Estimating disability prevalence in population surveys: an exploratory study using national surveys carried out in Ethiopia, Uganda, Tanzania and USA.

Abstract

The call for high quality disability data has gained momentum around the world. The objective of this paper is to analyse the consistency of estimates obtained with the Washington Group short set of questions, which are being recommended for use in national population surveys. Prevalence rates and their 95% confidence intervals were estimated using data from five surveys carried out in three low-income countries and one high-income country. Two surveys undertaken in the same year generated different disability prevalences: 16.1% and 25.6% using a broad cut-off; 4.0% and 6.7% using a narrower cut-off. In two longitudinal surveys, over 50% of persons identified as having disability in a wave were not identified as such in the next wave, independent of cut-off used. Data disaggregation by education and remunerative employment shows little differences between persons with and without disability, which is inconsistent with current evidence. The paper shows some significant inconsistencies that can have consequences for policy, international comparisons, for the monitoring the CRPD and SDGs, and most importantly, for persons with disability. It is important that advocates, policy makers, statistical offices and country authorities follow reliability and validity standards when recommending tools for disability data collection.