India’s Mahatma Gandhi National Rural Employment Guarantee Scheme (MNREGS) has generated a lot of controversy about its effectiveness as a safety net designed to benefit landless rural households. Muralidharan, Niehaus, and Sukhtankar discuss the results from a large-scale randomised evaluation that suggest MNREGS may be a surprisingly effective tool both for improving the welfare of the landless rural poor and increasing overall rural productivity.
The increased policy attention on reducing economic distress among farmers is welcome. Yet, it is as or more important to consider the welfare of the millions of landless rural households who depend mainly on wages, and not cultivation. They do not benefit directly from subsidies, loan waivers, minimum support price increases, or even income transfers to farmers. Further, the one component of India’s safety net designed primarily to benefit them – the Mahatma Gandhi National Rural Employment Guarantee Scheme (MNREGS) – is under pressure.
The MNREGS is the world’s largest public employment programme, with over 600 million eligible workers, and it has generated a commensurate amount of controversy. Proponents argue that it provides a lifeline to the rural poor in the lean season, raising rural wages, and enabling the creation of productive assets. Detractors argue that it is wasteful, plagued by corruption, and creates unproductive holes in the ground. Much good research has been done on these issues, but all are subject to important technical limitations, and policy views on MNREGS seem to be informed more by ideology and opinion than credible evidence.
To make progress, we use data from an unusually large-scale experimental evaluation where the government of (then unified) Andhra Pradesh randomised the roll-out of biometric smartcards for making MNREGS payments across nearly 20 million people. In prior work studying the impact of smartcards on MNREGS implementation quality, we found substantial improvements on several dimensions: Leakage fell by 41%, programme participation increased by 17%, the time lag between working and getting paid fell by 29%, the time to collect payment fell by 20%, and the variability in the payment lag fell by 39%. In other words, the use of smartcards substantially improved the effective presence of MNREGS on the ground and brought the implementation quality closer to what MNREGS architects had intended.
In addition to informing the ongoing debate on the role of biometric authentication in social programmes, the experiment also gives us a unique opportunity to answer a core question about MNREGS itself: What happens to the rural poor when MNREGS implementation is improved? The results are striking (Muralidharan et al. 2018). In treated areas, the incomes of MNREGS job card holders increased by 13% while overall poverty fell by 17%. These results from our survey data match those using the completely independent Socio-Economic and Caste Census (SECC), which also shows a significant reduction in poverty.
Some of these gains simply reflect the fact that corruption fell, and MNREGS earnings increased. But this turns out to be a relatively small part of the story. In fact, nearly 90% of the income gains we measure come not from the MNREGS itself but from increases in market earnings. In particular, we find a significant increase in market wages, perhaps because a better-implemented MNREGS forced private employers to raise wages to attract workers. Moreover, private employment did not fall as a result; once we account for spillovers into neighbouring sub-districts, we find that it actually increased.
How could both wages and employment go up at the same time? First, lower leakage could have improved public asset creation, thereby increasing productivity, wages, and employment. Second, if employers had monopsony power and were able to coordinate to keep wages low, then economic theory predicts that an increase in minimum wages can also increase employment. Finally, reduction in credit constraints (which we find evidence of) could have boosted private investments and productivity. Over time, the increase in rural wages may also speed up mechanisation of agriculture, which would further increase productivity as seen in historical evidence from the US (Hornbeck and Naidu 2014).
Overall, a better implemented MNREGA reduced poverty without ‘crowding out’ private sector economic activity. Regardless of the underlying economic mechanism, this is a crucial fact for policymaking. It has certainly shifted our own thinking. We began our own work on the MNREGS from a posture of scepticism, documenting corruption and seeking to mitigate it (Niehaus and Sukhtankar 2013). But our results from a highly credible, randomised evaluation suggest that MNREGS may be a surprisingly effective tool both for improving the welfare of the landless rural poor and also increasing overall rural productivity.
Policies that improve both equity and efficiency are quite rare, and a well implemented MNREGS may fit this category. Given what we currently know, the government should strengthen MNREGS rather than cutting it, or letting it slowly atrophy though weak implementation. It is good that the recent budget has increased the allocation for MNREGS. The government should now follow through to deliver on the full potential of this allocation, ensuring that funds reach projects and beneficiaries in a timely manner. Prioritising timely wage payments as well as asset quality will improve both short-term beneficiary welfare and long-term rural productivity.
This article first appeared in the Hindustan Times: https://www.hindustantimes.com/columns/strengthen-nregs-to-support-the-rural-economy/story-uOlHMjpeqcpe6fVXWekeRI.html
- Hornbeck, Richard and Suresh Naidu (2014), “When the Levee Breaks: Black Migration and Economic Development in the American South”, American Economic Review, 104(3): 963-990.
- Muralidharan, K, P Niehaus and S Sukhtankar (2018), ‘General Equilibrium Effects of (Improving) Public Employment Programs: Experimental Evidence from India’, Working Paper.
- Niehaus, Paul and Sandip Sukhtankar (2013), “Corruption Dynamics: The Golden Goose Effect”, American Economic Journal: Economic Policy, 5(4): 230-269.