Executive Summary
Australia faces persistent economic challenges. Productivity growth has slowed, business investment is uneven, and real wages have lagged expectations. In response, the Productivity Commission has proposed a ‘first step’ reform of company tax; a lower 20% rate for most firms, retention of 30% for the largest firms, and the introduction of a modest net cash-flow tax on all companies. The aim is to shift the burden of taxation away from the normal return to capital and towards what economists call ‘economic rent’.
The Productivity Commission’s proposal forms part of a long line of rent-tax ideas tracing back to the Meade Report and developed in Australia by Ross Garnaut and co-authors and Chris Murphy. Although presented by the Productivity Commission as a cautious first step, it adopts the same logic of taxing rents while exempting the normal return to capital. This paper therefore assesses the Productivity Commission’s approach within that wider intellectual framework, not merely as an isolated administrative measure.
Proponents argue that taxing rents is efficient. In theory, rents are surpluses above the normal return, so taxing them should not distort marginal decisions. In computable general equilibrium modelling prepared for the Productivity Commission, economic modeller Chris Murphy estimates that 54% of the corporate tax base is rent, drawn from land, minerals, and oligopoly positions. On this view, taxing rents more heavily would encourage new investment, raise wages, and increase consumer welfare. In a 2020 paper, Ross Garnaut and colleagues have gone further, setting out a blueprint for replacing company tax with a comprehensive cash-flow tax. This proposal is not entirely new. It revives the logic of the Resource Super Profits Tax proposed in 2010, itself a variant of the Brown tax. The Productivity Commission frames its package as modest and incremental, but the underlying idea is the same; shift the burden away from normal returns and towards so-called rents.
This paper takes those claims seriously, then tests them. It shows that the modelling rests on strong assumptions; that rents can be precisely identified, that capital is perfectly mobile, that dividend imputation can be ignored, that foreign and domestic tax burdens can be separated neatly, and that reforms can be judged by steady-state outcomes without considering the costs of transition. These assumptions drive the results, but they are not evidence. What the Commission and its modelling do not confront directly are incentive effects. Taxing ‘rents’ in practice risks taxing the entrepreneurial returns that drive innovation and investment. This paper argues that those incentives matter and cannot be set aside. Yet beneath the surface of theoretical neutrality lie assumptions that cannot survive contact with reality. Once the difficulties of classification, refund credibility, and adjustment are acknowledged, the supposed simplicity of a cash-flow tax gives way to uncertainty, instability, and risk.
Beyond the modelling lie deeper economic problems. The cash-flow tax requires the state to know the cost of capital for each firm, which is impossible. It also assumes that government can finance risky assets with risk-free financial instruments, which contradicts basic finance. It relies on the credibility of full loss refundability, which no government can deliver in practice. It assumes neutrality will hold over time, yet once projects are sunk governments face incentives to renege. And it treats entrepreneurial and innovative returns as if they were free surpluses, ignoring that these margins are the reward for risk-taking and investment in intangibles.
What would actually be taxed in practice are not unearned surpluses, but the returns that drive entrepreneurship and innovation. The neat theory of a rent tax cannot be delivered in the real world. A cash-flow tax in Australia would create new complexity, new disputes, and new risks. Instead of securing neutrality, it would undermine the very process of investment and innovation on which growth depends.
-
Introduction
Australia is facing slow labour productivity, uneven business investment outside of resources, and real wages have not kept pace with expectations. The Productivity Commission has been examining how to lift performance across these fronts. One strand considers how Australia collects company tax. In its July 2025 Interim Report, the Productivity Commission sketches an initial package. This includes reducing the company income tax rate to 20% for most firms, retaining the 30% rate for the largest firms, and introducing a modest net cash-flow tax as an additional instrument for both categories. The aim is to shift the burden away from the normal return to capital and toward what economists call ‘rent’; the promise is stronger investment, higher real wages, and faster productivity growth over time.[1] This type of proposal is not new to Australia — most recently Ross Garnaut and co-authors suggested a similar (but more sophisticated) version of this type of tax.[2] This type of tax is a variation of the so-called Brown-tax.[3] A previous attempt to introduce a Brown-tax in Australia was the 2010 ill-fated Resource Super-Profit Tax (RSPT). The continuity between that proposal and today’s cash-flow tax is important. Many of the criticisms that applied to the RSPT apply equally to the notion of a cash-flow tax.
The Productivity Commission’s proposal should therefore be understood as part of this rent-tax tradition rather than as a stand-alone administrative measure. Its analytical foundation lies in the Meade framework[4] and its modern Australian expositions by Garnaut and co-authors[5] and Chris Murphy[6]. Although the Commission presents its package as a limited first step within the existing company-tax system, it adopts the same conceptual logic – exempting the normal return to capital while taxing so-called economic rents. The analysis that follows evaluates the proposal within that broader intellectual lineage.
This paper takes the proposal of a cash-flow tax seriously. I begin by explaining what a corporate cash-flow tax is and what it is intended to do, and how the treatment of losses is supposed to work when so-called symmetry is achieved. I then present recent Australian modelling in support of the approach; those results now inform the Productivity Commission’s thinking. Only then do I turn to a critique of the proposal; first, where the modelling runs ahead of the evidence; then, the real-world economic problems that appear once we abandon textbook economics; finally, what would actually be taxed in practice were this type of tax adopted.
- Cash-Flow Taxation explained
A corporate cash-flow tax aims to tax the net cash generated by a firm’s real activities within the tax year. Cash receipts from sales and operating income are brought to account. Cash outlays on purchases and wages are deducted, and capital outlays are expensed immediately rather than depreciated over time. If there is a surplus of cash after all expenditures, including capital investment, have been incurred then the ATO taxes that amount by the tax rate. By contrast, if there is a cash deficit after all expenditures, including capital investment, have been incurred then the tax office refunds that amount multiplied by the tax rate.[7]
Generally, there are two definitions of the cash tax base in this form of taxation:[8]
- R-base: the tax includes only real transactions – cash receipts from sales minus cash payments for purchases, wages, and investment. All financial flows (new borrowing, repayments, interest, dividends, share issues, and buy-backs) are excluded for non-financial firms.
- R + F-base: the tax includes both real and financial transactions. It adds to the R-base by treating financial inflows (new borrowing or equity issues) as taxable receipts and financial outflows (repayments, interest, and dividends) as deductible payments.
The essential differences between the current company tax, the R-base, and the R + F-base can be shown in a stylised income statement.
Figure 1. Stylised comparison of taxable income under current company tax, R-base, and R + F-base cash-flow taxation.
| Item | Current company tax | R-base cash-flow tax | R + F-base cash-flow tax |
| Sales and operating receipts | Included as taxable income | Included as taxable income | Included as taxable income and financial inflows (new borrowing, new equity issues) |
| Operating expenses (wages, purchases) | Deductible | Deductible | Deductible |
| Capital expenditure | Depreciated over time | Fully expensed immediately | Fully expensed immediately |
| Interest payments | Deductible | Excluded | Deductible |
| Dividends paid | Not deductible | Excluded | Deductible |
| New borrowing / share issues | Not taxable | Excluded | Taxable inflow |
| Debt repayment / share buybacks | Not deductible | Excluded | Deductible |
| Losses | Carried forward subject to continuity rules | Fully refundable in theory (rarely in practice) | Fully refundable in theory (rarely in practice) |
| Tax base definition | Accounting profit adjusted for timing and deductibility | Net real cash-flow from production | Net real and financial cash-flow |
The difference between the R-base and the R+F-base is not just a technicality; it drives how the tax interacts with financing choices, how the boundary between real and financial activity is policed, and how the system fits with the rest of the tax structure. One of the long-standing problems in company tax is the deductibility of interest while dividends are not deductible; this creates a bias towards debt.[9] In theory, a pure R-base cash-flow tax with full expensing and full loss refundability could remove the debt–equity bias, since neither interest nor dividends would be deductible. In practice, however, refundability is never implemented, and the Productivity Commission’s proposal retains the existing company-tax element where interest remains deductible. Consequently, the debt bias would persist in any realistic or hybrid application of an R-base system. By contrast, an R+F-base includes all financial inflows and outflows; money raised from debt or equity is taxed as an inflow; repayments, interest, and dividends are allowed as deductions. In theory this achieves neutrality between debt and equity; in practice it means the tax must now account for every form of financial transaction.
This leads to the second difficulty. Modern firms routinely combine financial and non-financial activity. Retailers offer store credit; platforms embed payments; large firms use complex lease and supplier-financing structures. Under an R-base the tax office must decide what counts as real and what counts as financial — but that line is porous and open to dispute, so classification becomes the new margin for avoidance. An R+F-base sidesteps the boundary by pulling everything into the base, but at the cost of greater administrative burden, more complex accounting, and greater compliance costs.
The third point is fit with the wider tax system. If an R-base is adopted, the finance sector must be taxed separately because banks and insurers earn their income mostly through financial flows, and those flows are excluded from the R-base.[10] If an R+F-base is adopted, the need for such a parallel tax regime falls away, but the trade-off is a much more complex corporate tax base. Policymakers must choose between a simpler R-base that is conceptually neat but fragile at the boundary, or an R+F-base that is more comprehensive, but harder to operate.
The appeal of the cash-flow tax lies in its promise that only so-called ‘economic rent’ is being taxed. Economists use the term ‘rent’ to mean a return above the normal cost of keeping a resource in its present use; in plain terms it is a surplus over the investor’s opportunity cost. Land in a prime location, for example, earns a scarcity rent; an innovation that succeeds in the market earns a ‘quasi-rent’ for as long as it remains ahead of competitors. The argument is that these returns are not required to trigger the investment; they arise over and above the normal return that investors demand. If the normal return is left untaxed, then taxing the surplus should not distort marginal decisions. A project that only just breaks even before tax will still break even after tax. This apparent neutrality is the foundation of the theoretical case for the cash-flow tax.
Taxing rent is also attractive because it promises less interference with financing choices and investment planning. In a conventional company tax, the treatment of depreciation, interest deductibility, and dividend payments can distort choices about when and how to invest. By contrast, under a cash-flow system with immediate expensing and full loss treatment, the state is effectively a silent partner taking a constant share of both inflows and outflows.[11] That design, on paper, ensures that the tax falls only on the excess above the normal return and not on the normal return itself. For proponents this delivers the best of both worlds; it secures revenue while leaving the incentive to invest intact.
This is why proponents argue that a cash-flow tax is a rent tax. By expensing investment immediately and refunding losses, only the surplus above the investor’s opportunity cost remains taxable. The same logic lies behind the Allowance for Corporate Equity (ACE) and accelerated depreciation schemes used in other countries: each seeks to remove tax from the normal return. The difference is that ACE adjusts the base by allowing a notional return on equity, whereas the cash-flow tax adjusts timing by expensing investment. Both share the same ambition but by different mechanics.
In the Australian debate this argument has an extra appeal; it is often claimed that rents here are high and rising. Ross Garnaut and his co-authors, for example, point to widening operating margins in some sectors as evidence that surplus returns are large enough to be taxed without harming investment. Chris Murphy, in modelling prepared for the Productivity Commission, makes the same point in more formal terms; he estimates that about 54% of Australia’s corporate tax base is rent, drawn from land rents, mineral rents, and oligopoly rents in both financial and non-financial industries.[12] On this view, more than half of current company tax revenue already comes from rents, with the balance from normal returns to capital. The inference is that companies enjoy economic rents over and above the normal return, and that taxing those surpluses could raise revenue, and improve fairness, without affecting marginal decisions. For reformers, this makes the cash-flow tax particularly attractive; it promises to draw revenue from the growing pool of rents while leaving the normal return untaxed. The attraction is clear; if rents are both large and durable, then shifting the tax base towards those rents appears to offer a way to strengthen the system while avoiding the usual trade-off between revenue and investment.
- The Murphy Modelling
The Productivity Commission’s current thinking about company tax relies heavily on a set of computable general equilibrium (CGE) simulations prepared by Chris Murphy. CGE models are standard tools in policy economics, but they do need some explanation because their assumptions (often) drive the results. A CGE model is a very large mathematical description of the economy as a system of linked markets. It takes data on households, firms, industries, imports, and exports, and it calibrates that data so that the ‘model economy’ reproduces observed flows of goods, services, labour, and capital of the actual economy under investigation. Once that base case is established, the modeller can then simulate what happens if a policy is changed; for example, cutting the company tax rate, introducing an allowance for equity, or shifting the tax base towards economic rents. In short, the CGE model is designed to show how the economy might re-balance after a tax change, under the assumption that markets clear and resources are fully employed.[13]
Murphy previously used this approach in 2018 to investigate the company tax; he subsequently updated his work in 2025 to feed into the Productivity Commission’s Interim Report.[14] In the updated version he divided the company tax base into two parts; the normal return to capital, and economic rents. He defined rents to include land rents, mineral rents, and oligopoly rents in both financial and non-financial industries. This allowed him to quantify how much of the current company tax already falls on rents. His conclusion is striking; he estimates that 54% of the corporate tax base is rent, with the 46% balance falling on normal returns. In his earlier work the rent figure had been 41%, so the revised model suggests that rents have grown as a share of the base.
The model is then used to compare different reform options. A straightforward cut in the company tax rate reduces the burden on both rents and normal returns; an allowance for corporate equity leaves debt and equity financing more balanced but still taxes the normal return. By contrast, narrowing the base to target rents through a cash-flow tax removes tax from the normal return, while leaving rents being taxed. In Murphy’s simulations, that option produces the largest gains in both business investment and consumer welfare. The mechanism is simple; if normal returns are untaxed, more capital is attracted to Australia, more capital per worker means higher wages, and higher wages feed through to higher household consumption.
These results are then interpreted by the Productivity Commission as the intellectual foundation for its ‘first-step package’. In its Interim Report, the Productivity Commission explicitly cites Murphy’s modelling as evidence that moving the corporate tax base towards rents yields the strongest long-run gains. That modelling supports the proposed combination of a lower company income-tax rate for most firms, a retained 30% rate for the largest firms, and the introduction of a small net cash-flow tax. In the Productivity Commission’s words, narrowing the base towards rents is “the best way” to boost investment and productivity growth over time.
- From the Productivity Commission’s first step to the Garnaut blueprint
The Productivity Commission describes its recommendations as a first step, not a final design. Its Interim Report is explicit; the package of a lower 20% company tax rate for most firms, the retention of 30% for the largest firms, and the introduction of a 5% net cash-flow tax is meant to begin a transition. Subsequent steps would involve growing the size of the net cash-flow tax to fund broader reductions in company income tax, but only after evaluation of the initial reform. In other words, the Productivity Commission does not commit to a second step in detail; it points instead to an incremental process in which the cash-flow tax becomes more significant over time and the income-tax component less so. The long-term vision is a system that relies more on taxing rents and less on taxing the normal return to capital.
Ross Garnaut and co-authors provide the most fully worked-out version of what that vision would look like in Australia. In their 2020 proposal they set out a comprehensive plan for replacing company income tax with a corporate cash-flow tax for non-financial firms, alongside a separate Financial Sector Income Tax for banks and insurers. They also addressed the problem of losses by allowing refunds or, if refunds were not practical, carry-forwards with an uplift set at the project’s discount rate.[15] To protect the tax base, they proposed rules denying deductions for imports of services unless the corresponding receipts were taxed abroad. And they envisaged a gradual transition over a decade, with careful attention to the stock of existing assets and liabilities. Taken together, the Garnaut scheme represents a logical ‘end game’ if the Productivity Commission’s incremental path were carried through to completion.
It is important to be clear: the Productivity Commission has not endorsed the Garnaut plan as such. Its Interim Report confines itself to being a first-step package and a cautious statement that the net cash-flow tax could be expanded later. But the intellectual link is undeniable. The Productivity Commission relies on Murphy’s modelling to argue that rent taxation is efficient; Garnaut provides a detailed blueprint of how a rent-focused tax system could be structured. Observers of the Australian debate can reasonably infer that if the first step were judged a success, pressure would build for more ambitious reform along Garnaut’s lines.
The argument developed here therefore treats the Productivity Commission’s proposal as the applied embodiment of the rent-tax logic explored by Garnaut and Murphy. The fact that the Productivity Commission limits its first step to a small, non-refundable component does not change the underlying conceptual commitment to taxing rents while exempting the normal return. The institutional weaknesses identified below follow from that commitment, whether implemented fully or partially.
- Where the modelling runs ahead of the evidence
The Productivity Commission relies heavily on Murphy’s CGE results but those results, however, are only as sound as the assumptions that drive them. The problem is that a computable general equilibrium model is not an empirical discovery device; it is a framework that converts assumptions into numerical projections. The model can be a useful way of organising thinking, but it should never be mistaken for evidence. In this case, several of Murphy’s assumptions, in my view, are contestable.
First, the estimate that 54% of the company tax base is ‘rent’ depends on classifying land rents, mineral rents, and oligopoly profits as if they were observable surpluses, when in fact they may reflect innovation, intangibles, or entrepreneurial returns. In practice it is difficult to know how much of an observed margin reflects genuine scarcity or monopoly, and how much reflects returns to innovation, investment in intangibles, or entrepreneurial risk-taking (see below for more discussion on this point).[16] Treating the whole amount as rent risks overstating the non-distortionary part of the base.
A related problem lies in the Commission’s design itself. The proposed lower 20% company tax rate applies to firms with turnover below $1 billion, with larger firms remaining at 30%. In effect, size is used as a proxy for rent intensity. That is a questionable assumption. Many large firms — from BHP to CSL to Atlassian — remain highly innovative and capital intensive. Their surpluses often reflect risky investment in technology, logistics, or global expansion, not mere scarcity rents. Treating size as a marker of rent risks taxing the very entrepreneurial returns that drive investment and innovation in Australia.
Second, the model assumes perfectly elastic global capital supply, an assumption that abstracts from the uncertainty, risk, and regulatory complexity that shape real investment decisions. That assumption drives the results; if normal returns are not taxed, the model ‘imports’ more capital, which pushes up wages and welfare. But in the real-world capital is not perfectly mobile, investment decisions are made in the face of political and regulatory risks, and global capital markets are not entirely frictionless. If capital is not perfectly elastic, then the predicted wage and welfare gains might not materialise as the model suggests.
Third, Murphy treats dividend imputation as behaviourally neutral when calculating effective tax burdens.[17] For resident investors, the franking system already offsets company tax against personal tax, meaning that the normal return is not taxed at the headline rate. For those investors, the efficiency gains from moving to a cash-flow tax are therefore much smaller. By assuming that franking credits do not affect investment incentives, the model overstates the likely benefits. The interaction with dividend imputation also matters for any further reform along Garnaut’s lines. If a cash-flow tax were to replace company tax entirely, it is unclear whether franking credits would remain available; for resident investors this would be a material change, reducing the attractiveness of domestic equity and undermining the neutrality that proponents claim.[18]
A further complication lies in who the marginal investor actually is. In Australia’s listed sector, the marginal shareholder is now (very likely) an international investor who cannot benefit from dividend imputation. For that investor the effective rate is close to the statutory rate. Raising the combined burden on large firms from 30 to 35 per cent, even with cash-flow expensing, would materially reduce after-tax returns on Australian equity. As Richard Holden has observed, such a change would risk aligning Australia’s corporate tax position with countries like Colombia rather than its advanced-economy peers; an outcome hardly consistent with competitiveness.[19]
Fourth, the treatment of foreign ownership is stylised. The welfare results are calculated from the perspective of Australian households, yet a large share of company tax is paid by foreign shareholders. Reducing that burden may attract investment, but it also means Australian households forgo any fiscal benefit of having foreign shareholders paying part of the company tax. The model assumes that the extra capital more than compensates Australians for the loss of foreign-paid tax. That assumption is open to question.[20]
Finally, the model is solved for a steady-state outcome — a world in which the economy has fully adjusted to the tax change and a new equilibrium is reached. Real reforms do not work that way. Firms and households make decisions under uncertainty, transition can be messy and take decades to fully implement. By focusing only on the steady state, the simulations ignore costs, risks, and distributional effects along the path of adjustment. These steady-state projections also obscure the path by which any reform would occur. Transition would be slow, uneven, and politically fraught. Existing projects would straddle two regimes; losses would need to be recalculated; and balance sheets would be revalued. Accounting rules and tax law would diverge, generating uncertainty around reported profits, dividends, and solvency. The employment effects would be immediate in sectors exposed to capital revaluation. Far from a frictionless adjustment, the process would resemble the turmoil that followed the Resource Super Profits Tax; valuation shocks, abrupt policy reversal, and a loss of investor confidence. What is modelled as smooth equilibrium would in practice be prolonged disruption.
For these reasons, the results presented by the Productivity Commission should be read with care. The modelling is consistent with its assumptions, but it is not evidence that rents are as large as claimed, that capital is as mobile as assumed, or that the promised wage gains would actually arrive in practice.
- The Economic Problems of a Cash-Flow Tax
The weaknesses of the cash-flow tax run deeper than the assumptions of any modelling. Once we leave the world of neat neoclassical theory and confront the realities of economic decision-making, finance, and administration, the difficulties of a cash-flow tax begin to multiply.
The first issue is a knowledge problem. For the tax to work as advertised the authorities must be able to distinguish between normal returns and rents. In theory that means the Australian Taxation Office must know the cost of capital for every firm and every project. In practice that information does not exist outside the firm. The problem is even more acute for smaller businesses and start-ups. Their cost of capital is uncertain, often driven by founder expectations, venture funding conditions, or market sentiment. To suppose that a regulator could impute the right figure is implausible. Any attempt to do so would create disputes and distort behaviour. The only way forward is for the state to impose a regulatory cost of capital, much as happens under rate-of-return regulation in utilities.[21] The history of that type of regulation is not encouraging; it leads to disputes, gaming, and inefficient allocation of resources. The supposed neutrality of the rent tax dissolves when the calculation problem is recognised.[22]
The second problem is finance. The proposal assumes that government can share in both gains and losses by collecting a constant share of net cash inflows and refunding the same share of outflows. But as Modigliani and Miller showed in their famous theorems, the value of the firm depends only on its real assets, not on the mix of debt and equity used to finance it.[23] The government cannot finance risky projects with risk-free instruments. Yet that is what a symmetric refund system implies; the state writes a cheque for losses with the promise that taxpayers will be made good from future rents. That violates basic finance logic. The point became obvious in 2010 when the Rudd Government attempted to introduce the Resource Super Profits Tax. Investors quickly saw that the Commonwealth could not make itself a risk-bearing partner in resource projects without undermining its own fiscal credibility.[24]
The third problem is symmetry itself. In the textbook version losses are refunded in cash and in the same period. In practice governments do not behave that way. Losses are carried forward, subject to continuity tests, and given an uplift at a rate chosen by policy. When that uplift falls short of the project discount rate, part of the normal return would be taxed. Worse, the promise of full refundability is not credible in a downturn. No government will pay out tens of billions of dollars to failing firms during a recession. Neutrality on paper disappears when political economy intrudes.[25] The credibility issue runs deeper than fiscal optics. Once investors doubt that refunds will be honoured promptly, they treat the system as asymmetric from the outset. A firm expecting delayed or partial refunds must finance losses privately, increasing its effective cost of capital. That higher hurdle discourages the very investments the reform is supposed to encourage. In a downturn the state’s obligation to refund losses becomes a contingent liability on the budget; when credibility falters, neutrality collapses.
The fourth problem is time. Even if refundability were written into law, investors would still discount it. Once projects are sunk, governments face strong incentives to renege by either altering the uplift rate, limiting refunds, or redefining the base. This is the familiar problem of time inconsistency. A tax on rent that is neutral ex ante will not be seen as neutral if investors expect the rules to change ex post. That expectation depresses investment even if the statute promises symmetry.[26]
The fifth problem is the definition of rent itself. In public-finance theory rent is a surplus above opportunity cost. But in Austrian School and Schumpeterian terms, the very same margins are the reward for entrepreneurship. An innovator who brings a new product to market enjoys a quasi-rent until rivals catch up. Recent empirical work finds that rising mark-ups in advanced economies are often returns to innovation and intangible investment rather than monopoly power.[27] Taxing these quasi-rents as if they were surplus risk-free profits is to tax the reward for entrepreneurship. The neat distinction between ‘rent’ and ‘normal return’ becomes blurred once innovation is recognised.
The sixth problem is administration. Drawing the line between real and financial flows (R versus R+F) is not straightforward. Multinationals use complex financing chains. Intangible assets move across borders. If finance is excluded from the base, boundary disputes multiply; if finance is included, the administrative burden escalates. Garnaut’s own proposal acknowledged the need for a separate Financial Sector Income Tax, import-deduction restrictions, and a decade-long transition. These are not minor adjustments; they are large new structures with their own distortions.
The final problem is the open economy. The cash-flow tax presumes that location-specific rents can be neatly separated and captured — but in practice rents are not confined to a mine site or a parcel of land. They are embedded in global supply chains, brand value, and intellectual property. The more mobile the economy becomes, the harder it is to ring-fence ‘Australian rents’. At the same time, international initiatives such as the OECD’s Pillar Two regime constrain how far Australia can deviate from global tax policy norms.[28],[29] The neat textbook idea of a rent tax clashes with the messy reality of international taxation.
Taken together these problems show why the cash-flow tax remains a theoretical construct rather than a working system. The appeal of taxing rents is clear in abstract. But once we confront the calculation problem, the finance problem, the lack of credible symmetry, the incentives to renege, the entrepreneurial nature of so-called rents, the administrative complexity, and the open economy, the neat appeal dissolves.
- What would be taxed in practice
The promise of the cash-flow tax is that it falls only on rent. In practice the burden would be quite different. What the models label as rent is not an observable category in the real economy. The tax base would end up catching entrepreneurial returns, margins on innovation, and the uncertain rewards of intangible investment.
Take a new technology firm. In the first years it invests heavily in software, design, and marketing. Under a cash-flow system those outlays are expensed immediately, so the firm records tax losses. For the system to be neutral those losses would need to be refunded in cash. If they are not, the firm carries the losses forward with an uplift set by the authorities. If that uplift falls below the firm’s own cost of capital, part of the normal return is taxed. Once the firm succeeds and earns a margin, that margin is treated as rent. Yet the same margin is the temporary reward for entrepreneurship. To tax it as rent is to tax the very process of innovation.
Consider a miner. The project requires large upfront investment, financed with a mix of debt and equity. Under a two-sided cash-flow tax the government would notionally share in both outlays and returns. In practice, governments do not refund losses contemporaneously, so the neutrality breaks down. Moreover, the so-called resource rent is highly sensitive to commodity prices. A run of good years looks like rent, but it is the return for bearing commodity risk. The system would end up taxing the premium for risk-taking as if it were a surplus.
The financial sector illustrates the same point. Garnaut’s design recognises that an R-base cannot handle banks and insurers, so it proposes a Financial Sector Income Tax. That separate tax is not neutral; it relies on allocating margins between debt, equity, and risk-bearing activities. The neat cash-flow logic collapses at the boundary.
Even in ordinary retail and service industries the margins attributed to oligopoly rents are often the fruit of intangible investment; brand, customer networks, logistics, data.[30] These investments do not show up as tangible capital, but they are costly and risky. Once they succeed, the model records a rent. In reality, they are returns to entrepreneurial judgement under uncertainty.
The effect is that a cash-flow tax in practice would not fall only on true scarcity rents such as land in prime locations. It would fall on the uncertain, temporary surpluses that make entrepreneurship worthwhile. In Austrian terms these are not rents at all, but the reward for alertness and innovation.[31] In the empirical literature they are increasingly described as the return to intangible capital.[32] Either way, they are not free surpluses that can be taxed without consequence. They are the reason firms invest, experiment, and innovate.
The conclusion is stark. The promise of a rent tax — a levy on surplus that leaves incentives untouched — cannot be delivered in practice. What would actually be taxed are the fragile margins that drive entrepreneurship and innovation in the Australian economy.
-
Conclusion
The case for a corporate cash-flow tax is built on a clean theory and an optimistic model. In theory, it taxes only economic rents and leaves the normal return to capital untaxed. In the model, more than half the company tax base is rent, and removing tax from normal returns attracts more capital, raises wages, and improves welfare. On that logic, the Productivity Commission has proposed a cautious first step towards a cash-flow system, with Garnaut and others pointing to a fuller design as the long-run destination.But once the claims are tested against economic and institutional reality, the neat appeal breaks down. The transition would not be neat. Balance-sheet revaluation, asymmetric refunds, and fiscal exposure during downturns would test both the resilience of firms and the credibility of government commitments. The efficiency gains that appear in simulation would vanish once adjustment costs, uncertainty, and political risk are taken into account. The modelling assumes away dividend imputation, foreign ownership, capital-market frictions, and the problem of transition. It treats mark-ups as rents without asking how much is innovation or entrepreneurship. It relies on a steady-state world that never arrives. These assumptions are contestable, yet they drive the results that the Productivity Commission has relied upon.
Beyond the modelling lie deeper problems. To identify rents the authorities would need to know the cost of capital for every firm, which is impossible. To guarantee neutrality they would need to refund losses promptly and in full, which is not credible. To hold to their commitments they would need to resist the temptation to renege once projects are sunk, which no government has ever done. The cash-flow tax assumes government can finance risky assets with risk-free instruments, in defiance of basic finance logic. Above all, what the tax would actually fall upon are not free surpluses but the fragile returns to entrepreneurship and innovation.
The rhetoric of rent taxation suggests a painless source of revenue. The reality is different. In practice, a cash-flow tax would strike at the very margins that drive investment and innovation in the Australian economy. It would add complexity, create new uncertainties, and open the way to new forms of political risk. Far from being a superior alternative to company tax, it would deliver new distortions while taxing the entrepreneurial energy on which future growth depends.
The problems with Australia’s company-tax system are not mysterious. The rate is too high, the system is too complex, and the administrative burden weighs most heavily on productive enterprise.[33] From the government’s perspective, the desire to raise more revenue reflects not a defect in the tax base but a failure of fiscal discipline. The underlying problem is spending, not taxation: too much public money is absorbed by non-productive and non–value-adding activities.[34] Until that is addressed, the pursuit of exotic new “efficient” taxes such as a cash-flow tax will remain an exercise in political blindness — a refusal to confront the fiscal reality that Australia’s problem lies in wasteful spending, not taxation.
Endnotes
[1] Productivity Commission, From Productivity Drift to Lift: Building a More Resilient Economy, Interim Report (Canberra: Productivity Commission, July 2025), chs. 2–3.
[2] Ross Garnaut, Reuben Finighan, Michael Anthony, and colleagues, “Replacing Corporate Income Tax with a Cash Flow Tax,” Australian Economic Review 53, no. 4 (2020): 437–454.
[3] Brown, E. C. 1948, “Business‐income taxation and investment incentives”, in Income, employment and public policy, essays in honor of Alvin H. Hansen, Norton, New York.
[4] The Meade Report remains the canonical statement of the ‘expenditure tax’ or ‘cash-flow tax’ framework. It proposed replacing the corporate income tax with a tax on real and financial cash flows, designed to exempt the normal return to capital while taxing economic rents. Meade, J.E., 1978. The Structure and Reform of Direct Taxation: Report of a Committee Chaired by Professor J.E. Meade. London: George Allen & Unwin for the Institute for Fiscal Studies.
[5] Garnaut, as above.
[6] Chris Murphy, Tax Reform Modelling: Stage 1 Results (Canberra: Independent Economics, 2025).
[7] There is a lot more devil in the detail, as I set out below.
[8] Ross Garnaut, Reuben Finighan, Michael Anthony, and colleagues, “Replacing Corporate Income Tax with a Cash Flow Tax,” Australian Economic Review 53, no. 4 (2020): 437–454. Mervyn King, “The Cash Flow Corporate Income Tax,” in The Effects of Taxation on Capital Accumulation, ed. Martin Feldstein (Chicago; University of Chicago Press, 1987).
[9] In Australia this bias is partly mitigated by the dividend imputation system, which provides resident shareholders with franking credits that offset personal tax against company tax already paid. The effect is that, for resident shareholders who can fully use the credits, the after-tax return to equity is closer to neutrality with debt. The bias remains, however, for foreign investors who cannot benefit from franking credits.
[10] Garnaut and co-authors, for example, have a proposal for a ‘Financial Sector Income Tax’ for banks and insurers.
[11] Readers will recall that this, too, was a feature of the RSPT.
[12] Chris Murphy, Tax Reform Modelling: Stage 1 Results (Canberra; Independent Economics, 2025), p. 2; and Productivity Commission, From Productivity Drift to Lift: Building a More Resilient Economy, Interim Report (Canberra; Productivity Commission, July 2025), ch. 3, which cites Murphy’s estimate that 54 per cent of the company tax base is rent, up from his earlier estimate of 41 per cent in 2018.
[13] See, for example, Peter B. Dixon, Robert B. Koopman, and Maureen T. Rimmer. “The MONASH style of computable general equilibrium modeling: a framework for practical policy analysis.” In Handbook of computable general equilibrium modeling, vol. 1, pp. 23-103. Elsevier, 2013.
[14] Murphy, as above.
[15] Readers will recall that the RSPT had a very similar mechanism.
[16] See Chad Syverson, “Macroeconomics and market power: Context, implications, and open questions”. Journal of Economic Perspectives 33, no. 3 (2019): 23-43 and Jan De Loecker, Jan Eeckhout, and Gabriel Unger. “The rise of market power and the macroeconomic implications”. The Quarterly journal of economics 135, no. 2 (2020): 561-644.
[17] Murphy (2025, p. 3) explicitly states that ‘assessing the franking credits system is outside the scope of this report’. The CGETAX model adopts a small-open-economy framework in which the marginal investor is foreign and therefore cannot claim franking credits. Dividend imputation is recognised for accounting purposes but treated as behaviourally neutral in modelling investment responses.
[18] Franking credits could in principle be attached to cash-flow tax payments, but this would blur the line between a rent tax and an income tax. A cash-flow tax is designed to exempt the normal return and tax only rents, whereas dividend imputation treats company tax as a prepayment of personal income tax on total profits. Attaching franking credits risks hybridising the system, effectively making the cash-flow tax behave like an income tax for residents while remaining a rent tax for foreigners. This would undermine the neutrality claims that proponents rely on, change the incidence of the tax, and create fiscal exposure if refunds and franking claims coincided in downturns.
[19] Richard Holden, “Cash-flow tax plan would tie Australia’s company tax with Colombia.” Australian Financial Review, 31 July 2025.
[20] James Kelly and Robert Graziani, “International trends in company tax rates—implications for Australia”, Australian Treasury, 2004.
[21] There is a similarity here with the so-called “deeming rates” used in the welfare system, where government assumes a fixed rate of return on household financial assets when calculating pension entitlements. In both cases the state substitutes an administratively determined rate for the actual return, creating distortions, disputes, and incentives to game the system.
[22] Friedrich Hayek, “The Use of Knowledge in Society,” American Economic Review 35, no. 4 (1945): 519–530; Harold Demsetz, “Why Regulate Utilities?” Journal of Law and Economics 11, no. 1 (1968): 55–65.
[23] In corporate finance this point is captured by the Modigliani–Miller theorems, set out in every standard finance textbook. See, for example, Richard A. Brealey, Stewart C. Myers, and Franklin Allen, Principles of Corporate Finance, 14th ed. (New York: McGraw-Hill, 2022), ch. 17.
[24] Sinclair Davidson, “The Resource Super Profits Tax: A Policy Disaster,” Tax Policy Journal 6, no. 1 (2010): 1–15.
[25] For a discussion of these issues in the Australian context, see John Freebairn, “Resource Rent Taxes: Reflections on the Australian Experience,” Australian Economic Review 45, no. 4 (2012): 457–465.
[26] Finn Kydland and Edward Prescott, “Rules Rather than Discretion: The Inconsistency of Optimal Plans,” Journal of Political Economy 85, no. 3 (1977): 473–491.
[27] Jan De Loecker, Jan Eeckhout, and Gabriel Unger, “The Rise of Market Power and the Macroeconomic Implications,” Quarterly Journal of Economics 135, no. 2 (2020): 561–644.
[28] OECD, Tax Challenges Arising from the Digitalisation of the Economy: Global Anti-Base Erosion Model Rules (Pillar Two) (Paris: OECD, 2021).
[29] This is, of course, a violation of Australian sovereignty, but argument is beyond the scope of this discussion. For further discussion, see Chris Berg and Sinclair Davidson. “’Stop This Greed’: The Tax-Avoidance Political Campaign in the OECD and Australia”. Econ Journal Watch 14, no. 1 (2017).
[30] See Harold Demsetz, “Barriers to entry”. The American economic review 72, no. 1 (1982): 47-57.
[31] Israel Kirzner, Competition and Entrepreneurship (Chicago: University of Chicago Press, 1973).
[32] Jonathan Haskel and Stian Westlake, Capitalism without Capital: The Rise of the Intangible Economy (Princeton: Princeton University Press, 2017).
[33] See Davidson, S. and Heaney, R., 2012, January. Effective tax rates and the political cost hypothesis: a re-evaluation of Australian evidence. In Australian Tax Forum (Vol. 27, No. 1, pp. 79-105).
[34] Davidson, S., 2021. From debt to eternity. Agenda: A Journal of Policy Analysis and Reform, 28(1), pp.75-86.
