-A A +A
November 29, 2021

From: Paul Johnson

To: Competition Policy Watchers

Date: November 29, 2021

Re: A Competition Conundrum: Winner-Take-All Markets

The populist backlash against “bigness” generally and “big tech” in particular has prompted two strands of debate about competition law enforcement.

The first is broad and concerns foundational questions about what competition policy should strive to do; the second is narrower and concerns whether competition law enforcement is achieving its current objective, which is to protect and promote competition.

This note uniquely focuses on the second debate and considers how enforcement functions in winner-take-all markets. It questions whether competition laws are doing enough to constrain attempts by incumbents to eliminate nascent competition – cases where potential competition is prevented.

In winner-take-all competition, a single competitor will serve (most of) a market. Winner-take-all competition can occur in some digital markets due to very large economies of scale and scope or network effects. These benefits can only be achieved when firms get big, leading naturally to a very small number of very large firms. In those cases, the displacement of an incumbent by an entrant represents a paradigm shift, which can bring about enormous economic benefits such as new products, reduced prices, and improved quality.

At the same time, given the benefits incumbents enjoy from large economies of scale and scope or network effects, successful entry is difficult and, hence, unlikely.

Very difficult but very consequential entry contrasts with the usual situation (competition in a market) where an entrant may be likely to have some measure of success but the scale of the effects of success is much more modest.

To see the relevance of competition for the market to “prevent” cases, suppose that an incumbent with market power is engaged in an anticompetitive act that prevents successful entry by a would-be competitor. (Alternatively, the incumbent could be trying to acquire the competitor.) To stop that conduct, competition law in Canada and many other jurisdictions requires that, on a balance of probabilities, the entrant would have a substantial effect on competition but for the conduct. That legal requirement means that, absent the incumbent’s conduct, the competitor must be deemed “more likely than not” to succeed. In other words, if the incumbent’s conduct were stopped, the entrant must succeed with greater than a 50-percent chance.

The problem arises when successful entry is deemed unlikely, but the benefit of successful entry is very large. As described above, this is a key characteristic of winner-take-all markets. In short, the legal test and standard of proof on which anticompetitive effects are proven and met may be overly onerous when markets are characterized by a winner-take-all dynamic.

To illustrate the problem, suppose the entrant’s success is plausible, but not likely – say the entrant will have a one-in-10, or 10-percent, chance of success. Denote the economic benefits of successful entry by $X so that the mathematical “expected value” of entry is 10 percent of $X, which also represents the expected value of the economic harm of the anticompetitive act. But even as $X becomes arbitrarily large, the Competition Act does not condemn the incumbent’s actions because anticompetitive effects remain unlikely. This is a very curious – and I would argue perverse – result: $X could be a billion dollars so that, in expectation, the conduct causes $100 million of damage. Yet such conduct would never run afoul of the Competition Act.

Nevertheless, in Canada (and elsewhere), these calculations are not relevant in the eyes of the law. The Supreme Court in its Tervita decision made clear that an entrant’s success must be gauged by the balance of probabilities legal standard while simultaneously recognizing the difficulties of mustering evidence about unproven competitors:

“The evidence must be sufficient to meet the ‘likely’ test on a balance of probabilities, keeping in mind that the further into the future the Tribunal looks the more difficult it will be to meet this test …. Business can be unpredictable and business decisions are not always based on objective facts and dispassionate logic; market conditions may change…”

A number of potential solutions to this problem have been proposed and are discussed in a longer article I have written with John Pecman and Justine Reisler. Suffice it to say that there does not appear to be a consensus about a solution. But none of those solutions require regulating large parts of the economy like a public utility. That is fortunate because proposals to do so gloss over the complexities of micromanaging rapidly evolving industries.

The lack of consensus about a solution reflects the inherent uncertainty about new businesses and business models, which will prevent any solution from being perfect. We should not let the perfect be the enemy of the good if there is an opportunity to consider amendments to Canada’s Competition Act.

Paul Johnson is a member of the C.D. Howe Competition Policy Council and owner of Rideau Economics. From 2016-2019, he served as the T.D. MacDonald Chair in Industrial Economics at the Competition Bureau. 

To send a comment or leave feedback, email us at blog@cdhowe.org.

The views expressed here are those of the author. The C.D. Howe Institute does not take corporate positions on policy matters.