‘V’​ Shaped Recovery for all? An Analysis of State-Level GST Collection Data

Photo by Engin Akyurt on Pexels (https://www.pexels.com/@enginakyurt)


On the 24th of March 2020, the Government of India declared a nation-wide ‘lockdown’ to counter the spread of the novel coronavirus (SARS-CoV-2). The lockdown brought much of production and consumption to a virtual stop, triggering a stark contraction in the economy. The outbreak of the COVID-19 pandemic and the subsequent social distancing measures have sent many leading global economies, including India, into a dizzying tailspin.

The second advance estimate of the Ministry of Statistics and Programme Implementation indicates an eight per cent contraction in GDP in the 2020–2021 financial year. While there are a number of studies looking into the extent of economic recovery at a national level, very few studies provide a state-level perspective. In this article, I leverage the availability of data on Goods and Services Tax (GST) collection from Indian states for the period between August 2017 to December 2020, to analyse how the various states fared during the peak of the COVID-19 pandemic. Data on GST collections, albeit not perfect, are very good proxies that indicate the extent of consumption recovery in the Indian economy.

The Goods and Services Tax (Advantages and Disadvantages)

The GST replaced a great swathe of indirect taxes (such as the excise tax, service tax and Value-Added Tax (VAT)), and was one of the largest streamlining of India’s taxation system in recent history. The tax was enforced from the 1st of July 2017, after a great deal of deliberation among the various states and the centre.

GST is a ‘destination tax’, meaning that the tax for the goods/service accrues at the point of consumption. For example, if a product manufactured in Maharashtra, were to be consumed by an end-user in Tamil Nadu, the tax collected for the good would be split between the Central government and the government of Tamil Nadu. This implies, that the amount of collections reported by each state is a strong indicator of the extent of consumption activity in the state.

It is important to bear in mind that a wide variety of essential goods and commodities such as fresh food produce and public transport are exempt from GST. This implies that GST collections would be more ‘elastic’ to changes in income, than actual consumption.

To explain the concept of ‘elasticity’, let us take a simple example. Suppose, household A consumes two commodities, food and movies. When there is a sudden fall in the household’s income, the household would drastically reduce the amount it spends on movies as it is a non-essential good, while the change would not be as pronounced for food commodities. Since GST is charged on movies and not on essential food products such as fresh vegetables and meat, there would be a sharper fall in GST collections (in terms of percentage) than the overall reduction in consumption. Hence, changes in GST collections tend to overstate the extent of reduction in consumption. On the flip side, increased collections would mean that the consumption is actually increasing at a slightly higher rate than indicated by the GST collection trends (at least theoretically speaking).

Despite the potential bias, GST collections do provide us with a good indication of the health of the economy and its trajectory.

National Level Trends

At the national level, total GST collection plummeted in the month of April to around ₹19,602 crores (1 Crore = 10 Million) from around ₹78,694 in the month of March.

[Note: Total GST collection includes Central GST, State/UT GST, Integrated GST, GST on online services (such as Netflix), and cess levied on special goods.]

To understand the extent of the fall in GST revenue I also plot the average of the monthly collections across 2017 (from August), 2018, and 2019. The average collection in the month of April was around ₹85,517 crores in 2018 and 2019. April 2020 collections, hence, witnessed a mammoth 77 per cent decline from the previous years’ average collections. However, at the national level, the GST collections quickly rose within a couple of months, and by December 2020 had substantially exceeded past collections. The December 2020 collections according to the publicly available dataset is ₹87,146 crores. (Note: The collection estimates provided in the publicly available dataset do not match estimates provided in some other reports from the Ministry of Finance. It could be that the MoF has access to more updated data than that is publicly available.)

This recovery in GST collections has been widely cited to support the argument about the ‘V’ shaped recovery of the Indian economy.

The State-Level Trends

Before analysing state-level trends, it would be helpful to understand the relative contribution of states to the total GST collections. This would help us gain a better understanding of the relative economic position of states. It would also help to understand the per-capita collections in each state.

In terms of absolute GST collections, Maharashtra, Karnataka, Gujarat, Tamil Nadu, Uttar Pradesh, and Haryana lead, with each state collecting over ₹53,000 crores in 2020.

[Note: Total GST Collections is the total of IGST, SGST/UTGST, CGST, and cess on special goods collected in each state or Union Territory (UT).]

When adjusting for population, smaller states and UTs such as Dadra and Nagar Haveli, Sikkim, Goa, and Delhi lead, raising over ₹19,000 per person in GST (and cess). It is important to note that most of these smaller states (excluding perhaps Delhi) primarily rely on tourism and that the high GST collections per-capita is more because of tourists than resident consumption.

Population data sources: Census, UIDAI.

How has the pandemic impacted the various states of India?

To understand the impact of the pandemic, I look at the percentage decline in revenue from March 2020 to April 2020 (Figure 4). I also look at the average month-on-month change in GST collections in each state between May 2020 and December 2020 (Figure 5). States such as Ladakh (which reported no collections in the month of April), Jharkhand, Daman and Diu, Dadra and Nagar Haveli, Meghalaya, Chhattisgarh, Rajasthan, West Bengal, Uttar Pradesh, and Kerala saw a substantial reduction in GST collections. All the aforementioned states had over an 81 per cent decline in collections between March and April of 2020.

Some may argue that using the March 2020 collections as the basis for computing the decline in April 2020 is inaccurate. One may argue that since the March collections themselves were muted due to the ‘Janata Lockdown’ and other preventive measures that preceded the national lockdown, using March as the basis would understate the true decline in collections. Likewise, some could argue that the M-o-M average gain (Figure 5) is overstated due to the ‘base effect’, as April 2020 collections were abnormally low.

To tackle both these concerns, I estimate the change in 2020 GST collections for each month from the average of the 2019 and 2018 collections in the same month. For example, I compare the deviation in March 2020 GST collections from the average of the GST collections in March 2019 and March 2018. I compute the monthly deviation from past years’ average for each state. The light blue band in Figure 6 indicates the interval between the monthly deviation estimate of the state in the 2.5th percentile to the state in the 97.5th percentile of the monthly deviation distribution across states.

The variation in the growth/decline from past years’ average is quite large across states, specifically in the months of May and onwards. This gives us an indication that not all states experienced the fabled ‘V’ shaped recovery, and that several states were left behind.

To study the variation in recovery further, I construct an estimate that captures the deviation in aggregate GST collections between May and December of 2020, from the average aggregate collections between the months of May and December in 2018 and 2019. States that are reliant on tourism such as Daman and Diu, Lakshadweep, and Goa are some of the worst-hit and are yet to recover from the impact of the pandemic. Goa, for example, has seen a decline of close to 20 per cent in aggregate GST collections between May and December 2020, when compared to the previous years’ average. On the other hand, GST collections of states like Tamil Nadu and Maharashtra are almost on par with the previous years. Some states such as Madhya Pradesh, Rajasthan and Kerala have collected more than the previous years’ average collections between May and December, which is quite impressive.

The two-letter abbreviations of the states are the same as the ones used in registration plates, with the exception of Telangana, where, I use ‘TG’ instead of ‘TS’.

Unlike the post-April collection estimates, the decline in collections in April 2020 from the average collections in April 2018 and 2019 were not very heterogeneous among states. All but four states saw a decline of 60 per cent or above.

The biggest takeaway is that while all states were almost equally affected by the pandemic, not all states have been able to fully recover from it, with many, specifically tourism-dependent states left behind. The ‘V’ shaped recovery story, maybe a reality to some, but not for all.

How do we boost the economies of the states that are left behind? Would a massive infrastructure and spending boost have a multiplier effect that can restart the economy, specifically in the trailing states? The immediate future of the Indian economy lies in how we tackle these challenges ahead of us.


There is correlational evidence supporting the claim that historically high-output states are struggling to recover from the impact of the pandemic. There is a negative correlation (a Pearson’s correlation coefficient of -0.37) between a state’s historical output (measured by the per-capita nominal state domestic product in 2018–19) and its GST collections recovery in 2020 (measured by the deviation of aggregate GST collections between May and December of 2020 from previous years’ average). Note that the evidence is purely correlational and not causational.

NSDP data is unavailable for Andaman and Nicobar Islands, Dadra and Nagar Haveli, Daman and Diu, and Lakshadweep. Hence these UT’s have been excluded from this correlational plot. The UTs of Jammu and Kashmir and Ladakh are aggregated as one, to ensure comparability with 2018 and 2019 data (pre-abrogation of Article 370).

The shapefiles for the maps are from Prince Nihith

Author: Akshay Natteri Mangadu




Applied Econometrician

Love podcasts or audiobooks? Learn on the go with our new app.

Recommended from Medium

Back to School in the Age of Covid-19 & More.

Unfair B2B contracts — JJ Richards decision gives small businesses a glimmer of hope

MIT World Real Estate Forum: Bottom-Up Real Estate

Inclusive Development Index: Measuring What Matters

What you need to know before the Union Budget 2021!

Annual Financial Report of the Government of Canada Fiscal Year 2016–2017

Saved inflation

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store
Akshay Natteri

Akshay Natteri

Applied Econometrician

More from Medium

The Analysis: Ease of Doing Business by World Development Indicators

Pocock Blocked: Interim Checks of Experiment Design, Part 1

Railroad accidents 2010–2021 data analysis

Predicting Energy Consumption of Electric Vehicles

Image Source: https://www.homeandsmart.de/bmw-i3-elektroauto-vergleich