Do we need a new reform playbook?

15 October 2026

Deputy Chair Alex Robson delivered the keynote address to the CropLife Australia Members Forum on 14 October 2026.

Introduction

Productivity growth sets the speed limit for the economy. When it is stronger, wages and living standards can grow faster without adding to inflationary pressures. 

Over the past decade, Australians have had very little of it, and we have felt it.

Most of the ideas for lifting productivity are not new. What has changed, and what keeps changing, is the world those ideas have to work in.

Reform seems to be getting harder, and so the question I want to pose today is a simple one: do we need a new reform playbook? And what might that look like? I don’t have a complete answer, just the beginning of one. 

Today my remarks will begin with a bicycle and an English clergyman. And no, it's not the beginning of a lame dad joke.

I'll then set out three problems that make reform harder, each paired with the area of policy where it bites hardest: regulation, tax and, finally, government spending. 

Let’s begin by reminding ourselves of what productivity growth has actually delivered. 

In 1901 a bicycle cost about 473 hours of work, close to three months' wages. For a working Australian, that is roughly what a modest new car would cost today.

Now it costs about six hours. And today's bike is better.

Over a century, productivity growth turned a major purchase into something that costs less than an average day's pay.

In purchasing power terms, productivity growth effectively transformed a bicycle into a car.

Going from 473 hours to six means the cost of a bicycle, measured in work, fell by about 3.6 per cent a year on average for 125 years: a nearly 80-fold fall. Or, put differently: a nearly 8,000 per cent increase in productivity.

Had bicycle productivity instead risen by only 1.2 per cent a year, a bicycle would cost about 106 hours today: nearly three weeks of work, still a sizeable purchase.

That’s an important lesson: small differences in growth rates, held for long enough, become very large differences in how we live.

A corollary is that slower growth won't make us poorer. The more plausible outcome is that we will still become better off, just more slowly.

A much older, more frightening and pessimistic thought experiment was conducted by Malthus (1798). He described an economy from which productivity growth was permanently absent. Land was fixed but the number of mouths to feed was not, so more people meant less to eat for each.

Malthus and his modern followers were and are badly wrong. We are not heading towards the world he described. Why he was so wrong, and how we keep it that way, is what I want to discuss today.

There's a lot of discussion these days about a lack of economic reform in Australia. It can sometimes be tempting to put this down to a lack of political will or courage.

I think that is entirely the wrong place to look. Anyone who has spent time in a room with our leaders - of any political stripe - knows they have will and courage in spades. None of them are shrinking violets. 

I am confident that the constraint lies elsewhere. So a chapter on “How to get your reform mojo back” doesn’t need to be part of any new reform playbook. Instead, I want to focus on three realities that I think such a playbook would need to confront. 

Abundance, superabundance and anti-abundance

The first is what I'll call the abundance observation: much of the scarcity we observe is due to policy choices. Housing, transmission lines and new products are rarely short of capital or demand. They are short of permission.

A big part of the problem is too many independent veto points. Each veto holder puts a price on its own consent, and supply falls with every veto that is added. Economists call this the problem of the anticommons.

A separate problem is delay, which is akin to a tax that raises no revenue. But when the tax is levied in terms of time and compliance, no value is transferred; it is simply destroyed. And it rarely shows up directly in price statistics although there are exceptions to that -   housing being a good example. 

At the same time, some things are getting radically cheaper. AI may make intelligence itself close to free - an era of superabundance. But the catch here is what is complementary to intelligence. When one input becomes super-abundant, what it is used with becomes more valuable: land, energy, connections, approvals. So the cheaper artificial intelligence becomes, the more a grid connection or a zoned site is worth. 

And that creates anti-abundance. Scarcity creates rents, rents are capitalised into asset values, and asset owners become a constituency for scarcity. 

The bigger the rent, the harder it is defended. Super-abundance raises the stakes, so the fight over scarcity gets bigger, not smaller.

This is where regulation enters the picture. The details of a reform playbook for regulation are set out in the Commission's Pillar 1 work, so let me focus on how we think about it.

Regulation is spending and taxation by other means, except that its cost never appears in a budget. It is easy to impose but hard to see. Each rule can pass on its own merits while the stack as a whole fails, because rules are assessed one at a time but are borne all at once. 

Compliance costs are largely fixed, so they weigh most heavily on new and small firms, which means regulation can protect incumbents without anyone intending it to. 

Many small vetoes add up to a very big no: each may be sensible, but too little gets built. And rules drift over time. Written for one technology and one era, they can become barriers later, and prescriptive rules lock in old methods in a way that outcome-based rules do not.

In practice, that means leading with clear economy-wide regulatory objectives and targets. It means regularly assessing the stock of regulation, not just each new rule as it is proposed, with a sector-by-sector stocktake of the cumulative burden. 

It also means cutting veto points in approvals, with single entry points, statutory deadlines and fewer blockage points. And it also means sunsetting appropriately.

A modernised Concertina rule

The next reality starts with what economists call the second-best trap. With many distortions in place, removing one need not make us better off. Taken literally, that can be an argument for doing nothing.

Hatta's (1977) answer was that piecemeal reform can still be relied on if it follows a rule. The best known is the concertina rule: bring the highest distortion down towards the next highest and then take these two and reduce them to the third, and so on. The distortions collapse like a concertina. 

Under mild conditions, each step improves welfare. Australia's first waves of modern reform looked a bit like this: large, visible tariffs and price wedges, cut from the top down. Floating the exchange rate. Freeing up capital and product markets. Labour market reform. The ranking was relatively straightforward because many of the wedges could be identified and measured. 

The modern difficulty is that many of our remaining distortions are regulatory: approvals, licensing, compliance, delay. They are hard to rank, because their cost is value destroyed rather than a visible price gap. And the concertina rule works best when you can hear which note is being played the loudest.

Each remaining reform is also small on its own, often smaller than the random noise in the economy, so no single one is worth a political fight. 

This means that policymakers have to bundle reforms. This can work: the GST replaced a collection of nuisance taxes whose combined cost was large. But bundling the gains also means bundling the losers. Each reform's costs fall on an identifiable group, while the gains are spread thinly and arrive later. A bundle is only as strong as its ability to make the gains visible and to share them with those who bear the costs.

When the remaining reforms are small, scattered and hard to measure, someone has to turn a list into a coherent idea that people can see and support. That is a different and much harder task, and it gets harder as the agenda gets more dispersed. 

So part of the revised playbook is how and when to bundle, as well as better measurement and quantification, which are critical for reform. And identifying the reform dividends - how and why and when a reform will improve people's lives. 

You cannot rank what you cannot measure, and you cannot sell gains that people cannot see. Quantifying regulatory costs is what helps governments prioritise and apply a concertina rule to regulation, and that is what a stocktake and a burden-reduction target attempt. 

The same logic applies to tax reform. Switching between taxes with similar costs achieves little; the gain comes from moving away from the most damaging ones. 

Not all taxes are equal. Stamp duty taxes moving house, not building one, so it is a tax on worker mobility and on reallocation. Our remaining tariffs now cost more in nuisance and compliance than they raise in revenue. Taxes on the returns to saving compound with the time horizon. Personal income tax affects more than hours worked; it shapes investment in skills, innovation, and where talented people choose to work. 

Thresholds shape behaviour too: a business concession that ends at a size threshold can act as a tax on growth. 

And just like regulation, tax systems can drift, so the system must keep earning its place. In practice, every tax should be judged by its economic cost per dollar raised, and the mix should shift towards lower-cost bases over time. The bigger the gap between the tax being cut and the one replacing it, the bigger the gain. 

As part of a new reform playbook, governments could consider regularly estimating and publishing the economic cost of each tax per extra dollar raised, the productivity impacts of tax changes, their compliance and administrative costs, and the buoyancy of each tax and of the system as a whole: how revenue responds as the economy grows.

Baumol, amplification and government spending

Finally, there is the reality of William Baumol's cost disease and what it might mean for government spending and taxation.

Baumol's insight was that a string quartet takes the same four players and the same time as it did in Beethoven's day. 

In pure labour terms, productivity of the quartet is fixed, forever. You don't get a better performance by going from adagio to presto, or from pianissimo to fortissimo. And you don't raise productivity by sending the first violin home.

To match the bicycle's productivity gain in a pure accounting sense, a string quartet would have to play a 40-minute quartet in 30 seconds. Or one violinist would have to play all four parts in two minutes.

Those are obviously absurd prescriptions. The point is that if you are looking for productivity improvements in services, shedding labour and replacing it with capital is often not the right place to begin. 

In fact, the big gains in music productivity looked entirely different and came from somewhere else: recording, broadcasting and streaming. The labour needed to perform stayed the same, but the labour per listener collapsed. 

The key lesson is: in services, the productivity gains often come from changing how the service is delivered and altering the consumer experience, not from doing it faster or with fewer people.

But here's the kicker: the fiscal reality is that governments fund many services, so Baumol’s reality means that the cost of delivering the same services tends to rise faster than the economy. And if demand for those services is unresponsive to price, spending as a share of GDP can rise - even when there are no new programs.

This can lead to some unpleasant fiscal arithmetic: the extra spending has to be financed by taxes, and we know that the cost of raising a dollar rises more than proportionally with tax rates. So the cost disease, as bad as it is, can be amplified by the way we pay for it, via taxes.

This means that productivity in government services matters twice: directly, and through the taxes needed to fund them.

So what to do? First, diffusion of best practice across jurisdictions is a potentially massive prize: the gap between the best and the rest in service delivery is large and measurable.

Second: bending the curve. When you can see the cost disease coming down the pipeline, prevention, or bending the spending curve, takes on a new kind of urgency. But this kind of “prevention as fiscal self-insurance” demands greater fiscal discipline, not less, including not fully booking the expected future savings before they are realised, as tempting as that can be. It may require revisiting some of our fiscal rules and architecture.

Third, there is the issue of government spending more broadly - and I’ll close with this.

There is an ongoing debate about the size and growth of government spending and its effects on inflation and interest rates.

That debate matters, but I do not propose to enter into it today.

Instead I want to raise a point that seems to have gone missing recently, and to do so at the level of principles rather than specific programs.

The point applies to every level of government, of every persuasion, and it is this: the composition and quality of spending matter.

Two governments of the same overall size can leave citizens in very different positions and have very different economic impacts.

There are two basic questions to ask of every dollar spent. 

• What is it worth to the people who receive it? 
• And what does it add to the economy's productive capacity?

Put those questions together and spending falls into three broad categories: wasteful, unproductive and productive.

Wasteful spending 

Wasteful spending fails even before we count the cost of raising the revenue needed to fund it. What it is worth to the people who receive it and what it adds to productive capacity amount to less than the dollar spent. 

Spending that scores poorly on both questions takes resources from where they were worth more and puts them where they are worth less.

At the extreme, some spending doesn't just waste money; it can actually reduce output, for example subsidies that keep workers and capital in relatively low value uses. That kind of spending adds to demand while reducing what the economy can supply, and it shrinks the tax base that pays for everything else.

Unproductive spending

Unproductive spending is harder to spot. It may deliver real value, perhaps more than a dollar for every dollar spent, but not enough to clear the bar once the cost of raising the revenue is counted. 

And that bar is not fixed. It rises as spending and taxes rise: every additional program has to clear an ever higher bar, delivering more than a dollar of value for each dollar it spends. 

Because all programs draw on the same taxes, when big programs cost more, taxes go up, every dollar becomes more expensive to raise, and every other program has to clear a higher bar.

Some spending can cross the unproductive threshold over time. Downward drift works on both sides of the ledger: each dollar buys less service, while the cost of raising each dollar goes up. So spending that started out productive can quietly become unproductive.

Productive spending

Finally: productive spending. It clears the bar, and the best of it adds to the economy's productive capacity. And while spending that genuinely lifts productive capacity can itself generate some tax revenue, there is no magic pudding.

Public investment is where these questions can matter most, because well-chosen infrastructure can raise the productivity of everything that uses it. That puts a premium on project selection, with rigorous, published cost-benefit analysis before the project is committed to, rather than after. It also puts a premium on delivery, because delays and cost overruns eat away at a project's value after it is approved, shifting the benefit-cost calculus in the wrong direction. And ex-post evaluation also becomes more critical.

Conclusion

To conclude: three realities make the job of productivity-enhancing reform harder:

• Abundance raises the value of scarcity and the strength of those who defend it. 
• The remaining reforms are hard to rank and hard to sell. 
• And services cost more each year, while funding them costs more still. 

None of these realities is a reason for despair. They are not insurmountable. We just need to be clear eyed about them.

Each points to the same discipline: define the problem to be addressed, count the full costs and benefits, and don’t stop evaluating. In that sense, the new reform playbook is not very different from the old one. 

Malthus was wrong because human ingenuity and productivity growth kept pushing out the frontier. Our task is to keep proving him wrong, and we shouldn’t let the problems I’ve outlined today stand in our way.

References

Bailey, Martin J. (1971), National Income and the Price Level: A Study in Macroeconomic Theory, 2nd edition, McGraw-Hill, New York. 

Barro, R. (1981), "Output Effects of Government Purchases", Journal of Political Economy, 89(6): 1086–1121.

Barro, R. (1990), "Government Spending in a Simple Model of Endogenous Growth", Journal of Political Economy, 98(5, Part 2), pp. S103–S125.

Baumol, W. and Bowen, W. (1965) “On the Performing Arts: The Anatomy of Their Economic Problems,” American Economic Review, 55, no. 1/2: 495–502.

Baumol, W. (2012) The Cost Disease: Why Computers Get Cheaper and Health Care Doesn't, Yale University Press. 

Buchanan, J. and Yoon, Y. (2000) "Symmetric Tragedies: Commons and Anticommons," Journal of Law and Economics, 43 (1): 1–13.

Hatta, T. (1977) “A Theory of Piecemeal Policy Recommendations,” Review of Economic Studies, 44(1): 1-21. 

Heller, M. (2010) The Gridlock Economy: How Too Much Ownership Wrecks Markets, Stops Innovation, and Costs Lives, New York: Basic Books. 

Klein, E. and Thompson, D (2025) Abundance: How We Build a Better Future, New York: Avid Reader Press. 

Malthus, Thomas (1798) An Essay on the Principle of Population, as It Affects the Future Improvement of Society, with Remarks on the Speculations of Mr. Godwin, M. Condorcet, and Other Writers. London: Printed for J. Johnson, in St. Paul’s Church-Yard. 

Nordhaus, W. (1992) "Lethal Model 2: The Limits to Growth Revisited,” Brookings Papers on Economic Activity, 23, 2: 1–60. 

Nordhaus, W. (2006) “Baumol’s Diseases: A Macroeconomic Perspective,” NBER Working Paper No. 12218. 

Productivity Commission 2026, Trade and Assistance Review 2024–2025, Annual report series, Canberra 

Tupy, M. and Pooley, G. (2022). Superabundance: The Story of Population Growth, Innovation, and Human Flourishing on an Infinitely Bountiful Planet. Cato Institute.