Policy Learning

Introduction

Policymakers, analysts, and policy innovators are constantly looking for ways to improve environmental policy. This essay critically evaluates Wurzel et al.'s (2013: 234) assertion that ‘policy learning has played an important role in the recalibration of “old” instruments (to smarter forms) and the adoption of “new” ones’. This paper takes issue with this claim in two significant ways. First, it shows that isolating the role of policy learning in policy change is difficult. Second, it shows how other, more important factors either work in concert with or overpower the influence of policy learning in policy change. These findings call into question the validity of Wurzel et al.’s assertion (hereafter ‘the original assertion’) and provide insights into the policy learning discussion. In the next section, key definitions and context are laid out. The third and fourth sections examine evidence for the role of policy learning in policy change through the frames of factors and extent of policy change. The fifth section discusses and summarises key findings, and the sixth section concludes with a few policy recommendations.

Defining and contextualising key terms

Learning

‘Learning’ has been a buzzword in the policy literature for several decades now, creating a multiplicity of terms and concepts (for example, see Bennett & Howlett, 1992; Common, 2004; Fiorino, 2001; Heikkila & Gerlak, 2013; May, 1992; Nilsson, 2005; Sabatier, 1988). To visualise the relationships among these terms, a conceptual model is presented, which is termed the Competitive Landscape for Environmental Policy (CLEP). This model is then used to examine the interplay among the various factors and forms of environmental policy change. This essay employs Hall’s 1988 definition of learning (cited in Bennett & Howlett, 1992: 276) as a '…deliberate attempt to adjust the goals or techniques of policy in the light of the consequences of past policy and new information so as to better attain the ultimate objects of governance’. To illustrate its interplay with adjacent concepts, policy learning is placed at the centre of two axes, one showing ‘learning’ at input and product levels, and the other showing factors and extents of policy change. As policy ideas are ‘translated’ (Heikkila & Gerlak, 2013) into policies, they move generally from the top and left towards the bottom and right, as the arrows forming the axes suggest. Crossing either axis (starting from the top left) constitutes ‘policy change’, although policy learning may not be a factor.

Figure 1 – Competitive Landscape for Environmental Policy (CLEP)

(Adapted from Common, 2004; Heikkila & Gerlak, 2013; May, 1992; Nilsson, 2005; Obinger, Schmitt, & Starke, 2013; Radaelli, 2000)

In this model, the policy level is couched between individual and collective levels (input levels) and political and international levels (which surpass the scope of individual policies). Learning is couched between coercion and emulation, (which, along with learning, are) factors of diffusion and convergence (Obinger et al., 2013). An additional layer of complexity not explicitly visualised is the political, economic, and societal factors that influence policy change, which, as is shown in the cases below, are often the forces that push policy ideas across the axes into various forms of policy change.

‘Important’ and ‘policy change’

In this essay, ‘important’ is understood to mean ‘significant’ rather than ‘crucial, even if small’. Although it is crucial a policymaker ‘learn’ how a policy could change in order for a change to occur, such learning might not constitute a ‘significant’ role in why the policy changes. This essay examines whether policy learning has played an important (meaning significant) role in policy change. Also, for simplicity of discussion, the latter part of the original assertion—‘the recalibration of “old” instruments (to smarter forms) and the adoption of “new” ones’—is understood to mean ‘policy change’, as it is hard to imagine a realistic scenario in which an environmental policy would change in another way.

Learning’s interactions with other factors of change: coercion and emulation

Evidence-based policy and regulatory impact assessments

Notions of evidence-based policy (EBP) and smart regulation stem from the idea that rationality and evidence can contribute to effective policy design (Sanderson, 2002) and constitute a good starting point for a discussion of learning. However, the claim that policymakers use information effectively as an input in policymaking has been challenged (Bauler, 2012; Owens, 2005). Notwithstanding, Regulatory Impact Assessments (RIAs) are one tool policymakers have implemented in an attempt to evaluate policies’ effectiveness; as such, they appear to be aimed at promoting learning. However, in the European Union (EU), RIAs have been used by the European Commission (EC) as a top-down narrative tool, a platform from which the EC creates legitimacy and promotes norms and convergence (Radaelli, Dunlop, & Fritsch, 2013). Thus, rather than merely serving an evaluative purpose, RIAs may assert international pressure or promote coercion and emulation.

Multilateral environmental agreements

In a recent study, Schulze & Tosun (2013) find that in multilateral environmental agreements (MEAs) with potential trade partners, the EU coerces non-EU member states to adopt its environmental standards by offering them access to its markets and its exports. Their analysis, based on ‘the ratification behaviour of 25 non-EU Member States with regard to all 21 [EU] MEAs’, shows ‘the EU’s external influence at the ratification stage of environmental regime formation’ (Schulze & Tosun, 2013: 581) and concludes that the EU’s methods ‘impel these countries to join the EU’s preferred MEAs’ (ibid., p. 598). Policy change is clearly occurring, but merely accepting the terms and conditions of trade when coerced hardly constitutes policy learning. The examples of RIAs and MEAs would thus fall in the lower-left quadrant of the CLEP model (although extending towards the centre), showing an interaction of factors obscuring the role of learning in policy change.

Evidence of learning in extent of change: diffusion and convergence

Diffusion and convergence describe the processes by which policy disseminates, particularly internationally, and then begins to look similar as it becomes broadly adopted. Meseguer and Gilardi in 2009 defined policy diffusion as ‘…the process by which “policy choices in one country affect the policy choices in other countries”’ (as cited by Obinger et al., 2013: 113). Policy transfer, for its part, has been defined as ‘the process by which knowledge of policies, administrative arrangements, institutions and ideas in one political system (past or present) is used in the development of policies, administrative arrangements, institutions and ideas in another political system’ (Dolowitz, 2000: 3). Although subtly different concepts, policy diffusion and transfer are essentially interchangeable terms (Obinger et al., 2013). At first blush, these definitions may seem similar to that of policy learning; they are, however, broader concepts, which may include policy learning as an input. Thus, analysing the causes of diffusion and convergence (forms of policy change) can provide evidence of learning’s role in policy change.

Environmental policy integration

An example of diffusion is environmental policy integration (EPI), which as a principle is written into the European treaty. Sweden has been a leader in infusing principles of sustainability into its development goals and institutions {C}(Nilsson, 2005){C}; however, despite an informed and supportive public and rapid institutional change (and thus opportunity for EPI) both in government and the energy sector, meaningful EPI has been slow. ‘For some time, learning of a political nature, rather than conceptual learning, characterised the government’s energy ­– environment agenda, despite new knowledge and perspectives being available’ (ibid., p. 222). Consequently, even given Sweden’s favourable conditions, environmental policy learning resulted in ‘political learning, rhetoric, and “symbolic politics”’ (ibid., p. 223). This example spans the space between ‘policy diffusion’ and ‘political learning’ in the CLEP model, with some superficial coverage of policy learning. Nilsson’s conclusions suggest that, although the principles of EPI diffused from the European treaty to Swedish policy on paper, this did not result in substantive change. Thus, even when policy learning has occurred, policy change may not necessarily result.

Finnish climate policy

Hildén (2011) profiles another example of policy diffusion, this time of climate policy in Finland, as diffused from the United Nations (UN) via the EU’s agreement to support the UN Framework Convention on Climate Change (UNFCCC). Hildén finds that ‘Finnish climate policies have evolved rapidly….[,] driven in part by the emulation of solutions developed at the EU-level’ (Hildén, 2011: 1806). This is the clearest example examined so far of how policy learning has played a role in policy change; however, even Hildén acknowledges the role of ‘confounding factors’ in his model of the ‘policy cycle of climate change’ (ibid., p. 1800). This example occupies the triangular space bound by the corners of international learning, policy emulation, and policy diffusion in the CLEP model and provides some evidence for the role of policy learning in policy change. However, based on the rapid evolution and focus on policy emulation in the change in Finnish climate policy, the importance of the role of endogenous policy learning remains questionable.

New environmental policy instruments and cost-benefit analysis

Tews, Busch, & Jörgens (2003) provide examples involving the diffusion of four ‘New Environmental Policy Instruments (NEPIs)…[:] eco-labels, energy or carbon taxes, national environmental policy plans or strategies for sustainable development, and free-access-of-information (FAI) provisions’ (p. 569). Their analysis reveals the complexity of attempting to establish a causal relationship between learning and change. The authors conclude that ‘the adoption of environmental policy innovations is more likely if these policy innovations figure prominently on the global political agenda’ (ibid., p. 592). Here again, exogenous factors, such as the geopolitical agenda, are listed as the first factor in driving policy change. They find further that ‘the special features of a policy innovation can either facilitate or hinder its widespread adoption’ (ibid.). In these cases, policy learning may play a role in designing the ‘special features’ of the policy, but it is difficult to isolate that role (from, say, emulation) in the development of any such policy. Finally, as they break down the factors affecting adoption of each NEPI individually, their conclusions show further that the drivers of policy change are highly context specific (ibid.). Thus, while policy learning may play a role in some policy changes, the unique drivers of change in the four NEPI case studies show how generalising about policy learning’s role more broadly can be problematic, again placing in question the original assertion.

Livermore & Revesz (2013) examine the diffusion of a similar example to the NEPIs examined above: cost-benefit analysis (CBA). The diffusion of CBA has been somewhat more successful, with the formation of ‘[r]egional collaboration networks and professional associations’ (Livermore & Revesz, 2013: 309) for sharing best practices. However, many of Tews et al.'s (2003) conclusions apply to CBA as well. Thus, both NEPIs and CBA figure in the lower-right quadrant of the CLEP model, reflecting the international nature of their diffusion, with some spillover into convergence in the case of CBA.

Canadian ‘convergence’

Policy convergence is the theoretical next step after diffusion. The idea is that after policies (and even institutional structures) diffuse broadly, they begin to resemble each other (converge). And as policy transfer is ‘the outcome of international policy learning’ (Common, 2004: 41) and convergence the outcome of transfer (diffusion), examining the literature on convergence could shed light on the role of learning in policy change. A study involving a broad survey of provincial-level environmental policy analysts throughout Canada by Howlett & Joshi-Koop (2011) reveals how policies can change without transfer or convergence, even in cases where they might be expected. The authors find that ‘environmental policy analysts…fail to routinely use [external] information…in policy development and formulation’ (ibid., p. 91; see also Bauler, 2012; Owens, 2005), concluding that, ‘at least in Canada, the potential for trans-national learning to lead to policy convergence is limited, or at best, indirect and weak’ (Howlett & Joshi-Koop, 2011: 91). Thus, ‘it cannot simply be assumed that the existence of new evidence or information translates into policy development’ (ibid.). Even if there are lessons to be learned, and even in a country as institutionally capable as Canada, policy learning is not likely to be a factor in policy change (or maybe event to occur at all) if policy analysts are not making themselves aware of developments in policy outside their immediate purview. The study’s findings tenuously place it, if at all, narrowly on the top central axis of the CLEP model. This is because it illustrates how, when policies did change, it was the result of limited individual and small-order collective learning and innovation rather than from international (or even interprovincial) transfer. One could argue this is support for learning leading to change, but any link is weak.

Discussion and application

After developing his Advocacy Coalition Framework (ACF) to examine the role of policy learning in policy change, Sabatier (1988) concludes that ‘…changes in the core aspects of a policy are usually the results of…[external factors]…such as macro-economic conditions or the rise of a new systemic governing coalition’ (p. 134). This assertion is reasserted in Heikkila & Gerlak's (2013) theory on collective-learning and affirmed in every case study examined above. Thus, larger, exogenous, systemic ‘perturbations’ are likely required for significant policy change to occur, even in the presence of policy learning.

We can also apply CLEP analysis to instances where policy change has not yet occurred, such as with climate change legislation in the United States. In 2008, John McCain, Barack Obama’s Republican opponent in the presidential election, ran on a cap-and-trade agenda to address climate change. Cap-and-trade was seen as a conservative, business-friendly, market-based alternative to traditional command-and-control regulation. However, following the Climategate scandal, the economic crisis, abnormally cold weather, and the election of a Democratic president, congressional Republicans and conservative public opinion turned against any type of greenhouse-gas emission reduction legislation as unfounded in science, too expensive, and a partisan issue (Leiserowitz et al., 2012). In this case, it was not that policy learning had not occurred sufficiently for Republicans to understand (and even support) a new policy instrument in regulating greenhouse gas emissions, the issue was that external factors played a more important role than policy learning in the decision of whether or not to adopt a new policy instrument.

Taking the issue of climate change to the international context illustrates yet another example. Despite widespread academic and scientific consensus on the types of action that should be taken to mitigate the worst effects of climate change (IPCC, 2013; Stern, 2007; UNFCCC, 2013), the geopolitical forces have so far proven too complicated or otherwise unable to act in setting a global price for carbon. In this instance, it is not that appropriate policy instruments are unknown to global leaders, it is that geopolitical forces have (thus far) overpowered any amount of knowledge of appropriate responses. Thus, the CLEP model would predict that, rather than policy learning, international and political learning are needed to effectuate the implementation of a new policy instrument. Therefore, while policy learning may play a role by bringing to the fore the appropriate instruments, it is inadequate without sufficient amounts of political and international learning. Here again, the role of policy learning plays a secondary role in policy change.

However, even in a stable policy environment, policy learning’s role in policy change is hard to isolate based on its relationship to other factors. This has also been demonstrated in the case studies. Below, the CLEP model is reproduced with the case studies superimposed, demonstrating the complex relationships and overlap between policy learning and related factors in policy change.

Figure 2 – Competitive Landscape for Environmental Policy (CLEP) – cases superimposed

Concluding remarks

As has been shown, the definitions of ‘policy learning’ in the literature are diverse. The veracity of the original assertion that policy learning has played an important role in policy change is largely dependent on one’s definition of policy learning. But the overarching idea of trying to make smarter policies is worthy of sustained support and discussion (Sanderson, 2002). The CLEP model contextualises policy learning among a broad array of terms. In context, policy learning may play a role in various instances of policy change, but its relative importance compared to other factors is questionable, and its influence is difficult to isolate. A wide range of cases were examined where policy learning occurred, and yet policies changed in some cases but not in others. Examples from the literature have demonstrated that factors exogenous to policymaking subsystems likely play a more important role than policy learning in explaining policy change. This finding has at least two important implications for the discussion on policy learning. First, given the limited role of policy learning, any effort to improve the prospects of effecting policy change via policy learning might be better directed at other factors in policy change. Second, policies might be most improved by fitting them with characteristics of resilience in the face of large, exogenous disruptions. Rather than introducing a new policy with each high or low in the economic, political, or social cycle, policymakers can ask, ‘how might this policy fare under different economic or geopolitical circumstances’ or ‘is there flexibility that can be built into this policy instrument’? Recalibrating instruments and adopting new ones as necessary with such questions in mind might reduce future demands for policy change in the first place.

WORKS CITED

Bauler, T. (2012). An analytical framework to discuss the usability of (environmental) indicators for policy. Ecological Indicators, 17, 38. doi:10.1016/j.ecolind.2011.05.013

Bennett, C. J., & Howlett, M. (1992). The lessons of learning: Reconciling theories of policy learning and policy change. Policy Sciences, 25, 275–294.

Common, R. (2004). Organisational learning in a political environment. Policy Studies, 25(1), 35–49. doi:10.1080/0144287042000208224

Dolowitz, D. P. (2000). Introduction. Governance: An International Journal of Policy and Administration, 13(1), 1–4.

Fiorino, D. J. (2001). Environmental Policy As Learning: A New View of an Old Landscape. Public Administration Review, 61(3), 322–334.

Heikkila, T., & Gerlak, A. K. (2013). Building a Conceptual Approach to Collective Learning: Lessons for Public Policy Scholars. Policy Studies Journal, 41(3), 484–512. doi:10.1111/psj.12026

Hildén, M. (2011). The evolution of climate policies – the role of learning and evaluations. Journal of Cleaner Production, 19(16), 1798–1811. doi:10.1016/j.jclepro.2011.05.004

Howlett, M., & Joshi-Koop, S. (2011). Transnational learning, policy analytical capacity, and environmental policy convergence: Survey results from Canada. Global Environmental Change, 21(1), 85–92. doi:10.1016/j.gloenvcha.2010.10.002

IPCC. (2013). Human influence on climate clear, IPCC report says (p. 3). Stockholm. Accessed 7 February 2014. Retrieved from http://www.ipcc.ch/news_and_events/docs/ar5/press_release_ar5_wgi_en.pdf

Leiserowitz, a. a., Maibach, E. W., Roser-Renouf, C., Smith, N., & Dawson, E. (2012). Climategate, Public Opinion, and the Loss of Trust. American Behavioral Scientist, 57(6), 818–837. doi:10.1177/0002764212458272

Livermore, M. A., & Revesz, R. L. (2013). The globalization of cost-benefit analysis in environmental policy. New York: Oxford Univ. Press.

May, P. J. (1992). Policy Learning and Failure. Journal of Public Policy, 12(4), 331–354.

Nilsson, M. (2005). Learning, frames, and environmental policy integration: the case of Swedish energy policy. Environment and Planning C: Government and Policy, 23(2), 207–226. doi:10.1068/c0405j

Obinger, H., Schmitt, C., & Starke, P. (2013). Policy Diffusion and Policy Transfer in Comparative Welfare State Research. Social Policy & Administration, 47(1), 111–129. doi:10.1111/spol.12003

Owens, S. (2005). Making a difference? Some perspectives on environmental research and policy. Transactions of the Institute of British Geographers, 30(3), 287–292. doi:10.1111/j.1475-5661.2005.00171.x

Radaelli, C. M. (2000). Policy Transfer in the European Union: Institutional Isomorphism as a Source of Legitimacy. Governance, 13(1), 25–43. doi:10.1111/0952-1895.00122

Radaelli, C. M., Dunlop, C. a, & Fritsch, O. (2013). Narrating Impact Assessment in the European Union. European Political Science, 12(4), 500–521. doi:10.1057/eps.2013.26

Sabatier, P. A. (1988). An advocacy coalition framework of policy change and the role of policy-oriented learning therein. Policy Sciences, 21(2-3), 129–168. doi:10.1007/BF00136406

Sanderson, I. (2002). Evaluation, Policy Learning and Evidence-based Policy Making. Public Administration, 80(1), 1–22.

Schulze, K., & Tosun, J. (2013). External dimensions of European environmental policy: An analysis of environmental treaty ratification by third states. European Journal of Political Research, 52(5), 581–607. doi:10.1111/1475-6765.12011

Stern, N. (2007). Stern Review: The Economics of Climate Change. London. Accessed 7 February 2014. Retrieved from http://mudancasclimaticas.cptec.inpe.br/~rmclima/pdfs/destaques/sternreview_report_complete.pdf

Tews, K., Busch, P.-O., & Jorgens, H. (2003). The diffusion of new environmental policy instruments. European Journal of Political Research, 42(4), 569–600. doi:10.1111/1475-6765.00096

UNFCCC. (2013). UN Climate Change Conference press release (p. 4). Warsaw. Accessed 7 February 2014. Retrieved from http://unfccc.int/files/press/news_room/press_releases_and_advisories/application/pdf/131123_pr_closing_cop19.pdf

Wurzel, R. K. W., Zito, A. R., & Jordan, A. J. (2013). Environmental governance in Europe: a comparative analysis of new environmental policy instruments (p. 285). Cheltenham, UK: Edward Elgar.