GST re-labelling misleads and obscures the true revenue picture

GST's tax relabelling has blurred revenue trends, altered Centre-state fund sharing and raised fresh questions about transparency and fiscal federalism

“What’s in a name? That which we call a rose by any other name would smell as sweet.” Alas, Shakespeare’s Juliet was not right about names. Consider as an example the saga of the Goods and Services Tax (GST). 

Since September 2025, parts of the GST system have changed their names, with two real consequences. Observers of revenue performance have been misled, since headline numbers for the newly defined GST suggest that receipts are booming, even as overall performance has declined sharply. And cooperative federalism has suffered, as states are receiving a smaller share of the overall GST kitty, amounting to between ₹15,000 and 20,000 crore annually. Consider each in turn. 

Until September 2025, the GST system comprised two types of taxes: regular “GST” (hereafter designated with inverted commas) and the Compensation Cess. The two kitties were created by the GST Council because revenues from them would be deployed in different ways. But from the perspective of the consumer, the distinction was moot: they were both just indirect taxes that consumers paid on the goods and services they bought. 

In September 2025, the rate structure was simplified. Many goods moved from higher “GST” slabs to lower slabs, while SUVs and other large vehicles moved in the other direction from the 28 per cent slab to 40 per cent because the Compensation Cess on vehicles was abolished. But cesses continued on tobacco and pan masala. 

In February/March 2026, the Compensation Cess on tobacco and pan masala was abolished. These products largely moved to the 40 per cent tax slab (except beedis, which moved to the 18 per cent slab). These products that used to be subject to the Compensation Cess were now subject to a different tax called the Additional Excise Duty (AED). 

In effect, after September 2025, the labels were changed so that the system that was “GST” plus a wide-ranging Compensation Cess became one comprising a much bigger “GST” along with a smaller Additional Excise Duty. 

These changes have created confusion, as post-reform “GST” figures are not comparable to pre-reform ones. Yet most commentators continue to focus on “GST” because the main source of reporting, namely the GSTN, doesn’t publish data on the AED, which is levied by the central government. Nor does the central government budget explicitly provide this data. It is consequently easy for analysts to miss this tax altogether. 

So, if one wants to understand the revenue performance of the system as a whole (and compare this performance over time) one must look at taxes collected under all the labels. Figure 1 plots the two series. 

Following the September 2025 and February/March 2026 changes to the GST structure, revenues under the “GST” label decelerated to 7.5 per cent in FY26 but have since surged to 10.5 per cent in the first quarter of FY27. This has been touted as evidence of an immaculate (Laffer Curve) outcome: tax rates are cut but revenue soars. 

But this is misleading for two reasons. First, the “GST” rate on SUVs and tobacco products has been increased from 28 per cent to 40 per cent. More important, because it excludes revenues from the Compensation Cess pre-reform and from the AED post-reform. 

When the total GST system is considered, revenue growth declined sharply, from 9.4 per cent in FY25 to 5.8 per cent in FY26 even though the cuts were operational only for about half the fiscal year. And in the first quarter of FY27, revenue growth of the system declined further to 4.9 per cent. The implication is clear: the changes may have improved the system but they did lead to revenue losses. 

Before coming to a conclusion, however, an additional correction needs to be made. We need to adjust GST collections for refunds, which have been substantial, since it is net revenues that matter for government finances. Then, to make comparisons more meaningful, we need to express these net revenues as a per cent of GDP.   

The results are shown in Figure 2. Clearly, there have been improvements in GST administration because both collections and net revenues have picked up since the early years of the new system. But more recent trends are worrisome. Since FY24, net GST revenues have declined, especially following the recent rate cuts. In FY26, net revenues as a share of GDP were just 5.7 per cent, which is about 0.5 percentage points below the pre-GST average 

So much for the fiscal consequences. What were the implications for cooperative federalism? Although the law allows the centre to levy taxes on tobacco products, states expected that all the taxes under the Compensation Cess would be folded into the regular GST structure. Under this outcome, the states would get the (former) Compensation Cess revenue in two ways: 50 per cent of the total directly, as State GST; and 40 per cent of the Central GST indirectly, because the portion accruing to the Centre would merge into the divisible pool, which is shared according to the Finance Commission formula. 

The re-labelling has changed all that, as Ashok Bhattacharya of Business Standard has pointed out. Under the new arrangements, the Additional Excise Duty goes to the Centre, forming part of the divisible pool to be shared with the states. Meanwhile, a new tobacco machinery-related cess (the Health Security and National Security Cess), which would seem to violate the GST law de jure will also not be shared with the states. Our rough estimate is that re-labelling will deprive states out of ₹15,000- 20,000 crore, which will instead be taken by the Centre. 

Unfortunately, names matter. And the Centre gets to choose names. Merely by calling a GST tax a “duty”, it has altered the flow of GST revenue, obscured the performance of the system, and changed fiscal federalism. These changes have been almost invisible. The job of analysts has consequently become more difficult. But it has also become even more important.

 

( Abhishek Anand is a visiting fellow, Madras Institute of Development Studies; Josh Felman is principal, JH Consulting; and Arvind Subramanian is senior fellow, Peterson Institute for International Economics )