Building state capacity: evidence from biometric smartcards in India

Replication paper
Original publication: American Economic Review
Original researchers: Karthik Muralidharan, Paul Niehaus, Sandip Sukhantankar
Replication plan:

Atanda's replication plan

Replication researchers: Akinwande Atanda
Current status: Completed Replication Study

The Original Study

Most parts of the developing world lack the infrastructure to efficiently process and deliver payments to beneficiaries of anti-poverty programs. As a result, many poor are financially excluded, or receive only a portion of the funds intended for them. The paper “Building State Capacity: Evidence from Biometric Smartcards in India” by Muralidharan, Niehaus, and Sukhtankar (American Economic Review, 2016) studies the effect of implementing anti-poverty payments via Smartcards.  The authors worked with the government of the Indian state of Andhra Pradesh to randomize the rollout of Smartcards across 157 sub-districts with 19 million residents. They studied the effects for recipients of India’s two largest welfare programs (a work-for-payment scheme and national pension program). They found that Smartcards (i) decreased the time lag for recipients to receive funds, (ii) reduced leakages of benefits; and (iii) increased enrolment rates in the two programs.

The Replication

Improving human welfare has become a global objective, most especially in less developed countries with high incidence of poverty. Muralidharan, Niehaus, and Sukhtankar (2016) provided empirical evidence using a large-scale experiment to highlight that it is essential for government to build technical infrastructure to enable prompt payment transfers and reduce the theft of money meant for the poor by corrupt government officials, in order to improve welfare and financial inclusion of the poor. We conduct a pure replication of the original study using the same data and methods. Then, we will investigate the robustness of the original study to (i) outliers; (ii) alternative model specifications; and (iii) different estimation methods. Lastly, this replication study will test a new hypothesis to identify if there are moderating factors that contribute to the heterogeneous effects of biometrically-authenticated Smartcards in India.