Theories of the Policy Process

A wide variety of theories now exist to explain the policy process but none appear satisfactory for all purposes [16], [17], [18]. Historically, policy was assumed to occur in a ‘rational’ manner which involved passage through a number of distinct stages which together constitute a ‘policy cycle’ [19], [20]. While this model continues to underpin (often implicitly) the perspective of many health researchers and those involved in policy [21], it is not supported by empirical evidence and does not recognise the constraints on rational decision-making or the role of implementers in policy as it is enacted [17], [19]. Despite these limitations, the stages model presents a helpful heuristic device and many of the processes described above often do occur during the policy process, albeit in a non-linear manner.

The limitations of the stages heuristic resulted in the development of alternative models – three of which we draw upon in our analysis. Kingdon’s Multiple Streams model seeks to understand what factors explain the process by which the multitude of policy issues that could be considered by policymakers, results in the relatively few policy problems that are actually considered in detail [22]. He identified three factors, referred to as ‘streams’, which must come together for a policy to develop. The existence of a policy issue that is construed as worthy of policy intervention constitutes the problem stream. The policy stream refers to the availability of a solution that is viewed as having the potential to address the problem. The politics stream refers to the political climate operating which may either help or hinder the consideration of a specific policy issue (for example, including the political parties dominant at the time). These three streams can be ‘coupled’ by ‘policy entrepreneurs’; advocates of a proposed policy solution who are skilful communicators and are willing to sustain their efforts for prolonged periods of time.

The second theory, Punctuated-Equilibrium Theory, seeks to explain the observation that many areas of public policy exhibited little policy change (i.e. are in ‘equilibrium’) while a few areas experience rapid shifts in policy (‘punctuations’) [23]. It argues that the time constraints policymakers operate under result in them being unable to focus on all areas of public policy simultaneously. High-level policymakers therefore leave most areas to communities of experts while they focus on the areas that have become ‘hot topics’, attracting the attention of the media and a broader group of actors than previously engaged. Policy areas that are undergoing punctuations therefore experience an increased tendency for radical policy change as a result. In keeping with Kingdon’s Multiple Streams Model, Punctuated-Equilibrium Theory suggests that a change in the ‘framing’ (or perception) of a policy issue or a well-respected indicator can be crucial in triggering increasing interest in an existing policy problem.

The third perspective that informs this paper is Multi-Level Governance. In contrast to the other two theories, Multi-Level Governance does not provide a theory of the policy process as such but rather draws analytical attention to ongoing changes in institutional responsibilities, derived from experiences in Europe [24], [25], [26]. A multi-level perspective theorises a shift in power from one central governmental state authority (such as the UK Government) to a range of institutions that operate both above (for example, the European Union) and below the nation state (e.g. the devolved Scottish Parliament). It also theorises an increase in the involvement of non-governmental and private sector actors, signalling a move from ‘government’ to ‘governance’. Reflecting this academic perspective has been a growing interest in the impact of devolution – the increase of governmental institutions operating below the state as illustrated in the UK by the Welsh Assembly and the Scottish Government [27].

Ethics Statement

Ethical approval was obtained from the University of Glasgow’s College of Medicine and Veterinary Science research ethics committee. Informed consent was sought from each participant in a process approved by the ethics committee and documented in a written consent form.

A qualitative case study approach [28] has been adopted to investigate MUP’s development. We drew on two main sources of data: semi-structured interviews with policy actors and a review of relevant publicly available policy documents. Potential interviewees were mainly identified from the document analysis and usually initially approached by e-mail to request their participation in a research study on the relationship between evidence and alcohol policy. Of those contacted, seven potential participants were unwilling to participate (usually for reasons of lack of time) and interviews could not be scheduled with a further five. Thirty-six interviews were conducted by SVK (a public health doctor) with purposively selected policy actors (5 politicians, 10 civil servants, 7 advocates for MUP, 6 industry-related actors and 8 academics) directly involved in MUP debates in Scotland and/or the UK. In addition to sampling participants on the basis of the above five sectors, diversity was sought within each sector (political party membership for politicians, nature of industry-related actor and area of focus for advocacy groups) and in relation to supportiveness or hostility to MUP. Interviews typically lasted 45–60 minutes and were conducted either face-to-face (usually at the participant’s place of work) or over the telephone. Recruitment continued until adequate diversity was obtained in the sample and no major new themes emerged in the data [29]. Contemporaneous handwritten fieldwork notes were kept during the data collection and analysis process.

Given the small nature of the policy community we were seeking to access, we sought separate written consent for interview participation, audio-recording the interview, attribution to the interviewee’s broad sector (for example, civil servant, politician etc.) and the use of quotations [30]. All but four interviewees agreed to full participation. Following the interview, transcripts were amended by noting sections not for quotation and making limited changes to the text to reduce the risks of disclosure in consultation with interviewees. Finally, interviewees were given the opportunity to review a copy of the transcript and make limited suggestions for further amendments to ensure confidentiality.

Interview data were read repeatedly and thematic content analysis conducted by SVK [31]. Coding initially proceeded inductively with descriptive codes being used to organise the data with the assistance of NVivo 9. Following this, data relevant to specific political science theories were coded using sets of ‘tree’ codes for each political science theory and subcodes used to capture a key aspect of each theory. New codes were used to capture further inductive themes in the data and a broader set of theories (for example, theories relating to policy transfer and ideas [32], [33], [34]) were also during the coding process. A relatively parsimonious account of the policy’s development, and in particular one which helped explain the origins of MUP, was then sought through an iterative process of analysis. The principle of the constant-comparative method was used to help identify explanations for patterns within the data, while also paying appropriate attention to contradictions and tensions within the data.

The review of publicly available documents sought to include published documents, which reported on quantification of alcohol-related issues (either positively or negatively), the causes and/or consequences of alcohol use, identification or debate over proposed action(s) and advocacy for specific actions. The search was conducted online and by contacting the Scottish and UK governments, health service organisations and interest groups (including health actors, other civil society organisations and a wide variety of industry groups). Interest groups were identified by reviewing the actors responding to public consultations to MUP debates within Scotland [35] and the UK [36].

Documents submitted by policy stakeholders in response to the Scottish Parliament’s Health and Sport Committee consultation (conducted from November 2009 to January 2010) were retrieved from the Scottish Parliament website and systematically analysed using thematic content analysis to understand the framing of arguments presented by policy actors. Further details of the methodology are described elsewhere [37].

The Development of the Minimum Unit Pricing Policy in Scotland

Scotland is one of the four constituent countries of the United Kingdom and political decisions at the central government-level have been made at the Westminster Parliament in London since 1707 [38]. Throughout this time, Scotland maintained separate legal (including alcohol licensing) and education systems. In 1999, a separate Scottish Parliament was established with responsibility over some ‘devolved’ policy areas (such as health, education and social care) but with other areas ‘reserved’ to Westminster (including trade and foreign affairs). Devolution has allowed Scotland the freedom to pursue policy differently from England in some areas. For example, Scotland was the first country within the UK to pass legislation to ban smoking in public places [39]. The first two Scottish Parliamentary elections resulted in coalition governments between the Labour and Liberal Democrat Parties. The Scottish National Party (SNP) has been in government since 2007, initially with a minority but gaining a majority in 2011. Key events are highlighted in the timeline (see Table 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Timeline of Milestones in the Development of Minimum Unit Pricing in Scotland.

https://doi.org/10.1371/journal.pone.0091185.t001

Alcohol Harms in Scotland

There have been longstanding concerns about alcohol consumption and harmful drinking patterns in Scotland and the UK [40]. While many professionals perceived an increase in alcohol harms since the 1990s, there was initially a lack of epidemiological data to quantify the problem. Scotland introduced its first post-devolution alcohol strategy in January 2002 [41], two years before England. The ‘Plan for Action on Alcohol Problems’ (the Alcohol Plan) was the first report to bring together data on the alcohol market, alcohol consumption, social harms and health harms and demonstrated the considerable overall alcohol burden [41]. The Plan was introduced by a Labour-Liberal Democrat coalition following a commissioned review of effective and cost-effective measures to reduce alcohol harms [42]. It set out a broad range of measures to reduce alcohol-related harms including education, provision of services and licensing reform but maintained an emphasis on addressing problem drinkers and individual responsibility [41]. This framing of policy as a matter of addressing the minority of drinkers was reflected in two specific priority areas: addressing binge drinking and reducing drinking by children and young people [43].

A key component of the Alcohol Plan was the introduction of the Licensing (Scotland) Act 2005 which was fully enacted in September 2009 [44], [45]. As well as simplifying the licensing regime, this introduced five licensing objectives, including ‘the protection and improvement of public health’. This objective was seen as an important step forward in making public health considerations central to alcohol licensing. The Act also banned promotions encouraging excessive consumption in on-sales premises, such as ‘happy hours’ (when alcohol products are discounted for a time-limited period during the day). At the time of the Act’s drafting, there was recognition that further action might be required to address off-sales [43].

The Importance of Price for Alcohol-Related Harms

One important cause of alcohol-related harms that has been increasingly focused on is the relationship between increasing alcohol affordability and levels of consumption [46]. This relationship has been widely accepted with considerable supportive evidence existing [47], [48]. From a UK perspective, while alcohol prices have increased with inflation, increased living standards had made alcohol 66% more affordable in 2009 than in 1987, with a growing price differential between off-sales and on-sales prices [49]. Supermarkets have contributed to this increasing affordability by engaging in aggressive cost-cutting of alcohol [50], sometimes selling alcohol as a loss leader and/or below the cost of duty alone [51], [52]. Paralleling the increased affordability within off-sales, there has been a shift in consumption from the licensed trade to off-licenses (and particularly supermarkets) [53], [54]. It has been argued that this growing price disparity has led to increased consumption at home, including prior to leaving for a night out [55].

Towards the end of the Labour-Liberal Democrat coalition (1999–2007), a paper in the Lancet investigated trends in alcohol-related harms by analysing liver cirrhosis deaths and found a more than doubling in death rates in Scottish men from 1987–1991 to 1997–2001 [56]. This increase greatly exceeded changes in England and ran counter to trends across most of Western Europe, adding impetus to calls for further action.

Within Scotland, the importance of price as a key mechanism to tackle alcohol-related harms was highlighted by two Scottish organisations. The national agency for health improvement, NHS Health Scotland, was asked by the Labour-Liberal Democrat administration to review the alcohol evidence base to develop a logic model to inform a new Scottish Government strategy to address alcohol harms [57]. To develop the logic model, the agency reviewed research evidence but also considered plausible theory and existing policy [58]. This logic model highlighted the central importance of the “Reduction in individual and population consumption [emphasis added]” and the need to tackle alcohol affordability.

Price was also highlighted in the run-up to the 2007 election campaign by a new Scottish public health advocacy group: Scottish Health Action on Alcohol Problems (SHAAP), established by the Scottish Medical Royal Colleges and Faculties in response to concern about increasing alcohol-related harms [59]. The organisation was funded by Scottish Government with a commitment that it could operate entirely independently and a remit “to raise awareness about alcohol-related harm and to promote solutions based on the best available evidence.” [60].

A Move to a Population-based Approach

In May 2007, the incumbent Labour-led administration was replaced by a Scottish National Party (SNP) minority government. Until this time, the Labour Party was also in power in the UK Westminster Government and the presence of the same party within both governments may have served to curtail potential policy divergence. While both the Labour and SNP parties highlighted alcohol-related health harms in their election manifestos, their approaches to tackling the problem differed. In 2007, the Scottish Labour Party manifesto emphasised the importance of its “ground-breaking partnerships” with alcohol-related industries while also noting the importance of individual responsibility. The SNP manifesto placed less emphasis on individual responsibility but also advocated a broad partnership approach (including with alcohol industries). However, an important difference was the inclusion of price-based measures, noting it “is not acceptable that a bottle of water can be more expensive than alcohol” [61].

The Development of MUP in Scotland

In September 2007, SHAAP convened an expert group to investigate the relationship between alcohol price and harms. Its report built on the international evidence base [4], emphasising the need for a ‘whole-population approach’ to address alcohol-related harms [60]. The report described the epidemiology of alcohol harms in Scotland and noted price as a key driver. Recommendations to the Scottish Government from the expert workshop included the introduction of a floor pricing mechanism (with MUP being specifically discussed within the workshop), extending the ban on irresponsible alcohol promotions from on-licenses to off-licenses, advocating to the UK Westminster Government for increased alcohol taxation and for this to be linked to alcohol strength. Subsequently, SHAAP strongly advocated for MUP by regularly engaging with politicians, producing policy briefings and pro-actively communicating with mass media.

In parallel to the advocacy instigated by SHAAP, a team at Sheffield University was asked to carry out a review of the evidence base on alcohol pricing and promotion for the UK Government [47]. The Sheffield team carried out an econometric modelling exercise to compare different pricing interventions including duty increases, MUP (for a range of levels –20 pence per unit at the lowest level to 70 pence per unit at the highest) and restrictions on off-trade price promotions [11]. Importantly, this model indicated MUP most affected those at highest risk of harms when compared to taxation-based measures.

Similar efforts to use econometric modelling to estimate the impacts of MUP were pursued within Scotland. The first Scottish report, commissioned by SHAAP, provided estimates of changes in price and consumption, and also examined the potential impact on household expenditure by differing levels of income (given concerns about impacts on lower-income subgroups) [62]. Following the formal adoption of MUP in official Scottish Government policy in May 2009 [2], the Sheffield team was asked to produce another version of their model using Scottish data. This was first published in September 2009 [63] and subsequently updated twice. These peer-reviewed Sheffield models were not accepted by all MSPs, who drew on a critique of MUP funded by the brewer SABMiller [64].

Changing Scotland’s Relationship with Alcohol: The Second Scottish Alcohol Strategy

Following the SNP’s election, the Justice Minister, Kenny MacAskill, presented tackling cheap alcohol as a priority within Scottish Parliamentary debates. Initially, the Scottish Government intended to publish an alcohol strategy within the first few months of its administration. However, it realised a radical strategy was necessary and therefore published the discussion paper, ‘Changing Scotland’s Relationship with Alcohol’, in June 2008 instead and consulted on a number of proposals including MUP [65]. The paper explicitly made links to overarching government aims to deliver a “Wealthier and Fairer, Safer and Stronger, Healthier and Smarter Scotland”. Importantly, the approach moved from addressing problem drinkers to addressing population consumption [65]. This innovative ‘whole-population approach’ is reflected by one of the four key areas for action being to reduce overall alcohol consumption. In addition to MUP, a number of other controversial interventions were consulted on including: raising the minimum purchase age to 21 in off-sales; separate checkouts for alcohol sales; and introducing a ‘social responsibility fee’ for some alcohol retailers. Considerable Scottish Parliamentary and media debate occurred following the discussion paper, with the formal consultation receiving 259 responses from individuals and 207 responses from organisations [66].

Scottish Parliamentary Considerations of MUP

The passage of MUP into legislation in Scotland has not been straightforward. There has been a prolonged and heated political debate as to whether this legislative measure was the most ‘appropriate’ means of tackling alcohol-related harms in Scotland.

On the first attempt to introduce MUP, only the SNP (as a minority government) was openly supportive, with the other major political parties opposed. However, several of the measures included in the Bill (such as a ban on off-trade promotions) had broad support. Despite repeated attempts by the minority SNP Government to keep the MUP component of the Bill, this measure was ultimately withdrawn and the Bill passed without the MUP provision in November 2010.

In May 2011, Scottish Parliamentary elections resulted in the SNP (which had included MUP in its manifesto [67]) gaining an overall majority of seats in the Scottish Parliament [68]. A second MUP Bill was introduced and while the SNP majority government no longer required the support of opposition parties to pass legislation introducing MUP, two of the three opposition parties supported the Bill with the inclusion of a ‘sunset clause’ [69]. A change in leadership within the Liberal Democrats facilitated their change of position while the Conservative-led UK Government had started considering MUP themselves. During the latter stages of MUP’s second consideration for legislation, evidence related to reference pricing from Canada helped provide confirmation of public health benefits following the introduction of a comparable minimum price scheme [5], [6]. On 24^th May 2012, the Alcohol (Minimum Pricing) (Scotland) Bill was passed with 86 voting in favour of the measure, one vote against and 32 abstaining [70].

The Scottish Government initially planned to implement MUP in April 2013 but has been delayed after the Scotch Whisky Association followed through on its pre-legislation threats and challenged the legality of MUP on the grounds that it could distort trade within the European Union and questioned the Scottish Government’s legal competence to introduce the measure [71]. This legal challenge has been quashed but policy implementation will be delayed following an appeal [9].

Explaining the MUP Policy Process

To understand the process by which MUP developed, each of the previously described theories of the policy process will be applied. Rather than trying to identify an optimal theory, insights will be sought from each of the theories through a ‘multiple-lenses’ approach [72].

Lessons for Evidence-Informed Policymaking

Our analysis of the development of MUP raises a number of potential lessons for those seeking to improve the use of evidence within public health policy. First, there is some support for those seeking evidence-informed policy [82], [83], [84]. In particular, epidemiological data was important in highlighting the problem and systematic reviews and econometric modelling appear to have played an important role in allowing those seeking policy solutions to make use of price as a mechanism to address alcohol-related harms. Also, intermediary organisations served as a ‘bridge’ between the worlds of research and policy, assisting civil servants in developing an alternative policy image that would help further a public health approach in alcohol policy [85]. However, this case study also highlights the importance of the enlightenment function of evidence [86]. In this case, the change to a population framing that emphasises alcohol as ‘no ordinary commodity’, allowed a change in how alcohol policy is conceptualised and has been crucial for MUP’s development. There therefore remains a need for synthesis, but also longer term research that results in evidence helping policymakers think about policy issues in a new way [87].

Policy entrepreneurs, responsible for helping to combine the three streams of problem, politics and policy, have played an important role in policy development and they have drawn upon evidence in varied ways to assist in this process [73]. There continues to be considerable interest in the use of knowledge brokers to help provide evidence to those responsible for decision-making and the role of NHS Health Scotland in the case of MUP provides some support for such efforts [88], [89]. However, our analysis suggests that a wide variety of other factors were at least as important and so knowledge brokers alone are unlikely to be sufficient in fostering evidence-informed policy in many situations. Furthermore, the fact that advocates, civil servants and politicians all could operate as knowledge entrepreneurs suggests that knowledge linkage efforts should perhaps operate in a broader way – linking multiple communities rather than just bridging a divide between research and policy [90].

Adopting a Multi-Level Governance perspective has highlighted a number of issues that remain under-investigated within the public health literature. MUP arguably presents an example of ‘venue shopping’ where the Scottish Government provided an additional institution for public health advocates to attempt to influence policy rather than the UK Government, raising parallels with previous tobacco control policy developments [39]. The Scottish Government did not merely act as an alternative venue but the limited institutional powers actually fostered policy innovation, with MUP being a more effective public health measure than more traditional taxation-based measures used alone [91], [92]. To capitalise on the availability of a Scottish policymaking venue, local data [56], [93] which compared Scotland unfavourably to other jurisdictions were helpful in prioritising alcohol policy. This combined with the fact that health policy is a highly visible area for Scottish Government, given its limited powers [38]. The implications of this case study arising from Multi-Level Governance do not just apply within the UK but also illustrate the potential for evidence to influence policy in other political systems with multiple levels of political representation, including North America and Europe. However, the legal challenges at the European level highlight the importance of considering how those opposed to a public health policy can seek recourse to venues more advantageous to them [81].

Strengths and Limitations

Our study has a number of important strengths that are worth noting. We have drawn upon several sources of data to obtain a detailed understanding of a high-profile policy’s development. Furthermore, our data collection occurred in a timely fashion and we were successful in obtaining interviews with a number of actors directly involved in the policy process. Our study also makes use of a number of political science theories to better understand MUP’s development. This approach has allowed different aspects of the policy process to be understood while providing a relatively parsimonious account of the policy process.

A number of limitations should be noted. The literature on Kingdon’s multiple streams and Punctuated-Equilibrium Theory originated in the pluralist American tradition and although they have been applied to the UK setting [94], [95], debates continue about their relevance to this context. They have been critiqued for downplaying the importance of specific economic and elite interests. While in this article, with its focus on identifying potentially generalisable lessons for evidence-informed policy, we have not sought to detail the role of specific interest groups, we have begun to do so elsewhere [96], [97]. However, these two theories have been usefully applied to better understand the dynamics of the policy process and in particular allow an appreciation of the complex process by which MUP became a feasible policy solution. In addition, although our analysis in this paper has been informed by three political science theories, this has not resulted in a comprehensive account which explains all aspects of the policy process and other theories would yield additional insights. While the literature on multi-level governance has originated within Europe to help highlight the increasing diversity of policy actors and the multiple institutions involved in the policy process, an alternative body of literature on polycentric governance, points to the interplay and contestation that occurs between competing regulatory organisations, particularly given ongoing changes in institutional jurisdictions [98], [99]. This perspective could have been usefully employed for this case study, particularly to explore the latter phases of the policy’s development, including the ongoing legal challenges that have impeded MUP’s enactment. However, an adequate exploration of the contestation between institutions, including the incentive for Scottish politicians to demonstrate their leadership over the rival UK Westminster arena in the context of the Scottish independence agenda, is outside the scope of this article but we will examine this in subsequent publications [100]. Similarly, the literature on science and technology studies highlights the ambiguous nature of evidence and the processes by which research may be used strategically in policy debates [101]. For example, Latour has highlighted the malleability of evidence so that its meaning may be reinterpreted, often in strategic ways, during policy debates [102], [103]. In this case study it is clear that econometric modelling carried out by a team at Sheffield University has been particularly influential but that rival perspectives exist as to the validity and utility of this research. Again, this aspect of the policy debate will be explored in detail in subsequent analyses [104].