On the CCI Recommendations for Procurement Reform

Thoughts on "Building Winners: Strategic Procurement in the Age of Innovation", the latest report from Canadian Council of Innovators

Just this morning, Canadian Council of Innovators (CCI) released "Building Winners: Strategic Procurement in the Age of Innovation,” a policy guidance report that provides a thorough overview of the value of procurement as a flexible policy tool that is used in most, if not all of the world’s innovation-heavy economies to stimulate industrial investment in innovative technologies. In comparison, Canada’s latest budget actually reduces the procurement targets for innovation in a time where our economy is already badly lagging.

CCI’s core contention in their report is that supply-side approaches to innovation incentivization (for example tax credits like SR&ED) have proven to be less effective in recent decades than approaches that target individual firms in the early stages of technology development. In ignoring this, Canada neglects to use a key policy lever to support and directly incentivize innovation: government spending through procurement to drive demand for domestic innovation. The report makes an excellent argument for this view, and is a good summary of both the structural issues at play and what could be done to improve the situation.

This isn’t so much a summary of the CCI article as it is an editorialization. I encourage anyone interested to read both.

Building Winners

The key point is made very early in the report:

“Global evidence is clear that measures like economic complexity and labour productivity are the most reliable indicators that a country is on a path to a growing economy, because these metrics quantify how innovative an economy is. […] Countries with successful innovation economies have never had hands-off policies on innovation […].”

While innovative economies around the world use demand-side procurement approaches to incentivize innovation directly, Canada uses supply-side approaches such as tax credits and development cost reimbursement that are insufficiently hands-on to move the needle.

The rapid and accelerating pace of global change dictates that effectively supporting technology innovation requires, unsurprisingly, policy innovation. The CCI report gives 4 primary reasons for this.

Pace of Change

With less time available to develop an idea, the outcome of investment in those ideas becomes increasingly uncertain, prompting firms to stick with the status quo rather than take on risk.

It has been asserted by the government and many commentators that Canadian firms lack an appetite for risk and that this is a cultural issue. This is nonsense. As with any economic issue, all that matters is the related incentives. A near-complete lack of policy incentive for risk-taking in Canada, contrasted with places like the US where the government commits (via SBIR) to purchasing innovative products before they even exist, is directly responsible. The promise of demand during the development process through a commitment to purchase by the Canadian government fundamentally changes the calculus of risk-taking and directly incentivizes specific innovations.

Key takeaway: Procurement can change risk-taking behaviors by sharing the risk between public and private actors.

Scope of Change

Novel technologies like AI and quantum computing can fundamentally change how we interact with the world almost overnight, and the full utility of such technologies is often discovered in real time, well in advance of any related policy guidelines.

Large Language Models are the most extreme example of this, with OpenAI reaching 100 million global users in 2 months. Just as the actual impact and long-term use case for this technology has yet to establish itself, so to have related policy guidelines been unable to keep up. Canada had an opportunity to lead the world on the basis of the Montreal Declaration on the Responsible use of AI but has since failed to keep up. Procurement processes applied to novel technologies in the development stage give policy makers a direct first look. In a political climate driven by concerns over technological security, this approach has the benefit of allowing for a proactive and informed technological security response.

Key takeaway: Procurement can be the basis for proactive technology policy development and enhance technology security.

Winner Takes Most

Disruptive technologies are quickly consolidated into existing big tech firms, almost none of which are Canadian controlled. Something like 75% of Canadian AI IP has already exited the country to the south, and the floodgates remain wide open. As Jim Balsillie has repeatedly stated, the intangible economy is winner-take-most, and requires active policy guidance to ensure retention and long-term benefit from IP generated by our world-class research engine.

One of the best ways to retain IP is to provide a direct incentive to develop it here, using procurement approaches to purchase products from innovative Canadian companies (and thereby fund their domestic development of the associated IP) while they are still in the development phase.

Key takeaway: Procurement can incentivize domestic development and retention of innovative IP, leading both to economic benefits and (again) enhanced technological security.

Communication

In a small, open economy like Canada, we can’t just transplant SBIR-style policy because we lack the scale for it to work. The approach must be even more hands on, with different agencies that support different parts of the innovation pipeline actively communicating both with each other and with the entities they support to ensure effective matching of technology to government solution needs.

In other words, innovation requires deep connections and effective communication between all stakeholders at every stage of the innovation pipeline, and the required level of communication increases as the scale of the economy in question decreases. This network includes researchers, policy makers, lobby groups, funding agencies, private sector actors, and even the general public, and spans everything from basic ideation all the way through tech transfer, startup development, and eventual scaleup.

Building procurement mechanisms that can operate at all stages, from pre-incorporation ideation through product development in established firms, necessitates building bridges for communication between otherwise siloed funding mandates.

Key takeaway: Procurement frameworks can be the basis for and reason behind bridges that connect disparate parts of the innovation pipeline.

Buying Ideas

I have repeatedly said that it is not a good idea to import policy that works elsewhere without careful consideration of the context in which it works. The CCI report makes a similar point clear in the start of their recommended approach, stating

“We are a small economy reliant on international trade. It is the age of “moonshots” by the major economic powers, with industrial policy bets in the billions of dollars required just to ante up at the industrial policy table. Canada can’t easily match those subsidies, and our free trade agreements mean that we can’t keep the competitors out.”

One of the only policy levers that is used in nearly every successfully innovative economy, regardless of scale, is procurement and active involvement by the government in “buying domestic ideas”. By this, I mean establishing a contract to purchase and paying up front for the results, providing the input capital needed to complete the product, and guaranteeing at least a first customer.

This is inherently a failure-prone process: product development is hard, and rarely right the first time, but provided that both parties to the agreement adopt the mindset that the value of a success outweighs the cost of a failure and that there is value in the associated learning regardless of outcome, this is not a problem. The widespread use of this approach in successful innovation ecosystems globally makes clear that executed well, the calculus works, and it works because it represents a sharing of known risk between the public and private sectors.

As the CCI report states: “The traditional argument against engaging in industry policies, despite the acknowledged rationales, was ‘government failure.’ Government, it was argued, lacked the knowledge to capture minor positive externalities through targeted support—‘governments can’t pick winners’.”

It is perfectly true that the government cannot pick winners, and it should not try. In fact, it does not need to in order for procurement to be effective.

Deep tech investors in the private sector will tell you that investment in early technologies is a guessing game, to the point that investment decisions are often based more on the people involved in the attempt than on the technology itself. Even with niche-specific expertise, which the government lacks, winners are all but impossible to pick even for domain experts, and return on investment comes from playing the power law that dictates that even a single breakout success more than pays for the guaranteed majority of failures.

Well designed procurement frameworks that embrace the certainty of failures as an opportunity to learn have the potential to be the cornerstone of what CCI refers to as “firm-based industrial policy” that provides focused resources precisely when and where they are needed in order to strategically stimulate industrial risk-taking. Examples abound in the CCI report of technological innovations as fundamental as the computer chip that were catalyzed by similar mechanisms of shared risk-taking.

Specific Recommendations

CCI makes 4 key takeaways on how to improve the situation in Canada, drawing on examples that have worked elsewhere while carefully considering the differences in context. The report stops short of making specific policy recommendations, so I have taken the liberty of adding some, building on the foundation in the report.

Connected Systems

Where innovation ecosystems have been built successfully elsewhere, it is because different parts of the innovation pipeline communicate with one another. In the US, for example, the Bayh-Dole Act ensures that technologies coming out of universities are on the radar of early industrial support systems as soon as the patent is filed, if not sooner.

This is especially true at the stage of academic research and resulting technology development. In the status quo, academic researchers are so isolated from downstream industrial development and are so busy publishing papers that in many cases it does not even occur that research might have commercial value. Active, focused matchmaking between technologies in the research pipeline and the frameworks that can supports its subsequent domestic development is a necessity, not just for IP development, but also for its domestic IP retention. This is especially true in Canada, where our small, open economy guarantees that a laissez-faire approach to technology development will almost always result in early export of the associated IP.

In other words, for Canadian procurement to be effective, there needs to be communication between research institutions, tech transfer offices, firms that engage in early industrial development, and the government agencies that can provide demand for the results, communication that should start before the patent is even filed.

Playing to Win

The CCI report states that “In a strategic setting, firms forgo short-term profits for market share. States do as well, forgoing short-term consumer welfare considerations for the longer-term aim of developing the economy.”

Translated for those of us unencumbered by a need to be quite so diplomatic: the Canadian government does not take enough risk, and in so doing has traded avoidance of short-term loss for long-term economic stagnation. This problem is magnified in periods of rapid technological change, when new industries are established and the first to commercially exploit a technology has a strategic advantage over all others (for example: right now, and for the rest of the foreseeable future).

Addressing this problem requires an embrace of the power law: the long-term benefit of a small number of successes more than pays for the losses associated with a majority of short-term failures. This principle works even on small scales: VC funds successfully make this bet all the time, usually on less than 40 individual firms. An economic entity the size of Canada committing just 2% of its GDP to procurement can afford to make small, early stage bets on thousands of firms, all but guaranteeing a full sampling of the power law distribution, with returns paid out not in private profits but in the form of long-term economic benefit of having provided the seed to establish the next generation of high-growth domestic companies.

Be Opportunistic

Here, CCI makes the argument that a portion of what is spent procuring the goods and services on which the government relies can and should be mandated to be spent on novel versions. The argument is straightforward: new technologies arise unpredictably and the opportunity to exploit them is fleeting. Actively seeking them out increases the chances that opportunities will not be missed, and can lead to unexpectedly positive outcomes in the form of long-term economic development based on those novel technologies.

To make this feasible, there is significant training required. Governments are by nature slow to change course, and people are by nature slow to adopt novelty in the face of established solutions irrespective of improvement. Any realistic approach at enabling this requires more than just a mandate to do it: it needs dedicated training on how to implement and test novel solutions, as well as a shift in thinking around the impact of the inevitable duds.

Failure is only that which teaches us nothing.

Stop Spreading Peanut Butter

One of the problems that Canada faces in changing its approach to innovation is inertia of existing programs. Historically, when gaps are identified, it is easier politically and practically to put together a new funding body or agency to fill the gap than it is to rethink an old one that may no longer be serving its purpose, resulting in a plethora of increasingly narrow policy structures all competing for the same funding, and all of which are evidently collectively failing to deliver results in innovation.

A capstone agency that bridges the three siloed tri-council agencies to manage Canadian research funding is probably better than the status quo; better yet (but much harder) would be to rebuild the whole system so as to avoid the communication problem the capstone seeks to solve. The Canadian Innovation Corporation would probably have been an improvement over NRC IRAP in its current form; better yet (but much harder) would be to rebuild IRAP in a way that enables a degree of risk tolerance commensurate with the stage of development it seeks to support.

The same lack of political will for short-term risk-taking favors yet another gap-filing approach over rebuilding existing policy mandates and prevents these agencies from implementing anything that they might otherwise learn from their failures. Until that changes, we will continue to miss opportunities.

Conclusions

The CCI report does an excellent job framing the problem, drawing on examples of successful and unsuccessful policy initiatives implemented globally. While it does not make many specific recommendations on how it should be done, the case for procurement as a key part of the strategy to tackle the Canadian productivity problem is solid.

Every scaleup was once a startup, and every startup was once an idea. Without support that spans all stages of the process, innovations are lost. Procurement in its various forms is a tool that can span all stages of innovation, across all sectors, and can be focused on strategic economic development goals through supply-side incentives to address them. However, to work, this requires a hands-on approach built on a foundation of communication between different parts of the innovation pipeline, and a willingness to both take risks and to turn the inevitable associated failures into opportunities to improve associated policy mandates.