6. Approach to Policy Mapping and Delphi Methodology

Our research approach is summarised here and described in more detail below.  The approach follows a rationale of conducting a broad range mapping to identify implemented and proposed consumption-oriented policies across the Nordic countries. To allow for this broad scope, previous mappings, reports, and other publications have been utilised. Workshops with scientists, experts and practitioners were held to identify publications on policies. In a first step, a comprehensive mapping was conducted by the Swedish project team, resulting in a longlist compilation of policy options. 

In a second step, the draft mapping was reviewed by a cross Nordic panel representing Denmark, Finland, Iceland, Norway, and Sweden. The panel included experts in the consumption-based emissions and policy field (see Table 3 with project team). The Nordic expert team added examples from Nordic countries or, in some cases, examples beyond the Nordic context to the list. At the same time, they merged similar policies to avoid having too many policies resembling each other on the list. The cross-Nordic panel was also involved in the selection of particularly promising policies for the Policy Delphi. This selection of policies needed to be substantially shorter than the initial list.

Table 3. Project team and Nordic experts

In a third step, a Policy Delphi method (described below) was utilised to gather expert judgments and opinions on a reduced number of selected policies. The Policy Delphi was carried out in two rounds. In a first round, experts ranked policies according to their evaluated potential to reduce emissions and feasibility while substantiating their judgments with comments. In a second round, the experts could see the others’ anonymous rankings and comments and they were able to revise their rankings and reply to comments. For the second round of the Policy Delphi, the number of selected policy measures were further reduced to allow for an in-depth discussion among the experts. Policies were selected based on high rankings by the experts as well as strategic considerations related to political and media agendas, and whether the policy is of national or EU-level competence.

The research approach takes a broad scope, distilling a broad range of policy measures towards the most promising policy options.

The academic literature scoping review used selected keywords (search string "consumption based" OR "consumption-based") AND (carbon OR co2 OR ghg OR "greenhouse gas" OR "air pollution") AND (poli*) AND (Sweden OR Denmark OR Norway OR Finland OR Iceland OR “the Faroe Islands” OR Åland). The search was made on 4 October 2023 in the SCOPUS (n=33) and Web of Science (n=133) databases. Duplicates were eliminated to get a unique number of articles (n=113). The criteria for inclusion of articles was based on (1) an explicit empirical focus on one or multiple Nordic countries and (2) a substantive discussion on one or more (national) policies. Following the criteria, inclusion was based on a screening of titles and abstracts (n=25), and in-text analysis, which resulted in 16 reviewed articles. The articles were primarily read for policy evaluations discussing effectiveness (potentials to reduce emissions, cost-effectiveness, etc.) and feasibility (distributional effects, administrative constraints, distributional effects and policy acceptance).

Several previous policy mappings were analysed alongside the literature review. The Swedish research program Mistra Sustainable Consumption has previously conducted several mappings of policies for the focus areas of vacation travel, food, and home furniture. Svenfelt et al. (2022) previously published parts of these mappings. Moreover, the consumption-oriented policy mapping also draws on earlier reviews and compilations. For instance, Grubb et al. (2020) drew on an extensive mapping of consumption-oriented policy measures compiled within the EU Carbon-CAP project, employing a ranking methodology to assess the effectiveness and feasibility of 33 potential instruments (an approach similar to that used in this report). Ottelin et al. (2019) reviewed more than 100 papers covering consumption-based carbon footprints, searching for suggestions for policies. They identified about fifteen broad national policy recommendations.

In addition to the previous conducted mappings, an extensive number of reports have been scanned for policy proposals. See online appendix for a complete list of sources in the mapping. In Table 4 below, some key sources are listed. Three consecutive workshop sessions have been facilitated with (i) researchers, (ii) societal partners, and (iii) the international scientific advisory board of Mistra Sustainable Consumption. We utilized workshops to get feedback on our project design and to identify possible sources and reports on demand-oriented policy measures within the Nordic context.

The initial mapping of policies conducted by the Swedish project team was, in a second step, reviewed and complemented by a cross-Nordic researcher panel (representing Denmark, Finland, Iceland and Norway). In this step, the list of policy measures was complemented with examples from Nordic countries and, in some cases, examples beyond the Nordic context.

All identified policies have been compiled in a longlist (see Table 6). Similar policy proposals have been merged into generic categories with a common general description. Specific policy proposals, for instance, the exact design and level of a carbon tax, have not been included as a specific policy measure but as parts of a policy measure category.

Table 4. Key sources

In a third step, the longlist was again reviewed by the Nordic expert panel. The initial ranking by the cross-Nordic research panel was used to select a limited sample from the longlist (Table 6) of policy instruments for a Policy Delphi.

The Policy Delphi method is described by Löe and colleagues (2016) as a tool "(...) aimed to generate ideas, and to uncover and evaluate policy alternatives, through structured, critical collective debate among anonymous panellists.” The Delphi method was originally developed by the RAND Corporation and the US Air Force to capture expert opinions via iterations of questionnaires (Linstone, 1999). Several Delphi variations exist today. These include Classical Delphi, Decision Delphi, Ranking-Type Delphi, and Policy Delphi (Paré et al., 2013; Schmidt, 1997). Common principles for Delphi techniques include anonymous panels of experts engaging in multi-round structured dialogue, through which results from the initial round are synthesised before they are returned to the experts in subsequent rounds of discussion (Loë et al., 2016). Although early Delphis were used to find a consensus amongst experts, contemporary Policy Delphis are often used to identify expert perceptions and divergent opinions. The assumption is that expert views on policy measures can add relevant knowledge, although this approach cannot precisely estimate if a policy measure is, for instance, effective or feasible.

In this report, Nordic experts (Denmark, Finland, Iceland, Norway and Sweden) rank policy measures judged according to their potential to reduce greenhouse gas emissions and other air pollutants as well as according to their feasibility. The participants were given the following descriptions:

Potential to reduce emissions: if designed and implemented in a sufficiently strong form, is the potential low or high regarding reduction of emissions within the consumption categories where it would be implemented?

Feasibility: refers to a number of different aspects such as:

The experts are also asked to give qualitative comments on broader aspects (administrative feasibility, distributional effect, societal resilience, other barriers to implementation, enabling measures). In a second round, the same participating experts can see the anonymous assessments of the other participants and decide if they want to update their assessments. The Policy Delphi method effectively facilitates structured expert dialogue and reveals and evaluates policy options in complex and controversial areas (Löe et al., 2016).

The Policy Delphi assembled 23 participants in two rounds (the second round had one non-respondent). Responses for the first round was collected between the 6^th of January and 27^th of April 2024. Responses for the second round were collected between the 6^th of May and 13^th of June 2024.

A selection of a smaller number of policies for the second round was made based on high-rankings in both categories but also for strategic reasons where policies covering different categories of measures were chosen. Consideration was also given to the comments from round one, where the participants commented that some policies were very similar.

Table 5. Experts in the Policy Delphi