Rejigging statistics

The comprehensive overhaul of India’s national income statistics, which places India among the fastest-growing economies in the world, confounds economists, policymakers and the government.

Union Finance Minister Arun Jaitley and Reserve Bank of India Governor Raghuram Rajan interacting with media after the RBI board meeting in New Delhi on August 10, 2014.

GOOD tidings, even if only of the statistical kind, can be bewildering. Just as Union Finance Minister Arun Jaitley started preparing for his first full-fledged Budget, news came that the Indian economy was chugging along merrily at a world-beating rate of 7.4 per cent in the current year. But the problem for the Finance Minister is in reconciling this apparent acceleration in the pace of growth with other evidence on the ground.

For instance, how can an economy that is growing at its fastest in recent years also experience its lowest levels of inflation in about a decade? Or, how come the booming economy is not experiencing a growing demand for credit from the banking sector? These ought to perplex a Finance Minister in normal times, but for now the political captains of the Indian economy seem to be savouring the glad tidings ushered in by a statistical adjustment, instead of worrying about how they square with reality.

The latest revision to the national accounts, including fundamental changes in the methodology of calculating national income, has confounded not only economists and policymakers but also the government. While the changes in the methodology have been justified because they now adhere to international best practices as defined by the United Nations’ System of National Accounts, the problem for practitioners is their backward compatibility.

This is not a trivial issue. Any statistical set that is incompatible with past estimates makes it difficult to compare magnitudes of the past, which is the analytical basis for statisticians, policymakers, the government and citizens at large. This is precisely the problem posed by the revised numbers, especially of the gross domestic product (GDP). When the revised GDP numbers (with the new base set at 2011-12) were released in January, they caused a stir because they conflicted with the perception that the Indian economy had slowed down significantly in the past few years. But more consternation was in store when, on February 9, the Central Statistical Office (CSO) released advance estimates of the GDP, which indicated that the Indian economy would grow by 7.4 per cent in the current year (2014-15).

Soon after the release of the first set of new numbers in January, Reserve Bank of India Governor Raghuram Rajan, in a thinly veiled expression of surprise at the numbers, said: “We need to spend more time understanding the GDP numbers.” He added: “It is premature to take a strong view based on [the new] GDP numbers.” In particular, Rajan expressed surprise over the revised growth numbers for 2013-14, which was by common concurrence a bad year for the Indian economy. “Most of the data we have seen in 2013-2014, except inflation which was very strong, give us a sense that there was a slack in the economy,” he observed. The quizzical reaction of the Governor of the central bank warranted a briefing by the CSO Director General, Ashish Kumar. “Now they [RBI] have no doubts about it,” Kumar said after the meeting with the RBI Governor on February 18.

The new data, using a brand new methodology, showed that the Indian economy grew at 6.9 per cent in 2013-14, not at 4.7 per cent as estimated earlier using the old series (2004-05 base year). The growth rate for 2012-13 was also revised upwards to 5.1 per cent, instead of the 4.5 per cent estimated earlier.

However, within the same data set there lurked several other surprises, notable among them being the rate of capital formation. One would normally expect asset formation to gather pace during an economic upswing; however, the new data show that capital formation declined from 37 per cent to 33 per cent between 2012-13 and 2013-14. How could a smartly growing economy simultaneously experience a deceleration in asset formation? This was the obvious question raised by sceptics.

New method The CSO, in its note to the new statistical series, has identified three major changes in the way it computes national income and output. The first relates to how national GDP as a measure of what in economic parlance is referred to as incomes accruing to various “factors” of production—land, labour and capital and rewards to entrepreneurs —is calculated. The second pertains to the usage of market prices to compute GDP; the justification for this emanates from the logic that these are the prices at which economic agents actually transact in. The third major change has been the introduction of a new concept of “basic prices” in order to compute the extent of gross value added (GVA) in the new methodology. The concept of basic prices involves the netting out of taxes and subsidies in order to arrive at the level of GVA in the economy. Analysts have pointed out that the singling out of the government as a separate entity (and an economic agent) in order to compute GVA is not justified by any system of national accounts. Their logic stems from their understanding in economic theory that the government cannot be counted as a factor of production in its own right.

But by far the most significant change—which appears to have contributed significantly to the apparent inflation in levels of national income —has been the inclusion of new data sources. Foremost among them is the usage of data filed by companies to the Ministry of Company Affairs, instead of relying on the sample surveys conducted by the RBI and the Annual Survey of Industries, which is released by the Union Ministry of Statistics and Programme Implementation. The reliance on the new data source, which gathers data from over five lakh companies, has been justified as being more comprehensive and superior in sweep when compared with both the RBI’s sample and the relatively small coverage (about 2,500 companies) of the Annual Survey of Industries.

The usage of the new method and data source has resulted in dramatic changes in the rate of both savings and investment by Indian companies. Savings, as computed by the new method, were 30 per cent higher than those computed by the old method (2004-05 base) in 2011-12; for 2012-13 the new method yielded a savings rate that was 40 per cent higher! In effect, if one is to rest sound economic advice on the new data, the share of the private corporate sector in national savings has increased from about 29 per cent (according to the old series) to 40 per cent (as per the new data set).

Significantly, while the portion of GDP emanating from the manufacturing activity is set to increase by 6.7 per cent during the current year, growth estimated by another data source, the Index of Industrial Production (IIP), grew by only a little over 2 per cent in the nine months ending December 2014. Even the earnings and profits of corporates during the year have been pretty disappointing, as is evident in the response of the share markets to their announcements.

Problematic source

One possible explanation for the inflation in national income could perhaps lie in the very choice of data source, especially those emanating from the Ministry of Corporate Affairs. The data are, after all, an aggregation of data (primarily of a financial nature) that are furnished by corporate entities. This could result in the possible inflation of the data on two counts.

First, there are questions about the sanctity of the data submitted by corporates for purposes that are primarily of an accounting nature. The verification of such data could be problematic for several reasons, one among them being the competence and ability of the Ministry of Corporate Affairs to sift through this data for a purpose that it is not designed to perform. Second, there is a possibility that the data, because they are meant primarily to be for an accounting purpose, have been “financialised”, which could explain the inflation in the numbers on the level of national income.

Even if the new data set has resulted in better coverage—and, conversely, less “leakage” of statistics pertaining to economic agents—it still poses serious challenges. It is easy to understand the argument that the new method captures a wider picture of the economy, which explains the inflated numbers pertaining to the national income. But this is about the level of national income generated during a year. How does one explain the relative performance of the economy over time, as measured by the growth rate of the economy over time, especially when they are irreconcilable with other numbers pouring in from other sources of data?

In many ways, it is difficult to escape the conundrums that the new data pose. While industry has welcomed the pace of growth, its complaints of the government not doing enough to “ease” the process of doing business would ring hollow. For the monetary authority (the RBI), a cut in interest rates (a continuing clamour from industry) would be incompatible with the new numbers on growth. As for Arun Jaitley, reconciling the substantial slack in economic capacity (of men as well as material) with the rosy numbers will be a tough task indeed.