4. Players of the Game: Rationality, Choice, and Indeterminacy

A full analysis of the implications of methodological problems in game theory for the kind of approach used by North, that is, as a means to provide a social contract analysis of the emergence of social institutions, would constitute an important contribution to the study of the NIE in general and North’s approach in particular. Although I summarized it in my PhD thesis (Krul 2016), a proper treatment requires such length and complexity that it cannot be justified for full inclusion in this book. For some applications of game theory to the NIE and related economic theories, see Schotter (1981), Greif (1993, 1994, 1998), and Aoki (2001). For a valuable critical treatment of game theory from a philosophy of economics standpoint, see Hargreaves-Heap and Varoufakis (1995, 2004). A recent general treatment is Binmore (2007).

Here, too, a full explanation would require a lengthy discussion of the different ways a game theoretical result can be interpreted, which I must forego on this occasion.

Contrast North 2005 with, for example. North and Weingast 1989, where incentives do all the work and beliefs play virtually no role in the argument.

The implications of this use of intentionality for his evolutionary argument are discussed in the next chapter.

For North on the influence of Locke, see: Letter to David Nichols, May 23, 1991. For further discussion, see: Hough and Grier (2015: 50).

This is the so-called trembling hand. The assumption that a ‘trembling hand perfect equilibrium’ must exist, put simply that a plausible strategic interpretation must be robust to the possibility of error, is an important qualification of the rational choice informational framework of game theory (Selten 1975).

This again brings to mind game theory: the significance of repeated games, in which choices can signal trust and in which strategies of retaliation are possible, thereby enabling mutually beneficial equilibria in certain cases where one-shot games do not permit them.

See also the discussion of rationality and explanatory value in the NIE in Dequech (2006).

This approach is very much in line with the earliest divide between neoclassical and New Institutionalist approaches. Mäki argues that Ronald Coase and Milton Friedman’s early work can already be contrasted in this respect, in that Coase argued for realisticness of assumptions, whereas Friedman (not very coherently) argued against it (Mäki 1992, 1998). Of course, to what extent Friedman really was neoclassical can be disputed, but his famous essay on the methodology of economics is often seen as defining for the dominant postwar approach (Friedman 1953).

As Langlois and Csontos (1993) make clear, much of the argument as to whether neoclassical rational choice approaches and ‘behavioural’ approaches based (in part) on psychological data are really in conflict depends on what demands are placed on assumptions in economic theory—that is, whether the assumptions themselves should be empirical or not. While they argue against this demand, it appears in any case North supported it.

Strictly speaking, there can be uncertainty about this. One can have different types of agents all of whom have different preferences (and therefore different payoff distributions). But as long as the probabilities of encountering a given type of agent are known, this has the same effect as knowing the payoff distribution directly (Harsanyi 1967).

See Aumann (1976, 1987). This argument is not uncontroversial, however.

The following account draws extensively on Hargreaves-Heap and Varoufakis (1995).

For a thorough discussion beyond this off-hand comment, see Skyrms (1996, 2003) and some of the discussion in the following chapter.