Theory of Incentivizing: A Double-Edged Sword

Incentives are a vital element of economics. As such, their usage is widespread in many areas, such as education, health, and pro-social behaviour. The seminal importance of incentives arises from the fact that they are the mechanisms by which the individuals’ interests are aligned with the public interest. This is primarily achieved through giving monetary incentives that increases (decreases) the benefit an individual obtains from a socially beneficial (detrimental) activity e.g. providing financial incentives to individuals who use renewable energy. Accordingly, as provision of monetary incentives recalibrates the relationship between costs and benefits, one would expect it to work effectively. However, a quick glance into evidence provided by behavioural economics suggests that it rarely does, and in many cases, it even backfires.

In the study of Mellström and Johannesson (2008), authors tested the effectiveness of monetary incentives on blood donations, a great example of pro-social activity. To achieve this purpose, the authors conducted randomized field experiments in blood donation centres located in Sweden. The three treatment groups were offered different incentives for blood donation: no payment, $7 cash payment, and cash payment of $7 with charity option. Traditional economics predicts the willingness to donate to be the lowest for the first and highest for the second group, since in addition to the intrinsic reward of helping someone else, the second group receives further monetary reward, rendering the highest utility. Contrary to what is expected, the willingness to donate has actually decreased by 73.3% in the second group, and the charity-incentivised group’s willingness to donate blood was highest at 54%, followed by 47% for the non-incentivised group.

Explaining the Phenomenon: Image Motivation, Overjustification Effect & Decision Frames

The key insight for understanding this seemingly idiosyncratic phenomenon is to understand the fact that, in line with the tenets of sociology, our behaviours act as a signalling mechanism of our motives. While the price effect of incentives, i.e. decreasing cost or increasing benefits, can increase the take-up of an activity, this signalling effect may act as a barrier, and it is the relationship between these factors that determine the outcome for the incentivised economic activity. Reverting back to the blood donation example, regarding the monetary reward group, the presence of extrinsic monetary reward tainted the intrinsic image motivation of signalling the society that the individual is doing good for the sake of doing good. Accordingly, the addition of monetary incentive distorted the signal effect of this prosocial activity and made it unclear whether the activity is undertaken ‘to do good’ or ‘to do well’. Therefore, as the cost of sending the wrong signal to the society surpassed, i.e. overjustified, the additional monetary benefits, a decrease in the willingness to donate ensued. This occurrence is illustrated in the graph below (Figure 1), whereby the inclusion of incentive y decreases the aggregate supply of an activity to the point at which it is lower than it would have been without providing any incentives (the top polynomial curve).

Another important, yet subtle inference is that if the signalling effect is non-present, then the monetary incentives will work as predicted by mainstream economics. The conducted experiment of Ariely, Bracha, and Meier (2009) tested this hypothesis through providing monetary incentives to donators privately. In doing so, they have directly tested whether the overjustification effect that is widespread in social contexts can be overcome. The results reveal that monetary incentives do work in private contexts for prosocial activities, yielding a linear relationship between incentives and activity uptake, similar to the one depicted in the graph. This implies that the change in the framing of the decision-making context, i.e. from social to private, was the driving factor behind behavioral change.

A very important conclusion arises at this exact point: “The framing of the decision-making situation has a crucial impact on the activity uptake”.

Implications for Public Policy and NGOs

The inferences do not imply a doom scenario whereby it is virtually impossible to incentivise uptake of prosocial activities. Instead, they imply that careful consideration, planning, and execution can provide the desired outcome. Demonstrated through the blood donation example, the effect of incentives depends on their design, the form in which they are given, and the behavioural context (social/private) in which the activity occurs. The statistically significant increase in willingness to donate of the charity-incentivised group implies that the form of incentive should be in parallel with the form of intrinsic motivation, which were both social and altruistic in the blood donation example. In addition, the framing of decision-making context should be a function of the selected incentive form. Accordingly, the framing should be social if the form is altruistic, and private if the form is monetary to increase or reduce, respectively, the effect of incentives on image motivation.