Monday, December 11, 2023

Competition and Innovation: Incorporating a More Dynamic Perspective into Enforcement


This is the text of my 15-minute intervention last week at the OECD. Happy I wrote it down and read it because I was a bit overwhelmed, to be honest. Accompanying slides here

The best part, however, was the interaction we had with the OECD States/competition authorities (despite sitting very far from each other).


"Thank you immensely to the Chair and the Committee for the opportunity to be here with you today. I think that we all in this room might agree that the interplay between competition and innovation is increasingly vital in our economies, in which we all face numerous challenges in devising solutions to ever more complex problems.

These challenges are starkly highlighted by the ongoing issues surrounding climate change and the digital revolution. As we know, innovation can be the creative response that entrepreneurs make to changing conditions, to new challenges. To these entrepreneurs, “innovation is the ability to see change as an opportunity - not a threat.” And we as a society, as consumers are provided with choices on how to deal with these challenges, with these changing conditions. Furthermore, the critical role of innovation has been profoundly illustrated in our emergence from the health crisis brought about by COVID-19.

In competition policy circles like this one here, innovation isn't just a buzzword we throw around to sound smart while prices and output is what antitrust law is really about. There is a profound and increasing recognition that innovation is a  real factor in shaping competitive processes and often it should be taken even more seriously than price effects in competition policy enforcement.

We are very lucky here today to stand on the shoulders of this excellent background paper prepared by the OECD Secretariat. In my brief remarks today, I will focus primarily on the highly useful distinction between the approach that focuses on innovation incentives and the approach that instead focuses on the innovation impact or innovation effects.

The clear distinction drawn between a more static and a more dynamic perspective on innovation is particularly commendable and useful. It effectively highlights the progress made in terms of competition policy so far, as well as what remains to be done, both from a theoretical standpoint and in terms of competition policy enforcement.

Both approaches have their merits and limitations, and in a certain sense, they are complementary to each other. Understanding which approach to adopt depends on the crucial assessment by competition authorities of how competition manifests itself in the specific market, sector, or industry at stake.

Is the focus on current or potential competition in identified product markets or is the focus directly on innovation processes or spaces? This differentiation, as I see it, is the real focal point of the background paper, from which the brief considerations I will present in the following emerge.

Although this differentiation is absolutely crucial and we will return to it shortly, I believe the two approaches are driven by many insights that are common to both. I will mention three of these insights here.

The first is the belief that it is possible for competition enforcers to  identify situations, markets in which innovation is a significant dimension of rivalry between companies.

The second is that a longer-term or dynamic perspective is necessary to fully grasp innovation and its relationship with competition.

The third realization or insight is that a longer-term perspective must contend with uncertainty. Here we must decidedly forego the comfort of quantifications that offer the illusion of certainty in the application of antitrust law.

I've created a table, informed by the background paper and my previous research on this topic. It outlines key differences between the two approaches, in addition to the shared insights I've just mentioned.

The first approach focuses on the circumstance that rivalry in the product market is driven by product or process innovation. The second approach, on the other hand, brings directly to the fore the innovation efforts made by competing firms.Top of Form

This is clearly evident from the background paper, to which I would only like to add the idea that it is necessary to also consider the ecosystems where the relative orchestrators appropriate and manage the innovation efforts produced by the ecosystem participants.

The following point relates to the realization that the innovation concerns in the two approaches are distinct. On one side, there is a reduction or delay in introducing new features in a product market. On the other, there's a more direct concern regarding changes to innovation levels and intensity.

The third point is a focus on financial incentives to innovate by the first approach  and on capabilities to innovate in the second.

The fourth point is that, according to the first approach, future markets can be identified in a traditional manner. In contrast, based on the second approach, it is not possible to define markets that may emerge from innovation in such conventional terms.

The fifth aspect is that for the first approach, innovation is mostly of an incremental nature, while the second approach is perhaps more oriented towards innovation that is mostly of a disruptive nature.

The sixth aspect relates to different degrees of uncertainty. For the first approach, uncertainty is linked to the fact that products have not yet entered the market or potential outcomes from innovation are not clearly identifiable. For the second, the uncertainty is more radical as it stems from the circumstance that there is high uncertainty regarding future outcomes of innovation efforts.

The seventh and final aspect concerns theories of harm. For the first approach, profit-cannibalization narratives are particularly relevant, thus analogous to unilateral price effects. Relevant are also traditional foreclosure theories, which are then adapted to an innovation context. Instead, For the impact-based approach to innovation and competition, the theories of harm aim to highlight the direct impact on innovation of a behaviour or a transaction, not merely as a consequence of a reduction in competition.

Certainly, the distinction between these two approaches is valuable for analytical clarity and for promoting theoretical and other developments, but in enforcement practice, they are often also strategically interconnected, as noted in the background document.

So, we can distinguish cases where there is a more static approach to innovation and which aligns closely with the first approach. This involves determining product markets, even if they are in the future, even if the competition is only potential, and then also relying on conventional theories of harm based for instance on a profit cannibalization rationale.

And then there are other cases where the approach to competition is more dynamic, both in terms of determining a product market or an innovation space and in terms of theories of harm.

In the final part of my reflection, I concentrate on some challenges that I still see on the horizon from the perspective of further developing this more dynamic approach to innovation.

From seven, as in the previous table, the points I would like to briefly illustrate are now eight.

The first point is that it seems to me that it’s still necessary to better conceptualize the relevant rivalry between companies as a process of innovation and knowledge creation. From this perspective, it is clear that the traditional, orthodox industrial organization thinking and approaches may provide only limited assistance. Much more useful could be economic theories that focus on innovative, dynamic processes of variation and selection, such as evolutionary economics.

I am absolutely convinced that these alternative economic conceptualizations of dynamism and competition processes should receive greater attention from antitrust enforcers. The monopolization carried out by the industrial organization in the field of so-called competition economics has so far left little room for this and other theoretical contenders or even disruptors to emerge.

The second point is that there is still a significant need to develop a better understanding of innovation capabilities, especially in sectors where innovation processes are less structured. This might be the case because, for instance, the tools of appropriation in theses sectors or market phases might be different from traditional IP rights. A different understanding of innovation capabilities might be required based on the specific type of technology/knowledge production, the specific skills required (for instance, particularly valuable AI talent), and the specialized assets needed, such as data and specific technological tools required for innovation.

The third point is that if a better understanding of innovation capabilities is essential, there are additional innovation factors beyond capabilities that require attention. These include dynamic patterns, path dependencies, absorptive capacities, technological trajectories, and so on.

The fourth point is that greater efforts may be needed in certain contexts to assess the impact of the same innovation efforts on multiple sectors (e.g., digital) – cross-market.

A fifth, more practical challenge is that in some jurisdictions, it may be necessary to clarify in the law that "market definition" under certain circumstances is not necessary, or that the market can be determined in a radically different manner from traditionally employed methods.

A sixth, more challenging point from a theoretical and analytical perspective is that it seems necessary to better understand innovation processes within ecosystems. This, for instance, requires differentiation between the production of innovation and knowledge and the mechanisms of their appropriation. This pertains to contexts and ecosystems where both incremental and disruptive innovation hold significance.

A seventh point to consider is that it would be important to incorporate concerns about the "quality" of innovation into antitrust practice, not just focusing on innovation levels and intensity. From this perspective, close collaboration with data protection authorities and, in the near future, various types of AI regulators, including sector-specific ones, could be crucial.

Last but not least, there is the challenge related to the evidence required to ultimately capture the innovation dimensions mentioned in the background note. Some are already practiced, such as portfolios of R&D activities, commercial and licensing agreements with other companies. Other evidence, for example, to assess innovation capability, may include the analysis of links to universities, participation in scientific conferences, and so on.Top of Form

If the eight challenges, taken together, may seem a bit daunting, I would suggest that antitrust authorities join forces on many common fronts. A very practical and cost-effective suggestion is that one or more antitrust authorities, from different jurisdictions, could facilitate the organization of symposia, seminars, and workshops that bring together economists (including those from non-orthodox economic schools), management experts, behavioural scientists, scientists from different fields, and technologists. These gatherings would provide a platform for cross-disciplinary collaboration, offering diverse perspectives, approaches, and knowledge to enhance the understanding of innovation processes within specific industries and sectors. They could ideally also lead to the development of new assessment tools and methodologies immediately relevant to competition policy enforcement.

Many thanks for your attention."




And as you see here, fun we had!