Regular readers of Cafe Hayek know that I’m a big fan of Pierre Lemieux. His writings on trade are among the most insightful and important being offered today by anyone. And I’m eager for Pierre’s new book on trade to be published; it’s forthcoming from the Mercatus Center. I’ve read it and recommend it enthusiastically.
Here’s a new blog post, from over at EconLog, by Pierre. While I agree with much of it, I wish that Pierre had resisted the convention of ‘admitting’ that, as Pierre writes,
just as in the case of domestic trade and competition, some individuals will be economically harmed. In other words, more free trade is optimal in the Marshallian sense (maximizing net benefits across all individuals as benefit-cost analysis tries to do), but it is not Pareto-optimal (it is not generally true that it harms nobody at all). Moving from protectionism to free trade is likely to harm some individuals.
I’ve written on this particular matter at length. My objections to ‘admitting’ that ‘trade has losers’ can be found at the many links here. See also here. In this blog post, therefore, I’ll not rehearse my many reasons for objecting to this ‘admission’ that ‘trade has losers.’ But Pierre’s wording gives me the opportunity to make my central point in a way that I believe I’ve not yet done – which is this: Pierre is mistaken to write that trade fails to satisfy the Paretian criterion for efficiency.
Before I justify my objection to Pierre’s claim, here’s a quick explanation, for non-economists, of the meaning of Pareto optimality. A situation is Pareto optimal if it cannot be changed in a way that makes one or more persons better off without making at least one person worse off. (Incidentally, a Pareto-optimal situation – or, as economists typically say, a Pareto-optimal allocation of resources – while it can be a situation that is regarded by all as being ideal, need not be one that most people regard as ideal. This issue is beyond the scope of this post.) But when a situation is not Pareto-optimal, then the possibility exists to change that situation in a way that makes at least one person better off without making anyone else worse off. Such a change is called a Pareto-superior change, or ‘move.’ Most people would agree that if such a change can be brought about, it ‘should’ be brought about – or, at least, there are no good ethical or economic reasons to object to such a change.
A change is Pareto superior if that change makes at least one person better off without making anyone else worse off. In contrast, a change that makes almost everyone better off fails to satisfy the Paretian criterion for efficiency if that change makes even as few as one person worse off. Pierre’s statement that free trade is not Pareto optimal is, I think, meant to be a claim that free trade is not Pareto superior. That is, what Pierre is here saying is that, by changing the pattern of consumer expenditures today, free trade makes consumers and some producers better off but it also makes other people – most notably, some business owners and workers – worse off. This standard economic justification for free trade is that, while some owners and workers do suffer today, the total size of the gains to others – most notably, consumers, but also other business owners and other workers – is larger than the size of the total losses and, therefore, trade is economically ‘efficient’ and, hence, economically and ethically justified.
I do not dispute the textbook Marshallian demonstration that the total size of the monetized gains to the ‘winners’ from trade in any particular instance are greater than the total size of the monetized losses suffered by trade’s alleged ‘losers’ in that particular instance. But the case for free trade is much stronger than is implied by this formulation. Again, free trade is not only ‘Marshiallian’ efficient (or, Kaldor-Hicks efficient), it is also, at least plausibly, Pareto superior. That is, free trade does indeed arguably satisfy the Paretian criterion for efficiency.
How can my claim be correct? After all, by causing – or allowing for – changes in the pattern of consumer spending, trade does indeed make some owners and workers worse off. True. One way that I’ve argued in past posts in support of the proposition that free trade has no losers is to insist on a longer time period, beyond the mere ‘now,’ for assessing trade’s total effects. The longer the time period taken into consideration, the number of losers falls while the number of winners rises. The autoworker who loses his job today because of trade perhaps got that job to begin with because of trade (for example, because American automakers export some of the vehicles they manufacture). In such a situation, in what way is this autoworker a loser from trade? At the very least, the claim that he is a loser from trade requires for its justification something more than pointing out that he lost his job today because of trade.
Here, though, I wish to take a different, if related, approach. I want to argue that when considered from the standpoint of individuals’ expectations, trade is Pareto superior.
Consider Smith, who must choose which country to move to: Bastiatland or Trumpia. In Bastiatland free trade is the policy; in Trumpia, no foreign trade is allowed. Smith does not know what his occupation will be before he actually commits to move to one or the other of these two countries. But Smith does know enough economics to understand that, in part because of free trade, the ordinary denizens of Bastiatland have a higher standard of living than do the denizens of Trumpia. Smith also knows that, whatever his occupation in Trumpia, it will never be destroyed by imports. And he knows that, if he moves to Bastiatland, he is likely to get a job that might one day be destroyed by imports.
Nearly all economists – even those who support free trade only because of the ‘Marshallian’-efficiency reasons mentioned above by Pierre – will predict that an economically literate Smith will choose Bastiatland over Trumpia. I, too, make this prediction. I make this prediction even knowing that Smith knows that his risk of losing a job in Bastiatland is much higher than is his risk of losing a job in Trumpia. Smith will choose Bastiatland over Trumpia despite knowing that he is more likely to lose his job in Bastiatland. Smith chooses Bastiatland because, at the time of his choice, he assesses his expected stream of lifetime earnings – or, even, his expected stream of lifetime well-being – in Bastiatland to be higher than his expected stream of lifetime earnings or well-being in Trumpia.
If we agree that Smith would be perfectly rational to choose Bastiatland over Trumpia despite his knowledge of the greater risk of job loss in Bastiatland, this fact suggests that the appropriate comparison is between a policy of free trade and a policy of protectionism. Put differently, this fact suggests that it is inappropriate to judge the efficiency or welfare properties of free trade (or of protectionism) by looking only at particular outcomes that take place under each policy.
Of course if Smith, living now Bastiatland, loses his job to imports he will be disappointed. But if he chose to live in Bastiatland knowing of the risk of job loss, then the job loss is not a relevant Paretian welfare loss. In this case Smith must be presumed to have factored in the risk of losing a job when he chose to move to Bastiatland. Incurring this risk was part of the price that Smith voluntarily paid in order to attain what he obviously judged – what he expected – to be the more-valuable prospect of enjoying a larger expected lifetime stream of earnings in Bastiatland as compared to what he would have enjoyed in Trumpia. Smith’s cost is, therefore, not a loss. And if Smith’s cost (his actual loss of a job) is not a loss, then Smith cannot be said to be among the losers of Bastiatland’s policy of free trade. Free trade might well make everyone in Bastiatland better off, if only from an expectational standpoint. If so, then free trade is indeed the Pareto-superior policy.