How much tax are you drinking? — data & sources

What we took: every volumetric rate on the article, in dollars per litre of alcohol (LAL), from the column effective 3 August 2026. The ATO publishes both the current and the forthcoming column ahead of each indexation; we build on the forthcoming one and show the superseded February column only as the "what it was" comparison.

The 3 August 2026 column reflects a CPI indexation factor of 1.020. Draught subitems 1.2, 1.6 and 1.11 are identical in both columns — that is the freeze, visible in the ATO's own table.

What we took: the beer formula and the wording of the 1.15% threshold — "excise duty on beer is payable on the alcohol content above 1.15% by volume" — plus the definition of draught by container rather than by tap: a container over 48 litres, or between 8 and 48 litres and designed to connect to a pressurised or pump delivery system. That definition is why a keg and a stubby of identical beer sit in different tax brackets.

What we took: the 29% rate, unchanged since 1 July 2000 and never indexed; the half retail price method used to translate a shelf price into a WET figure; and the worked example we reproduce verbatim on the article — $120 wine, WET $34.80, WET-inclusive value $154.80, GST $15.48. That example is the ATO's own arithmetic and it is the cleanest available proof that GST is charged on top of the wine tax.

What we took: the provenance and the duration of the draught beer freeze — two years from 1 August 2025, with indexation resuming in August 2027. Confirmation that the February 2026 indexation did not apply to draught is visible in Source 1: subitems 1.2, 1.6 and 1.11 carry identical rates in both columns.

What we took: the 1 August 2006 rates that anchor the twenty-year comparison (packaged full-strength beer $37.90, packaged low $32.52, draught full $26.68, draught mid $20.39, spirits $64.21, brandy $59.94, RTDs $37.90); the 27 April 2008 alcopops step from $39.36 to $66.67 per LAL; and the evidence that the rate can pause but not fall — packaged full-strength beer held at $51.31 from 3 February 2020 through 1 February 2021, skipping the August 2020 indexation after CPI fell in the June 2020 quarter, then resuming from the old high rather than a lower base.

What we took: beer excise 2,729 · spirits 3,320 · other excisable beverages 1,789 · WET (net) 1,134 for 2024–25 actual, giving a total alcohol tax take of $8,972m; the forward estimates to 2029–30; and the tobacco excise line used for the alcohol-versus-tobacco comparison (2024–25: $7,767m; 2026–27 forecast: $3,560m).

What we took: beer 2,599 · spirits 3,208 · other excisable beverages 1,686 · WET 1,091 = $8,584m for 2023–24, the year that matches the latest AIHW consumption data.

What we took: the beverage shares of national pure alcohol available — wine 42%, beer 32%, spirits 23%, cider 3% — which is what makes "wine supplies the most alcohol and pays the least tax" a sourced statement rather than an opinion. Also 217.1 ML of pure alcohol and 9.8 L per capita (15+) for 2023–24, the largest year-on-year fall since the series began in 1960–61.

What we took: the definition of a standard drink — 10 g of alcohol, equal to 12.5 mL of pure alcohol — which is the denominator for every per-standard-drink figure on the article. Linked from the article as the neutral reference on standard drinks; we make no health claims of our own.

What we took: the 2014–15 effective-rate spread of $2.99 per LAL on cask wine to $79.38 on spirits and RTDs, with packaged full-strength beer at $36.08 and a weighted average of $36.50 — the last time anyone in government published this comparison; the origin of the brandy concession on 9 November 1979; and the finding that many small wine producers pay no net WET. This is the independent, non-industry benchmark for the anomaly, and it is eleven years old, which is itself part of the story.

What we took: corroboration of the anomaly — "the tax on medium- and full-strength packaged beer is now half as high again as that on even super-premium still wine, and it is nearly twelve times that on non-premium (cask) wine" — and the properly-hedged international statement we use in place of the industry's ranking claim: Australia "taxes its alcohol consumption more than most other affluent economies apart from Finland and Norway."

Funding disclosure, because it matters: Professor Anderson's work in this area has been supported by Wine Australia, the University of Adelaide and the Brewers Association of Australia — that is, by two of the industries whose tax treatment he is comparing. We use his figures because they are transparently derived and independently checkable, and we cross-check the anomaly against the PBO (Source 10), which has no industry funding. We are telling you about the funding because a page that asks you to check its working should not hide the provenance of its corroboration.

What we took: the recommendation that all alcohol be taxed volumetrically on a common basis — that is, that the wine/beer split this article describes be abolished. Recommendation 71 is sixteen years old and has not been implemented, which is the single most useful piece of context for anyone writing about the anomaly as though it were newly discovered.

What we took: nothing numerical. This is the page most commonly quoted in Australian coverage of beer tax — the source of the widely-repeated "42% of a carton", "almost $20 in tax per slab" and "fourth highest in the world" lines — and it publishes no methodology and no per-figure sourcing. We reference it because our independently computed carton figure lands within a whisker of their 42% claim, which is worth saying plainly: on the carton number they appear to be right, and we can show why, which they cannot. Their international ranking claim we do not repeat; we use Anderson's hedged formulation instead (Source 11).

Why this exists. Australia has published no official retail price series for alcohol since the ABS stopped collecting public-bar beer prices in June 2011. There is no successor series for packaged beer, wine or spirits either. A vendor's advertised price is checkable; a crowd-sourced average is not. So we capture, name and date.

What you should know about the sample, stated plainly. Dan Murphy's, BWS, Liquorland, First Choice, Vintage Cellars, Woolworths and Coles all block automated retrieval — 403 responses or bot-walls. Beer and mass-market wine prices now come from the major chains, captured by hand. Because those chains block automated retrieval, the prices in the first table below were read from Google Shopping listings on 31 July 2026, each showing the named retailer’s advertised price. We are precise about that wording deliberately: we saw a Google Shopping listing carrying that retailer’s advertised price. We did not fetch it from the retailer’s own website, and no page of ours says or implies we did. The second table is the original sample — smaller named Australian liquor retailers that publish retrievable prices. These are genuine advertised prices at genuine shops, but they are not the major chains and nothing on our pages implies they are. One retailer we did reach, Bottle Stop (formerly BoozeBud), was excluded from every average: it listed a wine that three other retailers sell for $9.00–$11.99 at $47.99, and ran 15–40% above the field on beer and spirits. That is a data-quality failure, not a price, and we would rather drop it and tell you than average it in quietly.

Packs we excluded, and why. Averaging across pack specifications is the fastest way to a wrong excise figure, because excise is charged on the litres of alcohol in the pack. Four founder-supplied listings were dropped for that reason: a Carlton Dry 4 × 6 × 330 mL at $53.95 (330 mL, not 375 — different volume, different excise), a Great Northern listing at $44.95 (a different product in the range, not Super Crisp), an XXXX Gold 30-pack at $52.90 and a case listing at $47.99. Only like-for-like 24 × 375 mL packs are averaged. Where one retailer listed bottles and another cans of the same beer at the same volume — Coopers Original Pale Ale — we did average them and the table says which is which.

Beer strengths are the brewer’s own, not a retailer’s. Every ABV in the beer rows was read from the brewery or brand site on 31 July 2026: Victoria Bitter 4.9% (Asahi Beverages media release, 27 February 2026 — victoriabitter.com.au publishes no strength for full-strength VB at all, and VB moved 4.9% → 4.8% in 2007 → 4.6% in 2009 → 4.9% in 2012, so any pre-2012 secondary source will hand you a wrong number); Coopers Original Pale Ale 4.5% (coopers.com.au/beer/original-pale-ale); Carlton Dry 4.5% (carltondry.com.au/our-beers — not the 3.5% on that site’s homepage, which is Carlton Dry Mid, a separate SKU); Great Northern Super Crisp 3.5% (greatnorthern.com.au/our-beers, where the range also lists Original 4.2%, Light 2.0% and Zero 0.0%); XXXX Gold 3.5% (xxxx.com.au). Each back-solves against the standard drinks printed on the pack. The sensitivity is not academic: Carlton Dry at the homepage’s 3.5% would understate the carton excise by $5.85, an error of 30%.

Wine ABVs vary within a brand, so we do not average them. [yellow tail] runs from about 11% (Pinot Grigio — the producer’s own store says 11.5%, retailers say 11.0% with 6.5 standard drinks, and 6.5 requires 11.0%, so we carry it as a range) through 12.0% (Jammy Red Roo) to 13.5% (Shiraz, verified on Casella’s own store). Jacob’s Creek Classic runs 12.2% (Sauvignon Blanc) to 13.9% (Pinot Noir); the Classic Shiraz we could not settle — the listing states 13.0% ABV and 8.2 standard drinks, which cannot both be true, since 8.2 requires 13.9% — so no per-standard-drink figure is published for it. 19 Crimes Red Blend is 13.5% and the Cali tier 14.0%. Because WET is a share of price, none of this changes the tax; it changes only the tax per standard drink, which is why the mass-market wine table carries no per-drink column and the per-drink figures we do quote name a single varietal.

Ownership, for completeness. Accolade Wines and Pernod Ricard’s wine business merged in 2025 to form Vinarchy, so Hardys and Jacob’s Creek are now the same company. [yellow tail] is Casella Family Brands; 19 Crimes is Treasury Wine Estates.

ABV notes. No retailer showed an ABV for the Wolf Blass Chardonnay, so the article uses 12.5% as a class assumption and labels those figures illustrative. Retailers disagreed on the Smirnoff Red listing — three said 40%, one said 37% — and the Australian product is 37%, which is what we used. The disagreement is immaterial to the headline: excise per standard drink on any spirit is the rate divided by 80, $1.38, whatever the strength.

Prices we could not capture. No 4 L cask at the $15 the older literature assumes — the floor we found was $19.00, and the article says so rather than carrying a stale price. No wine at the $30 price point; the wine we tried came in around $23, so that point is simply absent from the article. No gin from a retailer we trust, so the spirits figures use vodka and whisky instead. The nip of spirits, the two wine-by-the-glass prices and the pint and middy starting values are assumptions or volume-scalings, labelled as such on the article and never presented as observations.

What we took: the top five — 1 Great Northern Super Crisp · 2 Carlton Dry · 3 XXXX Gold · 4 Coopers Pale Ale · 5 Victoria Bitter — which is the product list for section 04 of the article. The two published renderings match position for position, one to ten (6 Hahn SuperDry, 7 Corona Extra, 8 Carlton Draught, 9 Tooheys, 10 Asahi). Great Northern Super Crisp has led for five consecutive years; the run starts in 2021, not 2020, as the 2022 edition of the same list makes clear by calling it “second year in a row”.

Three limits we state on the article rather than bury here. First, it measures off-premise retail — bottle shops. On-premise and tap volume data is limited and is not in it; the same report named Stone & Wood Pacific Ale the country’s top tap beer. Second, the 2025 edition does not state its metric, and the metric decides the answer: on Circana’s own retail liquor data ranked by dollar value rather than units (Drinks Trade × Circana, 28 June 2024, for calendar 2023), the leader was Hard Rated with Great Northern second. Only the 2023 edition of the News Corp list was explicitly volume-based, so we do not assert “by volume” for 2025. Third, one of the three commercial contributors of sales data is Asahi, which owns Great Northern. We use the list because it is the best public ranking that exists and because two outlets rendered it identically — not because it is audited.

What we would not use. We could not locate a 7NEWS article on this ranking and so do not cite one. We also rejected a widely-mirrored rewrite at australiatimes.com: it is dated 31 July 2026, self-declares that it was machine-generated, carries no attribution, and repeats the “since 2020” error. An AI rewrite of a real story is not a second source.

What we took: Victoria Bitter 4.9% (Asahi Beverages, 27 February 2026 — “VB Classic remains at 4.9% ABV with 1.4 standard drinks”); Coopers Original Pale Ale 4.5% (Coopers’ own spec table, 375 mL = 1.3 standard drinks); Carlton Dry 4.5%; Great Northern Super Crisp 3.5%; XXXX Gold 3.5%. Each was cross-checked by back-solving the standard drinks printed on the pack at 12.5 mL of pure alcohol per standard drink. Two traps are recorded on the article because they are live wrong-number risks for anyone checking us: carltondry.com.au’s homepage leads with 3.5%, which is Carlton Dry Mid rather than the flagship, and victoriabitter.com.au publishes no ABV for full-strength VB, so the only figure on that site is 3.5% for VB Mid. Untappd, BeerAdvocate and Wikipedia were returned repeatedly by search and contributed nothing to any figure here.

What we took: the Australian Hotels Association's recommended schooner price of $9.50 for VB, XXXX Gold and Carlton Draught, up from $9.10. This is the same figure, under the same "media-reported, not official statistics" label, that our Wages through the Ages page carries. Two caveats we could not close: the reporting names no AHA state branch, and schooner sizes differ between states — 425 mL in most, reportedly 285 mL in South Australia — so the figure supports a range rather than standing as a hard national price. The April corroboration confirms the 40-cent rise but quotes no absolute price. The upgrade path remains a dated price board photographed at a named venue.

What we took, page by page, because this is the section’s entire authority:

• p.442 — Recommendation 71, quoted verbatim on the article: “All alcoholic beverages should be taxed on a volumetric basis, which, over time, should converge to a single rate, with a low-alcohol threshold introduced for all products. The rate of alcohol tax should be based on evidence of the net marginal spillover cost of alcohol.”
• p.441 — which rate. “As the current full-strength packaged beer excise rate is closest to the estimates of the average spillover cost of alcohol in Australia, it would be appropriate for rates of tax on other products to converge at this rate over time.” This is why we model $65.03 and not a number of our own choosing.
• p.440 — the effective-rate ratios we reproduce. Spirits then faced effective rates about 2.2 times those on full-strength packaged beer, and about 1.3 times under a common rate. On the 2026 rates, comparing a 40% spirit with a 4.8% packaged beer, we get 2.23× and 1.28×.
• Footnote 18 — what a rate may be set on. The tax-relevant spillover cost excludes “private intangible costs (such as pain and suffering)”. That exclusion is why we qualify the $66.8bn social-cost figure rather than repeating it as a tax argument.
• p.436 — the description of the two-system split the article’s section 02 sets out.

The mechanic most write-ups get wrong, and it is on this page so you can check it: Recommendation 71 extends the 1.15% low-alcohol threshold to every product. It does not abolish it. That is why a single rate per litre of alcohol does not produce identical tax per standard drink — the fixed threshold is proportionally far larger in a weak drink than a strong one.

What we took: two things. First, the modelled scenario, in Treasury’s own words — the common rate “is assumed to be the current full strength packaged beer rate, including the 1.15 per cent alcohol excise exemption threshold”. That single sentence settles both the rate and the threshold treatment, and it is why our model is a reconstruction of a government costing rather than an invention. Second, the revenue estimates: a gain in 2025–26 of $1.33 billion under Option 1 (a fifteen-year transition) or $1.40 billion under Option 2 (five years). The caveat travels with the figure every time: these are 2009 projections built off a 1 July 2010 start date and the forward estimates of that year. They are a counterfactual, not a forecast of 2026, and the article says so wherever the number appears.

What we took: that the 2010 government did not adopt Recommendation 71 — the Office of Impact Analysis’s list of the seven Henry measures taken up does not include alcohol; that Re:think reproduced Treasury’s own chart of tax per standard drink and its worked example of a ginger beer taxed at $21.54 as a pre-mixed drink or $8.70 as a wine, and then asked no discussion question about alcohol at all, with the Tax White Paper it fed never published; and that the Senate Select Committee on Red Tape recommended a single volumetric rate again in 2017 as its Recommendation 1. What we did not use: the widely-repeated 2010 “wine glut” refusal quote. Its primary source URL is dead and we could not stand it up, so the article states the non-implementation as fact without it. A quote we cannot retrieve is a quote we do not print.

As a counterweight: S3 modelled roughly 14,900 full-time-equivalent jobs lost across the economy and a 6.7% fall in alcohol consumption. Those are their figures from their model. Ours has no behavioural response at all, so no employment, consumption or health number on the article is ever attributed to our arithmetic.

As corroboration: their S3 gives regular keg beer +3.69% where our arithmetic gives +3.7%, and a fall on premium bottled wine where we also get a fall. A CGE model built for a health foundation in 2015 and a spreadsheet built for a calculator site in 2026 land in the same place, which is the most useful check available on work this simple.

Where we differ we are the conservative one, and we say so rather than let it look like softening. Their cask figure is about +212%; ours is +152.3%. The difference is not about the tax — it is that our captured 2026 cask price of $21.00 is high relative to the tax on it, so the same dollar increase reads as a smaller percentage.

What we did not take: their revenue-neutral rate of $14.08 per litre of alcohol. It is wine only and it is 2015. There is no published revenue-neutral rate across all Australian alcohol; we looked, it does not exist, and we are not going to index a partial eleven-year-old figure forward and present it as one.

What we took: the academic case against a single rate, which is stronger than the industry version and which the article publishes for that reason. The authors find that beverage type predicts risky drinking behaviour independently of alcohol content — which, if it holds, means a single volumetric rate is not automatically the optimal one, because harm is not a pure function of ethanol. We also carry their other finding, which cuts the other way: cask wine is “significantly undertaxed relative to the damage it does”. Both are theirs. We do not adjudicate between them.

What we took: the $66.8 billion annual social cost figure — and, inseparably from it, the fact that $48.6 billion of that is the intangible component. That distinction is the whole reason we cite it. Henry’s footnote 18 says a tax rate should be set on spillover cost and excludes private intangible costs such as pain and suffering, so the most-quoted number in Australian alcohol policy is routinely used to make an argument its own framework says it cannot make. We take no position on how much anyone should drink; we are separating a rate-setting argument from a harm-accounting one.

What we took: the 2025 entry in the article’s short history of why Recommendation 71 never happened — the concessions grew rather than the base changing — and the Prime Minister’s characterisation of the freeze quoted in section 06.