India's China syndrome

Instead of the average growth of 7 per cent that the official data shows in the period between 2011-12 and 2016-17, Arvind Subramanian's estimate is a meagre 4.5 per cent

Those who are aware of the intense debate on the reliability of China’s economic data should find many familiar elements in the face off between former chief economic adviser Arvind Subramanian’s (AS) and the keepers of the faith in India’s official data machine. They include the Prime Minister’s Economic Advisory Council (the PMEAC), which has issued a rather stern riposte (“GDP estimation in India- Perspectives and Facts”, 19 June 2019) to AS’s claim that the new methodology for measuring India’s GDP with 2011-12 as the base year significantly overstates its growth rate (“India’s GDP Mis-estimation: Likelihood, Magnitudes, Mechanisms, and Implications”, Arvind Subramanian, CID Harvard University, June 2019). Instead of the average growth of 7 per cent that the official data shows in the period between 2011-12 and 2016-17, AS’s estimate is a meagre 4.5 per cent.

 
Where does China fit into all this? China data-sceptics have long claimed that if the indicators that should prima facie correlate with GDP growth are used as proxies, the growth rate would be very different. The most well-known of them is current premier Li Keqiang (LK). After dismissing official statistics as “man-made” he provided his own measure of growth (the Keqiang index) using a combination of three things — electricity consumption, railway freight, and bank lending. Estimates based on this index and its most sophisticated variants show much lower growth than the official releases. While the government currently pegs growth in the ballpark of 6.5 per cent, the sceptic estimates vary from 2.5 to 4.5 per cent.

 
There’s a twist to this tale though. Over the last few years, L-K type measures have come in for severe criticism. The new China, the critics claim, is very different from the old “industrial” China when investments and exports were the main engines of growth.  A change in economic policy direction (articulated in its 12th Five-year plan of 2011) drove a profound shift in China’s economic structure to a more inward-looking, consumer-driven economy. Thus, using an old economy measure like the LK index fails to capture this structural shift and produce a systematic downward bias in the growth rates. The defence’s case for India’s official data is similar.

 

At its core, AS’s argument revolves around L-K type proxies for GDP. Instead of Li Keqiang’s measly three, he produces a bagful of 17 and frets over the fact that while in the 2001-02 to 2011-12 period, they correlate well with GDP, these correlations break down in the subsequent period. AS also uses a four-variable index (credit, electricity, exports and imports) and pools this with similar indices for 70 countries. Since the relationship between GDP and this index bucks the general trend, AS concludes that the new methodology is flawed and biases GDP estimates up.
    
One can pick many nits in AS’s 17 as the PMEAC paper does. For one thing, many of them like cement and the IIP measure volume rather than value. It is not surprising that they moved closely with the old GDP series. The major grouse with the old estimation method was that it relied (partly due to data paucity) too much on volume measures when it should have used value-addition instead. The new methodology attempts to correct this and it is not surprising that the volume-GDP link weakens.

 
The PMEAC’s rebuttal is comprehensive but in the muddle of technical minutiae, the two central critiques of AS’s approach seem to lose their sting.
 
First, AS’s declining correlations captures a shift in the structure of India’s national income. India’s growth patterns over the last few years have been surprisingly similar to China’s with domestic consumption emerging as the principal driver, and investment as well as exports getting short shrift. While this is widely known, a couple of data points might drive this point home more convincingly. The average annual contribution or share of investment in the expansion of GDP from 2001 to 2011 period was 51 per cent. In the subsequent period it dropped to 28 per cent. The share of consumption (as a driver of GDP growth) went up from 48 to 58 per cent. The contribution of exports fell by as much as 24 percentage points. For China, this was intentional and driven by policy. For India, it was the product of local and global circumstance.

Second, both China and India saw a credit binge in the previous decade that spilled over to the current one. For India, it resulted in the well-known NPA crisis. For China, it manifested in investor concerns over stability and episodes of large capital flight. Regulators in both the countries began responding with policy measures from 2015 onwards. The move towards the Prompt Corrective Action (PCA) mechanism and Asset Quality Review (AQR) began at this time in India; China began its policy of “de-risking” its financial system.
 
Thus the second or allegedly “dodgy”’ phase of AS’s data series for India saw a major moderation in credit growth. This impacted investment and was responsible for the change in growth composition as well. Retail credit was unaffected and is consistent with the consumption boom story. However, with a relatively low share (around 18 per cent in 2012) in aggregate loans, it could not pull headline credit growth up. 

 
To add to this, Indian companies turned more to the markets for their funding needs. Between 2011 and 2017, while the outstanding credit of banks in India doubled, the value of outstanding corporate bonds trebled with a 30 per cent annual average growth in issuances. Working capital funding through commercial paper also saw a boom, and more so in the recent years on the back of a wave of liquidity created by demonetisation.
 
All these factors put together could explain why something apparently as simple as the relationship between credit and GDP growth could have changed substantially between the two periods. AS suspects a flaw in the estimation methodology whenever he encounters a change in the way economic variables relate to each other. In his paper he spares little time to listen to the stories that explain these changes, a habit rampant among China’s data sceptics. We wonder why.

  Barua is chief economist and Arora, a senior economist, at HDFC Bank