Note: I wrote this essay with Ursula Athenstaedt & Tony Evans.
Let the children have their way.
Let the children play.
Let the children play.
~ Carlos Santana
Trust is important, but it is also dangerous.
From time to time, the field of social psychology has to defend itself against the charge that its findings are obvious and that therefore the research that goes into finding those findings is a waste. Much of what we consider obvious now, however, was not obvious before study, which people tend to forget, thus falling prey to the hindsight bias, which itself has seemed more obvious since Baruch Fischhoff’s (1975) pioneering study. That the earth is round may seem obvious now, but it took a battle between science and the Vatican to make it so.
Now consider a simple question: Does trust in strangers increase or decrease with development during childhood? The answer is: it increases. Obviously, children naturally distrust strangers when they are little, but eventually learn that many (though not all) people are trustworthy. No. Wait! The real answer is that children become gradually more distrustful. At a young age, they do not know the meaning of betrayal, and many have not experienced it. As they grow older and learn that many (though not all) people are basically self-interested, children begin to comprehend the risk of being exploited, and they trust less.
Which view is correct? A plausible case can be made for either one. An educated opinion must be evidence-based, and empirically reliable and replicable evidence can only come from controlled studies.
With a team of graduate students, we (Evans, Athenstaedt & Krueger, 2013) conducted three studies over as many years in kindergartens and elementary schools in and around the city of Graz, Austria. The first 2 of these studies are described in the article, the third is summarized exclusively in this post. To study trust in children, we adapted the so-called Trust Game from behavioral economics. In a game with grown-ups, one player has an endowment of $10 and can choose between keeping it and giving it to the second player, in which case the amount is tripled. The second player then chooses between keeping all the money and splitting it with the first player. The socially desirable outcome is that the first player trusts the second player to share the gains and that the second player in fact does so. In grown-ups, rates of trust and reciprocity are highly variable and intermediate on average.
We created a trust game for children by using bags with ‘surprise toys.’ All children played the role of the first player; the second player was shown as a photograph. The voice of the data was loud and clear. [drumroll] Trust is more common in older (over 60%) than in younger children (about 25%). Now that you know, feel free to claim that this was obvious all along.
Consider a second question: are there age-related differences in simple altruism? If so, could it be that older kids are more willing to give their gift bags to others simply because they want to be nice? Using a method described by Cox (2004), we modified the trust game and told kids were that the recipients of their largesse could not give anything back even if they wanted to. We found that rates of giving were quite low (~20%) in all three studies and that the age effect was small. All the rates for trust and altruism are displayed in the figure (you may have to zoom in to see the print). We also found that trust and altruism were only weakly correlated (~.14, on average), suggesting that trust is not simply an extension of a benevolent social preference. The decision to trust, we believe, is a rather demanding cognitive task, which requires the taking of someone else's perspective and figuring out whether that other person will honor the norm of reciprocity or selfishly betray the trustor. Perspective-taking is a skill subject to maturation. Small children appear to stick with the default of not trusting strangers.
After having shown that trust does not reduce to altruism, we ruled out two other explanations of the basic age effect. First, we found that what really matters is the age of the trusting child, not the age of the trusted child. Varying the age of the recipient had no effect. Second, we found that young children’s limited ability to delay gratification cannot explain their low level of trust. The age effect of trust was strong when the children learned immediately whether their trust was rewarded (with interest). In the third and unpublished study, we found that the gender of the trusted child did not affect the results. The final goal of that study was to rule out a potential order effect. In the first 2 studies, children always played the altruism game after playing the trust game. In the third study, we used both sequences and found that it made no difference.
The exercise of ruling out various confounds and theoretical alternatives strengthened our belief in the uniqueness and theoretical integrity of the concept of trust. Psychologists and economists have recently come to agree on one thing: efficient market behavior and personal well-being very much depends on a healthy approach to trust (see, for example, the blog of my colleague Louis Putterman). No one can thrive when trusting everyone; and no one can thrive when trusting no one. Being able to tell when trust is worth the risk is a tricky challenge and a critical developmental task. Our Austrian participants taught us that the developmental trajectory points in a hopeful direction.
Note i. If the epigraphic lyrics of classic Latin rock seem a bit far-fetched, well, we trust that you will forgive us.
Note ii. If trust were norm-driven like trustworthiness (Bicchieri, Xiao & Muldoon, 2011), it would be easier to learn. Yet, in his excellent introduction to economics, Partha Dasgupta (2007), referring to Albert Hirschmann, asserts that “trust is a moral good, in that it grows with use but decays with disuse, which means that we don’t need to ‘economize’ on trust, in the way we need to with ‘bread and butter goods’ like bread and butter. Trust shares this feature with skills: the more one practices a skill, the better one gets at it” (p. 68, italics his).
Cox, J. C. (2004). How to identify trust and reciprocity. Games and Economic Behavior, 46, 260–281. doi.org/10.1016/S0899-8256(03)00119-2.
Evans, A. M., Athenstaedt, U., & Krueger, J. I. (2013). The development of trust and altruism during childhood. Journal of Economic Psychology, 36, 82-95. doi.org/10.1016/j.joep.2013.02.010,
Dasgupta, P. (2007). Economics: a very short introduction. Oxford, UK: Oxford University Press.
Fischhoff, B. (1975). Hindsight ≠ foresight: the effect of outcome knowledge on judgment under uncertainty. Journal of Experimental Psychology: Human Perception and Performance, 1, 288–99. doi: 10.1037/0096-1522.214.171.1248