Tobacco, rent and rates: Where voters part ways with their parties

The coalition government scrapped New Zealand’s generational tobacco-sales ban, but Vote Compass data has found it's the issue where National, ACT and NZ First voters are furthest from the parties they support.

While all three parties are opposed to a lifetime ban on selling tobacco to anyone born after 2009, the average response among people intending to vote for each of them is supportive of the idea.

The difference is greatest for National. Its supporters lean further towards supporting a tobacco ban based on age than those backing ACT or NZ First.

The findings come from an analysis of 185,371 Vote Compass participants between October 4 and 27, comparing people’s policy answers with the positions of the party they intend to support.

They show that choosing a party does not necessarily mean agreeing with its approach - including on policies it has helped overturn.

Find out where your views line up with party policies with Vote Compass

The National-led coalition repealed the generational tobacco-sales ban in 2024, before it took effect.

Introduced under Labour, the law would have prohibited tobacco sales to anyone born on or after January 1, 2009. Vote Compass tests the same principle, although its proposition uses the slightly different wording “born after 2009”.

Respondents answer on a five-point scale, from strongly disagree at one to strongly agree at five.

National, ACT and NZ First's stances are each positioned at 1 - strongly disagree.

National supporters’ average response was 3.66, compared with 3.39 for ACT supporters and 3.34 for NZ First supporters.

Those averages sit between neutral and somewhat agree.

The tobacco issue produced the largest gap between each coalition party and its supporters across the 30 policy questions examined.

The largest gap between party policy and supporters' views in the entire analysis concerned council rates.

Te Pāti Māori is recorded as strongly disagreeing that the Government should limit how much rates can increase each year.

Its supporters averaged 4.12 - just beyond “somewhat agree”.

Asked about its position, the party distinguished between helping households afford their rates and imposing a central-government cap.

“Te Pāti Māori does not support a blunt central government cap that could leave councils unable to fund essential infrastructure and services. We support addressing rates affordability alongside sustainable and fair funding for local government.”

Rates caps were also the biggest point of difference between the Greens and their supporters, whose average response was 3.68.

The Greens have opposed the Government’s proposal, arguing it would undermine services and fail to address the pressures on infrastructure funding.

Vote Compass asks about government-imposed limits on rates generally, rather than the details of a particular cap.

The biggest variation between an Opportunity policy stance and its supporters' outlook concerned how much landlords should be allowed to increase rents.

Vote Compass records the party as strongly opposing limits, while its supporters averaged 3.91 - close to somewhat agree.

Opportunity argues: “Rent controls are proven to reduce the overall availability of housing.”

Its published housing policy focuses on speeding up construction, improving the quality of homes and providing financing for affordable housing.

Labour’s largest gap between stated policy and supporter stance was on the proposition: “Only women who were assigned female at birth should be allowed to compete in women’s sports.”

The party is recorded as strongly disagreeing, while its supporters’ average response was 3.76, towards somewhat agree.

Labour supports guidelines for transgender inclusion in sport.

Their policy statement does not set out eligibility rules for every sport or distinguish between community participation and elite competition.

Try Vote Compass to see where your views line up with party policies

These are each party’s three largest policy to supporter gaps.

Generational tobacco ban: National strongly opposes a lifetime ban on selling tobacco to anyone born after 2009, while its supporters lean towards agreement on average.

Rent-increase limits: The party strongly opposes limits on how much landlords can raise rents. Its supporters lean slightly towards supporting them.

Dental care for low-income adults: National strongly opposes the proposition that the Government should cover these costs. Its supporters’ average answer is almost neutral.

Transgender women in sport: Labour strongly disagrees that only women assigned female at birth should be allowed to compete in women’s sports. Its supporters lean towards agreeing with the restriction.

Standardised testing: The party strongly opposes regularly testing primary-school students through standardised assessments. Its supporters lean slightly towards agreement.

Guaranteed minimum income: Labour strongly opposes guaranteeing a minimum income for everyone, regardless of whether they have a job. Its supporters’ average answer sits just on the agreement side of neutral.

Council rates: The Greens strongly oppose government limits on annual rates increases, while their supporters lean towards supporting limits.

Citizenship tests: The party strongly opposes requiring migrants applying for citizenship to pass a test on New Zealand’s laws and values. Its supporters lean towards agreement.

English-language tests: The Greens strongly oppose requiring all work visa applicants to pass an English-language test. Their supporters lean towards supporting the requirement.

Generational tobacco ban: ACT strongly opposes a lifetime ban on selling tobacco to anyone born after 2009. Its supporters lean slightly towards support.

Taxes on wealthier people: ACT’s position is that wealthier people should pay much less tax. Its supporters’ average is close to paying about the same, slightly towards more.

Rent-increase limits: ACT strongly opposes limits on how much landlords can raise rents. Its supporters’ average answer is effectively neutral.

Generational tobacco ban: NZ First strongly opposes a lifetime ban on selling tobacco to anyone born after 2009, while its supporters lean slightly towards support.

Dental care for low-income adults: The party strongly opposes the proposition that the Government should cover these costs. Its supporters lean slightly towards agreement.

Free tertiary education: NZ First strongly opposes restoring free tertiary education. Its supporters’ average is close to neutral, slightly on the disagreement side.

Council rates: Te Pāti Māori is recorded as strongly opposing government limits on annual rates increases. Its supporters’ average is slightly beyond “somewhat agree”. The party says it opposes a blunt central-government cap that could undermine essential infrastructure and services.

Citizenship tests: The party strongly opposes requiring migrants applying for citizenship to pass a test on New Zealand’s laws and values. Its supporters’ average is around “somewhat agree”.

English-language tests: Te Pāti Māori strongly opposes requiring all work visa applicants to pass an English-language test. Its supporters’ average is close to “somewhat agree”.

Rent-increase limits: Opportunity strongly opposes limits on how much landlords can raise rents, while its supporters’ average is close to “somewhat agree”. The party argues rent controls reduce housing availability.

Youth-offender punishments: Opportunity strongly opposes harsher punishments for young offenders who commit serious offences. Its supporters lean slightly towards agreement.

Council rates: The party strongly opposes government limits on annual rates increases. Its supporters lean slightly towards supporting limits.

Methodology

Results are based on 185,371 respondents to Vote Compass between October 4 and 27, 2026. Data were weighted to align the sample characteristics with those of the New Zealand adult population on the following characteristics: sex, age, ethnicity, education, and past voting record. A margin of error does not apply to non-probability samples. The modeled error estimate for a percentage in the overall sample is ±0.3 percentage points, 19 times out of 20. Results based on subsamples of the data should be expected to contain greater error.