Evolutionary approaches to the concept of drift in policy studies

Adrian Kay, Darrin Baines
2017 Critical Policy Studies  
Evolutionary approaches to the concept of drift in policy studies Within the important seam of policy studies dedicated to understanding how institutions and policies change over time, attention is increasingly being paid to cataloguing patterns of change in the absence of 'big' reform (Béland 2007 (Béland , 2010 Béland, Rocco and Waddan 2016). The ambition is to explain policy change by reference to factors which are endogenous to the institution or policy as a counterpoint to accounts that
more » ... y on exogenous shocks as correlates of observed 'big' changes. The concept of policy drift set out in Hacker (2004; 2005) has been widely referenced in this line of work as an intuitively appealing label for an empirical pattern seemingly observed across many different policy sectors: 'changes in the operation or effect of policies that occur without significant changes in those policies' structure' (Hacker 2004, 246). Whilst the policy does not change, its consequences do as the external environments in which the apparently stable policy operates shifts gradually. Although evolutionary approaches are not explicitly acknowledged in this line of inquiry, in this paper we argue that they have much to contribute. First, evolutionary thinking helps clarify drift as a distinct concept from those which are often co-listed as competitors for the analysis of cumulative, sub-surface and endogenous policy change: displacement, layering, conversion or exhaustion (Wincott 2013). Second, an evolutionary approach casts light on the role of agency in drift sequences, something poorly understood and articulated in the drift literature (Shpaizman 2017; Rocco 2017). It remains unclear whether policy drift should be understood as something caused directly by a political strategy employed by an influential policy actor, or, rather, as a policy sequence that is unintended by any particular actor in the policy process. This paper presents the argument, using an evolutionary metaphor, that it is the latter, leaving an important but necessarily separate question of the agency of policy actors: to recognise their interests in, and capacity to develop, appropriate responses to the consequences of policy drift. The third contribution of the evolutionary metaphor is to this question of agency: it extends the existing drift literature beyond a simple dichotomy of maintaining or reversing drift by identifying acclimatisation and adaptation as distinct agential responses to drift in the policy environment. The paper is structured to develop the argument that the nascent literature on evolutionary approaches in policy studies can help bolster our understanding of drift. The first section outlines, in brief and introductory terms, the application of evolutionary thinking to public 2 policy. The next section contributes to our analytical repertoire for drift patterns by using an evolutionary metaphor for drift to develop a three-step analytical framework to highlight limitations in the existing drift literature on questions of agency. The penultimate part of the paper develops the case of pharmaceutical services policy in the UK to illustrate empirically the value of this analytical framework for the application of the concept of drift in policy studies. The concluding section discusses the key points raised in the paper for future research into policy drift. EVOLUTIONARY APPROACHES IN POLICY STUDIES Although the terms 'evolution' and 'evolutionary' are widely used in the social sciences, for the most part they are employed as convenient short hands for slow and gradual change over time. To introduce evolutionary thinking more analytically into policy studies, evolution can usefully be stated as a descriptive label, or metaphor, for a process of change as distinct from biological evolution as a theory of change (John 1999; Kerr 2002; Cairney 2013). By doing so, one of the main criticism of evolutionary approaches in the social sciences -that biological analogies in the social world are misleading because there are no equivalents of natural selection in society -can be avoided. Instead, evolutionary perspectives are something useful for the analysis of all open, complex systems, of which natural systems are an important but not unique example (Campbell 1965). The idea of punctuated change has had a strong grip on thinking about evolution as a process in the political world of policy-making (Cairney 2013). For example, workhorse models of the dynamics of public policy by Kingdon (1984 Kingdon ( , 1993 and Baumgartner and Jones (1993 ) have used the notion of punctuated change; there are moments, windows or critical junctures where change is observed. Although these do not represent a full evolutionary theory or a detailed and contextualised application of evolutionary concepts in the social sciences, they do acknowledge their borrowings from evolutionary biology. Evolution serves well as a metaphor in policy studies. The claim that frameworks or theories act as metaphors often implies weakness and a lack of substance. However, as recent philosophers of social science have noted, metaphors can have a deeply constitutive, if subterranean presence, in the formation of concepts and analysis of change (Lakoff and Johnson 2003; Lewis 1996; Klamer and Leonard 1994). Indeed, metaphors are prior to any analogy; they establish broad and general mappings across conceptual domains and help structure understanding and assist in the perception of connections between different things. 3 In turn, this supports the formation of concepts to organise things into different categories. (Lakoff and Johnson 2003). In terms of existing studies of policy dynamics, evolutionary metaphors are a useful way of organising thinking about agency in a shifting policy environment by distinguishing two distinct processes: acclimatisation (learning by agents as they update their calculations about how to realise interests most effectively) and adaptation where agents' beliefs, identities, and interests are learned or constructed in their interaction with other agents and the broader policy environment in a process of self-organisation . In terms of adaptation, the literature on discursive institutionalism -although not explicitly evolutionary -is useful in highlighting how normative and cognitive change may occur as the problem situation itself is reconstructed and reframed in the interaction between agents and their environment (e.g. Carstensen 2015). This separate identification of adaptation in drift is a useful step in developing understanding the role of the agency in drift sequences that have not been reversed. Just as policy problems, policy solutions and criteria of 'success' or 'failure' are all constructed (Marsh and McConnell 2010), so recognition of drift and deliberation on its consequences underpin agent intentions to react to drift. Kaufman (1993, 644) argues that analysis of these sorts of dynamics of self-organisation should complement consideration of selection mechanisms in evolutionary thinking. In terms of a policy studies metaphor, the role of agents in evolution is not just ad hoc fiddling or myopic bricolage; rather it involves acclimatisation as well as adaptation as attitudes, beliefs and strategies change in the policy process as an emergent order from the operation of selection mechanisms, which operate to support or retard development of new policy interests, normative and cognitive ideas and strategies. This helps extend understand of the different forms of agency in drift sequences beyond simple reversal; a recognised gap in the drift literature (Shpaizman 2017; Rocco 2017). EVOLUTIONARY APPROACHES TO THE CONCEPT OF DRIFT Evolutionary approaches in the social sciences have often been elided with criticisms of functionalism (Kincaid 1996) . To circumvent this criticism requires describing selection mechanism that do not operate in a way that 'selects' behaviours or institutions that produce favourable consequences for some agent as that would amount to functionalism. This presents the dilemma of wanting an evolutionary metaphor to have some analytical value but also avoid functionalism. Kerr (2002) asserts that institutions or behaviours or ideas that have 4 consequences which are beyond the 'strict limits or 'selective ' pressures' (Kerr 2002, 351) set by the environment will not tend to occur. As a result, 'this forces individual actors or groups of agents to negotiate, and 'adapt' to, the context in which they are situated.' The basic argument is that because dysfunctional elements tend to disappear, persisting forms can be assumed to be adapted in the sense of not being dysfunctional. It amounts essentially to the assertion that non-dysfunctional institutions maintain themselves over time, because they do not transgress environmental limits or constraints. Some might see this position as vulnerable because there may be many non-dysfunctional alternatives to a given dysfunctional institution. Without the ability to say which of them will emerge, and at what time, rather than simply that one of them ultimately will, the predictive capacity of an evolutionary theory is limited. In terms of explanatory power, evolutionary theory is anaemic without functionalism. However, there is good reason for this lack of explanatory power. The ambition for dynamic theories of public policy is for theories, methods and concepts to produce intelligible narratives that involve highly contingent combinations of factors, and not universal theory of all policy change. In terms of evolutionary theory, there may be different selection pressures operating at different levels and over different time horizons. There is no single uniform and universal mechanism that uniquely selects and that can be accounted for by a covering law theory. Instead, there is a multitude of overlapping and potentially conflicting environmental pressures over the short, intermediate and long term that possibly change the impact of a public policy absent any formal policy change. The notion of selection as environmental pressure that produces a disposition or a tendency is similar to contemporary understanding of selection in the field of complex systems and evolutionary economics (Potts 2000; Dopfer 2012; Hodgson 2014) . In these terms, selection mechanisms are not universal fields that operate consistently over time but rather are often local, relative and operate over a specific period. Thus within the environment of the policy system there are dispositions, tendencies and constraints that limit what policy or policy proposal or idea or advocacy coalition (and so on) may be successful. Drift contributes initially to understanding policy change over time by bringing selection pressures -reinforcements and reactions -to the fore in a critique of the conventional account of stability in policy studies. This is the argument that institutions and policies maintain themselves over time because they do not set up large endogenous pressures for change, or, 5 in other words, they are functional. On this account, stability and functionality are in a constitutive relationship in public policy: functionality is an essential and necessary characteristic of any stable policy. The drift concept can be used to cast doubt on this argument in several respects. First, the basic insight of the stability account, that non-dysfunctional institutions and policy remain stable over time, is dubious: apparently stable policy may well be subject to drift. Second, even if dysfunctional institutions and policy do tend to disappear in the policy-making system because of some selection mechanism operating in the policy environment, there is no presumption that they will be replaced by stable, functional ones unless we can discover empirically a selection mechanism that winnows out dysfunctional institutions. In general, the policy drift literature honours the Hacker (2004 Hacker ( , 2005 account as seminal in the field;
doi:10.1080/19460171.2017.1414618 fatcat:jhffxr7e2bhepn3hcqnsdgggya