Does Affect Have An Effect? Willingness to Pay and Psychophysical Numbing

There has been a wealth of academic literature assessing various influences on citizens’ willingness to pay for public services. Most of the models either explicitly or implicitly assume that citizen preferences are either... [ view full abstract ]

There has been a wealth of academic literature assessing various influences on citizens’ willingness to pay for public services. Most of the models either explicitly or implicitly assume that citizen preferences are either fixed based on some exogenous set of characteristics (age and income, for example) or are at least driven by more or less predictable influences.

However, the behavioral economics and cognitive science literature suggests that this assumption may not hold. Starting with the work of Tversky and Kahneman (1974), several authors have established the fragile nature of preferences and how various “heuristics and biases” can affect people’s perceptions of the costs and benefits of activities.

In this research, we will examine the effect of one particular heuristic on citizen preferences, the so-called “affect heuristic” (Zajonc, 1980). First discussed in the early 1980s, the literature on the affect heuristic emphasized the simultaneity of emotions (affect) and cognitive processes, as opposed to the traditional model of decision making where cognition preceded emotion.

One of the cognitive effects that is produced by a reliance on affective responses to help simplify complex problem has been termed “psychophysical numbing” (Fetherstonhaugh et.al., 1997). This happens when individuals demonstrate a marked response to rather small changes in their environment while at the same time demonstrating smaller responses to reports of large changes. One of the corollaries of this effect is insensitivity to numbers. Fetherstonhaugh et.al. (1997, Study 3) demonstrate this effect in a study where respondents were asked to provide estimates of how many lives saved with a specified affliction would warrant a donation of $10 million to a medical nonprofit treating that affliction. Contrary to what would be expected if respondents weighed the benefits of their donation using “rational” cognitive processes, the amount of lives saved that justified the donation rose with the at-risk population.

We employ an experimental methodology to assess how willingness to pay for public services is influenced by numbers alone (which can be considered “low affect” prompts) versus numbers and pictures (increasing the affect of the prompt). Briefly, respondents will be randomly exposed to a question of how much they would be willing to pay per month in increased taxes to support a project. The design will be a 2x2 ANOVA design; the random assignment will consist of a low numerical impact value being presented alone, a high numerical impact being presented alone, a low numerical impact value combined with a picture that communicates the need for the service, and a high numerical impact combined with the same picture. If affect has an influence on willingness to pay we would expect that the numerical information combined with a picture will increase the willingness to pay for public services. Further, it should help mitigate the “insensitivity to numbers” problem by increasing the willingness to pay at a rate more than the increased benefit values alone. Given that the presentation is randomly assigned we do not expect any systematic differences, but we will collect basic demographic information to check for observational bias.