How many people in New Zealand are harmed by gambling, and how much do New Zealanders lose? These gambling harm statistics for New Zealand rest on two official sources. According to the New Zealand Gambling Survey 2023/24, 8.3% of people aged 15 and over (95% confidence interval 7.4% to 9.3%), about 359,000 people, were at some level of risk of gambling harm in the previous 12 months. For 2.4% (1.9% to 3.1%), about 104,000 people, the risk was in the survey’s “moderate-risk and problem” band. The Department of Internal Affairs (DIA) reports that gamblers lost $2.79 billion in 2023/24 across the four main types of gambling.
If these numbers describe your own situation or that of someone close to you, the Gambling Helpline is free on 0800 654 655 or by text on 8006. For support and tools, see gambling harm and support in New Zealand.
How is gambling harm measured in New Zealand?
Most figures come from the New Zealand Gambling Survey, a partnership between the Ministry of Health, which leads the project, and Health New Zealand | Te Whatu Ora, which analyses the results and publishes them in Kupe, an online results tool. According to the Ministry of Health, over 7,000 people aged 15 and over took part between late 2023 and early 2024. The data therefore describe 2023/24.
Harm is measured with the Problem Gambling Severity Index (PGSI), developed by Ferris and Wynne in 2001. It asks nine questions about the previous 12 months, such as “How often have you bet more than you could really afford to lose?” Answers score from 0 (never) to 3 (almost always), so totals run from 0 to 27. Kupe groups the totals into non-problem (0), low-risk (1 to 2), moderate-risk (3 to 7) and “problem gambling” (8 to 27).
The PGSI is a screening score based on what people say about themselves, not a diagnosis. The “any-risk” figure combines scores of 1 to 27. Kupe publishes the top two bands together, so a separate figure for the “problem gambling” band alone is not available.
Gambling harm statistics for NZ: how many people are affected?
Table 1. Gambling risk among all people aged 15 and over, New Zealand Gambling Survey 2023/24
| Category (Kupe’s name) | PGSI score | Share of people aged 15+ (%) | 95% CI | Estimated number of people |
|---|---|---|---|---|
| Non-gambler | n/a | 35.9 | 34.2 to 37.5 | n/a |
| Non-problem gambler | 0 | 52.4 | 50.7 to 54.0 | n/a |
| Low-risk gambler | 1 to 2 | 5.9 | 5.2 to 6.7 | n/a |
| Moderate-risk and problem gambler | 3 to 27 | 2.4 | 1.9 to 3.1 | 104,000 (81,000 to 132,000) |
| Any-risk gambler (low-risk, moderate-risk and problem combined) | 1 to 27 | 8.3 | 7.4 to 9.3 | 359,000 (320,000 to 401,000) |
Source: Health New Zealand, New Zealand Gambling Survey 2023/24, Kupe PGSI results, read 6 October 2026. CI is the 95% confidence interval.
The shares add up to 96.6%, not 100%. Kupe explains that 93 respondents (1.1% of the weighted population) answered “prefer not to say” or “don’t know” to at least one question, and 130 respondents (2.3%) did not receive the PGSI questions because of a routing error. The gap is a limitation of the survey. It does not show that the true figure is higher or lower.
Among people who gambled in the previous 12 months, any-risk is 13.0% (11.6% to 14.4%) and moderate-risk and problem is 3.7% (2.9% to 4.7%), because the base is smaller. The table shows how common risk and harm are across the population, not how gambling affects any one person.
Who carries more of the harm?
Table 2 breaks the two main shares down by group. Intervals are wide for small groups, so they matter as much as the percentages.
Table 2. Share of people aged 15 and over in each risk group, by population group, 2023/24
| Group | Any-risk gambler, % (95% CI) | Moderate-risk and problem gambler, % (95% CI) |
|---|---|---|
| Men | 9.7 (8.3 to 11.3) | 2.8 (2.0 to 3.8) |
| Women | 7.1 (6.0 to 8.4) | 2.1 (1.4 to 3.0) |
| Aged 15 to 24 | 7.8 (4.7 to 12.2) | 3.9 (1.6 to 7.8) |
| Aged 25 to 34 | 11.9 (9.2 to 15.1) | 3.0 (1.7 to 4.7) |
| Aged 35 to 44 | 7.6 (6.0 to 9.5) | 2.2 (1.4 to 3.4) |
| Aged 45 to 54 | 9.0 (7.0 to 11.2) | 2.4 (1.5 to 3.7) |
| Aged 55 to 64 | 9.5 (7.4 to 11.9) | 2.4 (1.4 to 3.8) |
| Aged 65+ | 5.0 (4.0 to 6.1) | 1.0 (0.6 to 1.7) |
| Māori | 14.4 (11.1 to 18.3) | 4.9 (2.8 to 7.9) |
| Pacific peoples | 16.1 (11.2 to 22.1) | 5.7 (2.5 to 10.8) |
| Asian peoples | 10.1 (7.9 to 12.7) | 1.9 (1.0 to 3.4) |
| NZ European/MELAA/Other | 6.6 (5.7 to 7.7) | 1.9 (1.4 to 2.6) |
| Area deprivation 1 to 3 (least deprived) | 6.7 (5.1 to 8.5) | 2.1 (1.1 to 3.5) |
| Area deprivation 4 to 7 | 7.2 (6.0 to 8.6) | 1.9 (1.3 to 2.7) |
| Area deprivation 8 to 10 (most deprived) | 12.2 (10.2 to 14.5) | 3.6 (2.3 to 5.2) |
| One or more disabilities, or a mental health, addiction or neurological condition | 13.3 (10.6 to 16.5) | 6.5 (4.3 to 9.3) |
| None reported | 7.4 (6.4 to 8.4) | 1.7 (1.2 to 2.3) |
Source: Health New Zealand, New Zealand Gambling Survey 2023/24, Kupe any-risk and moderate-risk and problem results, read 6 October 2026. MELAA is Middle Eastern, Latin American and African.
Kupe reports ethnicity as “total response”: a person who identifies with more than one group is counted in each, so the groups cannot be added together. Any-risk shares were 16.1% (11.2% to 22.1%) for Pacific peoples and 14.4% (11.1% to 18.3%) for Māori, against 6.6% (5.7% to 7.7%) for NZ European, MELAA and Other peoples. Kupe describes these differences as ethnic inequities and gives no causes, so this article does not either.
After adjusting for age and sex, Kupe puts the any-risk ratio at 2.12 (1.61 to 2.77) for Māori and 2.3 (1.64 to 3.23) for Pacific peoples, compared with people who are neither Māori nor Pacific, and marks both as statistically significant. A ratio above 1 means a group is more likely to be in the any-risk band. For Asian peoples the ratio is 1.2 (0.92 to 1.57). The interval includes 1, so the survey does not show a clear difference.
Head counts differ from shares. The estimated any-risk numbers are 93,000 for Māori, 50,000 for Pacific peoples, 80,000 for Asian peoples and 203,000 for NZ European, MELAA and Other peoples. The largest group by number has the lowest share.
Three more adjusted ratios have confidence intervals that do not include 1: 0.74 (0.59 to 0.93) for women compared with men, 1.44 (1.06 to 1.96) for the most deprived areas compared with the least deprived, and 1.95 (1.5 to 2.52) for people with one or more disabilities or conditions compared with those with none reported. The shares behind them are in Table 2.
By age, the 25 to 34 group has the highest estimate and the 65 and over group the lowest, but most intervals overlap, so the survey does not support a ranking. Estimates for the moderate-risk and problem band are less precise still, for example 5.7% for Pacific peoples (2.5% to 10.8%). These are associations, and a higher share does not show why.
What does harm look like in the survey?
Each PGSI question shows a different side of the harm. Among all people aged 15 and over, the most common responses about the previous 12 months, from the Kupe PGSI results read on 6 October 2026, were:
- felt guilty about the way they gamble: 4.2% (3.5% to 5.0%)
- bet more than they could really afford to lose: 3.6% (2.9% to 4.3%)
- went back another day to win back losses: 2.3% (1.8% to 2.9%)
- felt they might have a problem: 2.1% (1.6% to 2.7%)
- needed to gamble with larger amounts to get the same excitement: 2.1% (1.6% to 2.7%)
At the other end, 1.6% (1.2% to 2.2%) said gambling had caused health problems, including stress or anxiety, and 0.7% (0.5% to 1.1%) reported financial problems for themselves or their household. Borrowing money or selling something to gamble was the least common response, at 0.5% (0.3% to 0.8%). The data table does not state the exact answer threshold, so the figures are best read as people who reported the experience in the previous 12 months. How deposit limits and breaks work is covered in deposit limits and breaks at licensed online casinos.
Harm also reaches beyond the person who gambles. Kupe reports that 8.9% of people aged 15 and over (7.9% to 9.9%), about 381,000, experienced at least one form of household-level gambling harm in the previous 12 months. That means one or more of these: an argument about time or money spent on gambling, someone going without something they needed or bills not being paid, or someone taking time off study or work because of another person’s gambling. The share was 19.9% (15.7% to 24.5%) for Māori and 18.8% (13.6% to 25.0%) for Pacific peoples, against 7.3% (6.2% to 8.4%) for NZ European, MELAA and Other peoples. The survey asks about harm in the household or wider family, whether or not the person answering was part of it. A further 55.4% of people (53.7% to 57.1%) were concerned about the level of gambling in the community.
How much do New Zealanders lose to gambling?
Most people gamble in some form. In the previous 12 months, 64.1% of people aged 15 and over (62.5% to 65.7%) did, and 14.0% (12.9% to 15.0%) did so every week or almost every week. By activity, Kupe’s participation results show Lotto NZ products at 55.2% (53.6% to 56.8%), the TAB at 12.1% (11.0% to 13.2%) and pokies in a pub or club (class 4) at 11.9% (10.8% to 13.2%).
The money side comes from DIA, which publishes the amount lost by gamblers, described as the operators’ profits, for the four main types of gambling. The figures are actual dollars for each operator’s financial year-end. As at 6 October 2026, 2023/24 is the latest year published.
Table 3. Amount lost by gamblers by sector, $ million, not adjusted for inflation
| Sector | 2019/20 | 2020/21 | 2021/22 | 2022/23 | 2023/24 | Share of 2023/24 total (%) |
|---|---|---|---|---|---|---|
| TAB New Zealand | 315 | 385 | 380 | 376 | 371 | 13.3 |
| NZ Lotteries Commission | 631 | 694 | 654 | 710 | 792 | 28.4 |
| Gaming machines (outside casinos) | 802 | 987 | 833 | 1,070 | 1,037 | 37.1 |
| Casinos | 504 | 559 | 387 | 604 | 592 | 21.2 |
| Total | 2,252 | 2,625 | 2,254 | 2,761 | 2,792 | 100 |
Source: DIA gambling expenditure statistics, read 6 October 2026. The share column is calculated from DIA’s figures.
Gaming machines outside casinos, the electronic gaming machines (EGMs) known as pokies, are the largest item at $1,037 million. That is 37.1% of the 2023/24 total, based on DIA figures. On the same basis the total is 1.1% above 2022/23 and 24.0% above 2019/20, both nominal changes with no adjustment for inflation.
DIA’s page on minimising gambling harm says pokies “are one of the most harmful forms of gambling” and “the leading reason that people seek gambling-related help in Aotearoa”. It adds that one in five people (20%) who play pokies are “classed as low risk, moderate risk or problem gamblers”, a figure from the Health and Lifestyles Survey 2020. That is an older survey, so the figure should not be compared directly with the 2023/24 results.
Table 3 covers only the four main types of gambling that DIA reports and leaves out money spent with offshore online operators. It is not adjusted for population or inflation, so it is not a spending figure per person. It shows how much operators kept from gamblers, not how much any one person lost.
How common is online and overseas gambling?
The survey also measures online gambling. According to Kupe, 31.1% of people aged 15 and over (29.6% to 32.6%) took part in some form of online gambling, with New Zealand or overseas providers, and online Lotto at 25.8% (24.5% to 27.2%) accounts for most of it.
Kupe’s “overseas online gambling” indicator is 3.6% (2.9% to 4.5%), about 156,000 people. It counts anyone who answered yes to any of four questions, about online pokies, online poker and casino games, betting on e-sports through overseas websites and other online sports betting through overseas websites, as Kupe defines it. The share was 5.5% (4.1% to 7.1%) for men and 1.9% (1.4% to 2.6%) for women. It was 7.1% (4.6% to 10.5%) for Māori and 8.7% (5.3% to 13.1%) for Pacific peoples, with wide intervals for both.
Risk is higher among people who gambled on overseas sites. In that group, 43.2% (32.4% to 54.4%), about 68,000 people, were in the any-risk band, and 22.0% (13.5% to 32.7%), about 34,000, in the moderate-risk and problem band. For all people who gambled, the equivalent shares were 13.0% and 3.7%. This is an association, not a cause. The group is small and the intervals are wide.
The fieldwork ended before the Online Casino Gambling Act 2026 took effect, so these figures describe the period before licensing. They are a reference point for later data. Which operators hold a licence is covered in licensed online casinos in New Zealand, and what a person can do on a single site in how self-exclusion works at online casinos.
How many people know where to get help?
Kupe’s help-seeking results show that 39.0% of people aged 15 and over (37.4% to 40.6%), about 1,680,000, know where to seek professional support. The share was 29.9% (25.2% to 35.1%) for Pacific peoples and 27.8% (24.0% to 31.8%) for Asian peoples, against 42.1% (40.1% to 44.0%) for NZ European, MELAA and Other peoples and 41.2% (36.7% to 45.8%) for Māori.
Only 1.1% (0.8% to 1.5%), about 47,000 people, accessed help for themselves or someone else. The figure includes help sought for another person, so it cannot be set against the 104,000 people in the moderate-risk and problem band.
What the numbers do not tell you
- The survey measures gambling-related harm from what people said about themselves for the previous 12 months, and it covers people aged 15 and over.
- About 3.4% of the weighted population has no PGSI score.
- Fieldwork ran from late 2023 to early 2024, before the Online Casino Gambling Act 2026.
- DIA’s figures are nominal, cover only the four main types of gambling it reports and leave out offshore online operators.
- The survey and DIA use different bases, so combining them into an “average loss per person at risk” would mislead.
These figures describe the population. They are general information, not medical advice, and they cannot tell you whether your own gambling is a problem.
Where to get help if these numbers describe you
The Gambling Helpline is free: call 0800 654 655 or text 8006. It is run by Whakarongorau Aotearoa // New Zealand Telehealth Services. DIA also lists the Māori Gambling Helpline, the Pasifika Gambling Helpline, the Debt Gambling Helpline and the Youth Gambling Helpline as specialist services on its minimising gambling harm page. Safer Gambling Aotearoa, which is supported by Health New Zealand, lists local support services and has a “Test your gambling” page.
Frequently asked questions
How many people in New Zealand experience gambling harm?
The New Zealand Gambling Survey 2023/24 estimates that 8.3% of people aged 15 and over (95% confidence interval 7.4% to 9.3%), about 359,000 people, were at some level of risk of gambling harm in the previous 12 months. For 2.4% (1.9% to 3.1%), the risk was moderate or problem level.
What is the Problem Gambling Severity Index (PGSI)?
The PGSI is a nine-question screening score about the previous 12 months. Each answer scores 0 to 3, so totals run from 0 to 27. Kupe groups them as non-problem (0), low-risk (1 to 2), moderate-risk (3 to 7) and “problem gambling” (8 to 27). It is not a diagnosis.
How much do New Zealanders lose to gambling each year?
DIA reports that gamblers lost $2,792 million in 2023/24 across the TAB, NZ Lotteries Commission, gaming machines outside casinos and casinos. Gaming machines were the largest share, at 37.1% based on DIA figures. The amounts are not adjusted for inflation.
Are Māori and Pacific peoples more affected by gambling harm?
The survey reports higher any-risk shares for Māori (14.4%, 11.1% to 18.3%) and Pacific peoples (16.1%, 11.2% to 22.1%) than for NZ European, MELAA and Other peoples (6.6%, 5.7% to 7.7%). Kupe calls these ethnic inequities and gives no causes. The Pacific interval is wide, so that estimate is imprecise.
Where do NZ gambling harm statistics come from, and how recent are they?
The survey data come from the New Zealand Gambling Survey 2023/24, run by the Ministry of Health and Health New Zealand | Te Whatu Ora, with results in Kupe. The spending data come from DIA. Both were read on 6 October 2026, and the survey fieldwork ran from late 2023 to early 2024.
Sources
- Kupe: any-risk gambler, accessed 6 October 2026
- Kupe: moderate-risk and problem gambler, accessed 6 October 2026
- Kupe: Problem Gambling Severity Index, accessed 6 October 2026
- Kupe: community harm, accessed 6 October 2026
- Kupe: any gambling, accessed 6 October 2026
- Kupe: any overseas online gambling activity, accessed 6 October 2026
- Kupe: help-seeking, accessed 6 October 2026
- Ministry of Health: New Zealand Gambling Survey, accessed 6 October 2026
- DIA: Gambling expenditure statistics, accessed 6 October 2026
- DIA: Minimising gambling harm, accessed 6 October 2026
- Gambling Helpline, accessed 6 October 2026
- Safer Gambling Aotearoa, accessed 6 October 2026
- Safer Gambling Aotearoa: About us, accessed 8 October 2026