Should private health insurance rebates depend on age?

Evidence brief by Yuting Zhang. Updated 6 November 2026.

In short: We estimate that removing the extra rebate would lead about 15,000 to 42,000 older Australians to drop cover. That is a small share of the roughly 3.2 million affected, and it would save an estimated $730–940 million a year by 2028–29. The people most likely to drop cover are on low incomes, so the effect should be monitored by income. Full-rate pensioners with cover are under real financial strain, so any support should go to people on low incomes, whatever their age.

Our position: We support the Bill.

The policy question

People aged 65 and over receive a higher private health insurance rebate than younger people on the same income. On the base income tier, the rebate is about 24% for people under 65, 28% for those aged 65–69, and 32% for those aged 70 and over.

The government’s Bill would give people on the same income the same rebate, whatever their age, and direct the savings to aged care funding. Its first version was defeated in parliament. An unchanged version was introduced on 15 October 2026 and is not yet law.

The Senate committee recommended the Bill be passed. Coalition senators recommended it not be passed. The Greens sought an exemption for age pensioners.

How many people would drop cover?

Estimates differ widely, largely because they measure different things.

Source Estimate Type of evidence
Staub & Zhang (2026) 14,821–42,498 people Working paper using tax records
Government Impact Analysis About 44,000 people, roughly 1% of those affected Government modelling
NSW Health Minister Up to 80,000 people cancelling or downgrading in NSW alone State modelling, not available to check
Private Healthcare Australia poll 39% more likely to drop cover, which we calculate would be about 1 million people if all acted on it Industry poll of stated intentions

We give more weight to records of what people actually did than to stated intentions. When means-testing was proposed over a decade ago, industry polling predicted 1.6 million people would drop cover. Participation instead rose slightly, although that reform also included a stronger tax penalty.

What the evidence shows

Older Australians respond only modestly to price. Liu & Zhang (2023) used the 2005 age-based rebate increase as a natural experiment. They found price elasticities of −0.10 to −0.20, compared with the −0.2 to −0.5 typically found for the general population. A separate study with a different team and dataset found no statistically significant effect on take-up in this age group. Peer reviewed.

Most of the money goes to people who would keep cover anyway. Staub & Zhang (2026) find the savings would come mainly from people who keep their cover. This is why savings can be large while coverage falls only slightly. Working paper, not yet peer reviewed.

The effect on public hospitals looks small. In Victoria, a one percentage point rise in private cover reduced public elective waits by about 0.34 days (Yang, Yong & Zhang, 2024). Extrapolating from this suggests waits about three to eight hours longer. This is an extrapolation, not a direct estimate. A separate study also found savings from lower subsidies exceeded added public hospital costs. The studies are peer reviewed; the extrapolation is ours.

A pensioner exemption would be poorly targeted and miss struggling younger families. Our analysis of 2022 HILDA survey data found that about half of the insured pensioners it would protect are part-pensioners, and one in seven part-pensioners has more than $2 million in wealth, including the family home. Full-rate pensioners with cover are under more strain: they spend 6.3% of disposable income on premiums, nearly double the share for non-pensioners. An exemption would miss uninsured pensioners and younger families. At around $40,000 income per adult, people aged 25 to 44 spend almost three times as much of their income on housing, childcare and insurance as older Australians, and are nearly four times as likely to report financial hardship. Our own analysis, not peer reviewed.

What government could do

These are our recommendations, based on the evidence above (Zhang, Vergili & Kettlewell, 2026).

  1. Pass the Bill as it stands. Income is a better basis than age for targeting support.
  2. Target any concessions by low income, not age. Consider including deemed income in the income test, so benefits do not flow to very wealthy households.
  3. Address genuine hardship through means-tested support.
  4. Report monitoring by income. Price sensitivity is concentrated among lower-income older Australians. The government’s modelling also puts about 95% of the decline in the lowest income tier (our Senate submission).

What we don’t know

  • The hospital effect is less certain than the coverage effect. The waiting-time result comes from regional differences in Victoria, not a national change for older people. People who drop cover because of cost may use public hospitals more than healthy people who drop it.
  • Past responses may not predict this reform. The strongest evidence comes from the 2005 rebate increase, and this change goes in the opposite direction.

What this means for you

If you hold or are considering cover, you can estimate your own costs. If you are 65 or over and receive a rebate, it also shows your cost without the age-based rebate.

Sources

Our journal articles and working paper:

Our submission and articles in The Conversation:

Other studies:

Other sources:

Funding and interests

Yuting Zhang has received funding from the Australian Research Council (Future Fellowship FT200100630), the Department of Veterans’ Affairs, the Victorian Department of Health, the National Health and Medical Research Council, and Eastern Melbourne Primary Health Network, and in the past from US funders including the NIH, Commonwealth Fund, AHRQ and Robert Wood Johnson Foundation. Liu & Zhang (2023) is cited in the government’s Impact Analysis, and Professor Zhang gave evidence to the Senate committee.

This research uses The Person Level Integrated Data Asset (PLIDA) which we obtained from the Australian Bureau of Statistics. All findings, opinions and conclusions are those of the authors and do not necessarily represent the views of the Australian Government or any of its agencies.

This site counts page visits with GoatCounter, without cookies or personal data.