Impact of direct benefit transfer on leakage in cooking fuel subsidy

Increasing burden of fuel subsidies is a leading concern for the policy makers across the world. There are significant fiscal implications of growing subsidy burden with underlying capricious trends in international fuel prices. Substantial energy efficiency gains are also likely when fuel subsidies reforms are enacted. Nonetheless, any attempt to reform fuel subsidy policies invites univocal opposition from all the quarters. Particularly in developing countries, debates on fuel subsidies invoke stronger passions since a significant proportion of income is spent on fossil fuels.

Without affecting subsidy amount intended for beneficiary households, a great opportunity to decrease fiscal burden of fuel subsidies lies in minimizing leakage and diversion of subsidies. In India, varying estimates suggest 10 to 40% leakage in Kerosene and cooking fuel (Liquified Petroleum Gas- LPG) subsidies. With about 2% of GDP being spent on energy subsidies, it translates into a significant loss to the exchequer (Anand et al. (2013)). At current prices, the gap between subsidized domestic LPG price (~ USD 7 per refill unit) and non-subsidized LPG price (~ USD 20 per refill unit) provides great incentives to the distributors and delivery-men who exploit loopholes to divert the subsidized refills to commercial sector and make profit in illegal LPG refills market.

In order to contain the leakage in cooking fuel subsidy disbursement, government of India introduced Direct Benefit Transfer for LPG (DBTL) program in June 2013. DBTL policy decouples subsidies disbursement from over the counter LPG refills sales and facilitates direct subsidy transfer to the intended beneficiary households. Once DBTL is made mandatory in a district, all LPG refills sales take place at the market price, while households get the subsidy amount directly into their Universal Identity (UID) linked bank account conditional on LPG refill purchase.

This project aims to provide an empirical evaluation of DBTL policy exploiting phase-wise implementation of DBTL policy in a difference in difference framework. It will also provide micro-level empirical evidence on the success of UID in facilitating benefits transfer programs. Particularly, the study focuses on whether direct transfer of LPG subsidy helps in reducing leakage in presence of an underground market for LPG refills and if so, how much subsidy leakage can be avoided by providing it directly to the households. Household level LPG refills purchase data from across the country are employed covering about 1 million randomly selected LPG beneficiary households. In addition, LPG prices in underground LPG refills market are being tracked in about 90 districts since December2013.

Till recently, DBTL policy was being implemented in phases across the country and was introduced in 291 districts under six phases till January 2014. However, in a sudden development in late January this year, government of India announced to roll back the DBTL policy. This sudden change in policy provides an additional opportunity to look at the impact of removing the DBTL policy treatment on the subsidy leakage and on prices in the underground LPG refills market.

This study will contribute to ongoing debate in India on subsidy provision through direct cash transfer. Since DBTL policy is currently being reviewed, this analysis is likely to make a direct and timely contribution. Furthermore, the project will add to a small but growing empirical research on corruption in developing countries (see Olken and Pande (2012) for a review). Broadly, the setup relates to the point of tax incidence and evasion literature in public finance and allows us an opportunity to test dependence between point of provision of subsidy and subsidy leakage. The outcomes of this project will be particularly relevant to the context of carbon taxation policy where the role of point of incidence of taxes has not been given much attention (Kopczuk et al. 2013).

Outputs