Criteria for assessing the federal superclusters/global innovation clusters (ISE/GICs) program is based on cluster characteristics as set out in the Organisation for Economic Development and Co-operation (OECD) review of national cluster policy and the cluster policy scientific literature, policy experts David Wolfe (photo at left) and Travis Southin (photo at right) say in their study on the federal clusters program.

To what extent were the ISI/GICs integrated into the broader mix of federal innovation policies and aligned with provincial and local initiatives?

An academic interviewee observed that “the Federal Government had said these are going to be whole of government ventures, I have seen none of that . . . I have seen no discussion about where superclusters fit in terms of competition policy, trade policy, procurement policy . . . IP policy . . . none of them are aligned.”

Interviews with ISI/GIC members mentioned that they signed MOUs with various federal and provincial funding agencies to refer promising firms, co-fund research projects and provide technical vetting of applications if a project drew on their own expertise.

However, the National Research Council (NRC) was explicitly excluded as a member of any of the ISI/GICs and was compelled to launch a separate R&D grant program through its labs to support each cluster consortium – described by an NRC interviewee as clearly a second-best solution.

Other interviewees were more skeptical of the multi-level governance dimension of the GICs.

For their part, provinces criticized the GICs for not being coordinated with regional programs.

With respect to scale, successful cluster policy harnesses the existing strengths of strong local civic leaders and business champions to support their cluster-building ambitions. “The GIC's emphasis on national, technology-based consortia risks diluting the scale of investment in particular regional clusters and leading sectors,” Wolfe and Southin said.

Regarding autonomy, successful cluster initiatives build on existing industry strengths in specific localities and the efforts of firm members to foster established local network facilitators, or CMOs that promote the growth of the cluster.

While the GICs empower industry-led organizations to select cluster projects, the diffuse technology focus and broad geographic dispersion of many consortia may limit their ability to deliver the full benefits associated with cluster development, according to Wolfe and Southin’s study.

Policy alignment can be highly effective for cluster strategies, they noted. Cluster initiatives can leverage existing government programs and budgetary envelopes to access needed resources, or co-opt existing federal, provincial, or even private sector initiatives, and align their objectives to meet the objectives of the cluster strategy.

The ISI/GICs have only partially aligned with local and federal programs, through an informal, ad-hoc, and bottom-up process, rather than through institutionalized support channels, they said.

Most notable is the fact that, despite the express intent of Innovation, Science and Economic Canada (ISED) officials to design a program that was clearly distinguished from the NRC's previous national cluster initiative, “both programs tried to remedy Canada's lagging innovation performance by providing direct research funding to enhance the R&D performance of firms.”

While the NRC program prioritized public sector-led research through its institutes, and the ISI/GICs focus on industry-led research consortia, neither initiative has altered the downward trajectory in Canada's R&D performance, Wolfe and Southin said.

Both initiatives exhibit similar shortcomings: including a lack of alignment across existing federal programs and institutions, as recommended by the OECD report: weak integration with existing provincial and local initiatives on the ground, as would have been embodied in the Toronto-Waterloo Innovation Corridor proposal [which didn’t make the final cut to be a supercluster]; and failure to incorporate a central role for CMOs to build civic capital, promote linkages among geographically co-located firms and support peer-learning among cluster firms.

The failure of the Toronto-Waterloo Innovation Corridor group to influence the direction of the ISI/GIC program is symptomatic, but they were far from the only such group in the country. “Had local CMOs played a greater role in developing these programs, their ultimate design might have looked quite different,” Wolfe and Southin said.

Canada’s two national cluster programs are a “microcosm” of the problems in designing policies to remedy lagging innovation performance

In the end, the only reasonable way to assess the ISI/GICs is in terms of what they actually are – five industry-led research consortia with broad areas of technology focus, each based in a different region of the country with an extensive network of member firms extending beyond their regional base, Wolfe and Southin said.

As of June 30, 2025, ISED's website for the GICs reports $3.07 billion+ co-invested with industry ($1.17 billion in federal program funds, $1.9 billion from industry and other partners), 627 approved projects, 3280+ project partners (of which 1678 are SMEs), and 10,370 members in cluster networks.

Regarding IP rights generated (including patent applications, copyrights, trademarks and trade secrets), ISED’s website reports over 600 formal IP rights pursued related to foreground IP.

Despite those statistics, nearly a decade after the launch of the ISI/GICs, they haven’t significantly improved Canada’s innovation performance, based on the November 2025 Council of Canadian Academies (CCA) expert panel’s report on the state of science, technology and innovation in Canada.

The CCA’s report found that Canada still lacks an effective approach to support the development and commercialization – across the continuum from research to deployment – of the most promising areas that could improve national competitiveness and provide greater overall economic and social benefits.

“Canada's experience with two national cluster programs is a microcosm of the problems encountered in designing policies to remedy our lagging innovation performance,” Wolfe and Southin said.

After two and a half decades of a failure to reverse our declining innovation performance, it may be time to examine the shortcomings of the broader policy system, they said.

The key argument in their study is that the federal public service lacks organizational mechanisms to capture lessons learned from previous policy experiments and incorporate them in subsequent policy development.

The government's experience with the NRC's cluster initiative offered potential lessons for policy makers in the design of the ISI/GIC program.

Despite its stated intent to differentiate the ISI/GIC program from the CI program, both were predicated on the provision of increased government funding for research, without regard to how those new initiatives aligned with the existing innovation policy mix or with provincial and local efforts, Wolfe and Southin said.

“This experience suggests the innovation policy making system lacks the institutional capacity to draw lessons from prior initiatives in a reflexive manner.”

To a large extent, this reflects a failure along each of the three dimensions of policy learning identified by public policy and innovation policy scholars – government learning, lesson drawing, and social learning, they said.

With respect to government learning, the policy branch charged with formulating the ISI/GIC program did not draw upon the substantial body of cluster expertise built up in Industry Canada over the 2000s, and had a limited appreciation of the cluster policy literature.

With respect to lesson drawing, the linkages to external policy networks through both ISED's own consultation process and the Advisory Council on Economic Growth were governed by an unwillingness to support existing geographic and sectoral areas of strength – a key criteria of effective cluster policy.

Instead, the program targeted partnering with new cross-sectoral networks of industry actors and leveraging private sector investment. “This was achieved at the expense of accessing the accumulated knowledge of provincial agencies and local CMOs.”

To the extent that broader political considerations influenced the design and implementation of the NRC cluster initiative and the ISI/GIC initiatives, “Canada's preoccupation with balancing competing regional interests appears to have been the dominant factor,” Wolfe and Southin said.

In the case of the NRC cluster initiative, a central role was played by the Atlantic caucus of the governing party pushing for the implementation of the Atlantic strategy, of which NRC clusters constituted a key component.

The subsequent expansion of the program ensured that cluster initiatives were strategically located at NRC Institutes in each region of the country. The same regional distribution of cluster consortia is evident in the ISI/GIC program.

Federal clusters initiative failed to incorporate widely adopted mechanisms for cluster governance and peer networking

With respect to social learning, the federal government's approach seems locked into a conventional research funding approach to innovation policy with roots in the 1960s and 1970s, according to Wolfe and Southin’s study.

In terms of social learning, this approach missed the paradigm shift to a more interactive approach associated with innovation systems and cluster theory.

Canada has now experimented with two national programs aimed at promoting clusters across the country. Both targeted firm development through enhanced research funding – the first through public research institutes and the second through industry-led research consortia.

Neither program paid sufficient attention to the multiple mechanisms documented in the policy literature to support cluster dynamics, Wolfe and Southin said.

A key shortcoming of both was the failure to incorporate widely practised mechanisms for cluster governance and peer networking as integral components of the program, they said.  “The lack of cluster expertise inside the strategic policy branch at ISED likely reinforced this imbalance.”

The study’s findings indicate a lack of policy continuity and limited policy learning over the course of two decades in the federal public service, Wolfe and Southin said.

This is partly explained by the changes in government that occurred between the two initiatives and the ensuing changes in organizational structures, they added.

This is reinforced by the constant rotation of associate deputy ministers and policy staff across federal departments and the lack of institutional mechanisms for retaining and transferring policy learning.

“Overall, it suggests that there is at best a limited institutional capacity to support collective learning and knowledge exchange within the federal public service.”

Wolfe and Southin said their research indicates that within the central agencies and key line departments, the public service lacks a centre of expertise with responsibility for business innovation policy.

Both ISED and the Department of Finance have expertise with respect to research funding, consistent with the research funding focus observed in the NRC and ISI/GIC initiatives, they said.

Elements of business innovation policy are distributed across a range of branches at ISED and the tax policy branch at Finance, “but there is no central unit responsible for business innovation,” they noted.

Mechanisms for incorporating knowledge gleaned from federal agencies with strong regional and local expertise in how technology adoption and business innovation work in the private sector, such as the regional development agencies and the National Research Council-Industrial Research Assistance Program, are equally absent, Wolfe and Southin said.

The emerging consensus suggests that the ability to learn from past policy initiatives is critical for improving policy design and implementation.

The complexity of innovation policy-making, involving interactions among different actors and governance structures, requires policymakers to learn from previous policy experiments, from challenges encountered in implementing those policies, and to design policies that reflect lessons learned. 

“Given the overwhelming centralization of decision-making authority in the federal public service over the past three decades, this suggests that our case study of policy learning with respect to cluster innovation policy may be symptomatic of a larger institutional problem with the way Canada's version of the Westminster model of public administration has evolved.”

The Social Sciences and Humanities Research Council of Canada supported this research.