Want More Super If You Wait Until 70?

Chartered Accountants Australia and New Zealand has worked out what would happen if people could take NZ Super at 65 or wait each year until 70 for more money.

Its favoured idea assumes 40% of people who can get super wait until 70. It puts the saving at $98 billion over 20 years, and at $7.8 billion a year in 2048, according to RNZ. Peter Vial is head of Chartered Accountants New Zealand.

Under that idea, taking super at 65 would mean 95% of the current payment. Waiting until 70 would mean 107.5%. You would decide each year from 66 to 70. The age stays at 65. The payment amount changes.

Think of it like cutting a loaf. More thin slices if you start early. Fewer thick slices if you start late.

The fiscal starting point

NZ Super now costs the Government $26 billion a year. That is 5% of gross domestic product, the size of the economy. About 19% of all tax is spent on super.

More than 900,000 people now get NZ Super. In 10 years, that number is expected to pass 1.3 million.

By 2048, with no change, the cost is expected to reach $73 billion a year. That would be 8% of the economy. It would use 30% of all tax.

The population pattern is simple. In the 1960s there were seven working-age people for each pensioner. Now there are four. By 2065 there are expected to be two workers for each person on super.

Treasury's long-term book, He Tirohanga Mokopuna 2025, puts Government spending at $35,900 a person by 2065 in today's dollars, up from $18,300, with no policy change, according to the Treasury.

The design choices

The accountants tested more than one payment mix. The idea stays the same. Take it early and get less. Take it late and get more. The Government pays less over time because it pays for fewer years.

If only 20% waited, not 40%, the saving under the 95%-at-65 model would still be $64 billion over 20 years.

Another model would keep the age-65 rate at 100%. Waiting until 67 under that model would pay 105%. That version would save $72 billion over 20 years.

A third model would pay 130% for waiting until 70. It pays the most to late starters. It was costed with the other two.

Officials would need to check some details. How many people wait. The maths that sets the lower early rate and higher late rate. How it fits with KiwiSaver withdrawals, paid work after 65, and health benefits. None of those details were settled in the modelling release.

The political reception

Finance Minister Nicola Willis has said NZ Super needs changes, but they do not need to be as big as the OECD suggested, according to the NZ Herald. That was reported on 8 May 2026.

NZ First has said if elected, only New Zealand citizens could get NZ Super from 2029, according to 1News. That was reported on 6 September 2026.

There were earlier ideas about the age. Simon Bridges promised to slowly lift the age from 65 to 67. A Retirement Commission report said the age should stay at 65.

The broader context here is the Beehive maths every super idea faces. NZ Super for everyone has support across parties. The cost does not. A government that lifts the age pays for it at the next election. A government that does nothing leaves a bigger bill for the next Parliament.

Looking at what this means for decision-makers, the accountants' favoured model tries to find a middle path. It keeps 65. It cuts the early rate to 95%. It pays 107.5% for late starters. The savings rest on behaviour. Forty percent waiting until 70 is a big ask when 65 is the normal age. A straight age rise starts a campaign fight. Letting people choose to wait gives ministers talk of choice.

In my view, select committee checks, close looks by a small group of MPs, would focus on fairness as well as savings. People who can work to 70 gain. People in hard physical jobs, who live fewer healthy years, do not. Māori and Pacific results would be central. So would renters with no other savings, and women with broken work histories. Those fairness questions decide if choice feels like choice or like a cut.

Looking at Budget timing, there is also a timing point for Treasury and the Ministry of Social Development. Savings grow over 20 years. Admin and information costs come first. The $7.8 billion a year figure for 2048 helps long-term planning. It does not help a Budget set for the next four years.