Rethinking Behavior Change, Nudge-style
A prevailing regime by which groups, organizations, and institutions attempt to alter the behavior of its members and constituents is through imposing penalties and fines, which seek to deter certain behaviors. Parking tickets intend to prevent people from parking in certain areas, sometimes at certain times. Prison sentences, and the death penalty, are intended to serve as deterrents for serious legal violations.
However, fines often prompt behaviors different from what those trying to mould behavior (e.g., governments or organizations) intend. Many studies have shown that the death penalty/prison is not a deterrent to violent crime (see here). In a study of a daycare where several parents repeatedly picked up their children late from school, researchers found that the imposition of a fine for late pick-ups actually increased the number of parents picking up their children late. Additionally, when the fine was lifted, the behavioral change remained such that more parents still picked up their children late. Gneezy and Rustichini, the authors of the study, argue that parents saw the fine as a cost, which they were willing to pay, when previously there was a moral, not a financial, meaning to picking up children late.
An alternative approach to behavioral change that has received plenty of attention in the last several years is described by the behavioral economist Richard Thaler and the legal scholar Cass Sunstein in their 2008 book, Nudge: Improving Decisions About Health, Wealth, and Happiness. In it, Thaler and Sunstein argue, using copious evidence from cognitive psychology and behavioral economics, that our cognitive architecture creates systematic biases in decision making that cause problems in certain domains. Because we often rely on heuristics deriving from automatic processing of information (as opposed to deliberative processing, see Daniel Kahneman’s new book for far more details about this), we often err especially in domains of logic and statistics.
Enter: choice architects and their nudging solutions. Thaler and Sunstein argue that, however informal the policy and at whatever level it is enacted, the individuals who design program or policies—choice architects—can exert a good deal of influence over the kinds of decisions others make through “nudges.” These nudges are supposed to a) recognize common decision making errors and b) alter the decision making context in a way that acknowledges those biases. A nudge, for Thaler and Sunstein, is any aspect of design that “alters people’s behavior in a predictable way without forbidding any options or significantly changing their economic incentives” (p. 6). This caveat, that nudges do not shut off any behavioral options, allows Thaler and Sunstein to call their approach one of libertarian paternalism, whereby freedom of individual choice is preserved (the libertarian part) and choices are influenced such that the “choosers are better off,” according to their own standards (the paternalism part). So, a woman working in a school cafeteria who recognizes that students’ food choices are determined by the order and arrangement of the types of foods, and who changes the arrangement in a way that promotes more healthy eating behaviors is a choice architect employing a nudge toward a particular goal. And Sunstein, as the current administrator of the White House’s Office of Information and Regulatory Affairs, seeks to build these insights about human cognition into a variety of policies.
In a recent Sociology Compass article, Evan Selinger and Kyle Whyte, both professors of philosophy, raise a number of nudge issues. First, they suggest, many of the examples often cited as nudges do not actually meet the criteria Thaler and Sunstein set for nudges; they call these “mistaken nudges.” One of Thaler and Sunstein’s main points is that that nudges are modifications that do not change people’s financial incentives and do not add new costs to situations. But, Selinger and Whyte argue many of the programs that are touted as nudging behavior fail to meet this criterion. They often change financial incentives and expect individuals’ behavior to fall in line with those incentives (often referred to as “acting rationally”) in much the same paradigm of penalties and fines. As an example, Selinger and Whyte argue that the Toxic Release Inventory, which provides information about how much companies pollute, should not be considered a nudge since it actually increases the costs to companies of polluting. In general, Selinger and Whyte note there is some confusion about what constitutes a genuine nudge as defined by Thaler and Sunstein.
In addition to issues of definition, Selinger and Whyte review the ethical concerns other scholars have raised concerning nudges. Do nudges really preserve individual choice? Might they make use morally lazy by letting us rely on the infrastructure set up by others for our decisions? Will the widespread use of nudges lead to less practical wisdom, a devalued public sphere, and a more simplified public life? Others make a slippery slope argument that introducing behavioral changes through interventions might lead people to accept more definitive control from government in their lives. The philosopher Thomas Nagel has argued that some biases might actually derive from something that is otherwise socially useful, so it is worth figuring out which biases should be “worked with,” and which should be challenged. Some of these concerns seem overstated, perhaps relying on an overly abstracted concept of nudges and an imagined future that seems unlikely to occur.
The most important and significant criticism, from my perspective, is that choice architects get to choose which values and preferences they promote with nudges. Here it seems useful to distinguish between nudges that are intended to alter significant lifestyle behaviors in a way that requires privileging a goal (e.g., getting people to stop smoking), and nudges that intend to make the small-scale behaviors individuals are already compelled to do more efficient (e.g., getting people to pay their fines in a more efficient manner or to complete their tax forms correctly). Some nudges change behavior in some direction or towards some end, while other nudges adjust existing policy to take into account how individuals often behave. In the latter case, few would fault the government for trying to improve compliance on tax forms given that tax collecting is a basic task of the state. Using nudges to improve the efficiency and the rates of compliance for basic governmental tasks seems far less ethically problematic than using nudges towards ends about which people disagree.
A final concern of Selinger and Whyte is practical: They argue that Thaler and Sunstein fail to provide an adequate roadmap for implementing nudges, a process which has the potential to be very complicated. In particular, Selinger and Whyte point out that the meaning individuals attach to different nudges might vary dramatically, which has implications both for perpetuating potentially problematic associations (e.g., including a male voice in German cars to inform drivers when they are speeding, as drivers did not respond to female voices) and for the effect of nudges in different situations and populations. It is certainly important to understand variation in how individuals assign meaning to nudges; the upshot seems to be that policymakers and choice architects must fully understand the social context in which they are applying nudges, which likely requires a good deal of groundwork and pretesting before particular nudges are deployed.
What the critics of policies that are designed to address the cognitive underpinnings of decision making might overlook is that most policies currently “nudge” us in some direction simply by virtue of building in default choices (e.g., in the case of organ donation) and assuming particular models of decision making (and, by consequence, decision makers) in policies. If our behavior is currently being shaped by policies and programs based on long-existing structures independent of the intentional designs of others, is that a violation of democratic principles? We might ask what features of organizational structures and arrangements act as nudges for behavior independent of the intentions of others to guide our behaviors in such ways.
“Is there a Right Way to Nudge? The Practice and Ethics of Choice Architecture.” Evan Selinger and Kyle Whyte. Sociology Compass, 2011.