Contents lists available at ScienceDirect
Energy Research & Social Science
journal homepage: www.elsevier.com/locate/erss
Original research article
A contested transition toward a coal-free future: Advocacy coalitions and
coal policy in the Czech Republic
Petr Ocelíka,b,⁎
, Kamila Svobodovác
, Markéta Hendrychovád
, Lukáš Lehotskýa
,
Jo-Anne Everinghamc
, Saleem Alic,e
, Jaroslaw Baderaf
, Alex Lechnerg
a
Department of International Relations and European Studies, Faculty Social Studies, Masaryk University, Joštova 218/10, 602 00 Brno, Czech Republic
b
International Institute of Political Science, Faculty Social Studies, Masaryk University, Jostova 218/10, 602 00 Brno, Czech Republic
c
Centre for Social Responsibility in Mining, Sustainable Minerals Institute, The University of Queensland, St Lucia, Queensland 4072, Australia
d
Czech University of Life Sciences, Faculty of Environmental Sciences, Department of Land Use and Improvement, Kamycka 1176, Prague 16500, Czech Republic
e
College of Earth, Ocean and Environment, University of Delaware, 272 The Green APT 111, Newark, Delaware, 19716, USA
f
Faculty of Earth Sciences, University of Silesia, 41 - 200 Sosnowiec, Poland
g
School of Environmental and Geographical Sciences, University of Nottingham Malaysia, Semenyih, 43500, Malaysia
A R T I C L E I N F O
Keywords:
Energy policy
Energy transition
Coal phase-out
Policy networks
A B S T R A C T
Coal phase-out is an integral part of the ongoing energy transition to a decarbonized economy. Any such process
involves diverse actors that compete over the nature and pace of such transition. This research uses the Advocacy
Coalition Framework to analyze the conditions of policy change within an adversarial subsystem. It focuses on
the coal subsystem in the Czech Republic, a post-communist coal-dependent country with comparatively large
economically recoverable reserves. Using data from an organizational survey, exploratory social network analysis
techniques are applied to identify advocacy coalitions and deductive block-modeling is used to test hypotheses
on the subsystem’s functioning. The focus is on: (1) fragmentation of decision-makers, (2) targeting of
decision-makers, and (3) use of expert information. Two competing and ideologically distant coalitions were
identified: the Industry Coalition and Environmental Coalition. The results further show high fragmentation
among decision-makers, as indicated by their cross-coalition membership and the heterogeneity of their beliefs.
The targeting of decision-makers is practiced by principal members of both coalitions, i.e. environmental nongovernmental
organizations and industry, but also by research organizations. Lastly, expert information exchange
strongly overlaps with the identified coalitions and thus increases their cohesiveness. It is argued that
such subsystem configuration limits the potential for policy change through negotiated agreement or policy
learning. Policy brokers and policy venues are suggested as remedies to moderate the adversarial nature of the
subsystem.
1. Introduction
Energy transition to a low-carbon economy is one of the most
pressing and complex challenges that modern societies face globally. In
this context, coal has been recognized as a reliable and cheap source of
energy from the very beginnings of the industrial revolution and a
backbone of traditional energy industries [1]. Nevertheless, it is now
considered to be one of the major anthropogenic drivers of greenhouse
gas emissions [2]. Thus, the European Union's policies aim to phase out
coal production in the next 25 to 50 years and require member state
governments as well as regional authorities to engage in the energy
transition [3]. Besides regulation, the decline in coal demand is further
strengthened by increased competition from renewables and unconventional
natural gas resources, which are pushing coal out of the
North American market [4]. These trends generate landscape pressures
impacting member states where involved policy actors and their coalitions
compete in uncertain environments over the formulation of
specific transition pathways [5,6]. In this respect, coal dependent
countries that have not yet adopted phase-out strategies stand at a
decisive point when key decisions on how quickly and by what means
to transform their carbon intensive energy industries need to be made.
As Markard et al. [7] reported, research on the political dimension of
https://doi.org/10.1016/j.erss.2019.101283
Received 23 December 2018; Received in revised form 25 August 2019; Accepted 1 September 2019
⁎
Corresponding author at: Masaryk University, Jostova 218/10, 602 00 Brno, Czech Republic.
E-mail addresses: petr.ocelik@gmail.com (P. Ocelík), k.svobodova@uq.edu.au (K. Svobodová), hendrychovam@fzp.czu.cz (M. Hendrychová),
llehotsky@mail.muni.cz (L. Lehotský), j.everingham1@uq.edu.au (J.-A. Everingham), saleem@udel.edu (S. Ali), jaroslaw.badera@us.edu.pl (J. Badera),
Alex.Lechner@nottingham.edu.my (A. Lechner).
Energy Research & Social Science 58 (2019) 101283
Available online 03 October 2019
2214-6296/ © 2019 Elsevier Ltd. All rights reserved.
T
such changes has emerged only recently (see e.g. [8–11]). We aim to
contribute to this research using the Advocacy Coalition Framework
[12,13] to explore interactions among policy actors within the context
of a specific policy subsystem—an issue-defined network of actors within
juridical and geographic boundaries [14].
Pursuant to the Advocacy Coalition Framework [15], landscape
pressures such as macroeconomic trends or supranational regulation
are understood as external perturbations that impact policy subsystems
in terms of resource redistribution, available opportunities for minority
coalitions, or belief changes in the dominant coalition [14,16,17]. Of
course, not every external perturbation results in a major policy change
(cf. [14]). It is regarded rather as a necessary but not sufficient condition
of such change, and its effects are not independent of the subsystem's
internal functioning [19,20]. Thus, investigating the particular
mechanisms that mediate the effects of external perturbations is essential
to better understand of policy change [16,17,21]. More specifically,
we focus on three such mechanisms: fragmentation of political
authority [22], access to political authority [23,24], and use of expert
information [25]. First, a policy change needs to be accepted by policy
actors with formal decision-making authority. The degree of fragmentation
in terms of the decision-makers’ beliefs and their coalition affiliations
thus affects opportunities for such change [21,22,26]. Second,
other involved policy actors compete to access the authority in order to
influence policy decisions [23,27]. Third, use of expert information is an
integral part of the policy process with important impacts on potential
policy learning—a key pathway to policy change [25,28]. It is also
assumed that outcomes of the mechanisms depend on the type of policy
subsystem. In this research, the context is that of an adversarial policy
subsystem constituted by competing advocacy coalitions with incompatible
beliefs and prevalently within-coalition coordination patterns
[22]. As a specific case, we focus on the Czech Republic, where
the future of brown coal (henceforth “coal”) has been intensively contested
since the 1990s [29] (see Section 4). We assume that the Czech
Republic represents a typical case of coal phase-out in an adversarial
policy subsystem.
The overarching research question is the following:
RQ: How do policy actors and their coalitions interact to influence coal
policy in an adversarial subsystem?
To that end, we apply social network analysis to survey data collected
on the coal policy subsystem. Our contribution is twofold. First,
we connect Advocacy Coalition Framework with concepts from the
socio-technical transition literature thereby enabling us to theorize on
potential impacts of policy change, or its absence, within a broader
context of the Czech energy industry. Second, we address limitations of
the advocacy coalition detection techniques and offer an approach that
considers belief and relational elements of coalitions equally.
The article starts by introducing the Advocacy Coalition Framework
and then elaborates on theoretical arguments and derives hypotheses
focused on the mechanisms described above. Next, we describe the data
and methods, which rely mostly on descriptive and exploratory social
network analysis techniques [30]. The next section provides background
information on the Czech case. The results indicate the presence
of two antagonistic coalitions that compete for access to a fragmented
political authority and exchange expert information primarily among
their members. In conclusions, it is expected that such conditions will
facilitate the overlay of newly adopted policies on the existing core of
the regime, thus hindering any major policy change towards a rapid
phase-out.
2. Theory
2.1. Advocacy Coalition Framework
The Advocacy Coalition Framework is based on the assumption that
policy problems in modern democratic societies are contested by diverse
actors and coalitions of actors which advocate specific solutions.
Analysis of the advocacy coalitions is thus crucial to understanding the
policy process [14,28]. Coalition politics occur at the level of a policy
subsystem which is a subset of a political system defined by an issue area
[31,32]. An advocacy coalition (henceforth “coalition”) is thus defined
as a group of actors that (1) share policy core beliefs and (2) engage in a
coordinated activity [19,33]. The degree (dis)similarity among policy
core beliefs and the prevalent coordination patterns define different
subsystem types, specifically: unitary, collaborative, and adversarial
(see [25]). In this research, we focus on an adversarial subsystem defined
by competing coalitions with low-compatible policy core beliefs and
dominant within-coalition coordination patterns [25]. The first element,
policy core beliefs (PCBs), constitutes a set of normative assumptions
about how the subsystem ought to be organized. They are highly
salient and produce cleavages within a subsystem for some time [14].
The second element captures interactions, often informal, among actors.
The interactions are driven by, among other things, an exchange of information
that relates to substantive policy issues and political efficacy
[34], which is especially important in the case of complex problems as
well as under conditions of risk and scientific uncertainty [35,36].
Other drivers include alliance formation through which actors seek to
exchange various resources and access to political authority through
which actors influence other actors with decision-making competencies
[24,37]. Thus, the relational structures fundamentally condition the
characteristics and capabilities of the coalitions [31,38].
In this research, we focus on two types of interactions: political cooperation
and expert information exchange. The former rests on the assumption
that policy actors use different ways to translate elements of
their belief systems in order to influence the policy process before their
opponents can do the same. For this purpose, they seek allies and share
resources [14]. We therefore define political cooperation as an activity
that does not necessarily require a common objective or joint planning
and that might range from “support on a policy issue, support of other
organizations in/through international organizations or professional
associations, working together to find a solution of a policy problem …
[to] joint official statements, joint lobbying, co-organizing campaigns
and protests” (Appendix A). The Advocacy Coalition Framework further
assumes that the policy actors will seek allies among actors who hold
similar PCBs. The general expectation is that cooperation patterns
overlap with the clusters based on those beliefs [14,31]. Thus, relations
based on political cooperation are used to identify coalitions [39,40].
Expert information is defined as information of a technical, scientific, or
process character related to the coal subsystem. The distinction between
expert and political information is, of course, not clear-cut.
Especially in the case of evidence-based policy making, the production
and exchange of expert information is an integral part of the political
process that involves making claims about what the policy makers
should and should not recognize as evidence [41].
Coalitions are also defined by other attributes (such as coalition
membership and size, resources, or within- and cross-coalition activity)
which further influence their ability to achieve their objectives [32,42].
The distribution of reputational power within the subsystem is captured
as a reported “ability of the organization formally … and/or informally
… to change the results of the processes of creation, implementation,
and evaluation of the policies” (Appendix A; cf. [23,42,43]).
Applications of the Advocacy Coalition Framework to energy policies
range from identification of coalitions and their properties
[44–49], evaluations of external shocks and crisis policymaking
[16,17,50], energy transition [7], drivers of policy agreement [51],
policy change [42,52], and impacts of the institutional context on
policy outcomes [53] to related research on discourse coalitions
[18,54–57]. Ingold et al. [51] argue that issues of energy policy can be
characterized as a conflict between mostly right-wing actors with economically
liberal core beliefs against left-wing and environmental actors
(cf. [58]). This cleavage can be further strengthened in mature
subsystems [14] such as coal mining, where actors hold clearly articulated
ideological positions and the power structure of the subsystem
P. Ocelík, et al. Energy Research & Social Science 58 (2019) 101283
2
is well-established. As a result, two coalitions, one favoring economic
efficiency generated by the free market and the other supporting state
interventions to protect the environment, are usually present [32,51].
This is consistent with the expected coalition structure in adversarial
subsystems as hypothesized by Weible et al. [25]. Considering the
above, we formulated the following hypothesis:
H1. Two coalitions with distant policy core beliefs are expected to be
present in the policy subsystem.
Drawing on Weible et al. [25,59] we also focus on three particular
mechanism outcomes hypothesized to be present in an adversarial
subsystem: (1) fragmented political authority; (2) competitive access to
political authority; and (3) prevalently within-coalition patterns of expert
information exchange.
2.2. Fragmentation of political authority: divided we govern
The first investigated mechanism is the degree of fragmentation
within the political authority. Weible et al. [25] define (political) authority
as legitimized power to create and enforce institutional rules. In
adversarial subsystems, political authority is fragmented between and
also within the subsystems thereby facilitating coalition competition
[25]. Moreover, mining policies in general are of a regulatory character
and thus directly impact well-organized interests in society. As a consequence,
they tend to foster conflict rather than cooperation [60,61].
In this context, we focus on actors that have decision-making competencies
and provide institutional support throughout the policy process.
These capacities are distributed among ruling political parties that
function primarily as “transmission belts” between organized interests
and policy outcomes as well as competent government offices and ministries
mainly responsible for specific policy designs and their execution.
We call this actor group “decision-makers”. Considering the above, we
formulated the following hypotheses:
H2. The decision-makers are expected to be fragmented across
coalitions in the policy subsystem.
H3. The decision-makers are expected to have higher variability of
policy core beliefs in comparison to the other actor groups in the policy
subsystem.
Thus, the degree of fragmentation is captured at the level of decision-makers’
memberships (H2), i.e. whether different decision-makers
belong to different coalitions (or potentially to a residual actor group),
and at the level of policy core beliefs variability (H3). Thus, we expect
cross-coalition memberships and large variability of PCBs. This would
consequently limit policy change through policy learning or negotiated
agreements since coalitions are able to block each other and likewise
any single actor or coalition is prevented from controlling the decisionmaking
process [62].
2.3. Access to political authority: targeting the decision-makers
The second investigated mechanism is access to political authority
through targeting of decision-making actors. In adversarial subsystems,
coalitions compete to access this authority and influence policy decisions
to make them consistent with their own PCBs [22]. The targeting
is done mainly by the coalitions’ principal members who are central to
the coalitions and coordinate the majority of their activities [22]. In
line with Ingold et al. [51], we expect that the principal members of the
competing coalitions are mostly environmental NGOs (ENGOs) and
industry actors. In relation to ENGOs, this assumption is further substantiated
by the concept of transactional activism [63]. It assumes that
organized non-state actors rely mostly on their capacity to develop
enduring as well as temporary relations among themselves and towards
power-holding institutions [63–67]. As Čada and Ptáčková [64] argue,
the legitimacy of transactional NGOs is not based on active membership
but on epistemic sources such as scientific knowledge and expertise.
These NGOs then pursue their objectives through collaborative planning
with public authorities [64] and issue advocacy [67], among other
things. Thus, we expect that ENGOs use targeting of decision-makers by
extending political cooperation ties as well as through the provision of
expert information (cf. [63–65,67]). As Geels [68] suggests, industry
actors likewise engage in instrumental interactions, such as lobbying and
consultancy, to support policies that preserve their vested interests.
Thus, we expect to find the same patterns of political cooperation and
expert information exchange for industry actors as well. Considering
the above, the following two sets of hypotheses were formulated:
H4a. Environmental non-governmental organizations are expected to
send more political cooperation ties to decision-makers as compared to
the overall average.
H4b. Industry actors are expected to send more political cooperation
ties to decision-makers compared to the overall average.
H5a. Environmental non-governmental organizations are expected to
send more expert information ties to decision-makers as compared to
the overall average.
H5b. Industry actors are expected to send more expert information ties
to decision-makers as compared to the overall average.
Targeting aims at gradual policy change (ENGOs) or its prevention
(industry actors) and also counteracts parallel activities of the opposing
coalition. This would lead to further fragmentation among decisionmakers
and consequently limit the potential for policy change through
policy learning or negotiated agreement.
2.4. Expert information: tell me I am right
The third investigated mechanism is the use of expert information.
From an instrumentalist perspective, expert information is considered
crucial to the improvement and quality of policy outputs at the expense
of ideology and derived values [69,70]. As a result, policymaking is
seen as an expert consideration of feasible and efficient solutions within
the given constraints of a system (cf. [71]). It is visible even more in the
case of energy policy, which has traditionally been framed as a technical
issue requiring expert, apolitical information [72]. In this view,
expert information exchange could be seen as an activity that bridges
the ideological and political differences between competing actors and
coalitions. Contrary to this interpretation, the Advocacy Coalition
Framework literature suggests that actors and coalitions often engage in
expert debates to support their policy positions [15,22,41,73]. Moreover,
the importance of expert information increases under conditions
of uncertainty, risk or controversy [41,74,75]. In adversarial subsystems,
experts serve as principal allies or opponents of coalitions and
expert information tends to follow within-coalition interactions [25].
Considering the above, we formulated the following final hypothesis:
H6. The density of expert information exchange ties in the policy
subsystem is expected to be higher within the coalitions than across
them.
We expect that expert information exchange tends to overlap with
political cooperation and occurs more within than across coalitions.
This would increase the cohesiveness of coalitions rather than facilitate
brokerage and consequently limit opportunities for cross-coalition
policy learning [76].
3. Data and methods
The research is a single-case study – of the Czech coal policy subsystem
– with within-case variation [77]. It was selected as a typical
instance of adversarial policy subsystem that is characterized by presence
of coalitions holding low-compatible policy core beliefs that
compete to influence the policy-making process [59] (for more details,
see Section 4).
P. Ocelík, et al. Energy Research & Social Science 58 (2019) 101283
3
3.1. Data
The data was collected from an online survey conducted in the
second half of 2017. The definition of the network boundaries used a
standard combination of positional, decisional, and reputational approaches
[38,78]. The preliminary list of policy actors was formulated
based on a literature review [29,79] and previous research of the authors
[80–82]. In this step, the positional and decisional approaches
were applied. The resulting list was used in an expert survey for further
evaluation based on a reputational approach. We asked experts from
different fields ranging from industry, state institutions and ENGOs to
academia to evaluate the political influence of the listed organizations
in the coal subsystem. In addition, we conducted five expert interviews
along with the expert survey. Based on this data, we finalized a list of
involved policy actors that consisted of 83 organizations (for the
complete list, see Appendix A). The pretesting of the survey included
respondent debriefing and a pilot study. The pilot study did not use
samples from the target population due to concerns about overburdening
the respondents. We received 24 responses, which were then
analyzed to refine the four introduced Likert scales.
The questionnaire consists of three sections measuring (1) belief
systems, (2) interactions of the policy actors and (3) contextual information
(for more detail, see Appendix A). It is designed to a large
extent based on the questionnaire outline of the Comparing Climate
Change Networks project [40]. The first section includes measures of
core beliefs and policy core beliefs. The core beliefs scale was adopted
from previous research on mining policies [51,53]. The PCBs scales
were initially constructed based on the literature review and then further
developed based on expert consultations and the pilot survey. We
calculated the simple sum of all items for each dimension and rescaled
the scores from 0 to 1 to maximize clarity of interpretation and readability
[42]. The ordinal alpha [83] scores for the scales range between
0.78 and 0.82 (for more details, see Appendix B).
Relational data was collected for political cooperation, expert information
exchange, and political influence (see Appendix A). The resulting
networks consist of directed binary ties. The political influence
network was used to construct a reputational power measure. It simply
expresses the relative frequency of political influence indications for
each policy actor by others. This corresponds to the in-degree centrality
of the policy actor in the political influence matrix. The measure ranges
between 0 and 1 after normalization [42].
The respondents were representatives of the listed organizations.
The response rate was 82% (for more details, see Table 1). Of the 15
non-responses, 2 organizations claimed that they were not active in the
coal subsystem. Due to the risk of a low-response rate indicated in the
preliminary expert interviews, we decided to code the actors in a way
that enables recognition of only the type of actor but not their identity.
The only exceptions were ministries and the Office of the Government's
Office (with their approval).
3.2. Methods
Social network analysis (SNA) was used as a methodological framework
[30]. We have applied standard descriptive measures to evaluate
political cooperation and expert information exchange networks at the
global, sub-group, as well as node levels (see Appendix B).
To test our expectations regarding coalition structure (H1), we used
a combination of two exploratory techniques: (1) cluster analysis applied
to PCBs and (2) faction analysis applied to the political cooperation
network. First, we applied cluster analysis based on K-means to (1)
identify clusters of actors that share PCBs. To validate the solutions, we
used silhouette width, which measures the distance between observations
within one cluster compared to another and also within cluster
distances [84]. Values can range between −1 and 1, where positive
values indicate that the observations lie well within their clusters and
negative values indicate the opposite, i.e. potential misclassification.
Since coalitions are also defined by higher density of within-group interactions
as compared to cross-group interactions, we used the faction
analysis procedure in UCINET [85] to (2) detect such relationally cohesive
subgroups [42]. The procedure requires to define the number of
detected factions a priori. We estimated solutions with 2 to 6 factions
and evaluated them according to their substantive and theoretical interpretability
as well as statistical goodness of fit (see Appendix C). The
latter measure is a proportion of correctness that accounts for the total
number of errors associated with absent within- and present cross-faction
ties [42,85]. The coalitions were then identified as intersections of
the results of the (1) K-means cluster analysis of PCBs (2-clusters solutions;
silhouette = 0.37) and (2) faction analysis of the political cooperation
network (3-factions solution; correct proportion = 0.715).
We also applied other clustering techniques to check the robustness of
our results (for more details, see Appendix B). The final classification
results in the actors being classified into three subgroups: two coalitions
and a residual subgroup.
We used standard descriptive statistics to describe the coalitions’
attributes in terms of PCBs as well as their relational characteristics.
The same applies to the comparison of the decision-making actors with
other actor groups (H2). Additionally, we have used Levene's test [86]
for testing the difference of the policy dimension variance between the
decision-makers and the other actor groups as defined below in Fig. 2
(H3). Levene's test is a parametric median-based approach used to test if
k groups (sub-samples) have equal variance. Thus, the rejection of null
hypothesis (assumption of equality of the variances) between the decision-makers
and the other groups would support our expectations.
We also used deductive block-modeling based on Euclidean distances
through the structural equivalence UCINET procedure [85]. A
block model is a simplified representation of a network that consists of
groups of nodes (blocks) that have similar relations to others and patterns
of relations among nodes and blocks (social roles) [87,88]. Deductive
block-modeling then compares observed structures to hypothesized
models, allowing for the test of hypotheses about structural
configurations of a network. The presence or absence of a tie between
each pair of nodes is then regressed on a set of dummy variables representing
blocks; the last block is used as the reference category [89].
The statistical significance assessment is based on the Quadratic Assignment
Procedure [90], which uses 5000 trials of random matrix
permutations and estimated standard errors based on resulting simulated
sampling distribution. The model fit is given by a ratio of the
variance in the pairwise presence or absence of ties explained by the
differences among the blocks [89].
The ENGOs’ and industry actors’ targeting of decision-makers via
political cooperation and expert information exchange (H4 and H5) was
tested using two block models. Each model divided the corresponding
matrix into four blocks based on actors’ membership in (1) ENGOs, (2)
industry, (3) decision-maker groups, or (4) none of the above. The industry
group was included to control for its parallel activities. Thus, an
above-average overall density of the 1–3 (ENGOs to decision-makers)
Table 1
Response rate based on actor type.
Actor type Responded Total Response rate
(%)
Central and regional governance 16 16 100
Central and regional political parties 16 18 89
Environmental non-governmental
organizations
8 9 89
Research organizations 14 16 88
Professional associations and trade
unions
3 7 43
Industry 11 17 65
Total 68 83 82
P. Ocelík, et al. Energy Research & Social Science 58 (2019) 101283
4
block would support our expectations formulated in hypotheses H4a
and H5a. Likewise, an above-average overall density of the 2–3 (industry
to decision-makers) block would support our expectations formulated
in hypotheses H4b and H5b (for more details, see Appendix C).
The coalition-based patterns of expert information exchange (H6)
were tested using a block model that divided the expert information
exchange matrix into three blocks based on actors’ membership in the
(1) Industry Coalition, (2) Environmental Coalition, or (3) a residual
group. Thus, an above-average overall density of the within-coalition
blocks 1–1 (Industry Coalition) and 2–2 (Environmental Coalition) would
support our hypothesis (for more details, see Appendix C).
We used UCINET 6 [85] and R 3.3.1 [91] for data processing,
analysis, and visualization. The data were collected through the SurveyMonkey
platform [92].
4. Case description
The Czech Republic is a post-communist regime classified as a rather
consensus model of democracy with a bicameral parliament,
multiparty system, proportional representation, and prevailing practice
of executive power-sharing in coalition governments [93,94]. The decision-making
process on policy proposals is controlled primarily by the
ruling political parties relying on support from the majority in the
Chamber of Deputies and the competent Ministries (Ministry of Industry
and Trade, Ministry of Agriculture, Ministry of Environment),
which are primarily responsible for policy formulation and implementation.
Interest group representation, although formally resembling
a neocorporativist model, is not compromise-oriented and its
functioning is closer to neopluralism. Thus, the associated policy venues
are primarily used as vehicles for interest promotion and have only
consultative competencies [95,96]. This creates a situation of majoritarian
politics in a (semi)consensual institutional setting.
As a legacy of the command economy, Czech energy policy has been
formed mostly by technicians recruited from industry and thus is designed
for large centralized resources based on conventional fuels,
especially coal and nuclear power. The position of conventional resources
is further bolstered by a strong emphasis on energy self-sufficiency,
which values coal as a safe domestic source [97]. In media
discourse, coal has been framed predominantly economically and
technically through the performance of mining companies and the future
of the industry. Environmental issues, much less represented, are
mostly related to the direct mitigation of adverse local impacts [81].
This image of coal is further strengthened by substantial anti-environmentalist
elite framing [98,99] and skeptical public perception of
renewable energy [100].
The country is characterized by a coal-based economy [101] which
has regularly ranked among the largest world net exporters of electricity
with about 13 TWh generated by an equivalent of 16 million tons
of burnt coal [102]. Although there are reserves of bituminous coal,
brown coal, as well as lignite, only brown and bituminous coal continue
to be mined. There is a planned phase-out in 2023 of bituminous coal
production which is located in the Czech part of the Upper Silesian
Basin. Brown coal opencast mining is located in the North Bohemian
Basin (also frequently referred to as the Most Basin) and the Sokolov
Basin [79,81]. There are 871 million tons of recoverable brown coal
reserves owned by the state as well as private companies, thus creating
a strong incentive to expand or at least maintain production [82,103].
Moreover, this position has been consistently supported by the Ministry
of Industry and Trade, which has repeatedly proposed rescinding the
territorial ecological limits of surface coal mining (henceforth “the limits”)
established by the first post-communist government in the brown coal
regions of the Czech Republic. The limits were imposed primarily to
diminish the ecological burden wrought by surface extraction of brown
coal and the production of electricity in those regions as well as to
reduce the impact on urbanized areas and other infrastructure
[79,104]. They have been the focus of ongoing public and political
debate since their establishment in 1991 [105]. The question of the
limits soon became one of the major issues for the formation of the
Czech environmental movement, which now has professionalized and
started to rely on advocacy, lobbying, and expert knowledge for promoting
its agenda [80,106,107]. Evidence-based tactics have also been
used by proponents of rescinding the limits, including the Ministry of
Industry and Trade, arguing it is necessary to secure heating supplies
(cf. [108,109]). Thus, despite the fact that the State Energy Policy [110]
envisages a long-term phase-out of coal, the process has been protracted
and could be described as a contested transition.
5. Results
The Czech coal subsystem was clustered into two coalitions with
opposing PCBs (H1). In this section, we present the composition and
attributes of these coalitions, including their PCBs, as well as the results
of the evaluation of the three investigated mechanisms.
5.1. Usual suspects: industry coalition and environmental coalition
We identified two coalitions, the Industry Coalition (IC) and
Environmental Coalition (EC), based on the combination of the K-means
cluster analysis used for exploring PCBs and the faction analysis applied
to the political cooperation network. A 2-clusters solution was selected
for the K-means cluster analysis of PCBs (see Appendix B) based on the
highest silhouette width of 0.371 (within the range of 2–10 clusters).
The 3-factions solution identified three cohesive subgroups (factions)
with high proportion of correctness at 0.705 (see Appendix C).
Although increases in the number of factions produced solutions with
somewhat higher proportions of correctness, the first two factions in the
3-factions solution are to a large extent homogenous in terms of PCBs
and show high face validity in contrast to solutions with more than 3
factions.
These results support the first hypothesis, which assumes the presence
of two coalitions with distant PCBs. Their distance in terms of
PCBs is examined in Section 5.3.
The political cooperation matrix blocked by coalition memberships
(see Table 2) showed high within-group densities for both the Environmental
(0.255) and Industry (0.279) coalitions, whereas withingroup
density for the Residual Group (0.055) was not statistically different
from the overall network density (0.090). Although there are
disproportionately more interactions within the coalitions, the low fit of
the block model (0.067) shows that coalition membership is by far not
the only driver of political cooperation. This is not that surprising since
a number of other effects (be it network endogenous effects such as
triadic closure or exogenous effects such as different forms of
Table 2
Block model – coalition membership.a
1: IC 2: EC 3: RG
1: IC 0.279 0.065 0.137
2: EC 0.078 0.255 0.056
3: RG 0.062 0.013 0.055
a
Adj. R2
= 0.067, p < 0.001 (significant values in bold), obs. = 4556.
Table 3
Coalition membership based on actor type.
Industry coalition (n = 17) Environmental coalition (n = 18)
2 political parties (Ustí region) 8 ENGOs
3 political parties (central) 2 state agencies (central)
3 NGOs 2 political parties (central)
2 state agencies (central) 6 research organizations
1 regional agency (Ústí region)
6 companies
P. Ocelík, et al. Energy Research & Social Science 58 (2019) 101283
5
homophily or entrainment) are expected to be at play (cf. [51]). The
within-group interactions in the Residual Group as well as the betweengroup
interactions are not statistically different from the overall
average (network density) except for blocks 1–3 (the Industry Coalition
to Residual Group) and 3–2 (Residual Group to Environmental Coalition).
The 1–3 block consists of ties from the Industry Coalition to mostly decision-making
and administrative bodies (approx. 75% of the 1–3 ties).
The 3–2 block exhibits significantly lower density of ties compared to
the overall network density, indicating a lack of recognition of ENGOs’
and research actors’ relevance from the Residual Group.
5.2. Composition of the coalitions
There are 17 members in the Industry Coalition and 18 actors in the
Environmental Coalition (see Table 3). The properties of the coalitions in
terms of their reputational power, core beliefs, PCBs, as well as structural
characteristics and percentage of members included in the core of
the political cooperation network are summarized in Table 4. The same
indicators were used to describe the whole political cooperation network.
The Industry Coalition consisted of 17 actors and was dominated by
the industry sector represented by six companies, a professional association,
and a trade union. There are also three political parties at the
state level (two ruling and one opposition party) and two political
parties in the Ústí region (both ruling parties) as well as two central
state agencies, one of them being the Ministry of Industry and Trade, a
regional agency, and a regional council. The Environmental Coalition
consisted of 18 actors, mostly ENGOs (8) and research organizations
(6). The remaining members are two political parties from the state
level (one of them is an opposition party, the other one is not represented
in the Chamber of Deputies) and two state agencies, one of
them being the Ministry of Agriculture.
Relationally, both coalitions have comparable densities that are
about three times the overall average (0.090). The IC is highly centralized
(0.533) around a state-owned company (coalition out-degree
= 10), a ruling party (11), and a trade union (11), who are the
most active actors, as well as the Ministry of Trade and Industry, who
was the most popular actor (coalition in-degree = 12). The centralization
of the EC is markedly lower (0.375). Nevertheless, the EC
subgraph exhibited a core–periphery structure (core–periphery
fit = 0.733; see Appendix C). This might indicate different cooperation
patterns within the coalitions. Whereas the IC seems to be centralized
around the aforementioned actors and was clustered at the state and
regional level, the EC was formed around a clique of 5 ENGOs (core
members) that exchange resources and coordinate with the rest of the
coalition. At the network level, there was only a moderate fit with the
core–periphery structure (0.41; see Appendix C), corresponding with
the expected presence of two relationally cohesive groups – coalitions.
The core consisted of 23 organizations, eleven of which belong to the IC
(48% of the core) and six to the EC (26%). In terms of reputational
power, the IC had about a 2.5× higher score (0.478) than the EC
(0.183). The differences were statistically significant and could indicate
that although the EC was well-connected to the decision-making actors
(more below) than the rest of the network in general, the impact of its
interactions was limited. The core beliefs, expressing normative assumptions
about how society and the economy ought to be organized,
showed the expected ideological polarization. The core beliefs dimension
was measured through a composite variable rescaled to range
between 0 and 1, where the former indicates a laissez-faire position, the
latter an interventionist position, and 0.5 a neutral position. The IC
position (mean 0.384) was closer to an economically liberal view,
whereas the EC position (0.653) was closer to a mixed-economy approach
(for more information, see Appendix B).
5.3. Policy core beliefs
Comparison of the coalitions’ PCBs, specifically their policy, economic,
environmental, and process dimensions, supported the expectation
about their opposing positions. Compared to the entire network,
the lower standard deviations of the PCBs dimensions of the
coalitions supported the assumption of their belief homogeneity (cf.
[42]). The PCBs dimensions were composite variables rescaled to range
between 0 and 1, where the former indicates a strong pro-coal position,
the latter a strong anti-coal position, and 0.5 a neutral position (see
Appendix B).
The coalitions were markedly distinct in all four dimensions. In the
economic dimension, the Industry Coalition (mean 0.309) uniformly
supports the overall economic benefits of coal mining and use.
Nevertheless, the impacts of the decline in surface mining (henceforth
“mining”) on household heating supply and regional development are
evaluated differently. The IC position is rather homogenous in the
former but divided in the latter. The Environmental Coalition (mean
0.779) rejects the overall economic benefits of mining and its importance
for regional development and marginally recognizes the negative
impacts of mining decline on household heating supply. In the
environmental dimension, the IC (mean 0.413) is divided on the environmental
impacts of the mining and burning of coal and completely
rejects its impacts on climate change. In contrast, the EC (mean 0.914)
is nearly unanimous in its negative evaluation of coal's impacts on the
environment as well as climate change. In the policy dimension, the IC
(mean 0.249) unambiguously favors the importance of coal for the
Czech Republic's energy independence, with only a minor dissent from
the view that coal should be a key component of the future energy mix
and that the mining limits should be lifted. The EC (mean 0.798) to
some extent concedes that coal plays a role in energy independence but
is strongly against the rescinding of mining limits and a future energy
mix based on coal. In the process dimension, the IC (mean 0.331) is divided
on the question of trust among key players in the coal industry.
On the other hand, it largely agrees on the functionality of the mine life
Table 4
Coalition attributes.a,b
Reputational power PCB economy PCB environment PCB policy PCB process Density Degree centralization % core
IC 0.478a
0.309a
0.413a
0.249a
0.315a
0.279 0.533 48
(0.232) (0.222) (0.197) (0.146) (0.106)
EC 0.183b
0.779b
0.914b
0.798b
0.757b
0.255 0.375 26
(0.175) (0.161) (0.139) (0.142) (0.171)
RG 0.211b
0.480c
0.574c
0.451c
0.497c
0.055 0.174 26
(0.200) (0.234) (0.211) (0.233) (0.253)
Total 0.271 0.516 0.624 0.492 0.520 0.090 0.353 100
(0.233) (0.273) (0.266) (0.278) (0.258)
a
All variables range between 0 and 1. For details on PCB dimensions, see the Data and Methods section and Appendix B.
b
One-way ANOVA tests for reputational power and PCBs economy dimension significant at p < 0.05. One-way ANOVA tests for remaining three PCBs dimensions
significant at p < 0.001. Fisher's Least Significant Difference test was used to determine pairwise differences between the three groups for each dimension.
The groups with a different superscript letter are significantly different at p < 0.05 with Bonferroni correction. For more details, see Appendix B.
P. Ocelík, et al. Energy Research & Social Science 58 (2019) 101283
6
cycle and adequateness of the legal and regulatory framework. The EC
(mean 0.782) sees the relationships among key players as distrustful, is
divided on the issue of mine life cycle functionality, and sees the legal
and regulatory framework as insufficient (Fig. 1).
These results support the first hypothesis, which assumes the presence
of two coalitions with distant PCBs.
More specifically, the two coalitions significantly differ across all
four dimensions. The largest distance between the coalitions is in the
policy dimension, which captures the future role of coal including the
politically sensitive question of mining limits. Likewise, they markedly
disagree on the environmental dimension, which accounts for environmental
degradation including climate change impacts, as well as the
economy dimension, which evaluates the economic benefits of coal
mining. In the process dimension, both coalitions exhibit the largest
variance, thus indicating weaker within-coalition consensus.
5.4. Decision-making actors: far from consensus
The first investigated mechanism, fragmentation of political authority,
focuses on (H2) cross-coalition membership of the decision-making
actors and (H3) comparatively higher variability of their PCBs. As for
the former, three key decision-making actors (Ministry of Agriculture,
Ministry of Environment, and Ministry of Industry and Trade) have
been classified in different groups. The Ministry of Agriculture is a
member of the EC, the Ministry of Industry and Trade is a member of
the IC, and the Ministry of Environment is part of the Residual Group
(see Appendix B). Each of the ministries is controlled by a different
ruling party, their PCBs are markedly different, and they are influenced
Fig. 1. PCB comparison based on coalition membership.1,2
1
All variables range between 0 (strong pro-coal position) and 1 (strong anti-coal position). For details on PCB dimensions, see the Data and Methods section and
Appendix B.
2
Multiple comparisons test was used to determine pairwise differences between the three groups for each dimension. The groups marked with a different superscript
letter and different color are significantly different at p < 0.05. For more information, see Appendix B.
P. Ocelík, et al. Energy Research & Social Science 58 (2019) 101283
7
by opposing coalitions.
Thus, these results support the second hypothesis on fragmentation
of decision-makers based on their cross-coalition memberships.
We also applied a median-based Levene's test [86] to evaluate
whether the sub-samples (i.e., the decision-makers and the other six
remaining groups) have equal variances in the policy dimension of PCBs
which captures views on the future of coal in the Czech Republic and
seems to constitute a cleavage within the subsystem. The results show
that the decision-makers’ variance was larger and statistically different
(p < 0.05) only when compared to the ENGOs but not to the other
actor groups (see Appendix B).
These results do not support the third hypothesis on higher variability
of decision-makers’ PCBs in comparison to the other actor groups.
The descriptive measures (see also Fig. 2), nevertheless, indicate a large
spread of their PCBs with a range from 0.11 to 0.89 and a standard
deviation (0.23) for the policy dimension.
5.5. Targeting decision-makers: a shared practice
The second investigated mechanism, access to political authority,
Fig. 2. PCB scores based on actor type.1,2
1
All variables range between 0 (strong pro-coal position) and 1 (strong anti-coal position). For details on PCB dimensions, see the Data and Methods section and
Appendix B.
2
The blue dots represent the scores of individual organizations. The dotted line represents a neutral position (0.5); the green line represents the mean value.
Table 5
Block model – political cooperation.a
1: ENGOs 2: industry 3: decision-makers 4: remainder
1: ENGOs 0.786 0.063 0.333 0.048
2: industry 0.042 0.136 0.181 0.107
3: decision-makers 0.104 0.167 0.267 0.099
4: remainder 0.054 0.069 0.143 0.057
a
Adj. R2
= 0.095, p < 0.001 (significant values in bold), obs. = 4556.
P. Ocelík, et al. Energy Research & Social Science 58 (2019) 101283
8
focuses on ENGOs’ and industry actors targeting of decision-makers via
political cooperation (H4a and H4b) and expert information exchange (H5a
and H5b) ties. Thus, the resulting block model captures patterns of
interactions among the aforementioned groups and the rest of the
network (the remainder). The block model of the political cooperation
network (Table 5) shows that both ENGOs and the industry extend
disproportionately more ties to the decision-making actors, thus presenting
an indication of the assumed targeting. More specifically, the
densities of the 1–3 (ENGOs to decision-makers) and 2–3 (industry to
decision-makers) blocks are 0.333 and 0.181 in comparison to the
overall network density of 0.090. There are also disproportionately
more ties from decision-makers to the industry (3–2 block with density
0.167), which also supports the findings on privileged access of the
Industry Coalition. Additionally, there was evidence of targeting also in
the case of the remainder (4–3 block with density 0.143), which includes
research actors. Interestingly, the density of the 2–2 industry block
(0.136) is not statistically different from the overall network density.
This may indicate the co-existence of cooperative and competitive relationships
among the industry actors. The block model of the expert
information exchange network (Table 6) provided a similar picture. The
densities of the 1–3 (ENGOs to decision-makers) and 2–3 (industry to
decision-makers) blocks were 0.563 and 0.278, respectively, in comparison
to the overall network density of 0.173. There were also two
notable differences from political cooperation. Firstly, the 3–2 block
(decision-makers to industry) does not differ from the overall network
density. Secondly, there are disproportionately more ties from the industry
to the remainder (2–4), which also includes administrative
bodies. The lower fit of the block models (adj. R2
= 0.098 for each) is
not surprising given the actor heterogeneity of the coalitions. Although
targeting was only one of the drivers of interactions, it clearly contributes
to its structuring.
These results partially support the fourth (H4a, H4b) and fifth (H5b,
H5b) hypotheses, which assume targeting of decision-makers by ENGOs
and industry actors. The partial support of the hypotheses is due to the
comparable pattern of ties from the residual group (which includes nonprincipal
coalition members such as research organizations) to the de-
cision-makers.
5.6. Coalition-based patterns of expert information exchange
The third investigated mechanism, use of expert information, focuses
on (H6) prevalent patterns of expert information ties within the subsystem's
coalition structure. The block model (Table 7) of the expert
information exchange network shows that within-coalition densities of
the Industry Coalition (1–1: 0.478) and Environmental Coalition (2–2:
0.441) are almost three times higher than the network density (0.173).
The only other statistically different interaction is the 1–3 block, which
consists of ties from the Industry Coalition (1) to mostly decision-making
and administrative bodies in the Residual Group. The cross-coalition
densities (1–2 and 2–1), on the other hand, do not statistically differ
from the network density. As in the previous case, the lower fit (adj.
R2
= 0.094) of the block model indicates that coalition membership is
only one of the drivers of expert information exchange.
These results support the sixth hypothesis, which assumes a higher
density of expert information exchange within the coalitions than
across them.
6. Discussion
Sociotechnical transitions are inherently political as they involve
diverse actors that compete over their pace and direction [7,111]. Resulting
external perturbations (such as rapid growth of new technologies)
then impact more specific and semi-autonomous sectors (such as the
energy industry) in terms of opening opportunities for major policy
change. In this context, we focused on the conditions for policy change
in an adversarial subsystem, specifically, the case of the coal subsystem
in the Czech Republic. We examined the subsystem's coalition structure
and investigated particular mechanisms through which the present
coalitions and policy actors interact and thereby influence the subsystem
dynamics. In line with the theoretical expectations on mature
[19,31] and adversarial subsystems [25,59], we identified two opposing
coalitions. The Industry Coalition led by industry incumbents is
the dominant coalition with superior resources including direct access
to decision-making through two central and two regional ruling political
parties and the Ministry of Industry and Trade. It considers coal to
be as an integral part of the future energy mix and supports rescinding
the mining limits. The Environmental Coalition led by ENGOs, on the
other hand, challenges the status quo and promotes a rapid departure
from coal. It tries to offset its inferior position by focusing on the key
issue of mining limits and by using its transactional capacity [63].
As expected in adversarial subsystems [25,59], the two coalitions
differ markedly in their PCBs. In addition to the pro-status quo orientation
of the dominant coalition, this circumstance limits opportunities
for policy change through negotiated agreement [14]. In this
high-conflict environment, the situation is expected to be further affected
by the so-called devil shift—a systematic bias to overestimate the
capacities and bad intentions of political opponents which further
deepens distrust between competing advocacy coalitions [112] and
thus hinders policy learning opportunities [113]. As a result, any internally
driven major policy change is blocked by a competing coalition.
However, this situation does not generate a “hurting stalemate”
[112] but rather a series of partial incoherent policy modifications that
both introduce transition-oriented measures as well as reinforce the
vested interests of the industry.
In this context, the Environmental Coalition utilizes opportunity
structures provided by external perturbations, such as increasing
Europeanization of energy and environmental policies [114,115], to
further legitimize and boost its demands to the government. The
dominant Industry Coalition, on the other hand, has been relatively
successful in resisting these pressures and maintaining the status quo.
This is well-documented by recent notable events such as the rescission
of limits on the Bílina mine site of the majority state-owned company
North Bohemia Coal Mines in 2015 [109]. While the Ministry of Trade
and Industry tried to push the rescission of limits further in the proposed
Raw Material Policy, this effort has been rejected by the government
[116]. Likewise, although the updated State Energy Policy, a key
strategic document, proposes penalization of low-efficiency coal-fueled
Table 7
Block model – expert information exchange.a
1: IC 2: EC 3: RG
1: IC 0.478 0.121 0.209
2: EC 0.183 0.441 0.116
3: RG 0.153 0.066 0.112
a
Adj. R2
= 0.094, p < 0.05 (significant values in bold), obs. = 4556.
Table 6
Block model – expert information exchange.a
1: ENGOs 2: industry 3: decision-makers 4: remainder
1: ENGOs 0.982 0.167 0.563 0.140
2: industry 0.073 0.348 0.278 0.206
3: decision-makers 0.208 0.208 0.633 0.139
4: remainder 0.122 0.161 0.226 0.120
a
Adj. R2
= 0.095, p < 0.001 (significant values in bold), obs. = 4556.
P. Ocelík, et al. Energy Research & Social Science 58 (2019) 101283
9
power generation, the government has not approved an amendment
that would set related obligatory compensation payments [110,117].
Importantly, the key decision on the limits for the Czechoslovak Army
Mine has been postponed to 2020, thus keeping alive the opportunity to
substantially prolong coal mining and utilization [118]. Nevertheless, it
would be a simplification to describe the subsystem only in terms of the
two competing coalitions.
That said, we reflect on the early critique of the Advocacy Coalition
Framework's overreliance on belief systems at the expense of its relational
element [119,120] as well as strategically oriented motives of
actors’ interactions [121]. In other words, the assumption of shared
policy core beliefs and coordinated activity implies a restrictive definition
of coalitions. To obtain a more nuanced picture, it is useful to focus on
the relational element (factions) in order to investigate with whom
actors cooperate, regardless of their PCBs. Overall, the results are not
notably different from the coalition detection (see Appendix B), which
supports the initial expectation of two competing and ideologically
distant coalitions. However, there are also deviant cases that differ from
the average PCBs scores of the factions. These actors are tightly linked
with other members of the given faction but do not share their PCBs
(see Appendix B). As such, they may be regarded as potential brokers
that could facilitate dialogue with actors outside the faction and
thereby encourage policy learning. The Government's Office occupies
such a position within the Industry Coalition-led faction. Following this
argument, its position should be further strengthened and utilized. A
related methodological implication is that it might be problematic to
use only subgroup cohesion measures, such as faction analysis, to
identify coalitions (cf. [42,48]).
To provide a more comprehensive view on how the adversarial
subsystem functions, we have further complemented the coalition
analysis with a focus on three particular mechanisms: (1) fragmentation
of decision-makers, (2) targeting of decision-makers, and (3) use of
expert information.
First, the decision-makers are fragmented in terms of their prevailing
interactions as well as PCBs. The resulting cross-coalition memberships
of decision makers are expected to limit policy learning opportunities in
a high-conflict environment (cf. [19,33]) and generally impede longterm
planning as well as management. This is important especially
given the mixed membership of the three competent ministries. Moreover,
a large spread in PCBs is present also among state agencies responsible
for supervising and evaluating policies. In general, such political
authority fragmentation is considered to constitute an important
barrier to the transformation of policy learning into specific policy
changes [26,122]. As Poloni-Staudinger [123] shows, the institutions
and mechanisms of consensual democracy tend to improve the environmental
effectiveness of governments [94]. Their strengthening
through inclusively designed policy venues supporting cross-coalition
interactions and stakeholder engagement could therefore incentivize
policy learning (cf. [51]). The creation of institutions that would coordinate
and/or supervise transition-related activities of the relevant
state agencies seems to be a reasonable option.
Second, the targeting of decision-makers through political cooperation
and expert information is done by both ENGOs and industry as well as by
research organizations. The ENGOs’ activities are in line with transactional
activism [63], which aims to form partnerships with key government
actors [124]. Nevertheless, there is clear evidence that industry
actors, dominated by incumbents, rely on the same tactic [68].
Importantly, decision-makers extend disproportionately more political
cooperation ties exclusively to industry, which is expected to strengthen
regime-resistance [68]. The use of the targeting tactics to induce a
gradual policy change is thus contested by opposing actor groups that
include most of the principal members of competing coalitions [25].
Altogether, we expect that the targeting practiced by ideologically
distant actors that belong to different coalitions over time deepens
fragmentation among decision-makers and increases the level of conflict
within the subsystem.
Third, expert information exchange strongly overlaps with coalition
membership. This finding suggests that the use of expert information
increases cohesiveness within coalitions and shows a limited potential
for policy learning across coalitions. In contrast to the technocratic view
[125], this indicates a political nature underlying expert information
[25] amplified by conditions of high conflict and lack of established
expert forums [19,22]. As in the previous case, the involvement of
policy brokers and entrepreneurs is expected to enhance opportunities
for cross-coalition policy learning [126,127]. In this context, a pragmatic
approach to expert information [128] that requires critical interaction
between experts, decision-makers, and the public should be
promoted [129].
The study has important limitations that should be addressed in
future research. First, it is a single cross-sectional case study, which implies
limited generalizability of the results. The study primarily provides
novel empirical evidence (see [130]) to research on adversarial
policy subsystems [25,59] and, more broadly, policy dimension of energy
transition [7,10] that should be further integrated in comparative
studies (cf. [40,131]). Second, it uses mainly exploratory methods with
limited potential to account for confounding variables and other potential
sources of bias. Considering the above, there appear to be several
productive avenues for future research. Application of tie-formation
models could be used to identify drivers of cooperation that allow
for multi-theoretical and mechanism-based explanations [132], while a
longitudinal study would be a logical step enabling the examination of
such processes over an extended period of time [28,133]. A position
analysis could help to identify actors potentially critical for overcoming
polarization of the policy subsystem and fragmentation of political
authority [28,134], optimally executed as mixed-methods research in
order to uncover how key actors themselves understand their position
[135]. Likewise, taking into account the importance of landscape
pressures, research on external perturbations [16,18] as well as policy
transfer seems to be especially promising [136].
7. Conclusions
This study examined advocacy coalitions and interactions among
policy actors within an adversarial subsystem. More specifically, we
focused on the Czech coal subsystem, which faces increasing transitioninduced
pressures. We identified two ideologically distant coalitions
that compete over access to fragmented political authority and whose
expert information exchange tends to reproduce within-coalition coordination
patterns. Thus, we argue that the Czech coal phase-out will
not unfold through incremental adjustments [137] that result from
policy learning or negotiated agreements. Instead, it will continue to be
driven by external perturbations which, nevertheless, do not disrupt
established institutional and policy arrangements due to the adversarial
nature of the subsystem. As a result, we expect the overlay of new arrangements
on top of those already existing without transforming the
subsystem's core logic [138]. This process is further facilitated by the
substantial relational as well as ideological fragmentation of the decision-making
actors and political authority in general. As argued by
Geels et al. [139], such developments could nevertheless also indicate a
potential for more substantive policy changes if actors encounter new
problems and see new opportunities. The results of the IPPC's Special
Report [140] will likely bring such incentives for policy change, including
to reopen the discussion on the so-called Anti-fossil Act which
would authoritatively outline a transition pathway where a new round
of struggle between the Industry and Environmental coalitions could be
expected.
This study adds to the Advocacy Coalition Framework research by
analyzing the internal functioning of an adversary subsystem and
highlighting limitations of coalition detection techniques. It also makes
a theoretical contribution by complementing the Advocacy Coalition
Framework with mechanism-based explanations derived from related
literatures, particularly that on socio-technical transitions. Such
P. Ocelík, et al. Energy Research & Social Science 58 (2019) 101283
10
approach locates the study of policy change within the complex context
of socio-technical systems, thus substantially exceeding the issue-specific
focus of policy subsystems and addressing the assumption of their
interdependency (cf. [19]). The Advocacy Coalition Framework, on the
other hand, provides a conceptual and methodological toolkit wellsuited
for in-depth analysis of policy processes, a crucial part of the
socio-technical systems’ institutional structure (cf. [7]). Integrating the
two frameworks thus offers significant potential to enhance research on
energy transition by explicitly linking its macro-level (e.g., decline of
specific industries) and meso-level (e.g., related policy changes) pro-
cesses.
Acknowledgments
The authors would like to thank Nadja Kunz, Manuel Fischer, Juraj
Medzihorský, Filip Černoch, Tomáš Diviák, John Owen, Tomáš Vlček,
Jan Daniel, participants of the Research Seminar at the Masaryk
University, participants of the SMI Research Seminar at the University
of Queensland, and to anonymous reviewers for valuable comments as
well as to Colin Kimbrell who provided language editing services.
Importantly, the authors also thank to all the respondents for kindly
participating in the study. This research was supported by the Czech
Science Foundation (Grant 17–22978Y).
Supplementary materials
Supplementary material associated with this article can be found, in
the online version, at doi:10.1016/j.erss.2019.101283.
References
[1] V. Smil, Energy and Civilization: A History, The MIT Press, 2017.
[2] IEA, CO2 Emissions from Fuel Combustion 2018: Overview, (2018). https://
webstore.iea.org/co2-emissions-from-fuel-combustion-2018-overview (Accessed
15 December 2018).
[3] Europe's Energy Transition is Well Underway, European Commission, 2017.
https://ec.europa.eu/energy/en/news/europes-energy-transition-well-underway
(Accessed 15 December 2018).
[4] International Energy Agency, World Energy Outlook iea.org/weo, (2018), https://
doi.org/10.1787/weo-2017-en.
[5] S.H. Ali, D. Giurco, N. Arndt, E. Nickless, G. Brown, A. Demetriades, R. Durrheim,
M.A. Enriquez, J. Kinnaird, A. Littleboy, L.D. Meinert, R. Oberhänsli, J. Salem,
R. Schodde, G. Schneider, O. Vidal, N. Yakovleva, Mineral supply for sustainable
development requires resource governance, Nature 543 (2017) 367–372, https://
doi.org/10.1038/nature21359.
[6] F.W. Geels, J. Schot, Typology of sociotechnical transition pathways, Res. Policy
36 (2007) 399–417, https://doi.org/10.1016/J.RESPOL.2007.01.003.
[7] J. Markard, M. Suter, K. Ingold, Socio-technical transitions and policy change –
advocacy coalitions in Swiss energy policy, Environ. Innov. Soc. Transitions 18
(2016) 215–237, https://doi.org/10.1016/J.EIST.2015.05.003.
[8] T. Hoppe, F. Coenen, M. van den Berg, Illustrating the use of concepts from the
discipline of policy studies in energy research: an explorative literature review,
Energy Res. Soc. Sci. 21 (2016) 12–32, https://doi.org/10.1016/J.ERSS.2016.06.
006.
[9] E. Laes, L. Gorissen, F. Nevens, A comparison of energy transition governance in
Germany, The Netherlands and the United Kingdom, Sustainability 6 (2014)
1129–1152, https://doi.org/10.3390/su6031129.
[10] J. Markard, The next phase of the energy transition and its implications for research
and policy, Nat. Energy 3 (2018) 628–633, https://doi.org/10.1038/
s41560-018-0171-7.
[11] J. Meadowcroft, What about the politics? Sustainable development, transition
management, and long term energy transitions, Policy Sci. (2009), https://doi.
org/10.1007/s11077-009-9097-z.
[12] P.A. Sabatier, An advocacy coalition framework of policy change and the role of
policy-oriented learning therein, Source Policy Sci. 21 (1988) 129–168 https://
www.jstor.org/stable/pdf/4532139.pdf?refreqid=excelsior
%3Ae17c0dc9672b2d132e181c22b4c4616c (accessed April 3, 2019).
[13] P.A. Sabatier, C.M. Weible, The advocacy coalition framework: innovations and
clarifications, in: P.A. Sabatier (Ed.), Theor. Policy Process, Westpress, 2007, pp.
189–220.
[14] P.A. Sabatier, C.M. Weible, Theories of the Policy Process, (2014).
[15] P.A. Sabatier, Knowledge, policy-oriented learning, and policy change, Knowledge
8 (1987) 649–692, https://doi.org/10.1177/0164025987008004005.
[16] D. Nohrstedt, External shocks and policy change: Three Mile Island and Swedish
nuclear energy policy, J. Eur. Public Policy 12 (2005) 1041–1059, https://doi.org/
10.1080/13501760500270729.
[17] D. Nohrstedt, The politics of crisis policymaking: Chernobyl and Swedish nuclear
energy policy, Policy Stud. J. 36 (2008) 257–278, https://doi.org/10.1111/j.
1541-0072.2008.00265.x.
[18] A. Rinscheid, B. Eberlein, P. Emmenegger, V. Schneider, Why do junctures become
critical? Political discourse, agency, and joint belief shifts in comparative perspective,
Regul. Gov. (2019), https://doi.org/10.1111/rego.12238.
[19] P.A. Sabatier, The advocacy coalition framework: revisions and relevance for
Europe, J. Eur. Public Policy 5 (1998) 98–130, https://doi.org/10.1080/
13501768880000051.
[20] J.J. Pierce, H.L. Peterson, K.C. Hicks, Policy change: an advocacy coalition framework
perspective, Policy Stud. J. (2017), https://doi.org/10.1111/psj.12223.
[21] E.A. Albright, Policy change and learning in response to extreme flood events in
Hungary: an advocacy coalition approach, Policy Stud. J. 39 (2011) 485–511,
https://doi.org/10.1111/j.1541-0072.2011.00418.x.
[22] C.M. Weible, Expert-based information and policy subsystems: a review and
synthesis, Policy Stud. J. 36 (2008) 615–635, https://doi.org/10.1111/j.1541-
0072.2008.00287.x.
[23] H. Kriesi, S. Adam, M. Jochum, Comparative analysis of policy networks in
Western Europe, J. Eur. Public Policy (2006), https://doi.org/10.1080/
13501760500528803 1763.
[24] C.M. Weible, Beliefs and perceived influence in a natural resource conflict: an
advocacy coalition approach to policy networks, Polit. Res. Q. 58 (2005) 461–475,
https://doi.org/10.1177/106591290505800308.
[25] C.M. Weible, A. Pattison, P.A. Sabatier, Harnessing expert-based information for
learning and the sustainable management of complex socio-ecological systems,
Environ. Sci. Policy 13 (2010) 522–534, https://doi.org/10.1016/j.envsci.2010.
05.005.
[26] M. Fischer, Coalition structures and policy change in a consensus democracy,
Policy Stud. J. 42 (2014) 344–366, https://doi.org/10.1111/psj.12064.
[27] M.A. Zafonte, P.A. Sabatier, Shared beliefs and imposed interdependencies as
determinants of ally networks in overlapping subsystems, J. Theor. Polit. 10
(1998) 473–505, https://doi.org/10.1177/0951692898010004005.
[28] P.A. Sabatier, H.C. Jenkins-Smith, Policy Change and Learning : An Advocacy
Coalition Approach, Westview Press, 1993.
[29] T. Vlček, F. Černoch, Energy Sector and Energy Policy of the Czech Republic,
Masarykova Univerzita, 2013, https://doi.org/10.5817/CZ.MUNI.M210-6523-
2013.
[30] S. Wasserman, K. Faust, Social Network Analysis : Methods and Applications,
Cambridge University Press, 1994. https://books.google.cz/books/about/Social_
Network_Analysis.html?id=CAm2DpIqRUIC&redir_esc=y (Accessed 7 January
2019).
[31] K. Ingold, Network structures within policy processes: coalitions, power, and
brokerage in Swiss climate policy, Policy Stud. J. 39 (2011) 435–459, https://doi.
org/10.1111/j.1541-0072.2011.00416.x.
[32] C.M. Weible, T. Heikkila, K. Ingold, M. Fischer, Policy Debates on Hydraulic
Fracturing Comparing Coalition Politics in North America and Europe, Palgrave
Macmillan US, New York, 2016, https://doi.org/10.1057/978-1-137-59574-4.
[33] C.M. Weible, P.A. Sabatier, K. McQueen, Themes and variations: taking stock of
the advocacy coalition framework, Policy Stud. J. 37 (2009) 121–140, https://doi.
org/10.1111/j.1541-0072.2008.00299.x.
[34] P. Leifeld, V. Schneider, Information exchange in policy networks, Am. J. Pol. Sci.
56 (2012) 731–744, https://doi.org/10.1111/j.1540-5907.2011.00580.x.
[35] P. Cairney, M. Fischer, K. Ingold, Hydraulic fracturing policy in the UK: coalition,
cooperation and opposition in the face of uncertainty, Polit. Stud. Assoc. Annu. Int.
Conf. 2015, pp. 1–31, , https://doi.org/10.1057/978-1-137-59574-4_4.
[36] C.M. Weible, P.A. Sabatier, Comparing policy networks: marine protected areas in
California, Policy Stud. J. 33 (2005) 181–201, https://doi.org/10.1111/j.1541-
0072.2005.00101.x.
[37] D. Knoke, Who steals my purse steals trash, J. Theor. Polit. 10 (1998) 507–530,
https://doi.org/10.1177/0951692898010004006.
[38] M. Fischer, P. Sciarini, Drivers of collaboration in political decision making: a
cross-sector perspective, J. Polit. 78 (2016) 63–74, https://doi.org/10.1086/
683061.
[39] J. Broadbent, Comparative climate change policy networks, in: J.N. Victor,
A.H. Montgomery, M. Lubell (Eds.), Oxford Handb. Polit. Networks, Oxford
University Press, 2016, , https://doi.org/10.1093/oxfordhb/9780190228217.
013.38.
[40] T. Ylä-Anttila, A. Gronow, M.C.J. Stoddart, J. Broadbent, V. Schneider,
D.B. Tindall, Climate change policy networks: why and how to compare them
across countries, Energy Res. Soc. Sci. 45 (2018) 258–265, https://doi.org/10.
1016/J.ERSS.2018.06.020.
[41] P. Cairney, The role of evidence in theories of the policy process, Polit. EvidenceBased
Policy Mak. Palgrave Macmillan UK, London, 2016, pp. 13–50, https://doi.
org/10.1057/978-1-137-51781-4_2.
[42] A. Gronow, T. Ylä-Anttila, Cooptation of ENGOs or treadmill of production?
Advocacy coalitions and climate change policy in Finland, Policy Stud. J. 00
(2016) 1–22, https://doi.org/10.1111/psj.12185.
[43] M. Brockhaus, M. Di Gregorio, National REDD+ policy networks: from cooperation
to conflict, Ecol. Soc. 19 (2014), https://doi.org/10.5751/ES-06921-190414.
[44] D.J. Elgin, C.M. Weible, A stakeholder analysis of colorado climate and energy
issues using policy analytical capacity and the advocacy coalition framework, Rev.
Policy Res. 30 (2013) 114–133, https://doi.org/10.1111/ropr.12005.
[45] D.J. Hess, Coalitions, framing, and the politics of energy transitions: local democracy
and community choice in California, Energy Res. Soc. Sci. 50 (2019)
38–50, https://doi.org/10.1016/J.ERSS.2018.11.013.
[46] M. Hudson, What does Canute want? The “Monash Forum” and the Australian
P. Ocelík, et al. Energy Research & Social Science 58 (2019) 101283
11
climate deadlock, Energy Res. Soc. Sci. 49 (2019) 126–133, https://doi.org/10.
1016/J.ERSS.2018.10.029.
[47] H. Kriesi, M. Jegen, The Swiss energy policy elite: the actor constellation of a
policy domain in transition, Eur. J. Polit. Res. 39 (2001) 251–287, https://doi.org/
10.1111/1475-6765.00577.
[48] P.M. Wagner, T. Ylä-Anttila, Who got their way? Advocacy coalitions and the Irish
climate change law, Environ. Polit. 27 (2018) 872–891, https://doi.org/10.1080/
09644016.2018.1458406.
[49] S.J. Yun, D. Ku, J.Y. Han, Climate policy networks in South Korea: alliances and
conflicts, Clim. Policy 14 (2014) 283–301, https://doi.org/10.1080/14693062.
2013.831240.
[50] D. Nohrstedt, Do advocacy coalitions matter? Crisis and change in Swedish nuclear
energy policy, J. Public Admin. Res. Theory 20 (2010) 309–333, https://doi.org/
10.1093/jopart/mun038.
[51] K. Ingold, M. Fischer, P. Cairney, Drivers for policy agreement in nascent subsystems:
an application of the advocacy coalition framework to Fracking policy in
Switzerland and the UK, Policy Stud. J. 45 (2017) 442–463, https://doi.org/10.
1111/psj.12173.
[52] P.M. Wagner, T. Ylä-anttila, Can policy forums overcome echo chamber effects by
enabling policy learning? Evidence from the Irish climate change policy network,
J. Public Policy (2018).
[53] P. Cairney, M. Fischer, K. Ingold, Fracking in the UK and Switzerland: why differences
in policymaking systems don't always produce different outputs and
outcomes, Policy Polit. (2016) 1–22, https://doi.org/10.1332/
030557316X14793989976783.
[54] J. Osička, L. Lehotský, V. Zapletalová, F. Černoch, B. Dančák, Natural gas market
integration in the Visegrad 4 region: an example to follow or to avoid? Energy
Policy 112 (2018) 184–197, https://doi.org/10.1016/J.ENPOL.2017.10.018.
[55] B. Rennkamp, S. Haunss, K. Wongsa, A. Ortega, E. Casamadrid, Competing coalitions:
the politics of renewable energy and fossil fuels in Mexico, South Africa
and Thailand, Energy Res. Soc. Sci. 34 (2017) 214–223, https://doi.org/10.1016/
J.ERSS.2017.07.012.
[56] M.C.J. Stoddart, D.B. Tindall, J. Smith, R. Haluza-Delay, Media access and political
efficacy in the eco-politics of climate change: Canadian national news and
mediated policy networks, Environ. Commun. 11 (2017) 386–400, https://doi.
org/10.1080/17524032.2016.1275731.
[57] E. Schmid, B. Knopf, A. Pechan, Putting an energy system transformation into
practice: the case of the German Energiewende, Energy Res. Soc. Sci. 11 (2016)
263–275, https://doi.org/10.1016/J.ERSS.2015.11.002.
[58] M. Sotirov, M. Memmler, The advocacy coalition framework in natural resource
policy studies — recent experiences and further prospects, For. Policy Econ. 16
(2012) 51–64, https://doi.org/10.1016/j.forpol.2011.06.007.
[59] C.M. Weible, T. Heikkila, J.J. Pierce, Understanding rationales for collaboration in
high-intensity policy conflicts, J. Public Policy 38 (2018) 1–25, https://doi.org/
10.1017/S0143814X16000301.
[60] M. Fischer, Looking beneath the surface: differences in decision-making structures
across processes, Polit. Decis. Switz. Palgrave Macmillan UK, London, 2015, pp.
78–97, https://doi.org/10.1057/9781137508607_4.
[61] T.J. Lowi, Four systems of policy, politics, and choice, Public Admin. Rev. 32
(1972) 298–310. https://www.jstor.org/stable/pdf/974990.pdf?refreqid=
excelsior%3A83d4d0c28d9eee46a9b0aa6e0dc84e58 (Accessed 15 December
2018).
[62] J.J. Pierce, H.L. Peterson, M.D. Jones, S.P. Garrard, T. Vu, There and back again: a
tale of the advocacy coalition framework, Policy Stud. J. 45 (2017) S13–S46,
https://doi.org/10.1111/psj.12197.
[63] T. Petrova, S. Tarrow, Transactional and participatory activism in the emerging
European polity, Comp. Polit. Stud. 40 (2007) 74–94, https://doi.org/10.1177/
0010414006291189.
[64] K. Čada, K. Ptáčková, Possibilities and limits of collaboration between science and
NGOs in the Czech Republic, J. Clean. Prod. 49 (2013) 25–34, https://doi.org/10.
1016/J.JCLEPRO.2012.07.039.
[65] O. Císař, J. Navrátil, K. Vráblíková, Staří, noví, radikální: politický aktivismus v
České republice očima teorie sociálních hnutí, Czech Sociol. Rev. 47 (2011).
https://www.soc.cas.cz/publikace/stari-novi-radikalni-politicky-aktivismus-vceske-republice-ocima-teorie-socialnich-hnuti
(Accessed 15 December 2018).
[66] O. Císař, J. Navrátil, Promoting competition or cooperation? The impact of EU
funding on Czech advocacy organizations, Democratization 22 (2015) 536–559,
https://doi.org/10.1080/13510347.2013.869742.
[67] J. Mazák, T. Diviák, Transactional activism without transactions: network perspective
on anti-corruption activism in the Czech Republic, Soc. Mov. Stud. 17
(2018) 203–218, https://doi.org/10.1080/14742837.2017.1404904.
[68] F.W. Geels, Regime resistance against low-carbon transitions: introducing politics
and power into the multi-level perspective, Theory Cult. Soc. (2014), https://doi.
org/10.1177/0263276414531627.
[69] U. Beck, Risk Society: Towards a New Modernity, Sage Publications, 1992.
https://www.amazon.com/Risk-Society-Modernity-Published-association/dp/
0803983468 (Accessed 15 December 2018).
[70] C. Boswell, Knowledge, legitimation and the politics of risk: the functions of research
in public debates on migration, Polit. Stud. 57 (2009) 165–186, https://doi.
org/10.1111/j.1467-9248.2008.00729.x.
[71] F. Fischer, Technocracy and the Politics of Expertise, Sage Publications, 1990.
[72] P. Ocelík, J. Osička, V. Zapletalová, F. Černoch, B. Dančák, Local opposition and
acceptance of a deep geological repository of radioactive waste in the Czech
Republic: a frame analysis, Energy Policy 105 (2017), https://doi.org/10.1016/j.
enpol.2017.03.025.
[73] P. Leifeld, Reconceptualizing major policy change in the advocacy coalition
framework: a discourse network analysis of German pension politics, Policy Stud.
J. 41 (2013) 169–198, https://doi.org/10.1111/psj.12007.
[74] A. Lis, P. Stankiewicz, Framing shale gas for policy-making in Poland, J. Environ.
Policy Plan. 19 (2017) 53–71, https://doi.org/10.1080/1523908X.2016.1143355.
[75] A. Lis, A.K. Stasik, Hybrid forums, knowledge deficits and the multiple uncertainties
of resource extraction: negotiating the local governance of shale gas in
Poland, Energy Res. Soc. Sci. 28 (2017) 29–36, https://doi.org/10.1016/J.ERSS.
2017.04.003.
[76] H.C. Jenkins-Smith, C.L. Silva, K. Gupta, J.T. Ripberger, Belief system continuity
and change in policy advocacy coalitions: using cultural theory to specify belief
systems, coalitions, and sources of change, Policy Stud. J. 42 (2014) 484–508,
https://doi.org/10.1111/psj.12071.
[77] J. Gerring, What is a case study and what is it good for? Am. Pol. Sci. Rev. (2004).
https://www.jstor.org/stable/pdf/4145316.pdf?refreqid=excelsior
%3Ac3aea8bc4241197cda89133c004eb045 (Accessed 3 April 2019).
[78] P.V. Marsden, Network data and measurement, Annu. Rev. Sociol. 16 (1990)
435–463, https://doi.org/10.1146/annurev.so.16.080190.002251.
[79] T. Vlček, M. Jirušek, Key factors that drive the Czech Republic coal industry, Coal
Int. 263 (2015) 28–37. http://www.eru.cz/cs/ (Accessed 15 December 2018).
[80] F. Černoch, L. Lehotský, P. Ocelík, J. Osička, Ž. Vencourová, Anti-fossil frames:
examining narratives of the opposition to brown coal mining in the Czech
Republic, Energy Res. Soc. Sci. 54 (2019) 140–149, https://doi.org/10.1016/J.
ERSS.2019.04.011.
[81] L. Lehotský, F. Černoch, J. Osička, P. Ocelík, When climate change is missing:
Media discourse on coal mining in the Czech Republic, Energy Policy 129 (2019)
774–786, https://doi.org/10.1016/j.enpol.2019.02.065.
[82] T. Vlček, F. Černoch, The Energy Sector and Energy Policy of the Czech Republic,
Masarykova Univerzita, Brno, 2013, https://doi.org/10.5817/CZ.MUNI.M210-
6523-2013.
[83] B.D. Zumbo, A.M. Gadermann, C. Zeisser, Ordinal versions of coefficients alpha
and theta for likert rating scales, J. Mod. Appl. Stat. Methods 6 (2007) 21–29,
https://doi.org/10.22237/jmasm/1177992180.
[84] P.J. Rousseeuw, Silhouettes: a graphical aid to the interpretation and validation of
cluster analysis, J. Comput. Appl. Math. 20 (1987) 53–65, https://doi.org/10.
1016/0377-0427(87)90125-7.
[85] S. Borgatti, M. Everett, L. Freeman, UCINET 6 For Windows: Software for Social
Network Analysis, (2002).
[86] H. Levene, Robust Tests for Equality of Variances, (1960).
[87] A. Langle-Flores, P. Ocelík, O. Pérez-Maqueo, The role of social networks in the
sustainability transformation of Cabo Pulmo: a multiplex perspective, J. Coast.
Res. (2017), https://doi.org/10.2112/SI77-014.1 2017-Spring.
[88] H.C. White, S.A. Boorman, R.L. Breiger, Social structure from multiple networks. I.
Blockmodels of roles and positions, Am. J. Sociol. (1976). https://www.bebr.ufl.
edu/sites/default/files/SocialStructureFromMultipleNetworks.pdf (Accessed 15
December 2018).
[89] R.A. Hanneman, M. Riddle, Introduction to Social Network Methods, (2005).
http://faculty.ucr.edu/∼hanneman/nettext/ (Accessed 15 December 2018).
[90] D. Krackardt, QAP partialling as a test of spuriousness, Soc. Netw. 9 (1987)
171–186, https://doi.org/10.1016/0378-8733(87)90012-8.
[91] R Core Team, R: A Language and Environment for Statistical Computing, (2014)
http://www.r-project.org/.
[92] SurveyMonkey, (2018). www.surveymonkey.com.
[93] V. Hloušek, L. Kopeček, Česká republika - konsensuální nebo majoritní model
demokracie? Politol. Rev. 8 (2002) 3–21. https://is.muni.cz/publication/404403?
lang=cs (Accessed 15 December 2018).
[94] A. Lijphart, Patterns of Democracy: Government Forms and Performance in ThirtySix
Countries, Yale University Press, 1999.
[95] D. Ost, Illusory corporatism in Eastern Europe: neoliberal tripartism and postcommunist
class identities, Polit. Soc. 28 (2000) 503–530, https://doi.org/10.
1177/0032329200028004004.
[96] A. Pinková, Employer organisations and business GROUPS in the Czech Republic,
Polit. Cent. Eur. 11 (2015) 75–90, https://doi.org/10.1515/pce-2015-0002.
[97] J. Osička, F. Černoch, Anatomy of a black sheep: the roots of the Czech Republic's
pro-nuclear energy policy, Energy Res. Soc. Sci. 27 (2017) 9–13, https://doi.org/
10.1016/J.ERSS.2017.02.006.
[98] T.E. Shriver, A.E. Adams, R.L. Einwohner, Motherhood and opportunities for activism
before and after the Czech Velvet revolution, Mobilization 18 (2013)
267–288. http://mobilizationjournal.org/doi/pdf/10.17813/maiq.18.3.
t272388n60470456 (Accessed 15 December 2018).
[99] P. Vidomus, Oteplí se a bude líp: Česká klimaskepse v čase globálních rizik - Petr
Vidomus | KOSMAS.cz - vaše internetové knihkupectví, Kosmas (2018). https://
www.kosmas.cz/knihy/240774/otepli-se-a-bude-lip-ceska-klimaskepse-v-caseglobalnich-rizik/
(Accessed 15 December 2018).
[100] CVVM, Veřejnost o Spotřebě Elektrické Energie, o Obnovitelných Zdrojích a
Šetrnosti Výroby k Životnímu Prostředí – Květen 2017, (2017). https://cvvm.soc.
cas.cz/media/com_form2content/documents/c2/a4406/f9/oe170904.pdf
(Accessed 15 December 2018).
[101] IEA, Energy Policies of IEA Countries: Czech Republic 2016 Review, (2016).
https://webstore.iea.org/energy-policies-of-iea-countries-czech-republic-2016review
(Accessed 15 December 2018).
[102] IEA, Key World Energy Statistics, (2017). https://www.iea.org/publications/
freepublications/publication/KeyWorld2017.pdf (Accessed 15 December 2018).
[103] WEC, Coal in Czech Republic, (2016). https://www.worldenergy.org/data/
resources/country/czech-republic/coal/ (Accessed 15 December 2018).
[104] M. Sivek, T. Vlček, P. Kavina, J. Jirásek, Lifting lignite mining limits – correction
of the Czech Republic energy policy, Energy Sources Part B 12 (2017) 519–525,
P. Ocelík, et al. Energy Research & Social Science 58 (2019) 101283
12
https://doi.org/10.1080/15567249.2016.1219789.
[105] B. Frantál, Living on coal: mined-out identity, community displacement and
forming of anti-coal resistance in the Most region, Czech Republic, Resour. Policy
49 (2016) 385–393, https://doi.org/10.1016/j.resourpol.2016.07.011.
[106] O. Císař, K. Vráblíková, Transnational activism of social movement organizations:
the effect of European Union funding on local groups in the Czech Republic, Eur.
Union Polit. 14 (2013) 140–160, https://doi.org/10.1177/1465116512456311.
[107] A. Fagan, Environment and Democracy in the Czech Republic: The Environmental
Movement in the Transition Process, first ed., Edward Elgar, Cheltenham, 2004.
[108] Ministry of Industry and Trade, MPO předložilo studie k Problematice Těžebních
Limitů | MPO, 2015, (2015). https://www.mpo.cz/dokument161557.html
(Accessed 16 December 2018).
[109] Ministry of Industry and Trade, Vláda Rozhodla o Zrušení Limitů Těžby na Dole
Bílina | MPO, (2015). https://www.mpo.cz/dokument165352.html (Accessed 16
December 2018).
[110] Ministry of Industry and Trade, State Energy Policy, (2015). https://www.mpo.cz/
dokument161030.html (Accessed 16 December 2018).
[111] A. Cherp, V. Vinichenko, J. Jewell, E. Brutschin, B. Sovacool, Integrating technoeconomic,
socio-technical and political perspectives on national energy transitions:
a meta-theoretical framework, Energy Res. Soc. Sci. 37 (2018) 175–190,
https://doi.org/10.1016/j.erss.2017.09.015.
[112] P.A. Sabatier, S. Hunter, S. Mclaughlin, The Devil Shift: Perceptions and
Misperceptions of Opponents, (1987). https://about.jstor.org/terms (Accessed 15
December 2018).
[113] P.A. Sabatier, M.A. Zafonte, Policy knowledge: advocacy organizations, in:
N.J. Smelser, B. Baltes (Eds.), Int. Encycl. Soc. Behav. Sci. Elsevier, 2001. https://
philpapers.org/rec/SABPKA (Accessed 15 December 2018).
[114] J. Dyduch, Europeanization of the Energy Policy Within the European Union's
System of Governance, (2015). https://papers.ssrn.com/sol3/papers.cfm?
abstract_id=2633218 (Accessed 16 December 2018).
[115] K. Tews, Europeanization of energy and climate policy: the struggle between
competing ideas of coordinating energy transitions, J. Environ. Dev. 24 (2011)
267–291, https://doi.org/10.1177/1070496515591578.
[116] Ministry of Trade and Industry, Aktualizace “Surovinové Politiky České Republiky
v Oblasti Nerostných Surovin a Jejich Zdrojů” - 2016 | MPO, (2016). https://www.
mpo.cz/cz/stavebnictvi-a-suroviny/surovinova-politika/statni-surovinova-
politika-nerostne-suroviny-v-cr/aktualizace-_surovinove-politiky-ceske-republikyv-oblasti-nerostnych-surovin-a-jejich-zdroju_-2016-223504/
(Accessed 16
December 2018).
[117] Ministry of Industry and Trade, Zpráva o Plnění Nástrojů Státní Energetické
Koncepce ČR do Roku 2017, (2018). https://www.mpo.cz/assets/cz/energetika/
statni-energeticka-politika/2018/10/Zprava-o-plneni-nastroju-SEK-do-roku-2017.
pdf (Accessed 16 December 2018).
[118] G. Decree, Usneseni vlády České republiky ze dne 19. října 2015 č. 827, (2015) 2
https://apps.odok.cz/attachment/-/down/IHOAA3RKBT3O.
[119] E. Schlager, Policy making and collective action: Defining coalitions within the
advocacy coalition framework, Policy Sci. 28 (1995) 243–270, https://doi.org/10.
1007/BF01000289.
[120] A. Stritch, The advocacy coalition framework and nascent subsystems: trade union
disclosure policy in Canada, Policy Stud. J. 43 (2015) 437–455, https://doi.org/
10.1111/psj.12112.
[121] P. Cairney, T. Heikkila, A comparison of theories of the policy process, in:
P.A. Sabatier, C.M. Weible (Eds.), Theor. Policy Process, 2014, pp. 363–389.
[122] W. Voorberg, V. Bekkers, K. Timeus, P. Tonurist, L. Tummers, Changing public
service delivery: learning in co-creation, Policy Soc. 36 (2017) 178–194, https://
doi.org/10.1080/14494035.2017.1323711.
[123] L.M. Poloni-Staudinger, Are consensus democracies more environmentally effective?
Env. Polit. 17 (2008) 410–430, https://doi.org/10.1080/
09644010802055634.
[124] M. Jänicke, Ecological modernisation: new perspectives, J. Clean. Prod. 16 (2008)
557–565, https://doi.org/10.1016/J.JCLEPRO.2007.02.011.
[125] G. Marston, Tampering with the evidence : a critical appraisal of evidence-based
policy-making, Aust. Rev. Public Aff. 3 (2008) 143–163.
[126] D.C. Christopoulos, K. Ingold, Exceptional or just well connected? Political entrepreneurs
and brokers in policy making, Eur. Polit. Sci. Rev. (2014), http://dx.
doi.org/10.1017/S1755773914000277.
[127] M. Howlett, I. Mukherjee, J. Koppenjan, Policy learning and policy networks in
theory and practice: the role of policy brokers in the Indonesian biodiesel policy
network, Policy Soc. 36 (2017) 233–250, https://doi.org/10.1080/14494035.
2017.1321230.
[128] J. Habermas, Communication and the Evolution of Society, Beacon Press, 1979.
[129] M. Kowarsch, J. Garard, P. Riousset, D. Lenzi, M.J. Dorsch, B. Knopf, J.-A. Harrs,
O. Edenhofer, Scientific assessments to facilitate deliberative policy learning,
Palgrave Commun. 2 (2016) 16092, https://doi.org/10.1057/palcomms.2016.92.
[130] B.K. Sovacool, J. Axsen, S. Sorrell, Promoting novelty, rigor, and style in energy
social science: towards codes of practice for appropriate methods and research
design, Energy Res. Soc. Sci. 45 (2018) 12–42, https://doi.org/10.1016/J.ERSS.
2018.07.007.
[131] J. Goodman, J.P. Marshall, Problems of methodology and method in climate and
energy research: socialising climate change? Energy Res. Soc. Sci. 45 (2018) 1–11,
https://doi.org/10.1016/j.erss.2018.08.010.
[132] G. Robins, Social networks, exponential random graph (p ∗) models for, in:
D. Lusher, J. Koskinen, G. Robins (Eds.), Comput. Complex. Theory, Tech. Appl.
Cambridge University Press, Cambridge, 2012, pp. 2953–2967, , https://doi.org/
10.1007/978-1-4614-1800-9_182.
[133] M. Howlett, B. Cashore, The dependent variable problem in the study of policy
change: understanding policy change as a methodological problem, J. Comp.
Policy Anal. Res. Pract. 11 (2009) 33–46, https://doi.org/10.1080/
13876980802648144.
[134] M. Angst, A. Widmer, M. Fischer, K. Ingold, Connectors and coordinators in natural
resource governance: insights from Swiss water supply, Ecol. Soc. 23 (2018),
https://doi.org/10.5751/ES-10030-230201 art1.
[135] Y.R. Nooraie, J.E.M. Sale, A. Marin, L.E. Ross, Social network analysis: an example
of fusion between quantitative and qualitative methods, J. Mix. Methods Res. 11
(2018), https://doi.org/10.1177/1558689818804060 155868981880406.
[136] D.P. Dolowitz, D. Marsh, Learning from abroad: the role of policy transfer in
contemporary policy-making, Governance 13 (2000) 5–23, https://doi.org/10.
1111/0952-1895.00121.
[137] C.J. Bennett, M. Howlett, The lessons of learning: Reconciling theories of policy
learning and policy change, Policy Sci. 25 (1992) 275–294, https://doi.org/10.
1007/BF00138786.
[138] K. Thelen, How institutions evolve, in: J. Mahoney, D. Rueschemeyer (Eds.),
Comp. Hist. Anal. Soc. Sci. Cambridge University Press, Cambridge, 2003, pp.
208–240, , https://doi.org/10.1017/CBO9780511803963.007.
[139] F.W. Geels, F. Kern, G. Fuchs, N. Hinderer, G. Kungl, J. Mylan, M. Neukirch,
S. Wassermann, The enactment of socio-technical transition pathways: a reformulated
typology and a comparative multi-level analysis of the German and UK
low-carbon electricity transitions (1990–2014), Res. Policy 45 (2016) 896–913,
https://doi.org/10.1016/J.RESPOL.2016.01.015.
[140] V. Masson-Delmotte, P. Zhai, H.-O. Pörtner, D. Roberts, J. Skea, P.R. Shukla,
A. Pirani, W. Moufouma-Okia, C. Péan, R. Pidcock, S. Connors, J.B.R. Matthews,
Y. Chen, X. Zhou, M.I. Gomis, E. Lonnoy, T. Maycock, M. Tignor, T. Waterfield,
Global Warming of 1.5 °C. An IPCC Special Report, (2018). https://www.ipcc.ch/
site/assets/uploads/sites/2/2018/07/SR15_SPM_High_Res.pdf (Accessed 16
December 2018).
P. Ocelík, et al. Energy Research & Social Science 58 (2019) 101283
13