What GST 2.0 actually did in September 2025

India's GST Council, at its 56th meeting on 3 September 2025, approved what's been widely reported as a collapse from four tax slabs to two. The Central Board of Indirect Taxes and Customs issued the implementing notifications on 17 September, and the change took effect 22 September 2025 for services and for nearly all goods. The popular description — "5% and 18%, plus 40% on a few sin items" — is a reasonable shorthand, but it understates how many effective rates actually remain. There is still a 0% exempt category covering an expanded list of essentials (unbranded food grains, fresh produce, milk, life-saving medicines, individual life insurance); a 5% merit rate that absorbed most of what used to sit at 12%; an 18% standard rate that absorbed most of what used to sit at 18% and 28%; and a 40% rate — not a return of the old Compensation Cess, but a replacement for it — covering pan masala, cigarettes, aerated drinks, luxury cars and motorcycles above certain engine sizes, yachts, personal aircraft and real-money online gaming. On top of that, gold, silver, platinum, pearls, jewellery, imitation jewellery and coins stay at a separate 3% rate, rough diamonds and rough precious stones at 0.25%, and cut and polished diamonds at 1.5% (the central-tax halves of 1.5%, 0.125% and 0.75% in Notification No. 9/2025-Central Tax (Rate), G.S.R. 641(E), 17 September 2025, in force from 22 September 2025 (Gazette of India Extraordinary CG-DL-E-17092025-266209)), and tobacco products stay on the old rate-plus-cess regime for now, specifically because the Centre's compensation-cess-backed borrowing has to be repaid before that transition can complete. Four effective rates, three narrow legacy carve-outs, and one deliberately deferred category is not the same thing as two rates — and it is a long way from one.

The countries usually cited as "flat rate" aren't, either

"Singapore just has one GST rate" is the comparison that usually follows any discussion of simplifying India's slabs, with the UAE's 5% VAT often invoked alongside it. Both statements are true only in a narrow sense: each has one standard positive rate, and each also has a defined, real set of exemptions and zero-ratings sitting outside it.

New Zealand and Chile (a flat 19% VAT with no reduced-rate band) are the two systems most often held up internationally as the "cleanest" broad-based consumption taxes, and even they carry real carve-outs — residential rents and financial services for New Zealand; education, freight, some real estate and financial services for Chile. Exports are zero-rated everywhere as a matter of basic VAT design, not policy generosity; that's not really a carve-out so much as how a destination-based consumption tax has to work. But the exemptions on food, health, education and financial services are policy choices, made everywhere on this list, including in the Gulf states most often cited as India's simpler alternative. "One rate" internationally means one standard positive rate applied to most transactions, not a tax that touches every rupee, dirham or dollar of spending identically.

India already ran this experiment on paper, in 2015 — and said no

A committee chaired by then-Chief Economic Adviser Arvind Subramanian was asked, in June 2015, to calculate the Revenue Neutral Rate for the GST that was still being designed at the time — the single rate that, applied uniformly, would leave combined central and state tax revenue unchanged from the pre-GST system. Its December 2015 report calculated an RNR of 15 to 15.5 per cent, with a preference for the lower end. It then explicitly did not recommend adopting that single number as the actual applied rate. Instead it proposed a merit rate of roughly 12 per cent for essential goods, a standard rate of 17 to 18 per cent for most goods and services, and a demerit rate of around 40 per cent for luxury cars, aerated drinks, pan masala and tobacco — a tiered structure built explicitly to avoid taxing a poor household's food and a rich household's luxury car at the same rate. The 40 per cent demerit-rate figure the committee proposed in 2015 is strikingly close to the 40 per cent rate GST 2.0 actually adopted for sin and luxury goods a decade later — a rare case of an academic committee's specific number resurfacing almost unchanged in the eventual policy, even as the broader structure evolved around it.

Recent, on-record Indian commentary specifically on whether the country should go further, to a genuine single rate, is thin but not absent. A tax expert quoted by Business Today immediately after the September 2025 Council meeting put it directly: India's income levels don't yet justify a single rate, because a single rate would tax a poorer household's consumption basket at the same rate as everyone else's; a single rate is described there as a plausible long-run destination once compliance and per-capita income both rise further, not a near-term option. This piece did not find a broader survey of Indian economists, or a dedicated study from a body like NIPFP or ICRIER, modelling a full single-rate GST for India in the 2024-26 period — the debate available in the public record is about timing and feasibility, not a live proposal any GST Council member has put on the table.

Why the regressivity problem doesn't go away just because a rate is "flat"

The economic case against collapsing to one rate is not really about tax administration — it's distributional, and it's well documented outside India too. An OECD working paper on VAT regressivity found that consumption taxes are close to proportional, or even mildly progressive, largely because of the reduced rates and exemptions most countries apply to food and other necessities; strip those out in favour of one broad-based rate, and some regressivity reappears, because poorer households spend a larger share of their income on the exact things that used to carry a lower rate. This is precisely the logic behind Singapore's, the UAE's and the Gulf states' zero-rating of food, health and education — not generosity, but the standard fix for a tax that would otherwise land harder on people who spend most of what they earn.

There is a second option, and it's worth naming because it's the one actual alternative to rate-tiering that shows up in the recent international literature: a 2024 IMF working paper on "designing a progressive VAT" argues for a genuinely single-rate, broad-base VAT paired with real-time cash rebates to low-income households at the point of purchase, rather than differentiated rates, as the mechanism for protecting the poor. No country on the comparison list above actually runs this model at scale yet, and this piece found no indication India's GST Council has discussed it, but it is the honest answer to "couldn't you just have one rate and fix the poverty problem separately" — yes, in principle, with a delivery mechanism India hasn't built.

The more defensible structural difference is scale and political architecture, not a disputed inequality statistic. India's population is roughly 1.46 billion against Singapore's roughly 6.1 million and the UAE's roughly 11 million — India is about 133 times larger than either. And India's GST rate decisions run through the GST Council, where the Constitution's Article 279A gives the Centre one-third of the weighted vote and all states combined two-thirds, with any decision requiring three-quarters of the weighted vote — a multi-party bargaining structure in which states surrendered real fiscal autonomy in exchange for compensation-cess guarantees. Singapore's GST rate is set by a unitary Parliament with no equivalent negotiation; the UAE's VAT is set federally among seven emirates without India's kind of directly-elected, revenue-sharing state governments holding independent fiscal stakes in the outcome. A genuine single-rate GST for India would not just be a rate change — it would require reopening exactly the political settlement the GST Council was built to manage.

What would fix this

• The GST Council or Finance Ministry should commission and publish an updated, public Revenue-Neutral-Rate study for the post-GST-2.0 rate structure, the way the 2015 Subramanian committee did for the original design, so any future single-rate debate has real numbers behind it rather than a Singapore comparison that doesn't survive a rate-schedule check.
• If a single-rate proposal is ever seriously tabled, it should be evaluated alongside the IMF's point-of-sale-rebate alternative, not assumed to require rate-tiering as the only way to protect low-income households — India has the digital payments infrastructure (UPI) that a real-time rebate mechanism would need, even if it hasn't been built for this purpose.
• CBIC should publish a single, plain-language, consolidated statement of every current GST rate — including the gold, jewellery and rough-diamond carve-outs and the tobacco transition timeline — rather than leaving the exact current rate schedule to be pieced together from tax-advisory blog posts, which is what this piece and most public commentary currently has to do.
• The tobacco/pan-masala transition to the new rate structure should get a public completion date, tied to the actual compensation-cess loan repayment schedule, so "GST 2.0 is done" has a real endpoint rather than an indefinitely deferred category.

These are proposals from this blog, not recommendations any agency has adopted.

Sources and caveats

About this article: Researched, written and edited by Umashankar Triplicane Dwarakanathan, with AI research assistance; every figure is meant to trace to the primary source cited. See the Editorial Policy for how sourcing, AI use and corrections work.