Policy change &
the punctuated equilibrium theory
A longitudinal study of clean air policy-change in Sweden during 2011–2019
Gustaf Borelius
Bachelor’s thesis, 15 credits Department of Government Political Science C, Autumn 2020
Supervisor: Andreas Gottardis Word count: 13 941
Pages: 44
Innehåll
Introduction ... 3
1.1 Main Purpose & Research Question... 4
1.2 Demarcations, Method, & Time Span ... 5
1.3 Disposition ... 6
Theory - Policy Change and the Punctuated Equilibrium Theory ... 7
2.1.0 Brief Overview of Theories and Measurement of Policy Change ... 7
2.1.1 - Overview of the Punctuated Equilibrium Theory of Policy change ... 8
2.2 Previous Research on the Punctuated Equilibrium Theory of Policy Change ... 11
2.3 Operationalization... 13
Methodology... 19
3.1 Case selection ... 19
3.2 Research design & limitations ... 20
Results and Analysis... 24
4.1 Policy composition up until 2011 ... 24
4.2 The dynamics of Clean Air Policy 2011-2019... 30
4.3 Analysis ... 35
4.4 Discussion ... 37
Conclusions ... 40
References ... 42
Introduction
Understanding how policies change and why has become a central topic in the study of public policy in political science. The process of politics in general and public policymaking in particular involves complex systems of interacting actors with different preferences, goals, and ideas, operating within sometimes constraining and other times enabling institutional and normative environments. We know for a fact that policies do undergo changes, sometimes rather small and trivial but at other times in revolutionary ways that significantly affects large numbers of people. How are we to make sense of these intricate and interactive processes of how policies undergo change? Luckily, there are theories that, much like a burning torch, allow us to navigate in the fog and the bog of the policy-process. One such, the punctuated equilibrium theory, leads us to think that policy processes are generally characterized by incrementalism - minimal change to existing policies. However, the stability, or equilibrium, of this process can on rare occasions be “punctuated” due to an exogenous shock such as e.g an economic crisis or a nuclear meltdown. The result of this puncture is theorized to lead to rapid and dramatic change in public policy. With this in mind, it not only becomes interesting to descriptively study how policy changes over time, but also to link this potential change (or absence of change) in policy to expectations derived from the punctuated equilibrium theory, in order to assess how it “holds up”.
While this theory encompasses policy change in very broad terms, meaning the
incremental/punctuated pattern is thought to explain policy dynamics in everything from health care to international trade policy, some areas seem especially fruitful. One of them - environmental policy, has been hinted at to provide fertile ground for the broader study of policy change, and for relating the find ings to the punctuated equilibrium theory.1 Moreover, the dynamics of environmental policy is by no means irrelevant outside the literature or discipline. Gaining an understanding of the patterns of policy change in general, and environmental policy in particular, can be of utmost importance to policy makers, climate activists, corporations, and in the end therefore even to the environment itself. For example, this knowledge could possibly make policy entrepreneurs ask how to position themselves or act if a punctuation seems likely, in order to either maximize or counteract the cascading effects of its nature.
1 Baumgartner, Frank M. 2006 . Punctuated Equilibrium Theory and Environmental Polic y, s.24-46 i Repetto &
Gustave Speth (red), Punctuated Equilibrium and the Dynamics of U.S. Environmental Policy. Yale University Press. s.24, 43f
Furthermore, since the literature has been heavily anglo-centric and federalism-favoured, it is interesting to examine if these theorized patterns hold true in underexplored contexts such as the Swedish case. It might also be compelling to study environmental policy in Sweden, among other policy areas, as it may sometimes be greatly influenced by decisions and ambitions on an EU-level. This provides a rather unique context of measuring policy change where the policymaking is dyed in both “the national” and “the international”. With this in mind, the study will seek to examine environmental policy change in Sweden, with a focus on clean air-policy. This is because it is thought to be a typical case of environmental policy and thus widely representative of environmental policy in general. Because of the nature of clean air-policy in general still being very extensive, due to the fact that air pollution is a phenomenon that stems from a very large variety of causes and measures for these causes, this study will specifically limit itself to studying traffic-related change in clean air-policy.
Traffic will henceforth be understood in the context of motor-traffic, i.e “on the road”, thereby excluding air and maritime traffic.
1.1 Main Purpose & Research Question
This study has three explicit purposes. The first purpose is to respond to one of the heavyweights in the litterature - Frank M. Baumtgartner, and his invitation for further intensive analysis on the process of policy change, thereby further closing the (albeit slight) gap between quantitative and qualitative studies on the subject.2 The second purpose of this study is to apply an improved theoretical framework of capturing the elements of policy change, in order to give a more empirically nuanced and methodologically valid view of the phenomenon compared to much of the previous literature on policy dynamics. The third and final purpose will be to relate the empirical findings to those hypothesized by one of the dominant theories of policy change - the punctuated equilibrium theory, in order to produce a descriptive contribution to the literature regarding the explanatory capacity (or lack thereof) the theory provides. By doing so it could also inspire future research to focus on the causality of the findings, further putting the theory to the test.
In order to attain these three purposes, this study will attempt to answer the following two research questions:
In which ways has Swedish clean air policy changed or not changed over time?
2 Baumgartner, Frank M. 2006. s.37ff
To what extent can the change and/or absence of change of the Swedish clean air policy be linked to expectations derived from the punctuated equilibrium theory?
1.2 Demarcations, Method, & Time Span
As mentioned in the introduction of this study, a demarcation of this study will focus on traffic-related policy change in clean air-policy. This has been done because of feasibility reasons, stemming from the fact that the causes of air pollution, and therefore the means and ends of clean air-policy, are large and thus any attempt at intensively approximating an exhaustive account of clean air-policy in any “broad” fashion would be too time consuming.
Furthermore, choosing to study clean air-policy with a specific focus on pollution caused by traffic was based on the idea that it would serve as a typical and therefore representative case of clean air-policy in general. This line of thinking stems from the fact that traffic-related pollution should be suitable since it is (and for a very long time has been) one of the most significant sources of pollution and therefore obstacles for attaining cleaner air.3
The method utilized in this study can be summed up as a longitudinal text analysis, with the point of departure regarding material being on analyzing annual reports of the Swedish Environmental Agency on their assessments of the environmental quality objectives
(miljömålen). In addition to these annual reports, complementary materials such as various relevant governmental bills, reports, and EU-documents will be employed and analyzed as seen fit. This text-analysis will be conducted by using an analytical framework developed by Cashore & Howlett, that enables us to capture the elements of policy from our material in a systematic way to then asess possible changes. This is important because it gives us the prerequisites to answer our first research question, as well as link our results to the broader theory of the punctuated equilibrium theory - thereby also enabling an answer to our second research question.
Because this framework focuses on the more abstract ends or aims of a policy, as well as the means or tools for attaining them (as will be shown later); anchoring this study primarily in these mentioned annual reports is suitable as their purpose is to “present an overview of the
3 Naturvårdsverket. 2019. Frisk luft - underlagsrapport till den fördjupade utvärderingen av miljömålen (ISBN 978-91-620-6861-5). s.33
aims of the environmental aims” and to “assess how the efforts (to attain these aims) is progressing”.4
The time frame of this study covers the period 2011-2019, being selected based on theoretical considerations derived from the punctuated equilibrium theory (which is explained more in depth in the section of research design), as well as the insights that any analysis of policy development often requires studying a period of at least several years.5
1.3 Disposition
The continued disposition of this study will be as follows. Firstly, a theoretical introduction of policy change in general will be introduced, followed by a specific exposition of the punctuated equilibrium theory (PET) of policy change and some expectations we can derive from it. Following this, a review of previous research on the punctuated equilibrium theory of policy change will be presented, as well as some problems that plague the study of policy change in general. Subsequently, an operationalization section will present an analytical framework that demonstrates how to empirically capture the components of a policy, and therefore give us a tool for measuring policy change over time. Added to that, a smaller analytical framework will be presented that lets us assess the mode (or degree) of policy change through an interplay between its tempo and direction.
The third section of this study will cover the general methodology of the study, starting with an overview and selection and motivation of case. Followed by this will be an in-depth coverage on the chosen research design and its limitations, as well as a short methodological discussion on alternate routes that could’ve been chosen instead. In the fourth and last section, the results, analysis, and conclusion of this study will take place.
The results start off by using the main framework of this study (the one by Cashore and Howlett) in order to map out the components of clean air policy in 2011. It will then be attempted to trace how these components undergo change during the years, again through text analysis of mainly annual governmental reports, leading up until 2019. After summarizing how the clean air-policy has undergone change throughout this time span, an analysis with
4 Naturvårdsverket 2012 Steg på vägen - Fördjupad utvärdering av miljömålen 2012 (ISBN 978-91-620-6500-3) s.3
5 Cashore, Benjamin & Howlett, Michael, 2009. The Dependent Variable Problem in th e study of Policy Change:
Understanding Policy Change as a Methodological Problem. Journal of Comparative Policy Analysis 11(1): 33- 46. s.35
the help of the second analytical framework will be used in an attempt to classify the mode of the policy change. Note that at this point, the first research question will have been answered.
Lastly, a discussion regarding the results and analysis will be conducted, mainly by relating it to the punctuated equilibrium theory and its expectations of how the policy should have undergone change. Thus the second research question should have been answered.
Theory - Policy Change and the Punctuated Equilibrium Theory
2.1.0 Brief Overview of Theories and Measurement of Policy Change
The study of policy change can be treated as a core area of public policy, with a large interest in the description and explanation of changes in dominant policy patterns across diff erent fields.6 In the contemporary literature, there are several ways in which policy change can be measured, and multiple theories on why policy change does (or does not) occur; and so far, none of the approaches has achieved the position as dominant or generally accepted.7
The theories of how and why policy change occurs can be can according to Knill & Tosun be classified in accordance with their underlying causal logic.8 In their summary, there are on the one hand approaches that are anchored in a “linear-additive” view of causality, where a clear distinction between independent and dependent variables is assumed; where
independent variables mostly refer to macro-factors such as changes in government, socio- economic conditions, strength of societal interest groups, and so on.9 On the other hand, there are approaches based on a logic of “combinative causality'', where possible combinations of causal conditions able to generate a specific result are searched for.10
Theories with the underlying “combinative causality” logic for explaining policy change are generally characterized for being inclusive in integrating structures, institutions, and actors, although they often place specific emphasis on certain explanatory factors.11 The punctuated equilibrium theory can be categorized as a theory with the “combinative causality” approach, with a specific emphasis on institutional factors, as will be seen below.
6 Knill, Christoph & Tosun, Jale 2012. Public Policy: A New Introduction. London: Palgrave Macmillan. s.251
7 Knill & Tosun, 2012. s.251
8 Knill & Tosun, 2012. s.252
9 Knil & Tosun, 2012. s.252
10 Knill & Tosun, 2012. s.252
11 Knill & Tosun, 2012. s.253
There are also several ways to go about measuring policy change, with frameworks developed by e.g Sabaiter and Jenkins-smith that can be summarized as distinguishing between changes in core beliefs and changes in secondary aspects, with core beliefs being more difficult than secondary aspects to attain (thus indicating a paradigmatic policy change)12. However, in the study of policy change in relation to the punctuated equilibrium theory, a framework for measuring policy change developed by Peter Hall has been very popular - being the most often cited in the litterature and applied in empirical studies.
However, due to having issues such as conflating the elements of the dependent variable (amongst other things), Cashore & Howlett developed a framework based upon Hall’s, which serve as the instrument of measurement for this study. An overview of these issues, as well as Cashore & Howlett’s framework, will be covered in the operationalization section later.
2.1.1 - Overview of the Punctuated Equilibrium Theory of Policy change
The punctuated equilibrium theory was originally developed by Baumgartner & Jones, and strives to explain why political and policy processes usually are characterized by stability and incrementalism - an absence of noteworthy change from the status quo; but on rare occasions produce dramatic shifts away from this status quo.13 It is an attempt to combine the two most historically prominent streams of thinking in policy dynamics, Charles Lindblom’s work on incrementalism in 1959 and Peter Hall’s study of policy paradigms in 1989, breaking the long term orthodoxy of incrementalism.14 To begin with, the PET can be broken down into two central components that are fundamental in explaining the duality of change in the policy process: a positive and a negative feedback system.
The negative feedback system’s most important feature is that it can be described as
containing a self-correcting mechanism that works to maintain stability in a political system.
Similarly to Isaac Newton’s third law of motion, the self correcting mechanism reacts in the opposite way of external pressures trying to induce policy change, counterbalancing the push for change and thus works in maintaining the often prevailing status quo (and keeping the policy change in a form of an equilibrium).15 The PET is built on the idea that policymaking
12 Knill & Tosun, 2012. s.260
13 Knill & Tosun, 2012. s.111, 255f
14 Cashore & Howlett, 2009. s.34
15 Baumgartner, Frank R. & Jones, Bryan D. 2002. Policy Dynamics. Chicago: The University of Chicago Press.
s.8f; van der Heijden, Jeroen & Kuhlmann, Johanna, 2018. What Is Known about Punctuated Equilibrium
is mainly conducted in policy subsystems, outside the world of high politics; where these subsystems are defined as institutional arenas with capacity to handle many policy issues routinely, and are populated by experts, bureaucrats, and interest groups.16 These policy subsystems, also called policy monopolies, are in situations of negative feedback dominated by a stable set of policy images, that is, certain norms and ideas that limit how policy issues can be discussed.17 In addition to policy images, there are in the context of negative feedback given institutional arenas where policymaking takes place, in the litterature often called policy venues, that limit who can and cannot participate in the policy debate.18 In situations where policy image remains stable, or at least resistant to radical rethinking because of the self correcting nature of the political system; coupled with policy venues in which
institutional rules are clear in structuring who can be seen as a legitimate participant in the policymaking process and who can be defined as an outsider, we should expect stable policy monopolies and therefore an incremental patterns of policy change.19
However, these policy monopolies, though resistant to destabilization, are not invincible.
Through exogenous events (also called external shocks or perturbations), which according to Knill & Tosun bears the highest potential to alter both policy venues and images, stable policy monopolies can be broken.20
It is e.g through shining a strong light (which cannot be ignored) on a certain dimension of a policy that has previously been ignored or underemphasized by the prevailing policy image that these external shocks may cause a shift in the policy monopoly. This results in a punctuation, creating opportunities for both new actors to make legitimate claims of taking part in existing policy venues, which leads to an introduction of new considerations of norms and ideas regarding the policy issues (a redefinition of the policy image), which again can lead to even further opportunity for new actors to enter the policy venue, and so on... It is this
Theory? And What Does That Tell Us about the Construction Validation, and Replication of Knowledge in the Policy Sciences? The Review of policy research 35(2): 326-347. s.328f
16 Jennings, Farrall, Gray, & Hay, 2020. Moral Panics and Punctuated Equilibrium in Public Policy: An Analysis of the Criminal Justice Policy Agenda in Britain. Policy Studies Journal 48(1): 207-234, s.209
17 Baumgartner & Jones, 2002. s.12f; Jennings, Farrall, Gray, & Hay, 2020. s.209
18 Baumgartner & Jones, 2002. s.12; Knill & Tosun, 2012. s.111
19 Baumgartner & Jones, 2002. s.12
20 Knill & Tosun, 2012. s.111
self-perpetuating process that has been coined the positive feedback mechanism, and it can lead to a cascading and explosive (often called paradigmatic) change of the policies
themselves, through this destabilization of the policy monopoly. A hypothetical example of this could be for instance nuclear energy policy undergoing dramatic change through a nuclear reactor meltdown. Before this exogenous event, the prevailing policy image may be viewing nuclear energy in terms of cost-effectiveness and reliable energy production.
However, after this event the policy image may undergo a dramatic shift to nuclear energy being hazardous to both humans and the environment, paving way for new actors, such as e.g anti-nuclear environmentalists, to claim seats in the policy venue. As previously explained, we now have fertile conditions for a positive feedback process to take place, and in the context mentioned above it could probably lead to anything from tightened restrictions and safety protocols to plans of a full out decommissioning of nuclear power plants in a country.
policy image thought to change before policy venue, even though venue as can be seen can change policy image.
From the punctuated equilibrium theory, we can derive the following expectations or hypothises from the nature of the policy change, which we later can attempt to compare empirically:
- Incremental change should be the dominant form of policy change in the absence of an exogenous event
- If an exogenous event takes place, paradigmatic policy change should take precedence and follow relatively close in time
2.2 Previous Research on the Punctuated Equilibrium Theory of Policy Change Previous research on studying and exploring punctuated equilibrium theory (PET) in policy change have been quite vibrant. There has been much empirical support that change in policies in the long run generally follow the patterns of positive (paradigmatic) and negative (incremental) feedback that the PET theorizes. In an anthology of essays edited by Frank R.
Baumgartner & Brian Jones (the original architects of the PET-approach), with the
collaborating scholars using a then previously unprecedented data set, the authors draw the general conclusions that their essays have introduced a great variety of evidence in support of the ideas that government responses to public policy have been subject to positive and
negative feedback processes in the American context (in other words, support for the
punctuated equilibrium theory).21 As seen below, studies have discovered this to be the case in other nations, and in fact, findings have even emerged that policy making even on the international level exhibits policy change patterns consistent with the punctuated equilibria approach.22
In a meta-review conducted by Kuhlmann & van der Heijden regarding the application of the punctuated equilibrium theory on policy change, amongst other things two interesting
findings can be observed. Firstly, they find that studies focusing on the United States are dominant in the field and making up 59% of publications.23 Outside of the U.S, Anglo countries, especially the UK, make up a substantial remainder of studies (14%); however, PET is gaining traction in studying policy change in other european countries as well, such as Denmark (10%) and Belgium (9%).24 Kuhlmann & van der Heijden also finds that from the 86 articles part of the sample, 25 of them explicitly test hypotheses derived from the
punctuated equilibrium theory, where 17 of these studies confirm, 6 of them partially confirm, and 2 of them do not confirm theoretical expectations.25
21 Baumgartner, Frank R. & Jones, Bryan D. 2002. Policy Dynamics. s.291
22 Lundgren, Magnus; Squatrito, Theresa, & Tallberg, Jonas, 2017. Stability and change in internation al policy- making: A punctuated equilibrium approach. The Review of International Organisations 13(4) 547-572
23 van der Heijden & Kuhlmann, 2018. s.334
24 Ibid s.334
25 Ibid s.340f
Of the remaining 61 studies included in the sample, who refrain from explicitly testing hypotheses, 47 of them state that their empirical study confirms theoretical expectations, 8 of them partially do, and 6 do not.26 This can be interpreted as an indication that the punctuated equilibrium theory in general does a good job in explaining policy change.
Additionally, two notable observations one can make regarding the literature is that previous research is somewhat skewed toward the more extensive (quantitative) ways of gathering and analyzing data; as well as mostly focusing on analyzing the statistical distribution of many cases (and the appearance of this distribution) rather than regression techniques which are more common in the quantitative field of political science.27 While there does exist a respectable number of more intensive and case-like studies of single policies over time in order to understand the given contexts and details underlying the shift from incremental to paradigmatic policy change, such as in Cashore & Howlett’s study of forest policy changes in the U.S Pacific Northwest, these are to a certain extent in shorter supply.28 This first
observation is also buttressed by Kuhlmann & van der Heijden, stating that approximately 42% of approaches are of qualitative nature, 48% on quantitative, and the rest relying on a mixed-method design.29 Of course, as Baumgartner notes in his lessons for case analysis on the subject, the extensive analyses must be strengthened with intensive analyses containing these previously mentioned contexts and details in order to understand the process of policy change in a more complete fashion, as he invites for more of these types of studies.30
The literature of policy dynamics does, however, not go without its definitional and
operationazional disputes. In two of their articles, Cashore & Howlett criticises a large part of the previous literature on policy dynamics for conflating several distinct change processes present in specific elements of policy, in other words conflating “the dependent variable”.31
26 Ibid s.340f
27 Baumgartner, Frank M. 2006. s.26f
28 Cashore, Benjamin & Howlett, Michael, 2006. Behavioral Thresholds and Institutional Rigidities as
Explanations of Punctuated Equilibrium Processes in the Pacific Northwest Forest Policy Dynamics, s. 137 -161, i Repetto & Gustave Speth (red), Punctuated Equilibrium and the Dynamics of U.S. Environmental Policy. Yale University Press.
29 van der Heijden & Kuhlmann, 2018. s.339
30 Baumgartner, Frank M. 2006. s.36ff
31 Cashore & Howlett, 2009. s.37ff
Cashore & Howlett attribute this to the fact that scholars in the research area mainly use a
“three order model”-taxonomy devised by Peter Hall in order to measure policy change, which they mean lacks the capacity to disentangle these policies in a proper fashion. To overcome (or at least to alleviate) this “dependent variable problem”, Cashore and Howlett develop a more contemporary and improved taxonomy of policy composition based on Hall’s work, which they argue will lead to a more accurate picture of the actual patterns of change present in empirical cases.32
2.3 Operationalization
In order to study policy change over time empirically, and thus be able to compare our results to the punctuated equilibrium theory, an analytical tool is required which can systematically capture important components of a policy that is subject to change. As teased in the section summarizing previous research on the area, a taxonomy of policy components proposed by Cashore & Howell will be utilized in this study.33 This framework draws upon the original work of Hall’s “three order model” in classifying policy composition, where he has identified three policy elements subject to change: “The overarching goals on a high level abstraction that guide policy in a particular field, the policy instruments on a programme level used to attain these goals, and the settings of these instruments in an on-the-ground level.34 Hall saw the need to distinguish between these different elements in order to gain improved insights in the patterns of policy stability and development, because previous research on policy change before Hall’s model tended to conflate these elements into a single dependent variable.35 However, in light of new empirical evidence that has emerged since, Cashore & Howlett has argued that Hall’s efforts require recalibration in light of its own logic, as his three-order model still struggles with the problem of conflating important and seperate policy elements.36 Cashore & Howlett expands on Hall’s model by differentiating between policy aims and means, as well as between levels of abstraction: high level abstraction, specific on the ground
32 Cashore & Howlett 2009. s.39-42; see also Cashore, Benjamin & Howlett, Michael, 2007. Punctuating Which Equilibrium? Understanding Thermostatic Policy Dynamics in Pacific Northwest Forestry. American journal of political science 51(3):532-551. s.537.
33 Ibid.
34 Moore, Brendan, & Jordan, Andrew, 2020. Disaggregating the dependent variable in policy feedback research: an analysis of the EU Emissions Trading System. Policy sciences 53(2): 291-307. s.295
35 Cashore & Howlett 2009. s.36
36 Ibid, s.38
measures (less abstract), and specific on-the-ground measures (least abstract).37 By doing this, Cashore & Howlett distinguishes between six elements of policy that can undergo change instead of Hall’s three, which will be overviewed in the following.
Firstly, a policy can be overarchingly divided into policy ends or aims and policy means or tools. Policy ends or aims can in turn be conceptually subdivided into three categories or elements.38 The first, policy goals, in more abstract terms asks what general types of ideas govern policy development, such as e.g environmental protection or economic development.
The second, policy objectives, in less abstract terms asks what the policy formally aims to address, such as saving wilderness or species habitat. The third, policy settings, asks what the specific on-the-ground requirements of the policy are.
According to this model, policy means or tools can be subdivided into three elements as well.39 Policy instrument logic concerns what general norms guide the selection of
implementation tools, e.g preferences for the use of coercive instruments. Policy Mechanisms asks what specific instruments are utilized, such as tax systems. And finally, policy
calibrations ask what the specific ways in which the instrument is being used are, for example the use of mandatory vs voluntary regulatory guidelines or standards. An overview of these elements can be observed in table 1 below.
37 Moore & Jordan, 2020. s.295
38 Cashore & Howlett 2009 s.39; Henstra, Daniel 2011. The Dynamics of Policy Change: A Longitudinal Analysis of Emergency Management in Ontario. Journal of policy history 23(3): 399-428. s.406
39 Cashore & Howlett 2009 s.39
Table 1. Elements of policy object to change
High Level Abstraction Programme Level Operationalization (Less Abstraction)
Specific On-the- Ground measures
Policy Ends or Aims GOALS
What General Types of Ideas Govern Policy Development?
(e.g environmental protection, economic development)
OBJECTIVES What Does Policy Formally Aim to Address?
(e.g saving wilderness or species habitat, increasing harvesting levels to create processing jobs)
SETTINGS
What are the Specific On-the-ground Requirements of Policy?
(e.g considerations about the optimal size of designed stream-bed riparian zones, or sustainable levels of harvesting)
Policy Means or Tools INSTRUMENT LOGIC
What General Norms Guide
Implementation Preferences?
(e.g preferences for the use of coercive instruments or moral suasion)
MECHANISMS What Specific Types of Instruments are Utilized?
(e.g the use of
different tools such as tax incentives, or public enterprises)
CALIBRATIONS What are the Specific Ways in Which the Instrument is used?
(e-g designations of higher levels of subsidies, the use of mandatory vs
regulatory guidelines or standards)
Source : Adapted from Howlett, Michael & Cashore, Benjamin, 2009. The Dependent Variable Problem in the study of Policy Change: Understanding Policy Change as a Methodological Problem. Journal of Comparative Policy Analysis 11(1):
33-46
However, it is not enough to study if different elements of policy have changed over time. We would also like to be able to answer both the mode (or degree) of the change as well as the direction of it in order to properly answer in which ways the policy has changed over time.
Has the policy change been paradigmatic or incremental? Has the policy undergone a cumulative change away from the status quo, or do they represent a fluctuation consistent
with an existing policy equilibrium?40 In order to do this, our analysis must include a
framework that can separate between different types of incremental and paradigmatic change in our observations; as well as if the direction of change (as just recently mentioned).
Regretfully, there exists no tool for systematically d rawing the line when a change is incremental and when it is paradigmatic, which we will return to discuss. However, there does exist a taxonomy created by Durant and Diehl, and suggested by Cashore & Howlett in conjunction with the model above, that discerns the directionality of change as well as identifying four typical modes of change that can occur(see Table 2).41 By using this framework, we come a long way in answering in which ways the analyzed policy has undergone change as well as giving a more nuanced classification of the different modes of change (in contrast to the binary incremental/paradigmatic fault line typical in the litterature).
This framework will now be explained in more detail, before returning to the issue of determining the mode of change in practice, as teased above.
Table 2. Direction and typical modes of Policy Change
Tempo of Change
Directionality of Change Fast Slow
Cumulative “Classic” Paradigmatic Progressive Incremental In Equilibrium “Faux” Paradigmatic “Classic” Incremental (cells contain typical “modes” of change)
Source : Adapted from Howlett, Michael & Cashore, Benjamin, 2007. Punctuating Which Equilibrium? Understanding Thermostatic Policy Dynamics in Pacific Northwest Forestry. American Journal of Political Science 51(3): 532-551
40 Cashore & Howlett, 2007. s.537
41 Ibid, s.538
This above mentioned framework reconceptualizes policy change as an interplay of tempo and directionality, with tempo in this context being the speed of change of a policy and the directionality of change, meaning if the changes are (cumulatively) leading away from an existing policy equilibrium towards another or whether they represent a fluctuation consistent with an existing one.42 In doing so, this taxonomy identifies two commonly ignored or inaccurately juxtaposed modes of change that exist in addition to the “classic” or ordinary paradigmatic and incremental types traditionally found in the discourse.43 The first of these,
“faux” paradigmatic changes, are rapid and noncumulative; they are often misinterpreted as ordinary paradigmatic changes because they can deviate significantly away from the status quo, but then change back rather quickly into their original equilibrium. In other words,
“faux” paradigmatic change is in reality a form of incremental change. This type of change can also possibly be malicious to the analysis, as the time frame may very well begin or end in the middle of them and thereby leading in an over-or underestimation of the change in the result. We will return issues like these in the section of research design. Progressive
incremental changes are slow but cumulative changes that, in contrast to the name, result in major policy change over time. This type is often mistakenly categorized as classic
incremental change. The implications of these two additional types of change is that policy change ideally must be studied over longer periods of time in order to be able to adequately classify the different types of policy change, as they otherwise risk being incorrectly viewed as “classic” paradigmatic or incremental changes.
Now we return to the issues of determining mode of change in practice. An important observation the author has done is that there does not seem to exist a clear-cut definition between how much of a change is required for it to go from being incremental and instead becoming paradigmatic, and vice versa. This issue is emphasized by Cashore & Howlett, who notes that “... as have been pointed out for some time, neither a clear definition nor an
exhaustive taxonomy of change types currently exists, resulting in both incremental and paradigmatic change (and the distinction between them) remaining under-specified entities.44
42 Ibid, s.537
43 Ibid, s.537f
44 Cashore & Howlett 2009 s.40
The task of doing this seems to be up to the individual researcher, to some degree by using their intuition, knowledge, and common sense when analysing variations in the elements of policy change. Some guidance can be derived from the Hall’s logic, where changes in the high levels of abstraction (“goals” and “instrument logic”), often called third order changes, seem to be the hardest to achieve and thus imply a paradigmatic policy change (and therefore also to be the result of an exogenous event).45 First and second order changes are associated with the specific on-the-ground measures (“objectives” and “mechanisms”) and programme level operationalization (“settings” or “calibrations”) respectively, and changes in these are generally thought to be incremental and the result of endogenous activities within the policy subsystem.46 However, a slight change in the “goals” or “instrument logic” (or both) might not warrant a categorization as a paradigmatic change, and a complete shift in”objectives” or
“goals” might not be appropriately described as incremental. The consequence of leaving these considerations up to the judgement of the individual researcher will affect the reliability of this analysis in a negative way.
45 Ibid, s.37
46 Ibid, s.37
Methodology
3.1 Case selection
This study will cover the policy dynamics of environmental policy over time in the Swedish context. More specifically, it will concentrate its focus on Swedish clean air policy. Choosing a case within the environmental policy area as grounds for this study is inspired by
Baumgartners conclusions that it is a fertile ground for studying policy change, particularly in relation to the PET; this because it has shown consistent patterns of stability and change driven by the dynamics of the policy process, much like other areas of public policy.47 We can thus expect the results of an analysis of policy change on environmental policy such as clean air, atleast to a certain extent, to be representative of several other public policy areas - and therefore making some generalizations of the results increasingly plausible.
Because previous studies on environmental policy according to Baumgartner have shown consistent patterns of stability, coupled with the fact that this study aims to study policy change over several years, one can be fairly confident (or at least hopefull) in predicting that the result will produce some variation in the examined variables. By producing a descriptive result with at least some variation in the examined variables, the result not only becomes less drudging to read, it also contributes to set up future research explicitly focusing on the causal mechanisms that generate this variation - thus making a more significant contribution.48 There are multiple reasons. Firstly, clean air policy is interesting and fitting to study because of its intuitivity as a typical case of environmental policy. Moreover, gaining insights in this particular policy area can to some degree be seen as valuable in both other scientific
disciplines and real life actors, as this information could prove useful elsewhere for e.g policymakers or scientists of other disciplines concerned with both local and global challenges arising from air pollution.
But what is the reason for specifically choosing clean air policy in the Swedish context? The Swedish context is chosen mainly because it is a very uncommon national setting for analysis in the literature of policy dynamics in general and in PET-studies in particular, where there exists a heavy U.S bias.49 As mentioned in the introduction, its rather unique setting in both
47 Baumgartner, Frank M. 2006. s.24, 43f
48 Esaiasson, P., Gilljam, M., Oscarsson, H., Towns, A., och Wängnerud, L., 2017. Metodpraktikan: Konsten att studera samhälle, individ och marknad. Stockholm: Wolters Kluwer s.152
49 van der Heijden & Kuhlmann, 2018. s.335f
“the national” and “international” also plays a part. Braving new waters when analyzing policy change could potentially give new valuable insights, as well as setting up future research for doing so as well. However, using this logic, any country (or organization) underrepresented in the litterature could be motivated as an object of study. The final blow for specifically choosing Sweden therefore lies in its linguistic opportunity and practicality, with the author of this study being of Swedish nationality.
3.2 Research design & limitations
With reference to the chapter of previous research on the area, observations have been made that extensive analysis of policy change to a certain extent have been overrepresented. Along with this, Baumgartner, one of the architects of the PET, emphasizes that extensive analysis must be complemented by intensive in order to produce a fuller understanding of the process of policy change. As said, the purposes of this study will be to cater to this slight qualitative deficit in the PET-litterature, as well as aiming to give a more valid and nuanced picture of policy change than most previous literature. By carrying out a longitudinal case analysis of how policy changes over time with an improved framework, and relating our results to the punctuated equilibrium theory, we gain the prerequisites to answer our research questions and achieve the purposes of this study. As mentioned above, this will more concretely be done by analyzing the six elements of policy proposed by Cashore & Howlett over a selected time span.
Because we will be studying a specific policy by applying a systematic framework to an empirical material such as governmental reports and bills over different points in time, one could methodologically talk about this study as conducting a comparative within-case text- analysis. The text-analysis part can be emphasized in the pursuit of defining the policy contents empirically at specific points in time, while the comparative part can be understood in light of comparing the results of the analyzed material from the different points in time, and using it as grounds for speaking of policy change.50 Using text-analysis as a method also comes naturally from the descriptive nature of the study, particularly the first research
question, since it by definition is a form of descriptive and qualitative analytical method for systematically organising empirical content in order for it to tell us how things are or have
50 Esaiasson, Gilljam, Oscarsson, Towns and Wängnerud, 2017. s.98f
been (and in this case also the relative differences between them - how things have changed).51
This brings us to one of the most important and challenging considerations of this study - deciding upon the time frame. Ideally, in order to gather as convincing empirical evidence of policy change as possible, the number of observations should be many, as well as cover as long of a time period as conceivable.52 Also, because our theoretical expectations also make clear that paradigmatic changes linked to positive feedback processes are the exception (rarely occurring), and incremental change linked to negative feedback is the norm, any attempt to capture both by their nature should mean analysis of long periods of time to be the most desirable. However, due to the very nature of this study being a bachelor’s thesis, with limited time and resources, it is not possible to achieve many observations over a large span of time - at least not in a qualitative fashion. The time frame must therefore somehow be decided according to some other considerations, and thoughtfully. The main problem arising from an ill-considered time frame is the risk of censoring - i.e the incomplete recording of information at the beginning or the end of the event that is of interest.53 A hypothetical example of this can be recalled from the operationalization section concerning “faux''
paradigmatic change. To reiterate - faux paradigmatic change meant a fast but uncumulative, where the policy contents diverges dramatically away from equilibrium only to rather quickly move back, therefore being a form of incremental change. If the time frame of the analysis were to begin after the initial change, we would incorrectly categorize this change as paradigmatic (fast and cumulative) even though it in reality should be seen as incremental.
Note that this is just one example of censoring, and the “faux'' paradigmatic case is likely more malicious than other forms of censoring because we lack theoretical expectations of why they take place.
While we can never eliminate the risk of censoring complemetly. By choosing a time frame according to theoretical considerations, we can reduce it.54 For example, since we according to the PET can expect paradigmatic policy changes to be most likely to occur because of
51 Ibid, s.98f; Svensson, Torsten & Teorell, Jan, 2007. Att fråga och att svara: Samhällsvetenskaplig metod.
Stockholm: Liber. s.213
52 Baumgartner & Jones, 2002. s.258
53 Ibid, s.258
54 Ibid, s.258
exogenous events, we can attempt to position our time frame so that it doesn’t start or end right in the middle of one. We can even turn this insight to an advantage (to attain variation), by choosing a time frame that also includes some exogenous event we suspect could have affected the policy. In this case, the exogenous event we will position in accordance with our time frame is signing and ratification of the Paris Agreement, that formally entered into force november of 2016.
All things considered, this study will attempt to choose on a time frame that covers the period of this exogenous event (the Paris agreement) that is thought to likely have affected the contents of Swedish clean air policy in a major way (thereby in line with theoretical
considerations), while also covering a certain period before and after the change (in order to reduce risk of censoring). With this “certain period before and after” has been deemed appropriate to cover the time span between roughly 2011-2019.
However, if we are to hinge our time frame to the Paris Agreement, we must motivate in which way might be a suitable candidate as an exogenous event. But first, what is the Paris Agreement? According to the UNFCC website (United Nations Framework Convention on Climate Change), the Paris Agreement is a legally binding international treaty on climate change agreed upon in December of 2015 that aims at limiting the increasing global temperature below 2°C (but preferably 1,5℃) - mainly by minimizing the release of
greenhouse gases in the atmosphere but it also aims at supporting those who are affected by the effects of climate change.55 It is a deal that has been adopted by all countries of the world (with an exception of the U.S who formally withdrew from the treaty 20 November 2020 but with plans on rejoining) and becomes formally binding in November 2016.56
Most PET-scholars have relied on Kingdon’s idea that exogenous shocks galvanize mass political attention on an issue and highlight vital problems that require government attention - a first step in initiating a positive feedback mechanism (as covered in the theory section).57 In other words, the exogenous event should have the characteristics of being able to mount
55 UNFCCC. The Paris Agreement (Retrieved 22/12-20) Link: https://unfccc.int/process-and-meetings/the-paris- agreement/the-paris-agreement
56 Naturvårdsverket. Vad är Parisavtalet? (Retrieved 22/12-20), Link:
https://www.naturvardsverket.se/Miljoarbete-i-samhallet/EU-och-internationellt/Internationellt- miljoarbete/miljokonventioner/Klimatkonventionen/Parisavtalet/Vad-ar-Parisavtalet/
57 Beyer, Breunig, Radojevic, 2017. Punctuated Equilibrium Theory, in Zahariadis & Buonanno (eds) The Routledge Handbook of European Public Policy
heavy pressure on policy makers to do something. Most often in the litterature, these exogenous events are often described as sudden catastrophic events such as nuclear meltdowns or financial crises.
In this regard, the Paris Agreement is different because it is not a treaty that has suddenly sprung out of nowhere, but rather has been a rather predictable event at least some time before its agreement. However, because of its international consensus, coupled with the fact that the goals are relatively ambitious, it is thought that the treaty should do a good job at galvanizing mass political attention and mount pressure on national policymakers. With this in mind, despite its lack of “suddenness”, the Paris Agreement is thought to do well in increasing the likelihood of a positive feedback process to be initiated, and should therefore serve as a suitable exogenous event.
Lastly, a notable demarcation of this study is to only cover clean air policy related to pollution caused by traffic. The reason for this is that clean air-policy in general involves a too large and complex set of means and ends to be able to capture in a single study. Air pollution is not only thought to be caused by traffic, but by the energy sector, the industrial sector, agriculture and food production, construction, and so on. An exhaustive account of the Swedish clean air policy, as well as the changes in it, would necessarily have to include all of these, as well as the instruments and specific ways in which they are used to combat air pollution. This study therefore rests on the assumption that the changes in clean air policy related to pollution caused by traffic is representative of the changes in the overall clean air policy, as the measures against traffic-related air pollution is thought of as a “typical case” of clean air policy. It is a logical assumption to make, as air pollution related to traffic is one of the fundamental causes of polluted air.58 However, like all assumptions this one can of course in one way or another be met with some scepticism and criticism, and this is also welcomed.
However, in order to make this study feasible, it is an assumption that has been done.
58 Steg på vägen: fördjupad utvärdering av miljömålen 2012. s.191
Results and Analysis
Starting in the 2012 report for in depth assessment of the environmental quality objectives (miljömålen) by the Swedish Environmental Protection Agency (Naturvårdsverket) that covers 2011, we can begin to discern the different policy elements that make up the Swedish clean air policy related to traffic at this time with help of our framework. The means and ends of the policy will then be traced forward through time, noting any additions, removals, or changes in the elements. The first part of this section will be to map out the composition Swedish clean air policy of 2011 (note the above-mentioned report has its evaluation
anchored in 2011). The second part of this results section will then attempt to map out if and how the different components of the clean air policy have changed over time up to 2019. The third part will briefly summarize what has changed and what hasn’t, then turning to analyse the change. Lastly, a discussion regarding the results and the analysis of it will be conducted.
A disclaimer that may be wise to signal here is that all the citations from Swedish sources are directly translated by the author of this study.
4.1 Policy composition up until 2011 High level abstraction
Regarding the ideas that govern clean air policy development, four major ideas of problems governing air pollution can be identified - damage to human health, environmental health, cultural heritage, and economic development. Naturally, these problems are the same in the specific context of air pollution related to traffic. The report mentioned in the introductory part of this section notes that “Increased levels of air pollution cause significant damage to human health, to vegetation, and cultural objects”, as well as “decreased health and
environmental impacts costs society considerable amounts of money - with damage caused by high levels of polluting particles is estimated to cost up to 26 billion sek/year.59
Norms that guide implementation preferences are varied, and as noted in the above-
mentioned report “in order to reach the environmental quality objectives we must work on a broad front and use (all) the different instruments at our disposal”.60
59 Ibid. s.182f
60 Ibid. s.82
This holds true for traffic in particular, where coercive instruments such as different forms of taxation such as fuel and congestion tax, and regulations such as minimum requirements of nitrogen oxide emissions from new cars, and extended restrictions on studded tires, have been used.61 Non-coercive instruments used include e.g tax-exemption of biofuel, a green car premium, reduced vehicle tax cars classified as green, as well as investments in research of new and greener tech in the transport sector.62 What seems to be the most important norm that guides implementation of instruments seems to be efficiency, as described in the 2012 report in the context of environmental goals for a toxic-free environment (which is related to clean air policy) “In order to achieve a total (pollutant) exposure not harmful to humans or the ecosystem, high demands are placed on efficient instruments”, where “both coercive and non-coercive actions are necessary”.63
Programme level operationalization
What does the policy formally aim to address? In simple terms, it aims at reducing the levels of pollutants in urban areas in particular and in general the atmosphere.64 More specifically, as noted, it aims at reducing it to the point in which human health, animal and plant life, and cultural heritage is not harmed, and aims at doing this “within a generation”.65 An additional aim that merges with the previous is that “the levels of greenhouse gases in the atmosphere…
must stabilize to a level so that the human impact on the climate system does not become dangerous”, and that “the goal must be reached in such a way and haste that the biodiversity is preserved, the production of foodstuffs (livsmedelsproduktion) is secured, and that other goals for sustainable development is not jeopardized.66
61 Ibid. s.34, 95, 199
62 Ibid. s.98
63 Ibid. s.243
64 Ibid. s.180
65 Ibid. s.180
66 Ibid. s.162
What specific types of instruments are utilized? By use of different forms of regulatory and legislative instruments within the frame of e.g the legislation regarding environmental quality norms (lagstiftningen om miljökvalitetsnormer) and the Swedish Environmental Code
(Miljöbalken).67 Along with regulatory and legislative instruments, both economic incentives and deterrents such as different forms of taxation and subsidization are deployed in an
attempt to achieve the objectives of a cleaner air policy, which will be specified more in depth in the next section on “Specific on-the-ground measures”. Lastly, informational instruments (which can be viewed as the least coercive), such as informational campaigns directed towards households on e.g how they can change their traffic-habits are used.68 Specific on-the-ground measures
What can be interpreted as the specific on-the-ground requirements of the clean air policy? It is recognized in the 2012 report that while it is tough to determine what is “dangerous human impact on the climate system”, three specific requirements are important to attain the above mentioned aims.69 Firstly, there is a temperature goal to limit the global mean temperature increase to a maximum of two degrees celsius in the long run, in relation to pre-industrial level. The second goal, a concentration goal, where the concentration of greenhouse gases in the atmosphere must stabilize at the maximum of 400ppm carbon dioxide-equivalents. Lastly, there is an “intermediary goal” (etappmål) assigned until 2020, which states that greenhouse gas emissions must decrease by 40 percent compared to 1990, which translates to about 20 million tonnes of carbon dioxide-equivalents lower.70
The specific ways in which the instruments in which the instruments have been used will be accounted for in the following. Firstly, the central economic instruments will be explained, followed by the central legislative and regulatory instruments, and lastly the instrumental ones.
67 Ibid. s.88, 185
68 Ibid. s.100f; Gipperth, Lena & Pleijel, Håkan, 2008. Har Miljökvalitetsnormerna förbättrat utomhusluften?
s.45
69 Steg på vägen 2012 s.163
70 Ibid. s.173
From the first of August 2007, a congestion tax proposed by the Swedish government was accepted by the Swedish riksdag to be implemented in Stockholm, with the purpose of increasing mobility and the environmental conditions.71 The congestion tax is a fee or toll that “applies to certain vehicles passing in and out of the inner city between 06:30-18:30 on weekdays” (though some green cars are exempt from this), and it has been evaluated to both directly and indirectly reduce air pollution in the inner city significantly.72 Another specific form of a traffic-related economic instrument being deployed is the fossil fuel tax, which is a specific form of an energy and carbon dioxide tax. While the design of the tax can be quite complicated to describe in detail, the main points can be generally summarized as “a tax that generally is extracted by volume”, and is extracted in proportion to the amount of carbon dioxide emitted during combustion.73 Regarding subsidies that aim at reducing air pollution, it can be noted that “since 2004 biofuels used for motor related engine-operation are exempt from the fuel tax”.74Another central form of subsidy is a tax exemption for green cars.
Onwards from 1st of January 2011, owners of passenger cars that are either categorized as emission-class “Euro 5 or Euro 6” or are electric/hybrid” are entitled to a five year vehicle- tax exemption starting from the day the vehicle was first put into service.75
There are several distinct regulatory instruments in use on the area. To begin with, there are regulations for both maximum allowed emission levels for new vehicles that are sold, as well as regulations for maximum allowed emission levels for vehicles in general before they must be put out of service. According to the Swedish Transport Agency (Transportverket), these maximum allowed emission levels for new light passenger and commercial vehicles are specified in regulation 715/2007/EC of the European Parliament and Council (with date of effect from 2nd July 2007).76
71 Skatteutskottets betänkande 2006/07:SkU19
72 Steg på vägen 2012. s.95
73 Naturvårdsverket & Energimyndigheten, 2006. Ekonomiska styrmedel i miljöpolitiken - Rapport från Naturvårdsverket och Energimyndigheten (ISBN: 91-620-5616-6). s.113
74 Ibid. s.113
75 Transportstyrelsen. Femårig skattebefrielse. (Retrieved 18/12-20) Link:
https://www.transportstyrelsen.se/sv/vagtrafik/Fordon/Fordonsskatt/Femarig-skattebefrielse/
76 The Official Journal of the European Union, 2007. - Regulation 715/2007/EC (Retrieved 20/12-20) Link:
https://eur-lex.europa.eu/legal-content/EN/ALL/?uri=CELEX%3A32007R0715
In this regulation, it is stated that “manufacturers shall demonstrate that all new vehicles sold, registered or put into service in the Community are type approved in accordance with this Regulation”, where the regulation specifies the Euro 5 or Euro 6 emission-classes and their limits as the cut-off point.77 Another regulatory instrument on-the-ground is the possibility for municipalities to ban the use of studded tires on certain streets, where “decisions like these have already been made in Stockholm, Gothenburg and Uppsala”.78 The government has even granted the municipalities the power to implement these extended studded tire-bans
“to cover entire zones in urban areas”, although “a decision of an extended studded tire-ban is yet to have been made in any municipality”.79
Informational instruments deployed involves informational campaigns such as
“Klimatkampanjen”, that aimed at spreading information and increasing the knowledge of the causes and consequences of climate change (in which traffic is an important contributor); as well as more general forms of “communications” that supplement other forms of instruments (such as regulatory ones) by “spreading knowledge” and “creating acceptance” among the citizens.80
The main points of the traffic-related clean air policy up to 2011 is summarized in Table 3.a below, fitted in accordance with our analytical framework in order to give a clear view of its composition.
77 Ibid, Chapter 2 Article 4.1
78 Steg på vägen 2012 s.184
79 Ibid. s.184
80 Ibid. s.85, 101
Table 3a. Elements of Clean air policy related to pollution caused by traffic, 2011.
High Level Abstraction Programme Level Operationalization (Less Abstraction)
Specific On-the-Ground measures (Least
Abstraction)
Policy Ends or Aims
GOALS - Human health - Environmental health - Cultural heritage - Economic development
OBJECTIVES Reducing air pollution to the point that:
-Human health -Biodiversity -Cultural heritage -Production of foodstuffs
is not harmed, and generally that the human impact on the climate does not become dangerous
SETTINGS - Temperature goal:
temperature increase of maximum 2°c in the long run
- Concentration goal:
greenhouse gas
concentration to stabilize at 400ppm maximum in the long run
- Intermediate goal:
greenhouse gas emissions to be 40% lower by year 2020 in relation to 1990
Policy Means or Tools
INSTRUMENT LOGIC - Preference for efficient instruments
- Preferences for a mix coercive and non-coercive instruments
MECHANISMS - Economic instruments - Legislative and regulatory instruments - Informational instruments
CALIBRATIONS - Congestion tax - Fossil fuel tax
- Tax exemption on biofuel - Vehicle tax exemption of
“green cars” and electric/hybrid vehicles (miljöbilar)
- Regulations of minimum requirements of emissions from newly sold vehicles.
-Regulations on use of studded tires in certain streets or zones in urban areas
-Informational campaigns
& Communications
