The catchy phrase is as important in academic writing as it is in popular writing. In motivating their constitution-making stage in The Calculus of Consent, James Buchanan and Gordon Tullock assumed “that the individual is uncertain as to what his own precise role will be in any one of the whole chain of later collective choices that will actually have to be made.” A few years later John Rawls made the same assumption (albeit with different results), but phrased it more quotably as the “veil of ignorance.” Rawls’ terminology stuck. Buchanan and Tullock’s terminology remained just theirs.
Economists have long discussed the implications of transaction costs, decision costs, and information costs. Richard Thaler and Cass Sunstein combined those concepts with some behavioral economics and repackaged it all as “nudge.” In so doing they created an intellectual fad. This is not to begrudge or Thaler and Sunstein, or Rawls, their quotable notoriety. Quite the opposite. My students grasp the concept of “nudge,” and its implications, quicker than they grasp the concept of transaction or information costs. There is something of an irony, however, in that the ideas of transaction and information costs have no intrinsic relationship with behavioral economics. Indeed, imposing any sort of cost on individual behavior to “nudge” or channel behavior toward a given end is a straight-forward application of rational choice theory, which behavioral economics ostensibly undermines. (If behavioral economics taught us that fat people are nudged to eat properly by giving them cookies, then I would agree that it revolutionized microeconomics.)
But I come here to apply it, not to criticize it. That said, there are nudges, and there are elegant nudges. I mean, yes, deciding who must bear transaction costs, and under what circumstances, affects behavior and outcomes. But that’s no surprise. Calling the imposition of transaction costs a “nudge” is simply a nice way to say we’re using a stick, even if it’s a small stick, to whap people to get them to do what we want.
The more elegant subset of nudges are those that don’t impose any whap at all. It’s power without the stick. Setting the culinary agenda in the cafeteria line is the now well-known example of this sort of nudge. Some sort of food needs to go first. The manager can make it healthy food, or inexpensive food, or choose based on whatever the preferred criteria.
A different example exists in parts of constitutional (and legal) requirements. We often think that judges must enforce each part of a constitution or of a law or we won’t observe the expected behavior. Now, to be sure, it’s often true, even usually true, that we need an explicit enforcement mechanism to see the behavior we want to see. But sometimes we don’t need the stick, we don’t even need a small stick, yet we will still see the behavior the constitutional designers want to see. This is an interesting class of cases, even if it’s small.
One example is the U.S. Constitution’s Origination Clause. This clause provided that “All Bills for raising Revenue shall originate in the House of Representatives; but the Senate may propose or concur with Amendments as on other Bills.”
To be sure, contrary to speculation beforehand, the U.S. Supreme Court announced in 1990 that the Original Clause did not present the Court with a non-justiciable political question. So the Court would strike down legislation contravening the clause. Nonetheless, the Court’s interpretation of the Origination Clause is narrower than the House and the Senate’s interpretation, and it wasn’t clear before the 1990 case that the Supreme Court would enforce the clause at all. For two centuries, without judicial enforcement, the vast number of budget bills originated in the House of Representatives. And they still do, even when beyond the Court’s narrower construction of the Clause.
Why is this notable?
Voting sequence on legislation between chambers in the “strong” bicameralism existing in U.S. state and national legislatures is a version of the canonical “Battle-of-the-Sexes” game. In this game the two players have incentives to coordinate their actions, but there is also distributive tension between them. In the case of bicameral legislation, both chambers need to agree on enacting a piece of legislation. At the same time, the chamber that initiates the legislation can often derive a first-mover advantage relative to the second-voting chamber. As in Battle-of-the-Sexes, both chambers need to cooperate to enact legislation, but there is also distributive tension between the chambers regarding getting a bill closer to their respective preferences.
There are multiple equilibria in this game; either chamber can vote first. For most legislation, which chamber votes first is just a matter for the chambers to pound out between themselves. And if things get too complicated, or if preferences are too intransigent, the chambers use some sort of reconciliation procedure to work out their differences, like continuing to send amended versions of the bill back and forth between the chambers, or going to conference committee.
One way to solve the problem of multiple equilibria in a Battle-of-the-Sexes game, however, is to create a “focal arbitrator” so that both players implement the same equilibrium strategies. This person has no power other than to announce one equilibrium for both players to focus on. This is purely rhetorical. There is no cost extrinsic to the equilibrium payoffs in the game itself for ignoring what the focal arbitrator says. The focal arbitrator exercises power without a stick.
Focal arbitrators are typically conceived as people who announce which equilibrium players should focus on. In the case of the Origination Clause, however, the focal arbitrator is not a person, it’s a text. The constitutional text has served as focal arbitrator between the House and the Senate for over 200 years, successfully coordinating the sequencing of (almost all) legislative activity on money bills. The “nudge” is purely rhetorical, there is no cost, no stick, of any size.
That’s a truly elegant nudge. Nothing more than a vapor, but it’s gotten the job done.
To be sure, most behavior won’t fall into classes in which nothing more than rhetorical nudging will work by itself. Nonetheless, I tend to think we run to the stick too quickly, even if it’s a small stick. We sometimes miss how much behavior results from texts (and people) nudging as no more than focal arbitrators.