This project examined the effects of introducing biometric authentication into the payment process for NREGS participants and pension recipients in Andhra Pradesh (prior to its division into AP and Telangana). The authentication technology used (“Smartcards”) was a precursor to Aadhaar, but with important differences in the protocols used. Using a large-scale experimental evaluation, we found that the rollout of this reform (i) increased payments received by program beneficiaries without changing the amounts of money disbursed by government, (ii) did not significantly change measures of beneficiaries' access to the programs, and (iii) had large ripple effects on the rural economy more broadly, raising rural incomes by increasing both private sector wages and employment.

Key characteristics:

• Location: Andhra Pradesh, India

• Timeline: 2010 to 2012

• Sample: 8,000 households in 8 districts

Publications

• General Equilibrium Effects of (Improving) Public Employment Programs: Experimental Evidence from India

• Experimentation at Scale

• Building State Capacity: Evidence from Biometric Smartcards in India

• Implementing a Biometric Payment System: The Andhra Pradesh Experience

AEA Registry | Replication Data | Policy Briefcase | Media Coverage | more details

This project measured quality of implementation and beneficiary experiences during the Government of India’s pilots of Direct Benefit Transfers in lieu of food rations provided through the Public Distribution System in three Union Territories. We found that (i) implementation quality improved over time but remained a non-trivial challenge, with only ~⅔ of beneficiaries reporting that they received their benefits; (ii) beneficiary preference for DBT similarly increased over time, from ~35% of recipients to ~65% of recipients; (iii) the time and money cost of accessing and using DBT was higher than the cost of collecting PDS ration for those who used banks, but lower for those who used ATMs; and (iv) many beneficiaries were simply not aware that funds had been deposited in accounts in their name. Based on these findings, we suggested increased use of high-frequency monitoring of beneficiary experiences using mobile phones, and use of choice-based designs for future reforms to protect beneficiaries from potential adverse harms during initial rollout.

Key characteristics:

• Location: Chandigarh (CDG), Pondicherry (PDY), Dadra Nagar Haveli (DNH); India

• Timeline: 2016 to 2017

• Sample: 2,453 households in Chandigarh, 2527 households in Pondicherry, 1015 households in Dadra Nagar Haveli

Policy Paper

• "Direct Benefits Transfer in Food: Results from One Year of Process Monitoring in Union Territories"

This project evaluated the effects of requiring Aadhaar-based biometric authentication (ABBA) on the delivery of India’s largest social protection program, the Public Distribution System. We found that ABBA per se (i) did not change leakages, (ii) modestly increased transaction costs for beneficiaries on average, and (iii) substantially reduced benefits received for the minority of beneficiaries who had not linked (or “seeded”) their Aadhaar ID to the program database at baseline. Using the transaction data generated by biometric authentication to reconcile shipments of grain to Fair Price Shops did induce a large reduction in leakage, but also a significant reduction in benefits received by ration card holders. These results highlight the potential trade-offs between reducing corruption and protecting the entitlements of legitimate beneficiaries.

Key characteristics:

• Location: Jharkhand, India

• Timeline: 2016 to 2018

• Sample: 3,840 households across 132 sub-districts in 10 districts, randomized at the sub-district level, representative of 15.1 million beneficiaries in 17 of Jharkhand's 24 districts

Publications:

• Identity Verification Standards in Welfare Programs: Experimental Evidence from India

• Evaluating the Aadhaar-PDS link

AEA Registry | Replication Data | Policy Briefcase | Media Coverage

This project examined the impact of a cell phone-based monitoring system on the delivery of government-issued payments to farmers through the Rythu Bandhu Scheme in Telangana, India. We randomly informed 25% of the state’s frontline workers responsible for the disbursement of funds about the monitoring system and used a call center to reach out to farmers and collect information about their payments. Information from the calls and detailed administrative records were used to determine whether and when farmers received payments and consequently measure frontline worker performance. We found that phone-based monitoring (i) increased the likelihood of farmers receiving payments with the largest effect seen for farmers with the smallest landholdings (ii) increased the likelihood of farmers receiving payments in time to buy agricultural inputs for the planting season. We are now working with several state governments to apply this approach to measuring and improving last-mile service delivery in additional schemes.

Key characteristics:

• Location: Telangana, India

• Timeline: 2018

• Sample: 5.7 million farmers across 30 districts

Publications

• Improving Last-Mile Service Delivery Using Phone-Based Monitoring

AEA Registry | Replication Data | Policy Briefcase | Media Coverage | more details

This project is measuring the impact of cash transfers to pregnant women and lactating mothers on the health of their children as measured by the incidence of stunting and wasting, among other indicators. The transfers, which are being implemented by the Government of Jharkhand in eight districts, are unconditional in the sense that eligible mothers do not need to take additional steps after registering for the program in order to receive them, and framed in the sense that mothers are told the transfers are intended to help them feed themselves and their babies, and accompanied by information about how to purchase and prepare nutritious food. The project aims to inform ongoing discussion about how best to improve mother and child nutrition outcomes, and in particular the debate whether to provide cooked meals at anganwadi centers, take-home rations, or cash transfers.

Key characteristics:

• Location: Jharkhand, India

• Timeline: 2017 to 2020

• Sample: 6,000 pregnant women across 8 districts

• Status: On-going implementation and field operations

AEA Registry | Replication Data | Policy Briefcase | Media Coverage

This project is evaluating the take-up and impact of a reform giving beneficiaries in the Public Distribution System the choice to receive Direct Benefit Transfer (DBT) in lieu of their monthly ration each month. This approach is intended to de-risk the process of reform by reducing the chance that beneficiaries lose access to benefits due to implementation issues, while also generating direct feedback on which of the two proposed policy options (food or cash) beneficiaries themselves prefer. The project will also measure the impacts of providing beneficiaries with this choice on food consumption and nutrition outcomes. It builds on earlier pilot studies conducted with the governments of Bihar and of Rajasthan in which preferences varied widely, with 22% to 86% of beneficiary households opting for cash at least once as opposed to food depending on the context and time period.

Key characteristics:

• Location: Maharashtra, India

• Timeline: 2018 to 2021

• Sample: 1,148 households during the proof of concept

• Status: On-going proof of concept implementation

Publications:

• Assessing the Scope for Cash Transfers in lieu of the TPDS in Rural and Urban Bihar

AEA Registry | Replication Data | Policy Briefcase | Media Coverage

This project is evaluating the impact of a Direct Benefit Transfer for Electricity scheme in Punjab and Rajasthan, India. The scheme is being implemented jointly with the power corporations of each state and being evaluated by a research team led by Nicholas Ryan and Anant Sudarshan. The scheme replaces per unit subsidies for agricultural electricity with fixed cash entitlements payable to farmers. Farmers earn money if they consume less than the entitlement, giving them an incentive to conserve power. By providing this incentive, the scheme also aims to help reduce excess groundwater usage. The evaluation studies the impact of the scheme on (i) power consumption (ii) conservation of groundwater (iii) crop diversification and (iv) other farmer welfare outcomes, as well as drawing lessons for and insights into implementation.

Key characteristics

• Location: Punjab and Rajasthan

• Timeline: 2017- ongoing

• Sample: ~30,000 farmers in Punjab

• Status: Technical pilots complete; Full scale RCT in Punjab

AEA Registry | Replication Data | Policy Briefcase | Media Coverage

This project is designing and testing systems for measuring service delivery and beneficiary experiences in the Public Distribution System and the Mohalla Clinics in the NCT-Delhi region. It builds on our past experiences using phone-based monitoring as a cost-effective way to generate high-frequency data to improve frontline service delivery. We will place outbound calls to thousands of beneficiaries on a monthly basis and use the data to generate report cards for individual service delivery units, designed to identify unit-level service delivery issues and to help motivate frontline staff.

Key characteristics

• Location: National Capital Territory-Delhi

• Timeline: 2019-2021

• Sample: ~169,000 PDS beneficiaries per month; ~15,000 MC beneficiaries per month

• Status: On-going pilot

AEA Registry | Replication Data | Policy Briefcase | Media Coverage