Performing the Evidence-Based Drug Policy Paradigm

Introduction

To date, the focus of much of evidence-based policy scholarship has been on finding ways to bridge the divide between two communities: “policy decision makers” and “knowledge producers.” In the field of drug policy research where the rhetoric of evidence-based policy has been enthusiastically embraced, efforts have centered on the production of “gold-star” evidence and increasing its uptake in policy and practice through “research translation” activities (Lancaster, 2014). To overcome perceived barriers, some drug policy researchers have also been dedicated to analyzing policy processes so as to better understand how evidence is utilized in this highly politicized domain and develop more nuanced models which explain the contested relationship between evidence and policy decision-making (e.g., Monaghan, 2009). However, such pursuits are underpinned by a number of presuppositions. Firstly, it assumes that policy-relevant knowledge (or “evidence”) is fixed, stable, and inherently valuable in policy (Lancaster, 2014); and that it is “both discrete and isolatable” (Wood, Ferlie, & Fitzgerald, 1998). Secondly, the preoccupation in policy scholarship with “understanding how externally generated knowledge finds its way into policy” presumes that this thing we call “evidence” sits outside of the policy process thus perpetuating the “two communities” dichotomy (Freeman & Sturdy, 2014, p. 3).

A number of researchers and theorists working across a range of policy domains and disciplines have sought to problematize the notion of knowledge as something prior and external to policy. For example, Jasanoff (2004) argues for the “idiom of co-production” of knowledge and policy. From this perspective, policy and knowledge are not distinct, but rather co-constitutive. Jasanoff contends that “knowledge-making is incorporated into practices of state-making, or of governance more broadly […] States, we may say, are made of knowledge, just as knowledge is constituted by states” (p. 3). In the realm of clinical practice, Wood, Ferlie, and Fitzgerald (1998, p. 1737) draw on poststructuralist approaches to argue against seeing evidence and clinical practice as “diametrically opposed.” Instead, they see the boundaries between research and practice as “indeterminate” and the nature of evidence itself as “ambivalent” (Wood et al., 1998, p. 1737). Wood et al. (1998, p. 1737) suggest there is a need to move toward models that incorporate evidence and clinical practice in “a more immanent relationship.” Taking this theoretical perspective out of clinical practice and applying it to policy development in multiagency settings, Green (2000, p. 472) concludes that “such phenomena as ‘knowledge,’ ‘evidence’ and ‘practice’ are not natural or necessarily distinct, but are constituted through local and contingent practices.” Evidence here is not seen to be a “distinct entity” but rather “ambivalent and negotiable”; it is practice which “constitutes that evidence” (Green, 2000, p. 472). Analyzing the concept of evidence-based medicine (EBM) and its accretion into public policy, Harrison and Checkland (2009, p. 136) suggest that the notion of “EBM” (and by extension, I suggest, evidence-based policy) involves more than knowledge; “it also entails relationships between such knowledge, the technologies through which it is codified and disseminated (such as systematic reviews and the Cochrane websites) and practised (such as medical interventions) and the humans engaged in the EBM project.” From this perspective, “EBM might be seen as a network into which ideas, technologies and humans are ‘enrolled,’ but which has to be constantly sustained through continued ‘performance’ of their interrelationships” (Harrison & Checkland, 2009, p. 137).

Each of these writers sets forth a view which destabilizes the notion of “evidence” as a fixed and stable thing existing outside of or distinct from the policy process and focuses attention on interrelationship and practice. In doing so, many of these approaches invoke the language of “constitution,” “performance,” or “enactment.” While those working within the broad church of “social constructionism” have long highlighted the constructed nature of knowledge claims about the world (Hacking, 1999), the notion of “enactment” is somewhat different (Law, 2004). Teasing the two apart, Law (2004, p. 158) argues that:

Construction usually implies that objects start without fixed identities but that these converge and so gradually become stabilised as singular in the course of practice, negotiation and/or controversy. Enactment does not necessarily imply convergence to singularity, but takes difference and multiplicity to be chronic conditions.

Thus, the perspectives outlined above suggest that what we call “evidence” is not stable and distinct but rather constituted in policy and practice. The arguments put forward by STS scholars extend this and suggest that these ongoing enactments of “evidence” lead to ontological multiplicity. In the context of drug policy scholarship, the question then becomes: how do the practices of drug policy enact (and re-enact) “evidence” and the “evidence-based policy” paradigm?

Law and Singleton (2000) offer a way forward in this examination. In their paper, “Performing technology’s stories,” Law and Singleton examine the field of technoscience. They argue that accounts of technoscience, that is, the “stories” told within the field, do not simply offer descriptions but rather perform technoscience realities. That is, accounts (stories) told by people within the field enact technoscience in particular (and different) ways. A differing empirical account (story) is not necessarily wrong, it is “a particular and located enactment or performance of technological knowledge and practice that does equally particular kinds of work” (Law & Singleton, 2000, p. 767, emphasis original). Different people working within the field “perform alternative and different understandings of the character of technological knowledge” (Law & Singleton, 2000, p. 767). Law and Singleton draw on several empirical accounts in their analysis. They highlight the ways in which the telling of technoscience stories may frame technology, for example, around the notion of “project” (Law & Singleton, 2000, p. 768). In doing so, stories do not simply describe a technological “project.” More than this, a particular story acts to “breath [sic] life into a whole set of assumptions that we might think of as ‘projectness’” (Law & Singleton, 2000, p. 768). This implies, for example:

that technologies (in part) evolve under centralised control; that they need to be managed; that if they are fragmented then this is likely to be a problem; that they involve coordinated puzzle-solving; that they benefit from a coordinated perspective; that they indeed move through stages, have chronology; that they may have setbacks that need to be overcome [and so forth]. (Law & Singleton, 2000, p. 768)

In this paper, I take the work of Law and Singleton (2000) as a springboard for analysis. Through the lens of performance, this paper critically examines how “evidence” and the “evidence-based policy” paradigm are constituted in drug policy processes, enacted through the telling of policy stories. I argue that policy stories do not simply describe the drug policy process, but rather frame the notion of “evidence” and “evidence-based policy” in particular ways. Drawing on two Australian case studies and interviews with policy makers, advocates, researchers, and clinicians involved in the establishment of harm reduction programs to extend distribution of injecting equipment through peer networks and make naloxone available for administration by overdose witnesses, I ask: What do participants’ accounts of drug policy perform? And what might this imply? Through this analysis of participants’ accounts, I argue that what we call “evidence” is not fixed, but rather constituted by specific performances and practices. I suggest that these performances of the evidence-based drug policy paradigm are important, as they work to make and sustain (or, at times, interfere with) a set of assumptions about knowledge and rationales for policy action. This, in turn, raises questions about how the evidence-based drug policy endeavor might be reconsidered and remade in other ways.

Method

The analysis in this paper is based on two empirical case studies documenting Australian drug policy processes. The first case examined the processes leading to the establishment of two recently implemented programs to make naloxone available to potential overdose witnesses in Canberra, Australian Capital Territory (ACT) and Sydney, New South Wales (NSW). Naloxone (trade name, Narcan®) is a short-acting opioid antagonist, which temporarily reverses the effects of opioids and respiratory depression. It has been used for over 40 years by medical professionals, particularly in emergency medicine, and has been shown to be safe, reliable, and effective in these settings (Dietze & Lenton, 2010). For more than two decades, researchers have argued that naloxone should be widely available to potential overdose witnesses, particularly people who inject drugs, to help prevent morbidity and mortality associated with opioid overdose (Baca & Grant, 2005; Coffin & Sullivan, 2013; Darke & Hall, 1997; Kim, Irwin, & Khoshnood, 2009; Lenton, Dietze, Degenhardt, Darke, & Butler, 2009; Strang, Darke, Hall, Farrell, & Ali, 1996). Until recently, naloxone was not distributed for peer administration anywhere in Australia, despite the accumulated international descriptive and observational evidence. In December 2011, Australia’s first overdose prevention and management program providing naloxone on prescription to potential overdose witnesses was launched in Canberra (ATODA, 2013). Soon after, in June 2012, Sydney’s first naloxone program was established through the Kirketon Road Centre and the Langton Centre.

The second case examined the development of a pilot program to extend distribution of injecting equipment through peer networks (also called “secondary supply” or “peer distribution”) in Sydney, NSW. Multiple legislative and policy barriers restrict access to sterile injecting equipment because needle and syringe programs (NSPs) operate under a series of exemption laws (Australian Injecting & Illicit Drug Users League, 2010; Lancaster, Seear, & Treloar, 2015; Legal and Discrimination Working Party of MACBBVS, 2013; NSW Users and AIDS Association [NUAA], 2009). In most Australian jurisdictions, it is ordinarily illegal to distribute or to provide injecting equipment to another person unless designated by law as an “authorized” category of persons. In recent years, there have been calls from advocates and drug policy experts to remove legislative barriers to peer distribution of sterile injecting equipment as a way of improving coverage, reducing equipment reuse, and enhancing peer education activities (Australian Injecting & Illicit Drug Users League, 2010; Australian National Council on Drugs, 2013; Legal and Discrimination Working Party of MACBBVS, 2013; NUAA, 2009). In NSW, the legal provision and distribution of injecting equipment is made possible under exemptions outlined in Section 19 of the Drug Misuse and Trafficking Regulation 2011, whereby the director-general of the Department of Health may exempt “authorized persons” to “participate in an approved needle exchange program,” which until recently had been limited to approved authorized staff and workers of nongovernmental organization and government services including NSPs and through pharmacies. In October 2013, NSW Ministry of Health implemented a pilot project with NUAA for peer distribution of sterile injecting equipment (a commitment previously mentioned in the NSW HIV Strategy 2012–2015; NSW Ministry of Health, 2012). This pilot has been made possible not by amending legislation through parliament, but rather by the director-general amending the regulation to extend “authorized persons” to include clients of NUAA’s NSP.

To inform the case studies, semi-structured in-depth interviews were conducted with individuals identified as key actors in the development of these programs. This included policy makers, advocates, researchers, and clinicians who had contributed their expertise at various stages throughout the discussion and development of these harm reduction initiatives. In total, 19 interviews were conducted for the naloxone case study (across both the NSW and the ACT programs), and 10 interviews were conducted for the extended distribution of injecting equipment case study. ¹ Interviews were on average 1 hr in length (ranging from 30 minutes to over 2 hr) and were conducted either by phone or in person. The interview guide focused on eliciting participants’ perceptions of policy activity, the processes through which policy activity occurs, and the roles of multiple knowledges and “voices.” Of particular relevance for the analysis of policy “stories” in this paper, most participants were asked: “Tell me the story—from your perspective, how did the [naloxone program/extended distribution pilot] come to fruition?” Most participants offered lengthy narrative responses to this question. Participants were also asked about the knowledge or information they found compelling in thinking about the policy issue at hand. All interviews were audio-recorded and transcribed verbatim. Participants were given the opportunity to review their transcripts for the purposes of verifying accuracy, correcting errors, and providing clarifications. Interview transcripts were organized and coded with the assistance of NVivo 10 data management software. The transcripts were analyzed through a process of close reading, using the theory to guide the coding categories and identify kinds of “performances.”

In the following section, I focus on one particular kind of “performance” identified within this analysis: the enactment of the notion of evidence as “tool-like,” to be “used.” I draw out the particular notions of the character of “evidence” and “evidence-based policy” which are enacted by participants’ accounts and, in turn, consider what these performances might imply (Law & Singleton, 2000). In doing so, I raise a set of related critical questions about the boundaries this performance places around what we might think of as “evidence-based policy” and whether, at times, it might be practiced in ways that are counterproductive to its express aims (Race, 2011). I also suggest that framing “evidence” around the notion of “usefulness” may be seen as a kind of “coordinating mechanism” (Mol, 2002, p. 117), and while this insistence on “virtual singularity” (Law, 2004, p. 58) is in some ways productive, it may also obscure the ways in which the evidence-based drug policy paradigm may be otherwise.

Performatively Constituting “Evidence” as a “Tool” for “Use”

They used the evidence that I had collected in the best way that they possibly could to put their policy position paper out there and then they used it in the way—I mean they are expert advocates so they know how to use research evidence in particular kinds of ways and they used my research evidence in the way that they knew would get them what they aimed for. (Interview 21, Researcher)

This connects to another set of assumptions about the character of evidence which is given life through this account of drug policy. This account of “evidence” as a “tool” to be “used” also suggests that evidence must be fit for purpose. One cannot hammer a nail with a screw driver. Thus, by performatively constituting the thing we call evidence in this particular way, this account also invokes the notion of “usefulness” which necessarily implies an assessment of quality, practicality, suitability, and, perhaps most significantly, instrumentality. As one policy maker said:

I work with the researchers, with community based organisations, to get the evidence that’s needed. And I know sometimes their evidence might be different to what I need. Because I have an understanding of how things work, and how decisions are made, and what sort of evidence will be considered, and the weighting of different evidence bases. So for me, it’s presenting as strong a case as possible. (Interview 23, Policy maker)

By sustaining a view of “evidence” as having an instrumental purpose, these accounts also imply that a particular kind of knowledge is superior, or more “useful,” than another. The charge that “evidence” be “useful” potentially drives out more “contextual and interpretative forms of research” (Bacchi, 2009, p. 253). As the same policy maker stated:

You can’t make policy decisions based on interviews of ten people and these are the quotes that they provide. You know, you need to look at it from a population perspective, and epidemiology comes into it. Numbers do play a part of it. (Interview 23, Policy maker)

As Law and Singleton (2000, p. 768) note, “in the abstract there is nothing wrong” with focusing on the notion of “evidence” as a “tool” for “use,” but this is not an innocent description. It performs a particular notion which arguably suppresses the multiple other ways that “evidence-based drug policy” might be imagined or enacted. For example, an account of “evidence” as a “tool” for “use” (and its implied focus on instrumental knowledge) leaves little room for other voices and knowledges to be “made as” evidence. If the character of the thing we call “evidence” is performed as “instrumental usefulness” in the telling of policy stories, and this (as I have shown above) carries with it a range of assumptions about what “usefulness” entails, then this performance pushes aside the possibility that values and emotion, for example, may be valid knowledge(s) (evidence) in policy processes. This limitation was expressed by a number of participants, as they told their policy stories:

there were a number of older women, who were the family—the parents—of dead drug users who were responding [to a survey], [telling] their stories of their experience of overdose and how it had shaped their lives many years afterwards. Their kids had died when they were adults and they’d grown old and they were still—it was really moving, and the impact—I don’t know if this is a policy thing—but the impact on individuals of losing loved ones, it can never be—it gets really hard to—it’s a really big thing, and it’s really hard to capture that as a piece of evidence to inform policy. We do trials, and we do the things that we’ve talked about. (Interview 16, Clinician, emphasis added)

It is at this juncture that it is important to ask some different, but related, critical questions. How might framing “evidence” around the notion of “usefulness” be counterproductive? What happens when these performances of evidence-based policy “misfire” (Race, 2011, p. 411)? While the purported aim of evidence-based policy may be to provide a rational basis for action, what happens when enactments “can be seen to participate in processes that undermine some of its express aims” (Race, 2011, p. 411)? Could it be that this performance of “evidence” as a “tool” for “use” becomes a barrier to action even when players are agreed that action is required, due to the boundaries it places around other possible rationales for action (i.e., other ways of “knowing” that action is required)? The two case studies examined here provide fertile ground for such examination. In both cases, there has been local and international debate about lack of “high-level” evidence. In neither case is evidence of effectiveness available in the form of randomized controlled trials or systematic reviews. While observational studies exist, these studies do not conclusively satisfy the question of “what works” according to the “hierarchy of evidence.” ² However, both the availability of naloxone for administration by peers and peer distribution of injecting equipment potentially rest on human rights arguments, and the values-based understanding that people who inject drugs can and should play a role in enhancing their own health and the health of their community. In both cases, a strategy has been implemented involving the “piloting” and “evaluation” of the intervention to provide a rationale for action in each jurisdiction in the absence of existing “high-level” evidence. However, in telling their stories, for many participants, there was an awkward fit between what was “required” by the threshold of “evidence-based policy” and the shared understanding between most actors that implementing these programs was simply “common sense.” This perspective was strongly expressed by several clinicians in particular:

I had to come out fairly strongly and so I tried in the nicest possible way to say that now and then we just have to use common sense and do what’s right. And that has been my catchcry I guess. That at a certain level, if it gets to the stage where evidence is being used as a barrier, we have to call that for what it is and just say look this is just crap and we need to just get on with it and don’t be silly. (Interview 13, Clinician)

In some areas common sense can—must prevail. I don’t see what the benefit of a high order, high level random controlled trial would achieve. (Interview 16, Clinician)

Sometimes I think we make it hard for ourselves at the altar of wanting to have evidence-based policy, letting it get in the way of doing things that just make obvious common sense. (Interview 28, Clinician)

There is another possibility which could be explored here, that of “virtual singularity” (Law, 2004, p. 58). The seemingly incontrovertible idea of the “usefulness” of “evidence” in policy may be productive in that this “insistence on singularity” pushes multiplicity into invisibility, thus “avoiding the appearance and the experience of multiplicity” (Law, 2004, pp. 65, 66). The accounts told by participants in these case studies are stories about the singularity of evidence-based drug policy, which in their telling hide “the unfolding and uncertain nature of practices in favour of apparently stable and separate objects” (Law, 2004, p. 65). As participants give their different accounts of the drug policy process, they think they are describing the same thing. The idea that there is a single, useful thing we call “evidence” or “evidence-based policy” “makes it easier to create many different versions of [it] because it allows participants to assume that they are talking about (and making) a single [thing]” (Law, 2004, p. 66). This means that participants can assume that they share a “fixed” notion of what “evidence” is and its usefulness. As one policy maker said of the practice of evidence-based policy:

we use evidence to talk things out. (Interview 23, Policy maker)

Conclusion

This paper has pursued a mode of inquiry quite different to the dominant approaches within evidence-based policy scholarship which have hitherto sought to “bridge the divide” between “knowledge producers” and “policy makers.” Following the perspective of policy scholars who have sought to destabilize the notion of “knowledge” in policy, I have argued that what we call “evidence” is not fixed prior to its expression in accounts of policy. Rather, “evidence” and “evidence-based policy” are performatively constituted.

The argument I have raised here is also slightly different from that of other drug policy scholars who have drawn attention to the limits of the science of addiction and have sought to uncover the ways in which “facts” are discursively produced (e.g., Thomson & Moore, 2014). Here, my focus has been not on how the “truth” of “addiction” or “drugs” is enacted through scientific method but rather how the thing we call “evidence” itself is performatively constituted in policy practice and the telling of policy stories. Accounts of policy constitute the character of what we call “evidence” and its purpose in policy.

Participants’ accounts “are not simply describing” (Law & Singleton, 2000, p. 768) evidence-based drug policy making, but “performing a particular notion of the nature” of it. Following Law and Singleton (2000, p. 768), I have argued that as participants “tell a story” about drug policy, they breathe life into a range of assumptions underpinning what we call “evidence-based drug policy.” These accounts are not “innocent descriptions” (Law & Singleton, 2000, p. 769) of evidence-based drug policy processes; they are located enactments or performances that in their telling constitute the thing we call evidence as a particular kind of thing (here, “a tool for use”). In doing so, these performances imply that a particular kind of knowledge is more useful. Enacting “usefulness” as the hallmark of evidence provides justifiable rationales, legitimacy, and protection to actors within the policy process. The enactment of “evidence” as a “tool” for “use” is therefore important, as it makes and sustains a set of assumptions about knowledge, evidence, and rationales for policy action. These stories “prop up” an idea of policy making as a rational process of authoritative choice.

As I noted earlier, much of this seems incontrovertible. However, seeing these accounts not merely as descriptions but rather as a particular kind of performance opens up the possibility to interrogate the ways in which this familiar “talk” reproduces a range of troubling political assumptions which place limits around what gets to count as valid knowledge and which voices may be heard. And while these accounts may seem familiar that is only because “evidence” is enacted in this way, in stories and practices, over and over again across the drug policy (and indeed the broader evidence-based policy) field. These accounts seem familiar “because that is how they are performed” (Law & Singleton, 2000, p. 769). It is through enactment that such a reality is produced, but it depends upon a continued practice of crafting (Law, 2004). So if the notions of “evidence” and “evidence-based policy” seem solid, this is because they are being enacted as such. And we as researchers are complicit too. ³ As researchers in the field, we “mimic” in our writing the assumptions performed in these accounts and have our own stories to tell which prop up (or interfere with) these other performances (Law & Singleton, 2000). Like in Law and Singleton’s (2000, p. 769) study, the participants in my study gave me particular accounts when interviewed to “set the record straight and contribute to what they thought of as the definitive story.”

Most importantly, what the lens of performance offers is a window through which we can see that things “could be otherwise” (Law & Singleton, 2000, p. 769, emphasis original). Indeed, in the participants’ accounts analyzed here, I have identified some ways in which “evidence-based policy” is already being done otherwise. However, these “other” enactments (values, emotion, and common sense) are obscured, perhaps by an insistence on singularity. The idea that things could be otherwise raises critical questions about how the evidence-based drug policy endeavor might be reconsidered. Here, I have argued that the thing we call “evidence” does not sit objectively outside the policy process, but rather is constituted within it through practices and the telling of policy stories. Thus, if the reality of what we call “evidence” is enacted and not fixed prior to its expression in policy activity, then it can also be said that the reasons for enacting one version over another are political. Evidence and the evidence-based policy paradigm can be remade in other ways. Mol (1999) suggests that we can think of what is enacted and constituted as an ontological politics, which “suggests a link between the real, the conditions of possibility we live with, and the political.” This means that we might “interfere, to make some realities realer, others less so” (Law, 2004, p. 67). The question of how the evidence-based drug policy paradigm might be remade in other ways is a question which remains to be explored.