Source: Flickr, UNDP

Public works programs can form an important part of developing countries’ social safety nets, helping to protect poor populations’ access to income during times of shock. In order to be effective, however, public works programs need to be properly set up and implemented. In a new article published in World Development, researchers from IFPRI, the Indira Gandhi Institute of Development Research (IGIDR), the University of Pennsylvania, and Cornell University look at the Mahatma Gandhi National Rural Employment Guarantee Act (MGNREGA) to determine whether implementation failures are behind the decline in the program (in both person-work days and total expenditure) seen since 2009.

There remains much debate regarding this decline. Some experts argue that the program has done its job and that fewer people are utilizing it simply because fewer people now need it. However, other experts take the opposite view. Like many public works programs, MGNREGA is demand-driven, meaning that beneficiaries can choose whether or not to participate in the program. The authors posit that if the program suffers from widespread implementation failure – such as delayed wage payments or administrative rationing (denying employment to qualified applicants) – potential beneficiaries may choose not to use the program. This is known as the “discouraged worker effect”.

To test for this effect in the MGNREGA program, the study uses data from two rounds of the National Sample Survey (NSS; 2009-2010 and 2011-2012). These rounds covered 59,129 and 59,700 rural households, respectively. The household-level analysis used data from the 2011-2012 survey regarding whether any household member sought work from the MGNREGA program in the past year; this was then combined with data from the 2009-2010 survey regarding district-level rationing rates and delays in wage payments. This combination allowed the authors to determine whether failures in implementation of the MGNREGA program in earlier years led to reduced demand for the program later. The authors also conducted a district-level analysis that used the work-seeking rate as an indicator of demand in each district. Two econometric models were used to conduct these analyses – a probit model and a linear probability model (LPM).

The results show that administrative rationing has a negative impact on households’ demand for MGNREGA work. In the probit model, when administrative rationing rises by 10 percent at the district level, the probability that a household will seek work falls by 3.4-3.9 percent. This relationship is even stronger in the LPM, which shows a decrease in work-seeking of 8.4-9.2 percent when there is a 10 percent increase in administrative rationing.

The results regarding the effects of delayed wage payments are less consistent. In the probit model, there is no evidence that a delay in payments has any impact on households’ likelihood of seeking work. The LPM, on the other hand, finds strong evidence that delayed payments do in fact decrease work-seeking behavior; however, even in the LPM, delayed wages appear to have less influence over households’ work-seeking choices than does administrative rationing.

The study also examines possible reasons for administrative rationing – whether such rationing is politically motivated or is due rather to a lack of administrative capacity at the district level. The results suggest that political motives have little to do with administrative rationing. Rather, rationing tends to happen when district-level authorities are overwhelmed by other demands, such as dealing with weather shocks like drought, or when banking infrastructure is not sufficient to support the needs of the program.

The authors suggest that administrative rationing can be addressed in part by ensuring that enough well-trained staff are available at multiple administrative levels so that the program can be scaled up when shocks, such as drought, hit an area and drive up demand for the program’s benefits. In addition, issues related to delayed wage payments can be addressed by investing in district-level banking infrastructure to help ensure that payments can be disbursed in a timely manner.

The authors also emphasize that understanding and addressing implementation failures is essential in improving the functioning of public works programs like MGNREGA. While the recent decline in MGNREGA usage has been interpreted as a sign that the program had served its purpose and is no longer as widely needed, this study’s results suggest the true causes behind that decline may warrant further investigation.

Photo credit:Flickr, UNDP