Did Indian growth really slow down during 2016-18?
By Jayanta Kumar Mallik
The National Accounts Statistics (NAS) numbers released recently have elicited sharp reactions. Questions are raised about the credibility of the back series GDP data, 2004-12. The current series data for Q2FY19—7.1% GDP growth at 2011-12 prices, lower than market expectations—raises concerns about the growth prospects of Indian economy (‘Welcome to the new Hindu rate of growth, of 7%’, FE, December 3; https://goo.gl/sdMQ7n). The quarterly estimates are built on thin information; perspectives can be drawn from annual data. GDP growth slowed down during 2016-18: from 8.2% in 2015-16 to 6.7% in 2017-18. This, according to popular notion, was due to the short-run (and lingering) impact of demonetisation announced on November 8, 2016, and GST introduced in July 2017. There is no credible evidence to support this position. The slowdown is a display of weak spots in macro data: the gross value added (GVA) in storage and telecom appear to be underestimates, while the numbers in construction and insurance are extremely peculiar; total GVA excluding these four (constituting a little over 90% of the total) recorded a growth of 8.1% in 2016-17 (8.4% in 2015-16, 7.2% in 2014-15). Estimation problems—in the current series—need attention, if the NAS numbers are to be rendered useful.
According to NAS 2018, the GVA in construction during 2012-17 had grown at 2.5%, on average, in sync with its output (2.6%). The slow and harmonious growth of output and value-added in this sector is notable. Is this realistic in a country of 1.3 billion people—it’s not a totalitarian economy—growing at a rate of 7%? The presumption that the withdrawal of the two currency notes from circulation had muted the growth in the Indian economy remained rife in intellectual discourse for two full years. The growth of 1.3% in construction GVA in 2016-17 was perhaps a good find for the orthodoxy that the cash-intensive activity in this sector was stifled because of demonetisation. What had happened to this activity during the four years before demonetisation? In 2012-13, for instance, it grew by only 0.3%. Construction GVA had grown at 10.1%, on average, during 2003-12 (10.6% using back series numbers).
GVA in storage and telecom appear to be underestimates, while construction and insurance are extremely peculiar cases: In 2016-17, growth in five out of 11 sectors (agriculture, forestry and fishing; mining and quarrying; electricity, gas, water supply and other utility services; public administration and defence; and other services) was either higher than or on par with that in 2015-16. In agriculture, output of major crops recorded sharp growth both in kharif and rabi seasons. Illustratively, foodgrain output in kharif (138.3 million tonnes) and rabi (136.8 million tonnes) of 2016-17 were new records, and they attained new peaks (140.7 million tonnes and 144.1 million tonnes, respectively) in 2017-18, belying the thesis that demonetisation had muted the activity in this sector. The growth in three others (manufacturing; trade, repair, hotels and restaurants; and real estate, ownership of dwelling and professional services) remained high at around 8%, albeit a moderation from 2015-16. The growth dipped sharply in three sectors: construction; transport, storage, communication and services related to broadcasting; and financial services . These three need a closer look.
Construction (along with capital formation) has been tracking movements in steel prices. This explains subdued growth during 2012-17 as also the pick-up witnessed in 2017-18. The transmission process has been explained in a paper submitted to a journal, the gist of which is given here. The estimates of GVA in pucca (accounted) construction, the main building block of gross fixed capital formation (GFCF), are prepared from limited information on “basic construction material” using fixed ratios of a benchmark year. Iron and steel, with two-thirds of the weight assigned to basic construction material, occupies a key position in the estimation. In tandem with global prices, domestic prices of iron and steel in the wholesale price index (WPI) rose by 161% during 2003-12 (more than twice the increase in the headline WPI) and fell by 15% during 2012-17 (as against 12% increase in headline WPI). The surge in the investment rate (gross domestic capital formation as percentage of GDP) during 2003-08 and the slide during 2012-17 had tracked the movements in steel prices.
Transport, storage, communication and services related to broadcasting: In this sector, GVA declined in railways and storage; and in communication and services related to broadcasting, the growth rate dipped to 2.5% with the telecom GCA contracting by 3.9%. The railways have been losing business to airlines and road transport operators. The decline in storage appears to be a portrayal of the activity of government-owned entities, non-cognisant of the tremendous growth in warehousing in the private sector. Warehouse transaction volume in eight major cities (Ahmedabad, Bangalore, Chennai, Hyderabad, Kolkata, Mumbai, NCR and Pune) recorded a growth of 85% in 2017. Warehouse space in these eight cities grew by 15.3% in 2017 (14% in 2016). In comparison, average capacity of the facilities of the Central Warehousing Corporation (CWC) declined by 13.4% in 2016-17. The GVA in telecom also fail to reflect the activity in the private sector. Tele-density (telephones per 100 persons) in the country improved to 93% at end-March 2017 from 83.4% a year ago, with the total number of telephone lines growing by 12.8% in 2016-17. Growth happened in wireless lines (13.2%), while the number of wire lines declined by 3.3%. Wireless lines accounted for 90% of telephone lines provided by the private sector players as against 10.3% in the public sector.
In this segment, GVA growth in monetary financial institutions (mainly representing banks) slipped to 3.4% in 2016-17 (5.6% in 2015-16), and GVA contracted by 18.7% in insurance and pension funds with life insurance GVA falling by 35.4%. In banking, it is well known that “it is not business as usual for lenders and borrowers”. It is difficult to understand the contraction in insurance. Details used in the estimation of GVA in this sector are not available, unlike in construction where even minute details on items like fixtures and fittings are available in NAS. In any case, the decline in insurance GVA is not a reflection of the current business nor does it help us in understanding the economic slowdown. In life insurance, for instance, sum assured increased by 15.9% in 2016-17 (12.5% in 2015-16), premium income grew by 14% in 2016-17 (11.8% in 2015-16), total funds under management grew by 14.1% in 2016-17 (11.3% in 2015-16), and profits after tax rose by 4.2% in 2016-17 (-2.6% in 2015-16).
It would make sense to exclude construction and insurance (peculiar cases) and storage and telecom (underestimates) and compute growth of the rest. This measure, constituting a little over 90% of the total GVA, recorded a growth of 8.1% in 2016-17 as compared with 8.4% in 2015-16 and 7.2% in 2014-15.
Imbalances in demand-supply growth, juxtaposed with the commodity prices, suggest that the problem of underestimation was more acute in 2017-18: The growth rate of domestic demand (consumption plus investment) improved to 7.7% in 2017-18 (6.7% in 2016-17). In the supply side, the growth in commodity production (comprising agriculture, forestry and fishing; mining and quarrying; manufacturing; and electricity, gas and other utility services) fell to 4.7%, and the entire load of domestic demand was on services (see chart). The perceptive reader would get a sense of the sustainability issues embedded in these imbalances. However, before such concerns could be raised, it would necessitate a note that the uneven growth in commodity production and domestic demand should have led to a surge in commodity prices; that did not happen, thankfully. In 2017-18, WPI of all commodities increased by only 3%, and the consumer price index (CPI)—Base: 2012=100—representing select commodities and services rose by 3.6%.
The demand-supply gap is a manifestation of the gaps in estimation. GVA numbers for 2017-18 in three commodity sectors appear to be underestimates. The 3.4% growth in agriculture draws on the third advance estimates of crop production: upward revisions in the fourth advance estimates (notably, in foodgrains to 285 million tonnes from 280 million tonnes in the third) would yield a higher growth. The deceleration in mining GVA growth is not in sync with its activity. According to Indian Bureau of Mines, value of mineral production (excluding fuel, atomic and minor minerals) grew by 25.1% during 2017-18 (April-February) as compared with 14.5% in the same period a year ago, and output volumes of fuel minerals generally showed improved growth. The deceleration in manufacturing is negated by other data. The IIP show acceleration in manufacturing growth from 2.8% in 2015-16 to 4.4% in 2016-17 and further to 4.6% in 2017-18. Gross capital formation in this sector had grown by 25% in 2016-17 on top of 7.9% in 2015-16. Non-oil imports, an indicator of demand for industrial inputs, grew by 20% in 2017-18 in dollar terms (-0.2% in 2016-17).
Most of the slowdown in 2016-17 was in the four sectors (construction, storage, telecom and insurance) that suffered from underestimation or exhibited peculiarity. The underestimation, in some cases, represents the inability of macro data to fully reflect the transformation in the Indian economy characterised by faster growth of activity in the private sector as compared to that in the public sector. Growth has been strong where demonetisation was supposed to have hurt the most—in agriculture, for instance. The estimation problems need attention, if NAS numbers are to be rendered useful. The details that have gone into the computation of GVA in insurance should be published to enable meaningful research work.
While activity-wise macro data lend no support to conventional wisdom, studies using alternative micro-level data (such as ‘nightlight’, which presumably better captures informal sector activity) show that demonetisation had adversely affected economic activity. This evidence is not a surprise; but the problem comes when it is used to gauge the macro-level impact. Measurement of the impact on GDP growth of disruptions attributable to demonetisation would be appropriate when resources are fully employed; else, producing units can make good the temporary loss by deploying extra resources/efforts at a subsequent period. This point is clear from the data on agricultural production. This is not to belittle the research using innovative data, and the least to undermine the need for strengthening the statistical system. Data from conventional sources presented here suggest that growth numbers would be higher than those at present.
– The author works with RBI. Views are personal