A degree of targeting is useful in ensuring that policies are effective in reducing poverty. But we have to be careful how this is done.
With the right policies,India has a good chance of seeing accelerated poverty reduction in the coming decades. As I have previously argued,this will require that India does a better job in reaching the countrys many poor people through its social policies. However,that does not necessarily mean that those policies need to be better targeted in the usual sense of concentrating benefits on the poor by avoiding leakage to the non-poor. One of the most common mistakes in social policymaking across the globe is to confuse better targeting with greater impact on poverty. They are not the same thing. Targeting can help,but it is not the objective.
Targeted policies typically impose costs on poor people,and policymakers do not always take proper account of those costs. One such cost is the loss of benefits,in kind or cash,when incomes rise from other sources,which determine the incentives faced by people in trying to do better economically. Think about what happens when we take targeting to the limit. This would entail a perfectly targeted set of transfers to poor families meaning that the transfers bring everyone up to the desired minimum income. This would impose a 100 per cent marginal tax on recipients. There would probably be little or no point in taking on extra work to earn an extra Rs 500 (say) if your transfer receipts fall by Rs 500. Such high costs of targeting will undermine the incentives for poor people to escape poverty by their own means. This is unlikely to be the best we can do from the point of view of poverty reduction.
Consider,instead,the other extreme of no targeting universal basic schooling or healthcare or,when talking about transfers in cash or kind,a basic income scheme in which everyone gets the same amount,whether poor or not. Uniform provision entails zero marginal cost in taking on extra work,or some other income source. The incentives for promotion from poverty
Most social policies fall between these extremes in practice. Starting from a basic income scheme,it is likely that some degree of targeting will enhance the impact on poverty that can be obtained with given public resources. But we have to be careful about how this is done. Some of the ways we can achieve better targeting entail costs that are not always obvious,but matter greatly to poor people.
Consider,for example,the Mahatma Gandhi National Rural Employment Guarantee Scheme (MGNREGS). This is probably the most ambitious direct intervention against poverty that any developing country has ever tried. The scheme promises 100 days of work per year to all rural households whose adults are willing to do unskilled manual labour at the statutory minimum wage notified for the programme. Work is to be made available to anyone who asks for it within 15 days of completing an application to work,failing which the state government is liable to pay an unemployment allowance. Over 50 million households participate.
It is likely that a scheme such as MGNREGS will be better targeted to poor people than a basic income scheme. But that does not mean that MGNREGS will have greater impact on poverty. That depends on many other factors,including both benefits and costs to poor people from the scheme. MGNREGS can have huge benefits if it works well. It can provide much needed social insurance,as long as people can get help from the scheme when they need it. Alas,the extensive rationing found on MGNREGS (especially in some states) is not encouraging in this respect. Creating assets that benefit poor people can also be a big plus,though there have been concerns as to whether MGREGS gives sufficient attention to such asset creation.
Workfare schemes such as MGNREGS illustrate well the point that even a well targeted scheme can be dominated by untargeted transfers when one takes account of all the costs involved,including the other income forgone by participants. There has been evidence of non-negligible dead-weight losses to participants in such schemes in India,so much so that a basic income scheme would have been more cost-effective in simply transferring money to poor people. (Puja Dutta,Rinku Murgai,Dominique Van De Walle and myself show this in our forthcoming book,Rozgar Guarantee?) Of course,MGNREGS tries to do more than simply transfer money to poor people. But how we assess the programme depends a lot on those other benefits,and it would be fair to say that they are not as yet fully evident in MGNREGS. The scheme needs reform to achieve its potential.
Some degree of targeting will no doubt remain useful in assuring that social policies are effective in reducing poverty. But more universal provisioning in education,health and social protection deserves a closer look in India. The cost to the budget,and how it is financed,is of course hugely important. But there appears to be scope for financing a decent level of basic income in India by cutting other spending,especially subsidies favouring the non-poor,as Pranab Bardhan and others have argued. The idea has spanned both rich and poor countries. The negative income tax advocated by Milton Friedman and others can be thought of as a basic income scheme financed by a progressive income tax. This type of scheme would dominate many policies found in practice today. It would clearly yield a better incidence than subsidies on the consumption of goods that the rich consume more than the poor. And a basic income scheme could have broad political support,which would help assure its sustainability. In terms of its impact on poverty,it may also dominate seemingly well targeted schemes,once all the costs to poor people are considered.
The writer,a former director of the World Banks research department,holds the Edmond D. Villani Chair of Economics at Georgetown University,US