Does crop output depend upon rain alone?

DEBATE

Surjit S Bhalla,
Managing Director
Oxus Investments Pvt Ltd

The share of agriculture in India's GDP has come down from over 50 per cent in the 1950s, to less than 20 per cent today. Yet, what is striking is that the rate of growth of agricultural output has stayed constant at about 2.7 per cent a year.

Rainfall data has been collected for over a 130 years on 306 sub-stations in India. The agricultural months for which rainfall is most important are June to September; the four-month deviations of each rainfall centre from its 130 year average has been used to form a composite rainfall index.

This index is used to test for the dependence of Indian agriculture on the volume and volatility of rainfall. If the dependence of agriculture on rainfall has declined, then one would find that variations in rainfall would explain less and less variation in agricultural output. This is not the case.

For the period 1950-76 (the halfway point), rainfall alone (actually rainfall and rainfall lagged one year) explained as much as 48 percent of the variation in output; for 1977-2003, the variation explained increases marginally to 53 per cent.

For these two periods, agricultural growth accelerated from 2.4 to 3.1 per cent a year. However, the average rate of growth is very sensitive to the exclusion/inclusion of particular set of years.

The fact of equal, or higher, dependence of agriculture on rainfall does not mean that agriculture has not progressed. It has done so by leaps and bounds. The number of cultivators has stayed roughly constant at about 90 million for the past decade or so. The land area under cultivation has stayed the same, as has the number of agricultural labourers.

Given that output growth is marginally higher, this means that labour and land productivity has increased significantly. However, the fact remains that output growth in agriculture (perhaps reflecting a stylised fact) has stayed constant at around 2.5-3 per cent per annum.

The rainfall data can also help to illuminate some of the political party generated controversies. The Congress contends that it is the saviour of the common man, the party that has taken Indian agriculture to new heights.

During the five-year Congress period (rainfall years June-September 1991 to June-September 1995), agricultural output growth averaged only 2.3 per cent a year, with average rainfall slightly below normal.

During the six-year NDA period, (rainfall years June-September 1998 to June-September 2003), average agricultural growth was a higher 2.8 per cent, average rainfall being the same. But if just one year is added to the Congress period (1996-1997), then average output growth for the six year period (1991 to 1996) jumps to 3.4 percent per annum "" one of the best six year averages in modern Indian history. So, don't derive any lessons about Indian agriculture without first looking at the rainfall pattern, and Indian agriculture's dependence on it.

Surinder Sud,
Agriculture Editor,
Business Standard

Crop production does not rely on rainfall alone; it is influenced by several factors. Rainfall is just one of them. And in this case, too, it will be far-fetched to see a direct, one-to-one correlation between the amount of rainfall and crop production.

This is borne out by the past 20-year data on the country's cumulative rainfall in the four-month monsoon season and the total foodgrain production.

The quantum of rainfall in each of the two major drought years of 1987-88 and 2002-03 was the same "" 81 per cent of the normal. But the foodgrain production was 140.35 million tonnes in 1987-88 and 174.77 million tonnes in 2002-03, a difference of 34.42 million tonnes or 24.5 per cent.

It can, of course, be argued that the two years are quite some time apart and a good deal of progress must have been made during this time gap. But even if we take two consecutive years, 2000-01 and 2001-02, the picture that emerges is similar.

In both these years, too, the total rainfall was the same "" 92 per cent of normal. But the foodgrain output was merely 196.81 million tonnes in 2000-01, while it was as high as 212.85 million tonnes the very next year with the same level of rainfall.

Again, the rainfall in 1993-94 as well as 1995-96 was 100 per cent normal. But the foodgrain output dropped from 184.26 million tonnes in 1993-94 to 180.42 million tonnes two years later in 1995-96 with the same amount of rainfall.

Also, 1994 saw the maximum rainfall in past two decades, 110 per cent of normal, but the grain production was 191.5 million tonnes. But two years later, in 1996, the country bagged a far higher harvest of 199.44 million tonnes with a comparatively lower rainfall of 103 per cent of normal.

There, indeed, is no dearth of such examples to demolish the notion of a direct link between rainfall and crop production. Actually, the distribution of rainfall and other climatic factors accompanying it matter much more than its quantum alone.

Even deficient monsoon rainfall, if distributed evenly over the season and not accompanied by other adverse climatic conditions, can lead to a better harvest than above-normal but unevenly distributed rainfall.

Besides, the temperature regime throughout the crop growing season also makes a perceptible difference to the final crop yields. The wheat crop, for instance, was the victim of prolonged cold and foggy spell in 2002-03, and sudden rise in temperature in March in 2004-05.

The other critical factors that influence crop output include use of inputs, including seed, irrigation, fertilisers, pesticides and credit; incidence of diseases and pests; and technology. This apart, factors like timeliness and precision of farm operations and overall management, too, impact the crop yields.