■

• Engel, C. (2007). 'Institutions for Intuitive Man', MPI Collective Goods Preprint no. 2007/12.

By its critics, the rational choice model is routinely accused of being unrealistic. One key objection has it that, for all nontrivial problems, calculating the best response is cognitively way too taxing, given the severe cognitive limitations of the human mind. If one confines the analysis to consciously controlled decision-making, this criticism is certainly warranted. But it ignores a second mental apparatus. Unlike conscious deliberation, this apparatus does not work serially but in parallel. It handles huge amounts of information in almost no time. It only is not consciously accessible. Only the end result is propelled back to consciousness as an intuition.

It is too early to decide whether the rational choice model is ultimately even descriptively correct. But at any rate institutional analysts and institutional designers are well advised to take this powerful mechanisms seriously. In appropriate contexts, institutions should see to it that decision-makers trust their intuitions. This frequently creates a dilemma. For better performance is often not the only goal pursued by institutional intervention. Accountability, predictability and regulability are also desired. Sometimes, clever interventions are able to get them both. Arguably, the obligation to write an explicit set of reasons for a court decision is a case in point. The judge is not obliged to report the mental processes by which she has taken her decision. Justification is only ex post control.

Intuitive decision-making is even more desirable if the underlying social problem is excessively complex (NP hard, to be specific), or ill-defined. Sometimes, it is enough for society to give room for intuitive decision-making. For instance, in simple social dilemmas, a combination of cheater detection and punishing sentiments does the trick. However, intuition can be misled. For instance, punishing sentiments are triggered by a hurt sense of fairness. Now in more complex social dilemmas, there are competing fairness norms, and people intuitively choose with a self-serving bias. In such contexts, institutions must step in so that clashing intuitions do not lead to social unrest.

• Engel, C. (2004). 'Social Dilemmas, Revisited from a heuristics perspective', MPI Collective Goods Preprint no. 2004/4.

The standard tool for analysing social dilemmas is game theory. They are reconstructed as prisoner dilemma games. This is helpful for understanding the incentive structure. Yet this analysis is based on the classic homo oeconomicus assumptions. In many real world dilemma situations, these assumptions are misleading. A case in point is the contribution of households to climate change. Decisions about using cars instead of public transport, or about extensive air conditioning, are typically not based on ad hoc calculation. Rather, individuals rely on situational heuristics for the purpose. This paper does two things: it offers a model of heuristics, in the interest of making behaviour that is guided by heuristics comparable to behaviour based on rational reasoning. Based on this model, the paper determines the implications for the definition of social dilemmas. In some contexts, the social dilemma vanishes. In other contexts, it must be understood, and hence solved, in substantially different ways.