Monday, April 29, 2013

In India, No “Statistics” for the Statistician


The Ministry of Statistics and Programme Implementation does a silent job of collecting data on various fields from the entire country. Some of the data have predominant significance in an individual’s life - like that for Consumer Price Index [Urban] where data is taken on the selling price of certain goods [pre-fixed]. The concerned data help in determining the Dearness Allowance (DA) of government employees. Some other data which also gets collected may not be of direct use to individuals, but in long term, it helps in policy formulation and taking important decisions relating to the holistic development of the country. These include the area enumeration of different villages in India, getting idea regarding the amount of crop yield from a particular field for all seasons – viz. Rabi, Kharif and Jayat Rabi, socio-economic survey, urban frame survey, Annual Survey of Industries, among others. All these surveys are done by the Field Operation Division of National Sample Survey Office under the aforementioned Ministry. The surveys are carried out simultaneously by the different Regional and Sub-Regional Offices all over India.

One of the major problems that the investigators face is the lack of response since there is no law binding the informant for supplying information. Suppose, let’s take the experience of a crop-cutting survey. Now any state of India has hundreds of villages and it is the responsibility of Field operations Division (FOD), Faridabad to allot certain villages for each round in such a manner such that after every five years, the same village gets repeated again. A tedious task no doubt, given the size of most of the Indian states. But then one may sideline this as a routine work, saying that they are just repeating the procedures which they are following from time immemorial. Next job is to pick up a field through random number. Suppose village A in state X got selected and there are 1000 plots in which crop is being cultivated in that village, then if the random number being allotted is 700, then choose the 700th plot for your study. If random number allotted is 1500, divide 1500 by 1000 and the remainder i.e. 500 will give you the required plot number. Its statistics so far – probably no big deal, and can be done by sitting at the village Patwari’s office.

Even if the chosen plot does not grow the required crop, say wheat about which information is needed - then go for the next field, not that complicated yet. But now surfaces the typical problems. One fine morning you happily go to the field and ask the farmer for his help. The poor Indian farmer who used to be very naïve would previously agree to help unconditionally - making life easier for the investigator. But now the farmer, who has also started feeling that he has missed his share of pie for quite some time asks the question-

              mujhe kya milnewala hain?

And well, the investigator knows the answer only too well. Rs 50, may be, and that too after a year. For that the farmer has now to cut the crop in the investigator’s presence, manually thresh it and give the exact measurement to the investigator. The most difficult is the crop cutting part - it is not the usual crop cutting that the farmer does at regular interval of time. You take a square of side 5m or an equilateral triangle of the same side length or a circle with radius 5m, that too after measuring the plot by actual walking, then subtracting 7 from both length and breadth and then considering random numbers. Then you have to separate the crop of that small area and dry it up. After say 8-9 days, the investigator once again goes. This time that crop should be manually threshed - even a difference of 100 gm can now bring lots of changes in the estimate of the total crop yield of that particular crop of the country. But what has the farmer got to gain from all these statistical jargons?

He gives time for crop cutting - but he has nothing to bind him to that date. If he feels he may cut it before or might cut it after - depending on his experience regarding the crop as well as weather condition. But the poor investigator has nothing to do - he has to get the information and that too for the plot chosen, he cannot force the farmer for any of the things, but cannot also return empty-handed. Even if the farmer refuses for manual threshing - the investigator might as well has to thresh the crop on his own, since his job is to get the data and he knows that the data would be needed to calculate the national figure. Now this particular yield rate will give an estimate of the amount of that particular crop grown in that season in the entire country. And then if any of the data goes wrong - the entire nation, powered by the now 24X7 news channels will attack the statistical machinery of the country and start throwing questions like how can there be such under production? How can China do it and not India? But does the process look very simple?

The process of data collection is not very complicated apparently. The Sample Design and Research Division (SDRD) of NSSO located in Kolkata is carrying out the research as to what are the items relevant these days on which the data should be collected. The Industrial Statistics (IS) wing of the Central Statistical office (CSO) has left no stone unturned in identifying the different industries in the country, the varied raw materials needed for the output of the various products that rule the market; and the FOD ensures that with the inadequate manpower it has, it must collect data for all the proposed schemes that will help in policy making. Now-a-days, the NSSO collects data for six years in a decade and has left four years for different central ministries who might request this particular department to collect data as per their requirement. Thus it is quite evident how much importance this particular ministry has.

But how much awareness do the common people have with regard to this is a big question. The people also feel that the time in which they are providing the data is simply meaningless since they are not going to gain anything from the data they provide and there is no law binding on them which compels them to provide the data.

At least, apparently no one sees any monetary gain in this.

Hence even if a person having three cars within his residential complex demands that his monthly per-capita expenditure is Rs 3000 for five persons, no one can challenge his data and no one can compel him to give the actual data. Similarly for the farmer, first of all, demarking the area with the precision as demanded by a statistician is itself a herculean job. Then that much crop needs to be separated and should be threshed separately manually in front of the Superintendent Officer and the village Patwari. Thesedays, on request from the Ministry of Commerce, the NSSO is collecting data on Wholesale Price Index Number. The Commerce Ministry has done everything - right from fixing the commodity basket at 2004-05 base to designing the schedule and selecting the quotations and the units.

What NSSO does is to go to the respective units, that too every month and request the people to fill up the schedule and upload it in the web portal. The Commerce Ministry has provided each unit with username and password and given a master password to the ROs for continuous monitoring. But however much they have tried to make life smooth; people would not do the job and give thousands of excuses for not doing it - reason being nothing is forcing them to provide the data.

But suppose if the Income tax Department demands certain data from the same unit - they’ll postpone all of their jobs at hand and start preparation for providing data to the IT Department. Reason? Well, the IT Department has power which NSSO doesn’t have. The entire work of NSSO is persuasion based. No one has any power to force the people to provide information -however useful the information might be. The only exception is in the Annual Survey of Industries where non-responses is comparatively less as people are empowered with the Statistical Act 2008. But in most of the cases, people do not have the necessary awareness. The importance that these data carry for the entire nation is something that is known only to a handful of persons who also do not have the necessary authority to make people part with the data.

Hence some sort of autonomy must be provided to the Ministry as data is the basis for everything. Instead of blaming the CSO or NSSO for some wrong figures, kindly give them the power so that data collection becomes a smoother and easier process.

--------------------------------------
The anonymous author spells out the ground realities which speak of the travails of the official statistician.