In the midst of the din over Charlottesville, let’s not overlook the Trump administration’s controversial stance on free trade. This smoldering problem has risen to the surface in the delicate negotiations that the United States now is undertaking with China over the status of American intellectual property rights, and with Mexico and Canada over the North American Free Trade Agreement. Before looking at the particulars of these two disputes, it is instructive to set out the intellectual case for free trade, which Trump has consistently misunderstood.

The basic insight here is that ordinary contracts between private parties are not neutral; they produce gains for each party. If I swap my horse for your cow, it is tempting, but wrong, to say that no value has been added for the parties because we have the same horse and cow after the transaction that we had before it. So why worry, the argument goes, if the trade does not take place? This facile argument ignores that these transactions are costly to complete. Why would two parties waste money to organize a trade from which neither side has gained? Even this simple trade is a positive-sum game that produces for each side a net advantage that exceeds the costs of putting the deal together. I could desperately need a cow for milk or breeding, and you might need the horse to pull a plow. The trade allows both of us to get greater value by the more efficient deployment of existing resources. That short-term advantage has long-term effects, by letting me breed horses while you breed cows.

But there is a catch. In general, bartering is inefficient because it is rare that any random pair has the goods that the other side wants in just the right amount. But introduce money, so that the sale of each animal no longer depends on the purchase of the other, and the gains from trade become ever more salient.