Working Paper: NBER ID: w30773
Authors: Alex Reesjones; Ao Wang
Abstract: We present a general approach to experimentally testing candidate reference points. This approach builds from Prospect Theory’s prediction that an increase in payoffs is perfectly offset by an equivalent increase in the reference point. Violation of this prediction can be tested with modifications to existing econometric techniques in experiments of a particular design. The resulting approach to testing theories of the reference point is minimally parametric, robust to broad classes of heterogeneity, yet still implementable in comparatively small sample sizes. We demonstrate the application of this approach in an experiment that tests the role of salience in setting reference points.
Keywords: Prospect Theory; Reference Points; Behavioral Economics
JEL Codes: C14; D9
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| utility derived from consumption is evaluated relative to a reference point (D11) | individuals assess gains and losses based on this reference (G41) |
| uniformly increased payoffs in an experimental choice (D81) | expected choice probabilities reflect this change (D81) |
| salient goals (L21) | serve as reference points (Y20) |
| non-salient goals (L21) | do not influence decision-making (D91) |
| average earnings as a reference point (J31) | fail test (C52) |