prevalent in young adults, but no data are available to directly tie particular kinds of injuries—motor vehicle accidents or firearm injuries, for instance—to functional limitations or disabilities. This kind of information is clearly needed to develop effective programs for the prevention of disability.
In summary, taking into account all of the data discussed above, it is clear that the number of persons with disabilities depends on the definition of disability. Combining the number of noninstitutionalized people with any activity limitation estimated in the NHIS and the population of all ages in nursing homes, it appears that 35 million people live with disabling conditions. SIPP data point to about 46 million people with some type of work or functional limitation. With definitions focused only on inability to work or to carry out other major activities, or on receipt of disability benefits, the number of people with disabilities is substantially smaller. By any definition, however, people with disabling conditions, on average, are older and have lower incomes than others.
Because disability has many dimensions, different measurement concepts are necessary. Data are needed on (1) the clinical conditions that lead to functional limitations, (2) the impact of these limitations on the activities that individuals are able to carry out, and (3) the social and economic impacts these individuals experience because of functional limitations. Each of these aspects of disability can legitimately be measured in different ways, and because social programs are tied to some measures of disability, different definitions to match eligibility requirements are necessary. Although it would be extremely costly and technically difficult for any single data system to deal with all of these concepts and measurement systems simultaneously, it is important that attention be paid to improving the quantity, quality, comprehensiveness, and relevance of data on disability in the United States for consumers and for setting policy.