Above: Nirmala Sitharaman taking charge as finance minister. The government’s Economic Advisory Council has rebutted Arvind Subramanian’s claims on India’s GDP figures/Photo: PIB
The government is in denial as it refutes ex-chief economic adviser Arvind Subramanian’s assessment of GDP figures. But it should develop a new generation of reforms and rely less on raw wisdom
By Sanjiv Bhatia
There has been a lot of controversy around India’s GDP numbers. Many economists, including myself, have pointed out (https://www.indialegallive.com/cover-story-articles/focus/gdp-figures-the-politics-of-economics-63338) troubling issues with India’s growth numbers. In a recent article, India’s ex-chief economic adviser Arvind Subramanian stoked the controversy by claiming that there was a 95 percent probability that India’s GDP growth over the 2011-2016 period was around 4.5 percent, almost 2.5 percent per year lower than the government’s claim of 7.1 percent over that period. This suggests that over six years, the overestimation in GDP could be as high as $500 billion.
This has severe implications for the country’s credit rating, its credibility with foreign investors, the ability of Indian companies to borrow overseas and for the stock market. As Subramanian so aptly states in his conclusion: “The heady narrative of a guns-blazing India must cede to a more realistic one of an economy growing solidly but not spectacularly.”
This research, coming from a person as close to the real numbers as anyone can possibly be, reinforces the observations made about India’s actual GDP by myself and a few other economists.
The facts are irrefutable—underlying figures on consumption, investment, credit growth, employment and exports do not support an economy growing at seven percent.
The PM’s Economic Advisory Council recently submitted a rebuttal to Subramanian’s claims. But in a typical “shoot the messenger” style, it offered no alternative hypothesis, or new data. Instead it is laced with opinions and denials. This is not reassuring. The government can’t perpetually be in denial mode. Erroneous GDP calculations are putting the PM’s credibility on the line when he speaks of India being a $5 trillion economy by 2024—an event that would require GDP to grow at 14 percent every year over the next five years (the probability of which is close to zero given that India has had just one double-digit growth year in the past 70 years). As a result of this hubristic assessment of India’s economic growth, programmes to distribute wealth are higher on the government’s agenda than reforms to generate wealth. The reality that wealth must first be created before it can be distributed could hit India soon.
Policies driven by a focus on redistribution will be a crucial mistake. India’s first generation of economic liberalisation reforms, introduced in 1991, changed India to a wealth-creating economy, and over the next two decades, produced a remarkable increase in economic well-being. Per capita income increased by 650 percent, poverty decreased from 47 percent to 17 percent, real estate boomed and the stock market rose by over 1,200 percent. But the marginal benefits from those reforms have started to decline and a new generation of reforms is now required for India to continue its transition from a low middle-income economy to an upper-middle-income economy. That process starts with accepting the reality that India’s economic growth is overestimated and overhyped.
GDP has now become the standard measure of economic progress. But despite its ubiquitous use, there are doubts about both the accuracy of its measurement and its usefulness as an indicator of economic growth. There are three reasons why GDP numbers can be wrong. Firstly, GDP is an artificial construct and its measurement is fraught with error. Secondly, bad data and poor methodology can induce errors in estimation. Finally, there is outright fudging of data by governments and that makes GDP growth numbers untenable.
A construct like GDP is subject to methodological, estimation and measurement errors that are so profound that its stated value is quite abstract. So, when politicians become giddy about its value (India is the fastest growing economy in the world), and analysts give their estimate of GDP growth (GDP will grow at 7.25 percent this year), discount all that as sheer hype. Those estimates are entirely meaningless and will undoubtedly change with subsequent revisions (the US, for example, just recently revised its 1929 GDP number).
There are many challenges with calculating GDP. For instance, it is difficult to measure the contribution of innovation to GDP. How does one measure the value of a free Google Maps even though the travel time it saves can be used productively? Or the value of the voice recognition software that I am using to write this blog, which makes me significantly more productive? Or a free phone call on Skype which adds nothing to GDP while a paid international call on Airtel gets counted in GDP? There is now a strong consensus that GDP misestimates the digital economy and that the economic growth of innovative countries like the US is far higher than their stated GDP value. Conversely, the stated growth in traditional economies like India is far higher than actual economic development.
Another problem is that many activities go unreported in the calculation of GDP. Housekeeping, cleaning, cooking and other such duties performed by members of the family do not get counted in GDP. But if a maid cleans the house, and reports her income, it adds to GDP. This problem is particularly acute in a country like India where 90 percent of the employment is in the informal economy where payments are in cash or barter. Another problem with GDP is in the measurement of services. GDP works well in a production economy (a factory making widgets), but in a service economy like India how, for example, does one measure the productivity of a doctor? Is the doctor more productive if he sees more patients in a day or fewer patients but with better outcomes?
The problem of estimating GDP is compounded by the lack of good, unbiased data collection and statistical analysis. The quality of India’s data has always been suspect, and the recent merger of the Central Statistics Office (CSO) and the National Sample Survey Office (NSSO) into one unit, answerable to the government instead of Parliament, has accentuated concerns about the sanctity of the country’s data. While Subramanian astutely refrains from directly accusing the government of fudging the data, he acknowledges that it was change in data sources and methodology that is responsible for the overestimation of growth.
GDP can be calculated separately using three different methods: the supply or production method, the demand or expenditure method and the income method. The final number should be identical under these different methods. A simple example illustrates this. Assume a small island economy with a factory that produces ten shirts and sells them for Rs 10 each. Under the production method, the GDP of this island would be the production of ten shirts at Rs 10 each for a total of Rs 100. The expenditure method would calculate the amount paid by the residents to buy these shirts—again Rs 100. And as the expenditure of the residents becomes income for the factory, the income method also gives a GDP of Rs 100.
In most developed economies like the US, GDP is calculated using the expenditure method. The expenditure approach calculates the spending by the different groups that participate in the economy. So, GDP = C + G + I + NX, or (consumption by the citizens + government spending + investment by businesses + net exports). In India, GDP is measured from the production side by calculating the gross value added (GVA) from the production of goods and services using data from financial accounts of companies, tax data, and other proxies, and then adjusting it for inflation using price deflators.
Subramanian gives three reasons for the overestimation of GDP in the post-2011 period. First, he finds that the price deflators used to account for inflation in the calculation of real GDP are inappropriate—an argument supported by Gita Gopinath, chief economist at the IMF. Lower oil prices post 2013 and reduced input costs make the value added appear larger than it was. Secondly, Subramanian finds that the value added by the manufacturing sector was overestimated in the post-2011 period. He blames that on a shift in methodology—using data from the MCA-21 database of companies maintained by the Ministry of Corporate Affairs (MCA) instead of data from the Index of Industrial Production (IIP) and the Annual Survey of Industries (ASI). It was recently discovered that almost a third of the companies in the MCA-21 database are closed, non-traceable or misclassified. As a result of this GVA from manufacturing was being overestimated. The third reason Subramanian gives for GDP’s overestimation is the inflated value of the contribution of the informal sector. As it is hard to obtain reliable data from the informal sector, statisticians use a ratio of the value added by the formal sector as the imputed contribution of the informal sector. This ratio has remained constant despite clear evidence that the adverse impact of demonetisation was disproportionately higher on the informal sector.
One can always quibble with statistics, and I too have issues with Subramanian’s methodology, but his conclusions cannot be brushed away. For example, his use of correlation analysis on a small sample size of seven observations (the post-2011 period) is of concern. Correlation estimates on such small samples are notoriously unstable and fraught with errors—which may explain why Subramanian finds a negative correlation (instead of a low but positive correlation) between GDP and 11 of his 17 economic variables for the 2011-2016 period. A negative relationship between IIP and GDP, for example, makes no sense economically or statistically.
What is more relevant, however, is the conclusion that Subramanian should have arrived at but didn’t. He rightly concludes that India’s GDP has been overestimated, but blames that on flawed methodology. But what should be of more profound concern is the deteriorating trend in all the major economic indicators post 2011. This trend is a statistical fact independent of any methodological problem in estimating GDP. In an article titled “India’s Wasted Decade” (https://contractwithindia.com/indias-wasted-decade), I examine a wide range of economic and human development indicators and find clear evidence of deterioration in all these indicators in the post-2011 period. It suggests that India’s original liberalisation reforms of 1991 have run their course and that it is time for Economic Reforms 2.0.
Subramanian has boldly put aside his partisan leanings and stated the obvious—India’s economic growth is moribund. Many of us have said this in opinion pieces, but coming from someone who was in the driver’s seat, should be a wake-up call. Modi would be well advised to bring together experts from around the world to develop a new generation of reforms based on less government interference, free markets, increased privatisation of production resources, modernisation of capital markets, less regulation and greater economic freedom. The economy is too complex a system to rely on raw wisdom.
—The writer is a financial economist and founder, contractwithindia.com
You might also be interested in: