A study to understand how availability of public goods in Bihar can be incorporated to improve poverty measures

Publicly provided goods and in-kind transfers vary substantially across Indian villages, yet they have never been explicitly incorporated in the official poverty measures. Since public provision influences household consumption decisions and levels, poverty numbers based solely on private consumption data are necessarily biased. This policy brief proposes a method for adjusting the distribution of consumption based on the local availability of public goods. The analysis is based on primary data collected from 40 villages in Bihar in 2012. Kjelsrud and Somanathan consider three types of public services: schooling, health care and subsidised food grains through the Public Distribution System (PDS). They show that accounting for the use of these services leads to a narrowing of the consumer expenditure distribution and lower inequality in Bihar because the poor utilise the public facilities more intensively than other households. However, Kjelsrud and Somanathan also find that such accounting leads to a rise in the regional dispersion in poverty rates. So while the targeting within villages leads to a fall in the overall inequality, facilities are not always located in the poorest villages. Finally, this study shows that the changes in measured poverty from all three types of public services are positively correlated and there is clustering in the location of publicly provided goods. It is likely that this clustering has implications for growth since investments in education, health and nutrition are estimated to have high marginal returns in the literature and have also been shown to be an important historical precondition for growth many parts of the world.