The power of transformative ideas
Once we allow for institutional ambiguity and cross-policy incoherence being exposed by endogenous and external pressures, we are obliged to pay attention to how policy actors make sense of the policy environment they inhabit and how they weigh the policy options available to them (Hall, 1989, 1993; Ross, 2000, 2008; Campbell, 2002; Schmidt, 2002, 2008, 2010; Beland, 2005, 2007; Somers and Block, 2005; Taylor-Gooby, 2005; Starke, 2008; Beland and Cox, 2011). While the scholars of incremental institutional change continue to abide by material interest-based preferences, ideational accounts quintessentially allow for more fluid cognitive orientations and normative preferences, often conceptualized as 'switchmen', to use Max Weber's phrase (Weber, 1964: 280), 'policy paradigms' (Hall, 1993,1989), 'focal points' (Goldstein and Keohane, 1993), 'frames of reference' (Jobert, 1989), 'strategic hooks' in the battle for political control (Blyth, 2002), 'collective memories' (Roth- stein, 2005), and 'discourses' (Schmidt, 2008). Scholars in the tradition of the 'new politics' of the welfare state and also researchers in the school of incremental but transformative change quintessentially reason from institutions to actors. The central focus is on how existing institutions provide powerful incentives to policy actors to replicate, rather than adjust and adapt, prevailing institutional regimes. A similar critique can be levelled against the 'varieties of capitalism' literature as it reasons from the perspective of institutional advantages of liberal and coordinated market economies (Hall and Soskice, 2001; Hall, 2005; Eichengreen, 2007b; Hancke et al., 2007; Martin and Thelen, 2007; Martin and Swank, 2008).
Taking key policy actors as the point of departure, almost 40 years ago, Hugh Heclo (1974) was the first to direct our attention to the dynamics of social learning in the welfare state, driven by the complex interplay of political competition and expert consultation in the policy process (Heclo, 1974: 320). By diachronically comparing British and Swedish welfare state innovation, Heclo was able to underscore the critical importance of high-level expert elite networks of 'policy middlemen' in providing a 'vital new source of domestic and international intelligence' to the design of modern social policy (1974: 319). Fundamental to Heclo's conception of policy learning is uncertainty: 'Politics finds its sources not only in power but also in uncertainty—men collectively wondering what to do' (Heclo, 1974: 305). Consequently, he defined policy learning as ‘relatively enduring changes in thought or behavioural intention that result from experience and/or new information with the attainment or revision of policy objectives' (1974: 306). This definition underscores three important dimensions: reflexivity, information feedback, and political motivation. Policy actors are necessarily ‘reflexive', suggesting that they are able to creatively diagnose problems and envision policy alternatives for emergent problems (Mantzavinos, 2001). Political actors appreciate the policy environment in relation to problem situations and stated policy objectives under conditions of what Herbert Simon has called 'bounded rationality'.
'Bounded rationality' suggests a world too complex for actors comprehensively to establish the most appropriate means to ends without much delay (Simon, 1985).
The emphasis on information feedback denotes that learning is not a passive exercise of deduction from experience, but an active mental process of interpreting pressures and modifying cognitions, beliefs, and desires. In the terminology of Fritz Scharpf's actor-centred institutionalism, information feedback critically bears on both cognitive and normative orientations (Scharpf, 1997). Cognitive orientations are causal propositions about the world held to be true, which involve a 'programmatic set of statements about cause and effect relationships and a method for influencing those relationships' (Weir, 1992). They guide policymakers by specifying causalities upon which they build their collective action strategies. Normative orientations are preferences for what the world should be like and thus can be considered as principled desires and justifications of policy actors. Together, cognitive and normative orientations constitute 'compelling reasons' for making particular choices (Hedstrom, 2005). Policy choice is never politically neutral. It inevitably involves normative judgement over which combination of policy measures is most appropriate in the light of political objectives, moral aspirations, and positional interests.
Policy learning, in most descriptions, is based on a process of trial and error, in which only the most pressing problems are diagnosed and alternatives are debated and adopted until 'satisfycing' solutions are agreed to (Simon, 1945). The emphasis is on widening the repertoire of choice through information processing, 'to detect and correct errors' (it is hoped) in order to 'improve the functioning of systems' (Olsen and Peters, 1996: 4). Policy actors are strategic, seeking to realize certain political objectives. But, uncertainty, reflexivity, and bounded rationality complicate the search for feasible best alternatives. The search for alternative policy solutions is initiated when existing practices are perceived to be unsatisfactory for solving impending problems (Cyert and March, 1963). Most proponents of policy learning (Deutsch, 1963; Heclo, 1974; Etheridge, 1981; Rose, 1991; Bennett and Howlett, 1992; Hall, 1993; Sabatier and Jenkins-Smith, 1993; Olsen and Peters, 1996) argue that policymakers 'draw lessons' from policy experience and apply new knowledge when dealing with protracted policy failure. Misconceptions are liable to lead to adverse unintended consequences. These errors are not fatal, however, as long as information feedback remains possible, allowing mistakes and unintended consequences, as catalysts, to be corrected in time. Quintessential is that reflexive actors faced with problem situations, characterized by high levels of uncertainty, are highly unlikely to continue to abide by standard conceptions of utilitarian rationality based on fixed preferences and stable material interests.
The fundamental political question to answer is, what triggers or motivates actors to learn? Where does the readiness to subject pre-established policy ideas to new insights and critical information of policy experience come from? Because democratic politics embodies a struggle over ideas and preferred courses of action, I maintain that evaluation, assessment, deliberation, judgement, and learning are inherent to the political process. Policy problems must be brought to public attention, which also depends on the nature of the political process and the role of the media. In affluent democracies, as Bryan Jones puts it: 'Elections are the mechanisms that enforce that responsiveness' (Jones, 2001:170). Through electoral competition, proponents of rival policies will be self-interestedly motivated to find fault with existing or proposed policies. Particularly when existing policy repertoires fail to meet the popular expectations of the electorate, voter dissatisfaction can be a critical driver of policy reform. As long as policymakers are 'ambitious', they are pressed to attend to information feedback from policy experience and flow of ideas. Since most political judgements are largely retrospective, elected leaders have a strong motive to solve problems 'before they fester and grow' (2001: 170). Crucial is that political decisionmakers only want to learn something they do not already know if they are pressed to do so. Karl Deutsch once famously defined power as 'the ability to afford not to learn' and 'the ability to talk instead of listen' (Deutsch, 1963:111). Without the chance of losing power and authority, there is no existential political incentive for policymakers to learn.
However much policy choice is informed by 'puzzling', the question of which ideas generate the greatest impact is ultimately the outcome of the 'contest of power' (Heclo, 1974: 306). For ideas 'to have a political impact, such views had to become attached to some institution or popularly organized group', Heclo conceded (1974: 319). Moreover, policy actors never reflect only on substantive policy issues, they also reflect on the causal links between policy solutions and their power positions. In other words, there is as much 'powering' to 'puzzling' as there is 'puzzling' over 'powering' (Ferrera and Gualmini, 2000). For policy learning to materialize there must be cognitive change on the part of policy actors' understandings of policy pressures as well as behavioural adaptation as a result of this new knowledge and information. To be sure, policy learning and political decisionmaking are often disjointed because of the short time horizons, multiple goals, power asymmetries, and unclear accountability rules in the world of politics (Pierson, 2004). Improved understanding may trigger political motivation to do something, but not when solutions are seen as politically unrewarding. Problems with long gestation time, such as rising pension and healthcare costs, are often ignored for quite a while. Heclo also conceded to the possibility of 'non-learning', acknowledging that policymakers might be unwilling or unable to adapt behaviour to new insights, experience, and information (1974: 312). This could be called 'bounded irrationality'. A good example is the profound fear of inflation on the part of German policymakers, deeply rooted in the experience of hyperinflation in the 1920s, making them particularly wary about strategies of 'quantitative easing' and the issuance of Eurobonds as a resolution to the contagious eurozone sovereign debt crisis, so as to restore macroeconomic confidence in the wake of the financial crisis of 2008 and 2009 by allowing vulnerable economies to borrow at sustainable interest rates as they reform. When problems aggravate existing rather than future ways of living, on the other hand, the motivation to find quick political answers may outpace deeper understanding of looming policy problems. A typical example would be the Continental welfare regimes' shift towards early exit in the face of mounting (youth) unemployment in the 1980s and 1990s, which later proved so difficult to reverse. Ideas that have a comparative political advantage are those that offer the best causal narrative or explanation of the current situation and are at the same time compatible with widely supported policy objectives. Ideas that have a history of past usefulness and success will more likely be 'apportioned with credit', which in turn reinforces their overall political 'strength'.
Over the 1990s, empirical analyses of how cognitive and normative ideas shape processes of policy change moved to centre-stage in political institutionalism. Quite a few studies have been devoted to the shift in macroeconomic policy from Keynesian demand management to neoclassical policy analysis since the late 1970s (Hall, 1993; Berman, 1998; Braun and Busch, 1999, 2006; Hay, 2001, 2002; Blyth, 2002, 2011; McNamara, 2002). The most famous example is Peter Hall's analysis of the shift from Keynesianism to monetarist macroeconomic policy in Great Britain in the late 1970s (1993). The aftermath of the oil shocks together with accumulated policy failures with demand management prompted an 'unfreezing' process in which Keynesianism lost its prominence while monetarism provided the new intellectual template for emancipating the pursuit of low inflation and sound public finances as core policy objectives. Kathleen McNamara similarly analysed how monetarism prepared the ground for the construction of the European Monetary Union. The long experience of stagflation created a political opportunity for the centre-right to make way for a limited role of government in economy and society. Germany's economic success in the 1980s with a pragmatic version of monetarism, emphasizing a strong and stable currency, ultimately provided many European governments with a powerful example to follow (McNamara, 1998; Dyson and Featherstone, 1999).
Over the past decade or more, many social policy scholars have turned to transformative ideas and discourse analysis to explain social reform (Visser and Hemerijck, 1997; Hemerijck and Visser, 2000; Cox, 2001; Schmidt, 2002; Taylor-Gooby, 2005; Beland, 2005; Stiller, 2007; Fleckenstein, 2008). What distinguishes these studies from the analysis of macroeconomic regime change is their strong emphasis on the political construction of change imperatives and discourse, rather than the particular substance of labour market deregulation, activating social insurance, pension reform, and social services innovation, in response to structural change. Vivien Schmidt is of the opinion that political discourse matters more in welfare reform exactly because it relates to broadly shared norms and values about fairness (2002). Robert Cox (2001) found that by deliberately invoking the urgency of reform, policymakers in the Netherlands were more successful than their German neighbours in the 1990s. Daniel Beland (2007) revealed how Third Way political discourses served New Labour to shift policy attention away from burning issues of income inequality towards employment, education, and equality of opportunity. Another important insight that emerges from country-specific trajectories of social reform is the role of expert committees of policy evaluation, which sometimes break new ground for policy recommendation, in the wake of protracted policy failure. Cases in point are the 1993 Buurmeijer Commission in the Netherlands, which prepared the overhaul of Dutch social insurance administration; the 1997 Swedish non-partisan expert pension reform committee, whose recommendation stood at the basis of the 1998 Swedish pension reform; and the 2002 Hartz Commission that precipitated Gerhard Schroder's Agenda 2010 (Clasen and Clegg, 2011).
One of the characteristic institutional developments of the post-war decades of welfare state expansion is that affluent political economies have increasingly been endowed with an array of institutions that provide policy 'intelligence' of various kinds. Institutions of policy analysis have become the new 'policy middlemen' of advanced welfare states. As extra-political agents of change, institutions of policy analysis have to tread carefully to ensure that reform proposals are both intellectually and politically accepted (Baccaro, 2003; Culpepper, 2005). Preparing the ground for social policy reorientation is often a stronger attribute of think tanks, expert committees, advisory councils, planning agencies, and strategy units than of political parties or organized interests who have their ideological roots and interest-based constituencies to cater for. We therefore need a better understanding of how policy-relevant ideas, their cognitive contents and normative underpinnings, are created, selected, modified, disseminated, and sometimes ignored in the policy process. Epistemic privilege carries competitive advantage only if it is able to muster political support! It is imperative for students of social policy change to penetrate the interface of policy expertise and political deliberation, once again.