The nature of human behaviour assumed in most economic models is based on the ‘Homo Economicus’ assumption – everyone is out for him or her self. This makes economic theory easy, or at least easier, since it minimises the joint-ness of decision making.
The 'Homo Economicus' assumption, however, fails rather consistently in laboratory experiments. Notions of trust and public-minded-ness turn out to be much more common in human behaviour than they are in the theory. Economic outcomes such as growth, economic development, international trade, and labour market behaviour have been shown to depend upon such values.1 Recent advances in experimental economics are developing a set of results that should help economic theorists develop models that are better able to explain real-world economic behaviour in a variety of settings.
In CEPR Policy Insight No. 9, we describe an experiment that tests whether public-minded-ness, or pro-sociality are transmitted from parents to children.
The experiment is based on a rather straightforward public good game. Participants gets some tokens and they decide how many to keep and how many to put in the ‘pot’. Whatever is in the pot after each participant has made his/her choice (without knowing yet what the others have chosen) is doubled and divided evenly among the participants. Obviously, the participants would be best off if they could all commit to putting all their tokens in the pot – everyone would double his/her money. But each participant has an incentive to put in nothing.: indeed, this is the best action whatever the others decide.
As usual in this type of games, in our experiment we observe more pro-sociality than simple game theory predicts. The interesting part of the experiment, however, comes when we compare the behaviour of children with that of their parents. The surprising result is that there is not correlation between the number of tokens contributed by parents and their children. The results of a regression of children’s contributions on parents’ contributions show a coefficient very close to zero both in terms of absolute magnitude and in terms of statistic significance. The result survives to a host of robustness checks..
We did, however, find some systematic patterns that suggest some value-transmission channels. In particular,
- boys are less cooperative than girls;
- older children are less cooperative than younger ones; and
- above all, children from larger families tended to contribute significantly less than children from smaller ones.
This last result, together with the absence of correlation between parents’ and children’s behaviour appear to be consistent with part of the psychology literature, which emphasizes the importance of peer effects rather than family influences in the socialization process.
More work to be done
The consistent finding that models based on a notion of self-interest are rejected in the laboratory indicates to the economic theorist the need to develop new equilibrium concepts that can explain the observed behaviours. A dialectic between experimental economics and game theory could be very fruitful in developing models that better explain economic outcomes.
1 See Guiso, Sapienza and Zingales (2006) for a review on the importance of culture on economic outcomes.