As a consistent libertarian, there are few values that I hold in greater esteem than free trade, in domestic and foreign markets alike. As a matter of principle, we should be deeply suspicious of any competitor who portrays himself as the victim of predatory pricing or, in the international arena, of the "dumping" of cheap imports in domestic markets.

But all sales are not created equal -- or beneficial. Fraud in the marketplace is inconsistent with the principle of mutual advantage, and the sale or importation of dangerous or misrepresented goods poses hidden health threats to innocent parties.

There are U.S. statutes that impose a prohibition against "parallel importation," that is, the resale in the United States, typically at below domestic prices, of drugs that were originally sold exclusively for consumption in foreign markets. If competition is served, so the argument goes, by allowing goods of foreign origin to freely compete with domestic goods, then why treat the reimportation of domestic goods any differently?

Even if we assume the reimported goods are identical to domestic goods, parallel importation still counts as a perversion of the basic principle of free trade.

There is a critical distinction between restrictions imposed by contract with the initial sale of the drug and restrictions imposed by law upon the resale of goods. Quite simply, if restrictions on parallel importation were statutory devices designed to thwart international competition, then they should be rejected -- but they aren't.

Patented goods are subject to a lawful monopoly created by the state in order to induce their creation. No one thinks that new pharmaceutical drugs will be invented by private firms that cannot receive a rate of return sufficient to recover not just the cost of fabricating and selling each pill, but also the huge front-end costs that reach (when dead ends are taken into account) under anyone's estimate in the hundreds of millions of dollars for each new product that reaches the market.

The legal monopoly granted by the patent is the only mechanism that allows the producer to recover those fixed costs, for without it new competitors could produce the same generic compound at a fraction of the price, driving the first drug out of the marketplace.

Once the patent monopoly is granted, the question arises how the patented product should be priced? The state could insist that the fairest thing is to charge the same price to all, which is usually lower than the highest price otherwise would be. Under that approach, people who want the drug and are willing and able to pay for it will get a discounted price, while those with less means will have to pay more or might not be able to purchase it at all.

By contrast, allowing price discrimination (i.e., variation in prices) means the low-demanders can pay a lower price -- but only because the high-demanders are charged a higher price.

In the international context, this practice allows American firms to sell cheaply or even give away their AIDS drugs to struggling African nations, but only if those countries are prohibited from reselling the drugs for a nifty profit in the United States or some other nation.

Price discrimination (and contractually banning resale) increases the potential gains from the patent, which in turn should result in greater efforts to find more patentable drugs -- a win for the poor, a win for patients and a win for the drug companies.

American companies must preserve their ability to price discriminate and seek by contract to limit the resale of the goods back in the United States. This restraint is not imposed by government. It is not antithetical to free trade; it is part and parcel of free trade. If it were possible to enforce contract provisions that required foreign buyers to pay in damages an amount equal to the difference between the local price of a given drug and the price in the U.S., the profit from ignoring the contract would be eliminated -- and so would reimportation.

All too often, these contractual restrictions are worthless because it is hard to trace drugs that pass through several vendors. Thus imposing statutory restrictions on reimportation of patented products is an effective substitute for a valid, if ineffective, contractual restraint.

To eliminate the ban on parallel imports will have at least two undesirable consequences. First, it will restrict needed export sales: if fewer drugs are exported to Africa, then fewer will be reimported. The upshot could be shortages overseas.

Second, it will sap the incentive to innovate at home, for reimportation is just a costly way (two shipments, not one) to avoid a price discrimination regime that is legal and proper under domestic law. It will not do for American law to let foreign pricing practices dictate our own pricing strategies.

Banning parallel imports, alas, fixes the persistent problem of foreign free-riding on American innovation, but the only way to counter that misguided effort is through tough negotiations in which the American government (which shamelessly sponsors export cartels for goods that can be competitively priced at home) should use some of its political power for more sensible ends.

It would be most unwise to imitate the practices of foreign nations in order to undo the domestic patent monopolies that have spurred a level of investment that make this nation the dominant, if unloved, force in pharmaceutical innovation.

Richard A. Epstein is James Parker Hall Distinguished Service Professor of Law, the University of Chicago, and Peter and Kirstin Bedford Senior Fellow, the Hoover Institution.