Elected governments have to meet the expectations of a population alerted as never before to their unfilled needs from government. Disease, lack of sanitation, illiteracy and lack of access to transportation crowd into their lives like smog, and they want it waved away. The magnitude of what needs to be done on multiple fronts is huge, and the intensity with which the voter demands action is commensurate with the deprivation to which they have long been subjected.
Because of fiscal constraints, no state government can provide its entire population with any particular service all at once. Since universal coverage can only be an eventual target, what matters is the spatial manner in which that target is approached. This is because the public goods that lie at the core of what governments are supposed to provide are limited in their spatial reach. A functioning public toilet in one slum area is not accessible to residents of another. So, while moving towards a target, when coverage is nowhere near universal, it is important to ensure equality of access in all areas of the state, at whatever fractional level.
Although there has been some prominence lately in national discourse on inequalities between states, drummed up principally by Bihar, the issue of fair allocation of fiscal resources flowing from Centre to states is largely resolved (which is not to deny that there remains immense scope for improvement). But what has not attracted as much attention is the very considerable inequality existing within states. This issue arises if at all in the context of states with a Maoist problem, or in the current Andhra partitioning; but no one actually questions budgetary allocations between districts within states for the very simple reason that these are not known.
Most public services in states have not been functionally transferred to local bodies, and are still done by line departments of state governments. The allocation of funds between districts would of course be subordinated to some kind of administrative guidelines, if not very tight formulae, but these are not hung out in the courtyard for all to see. They lie buried in files and could only perhaps be uncovered by an application under the Right to Information.
But that is not necessary, as it turns out. There is a Census Directory that comes out separately for villages and towns in all states some years after the Census. The 2011 Census Directory is not yet out, but that for 2001 was out in 2007, and even though the data pertain to 2001, they yield a lot of information on the cross-district patterns of allocation at less-than-universal levels of provision.
The great problem in all states in India is that districts are so very uneven in size. In a sense, this mirrors the problem at national level as well, where the government has to divide fiscal flows equitably between states varying massively by population and area. But because of the formulaic and very public way in which it is done at the Centre, the allocation is known.
At state level, likewise, districts vary enormously in size by population, and even more by number of village habitations. In an exercise I did for 20 large states using the 2001 Village Directory, I found that the percentage of villages served with a primary school or medical facility was not lower in districts with a larger population (except in two states). Clearly, budgetary allocations do accommodate variations in district population for the most part. But, in 18 of the 20 states, there was systematically lower village coverage in districts where the population was splintered into larger numbers of villages. Gujarat and Tamil Nadu were the two exceptions where there was no fiscal penalty for village splintering.
In nine of the 18 states, more splintered districts had a higher percentage of tribal population; in four states, they had a higher percentage of scheduled castes. There you have it - what a formula failure can do for systematic deprivation of a demographic group, even while avowed policy at both state and central levels is to have special budgetary provisions for disadvantaged castes and tribes. These supplementary budget provisions for scheduled castes and tribes can do nothing to correct a fundamental defect in the formula used for mainstream budgetary allocations.
Budgetary allocation formulae are the kind of deep structural underpinning that very few state governments bother about particularly. Not surprisingly, in the recent round of state elections, a large minority among voters in tribal constituencies rejected all the candidates on offer.
Fortunately, this error can be very easily corrected by states adopting adjustment for number of villages rather than population in budgetary allocations between districts. Even better would be a reorganisation of districts in all states into jurisdictions of roughly even size by number of villages. District boundaries are not as sacrosanct as state boundaries and should in principle pose fewer obstacles to resizing. The only objections to this might come from economists, for whom geographical redefinition introduces an unwanted discontinuity in time-series data at district level.
But how important are public services anyway for the re-election prospects of a party in power? Is the alternative of giving out subsidised foodgrain a better alternative? From the recent poll results, the best inference would seem to be that subsidised foodgrain does work in favour of the party in power, if it is effectively delivered, as in Chhattisgarh, but not if it is promised but not delivered, as in Delhi and Rajasthan.
The writer is a retired professor of economics