Strategic thinking – the intrinsic ability to anticipate actions and act accordingly – is a cornerstone of rational decision-making. It is required to predict and internalise future choices in inter-temporal decisions, and to best respond to anticipated moves of others in games of strategy. This ability is of paramount importance in our day-to-day lives. It guides us through our education and career choices and helps us avoid being manipulated by others or suffering abusive relationships. Strategic thinking is not only important to make adult decisions, but it is also critical in the day-to-day decisions of children and adolescents. How to avoid risky options? How to react to an abusive ‘friend’ or a bully? These are all too common questions that children and adolescents have problems addressing. A natural explanation is that young brains are not developed and that thinking abilities required to strategise are not yet in place. However, little is known about their development.
From a conceptual perspective, acting strategically requires people to put themselves in the shoes of others, an ability loosely referred to as ‘theory of mind’, and to think logically about their own as well others’ courses of action. Developmental psychology literature demonstrates that very young children are self-centred and unable to take the perspective of others. At around five years of age, their theory of mind ability starts to develop (Perner 1991). Children become less self-centred and start adapting their behaviour to norms and rules in their environment. They move from a situation in which they neither infer nor care about what others think to a situation in which they attribute beliefs to others and empathise with them. As for logical thinking, its development occurs in stages (Piaget 1972). Children develop the ability to think logically about what they observe (inductive logic) between the ages of eight and twelve (Feeney and Heit 2007) and they start developing the ability to reason abstractly (hypothetical and counterfactual thinking) around twelve years of age (Piaget 1972, Rafetseder et al. 2013).
Recent game theoretic studies in children and adolescents have shown that children are able to strategise to some extent or after a certain age, but the results seem to be largely dependent on the task at hand (Sher et al. 2014, Blake et al. 2015, Sutter 2007, Fehr et al. 2008). The objective of our research is to study how the ability to take on a perspective and the ability to think logically contribute to strategic thinking as a function of features of the task or context, and how the development of those abilities maps onto the observed changes in behaviour until adulthood. We first took that line of inquiry to preschool classrooms by asking children to complete a series of tasks varying in three dimensions (Brocas and Carrillo 2017):
- Context: Some tasks were individual decision-making tasks in which subjects were required to anticipate their own future moves while others were games against the experimenter requiring the anticipation of own and opponent future moves.
- Complexity: Some tasks required few steps of anticipation while others required more steps.
- Logical thinking: Some tasks required basic inductive logic while others required to think ahead and to apply backward induction reasoning. The study revealed that most participants in our sample were able to perform inductive logic but fewer and fewer were able to think ahead as the number of steps to anticipate increased. Participants were also more capable of taking their own perspective than that of the experimenter.
The fact that performance in our tasks was inversely related to difficulty suggests that learning to strategise is part of the overall development of executive functions. This development is gradual, allowing children to perform some simple aspects of strategic thinking early in life and to later extend this ability to more complex situations. As we grow, we acquire more sophisticated logical thinking abilities, helping to solve more sophisticated problems of strategic thinking. We investigated this hypothesis by asking children, adolescents, and adults to play an alternating dictator game (Brocas et al. 2017). Each dictator game featured the choice between a fair split (4,4) and a selfish split (6,1) between oneself and a partner. Controlling for altruism, defined as the willingness to sacrifice own payoff to benefit others and measured with another task, we focused on developmental changes in strategic giving, defined as the willingness to forego a current payoff as a means to encourage a mutually profitable long-term relationship. Strategic giving relies on abstract and counterfactual thinking to assess the future cooperative gains of such a relationship. Our objectives were to determine at what age participants were able to manifest this ability and what their choices were looking like until that point.
The evolution of altruism and strategic thinking
We observed first that altruism was hump-shaped. It increased monotonically with age for school-age children, and decreased for older participants. We then studied how subjects reacted to the choices of their partners. We found a significant increase with age in the probability of reciprocating – choosing (4,4) as a response to (4,4) by the partner – and a stable and low probability of forgiving – choosing (4,4) as a response to (6,1) by the partner. Finally, older school-age subjects who chose (6,1) early in the game could be induced to switch to (4,4) in the next round if the partner chose (4,4). This reversal did not occur for youngest participants or for adults. We last looked for evidence of the ability to anticipate future choices. In principle, a participant who understands the strategic benefits of cooperation should choose (4,4) in the first round to initiate such cooperation. This should be true in particular for non-altruistic players who are able to think ahead. We found a sustained increase across age groups in the probability of choosing (4,4) in the first round among non-altruistic subjects, suggesting that as we age, we become less motivated by other regarding concerns, but more capable of giving strategically.
Overall, the changes in behaviour observed over age resulted from the combination of the evolution of altruism and the evolution of strategic thinking. Young children were typically selfish and myopic. Older children were increasingly more altruistic but also more strategic. The first level of strategic thinking consisted of adapting to the choices of their partner, which relies on inductive reasoning. Consistent with the existing developmental literature, children in our sample were developing this ability gradually from age eight. The second level of strategic thinking consisted of anticipating the strategic gains of efficient cooperation. This necessitates the ability to reason abstractly (hypothetical and counterfactual thinking), which is known to start developing at around twelve years of age. Interestingly, the age at which we noticed an improvement of strategic giving corresponds closely to that time period.
The results taken together reveal that strategic behaviour is multifaceted and depends on a network of interacting abilities that develop gradually. Performing well in strategic environments requires one to have acquired different types of logical reasoning abilities and to be able to apply them in specific contexts. Some environments necessitate basic logical arguments while others necessitate abstract theories. Understanding how the development of these underlying abilities impacts the development of strategic thinking is of paramount importance to assessing how children and adolescents react in their own environments. A child who is not yet able to think abstractly takes its environment at face value. He does not anticipate possible futures and he is prone to commit to relationships he should avoid instead. This child can learn only through exposure, but exposure to the consequences of wrong choices may be irreversible. At the same time, this child may not be convinced to act otherwise because he is not yet ready to think otherwise. Studying the underlying abilities necessary to perform strategic thinking will help us understand not only why children and teens make the decisions they make but also why it is so difficult for them to avoid.
Blake, P R, D G Rand, D Tingley, and F Warneken (2015), “The shadow of the future promotes cooperation in a repeated prisoner’s dilemma for children”, Scientific reports 5.
Brocas, I, and J Carrillo (2017), “The Determinants of Strategic Thinking in Preschool Children”, CEPR Discussion Paper 12253.
Brocas, I, J Carrillo and N Kodaverdian (2017), “Altruism and strategic giving in children and adolescents”, CEPR Discussion Paper 12288.
Feeney, A, and E Heit (2007), Inductive reasoning: Experimental, developmental, and computational approaches, Cambridge University Press.
Fehr, E, H Bernhard and B Rockenbach (2008), “Egalitarianism in young children”, Nature 454, 1079–1083.
Perner, J (1991), Learning, development, and conceptual change Understanding the representational mind, Cambridge, MA: The MIT Press.
Piaget, J (1972), “Intellectual evolution from adolescence to adulthood”, Human development, 15 (1),1–12.
Rafetseder, E, M Schwitalla and J Perner (2013), “Counterfactual reasoning: From childhood to adulthood”, Journal of experimental child psychology, 114 (3), 389–404.
Sher, I, M Koenig, and A Rustichini (2014), “Children’s strategic theory of mind”, Proceedings of the National Academy of Sciences, 111, 13307–13312.
Sutter, M (2007), “Outcomes versus intentions: On the nature of fair behavior and its development with age”, Journal of Economic Psychology, 28, 69–78.