The Pareto Fallacy

For no reason I can really understand, various technocrats on both ends of the aisles have convinced themselves that the relative rarity of their political preferences in the electorate is attributable to a simple lack of technical expertise among voters. Matt attributes this to what I'll call the Pareto-fallacy (named for the concept of Pareto optimality): The idea that because a certain policy could enhance widespread well-being through progressive and equitable distribution of its benefits, it will. Too often, it won't. We've a world of rational, self-interested actors where certain individuals and groups have far more power than others and that has the distributionary outcomes you'd expect. That we can construct models where power is flattened and Pareto works isn't particularly important.

This has been a nifty trick of the free trade movement: They've transformed debates over specific legislation (NAFTA, CAFTA) into debates over abstract theory. So the economists line up against protectionism, the pundits explain the pitfalls of restricting trade, and nobody actually reads or analyzes the likely outcomes of the particular trade changes proposed by the bill. It's a neat way to pass legislation, but not quite optimal. Congress passes bills, not models.

I've been thinking about this stuff over the weekend thanks to Greg Mankiw's republishing of a suppressed Fortune column he wrote explaining why it's rational for many Americans not to vote. "Sometimes," he writes, "the most responsible thing a person can do on election day is stay at home...If you really don't know enough to cast an intelligent vote, you should be eager to let your more informed neighbors make the decision."

This makes a superficial sort of sense if you assume a commonality of interests between educated and uneducated, poor and rich, black and white, etc. And, to some degree, there certainly is such a commonality. But not on everything. If you receive expansive health care benefits with very little cost-sharing from your employer, for instance, your immediate interests are different from an low-wage, uninsured worker. A system that advantages them will require greater buy-in from you. If you're the owner of a profitable retail outlet with very low labor costs, you don't want a sharp raise in the minimum wage. If you're the low labor cost in question, you do.

In both these cases, the better-off, more educated end is, statistically, better informed and likelier to vote. Their interests, however, do not directly coincide with many of those who aren't voting. Obviously, many of us vote against our economic interests (in both directions) and evaluate on metrics distinct from our direct self-interest. But it's not rational for the country's non-voters to expect that our engagement in the political process will bring about a better outcome for them.

Crossposted, in true Pareto efficiency fashion, at Tapped.