Wednesday, February 17, 2016

Why 7.6% growth is hard to square

This year, the gross domestic output (GDP) is forecast to grow at 7.6 per cent in real terms, that is, net of inflation. It is higher than China’s growth rate projection of 6.9 per cent, making India the world’s fastest-growing large economy. The growth rate has steadily climbed from 2012-13, when it had plummeted to 4.5 per cent. Why, then, is no one celebrating the turnaround or applauding the policymakers for their success? Apparently, few trust the official figures. Why?
Last year, a new series of National Accounts Statistics (NAS) was released with 2011-12 as the base year (replacing the earlier series with 2004-05 base year) — roughly a 10-yearly routine for the Central Statistics Office (CSO) to keep up with the economy’s structural changes using newer databases and improved estimation methods. This time around, the CSO has also incorporated the United Nations System of National Accounts (SNA 2008) — a welcome effort to keep up with international accounting standards.
A game of numbers
Usually, the base year revision changes the absolute level of GDP (and its major constituents), but its growth rates invariably remain the same. But it is different this time. While the absolute GDP size for 2011-12 with the new base year was smaller by 2.3 per cent, its growth rates for the following two years are significantly higher (compared to the older series). The difference is sharper by sectors (or industry): for instance, manufacturing sector growth rate for 2013-14 is 5.3 per cent in the new series, compared to negative 0.7 per cent in the older series. Disconcertingly, such a high growth rate is out of line with many other economic indicators such as credit flows, output expansion in major industries or capacity utilisation in critical industries such as steel or cement. In other words, the rosy official estimates did not pass the “smell test”. Hence the widespread scepticism of the new GDP numbers, shared by some policymakers as well.
India has thus not only surpassed China’s official growth rate but now shares its dubious distinction of inflated output estimates. China’s Premier, Li Keqiang, when he was the governor of Liaoning province, said famously about his country’s GDP that it is man-made and hence unreliable; it was meant for reference only. He further added that the true measures of China’s economy are growth in bank credit, rail freight and electricity output.
Why have the GDP estimates become unreliable after the revision? To be sure, early on, Indian national income estimates were not without blemishes. But the revision seems to have worsened the situation with widely questioned figures. Though the absolute GDP size for 2011-12 in the new series is marginally smaller (than that in the old series), its institutional composition has changed significantly. The private corporate sector’s (PCS) share in the GDP has expanded to 34 per cent now (23 per cent in the older series); and household (unorganised or informal) sector’s share in the GDP has shrunk to 45 per cent in the new series (from 56 per cent earlier). How could this happen? It is the result of changes in methodologies and the databases used.
Arithmetically, the sharp compositional change in favour of the fastest-growing sector would mean higher GDP growth rate, everything else remaining the same.
Private corporate sector
The revised NAS has used the Ministry of Corporate Affairs MCA-21 database of about 5.2 lakh companies to estimate PCS’s contribution to domestic output. It is then “blown up” (scaled up) to over 9 lakh “active companies” that claimed to have filed their financial returns at least once during the previous three years. Detailed investigations suggest shortcomings in these procedures, leading to an overestimation of the size and growth rates of PCS in the new GDP series — a tentative result that can be verified only if the MCA-21 database is made available for independent verification.
The CSO is, however, convinced of the superiority of the revised estimates compared to the earlier one based on Reserve Bank of India’s purposive sample of about 4,500 high paid-up capital companies. Prima facie, it is hard to dispute the CSO’s contention. But the truth lies in knowing the structure of PCS and the MCA-21 database.
There are close to a million registered companies (2014 figure), but their distribution is extremely skewed: about 65,000 (6.4 per cent) are public limited companies; those listed on the Bombay Stock Exchange are about 5,000. The Centre for Monitoring Indian Economy’s (CMIE) Prowess database consists of about 26,000 non-financial private corporations accounting for about 18 per cent of the GDP. The top 100 companies accounted for nearly one-half of gross value added of the CMIE estimate. Could the majority of registered companies, producing output sporadically (if at all), account for the rest of output of the PCS? Probably not, we would contend. Hence official GDP of the PCS overestimates the sector’s output.
One suspects that a large proportion of the private limited companies are tax hedges and are used to ensure promoters’ control over productive enterprises via benami (illegal) holdings, and/or to avoid laws and regulations. It is widely known that behind most large public limited companies there exist numerous private limited companies and unincorporated businesses, which promoters often use to optimise private returns for the entire group. Hence, a company (as a legal entity) exists mainly as a cog in the wheel supporting family business. The spate of scams in India in recent years has offered ample insights into such structures of private business, which effectively use a large array of private limited companies to siphon off surpluses into untraceable corporate and even non-corporate entities, via intra-group and inter-corporate transactions.
If the foregoing account is a fair characterisation of the PCS, then blowing up the sample estimates for the universe of active companies may be statistically valid, but its economic implications could be suspect. Therefore, the revised method could have contributed to an overestimation of the corporate sector’s contribution to the GDP.
Household sector
Since, by definition, the household sector’s production is not directly recorded in audited balance sheets, they are captured by large, nationwide sample surveys.
Conventionally, the sector’s output is estimated as a product of output per worker and the number of workers employed — an imperfect but widely accepted method for the unorganised sector. Usually, during the NAS revision, latest survey data are utilised to capture the most recent output trends. The shortcomings of such a method are widely known, but for lack of anything better, the simple yet sound practice was followed.
But the recent revision introduced a new procedure under the assumption that the older method overestimated the contribution of self-employed workers. The changed methodology drastically reduced output per worker in the unorganised sector, leading to the shrinkage of this sector’s output in the GDP.
In other words, methodological changes enlarged the size of the PCS and contracted the size of the household sector (keeping the overall size of GDP almost the same). They together seem to have contributed to render the growth rates in new GDP unreliable.
To conclude, the new NAS has forecasted a GDP growth rate of 7.6 per cent for the current year, making India the world’s fastest-growing economy. Few seem elated at the prospect, as the growth estimate seems out of line with other economic indicators. This is a continuation of the doubts expressed on the revised NAS published last year. After the revision, the size of the fast- growing private corporate sector has got enlarged and that of the household sector contracted. These changes are the result of questionable changes in the methodologies and databases used in the revision. We have contended that they could have seriously affected the GDP growth rates and its constituents. Hence, serious doubts about the GDP estimates persist.
(R. Nagaraj is a Professor at Indira Gandhi Institute of Development Research, Mumbai.)

No comments: