PM’s ‘revdi’ remark: We need to disentangle good subsidies from bad

Freebies typically conjure up images of free televisions distributed by the late J Jayalalithaa, free cycles distributed by Nitish Kumar or laptops distributed by Akhilesh Yadav. While these are the most highly visible and discussed freebies, they are fiscally insignificant compared to the much larger subsidies on food, fertiliser and petroleum. Though curbed in recent years, these “visible” subsidies in government budgets remain a major source of fiscal stress. Then there is a range of “invisible” subsidies, especially in state government budgets, not always recognised as such, but which are also very large.

Technically, a subsidy is the unrecovered cost of any service (or good) provided by the government. The deficit between the receipts and expenditure of a government department in providing a service is the unrecovered cost of providing that service, that is, a subsidy, even if not recognised as such in the budget. Examples include the unrecovered cost of providing public education, healthcare, irrigation, power, water supply and sanitation. It will be immediately evident that not all subsidies are equally undesirable. A final category of freebies is pure cash grants for poor households. These are discussed below.

Returning to subsidies, the total volume consists of all visible subsidies plus all the invisible subsidies implicit in the provision of social and economic services. In an exercise undertaken in 2020, the late Satadru Sikdar and I found that the volume of subsidies as a proportion of GDP comes down with rising per capita incomes, but very gradually (“Merit goods and the fiscal space for reviving growth”, Economic & Political Weekly, February 1, 2020). The total volume of subsidies came down from 13 per cent of GDP way back in 1987-88 to a little over 10 per cent by 2015-16, almost 30 years later. This proportion is unlikely to have changed very much today given the slow pace of change. State governments provide the bulk of these subsidies, mainly for social services like education and health. The central government accounts for less than 30 per cent of total subsidies, provided mainly for economic services including food.

In the mixed bag of subsidies some are less warranted than others. From the total volume we separated out a very small number of “merit subsidies” which might be warranted in public interest. All governments have provided a food subsidy for poor households by bi-partisan consensus for decades. It has been increased by the present government as relief during the Covid-19 crisis. Then there is basic education and health services which have large benefits for society beyond the benefit accruing to the immediate recipient of the service, what economists call “externalities”. Finally, we have separated out expenditure on water supply and sanitation, where again the benefit to society is much larger than that accruing to the immediate recipient of the service — for example, prevention of infectious diseases. These four “merit” subsidies account for only a third of total subsidies. Thus, two-thirds of total subsidies, about 6 per cent of GDP, are unwarranted freebies which should be eliminated.

As explained in the article cited above, if central and state governments could step beyond their business as usual budgets and take bold measures to phase out these unwarranted freebies, along with much of the tax exemptions and concessions, which amount to about 5 per cent of GDP, that would free up huge fiscal space. This would enable a massive reduction in the combined fiscal deficit of the Centre and the states, while at the same time stepping up required expenditure on education, health and infrastructure. The myth of restricted fiscal space simply reflects the missing appetite for deep fiscal reforms which could radically change the structure of central and state government finances.

Finally, I turn to the question of income support schemes. There is a growing demand in many advanced countries, which already have large social security schemes, to provide a minimum “Universal Basic Income” for all. These demands are supported by global corporate leaders like Mark Zuckerberg as well as leading economists and think tanks across the ideological spectrum. Against that background, providing a small safety net for the poor in countries like India, which have no social security system, is the least that any caring government can do. Especially, when the government is foregoing much larger sums of revenue as tax concessions for the rich and banks are writing off vast sums of non-performing loans also provided to the rich. MGNREGA is the largest and longest-standing income support programme in India for the unemployed in rural areas. But it is often not regarded as such as it entails payment against performance of work. There are also pure cash grant schemes such as the PM-Kisan of the central government and some schemes now being piloted by a few state governments.

The usual complaint against such schemes is that they artificially raise rural wages, reduce the incentive to search for work, and that the poor blow up these freebees on liquor etc. Since MGNREGA and similar schemes in the states pay much less than the minimum wage, they obviously cannot raise rural wages beyond what is the legal minimum wage anyway. As for the somewhat patronising critique that such support will induce the poor to not search for work or binge on alcohol, there are now dozens of randomised control trials (RCTs) across the world, led by Nobel Laureates Abhijit Banerjee, Esther Duflo and Michael Kramer, that disprove such beliefs. I should specifically mention the RCT conducted by the well-known women’s voluntary organisation SEWA in Madhya Pradesh. It showed that the very small cash support provided in the villages where the intervention was carried out was spent primarily on seeking better education for children, repairing dwelling huts and supplementing the very meagre diets of these poor households.

The writer is Chairman, Centre for Development Studies