If, as an NIPFP paper on the costs and benefits of Aadhar estimates, the leakage in MGNREGA on account of ghost workers and manipulated muster rolls is only 12%, then this is more a cause for celebration than one for dismay. The paper assumes this level of leakage (between 7-12%) in other public programmes—PDS and fertiliser, LPG and education subsidies—as well. Though it would be great to believe that the level of theft in these programmes is so limited, reality does not allow us that luxury. Take the example of MGNREGA, which is targeted at the poor population. As economist Surjit Bhalla finds after examining NSS 2009-10 data (http://goo.gl/auMse), at an all-India level, 20% of the non-poor population were beneficiaries of MGNREGA; this number was 54% in Rajasthan and just 3% in Maharashtra. Non-poor beneficiaries of MGNREGA, though eligible for work under the scheme, nevertheless represent a leakage in a system aimed at the poor.
In the case of PDS, the NIPFP paper is similarly lenient. It takes the amount of PDS kerosene that does not reach the intended recipients as 38% (a number calculated independently by NCAER and Nandan Nilekani back in 2005), but then reduces 25% from this on account of the supposed improvements made to the system over the years. In the case of PDS foodgrain leakage, the report estimates non-existent beneficiaries account for 16% of the total amount of subsidised grain. This again pegs the leakage at an optimistic but unrealistic level. As calculations made by Rajesh