Do Financial Incentives Reduce Intrinsic Motivation for Weight Loss?: Evidence from Two Tests of Crowding Out

Background: Financial incentives have been used successfully to promote health behaviors, however there is concern that incentives may be counterproductive if they “crowd out” pre-existing intrinsic motivation and lead to a decrease in performance once incentives are removed to a level lower than had they never been introduced. Existing evidence of crowding out may not be generalizable to the context of health behaviors and there is little empirical evidence on the direct effects of financial incentives on motivation in this context.

Objectives: To determine whether financial incentives crowd out intrinsic motivation for weight loss using direct measures of motivation in two randomized controlled trial settings.

Methods: We measure intrinsic motivation directly in two randomized controlled trials of financial incentives for weight loss with different incentive structures (cash rewards versus deposit contract). In both experiments, substantial financial rewards were offered to voluntary participants who were overweight for meeting weight loss goals over six months, followed by a three-month no-incentive follow-up period. We use the Treatment Self-Regulation Questionnaire, a survey tool developed in the psychology literature, to obtain consistent measures of intrinsic motivation at baseline, the end of the incentives period (six months), and the end of the follow-up period (nine months). We test for differential changes in these motivation scores over time by study arm. Crowding out would be demonstrated by a relatively greater decrease in intrinsic motivation in financial incentive arms than in the control arms. 

Results: We find that participants were highly motivated at baseline and that incentives did not crowd out this intrinsic motivation whether they were designed as deposit contracts or cash rewards. The average intrinsic motivation level at baseline was 6.36 (SD 0.89) in study 1 and 6.47 (SD 0.76) in study 2 on a scale from 1 (lowest) to 7 (highest). There were no statistically significant differences in the change in intrinsic motivation over time, comparing any of the incentive arms to the respective control group in each study using Wilcoxon rank-sum tests and correcting for multiple comparisons as appropriate. This lack of crowd out is consistent with behavioral evidence of the long-run effects of incentives seen in the field. Results are consistent across subgroups of potential interest (e.g., those who earned relatively more money via financial incentives) and robust to correction for censoring and sensitivity analyses in handling of missing data.

Conclusions: Direct comparisons of intrinsic motivation levels among participants who received financial incentives for weight loss and those who did not show that motivation levels started high and did not change differentially for participants in incentive versus no-incentive arms over time, inconsistent with the crowding out hypothesis. This finding is important given the growing use of incentives to promote behavior change by policymakers and employers and the policy debate on whether financial incentives are likely to be counter-productive in domains like health.