This paper argues that survey-based median household consumption expenditure (or income) per capita should be incorporated into standard development indicators, as a simple, robust, and durable indicator of typical individual material well-being in a country.
Using household survey data available for low- and middle-income countries from the World Bank’s PovcalNet tool, the study shows that as a measure of income-related well-being, it is far superior to the commonly used GDP per capita as well as survey-based measures at the mean. Survey-based median measures are “distribution-aware”, i.e. when used as the denominator of various widely available indicators such as mean consumption expenditure per capita they provide a “good-enough” indicator of consumption (or income) inequality.
Finally, as a post-2015 indicator of progress at the country-level in promoting shared development and reducing inequality, the paper proposes that the rate of increase in median consumption per capita after taxes and transfers exceed the rate of increase in average consumption in the same period.
Key findings:
- Median consumption per capita is for developing countries as good or better as a measure of material well-being as the poverty rate (and clearly better than GNI per capita) for countries or groups of people within countries,45 combining information on the poverty rate and the poverty gap while also reflecting well the level of development broadly conceived in social and political terms; is “good enough” as a measure of inequality, when compared to mean consumption per capita, capturing for all practical purposes information equivalent to widely used measures of inequality including the Gini coefficient.
- This measure is “good enough” as a measure of changes in inequality over time when change at the median is compared to the mean, capturing the same information as the new World Bank measure of “shared prosperity”. Compared to the typical measures, a focus on the median and on changes in material well-being at the median is not only good enough as outlined above but has at least three additional advantages. In contrast to the typical measures using a poverty line, it is durable i.e. it will have political salience in all developing countries, including the uppermiddle-income developing countries, even as average per capita income increases in the next several decades. It better reflects broader measures of development. And in contrast to the World Bank measure of shared prosperity, the median when compared to the mean reflects changes at the top of the distribution (and without imposing any prior about at what percentile at the top).
- This good enough measure relies heavily on availability from household surveys of data on consumption (or income), in PPP terms, and ideally if not at annual frequency then at least every few years. The work at the World Bank, the United Nations and other development institutions to support the collection and management of these data should be applauded, particularly in low-income countries that can benefit from technical and financial support for collection and analysis of this basic data. Greater attention should be given to making these data fully accessible to the research community. The post-2015 development framework should include attention to the logic of creating the global public good such data represent in permitting broader learning about the determinants and consequences of development broadly conceived.