Mike Munger recently wrote: It’s about a fundamental misunderstanding of economic competition held by many legislators and regulators. competition Competition is about having more firms, not fewer. This misunderstanding arises from Munger’s confusion of the textbook definition of “perfect competition” (which he calls a “silly concept”) with the actual competition that takes place in the world outside the classroom blackboard. Munger gives the following example:
Senator Elizabeth Warren recently argued that the Biden administration should block Capital One’s acquisition of Discover through the FTC. Her logic is that it’s “perfect competition” and that two small companies are better than one medium-sized company. But the folly of this approach can be seen in the massive industry where Visa, MasterCard and American Express control 98% of credit transaction volume. A merger between Capital One and Discover would create more competition in the industry, not less. The newly created company would have the financial muscle and scale to move the credit card industry away from outdated ways of doing things.
So that got me thinking about other cases where “more companies = more competition = better” doesn’t apply. Two big examples that come to my mind are banking and mobile operating systems.
First of all, The Great DepressionIn Canada, the banking system was largely unregulated. Out of that system came a relatively small number of very large banks operating across the country. In the United States, there was a highly regulated banking system that was heavily biased toward a “unit bank” system rather than a branch banking system (among other things), which meant that banks were limited in how far they could expand geographically (they were often prohibited from operating across state lines) and therefore limited in size. Out of that system came a system of tens of thousands of relatively small banks across the country.
From the perspective of “more companies = more competition = better,” it may seem like the United States, with its vast number of banks, was in a better position than Canada, which was “dominated” by a few huge banks. But in fact, the opposite was true. Because the banks were so numerous and small, they had very little diversification of their holdings and were largely tied to local economic conditions. Large, highly diversified banks are better able to absorb economic shocks than small, less diversified banks. This is one reason why over 10,000 banks failed in the highly regulated US banking system during the Great Depression, while none failed in the lightly regulated Canadian banking system.
The second example that comes to mind is mobile operating systems. It’s not uncommon to see some level of concern about the fact that mobile operating systems are currently a virtual duopoly between Android and iOS. Aren’t more mobile operating systems on the market better, since it would increase competition? No, not necessarily. You might wonder why this is the case, and as Munger himself says, the answer is transaction costs.
Let’s take a rather extreme example. Imagine if the genie snapped his fingers, tomorrow there would be 10,000 different mobile operating systems on the market. Wouldn’t that be good for competition? One of the nasty features of the real world that the introductory chalkboard model doesn’t take into account is transaction costs. If you want to create and sell an app that simulates coin tosses to improve the lives of indecisive people, you simply can’t program and format it for 10,000 different OSes. The transaction costs are too high. The same goes for all programmers and app developers. The more operating systems there are on the market, the more complex and expensive it becomes to deliver applications and programs to everyone in the market. Over the past 20 years, the variety of operating systems has increased. Symbian was one of them. BlackBerry had its own operating system, called (not very imaginatively) BlackBerryOS. WebOS was another. Windows Mobile was doing well. All of these went away, and Android and iOS were at each other’s throats. Would it be better if all the obsolete mobile operating systems still existed, with more and better competition? Maybe in a hypothetical world of perfect competition. But in the real world, where transaction costs exist, it’s not at all clear that this is the case. More operating systems means more transaction costs associated with producing everything that uses those operating systems. This is very likely to make the mobile OS market less productive, rather than more productive.
What is the “optimal” number of operating systems? I don’t know. Neither do you. The answer can’t be derived from a chalkboard, an armchair, or tea leaves. But our best chance of finding the answer is when the market is freed up enough to allow participants to engage in real-world market competition.
