Preempting Competition

by S.M. Oliva
Mises Economics Blog
Jun. 17, 2010

Last August the British conglomerate Rio Tinto sold off its Alcan Packaging division. Some of these assets were sold to Australia-based Amcor Limited in a transaction valued at just over $2 billion. The U.S. Department of Justice objected to one part of the sale -- Alcan Packaging Medical Flexibles, valued at a mere $65 million. Rather then litigate, Amcor agreed to re-sell these assets to a buyer approved by the DOJ.

The principal asset is a manufacturing plant in North Carolina that produces “vented bags for medical use.” The DOJ claims there are only three significant companies that make this exact product, and since Amcor was already one of them, it would be wrong for them to also own Alcan’s plant. By the DOJ’s estimate, the combined Amcor-Alcan would “control” 60 percent of the market. Sure, that’s not a monopoly, but with the third company, the combined market share rises to 95 percent. And this product is too specialized to justify any new entrants into the market, according to the DOJ.

(Actually, it appears the main reason this product is difficult to produce is that it requires a material that happens to be patented by DuPont; but that’s a subject for another post.)

In articulating the “competitive harms” of permitting the Amcor-Alcan merger, the DOJ offers this defense of preemptive intervention:

[B]y reducing the number of significant competitors in the U.S. market for vented bags for medical use from three to two, Amcor and the one other competitor would gain the incentive and likely ability to raise prices through coordinated interaction. The fringe competitors would be unable to render the coordination unprofitable by repositioning or expansion. Coordination would be more likely because, for example, the merger would make customer allocation easier. Each competitor could be reasonably certain as to the identity of the other’s customers, making cheating easier to detect and discipline and, because each competitor is at or near capacity, the ability of each profitably to expand sales and steal business from the other would be limited.

In other words, allowing the merger equals a greater likelihood of illegal collusion or cartel activity. It’s strange the DOJ would make this claim. Most DOJ-prosecuted cartels involve industries where three or more competitors engaged in collusion. Major cases like the computer memory and LCD display markets involve at least seven or eight companies. Reducing the number of major competitors from three to two should have little or no impact on the probability of future collusion. It might impact the DOJ’s ability to learn about collusion, since there’s one fewer company that can squeal in exchange for “amnesty,” but that’s a different argument.

Even with the DOJ-ordered divestiture, there’s nothing to say the three companies won’t still collude. Since the Alcan plant will go to a buyer selected by government fiat, rather then the market process, there’s no way to know ex ante how qualified the new buyer will be to even compete in the industry. Knowing that the government now expects at least three competitors, might the new Alcan owners be tempted to collude with Amcor or the third company to guarantee a certain level of customers and sales?

But let’s return to the larger issue. The DOJ justifies overruling the market because it believes -- without any evidence -- that there’s an increased risk of future illegal behavior. That’s about as non-objective a legal principle as you can state. It applies the same “preemption” logic used to justify the Iraq war to routine business transactions: We must act now because although we don’t know what will happen in the future, we think it might be bad.