New Paternalism on the Slippery Slopes, Part 4: Context Dependence

New paternalists have also relied on the notion of context dependence to justify their policies. But as with hyperbolic discounting, they unjustifiably assume the existence of an inconsistency of preferences gives the policymaker license to choose among the inconsistent preferences. That assumption is the paper’s next target (pp. 703-704):

For a variety of decisions, people are subject to what behavioral economists call context-dependence. This means that how they choose among two or more options depends on seemingly irrelevant aspects of how the situation is described. For example, medical patients are more likely to assent to a treatment with a 90% survival rate than one with a 10% death rate, even though these are the same. In this case, people seem to favor “positive” over “negative” framing. People also seem to prefer options framed as the existing or a baseline position; this may be called status-quo bias. Another example of the power of framing is the persistent difference between willingness-to-pay (WTP) and willingness-to-accept (WTA), meaning that people will demand more money to part with an item than they will pay to acquire it, even when the item’s value is a trivial portion of their wealth or income.

The phenomenon of context-dependence underlies various new paternalist proposals. All of Sunstein and Thaler’s proposals for new contractual defaults, for example, rely on the difference between WTP and WTA. Although such defaults leave all contractual options open (at least for the most modest proposals), employees may be less willing to part with a given term (such as guaranteed paid vacation) than to bargain for its inclusion. If there were no difference between WTP and WTA, and if transaction costs were zero, then the realized terms of contract would be the same regardless of the default.

The problem with context-dependence is similar to that of hyperbolic discounting: the new paternalist argument relies on an internal inconsistency to justify intervention. There is no theoretical basis for choosing which behavior represents the individual’s “true” best interest as he sees it. Which better represents a person’s real preferences: what he is willing to pay for something or what he is willing to accept to part with it? There is no theoretically correct answer to this question, as Sunstein and Thaler admit: “If the arrangement of the alternatives has a significant effect on the selections the customers make, then their true ‘preferences’ do not formally exist.”

In the absence of a true underlying preference as the correct standard, what standard should be used? Sunstein and Thaler decline to answer that question: “We are not attempting to say anything controversial about welfare, or to take sides in reasonable disputes about how to understand that term.”

In short, there is no standard provided by behavioral economic theory. The answer to the “what standard” question will depend on policymakers’ own particular notions of welfare and well-being, as well as the weight they attach to autonomy. Notably, behavioral economics does not necessarily place any weight on autonomy, despite Sunstein and Thaler’s obeisance to the value of individual choice. Policymakers who adopt the new paternalists’ approach need not share their belief in choice. The new paternalist paradigm places them on a gradient from policies that only mildly restrict choice to policies that restrict or abolish it.

Additionally, we should realize that the supposedly beneficial default rules that S&T favor, such as a presumption of paid vacation, are not necessarily costless. Wages will adjust to account for the value of additional benefits. Thus, a new default rule does not simply give workers something they lacked before; it gives them something in exchange for losing something else.

(As usual, full citations are available in the full paper. Cross-posted at ThinkMarkets.)