GDP and consumption growth did very well from FY12 to FY18; GDP growth slowed beyond expectations starting late 2018. Let us not mix and match the two events.
If politics is mixed with economy, and the two are combined with ideology, then what we have is what can mildly be called a “potent” mixture. That is the case today with discussions about the Indian economy a day before the 71st anniversary of the Republic.
Let us discuss some known facts, at least as we see them. As the reader will soon observe, we belong to the camp which believes that GDP growth was 7% for the first five Modi years (2014-2018); and we believe that GDP growth has fallen woefully short of predictions, and expectations, and analysis, to be only 5% in FY20. This is an unprecedented fall in a non-crisis year, and all the experts who were shouting from the rooftops for a tight monetary policy, and a tight fiscal policy, should soul-search for where they went wrong—and believe us they did, all the way from international organisations to domestic banks and institutions.
While this soul-searching is going on, we want to address the confusion that has erupted because of the steep slide in GDP growth in FY20. Many commentators have conflated the two distinct episodes and are presenting it as one. Their contention is that the recent slowdown in growth has been a permanent fixture of the Indian economy, especially since PM Modi took over the reins in May 2014. You have heard experts say that India is already back to its pre-liberalisation era of slow 3-5% GDP growth. A twist is added by some to say that we are back to Raj Krishna’s quip about the Hindu rate of economic growth of 3-4%—Hindu rate of growth, BJP’s hindutva, PM Modi, the CAA protests, etc—you get the drift.
Now, couple all of this with the fact that the Modi government has dumped the NSS consumer expenditure survey for FY18. Incidentally, we are in full agreement with the government’s decision to treat the written report with the disdain it deserves—but we are NOT in agreement with not releasing the unit-level data. As we have mentioned several times before, the unit-level data must be released, if only to document the reality that there is something seriously wrong with the NSS architecture for collecting data.
As big data analysts, we do believe that truth is hidden somewhere in the muddled data, and that the responsibility of a good researcher is to extract the information from the noise. Towards this, we bring a diverse set of statistical facts to the table—from night-lights data to national accounts data to data brought out by organisations, e.g., TV sales, internet subscriptions, automobile sales, airline passengers etc. (Our forthcoming paper—Poverty and Cash Transfers: The way forward for India contains a detailed account of data sources and implications for assessments about growth and poverty in India.)
Is 2019 different from 2014-2018? A resounding yes, according to the night-lights data. This data is available with the Earth Observation Group at NOAA. It comprises of low-light imaging data collected by satellite and filtered to measure the quantity of artificial (i.e., human-generated) light in an area. Night-lights data enables us to capture the economic activity between 2012 and 2017. The first four years, for the first four months (2019 night-lights data is only available till April), show an average growth of 12.5% per year; GDP growth in the first two quarters (months January to June) from 2014 to 2018 averages 7.1% annually. In the first two quarters of calendar 2019, GDP growth averaged 5.4%, and night-lights averaged 5.8% growth. Both very distinct measures of economic activity suggest that 2019 was a bad year, and that the prior years had significantly higher economic activity.
Some political economy experts believe that demonetisation is responsible for the 2019 slowdown while some others believe that the GDP data are flawed because such data do not adequately capture activity in the informal sector. The months immediately following demonetisation (November, December, and January) do show an average growth of minus 20%; the next 16 months averaged 23%. This strongly indicates that something happened to cause growth to sharply decelerate in late 2018, but it was not demonetisation or the GST. We believe the evidence is supportive of the fact that the ultra-tight monetary policy followed by the Patel-Acharya duo at RBI, and the NBFCs crisis (itself exaggerated by tight monetary policy) are factors that precipitated our recent growth slowdown.
Was 2017-18 a very bad year for economic activity? The Consumer Expenditure Survey of the NSSO takes a snapshot of economic activity in a particular agricultural year (July to June), and computes a level estimate of consumption in that year. When that estimate is compared to 2011-12, the NSSO finds that real per capita consumption declined by 3.7% over a span of six years. This is almost unheard of in most mature economies; even in our worst GDP growth year in recent memory (1991 or 2009), per capita GDP growth declined by 0.5% (1991) or increased by 2.5% (2008).
We strongly believe that misdiagnosis of what happened in 2017-18, or in 2019, can lead to faulty policy prescriptions. As it happens, 2017-18 was a very good year for economic activity. The night-lights data consistently finds 2017-18 to be the best economic activity year in the last seven years.
We have attempted to construct an alternate consumption series for 2017-18 using various sources of data. Rice, wheat, and cereal production (accounting for about 8% of the consumption basket in 2011-12) does show an average per capita annual decline of 0.02% (i.e., flat consumption) between 2011-2017. But, there is an inversion of let them eat cakes in the aam aurat basket—woman does not live by bread alone; per capita consumption of meat, eggs, pulses, edible oils, and even turmeric shows a healthy per capita increase between 2011 and 2017. The food and tobacco consumption components (accounting for 42% of the basket) show an annual real per capita increase of 2.6% over 2011-12.
But, this is just the beginning of the increase between 2011-12 and 2017-18. In the aggregate, for an additional 27% of consumption expenditures (e.g., TV, internet, laptops, petrol, electricity, education, health, airline travel, etc) there is an annual 5.7% increase in per capita consumption. Overall, and we have assembled data for about 75% of expenditures, we obtain a 3.5% annual increase in per capita consumption (this contrasts with about a 4% annual increase according to national accounts).
It is unlikely that the addition of 25% of consumption items we have missed (appliances, toiletries, restaurant and hotel, and delivery consumption, etc) would have grown at less than the average rate of growth of the included 75%). At the end, we just might be reaching the conclusion that the national account estimates of consumption and GDP growth have been fairly accurate. No doubt, some political economists will contest some of our conclusions. But, let the debate be mature and based on verifiable facts.
Our analysis has implications for poverty analysis, and analysis of policies that the Modi government has followed. Our very conservative estimate is that absolute poverty in 2017-18 (not before, not after) in India was in the low-to-mid single digits, a decline of some 6 to 8 percentage points from the 14% level achieved in 2011-12. This is indeed a happy note as we celebrate 70 years of the creation of the Republic of India.
Bhalla is Executive director, IMF & Bhasin is an independent economist. Views are personal