Accuracy of Targeting and Implications of Rationing in National Rural Employment Guarantee Scheme: Evidence from West Bengal, India
Posted: 16 Sep 2012
Date Written: July 15, 2012
One of the most significant feature of the National Rural Employment Guarantee Scheme (NREGS) has been the demand based principle, which allows for no rationing of the labourers, who seek work under the scheme. However, studies have found that rationing has been one of the significant problem of this programme, where households have ended up not getting work, despite of seeking. This papers attempts to study whether under the system of rationing, the households, who got work are poor and from the disadvantaged sections. It also attempts to find if the households, who sought but did not get work are poor and disadvantaged. Since studies have shown that the scheme has been very poorly implemented in West Bengal, we study the above for the state. It uses National Sample Survey data for the 66th round (2009-10) and makes use of various parametric as well as non parametric techniques including bivariate probit model with sample selection along with logistic regressions. The study finds that the poor have been relatively well targeted by the programme in the sense that the better off households did not work. However it finds SC having significantly negative probability of getting work controlling for all the other factors. The paper then emphasizes on the implications of rationing, which suggest that substantial number of poor are denied work despite of their seeking. Further, the study goes on to analyse if the number of days worked by the households, who participated in the programme are biased towards the non poor. In other words, it attempts to explore if the non poor, who worked, got more number of days to work for the scheme compared to the poor households. Along with the non parametric test, count regressions with Heckman sample selection have been used. Preliminary results suggest that the poor and disadvantaged households do not get work for higher number of days compared to the non poor. Further, the paper also presents some qualitative evidences collected from through household interviews and focus group discussions in different districts of West Bengal to support the results obtained from the quantitative approach.
Keywords: Targeting, Rationing, NREGS, West Bengal, Stochastic dominance, FGT, Bivariate Probit, Sample Selection
JEL Classification: I38, I32, J22, C14, C21
Suggested Citation: Suggested Citation