If we start from the premise that institutions are socially constructed (Berger and Luckmann 1966) and that economic activity is socially instituted (Polanyi 1957) and a situated process (Granovetter 1992), we may argue that economic phenomena are, in one form or another, also institutional phenomena. The intertwined relationship between institutions and human (including economic) activity has captured the imaginations of economists, political scientists, and sociologists since the late nineteenth century. In economics, attempts to adopt the "scientific method" were challenged unsuccessfully by economists of the German historical school led by Gustav Schmoller in the early 1900s. Drawing on the ideas of Immanuel Kant and Georg Wilhelm Friedrich Hegel, the group asserted that simplistic assumptions about the "rational man" and economic equilibria were unfounded and the quest for a set of universal laws for economics fruitless. Following the lead of their German contemporaries some American institutionalists, notably Thorstein Veblen, were also suspicious of abstract universal principles, interested in solving practical problems, and aware of the role of events and historical contingencies (Scott 2001).
Early institutionalists pointed to pervasive market power and to indeterminacy even under perfect competition; the role of social institutions in shaping individual preferences (and hence the importance of institutions as the subject of economic analysis); the usefulness of pragmatic and psychologically realistic models of economic motivation (as opposed to utilitarianism); and the centrality of time and space in understanding the evolution of the economic system (Scott 2001). Both the "old" and "new" strands of institutionalism in economics emphasize the importance of institutions and draw attention to the "evolutionary" aspects of economic activity (Hodgson 1994a). The question for most institutionalists in economics has been "not how things stabilize themselves in a 'static state,' but how they endlessly grow and change" (1988, 130). There was significant interest in institutionalism in the interwar years, particularly in the United States. In the years after the First World War there was widespread recognition of the need for "improved economic data and policy analysis." There was also recognition of the potential role of government in the reconstruction of the economy (Rutherford 2001, 178). Institutionalists did much to improve the statistical work of government agencies and develop monetary and financial data, including work on money flows which later became the "flow of funds" accounts (179).
Recent years have witnessed a return to institutionalism in social scientific inquiries (Jessop 2000). In economics the renewed interest in institutions has resulted in three significant developments. First, "new institutional economics" has emerged as a reasonably coherent subdiscipline of mainstream economics, exemplified in works by Douglass North (1990) and Oliver Williamson (1985, 2000). Second, "old institutional economics" in the tradition of Thorstein Veblen (1857-1929) has received a significant boost from works by F. Gregory Hayden (1993, 1994), Geoffrey Hodgson (1988, 1993a, 1993b, 1999a, 1999b), Walter C. Neale (1987, 1994), Malcolm Rutherford (1994, 2001), Warren Samuels (1989), Richard Scott (2001), and numerous others. Third, interest in the institutionalist perspective has once again underlined the methodological weaknesses of the old institutionalist school and the narrowness of the methodological base of the new. The old institutionalists are generally commended for drawing attention to the complex and "instituted" nature of the economy but criticized for vagueness on how best to incorporate complexity into economic analysis. The new institutionalists may be praised for bringing institutions into economic analysis but criticized for remaining largely within the limited bounds of the neoclassical conceptual framework.
For institutionalists, key to understanding the processes of growth and change are the institutions of the economy, as well as individual preferences. But understanding institutions requires appreciation of complexity, continuity, and evolution in historical time. Conducting institutional analysis consistent with the German historical school and its American counterparts must go beyond arguing over the semantics of what comprises institutions. The task requires carefully organized categories of institutions that reveal the levels, scales, and systems around and through which institutions are woven and methods to operationalize the rich and diverse concepts developed by institutionalists. This paper starts from the premise that institutions define inter-relations at the social, organizational, and societal levels; that institutions are context specific and operate at, through, and across different territorial scales; and that institutions function and affect phenomena in the social, economic, and political systems. As with most terms, multiple meanings are associated with levels, scales, and systems. Among other tasks this paper attempts to offer working definitions for these terms to facilitate operationalizing key institutionalist concepts.
Levels, scales, and systems are of course socially constructed and for the most part arbitrary. There is not much benefit in squabbling over exactly where the social system ends and the cultural, or the political or economic, begins. Rather, the point is to recognize that the multiple facets of the whole that constitute the "real world" are reasonably distinguishable and thus can be more closely examined. By extension, scales of governance, for example, "the region" or "the nation," are socially constructed rather than ontologically pre-given (Brenner 1998, 460). The questions of where one would start to examine institutions and what the implications are in terms of policy, action, or research are really functions of one's geographical, social, economic, and political positioning.
This paper argues that institutional analysis is most effective when conducted in recognition of the levels of inter-relation, scales of governance, and integrity of systems. Fixing the problems of competitiveness and growth in the economic system by safeguarding the bottom line and "shareholder interests" will not do much for the sustainability of the total system comprising social, economic, and ecological systems. Taking action on social issues and persuading governments to part with concessions to remedy undesirable social situations will not necessarily fix the causes for many social issues, often created by failures in the economic system. Similarly, maintenance and advancement of ecological integrity cannot be realistically accomplished in isolation from the social and economic systems as what occurs in each of these systems often has quite significant implications for ecological integrity: ecosystems "contain" the economic and social systems.
This paper examines the multiple meanings associated with the term "institution," briefly reviews the old and new strands of institutionalism in economics, outlines three reasonably distinct categories of institution, and uses the categories as the foundation to define five types of institution to capture institutions as phenomena operating at multiple levels of inter-relation and scales of governance and through different systems. A conceptual framework is then outlined for studying technological transitions (1) as evident in two empirical cases: (1) the pulp and paper industry in Europe and (2) the Dutch waste arena (subsystem). The concluding section synthesizes the contributions of this paper and outlines the implications of further research and policy making centered on effecting change.
Multiple Roles and Meanings of "Institutions"
Much of the analysis claiming recognition of the importance of institutions in economic activity does little more than throw in institutions as an add-on "factor" or a "filter" to be accounted for in schematics representing causal flows in a given situation. A popular tendency is to view institutions as constraints that bear upon a "noninstitutional" economy or market, failing to see economies and markets as collections of institutions (Hodgson 1999a, 145). The main characteristic of the institution is the "permanency" or "invariability" relative to the individual as a unit of analysis (1988, 1999a). "The institution" is therefore more akin to spatial and temporal inquiry than "the individual" with a fixed set of preferences (Williamson 1994). The focus by some economists on "the institution" represents a significant departure from the standard rational choice theory of neoclassical economics. The new focus underlines the important roles of history and culture in determining economic actors' operative goals, values, and views of the choice context insofar as actions involve coordination with or will induce responses from other actors (Nelson 1994).
Institutions may be defined as the set of conventions and rules of action that prevail in the economy, are embedded in the local social structure, and show a marked regional differentiation (Kratke 1999, 683). Institutions are "proceduralist" rather than "consequentialist," influencing behavior that occurs in a particular situation independently of an individual's goal orientation (Elster 1989, cited in Setterfield 1993, 756). Veblen famously referred to institutions as "settled habits of thought common to the generality of men" (1919, 239). The evidence for an institution is "the regularities of people's actions and their responses to questions about what they are doing" (Neale 1994, 404). John R. Commons (1924) defined an institution as collective action exercised by different types of organizations-such as the family, the corporation, the trade union, and the state-in control of individual action. Wesley Mitchell described an...