information.
She, P and D.C. Stapleton (2006). An Inventory of Disability
Information for the Population Living in Institutions. Rehabilitation Research
 The statistics presented in this section are based on the 1990 and 2000 Census data.
 2000 Census data on two additional disability types, employment disability and go-outside-home disability, are not valid. Based on the 2003 ACS, the prevalence of disability in the non-institutional population in 2000 would have been 1.5 to 2.0 percentage points higher had these two disabilities been included.
 Non-institutional GQs include group homes, college dormitories, military quarters, and a lengthy list of other non-household living situations.
 Due to a data collection problem, the 2000 Census statistics may overstate the share of the population with disabilities that lives in institutional or other GQs.
 Based on the 2003 ACS, disability prevalence among all persons aged 25 to 61 not living in GQs is 11.9 percent, including 7.5 percent with a physical disability and 4.0 percent with a mental disability. Young adult males account for a disproportionately large share of the incarcerated population; only about one percent of the incarcerated population is over age 65.
 Based on the 1996 National Survey of Homeless Assistance Providers and Clients, 80 percent of homeless clients are 25 to 54 years old.
 As of 2006, ACS data collection excludes the following GQs: domestic violence shelters, soup kitchens, regularly scheduled mobile food vans, targeted non-sheltered locations, natural disaster shelters, transient locations (such as RV campgrounds, marinas, and military hotels), dangerous encampments, and maritime vessels.