If you were given a small reward every time you worked out, would you be more inclined to stick to a permanent exercise regimen? How much would that incentive have to be? Would regular exercise be beneficial for your health and ultimately to the greater society?

Monetary incentives are being increasingly introduced by policymakers into all kinds of interventions across a wide range of sectors including health, education, and the environment. The rationale for this trend is that we often make choices without accounting for how these choices may affect other people (and sometimes even ourselves). Obesity is one such example. It imposes high healthcare costs on society. Economists argue that well-designed monetary incentives can drive people to alter their choices so that everyone can benefit.

​In health, the bulk of existing literature concludes that financial incentive schemes promoting healthy behavior are quite effective in the short run. Think about all the times we’ve rushed to our favorite exercise class lest we be fined for a no-show or were rewarded for showing up for ten classes in a row. Now what happens when those incentives eventually disappear altogether?

There are two theories on the long-term effects of monetary incentives. The motivation crowding theorists say financial incentives can lower helpful behavioral motives by signaling that changing current behavior is difficult or unattractive.  As a consequence, for instance, people go to the gym less often than they would have without monetary rewards. The opposition (the habit-formation crowd) claims there is a positive correlation between past and present consumption which implies that a short-run monetary incentive for healthy behavior will translate into habitual behavior.

A new paper by Boris Augurzky, Thomas K. Bauer, Arndt Reichert, Christoph M. Schmidt, andHarald Tauchmann experiments with the latter theory using data from a randomized experiment investigating monetary rewards for weight reduction in obese people. A sample size of 700 participants was randomly assigned to three experimental groups; one group received €300 to achieve individually assigned contractual target weight losses between 6 and 8 percent within four months. Another group received €150 euros for the same objective. All participants (with or without weight loss rewards) who achieved 50 percent of the contracted weight loss in the first phase, were randomly assigned to a second weight-maintenance intervention (with groups receiving €500 and €250 respectively to maintain their weights). This came as a surprise to those assigned to the two intervention groups because participants had not been informed about the second weight‐maintenance intervention, thus eliminating any anticipation during the weight-loss intervention.

So what were the results of the experiment?

At the end of the first phase, all groups on average weighed less than at the start of the experiment with statistically significant weight losses in both treatment groups. A similar result was observed following the second intervention which showed that weight loss effects continued. The paper is the first to show that monetary rewards can have lasting effects on body weight.[1] The results show that the effects remained quite stable over time and was mainly due to differences in the speed of weight (re-)gain between the treatment groups and the control group. The effects of the weight-maintenance rewards, in contrast, do not persist over time. This is an important finding, too. In fact, this paper is also the first of its kind to examine weight-maintenance rewards in a large-scale experimental study.

So did the amount of the reward matter?

This experiment does not reveal a statistically significant difference between the lower and higher rewards, despite remarkably better results from the higher weight loss reward.

And what does that mean for the design of future similar field experiments?

To determine the rewards, prior studies provided guidance and that’s how these particular amounts were determined. Considering the socio-economic background of participants, the amount did matter to them, but it may not be so for higher income levels. Women, for instance, who in the sample tend to be in employment less often than men, are affected more strongly by the higher reward. This does however not apply to men. Any future experiment would need to take into account the monthly wages of participants.

How would this research and the subsequent findings be operationalized into World Bank projects?

About 62 percent of obese people reside in developing countries. In fact, more people die from obesity than from being underweight in many of those developing countries. And 50 percent of the obese reside in China, India, Russia, Brazil, Mexico, Egypt, Pakistan and Indonesia (Ng et al., 2014). In both the developed and developing world, obesity has become a public health challenge, taxing and overburdening health systems.
As with most field experiments, the empirical results may not be externally valid. Hence, one should be careful in extrapolating these results to a wider population. Yet, through this paper and other subsequent studies and experiments that can be incorporated into World Bank projects, the hope is to contribute to the dearth of knowledge on effective interventions to improve public health in developing countries, especially with regards to obesity. The study also relates to the literature on conditional cash transfer programs and specifically ones covering nutrition supplement components even though these have primarily focused on fighting childhood undernutrition (e.g., Leroy et al., 2009).

Overall, the results of the experiment suggest that habit formation tops any potential adverse effects of monetary incentives – proving the age old adage that we are indeed creatures of habit, occasionally lured into successfully sustaining healthy habits with financial rewards.

[1] After correcting for differences in the rates of participants receiving weight-maintenance rewards across treatment groups of the first intervention (weight-loss rewards).

This article is published in collaboration with The World Bank. Publication does not imply endorsement of views by the World Economic Forum.

To keep up with the Agenda subscribe to our weekly newsletter.

Author: Arndt Reichert is an Economist in the Development Impact Evaluation (DIME) team within the Development Research Group at the World Bank and coordinator of DIME’s energy program. Anushka Thewarapperuma is an Operations Officer in the Development Impact Evaluation unit (DIME) of the World Bank’s Development Research Group (DECRG).

Image: A woman is seen jogging. REUTERS/Brian Snyder.