How reliable are India’s agriculture statistics?
New Delhi: The Narendra Modi-led government at the Centre has set a target of doubling farm incomes by 2022.
A severe handicap is not what steps to take to achieve the target but knowing what current levels of farm incomes are.
India does not collect data on farm incomes in regular intervals, except for one off surveys by the National Sample Survey Organisation (NSSO).
The last such survey was published in December 2014 using data from a year before.
This means, the policymakers have to use proxy indicators like production estimates and prices to arrive at average incomes for the base year. But then, how good are these estimates?
Over the past year, the robustness of India’s crop production statistics, the most basic indicator of the farm sector performance has increasingly looked doubtful. It began with the government projecting a robust production of winter crops like chana (gram) and wheat for the 2015-16 crop year.
Later, tightening market supplies and rising prices showed these estimates were way off the mark.
For wheat the government estimated a 93.5 million tonnes output for the 2016 spring harvest despite lower sowing and severe moisture stress following consecutive years of drought.
Private traders estimated at least 5 million tonnes lower crop. They were proved right when wheat prices started rising and the government was forced to abolish the 10% import duty (in December) to keep a lid on prices.
Similarly, the government estimated sugar production during the 2016-17 season (October to September) at 22.5 million tonnes last week, higher than some market estimates of 20 million tonnes and the 21.3 million tonnes estimated by the trade body Indian Sugar Mills Association, or ISMA.
The lowest sugar production in seven years may force the government to lower the current 40% import duty, if its output estimates again fall short and see a downward revision in the coming months.
The situation was similar for chana. A market intelligence report from research division of the NCDEX commodity exchange (released in January) pointed out that while the government estimated a 7.2 million tonnes crop, market estimates were a modest 6.5 million tonnes—an overestimation of nearly 10%. Reliable production estimates from the government will help in better planning and check volatility in prices, the report said.
Despite the optimistic estimates, agriculture growth was a dismal 1.2% in 2015-16, but better than the 0.2% contraction in 2014-15. Both were drought years.
In 2016-17, following a normal monsoon, the first advance estimates of national income released earlier in January projected a 4.1% growth for the farm sector—a rebound based on 9% higher production of Kharif foodgrains and optimistic winter crop production targets.
Interestingly, these estimates do not take into account the churnings in the horticulture sector, post demonetisation of high value currency in November.
Due to the cash crunch and higher market arrivals, wholesale prices of vegetables crashed by 24% in November and an even steeper 33% in December. The horticulture sector contributes nearly 30% to the value of India’s crop sector (which in turn contributes 60% to the agriculture GDP, rest coming from forestry and fisheries) but the first advance estimates of agriculture growth did not account for what is happening to nearly a fifth of India’s farm economy.
The reason? There is no data yet on either area or production estimates for the horticulture sector in the 2016-17 crop year. This is despite horticulture production surpassing that of foodgrains since 2013, implying its growing importance in India’s farm economy.
To add to all of this, the estimates of weekly progress of sowing during the kharif and rabi crop seasons, even now, are based on ‘eye estimates’ reported by states (and not based on actual field surveys or satellite images).
In 2001, the National Statistical Commission had asked the government to improve the quality of India’s agriculture data such as sowing and horticulture statistics but there is almost no progress, said Himanshu, associate professor at Jawaharlal Nehru University, Delhi.
“The danger of unreliable data is that one does not know what to plan for consumers and farmers. In today’s day information is key to better policies but the government have not invested enough to gather robust data,” he said.
Therefore, the Union budget could do well by allocating more funds to gather reliable data on the agriculture sector.
This will not only help in better policy decisions such as aggressive government procurement when there is a glut or prompt trade policy decisions to keep food inflation at bay but also help achieve long-term goals such as doubling of farm incomes.
However, a reading of the 2016 union budget show that despite the glaring lapses in data, budgetary allocation for the department of economics and statistics (under the agriculture ministry) was slashed by nearly 30% year on year—from Rs327 crore in 2014-15 and Rs309 crore in 2015-16 (revised estimates) to a budgeted Rs236 crore in 2016-17.