Estimating disability prevalence and disability-related inequalities: Does the choice of measure matter?

• Published in: Social science & medicine (1982) Vol. 272; p. 113740
• Main Authors: Amilon, Anna, Hansen, Kasper M., Kjær, Agnete Aslaug, Steffensen, Tinne
• Format: Journal Article
• Language: English
• Published: England Elsevier Ltd 01.03.2021; Pergamon Press Inc
• Subjects: Denmark; Disability; Disability measures; Disability prevalence; Employment; Estimates; Health problems; Health status; Illnesses; Inequality; Living conditions; Mental disorders; Mental health; People with disabilities; Polls & surveys; Prevalence; Respondents; Social inequality; Social policy; The global activity limitation indicator; The Washington group short set of questions on disability; Voter behavior; Voter turnout; The Washington group short set of questions on disability; Voter turnout; Employment; Disability prevalence; The global activity limitation indicator; Disability measures; Denmark; Inequality
• ISSN: 0277-9536; 1873-5347; 1873-5347
• DOI: 10.1016/j.socscimed.2021.113740

Different measures for quantifying the percentage of people with a disability in surveys result in diverging estimates of prevalence and disability-related inequalities. Thus understanding the implications of using different disability measures is of vital policy importance. This study is the first to investigate the within-survey variation in disability prevalence based on two internationally recognized measures: the Washington Group Short Set (WGSS) and the Global Activity Limitation Indicator (GALI). It is also the first to examine the disability-related inequality in voter turnout, based on official validated voter records. We use data on 11,308 25-54-year-old respondents from the 2016 wave of the Survey of Health, Impairment and Living Conditions in Denmark (SHILD) to estimate the disability prevalence based on the WGSS and the GALI. Moreover, we investigate health characteristics of individuals with a disability according to the two measures and inequalities in two central social policy success parameters: voter turnout and employment. The WGSS estimates higher disability prevalence (10.6%) than the GALI (5.5%). Only 2.5% of the sample are in both groups, implying that largely, different individuals are defined as having a disability depending on which measure is used. The health profiles of the two groups also differ, as people with a GALI-defined disability are significantly more likely to report a severe mental illness or a major physical health problem. The GALI estimates indicate larger inequalities between people with and without a disability than the WGSS for the probability of being employed, whereas there are no significant differences for voter turnout. The choice of disability measure strongly influences within-survey estimates of disability prevalence, the health profile of the defined groups, and inequalities in outcomes. The WGSS underrepresents the number of people suffering from severe mental illness. Estimated inequalities in employment are larger for the GALI than for the WGSS. •The estimated prevalence of disability varies by choice of disability measure.•Different people are defined as having a disability subject to which measure is used.•The choice of measure strongly influences the health profile of the defined groups.•Inequalities between people with and without disability vary across measures.