With trust-busting all the rage these days, especially on the Liz Warren left, this seems like a good time to repost something I wrote fifteen years ago:
The right way to look at antitrust law is not whether economic theory suggests that particular monopolies can and will exist that can, in theory, be corrected through government action, to the benefit of both consumer welfare and economic efficiency. The real question is whether, once one establishes a regime of antitrust law, whether that regime is going to do more overall harm than good.
Sure, the antitrust authorities may identify "real" monopolists here and there. But they will also likely often misapply economic theory, or apply the wrong economic theory, and break up "monopolists" that were economically efficient and serving consumer welfare.
Meanwhile, various companies will try their darndest to persuade the antitrust authorities that their competitors are "monopolists" that should be reined in. Every time they are wrong and the authorities don't believe them, their lobbying efforts (and the contrary lobbying efforts of the alleged monopolist) are a deadweight lost. Every time they are wrong and the antitrust authorities do believe them, further damage is done to the economy (plus the dead weight lobbying losses).
Moreover, there is no reason to assume that antitrust authorities will always be pursuing the common good. Antitrust enforcement, and antitrust law itself, will develop through political mechanisms, and the authorities will be sorely tempted to punish the enemies of their political allies and reward their friends. Demagogic politicians will order their governments to go after large companies for political gain, not for sound economic reasons. Much of American antitrust law for many years was based on no economic theory whatsoever, but on general mistrust of "bigness," a desire to protect mom and pop businesses from more efficient competition, and demagoguery starting with the American king of the genre, Teddy Roosevelt. And so on.
The losses from antitrust law can be huge: the costs of rent-seeking and countering rent-seeking alone can dwarf any gains from good antitrust enforcement, and antitrust enforcement won't necessarily be good. The question, to this extent, is whether you would rather have market outcomes be governed by Wall Street or K Street. Meanwhile, the losses from monopoly are not likely to be huge. The monopolist sells at a monopoly price, and gains monopoly prents, but the product is still, after all, on the market. And the available monopoly rents should incentivize competitors try to break the monopoly.
So, in my view at least, the libertarian case for not having antitrust law is not that the market is perfectly efficient, but that regulation is worse. I'd be content to leave antitrust law to price-fixing, as the Reagan Justice Department did, but that leaves the regulatory superstructure in place, to create further mischief in the future. So abolish the Antitrust Division, reduce the FTC's functions to interstate fraud, and give it jurisdiction over price-fixing, too. And then be content with the results with an imperfect market.