Do the right thing..... but for how long
Cécile Bazart  1@  
1 : Laboratoire Montpelliérain d'Économie Théorique et Appliquée  (LAMETA)  -  Website
Centre international de hautes études agronomiques méditerranéennes [CIHEAM], Institut national de la recherche agronomique (INRA) : UR1135, CNRS : UMR5474, Université Montpellier I
34960 Montpellier cedex 2 -  France


1 Corresponding author : LAMETA, université de Montpellier, Rue Raymond Dugrand Site Richter, CS79 606 34960 Montpellier cedex 2.

This work focuses on the effect of information on motivation to act pro-socially in the short and the long run.

There is a large set of literature dealing with motives to do the right thing. Part of it focuses on social motivation and social norms. Bicchieri (2006) showed that people have conditional preferences for doing the right thing, even under complete anonymity: either conditional on what the others do (the so-called descriptive norms) or conditional on what the others think is socially acceptable to do (normative norm). The underlying question is the following: can public/private organizations encourage behaviour change by using descriptive and normative norms? On the one hand, Bicchieri and Xiao (2009) state that when behaviour is not observable, only descriptive norms matters with 60% of dividers of a dictator game, in a previous session, sharing the amount equally or saying dividers should share equally. But on the other hand, some, like Raihani & McAulie (2014) defend normative norm effect on decision. In their experiment it is suggested that Player 1 give 40% (in a different condition 20%) of the endowment to Player 2. Still, they found the opposite. Again, normative norms do not show an effect on people's choices in a modified dictator game where dividers receive advice from a group of 3 subjects who are unaffected by the divider's decision.

In line with these researches we firstly question the effect of norms to motivate pro-social behaviours. We, secondly, question the time lag between the moments when information is displayed to subject and decision time and its impact on the efficiency of the measure. As a consequence we will be able to discuss the advantages of policies using communication on norms to guide decisions, their effect and the term of their efficiency.

To this purpose we built an experiment using a dictator game to avoid a strategic environment. The first treatment is a baseline organized in 2 parts. In a first part subjects play a one-shot dictator game, then played a 10 minutes distraction task. We used a slider task (40 sliders min out of 60). Then subjects started playing part 2 with 9 rounds of dictator games. Finally we implemented a norm elicitation block as in Krupka and Weber 2013. In this context we introduce in part 1 a descriptive and, respectively, a normative message. The Descriptive norm was implemented by displaying the following information to subjects: "previous experiments conducted under identical conditions show that most often players A give 5 EUROS" (treatment 2). In treatment 3 a normative information was displayed: "previous experiments conducted under identical conditions show that most often players A consider that giving 5 EUROS is socially acceptable". Again short and long run as repeated decisions to give were observed.

Our results show that on average over the 10 points of initial endowment subjects give on average: 3.8 in the baseline treatment; 3.4 in the descriptive treatment and 3.8 in the normative one. Over time (part 2) we find that the average amount is lower in the descriptive treatment than in the baseline treatment. Conversely we observed a higher level in amounts sent in the normative treatment by comparison to the baseline treatment. Still our econometrical analyses do not allow to conclude on any significant difference between the baseline, the descriptive and the normative treatments. But if we analyse compliance to the norm of giving 5 then the Normative norm seems to significantly increase the probability that player A would comply with the give 5 euros norm compared to the Baseline. Last we compare the initial design including the distractive task and a set of treatments excluding any distracting task. In all treatments we observe a higher level in amounts sent to receivers.


Bicchieri, C; 2006, "The Grammar of Society: The Nature and Dynamics of Social Norms" pp. xvi, 260, Cambridge and New York: Cambridge University Press.

Bicchieri C, Xiao E; 2009, "Do the right thing: but only if others do so", Journal of Behavioral Decision Making, 22, pp 191–208.

Raihani, N.J., McAuliffe, K. 2014, "Dictator Game giving: the importance of descriptive versus injunctive norms", PLoS ONE 9(12).

Schram A, Charness G, 2015, "Inducing Social Norms in Laboratory Allocation Choices", Management Science 61(7), pp. 1531–1546.


  • Other