European Policy Analysis Volume 2, Number 1, Spring 2016 | Page 25

European Policy Analysis

intervene on different decisional levels
(Ingold 2014; Ingold and Fischer 2014;
Ingold, Balsiger, and Hirschi 2010; Prell,
Hubacek, and Reed 2007). To account for
this structural complexity, different policy
scholars have adopted a network approach
(Ingold 2010; Newig and Fritsch 2009):
in order to better reconstruct decisionmaking processes and stakeholder
intervention ( Knoke et al. 1996;
Krackhardt 1990; Knoke 1990; Kriesi
1980),various studies have proven that
social network analysis (SNA) provides
an impressive toolbox for the empirical
analysis of social network structures and
their relevance for opportunities and
behavioral choices of persons integrated
in policymaking. We apply SNA and the
methods used here in a descriptive way as
we are not interested in the investigation
of direct causal links or chains. The aim
of this article is to explore and understand
the potential differences in policy output
production across two decisional levels.

Comparable datasets on policy
networks are rare, not least because
gathering data at various points in
time is highly demanding and resource
consuming. In that sense, the dataset at
our disposal is exceptional. It is comprised
of comparable network data in a policy
domain collected at two different periods.
The first dataset covers the decisionmaking process on policy instruments
within the context of Swiss national
climate politics between 2002 and 2005.
It was gathered through face-to-face
interviews in 2004 and 2005 (see Ingold
2008; 2010). The second dataset on the
preparatory phase of the Swiss position at
the Conferences of the Parties (COP) 16
in Cancun in 2010 stems from a written
survey sent out by post in the beginning
of 2011.

To identify key actors involved in
the respective policymaking processes,
we relied on a combination of positional,
decisional, and reputational approaches.
In line with Knoke et al. (1996), formal
organizations, rather than individuals, are
the unit of analysis. Actors in this research
were therefore defined as organizations
participating in the policymaking
processes and, following the decisional
approach, actors formally implicated in
climate policymaking were identified. The
first list of actors was then complemented
with actors holding an overall strategic
position or being men tioned as very
powerful during initial expert interviews.
This left us with a set of 35 actors for the
national decision-making process and
50 representatives of these organizations
were interviewed. For the preparatory
phase of the COP 16, questionnaires were
sent to 22 actors and the response rate
of this survey was 70% (complete actors
list in appendix). Both surveys were thus
based on questionnaires designed in the
same way and containing batteries of
questions to investigate actors’ relational
profiles and policy preferences.

Based on a list of all actors
participating in the respective decisionmaking process, interviewees were asked
to identify those actors with whom
they collaborated intensely (relational
profiles). Furthermore, actors were
asked to rank the policy options under
discussion in the respective process
(policy preferences). For the national
decision-making process, they ranked the
following policy instruments evaluated
during the preparliamentary phase of
2004: voluntary agreements, CO2 tax,
climate penny, and tradable permits. For
the preparatory Cancun negotiations,
they had to give their opinion on the
25