From: Elisa Kearney and Paul Johnson
To: Canadian competition law observers
Date: December 14, 2023
Re: Unpacking some of the Limitations of the Competition Bureau’s Report on Competitive Intensity in Canada
The Competition Bureau’s new report on competition may have drawn flattering headlines, but what makes a good newspaper headline does not make sound policy. We urge caution before drawing conclusions from the report, and encourage further detailed study by independent economists and other academics of the Statistics Canada data on which the report is based.
Further study would be valuable because the relationship between concentration and competition highlighted by the report is frequently nuanced and subtle.
The Competition Bureau itself offered caveats along these lines, observing that the data’s “structure and source do impose some limitations on how [the Bureau] can use it” to make conclusions about the state of competition in Canada and how it has changed over time and that the findings are “not intended to motivate law enforcement activities under the Competition Act.” By extension, caution is warranted in interpreting the results and using the results to inform policy.
One important reason for this caution is that the report’s results are ambiguous on some important dimensions. The report does not show that competition has been “shrinking” because Canadian industries have become more concentrated. In fact, the report actually shows that more Canadian industries have seen decreases in concentration (using the Herfindahl-Hirschman Index, or HHI) than have seen increases in concentration (88 versus 59) and the number of industries where concentration has seen “no meaningful change” is even larger (164).
Instead of focusing on the overall decrease in concentration in Canadian industries, the report focuses on the most concentrated Canadian industries, which it defines to be industries with HHI greater than 2500, which is the US threshold for declaring an “antitrust market” highly concentrated while assessing mergers. But “industries are not markets” so the choice to apply this threshold to a study of four-digit NAICS industries may not be appropriate.
There are either 26 or 27 such “highly concentrated” industries in Canada, says the report, accounting for 8 or 9 percent of the industries it studied. Among those industries, nine saw increases in concentration, two saw decreases in concentration, and 13 – half of the “highly concentrated” industries – saw no meaningful change. It is dangerous to draw firm conclusions about the state of Canadian competition based on these results. Are these industries particularly important for Canadians? Does the even smaller number of such industries that saw increases in concentration (nine out of 314 industries studied) form a basis on which to construct good public policy?
Unfortunately, we cannot answer those questions because the report contains no information on which Canadian industries it found to be “highly concentrated.” This is a critical omission, and suggests that any conclusion drawn from the report’s findings be attached with strong caveats for at least three reasons:
- First, the data on which the report relies “does not account for products and services bought by Canadians from foreign firms (imports).” Consider the implications if the nine highly concentrated industries with increasing concentration included industries like “semiconductor and other electronic component manufacturing” (NAICS 3344). Because the vast majority of semiconductors are manufactured abroad, an increase in the concentration of Canadian semiconductor manufacturing reveals little, if anything, about competition in Canada. The same conclusion should be drawn regarding any industry where imports are important, like motor vehicle manufacturing (NAICS 3361).
- Second, the report’s analysis does not appear to consider the implications of regulation and almost certainly includes industries that are concentrated as a result of public restraints on competition. In that light, suppose that “general medical and surgical hospitals” (NAICS 6221) is among the nine highly concentrated industries with increasing concentration. Given the very significant governmental role in healthcare in Canada, such increasing concentration does not represent a competition problem that reform to the Competition Act should or could address. Similar conclusions should be drawn regarding any other heavily regulated industries (e.g., electric power generation, transmission and distribution (NAICS 2211)). It would also be important to understand whether, among the nine industries with increasing concentration, there are industries where the government limits competition through foreign ownership restrictions and their effect.
- Third, the report identifies an important caveat that bears re-emphasis: “industries are not markets.” In other words, while the report measures statistics within an industry at a national level, competition may play out locally so that reported concentration does not reflect any valid measure of competition. Consider, for example, “taxi and limousine service” (NAICS 4852) where entry by ride-sharing platforms has upended the industry during the time period considered in the report. That disruption has been lauded for bringing more competition to local markets. That increased local competition has likely resulted in less concentration at the local level but more concentration at the national level as Canadians increasingly ride with a small number of ride-sharing platforms like Uber and Lyft.
It may well be that neither imports nor regulation are important to the “highly concentrated” industries on which the report focuses. Similarly, the report’s results may not turn on discrepancies between local markets and national-level measures of concentration. But we simply cannot know without information beyond that which is contained in the report.
In addition to calling on economists and other academics to examine the Statistics Canada data and augment the findings of the report, we call on the Competition Bureau and, where possible, others to do more regular assessments of the effectiveness of its competition interventions for the economy, including ex post evaluations of specific enforcement actions.
Measuring the effects of competition interventions – or the effects of the absence of competition interventions – is a complement to market studies and an under-utilized tool in the Bureau’s toolbox. More importantly, it is also the most direct way to assess whether Canada’s competition enforcement regime is “fit for task.”
The Bureau estimates that the dollar value of annual consumer savings from its actions to stop anti-competitive and deceptive marketing activity is $50 for every dollar spent. But that estimate appears to be largely driven by “default” assumptions about the benefits of the Bureau’s interventions; the methodology and calculations provided are not clear.
By way of contrast, Britain’s competition regulator estimates the price, quality or other negative effect avoided due to its intervention in an annual impact assessment report. That process is reviewed by an external expert. The US Federal Trade Commission has undertaken retrospectives for a range of consummated mergers over 35 years “to understand whether the agency’s threshold for bringing an enforcement action in a merger case has been too permissive, thus allowing too many potentially harmful mergers to go through,” and “to assess the performance of tools that agency economists use to predict the effects of proposed mergers prospectively.” As an independent enforcement agency such introspection is also an essential component of safeguarding accountability and enhancing public confidence.
Elisa Kearney is a partner in the Competition Law and Foreign Investment Review practice of Davies Ward Phillips & Vineberg LLP and the Chair of the C.D. Howe Institute Competition Policy Council and Paul Johnson is the owner of Rideau Economics and a former TD MacDonald Chair in Industrial Economics at the Competition Bureau.
To send a comment or leave feedback, email us at firstname.lastname@example.org.
The views expressed here are those of the authors. The C.D. Howe Institute doesn’t take corporate positions on policy matters.