| Below the line
Arun Shourie's new book, Falling over Backwards, is both engrossing and disturbing. It is a book against reservations and against judicial populism. Here is a quote from the book: 'The individual, and not the group should be the unit of State policy.' I don't want to focus on reservations right now, but on poverty and backwardness. Not even the most ardent of reformers will argue that there shouldn't be anti-poverty programmes. But every reformer will argue that these should be targeted, and only the deserving should benefit from subsidies. And the crucial point is that you can target individuals, you cannot and should not target groups or collective entities.
In the entire discourse on reforms, there is an argument that trickle-down benefits of growth haven't percolated through to backward and deprived segments. That's almost tautologically true, especially if we interpret benefits in a relative rather than an absolute sense. But consider the way we interpret deprivation. India Shining happened in urban India, India Whining happened in rural India, and that explains the outcome of the 2004 general elections. Ipso facto, everyone in rural India is backward and there must be Bharat Nirman and the National Rural Employment Guarantee Act as a new deal for rural India.
Alternatively, the backward states are Bihar, Madhya Pradesh, Rajasthan, Uttar Pradesh, with three of these states now divided and Orissa added to the BIMARU fold. We must do something for backward states, through a Rashtriya Sam Vikas Yojana. Or, out of India's 600-plus districts, 200 are backward and there must be a backward regions grant fund. Or, the entire unorganized sector (7 per cent of the labour force) is exploited and weeping, and there must be social security for everyone in the unorganized sector. Or, there must be positive affirmation, through reservations or otherwise, for all scheduled castes and scheduled tribes and the other backward classes.
The moment we interpret poverty and backwardness as a collective concept, including a geographical one, there are two kinds of errors we can commit, reminiscent of Type I and Type II errors in statistical hypothesis testing. By excluding poor individuals who don't belong to the collective, we commit an error of exclusion. By including rich individuals who belong to the collective, we commit an error of inclusion, the so-called 'creamy layer' idea. If one switches identification from collective groups to individuals, neither of these errors need be committed. Indeed, in every country that has a decent anti-poverty programme, individuals have special identity cards ' call them social security cards, if you will.
What causes poverty or backwardness' Opinions may differ. But figuring prominently in the list of variables will be lack of physical infrastructure (roads, electricity), inadequate access to social infrastructure (health, education), inadequate access to land and credit, and so on. We tend to think that these are public goods (and services) and must be provided by the state. Economists have a definition of a public good, a public good being one where my consumption does not reduce your consumption (perhaps clean air) and you can't prevent me from consuming the good (perhaps law and order). Interpreted thus, none of the items listed are public goods. They are all private goods. They may be merit goods in the sense of requiring subsidies by the state, but they are not public goods. Nor does state subsidization necessarily imply state provisioning.
Having said this, whether it is state provisioning or state subsidization, in how many of the variables just listed are we talking about collective private goods' With the exception of physical infrastructure, where there is a collective or geographical entity that is backward and deprived, none. The deprivation is individual-based, not community-based. Yet, we stick to the community-based identification, because it is easier to handle. Not only easier to implement, but also easier to handle politically because community-based vested interests can be pretty powerful. It is not very surprising that despite having done several things in the name of the poor, the lot of the truly poor hasn't changed. The benefits have been reaped by the so-called 'creamy layer', not only in reservations, but elsewhere also.
Below-the-poverty-line identification is a case in point. How do we decide what percentage of the Indian population is BPL' For statistical purposes, the poverty line figure is based on National Sample Survey data and we now have a figure of 22 per cent in 2004-05. But that's neither here, nor there. Not only is this based on a survey rather than complete enumeration, NSS data don't enable us to specifically identify BPL households. To do that, one needs specific household-level indicators, like the 13 new parameters the Planning Commission has been talking about. Such parameters also enable us to broad-base the notion of poverty, away from income (or expenditure) poverty.
One can consider ownership of specific assets like pucca houses or land, or consumer durables or access to piped water or road connectivity. One can throw in educational levels of the family (or drop-out rates) and variables like old age and women-headed households and number of children per household. Female work-participation rates, consumption loans taken and dependence on manual labour are other possibilities ' perhaps even a rural/urban classification, and if we want to, even caste.
There will never be much debate on determining what variables should be included in the identification of poverty. However, once these variables are identified, there has to be a process of aggregation, and this requires assignation of weights to these variables. In the last resort, this is subjective and since index values are often critically dependent on weights, there is much more scope for debate here. But it should be possible to reduce the subjectivity and test for robustness of index values regarding choice of both variables and weights.
In any event, just because something is difficult, one ought not to shun the attempt, which is what policy-making invariably tries to do. A case in point is the recent report on social security for unorganized sectors submitted by the national commission for enterprises in the unorganized sector. It has differential packages for BPL and above-the-poverty-line households, but makes no attempt to even begin defining BPL.
This is not a problem that is typical to Indian policy-making. For example, poverty lines have also evolved in the United States of America. More accurately, there is a difference in the US between a poverty threshold and poverty guidelines. The poverty threshold is a bit like what we call the poverty line, but has now evolved to capture family size, the sex of the head of the household, the number of children who are less than 18 years of age and farm or non-farm residence. This is what the Census Bureau uses to statistically state what percentage of the US population is below the poverty line (12.7 in 2004).
This is the counterpart of what we obtain through NSS data, courtesy the Planning Commission, but before refinements in the US poverty threshold took place to reflect adjustment variables like sex of the head of the household. But in parallel, the department of health and human services issues poverty guidelines to determine whether a household should be eligible for assorted social security benefits and these are based on the refined poverty threshold. To state it differently, we are still on the unrefined poverty threshold. For anti-poverty programmes, we need to transit to the individual-based refined poverty threshold or poverty guideline. If we can successfully do this, most of the pointless debate will disappear, including that on reservations.