Measuring the Wrong Thing, Fixing the Wrong Problem: Productivity, Interest Rates and Australia's Living Standards

A personal introduction

I have spent more than 45 years building and scaling businesses, and my companies have placed in the BRW Fast 100 four times. In none of them was productivity a number on a spreadsheet. It was better tools, better systems, better-trained people, and the confidence to invest in them.

That is why the current productivity debate troubles me.

On 29 September the Reserve Bank raised interest rates for the fourth time this year, and the Governor told the country that productivity is doing nothing. Weak productivity has become one of the main justifications for pushing rates higher and unemployment up. If the measure is wrong, or only half right, we are making some of the most consequential economic decisions in the country with a faulty instrument.

There is good reason to think it is at least half wrong. The standard measure divides GDP by hours worked. It cannot see a better hip replacement, a safer aged care home, or software that does in seconds what used to take a week. As more of our economy moves into care and knowledge work, more of what we produce falls outside the ruler.

But I don't accept the comfortable view either. Outside the data centre boom, business investment has been weak for more than a decade. Competition in key markets is thin, housing is too slow and too expensive to build, and too many good Australian ideas never reach scale. Those problems are real, and every rate rise makes them worse.

This paper looks honestly at both sides: what the numbers say, where they mislead, and what Australia should do about it. It is a companion to my earlier paper on interest rates, because the two problems are now tied together. We are using a broken ruler to justify a blunt instrument, and the people paying for it are the same households in both cases.

Australia can do better on both counts. The tools exist. What we need is the willingness to measure honestly, invest seriously and stop treating productivity as a slogan.

Mark Falzon

Executive summary

Australia has two productivity problems, and it is treating them as one. The first is real: outside a narrow boom in data centres, the economy is not investing, competing or building enough to lift output per worker. The second is a measurement problem: the standard ruler cannot see much of what a modern care and knowledge economy produces. Policy is now leaning heavily on the flawed number, using it to justify higher interest rates, while doing too little about the real problem. Higher rates then make the real problem worse.

Current reality

• On the official measure, productivity has stalled. GDP per hour worked fell 0.2% over the year to June 2026 and is back at its December 2019 level. The Productivity Commission describes the trend as the weakest in 60 years, against a pre-pandemic 20-year average of 1.2% a year.
• The averages hide a two-speed picture. Market sector productivity has grown 7.5% since 2014. Non-market productivity (health, education, care, public administration) has fallen 2.5% and is below its 2007 level. Yet when the Productivity Commission adjusted health care for quality, it found productivity growing about 3% a year.
• Investment is narrow. Business capital spending is up 10.7% on a year ago, but much of it is imported server equipment for data centres. The Productivity Commission Chair notes that investment more broadly has never recovered from the GFC.
• The flawed number is setting interest rates. The RBA assumes trend productivity of 0.7% a year, which caps the wage growth it regards as safe at about 3.2%. Unit labour costs (wages adjusted for productivity) are running at 3.6%, and the RBA reads that as inflationary pressure. If true productivity is higher than measured, labour costs are overstated, and rates are higher than they need to be.
• Rate rises feed the slowdown. Higher rates cut business investment, stall housing construction and slow the movement of workers into new, more productive firms. A weak productivity number justifies higher rates, which produce weaker productivity.
• The cost is large. Former Treasury Secretary Ken Henry has estimated that slow productivity growth has cost workers around $500,000 in forgone pay rises over the past quarter century.
• Reform is happening, but slowly. Government reports more than 60 reforms since the August 2025 Economic Reform Roundtable, and the Productivity Commission made 47 recommendations in December 2025. As its Chair puts it, productivity is a game of inches.

The vision: measure honestly, invest seriously, share the gains

Australia should stop arguing about a single headline number and start managing productivity like a business manages performance: with better measures, clear targets and sustained investment. That means fixing the ruler, so the care and knowledge economy is counted properly; keeping a flawed figure from driving interest rate decisions on its own; and attacking the real barriers of weak investment, thin competition, slow housing construction and slow technology adoption. It also means making sure the gains flow to workers as higher real wages, because productivity that does not lift living standards will not keep public support.

Next steps: what Australia should do now

The rest of this paper sets out the evidence and the detail behind each step.

1. Where Australia's productivity stands in 2026

On the official measure, Australian labour productivity has gone nowhere for six years. GDP per hour worked in June 2026 was at the same level as in December 2019: output and hours worked have both grown about 14% since then, and almost nothing has been gained per hour (Aard, from ABS data).

ABS Australian System of National Accounts 2024-25 (financial years); ABS National Accounts June quarter 2026 (2025-26 = year to June quarter); averages from PC and EY; assumption from RBA SMP August 2026

The recent run sits well below both the pre-pandemic average and the 0.7% a year the RBA now assumes in its forecasts. The pandemic years are distorted (hours collapsed faster than output in 2020-21, then recovered in 2022-23), but the picture since is consistently flat to negative.

Three things stand out

The slowdown predates the pandemic. The Productivity Commission Chair, Danielle Wood, dates the real break to 2016, when productivity essentially flat-lined. This is a decade-long problem, not a post-COVID hangover.

The two sectors are moving in opposite directions. Market sector productivity has grown steadily over the past decade. The decline is concentrated in the non-market sector, which has also grown rapidly as a share of the economy. That combination drags the headline down even if no individual worker is less productive (a composition effect), and it is exactly where measurement is weakest.

Investment is booming in one place only. Capital spending is up strongly, but the ABS attributes most of the surge to server racks and processing equipment for data centres, much of it imported. Beyond data centres and renewable energy projects, the investment outlook is subdued, partly because of higher interest rates (AAP). Imported servers add capacity, but on their own they do little for the productivity of the wider workforce.

2. Why productivity matters: wages, inflation and interest rates

Productivity is not an abstract statistic. Over time it is the only sustainable source of higher real wages, and in the short run it sets how much wage growth the RBA will tolerate before it raises rates. That second link is why the measurement debate matters so much right now.

The arithmetic the RBA uses

The RBA's own framework says long-run wage growth consistent with full employment and inflation on target equals the 2.5% inflation midpoint plus productivity growth. At that rate, labour costs per unit of output (unit labour costs) rise in line with target inflation and firms face no extra pressure to raise prices (RBA).

That makes the productivity assumption a direct input into how high interest rates go.

In August 2025 the RBA cut its trend productivity assumption from 1.0% to 0.7%. That one decision lowered the bar for "safe" wage growth by 0.3 percentage points and lowered its estimate of how fast the economy can grow without inflation. Both feed directly into the case for higher rates.

Why mismeasurement matters

Unit labour costs are calculated by subtracting measured productivity from labour cost growth. If productivity in the care and knowledge sectors is understated, unit labour costs are overstated by the same amount, and the economy looks more inflationary than it is. The Productivity Commission's own work found quality-adjusted productivity in health care growing about 3% a year, far above what the headline measure records for the non-market sector (Cambridge, citing the PC).

It would be wrong to claim this explains the whole gap: the market sector, which is measured reasonably well, has also slowed. But even a modest understatement, applied across a non-market sector that now employs a large and growing share of Australians, is enough to change how tight the labour market looks. And that is exactly the judgement on which the September rate rise rested.

Why it matters for living standards

Over the long run the link is simpler. Real wages can only rise sustainably if each hour of work produces more. Former Treasury Secretary Ken Henry estimates that the productivity slowdown has cost workers around $500,000 in forgone pay rises over the past 25 years (ABC). Economist Chris Richardson estimates that if productivity growth stays at 0.7% rather than Treasury's 1.2%, the annual budget balance could be about $40 billion worse by 2034-35 (same source). Real net national disposable income, the broadest measure of material living standards, did not grow at all in the June quarter.

The RBA's own submission to the Senate Select Committee on Productivity puts it plainly: productivity makes goods and services cheaper, supports real wage growth and drives growth in real incomes per person (RBA). On that, every side of the debate agrees.

3. Why productivity has stalled: separating the ruler from the reality

Both camps in this debate are partly right. Some of the slowdown is an artefact of how we measure. Some of it is real, and the real part is where policy should focus.

Where the ruler is broken

Quality is invisible. In health, education and care there is no market price, so output is largely measured by inputs or activity (patients treated, students enrolled, hours of care). A treatment that cures faster, a school that teaches better or an aged care home that keeps residents safer can all show up as flat or falling productivity. The Productivity Commission concedes traditional measures miss quality, and found that adjusting for it lifts health care productivity growth to about 3% a year.

The economy has shifted toward what is hardest to measure. Non-market health, education and care services have grown particularly fast over the past decade, with the NDIS and aged care expansion at the centre. This shift lowers the headline average even if productivity within every sector were unchanged. The RBA itself acknowledges the growing non-market share as one of the factors behind the slowdown.

Digital and intangible output is undercounted. Free online services, better software and improvements in quality that do not change prices are poorly captured everywhere. International researchers such as Diane Coyle in the UK have documented the gap between measured output and what a digital economy delivers.

Mining distorts the average. The RBA notes sharp falls in mining productivity as companies tap lower-grade deposits. That is a real cost for those firms, but it says little about the productivity of the rest of the workforce.

Where the problem is real

Weak investment. Danielle Wood points to business investment that has not recovered since the GFC, meaning less technology and less capital per worker (CommBank). The current data centre surge is real but narrow and import-heavy.

Fading business dynamism. The RBA's research shows fewer workers moving to new and young firms, which slows the shift of labour and capital to more productive businesses. Lower dynamism is linked to weaker competition in concentrated markets (RBA).

Housing and construction. Construction productivity has lagged for decades. Slow approvals, fragmented building codes and high costs mean Australia builds too few homes too slowly, which also keeps workers from moving to where the best jobs are.

Slow technology adoption. Australia is good at buying technology and less good at using it. Many small and medium firms trail the frontier in digital and management practices, and the gap between the best firms and the rest is wide.

Regulation and energy. Overlapping approvals, state-by-state rules and an energy transition that has raised costs during the changeover all weigh on output per hour.

Interest rates themselves. Higher rates raise the hurdle for new investment, stall housing construction and discourage the start-ups and expansions that drive dynamism. A tool used to compensate for weak productivity also suppresses the investment that would lift it.

What the optimists get right, and what they get wrong

Recent commentary, including an Independent Australia analysis published on 12 October 2026, argues that productivity is actually strong and only the measure is broken (Independent Australia). It is right that the ruler is broken in important places, right to challenge the RBA's reliance on it, and right that strong employment and solid market sector performance contradict the picture of a failing economy.

Some of the evidence it relies on does not hold up, though, and the case is stronger without it. Strong job creation is not proof of productivity: if output and hours both grow 2%, employment rises while output per hour stays flat. Record luxury spending and imports reflect income and wealth at the top as much as productivity across the workforce. And the investment surge is real but concentrated in data centres. The credible position is that Australia's productivity is understated and also underperforming. Both need fixing, and the first should not be used to excuse the second.

4. Australia's history of productivity reform

Australia has lifted productivity before, and the lesson of each period is the same: big gains came from sustained, bipartisan reform that opened markets and encouraged investment, not from any single measure. Figures below are approximate and drawn from the historical record.

What the record says

The 1990s show what works. National Competition Policy succeeded because the Commonwealth paid the states to reform, the reforms were monitored independently, and the program ran for a decade across governments of both colours. The revived National Competition Policy, which again pays states to fix poorly designed regulation (Inside Small Business), is the closest thing Australia has had to that model in 30 years.

The 2016 to 2025 period shows what fails: reviews that are commissioned, praised and shelved. The Treasurer says more reform has been delivered in the year since the Roundtable than in the previous quarter century, with more than 60 reforms under way (ALP). That claim will be judged by the productivity data over the next five years, not by the count of announcements.

5. Global practice

Australia is not alone. Most advanced economies have seen productivity slow since the mid-2000s, and several have done more than Australia to measure it properly and to lift it. The examples below are drawn from the historical record (approximate).

The common thread

Countries that have done better share three habits. They measure the services sector honestly, so they know where the problem actually is. They keep markets contestable, so new firms can challenge old ones. And they treat technology adoption by ordinary firms as the main game, rather than invention by a few. None of them relies on interest rates to compensate for weak productivity, because no central bank can raise output per hour by cutting demand.

6. AI and emerging technology

AI is the largest productivity opportunity Australia has had in a generation, and right now it is showing up in the statistics mainly as a cost. How the country handles the next five years will decide which of those it turns out to be.

The opportunity

The Productivity Commission estimates AI could add about 4% to labour productivity over the decade (CommBank). That is more than all the labour productivity growth recorded across the whole economy since 2014 (3.6%). The gains are largest in exactly the sectors where productivity is weakest or hardest to measure: health administration, aged care documentation, legal and financial services, education and public administration.

Why it isn't showing up yet

The build comes before the benefit. Data centre investment has risen for eight straight quarters and dominates business investment. That spending adds to demand and, as the RBA notes, to capacity pressure now. The productivity payoff comes later, when firms across the economy use what is being built.

Adoption is uneven. Large firms are moving quickly; most small and medium businesses are not. Productivity gains from general-purpose technologies have historically come from diffusion through ordinary firms, which takes skills, management capability and confidence to invest. That is the gap Australia most needs to close.

The ruler may miss it. A clinician who uses AI to cut paperwork and spend more time with patients produces better care, not more measured output. A lawyer who produces the same advice in half the time may bill fewer hours. Much of AI's early value will show up as quality and time saved, which is precisely what the current measures fail to see. Without better measurement, AI could make the economy more productive while the headline number barely moves.

The risks

AI will also displace work, and the costs will fall unevenly. If productivity gains flow mainly to capital and a few large firms, the result will be higher profits without higher wages, which is the pattern many Australians already suspect. Policy needs to support transitions for affected workers and ensure the gains are shared, or public consent for the technology will not last.

What it means for interest rate policy

The RBA currently assumes productivity will recover only to 0.7% a year. If AI delivers anything close to the Productivity Commission's estimate, that assumption is too pessimistic, the economy can grow faster without inflation, and current rate settings are tighter than they need to be. At a minimum, the RBA should publish how its forecasts change under an AI-driven productivity scenario, so the public can see how much of the current tightening rests on a pessimistic assumption.

7. The recommendations in detail

The recommendations fall into two groups. The first fixes how productivity is measured and used in policy, so a flawed number stops driving interest rates. The second attacks the real barriers. Neither works without the other.

Fixing the ruler

1. An official quality-adjusted productivity series. The ABS and the Productivity Commission should publish, alongside the headline measure, productivity for health, education and care adjusted for quality and outcomes, following the UK model. The Productivity Commission has already shown it can be done for health care. Publishing it officially, every year, would give policymakers and the public a truer picture of where the economy is and is not improving.

2. Stop a single number from setting interest rates. The RBA should publish how its assessment of labour market tightness and its rate settings change under alternative productivity estimates: the current 0.7% assumption, the pre-pandemic 1.2% average, a quality-adjusted measure and an AI scenario. It should give less weight to headline unit labour costs while the non-market sector distorts them, and say so openly. This sits naturally with the Supply Shock Protocol and distributional reporting proposed in my interest rate paper.

Attacking the real barriers

3. Tax that rewards investment. Redesign business tax so it rewards firms that invest, with faster write-offs for productive capital and particular attention to small and medium firms, which face the highest effective cost of capital. The Productivity Commission has put forward options for a company tax redesign along these lines; government should act on them.

4. Competition with teeth. Fully fund and implement the revived National Competition Policy, using the 1990s model of payments to states tied to verified reform. Use the new mandatory merger regime to stop further concentration in markets that are already thin, and give the ACCC the resources to act on it.

5. AI and digital adoption for small business. A national program focused on diffusion, not invention: co-funded adoption grants, practical training for owners and managers, and sector-specific playbooks, along the lines of Singapore's model. The goal is to get the tools already available into the hands of the hundreds of thousands of firms that are not yet using them.

6. Build faster and cheaper. Housing and construction need a national effort: planning reform with firm approval deadlines, a single harmonised building code, recognition of modular and prefabricated building across all states, and skills programs for the trades. Faster building also lets workers move to where the best jobs are.

7. Pay for outcomes in care. The Productivity Commission's five pillars inquiry called for breaking down the siloed approach to the care economy. Funding for health, aged care and disability services should reward outcomes rather than activity, and should fund the technology that frees care workers to spend more time on care.

8. Strengthen business dynamism. Build on the planned restrictions on non-compete clauses for most workers, simplify the rules for starting, growing and closing a business, and make it easier for capital and workers to move to the firms that use them best.

9. Cheaper, reliable energy. Deliver the energy transition at least cost, as the Productivity Commission has urged, so business has the affordable, reliable power that productive investment depends on.

10. A Productivity and Living Standards Statement. Treasury and the Productivity Commission should publish an annual statement that tracks headline and quality-adjusted productivity, real wages, the labour share of income and progress on every reform. Productivity has to be seen to lift living standards, or public support for reform will not last.

8. Risks, counter-arguments and who shares the gains

Who shares the gains

Productivity reform has a credibility problem with ordinary Australians. For years they have been told that productivity is the path to higher wages, while profits grew and real wages went backwards. The current period, with record bank profits, falling real wages and an RBA citing weak productivity as a reason to push unemployment up, reinforces that suspicion.

The fix is to make the gains visible and shared: higher real wages, cheaper services, shorter waits for care, more affordable housing. Reform that delivers those things will have support. Reform that is used mainly to justify wage restraint and higher interest rates will not, and should not.

9. Conclusion

Australia's productivity debate has become a contest between two slogans: "productivity is broken" and "the measure is broken". Both are partly true, and neither is a policy. The official number understates what a modern care and knowledge economy produces, and it is being given too much weight in decisions about interest rates and wages. At the same time, investment outside data centres is weak, competition is thin, housing is slow to build and technology spreads too slowly through ordinary firms.

The way forward is to separate the two problems and act on both. Fix the ruler, so the country knows where it is really improving. Stop a single, flawed figure from driving rate rises that suppress the very investment productivity depends on. And get on with the reforms that lift output per worker: tax that rewards investment, competition with teeth, faster building, practical AI adoption for small business, and care funded for outcomes.

Australia did this in the 1990s, when sustained, funded and bipartisan reform delivered some of the strongest productivity growth in the developed world. The country now has better data, better tools and a technology wave that could lift productivity more than anything since. The difference between a decade of stagnation and a decade of rising living standards is whether we measure honestly, invest seriously and make sure the gains reach the people doing the work.

Open questions for further work

• How large is the measurement gap across health, education, aged care and disability services combined?
• How would the RBA's assessment of unit labour costs change under a quality-adjusted productivity measure?
• How much of current business investment, outside data centres, is going into productivity-raising capital?
• What share of small and medium businesses are using AI in core operations, and what is holding the rest back?
• How has the labour share of income moved since 2016, and by sector?

Sources

Data accessed 16 October 2026. Historical figures in sections 4 and 5 are approximate and drawn from the public record; they should be checked against primary sources before publication.

• ABS, Australian National Accounts, June quarter 2026
• ABS, Australian System of National Accounts 2024-25
• ABS, Private New Capital Expenditure, June 2026
• Productivity Commission, Productivity update, September 2026
• Productivity Commission, Quarterly productivity bulletin, March 2026
• Financial Standard, Productivity Commission makes 47 recommendations
• CommBank, interview with Danielle Wood, June 2026
• RBA, Statement on Monetary Policy, August 2026, Economic Conditions
• RBA, Drivers and implications of lower productivity growth, August 2025
• RBA, Submission to Senate Select Committee on Productivity in Australia, February 2026
• Cambridge, Is the public sector dragging down living standards?
• EY, Australian National Accounts June 2026
• Aard, How productive is the Australian economy?
• Bill Mitchell, Australian national accounts
• AAP, Data centres drive record $6 billion investment boom
• ABC, Chalmers's economic reform roundtable, August 2025
• ALP, 12 months of economic reform
• Inside Small Business, One year on, has the roundtable paid off?
• Independent Australia, A picture guide to how productivity is making Australians better off