Duty-Recorded Tobacco in Sri Lanka: Production, Taxation and the Limits of Industry Claims

Recorded cigarette output and excise moved in opposite directions

Treasury’s provisional 2025 cigarette-production figures show a slight rise, while recorded cigarette excise receipts fell. Treasury’s Final Budget Position Report 2025 reports a provisional 1,935 million sticks, against 1,923 million in 2024. Its revenue accounts record Rs 101.173 billion in cigarette excise, down from Rs 117.145 billion. The resulting 0.62% production increase is provisional and coexists with a 13.63% fall in cash receipts.

That divergence is the central analytical problem, not evidence that one measure must be wrong. Production, taxable releases, inventories, product mix and cash collection occupy different points in the system. The inspected sources do not supply the full bridge. They cannot support a precise illicit substitution estimate or a claim that a tax change caused the difference.

Evidence checked through: 20 October 2026. Coverage is limited: duty-recorded products, fiscal accounts, dated issuer disclosures, operative product restrictions and survey demand context are examined. National retail value, representative margins, company shares and a current illicit-market share remain unavailable.

The boundary separates cigarettes, beedi and prohibited products

The principal market boundary is licensed and duty-recorded manufactured tobacco products released for Sri Lankan domestic consumption. Production is a separate supply indicator, not a substitute for domestic releases. Manufactured cigarettes and beedi have different tax treatments. Smokeless products and newer products enter the analysis as observed uses or regulatory alternatives; they are not silently added to the legal cigarette denominator. Export movements and imported supply require separate treatment from domestic manufacture.

The 2016 prohibited-products regulation covers manufacture, import, sale and offering for sale of smokeless tobacco products or mixtures containing tobacco, specified flavoured, coloured or sweetened cigarettes, and electronic cigarettes containing tobacco. That wording does not by itself determine every enforcement interpretation of tobacco-free nicotine devices or heated products. A survey response about use is not proof of lawful supply.

This audit is for understanding recorded supply, taxation and market constraints. It does not convert health exposure into a sales opportunity or evaluate a company’s investment merits. The original June 2025 publication and August 2026 correction remain part of its history.

The fiscal total reconciles; the supply-to-tax bridge remains open

Source: Treasury Final Budget Position Report 2025, Table 4.8 (printed p. 86) and detailed revenue accounts. Table 4.8 supplies the cigarette quantities without a provisional marker; Figure 4.3 on printed p. 85 labels the 2025 cigarette-production and excise series provisional. The 2025 production level and its calculated growth therefore retain that qualification. Table 4.7 on the same page also marks the 2025 excise aggregate provisional. Changes use (later / earlier − 1) × 100; quantities are reported rounded, receipts are calculated from the rupee amounts. The summed codes reproduce the published rounded cigarettes/tobacco aggregate. Customs’ 2024 table independently reproduces the cigarette-revenue code. None of these amounts is retail sales, VAT-inclusive consumer spending or a total of all taxes associated with tobacco.

The product distinction is consequential. The separate Tobacco Tax is not a second measure of the cigarette excise series. Nor can its receipts simply be divided by the current beedi rate to obtain a national production total: rate changes, payment timing, assessments and other accounting differences must first be reconciled.

The 2026 Budget report records a dated cigarette excise schedule as at 31 October 2025: Rs 19,350, 50,150, 71,450, 81,000 and 90,050 per 1,000 sticks for successive length bands of up to 60 mm, over 60–67 mm, over 67–72 mm, over 72–84 mm and over 84 mm. These are specific excise amounts. Without the relevant product prices, VAT and levy bases, they cannot be converted into a single retail tax percentage.

Parliament’s July 2025 account identifies Gazette 2430/16 of 1 April 2025, raising beedi tax from Rs 2 to Rs 3 a stick effective 2 April 2025. The later Final Budget report’s Box 1.3 gives 2 April 2026, but its measures appendix and the parliamentary account identify 2025. The conflicting date is preserved, not silently adopted.

The January–August 2026 Fiscal Review groups collection by agency as well as tax. Its rounded Customs chart includes cigarettes, while the Excise Department chart includes a separate tobacco component. Those pie percentages cannot provide an exact updated cigarette tax total. The latest complete reconciled annual comparison here therefore remains 2025.

Issuer accounts explain one business, not the entire market

Ceylon Tobacco Company PLC supplies a dated example of cigarette manufacturing and sales. Its 2025 annual report, Note 4, exposes the difference between turnover and revenue after government levies.

The company’s value-added statement allocates Rs 173.832525 billion to the state. That amount equals levies of Rs 149.106655 billion plus income-tax expense of Rs 24.725870 billion. It is therefore an accrual-based allocation combining tax categories, not an excise-only cash receipt.

CTC’s excise accrual exceeds Treasury’s 2025 cigarette receipts by Rs 10.825049 billion. In 2024 the comparison runs the other way: Treasury receipts exceed CTC’s reported excise by Rs 10.491917 billion. Different recognition, settlement, product and entity boundaries must be bridged before attribution. The change in sign is counterevidence to treating the residual as a stable competitor share or illicit-tax gap.

The unaudited results issued on 13 August 2026 add useful counterevidence to a simple growth narrative. First-half net revenue rose from Rs 31.128 billion to Rs 32.635 billion, 4.84%, but second-quarter net revenue fell from Rs 16.980 billion to Rs 16.565 billion, 2.44%. Half-year profit rose 2.49%, while net operating cash flow fell from Rs 11.658 billion to Rs 7.299 billion, 37.39%. Timing and working-capital effects matter; profit and cash are not interchangeable. None of these company results measures representative outcomes for manufacturers, farmers, distributors, workers or retailers.

Product categories and routes matter more than an unsupported share

A physical route links tobacco inputs, manufacture, packaging and fiscal release to distributors, retailers and users. Imports and unrecorded supply create additional routes, but their current quantities and transaction economics are not established here. Tobacco-company activity reports can describe their own supply and distribution relationships; they cannot establish a national outlet census or the bargaining power of every supplier and retailer.

Beedi illustrates the danger of equating administrative records with a complete industry population. At the July 2025 parliamentary meeting, officials reported 1,140 licences in 2024 and 840 renewals with 300 still pending in 2025. Pending renewal is not proof that 300 businesses closed. The same discussion attributed revenue weakness to smuggling, while committee members questioned the tax response. Those are competing institutional explanations, not a causal study.

Cigarettes, beedi, prohibited smokeless products and other nicotine products cannot be ranked on a common legal-supply denominator. GATS records overlapping use, and the tax system distinguishes products. These facts support a qualitative account of alternatives and access constraints. They do not establish substitution elasticities, a market concentration ratio or numerical competitive-force ratings. The original unsupported company-share claim remains excluded.

National survey evidence supplies a historical demand basis

The GATS 2020 fact sheet, re-inspected for this edition, reports a national household survey of people aged 15+, with fieldwork from December 2019 to July 2020. It completed 6,770 interviews. Current use includes daily and less-than-daily use.

Categories overlap. These are weighted prevalence estimates, not purchasers, legal sales, sticks or expenditure. The very low female-smoking estimate also requires caution about small numbers and self-report. Among current smokers, 51.6% planned to quit or were thinking about quitting; the existence of quit interest does not establish successful cessation or adequate service access.

A user may buy individual units, share purchases or use multiple products. The evidence supports distinguishing users, payers, retailers and health-service providers, but does not provide a current national purchasing journey or spending basket. Price sensitivity cannot be estimated from one historical price observation and a later tax receipt.

The earlier edition’s STEPS 2021 and GSHS 2024 references retain their separate populations and access dates. STEPS covers adults aged 18–69; GSHS covers enrolled students in grades 8–12. Neither is a new observation of the GATS population. The school survey’s reported e-cigarette use is a prevention and enforcement signal, not evidence of a lawful product market.

Rules constrain both traditional products and transitions

The NATA Act restricts sales to people under 21 and regulates advertising, sponsorship, free distribution and smoking in enclosed public places, with specified exceptions. Product restrictions and fiscal schedules are distinct controls: payment of a tax does not override a prohibition, and observation of use does not show enforcement success.

Treasury, Customs and Excise shape fiscal administration; NATA and the health authorities shape prevention and product controls; Parliament scrutinises implementation. Consumers, workers and farming communities have interests not captured by issuer profits. The earlier edition’s WHO FCTC Article 5.3 guidance remains relevant to transparent, necessary interaction with regulated firms and protection of policy from vested interests.

Technology and product-transition claims require evidence at the point of use. A company investment in machinery does not establish sector-wide productivity; a newer nicotine product is not automatically legally supplied or less harmful. The present evidence supports monitoring product identification, enforceable classification, traceability and survey exposure. It does not support an adoption forecast or a recommendation to substitute one product for another.

Illicit-trade methods answer different questions

The WHO South-East Asia review, retained from the August edition, distinguishes a 2018 study in six purposively selected districts from a 2020 tax-gap model. The first used discarded butts, empty packs and test purchases; those units and the non-probability geography prevent a national-share interpretation. Reported availability is a further, different measure.

The tax-gap exercise compared modelled revenue due with recorded revenue. Its approximately 0.2% result depends on consumption self-report, assumed prices and a generalised tax rate; it is not an estimate that 0.2% of cigarettes were illicit. Seizure counts measure detected activity and change with targeting and enforcement effort as well as underlying supply.

CTC’s 2025 annual report describes an estimated 0.5 billion illicit sticks. The inspected discussion does not provide a reproducible sampling frame, product-identification protocol and uncertainty calculation. This remains an interested-party estimate. It is not added to Treasury production to manufacture a national denominator. The WHO method discussion retains the August edition’s source-access date and its historical observation periods.

Interpretation: the existence of unrecorded supply is compatible with several different quantities. To narrow that uncertainty, independently assessable sampling and administrative reconciliation are needed. Neither a company claim, a tax shortfall nor a selected enforcement operation supplies the missing national measure.

A bounded outlook follows the unresolved mechanisms

As assessed on 20 October 2026, this outlook covers the period to 20 April 2027, with earlier reassessment on material rule or evidence changes.

The current evidence supports a mixed baseline: recorded cigarette output stabilised in 2025, cash cigarette excise fell, separate Tobacco Tax rose, and one issuer’s 2026 half-year and quarter moved differently. It does not support a single growth or decline forecast. If a later tier-specific release and settlement account explains the divergence, fiscal interpretation should change accordingly. If a new national survey shows lower comparable use, that strengthens a demand conclusion that production alone cannot establish.

A source-backed opportunity is better fiscal and product transparency: it can reduce reliance on interested-party estimates and expose classification or collection problems. Material threats to a reliable reading include stale consumption evidence, inconsistent licence counts, unreconciled cash and accrual measures, and missing independent illicit sampling. These are analytical and administrative conditions, not evidence that every firm faces the same commercial result.

Requirements for lawful participation and credible measurement

Three requirements follow from the product, fiscal and channel findings. Correct classification must connect product composition and cigarette length with legal status, labelling and duty; the observable test is consistency across product, release and tax records. Traceable distribution must distinguish licensed or otherwise lawful supply from detected unrecorded products; the test is an auditable chain of origin and sale, not a claimed outlet count. Financial reconciliation must separate turnover, levies, income-tax expense and cash settlement; the test is a bridge that explains timing and perimeter differences.

These are analytical critical requirements for the defined recorded system. They are not assertions about any company’s capabilities. A changed product rule, independently verified supply-chain account or new fiscal reconciliation would alter their application. No country ranking, corporate diversification programme, investment recommendation or numerical attractiveness score is carried forward.

Coverage, responsibility and future evidence

Production and fiscal quantities are supported more strongly than current national customer behaviour, channel economics or illicit volume. The limited scope excludes full retail valuation, shares, representative profitability and causal tax effects. The 2025 issuer accounts and the 2026 interim do not repair those national gaps; historical health-burden models retained in the source history are not current observed outcomes.

The next checks are the next complete fiscal release, product and length-tier quantities, any new duty or product notifications, beedi licence reconciliation, the next CTC filing, and new national survey or independently documented illicit research. Updated access dates must not be mistaken for updated observations. A future accounting bridge should resolve the current residuals explicitly, preserving the old edition and its corrections.

The original publication date, migration provenance and historical links remain intact. The August correction’s removal of unsupported market sizes, company shares, competitive scores and corporate strategy remains in force. This edition adds inspectable supply and accounting analysis while keeping the conclusions within the evidence.