Australia loses more money gambling per person than almost any country on Earth. Around $1,500 per adult every year. The system is not broken. It is working exactly as designed.
Australia’s total gambling losses have grown from under $1 billion in 1975 to over $32 billion in 2023–24. The introduction of poker machines outside casinos in the 1990s drove an explosion, and losses have barely paused since. Toggle between a state breakdown and the national total line below.
Source: QGSO, Australian Gambling Statistics, 40th edition, 2025. Values are nominal dollars which are not adjusted for inflation. Hover over any year for exact figures. WA’s modest share reflects its ban on gaming machines outside Crown Casino.
NSW residents lost $2,007 per person in 2023–24 which is 32% above the national average. The red bar tells the story: gaming machines. Western Australia, which bans pokies outside one casino, clocks just $896. The difference is not culture. It is policy.
Source: QGSO, Australian Gambling Statistics, 40th edition, 2025, Summary Table E. Per capita figures are for the adult population. WA’s gaming machines figure is $0 by law and only one casino is permitted. NT figures reflect a very small population base.
For decades, poker machines were Australia’s defining gambling problem. Something changed around 2015. As states quietly tightened pokies regulations, the sports betting industry,now entirely online exploded. Wagering’s share of household income has risen 60% since its 2013–14 low point, driven by smartphone apps, in-play betting, and wall-to-wall odds discussion on sports broadcasts.
Source: QGSO, Australian Gambling Statistics, 40th edition, 2025, Product Tables: Gaming Machines Sheet 9 & Wagering Sheet 9. Household income benchmarks from ABS. COVID-19 temporarily closed venues in 2019–20; online wagering surged as physical gambling was disrupted.
NSW residents devote roughly 1.6% of household income to poker machines which is almost double Victoria’s share and triple South Australia’s. Use the dropdown to switch between pokies and online betting to see how state patterns differ across product types. WA’s pokies line is zero throughout.
Source: QGSO, Australian Gambling Statistics, 40th edition, 2025. NT excluded from wagering comparisons due to population-size distortions (single TAB operator, very small denominator). WA gaming machines are zero by state law.
Each bubble is a state. The horizontal axis shows what share of household income goes to poker machines; the vertical axis shows wagering (online betting). Bubble size reflects the combined gambling burden. Press Play to watch 26 years of change: pokies losses drift left as regulation tightens, while wagering bubbles float upward, the fingerprint of the mobile betting revolution. WA’s bubble hugs the left edge throughout: no pokies, only wagering.
Source: QGSO, Australian Gambling Statistics, 40th edition, 2025. X-axis = gaming machine losses as % of household income; Y-axis = wagering losses as % of income. Bubble size = combined pokies + wagering burden. WA has no licensed gaming machines outside Crown Casino (x ≈ 0 throughout). NT excluded due to wagering data anomalies from single-operator distortion.
Press Play to watch each state’s total gambling losses grow in dollar terms since 1998–99. NSW has led the entire time. Queensland accelerated from 2020 as online wagering surged. Victoria’s dip in 2020–21 captures Melbourne’s extended COVID lockdowns, then a sharp rebound once venues reopened.
Source: QGSO, Australian Gambling Statistics, 40th edition, 2025. Values are nominal (not inflation-adjusted). NT excluded due to data anomalies in early years. Victoria’s 2020–21 dip reflects extended COVID-19 lockdowns; Queensland’s post-2020 surge reflects online betting growth.
These five charts tell a story of a country caught inside a system that profits from addiction. Australia’s gambling losses are not a personal failing, they are the predictable outcome of deliberate policy choices: 87,000 poker machines in NSW clubs and pubs, a $4-billion-a-year revenue stream for state governments, and the deregulation of online sports betting that put a casino in every smartphone.
NSW’s dominance is not accidental. Its poker machines are protected by one of the most powerful lobbying forces in the country. The clubs channel some winnings back into communities, creating a dependency that makes reform politically costly.
Western Australia’s story is the counterfactual. By banning gaming machines outside a single Crown Casino, WA residents lose barely half what NSW residents lose per head. The experiment works. It simply requires political will.
Meanwhile, the wagering surge reveals a new frontier. The next generation of problem gamblers may not be the retiree feeding notes into a machine in Wollongong, it may be the 22-year-old watching the NRL with a live-betting app, placing in-play bets between quarters.
The house always wins. The question is who pays.
References
Queensland Government Statistician’s Office, Queensland Treasury. (2025). Australian Gambling Statistics, 40th edition, 1998–99 to 2023–24. Queensland Government. https://www.qgso.qld.gov.au/issues/2646/australian-gambling-statistics-40th-edn-1998-99-2023-24.pdf
Queensland Government Statistician’s Office, Queensland Treasury. (2025). Total gambling expenditure (nominal value) by state and territory, 1975–76 to 2023–24. Queensland Government. https://www.qgso.qld.gov.au/statistics/theme/society/gambling/australian-gambling-statistics
Australian Bureau of Statistics. (2023). Household income and wealth, Australia, 2021–22 (Cat. No. 6523.0). ABS. https://www.abs.gov.au/statistics/economy/finance/household-income-and-wealth-australia/latest-release
Markham, F., & Young, M. (2015). ‘Big Gambling’: The rise of the global industry-state gambling complex. Addiction Research & Theory, 23(1), 1–4. https://www.tandfonline.com/doi/full/10.3109/16066359.2014.929118
Young, M., & Markham, F. (2017). Coercive commodities and the political economy of involuntary consumption: The case of Australia’s gambling industries. Environment and Planning A, 49(12), 2762–2779. https://journals.sagepub.com/doi/10.1177/0308518X17734546
Generative AI Acknowledgement
Claude (Anthropic) assisted with R/plotly code development, data structure inspection, and chart syntax. All narrative framing, analytical interpretation, and editorial decisions were made by me. Data sourced directly from QGSO open data under Creative Commons Attribution 4.0 International licence.