Working Paper: CEPR ID: DP14751
Authors: Shilpa Aggarwal; Rebecca Dizon-Ross; Ariel Zucker
Abstract: How should the design of incentives vary with agent time preferences? We develop two predictions. First, “bundling” the payment function over time – specifically by making the payment for future effort increase in current effort -- is more effective if individuals are impatient over effort. Second, increasing the frequency of payment is more effective if individuals are impatient over payment. We test the efficacy of time-bundling and payment frequency, and their interactions with impatience, using a randomized evaluation of an incentives program for exercise among diabetics in India. Consistent with our theoretical predictions, bundling payments over time meaningfully increases effort among the impatient relative to the patient. In contrast, increasing payment frequency has limited efficacy, suggesting limited impatience over payments. On average, incentives increase daily steps by 1,266 (13 minutes of brisk walking) and improve health.
Keywords: Incentives; Time Preferences; Behavioral Change; Diabetes; Randomized Controlled Trial
JEL Codes: D03; D91; I12
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| time-bundled contracts (D86) | compliance among individuals who are impatient over effort (D11) |
| payment frequency (G35) | compliance among individuals who are impatient over payments (D15) |