The Hindu Business Line : Data issues: Who said they hardly matter?

Instances where data needs are not met are many and serious. This reflects the limited appreciation of the information requirements of economists and policy-makers alike. There is a need for a statutory Statistical Commission with sufficient powers to consult, and enforce discipline on information collection, dissemination and transparency practices, says A. Vasudevan.

WE OFTEN take it for granted that statistics on the Indian economy are comprehensive and sound, but very rarely ask ourselves whether this is true. Data furnished by different official and private agencies are accepted without questioning their quality and coverage. Generation of newer sets of data for policy or research purposes is limited.

The instances where data needs are not met are large and serious. For instance, it is just not possible for one to know the current unemployment rate or the unemployment rate year by year. All we have is information based on National Sample Surveys at intervals of time and again with methodologies that are not standardised.

We also do not have a representative consumer price index for the country as a whole. What we have is the Consumer Price Index for industrial workers but it is neither representative nor available on a more frequent basis than a monthly one. As a result, inflation is often measured in terms of the Wholesale Price Index of `all commodities'.

Data availability on a continuous or time series basis is also a problem for researchers. For example, data on retail orders for consumer durables or orders for new construction or for new machinery are not available on a monthly basis or even on a quarterly basis. Such data would give an idea of the state of the consumer demand.

The data on inventory positions as also on capacity utilisation of different industries that at one time were made available are hardly reported nowadays.

In the financial sector, a frequently asked question is about the market liquidity measurement. For this purpose, one needs not only bid-asked spreads and yield spreads, but also indicators such as quote sizes, trade sizes, trade volume, and trading frequency transaction wise for each day for each security, be it of the government or of private corporate body. These data are not easily available.

On the retail payment side, rarely one gets across data on domestic and international credit and debit cards. Reliable information on the derivatives markets, and on insurance and mutual funds and pensions on a continuous basis is rarely provided.

No doubt, it is not necessary to publish all the data collected. But agencies do need to inform the interested researchers about the types of data they collect and how to access them. Still better would be if agencies on their own place in public domain all the information about the data that they collect, the data they publish and the methodologies followed in the collection of data and in processing them.

This is necessary now than ever before because the expectations factor has become critical for ordinary economic agents to taking decisions on spending and investment. From the point of view of the policy-makers too, knowledge about the expectations would help determine the strategies required to either promote or counter expectations.

The need for diverse and quantified information is what most researchers experience. But most research workers have simply accepted the data provided by the agencies and rarely asked for specific data to address the specific research problems that they are wrestling with. This has to an extent been a disincentive for agencies in taking interest in generating new types of data. One suspects that official bodies do not bother to collect new types of data unless they encounter serious problems in addressing major policy issues or the top executives specifically ask for them.

This appears to be true, and sadly enough, reflects the limited appreciation of the information requirements by the economics and statistical wings of the agencies. Otherwise how does one explain the strong growth of the services sector and this sector's disconnect with the factors that influence other real sector activities and with the forces that operate in the financial and business world?

Let us not forget that till almost the end-1970s, many Indian economists measured the linkages between agriculture and industry, and worked out the influence of projected industrial growth on agricultural growth. Though the services sector has posted high growth rates in recent years, averaging about 7 per cent a year, it is not clear why the sector's dynamics did not cast an influence on the rest of the economy.

Some attribute this phenomenon to the unwillingness of producers in agriculture and industry to absorb and adopt new technologies. However, it is not clear why this should be so. After all, Indian farmers adopted new technologies under the Green Revolution of the late 1960s. Industries too have often exhibited readiness to emulate new technologies and introduced new and qualitatively improved products: some of the examples of technology-emulation are in the areas of manufacture of automobiles and of plastics in India. One can in fact cite a number of examples of man-made materials being used in manufacturing either singly or in blend with natural resources.

In the above context, one needs to also revisit the entire gamut of financial statistics and the manner of their presentation because it has become difficult for most research workers (and policy-makers as well) to reconcile them with the statistics on national accounts.

One is also puzzled by the financing patterns of the corporate sector from the balance-sheet data of the corporate sector. The growth rates of companies are often seen in terms of profits earned but the demand for the products of the companies and the costs incurred in the financing strategies that were followed hardly give definitive clues on how the ultimate profit outcomes have emerged.

The coverage and quality of data that are published in certain cases — what the economists call the data uncertainty — creates serious problems of inter-temporal comparisons.

Although this is used as an argument for gradualism in policy making — for example, interest rate smoothening — this could be partially avoided by communications about policies regarding minimum responses and the number of revisions needed to overcome different proportions of responses. Transparency of this sort is most needed in respect of data on national income and the indices of industrial production and wholesale prices.

It is now time also to undertake a number of surveys in different areas of activity in different regions, but given the official agencies' limited capacities, they could be entrusted to research institutions and other credible organisations in different regions.

Researchers interested in North-East States, for example, find very little of worthwhile data. But once this suggestion is accepted, it would be necessary to have some standardisation (with flexibility of course) of methodologies to be worked out by experts and enforced by the sponsoring agencies.

For this to be realised, there is a need for a Statistical Commission to be set up statutorily with sufficient powers to consult, and enforce discipline on information collection, information dissemination and transparency practices.

They also need to continuously research into the statistical needs of the country and of their comparability with the systems abroad, particularly those that are agreed upon under the United Nations systems. This would eliminate some of the current deficiencies and help improve forecasting abilities and policy analyses.

(The author, former Executive Director of the Reserve Bank of India, can be accessed on asurivasudevan@hotmail.com)