A quote from Article 280 of the Constitution is useful: “It shall be the duty of the [finance] commission to make recommendations to the president as to (a) the distribution between the Union and the states of the net proceeds of taxes, which are to be, or may be, divided between them… (b) the principles which should govern the grants-in-aid of the revenues of the states.” Splice that with Article 275, on grants-in-aid, and the idea is clear. After catering for the Union government’s responsibilities set out in the Seventh Schedule, there is a kitty of the net proceeds of taxes.
A finance commission will determine how this is to be shared between the Union and the states. Once the aggregate share of the states is determined, a finance commission will also determine the shares of different states out of the total. This is the so-called vertical and horizontal calculation, and pertains to clause (a). It has nothing to do with grants-in-aid. This may not be the best way to put it, but clause (a) follows the principle of equality. Once a formula is set (and different finance commissions have used different formulae), the shares of different states of the aggregate state kitty are mechanically derived. To extend this analogy, clause (b) is more about equity or fairness, the idea behind grants-in-aid. Irrespective of where a citizen resides, she must have access to the same minimum level of goods and services.
Let’s call these public goods and services, assuming they have to be delivered by the government. I don’t think any finance commission, even the 14th, has been able to address the conceptual difference between clauses (a) and (b) cleanly. Though there are separate provisions for grants-in-aid, the formulae that have been used, particularly from the Eighth Finance Commission onwards, have had equity indicators built into them. I am not especially picking on Union finance commissions. The charge is also applicable to state finance commissions, and to expenditure through Centrally sponsored schemes or schemes with Central assistance. At all three tiers of government, goods and services have to be provided through public expenditure. But even after we have agreed on the minimum threshold level of goods and services, costs aren’t uniform across the country.
Think of the special-category states. What’s “special” about them, and why has this concept been around since 1969? Because of their strategic location along the borders? But this is a slightly different argument and I am not talking about the Gadgil-Mukherjee formula. As the terminology suggests, this is formula-driven. Nor do I have in mind the states’ low resource base. I have in mind Central assistance. A major argument behind this has been hilly and difficult terrain.
Hilly and difficult terrain isn’t specific to the 11 special-category states. Other states also grapple with hilly and difficult terrain. Indeed, at one level, particularly for large and heterogeneous states, it isn’t even a state-level problem. It’s a district-level problem, or perhaps even more localised than that. To address equity issues under clause (b), one needs to know how much it costs to deliver (per capita per unit, say) goods and services in different parts of the country. This may be difficult to estimate, but that’s no reason not to even attempt it. Sure, surrogate indicators have been used to measure cost disability, such as the deviation of physical or social infrastructure from the average, area and population, and weights have been attached to these indicators in formulae. But these surrogate indicators measure the baseline gap; they don’t quantify cost.
Take Chhattisgarh, which has a geographical area of 1,35,190 sq km. Does this figure give me a sense of how much it costs to deliver goods and services in the state? Probably not, because the hilly areas of its north and south are different from the plains in its middle. With all the data now available, including through satellite imagery, can one not factor in elevation and work out Chhattisgarh’s topographical area? That would be a far better indicator of cost than geographic area.
A tangential thought: the Central Ground Water Board has a watershed atlas, with a basin map, divided into 34 basins. If one wishes to zero in on intra-state differences, this would probably give us a decent idea of region-wise costs. It’s probably also possible to disaggregate the costs to get at the sub-basin-level figures. For understandable administrative reasons, development discussion has tended to focus on state-level considerations. Switching the discourse to the cost of delivery permits better appreciation of intra-state differences.
For instance, the 127 agro-climatic zones identified under the National Agricultural Research Project give us a better sense of agricultural issues. I am not advocating the usage of topographical area alone. Plausibly, per capita and per unit, it costs less to deliver goods and services in a village with a population of 10,000 than in one with a population of less than 200. (In Census 2001, there were around 90,000 villages with a population of less than 200, with large absolute numbers in Himachal Pradesh, Uttarakhand, Rajasthan, Uttar Pradesh, Jharkhand and Odisha. We don’t yet have the numbers for Census 2011.) With development, for various reasons, some smaller villages disappear. When Gandhiji wrote about India living in its villages, he mentioned the number of villages too. There were seven lakh. The number is 6,40,000 now, not all inhabited. Let’s avoid that digression. The limited point is that building a road in Arunachal Pradesh is not the same as building a road in West Bengal — purely because of cost considerations.
The writer is member, Niti Aayog. Views are personal.