This paper — jointly published by HelpAge International and Development Pathways — discusses key challenges in measuring levels of poverty among older people, at the individual level, based on survey data that are collected on households as collective units. Throughout this process, poverty analysts inevitably make a range of methodological choices and assumptions. However, what is not well known among those outside of the field, is that seemingly minor methodological choices can heavily influence the value of poverty rates and welfare rankings of different population groups in society.
Using data from income and expenditure surveys from 32 countries in Africa and Asia, the paper illustrates that measures of old age poverty are highly sensitive to the methodological assumptions that underpin poverty statistics. It shows that the common per capita approach that is widely used in applied research and publications from influential actors like the World Bank and UN agencies tends to underestimate levels of poverty among older people. Moreover, the standard approach is likely to overstate the welfare of older people vis-à-vis children, does not factor in the additional cost of living with a disability faced by many older people, and ignores intra-household inequality in the distribution of resources (e.g., between male and female household members).
The findings have important implications for social protection policy, as all too often poverty rates are used to prioritise or deprioritise interventions for specific subgroups of the population, without adequately testing their sensitivity to methodological choices. The paper encourages policy makers and development practitioners to better recognise data limitations and that poverty measurement is not an objective, value-free science. It also calls on the wider statistical community to develop practical methods that can factor in the costs of disability and the intra-household allocation of resources in poverty measurement.
Read the paper here.