Using the Rajan index to fix allocations will distort states incentives
During the course of his maiden speech as RBI governor,Raghuram Rajan mentioned that he was not seeking Facebook likes. He actually received quite a few. The RBIs mid-term policy review on September 20 must have given his admirers much heartburn and possibly earned him some dislikes. The spotlight has turned on him again,with the release of the Rajan committee report on evolving a composite development index of states. If he has a Facebook page,it must be full of either angry or sarcastic comments. Only a handful of people like Nitish Kumar,whose demand for the backward status tag for Bihar is now going to be conceded,or staunch opponents of Narendra Modi,who are delighted that the committee has declared that Gujarat is not highly developed,are happy with the report. All others find a plethora of defects in it.
The general sense of dissatisfaction felt by a majority about the Rajan committees index of backwardness is not surprising at all. If ten people are asked to define a backward state,they are very likely to give ten quite different definitions. This is because it is impossible to be very precise about what we mean by backwardness. Any attempt to make an imprecise or inexact concept too precise must inevitably invite criticism,because there cannot be any unique way of doing so. One can even argue that exercises of this kind are actually quite harmful when they have important policy implications. Unfortunately,this committees report may well play a significant role in the future,by influencing Centre-state resource transfers.
The first attempt to define some notion of backwardness in India was pretty straightforward. Three states (Assam,Jammu and Kashmir and Nagaland) were put in a special category under the Gadgil formula in 1969. The list soon expanded to contain 11 states. A common feature of these states was that they were all handicapped by adverse geographical characteristics such as hilly terrain or locations far away from large population centres. These features increased the cost of delivering public services. Moreover,these states were less than ideal locations for entrepreneurs,both because of poor infrastructural facilities as well as large distances from major sources of demand.
Clearly,these states were handicapped in so far as industrialisation was concerned. Importantly,the underlying reason for their handicap was beyond their control. It made sense for the Centre to compensate them for this handicap. Ninety per cent of Central assistance to these states is treated as a grant,with the remaining 10 per cent constituting loans that need to be repaid. In contrast,as much as 70 per cent of Central assistance to other states is treated as a loan. In addition,special category states also get significant excise duty concessions,so as to enable them to entice entrepreneurs to set up shop.
Large gains are involved,so its not surprising that Nitish Kumar demanded that Bihar also be included in the list of states requiring special assistance. Bihar does not suffer from the handicaps associated with the current set of special category states. But,for obvious political reasons,the UPA government has to humour Nitish and so the Rajan committee was set up to evolve a composite index of backwardness.
How best could the committee have performed this task? Is a state backward because it falls far short of accepted levels of income or social indicators such as health or education? What about other indicators such as the level of infrastructure,or a measure of the exogenous constraints imposed by unfavourable location or topography?
More generally,should the components of an index of backwardness depend on the inputs going into a development process,or should they be outputs such as per capita income and the like? The answer must depend on how the index is going to be used. If the index is to be used in order to compensate the more backward states,then output variables should not be included as components of the index because it would have very bad incentive properties. States would have every incentive to remain backward by taking the easy way out impose low taxes and give handouts whenever possible to win cheap political support,since the Central dole will be relatively high. From this perspective,the earlier criteria involved in the selection of special category states was much better,since these were not characteristics that the state could influence.
The thought process underlying the choice of components by the Rajan committee is quite opaque. The committee honed in on 10 variables,such as per capita household expenditure,poverty ratio,female literacy,education,health,financial inclusion,a connectivity index,etc. This is a curious mix of both inputs and outputs,defying any logical explanation for why these and only these should make up an index of backwardness. Moreover,several of the variables are highly correlated. The most glaring example of this is the pair of poverty ratio and per capita household expenditure. Surely,the latter plays an important role in determining the former. So,the inclusion of both involves a large measure of double counting.
Several commentators have also criticised the construction of the index,because the composite index gives equal weight to each component. This is perhaps the least objectionable part of the construction. Clearly,there is no compelling reason to decide that one component is more important than another,and so equal weighting is the best option. A potentially more problematic issue is whether the variables have been normalised properly. For instance,female literacy is a percentage and so has to range between zero and a hundred. On the other hand,per capita incomes are several hundreds of rupees. Unless the latter has been scaled down,the index would be heavily influenced by figures of per capita incomes.
Perhaps the committee,composed of several distinguished academics,has made this rather obvious adjustment. Unfortunately,the overall task has been performed so shoddily that one cannot assume that this has been done.
The writer is professor,Department of Economics,University of Warwick,UK