This paper advances the theoretical framework of the “stagist heuristic” framework or sometimes known as the ‘process model’ in attempting to explain and analyze the policy activities which led to the enactment of Quebec’s Tobacco Act  of 1998. The main premise of this paper is to evaluate the usefulness of the process model in understanding the policy making process through a comparative study between this model and the Advocacy Coalition framework (ACF). I employ, and borrow, the case study of Bretton et al., (2008) that offers an alternative outlook to the developments of the Tobacco Act using the Advocacy Coalition Framework. This paper concludes with a discussion of the models which satisfactorily reflect the reality of how policies are formulated and enforced.
This paper is organized into 4 parts. Part one sketches the theoretical perspectives of the ‘stagist heuristic’ model and examines the factors and processes leading to the adoption of the Tobacco Act. Crucially, this part will highlight the critical role of policy actors in affecting policy processes and outcomes. Part Two provides a critical analysis to the effectiveness of the model by elaborating the advantages of the model. Part Three will go on to provide criticisms of the model by comparing it with the advocacy coalition framework used in analyzing the Tobacco Act of 1998. This part will present the many criticisms of the stagist model, using mainly contributions offered by Lindblom & Woodhouse (1993) and Sabatier (1999). Finally, Part Four concludes with a brief overall assessment of the framework, considering in particular, its status as an analytical tool for understanding policy making in the real world.
In the context of this paper, policy analysis is defined as ‘a set of interrelated decisions taken by a group of political actor or group of actors concerning the selection of goals and the means of achieving them within a specified situation where those decisions should, in principle, be within the power of those actors to achieve’ (Jenkins, 1978: 35). From Jenkins’s (1978) definition above, which acknowledges public policy as ‘a set of interrelated decisions’ taken by numerous individuals and organizations in government, I will form the basis of this paper. I will focus solely in understanding the ‘processes’ or what Jenkins (1978) referred to as “interrelated decisions” leading to the adoption of the Tobacco Act. His definition also correlates to Lasswell’s conceptualization of “knowledge of” rather than “knowledge in” policy making, in which the latter (i.e. knowledge in), is more substantive and prescriptive (Dunn, 1981; Hogwood & Gunn, 1984, Hill, 1993).
The policy domain is inherently complex, and so analysts have made use of various models of simplification to comprehend the overwhelming situation and to understand it more thoroughly. Through the lens of the ‘stagist heuristic’ model, policy analysts have been able to synthesize the complexity of such process into a series of functional phases, which frame this overtly political process as a continuous process of policy making.
1.1. The ‘Stagist Heuristic’ Framework
As pioneered by Lasswell (1956), and modified by Jones (1970), Mack (1971), Rose (1973), Anderson (1975), Jenkins (1978), Brewer & De Leon (1983) and Hogwood & Gunn (1984), this ideal-type framework adopts a technocratic approach to public policymaking, embracing linear and logical progression from agenda setting and concluding with policy evaluation and termination. The chronological orders of the policy life cycle are commonly categorized as problem definition, agenda-setting, policy formulation, implementation and finally evaluation (Dunn, 1981; Hogwood & Gunn, 1984; Sabatier, 1999; Dye, 2002; Colebatch, 2002).
1.1.2. Problem recognition and definition.
Hitherto, the greatest impetus to the developments of policy science crystallizes on a response to a myriad of social problems within, what Lasswell terms as “policy orientation” (cited in Dunn, 1981; Hogwood & Gunn, 1984; Howlett & Ramesh, 2003). Similarly, the process model presupposes the recognition of problem triggered by a ‘felt existence of problems or opportunities’ (Dunn, 1981). A problem is defined as an ‘unrealized value, need, or opportunity which, however identified, may be attained through public attention’ (Dunn, 1981: 44) which needs to do something about as pointed out by Wildavsky (1979) ‘a difficulty is a problem only if something can be done about it’ (Wildavsky, 1979:26).
However, problem recognition and definition are not straight forward activities. According to Birkland (2007) because a problem is a process of social construction, as mirrored by Dunn (1981) who states how the ‘problem is in the eye of the beholder’ (Dunn, 1981: 27), it depends on subjectivity of interpretations held by various stakeholders. And so, the majority ruling may be ill-defined and, at times, may even be “misframed”  (Baker, 1977). In addition, as Steiss & Deneke (1980) suggests, problems are ‘seldom mutually exclusive’ because ‘they often exist in a hierarchical relationship to one another, and the solution of one may depend on the solution of another, either higher or lower in the hierarchy’ (Hogwood & Gunn, 1984: 124); therefore may often lead to a further redefinition and modification of the problematical situations (Wohlestetter, 1976; Wildavsky, 1979; McRae & Wilde, 1985), which, in turn, lead to the creation and realization of more problems (Wildavsky, 1979), which I go on to address in the following paragraph.
Quebec’s Tobacco Act was primarily enacted as a response to the growing concerns of the public towards the issue of passive smokers or secondary smokers. As reported by Breton et al. (2008), the Tobacco Act was enacted to ‘protect the fundamental right of non smokers to enjoy a smoke free environment than by the harms to health’ (Breton, et al., 2008: 1682). However, the definition of the issue leads to the discovery of more social problems. On one hand, problems such as addictiveness of smoking, prevalence of youth smokers (which have significantly lead to raising educational awareness of the hazards of smoking), how the majority of the population are non smokers and, finally, the financial burden to the public health care system are brought to attention. On the other hand, protesters of the bill have contested the lethality to passive smokers, arguing that such intervention might impede the competitiveness of the tobacco industry (through the implementation of tax) and, thus, affect the economy of the province.
In liberal democracies, such as Quebec, problem identification and definition are conceptualized as highly pluralistic, involving diverse policy stakeholders such as the public (population, retailers), individuals (Minister of Health), organizations (e.g. Quebec Division of Cancer society, Ministry of Environment, Ministry of Health, Hospital industry), interest groups (e.g. Tobacco manufacturers, Non Smokers Right Association, Tobacco workers union, Events Rallying for the Freedom of Sponsorship group), the media, policy communities (Regional Public Health Directorates [RPHD], columnists and journalists, Quebec Coalition for Tobacco Control [CQCT] and also mentioned U.S administration) (Dunn, 1981; Sabatier, 1991; Kingdon, 1995; Dye, 2002; Howlett & Ramesh, 2003); the actual agenda setting is characterized by different patterns in terms of actor composition and the role of public. There are outside initiation as well as inside initiation  (May, 1991 cited in Fisher, Miller & Sidney, 2007); mobilization and consolidation  (Howlett & Ramesh, 2003). In this case, the tobacco control adheres to Howlett & Ramesh (2003) concept of consolidation; whereby due to the impending agitation of the issue amongst the public and subsequent contraband crisis of cigarettes smuggling in the US, policy elites (prominently the new Minister of Health and the National Assembly [NMA]) have seized the opportunity for government legitimacy in tobacco control by, effectively, propagandizing the issue to the public via ‘regularly intervening in the media on different aspects of tobacco control and visiting MNA’s cabinets’ (Breton et al., 2008: 1685).
1.1.3. Agenda setting
Next, I explore the agenda setting phase which Birkland (2007) defined as ‘the process by which problems and alternative solutions gain or lose public and elite attention’ (cited in Fisher, Miller & Sidney, 2007: 63). The elevation from systemic agenda into institutional agenda  is usually dominated by power struggles between groups competing to elevate or block issues from reaching the institutional agenda (Cobb & Ross, 1997); acting singly or, more often, by building strategic coalition with others (Sabatier, 1991; Lindblom & Woodhouse, 1993).
From the case study, this process is signaled by the agenda of the new Minister of Health and the tabling of the bill by the Council of Ministers at the legislative meeting. As part of his strategy in building a winning coalition, the Minister successfully gathers allies and supporters for the bill by establishing the CQCT; embracing Sabatier’s ACF model. In addition, Breton et al. (2008) mentioned that the Minister of Health has also announced plans to include in the bill provisions that prohibit active involvement of tobacco companies in sponsoring arts and sports events. Spearheaded by the centralized coordination of the CQCT, the winning coalitions which compose of Non Smokers Association and various municipalities through representatives from the RPHD, effectively debated the bill and gathered political support from the Council of Ministers at the parliamentary commission meetings, which resulted in the official adoption of the bill on February, 1998.
1.1.4. Policy formulation and decision-making.
In the traditional stage model of the public policy process, policy formulation is part of the pre-decision phase of policy making in which the political interchange described by Lindblom (1993) as “competition of ideas” emanates. It involves identifying and/or crafting a set of policy alternatives to address a problem, and narrowing that set of solutions in preparation for the final policy decision. This approach to policy formulation, embedded in a stages model of the policy process, assumes that participants in the policy process recognize and define a policy problem, consequently moving it onto the policy agenda.
During this stage of the policy cycle, expressed problems, proposals and demands are transformed into government programs. At the same time, studies of policy formulation have been strongly dominated by the effort to improve practices within governments by introducing the techniques and tools of rational decision making. In all political systems people gather facts, interpret them and debate issues. This stage is when the Minister establishes centralized command through CQCT to formulate the policy to tackle the issue of public smoking. In addition, the continuous dialogue and consultation involved in an agreement of the bill with NGOs, municipalities, health institution, local and regional organizations as well as oppositions falls into this stage. Crucially, the bill was also amended to streamline the phasing out of tobacco industry sponsorship but ‘offered no alternative solution to youth smoking and did not contest the actual harms on health of tobacco use’ (Breton, et al., 2008: 1686).
Brewer & DeLeon (1983) usefully define decision making as ‘the choice among policy alternatives that have been generated and their likely effects on the problem estimatedaˆ¦it is the most overtly political stage in so far as the many potential solutions are winnowed down and but one or a select few picked and readied for use.’ (Howlett & Ramesh, 2003:162). The models on decision making are classified as rationalism which asserts utility maximization to complex policy problems in which ‘policy relevant information was gathered and then focused in a scientific fashion on the assessment of policy options’ (Howlett & Ramesh, 2003:166); and incrementalism which describes policy making as a political activity of maintaining the status quo through gradual and continuation of past policies.
Dye (2002) defined implementation as ‘the implementation of policies through organized bureaucracies, public expenditures, and the activities of executive agencies’ (Dye, 2002: 15). Intra- and inter- organizational coordinating problem and interaction of field of agencies with the target group ranked as the most prominent variables accounting for implementation failures (MacRae & Wilde, 1985; Howlett & Ramesh, 2003). Another explanation focuses on the policy itself, acknowledging that unsuccessful policy implementation can be, though by no means the only, result of bad implementation, but also bad policy design, based on wrong assumptions about the cause-effect relationship (Hogwood & Gunn, 1984; Fisher et al., 2007: 52).
The study of implementation is dominated by the concept of top-down centralized implementation and bottom-up implementation. The top-down school or the “vertical” dimension represented, for instance, by scholars like Van Meter & Van Horn (1975), Hood (1976), Gunn (1978), Nakamura & Smallwood (1980) and Mazmanian & Sabatier (1983), conceive of implementation as ‘the hierarchical execution of centrally defined policy intentions’ (Fisher et al., 2007:89). Proponents of the bottom-up or “horizontal” approach include Lipsky (1971, 1980), Ingram (1977), Elmore (1980), and Hjern & Hull (1983) who have emphasized the fact that implementation consists of everyday problem strategies of “street-level bureaucrats” (Pressman & Wildavsky, 1973; Colebatch, 2002; Fisher et al., 2007). In this case, the policy implementation correlates closely with the top-down approach because the implementation is based on the commitments and directives from the top echelon of the government i.e. Minister of Health and Ministry of Health.
Finally, evaluation is the post hoc analysis of policies and programmes carried out by ‘government agencies themselves, outside consultants, the press, and the public’ (Dye, 2002:15) through ‘collecting, testing, and interpreting information about the implementation and effectiveness of existing policies and public programmes’ (Majone, 1989: 167). The plausible normative rationale is that policy making should be appraised against intended objectives and impacts form the starting point of policy evaluation, which forms the basis for justifying government actions for continuation or termination of public programmes and enables accountability of government offices especially in democratic setting (Majone, 1989). However, from the case study, it is unclear whether any form of evaluation was carried out or not.
Despite depicting the developments of the Tobacco Act in a series of stages, as mentioned by Hogwood & Gunn (1984) in ‘the dividing lines between the various activities are artificial and policy makers are unlikely to perform them consciously or in the implied logical order’ (Hogwood & Gunn, 58), Lasswell, as Hudson & Lowe (2004) note, did not conceptualized these stages as ‘real’, in the sense that they encompass clear beginnings and ends. Rather, ‘their function being merely analytic-to help us explore different dimensions of the policy process. He [Lasswell] is more concerned with the value systems, institutions and wider social processes that shaped policy in the real world’ (Hudson & Lowe, 2004: 5). Therefore the process model does provide valuable descriptive analysis of the policy process.
As explained above, the process model helps to disaggregate an otherwise seamless web of public policy transactions, as each segment and transition are distinguished by differentiated actions and purposes. Furthermore, the cumulative analyses of the various stages, arguably, contribute to the disentangling of the intricate political and social interdependencies, manifested in the policy arena, to bring about an ordered and manageable ‘system’  (DeLeon, 1983).
Furthermore, this process framework has significant ‘strategic’ implications. Firstly, by analyzing the policy actors and processes in discrete stages, it assists in identifying how stakeholders may support or resist health policies (ODI, 2007); and therefore develop strategies in building winning coalitions as mentioned by Easton (1979) which states how the process model ‘lend themselves to the identification and study of interactions, not only among the various stages in the process but also among various participating organizations and between organizations and the lager social and economic environment’ (cited in Hogwood & Gunn, 1989: 25). Although this might be more applicable to the ACF, such advantage also applies to the process model especially during the agenda setting phase. As described above, in the agenda setting phase, the process model highlighted and identified various policy stakeholders and analyzed the relationship of ‘policy advocacies’ which resisted (Tobacco Manufacturers and Tobacco Worker’s Union) against those whom supported the bill (Minister of Health and Non Smokers Association); thereby enabling the assessment of ‘the cumulative effects of various actors  , forces, and institutions that interacts in the policy process and therefore shape its outcome(s)’ (Jann & Wegrich  cited in Fisher, Miller, & Sidney, 2007: 44).
Secondly, it also helps in identifying and addressing various obstacles that undermine successful implementation of policies (ODI, 2007). The process model follows the assumption of how public policy making is a goal oriented process aimed ‘to reach a goal or realize an objective or a purpose’ (Anderson, 1984 cited in Colebatch, 2002:85), henceforth policy makers are able to identify constraints, which in this case, a negotiation with oppositions and gathering public support for the bill ensured the successful adoption of the bill.
Finally as pointed out by Hogwood & Gunn (1984), ‘the process framework is rather flexible in the sense that it enables us to systematize existing knowledge without precluding the integrating of future insights (about stages, influences, interactions, etc) to the framework’ (Hogwood & Gunn, 1984: 25). In other words, it improves the prospects of technical evidence considered during policy formulation leading to evidence based policy making. The most common method in the British government in gathering technical information for systematic analysis of policies is through trial and error achieved by carrying out a pilot test before actual implementation of policies.
3. Criticisms: A better understanding in policy making.
On the contrary, Parkinson (2008) in his lecture, quite rightly so, argue that the process model resembles a mechanistic tool that describes checklists of “parts” present in the policy making arena; parallel to Nakamura’s (1987) notion of a “textbook approach” (Sabatier, 1999).
Henceforth, the top down legalistic framework is an artificial portrayal of the policy process (Dunn, 1981; Sabatier, 1999) as stated by Lindblom (1993) that ‘deliberate, orderly steps are therefore not an accurate portrayal of how the policy process actually works. Policy making is, instead, a complexly inter-active process without beginning or end’ (Lindblom & Woodhouse, 1993: 11). In other words, these processes do not evolve in a pattern of clear cut sequences; instead the stages are constantly meshed and entangled in an ongoing process which is more accurately resembles a “primeval soup” (Kingdon, 1995; Howlett & Ramesh; 1995). Therefore, the process model leads to the imposition of hypothetical explanation of future events which may be inappropriate or misleading with ‘actions occurring fitfully as problems become matched with policy ideas considered to be in the political interests of a working majority of the partisans with influence over the policy domain’ (Lindblom & Woodhouse, 1993: 10).
3.1. Rationalization of processes.
Hogwood & Gunn (1984) question the coherence and rationality of the process model as a “blueprint” for action by ‘giving rational explanation or justification of past acts, even when the acts in question do not lend themselves to such treatment’ (Hogwood & Gunn, 1984: 26). Furthermore, Lindblom (1993) also argue that the stages are not hierarchical which proceeds from agenda setting and concluding with evaluation; rather they often overlap loop” with each other as analysis proceeds. This is further elaborated below.
Firstly, Lindblom (1993) argue that ‘there may not even be a stage when problem definition occurs, since participants often vary widely in their ideas about the “problem” (Lindblom & Woodhouse, 1993:10). He explains that this is because ‘policy sometimes is formed from a compromise among political participants, moreover, none of whom had in mind quite the problem to which the agreed policy responds’ (Lindblom & Woodhouse, 1993:10).
Secondly, Lindblom & Woodhouse (1993) also point out the inaccuracy to suggest that the “decision-making” phase exist. As suggested by Heclo (1972) ‘a policy can consist of what is not being done’ (Hogwood & Gunn, 1984: 21) and, thus, equally important, are the decisions to keep issues, that would be inconvenient, firmly off the agenda for political success in winning the disputes that arise. In other words, policy may emerge without any explicit decision, by failure to act as or the power of “nondecision making” (Bachrach & Baratz, 1962; Heclo, 1972). Bachrach & Baratz (1962) which exhibits the existence of institutional bias so that key groups are excluded in what is termed as the “three dimensional view of power”, in which power is used to exercise ‘to control over the agenda of politics and of the ways in which potential issues are kept out of the political process’ (Lukes, 2005: 25). Furthermore, stating decisions are taken exclusively in the “decision-making” phase is rather inaccurate, because in reality, decisions are constantly being made regardless of the stages you are in. For example, during the policy formulation, policy makers makes decisions on which alternatives to adopt for consideration and hence to implement; and during the implementation stages, policy makers make decisions on the choices of policy instruments to be utilized (Hill, 1993; Howlett & Ramesh, 2003).
Thirdly, Lindblom & Woodhouse (1993) also argued that implementation and evaluation cannot be separated from the other steps. As mentioned by Lindblom & Woodhouse (1993) ‘an attempt to implement one policy almost always brings new problems onto the agenda, meaning the implementation and the step called agenda building collapse into each other’ (Lindblom & Woodhouse, 1993: 10). An example from the case study is that during the implementation of the Quebec Tobacco Act, to include taxation on tobacco and a ban on tobacco sponsored arts and sports event, subsequently led to the discovery that such measure might impinge the competitiveness of tobacco industries and affect the economy of the province.
Finally, policy evaluation often regarded as the end of the line, does not actually constitute a “step” in policy making unless it throws light on possible next moves in policy, in which case ‘evaluation becomes intertwined with all other attempts to appraise and formulate options for reshaping government activity’ (Lindblom & Woodhouse, 1993: 10). Moreover, I think that the evaluation phase overlaps with the agenda setting phase and the policy formulation phase. During the agenda setting and policy formulation phase, policies are also evaluated needed to persuade and influence people in adopting and supporting the bill.
3.2. Multiplicity of interactions.
On the other hand, Sabatier (1999) note the framework oriented scholars towards looking at just one stage at a time, thus neglecting the entirety of the process by stating that ‘they portrayed a disjointed, episodic process rather than a more ongoing, continuous one’ (Sabatier, 1999: 23). In addition, Sabatier & Jenkins Smith (1999) set out 5 major deficiency of the heuristic approach; it provides little description of how policy moves from one stage to another; it cannot be tested empirically; it is essentially a ‘top down’ which fails to take account of ‘street-level’ and other actors; and it disregards multiple levels of governmental interactions. Finally, it does not provide an integrated view on the gathering of policy related information, apart from the ‘evaluation’ phase (Parsons, 1995; Sabatier, 1999) as pointed out by (Majone, 1989) the effectiveness in solving social problems centres in bringing more information and systematic analysis into the policy making process.
From the case study, it is clear that the process model is limited in its capacity to provide institutional analysis of government interactions because it is primarily conceived to provide systematic analysis of the overall policy making process, unlike “institutionalism” perspective which focuses on the role and relationship of government institutions which regards public policy as an institutional output of the mechanisms of the government where it is ‘authoritatively determined, implemented and enforced by these institutions’ (Dye, 2002:12). Furthermore, the process model adheres to the view that policy making is a hierarchical top down process which initiates from agenda setting and finally ends with evaluation stage and therefore only takes account of authorized decision makers. Finally it is also rather limited in empirical research on each stage and only makes an attempt to describe systematic gathering of information in the evaluation phase only. However, on the other hand, I would have to disagree with Sabatier (1999) in that the process model does not provide clear distinction between the stages and the progression from one phase to another. I think the primary distinction of the stagist model lies in the context of “policy environment” and “policy stakeholders” involved  . Henceforth taking the definitions which I presented above of each stage and the ones offered by Dye (2002: 14-15), the demarcations between the stages are summarized in the table below:
Policy Stakeholder (i.e. who are involved)
Policy Environment (i.e. where does it take place)
1. Problem identification
Individuals, public and private organizations, interest groups, think tanks, mass media and policy communities.
Public debates, consultation with public, and sometimes top level government officials identify it themselves.
2. Agenda setting
Public officials acting as “gatekeepers” as well as involvement of “policy entrepreneurs”.
Mostly done by the Executive branch of the government and in government offices.
3. Policy formulation
Primarily done by government officials in Executive agencies, but may also involve interest groups, congressional committees, and think tanks.
Again done in “Executive” government offices but may also involve the “Legislative” branch of the government i.e. Parliament or Senate.
Primarily “street-level” bureaucrats and occasionally involving public participation.
Carried out in formal government institutions.
Done by government agencies but may also involve the public through medias, consultants and think tank organizations. Also very important is the use of citizen juries to evaluate public programmes.
Evaluation is carried out in government offices, but also may be carried out in NGO organizations (such as EU, UN etc) and non-governmental institutions.
In addition, Breton et al., (2008) successfully utilize the ACF to explain how the interactions of multiple policy advocacies have impacted policy change, which is another major deficiency of the stagist heuristic model. As mentioned by Majone (1989), ‘both continuity and change are inherent in the conception of policy’ (Majone, 1989: 35) and therefore should be accounted for in the models in its capacity to comprehensively capture the policy making process.
For example, the ACF manage to show how the changes in the external events directly impact the core beliefs of tobacco subsystems and hence resulted in the adoption of Tobacco Act (Refer to Fig.2 in Breton et al., 2002: 1683). However unlike the ACF, process model does not provide description on how policies are impacted by change. Moreover the process model assumes that every policies starts from scratch i.e. always starts by identification of problems. Conversely, policies may be enacted not from new problems or opportunities that emerge, rather continuation of past policies in which case, the “problem identification” phase may be invalid.
In conclusion, the process model provides valuable insights in directing analyst’s attention to critical features in the policymaking process, and on elucidating the policy process paradigm. Furthermore, although the ACF model is conceived to account for the entire policy process, it is limited in its capacity to explain only the policy formation (i.e. agenda setting and decision making). In other words, both models differ in their level of analysis, which I hope have been successfully demonstrated above.
On the other hand, the idea of breaking down the making of public policy into phases, may well impose stages on reality that is infinitely more complex, fluid and interactive; but to adopt a cyclical metaphor, it is not necessarily an unreasonable or unrealistic way of looking at what happens when public policy is made. Nonetheless, the process model does still provide some useful insights in public policy making. In my opinion, the most important thing is not to look at one “best” model to explain a particular policy rather a combination of models is needed as pointed out by Dye (2002: 12):
‘These models are not competitive in the sense that any one of them could be judged “best”. Each one provides a separate focus on political life, and each can help us to understand different things about public policy. Although some policies appear at first glance to lend themselves to explanation by one particular model, most policies are a combination of rational planning, incrementalism, interest group activity, elite preferences, game playing, public choice, political processes, and institutional influences.’