≡ Menu

Some Control

Andy Kessler has a pretty good essay in today’s Wall Street Journal on the Reagan administration’s mid-1980s effort to protect U.S. producers of semiconductors against the perfidious Japanese who were alleged to be selling semiconductors to Americans at prices below cost. Here are two key paragraphs:

But not all jobs are equally desirable. It’s profits, not sales, that create wealth. We should invest along the productivity fabric. Jobs for jobs’ sake destroys wealth. Saving Detroit was a mistake. Should Nike shoes really be made in Oregon?


Tax cuts create, tariffs destroy. The “new Nafta” looks OK but is hardly free trade. The market will determine which jobs to keep and which to toss to the sea. Mercantilism has failed again and again, from British Corn Laws to Japanese chips. Show me the margin. That’s where jobs, and Hogs, will be created.


But my main purpose in writing this post isn’t to applaud Kessler’s case for unilateral free trade (although applaud it I do). My main purpose, instead, is to pick a nit with a piece of language used by Kessler. I pick this nit because this piece of language is ubiquitous and misleading.

Early on in his essay Kessler writes

In the mid-1980s, when Donald Trump was in casino mode and hosting prizefights, the Japanese were knocking out U.S. memory-chip makers. Hitachi instructed salesmen to “quote 10% below their price. If they requote, go 10% again. Don’t quit until you win.” The company later claimed it was never official policy, but the Japanese controlled 90% of the market for some crucial chips. Something had to be done, right? Well, no.

The offending piece of language here is the word “control.” As used here, it clearly – when you think about it for more than one moment – is inaccurate. If Hitachi actually did control 90 percent of the market for some crucial chips, it would not have had to cut its prices to continue to make 90 percent of industry sales.

Put differently, Hitachi then made 90 percent of industry sales because it cut its prices substantially – which means that consumers had viable options. If any of the relevant parties in this market back then had any ability remotely deserving the label “control,” those parties were memory-chip buyers, for it was these buyers – who controlled how they spent their own money – whose ability to not buy Hitachi’s chips obliged Hitachi either to cuts its prices or to forego sales.

It’s commonplace both in public and academic-economic discourse to describe the current market share of a firm as the portion of the market that is “controlled” by that firm. Even worse – if worse here is possible – is the equally commonplace description of a group of firms as “controlling” that portion of a market to which those firms collectively, although not as a unit, sell.

Language matters. “Mere rhetoric” or “mere semantics” aren’t mere; they’re major. A firm that convinces consumers to buy some of its outputs by charging attractively low prices or by otherwise offering attractive deals does not “control” that, or any, portion of the market. If there is control in these situations, it belongs to consumers. (Protectionism, therefore, is nothing but an effort to strip control from consumers and to give it to politically prominent producers.)