Government of India knows farm output is measured wrong.But government agencies bitterly disagree on correctives.Farmers and consumers are paying the cost of an endless debate
Singhari Ram Porte of Meu village in north Chhattisgarh is a patwari,a littleknown group of around 2 lakh spread across India.The patwaris are primarily land record officers who collect land revenue on the states behalf.But they have a more important job.After the onset of the agricultural year in July,they provide a key input to the economys managers the sowing patterns of farmers for the two seasons,rabi and kharif,and the expected harvest.The information is critical to the advanced estimates of harvest the government provides for the rabi and kharif crops in September,January,April and June.Each estimate is revised with more data,but Porte and Cos word is the main ground-level information that goes into the governments food management decisions on whether to export or import grain.For over 50 years,their art of reading Indias farm leaves has indirectly impacted the money the farmers produce fetches and the price consumers in villages and cities pay for grain.
Debate on Data
But lately,the patwaris job as information providers has come under scrutiny because farm data has increasingly become prone to errors.A recent review of crop data by the National Sample Survey Organisation found that less than half of Indias villages report cropping patterns and yield expectations on time.Among the villages that did,a quarter of the farms had wrong data for early kharif estimates,39% farms had the late kharif estimates wrong and the rabi season data was off the mark for 34% farms.
Planning Commission member Abhijit Sen says almost every year,the final estimates are around 10% higher than the first estimates.In 2010-11,the first and final estimates of rice harvest had a difference of nearly 15%.And the final estimates come out one and a half years after sowing,says Sen.Now patwaris find themselves at the heart of a heated debate among agriculture ministry mandarins and farm economists on a cure for shoddy data.As the government,farmers and consumers have learnt the hard way,wrong data can have disastrous results.
When the government underestimates output,it ends up banning exports and the resulting oversupply of those items leads to a fall in prices, says Ashok Gulati,chairman of Commission of Agricultural Costs and Prices,which advises the government on farm pricing.Gulati is convinced that inaccurate farm statistics distort market prices.
That is no laughing matter for a government which has admitted that inflation is a tax on the poor.With India bracing for an ambitious food security programme,economists say even a 5% variation can lead to dangerous policy choices.
For one such instance,rewind two years.The government allowed onion exports,enthused by rose-tinted output estimates.But the move boomeranged as the onion crop fell short of expectations and prices shot up.A red-faced UPA government was forced to prevent more onions from reaching ports.
It was the same story with the import of pulses,the favoured protein source for many Indians.Looking to curb the massive shortage of pulses over the last decade,the government spent nearly 800 crore in subsidies to import and stabilise domestic prices.Five years later,importing agencies were confronted with losses of 1,244 crore.They eventually sold videshi pulses to four large private firms instead of the poor.
The statistically significant scandal was criticised by the Comptroller and Auditor General.In a damning report tabled in Parliaments recent winter session,it pointed out that the main objectives of the endeavour availability of pulses and at stable prices remained largely unfulfilled.
The shoddy data work is not due to lack of resources or energy.
From the stage where patwaris like Porte provide primary inputs on expected crop patterns to the conference rooms in New Delhis Krishi Bhawan where production estimates are aggregated,four different government schemes and as many as 16 departments are involved.
The agriculture ministry holds a weekly crop weather watch meeting where inputs from as many as 11 departments are taken.That apart,the data sent by states is validated by independent agencies such as Institute of Economic Growth and even the Indian Space Research Organisation (Isro).
The Root Cause
So why then is farm data quality poor It is because the fundamental flaw in the system remains unattended.
The patwaris must traverse 20% (or 1.2 lakh) of Indian villages and record the quantum of wheat,rice,sugarcane or 24 other crops sown in each farm twice a year.They must also conduct 8 lakh experiments on crops to check productivity levels.
Thats no mean feat,but the patwaris have more on their plate.Porte says they are roped in for other jobs in the village at the ration shop,the anganwadi and government schemes like mid-day meals.It is difficult to record output as there isnt enough time to go to the 6-8 villages assigned to us, he admits.
But the system,followed by most states,has run unchanged in all its 50 years.
In some states,15 villages are assigned to each patwari, an agriculture ministry official told ET on Sunday,adding that it is not humanly possible for them to go to each village.Sen believes the problem is not in the system itself,but how it is used.District authorities,block development officers and states end up revising figures on their own if they dont like them.It isnt scientifically done.
In late December,Rahul Gandhi belatedly backed foreign direct investment in retail.His support was directed at the potato farmers of Farrukhabad in Uttar Pradesh.They preferred dumping their harvest because of a potato glut that hammered prices.Astute observers blame the glut on,what else,poor output estimates.Wrong farm data also tinkers with Indias growth numbers.If [farm] data is misreported,although it wont change the direction of growth rates as long as we are missing data year after year,the magnitudes would be affected, says Indias chief statistician TCA Anant.In 2010-11,Indias farm sector growth rate was revised upwards by 22% thanks to wide variations between initial and final estimates.
Fodder for Thought
Experts say its high time the government took better stock of the fields,as 70% of Indians subsist on farms and the other 30% are wary of persistently high food prices.
For their part,government bodies have long highlighted the ills of farm data.
In 2001,the National Statistics Commission headed by former RBI governor C Rangarajan had asked the government to improve the functioning of the patwari system.This was to be implemented in the 10th Five Year Plan.
Six years later,while working on the 11th Plan,an official group on the farm sector under economist VS Vyas reminded the government of its unfulfilled commitment to fix the patwari system.
But all the 11th Plan gave farm statistics was a scheme called Forecasting Agriculture output using Space,Agri-metereology and Land-based observations (FASAL).This too,officials say,has been a failure.
In February 2009,the government created another panel this time under noted agriculture economist A Vaidyanathan.His report submitted last February clearly shows that the system does not deliver complete,timely and reliable data.His panel had asked the government to create an independent pool of workers gathering farm data from just 15,000 villages and assimilating it through a new autonomous body called National Crop Statistics Centre (NCSC).
In a recent meeting,Planning Commissions Sen had asked the agriculture ministry for urgent action on the Vaidyanathan panels recommendation.But there is a certain reluctance,Sen said.
Science of Getting it Right
The ministry isnt sure if the massive reduction in sample size from 1.2 lakh villages to 15,000 would throw up good data for the 27 crops that they need to track.
Its all complicated math and the first stage in implementing the committees recommendations would be to find the appropriate sample size for estimating each crop as 15,000 is just a ballpark number, says agriculture secretary PK Basu.
While this issue has been referred to an agricultural statistics research institute that will take six months to suggest the right sample size,the ministry is keen to hire independent staff for 1.2 lakh villages to tide over the patwari problem.
This year,the government is due to create a National Crop Forecasting Centre (NCFC) where Isros technologies for using remote sensing and satellite images to estimate crop output will be transferred.The Vaidyanathan panel had suggested that these activities be subsumed under NCSC.
While nomenclature battles can be resolved over time,the art of using science for better farm statistics has a long way to go though.One key element of FASAL,which is to be extended into the 12th Five Year Plan,was sophisticated forecasts based on a crop weather model by the weatherman.A single such forecast is yet to come.
Isros Space Applications Centre was to develop the methodology for forecasting 11 crops by remote sensing techniques,but has managed to do so for only four.Moreover,the satellite images cant always tell the difference between elephant fodder and sugarcane.The government is now keen that Isro launch a satellite dedicated purely for agricultural applications with the capacity to deliver high-resolution images up to 1 sq m.
At Isros current pace,another 35 years may be needed before all 27 crops can be correctly identified through satellite images.There is a real risk that the satellite may well be past its expiry date by then.