top of page

How Innovation Ecosystems Lift Productivity: The Firm, the Place and the Nation

5 hours ago
8 min read

John H Howard, 17 November 2026

People who follow the Acton Institute’s research on innovation ecosystems often ask a simple question. How do innovation ecosystems lift productivity?

The question usually refers to the national economy. Commentators now debate Australia’s weak productivity record, sometimes suggesting that the country needs to get better at “making things”. The Institute’s research starts from places and how they are governed, with productivity as the principal economic outcome.

The short answer is that productivity rises inside firms and service organisations when managers adopt, apply and use new knowledge and technology. An innovation ecosystem builds that capacity by making the necessary complements cheaper, quicker, and less risky to assemble than any one organisation could manage alone.

The longer answer has three levels: the firm, the place and the nation. The recently published book Hidden in Plain Sight gives the general account (Howard, 2026a). Its second volume tests that account against twenty Australian districts and regions (Howard, 2026b). The book Turning AI into Productivity applies it to one technology (Green & Howard, 2026).

The firm

A technology produces little by itself. A firm gains when managers combine it with skills, data, redesigned processes, and the ability to take a pilot into production. This is the Complementarity Principle, which Making Sense of AI in 2026 developed at the firm level (Howard, 2026c).

The complement that takes longest to put in place limits what a firm gains. For example, two firms using the same AI model can obtain very different results, because they may differ in everything around the model.

The European field visits for Turning AI into Productivity showed where firms first gain: in specific tasks, in firms with absorptive capacity, and in firms connected to institutions that share the risks of adoption (Green & Howard, 2026).

Engineers in Dortmund were using AI to improve assembly quality. Engineers in Munich used it to cut the time it took to program industrial robots. In both cases, the tools enhanced skilled work inside established engineering workflows.

Management capability was the most common missing complement. Firms ran successful pilots and then did not proceed to production.

The place

Many of the complements a firm needs are collective goods. A skilled labour market, shared research facilities, data infrastructure, translation institutes and business networks are more than a single organisation can provide, particularly a small or medium-sized firm.

Firms and institutions bring those complements together in places. Hidden in Plain Sight identifies three channels through which a place creates value (Howard, 2026a). Table 1 provides a summary.

Table 1: Three channels of value and their effect on productivity

Channel

What happens in the place

Effect on productivity

Complementarity

Firms, universities, research institutes and intermediaries gain access to capabilities they could not build alone

More output from the same inputs

Conferral

Founders, anchor firms, political leaders and a sustained civic voice give the place standing and credibility

Talent, capital and institutions commit earlier than measurable fundamentals alone would secure

Resilience

A place keeps its skills, institutions, infrastructure and relationships through a shock

Productive capacity remains available for the next cycle of investment

Source: Adapted from Howard (2026a), Chapters 3 to 6.

Eindhoven illustrates the third channel. The complements assembled around Philips in systems engineering, supplier depth and technical education remained in place after the company contracted in the 1990s, which helps to explain the growth that followed.

Firms and institutions rarely assemble the complements unprompted. Someone with a mandate has to identify the missing complement and take responsibility for supplying it. For that reason, both volumes of Hidden in Plain Sight give close attention to governance.

Volume II of Hidden in Plain Sight applies the framework to twenty Australian districts and regions (Howard, 2026b). Across those cases, the complement districts most often still lack is custody: one organisation responsible for the district as a whole, with the acceptance of landowners, funders, and planning authorities.

In the twenty cases, capability tends to accumulate where the four domains are aligned, and a custodian has a lasting interest in the district. It tends to remain underused where either condition is absent, which is the productivity problem at the scale of a place.

Productivity is the principal economic outcome of a well-functioning ecosystem. Community, public health, socio-cultural and public policy outcomes accompany it. Across the cases examined for the book, those wider outcomes usually accompanied the productivity outcome and often came before it.

The nation

The connection between a place and the national economy is harder to establish. Three observations from current official data help.

Sector shares

According to the latest Reserve Bank of Australia Snapshot of the Composition of the Australian Economy (November 3, 2026), Manufacturing accounts for 5.4 percent of Australian output. Health and education account for 14.7 percent, Finance for 7.8 percent, Construction for 7.7 percent, and Mining for 9.2 percent (Reserve Bank of Australia, 2026).

The Australian Industry Group reports that Manufacturing accounted for about 917,000 jobs in the June quarter of 2026, or around five percent of all jobs in Australia. The Group added that Manufacturing is the fastest adopter of new technologies, promising productivity gains amid supply-side pressures (Australian Industry Group, 2026). 

On a rough calculation, a 10 percent productivity gain across the Manufacturing sector, applied to its 5.4 percent share of output, would raise the economy-wide level by a little over half of one per cent. That gain would be worth having, but it is unlikely to be enough on its own. The Treasurer and the Productivity Commission both look across the whole economy for the larger gains.

The findings in Turning AI into Productivity are consistent with the Treasury expectation that AI will support long-run productivity growth where firms invest in the processes, management practices and skills that put it to use (Treasury, 2026). The Commission’s recommendations cover business investment, regulation, skills, data and digital technology, energy and the care economy (Productivity Commission, 2025a).

Where the statistics record the weakness

Labour productivity across the economy was flat in the June quarter of 2026 and 0.2 per cent lower over the year, as hours worked grew slightly faster than output. Market sector productivity rose by 0.2 per cent in the quarter but was still 0.1 per cent lower than a year earlier (Productivity Commission, 2026).

The weakness is more persistent in the non-market sector, which is largely health, education and public administration. Productivity there fell by 0.1 per cent in the quarter and 0.3 per cent over the year, which leaves it below the level of March 2007. Across the whole economy, productivity is approximately one percent above its 2015 to 2019 average (Productivity Commission, 2026).

In the twenty Australian cases assessed in Volume II of Hidden in Plain Sight, health and biomedical activity around teaching hospitals and universities is prominent, with Westmead, Monash Clayton and Brisbane among the examples (Howard, 2026b). The national productivity problem and the ecosystem agenda therefore overlap in services more than a factory narrative would suggest.

The Productivity Commission has said that measurement difficulties mean that analysts should not judge the non-market sector on its productivity record alone, while maintaining that the sector still requires reform (Productivity Commission, 2025b). A measurement problem and a performance problem may both be present.

The capacity to adopt

Treasury has observed that growth at the productivity frontier has slowed, which it regards as consistent with slower technology diffusion (Treasury, 2025). Accounts of diffusion argue that firms lack the “absorptive capacity” to recognise, assimilate, and apply new knowledge.

Many firms, especially smaller ones, face substantial barriers to building absorptive capacity, including limited management capability, shortages of technical skills and poor access to data, research facilities and translation services.

The ecosystem account explains how firms overcome these barriers by recognising that many complements are collective goods available in ecosystem places. In the absence of other constraints, firms may be able to build capacity in these places.

Hidden in Plain Sight notes that an early application of Industrial AI in Australia is to improve productivity in the industries the country already has: manufacturing, bio-manufacturing, mining, ports, agriculture, energy, and health (Howard, 2026a). These are also sectors where returns depend heavily on complements.

What the answer does not claim

The firm and place answers to the question “how do innovation ecosystems lift productivity” draw on a case base of 110 ecosystems (now 140) in 35 countries, together with field visits. The national answer is an inference from the mechanism and from sector shares.

Both books acknowledge the limit. The field observations in Turning AI into Productivity do not resolve the question of aggregate productivity (Green & Howard, 2026). Hidden in Plain Sight records that the evidence remains contested on whether place-based programs generate net national gains or move activity between regions (Howard, 2026a).

The cases explain how a lift occurs without measuring its size.

Volume II assesses capability, interaction and governance in the twenty Australian cases. It does not measure productivity outcomes (Howard, 2026b). But the twenty cases offer a starting point for that task.

The custodian of a district could record a small set of outcome indicators against a baseline. Over several years, the record would begin to show how much difference a functioning ecosystem makes.

What decision makers can do

Several actions follow for people with roles in government, universities, firms and precincts.

  • Diagnose before funding. Identify the complement that limits the use of capability already present in a place.

  • Give one organisation custody of the district, with the mandate and the tenure to supply the missing complement. Two tests apply. Does its own standing depend on the coherence of the district? Do the landowners, funders and planning authorities accept its coordination?

  • Fund translation and intermediation as continuing functions and assess them on results taken into use.

  • Build management capability into the design of adoption programs.

  • Record outcomes in health, software and other services alongside manufacturing, over periods long enough for complements to accumulate.

  • Agree a small set of outcome indicators for each district at the start, so that later evaluation has a baseline.

  • Align place-based ecosystems with national priorities through a standing committee of Cabinet or National Cabinet.

Australia’s productivity challenge may be, in large part, a challenge of building the capacity to adopt across a services economy. Innovation ecosystems are where firms and service organisations build that capacity, in hospitals, universities and software firms as well as in factories.

The Acton Institute for Innovation books referred to in this Insight are available through Amazon Publishing

Hidden in plain sight: How places create value from innovation: Paperback and Kindle

Making sense of AI in 2026: A framework for policy and practice: Paperback and Kindle

Turning AI into productivity: The role of innovation and industrial ecosystems: Paperback and Kindle


References

Australian Industry Group. (2026). Manufacturing in Australia: Performance and outlook 2026. https://www.australianindustrygroup.com.au/resourcecentre/research-economics/manufacturing-in-australia-2026/

Green, R., & Howard, J. H. (2026). Turning AI into productivity: The role of innovation and industrial ecosystems. University of Technology Sydney.

Howard, J. H. (2026a). Hidden in plain sight: How places create value from innovation. Acton Institute Publishing.

Howard, J. H. (2026b). Hidden in plain sight, Volume II: Twenty Australian innovation districts and the governance question. Acton Institute Publishing.

Howard, J. H. (2026c). Making sense of AI in 2026: A framework for policy and practice. Acton Institute Publishing.

Productivity Commission. (2025a). Meeting the productivity challenge. https://www.pc.gov.au/inquiries-and-research/five-productivity-inquiries/meeting-productivity-challenge

Productivity Commission. (2025b). Quarterly productivity bulletin: December 2025. https://www.pc.gov.au/ongoing/productivity-insights/bulletins/quarterly-bulletin-december-2025

Productivity Commission. (2026). Productivity update: September 2026. https://www.pc.gov.au/ongoing/productivity-insights/update-september-2026

Reserve Bank of Australia. (2026). Composition of the Australian economy: Snapshot. https://www.rba.gov.au/education/resources/snapshots/economy-composition-snapshot

Treasury. (2025). Economic Reform Roundtable: Productivity overview. Australian Government. https://treasury.gov.au/sites/default/files/2025-08/productivity-overview.docx

Treasury. (2026). Intergenerational report 2026: Australia’s future to 2066. Australian Government. https://treasury.gov.au/sites/default/files/2026-09/p2026-797917.pdf


The Acton Institute for Innovation is an independent policy research institute based in Sydney. Innovation Insights presents the Institute’s own analysis. www.actoninstitute.au

Comments


bottom of page