Modelling social apsects
Following the openmod Workshop in Berlin in January 2020, and especially the session “E.4 Let’s connect: how to model social aspects?“, it was decided a thread on the forum regarding the modelling of social aspects would be of great interest. You can find elements of context hereunder, based on collaborative notes of what was discussed during this session. Feel free to add any comment or contribution, this is thought as an item opening.
Presentation by Diana Süsser (IASS Potsdam)
Presentation as an introduction to the session, (see nextcloud: E4_How to model social aspects_download-version.pdf)
While Social sciences are crucial in the research around climate change mitigation, they get way less fundings than technical sciences (see sources).
QTDIAN is a model library for social scenarios and qualitative storylines. It is based on:
- Socio-technical transition literature (multi-level, acceptance)
- Collection and ranking of users’ needs (interviews, workshops)
And uses the agent based model ATOM from the SENTINEL project.
Topics of general discussion
Processes of adoption of new technologies are not well understood. Various elements can greatly influence decision-making and acceptance:
- “Windows of opportunity” and “change agents” (Zugpferde) are very important for decision making and acceptance ; strong narratives can be helpful for technology shifts in windows of opportunities (e. g. when different infrastructures like grids reach their lifetime simultaneously).
- Constellation of strong leaders and existence of opponents have essential influence on adoptions rates.
- A critical mass of people is often necessary for a perception change.
- Lifecycle Assessment (LCA) models deal with resource impacts; they in turn can have influence on acceptance of technologies (e.g. critical resources in the production of e-mobility batteries); misinformation of the public often plays a role for acceptance. Other resource availabilities are also important, like scarcity of craftsman
- Visibility of technologies (e. g. electric car versus building retrofit or heating system) makes a difference in adoption as social comparison is an important factor of end-user demand
Different models with social aspects like acceptance have been developed, but they all lack a reliable database from what we know so far, and therefore are of small use (without reliable data social models can be considered as “dummy” models). That´s one of the central reason why they did not find their way into mainstream so far. Good databases are required.
There can actually be different ways to include social aspects in (transformative) research:
- A) including social aspects in the modelling, as discussed hereinabove
- B) including models into society (societal processes)
- Civil society should be involved in defining the scenarios; they should also be involved in the analysis. The output of models has to be prepared to make it manageable and that is already the first step of analyses. Iteration cycle of model results - reflection of results through civil society - new modeling/results would be desirable. Therefore, models must be easy to understand, transparent in their assumptions (what is based on empirical findings and what are just assumptions) and simple to use.
- There could be a local use of energy models to make decisions with society and empower it. Once again, some guidance for the users as well as making assumptions and wording very clear in models is essential in order to make sure everyone mean the same thing, for being transparent and reproducible.
- Something to keep in mind: on local level discussions the “big picture” is often missing (population explosion, other developments, …)
- also see “STEmp-Tool” & “en-roads” in the sources below, as well as the openmod forum thread An open energy system analysis community
- How to model effects such as the ones of misinformation; the citizen empowerment leading to sufficiency?
- Are linear optimization models suitable for modeling social feedbacks and interactions? Are non-linear models better options and which model types are out there? Logics of interaction and target conflicts may better be seen in simulation models than in optimization models; easier to discuss with decision-makers (politicians, etc.) ; relevant for transformative research. “System Dynamics” models could be of use.
- How do you obtain reliable social databases? Surveys?
- Problem of calibration for social models is critical, how to deal with it?
- IPCC Socioeconomic Pathways include many social parameters; they are not that much used in energy models as the climate modeling and energy modeling communities (Integrated Assessment Models, IAMs) share only few overlap so far - how could this be changed?
First ideas, food for thoughts:
- Influence could be modelled via scenario-/storyline-based social diffusion curves with different phases/levels of acceptance.
- Different user groups / social dynamics / drivers / barriers should be reflected
- Critical mass for acceptance change: modeling with thresholds?
- Useful theories/concepts for social modeling:
- Behavioural Economics
- Social network theories (concept of trust)
- Theory of bounded rationality
- Theory of planned behaviour
- Hybrid agent-based models for decision-making
- Often misuse of models in policy; we need to reflect what are interpreted as results of a model
Sources, collection of existing research / models
- On social sciences : socio-political-technical features are excluded in current models (Koppelaar et al., 2016); more work need to integrate social aspects in quantitative modelling (Pfenninger et al., 2014)
- ZNES (Europa-Universität) Flensburg (Marion Wingenbach, Öko-Institut e.V. Freiburg): 3 years project, collecting data; developing acceptance factors for wind-energy using oemof framework … (PhD thesis). The developed “burden-level” in the ZNES-project could be developed further by including more factors based on a broader analysis of acceptance issues in German society (better understanding of regional distribution of acceptance: role of democratic and financial participation, influence of local natural landscape) 
- en-roads: getting people involved in the global transition 
- After the projekt the developed burden level for wind energy and delay times for grid expansion projects have been integrated in oeko-institute model “powerflex”  ; 
- Analysis how a distribution of wind energy on the basis of an equal burden level influence costs of energy transition in PhD thesis 
- Reiner Lemoine Institute: StEmp-Tools: Stakeholder Empowerment tools are made to discuss transformation on a local level 
- Projekt name: “VerNetzen”: Sozial-ökologische und technisch-ökonomische Modellierung von Entwicklungspfaden der Energiewende 
- Diana Süsser developed an ABM on wind energy diffusion (including influence of strong leaders in her PhD thesis)
- WS on the human in the energy model in 2018: “Was ist die Energiewende ohne uns? Die Menschen im Energiesystemmodell” 
- NIWA: New Zealand Project with Maori about hydrology / indigenous science on acceptance of water extraction from rivers
- Eco-SESA: Cross Discplinary Program on local renewable energy production & consumption and the involvement of citizen 
- OMEGAlpes : Open Energy Model Generation Tool for MILP Optimization . Social scientists included in the development Lou Morriet PhD thesis 
- Concept of Social energy nodes Gilles Debizet 
- IÖW: model relation between RE-job creation, wealth distribution and acceptance, included through market acceptance factor
- Anthropology of energy, in rural areas of Germany, Laure Dobigny (in french)  ; (Another article in english) 
- German Sociological society and German Physical Society has a modelling team with social focus; organize a yearly conference on social modeling topics
- Alex Pentland: social physics 
- Cambridge Analytics and Facebook conducted empirical science on acceptance and delivered a reliable empirical data base
- Atom model (part of SENTINEL project) is an agent based model on decision making regarding energy technologies in households
- MIDAS - collection of knowledge including models which are used by European Commission and European Parliament; restricted use for other parties (to check)  --> could be interesting to look for open access models in the social context there
- Center for Research in Social Simulation (CRESS) 
- PAC-scenarios of RGI-network could be a good starting point, as they should be compatible with 1.5° climate target