The concept of institutions and the related theoretical issues of their normativity and enforcement are still a matter of controversy among social scientists, including economists. The present article establishes a conceptual distinction between institutions, social norms, and decision-theoretic norms, arguing that some institutions may not be norms of any kind. Using the concepts developed previously, it then examines whether and how institutions have been treated as norms in institutional economics and in sociology.
Institutions, Social Norms, and Decision-Theoretic Norms: A Brief Conceptualization
The appropriate concept of institutions (1) depends on the purposes of the analysis. Bearing this in mind, one may broadly define institutions as socially shared patterns of behavior and/or thought.
Institutions that do not require the threat of external sanctions to exist are called here self-enforceable, in a strong sense. (2) The individual enforces the institutional pattern upon him or herself, with no need for sanctions external to that person. Self-enforceability does not necessarily mean that the threat of sanctions does not exist; it only means that this threat is not required for the existence of a given institution. When this threat does not exist, the institution is self-enforcing.
Social norms are social standards of behavior and/or thought that (a) indicate what people should or should not do or think under some circumstances and (b) are at least in part enforced upon individuals by external pressure. If a social norm exists, the threat of sanctions also exists. Moreover, the threat of sanctions helps maintain the norm. Some individuals may internalize the norm, in other words, they may come to believe that following it is the right thing to do, regardless of sanctions.
Legal norms are social norms enforced by the legal system. Informal social norms are enforced by the approval or disapproval of other people in the group or community. They may be moral or epistemic norms. For the individuals who internalize a norm, the latter is self-enforceable. Although following a social norm frees one from the risk of sanctions, a social norm is not necessarily followed in order to avoid sanctions or, more generally, for instrumental reasons (for an extreme version of this argument, see Elster 1989).
A convention is an institution with at least two additional characteristics that other institutions may not have: (a) when followed consciously, a convention is followed at least in part because other people are following it and not, or not only, because there is an external pressure to comply, and (b) it is to some degree arbitrary.
In the view defended here, not every informal institution (even excluding organizations) is an informal social norm. In other words, not every institution needs to be enforced by the threat of sanctions and/or internalized. Some conventions may be followed without the threat of sanctions and without any sense of moral obligation.
An institution may be said to be a decision-theoretic norm (i.e., to have a normative content in the decision-theoretic sense) if, even in the absence of an external pressure, it is always in the interest of an individual to follow that institution when he or she expects (nearly) everybody else to do the same. From this perspective, a deviant individual would harm his self-interest. When individuals know how to promote their self-interest (avoiding, for example, the kinds of mistakes discussed by behavioral economists), decision-theoretic norms are self-enforceable.
In the view defended here, not all conventions and consequently not all institutions (even excluding organizations) are decision-theoretic norms. Someone could agree with this while treating all conventions as social norms. Similarly, someone who treated all conventions as decision-theoretic norms could agree with the statement made earlier in this paper that not all conventions are social norms. Based on arguments developed in Dequech 2003, 2005b, 2006, I suggest, however, that some conventions are neither social norms nor decision-theoretic norms. If they are not social norms, they must be self-enforcing, but they may be so without being decision-theoretic norms. Examples include some conventions with which innovators and speculators break.
Institutions and Norms in Institutional Economics and Sociology
I shall now apply the concepts of social norm and decision-theoretic norm to examine whether institutions have been treated as one and/or the other kind of norm in institutional economics and in sociology. This question clearly applies to institutions other than collective agents (organizations).
There is no consensus among institutional economists regarding the conceptual relation between institutions and social norms. This is true even within the subset of "old" or "original" institutionalists. (3) Some of these institutionalists seem to treat all institutions as social norms. J. Fagg Foster (1981, 907-908), followed by Paul Bush (1987, 1076) and Marc Tool (1993, 122), characterized institutions as (sets of) "prescribed patterns of correlated behavior." For Walter Neale (1987, 1188), if people do not follow the rules, they are sanctioned. Yngve Ramstad (1989, 763) argued that institutions are the working rules and customs forming "the 'rule structure' within which individuals must confine their activities subject to sanctions." (4)
For Geoffrey Hodgson (1998, 179), in a weaker statement, "institutions incorporate values, and processes of normative evaluation. In particular, institutions reinforce their own moral legitimation: that which endures is often--rightly or wrongly--seen as morally just." Hodgson did not maintain that this is always necessarily the case. Furthermore, even if all institutions were value laden and legitimate, this would not make all of them social norms (Dequech 2005c). More recently, Hodgson (2006) has argued that institutions, as rules, are normative or "immanently normative" and, similarly, have "actual or potential normative content" (3, 5). He has also stated that in the case of some institutions "normative issues are often secondary to the straightforward incentives and disincentives involved" (2004, 433). (5) In either statement, the meaning of the term normative...