Economists' Learned Helplessness — Confessions of a Supply-Side Liberal

*“Top 10 Unethical Psychological Experiments”*

As economists, it is important for us to pay attention to the unintended side effects of our usual initial working assumption that people are fully optimizing–doing the best they can given the situations they are in. To the extent we treat this not just as an initial working assumption, but as if it were the God’s truth, we are in danger of missing opportunities for helping people make better decisions.

Fortunately, economists don’t routinely assume that those in government are always optimizing. Instead, a routine starting point for policy analysis is to act as if those in government want to improve the general welfare (along with some more self-interested motives), but don’t always know how. In an excellent Project Syndicate essay, “The Tyranny of Political Economy,” Dani Rodrik writes about the learned helplessness that can result if one follows to a logical conclusion the assumption that those in government are already fully optimizing–often in a self-serving way–subject to their constraints. He argues that new ideas and advice can make a difference.

Dani Rodrik worries that the logic of optimization is leading economists to doubt, on principle, whether policy advice can make any difference. I worry that the logic of optimization is leading economists to doubt, on principle, whether advice to households or firms can make any difference. If we assume people are already optimizing, where in fact they are not, then we will be blind to opportunities to help. If individuals are optimizing 95% of the way, the approximation that they are optimizing 100% of the way could well be an appropriate simplification in building a larger model, but when focusing attention on that decision, it still leaves a 5% leeway for improvement. That 5% improvement in decision-making could correspond to a large increase in welfare–an especially important opportunity because the increase in welfare from better decision-making would require no coercive action, but only persuasion based on the hearer’s appropriate self-interest.