By Sebastian Galiani*
No doubt, we economists believe that incentives matter. Recently, it has been argued that incentives could be used to encourage desirable habits, such as getting regular exercise or having a healthy diet. This has become particularly important in Latin America given its demographic and epidemiological situation (Galiani and Weinschelbaum, 2014).
If the welfare created by consuming or undertaking some activity today depends on that same level in the past, or in more technical terms, if the marginal utility of consumption today is positively correlated with past consumption, then providing monetary incentives to undertake an activity could lead to us to do that same activity in the future, thus changing behavior.
This hypothesis is known as “habit forming.” However, it has also been noted that granting of rewards can be counterproductive, as it provides an extrinsic motivation for a task or activity that can displace the intrinsic motivation.
This hypothesis is known as “crowding-out”. This effect could be particularly important in the medium term, when economic incentives are withdrawn and the individual may have lost her initial motivation for the activity in question.
Charness and Gneezy (2009) (hereinafter C&G) exploit an experimental design to evaluate the impact of providing financial incentives in forming the habit of going to the gym. To do this, the authors compare the behavior of three groups of college students.
On a first instance, 120 students received a booklet with information about the benefits associated with physical activity. Then, they were randomly assigned to a control group, which did not have any additional requirements, and two possible treatment groups: individuals belonging to “Group 1” were offered a payment of $25 to attend the College gym once.
“Group 2” individuals were offered the same, ie a payment of $25 for attending a gym once, but also were offered an additional payment of $100 for attending the gym eight times during the next four weeks.
The results of this experiment show that the control group slightly decreased their gym attendance from 0.59 visits per week for the eight weeks prior to the intervention to 0.56 visits per week for the seven weeks after the completion of the intervention.
The result, however, is different in the treated groups. Group 1 increased from 0.70 visits per week to 0.76 visits per week after the intervention, while group 2 increased from 0.6 visits per week to 1.24 visits per week, ie attention to the gym doubled in those individuals who received additional monetary incentive.
Additionally, the authors find that the effect comes from those who did not regularly attend the gym in the pre-intervention period. However, it is not surprising that the incentives generated a strong effect during the treatment period, but does this effect persist over time? Yes, at least seven weeks after the intervention.
Acland and Levy (2013) examine the validity of the hypothesis of habit formation over a longer period. Specifically, these authors recruited 120 students who reported not going to the gym regularly and replicated the C&G experiment, using a time horizon spanning 37 weeks prior to the experiment to 33 weeks after completion.
In this case, students were randomly assigned to a control and a treatment group. All participants of the experiment were offered $25 to attend the college gym once during the week (control group that is equivalent to Group 1 in the C&G experiment).
This initial offering, called “Learning Week”, intended to encourage individuals to overcome any fixed cost associated with learning how to use the gym. In turn, the treatment group received an additional offer of $100 to attend the gym twice a week in the four weeks following the “Learning Week” (equivalent to group 2 of the C&G experiment).
The results obtained in this study show that in the 8 weeks after completion of treatment (immediate post-treatment) attendance to the gym increased by 0.25 weekly visits. However, after 4 weeks of winter break and the following 21 weeks (distant post-treatment), no significant difference was found in the assistance to the gym between the control group and the treatment group, which suggests that the effect of habit formation remained only in the short term.
In conclusion, the available evidence does not allow us to state that the use of financial incentives can change habits in the long run. This is an important issue in Latin America, not only in relation to habits that affect population health, but also on other behaviors that seek to be affected by the extensive social support network developed in the region. More experimental research in this area is needed.
* Sebastian Galiani is a Professor of Economics at the University of Maryland and Visiting Professor at the Universidad de San Andres, Argentina
Leave a Reply