%20-%20https://sketch.com%20--%3e%3ctitle%3e3ECBBF50-FD08-412A-943D-2CFC5E025A16%3c/title%3e%3cdesc%3eCreated%20with%20sketchtool.%3c/desc%3e%3cg%20id='Stylesheet'%20stroke='none'%20stroke-width='1'%20fill='none'%20fill-rule='evenodd'%3e%3cg%20transform='translate(-926.000000,%20-668.000000)'%20id='BUTTONS'%3e%3cg%20transform='translate(100.000000,%20597.000000)'%3e%3cg%20id='BTN-/-RIGHTS'%20transform='translate(816.000000,%2065.000000)'%3e%3cg%20id='ICON-/-Rights'%20transform='translate(10.000000,%206.000000)'%3e%3cpath%20d='M13.5752841,0%20L1.34943182,0%20C0.59375,0%200,0.586666667%200,1.33333333%20L0,22.6666667%20C0,23.4133333%200.620738636,24%201.34943182,24%20L17.6505682,24%20C18.40625,24%2019,23.3866667%2019,22.6666667%20L19,5.36%20L13.5752841,0%20Z%20M13.5752841,1.89333333%20L17.0838068,5.36%20L13.5752841,5.36%20L13.5752841,1.89333333%20Z%20M1.34943182,22.6666667%20L1.34943182,1.33333333%20L12.1988636,1.33333333%20L12.1988636,5.36%20C12.1988636,6.10666667%2012.7926136,6.69333333%2013.5482955,6.69333333%20L17.6235795,6.69333333%20L17.6505682,22.6666667%20L1.34943182,22.6666667%20Z'%20id='Shape'%20fill='%23595959'%20fill-rule='nonzero'%3e%3c/path%3e%3ccircle%20id='Oval'%20stroke='%23595959'%20stroke-width='1.5'%20fill='%23595959'%20cx='9.5'%20cy='13.5355339'%20r='5.5'%3e%3c/circle%3e%3cpolyline%20id='Path-2'%20stroke='%23FFFFFF'%20stroke-width='1.5'%20transform='translate(9.500000,%2013.035534)%20rotate(45.000000)%20translate(-9.500000,%20-13.035534)%20'%20points='8%2016.0355339%2011%2016.0355339%2011%2010.0355339'%3e%3c/polyline%3e%3c/g%3e%3c/g%3e%3c/g%3e%3c/g%3e%3c/g%3e%3c/svg%3e)
Lowndes argues that approaches to explaining institutional change depend on how we understand the relationship between agents and institutions.Footnote 1 All versions of institutional theory argue that institutions matter because they shape the choices, behaviour and even the interests and identities of agents. However, perhaps the most prominent approaches within institutionalism have adopted rather a determinist view about the extent to which institutions shape agents, resulting in a highly constrained view of agency. This ‘sticky’ view of institutions has led to difficulties in explaining how institutions change. The three established versions of ‘new institutionalism’ (rational choice, historical and sociological institutionalism) are now under criticism from a new approach – constructivist institutionalism – with the critics claiming that the established positions are unable to explain institutional change effectively, largely because the agents in question are said to be highly constrained by their institutional environments. Prominent constructivist institutionalists, such as Colin Hay, Mark Blyth and Vivien Schmidt, thus offer alternative accounts featuring interpretive agents operating in relatively fluid ideational and discursive contexts to explain institutional change.Footnote 2
This article argues that the alleged problems in explaining institutional change are exaggerated. A key argument of this article is that a suitably tailored version of historical institutionalism (HI) can accommodate a constructivist approach to produce a sophisticated and more rounded account of how interpretive agents interact dialectically with institutional and wider structural contexts and produce change. It is argued that the constructivists in question have a somewhat confused understanding of constructivism, excessively privilege agency, and lose sight of the significance of institutional and wider structural variables which inevitably shape agency and institutional change processes. The constructivist approach under review is somewhat ironic because institutionalism re-emerged in recent decades as a critique of overly agent-centred approaches – such as behaviouralism – arguing that we needed to bring institutions ‘back in’. The constructivists in question now run the risk of taking institutions ‘back out’.
Parsons argues that some versions of institutionalism ‘overlap’ with constructivism.Footnote 3 This article refines Parsons's view and argues that constructivism (properly understood) is integral to a suitably tailored version of historical institutionalism. More broadly, the challenge is how to describe and explain contingent degrees of agent-centred discretion (arguably the ultimate propellant of institutional change) within a context of constraint, conditioning and empowerment associated with institutionally embedded agents. I argue that such empowerment and discretion can be potentially enhanced when agents (as they typically do) operate not only across multiple institutional environments but also in wider structural environments which are often changing and which can both constrain and empower them.
I first outline the constructivist institutionalist critique of historical institutionalism. The constructivist institutionalism in question is then outlined and critiqued. I then argue that an account of HI that places active and interpretive agents at the centre of its analysis and locates agents dialectically interacting with institutional and wider structural contexts can absorb core elements of constructivism and explain institutional change. I conclude with empirical illustrations of the key claims about how agents operate within institutions with contingent degrees of ‘bounded discretion’ and how, in turn, institutional and wider contexts can shape and empower agency in change processes. Empirical illustrations are made using examples mainly from Australian politics, especially from economic policy and central banking.
The problematic constructivist critique of historical institutionalism
In the new institutionalist literature, institutions are often depicted as inertial, rule-bound and resistant to change. Pierson argues that: ‘Institutional arrangements in politics are typically hard to change’, and that ‘Actors find that the dead weight of previous institutional choices seriously limits their room to maneuver’.Footnote 4 And when change does occur, much is made of the alleged path-dependent nature of institutional change, pointing to the effects of institutional legacies, sunk costs, decision branches, increasing returns or lock-in effects.Footnote 5 Institutional life is thus seen as operating within the grooves of established institutional paths or under the ‘self reinforcing persistence of institutional structures’, as Schwartz puts it.Footnote 6 Given this ‘sticky’ view of institutions, it is often claimed that historical institutionalism has tended to be better at explaining institutional stasis and continuity rather than change. Hay and Wincott thus worry that a ‘latent structuralism’ can be discerned in much of historical institutionalism, while more recently Olsen argues that the approach is ‘overly structuralist and does not grant purposeful actors a proper role’.Footnote 7 Crouch goes so far as to argue that ‘neo-institutionalism will become the new dismal science, endlessly demonstrating that actors … are doomed to keep re-enacting their past legacies’.Footnote 8
These sorts of criticism are not new. Pioneering HI scholars such as Thelen and Steinmo argued long ago that a ‘critical inadequacy of institutionalist analysis has been a tendency towards mechanical, static accounts that largely bracket the issue of change and sometimes lapse inadvertently into institutional determinism’.Footnote 9 In his work on the institutional shaping of national tax regimes, Steinmo also noted that existing institutional accounts tended to be seen as ‘uncomfortably static’, tending towards ‘institutional determinism’.Footnote 10 In such accounts, exogenous shocks or crises which break the bonds of institutional constraint have often been favoured as explanations of institutional change. As Krasner argued: ‘Change is difficult … Institutional change is episodic and dramatic rather than continuous and incremental. Crises are of central importance’.Footnote 11 Yet Thelen and Steinmo are rightly critical of such exclusively exogenous accounts of change: ‘the problem with this model is that institutions explain everything until they explain nothing. Institutions are an independent variable and explain political outcomes in periods of stability; but when they break down, they become the dependent variable, whose shape is determined by the political conflicts that such breakdown unleashes’.Footnote 12 In other words, Krasner's model of institutional change is not an institutional theory of change per se and essentially ‘black boxes’ internal institutional dynamics. The key missing element, of course, is a role for agency and some kind of plausible analysis for the interplay between agency and institutional contexts.
These types of criticisms now act as the springboard for the current constructivist critiques of institutional theory. The tendency to see agents as deeply embedded or constrained by institutional environments is held to be the main problem. Indeed, rational choice theories do tend to see institutions as arrangements based on relatively stable choice equilibria, or they assume largely a priori accounts of agency and preferences verging on a ‘calculating automatons’ approach.Footnote 13 The norm-bound, ‘over-socialized’ and limited conceptions of agency often found in sociological institutionalism also make explaining change difficult.Footnote 14 Thus, Fligstein argues that: ‘The critique of both the sociological and rational choice perspectives that I want to make suggests that neither opens up the problem of action and gives real people much leeway in creating their social worlds’.Footnote 15
Constructivist institutionalists like Colin Hay see these problems also afflicting HI; indeed, he argues that HI ‘compounds’ them by also emphasizing ‘path-dependent lock-in effects’, which are said to further inhibit change.Footnote 16 Mark Blyth also argues that historical institutionalism sees institutions as ‘constraining rather than enabling political action’.Footnote 17 Vivien Schmidt, who refers to herself as a ‘discursive’ institutionalist, similarly claims that the established institutionalisms have produced a view of institutions that is ‘overly sticky’, constrained by fixed preferences, fixed norms or what she sees as the almost determinist ‘history-based logic of path dependency’.Footnote 18 She argues that institutions in this view are resistant to change, acting ‘mainly as constraints’, leaving the problem of ‘How do we explain institutional change?’ and, indeed, ‘How do we explain agency?’Footnote 19 From this vantage point, Schmidt sees the established institutionalisms as ‘subordinating’ agency to structure, leaving us with ‘unthinking actors’ who are in an important sense ‘not agents at all’.Footnote 20 More recently, she complains about ‘historical institutionalism's stickiness and lack of sentient agents’, and that historical institutionalism ‘is still missing the “micro-foundations” of macro-historical change’.Footnote 21
The solution to these problems, according to the new constructivists, is to zero in on agency, but especially the subjective ideational and inter-subjective discursive realm, seemingly a more fluid and flexible environment in which to effect change, largely because this move ostensibly allows agents to ‘construct’ their realities and fields of action, apparently unimpeded or less impeded by institutional constraints. Schmidt thus sees constructivist institutionalism as a better alternative, because it ‘puts agency back into institutional change’.Footnote 22 For Hay, a constructivist focus means an emphasis on ‘strategic actors’, ‘who must rely upon perceptions’ of their environment and whose ‘desires, preferences and motivations’ are ‘irredeemably ideational’.Footnote 23
A slightly different approach to the same end comes in the work of Mark Blyth.Footnote 24 Whilst not explicitly labelling himself as a constructivist institutionalist, he does argue that much of the existing research on the role of ideas, including in historical and rational choice institutionalism, has tended to use ideas as ‘fillers’ or auxiliary hypotheses or as important only when they are congruent with pre-existing institutions or a nation's wider political culture.Footnote 25 Blyth claims that neither approach gives sufficient weight ‘to ideas as explanatory factors in their own right’.Footnote 26 Whilst he worries about descending into an ‘ideational essentialism’, this is more or less what happens: the key problem being a lack of empirically grounded theorizing about how agents and their ideas actually connect with institutions or indeed wider structures.Footnote 27 His main argument is that in the throes of major crises, institutional contexts largely dissolve amidst high levels of catharsis or ‘Knightian uncertainty’, with ideas subsequently playing a critical constructivist role in interest definition and the shaping of institutional change.Footnote 28 Under such conditions, according to Blyth, there occurs a ‘separation of ideas from their current institutional moorings’.Footnote 29 Blyth argues that ‘institutional change only makes sense by reference to the ideas that inform agent's responses to moments of uncertainty and crisis’.Footnote 30 ‘Ideas give substance to interests and determine the form and content of new institutions’.Footnote 31 Hence, according to Blyth, the mechanisms of crisis and uncertainty largely serve to erase existing institutional conditions: a situation analogous to the logic of Krasner's argument above, where suddenly institutions ‘explain nothing’. As Jacobsen puts it, ‘all that is solid melts into ideational air’.Footnote 32 Under such conditions, the approach largely conflates interests with ideas and also separates ideas and agents from institutions. Blyth calls this, ‘ideas all the way through – that is, a situation where ideas permeate all aspects of materiality and determine agents’ orientations to social objects’.Footnote 33 This is an ideationally primitive account of institutional life and change.
What role, if any, do institutions play in such accounts and is the new constructivist institutionalism ‘institutionalist’ in any significant sense? Here, the new constructivists sometimes offer qualifying statements. Hay, for example, argues that ‘institutional change does indeed occur in a context which is structured’, and like Blyth he warns that it is important not to subscribe ‘to a voluntarist idealism in which political outcomes might be read off, more or less directly, from the desires, motivations and cognitions of the immediate actors themselves’.Footnote 34 Yet at other times he argues that the outcomes of political struggles ‘can in no sense be derived from the extant institutional context itself’.Footnote 35 Hay similarly argues that ‘preference sets or logics of conduct cannot be derived from the [institutional] setting in which they are located’.Footnote 36 And for Hay and Rosamond it is ‘the ideas that actors hold about the context in which they find themselves rather than the context itself which informs the way in which actors behave’.Footnote 37 Institutions also recede when Blyth, as noted above, talks about agents losing ‘institutional moorings’ amidst change processes. Blyth cautions that ‘this is not to say that only ideas matter, nor that institutional change is purely an ideational affair’.Footnote 38 ‘But’, he continues, ‘ideas certainly do matter in periods when existing institutional frameworks … fail’.Footnote 39 What this notion of ‘failure’ actually means is not explained, but it is an outcome one assumes that is tantamount to institutional erasure in crisis moments. ‘At these junctures, it is ideas that tell agents what to do and what future to construct’;Footnote 40 ‘In sum’, he argues, ‘what is critically important in understanding agent's behaviour are the ideas held by agents, not their structurally [or institutionally] derived interests’.Footnote 41 Ultimately then, what Blyth is doing is constructing a special case (crises), where only ideas appear to matter in informing agents about their interests and strategies in shaping institutional change.
Is there a crisis ‘let out’ where institutional effects on agents recede? The answer, arguably, is no, for two reasons. First, Blyth's claim that institutions somehow ‘fail’ during crises is unconvincing. Does an institution such as a central bank, even amidst a major macroeconomic crisis, simply disappear from casual view? Not according to Blyth's own empirical analysis of the fall of post-war embedded liberalism in the United States when the institutional clout of the Federal Reserve is emphasized as important in shaping institutional change.Footnote 42 The same applies when Blyth talks about the importance of the institutional and organizational resources that underpinned the new business mobilization that helped sponsor the rise of neoliberalism in the same era in the United States. These institutional underpinnings were important in supporting and mobilizing new business ideas about policy and institutional change. Yet despite these empirical acknowledgements, Blyth's theoretical analysis steadfastly privileges ideas and does not contain a theoretically grounded analysis of how ideas and institutions interact. True, at one point, Blyth claims that ‘ideas, interests and institutions’ interact ‘synthetically’, but this insight does not form the basis for his theoretical account.Footnote 43 Secondly, although Blyth emphasizes the role of ideas during crisis periods, the main problem with this selective view is that it can plausibly be argued that ideas always matter because they are always important in agent's interpretations of reality and in helping to forge interest definitions and change strategies within institutional contexts.Footnote 44
For her part, Schmidt states her approach: ‘risks appearing highly voluntaristic unless structural constraints derived from the three older institutionalisms are included’.Footnote 45 Generally, however, she sees institutions merely as arenas which ‘frame the discourse’.Footnote 46 ‘The “institutionalism” in discursive institutionalism’, she writes, is not about ‘the external rule following structures of the three older institutionalisms’, but is instead reduced to the ‘constructs of meaning which are internal to “sentient” (thinking and speaking) agents’.Footnote 47 For Schmidt, institutions simply appear as a ‘meaning context’, as providing ‘background information’, or as ‘contingent (the result of the agent's thoughts, words and actions)’.Footnote 48 Schmidt also argues that the deliberate nature of discourse allows agents to ‘conceive of and talk about institutions as objects at a distance, and to dissociate themselves from them’.Footnote 49
There are a range of problems with the new constructivist institutionalist accounts above. First, there is a degree of confusion in relation to the several varieties of extant constructivism.Footnote 50 The accounts above vary from postmodern accounts, where ideas, inter-subjective meanings and discourse are primitive and wholly define or constitute social and institutional life, to more ontologically realist accounts, which admit that institutions and wider structures can have real effects.Footnote 51 Here, institutions and wider structures take on a congealed reality of their own and cannot simply be reduced to real-time sets of ideas or inter-subjective meanings held by agents. This confusion accounts for why these scholars oscillate from ideationally primitive accounts to periodically worrying about ‘ideational voluntarism’ or losing sight of the ‘structural constraints derived from the three older institutionalisms’.Footnote 52 Secondly, the rejection or sidelining of HI is an unfortunate step. Hay sees HI as increasingly ‘hollowed out’, allegedly by prominent scholars such as Paul Pierson and Peter Hall, who he suggests have opted for rational choice, calculus-based approaches to analysis.Footnote 53 Schmidt agrees and argues that ‘historical institutionalists have (mostly) turned to rational choice institutionalism’.Footnote 54 As argued below, the claim that HI has been ‘hollowed out’ or that HI scholars have defected to rational choice is far from being wholly correct. Thirdly, and relatedly, the constructivists in question have misjudged the resources available within HI. In rejecting what they see as excessively ‘sticky’ HI, they have lost the opportunity to ground HI in a more agency-based approach that is capable of absorbing constructivist insights and offering a rounded account of institutional change. These constructivists have ended up with a form of analysis which relegates institutions to a vague or almost meaningless role, leaving the question about how such an account can be meaningfully described as institutionalist. Blyth criticizes historical institutionalism for its alleged ‘lack of explicit theorizing about the relationship between ideas, interests and institutions’, and yet it is not clear that granting primacy to ideas achieves this either.Footnote 55
In Defence of an agent-centred historical institutionalism
The constructivists in question wish to put ‘agency back into institutional change’, but the notion of agency is actually well established within an important strand of HI thinking and a further problem with the new constructivist assault is to overlook this.Footnote 56 It is often assumed that HI is a monolithic whole, but this view is incorrect. In fact, HI is best seen as constituted by two different types of approach. The first of these, as outlined above, features strong elements of institutional stickiness and path-dependency, with the resultant emphasis on exogenous shocks to explain change.Footnote 57 Arguably, this is an overly structuralist account which downplays agency, and it has quite rightly been the focus of criticism by, amongst others, the new constructivists. However, there is also a second strand of HI analysis, which escapes these criticisms and which can provide a robust account of institutional change. This is an approach that focuses on active agency within institutional settings and that sees the agents in question as being shaped (though not wholly determined) by their institutional environments. In this manner, unlike the new constructivism, this approach does not lose sight of the importance of institutional dynamics, and nor, contra the constructivists' views about HI, does it lose sight of agency. As Crouch puts it, the challenge ‘is to devise theories of action that retain all the insights of neo-institutionalism concerning the constrained nature of human action, while also being able to account for innovation’.Footnote 58
It is true, of course, as many constructivists argue, that the effects that institutions have on agents are substantially constituted by agents themselves through ideational processes and choices. Hence, Schmidt approvingly quotes John Searle when he states: ‘Institutional facts are those things which exist only through collective agreement about what stands for an institution’.Footnote 59 The problem with this formulation, however, is that it perceives only one dimension of the two-way dialectical interaction between agents and institutions. Institutions also confront agents with sets of what Searle calls ‘constitutive rules’ and with other constraining and enabling conditions which can become, as Adler puts it, ‘diffused and consolidated’ over time and which shape behaviour.Footnote 60 It may be possible under extraordinary (say revolutionary) conditions for agents to collectively overturn or deny institutions, but more ordinarily, institutions confront agents in the here and now as embedded, already structured terrains. Hence (whatever I think), I will typically go to jail if convicted of murder under the law. Institutions are thus ontologically prior to the individuals who populate them at any given time. The temporal dimension is important here. Institutions have properties that help structure thought and behaviour at one remove from the immediacy of thought or action by agents at any given point in time. Institutions can thus shape or sometimes even impose behaviour. This is what gives institutions causal properties and why, at bottom, we pursue ‘institutional’ analysis. By essentially eschewing a meaningful institutional analysis, recent constructivist or discursive institutionalists place almost all explanatory weight on agency and lose sight of institutions. In fact, they go further by eschewing a meaningful contextual analysis of agency and instead conflate agents with the ideational. In a parallel critique of the postmodern institutional interpretivism of Bevir and Rhodes, McAnulla argues that such a move ‘leaves us with an implausibly narrow view of how individual beliefs and actions are generated’, whilst Adler more generally argues that constructivists (especially postmodern constructivists) ‘concede too much to ideas’.Footnote 61
Ultimately, however, in a more dynamic historical sense, the approach here does not give primacy to agents, institutions, structures or ideas, but instead holds each to be mutually constitutive in a dialectic manner.Footnote 62 Agents thus confront institutions as a ‘distinct strata of reality’ that must be dealt with as entities in the here and now and perhaps changed over time.Footnote 63 In other words, institutions are more than just real-time ideational artefacts but are more like inherited sets of rules and duties that need to be navigated and negotiated. Moreover, ideas do not operate in a vacuum but are instead ‘embedded in a historical context and need institutional support to be effective’.Footnote 64 Hence, and this is what is missing from the recent constructivist institutionalist accounts, the impact of such institutional (or indeed wider) structural environments means that agents’ choices are not made on a tabula rasa. Agents and environments interact and mutually shape one another over time. When Hay and Rosamond argue that what matters is the ‘ideas that actors hold about the context in which they find themselves rather than the context itself’, they adopt an either/or approach and overlook this crucial dialectical interaction.Footnote 65 This means that at one level, institutional or structural environments are analytically distinguishable from agents and are ‘out there’ to the extent that they are not just constituted by real-time subjective ideational construction. Institutional or structural environments can exert real (though always interpreted) effects by imposing costs or benefits on agents, by shaping actor interpretations and preferences, the scope of ‘bounded discretion’ of agents in institutional life, and the resources and opportunities that are available to actors.
This alternative ‘non-sticky’ or more ‘flexible’ approach within HI thus emphasizes agency, the dialectical interaction between agents and institutions, and rejects forms of institutional analysis which over-condition actors. Thelen and Steinmo have rightly argued that we need to unpack the institutional black box and focus on the activities of ‘strategic actors’, and how they are capable of acting on ‘openings’ provided by shifting contexts.Footnote 66 Indeed, Hay himself approvingly quotes Thelen and Steinmo, when they emphasize the role of agency and suggest that institutions not only constrain behaviour but are also the ‘outcome (conscious or unintended) of deliberate political strategies’.Footnote 67 Building on the earlier work of Thelen and Steinmo, HI scholars, such as John Campbell, have also provided a useful account of agent-centred institutional change.Footnote 68 Campbell rejects institutional determinism and emphasizes the importance of agency, but also warns against ‘reducing our accounts of institutional change to interpretive frameworks’.Footnote 69
However, some scholars, like Schneiberg, worry about abandoning ‘institutional determinism’ and the emphasis on the ‘constraining power of context’ that allegedly gives HI its ‘analytical edge’.Footnote 70 He argues: ‘Either we preserve institutional insights about path dependence and the constraining power of context and deny the prospects for fundamental or qualitative transformation. Or we preserve observations about fundamental change and new path creation and deny our insights about path dependence and the explanatory power of the institutional context’.Footnote 71 Arguably, such a choice is unnecessary if we reject the notion that ‘institutional determinism’ within HI provides a useful ‘analytical edge’. As just argued, determinist institutional theory has reached an impasse which is unable to provide a rounded account of agency within institutional life and which is unable to explain change other than through exogenous shocks or highly constrained notions such as path dependency.Footnote 72 A more nuanced and balanced approach that moves beyond such totalizing accounts is needed, one that can model agents operating with variable and contingent degrees of constraint and empowerment within institutional settings. As Hall argues, we need to ‘understand how institutions that are to some extent plastic can nonetheless contribute to the structuring of the political world’.Footnote 73
This type of HI approach with its focus on institutionally situated, active agents needs to be fleshed out in order to outline the sorts of capabilities and resources that actors have or which they can garner from their institutional environments, and which may enable them to effect change. To do this, we focus on three sets of capabilities and resources, all of which provide useful agent-centred micro-foundations for institutional analysis.
First, agents interpret and construct the experience of their institutional situation using subjective and inter-subjective cognitive and normative frameworks and discursive processes. An HI approach can easily integrate this constructivist notion of interpretive agency and give full recognition that ideas, language and the inter-subjective discursive processes provide the crucial building blocks for establishing meaning and understanding and thus lead to purposeful action in politics and institutional life.Footnote 74 In their original contribution, Thelen and Steinmo emphasized the importance of an institutional analysis that explores the ‘interaction between ideas, interests and institutions’.Footnote 75 Other HI scholars such as Hall, King and Weir have also highlighted how institutions also play an important role in shaping the absorption and diffusion of policy ideas.Footnote 76 In Hattam's detailed analysis of labour movement strategy in the United States, for example, she emphasizes the impact of institutional shaping, especially in relation to the impact of the judicial system, but she also adds an ‘interpretive leg’ to her analysis by arguing that ‘only by considering the meaning and significance that workers brought to their particular institutional environments can we begin to decipher the distinctive pattern of working-class formation in the United States’.Footnote 77
Not only is HI compatible with key constructivist insights but it also offers a solution to a long-standing problem with constructivism by offering a more rounded account of agency. Constructivism's strength lies in insisting that ideas and inter-subjective meanings inform and shape the interests and choices of agents. But constructivism's account of agency is somewhat truncated.Footnote 78 The tendency within constructivism to focus heavily on ideational structures (or on social wholes) breeds its own kind of ‘structuralism’, a move somewhat analogous to the norm-bound, ‘over-socialized’ conceptions of agency often found in sociological institutionalism.Footnote 79 Constructivists also say too little about where ideas, meanings or norms come from, how they are shaped and how they change. At present, within much constructivist thinking, ideas are largely assumed to emerge out of other ideational processes, such as learning, emulation or socialization. Constructivists often seem intent in bracketing off other sources of influence. Yet without a clearer view of situated agency, the full gamut of factors and contexts that are likely to shape ideas and interests is likely to remain obscured. As Checkel argues, ‘without more sustained attention to agency, these [constructivist] scholars will find themselves unable to explain where their powerful social structures (norms) come from in the first place and, equally importantly, why and how they change over time’.Footnote 80 Interaction between agents and their contexts can be important in this respect. Moreover, institutional and wider contexts are not only shaped by ideas but in turn play a role in shaping ideas. This dialectical process of mutual shaping is illustrated empirically later in this article.
Secondly, despite the role of institutional shaping, agents still have contingently variable degrees of agential space or ‘bounded discretion’ within institutional settings and can change institutions over time.Footnote 81 To a point, this is analogous to the ‘choice-within-constraints’ approach found in at least some versions of rational choice institutionalism that adopt context-rich and nuanced specifications of ‘thick rationality’.Footnote 82 Institutions matter because of the ways they reflect, refract, restrain and enable human behaviour, whilst in turn, it is the behaviour of agents that reproduces or transforms institutions over time. Institutional life, then, is not about dull conformity or blind compliance; agents are not simply ‘rule bound’, nor completely locked into the trajectories shaped by path dependency. As Mahoney and Thelen suggest, rule imprecision, ambiguity and degrees of discretion during rule enforcement or implementation all open up space for agents.Footnote 83 Ideas also matter because institutions define roles not final behaviours, and roles always need to be interpreted. Streeck and Thelen argue that within institutional settings: ‘applying a general rule to a specific situation is a creative act … the meaning of a rule is never self-evident and is always subject to and in need of interpretation’.Footnote 84 As is well known, actors can also be very creative in interpreting or bending rules or in finding loopholes to suit their agendas. Sometimes, whole careers in law or tax accountancy can be devoted to such tasks. This underlines the argument made by North that the informal institutional arrangements worked out on the ground by agents are an important component of institutional life.Footnote 85 In applying this reasoning to the China case, Tsai argues that ‘adaptive informal institutions’ or ‘informal coping strategies’ can play an important role in endogenous institutional change: ‘the etiology of formal institutional change may lie in the informal coping strategies devised by local actors to evade [or modify] the restrictions of formal institutions’.Footnote 86 As Bell and Feng argue in explaining the rising authority of the People's Bank of China in recent decades, the leaders of the bank turned such practices of informal institutional navigation into an art form in a range of policy arenas.Footnote 87 It is also the case that agents often confront complex institutional agendas or operate in multiple institutional settings. This can confront agents with multifarious or even competing rules and agendas or ‘friction among multiple political orders’, all of which potentially opens up space for discretionary behaviour.Footnote 88 Agents thus have a degree of ‘bounded discretion’ within institutional environments, implying scope for contingent degrees of discretion and manœuvrability. This is a point well understood in the voluminous literature on policy implementation and the discretion afforded to ‘street-level bureaucrats’ or other types of ‘rule takers’. It is for these reasons that Crouch calls for a ‘post-determinist’ institutional analysis.Footnote 89
Thirdly, a further innovation is to see institutions not just as sources of constraint but also as having important empowering and enabling effects which interpretive agents may be able to exploit. As Scharpf puts it, institutions both ‘enable and constrain’ actors.Footnote 90 Agents actively interpret their situation and weigh up the costs and benefits of change, based partly on institutionally contingent assessments of resources and capabilities. Institutional dynamics also involve power struggles as actors exploit their institutional positions and deploy resources to win battles and reshape their institutional environments. Indeed, Mahoney and Thelen argue that ‘what animates change is the power-distributional implications of institutions’.Footnote 91
Implications of this Approach
This emphasis on the importance of interpretation, agency, discretion and empowerment within institutional settings has a number of implications for how we approach a series of conceptual debates and common understandings within institutional analysis.
For example, the approach advocated here questions the high levels of determinism typically associated with path dependency perspectives and focuses more directly on underlying historical mechanisms, such as institutional legacies or increasing returns, and assesses how agents deal with such contingencies when they shape and drive institutional change. As Johnson argues, ‘path contingency’, rather than the more determinist path dependency, may be a better way of framing the issues at stake.Footnote 92 Harty also makes a useful contribution by arguing that too often institutionalists have emphasized the costs of institutional change implied by shifting off the paths associated with increasing returns.Footnote 93 As noted above, one escape route in terms of explaining change in such sticky accounts has emphasized exogenous shocks which reduce the costs of change and provide ‘windows of opportunity’.Footnote 94 But this, according to Harty, has left institutional accounts with a ‘coherence’ dilemma: institutional logics are used to explain institutional persistence but not change, with the latter explained by reference to exogenous factors.Footnote 95 Harty argues that this dilemma can be addressed by introducing an agency-centred, ‘resource-based’ approach which investigates how agents within institutions might use rules, institutional sources of authority, expertise or other institutional resources to help effect change and offset the potential costs of moving off the path.Footnote 96
Another implication of a more explicitly agent-centred HI approach is that it allows us to move beyond a series of dualisms that we find in constructivist and ‘sticky’ HI approaches to institutional change. For example, both approaches set up a dualism between exogenous and endogenous sources of change. The constructivists, for example, accuse HI of not having an endogenous account of change. Yet the more flexible, agent-centred HI approach advanced here can embrace endogenous change dynamics involving agent-centred strategic action set within a field of wider exogenous, though institutionally mediated, forces. Endogenous and exogenous factors can thus be linked. Moreover, even a supposedly exogenous account of change – such as the impact of a crisis – cannot in any reasonable sense be thought of as just that. Only agents operating within institutions can produce change; hence, it is hard to see how exogenous and endogenous change dynamics can be separated. As Blyth explains, ‘Theoretically, no exogenous factor can in and of itself explain the specific forms that institutional change takes’.Footnote 97 Only agents can do this and accounts of institutionalism, including historical institutionalism, should have accounts of how agents act within institutions. It thus seems unwise to draw a sharp line between exogenous and endogenous factors in institutional analysis. Exogenous forces can shape endogenous process. It could even eventuate, as Harty points out, that an exogenous shock might itself be precipitated by factors that are endogenous to the institutional system.Footnote 98
It is also important to move beyond a further dualism that is a feature of both constructivist institutionalism and sticky HI approaches. Both approaches set up a distinction or dualism between change and stasis.Footnote 99 The constructivist approach sets up its own allegedly ‘dynamic’ account of institutional change against the allegedly ‘static’ account given by historical institutionalists. Hay, for example, emphasizes ‘new institutionalism's characteristic emphasis on inertia’, proffering constructivist institutionalism as a means to ‘overturn’ this.Footnote 100 This dualism is reflected in the tendency to see institutional change as a crisis-induced rapid shift or as an abrupt episode of ‘punctuated equilibrium’, typically followed by a period of stasis: hence, there is sequencing between change and stasis. Katznelson links this to agency when writing about periods of crisis as those when ‘constraints on agency are broken’, thus opening up new opportunities.Footnote 101 By setting up a dualism between change and stasis, such approaches also set up a dualism between agency and structure. Crises allow for agency, while ‘normal’ institutional life does not, or at least, radically limits it. Arguably, however, evolutionary and more radical change can be handled similarly. The level of actor discretion may increase under crisis conditions when existing arrangements might be seriously called into question, but as Campbell argues, agency-based processes ensure that change and innovation also proceed in evolutionary ways as well through ongoing actor engagement with institutional environments.Footnote 102 This is certainly a more agency-based view of politics and change than granted under more determinist path-dependent approaches to institutional analysis. As is argued by Thelen, Streeck and Thelen, Campbell, Mahoney and Thelen, institutions are constantly changing in an evolutionary mode where small progressive changes can end up producing substantial changes over time.Footnote 103 Also, what looks like ‘institutional stability’ may actually involve or rely on ongoing processes which ‘include a major dose of institutional adaptation’.Footnote 104 Hence, if we see various forms of incremental institutional adaption, such as layering, conversion or bricolage as constituting institutional change, then, as Harty points out, the line between stability and change is quite blurred.Footnote 105 Moreover, in this view, it is the stable components of institutions that may provide the context in which other parts change.
Agents, institutions and structures
The Dialectics of Institutional Change
A further element of an appropriate HI account is that it must be able to model the interaction between agents and institutions and wider structures and engage in a meaningful way with the agency-structure debate. To explain change and to move beyond sticky versions of HI, we need to model agents both as partially constrained by their immediate institutional contexts and also as operating in institutional and structural settings that constantly evolve and potentially open up new opportunities for agents. Thelen has argued, ‘institutional arrangements cannot be understood in isolation from the political and social setting in which they are embedded’, whilst Pontussen has criticized historical institutionalism for focusing too narrowly on ‘intermediate level institutions’ and ignoring wider structural environments.Footnote 106 Hence, it is useful to distinguish between institutional and wider structural contexts, in part because a good deal of institutional theory tends to conflate these or use the terms interchangeably. Distinguishing between institutions and structures is also a useful contribution to agency/structure debates, particularly in showing how institutions can act as key mediating influences between agents and wider structures. To the extent that structural factors have been considered in institutional theory, they have been modelled as crises which propel change, as external political factors which reinforce path dependency, or as changing ‘environmental conditions’ that alter institutional power distributions.Footnote 107 However, the scope of such approaches can usefully be expanded and also sharpened.
Structural contexts can include the broader political, economic or social environments that operate in a ‘strategically selective’ manner, establishing incentives or disincentives or other rationales for action that may lead agents to favour certain developments or choices over others.Footnote 108 Although ‘structuralist’ accounts in political analysis are often handled in a way which implies deterministic patterns of constraint, in reality, as shown empirically below, structures can both help constrain and empower agents. Social, economic and political structures, such as the age distribution in a population, the structure of a national economy, or the distribution of power in a given polity, shape and are shaped by agents over time. The pattern of constraint or enablement embedded in such relationships can also change systematically over time. Bell and Feng, for example, show how economic transition and changing power structures in the Chinese party state helped shape institutional change, which underpinned the rising authority of the People's Bank of China in recent decades.Footnote 109 Since structural factors are typically the result of embedded historical processes, they arguably form a broader background context in which specific institutions operate. As Pontussen puts it, ‘the centralisation of state power might be considered a structural feature that underlies various political institutions’.Footnote 110 Structures thus have institution-like effects in that they shape the options and strategies available to agents. And as noted above, institutions will also typically mediate structural effects. Hence, from the perspective of given agents, the impact and indeed the meaning of structural forces will be shaped by the institution in which they are located. Structural changes such as tightening labour markets leading to strong inflationary impulses in an economy, for example, could mean very different things to inflation-fighting central bankers, compared to, say, a government carrying a heavy debt burden and potentially interested in monetarizing such debt.
In dealing with the dialectical interactions between agents, institutions and structures, the approach here (1) models agents, institutions and structures as being analytically distinct in the sense that each has properties that are not simply reducible to the other at any given point in time; (2) models agents as operating in institutional and structural contexts that are pre-given at any given point in time; (3) models agents, institutions and structures as operating in a dialectical, mutually constitutive relationship over time; and (4) sees institutional and structural effects as ultimately mediated and actualized by agency.
First, institutions and structures at any given point in time are not simply reducible to the actors that inhabit them.Footnote 111 The temporal dimension is crucial here. Adler argues that social analysis of current activity must take account of ‘antecedent conditions’, whilst McAnulla argues that ‘current activity and reflexivity always take place in a pre-structured context’.Footnote 112 This analytical distinction is important because it allows us to examine the mutual conditioning and dialectical interplay between agents and the wider contexts of action over time.Footnote 113 This is what gives institutions and wider structures potentially causal properties.
Secondly, Archer's ‘morphogenetic’ account of such dialectics sees agents, institutions and structures as mutually constitutive over time, producing emergent new properties via repeated cycles of interaction that are distinguishable from original elements – a process akin to political chemistry.Footnote 114 As Archer puts it: ‘Structures [and institutions], as emergent entities are not only irreducible to people, they pre-date them, and people are not puppets of structures because they have their own emergent properties which mean that they either reproduce or transform social structures rather than create them’.Footnote 115 In other words, structures (or institutions) are produced by the thoughts and actions of agents over time, but, at any given time, as argued above, institutions are not simply reducible to agents or what they think (thus avoiding the trap of intentionalism or voluntarism). As Wendt argues, ‘Social structures have an inherently discursive dimension in the sense that they are inseparable from the reasoning and self-understandings that agents bring to their actions’.Footnote 116 However, as Wendt also points out: ‘This discursive dimension does not mean that social structures are reducible to what agents think they are doing’.Footnote 117 Culpepper makes a similar point by arguing that whilst ‘joint belief shifts’ are important in institutional change dynamics, ‘this is not to say that what is happening [regarding agents] is only in their minds’.Footnote 118 Indeed, it is often the case that ‘institutional change is made possible by changing external conditions’.Footnote 119
Thirdly, this suggests that within these dialectical relations, the impact of structures or institutions is ultimately mediated or actualized by agents. As Archer puts it, ‘structures only exert an effect when mediated through the activities of people. Structures are only ever relational emergents and never reified entities existing without social interaction’.Footnote 120 For example, as Archer argues: ‘a top heavy demographic structure simply cannot constrain a generous pension policy unless and until some group, which is in a position to introduce it, does in fact advocate such a policy’.Footnote 121 This insistence on the role of active agency accords with our approach to institutional analysis and helps militate against arguments about structures degenerating into a deterministic form of ‘structuralism’.
The account being developed here thus has embedded agents as a key component of the analysis, albeit agents who are dialectically engaged in shaping and being shaped by their relevant contexts over time. Institutional or structural environments can exert potentially important, though always agency actualized, effects by imposing costs or benefits on agents, by shaping actor interpretations and preferences, the scope of bounded discretion and the resources and opportunities that are available to actors. The agents in question, as argued above, are interpretive agents using ideas to help define their interests, motives and strategies for action. They operate within fields of bounded discretion utilizing institutional resources to help enable or empower action. The emphasis on interpretive agency here implies that we question the argument of McAnulla that ‘causal mechanisms can exist independently of our knowledge of them’,Footnote 122 as it reifies structures. Structural forces may not always be immediately observable or even observable to all actors in a system. Nevertheless, they must at least be knowable and able to be appraised by relevant agents (even if incorrectly) if they are to have an effect in helping to shape behaviour or strategy.
Some Empirical Illustrations
In this section, some empirical examples of the institutionally embedded though nevertheless actor-centred phenomena outlined above (interpretation, bounded discretion and the use of enabling resources and capabilities) are explored. It is shown how agents, institutions and structures dialectically interact, with agents mediating and actualizing institutions and structures and with these, in turn, helping to shape the ideas and the scope of bounded discretion available to actors – all as part of a process of shaping institutional change. To do this, we draw on empirical examples from the politics of economic policy and central banking, largely from Australian experience.
The focus of our empirical account is mainly on agents and policy makers within the Reserve Bank of Australia (RBA), and here we are able to link interpretive agency and bounded discretion and illustrate how these were in turn shaped by institutional and/or structural environments, first in relation to the shift towards central bank independence in Australia.
After the Second World War, central banking in Australia operated under legislation that stated that the RBA was independent and that it controlled monetary policy. However, within the federal government and the Treasury, prevailing preference regarding monetary policy saw the government holding the whip hand, essentially dictating monetary policy to the RBA. This view and approach was finally overturned by the government (and the RBA) during the harsh policy-induced recession of the early 1990s. Realizing the dangers of monetary policy during and after the recession, the government gradually let the RBA take the lead in formulating monetary policy, a shift that formally recognized the independence of the RBA in a memorandum of understanding signed by the government with the RBA in 1996.Footnote 123 This illustrates changing interpretations about the appropriate locus of authority over monetary policy and the degree of discretion agents had in working in and around formal institutional and legislative arrangements. Nevertheless, it also illustrates that the shift to central bank independence was greatly facilitated by the formal legislative environment. In particular, the agents in question could pursue a relatively bold independence agenda with low political transaction costs because the necessary legislation regarding independence for the RBA was already in place. For their part, especially in the wake of the damaging recession, federal politicians had a strong (institutionally derived) political incentive to hand authority to the RBA, in part to help establish a political buffer between themselves and the newly menacing challenges of monetary policy. As a leading journalist has put it: ‘independence was a gift under duress from the politicians who felt, post recession, that distance from the interest rate levels wasn't such a bad option’.Footnote 124 In these ways, the institutional environment helped shape the preferences and agendas of key agents in government and the RBA. Indeed, given the recessionary crisis and state of the RBA's credibility in the early 1990s, it would have been very unlikely that the government would have publicly supported independence or been brave enough to legislate to this effect at that time. But the fact that the legislation already existed made it possible for the authorities to implement institutional change on an informal basis and tacitly give the RBA its head during this period.
Our empirical story also revolves around other major policy and institutional changes during the 1980s and 1990s. For example, the Australian dollar was floated in late 1983 in the wake of sustained market pressure. In principle, this implied a more relaxed policy approach to the value of the dollar on the part of policy makers. Henceforth, markets would price the dollar, factoring in views about relevant economic fundamentals and sentiments about the soundness (or otherwise) of government policy settings. The theory behind the float was also partly shaped by the view that market shifts in the dollar would shoulder the burden of economic adjustments. Indeed, orthodox economic theory suggested that a flexible dollar would help deal with problems such as current account deficits that had historically been of concern to policy makers in Australia through the equilibrating mechanism of import/export price adjustments.
In practice, however, such sanguine outcomes did not occur and after the float the authorities remained fixated on the level of the current account deficit (CAD), which widened substantially in the 1980s, doubling from earlier averages to 4 per cent of gross domestic product (GDP) and peaking at 6 per cent in 1986 and again in 1989. Both episodes constituted current account ‘crises’ in the eyes of policy makers.Footnote 125 This fixation was primarily driven by the fear of the structural impact of a falling dollar if financial markets lost confidence in domestic policy settings and the management of the CAD. By 1985–86, this scenario was unfolding as market reactions produced a substantial depreciation of the dollar. In 1986, official concerns were such that Treasurer Keating famously declared Australia could become a ‘banana republic’ if solutions were not found to the CAD and dollar problems and underlying structural issues.
Through the 1980s, monetary policy, the key discretionary macroeconomic policy instrument, remained focused on the CAD.Footnote 126 At one level, this fixation illustrates the importance of constructivist insights as well as insights about bounded discretion. Despite the fact that the institutional regime had been altered fundamentally by the 1983 float, the mindset of the policy authorities remained locked in a prior era of fixation on managing the current account and the value of the currency. In essence, there was a lag in ideational construction appropriate to the new flexible rate environment. This can partly be explained by cognitive factors, especially the slowness of the authorities in learning how to handle a floating rate regime. But perceptions were also shaped by wider structural factors. For one, there were problems with the degree and speed with which price adjustments under a floating rate regime would help stabilize the current account. Compounding this were concerns about the falling terms of trade as well as about the structural weakness of the economy, especially weakness on the export front and the economy's high import propensities. This structural position helped highlight concerns about the CAD and helped shape the relations between policy makers and another structural dimension – the power of financial markets and their control over the dollar. Here, the more the authorities worried about the CAD, the more vulnerable they felt in relation to the markets and their expression of structural power. Indeed, Bernie Fraser, the governor of the RBA, devoted an entire speech in 1990 to affirming the commitment to tackle the CAD, arguing that worsening current accounts and growing external debt made Australia excessively vulnerable to adverse market reaction. ‘Confidence, as we all know, is a fragile thing,’ he warned. ‘Even countries without large foreign debts can be subject to adverse re-assessments by international markets.’Footnote 127
However, this mindset changed markedly soon after Fraser's speech, partly reflecting the impact of new policy ideas and also the way in which ideas were being shaped by the impact of changing institutional and wider structural conditions in the economy. An important structural shift was provoked by the accidental policy-induced recession of the early 1990s in the wake of the high interest rates of the late 1980s that were aimed at quelling a post-deregulation asset price boom and a ballooning CAD. This is a further illustration of the links between interpretive agency and bounded discretion and again how these can be shaped by changing environments. The early 1990s recession may have been a policy-induced blunder but the recession was also, as Harty points out, an exogenous shock partly precipitated by factors that were endogenous to the institutional system itself.Footnote 128 Central here had been the RBA's inept handling of the process of post-1983 financial deregulation that had produced wild asset price inflation by the late 1980s. Also at work was a process of ‘ideational path dependency’, where previous experience had shaped ideas along a given path and which led policy makers to think that high official interest rates could be followed by a soft landing.Footnote 129 The high official interest rates of 1986 amidst the ‘banana republic’ crisis and the subsequent soft landing of 1987 had conditioned policy makers to expect the same outcome after the high rates of 1989. Instead, the economy crashed into a deep recession. These institutional and policy legacies ultimately helped produce the recessionary crisis and, amidst this, the RBA's leaders developed novel interpretations and exercised considerable discretion in fashioning a new approach to monetary policy, one focused much more than previously on fighting inflation. Indeed, a radical decision was made within the RBA in late 1990 to chase the economy down and prolong the recession in order to exact a king-hit against inflation.Footnote 130 Importantly, the RBA's legislative mandate was flexible enough to afford this degree of discretion, but just as importantly, the structural context of the economic crisis provided incentives to take bold moves in this direction in order to try and salvage at least some policy gains from the recession.Footnote 131
The recession also became a battleground within the RBA and more broadly within the elite policy circle and provides a key insight into the way in which institutions provide resources that can empower (and not just constrain) agents and hence facilitate agency and institutional change. Amidst the recessionary crisis, the monetary policy authorities broadly decided to attack inflation. But importantly, the politics of this process was mediated by pre-existing institutional legacies. The monetary hawks within the RBA (and the Treasury and opposition) wanted an even more determined push on inflation than that which actually occurred. Importantly, those pushing for a more ‘balanced’ approach (including the then Governor Bernie Fraser) were able to utilize the RBA's institutional legacy and statutory foundations to support their position. In particular, the RBA's post-war Keynesian-inspired dual-goal legislative charter, which mandated the goals of low inflation and full employment, was used strategically by Governor Fraser to help counter the inflation hawks around him and to steer a (somewhat) more moderate course for monetary policy. In other words, pre-existing institutional resources were used strategically to help empower relevant actors. In reflecting on the pace of interest rate reductions in the early 1990s, Fraser stated in an interview:
We should have moved faster to cut rates yet there was a lot of resistance to moving even as fast as we did. This is where the dual goals become important because if we didn't have them and were stuck with inflation as the sole objective, as in the New Zealand or European model, it would have been easier for those in the Bank who were uncomfortable about moving as rapidly or as often as we did to lower interest rates to point to such a single objective (i.e. a low inflation one). I was able all the time to counter with the argument that we were also legally charged with a concern about growth and employment. Without the dual goals I don't have any doubt at all (particularly given the media and political criticism of many of the rate cuts) that it would have been very much harder to make those rate reductions. I think it made an enormous difference to have those dual goals.Footnote 132
Fighting the CAD with monetary policy in the run-up to the recession had clearly proved to be a risky business and this had a sobering effect on this front on policy makers. But, just as significantly on the ideational front, an influential paper published in 1989 by a senior economist (John Pitchford) had suggested that the authorities should simply stop worrying about the CAD.Footnote 133 Pitchford argued that the CAD was primarily driven by private external debt obligations that were largely servicing productive domestic investment and that the authorities should not be trying to second-guess such market outcomes or potential market reactions to them.Footnote 134 Pitchford's sanguine ‘consenting adults’ view about external private debt and the CAD reflected the increasing salience of neoliberal views about markets and economic policy; views that were becoming increasingly influential in the top policy circle in Canberra and at the RBA in Sydney.Footnote 135 In fact, the RBA led the push to revise official thinking. As the RBA's then deputy-governor put it, the ‘mind-set’ within the Bank began to change.Footnote 136 The RBA finally convinced Treasurer Keating, who was soon to become prime minister, to stop fixating on the CAD. Partly as a result of these interventions, the authorities stopped anguishing about the CAD and stopped trying to use monetary policy to target it, instead shifting the monetary policy focus to fighting inflation.Footnote 137 The ideas of policy makers were thus a crucial element in interpreting reality and in shaping policy options.
However, and crucially, this constructivist take on developments is not the whole story. We also need to model agents and their constructions of reality as also a product of interactions with relevant institutional and wider structural environments.
Hence, by the mid-1990s, it was apparent that an important structural shift was occurring in the Australian economy. First, there was a shift towards far stronger terms of trade for Australia as commodity exports (especially to a booming China) strengthened. This was enough to reverse a process of declining terms of trade that had been under way for decades in Australia. It was a profound structural shift that further emboldened policy makers regarding the CAD and helped to alter mindsets and the stance of economic policy. Even during episodes when the CAD rose to historically high levels during the 1990s, the policy authorities remained sanguine, pointing out that mounting debt servicing obligations could be handled by rising export incomes and the strength of Australia's terms of trade. Secondly, a series of major institutional changes including progress in reducing inflation, and a series of major microeconomic reforms in tariff protection, the labour market and other areas helped increase productivity growth, make the economy more flexible and resilient to shocks, and better placed to deal with currency fluctuations, in part by reducing the ‘pass through’ effects of a falling currency on domestic prices and inflation. In this manner, the lowering of concerns about the CAD from the early 1990s placed the government and the authorities in a new and more powerful position in relation to broader structural forces, especially the power of the currency markets and their control over the dollar. This is an important illustration of how changing mindsets and material conditions can alter both perceptions about, and indeed the constitution of, prevailing power structures.Footnote 138
These new relationships were illustrated graphically during the Asian financial crisis of 1997–98, which in effect proved to be a contest between the authorities (especially at the RBA) and the global currency markets. In this context, the RBA's leaders developed novel interpretations, exercised policy discretion and challenged earlier mental straitjackets (especially fixation on the dollar). Market concerns about the potential fallout of the crisis in Australia, particularly on the export front, were reflected in a depreciating currency which fell from the high 70 cent mark against the US dollar to around 56 cents. The RBA's leaders could have attempted to defend the currency, ward off potential imported inflation, and attempt to appease the markets by adopting higher interest rates, but they chose not to.Footnote 139 This was in contrast to the more orthodox approach of defending the currency through higher interest rates, as adopted in New Zealand (and which led to a domestic recession). Crucially, interest rates were not raised during the crisis and domestic growth and employment were protected. John Edwards, a well-known former Sydney-based bank chief economist, comments that Governor Macfarlane did well: ‘A more easily rattled Governor, someone with less monetary experience, someone with more reliance on models and theories and less on accumulated wisdom, would quite easily have cost Australia billions of dollars in lost output and a hundred thousand jobs’.Footnote 140
In fact, the RBA was well aware that a good deal of the market making was the result of aggressive, speculative action by hedge funds that had already wreaked havoc in East Asia. The RBA's leaders were anxious not to see this happen in Australia and were keen to stand up to the hedge funds and their efforts at short-selling the dollar. As it turned out, the RBA's views about the economy's strong fundamentals prevailed and the hedge funds lost their bets. In this context, the net effect of the RBA's responses was to cushion the impact on the domestic economy and absorb the short-term shock on the exchange rate. This is exactly the flexible response that a floating rate regime is supposed to facilitate, but it had taken fifteen years since the dollar float for this mindset to be fully locked in.
Conclusion
Some time ago, Kathryn Sikkink complained that political scientists, ‘whose entire existence is centered on the production and understanding of ideas, should grant ideas so little significance for explaining political life’.Footnote 141 Things have now changed and certainly Hay, Blyth and Schmidt are right to highlight the importance of constructivist insights. They are also right to critique sticky versions of HI and to argue for a greater emphasis on agency. However, their heavy ideational emphasis, at times veering towards postmodernism, ends up displacing a focus on ‘situated agents’ operating within institutional settings which both enable and constrain such agents. Hence, what Colin Hay has called ‘voluntarist idealism’ is not the way forward.Footnote 142 Agents cannot simply ‘make up’ their realities in ways that are disconnected from the environments they inhabit at any given point in time. Try telling this to the authorities in the United States as they faced the real prospect of a complete collapse in the American banking system if swift bail-out action of some kind was not taken. Agents interpret their realities and make choices, but not in ways wholly of their own choosing. As Marx once argued, ‘Men make their own history, but they do not make it as they please; they do not make it under self-selected circumstances, but under circumstances existing already, given and transmitted from the past’.Footnote 143 Institutional and structural environments are central to this process and inevitably help shape agents’ identities, interests, calculations and choices. Grounding a constructivist approach within a suitably tailored historical institutionalism which is sensitive to the impact of institutions and wider structures and which can produce more rounded accounts of agency is arguably the way forward.
We should thus reject the claim that HI is incapable of explaining institutional change or inevitably ends up with sticky, overly static accounts of institutional life. The fact that there are two extant versions of HI is often overlooked. HI has both a sticky and a more flexible, agent-centred version. The latter can easily accommodate an agential focus and key constructivist insights, but as argued here, it is crucial to locate agents operating within a dialectical relationship with their institutional and wider structural environments. This is why institutional approaches were developed in the first place. Whether the synthesis proposed here of modern constructivism and an agent-centred historical institutionalism ends up being labelled as historical institutionalism or as constructivist institutionalism is not particularly important. What is important is the appropriate synthesis of explanatory elements.
