I have mused that one way to construct an argument against the permissibility of above-cost mixed bundling (and against applying a predatory-pricing analysis to mixed bundling) is to look at the cross subsidy from the monoply market for Wm to the competitive market for Wc. Dan Crane responded that cross subsidies using the monopoly profits shouldn’t matter:

When A offers the bundle at a $1.50 discount, it is sacrificing a profit of $1.50 per sale in order to “exclude” B from the market. Money is fungible. That sacrifice is identical to the sacrifice A would make if it took $1.50 out of a bank account to pay for single-product predation. It is no more plausible to think that A would sacrifice the $1.50 of monopoly profits to effectuate an above-variable-cost price that B must meet than to think that A would withdraw that money from its bank account to engage in single-market predation. The recoupment doesn’t happen any sooner. A can only charge a monopoly price for Wc if it first drives out B.

There is more to the story, I think, and I’m not quite ready to give up the argument. A big part of the problem in antitrust is that we need not only to decide what conduct is anticompetitive, but also what conduct can be defined legally so that antitrust enforcement does not sweep up competitive conduct.

Looked at it from that perspective, the question is not whether direct subsidies from a bank account would have the same economic effect as cross-subsidies through mixed bundling. The question is whether mixed bundling can be regarded as anticompetitive and whether it can be delineated clearly enough to make enforcement possible. The fact that A has a monopoly on Wm (graph here) is anticompetitive in the sense that the monopoly produced dead-weight loss and the other costs of monopoly (such as the monopolist’s inefficient rent-seeking behavior to attain or maintain the monopoly, and the cost of patroling the markets against such behavior). Cross-subsidies through mixed bundling are part of the social cost of monopoly. It is true that we do not forbid monopoly profits; but the reason is not because monopoly profits aren’t anticompetitive; the reason is that the prospect of monopoly profits creates desirable incentives to innovate. Once monopoly is attained, certain conduct permissible in a competitive market becomes unlawful. The same logic can be applied to mixed bundling. The monopolist itself, by virtue of the bundle links the monopoly and competive market and enables enforcement against the specific use of monopoly profits. The use of monopoly products becomes cognizable by the antitrust laws through the bundle itself. It’s not a perfect argument, surely, because it must needs acknowledge that only one part of the anticompetitive concern is addressed. On the other hand, the argument leads to a clear rule: No mixed bundles with monopoly products.

Mixed-bundling is different from the direct subsidy of Wc from a bank account. In that case we cannot tell whether the direct subsidies create anticompetitive concerns until the price of Wc drops below cost, at which point no other economic explanation but the expectation of recoupment is available. Some bank-account subsidies of Wc may be anticompetitive (that is, aimed at exclusion and motivated by the hope for recoupment), but we cannot tell—and we generally like lower prices.

Another interesting point about this argument against mixed bundling is the comparison with non-bundled subsidies. Assume that Wm is a particular kind of table (monopoly price of $100) on which A has a monopoly and that Wc is a chair (competitive price of $50, cost of $30) that goes with the table. My hypothetical rule would forbid a mixed bundle of table and chair for $140. Yet does it make sense to forbid mixed bundles if A could just lower the price of the chair from $50 to $40, but not bundle it with the table? The economic effect at first blush seems the same because a consumer buying a table and chair from A would still pay $140, instead of $150, when buying a table from A and a chair from B. But without the bundle, A is lowering the price for chairs also for those consumers not interested in the table, creating increased demand for chairs and for tables as a complementary product (and incidentally a greater incentive for other firms to enter the table market in competition wth A). The effect then is that A may find itself selling so many chairs that the cross-subsidies become economically unfeasible and the market will self-correct (that is, A will raise the price for chairs again).

I am not sure that this argument for the illegality of mixed bundles quite works. But I do think that the fungibility of money isn’t a sufficient counterargument against the illegality of discernible cross-subsidies using monopoly profits.

Technorati Tags: antitrust, monopoly, economic analysis, mixed bundles, cross subsidy

Comment feed: RSS 2.0. You can leave a response, or trackback from your own site.