1. Five years of state-wise GST, and the year it broke pattern

Domestic GST collection grew every year from FY2021-22 through FY2024-25 — then fell in FY2025-26, the first year-on-year decline in this window. National domestic GST collection: ₹10.98 lakh crore (2021-22) → ₹13.25 lakh crore (2022-23, +20.7%) → ₹15.23 lakh crore (2023-24, +14.9%) → ₹16.76 lakh crore (2024-25, +10.0%) → ₹16.32 lakh crore (2025-26, −2.6%). Growth decelerated every single year even before the reversal, so the FY25-26 dip is the continuation of a slowing trend, not a sudden shock — but it is still a reversal, not just slower growth.

All figures ₹ crore, domestic GST collection only (excludes import IGST). Source: GSTN Statistics portal, "Collections-Statewise" workbooks, FY2021-22 through FY2025-26. Note: Uttar Pradesh is the only state in this top-10 list whose FY2025-26 collection fell below its FY2023-24 level. Every other state stayed above its FY2023-24 level, but six of the remaining nine (Karnataka, Gujarat, Tamil Nadu, West Bengal, Telangana, Rajasthan) still dipped below their own FY2024-25 peak — only Maharashtra, Haryana and Delhi grew or held flat both years running.

Nine of the 24 states and union territories for which mid-year figures are separately available saw a real contraction in the April–November 2025 period compared to the same months a year earlier — Tamil Nadu (−2.3%), Kerala (−4.1%), Odisha (−7.4%) and Madhya Pradesh (−2.7%) among the larger states. The proximate cause is well documented: GST 2.0, the rate-rationalisation that collapsed India's multi-slab structure into two broad rates (5% and 18%, plus a 40% rate on select items), took effect 22 September 2025. Net GST collections fell 4.25% in November 2025 alone — the first full month after the change — against the same month a year earlier. Import-linked GST collection tells a different story entirely: it rose 11.7% over the same year, from ₹5.33 lakh crore to ₹5.95 lakh crore, which is why India's combined gross GST growth (5.57% for the year) looks merely slow rather than negative — the domestic decline was real, and imports masked it in the headline number.

2. FY2025-26 in full: collection, settlement, and each state's effective take

GST is collected in three components: CGST and SGST (split evenly between Centre and the state where the transaction is registered) and IGST (collected on inter-state and import transactions, then settled to the destination state under the consumption principle). Because IGST settlement flows to whichever state actually consumes the good or service, a state's "effective" GST revenue — SGST collected plus its IGST settlement share — is a better proxy for what a state government actually gets to spend than SGST alone.

SGST and IGST-settlement figures are FY2025-26 (up to March 2026), from GSTN's "Collections-Statewise" and "Settlement of IGST to States/UTs" workbooks respectively. Note the reordering versus raw collection: Uttar Pradesh overtakes Gujarat once IGST settlement is added, because UP is a large consuming state even though it isn't among the largest GST-collecting states — the destination-based settlement mechanism is specifically designed to correct for exactly this gap.

3. What comes back: tax devolution runs on a completely different formula

Tax devolution — the states' constitutionally mandated share of the Centre's divisible tax pool (41% under the 15th Finance Commission) — is distributed by a formula weighted on population, income distance (favouring lower-income states), area, forest cover and demographic performance. It has nothing to do with where GST is actually collected. On 1 August 2026 the Union Government released an advance instalment of ₹1,09,019 crore, on top of the regular monthly transfer, with the following state-wise breakdown:

Devolution figures: PIB, "Union Government releases tax devolution of ₹1,09,019 crore to State Governments," 1 August 2026 (one advance instalment, not the full-year devolution total). GST share is each state's share of the matched-state effective-GST total (Section 2). Ratio >1× = the state's devolution share exceeds its GST-contribution share; <1× = the reverse.

Read the ratio column plainly: Haryana, Maharashtra, Gujarat, Tamil Nadu and Karnataka each get back a devolution share that is roughly a third to a half of their GST-contribution share — independent reporting frames this as these states "recovering roughly 35 to 55 paise for every rupee of contribution capacity," and separately confirms Maharashtra's 6.44% devolution share against its much larger economic footprint, which matches this piece's own figure exactly. Uttar Pradesh and Bihar sit at the other end — UP's 17.62% devolution share in this instalment is close to its formula-fixed 17.93–17.96% share under the 15th Finance Commission's 2021-26 award, more than double its 8.72% share of effective GST revenue.

Why some states get a bigger share: the constitutional mechanics

Devolution and IGST settlement are not the same constitutional mechanism, even though this piece has discussed both together. IGST settlement (Section 2) flows from Article 269A, added by the 101st Constitutional Amendment (2016) that enabled GST: inter-state IGST is levied and collected by the Union, then apportioned to states "in the manner as may be provided by Parliament by law" on the recommendation of the GST Council — a destination-based mechanical correction, not a redistributive one. It sends revenue to whichever state actually consumed the good or service, nothing more.

Tax devolution (this section) runs on entirely different constitutional footing. Article 270 requires that taxes in the Union List — income tax, corporation tax, Union excise, and since the GST amendment, CGST too — be "levied and collected by the Government of India" but "distributed between the Union and the States" rather than kept entirely by the Centre. Article 280 is what actually decides the numbers: it requires the President to constitute a Finance Commission every five years, whose job is to recommend both the vertical split (Centre's share vs. states' combined share — currently 41%, set by the 15th Finance Commission for 2021-26) and the horizontal split among the states themselves.

Here is the part that surprises most readers: the Constitution does not specify the horizontal formula. Article 280 does not say "weight by population" or "favour lower-income states" anywhere in its text — it simply mandates that a Finance Commission be appointed and make recommendations. The actual formula (population, income distance, area, forest cover, demographic performance, and the exact weight given to each) is decided fresh by each Finance Commission as a policy judgement, not fixed constitutional law, and it has changed from one Commission to the next. What Article 280 guarantees is a periodic, independent, formula-based recommendation process — not any particular formula. The 15th Finance Commission's specific choice to weight income distance heavily (which is what pushes UP and Bihar's shares well above their GST-contribution shares) reflects that Commission's judgement about equalising fiscal capacity across states with very different tax bases, consistent with the redistributive intent Finance Commissions have applied since the First Finance Commission in 1951 — but it is a policy choice made under the Constitution's framework, not a rule written into the Constitution itself.

4. National devolution totals, for context

Union Budget documents put devolution to states at approximately ₹12.47 lakh crore (Budget Estimate, FY2024-25, later revised ₹39,674 crore higher), ₹14.22 lakh crore (Budget Estimate, FY2025-26, +10.5% over the FY24-25 revised figure), and ₹15.26 lakh crore (Budget Estimate, FY2026-27, +9.6% further). These are Budget Estimates, not final actuals, and this piece could not locate a single clean, consistently-sourced actuals series back to FY2021-22 to match against the five years of GST data above — that comparison is flagged as a gap rather than papered over with a constructed number.

Related on this blog

See also: India's Two Tracks of Health Cover, and Where They Actually Meet

Sources

• GSTN Statistics portal, gst.gov.in/download/gststatistics — "Collections-Statewise" workbooks, FY2021-22 through FY2025-26, and "Settlement of IGST to States/UTs," FY2025-26
• Department of Revenue, Ministry of Finance — departmental structure and links to GST/state-tax statistics sources used in this piece
• Constitution of India, Articles 270, 269A and 280, and the 101st Constitutional Amendment Act, 2016 — constitutional basis for tax devolution, IGST apportionment, and the Finance Commission respectively
• PIB, "Union Government releases tax devolution of ₹1,09,019 crore to State Governments," 1 August 2026 — state-wise devolution instalment table
• Business Standard, "Like-for-like GST growth declines to five-year low of 5.57% in FY26," and "Tax turbulence: Revenues take a hit in first full month of GST 2.0" — national and post-GST-2.0 collection trend, independently confirming this piece's own computed figures
• The Secretariat, "India's 41% Tax Devolution Faces Political Test as States Ask What They Get Back" — independent confirmation of Maharashtra's 6.44% devolution share and the "35–55 paise per rupee" framing
• Union Budget documents (Budget Estimates, various years), as referenced in PRS India budget analyses — national devolution totals cited in Section 4

This analysis is based on official GSTN and PIB data as cited above. Figures marked as a single instalment (Section 3) are not full-year totals and should not be read as such; the national devolution figures in Section 4 are Budget Estimates, not audited actuals. GST collection figures reflect domestic collection only unless stated otherwise. This piece does not take a position on what the "correct" devolution share should be — that is a live political and economic policy debate, not a factual question this data can settle.

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.