Australia enters 2026 with an economy running close to its speed limit - and that speed limit has fallen. The collapse in multifactor productivity (MFP) growth, structural shifts in the composition of the economy, and a more fraught global backdrop all point to an extended period of modest growth, sticky inflation, and a Reserve Bank less willing to provide accommodation than at any time since the global financial crisis.
The productivity problem
The most significant development in Australia's economic landscape is the deterioration in MFP - the residual productivity gains that come from efficiency improvements, innovation, and better allocation of resources. Our estimates, derived using an HP filter on national accounts data, show MFP growth has turned negative since 2022, a marked departure from the modest but positive growth that prevailed through most of the 2010s.
Several factors are driving this decline. Capital shallowing is perhaps the most important: in FY23, the capital-to-labour ratio fell by 4.9% - the largest decline on record. Strong employment growth, driven in part by high immigration, has not been matched by business investment. Workers have less capital to work with, and productivity suffers as a result.
The composition of the economy has also shifted in ways that weigh on aggregate productivity. The care economy - health, aged care, disability services, childcare - has grown from around 10% of employment in the early 2010s to roughly 15% today (Maltman and Rankin 2024, e61 Institute). The shift continues: employment in health care and social assistance grew by 112,000 (or 4.9%) over the year to August 2025 (Jobs and Skills Australia), and the sector now accounts for 57% of job gains over the past year (Indeed Hiring Lab). These are labour-intensive services where measured productivity gains are inherently difficult to achieve, and where regulatory requirements often prescribe how services must be delivered.
Meanwhile, Australia has lost high-productivity, capital-intensive, trade-exposed sectors. Manufacturing now accounts for just 5% of GDP - among the lowest share in the OECD. The mining boom of the 2000s, and the associated surge in the Australian dollar, accelerated this structural shift.
More recently, soaring energy costs on the East Coast have made it harder for remaining trade-exposed industries to compete. The green energy transition, while perhaps necessary, adds to near-term cost pressures - capital is being redirected toward replacing existing energy infrastructure rather than expanding productive capacity. The result is an economy increasingly tilted toward services with low productivity growth potential.
A note on AI
Artificial intelligence will eventually reshape productivity dynamics, particularly in knowledge-intensive industries. But technological revolutions take time to diffuse through the broader economy - the productivity gains from electrification and computing took decades to materialise at scale. For the forecast horizon relevant to this outlook, it is difficult to see AI delivering a broad-based productivity uplift that materially changes the supply-side constraints described above. It remains a reason for long-run optimism, not a near-term solution.
A lower speed limit
The practical implication is that Australia's potential growth rate has fallen. With weak labour force growth and near-zero or negative MFP, we estimate potential growth at around 2-2¼%. This means that GDP growth of approximately 2% is not underperformance - it is roughly "speed limit" growth, leaving the output gap close to flat.
The decline in potential growth has a direct corollary: a lower neutral interest rate. If trend growth has fallen from ~3.5% to under 2%, the real rate consistent with balanced growth has fallen with it.
This matters for policy. In a world where potential growth was 3% or higher, the economy could absorb demand shocks without generating inflationary pressure. That buffer has shrunk. The economy is now more inflation-prone for any given level of demand growth, and policymakers have less room to run the economy hot in pursuit of full employment.
What this means for wages
Productivity growth is the only sustainable source of real wage growth. Historically, around 82% of MFP gains have flowed through to wages. With MFP now flat or negative, that engine has stalled. Workers can still secure nominal wage increases, but without productivity growth to back them, those gains either erode through inflation or squeeze margins - neither of which is sustainable. The productivity slump is not an abstraction. Real hourly compensation lost significant ground during the inflation surge and has been slow to recover - without productivity growth, further gains will be hard to sustain.
Inflation dynamics have changed
Our modelling suggests the post-COVID Phillips curve is steeper (for both prices and wages) than in the decade prior. Sectoral capacity constraints matter more, and with little productivity buffer, wage gains pass through more directly to prices.
This makes sub-3% inflation difficult to achieve without generating more slack in the labour market. The combination of sticky services inflation, constrained supply capacity, and modest but persistent demand growth points to inflation lingering at the top of the RBA's 2-3% target band rather than settling comfortably in the middle.
Unemployment, currently around 4.3%, is likely to drift toward our estimate of NAIRU at approximately 4.6%. At-potential growth implies a gradual rise rather than a surge. But there is meaningful uncertainty around both the NAIRU estimate and the level of potential growth. If potential is closer to 1¾ than 2¼%, even 2% GDP growth could prove mildly inflationary - meaning either inflation stays sticky for longer, or unemployment must rise above NAIRU to bring inflation sustainably back to target.
Monetary policy: back to Taylor
The post-GFC decade saw central banks, including the RBA, lean heavily on unconventional policy and forward guidance (and models like Holston-Laubach-Williams). The "lower for longer" mindset, followed by "higher for longer" during the post-COVID inflation surge, reflected an approach that emphasised signalling and patience over mechanical rules.
We expect this to shift. With inflation sticky, the output gap near zero, and limited productivity tailwinds, the RBA is likely to return to a more orthodox, Taylor-rule-guided approach to rate setting. Policy is currently mildly expansionary relative to a Taylor-style benchmark. In that framework, the bias is toward tightening, not cutting - unless unemployment rises faster than expected.
This represents a meaningful change in the reaction function. Markets pricing in aggressive rate cuts may be disappointed. The bar for easing is higher when inflation is sticky and supply constraints are binding.
Central case: convergence to potential, but no relief on inflation or rates
We expect the economy to converge gradually toward its (lower) potential growth rate, with unemployment drifting up toward NAIRU over the course of 2026. This is not a recession scenario - it is the economy settling into a constrained equilibrium.
Growth of around 2% will feel underwhelming but is consistent with the supply side. The labour market will loosen, but only modestly. Unemployment rising from 4.3% toward 4.5-4.6% reflects at-potential dynamics, not a demand collapse.
The problem is inflation. With the Phillips curve steeper and productivity providing no buffer, inflation is likely to remain sticky at the top of the 2-3% target band. The RBA will face persistent pressure to demonstrate credibility, but without a sharp rise in unemployment, the final leg of disinflation will be slow.
In this environment, the best case for interest rates is that they do not change materially over the next six months. The worst case is that the RBA is forced to tighten by another 25-50 basis points to defend the inflation target. Rate cuts remain unlikely absent a sharper-than-expected deterioration in the labour market or a benign external shock.
A less helpful global backdrop
Australia has historically benefited from positive external shocks - particularly commodity booms driven by Chinese demand. That tailwind looks unlikely to recur.
China is caught in a bind. US tariffs constrain its export-led growth model. Domestic headwinds from property sector deleveraging and structural rebalancing away from investment limit the scope for stimulus. Beijing is unlikely to deliver the kind of infrastructure-driven spending surge that lifted commodity prices in 2009 or 2016. This removes one historical source of positive supply shocks for Australia.
At the same time, global disinflation cannot be relied upon to do the RBA's work. Geopolitical tensions - whether in Taiwan, Ukraine, or the broader US-China relationship - pose upside risks to commodity prices and supply chains. The terms of trade are more likely to be a headwind than a tailwind from here, and the global environment offers few easy wins for Australian policymakers.
Risk balance
The central case is an economy growing near potential, inflation sticky around 3%, and rates on hold or modestly higher. The risks are asymmetric.
On the benign side, a global disinflation shock - perhaps from faster resolution of supply chain issues or weaker-than-expected global demand - could deliver inflation progress without requiring domestic weakness.
On the adverse side, persistently poor productivity and participation would lower the speed limit further, keeping policy tight and leaving the economy vulnerable to stagflationary dynamics. In that world, the RBA faces an uncomfortable choice between tolerating above-target inflation or engineering a more significant rise in unemployment.
Australia's economic challenge has shifted. The binding constraint is no longer weak demand - it is the economy's limited capacity to grow without generating inflation. With productivity weak and the labour market tight, rate cuts offer little upside. Until supply improves, policy settings will remain constrained and the margin for error thin.
Usual caveats: I'm not your financial advisor, this isn't investment advice, and my model could be wrong. Do your own research before making any decisions.
No comments:
Post a Comment