Public sector pension indexation: how PSS, CSS, MilitarySuper and DFRDB pensions grow over a long retirement

For retirees on lifetime defined benefit pensions from Commonwealth public service schemes, the indexation methodology is the principal mechanism by which the pension's real purchasing power is preserved — or eroded — over time. Without indexation, a $50,000-a-year pension would lose roughly half its real purchasing power after 25 years of typical inflation. With indexation, the pension keeps pace with inflation to varying degrees depending on the formula written into the scheme's governing legislation. The Public Sector Superannuation Scheme (PSS), the Commonwealth Superannuation Scheme (CSS), MilitarySuper, the Defence Force Retirement and Death Benefits scheme (DFRDB), and the various state public sector schemes each have their own indexation provisions, and the differences are material over a 20-30 year retirement. Two ex-public servants who retire with similar starting pensions can find their pensions diverging substantially over time depending on which scheme they're in.

The starting point for understanding the differences is the Consumer Price Index (CPI) — the standard ABS measure of general inflation across a typical urban consumer basket. Most Commonwealth public sector schemes index their pensions by CPI: each year (or twice a year, depending on the scheme) the pension increases by the percentage change in CPI over the relevant reference period (Commonwealth Superannuation Corporation — Indexation, https://www.csc.gov.au/Members/Forms-and-resources/Indexation, accessed 6 May 2026). CPI indexation is mathematically clean, predictable, and transparent. It's also limited: the typical urban consumer basket isn't the typical retiree's basket. Health costs, utilities, residential services, and other categories that weigh heavily in retiree spending can grow faster than general CPI in some periods. The result is that a CPI-indexed pension may slip in real terms relative to the actual cost of living a retiree experiences.

The Pensioner and Beneficiary Living Cost Index (PBLCI) is the ABS index designed to address this — it weights the consumer basket toward the spending patterns of pensioners and welfare recipients (ABS — Selected Living Cost Indexes, https://www.abs.gov.au/statistics/economy/price-indexes-and-inflation/selected-living-cost-indexes-australia, accessed 6 May 2026). PBLCI typically grows at a similar rate to CPI on average but can outpace it in periods when pensioner-heavy categories run faster. The Australian Age Pension uses a layered methodology under section 1192 of the Social Security Act 1991: it's adjusted in March and September each year by the higher of CPI or PBLCI growth, then benchmarked to ensure the maximum single rate sits at or above 27.7% of Male Total Average Weekly Earnings (MTAWE), with the partnered combined rate at 41.76% of MTAWE (DSS Social Security Guide 5.2.2.30 — Indexation of pensions, https://guides.dss.gov.au/social-security-guide/5/2/2/30, accessed 6 May 2026). When wage growth outpaces both CPI and PBLCI, the MTAWE benchmark pushes the Age Pension higher than a pure inflation-indexed pension would go. Over decades, the MTAWE benchmark is the structural feature that ensures Age Pension keeps pace with broader wage growth, not just price growth.

For most Commonwealth public sector pensions, the MTAWE benchmark doesn't apply. The schemes index by CPI without the wage-growth backstop. This is a structural difference rather than a grievance: Age Pension is calibrated to wages because it's a baseline social security payment that should keep pace with community living standards; public sector pensions are pre-funded retirement entitlements indexed to preserve real purchasing power. The two schemes are doing different jobs. The practical consequence, however, is that public sector pensions tend to grow more slowly than Age Pension over long retirements, and the relative position of an ex-public servant compared to an Age Pensioner shifts over time.

Within the public sector schemes, the CSC website confirms the current indexation arrangements. The Commonwealth Superannuation Scheme (closed to new members on 30 June 1990) indexes by CPI twice yearly (January and July). The Public Sector Superannuation Scheme (closed to new members on 30 June 2005) similarly indexes by CPI twice yearly. MilitarySuper indexes by CPI for its various member categories. The Defence Force Retirement and Death Benefits scheme — closed to new members on 1 October 1991 — was historically CPI-indexed only, but the Defence Force Retirement Benefits Legislation Amendment (Fair Indexation) Act 2014 introduced the "Fair Indexation" reform from 1 July 2014, applying the better of CPI, PBLCI, or 27.7% MTAWE growth (the Age Pension indexation methodology) to DFRDB recipients aged 55 and over (https://www.legislation.gov.au/Details/C2014A00091, accessed 6 May 2026). For DFRDB pensioners who qualified for the 2014 reform, the indexation has been more generous since then than the original CPI-only methodology would have produced. Similar reforms have been advocated for PSS, CSS, and MilitarySuper recipients but have not been legislated. The Public Sector Superannuation Accumulation Plan (PSSap) is an accumulation scheme rather than a defined benefit, so the indexation question doesn't arise — the balance grows by investment returns. State public sector schemes vary widely; New South Wales, Victoria, Queensland and others each have their own provisions, and the indexation should be confirmed scheme-by-scheme.

The cumulative impact of indexation differences over a long retirement is substantial. For an illustrative starting pension of $60,000 at age 60, the difference over 25 years between CPI-only indexation (averaging 2.5% a year) and a CPI/PBLCI/MTAWE methodology (averaging 3.2% a year) compounds to roughly $250,000 in cumulative payments — about a 12% difference in lifetime value. A 0.7% per year difference in indexation rate doesn't sound large in any single year, but over 25 years of compounding it produces a quarter-million-dollar gap. For pensioners drawing for 30 years rather than 25, the gap widens further. The indexation methodology isn't a footnote — it's a primary driver of lifetime income.

A specific point about high-value DB pensions: the post-2017 super reforms introduced the Capped Defined Benefit Income Stream (CDBIS) framework, capping the favourable tax treatment for DB pension income above the defined benefit income cap. The cap is set at the general transfer balance cap divided by 16, which for FY25-26 with the general TBC at $2.0 million produces a defined benefit income cap of $125,000 per year (ATO — defined benefit income cap, https://www.ato.gov.au/individuals-and-families/super-for-individuals-and-families/in-detail/defined-benefit-income-cap-tool, accessed 6 May 2026). The cap is indexed because the general TBC is, but it indexes in $100,000 increments aligned to the general TBC's CPI-driven movements, which may not keep pace with senior DB pensioners' actual pension growth — particularly for DFRDB recipients on the post-2014 enhanced indexation. Over time, more of a senior ex-public servant's pension income may fall above the cap, producing higher tax than at retirement. For ex-Senior Executive Service officers, senior military officers, and other high-end DB pensioners, the cap interaction with indexation is part of the income-tax projection.

What do worked planning examples show?

These two cases show how the indexation question lands differently depending on the scheme and the member's position. Illustrative only — not personal advice — using FY25-26 figures.

Case 1 — Anne, 62, recently retired CSS pensioner. Anne retired from the Commonwealth public service after 35 years of service and is drawing a CSS lifetime pension of $72,000 a year. The CSS pension is indexed by CPI twice yearly (January and July) per the CSC indexation schedule. On these facts, the rational planning step is to project the pension over Anne's expected life expectancy — say to age 90 — using realistic CPI assumptions (2.5% to 3.0% a year), recognising that the projection will produce slower nominal growth than an Age Pension trajectory (which uses CPI/PBLCI plus the 27.7% MTAWE benchmark under SSA 1991 s.1192). Anne should also model her likely Age Pension entitlement over time: as her CSS pension grows by CPI while the Age Pension thresholds index by CPI and the Age Pension itself grows on a more generous methodology, her relative position may shift, potentially making her eligible for a part Age Pension at some point in her 70s or 80s. The CSS pension is also a "non-100%-deductible" capped defined benefit income stream subject to the post-2016 10% deductible-amount cap on the taxable component for CSS/PSS, which becomes relevant once she turns 60 if she hasn't already. The trap to avoid is using a single generic inflation rate for both her CSS pension and the Age Pension thresholds — they index differently, and the divergence matters for the long-term picture.

Case 2 — Bruce, 68, DFRDB pensioner with the 2014 indexation upgrade. Bruce served in the ADF and is drawing a DFRDB pension of $58,000 a year. He was over 55 on 1 July 2014 and so qualified for the Fair Indexation reform under the Defence Force Retirement Benefits Legislation Amendment (Fair Indexation) Act 2014; his pension now indexes by the better of CPI, PBLCI, or the 27.7% MTAWE-benchmarked rate. On these facts, his pension trajectory is closer to Age Pension methodology than to CPI-only schemes. The planning angle is to project his pension under realistic blended-indexation assumptions (3.0% to 3.5% a year average), recognise that the methodology produces materially more generous growth than CPI-only, and weigh that against any commutation options the scheme allows — a commutation gives a fixed lump sum that doesn't index, and against a generously-indexed pension the commutation factor may understate the long-term value. The trap to avoid is assuming all DFRDB pensioners are on the same methodology — recipients who were under 55 on 1 July 2014 may not have been eligible for the Fair Indexation reform at that point and the indexation history varies by individual circumstance. DFRDB recipients should also note that DFRDB is excluded from the 10% deductible-amount cap that applies to CSS and PSS post-1 January 2016, which has its own income-tax implications.

For ex-public servant retirees, the indexation methodology written into the scheme's legislation determines how their pension evolves over a 20-30 year retirement. Two schemes that look similar on the day of retirement can diverge substantially over time. The structural difference between price-indexed schemes (most public sector pensions) and the wage-benchmarked Age Pension produces a relative-position shift over decades, with implications for Age Pension entitlement, tax exposure under the CDBIS $125,000 cap, and the value calculation when commutation choices are available. For most retirees on these schemes, the planning conversation that matters is realistic projection over the expected horizon — using the actual scheme indexation methodology, not a generic inflation rate — and understanding what that means for income, tax, and Centrelink interaction in their later years.

Sources

• DSS Social Security Guide
• Australian Bureau of Statistics — Selected living cost indexes australia
• csc.gov.au — Indexation
• Federal Register of Legislation — C2014A00091
• Australian Taxation Office (ATO) — Defined benefit income cap tool