It’s been almost two years since the transition to the GST and the system is beginning to stabilise – tax rates have broadly remained unchanged for several months now and compliance procedures too have become simpler. We can now start to use and interpret the GST data – although the government is releasing very limited data. One of the advantages of GST is that the tax collections closely reflect underlying final sales in the economy and are thus a very important gauge of the economic momentum. Prior to GST, the excise and service tax collections were much poorer at this because a large part (more than 50 percent) of excise collections were specific duties levied on petroleum and tobacco. In contrast, all the GST (barring some items of the compensation cess) is an ad-valorem levy covering both goods and services. And the value-added nature of the tax means that it is effectively a tax on end consumption and thus tax collections reflect final sales. To be fair, there are limitations in this data too – the GST base is still narrow as I discuss later, there are still tax rate changes which distort YoY comparisons, there are different rates which bring in the mix effect and we are still relatively new into the GST transition. But despite these limitations, there isn’t any other cross-sector near-real-time data on the health of the economy.
So, at the risk of over-interpreting data, here is my conjecture of what the data of the last couple of months is telling us:
Not all GST collections are alike: Another point about the GST collections data is that while most analysts focus on the overall GST collections, which treats all components at the same level, there is a big difference in the interpretation of its different components. On one hand is CGST and SGST collections, which reflect end sales (a proxy for retail sales) while the IGST collections reflect only inter-state sales and thus largely capture B2B or ‘supply chain’ movements. This could be either import of goods or supply of goods from a centralised manufacturing plant to distribution centres across the country. The compensation cess, on the other hand, captures only a narrow section of the economy. Growth still seems to be reasonable: Total CGST and SGST collections have increased around 12.5 percent YoY during March and April (months of tax accrual but paid in April and May respectively). This is either inline or slightly higher than nominal GDP growth. Also worth noting is that inflation, measured either by the CPI or WPI, is running at 3-4 percent. This either implies that the underlying ‘real’ growth in the economy (at least the sectors that GST is levied on) is a healthy 8-9 percent or that there is a material improvement in compliance due to pressure on tax collections and thus tax evasion has reduced. The latter though seems unlikely given that the tax department gets into overdrive to collect revenues only towards the end of the year and these are the first two months of the year (the government accounts for tax collections on cash and not accrual basis) and also that these were the two months before elections. Further, there have been some tax rate reductions in the past 12 months, most notably on several consumer durable goods. So, this 12.5 percent nominal growth in revenues (and therefore the implied ‘real’ growth) is, if at all, a modest understatement of what like-to-like revenue growth would have been but for the rate reductions. Yes, this contradicts with data from sectors such as car sales. But from the perspective of the overall economy, car sales are a relatively small sector. Ballpark calculations suggest car sales will account for less than 10 percent of CGST/SGST collections. The slowdown in ‘supply-chain’: Where the GST data does suggest a slowdown is in the IGST collections which have grown just 5 percent over the recent two months. This is much below nominal GDP growth and only modestly above inflation implying that in real terms there is no ‘supply chain’ revenue growth, possibly an inventory adjustment. So, the GST data seems to suggest that what is happening in the economy currently is more a supply chain slowdown or adjustment even as end sales growth still seems reasonable. Widening of the tax base: An average of 72 lakh GSTR 3B returns were filed for the months of March and April 2019 (the returns being filed in April and May respectively) on a provisional basis. This is an increase of 18 percent on a YoY basis on the number of returns being filed. Given that the increase in tax revenues is less than this, this implies that the average tax paid per taxpayer has fallen. While this suggests a fall in compliance, what is more likely to have happened is that a lot of smaller businesses which were either outside the tax net or were not GST compliant are now filing returns. These smaller businesses would make a very small addition to the tax pool (due to their size) and hence their addition brings down the average tax paid per taxpayer. But their being in the GST net is essential for the inbuilt anti-tax avoidance mechanism of GST to work. The wider the GST tax base, the lesser the scope for tax evasion. The tax base is still narrow: One of the premises behind the GST was that a low rate of GST cast over a very wide tax base will reduce tax evasion thus increasing the government’s ability to invest through higher revenues and at the same time drive productivity and growth by reducing distortions in the economy. This objective of the GST has sadly not been met and the low (value) base of GST exemplifies this. The total GST collections excluding the cess have averaged $11bn in the last few months and if we assume a blended average rate of about 15 percent on this, the value of taxable turnover being reported is a little less than $900bn. This is less than a third of the GDP – the remaining is a combination of (mostly) exemptions and some evasion. And it is exemptions that to an extent aid evasion by cutting off the chain of input credit. And this is contrary to the basic premise behind the GST. Ashutosh Datar is a Mumbai-based independent economist. Read Ashutosh Datar's columns