The new GDP data series is an improvement over the old, at least in terms of methodology and international comparability.
Niels Bohr, the Nobel-winning physicist and one of the key members of the Manhattan Project, which led to the production of the first atomic bomb, once famously said, “Prediction is very difficult, especially if it is about the future”. In India, it seems that even the past is uncertain. For example, until January 29, 2015, official data showed that India’s GDP grew by 4.5% and 4.7% during FY13 and FY14, respectively. The official data release on January 30, 2015 has now revised these numbers upwards to 5.1% and 6.9%, respectively.
The consensus expects (as per RBI’s December 2014 survey of professional forecasters on macroeconomic indicators) FY15 growth to be 80 basis points (bp) better than FY14 and FY16 growth to be 100 bp better than FY15. If one adds up these forecasts with the newly released data, then India should grow by 7.7% in FY15 and 8.7% by FY16. In fact, the recent collapse of international commodity, especially crude oil prices (not incorporated in the December 2014 consensus GDP forecasts), should add more to India’s GDP in FY15 and more so in FY16. So India seems all set to surpass 8.7% GDP growth, the average rate the country registered during the heydays of the Indian economy, in 2004-08. This is particularly credible because India’s past heydays coincided with a booming global economy, while the current global growth situation and outlook is largely sombre.
While one is free to be sceptical of the sharp revision in data and the implications, both past and future, one need to keep in mind a few facts. This is not the first time that with the base-year change the past data has been revised sharply. For example, when the base-year was changed from FY00 to FY05, the GDP growth for FY00 got revised from 6.1% to 8%.
Many macroeconomic data, such as foreign trade, money supply and government finance are sum totals of actual transactions. While there can be misreporting or calculation errors in such data, these are actual numbers and not estimates. In the case of GDP and many of the underlined parameters, such as agricultural or manufacturing production, the data are estimates rather than actuals. Such estimates are generally based on past sample surveys. When new surveys throw up structural changes, the past estimates get revised.
There are differences between routine data revisions and base-year changes. When any data is first released, it incorporates fewer sample points/returns submitted by the designated agencies. As more response flows in, the data gets revised. Generally these revisions are relatively small. In the case of base-year change, the whole structural model for estimation gets revised, resulting in potentially large changes in the past data.
Large data revisions, either due to the late receipt of statistical returns or change in the structure of the estimation process, is by no means unique to India. Data, even for the advance industrialised countries with much stable statistical systems and steady economic structure, also get revised, sometimes by a large margin.
India publishes two sets of real GDP data—demand-side and supply-side. While in most of the countries, the demand-side GDP is taken as the representative rate, in India it is the supply-side GDP. Indian statistical agencies, in the past, suggested that the country’s statistical framework for estimating the supply-side GDP is more robust than the demand-side GDP and, consequently, the supply-side number becomes the representative rate in India. Methodologically, the difference between the demand and supply-side GDP is the treatment of indirect taxes and subsidies. While these two measures of GDP growth is expected to move in tandem with little difference, these are unlikely to be exactly equal. Once again, the base-year changes can magnify the differences between these two growth rates. For example, in the new-series (base FY12), the demand-side GDP growth rates at 5.1% and 6.9% in FY13 and FY14, respectively, are higher than the same from the supply-side at 4.9% and 6.6%, respectively. It seems, in line with the international practices, India now wants to move towards demand-side GDP as the main growth indicator, optically the GDP growth for FY13 would get revised from 4.5% to 5.1% and for FY14, it will get revised from 4.7% to 6.9%.
A comparison of the new with the old data suggests certain interesting points. The new series suggests that the old series slightly overestimated the absolute size of the economy in FY12 and FY13, while the size of the economy is largely the same for FY14 under both the series. These hold true both from the demand- and supply-side.
The implicit GDP deflator suggests that inflation was underestimated for FY13, while it was overestimated for FY14. Once again, these hold true for both the demand- and supply-side GDP.
The new series suggest that the old series underestimated the size of agriculture and industry, while it considerably overestimated the size of services sector. In terms of individual components, the extent of underestimation for agriculture, manufacturing and real estate, etc, services were considerable. On the other hand, overestimation of financial services, transport & communication and trade & hospitality (all components of the services sector) were noticeable. Part of this, however, is due to change in methodology by the new series to report activities on a consolidated (principle activity) rather than segmental basis, as was the practice earlier.
On the demand-side, the new series show that the shares of both the major components of GDP—private consumption and investment—have declined. This has been made good largely by the falling share of imports (a negative entry in GDP) and also to some extent rising share of discrepancies (unexplained component).
Interestingly, the new series suggest underestimation of private consumption and overestimation of investment growth by the old series.
Despite the substantial jump in FY14 GDP growth in the new series, the real problem facing the economy—slowdown of investment growth—continues. In fact, the new series suggest that the investment situation in FY14 was more precarious than what was captured by the old series.
The real reason for the investment quandary, as depicted by the new GDP series, seems to be the sharp draw down of inventories and fall in valuables (mostly gold, which forms a major part of the physical capital formation by the households). On the other hand, fixed capital formation seems to have fared better in FY14 than what was captured by the old series. This, to some extent, rekindles hope for a noticeable recovery in FY16.
Whether India will be able to finish FY15 with close to 8% growth and whether we would reach 9% growth in FY16 and become world’s fastest growing economy remains uncertain. No one was talking about such possibilities, at least publicly, before the revised GDP data was released on January 30. Whether the Indians and the global community at large would find the major revision of past numbers credible also remains doubtful. We, however, feel that the new series is an improvement over the old series at least in terms of methodology and international comparability. With the current conducive business climate and high expectations, GDP growth over 8% in FY16 certainly does not look outside the realm of possibility. Meanwhile, we will remain busy revising the structural forecasting models and updating all the past data, particularly those expressed as percentage of GDP. Despite the uncertain past, we seem to be moving towards a bright future.
The author is chief economist, Anand Rathi Financial Services Ltd