The devil is in the data

Updated: Dec 11 2013, 08:47am hrs
As the agriculture and food economy move on to the fast-track information and statistics get the spotlight both for the needs of public policy and private investment and market functioning, the concepts like small samples, on farm physical investment, agricultural sectoral terms of trade, arrivals and surpluses, quality differentials get to the fore. One would have thought that agriculture would mean simplicity in statistics. Not so. When a sector is a part of a larger entity, life gets more complicated, since the interaction between the two becomes a further issue. The National Statistical Commission Committee on Agriculture that I chaired, therefore, begins with terms of trade of agriculture with the rest and inter-sectoral surpluses starting with the work of KN Raj, then me, Sudipto Mundle, Ashoka Mody ending with the recent estimates of BB Bhattacharya. When you get out of the comfortable world of rice and wheat and get into pulses, vegetables, fruits, spices, milk, eggs, chicken, fish and meat, you get into the world of small samples. India has a long tradition of small sample theoretical and estimation work, and the Indian Agricultural Statistics Research Institute and its pioneer Professor Panse cannot be ignored and will be if it is agreed to spearhead the creation of an information base here.

A lot of work will be needed since these are the crops that are at the root of what is called food inflation, which apart from policy failures is the only thing that will hold back India from a 7%-plus annual growth rate.

To fight food inflation you will not only need small samples on output but also demand levels and changes of individual crops, with tree crops and in vegetables, fruits and forest products and non-crop based commodities like milk etc in annual time. That is, in fact, more complicated and at present we measure demand only in NSS quinquennium years and interpolate in intervening years. Interpolated data are no use to understand fluctuations and develop policies. And so demand data will have to be presented annually and prices tracked as it were on a real-time basis. The committee has made detailed recommendations on this complicated issue and suggested annual surveys.

There is a lot of confusion on agricultural markets. The committee has defined agricultural markets and suggested that if such market purchase sale behaviour is seen then every year data on such transactional outcomes for agricultural outcomes in summary form is needed and periodically details must be collected. Such markets in large villages and what are called census towns are gaining importance. It has also evaluated the existing market price data and suggested both more detailed and regular collection. There are also prices emanating increasingly from trade on a global plane. This is the consequence of imports for fighting food inflation. There are both classificatory and timing issues and the committee has suggested an advisory committee of users and data producers to build and maintain systems for information needs of both policy makers and businesses.

Agri-climatic regionalisation is needed for details of plans for the agricultural sector, like for example the RKVY where the schemes have to be in the resource frontier of the area, or again the NABARD concept of financial products for developed for each area. The committee has detailed a comprehensive strategy for information requirements for these needs.

The committee has a detailed section on inputs. It notes the variance between land use data of areas cultivated under irrigation and MOWR data, and has asked for reconciliation after taking care of coverage differences. It has asked for crop-wise data collection for other inputs. It has analysed the need for data on agricultural storages and recent trends of stocks providing the base for financial collateral and companies engaged in such agricultural services.

The committee has taken an economists view on futures and suggested data collection for the futures markets. Its view is that such information would help the farmer for making acreage and input allocation decisions and if markets are thin the information could be used for data-based regulation of futures markets and strategies for their development.

The committee has given examples of data networks and portals being used after being developed in the recent past and has suggested PPPs for this purpose.

The author is a former Union minister