What Institutions Are

In everyday usage the word institution is loose. It can name a specific organization — 'a venerable institution like Harvard' — or an abstract practice — 'the institution of marriage.' Sociology uses the term in a more precise sense: institutions are the relatively stable patterns of rules, roles, and expectations that organize social activity around a core social function. The core functions are a short list: biological reproduction and the socialization of children (family); transmission of knowledge and skill (education); production and distribution of goods (economy); collectively binding decision-making (polity); meaning-making and the management of ultimate concerns (religion); and the enforcement of norms and resolution of disputes (law). A society without stable arrangements for these functions is not possible; a society is in part identified by the specific shapes its arrangements take.

The classical analytical treatment is Berger and Luckmann's The Social Construction of Reality (1966). Their account begins with habitualization: any action repeated often enough that it becomes a pattern available for reproduction with economy of effort. Habitualization becomes institutionalization when the habitualized actions are reciprocally typified — when an actor recognizes that others perform actions of the same type and that the types are available to anyone positioned appropriately. The classic example is the morning handshake: once two people establish it as their greeting, a third person entering the scene recognizes it as the local rule, and new members are inducted into what is now a local institution (Berger and Luckmann 1966, pp. 53-58). Institutions, in this view, are not abstract normative frameworks imposed from outside; they are the sedimented residue of repeated interaction that has acquired an objective quality — a thereness — for those who encounter it.

A key distinction, sometimes blurred in introductory texts, separates organizations from institutions. Harvard, Toyota, and the US Department of Education are organizations: specific, bounded, legally constituted entities with addresses and budgets. Higher education, the corporation, and public schooling are institutions: the rule-governed patterns within which organizations like these operate. The institution of higher education predates any particular university and will outlast it; the rules governing what a university is, what counts as a degree, how faculty are hired, and how research is legitimated are the institutional facts against which any specific organization is evaluated. When analysts speak of 'institutional change,' they typically mean change at this pattern level rather than change within any single organization (North 1990, pp. 3-5).

The foundational sociological treatment of institutional differentiation is Weber's Economy and Society (1922/1978). Weber traced the historical process by which originally fused spheres — economy, polity, religion, law — became analytically and practically distinct in the modern West, each with its own characteristic logic, its own typical actors, and its own forms of legitimate authority (Weber 1922/1978, vol. 1, pp. 54-56). The economy, on Weber's account, came to operate by a logic of calculative rationality enforced through markets and contracts; law by a logic of formal rationality administered through specialized courts; religion by a logic of meaning-making increasingly confined to private salvation rather than public coordination. This differentiation thesis — that modernity is characterized by the separation and specialization of institutional orders — has remained central to sociology, debated and qualified but not displaced (Alexander 1990, pp. 1-8).

Two further features of institutions matter for the analysis that follows. First, institutions are normative as well as regulative: they do not merely restrict action but supply the vocabularies and expectations through which actors understand what they are doing. A marriage is not only a legal contract; it is a culturally thick understanding of what two people are to each other. Second, institutions are carriers of power: the rules and roles are never neutral but always distribute access, authority, and resources in particular ways, with particular beneficiaries and particular excluded parties (Bourdieu 1977, pp. 78-82). Institutional analysis therefore cannot be separated from the analysis of inequality, a point that recurs throughout the following pages.