In India, most public services provided to citizens are delivered by the governments of the states they live in. The theory of federalism tells us that State governments are likely to have a better idea of their particular conditions, needs and priorities than the Union Government. The 14^th Finance Commission (FFC) recognised this, and hence one of its major concerns was to “provide enhanced fiscal flexibility to the states to meet their expenditure needs and make expenditure decisions in line with their own priorities.”[i]

In an earlier blog, Avani has talked about how successful the Commission’s recommendations have been at meeting this objective. Most states now receive more money from the Union, and so we can conclude (with some caveats, as Avani describes) that in some measure most states have more fiscal autonomy.

The question which naturally follows is that of how states are using this fiscal autonomy. This is a particularly pressing concern for the social sector, since any changes in budgets have had a history of impacting “non-plan” expenditures: mainly recurring expenditure for public services! In addition, there was much distrust of State governments: a feeling that they would prioritise infrastructure over the social sector. Yamini talks about how even government officials at the states felt that the social sector would be squeezed.

When last year’s Union Budget revealed that Centrally Sponsored Schemes (CSSs) were being cut, and that states were expected to make up the deficit, much ink was poured and many hands were wrung in despair. As Jean Dreze put it, “anyone with a minimal understanding of Centre-State relations is likely to hear alarm bells.”[ii]; this feeling was more generally shared by many working in the fields of education, health, women, children, and so on. [iii] [iv] [v] [vi]

When my colleagues at AI and I were looking at the budgets of 19 states, this was at the top of our minds. What do the numbers in states’ budgets tell us about their plans for social sector spending after the FFC report?

Click here for full summary

The good news is that every single state intended to spend more money on the social sector in the Revised Estimates (RE) of 2015-16, as compared to their actual expenditures in 2014-15. The highest increases were visible in many of the poorest states including Bihar (46%) Chhattisgarh (49%) and Jharkhand (53%). Telangana nearly doubled its social sector expenditure, with an 86% increase.

Interestingly, even when looking at per-capita social sector expenditures in 2015-16 RE, some poorer states planned to spend more than wealthier states. For instance, Chhattisgarh intended to spend Rs. 9,877 per-capita, compared with Kerala at Rs. 8,590 per-capita. Similarly, Odisha intended to spend Rs. 200 more per-capita than Punjab.

Another useful way of looking at these numbers is to understand the share of the social sector in the total expenditure. This gives an understanding of the priority of the social sector in the state. Put another way, if all states are spending more money overall, the social sector will also increase. However, the share gives an idea of whether the increase in social sector spending is more or less than the increase in overall expenditure.

Even by this measure, we find that 16 of the 19 states we studied have a higher share of social sector expenditure. Sharp increases are seen in Andhra Pradesh, Uttar Pradesh, Telangana and Madhya Pradesh, and modest increases in other states. The 3 states which haven’t seen an increase are Kerala, where the share is constant, and Odisha and West Bengal, where the share has fallen by 2 and 1 percentage points.

Within the social sector, of course, states could choose to spend on different sectors. We would expect, if states are using their autonomy, to see changes not just in the aggregate level of social sector spending, but also in particular sectors.

For a few states where we looked at detailed sectoral budgets, we find some evidence of changes. In 2015-16 RE, Karnataka increased allocations for housing by 91%, while Maharashtra doubled allocations for water supply and sanitation, while total social sector expenditure grew by 18% and 23% respectively. Similarly, Uttar Pradesh, doubled investments in the budget head “Welfare of Scheduled Castes, Scheduled Tribes and other Backward classes”, while Bihar increased investments in “Labour and Labour Welfare” by 78% in 2015-16 RE compared with the previous year. There is some evidence that Bihar is changing patterns also in the current year’s budget: the state is reducing expenditure on education by 10%, while sharply increasing allocations to medical and public health by 83% in the budgeted estimates of 2016-17.

This data comes with some caveats. The biggest problem is that we are comparing revised estimates of expenditure in 2015-16 to money actually spent in 2014-15. Many states find it very difficult to spend allocated money on the social sector, particularly through CSSs, and often large amounts of money languish unspent. It’s likely that when audited expenditure numbers for 2015-16 come in, the amount of money actually spent on the social sector will fall. In addition, money which states are spending need not spent wisely, or even with any effect at all. Often outcomes such as children learning, or fall in mortality, seems to improve very little despite changes in the amount of money spent.

How can we get a more reliable estimate of expenditures? What can we do to ensure that social sector spending grows not just in the aggregate, but also in every state in areas of priority? The next couple of blogs will attempt to answer the way forward for both states and the Union.