A discussion paper evaluating a new, forward-only retirement system funded by non-marketable government bonds, run alongside the existing superannuation pool rather than replacing it.

Prepared September 2026

By Deep T, Edited, Gunnamatta

Executive summary

The original 3 blog critique argued that Australia’s compulsory superannuation system fails on ten counts: it subsidises the wealthy, excludes carers and the self-employed, concentrates the nation’s wealth in a handful of mega-funds, slows the circulation of money through the economy, and burdens young Australians disproportionately, among other flaws.

The system proposed in response does not attempt to fix compulsory super or unwind what it has already built. It proposes a parallel system. That system is a Digital Pension Scheme (DPS)

The Digital Pension Scheme (DPS) operates as a government-issued, non-marketable bond which sits as collateral behind a digital account opened for every citizen. That account is topped up each year (payable fortnightly? Monthly?) by an actuarial calculation of what the person needs in retirement, and separately accrues interest linked to inflation and productivity growth. This would mean compulsory superannuation contributions are abolished, so that wages that would have gone into a super fund are instead paid directly to the worker, who is then free to spend, save, invest, or allocate into a genuinely voluntary superannuation fund. This would imply that the concessionary approach to taxation on voluntary superannuation contributions would shift to the earnings stage only, so that the concessions no longer reward high earners disproportionately just for redirecting salary.

This means that the DPS saves significant amounts of tax subsidies and means that how these tax savings are spent is important in relation to money supply and inflation. At least initially the extra income should be used to pay down government debt.

There also needs to be equity, as well as equality, built into the system so that those who contribute are rewarded for their contribution to the Australian economy and no one is left out.

The DPS plus how tax savings are spent is a materially stronger economic proposal than the compulsory super scheme currently in place, as well as addressing the inequities of that system. It also holds up well against the closest real-world precedent, the United States Social Security Trust Fund, which has operated on exactly this basis (special, non-marketable Treasury securities, invisible to public debt markets) for nearly ninety years. In addition to this an actuarial approach to determining pensioner needs offers some scope to address housing inequality through a DPS, and with governments likely facing a significant shortage of housing, as well as a greater presence of renters as opposed to mortgagees and outright owners for the foreseeable future.

1. What problems are we trying to solve?

The originating paper’s ten problems, in short:

1. The Taxpayer Is Subsidising Wealthy Superannuants
2. Super Doesn’t Create Money — It Reallocates Control Over It
3. It Favours the Wealthy and Penalises Everyone Else
4. It Fails Non-Working Wives, Carers, Entrepreneurs, and Divorcees
5. Fund Management Is Dominated by a Handful of Mega-Funds
6. Fund Managers Operate Under a Narrow Mandate That May Diverge From the National Interest
7. The Government Corrals the Wealth but Lacks a National-Interest Investment Architecture
8. Compulsory Super Slows the circulation of Money
9. Fund Managers Chase Asset Appreciation, Not Productivity
10. The System Stifles Opportunity for Young Australians

The system proposed here addresses the 10 points above but does not try to reform the existing pool to fix these things. It tries to make the existing pool increasingly irrelevant by offering a better-designed alternative going forward, and lets workers and, implicitly, the market decide which system serves them if both are ever allowed to coexist as genuine choices.

In addition to the 10 specific problems, there is one fundamental issue that is also being solved. Compulsory superannuation embeds funds management allocating citizens funds to invest, these managers get paid to make decisions about allocations. The actual investment decisions are focussed on increasing the individual fund, meaning that broader economic implications are not prominent considerations. In circumstances where Australian citizens are just as capable of investment and spending without those managers the DPS allows this to occur. The advent of a DPS also enables a more considered approach to investment and the ‘national interest’.

2. The proposed system, as now specified

2.1 A DPS account for every citizen, opened going forward

From an implementation date, every citizen would have a government-administered digital DPS account. Unlike compulsory super, this account’s starting trajectory is not the accidental output of years of proportional wage contributions — it is set directly by the government, using an actuarial calculation of what an adequate retirement income for that person, at that age, would require, discounted to a present value and credited to the account today. The account could be an account at the central bank, or, structured as digital money or a stablecoin held on Web 3 that could not be spent or cashed in until eligibility protocols are met.

2.2 Non-marketable government bonds as collateral

The bonds backing these accounts are not issued into the public debt market, sold to bond investors, or auctioned in the way ordinary Treasury bonds are. They are a special, non-tradeable class of government security, held only inside the DPS account system itself, similar in legal character to how the Australian government could, if it chose, create a dedicated instrument the way the United States already does for Social Security.

The closest working precedent: the US Social Security Trust Fund

The US Social Security Trust Fund has operated on precisely this model since the 1930s. When the Trust Fund runs a surplus, it is required by law to invest that surplus in special-issue Treasury securities available only to the Trust Fund — never sold to, or tradeable by, the public, because they exist entirely outside the marketable Treasury bond system, they have no direct effect on bond yields, auctions, or investor demand for ordinary government debt — precisely the property the proposal is relying on. This is a genuine, decades-long precedent for exactly the mechanism described, and it substantially supports the claim that issuing bonds this way need not disturb bond markets during the accumulation phase.

When the Trust Fund needs cash to pay benefits, it redeems these special-issue securities, and the Treasury must find the money to honour them — from current tax revenue, by cutting other spending, or, if neither is available, by issuing new marketable debt at that point. The special-issue bonds do not eliminate the eventual funding requirement; they simply move the moment it becomes a market-facing event from today (when the account is credited) to later (when the account is drawn down).

2.3 Annual growth: actuarial top-up plus inflation/productivity interest

Each account grows two ways every year. First, an actuarial recalculation — the same discipline used by any defined-benefit pension scheme or life insurer — reassesses what the person’s account balance should be, given updated life expectancy, updated assumptions about adequate retirement income, and time remaining to retirement, and tops the account up to that revised figure. Second, the existing balance separately accrues interest at a rate tied to inflation and productivity growth in the economy, similar in spirit to the “notional rate of return” used in Sweden, Italy, Poland and Latvia’s Notional Defined Contribution (NDC) pension systems, where individual notional accounts are credited annually at a rate linked to average wage or GDP growth rather than to any real investment return, because there is no real invested asset behind them either.

This two-part design is more sophisticated than a simple indexation rule and closer to how a well-run defined-benefit fund would be actuarially managed — except that the assets backing it are the government’s own credit rather than a pool of invested securities.

The actuarial approach is further enabled by very large volumes of general consumption data – including groceries, energy, fuels, insurance and medical costs – which is increasingly logged digitally on a daily basis, with more Australians making payments online or by card, and with most Australian service providers and vendors invoicing digitally. This offer considerable scope to make effective determinations based on actual outlays and invoices relating to living expenses, protecting both individual pensioners and taxpayers.

2.4 Abolishing compulsory contributions: money released into the economy

This is the change with the most immediate and largest effect. Under the current system, the Superannuation Guarantee compels employers to direct 12% of ordinary earnings into a fund the worker cannot touch for decades. Under the proposed system, this compulsion ends entirely for accounts covered by the new pillar: that 12% (or whatever the applicable rate is at the time) is paid directly to the worker as normal income, available immediately for consumption, debt repayment, housing deposits, business investment, or voluntary saving, exactly as the proposal specifies. Because the worker’s future retirement provision is now being separately built by the government’s own actuarial account rather than by this wage deduction, there is no gap left by removing it — provided the actuarial calculation is genuinely adequate, which is the condition the whole system’s fairness rests on.

2.5 Voluntary superannuation, tax-adjusted

Superannuation continues to exist for anyone who wants to save beyond the new universal account, but purely voluntarily, and with the tax treatment redesigned: contributions are taxed as ordinary income when made, with the concessional tax rate applying only to the earnings the voluntary fund subsequently generates. This removes the single largest driver of the current system’s regressive subsidy — the upfront tax break simply for redirecting salary into super — while still rewarding people for choosing to save and invest beyond the government-provided base.

3. How this maps to the ten problems

This mapping is this paper’s own assessment for the purpose of evaluating the proposal, not an official government or industry analysis. Problems 5, 6 and 9 are marked ‘not applicable’ rather than ‘solved’ because the new pillar does not reform the existing mega-fund system — it simply operates independently of it, by design, per the scope correction in Section 1.

4. Fiscal and monetary realism

4.1 Why non-marketable bonds genuinely avoid disturbing bond markets today

This is the proposal’s strongest and most defensible claim, and the Social Security precedent backs it up directly. A special-issue, non-marketable government bond is not auctioned, not held by private investors, and not priced by the market — it is an internal government instrument, and its existence has no more effect on Australian Government Securities yields or investor demand than an internal government spreadsheet entry would.

4.2 What does not disappear: the eventual funding requirement

What the non-marketable design defers rather than removes is the moment the government actually needs real resources to pay a retiree drawing down their account. At that point, exactly as with Social Security’s special-issue redemptions, the government must find the money from current tax revenue, from other spending cuts, or by issuing new, real, marketable debt. The size of that future liability is entirely a function of how generously the actuarial calculation in Section 2.3 is set, and this is a genuine policy lever rather than a fixed cost: setting the target at, say, an Age-Pension-equivalent adequacy level is a vastly smaller commitment than replicating today’s average superannuation balances, which are inflated by decades of concessional voluntary contributions from people who did not need the help. Whatever adequacy target is chosen it must be stated explicitly in legislation, in dollar and GDP-share terms, and reported on at least monthly so the true scale of the future commitment is never allowed to drift by accretion the way superannuation’s own tax concessions did after 1992.

Whilst compulsory superannuation slows economic activity, the DPS greatly increases economic activity, growing the tax base, increasing business activity and innovation, and providing more opportunities to younger generations currently facing housing and cost of life implications making compulsory superannuation of questionable benefit. All of which increases the government collections today, reduces government debt increases productivity and innovation and, in the future, to fund pension drawdowns when needed. Importantly though the pension liabilities are known, calculable and predictable.

4.3 Indexation and actuarial risk

Crediting interest at a rate tied to inflation and productivity is sound in principle but exposed to the same risk the NDC countries have already encountered: if measured productivity growth disappoints for a sustained period (as Australia’s own productivity growth has done for much of the past decade), the notional crediting rate will disappoint alongside it, and retirees relying on the new pillar will receive less than the actuarial ‘projections’ implied. Sweden’s system responded to an equivalent problem by building in an automatic balancing mechanism that reduces crediting rates when the system’s long-run finances come under strain. rather than leaving it to a future government to renegotiate under pressure.

4.5 The “digital money” implementation question

As before, the technology used to record these balances is a separate question from the economic design. A conventional government database, of the kind that already runs HECS-HELP and the Age Pension, would achieve the same economic outcome as a blockchain-based or stablecoin-style ledger. Where a genuinely digital-currency-style implementation could add real value is if the government wanted account balances to be more portable, programmable, or interoperable with other digital payment rails but this is an implementation choice layered on top of the actuarial and fiscal design, not a substitute for it.

5. Comparable real-world precedents

6. Risks and open questions

• The actuarial adequacy assumption is the whole game: every one of the ‘solved’ outcomes in Section 3 depends on the government’s actuarial top-up genuinely being adequate. If it is set too low — whether through honest miscalculation or future budget pressure — citizens under the DPS would be worse off than they would have been under compulsory super.
• Deferred rather than eliminated fiscal cost: as Section 4.2 sets out, the choice to use non-marketable bonds is a genuine and precedented way to avoid disturbing capital markets today. Future taxpayers or future governments will face the real cost when today’s citizens retire, and that cost should be projected, published, and stress-tested (as the US Congressional Budget Office and Social Security Trustees already do for the US system) rather than allowed to build up unquantified.
• Productivity-growth dependency: if Australia’s productivity growth remains weak for a sustained period, as it has for much of the past decade, the inflation-plus-productivity crediting rate could disappoint relative to what today’s actuarial projections assume, unless an automatic balancing mechanism (Section 4.3) is legislated from the start.
• Employer and labour-market interaction: releasing the 12% super guarantee as direct wages assumes it flows straight through to workers rather than being absorbed, renegotiated, or offset elsewhere in enterprise bargaining and award structures would need explicit legislative protection to ensure the released money actually reaches workers rather than employers..
• Loss of a productive-investment role: because the new pillar holds no real assets, it cannot contribute to the national-interest investment architecture that Problem 7 identifies as missing even from today’s system. If that remains a goal, it has to be pursued through a different vehicle entirely, run independently of this proposal.

Conclusion

Properly specified as a new, forward-only, non-marketable-bond-backed pillar that leaves the existing $4.4 trillion system untouched, this proposal is considerably more fiscally defensible than a transition-based version would be, and its central technical claim — that issuing government bonds this way need not disturb bond markets — is well supported by nearly ninety years of the US Social Security Trust Fund operating on precisely this basis. It would fully solve the money-velocity and young-Australians problems that the originating critique identified as the most damaging, and substantially improve the regressivity and carer-exclusion problems, for anyone covered by the new system. Its long-run soundness rests entirely on a single discipline the paper cannot make optional: the actuarial adequacy calculation behind each account’s annual top-up must be genuinely sufficient, independently reviewed, and stress-tested in public.