Partial equilibrium intuitions about choice
I think it’s fair to say that economists, in general, are disposed to favor “choice”. It is easy to understand why. If you model the world as being composed of rational and well-informed optimizers of their own welfare, giving a person a new alternative cannot possibly harm her, and may well make her better off. In the financial world, the value of an option (which is nothing more than a choice, the right but not the obligation to take some action) is never negative. On the contrary, when they are priced, options often turn out to be very expensive. An extra choice can’t hurt, and may turn out to help quite a lot.
There are behavioral and psychological critiques of this intuition. If people are not well-informed and rational, the availability of poor alternatives can increase the likelihood of costly mistakes. A bewildering array of alternatives may impose a cognitive burden on the chooser, the “cost” of which may outweigh the benefit of a marginally better result. Outcomes that are objectively identical may be subjectively worse when they are the result of open choice than when they result from compulsion or acts of God. Human beings may unpleasantly second-guess or hold themselves accountable for choices whose outcomes are less than perfect, but stoically reconcile themselves to imperfections they could not have prevented.
I want to put those critiques aside though, and stick to the world of conventional economics with its rational, well-informed agents. It is true, in this world, that giving any one person a choice never makes her worse off. But it does not follow, unfortunately, that giving everyone an extra choice will make everyone, or even anyone, better off. That is, to use one of the stylized insults of economists, “partial equilibrium thinking”. If you give me an extra choice, I will only pursue it if it benefits me. But if you give my customer an extra choice, that may very well harm me. If you give everyone a new choice, where the benefits conferred by our own freedom and the costs imposed by the choices of others take us is anybody’s guess.
This fact should be elementary to economists, but somehow it isn’t, at least not when it comes to policy debates. All economists encounter the “prisoner’s dilemma” prenatally. In a prisoner’s dilemma, what is clearly the best “partial equilibrium” choice for every participant — the best choice holding everybody else’s behavior constant — leads to a poor “general equilibrium” outcome when everybody does it. The prisoner’s dilemma is a situation in which all parties would be made better off if everybody involved had an attractive option taken off the table. Another common example is the “tragedy of the commons“. These situations are not at all rare in real life.
It never, ever, follows that creating a new option for people in an economy must make everyone, or even anyone, better off. Economists who worship at the alter of the first welfare theorem and sloppily equate more choice with “more complete” markets need to recall the Theory of the Second Best (ht Yves Smith, long ago). Markets are either complete or they are not. If they are not complete, the kind of intervention often described as “completing markets” (creating new choices, inventing new contracts) might help, but might also lead to very poor outcomes. For example, “more complete” financial markets have recently served to enable banks and institutional investors to customize payoff distributions in order to extract maximum value from government guarantees and foreseeable bailouts.
Partial equilibrium intuitions about choice are particularly destructive in circumstances where there are economies of scale to participation. The Prisoner’s Dilemma is a simple example of that: the benefits of not ratting out accomplices to a crime increase dramatically if everybody stays quiet than if only some people do. But there are many other examples where restricting choice can be Pareto improving when economies of scale obtain.
This rant was most immediately inspired by a lazy style of libertarian argument. The availability of sweatshop work must be a good thing, because workers must find the conditions, however abysmal, to be better than their next best alternative. Open borders are a great way to help the people of poor countries, because working in rich countries confers huge benefits on migrants. Complaints about the effect of “brain drains” are immoral, because they amount to forcing individuals who would benefit from migration to suffer for the questionable benefit of their compatriots.
In practice, I often agree with the lazy libertarians on these issues. I think China has done well for itself and improved the welfare of its people in part by tolerating “sweatshop” labor. My strong prejudice is to support open borders as much as possible. But we can’t think about these questions without considering counterfactuals, weighing the equilibria that would obtain if choice were restricted against the immediate benefits that those given an extra option are able to reap. What is seen and what is unseen, and all of that.
I’m sure that some anti-sweatshop sentiment is more about the narcissistic self-regard of the liberal rich than improving the welfare of the global poor. But the better hippies have thought these issues through and are not idiotically trying to remove workers’ best available option in the name of guilt-free lattes. Organized sweatshop labor displaces other arrangements, some of which may not be as miserable as stereotypes of “subsistence farming” suggest. If there are economies of scale to participation in traditional ways of life (and there almost certainly are), then the fact that people are willing to abandon their villages for sweatshop work tells us very little about whether welfare has been improved or harmed by its introduction. Further, the crappiness of the alternatives faced by potential workers is not independent of the existence of sweatshop work. In countries whose elites do well by arranging the provision of “flexible” labor, the awfulness of alternatives to sweatshop work might be contrived. The notion of an “oil curse” leading to corrupt political arrangements is uncontroversial. Surely a “labor curse” is just as plausible, and the details of its operation would be more pernicious. Arguably, China has done well with sweatshop labor because its elites have perceived “social stability” to be fragile, and have worked to deliver economic development rapidly and broadly to keep the revolutionaries at bay. The sweatshop model might not deliver the goods so well in countries whose leaders are less wary of their publics.
It is not incoherent to argue that a country might benefit from retaining talented people, and it is not even incoherent to argue that individuals who would choose to emigrate might in fact be better off themselves if they as well as all their compatriots could be persuaded to stay and contribute to development at home. Most of us view freedom as a per se good, and for myself, I’d have a very hard time arguing for emigration restrictions anywhere. Model risk is a bitch. That you can tell a story doesn’t mean the story is true, and when the cost of error is uselessly confining people, we should subject our fairy tales to pretty strict scrutiny. Fortunately, the existence of choice is not binary. We can think of “no choice” as a choice where one alternative is accompanied by either an infinite cost or infinite payoff. (That is, I have “no choice” but to stay in-country if the cost of migration or the benefit of staying is infinite.) A state that forbids emigration at pain of jail or death attaches a large negative payoff to trying to leave. But a country might attach a modest cost to emigration, or perhaps subsidize the retention of talented people. This sort of “nudge” does much less damage to norms of personal freedom, and may well contribute to the welfare of both the people affected and the polity as a whole. Indeed, in the US, the same sort of people (like me!) who support open borders are enthusiastic about interventions intended to retain foreign-born entrepreneurs and graduate students by offering them valuable immigrant visas. Whether you want to call this proposal a subsidy or elimination of a cost, it amounts to using the instruments of the state to reshape people’s choice space in ways that are arguably good for them and good for the polity. And ultimately, that is something a state ought to strive to do.
Does this sort of policy translate to “more” freedom or “less”? You can’t say. Freedom is not a scalar quantity. Sometimes actions of the state render one alternative overwhelmingly preferable to any other, and so clearly restrict choice. But the opposite tactic — having the state reshape people’s choice space so that alternatives that become evenly matched and force people to make agonizing tradeoffs, hardly serves the cause of freedom. And in a world of prisoner’s dilemmas, laissez faire policy, leaving the “natural” choice space undisturbed, just turns notional freedom into a figleaf for predictably bad choices and outcomes. People often can and do develop means of cooperating and coordinating to avoid prisoner’s dilemmas without the assistance of states at all, or with forms of assistance that libertarians find unobjectionable, like enforcement of contract. That’s awesome. But the world is full of hard problems with very serious consequences not all of which resolve themselves. It is reasonable that ones enthusiasm for state intervention into the choice space of individuals is conditioned by how prone to corruption and error one thinks the state to be. But it is either simpleminded or cynical to rule out such intervention based on economistic arguments about how choice always improves welfare. That’s simply untrue.
I’m going to end this with a bit from the always wise Nick Rowe:
[P]eople can solve partial equilibrium problems a lot more easily than they can solve general equilibrium problems. When a shock hits, each individual can solve for how his own reaction function should shift in response to that shock. But he can’t easily solve for the new Nash equilibrium point, because that requires him to figure out how every individual’s reaction function has shifted, solve for the new intersection point of all those reaction functions, assume everyone else will do the same, and expect everyone else will move to the new Nash equilibrium too. That’s hard.
If its hard for people to solve general equilibrium experiments, monetary policy should try to ensure they don’t have to solve general equilibrium experiments. Instead, monetary policy should try to ensure that the general equilibrium solution is as close as possible to the partial equilibrium solution, which individuals have a better chance of solving.
Rowe is writing about monetary policy in particular, from (I think) a New Keynesian perspective that assumes the existence of a unique stable long-term equilibrium that is where we want to be. But let’s generalize his ideas to policy in general and a world with a great multiplicity of potential equilibria. Rowe suggests that policymakers should look to the general equilibrium they hope will obtain, and shape the choice space so that decisions made by individuals holding the rest of the world constant move the world towards that equilibrium. In a world with many potential general equilibria (let’s call them “visions of the future”), policymakers must first understand the space of feasible equilibria and choose the one towards which the choice space should be shaped to grope. Choosing a vision of the future and designing policy that moves the polity towards a not-inevitable but intentional and hopefully positive state of affairs? I think that Rowe may have done the impossible and translated the concept of “leadership” into terms that economists can understand. Somebody ring up Sweden!