Summary
Highlights
The Multidimensional Poverty Index (MPI) is a key measure of multidimensional poverty, calculated annually by the United Nations Development Program and presented in the Human Development Report. It focuses on identifying who is poor by using cutoff levels within each dimension of deprivation, similar to a poverty line for income.
The MPI focuses on three equally weighted dimensions of deprivation: health, education, and standard of living. Within health, indicators include child mortality and malnutrition. Education indicators are school attendance and years of schooling. The standard of living dimension has six indicators, each contributing 1/18th to overall deprivation, including lack of electricity, unsafe drinking water, inadequate sanitation, poor flooring, unimproved cooking fuel, and absence of multiple assets (telephone, radio, TV, bicycle, motorbike).
The MPI approach considers possible interactions and complementarities between deprivations by building the index from household measures rather than averaging averages. This allows for an understanding of how multiple deprivations can reinforce each other, as seen in the links between asset deprivation and health or education.
The MPI is calculated as the product of the headcount ratio (H) – the fraction of people who are multidimensionally poor – and the average intensity of deprivation (A) among those who are poor (MPI = H x A). This adjusted headcount ratio satisfies desirable properties like dimensional monotonicity, meaning an increase in deprivation for an already poor person is reflected in the index.
An example demonstrates MPI calculation with three people and three dimensions (health, wealth/standard of living, and education) using a deprivation cutoff of 0.5. The example shows how to determine who is multidimensionally poor, calculate the headcount ratio (H), and the average intensity of deprivation (A), ultimately leading to the MPI value. In this example, with a cutoff of 0.5, H is 2/3 and A is 5/6, resulting in an MPI of 5/9 (or approximately 0.555).
The MPI provides a more comprehensive understanding of poverty than income-based measures. Countries like Bangladesh are shown to be less poor by MPI than by income, while Pakistan is poorer. Similarly, Ethiopia is more MPI-poor than predicted by income, contrasting with Tanzania. Latin American countries often rank worse on MPI than income poverty, with Colombia being an exception. There is significant variation in multidimensional poverty across developing countries, with some nations like Nigeria showing almost 90% of their population in multidimensional poverty.
The video concludes by pointing out that even within a single country, like Cameroon, there can be substantial regional differences in the intensity of poverty, emphasizing that the MPI not only measures who is poor but also the depth of their deprivation.