According to NSS 2007-08, a total of 477 million mandays of work were created under non-NREGA public works and a total of 492 million under NREGA public works. This is the first time the NSS ever made a distinction between NREGA and non-NREGA public works.
In 2007-08, the government spent Rs 10,800 crore on NREGA and created, according to the NSS, a total of 492 million persondays of employment. At the average wage of Rs 77.85 reported by the NSS, thats a total wage bill of Rs 3,833 crore. Yet, going by the NREGA norm of a 70% wage component, the government should have given out Rs 7,560 crore as wages. So, the net effectiveness of the programme was just above 50%. Looked at another way, the effectiveness was about a third, since the government spent Rs 10,800 crore to deliver Rs 3,833 crore of wages.
Compare this with the non-NREGA public works expenditure. A total of Rs 16,538 crore was made in 2007-08 for non-NREGA public works. Since wage components are not stipulated for all of them, we have made some assumptions. The wage component of all projects has been taken to be 60%, except in the case of rural housing and rural roads where it has been taken to be 20%. In which case, the wage component of the budget allocation was Rs 5,540 crore. Compare this with the Rs 3,546 crore that the workers should have got, and the effectiveness is around 64%.
Since the NSS does not make a distinction between central-government sponsored public works and state-government sponsored public works, we have assumed all non-NREGA works are centrally sponsored. Once you factor this in, it may change the efficiency levels, but not enough to change the order of magnitude.
If you assume that the expenditure has not resulted in any significant asset creation, the effectiveness of expenditure drops dramatically Rs 16,538 crore was spent to deliver Rs 3,546 crore of wages, or an effectiveness of 21%.
Interestingly, 56% of the beneficiaries of the non-NREGA public works were poor (according to the new official Tendulkar poverty line which is some 20-25% higher than the old Indian poverty line) versus 52% in the case of NREGA works. The data also show and this is the good news that the effectiveness of government programmes has risen, from an efficiency level of 35% in 1993 to 51% in the case of non-NREGA programmes. The larger point remains that it is far more efficient to transfer cash in the case of NREGA, the government spent Rs 10,800 crore to transfer Rs 3,833 crore to workers in 2007-08.