By Jean Dreze
Posted May 11th 2011
"Conditional cash transfers" (CCTs) are a new buzzword in policy circles. The idea is simple: give poor people cash conditional on good behaviour such as sending children to school. This helps to score two goals in one shot: poor people get some income support, and at the same time, they take steps to lift themselves out of poverty.
CCT enthusiasm, however, is often based on a superficial reading of the Latin American experience. In Brazil, Mexico and other pioneers of this approach, CCTs were used to bring into the fold of health and education services a fringe of marginalised households, in a situation where a large majority of the population was already covered by extensive social insurance systems. CCT is basically an incentive and, predictably enough, it often works: if you pay people to do something that benefits them anyway, they tend to do it. It is the same principle as scholarships for disadvantaged children. Incidentally, there is no evidence that scholarships — that is, conditional cash transfers — work better than “conditional kind transfers” like school meals or free bicycles for girls who complete Class 8. In fact, I submit that the latter would win hands down in any sensible and sensitive evaluation of the two approaches. Be that as it may, I am not questioning the potential effectiveness of CCTs in their limited capacity of “incentive”.
CCT enthusiasm, however, is often based on a superficial reading of the Latin American experience. In Brazil, Mexico and other pioneers of this approach, CCTs were used to bring into the fold of health and education services a fringe of marginalised households, in a situation where a large majority of the population was already covered by extensive social insurance systems. CCT is basically an incentive and, predictably enough, it often works: if you pay people to do something that benefits them anyway, they tend to do it. It is the same principle as scholarships for disadvantaged children. Incidentally, there is no evidence that scholarships — that is, conditional cash transfers — work better than “conditional kind transfers” like school meals or free bicycles for girls who complete Class 8. In fact, I submit that the latter would win hands down in any sensible and sensitive evaluation of the two approaches. Be that as it may, I am not questioning the potential effectiveness of CCTs in their limited capacity of “incentive”.
What is remarkably dangerous, however, is the illusion that CCTs can replace public services by enabling recipients to buy health and education services from private providers. This is not how CCTs work in, say, Brazil or Mexico. In Latin America, CCTs are usually seen as a complement, not a substitute, for public provision of health, education and other basic services. The incentives work because the services are there in the first place. In India, these basic services are still missing to a large extent, and CCTs are no substitute.
Consider, for instance, healthcare. In Brazil, basic health services such as immunisation, antenatal care, and skilled attendance at birth are virtually universal. The state has done its homework — almost half of all health expenditure in Brazil is public expenditure, compared with barely one quarter (of a much lower total) in India. In this situation, providing incentives to complete the universalisation of healthcare seems quite sensible. In India, however, public health services are virtually non-existent, and it would be very unwise to think that CCT-type programmes like the Rashtriya Swasthya Bima Yojana (RSBY) can fill the gap.
Another contextual difference, mentioned earlier, is that Latin American countries tend to have highly developed social insurance systems, with wide coverage. “Targeting” CCTs to marginalised groups in such a situation makes some sense, because the bulk of the population is already covered and the rest is (relatively) easy to identify. In India, however, large sections of the population are in dire need of social support, and the experience with targeting is quite sobering. Indeed, every known method of identifying “BPL” (below poverty line) households involves large exclusion errors. This is an unresolved issue for any targeted CCT initiative in India.
Another contextual difference, mentioned earlier, is that Latin American countries tend to have highly developed social insurance systems, with wide coverage. “Targeting” CCTs to marginalised groups in such a situation makes some sense, because the bulk of the population is already covered and the rest is (relatively) easy to identify. In India, however, large sections of the population are in dire need of social support, and the experience with targeting is quite sobering. Indeed, every known method of identifying “BPL” (below poverty line) households involves large exclusion errors. This is an unresolved issue for any targeted CCT initiative in India.
In short, a nuanced approach is required to the design of social security transfers. CCTs are useful in some circumstances: scholarships are one example. In other situations, like pensions for widows and the elderly, there is a case for unconditional cash transfers. Conditional transfers in kind, such as midday meals in primary schools, also have a role.
Finally, there is a place for unconditional transfers in kind, such as the Public Distribution System (PDS).
A wholesale transition from the PDS to cash transfers in rural India would, in my view, be misguided and at the very least premature. For poor people, food entitlements have several advantages over cash transfers. First, they are inflation-proof, unlike cash transfers that can be eroded by local price increases, even if they are indexed to the general price level. Second, food tends to be consumed more wisely and sparingly; cash, on the other hand, can easily be misused. Third, food is shared equitably within the family, while cash can easily be cornered by selfish individuals. Fourth, the PDS network has a much wider reach than the banking system. In remote areas, where the need for social assistance is the greatest, banking facilities are simply not ready for a system of cash transfers (as it is, they are unable to cope with NREGA wage payments). Last but not least, cash transfers are likely to bring in their trail predatory commercial interests and exploitative elements, eager to sell alcohol, branded products, fake insurance policies or other items that would contribute very little to people’s nutrition or well-being.
Finally, there is a place for unconditional transfers in kind, such as the Public Distribution System (PDS).
A wholesale transition from the PDS to cash transfers in rural India would, in my view, be misguided and at the very least premature. For poor people, food entitlements have several advantages over cash transfers. First, they are inflation-proof, unlike cash transfers that can be eroded by local price increases, even if they are indexed to the general price level. Second, food tends to be consumed more wisely and sparingly; cash, on the other hand, can easily be misused. Third, food is shared equitably within the family, while cash can easily be cornered by selfish individuals. Fourth, the PDS network has a much wider reach than the banking system. In remote areas, where the need for social assistance is the greatest, banking facilities are simply not ready for a system of cash transfers (as it is, they are unable to cope with NREGA wage payments). Last but not least, cash transfers are likely to bring in their trail predatory commercial interests and exploitative elements, eager to sell alcohol, branded products, fake insurance policies or other items that would contribute very little to people’s nutrition or well-being.
Of course, cash transfers have their advantages too: they have lower transaction costs, are (potentially) more convenient for migrant labourers, and may be easier to monitor. Sometime in the future, when the banking system has a wider reach and the food security problem has been resolved, a cautious transition to cash transfers may be advisable. Indeed, I am not averse to the idea of a “universal basic income”. But this is a somewhat futuristic idea, and for the time being, food is best.
The most common argument for cash transfers is that cash makes it possible to satisfy a variety of needs (not just food), and that people are best judges of their own priorities. Fair enough. But if people are best judges of their own interest, why not ask them whether they prefer food or cash? In my limited experience, poor people tend to prefer food, with a gradual shift from food-preference to cash-preference among better-off households. Further, poor people tend to give very convincing reasons for preferring food. I am more inclined to listen to them than to the learned champions of cash transfers.
The writer is an honorary professor at the Delhi School of Economics and a member of the National Advisory Council
The writer is an honorary professor at the Delhi School of Economics and a member of the National Advisory Council
