Bibek Debroy: The backward districts paradox

There is a paragraph in the Budget speech that should have intrigued us. To quote, "An Inter-Ministerial Group (IMG) has identified 170 backward districts on certain socio-economic variables.... I propose to accept the recommendations of the IMG, and I am happy to announce the establishment of a Backward Regions Grant Fund (BRGF). An allocation of Rs 5,000 crore has been made in the Plan for 2005-06, and an equal amount will be allocated every year in the next four years. Consequent upon the establishment of the Fund, the existing Rashtriya Sam Vikas Yojana (RSVY), envisaged to end in 2006-07, will be wound up with suitable transition arrangements that will protect every district now covered under RSVY."

Regional disparities and the problem of backward districts have occupied us for a very long time, going back to the 1960s. Had earlier interventions worked, the problem of backward districts should have disappeared.

Indeed, there is a moot point about whether we should target poor people or backward regions. Relatively prosperous people also inhabit backward districts. When we generalise and argue that a geographical region as a whole is disadvantaged, there is always the probability of not-so-deserving people becoming recipients of subsidies and subsidies to backward districts predominantly accruing to better-off segments.

And there is also the probability of poor people in not-so-backward districts being ignored. Had we been able to identify and target the poor properly, we should avoid any generalisations on backward districts. It is because the poor can't be identified easily, that we use the short-cut of identifying backward districts and assuming that most people in backward districts are poor.

However, any such identification is a function of criteria used. And what should intrigue us is that the number of backward districts, as defined by IMG, has increased to 170. Whereas, if you take a vote among knowledgeable people, most people will suggest that the number of backward districts is 150.

What is going on? In recent times, the identification of backward districts is partly associated with RSVY, mentioned in the Budget speech. This was a Tenth Plan (2002-07) programme and under RSVY, the Planning Commission identified 100 backward districts. But Delhi, Goa, Bihar and Orissa weren't included in this identification. Delhi and Goa were left out because they didn't have any backward districts. Bihar and Orissa were left out because there were special components for these states under RSVY.

Identification of backwardness was based on three criteria: value of output per agricultural worker, agricultural wage rate and percentage of SC/ST population. Thirty-two more districts were added because they were affected by "Left Wing Extremism".

As far as I can make out, and I may be wrong, after including Bihar and Orissa, RSVY covered 147 districts. Add 32 districts covered by extremism and we have 179 backward districts.

To complicate matters, we then had the National Food for Work Programme (NFFWP) under the United Progressive Alliance government and in consultation with the Planning Commission, the rural development ministry identified 150 backward districts for this, which explains why most people think the number of backward districts is 150.

This identification doesn't seem to have been as transparent as the one under RSVY. The original three criteria used under RSVY still remained, but the poverty ratio (head count ratio) seems to have been included, giving us 212 backward districts.

But because the number of districts was too high, some subjectivity seems to have then been introduced. Exclude Goa. Exclude union territories. Allow states some subjectivity in adding more backward districts identified on the basis of non-transparent criteria and then scale down the figure to 150. And finally, we have the IMG, which has identified 170 backward districts on the basis of "socio-economic variables".

What these variables are, we don't quite know. The Planning Commission's website mentions the IMG, but doesn't give us the report, or the criteria used. If you ask around, you will be told that the three original criteria used in RSVY remain and that extremism has also been factored in, which RSVY also eventually used. And then you will be told that political considerations have also been used by IMG. Such that each state should have at least a certain number of backward districts and that special category states should have more.

This is all very unsatisfactory. First, why can't a common set of indicators be used, delinked from political considerations, for identification of backward districts? Why do we need to jump around from a RSVY identification to a NFFWP identification and then to a BRGF identification?

Whatever criteria are used, there will be consensus that there is a core set of around 100 backward districts, uniform across all identifications. Second, why should the criteria not be input-type variables such as access to physical infrastructure (roads, drinking water, sanitation) and social infrastructure (education, health)?

Indicators like agricultural productivity and agricultural wage rates are outcomes. Percentage of ST population is an indicator of backwardness, but certainly not percentage of SC population. Third, the problem is not identification, but improvements in efficiency of government expenditure. Why didn't RSVY work? If we don't answer that question, we won't know why BRGF should work.