Drought relief: cash transfers better than rural works

I became a journalist in 1965, when two successive droughts killed thousands and forced India to beg for US food aid. India’s share of global food aid was so large then that a best-selling book claimed that India was unviable and should be left to starve, conserving food aid for viable countries.

How distant those dark days seem! This year another drought has struck, but we have no panic or talk of food aid. India withstood a terrible drought in 2002, so Indians are confident that the government knows how to tackle droughts. That’s a revolutionary change from the 1960s.

Most people think that the green revolution ended mass starvation. Not so. The green revolution improved yields, and made India self-sufficient. Yet it did not raise yields sufficiently to increase foodgrain availability per head. Remarkably, grain consumption per head in later years rarely reached the 1964 level.

Why, then, did droughts cease to cause mass starvation? Because of better food distribution, not production. In subsequent droughts, enough food got through to the worst hit.

Rural employment was the key to success. Maharashtra first experimented with an Employment Guarantee Scheme when hit by successive droughts in the early 1970s. Simultaneously, countries of the Sahel region in Africa suffered repeated drought. Food aid was rushed to the Sahel, while very little came to India or Maharashtra. Yet there was mass starvation in the Sahel, and none in Maharashtra.

Economists Amartya Sen and Jean Dreze explained the paradox. In Maharashtra, grain availability per head was just half that of the Sahel. But employment schemes enabled hungry Maharashtrians to earn just enough to stave off death. Maharashtra suffered mass hunger, but not mass starvation.

The opposite was true in the Sahel. Food aid piled up in ports and godowns, but there was no mechanism to get it to those starving. NGOs, government agencies and free kitchens did not have the necessary reach.

In Maharashtra, the market provided the reach. Once the needy obtained purchasing power through employment schemes, the market drove grain to where the money was. This reached the needy more efficiently and completely than free government kitchens.

Other states did not follow Maharashtra’s lead in enacting employment guarantee legislation. But when a drought occurred, they would rush to create emergency employment schemes. Gujarat, for instance, created millions of man-days of work to successfully combat the 1987 drought.

This approach aimed to relieve distress, not improve economic efficiency. Yet economic efficiency improved too. Research across the world shows that, other things equal, small farmers are more efficient than large ones, mainly because they maximize the use of family labour.

Markets usually drive assets to those who use them most efficiently. When industries suffer a recession, inefficient industries fail and are acquired by efficient ones. But this does not happen in rural India in a drought: efficient small farmers do not acquire the land of inefficient absentee landlords. On the contrary, many small farmers make distress sales of their land and cattle to large farmers.

Why such perverse results? Because rural credit and insurance markets are missing or incomplete. Big farmers dominate access to credit, and enjoy the lion’s share of huge rural subsidies for electricity, water, fertilizers and credit. In these conditions, small-farm efficiency does not translate into superior yields or income.

Rural employment schemes plug the financial gaps of small farmers and thwart distress sales. This not only relieves suffering but improves economic efficiency: small landholders are efficient.

How should rural employment be tailored to meet needs in the current drought? First, the financial allocation must be raised: the budget has already done this. Second, the spending share of the states should be reduced in the worst-hit areas, enabling them to tap more central money. In 2008, Bihar provided just 23 days work per applicant, Chhattisgarh 30 days, Jharkhand 44 days and Uttar Pradesh 26 days. These states are badly hit by drought this year, and must expand person-days of work massively.

Third, payments to workers must be prompt. Many states routinely delay payments, hitting the neediest. Dreze rightly calls this deplorable.

A radical solution would be for states to stop rural works and simply pay the legal compensation to job-card holders. This will cost less than rural works, since money will not be spent on materials or supervision: it will go entirely to the needy. In effect, cash transfers will replace rural works. Activists like Dreze will demur: they think rural works create durable assets, despite evidence to the contrary. I believe that in a major drought, we should focus on getting cash to the needy, not on building mud roads that vanish in the next monsoon.

What do you think?