Emulation and the Transformation of Social Norms

suppose we want to get other people to do (or not to do) something. There are various strategies we might employ. We could take those people’s preferences as given and try to alter their behavior by altering their incentives through sanctions or rewards. Alternatively, we could try to persuade them to alter their preferences. Both of these strategies can be problematic, however. The latter strategy is highly ambitious in one way (changing people is hard), the former in another (it can be hard to make sanctions stick).

A third strategy is less ambitious and for that reason more promising. This is the strategy of “informing people’s preferences.” Here, we take people’s preferences as given and simply provide people with information about how best to satisfy those preferences. The strategy is a very familiar one; public information campaigns do it all the time.

This article addresses one particular variant on the latter strategy, which plays upon one very particular preference: a person’s preference to do the same as others whom that person esteems. Call it a “preference to emulate.” Such a preference may not be absolutely universal. But commentators ranging from Adam Smith (1759, pt. 1, sec. 1, chap. 3) to Thorstein Veblen (1925, chap. 2) have historically thought that the preference to emulate, or something like it, was pretty nearly so. Contemporary social-psychological experiments also confirm that this preference, or something like it, is strong (Bicchieri and Xiao 2009).¹

The emulation preference is certainly common enough to prove very useful to social planners and public policy makers, if only they can harness it successfully. Playing on the emulation preference in order to alter people’s behavior would appeal to them for many reasons. Compliance is likely to be relatively high, costs low, and intrusiveness minimal. The strategy of playing on people’s emulation preferences is referred to herein as “the emulation nudge.”

THE EMULATION NUDGE AT WORK

Consider, first, a few examples of the emulation nudge and how it has recently been deployed in public policy (Miller and Prentice 2016; Sunstein 2016a, 723–25). These examples provide the empirical base upon which the subsequent discussion will build.

The first is a tax example. Around the world, tax authorities have long been anxious to find ways of increasing voluntary compliance (OECD 2010; Walsh 2012). In February 2011, Her Majesty’s Revenue and Customs agency in the UK—in conjunction with the Cabinet Office’s “nudge unit”—launched an experiment “to establish the impact of altering the messages sent in letters to encourage tax debtors to pay the tax owed.” It was a large-scale trial, involving some 140,000 taxpayers. One version of the letter sent to tax delinquents “informed people that the majority of people in their area had already paid their tax, and ... reminded people about the importance of paying tax for their local services.” Receiving one form of that letter resulted in 83 percent of debtors paying their tax, compared to the 67.5 percent who paid up in response to a letter containing no such nudge (UK BIT 2011, 16; see also Hallsworth et al. 2014). (A letter that referred to what the majority of people “in your town” did evoked far more response [the 83 percent paying] than did the letter that simply reported that a majority nationwide [72.5 percent] did.) The “nudge unit” estimates that, “if applied equally successfully to all new debt, 98.3 percent of people in total would either pay unprompted or through letters alone,” largely obviating the need for more heavyhanded tax collection methods (UK BIT 2011, 16).

A second example concerns energy consumption. A series of experiments tested the effects of sending residential utility customers letters comparing their own electricity usage with that of other people in their neighborhood. A “smiley face” appeared alongside that chart if the customer had used less than their neighbors and a “frowning face” if the customer had used more. Such letters resulted, on average, in a 2 percent reduction in energy consumption. To achieve that same result through price hikes...