British Politics, 2007, 2, (91–99) r 2007 Palgrave Macmillan Ltd 1746-918x/07 $30.00 www.palgrave-journals.com/bp

Debate

An Alternative Institutional Theory to Path Dependence: Evaluating the Greener Model Fiona Ross Department of Politics, University of Bristol, Bristol BS8 1TU, UK E-mail: fi[email protected]

The popularity of path dependence theory is based on a swell of studies that have embraced a loose set of premises, many of which are unfalsifiable. Although scholars have devised some innovative means of broadening the theory, they have yet to develop a flexible model that is also systematic and theoretically elegant. This paper argues that layering morphogenetic social theory on top of a loose version of path dependence cannot resolve a number of critical difficulties that impair the explanatory and descriptive utility of non-economic specifications. However, morphogenetic social theory does promise an alternative and improved lens through which we can understand institutional development in the absence of path dependence theory. British Politics (2007) 2, 91–99. doi:10.1057/palgrave.bp.4200043 Keywords: path dependence; morphogenetic social theory; institutions; NHS

Introduction In his ambitious article, ‘Path Dependence, Realism and the NHS’, Ian Greener seeks to synthesize a broad understanding of path dependence theory with morphogenetic social theory, a perspective imported from the field of sociology and most closely associated with the work of Margaret Archer. Greener’s mission is to use morphogenetic theory to render non-economic conceptions of path dependence more systematic and theoretically satisfying. Unlike path dependence, morphogenetic social theory is not about the returns generated by the institutions themselves or the institutional properties that feature prominently in institutionalist analysis. Rather, it focuses on how dominant interest groups interact within inherited structural and cultural situational logics (based on a combination of pre-existing interests and ideas) and in turn impart their own legacy on these conditions. To evaluate whether loose formulations of path dependence can be theoretically enriched through morphogenetic social theory, we need to answer three questions: (a) why is path dependence, when rigorously

Fiona Ross An Alternative Institutional Theory to Path Dependence

92

(and appropriately) specified, so unattractive, and why is broadening it theoretically problematic? (b) can morphogenetic theory resolve the core problems with path dependence, or are we offered a flawed synthesis with little analytical purchase? and (c) can morphogenetic theory serve as an alternative and improved lens through which we can understand institutional development in the absence of path dependence theory? In the first section of this article, I elaborate on a number of difficult problems with path dependence theory that lie at the centre of Greener’s attempt to salvage it. Following this discussion, the paper analyses the fusion of morphogenetic theory with broad formulations of path dependence. The third section discusses the value of morphogenetic theory for institutional analysis in the absence of path dependence theory. The conclusion argues that morphogenetic social theory promises to offer an interesting alternative model of institutional development that may prove to be better suited to the political realm than path dependence.

Some Problems with Path Dependence Path dependence theory has been subjected to far-reaching critiques in economics (Liebowitz and Margolis, 1990, 1995; Arrow, 2004) and politics (Schwartz, 2002; Peters et al., 2005). Most charges levelled concern the weak empirical applicability of its assumptions, notably the lack of evidence for increasing returns and the inability of the theory to account for change effectively. Both shortcomings have led scholars to change radically the core assumptions of path dependence theory, in turn incurring new theoretical and methodological challenges. In economics and politics, scholars have found increasing returns arguments unconvincing, with Liebowitz and Margolis (1995) and Arrow (2004) rejecting the plausibility of the concept. A strengthening of institutions across time implies that their efficiency is heavily insulated from contextual developments, an assumption that is inconsistent with a vast body of empirical research. Much theory and evidence in political science, as Schwartz (2002, 4) points out, suggests that decreasing returns commonly characterize institutions and that lock-out may be a more pervasive feature than lock-in (but see Deeg, 2001). Indeed in public policy, where ideas, needs and client groups change, it would be truly surprising if institutional development was marked by increasing returns. It would be remarkable, for example, if the 60-year-old NHS, designed in another era with different needs and expectations, did not generate decreasing returns. Even the indicators of institutional returns have changed across the Service’s lifespan. As Klein (2001, 113) reminds us: ‘The NHS was born into a working class society only slowly emerging from war, British Politics 2007 2

Fiona Ross An Alternative Institutional Theory to Path Dependence

93

where rationing and queuing were symbols not of inadequacy but of fairness in the distribution of scarce resources’. In seeking to radically change the Service, particularly after the 2001 election, New Labour’s explicit goal was to prevent further decreasing returns from undermining the NHS in its entirety. Alan Milburn, the Secretary of State for Health between 1999 and 2003, claimed that his fear was of complete lock-out of a publicly funded system if the government failed to address the plethora of problems afflicting the NHS. The second dominant charge levelled against path dependence theory concerns its weak ability to account for change (Crouch and Farrell, 2002), particularly endogenously generated change (Deeg, 2001; Peters et al., 2005). Despite being imported from economics into politics in an attempt to infuse historical institutionalism with a source of agency, path dependence, as Schwartz (2002) has pointed out with lucidity, does the precise opposite: by focusing on endogenously induced stability, actors’ only room for influence is to reproduce exogenously created paths. Although the theory allows for incremental adaptations, it does not accommodate actors creating pathbreaking change of their own volition and doing so would describe a fundamentally different history of institutional development. The pattern of stability and change that path dependence theory seeks to describe and explain would cease to resemble an evolutionary model with clearly defined path beginnings and ends, and would instead become one of incremental and nonincremental adaptation to forces internal and external to the institution. As Schwartz (2002, 16) notes, the relevant metaphor would be ecological rather than evolutionary and the generic observation would be one of ‘history matters’ rather than the theoretical statement that path dependence theory seeks to make. This observation lies at the heart of Greener’s endeavour. Of course, Greener is by no means alone in attempting to find solutions to the many problems with path dependence theory. The empirical limitations of the rigorously specified economic variant of increasing returns have led a number of social scientists to find interesting ways of conceptualizing the impact of past legacies and early choices on later institutional developments (Mahoney, 2000; Thelen, 2002; Ebbinghaus, 2005). In an attempt to salvage path dependence theory, social scientists have broadened their models significantly by changing its assumptions regarding the mechanisms of path maintenance, and by expanding the possibilities for endogenous change (Mahoney, 2000; Thelen, 2002). Rejecting a positive feedback model as ‘too deterministic and inflexible’ to account for institutional change, Ebbinghaus (2005, 2–7) recommends a loose ‘branching pathways’ and ‘open’ framework. These adaptations have departed from the well-specified, if empirically weak, theory developed in economics, thus changing radically its assumptions about the mechanisms of path creation, replication and change. British Politics 2007 2

Fiona Ross An Alternative Institutional Theory to Path Dependence

94

Although utterly convincing empirically, on these accounts of institutional development the search for path beginnings and path ends is likely to be a spurious exercise: the past determines the future in a fundamental, often unpredictable way, leading to reactive, reinforcing and branching sequences. In this formulation, the causes and mechanisms of path creation are indistinguishable from the causes and mechanisms of path maintenance. In changing the core assumptions of path dependence to include all manner of routes to stability and change, the sense in which institutions are ‘dependent’ is unclear, and the concept of a ‘path’ becomes less meaningful. Likewise, the looser our models, the less falsifiable they become. In its broadest formulations, path dependence constitutes, at best, a heuristic device. As Greener correctly acknowledges, these studies suffer from a ‘diversity of assumptions’, with the result that they offer little in the way of theory. And, by stretching its assumptions to the point of being unfalsifiable, path dependence has achieved a damaging pseudo-theoretical dominance in institutional analysis. Indeed, a review of academic journals over the past decade could easily lead to the conclusion that there is a single logic to institutional processes and it is a path-dependent one. Crouch and Farrell (2002) go as far as to refer to the ‘new determinism’. Most superficially, path dependence has served as a theoretical hanger for a range of articles demonstrating why policy or institutions have not changed in the face of compelling pressures. Commonly, it gives weight to observations of continuity. Of course, in other cases there is also more substance to the appeal of path dependence than legitimating studies of institutional and organisational inertia. Although there is nothing new in noting that history matters, as a heuristic tool, path dependence has focused our attention on more specific ways to think about how precisely the past matters: we have a greater sense about the way in which early choices can exert larger consequences down the line; as a concept it helps social scientists consider how contingent events, particularly small ones, impart long-lasting consequences and it invites us to think more critically about how seemingly independent political choices at a given moment in time have longer historical antecedents (see Hacker, 1998, 2004; Tuohy, 1999). The concept of path dependence also asks us to engage seriously with issues of timing and sequencing (Myles and Pierson, 1999; Pemberton, 2002). Timing is central to path-dependent processes because when events occur, something that is usually unpredictable has decisive consequences for both path production/termination and path reproduction (Pierson, 1998; Myles and Pierson, 1999; Pierson, 2000, 253, 256; Hacker, 2004). Small events that will have little effect in disrupting a path if they occur late in a sequence can have a critical impact if they occur earlier. The right time for events to ‘matter’ is before paths become institutionalized (see Pierson, 2000; Schwartz, 2002). British Politics 2007 2

Fiona Ross An Alternative Institutional Theory to Path Dependence

95

In short, despite the problems identified earlier, path dependence theory refocuses our attention away from the power of untamed strategic decisions taken at a single point in time. In that sense it is an antidote, possibly an overreaction, to rational choice analysis and behaviourism. It has also generated some counter-intuitive insights that political scientists can take on board in exploring their cases. These are real accomplishments but they do not constitute a theory of institutional development. They are aspects of path dependence theory, but they are insufficient observations to provide evidence of path dependence. In loosening the well-specified (and thus testable and falsifiable) assumptions of increasing returns arguments to render path dependence adequately flexible to capture the empirical data, scholars have reduced its theoretical value, hence Greener’s efforts to render these somewhat amorphous models more systematic through morphogenetic social theory. Can morphogenetic social theory solve these weaknesses, rendering a noneconomic formulation of path dependence theory both more dynamic and more systematic, or otherwise? Alternatively, does morphogenetic social theory offer us a new and improved lens through which we can understand institutional development, possibly adding a source of agency to historical institutionalist analysis and capturing variation within and across institutional settings but not salvaging path dependence theory?

Mixing Theories: Morphogenetic Social Theory and Path Dependence In morphogenetic social theory, the relevant actors are significant vested interests. History matters in that groups inherit both structural and cultural properties, giving rise to situational logics, based on a combination of necessary/contingent structural relationships and complementary/incompatible cultural (ideational) ones. For example, at the founding of the NHS stability in the Service was secured structurally by the necessary relationship between the state and consultants (although the NHS also had structural tensions in the ‘necessary but incompatible relations between the state and the GPs’). Ideas, the cultural logic, also promoted stability owing to what Greener describes as the ‘necessary relation between Keynesianism, Fabianism and the medical model of care’. These logics are not rigid and binding, rather they shape and structure the interests, relationships and balance of power between groups. In turn, these interactions impart their own legacy on structural and cultural conditionings, recreating or replicating the situational logics in place, and thus contributing to the subsequent morphogenetic cycle. Despite the very obvious relevance of these relationships and ideas for political phenomena and the possibilities that morphogenetic social theory offers for institutional analysis (discussed below), there are at least three problems with layering morphogenetic theory onto path dependence theory to British Politics 2007 2

Fiona Ross An Alternative Institutional Theory to Path Dependence

96

improve the theoretical capacity of the latter. First and fundamental, we have no explanation of how the core tenets of the two models are compatible, how their assumptions are theoretically consistent, or the precise means by which these two theories could be rendered compatible. For example, it is not clear that the cycles of morphogenetic theory map onto the concept of paths: they are relational and ideational inheritances that do not necessarily contribute to bounded episodes in history. The morphogenetic cycle of events does not match the concept of distinctive paths with clear creation and break points. Rather, it is a more ecological cycle of institutional development. In this sense, morphogenetic social theory offers the potential to move us away from fruitless debates over path beginnings and ends: cycles feed into each other. Second, and related to the first point, path dependence cannot simply be rendered systematic by layering a more dynamic framework on top of it. Its assumptions, and those of the synthesized model, require clear specification. Despite the many strengths of Greener’s analysis, the concept of path dependence is applied as loosely in this synthesis as it is elsewhere. Path dependence cannot simply be theoretically enriched by charting a number of factors that encourage/hamper stability/change within organizational settings. It also requires a clear specification about how path dependence theory is being applied. In Greener’s case study, path dependence at different times refers to stable conditions, stable relationships and mutual interests. Path dependence is taken as the default position and the conditions for necessary and sufficient path dependent processes are never established. We are not offered an integrated model of institutional creation, path maintenance and change. Consequently, the synthesis does not help clarify the fundamental problem of path boundaries that have plagued looser formulations of the theory. Likewise, Greener does not address the weighting of the structural and cultural dimensions of morphogenetic theory: are both dimensions necessary and sufficient conditions to support continuity and change? If so, ideas carry the same explanatory weight as interests; an issue that requires explicit empirical investigation in the case study. Third, we have no understanding of how this very different model of morphogenetic theory, introducing power, interests and ideas, relates to the range of standard institutional properties that historical institutionalists view as so essential to their models. It is unclear how situational logics interact with the institutional processes and characteristics that feature prominently in historical institutionalism. Although the case study of the NHS very nicely demonstrates necessary and contingent structural interests (and to a lesser extent cultural compatibilities and incompatibilities) and notes how these logics can change, the analysis does not link the explanatory power of these variables to other institutional properties. The case study does not establish that the relational and ideational conditions are in any sense sufficient to facilitate British Politics 2007 2

Fiona Ross An Alternative Institutional Theory to Path Dependence

97

change nor are we told how they might relate to other necessary conditions. We certainly have contrasting accounts of the NHS that are equally plausible (Hacker, 1998, 2004; Tuohy, 1999; Freeman and Moran, 2000).

An Alternative Account Although some of these issues might be resolved through a detailed analysis of both theories, others are more fundamental and the problems with path dependence are probably larger than the scope of morphogenetic social theory. Of greater value, Greener has exposed us to a possibly more interesting and dynamic alternative model of institutional development, one that alerts us to two potentially important sets of variables that political scientists have not effectively brought to the fore of their institutional analyses. In many respects, the core units of morphogenetic social theory are better suited to the political realm than those of path dependence theory: structural interests, institutional relationships, competition, power and ideas. These elemental political traits are poorly handled by path dependence theory, where there is little significant room for political competition (reduced by first-mover advantages and reinforcing institutional rules), partisanship or think tanks, all of which might encourage actors to think critically and creatively. In path dependence arguments, politicians think small and repetitively, regardless of institutional incentives for policy departure, or conflict over ideas (Peters et al., 2005). Likewise, although the use of power enters the path dependence equation at the margins (in exacerbating first-mover advantages), it generally describes a consensual scenario where the incentives for institutional maintenance are so great that differences between actors generate only on-path adjustments. Even incremental changes have no cumulative impact, a premise that is inconsistent with much policy research (Moran, 1994), and extraordinary given the time span of historical analyses (Peters et al., 2005, 1277). In morphogenetic social theory, new equilibria can be sustained by the astute use of power (see Deeg, 2001; Schwartz, 2002). As Peters et al. (2005, 1278) contend in their critique of path dependence theory: Political conflict y is not just a feature of formative moments but just as often occurs during path-dependent periods, whenever path dependency is sustained by a dominant political coalition successfully fending off all attempts by minorities to alter the political course. Indeed, in many respects, morphogenetic social theory promises to offer a better fit with political data than path dependence. Because of the muchimproved capacity for political agency, the model is not unduly weighted towards institutional continuity. Perhaps ironically, morphogenetic theory as applied by Greener serves to illustrate the comparatively limited conditions British Politics 2007 2

Fiona Ross An Alternative Institutional Theory to Path Dependence

98

under which we would anticipate path-dependent effects/institutional stability: institutions are just as likely to weaken over time as they are to strengthen. Morphogenetic social theory also offers significant promise for comparative institutional analysis: drawing on Archer’s model we can generate hypotheses about the type of structural interests and cultural understandings in given institutional situations that exert stabilizing and de-stabilizing tendencies. It incorporates change in situational logics by considering when and why compatible ideas become incompatible and when and why necessary relationships become contingent. The model can account for shifting degrees of institutional security, allowing us to generate predictions about the structural and cultural conditions under which institutions change; a far more satisfactory scenario than exogenous or contingent events. In short, though morphogenetic social theory cannot remedy the larger theoretical problems afflicting loose models of path dependence, it offers innovative ways of rendering institutional analysis more dynamic and systematic. Likewise, it can potentially expand our understanding of how power resources change within organizational settings. Layering morphogenetic theory onto path dependence runs the risk of limiting and constraining its possibilities for political science.

Concluding Comments Theorizing a topic as large as institutional development is difficult. In his innovative attempt at improving the theoretical foundation of non-economic formulations of path dependence, Greener has introduced us to a potentially important and exciting new lens through which we can systematically compare institutions, unravel different degrees of tension within existing systems and predict their comparative stability across time and space. Yet linking morphogenetic social theory to path dependence threatens to do more harm than good. It cannot resolve the fundamentally problematic issue for broader models concerning the beginnings and ends of paths and it is unclear what the concept of paths actually contributes to this more dynamic and political theory: in morphogenetic social theory, institutional development is more ecological and less discrete than implied by the concept of a path. Likewise, the sense in which institutions are genuinely dependent in morphogenetic theory is far from obvious. What Greener has offered us is an alternative model of institutional development, one whose assumptions are better suited to the political realm than path dependence and one that can possibly move us away from spurious debates over the concept of a path and its empirical verification. At present, path dependence is an appealing heuristic device for thinking about historical sequencing and policy inheritances; it is less convincing as a theoretical framework for understanding the perennial challenge of continuity and change and structure and agency in delivering these conditions. Morphogenetic social theory promises to do a better job. British Politics 2007 2

Fiona Ross An Alternative Institutional Theory to Path Dependence

99

References Arrow, K.A. (2004) ‘Path Dependence and Competitive Equilibrium’, in T. Guinnance, W.A. Sundstrom and W.C. Whatley (eds.) History Matters: Essays on Economic Growth, Technology, and Demographic Change, Stanford, CA: Stanford University Press. Crouch, C. and Farrell, H. (2002) ‘Breaking the path of institutional development? Alternatives to the new determinism’, Discussion Paper, 02/5, Max Planck Institute for the Study of Societies. Deeg, R. (2001) ‘Institutional chance and the uses and limits of path dependency: The case of German finance’, Discussion Paper, 01/6, Max Planck Institute for the Study of Societies. Ebbinghaus, B. (2005) ‘Can Path dependence explain institutional change?’, Max Planck Institute for the Study of Societies Discussion Paper, 05/2. Freeman, R. and Moran, M. (2000) ‘Reforming Health Care in Europe, in M. Ferrera and M. Rhodes (eds.) Recasting European Welfare States, London: Frank Cass. Hacker, J.S. (1998) ‘The historical logic of national health insurance: structure and sequence in the development of British, Canadian, and US medical policy’, Studies in American Political Development 12: 57–130. Hacker, J.S. (2004) ‘Dismantling the health care state? Political institutions, public policies and the comparative politics of health care reform’, British Journal of Political Science 34: 693–724. Klein, R. (2001) The New Politics of the NHS, (4th edn) London: Prentice-Hall. Liebowitz, S.J. and Margolis, S.E. (1990) ‘The fable of the keys’, Journal of Law and Economics 33(1): 1–25. Liebowitz, S.J. and Margolis, S.E. (1995) ‘Policy and path dependence: from QWERTY to Windows 95’, Regulation: The CATO Review of Business and Government 18(3): 33–41. Mahoney, J. (2000) ‘Path dependence in historical sociology’, Theory and Society 29(4): 507–548. Moran, M. (1994) ‘Re-shaping the healthcare state’, Government and Opposition 29(1): 48–63. Myles, J. and Pierson, P. (1999) ‘The comparative political economy of welfare reform’, Paper presented at the Robert Schuman Centre (RSC), European University Institute, Florence, Italy. Pemberton, H. (2002) ‘Storing up problems for the future? Path dependence and Britain’s present pensions crisis’, Paper presented at the Annual Meeting of the American Political Science Association, Boston, MA, USA. Peters, B.G., Pierre, J. and King, D. (2005) ‘The politics of path dependency: political conflict in historical institutionalism’, Journal of Politics 67(4): 1275–1300. Pierson, P. (1998) ‘Not just what, but when: issues of timing and sequencing in comparative politics’, Paper presented at the Annual Meeting of the American Political Science Association, Boston, MA, USA. Pierson, P. (2000) ‘Increasing returns, path dependence, and the study of politics’, American Political Science Review 92(4): 251–267. Ross, F. (2007a) ‘Policy Histories and Partisan Leadership in Presidential Studies: The Case of Social Security’, in G. Edwards and D. King (eds.) Politics and Polarization: The George W. Bush Presidency, Oxford: Oxford University Press. Schwartz, H. (2002) ‘Down the wrong path: path dependence, markets, and increasing returns’, Typescript, University of Virginia. Thelen, K. (2002) ‘How Institutions Evolve: Insights from Comparative Historical Analysis’, in J. Mahoney and D. Rueschemeyer (eds.) Comparative Historical Analysis in the Social Sciences: Innovations in Theory and Methods, Cambridge: Cambridge University Press. Tuohy, C.H. (1999) Accidental Logics: The Dynamics of Change in the Health Care Arena in the United States, Britain, and Canada, Oxford: Oxford University Press.

British Politics 2007 2