Risky Business: Policy Uncertainty, Institutional Insularity, and Learning in Canadian Provincial Energy Policy Making 2006-2016
Total Page：16
File Type：pdf, Size：1020Kb
Risky Business: policy uncertainty, institutional insularity, and learning in Canadian provincial energy policy making 2006-2016. Heather Millar [email protected] Department of Political Science University of Toronto 100 St. George Street Toronto, ON, M5S 3G3 Prepared for the International Workshop on Public Policy University of Pittsburgh June 26th – 28th, 2018 Draft Working Paper: comments are welcome, please do not cite without author's permission. Abstract Policy learning scholarship has moved beyond describing typologies of learning to examine the dynamics of policy learning at micro, meso, and macro levels. Scholars have begun to theorize individual processes of cognition (Moyson 2016), linkages between individual and collective process of learning (Heikkila and Gerlak 2013), and organizational capacities necessary for policy change (Borrás 2011). Despite these advances, recent reviews have identified the need to specify necessary conditions of learning, particularly as to the role of knowledge claims and institutional structures in learning processes. This paper builds on Dunlop and Radaelli’s (2013) causal typology of modes of learning to operationalize measures for problem tractability and actor certification. The paper contends that problem tractability can be analyzed by measuring policy complexity and ambiguity, generating a typology of risk narratives that align with different modes of learning. To measure actor certification, the paper uses institutional insularity, suggesting that closed networks are more likely to support hierarchical and epistemic learning, while open structures facilitate more plural modes of bargaining and reflexive learning. Drawing on cases of hydraulic fracturing regulation in three Canadian provinces (2006-2016), the paper demonstrates that changes to institutional structures and/or risk narratives transforms dominant learning modes in each jurisdiction, resulting in changes to regulatory structures. The paper finds that while perceptions of complex risk can trigger processes of epistemic learning, resulting in result in more stringent regulatory frameworks, institutional openings generated by public consultations can prompt processes of reflexive learning or bargaining that foster radical precautionary regulatory change. Introduction Policy studies scholarship has long identified learning as a key mechanism of policy change. From early studies on lesson drawing (Rose 1991), social learning (Hall 1993; May 1992), and policy oriented-learning (Sabatier 1988) to more recent examinations of learning in policy diffusion and transfer (Stone 2017; Dobbin, Simmons, and Garrett 2007), research has documented the ways in which policy actors often use other jurisdictions as “laboratories of democracy” to inform policy formulation and design (E. R. Graham, Shipan, and Volden 2013). Yet reviews of the field continue to bemoan a lack of clarity regarding conceptualization and measurement, namely: what is learned, who is learning, and the conditions under which learning occurs (Bennett and Howlett 1992; Moyson, Scholten, and Weible 2017; Pierson 1993). Recent theoretical work by Dunlop and Radaelli (2013, 2018a, 2018b) has begun to address this gap by systematically mapping the field of policy learning on the dimensions of problem tractability and actor certification. Problem tractability refers to the perceived uncertainty of a policy problem, while actor certification refers to the authority of information sources – for example whether there is one certified “teacher” or many. Together, these two conditions generate four varieties of learning: epistemic learning, reflexive learning, learning through bargaining, and learning in the shadow of hierarchy (Dunlop and Radaelli 2018a, 4). The typology is explanatory in that movement on either of the two axes – problem tractability or actor certification – helps explain the transition of a polity from one mode of learning to another. The framework thus provides a causal logic for explaining why different modes emerge at different times in a given policy area. As such, the framework provides an opportunity for scholars to develop testable propositions about sequences of learning and the causal drivers of policy change (Dunlop and Radaelli 2018a, 15). This paper contributes to this policy learning framework in two ways. First I operationalize measures for problem tractability and actor certification. To measure problem tractability I build on findings from ideational studies on uncertainty and risk perception (Blyth 2009, 2013; Klinke and Renn 2012; Slovic 1987) to develop a typology of “risk narratives.” Drawing on concepts of policy complexity and ambiguity (Pierson 1993; Cairney, Oliver, and Wellstead 2016; Klinke and Renn 2012), I generate a four-part typology of linear, complex, uncertain, and catastrophic risk, suggesting that risk narratives help refine our measurement of problem tractability. To measure actor certification I propose an aligned concept of institutional insularity (Pierson 1993), proposing that open institutional contexts are more likely to generate contested legitimacy than those with a limited set of decision makers (Skogstad 2008). Second the paper’s findings suggest that identifying the sequencing of different modes of learning can help explain the content of policy designs, explaining why one jurisdiction implements more substantive policy change than another at different points in time. This paper probes the plausibility of these propositions by examining processes of learning in provincial regulation of hydraulic fracturing in Canada. Hydraulic fracturing, or “fracking” as it is more commonly known, is an unconventional method of oil and gas production that has fundamentally transformed the oil and gas industry in North America (Neville et al. 2017). Through the use of new technologies enabling the combination of horizontal drilling and multi- Millar - Learning & Hydraulic Fracturing Policy 1 stage hydraulic fracturing1 both US state governments and Canadian provinces have been able to access previously untenable reserves of oil and natural gas, offsetting the decline of conventional reserves (NEB 2009; USEIA 2013). Despite these economic benefits, the extraction process has garnered a range of environmental critics from both academic and activist communities, as groups have raised the potential for groundwater contamination, habitat fragmentation, air pollution, chemical spills, and seismic activity, among others (The Council 2014; Neville et al. 2017). In both Canada and the US, sub-national governments have responded to these challenges by implementing a variety of regulatory frameworks, ranging from single issue regulation, to comprehensive frameworks, to moratoria and bans (Carter and Eaton 2016; Rabe 2014; Olive and Delshad 2017). As such, hydraulic fracturing is a policy field with variation in policy uncertainty and institutional insularity, as well as a range of regulatory outcomes, making it a useful test case for theory generation. The study examines processes of regulatory development from 2006-2016 in three jurisdictions: New Brunswick, Nova Scotia and British Columbia. Drawing on evidence collected through in- depth process tracing (Beach 2016) I find that different risk narratives, together with different institutional contexts triggered different sequences of learning, resulting in more or less stringent regulatory designs. In provinces with more insular institutional structures, narratives of linear and complex risk prompted processes of hierarchical and epistemic learning respectively, resulting in single issue regulation on the one hand and more comprehensive regulation on the other. Alternatively, in provinces with more open institutional structures, narratives of uncertain and catastrophic risk triggered respective processes of reflexive learning and bargaining, resulting in moratoria and legislated bans. I find that while linear risk narratives and hierarchical learning can “lock-in” policy stasis, uncertain risk narratives and reflexive learning generate political uncertainties that are difficult for government officials to manage. The paper proceeds in the following way. The next section reviews Dunlop and Radaelli’s (2013) theoretical framework of policy learning and proposes measures for the concepts of problem tractability and actor certification. I outline a typology of risk narratives, proposing that problem tractability can vary with regard to both policy complexity and uncertainty. I also review the concept of institutional insularity, suggesting that closed networks are likely to generate higher levels of actor certification than open networks. Section three demonstrates the applicability of these measures, examining the lock-in of hierarchical learning in British Columbia, the movement from epistemic learning to bargaining in New Brunswick, and from reflexive learning to bargaining in Nova Scotia. The paper concludes with a discussion of the implications of the case findings for further theory building and development. Modes of learning Building on rich history in policy studies, research on policy learning has entered into a new stage of theoretical development. Moving beyond early conceptual work establishing types of learning (Hall 1993; May 1992), policy scholars have begun to theorize relationships between 1 Other terms for the process include “hydro-fracking,” “fraccing,” and “high volume hydraulic fracturing.” For the purposes of this paper I use the term “hydraulic fracturing”