National estimates of disability prevalence produced using 2021 SIPP data may be higher than estimates from other federal surveys that use the same disability measure. In general, SIPP data suggest that disability prevalence has increased in the U.S. over time since 2008, particularly among adults aged 40 and over. This pattern is not consistently observed in other federal surveys. Disability estimates from the American Community Survey, for example, suggest that disability prevalence rose only very slightly among adults aged 40 and older between 2008 and 2014 and then declined between 2014 and 2018.

It is difficult to determine the exact cause of rising disability rates in SIPP. The Census Bureau has investigated several possible explanations for rising disability rates in SIPP. More information can be found in the working paper titled Understanding Changes in the Disability Prevalence in the 2014 Survey of Income and Program Participation: Three Explanations Considered, available at https://www.census.gov/library/working-papers/2018/demo/SEHSD-WP2018-04.html

Overall, higher disability estimates in SIPP do not appear to result from errors in data collection or data processing. There is some evidence that differences in SIPP sample composition between panels have contributed to changes in disability prevalence estimates. Members of the 2014 SIPP Panel, for example, reported poorer self-rated health, on average, than members of the 2008 SIPP Panel. Survey instrument changes are another potential explanation for the jump in disability prevalence, although it should be noted that the question items used to estimate the proportion of the population with any disability (RDIS) have remained consistent since 2008.

Data users should be aware of these concerns regarding 2021 SIPP disability estimates. SIPP 2021 disability data can be used to examine associations between disability and other variables over time, as well as to examine group differences in disability. However, we encourage caution

if using these data for population estimates of disability, given that SIPP disability estimates do not necessarily align with estimates from other federal surveys.