Multi-country evidence on societal factors to include in energy transition modeling

Multi-country evidence on societal factors to include in energy transition modeling | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Multi-country evidence on societal factors to include in energy transition modeling Vivien Fisch-Romito, Marc Jaxa-Rozen, Xin Wen, Evelina Trutnevyte This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-4312891/v3 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 21 Feb, 2025 Read the published version in Nature Energy → Version 3 posted You are reading this latest preprint version Show more versions Abstract Integrated assessment and energy system models are challenged to account for societal transformation dynamics to produce feasible low-carbon pathways. Yet, empirical evidence is lacking on which factors should be incorporated, how and to what extent this would improve the quality and relevance of modeled pathways. Here, we include six societal factors related to (i) infrastructure dynamics, (ii) actors and decision making and (iii) societal and institutional context into an open-source simulation model of the national power system transition. We apply this model for 31 European countries and, using hindcasting (1990–2019), quantify which societal factors improved the modeled pathways. We find that, if well-chosen and in most cases, incorporating societal factors can improve the hindcasting performance by up to 24% in terms of modelled installed capacity of individual technologies, but there are also situations where hindcasting performance can become worse. The combinations of most relevant societal factors differ among countries and model outputs, but infrastructure lock - in, public acceptance and investment risks contribute more strongly and frequently to model performance improvement. Our study hence paves the road to evidence-based choice of societal factors to be included in energy transition modeling in a systematic and transparent way. Scientific community and society/Energy and society/Energy supply and demand Earth and environmental sciences/Environmental social sciences/Socioeconomic scenarios Earth and environmental sciences/Environmental social sciences/Climate-change mitigation Scientific community and society/Energy and society/Energy economics societal transformation dynamics energy systems modelling integrated assessment modelling socio-technical transitions hindcasting Figures Figure 1 Figure 2 Figure 3 1 Introduction Integrated assessment models of climate change and energy system models are widely used to quantify low-carbon pathways at a global and national scales. For the power sector, which needs to decarbonize earlier than other sectors [ 1 ], the models represent electricity supply, end-use technologies, service demands and their interlinked dynamics to assess the sector’s transformation to enable broader carbon neutrality. In practice, this transformation depends on the co-evolution of technical, social, economic, institutional or political factors which may induce lock-ins or tipping point dynamics and change feasibility space for the emergence of new technologies [ 2 , 3 ]. Modelers are now increasingly requested to account for societal transformation dynamics in their modeled pathways to inform policy makers about intervention designs, their feasibility and impact [ 4 – 8 ]. Social scientific knowledge on energy and climate mitigation has so far evolved in parallel to modeling and analysis of pathways [9] with rare links done either ex-ante with informed storylines [ 10 ] or ex-post by evaluating pathways outputs on different societal indicators [ 11 ]. The few articles that have attempted to represent societal factors of power sector transitions in technology-rich models have focused on one factor at a time, such as cost of capital [ 12 ], actors behaviour [13, 14], or social acceptance [15]. As a result, critiques of models have emerged in the meantime about their poor ability to produce solutions-oriented outcomes or to inform about feasibility [9, 16, 17]. Analyses that endogenize multiple societal factors into models could contribute to responding to these critiques, but such models are at early stage of development [18–20]. Most of all, it is unclear whether some societal factors should be given priority over others for inclusion in models and how these factors should be endogenized (see for instance recent methodology description for the integration of capital costs [ 21 ]). Critiques have pointed out the limited performance of power sector simulation models in capturing real-word transitions [22, 23] and it is also unclear to what extent the incorporation of societal factors would improve it or not. This relates to model evaluation which is an emerging topic in the energy and integrated assessment modelling community to improve robustness and usefulness of models for policy makers [ 24 ]. Hindcasting — assessing model ability to reproduce past dynamics — is one relevant evaluation method to assess the appropriateness and credibility of a model structure [ 24 ], but few hindcasting exercises have been done so far [ 25 – 27 ] and none has considered modeling societal factors. Here, we provide first-of-the-kind empirical evidence, using an open-source techno-economic simulation model of the power system transition, where we incorporate, one by one and as a combination, six societal factors related to (i) infrastructure dynamics, (ii) actors and decisions making and (iii) social and institutional context (Table 1 ). We set up this model for 31 European countries (EU27, Switzerland, Iceland, Norway, and the UK) to allow generalizable insights and conduct a hindcasting evaluation of power capacity expansion, renewable generation and CO 2 emissions from 1990 to 2019. This approach allows us to conclude on which societal factors and how should be ideally incorporated in future modeling. Table 1 Societal factors included in the model one by one and as combinations. Methods section provides more details on model implementation of these factors. Aspects Societal factors included Approach how this factor is represented in the model Infrastructure dynamics Lock-in (L) Power capacity lifespan is extended if this capacity is economically profitable to keep using in the future at the same utilization rate. Fast transition (T) Existing capacity is retired before the end of the lifespan if it is economically profitable to build and use alternative technologies at the same utilization rate. Actors and decision making Actors heterogeneity (A) Lower cost sensitivity is assumed for investment choices [20]. Investment risks (I) Different weighted average costs of capital (WACC) are used per countries and technologies [ 28 ]. Social and institutional context Public acceptance (P) Incumbent technologies cannot be built in the case of low public acceptance [29]. Governance (G) Electricity market liberalization facilitates renewable introduction [30], while state-owned utilities show higher tendency to invest in renewable [ 31 ]. 2 Results 2.1 Societal factors in the model tend to improve the hindcasting performance for capacity expansion We find that incorporating societal factors in the model one by one or as combinations mostly improves the hindcasting performance in terms of installed capacity dynamics for most of the 31 European countries studied (Fig. 1 ). Only in the cases of Iceland and Latvia, the model version with the best hindcasting performance (i.e. lowest error) is the techno-economic one without societal factors incorporated. For the rest, the model version with the best hindcasting performance includes at least one societal factor. We obtain high heterogeneity between countries in terms of hindcasting performance gain compared to the techno-economic model. Minor gain in hindcasting performance is achieved for Lithuania, Norway, Slovenia and Switzerland (less than 1% of symmetric mean absolute percentage error (SMAPE)) and low gain (less than 5%) for another nine countries (Belgium, Bulgaria, Croatia, Cyprus, Greece, Malta, Poland, Portugal, Romania). Countries with minor gains are characterized by low cumulative installed capacities in 1990, below 10 GW, and a limited diversity of technologies, or by a high proportion of hydroelectric power plants, which induces inertia in the evolution of the system due to long lifetimes. Values for Austria, Denmark, France, Germany, Hungary, Ireland, Italy, Luxembourg, Netherlands and Slovakia lie in between 5% and 15%. We obtain the highest hindcasting performance gain of more than 15% of SMAPE for six countries (Czech Republic, Estonia, Finland, Spain, Sweden and United Kingdom). This is mainly because, contrary to the techno-economic model, the inclusion of some societal factors in the model allows to better capture the increase in installed capacities of decentralized renewable technologies (e.g. biomass, offshore and onshore wind and solar PV) observed in these countries from 2010. Even if the inclusion of one or several societal factors can improve the hindcasting performance for most of the countries, there are many cases when including some specific factors or their combinations, the hindcasting performance gets worse than the performance of the techno-economic model (Fig. 1 .). Societal factors hence need to be included in an evidence-based way to avoid worsening the model’s performance and hindcasting is a relevant approach to do so. We also observe a large spread in terms of absolute hindcasting performance (Fig. 1 ). On the one hand, for Belgium, Lithuania, Luxembourg, Malta, Netherlands, Portugal and Switzerland, the model version with the worst hindcasting performance has a better performance than the average of the whole simulation set, all countries considered. Except for Luxembourg and Netherlands, low performance gains are obtained for these countries as well, pointing to a tiebreak between additional modeling efforts and actual gain. On the other hand, for Bulgaria, Norway and Sweden, the model version with the best hindcasting performance gives more errors than the average value of the whole hindcasting simulations set. For this second group of countries, this suggests a potential for improving the model structure and the way that societal factors are incorporated in order to improve the model performance to capture capacity dynamics. 2.2 The contribution of the different societal factors to model hindcasting performance differs between countries and factors The selection of societal factors leading to the best hindcasting performance differs between countries (Fig. 1 ). First, there is a difference in terms of number of societal factors incorporated, ranging from one for Croatia, Cyprus, Lithuania, Malta, Norway, Romania and Switzerland to four for Austria, Denmark, Germany, Italy and Luxembourg. There is no country where the integration of more than four factors leads to an additional gain in performance. This suggests that increasing the complexity of the model does not necessarily increase its accuracy, supporting previous findings on complexity of technology representation in models [ 32 ]. Second, the best combinations differ in terms of societal factors incorporated. Each societal factor is present for at least four countries in the model version giving the best hindcasting performance which suggests that all societal factors analyzed are potentially relevant to incorporate in integrated assessment and energy systems models. Lock-in , Public acceptance and Investment risks are the societal factors most present in the model versions giving the lowest errors, with respectively 14, 14 and 12 occurrences over 31 countries (Fig. 1 ). Model hindcasting performance is more sensitive to the incorporation of some societal factors in the model than others, both positively and negatively. Actors’ heterogeneity has the highest influence in most countries (Fig. 2 b) which is expected since this factor accounts for cost sensitivity which in turn drives the shares of new investments for all technologies. The second most influencing factor overall is the Fast transition factor which represents more than 20% of the variance for ten countries (Fig. 2 b) but tends to decrease the hindcasting performance for them (Fig. 2 a). Public acceptance is the third most influential factor representing more than 10% of the hindcasting performance variance in six countries (Austria, Germany, Netherlands, Portugal, Romania, Switzerland), but with a negative effect on model hindcasting performance for Switzerland and Romania (Fig. 2 a and 2 b). This highlights the importance to be cautious when introducing or omitting these three factors because they are not always beneficial. Interactions of certain specific societal factors also have a high influence on model hindcasting performance for Italy and Netherlands where they represent more than 25% of the variance (Fig. 2 b). In the case of Italy, this is mainly due to a synergistic effect of Fast transition and Actors heterogeneity that gives an overall hindcasting gain greater than the sum of the effects of each factor incorporated individually (Supplementary Fig. S1 ). Conversely, for the Netherlands, we obtain antagonistic effect with the combination of Public acceptance and Governance and with Actors heterogeneity and Governance , which give lower performance than the sum of the effects of each factor incorporated individually (Supplementary Fig. S1 ). Remarkably, the integration in the model of any societal factor, one by one or in combination, has limited influence on the model hindcasting performance for Lithuania and Portugal (Fig. 2 a). Lithuania is characterized by an oversized supply system compared to its electricity demand with low technology diversity, creating a locked dynamics in all modeled pathways. For Portugal, the much lower costs for coal and wind power compared with other technologies drive the capacity dynamics in the model. 2.3 Societal factors in the model induce a hindcasting performance trade-off among model outputs For most countries the combination of societal factors giving the lowest deviation from historical data for capacity dynamics is not the same as giving the lowest deviation for CO 2 emissions and the share of renewable electricity production (Fig. 3 ). We hence obtain a trade-off in hindcasting performance between model outputs but to a different extent between countries. For a first group of sixteen countries composed of Austria, Bulgaria, Croatia, Cyprus, Czech Republic, Estonia, France, Germany, Hungary, Italia, Malta, Norway, Poland, Portugal, Spain and United Kingdom, the model version giving the best hindcasting performance for installed capacity also improves the model performance for the two other outputs compared to the techno-economic model. For the second group of twelve countries composed of Belgium, Denmark, Finland, Greece, Ireland, Luxembourg, Netherlands, Romania Slovakia, Slovenia, Sweden and Switzerland, the model version giving the best hindcasting performance for installed capacity gives more errors than the technoeconomic model for at least one of the other outputs. This suggests the need to use a combination of societal factors, like a country-specific pareto-optimal set of factors, that altogether would allow to improve the hindcasting performance for the three metrics, and other relevant model metrics too. For Belgium, Denmark, Finland, Ireland, and Slovakia, there are combinations of societal factors that are relevant to improve the performance for the three model outputs. Conversely, this is not the case for Switzerland, Greece, Romania and Slovenia, suggesting the need for these countries to better represent the societal factors or to further modify the original structure of the model. 3 Discussion This study provides new and systematic evidence to further incorporate societal factors into integrated assessment models and energy system models to construct low-carbon transition pathways. By doing a multi-country hindcasting exercise with a simulation model of the power sector, we find that incorporating societal factors related to (i) infrastructure dynamics, (ii) actors and decisions making and (iii) social and institutional context has the potential to improve the model performance to capture the real-world dynamics of installed capacities of individual technologies in the power sector. This finding is valid for most of the European countries studied but the combination of societal factors incorporated to increase the hindcasting performance differs between countries. Actors heterogeneity influences the model outputs the most but Lock-in of infrastructure, Public acceptance and Investment risks are most likely to be associated with a best hindcasting performance. Modellers should hence focus their future research on these four societal factors for integration into models. We show that the gain in model performance induced by the integration of societal factors to capture power system dynamics depends strongly on the national energy context. Low performance gains are obtained for countries with small cumulative installed capacities and with high shares of hydroelectric production in their electricity mix. Conversely, high performance gains are obtained for countries who tend to have high cumulative installed capacities and that have experienced in the last decade rapid growth in decentralized renewable technologies. We also highlight that a performant model structure for installed capacities is not necessarily relevant to capture other output dynamics, such as renewable generation or CO 2 emissions with a potential accuracy trade-off. Going further, these elements suggest that there is no one-size-fits-all approach and that models that represents national energy transitions should be more country-specific in building their structure and should be evaluated on different outputs to choose one or several adequate model structures. Hindcasting exercises are key here to provide evidence for model construction. Our hindcasting study was limited to six societal factors frequently mentioned in the literature as influencing the installed power capacity dynamics. In the future, testing the impact of other societal factors or other model implementations of the same societal factors could be relevant, not only for countries where societal factors decreased the model performance but also to improve the model performance simultaneously on multiple outputs for countries where we obtain a high-performance tradeoff. Second, including the factors influencing electricity demand would be interesting in order to better capture the co-evolution of electricity demand and supply that has led to high overcapacity for some European countries [ 33 ]. Capturing this co-evolution is all the more important at a time when energy security and demand reduction are back in the spotlight because of the conflict in Ukraine. Finally, future work should include societal factors and test the hindcasting performance of the whole energy system models and eventually of integrated assessment models of climate change. Energy modelers are always torn between proposing the most realistic model structure possible for policy relevance and at the same time making simplifications to be able to interpret and communicate the results. This study confirms that increasing model complexity (i.e. incorporating more societal factors) is not equal to increasing model accuracy and performance [ 32 ]. The hindcasting methodology could be replicated during any model building phase [25, 58] in order to obtain a parsimonious model using as few explanatory variables as possible, while still ensure that the model is ’good enough’ for its specific purpose [34, 35]. Moreover, this study highlights that the modellers’ decisions to incorporate or not to incorporate societal factors are impactful for results. The hindcasting methodology allows to disentangle the societal factors that most influence the model performance by themselves or in interaction with other factors. The exercise done here hence paves the way to reduce subjectivity of modeller’s decisions [ 36 ] and to be transparent on the structural uncertainties of the model [ 24 ]. We also provide tools and datasets in this paper and elsewhere [56, 58] to enable informed choice. This study supports that further efforts should be devoted to two aspects on model building methodology. First, introducing societal factors in integrated assessment and energy system models can significantly improve their performance which calls for a stronger cooperation of this community with social sciences to have representations of societal factors more empirically based. Second, integrated assessment and energy system models have been mostly ’diagnosed’ so far through prospective intercomparison exercises to identify their main assumptions, characteristics, and behaviour [ 37 , 38 ]. To go further, we argue this should be completed by common hindcasting exercises, as done in this study, in order to highlight their (in)ability to capture sectoral and national dynamics and adapt the model structure. Reproducing past trends in hindcasting is not the guarantee of a better model performance because energy-economy-social systems do not exhibit structural constancy over time [ 35 , 39 ]. However, a scenario produced from a model that is good enough to capture past trends can bring insights for discussing not only the techno-economic, but also socio-political feasibility of achieving climate targets [40, 41] and highlight how current socio-technical dynamics potentially hinders climate ambition in the short term and what policies are needed [ 42 ]. 4 Methodology In our paper, we set-up an open-source model of the power system transition (Section 4.1 ) and followed a two-step modelling approach. In the first step, we developed a set of hindcasting simulations of national power generation and capacity expansion in 31 European countries (EU27, Switzerland, Iceland, Norway, and the United Kingdom) from 1990 to 2019. Each simulation corresponds to a model version where one or a combination of several societal factors from Table 1 are included (Section 4.2 ). Section 4.3 details the data sources. In the second step, from each scenario we extracted the trends in terms of dynamics of installed capacity, generation and CO 2 emissions and compared them to historical pathways to evaluate the hindcasting performance. Section 4.4 describes the metrics used to evaluate hindcasting performance. 4.1 The STONES model For the purpose of this study, we built an open-source simulation model of the power system, called Socio-Technical Outlook of National Energy System (STONES) ( https://github.com/VivienFischRomito/STONES_model_open ). STONES is a bottom-up, technology-rich, simulation model with a recursive dynamic of one year time step. The model can be considered as a socio-technical energy transition model since it represents technoeconomic detail, actor heterogeneity and transition pathway dynamics [ 43 ]. It includes 14 different electricity production and storage technologies. The model takes as inputs national energy systems data such as costs, resource potentials, and flexibility constraints for each technology. The outputs are the installed capacities of individual technologies and electricity production for every year for all technologies. The STONES model is composed of three distinct modules used consecutively at each time step: the electricity demand module, the dispatch module and the capacity expansion module, all described in detail in the Supplementary Information Section 1 . The electricity demand module builds load duration curves for the current year and the expected ones in the future. It allows to obtain hourly electricity demand values for the observed year to feed the dispatch module, and for five years ahead to feed the capacity expansion module. The model uses as inputs the hourly electricity demand profiles and the hourly load factors of non-dispatchable renewable technologies for the current year. The annual time series are aggregated using clustering into six representative days to reduce calculation time because of computational constraints. The dispatch module determines the system-wide cost-optimal dispatch by minimizing the sum of variable costs for all technologies while satisfying hourly electricity demand constraints over the six representative days obtained from the electricity demand module. This optimization is done satisfying different constraints on electricity production, adequacy and transmission capacities. The capacity dynamic module determines the evolution of installed power capacity to satisfy the projected electricity demand five years later. As main inputs, the model uses technology costs, age of installed capacity and expected load duration curves obtained from the electricity demand module. First, the module estimates the total capacity needed to satisfy the expected peak demand five years later based on the expected capacity retirement over the following years. Then, since only a part of the installed capacity generates power at full load all year, the module divides the load duration curves into load investment segments (i.e. peak load, intermediate load, base load) to make the link with dispatch decisions, following previous approaches [ 44 – 46 ]. The total needed capacity is then split between technologies using a multinomial logit equation, a common approach in energy simulation models [7, 45]. To do so, for each investment segment, the technologies are compared based on their annualized levelized costs. Finally, the capital stock evolves at each time step and for each technology, based on new investments made and retirement of the existing capacity according to the lifespan. 4.2 Representation of societal transformation factors in the model In the STONES model, we integrated six societal transformation factors related to three influencing aspects of energy transition, namely (i) infrastructure dynamics, (ii) actors and decision making and (iii) social and institutional context (Table 1 ). Infrastructure dynamics influence the transition speed of existing capital towards new technologies. On the one hand, power infrastructure has long lifespan and is characterized by lumpiness and a path-dependent evolution [47, 48]. This Lock-in effect factor incentivizes the usage of existing capacity for a longer time. On the other hand, the market environment can incentivize earlier retirement of capacities, inducing a more flexible infrastructure dynamics and allowing a Fast transition towards alternative power technologies. We incorporate these two factors in the model for coal, gas, and nuclear technologies since they have represented an important share of the electricity mix in the last decades and because capacity age data were available. We represent Lock-in in the model by allowing a lifetime extension of ten years. To do so, at the expected lifespan end of incumbent capacity, the model compares the costs of continuing to use the incumbent capacity (i.e. annual operational costs) with the costs of building and using new capacity of alternative technologies (i.e. annual levelized costs). To represent Fast transition , a similar comparison is done in each year in the model. The cost comparisons are done for both societal factors assuming that potential alternative technologies would produce electricity in a similar pattern to avoid, for instance, inconsistent replacement of a base load technology by a peak technology. We use a multinomial logit equation for this comparison to allow partial replacement/extension of the capacities. When incorporating the aspects of actors and decision making, we modified the perceived technology costs and hence influenced the direction of power capacity transition. First, there are various actors involved in the process of capacity investment decision making and some of them allocate a lower weight to technology cost for their decision. We represent Actors heterogeneity in the model by decreasing the cost sensitivity of investment choices [ 14 ]. Second, Investment risks quantified by the cost of capital appear as an important driver of investment choices, especially for technologies with high upfront costs, such as renewable generation [49], and it evolves over time and differ between countries [ 50 ]. Following the methodology of Steffen [ 12 ], we represent risk in the model by replacing the discount rate with the weighted average cost of capital in the levelized cost calculation differentiated by country, technology and observed year. The social and institutional contexts create non-linear change in energy transition by either continuing past trends of investments or, with changes beyond a certain threshold, speeding up the uptake of new technologies. First of all, public acceptance influences the emergence of renewable generation in the energy landscape. Wind energy diffusion has recently slowed down in some countries due to public acceptance issues [51]. Negative public opinion about incumbent technologies can lead to a technology phase-out such as nuclear power in Germany during the 2010s, creating a window of opportunity for the emergence of renewable technologies. To represent this stylized fact, we assumed a technology cannot be built in the model for a given year when more than 60% of the population has a negative perception of it, following Cotterman et al. [29]. For wind power, local acceptance of wind projects is a more important factor than overall public opinion because of potential disamenity costs [15, 52]. Based on data from previous studies [15, 53], we assumed that local acceptance is 35% lower than overall public acceptance for wind power and then used this figure to apply the same approach as for other technologies to represent power technologies rejection. Conversely, widespread positive opinion about renewable generation technologies over the population creates pressure on governments for the adoption of support policies [ 54 ]. To represent the variation in perception of adoption costs, we assumed that wind and solar investment costs are reduced by 20% if more than 80% of the population has a positive perception of these technologies. Secondly, Governance of national power infrastructure and markets tends to modify the investment patterns. Energy liberalization has increased public support for renewable energy in Organisation for Economic Co-operation and Development (OECD) countries, mainly because of reduction in entry barriers [30]. State-owned utilities show higher tendency to invest in renewable generation [ 31 ]. We represent these trends in the model by modifying as well the investment cost for solar and wind technologies in relation to two governance indicators obtained from the OECD Product Market Regulation database [ 55 ]: legal or administrative barriers for third-party access to competitive electricity markets (entry barriers) and ownership or direct control of transmission and generation enterprises by state entities (public ownership). Indicator values range from 0 to 6. When entry barriers are low ( = 4), we assumed a 20% decrease of investment costs. Conversely, when entry barriers are high ( > = 4) and public ownership is low ( < = 2), we assumed a 20% increase of investment costs. 4.3 Data sources Most of the historical data about costs, generation and capacity dynamics come from the same data package [ 56 ]. It covers the period of 1990–2019 for the 31 European countries used in this study. Original data sources were identified using a literature review focused on open-access sources [ 57 ], energy company statistics, and national agency statistics. These original data sources were then harmonized to enable their use as input parameters for power system modeling. This approach produced four types of processed data files for each country: (i) a country file documenting nationally-aggregated variables, such as annual time series of electricity demand, transmission and distribution losses, and peak load; (ii) technology files describing techno-economic variables for each major generation technology in the country’s electricity mix, such as annual time series of investment costs, installed capacities, and actual generation; (iii) resource files describing CO 2 emissions intensities and annual time series of costs for each generation fuel or input resource; and (iv) load profiles describing 24-hour national load curves for each available year. The data package includes an annotated list of all used original data sources, describes harmonization and processing steps for each variable, and provides the final processed data files in comma-separated format. Other data used and sources are described in the Supplementary Table S1 . 4.4 Hindcasting perfomance For each country, we ran hindcasting simulations using model versions with different combinations of societal factors included, following the model development methodology of Wen et al. [ 25 ]. By each societal factor being either represented or not, if all data were available, we obtained 64 (2 6 ) hindcasting simulations per country. For each hindcasting simulation, we quantified for different outputs (installed capacity, CO 2 emissions and renewable generation share) the deviation between simulation outputs and historical values over the 1990–2019 period. To do so, we first calculated the Symmetric Mean Absolute Percentage Error (SMAPE) over the years which is a relevant indicator to evaluate hindcasting performance in terms of the absolute magnitude of the deviations [58]. This indicator is also an absolute indicator without cancelation effect which allowed us to sum the values obtained over the different technologies for installed capacity to obtain an overall accuracy indicator. We then quantified the relative performance gain or loss brought by the inclusion of societal factors in the model compared to the techno-economic model version without societal factors included. We finally identified the societal factors that contribute the most to the variability of hindcasting performance. To do so, we performed analysis of variance (ANOVA) on hindcasting performance in order to compare the contributions to variance of societal factors and their interactions. We also captured the effect direction of each societal factor inclusion by comparing the average values of performance over the simulation subsets with and without the considered societal factor included [ 59 ]. Declarations Acknowledgements This work received funding from the Swiss National Science Foundation Eccellenza Grant for the project "Accuracy of long-range national energy projections" (Grant no. 186834, VFR, XW, ET) and from the Swiss State Secretariat for Education, Research, and Innovation (SEFRI) for the project “Net-zero pathway research through integrated assessment model advancements (PRISMA)” (Project no. 101081604, ET). We thank Bjarne Steffen and Friedemann Polzin for having provided the dataset of weighted average cost of capital. Contribution Methodology, VFR, ET; Software, VFR; Investigation, VFR; Data curation, MJR, XW, VFR; Conceptualization, VFR, ET; Writing – Original Draft, VFR; Writing – Review & Editing, all authors; Visualization, VFR; Supervision – ET; Funding Acquisition – ET. Competing interests The authors declare no competing interests References Méjean, A., Guivarch, C., Lefèvre, J. & Hamdi-Cherif, M. The transition in energy demand sectors to limit global warming to 1.5 C. en. Energy Efficiency 12, 441–462. issn: 15706478. https://doi.org/10.1007/s12053-018-9682-0 (Feb. 2019). Jewell, J. & Cherp, A. On the political feasibility of climate change mitigation pathways: Is it too late to keep warming below 1.5C? en. WIREs Climate Change 11. _eprint: https://onlinelibrary.wiley.com/doi/pdf/10.1002/wcc.621, e621. issn: 1757-7799. https: //onlinelibrary.wiley.com/doi/abs/10.1002/wcc.621 (2020). Geels, F. W., Sovacool, B. K., Schwanen, T. & Sorrell, S. Sociotechnical transitions for deep decarbonization. en. Science 357, 1242–1244. issn: 0036-8075, 1095-9203. https: //www.science.org/doi/10.1126/science.aao3760 (Sept. 2017). Pianta, S. & Brutschin, E. Emissions Lock-in, Capacity, and Public Opinion: How Insights From Political Science Can Inform Climate Modeling Efforts. en. Politics and Governance 10, 186–199. issn: 2183-2463. https://www.cogitatiopress.com/politicsandgovernance/ article/view/5462 (Sept. 2022). Krumm, A., Süsser, D. & Blechinger, P. Modelling social aspects of the energy transition: What is the current representation of social factors in energy models? en. Energy 239, 121706. issn: 0360-5442. https://www.sciencedirect.com/science/article/pii/ S036054422101954X (Jan. 2022). Peng, W. et al. Climate policy models need to get real about people — here’s how. en. Nature 594. Bandiera_abtest: a Cg_type: Comment Number: 7862 Publisher: Nature Publishing Group Subject_term: Climate change, Economics, Policy, Renewable energy, 174–176. https://www.nature.com/articles/d41586-021-01500-2 (June 2021). Keppo, I. et al. Exploring the possibility space: taking stock of the diverse capabilities and gaps in integrated assessment models. en. 16. Publisher: IOP Publishing, 053006. issn: 1748-9326. https://doi.org/10.1088/1748-9326/abe5d8 (Apr. 2021). Trutnevyte, E. et al. Societal Transformations in Models for Energy and Climate Policy: The Ambitious Next Step. en. One Earth 1, 423–433. issn: 2590-3322. https://www. sciencedirect.com/science/article/pii/S2590332219302246 (Dec. 2019). Hirt, L. F., Schell, G., Sahakian, M. & Trutnevyte, E. A review of linking models and socio-technical transitions theories for energy and climate solutions. en. Environmental Innovation and Societal Transitions 35, 162–179. issn: 2210-4224. https://www.sciencedirect.com/science/article/pii/S2210422420300368 (June 2020). Van Sluisveld, M. A. E. et al. Aligning integrated assessment modelling with sociotechnical transition insights: An application to low-carbon energy scenario analysis in Europe. en. Technological Forecasting and Social Change 151, 119177. issn: 0040-1625. https://www.sciencedirect.com/science/article/pii/S0040162518300593 (Feb. 2020). Brutschin, E. et al. A multidimensional feasibility evaluation of low-carbon scenarios. en. Environmental Research Letters 16. Publisher: IOP Publishing, 064069. issn: 1748-9326. https://doi.org/10.1088/1748-9326/abf0ce (June 2021). Steffen, B. Estimating the cost of capital for renewable energy projects. en. Energy Economics 88, 104783. issn: 0140-9883. https://www.sciencedirect.com/science/ article/pii/S0140988320301237 (May 2020). Verrier, B., Li, P.-H., Pye, S. & Strachan, N. Incorporating social mechanisms in energy decarbonisation modelling. en. Environmental Innovation and Societal Transitions 45, 154–169. issn: 2210-4224. https://www.sciencedirect.com/science/article/pii/ S2210422422000934 (Dec. 2022). Li, F. G. N. Actors behaving badly: Exploring the modelling of non-optimal behaviour in energy transitions. en. Energy Strategy Reviews 15, 57–71. issn: 2211-467X. https://www.sciencedirect.com/science/article/pii/S2211467X17300020 (2021) (Mar. 2017). Koecklin, M. T., Longoria, G., Fitiwi, D. Z., DeCarolis, J. F. & Curtis, J. Public acceptance of renewable electricity generation and transmission network developments: Insights from Ireland. en. Energy Policy 151, 112185. issn: 0301-4215. https://www.sciencedirect. com/science/article/pii/S0301421521000549 (Apr. 2021). Stern, P. C., Dietz, T., Nielsen, K. S., Peng, W. & Vandenbergh, M. P. Feasible climate mitigation. en. Nature Climate Change 13. Number: 1 Publisher: Nature Publishing Group, 6–8. issn: 1758-6798. https://www.nature.com/articles/s41558-022-01563-7 (Jan. 2023). Braunreiter, L., van Beek, L., Hajer, M. & van Vuuren, D. Transformative pathways – Using integrated assessment models more effectively to open up plausible and desirable low-carbon futures. en. Energy Research & Social Science 80, 102220. issn: 2214-6296. https://www.sciencedirect.com/science/article/pii/S2214629621003133 (Oct. 2021). Moore, F. C. et al. Determinants of emissions pathways in the coupled climate–social system. en. Nature 603. Number: 7899 Publisher: Nature Publishing Group, 103–111. issn: 1476-4687. https://www.nature.com/articles/s41586-022-04423-8 (Mar. 2022). Freeman, R. Modelling the socio-political feasibility of energy transition with system dynamics. en. Environmental Innovation and Societal Transitions 40, 486–500. issn: 2210-4224. https://www.sciencedirect.com/science/article/pii/S2210422421000800 (Sept. 2021). Li, F. G. N. & Strachan, N. Take me to your leader: Using socio-technical energy transitions (STET) modelling to explore the role of actors in decarbonisation pathways. en. Energy Research & Social Science 51, 67–81. issn: 2214-6296. https://www.sciencedirect. com/science/article/pii/S2214629618306029 (May 2019). Lonergan, K. E. et al. Improving the representation of cost of capital in energy system models. English. Joule 0. Publisher: Elsevier. issn: 2542-4785, 2542-4351. https://www. cell.com/joule/abstract/S2542-4351(23)00073-9 (Mar. 2023). Creutzig, F. et al. The underestimated potential of solar energy to mitigate climate change. en. Nature Energy 2. Number: 9 Publisher: Nature Publishing Group, 1–9. issn: 20587546. https://www.nature.com/articles/nenergy2017140 (Aug. 2017). Gilbert, A. Q. & Sovacool, B. K. Looking the wrong way: Bias, renewable electricity, and energy modelling in the United States. en. Energy 94, 533–541. issn: 0360-5442. https://www.sciencedirect.com/science/article/pii/S0360544215015133 (Jan. 2016). Wilson, C. et al. Evaluating process-based integrated assessment models of climate change mitigation. en. Climatic Change 166, 3. issn: 1573-1480. https://doi.org/10.1007/ s10584-021-03099-9 (May 2021). Wen, X., Jaxa-Rozen, M. & Trutnevyte, E. Hindcasting to inform the development of bottom-up electricity system models: The cases of endogenous demand and technology learning. en. Applied Energy 340, 121035. issn: 0306-2619. https://www.sciencedirect. com/science/article/pii/S0306261923003999 (June 2023). Glotin, D., Bourgeois, C., Giraudet, L.-G. & Quirion, P. Prediction is difficult, even when it’s about the past: A hindcast experiment using Res-IRF, an integrated energy-economy model. en. Energy Economics. Eighth Atlantic Workshop on Energy and Environmental Economics 84, 104452. issn: 0140-9883. https://www.sciencedirect.com/science/ article/pii/S0140988319302336 (Oct. 2019). Trutnevyte, E. Does cost optimization approximate the real-world energy transition? en. Energy 106, 182–193. issn: 0360-5442. https://www.sciencedirect.com/science/ article/pii/S0360544216302821 (July 2016). Polzin, F. et al. The effect of differentiating costs of capital by country and technology on the European energy transition. en. Climatic Change 167, 26. issn: 1573-1480. https: //doi.org/10.1007/s10584-021-03163-4 (2021) (July 2021). Cotterman, T., Small, M. J., Wilson, S., Abdulla, A. & Wong-Parodi, G. Applying risk tolerance and socio-technical dynamics for more realistic energy transition pathways. en. Ap plied Energy 291, 116751. issn: 0306-2619. https://www.sciencedirect.com/science/ article/pii/S0306261921002609 (June 2021). Nicolli, F. & Vona, F. Energy market liberalization and renewable energy policies in OECD countries. en. Energy Policy 128, 853–867. issn: 0301-4215. https://www. sciencedirect.com/science/article/pii/S0301421519300187 (May 2019). Steffen, B., Karplus, V. & Schmidt, T. S. State ownership and technology adoption: The case of electric utilities and renewable energy. en. Research Policy 51, 104534. issn: 0048-7333. https://www.sciencedirect.com/science/article/pii/S0048733322000610 (July 2022). Priesmann, J., Nolting, L. & Praktiknjo, A. Are complex energy system models more accurate? An intra-model comparison of power system optimization models. en. Applied Energy 255, 113783. issn: 0306-2619. https://www.sciencedirect.com/science/ article/pii/S0306261919314709 (Dec. 2019). Moret, S., Babonneau, F., Bierlaire, M. & Maréchal, F. Overcapacity in European power systems: Analysis and robust optimization approach. en. Applied Energy 259, 113970. issn: 0306-2619. https://www.sciencedirect.com/science/article/pii/S0306261919316575 (Feb. 2020). Schwanitz, V. J. Evaluating integrated assessment models of global climate change. en. Environmental Modelling & Software 50, 120–131. issn: 1364-8152. https://www. sciencedirect.com/science/article/pii/S1364815213001965 (Dec. 2013). Oreskes, N. Evaluation (Not Validation) of Quantitative Models. Environmental Health Perspectives 106. Publisher: [National Institute of Environmental Health Sciences, Brogan & Partners], 1453–1460. issn: 0091-6765. https://www.jstor.org/stable/4641214 (1998). Muttitt, G., Price, J., Pye, S. & Welsby, D. Socio-political feasibility of coal power phaseout and its role in mitigation pathways. en. Nature Climate Change 13. Number: 2 Publisher: Nature Publishing Group, 140–147. issn: 1758-6798. https://www.nature.com/ articles/s41558-022-01576-2 (Feb. 2023). Harmsen, M. et al. Integrated assessment model diagnostics: key indicators and model evolution. en. Environmental Research Letters 16. Publisher: IOP Publishing, 054046. issn: 1748-9326. https://dx.doi.org/10.1088/1748-9326/abf964 (May 2021). Krey, V. et al. Looking under the hood: A comparison of techno-economic assumptions across national and global integrated assessment models. en. Energy 172, 1254–1267. issn: 0360-5442. https://www.sciencedirect.com/science/article/pii/S0360544218325039 (Apr. 2019). DeCarolis, J. F., Hunter, K. & Sreepathi, S. The case for repeatable analysis with energy economy optimization models. en. Energy Economics 34, 1845–1853. issn: 0140-9883. https://www.sciencedirect.com/science/article/pii/S0140988312001405 (Nov. 2012). Vinichenko, V., Vetier, M., Jewell, J., Nacke, L. & Cherp, A. Phasing out coal for 2 C target requires worldwide replication of most ambitious national plans despite security and fairness concerns. Environmental Research Letters (2023). Cherp, A., Vinichenko, V., Tosun, J., Gordon, J. A. & Jewell, J. National growth dynamics of wind and solar power compared to the growth required for global climate targets. Nature Energy 6. Publisher: Nature Publishing Group UK London, 742–754 (2021). Gambhir, A. Powering past coal is not enough. en. Nature Climate Change 13. Number: 2 Publisher: Nature Publishing Group, 117–118. issn: 1758-6798. https://www.nature. com/articles/s41558-022-01574-4 (Feb. 2023). Li, F. G. N., Trutnevyte, E. & Strachan, N. A review of socio-technical energy transition (STET) models. en. Technological Forecasting and Social Change 100, 290–305. issn: 0040-1625. https://www.sciencedirect.com/science/article/pii/S0040162515002267 (Nov. 2015). Wise, M. et al. Representing power sector detail and flexibility in a multi-sector model. en. Energy Strategy Reviews 26, 100411. issn: 2211-467X. https://www.sciencedirect. com/science/article/pii/S2211467X1930104X (Nov. 2019). De Boer, H. S. ( S. ) & van Vuuren, D. ( P. ) Representation of variable renewable energy sources in TIMER, an aggregated energy system simulation model. en. Energy Economics 64, 600–611. issn: 0140-9883. https://www.sciencedirect.com/science/article/ pii/S0140988316303528 (May 2017). Després, J. et al. Storage as a flexibility option in power systems with high shares of variable renewable energy sources: a POLES-based analysis. en. Energy Economics 64, 638–650. issn: 0140-9883. https://www.sciencedirect.com/science/article/pii/ S0140988316300445 (May 2017). Fisch-Romito, V., Guivarch, C., Creutzig, F., Minx, J. C. & Callaghan, M. W. Systematic map of the literature on carbon lock-in induced by long-lived capital. en. Environmental Research Letters 16. Publisher: IOP Publishing, 053004. issn: 1748-9326. https://doi. org/10.1088/1748-9326/aba660 (Apr. 2021). Fouquet, R. Path dependence in energy systems and economic development. en. Nature Energy 1, 16098. issn: 2058-7546. https://www.nature.com/articles/nenergy201698 (July 2016). Hirth, L. & Steckel, J. C. The role of capital costs in decarbonizing the electricity sector. en. Environmental Research Letters 11. Publisher: IOP Publishing, 114010. issn: 17489326. https://doi.org/10.1088/1748-9326/11/11/114010 (Nov. 2016). Egli, F., Steffen, B. & Schmidt, T. S. A dynamic analysis of financing conditions for renewable energy technologies. en. Nature Energy 3. Number: 12 Publisher: Nature Publishing Group, 1084–1092. issn: 2058-7546. https://www.nature.com/articles/s41560-0180277-y (Dec. 2018). Cousse, J., Wüstenhagen, R. & Schneider, N. Mixed feelings on wind energy: Affective imagery and local concern driving social acceptance in Switzerland. en. Energy Research & Social Science 70, 101676. issn: 2214-6296. https://www.sciencedirect.com/science/ article/pii/S2214629620302516 (Dec. 2020). Ruhnau, O., Eicke, A., Sgarlato, R., Tröndle, T. & Hirth, L. Cost-Potential Curves of Onshore Wind Energy: the Role of Disamenity Costs. en. Environmental and Resource Economics. issn: 1573-1502. https://doi.org/10.1007/s10640-022-00746-2 (Nov. 2022). Harold, J., Bertsch, V., Lawrence, T. & Hall, M. Drivers of People’s Preferences for Spatial Proximity to Energy Infrastructure Technologies: A Cross-country Analysis. The Energy Journal 42. issn: 01956574. http://www.iaee.org/en/publications/ejarticle. aspx?id=3703 (Oct. 2021). Stokes, L. C. & Warshaw, C. Renewable energy policy design and framing influence public support in the United States. en. Nature Energy 2. Number: 8 Publisher: Nature Publishing Group, 1–6. issn: 2058-7546. https://www.nature.com/articles/nenergy2017107 (June 2017). OECD. Product Market Regulation Network Sectors indicators: Methodology for calculating the 1975-2018 time series. en (2018). Jaxa-Rozen, M., Wen, X. & Trutnevyte, E. Historic data of the national electricity system transitions in Europe in 1990–2019 for retrospective evaluation of models. en. Data in Brief 43, 108459. issn: 2352-3409. https://www.sciencedirect.com/science/article/pii/ S2352340922006540 (Aug. 2022). Eurostat. Eurostat Dissemination Database. Publications Office of the European Union. https://ec.europa.eu/eurostat/data/database (2023). Wen, X., Jaxa-Rozen, M. & Trutnevyte, E. Accuracy indicators for evaluating retrospective performance of energy system models. en. Applied Energy 325, 119906. issn: 0306-2619. https://www.sciencedirect.com/science/article/pii/S0306261922011667 (Nov. 2022). Guivarch, C. & Monjon, S. Identifying the main uncertainty drivers of energy security in a low-carbon world: The case of Europe. en. Energy Economics 64, 530–541. issn: 0140-9883. https://www.sciencedirect.com/science/article/pii/S014098831630086X (May 2017). Additional Declarations There is NO Competing Interest. Supplementary Files FischRomitoetalSupplementarymaterial.pdf Supplementary Information Cite Share Download PDF Status: Published Journal Publication published 21 Feb, 2025 Read the published version in Nature Energy → Version 3 posted You are reading this latest preprint version Show more versions Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-4312891","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[{"code":1,"date":"2024-04-26 02:07:49","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}},{"code":2,"date":"2024-04-26 22:13:57","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"articleType":"Article","associatedPublications":[],"authors":[{"id":302855772,"identity":"40e6a28f-5081-4a46-b998-395d68e03a91","order_by":0,"name":"Vivien Fisch-Romito","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABEklEQVRIie3PMUvEMBTA8VcCb0ov6ysO9xXiUh3K9YO4nBQ6eZOjcBQCcZKbbxC/xg0OKQFd+g1u6dRNOEdBwVRqHUwFN5H8hzak70cTgFDoL8YAjHsJBMCPDfeMWgAa1tMk0SNBYPJHMpRU4wACEnwuPclHfmyed0AzYbs2us/OZnP1cBXvThcbYF3rI5bLetsAIZUnErpypRHLfdxQsa3Q7XgOpHBpYw1rJI5Hr8Y6wtN9rKmQho8n/EbetPuLaNzAQC57kk8QwZixUU/g4oswRxYSpkl9o6m/S+pIf5eySG41Lcli6iMoanV40RnNN7ZzJFvdKVsfnvQ6F9eq85Ehz6fzik3P+8t/C0KhUOjf9g4dp1FVuHKPZAAAAABJRU5ErkJggg==","orcid":"https://orcid.org/0000-0002-2419-196X","institution":"University of Geneva","correspondingAuthor":true,"prefix":"","firstName":"Vivien","middleName":"","lastName":"Fisch-Romito","suffix":""},{"id":302855773,"identity":"d743078b-d89f-4f2c-80b4-a3cf4a37127c","order_by":1,"name":"Marc Jaxa-Rozen","email":"","orcid":"","institution":"University of Geneva","correspondingAuthor":false,"prefix":"","firstName":"Marc","middleName":"","lastName":"Jaxa-Rozen","suffix":""},{"id":302855774,"identity":"fee8be80-82fc-4271-996f-c9405e4338fc","order_by":2,"name":"Xin Wen","email":"","orcid":"https://orcid.org/0000-0001-6479-0126","institution":"University of Geneva","correspondingAuthor":false,"prefix":"","firstName":"Xin","middleName":"","lastName":"Wen","suffix":""},{"id":302855775,"identity":"bd2f29c2-c432-4e2e-baa3-b97f1d451576","order_by":3,"name":"Evelina Trutnevyte","email":"","orcid":"https://orcid.org/0000-0002-1716-6192","institution":"University of Geneva","correspondingAuthor":false,"prefix":"","firstName":"Evelina","middleName":"","lastName":"Trutnevyte","suffix":""}],"badges":[],"createdAt":"2024-04-23 15:04:23","currentVersionCode":3,"declarations":{"humanSubjects":false,"vertebrateSubjects":false,"conflictsOfInterestStatement":false,"humanSubjectEthicalGuidelines":false,"humanSubjectConsent":false,"humanSubjectClinicalTrial":false,"humanSubjectCaseReport":false,"vertebrateSubjectEthicalGuidelines":false},"doi":"10.21203/rs.3.rs-4312891/v3","doiUrl":"https://doi.org/10.21203/rs.3.rs-4312891/v3","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41560-025-01719-7","type":"published","date":"2025-02-21T05:00:00+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":56618973,"identity":"4ee07738-69c3-46aa-9987-27cadf5015e8","added_by":"auto","created_at":"2024-05-16 17:46:22","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":85909,"visible":true,"origin":"","legend":"\u003cp\u003eComparison of the hindcasting performance of the techno-economic model and all model versions with societal factors included in 1990–2019 in terms of installed capacity. Hindcasting performance corresponds to the values of symmetric mean absolute percentage error over the years (SMAPE) summed over all technologies. The lower (higher) is the value, the better (worse) is the hindcasting performance. Red dots refer to the techno-economic model version without societal factors incorporated. Black bars refer to the range of model version with societal factors included one by one and as combinations. Countries are ordered by gain in hindcasting performance induced by the inclusion of societal factors compared to the techno-economic model. The blue line represents the average hindcasting performance of all combination cases of countries and societal factors. For each country, the combination of societal factors leading to the best hindcasting performance is highlighted for each country. Each letter refers to one societal factor incorporated: L — Lock-in; T — Fast Transition; A — Actors heterogeneity; I — Investment risk; P — Public acceptance; G — Governance.\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-4312891/v3/bb7a9db38fa46ffc0b478b28.png"},{"id":56618974,"identity":"66cead2a-4d27-4d0e-aa19-8e00a26ba594","added_by":"auto","created_at":"2024-05-16 17:46:22","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":152598,"visible":true,"origin":"","legend":"\u003cp\u003eInfluence of societal factors on model hindcasting performance for capacity dynamics. Upper figure (a) shows the direction of the effect of each societal factor. Bottom figure (b) shows the analysis of variance (ANOVA) on model hindcasting performance. Hindcasting performance corresponds to the symmetric mean absolute percentage error over the years (SMAPE) summed over all technologies. In figure (a), triangle represents the average relative model performance across the hindcasting simulation subset with the considered societal factor included. Point represents the average relative model performance across the hindcasting simulation subset without the considered societal factor included. Only the three most influencing factors are represented for readability. All factors are represented in the Supplementary Fig. S2.\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-4312891/v3/744388f403b98ac43b689396.png"},{"id":56620368,"identity":"eb7b8fbc-c1d7-4ca5-8d9d-c16def3ef05f","added_by":"auto","created_at":"2024-05-16 17:54:22","extension":"jpeg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":268824,"visible":true,"origin":"","legend":"\u003cp\u003eComparison of the hindcasting performance (relative to the techno-economic model) for installed capacity, share of renewable electricity production and \u003cem\u003eCO\u003c/em\u003e\u003csub\u003e2 \u003c/sub\u003eemissions. Each dot refers to a model version with one combination of societal factors incorporated. Black dot represents the model version with the best hindcasting performance for installed capacity).\u003c/p\u003e","description":"","filename":"floatimage3.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-4312891/v3/343895b8ea6544b8bf8e869d.jpeg"},{"id":76912823,"identity":"020b2e69-98d8-472d-8674-e03ff8a7fbe3","added_by":"auto","created_at":"2025-02-22 08:05:30","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1133767,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4312891/v3/0815928c-4d5b-4a9f-b858-70c6f9dcad39.pdf"},{"id":56618975,"identity":"4311f5f0-5670-4c44-a721-fff2c48a241e","added_by":"auto","created_at":"2024-05-16 17:46:22","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":779052,"visible":true,"origin":"","legend":"\u003cp\u003eSupplementary Information\u003c/p\u003e","description":"","filename":"FischRomitoetalSupplementarymaterial.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4312891/v3/af5bc27929a65e53a09aa8f1.pdf"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Multi-country evidence on societal factors to include in energy transition modeling","fulltext":[{"header":"1 Introduction","content":"\u003cp\u003eIntegrated assessment models of climate change and energy system models are widely used to quantify low-carbon pathways at a global and national scales. For the power sector, which needs to decarbonize earlier than other sectors [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e], the models represent electricity supply, end-use technologies, service demands and their interlinked dynamics to assess the sector\u0026rsquo;s transformation to enable broader carbon neutrality. In practice, this transformation depends on the co-evolution of technical, social, economic, institutional or political factors which may induce lock-ins or tipping point dynamics and change feasibility space for the emergence of new technologies [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. Modelers are now increasingly requested to account for societal transformation dynamics in their modeled pathways to inform policy makers about intervention designs, their feasibility and impact [\u003cspan additionalcitationids=\"CR5 CR6 CR7\" citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e8\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eSocial scientific knowledge on energy and climate mitigation has so far evolved in parallel to modeling and analysis of pathways [9] with rare links done either ex-ante with informed storylines [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e] or ex-post by evaluating pathways outputs on different societal indicators [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e11\u003c/span\u003e]. The few articles that have attempted to represent societal factors of power sector transitions in technology-rich models have focused on one factor at a time, such as cost of capital [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e12\u003c/span\u003e], actors behaviour [13, 14], or social acceptance [15]. As a result, critiques of models have emerged in the meantime about their poor ability to produce solutions-oriented outcomes or to inform about feasibility [9, 16, 17]. Analyses that endogenize multiple societal factors into models could contribute to responding to these critiques, but such models are at early stage of development [18\u0026ndash;20].\u003c/p\u003e \u003cp\u003eMost of all, it is unclear whether some societal factors should be given priority over others for inclusion in models and how these factors should be endogenized (see for instance recent methodology description for the integration of capital costs [\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e21\u003c/span\u003e]). Critiques have pointed out the limited performance of power sector simulation models in capturing real-word transitions [22, 23] and it is also unclear to what extent the incorporation of societal factors would improve it or not. This relates to model evaluation which is an emerging topic in the energy and integrated assessment modelling community to improve robustness and usefulness of models for policy makers [\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e24\u003c/span\u003e]. Hindcasting \u0026mdash; assessing model ability to reproduce past dynamics \u0026mdash; is one relevant evaluation method to assess the appropriateness and credibility of a model structure [\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e24\u003c/span\u003e], but few hindcasting exercises have been done so far [\u003cspan additionalcitationids=\"CR26\" citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e27\u003c/span\u003e] and none has considered modeling societal factors.\u003c/p\u003e \u003cp\u003eHere, we provide first-of-the-kind empirical evidence, using an open-source techno-economic simulation model of the power system transition, where we incorporate, one by one and as a combination, six societal factors related to (i) infrastructure dynamics, (ii) actors and decisions making and (iii) social and institutional context (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). We set up this model for 31 European countries (EU27, Switzerland, Iceland, Norway, and the UK) to allow generalizable insights and conduct a hindcasting evaluation of power capacity expansion, renewable generation and \u003cem\u003eCO\u003c/em\u003e\u003csub\u003e2\u003c/sub\u003e emissions from 1990 to 2019. This approach allows us to conclude on which societal factors and how should be ideally incorporated in future modeling.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eSocietal factors included in the model one by one and as combinations. Methods section provides more details on model implementation of these factors.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"3\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAspects\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSocietal factors included\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eApproach how this factor is represented in the model\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eInfrastructure dynamics\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eLock-in (L)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003ePower capacity lifespan is extended if this capacity is economically profitable to keep using in the future at the same utilization rate.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eFast transition (T)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eExisting capacity is retired before the end of the lifespan if it is economically profitable to build and use alternative technologies at the same utilization rate.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eActors and decision making\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eActors heterogeneity (A)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eLower cost sensitivity is assumed for investment choices [20].\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eInvestment risks (I)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eDifferent weighted average costs of capital (WACC) are used per countries and technologies [\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e].\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eSocial and institutional context\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePublic acceptance (P)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eIncumbent technologies cannot be built in the case of low public acceptance [29].\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eGovernance (G)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eElectricity market liberalization facilitates renewable introduction [30], while state-owned utilities show higher tendency to invest in renewable [\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e].\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e"},{"header":"2 Results","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\n\u003ch2\u003e2.1 Societal factors in the model tend to improve the hindcasting performance for capacity expansion\u003c/h2\u003e\n\u003cp\u003eWe find that incorporating societal factors in the model one by one or as combinations mostly improves the hindcasting performance in terms of installed capacity dynamics for most of the 31 European countries studied (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e). Only in the cases of Iceland and Latvia, the model version with the best hindcasting performance (i.e. lowest error) is the techno-economic one without societal factors incorporated. For the rest, the model version with the best hindcasting performance includes at least one societal factor. We obtain high heterogeneity between countries in terms of hindcasting performance gain compared to the techno-economic model. Minor gain in hindcasting performance is achieved for Lithuania, Norway, Slovenia and Switzerland (less than 1% of symmetric mean absolute percentage error (SMAPE)) and low gain (less than 5%) for another nine countries (Belgium, Bulgaria, Croatia, Cyprus, Greece, Malta, Poland, Portugal, Romania). Countries with minor gains are characterized by low cumulative installed capacities in 1990, below 10 GW, and a limited diversity of technologies, or by a high proportion of hydroelectric power plants, which induces inertia in the evolution of the system due to long lifetimes. Values for Austria, Denmark, France, Germany, Hungary, Ireland, Italy, Luxembourg, Netherlands and Slovakia lie in between 5% and 15%. We obtain the highest hindcasting performance gain of more than 15% of SMAPE for six countries (Czech Republic, Estonia, Finland, Spain, Sweden and United Kingdom). This is mainly because, contrary to the techno-economic model, the inclusion of some societal factors in the model allows to better capture the increase in installed capacities of decentralized renewable technologies (e.g. biomass, offshore and onshore wind and solar PV) observed in these countries from 2010.\u003c/p\u003e\n\u003cp\u003eEven if the inclusion of one or several societal factors can improve the hindcasting performance for most of the countries, there are many cases when including some specific factors or their combinations, the hindcasting performance gets worse than the performance of the techno-economic model (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e.). Societal factors hence need to be included in an evidence-based way to avoid worsening the model\u0026rsquo;s performance and hindcasting is a relevant approach to do so. We also observe a large spread in terms of absolute hindcasting performance (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e). On the one hand, for Belgium, Lithuania, Luxembourg, Malta, Netherlands, Portugal and Switzerland, the model version with the worst hindcasting performance has a better performance than the average of the whole simulation set, all countries considered. Except for Luxembourg and Netherlands, low performance gains are obtained for these countries as well, pointing to a tiebreak between additional modeling efforts and actual gain. On the other hand, for Bulgaria, Norway and Sweden, the model version with the best hindcasting performance gives more errors than the average value of the whole hindcasting simulations set. For this second group of countries, this suggests a potential for improving the model structure and the way that societal factors are incorporated in order to improve the model performance to capture capacity dynamics.\u003c/p\u003e\n\u003cp\u003e2.2 The contribution of the different societal factors to model hindcasting performance differs between countries and factors\u003c/p\u003e\n\u003cp\u003eThe selection of societal factors leading to the best hindcasting performance differs between countries (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e). First, there is a difference in terms of number of societal factors incorporated, ranging from one for Croatia, Cyprus, Lithuania, Malta, Norway, Romania and Switzerland to four for Austria, Denmark, Germany, Italy and Luxembourg. There is no country where the integration of more than four factors leads to an additional gain in performance. This suggests that increasing the complexity of the model does not necessarily increase its accuracy, supporting previous findings on complexity of technology representation in models [\u003cspan class=\"CitationRef\"\u003e32\u003c/span\u003e]. Second, the best combinations differ in terms of societal factors incorporated. Each societal factor is present for at least four countries in the model version giving the best hindcasting performance which suggests that all societal factors analyzed are potentially relevant to incorporate in integrated assessment and energy systems models. \u003cem\u003eLock-in\u003c/em\u003e, \u003cem\u003ePublic acceptance\u003c/em\u003e and \u003cem\u003eInvestment risks\u003c/em\u003e are the societal factors most present in the model versions giving the lowest errors, with respectively 14, 14 and 12 occurrences over 31 countries (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e).\u003c/p\u003e\n\u003cp\u003eModel hindcasting performance is more sensitive to the incorporation of some societal factors in the model than others, both positively and negatively. \u003cem\u003eActors\u0026rsquo; heterogeneity\u003c/em\u003e has the highest influence in most countries (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003eb) which is expected since this factor accounts for cost sensitivity which in turn drives the shares of new investments for all technologies. The second most influencing factor overall is the \u003cem\u003eFast transition\u003c/em\u003e factor which represents more than 20% of the variance for ten countries (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003eb) but tends to decrease the hindcasting performance for them (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003ea). \u003cem\u003ePublic acceptance\u003c/em\u003e is the third most influential factor representing more than 10% of the hindcasting performance variance in six countries (Austria, Germany, Netherlands, Portugal, Romania, Switzerland), but with a negative effect on model hindcasting performance for Switzerland and Romania (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003ea and \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003eb). This highlights the importance to be cautious when introducing or omitting these three factors because they are not always beneficial.\u003c/p\u003e\n\u003cp\u003eInteractions of certain specific societal factors also have a high influence on model hindcasting performance for Italy and Netherlands where they represent more than 25% of the variance (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003eb). In the case of Italy, this is mainly due to a synergistic effect of \u003cem\u003eFast transition\u003c/em\u003e and \u003cem\u003eActors heterogeneity\u003c/em\u003e that gives an overall hindcasting gain greater than the sum of the effects of each factor incorporated individually (Supplementary Fig. \u003cspan class=\"InternalRef\"\u003eS1\u003c/span\u003e). Conversely, for the Netherlands, we obtain antagonistic effect with the combination of \u003cem\u003ePublic acceptance\u003c/em\u003e and \u003cem\u003eGovernance\u003c/em\u003e and with \u003cem\u003eActors heterogeneity\u003c/em\u003e and \u003cem\u003eGovernance\u003c/em\u003e, which give lower performance than the sum of the effects of each factor incorporated individually (Supplementary Fig. \u003cspan class=\"InternalRef\"\u003eS1\u003c/span\u003e). Remarkably, the integration in the model of any societal factor, one by one or in combination, has limited influence on the model hindcasting performance for Lithuania and Portugal (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003ea). Lithuania is characterized by an oversized supply system compared to its electricity demand with low technology diversity, creating a locked dynamics in all modeled pathways. For Portugal, the much lower costs for coal and wind power compared with other technologies drive the capacity dynamics in the model.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec4\" class=\"Section2\"\u003e\n\u003ch2\u003e2.3 Societal factors in the model induce a hindcasting performance trade-off among model outputs\u003c/h2\u003e\n\u003cp\u003eFor most countries the combination of societal factors giving the lowest deviation from historical data for capacity dynamics is not the same as giving the lowest deviation for \u003cem\u003eCO\u003c/em\u003e\u003csub\u003e2\u003c/sub\u003e emissions and the share of renewable electricity production (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003e). We hence obtain a trade-off in hindcasting performance between model outputs but to a different extent between countries. For a first group of sixteen countries composed of Austria, Bulgaria, Croatia, Cyprus, Czech Republic, Estonia, France, Germany, Hungary, Italia, Malta, Norway, Poland, Portugal, Spain and United Kingdom, the model version giving the best hindcasting performance for installed capacity also improves the model performance for the two other outputs compared to the techno-economic model. For the second group of twelve countries composed of Belgium, Denmark, Finland, Greece, Ireland, Luxembourg, Netherlands, Romania Slovakia, Slovenia, Sweden and Switzerland, the model version giving the best hindcasting performance for installed capacity gives more errors than the technoeconomic model for at least one of the other outputs. This suggests the need to use a combination of societal factors, like a country-specific pareto-optimal set of factors, that altogether would allow to improve the hindcasting performance for the three metrics, and other relevant model metrics too. For Belgium, Denmark, Finland, Ireland, and Slovakia, there are combinations of societal factors that are relevant to improve the performance for the three model outputs. Conversely, this is not the case for Switzerland, Greece, Romania and Slovenia, suggesting the need for these countries to better represent the societal factors or to further modify the original structure of the model.\u003c/p\u003e\n\u003c/div\u003e"},{"header":"3 Discussion","content":"\u003cp\u003e \u003cdiv class=\"BlockQuote\"\u003e \u003cp\u003eThis study provides new and systematic evidence to further incorporate societal factors into integrated assessment models and energy system models to construct low-carbon transition pathways. By doing a multi-country hindcasting exercise with a simulation model of the power sector, we find that incorporating societal factors related to (i) infrastructure dynamics, (ii) actors and decisions making and (iii) social and institutional context has the potential to improve the model performance to capture the real-world dynamics of installed capacities of individual technologies in the power sector. This finding is valid for most of the European countries studied but the combination of societal factors incorporated to increase the hindcasting performance differs between countries. \u003cem\u003eActors heterogeneity\u003c/em\u003e influences the model outputs the most but \u003cem\u003eLock-in\u003c/em\u003e of infrastructure, \u003cem\u003ePublic acceptance\u003c/em\u003e and \u003cem\u003eInvestment risks\u003c/em\u003e are most likely to be associated with a best hindcasting performance. Modellers should hence focus their future research on these four societal factors for integration into models.\u003c/p\u003e \u003cp\u003eWe show that the gain in model performance induced by the integration of societal factors to capture power system dynamics depends strongly on the national energy context. Low performance gains are obtained for countries with small cumulative installed capacities and with high shares of hydroelectric production in their electricity mix. Conversely, high performance gains are obtained for countries who tend to have high cumulative installed capacities and that have experienced in the last decade rapid growth in decentralized renewable technologies. We also highlight that a performant model structure for installed capacities is not necessarily relevant to capture other output dynamics, such as renewable generation or CO\u003csub\u003e2\u003c/sub\u003e emissions with a potential accuracy trade-off. Going further, these elements suggest that there is no one-size-fits-all approach and that models that represents national energy transitions should be more country-specific in building their structure and should be evaluated on different outputs to choose one or several adequate model structures. Hindcasting exercises are key here to provide evidence for model construction.\u003c/p\u003e \u003cp\u003eOur hindcasting study was limited to six societal factors frequently mentioned in the literature as influencing the installed power capacity dynamics. In the future, testing the impact of other societal factors or other model implementations of the same societal factors could be relevant, not only for countries where societal factors decreased the model performance but also to improve the model performance simultaneously on multiple outputs for countries where we obtain a high-performance tradeoff. Second, including the factors influencing electricity demand would be interesting in order to better capture the co-evolution of electricity demand and supply that has led to high overcapacity for some European countries [\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e]. Capturing this co-evolution is all the more important at a time when energy security and demand reduction are back in the spotlight because of the conflict in Ukraine. Finally, future work should include societal factors and test the hindcasting performance of the whole energy system models and eventually of integrated assessment models of climate change.\u003c/p\u003e \u003cp\u003eEnergy modelers are always torn between proposing the most realistic model structure possible for policy relevance and at the same time making simplifications to be able to interpret and communicate the results. This study confirms that increasing model complexity (i.e. incorporating more societal factors) is not equal to increasing model accuracy and performance [\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e]. The hindcasting methodology could be replicated during any model building phase [25, 58] in order to obtain a parsimonious model using as few explanatory variables as possible, while still ensure that the model is \u0026rsquo;good enough\u0026rsquo; for its specific purpose [34, 35]. Moreover, this study highlights that the modellers\u0026rsquo; decisions to incorporate or not to incorporate societal factors are impactful for results. The hindcasting methodology allows to disentangle the societal factors that most influence the model performance by themselves or in interaction with other factors. The exercise done here hence paves the way to reduce subjectivity of modeller\u0026rsquo;s decisions [\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e] and to be transparent on the structural uncertainties of the model [\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e24\u003c/span\u003e]. We also provide tools and datasets in this paper and elsewhere [56, 58] to enable informed choice.\u003c/p\u003e \u003cp\u003eThis study supports that further efforts should be devoted to two aspects on model building methodology. First, introducing societal factors in integrated assessment and energy system models can significantly improve their performance which calls for a stronger cooperation of this community with social sciences to have representations of societal factors more empirically based. Second, integrated assessment and energy system models have been mostly \u0026rsquo;diagnosed\u0026rsquo; so far through prospective intercomparison exercises to identify their main assumptions, characteristics, and behaviour [\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e37\u003c/span\u003e, \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e38\u003c/span\u003e]. To go further, we argue this should be completed by common hindcasting exercises, as done in this study, in order to highlight their (in)ability to capture sectoral and national dynamics and adapt the model structure. Reproducing past trends in hindcasting is not the guarantee of a better model performance because energy-economy-social systems do not exhibit structural constancy over time [\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e, \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e]. However, a scenario produced from a model that is good enough to capture past trends can bring insights for discussing not only the techno-economic, but also socio-political feasibility of achieving climate targets [40, 41] and highlight how current socio-technical dynamics potentially hinders climate ambition in the short term and what policies are needed [\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e42\u003c/span\u003e].\u003c/p\u003e \u003c/div\u003e \u003c/p\u003e"},{"header":"4 Methodology","content":"\u003cdiv class=\"BlockQuote\"\u003e\n\u003cp\u003eIn our paper, we set-up an open-source model of the power system transition (Section \u003cspan class=\"InternalRef\"\u003e4.1\u003c/span\u003e) and followed a two-step modelling approach. In the first step, we developed a set of hindcasting simulations of national power generation and capacity expansion in 31 European countries (EU27, Switzerland, Iceland, Norway, and the United Kingdom) from 1990 to 2019. Each simulation corresponds to a model version where one or a combination of several societal factors from Table\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e are included (Section \u003cspan class=\"InternalRef\"\u003e4.2\u003c/span\u003e). Section \u003cspan class=\"InternalRef\"\u003e4.3\u003c/span\u003e details the data sources. In the second step, from each scenario we extracted the trends in terms of dynamics of installed capacity, generation and \u003cem\u003eCO\u003c/em\u003e\u003csub\u003e2\u003c/sub\u003e emissions and compared them to historical pathways to evaluate the hindcasting performance. Section \u003cspan class=\"InternalRef\"\u003e4.4\u003c/span\u003e describes the metrics used to evaluate hindcasting performance.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec7\" class=\"Section2\"\u003e\n\u003ch2\u003e4.1 The STONES model\u003c/h2\u003e\n\u003cdiv class=\"BlockQuote\"\u003e\n\u003cp\u003eFor the purpose of this study, we built an open-source simulation model of the power system, called Socio-Technical Outlook of National Energy System (STONES) (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://github.com/VivienFischRomito/STONES_model_open\u003c/span\u003e\u003c/span\u003e). STONES is a bottom-up, technology-rich, simulation model with a recursive dynamic of one year time step. The model can be considered as a socio-technical energy transition model since it represents technoeconomic detail, actor heterogeneity and transition pathway dynamics [\u003cspan class=\"CitationRef\"\u003e43\u003c/span\u003e]. It includes 14 different electricity production and storage technologies. The model takes as inputs national energy systems data such as costs, resource potentials, and flexibility constraints for each technology. The outputs are the installed capacities of individual technologies and electricity production for every year for all technologies. The STONES model is composed of three distinct modules used consecutively at each time step: the electricity demand module, the dispatch module and the capacity expansion module, all described in detail in the Supplementary Information Section \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e.\u003c/p\u003e\n\u003cp\u003eThe electricity demand module builds load duration curves for the current year and the expected ones in the future. It allows to obtain hourly electricity demand values for the observed year to feed the dispatch module, and for five years ahead to feed the capacity expansion module. The model uses as inputs the hourly electricity demand profiles and the hourly load factors of non-dispatchable renewable technologies for the current year. The annual time series are aggregated using clustering into six representative days to reduce calculation time because of computational constraints.\u003c/p\u003e\n\u003cp\u003eThe dispatch module determines the system-wide cost-optimal dispatch by minimizing the sum of variable costs for all technologies while satisfying hourly electricity demand constraints over the six representative days obtained from the electricity demand module. This optimization is done satisfying different constraints on electricity production, adequacy and transmission capacities.\u003c/p\u003e\n\u003cp\u003eThe capacity dynamic module determines the evolution of installed power capacity to satisfy the projected electricity demand five years later. As main inputs, the model uses technology costs, age of installed capacity and expected load duration curves obtained from the electricity demand module. First, the module estimates the total capacity needed to satisfy the expected peak demand five years later based on the expected capacity retirement over the following years. Then, since only a part of the installed capacity generates power at full load all year, the module divides the load duration curves into load investment segments (i.e. peak load, intermediate load, base load) to make the link with dispatch decisions, following previous approaches [\u003cspan class=\"CitationRef\"\u003e44\u003c/span\u003e\u0026ndash;\u003cspan class=\"CitationRef\"\u003e46\u003c/span\u003e]. The total needed capacity is then split between technologies using a multinomial logit equation, a common approach in energy simulation models [7, 45]. To do so, for each investment segment, the technologies are compared based on their annualized levelized costs. Finally, the capital stock evolves at each time step and for each technology, based on new investments made and retirement of the existing capacity according to the lifespan.\u003c/p\u003e\n\u003c/div\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e\n\u003ch2\u003e4.2 Representation of societal transformation factors in the model\u003c/h2\u003e\n\u003cdiv class=\"BlockQuote\"\u003e\n\u003cp\u003eIn the STONES model, we integrated six societal transformation factors related to three influencing aspects of energy transition, namely (i) infrastructure dynamics, (ii) actors and decision making and (iii) social and institutional context (Table\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e). Infrastructure dynamics influence the transition speed of existing capital towards new technologies. On the one hand, power infrastructure has long lifespan and is characterized by lumpiness and a path-dependent evolution [47, 48]. This \u003cem\u003eLock-in\u003c/em\u003e effect factor incentivizes the usage of existing capacity for a longer time. On the other hand, the market environment can incentivize earlier retirement of capacities, inducing a more flexible infrastructure dynamics and allowing a \u003cem\u003eFast transition\u003c/em\u003e towards alternative power technologies. We incorporate these two factors in the model for coal, gas, and nuclear technologies since they have represented an important share of the electricity mix in the last decades and because capacity age data were available. We represent \u003cem\u003eLock-in\u003c/em\u003e in the model by allowing a lifetime extension of ten years. To do so, at the expected lifespan end of incumbent capacity, the model compares the costs of continuing to use the incumbent capacity (i.e. annual operational costs) with the costs of building and using new capacity of alternative technologies (i.e. annual levelized costs). To represent \u003cem\u003eFast transition\u003c/em\u003e, a similar comparison is done in each year in the model. The cost comparisons are done for both societal factors assuming that potential alternative technologies would produce electricity in a similar pattern to avoid, for instance, inconsistent replacement of a base load technology by a peak technology. We use a multinomial logit equation for this comparison to allow partial replacement/extension of the capacities.\u003c/p\u003e\n\u003cp\u003eWhen incorporating the aspects of actors and decision making, we modified the perceived technology costs and hence influenced the direction of power capacity transition. First, there are various actors involved in the process of capacity investment decision making and some of them allocate a lower weight to technology cost for their decision. We represent \u003cem\u003eActors heterogeneity\u003c/em\u003e in the model by decreasing the cost sensitivity of investment choices [\u003cspan class=\"CitationRef\"\u003e14\u003c/span\u003e]. Second, \u003cem\u003eInvestment risks\u003c/em\u003e quantified by the cost of capital appear as an important driver of investment choices, especially for technologies with high upfront costs, such as renewable generation [49], and it evolves over time and differ between countries [\u003cspan class=\"CitationRef\"\u003e50\u003c/span\u003e]. Following the methodology of Steffen [\u003cspan class=\"CitationRef\"\u003e12\u003c/span\u003e], we represent risk in the model by replacing the discount rate with the weighted average cost of capital in the levelized cost calculation differentiated by country, technology and observed year.\u003c/p\u003e\n\u003c/div\u003e\n\u003cp\u003eThe social and institutional contexts create non-linear change in energy transition by either continuing past trends of investments or, with changes beyond a certain threshold, speeding up the uptake of new technologies. First of all, \u003cem\u003epublic acceptance\u003c/em\u003e influences the emergence of renewable generation in the energy landscape. Wind energy diffusion has recently slowed down in some countries due to public acceptance issues [51]. Negative public opinion about incumbent technologies can lead to a technology phase-out such as nuclear power in Germany during the 2010s, creating a window of opportunity for the emergence of renewable technologies. To represent this stylized fact, we assumed a technology cannot be built in the model for a given year when more than 60% of the population has a negative perception of it, following Cotterman et al. [29]. For wind power, local acceptance of wind projects is a more important factor than overall public opinion because of potential disamenity costs [15, 52]. Based on data from previous studies [15, 53], we assumed that local acceptance is 35% lower than overall public acceptance for wind power and then used this figure to apply the same approach as for other technologies to represent power technologies rejection. Conversely, widespread positive opinion about renewable generation technologies over the population creates pressure on governments for the adoption of support policies [\u003cspan class=\"CitationRef\"\u003e54\u003c/span\u003e]. To represent the variation in perception of adoption costs, we assumed that wind and solar investment costs are reduced by 20% if more than 80% of the population has a positive perception of these technologies. Secondly, \u003cem\u003eGovernance\u003c/em\u003e of national power infrastructure and markets tends to modify the investment patterns. Energy liberalization has increased public support for renewable energy in Organisation for Economic Co-operation and Development (OECD) countries, mainly because of reduction in entry barriers [30]. State-owned utilities show higher tendency to invest in renewable generation [\u003cspan class=\"CitationRef\"\u003e31\u003c/span\u003e]. We represent these trends in the model by modifying as well the investment cost for solar and wind technologies in relation to two governance indicators obtained from the OECD Product Market Regulation database [\u003cspan class=\"CitationRef\"\u003e55\u003c/span\u003e]: legal or administrative barriers for third-party access to competitive electricity markets (entry barriers) and ownership or direct control of transmission and generation enterprises by state entities (public ownership). Indicator values range from 0 to 6. When entry barriers are low (\u0026thinsp;\u0026lt;\u0026thinsp;=\u0026thinsp;2) and public ownership is high (\u0026thinsp;\u0026gt;\u0026thinsp;=\u0026thinsp;4), we assumed a 20% decrease of investment costs. Conversely, when entry barriers are high (\u0026thinsp;\u0026gt;\u0026thinsp;=\u0026thinsp;4) and public ownership is low (\u0026thinsp;\u0026lt;\u0026thinsp;=\u0026thinsp;2), we assumed a 20% increase of investment costs.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec9\" class=\"Section2\"\u003e\n\u003ch2\u003e4.3 Data sources\u003c/h2\u003e\n\u003cdiv class=\"BlockQuote\"\u003e\n\u003cp\u003eMost of the historical data about costs, generation and capacity dynamics come from the same data package [\u003cspan class=\"CitationRef\"\u003e56\u003c/span\u003e]. It covers the period of 1990\u0026ndash;2019 for the 31 European countries used in this study. Original data sources were identified using a literature review focused on open-access sources [\u003cspan class=\"CitationRef\"\u003e57\u003c/span\u003e], energy company statistics, and national agency statistics. These original data sources were then harmonized to enable their use as input parameters for power system modeling. This approach produced four types of processed data files for each country: (i) a country file documenting nationally-aggregated variables, such as annual time series of electricity demand, transmission and distribution losses, and peak load; (ii) technology files describing techno-economic variables for each major generation technology in the country\u0026rsquo;s electricity mix, such as annual time series of investment costs, installed capacities, and actual generation; (iii) resource files describing \u003cem\u003eCO\u003c/em\u003e\u003csub\u003e2\u003c/sub\u003e emissions intensities and annual time series of costs for each generation fuel or input resource; and (iv) load profiles describing 24-hour national load curves for each available year. The data package includes an annotated list of all used original data sources, describes harmonization and processing steps for each variable, and provides the final processed data files in comma-separated format. Other data used and sources are described in the Supplementary Table \u003cspan class=\"InternalRef\"\u003eS1\u003c/span\u003e.\u003c/p\u003e\n\u003c/div\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e\n\u003ch2\u003e4.4 Hindcasting perfomance\u003c/h2\u003e\n\u003cdiv class=\"BlockQuote\"\u003e\n\u003cp\u003eFor each country, we ran hindcasting simulations using model versions with different combinations of societal factors included, following the model development methodology of Wen et al. [\u003cspan class=\"CitationRef\"\u003e25\u003c/span\u003e]. By each societal factor being either represented or not, if all data were available, we obtained 64 (2\u003csup\u003e6\u003c/sup\u003e) hindcasting simulations per country. For each hindcasting simulation, we quantified for different outputs (installed capacity, \u003cem\u003eCO\u003c/em\u003e\u003csub\u003e2\u003c/sub\u003e emissions and renewable generation share) the deviation between simulation outputs and historical values over the 1990\u0026ndash;2019 period. To do so, we first calculated the Symmetric Mean Absolute Percentage Error (SMAPE) over the years which is a relevant indicator to evaluate hindcasting performance in terms of the absolute magnitude of the deviations [58]. This indicator is also an absolute indicator without cancelation effect which allowed us to sum the values obtained over the different technologies for installed capacity to obtain an overall accuracy indicator. We then quantified the relative performance gain or loss brought by the inclusion of societal factors in the model compared to the techno-economic model version without societal factors included. We finally identified the societal factors that contribute the most to the variability of hindcasting performance. To do so, we performed analysis of variance (ANOVA) on hindcasting performance in order to compare the contributions to variance of societal factors and their interactions. We also captured the effect direction of each societal factor inclusion by comparing the average values of performance over the simulation subsets with and without the considered societal factor included [\u003cspan class=\"CitationRef\"\u003e59\u003c/span\u003e].\u003c/p\u003e\n\u003c/div\u003e\n\u003c/div\u003e"},{"header":"Declarations","content":"\u003ch2\u003eAcknowledgements\u003c/h2\u003e\n\u003cp\u003eThis work received funding from the Swiss National Science Foundation Eccellenza Grant for the project \u0026quot;Accuracy of long-range national energy projections\u0026quot; (Grant no. 186834, VFR, XW, ET) and from the Swiss State Secretariat for Education, Research, and Innovation (SEFRI) for the project \u0026ldquo;Net-zero pathway research through integrated assessment model advancements (PRISMA)\u0026rdquo; (Project no. 101081604, ET). We thank Bjarne Steffen and Friedemann Polzin for having provided the dataset of weighted average cost of capital.\u003c/p\u003e\n\u003ch2\u003eContribution\u003c/h2\u003e\n\u003cp\u003eMethodology, VFR, ET; Software, VFR; Investigation, VFR; Data curation, MJR, XW, VFR; Conceptualization, VFR, ET; Writing \u0026ndash; Original Draft, VFR; Writing \u0026ndash; Review \u0026amp; Editing, all authors; Visualization, VFR; Supervision \u0026ndash; ET; Funding Acquisition \u0026ndash; ET.\u003c/p\u003e\n\u003ch2\u003eCompeting interests\u003c/h2\u003e\n\u003cp\u003eThe authors declare no competing interests\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eM\u0026eacute;jean, A., Guivarch, C., Lef\u0026egrave;vre, J. \u0026amp; Hamdi-Cherif, M. The transition in energy demand sectors to limit global warming to 1.5 C. en. \u003cem\u003eEnergy Efficiency \u003c/em\u003e12, 441\u0026ndash;462. issn: 15706478. https://doi.org/10.1007/s12053-018-9682-0 (Feb. 2019).\u003c/li\u003e\n\u003cli\u003eJewell, J. \u0026amp; Cherp, A. On the political feasibility of climate change mitigation pathways: Is it too late to keep warming below 1.5C? en. \u003cem\u003eWIREs Climate Change \u003c/em\u003e11. _eprint: https://onlinelibrary.wiley.com/doi/pdf/10.1002/wcc.621, e621. issn: 1757-7799. https: //onlinelibrary.wiley.com/doi/abs/10.1002/wcc.621 (2020).\u003c/li\u003e\n\u003cli\u003eGeels, F. W., Sovacool, B. K., Schwanen, T. \u0026amp; Sorrell, S. Sociotechnical transitions for deep decarbonization. en. \u003cem\u003eScience \u003c/em\u003e357, 1242\u0026ndash;1244. issn: 0036-8075, 1095-9203. https: //www.science.org/doi/10.1126/science.aao3760 (Sept. 2017).\u003c/li\u003e\n\u003cli\u003ePianta, S. \u0026amp; Brutschin, E. Emissions Lock-in, Capacity, and Public Opinion: How Insights From Political Science Can Inform Climate Modeling Efforts. en. \u003cem\u003ePolitics and Governance \u003c/em\u003e10, 186\u0026ndash;199. issn: 2183-2463. https://www.cogitatiopress.com/politicsandgovernance/ article/view/5462 (Sept. 2022).\u003c/li\u003e\n\u003cli\u003eKrumm, A., S\u0026uuml;sser, D. \u0026amp; Blechinger, P. Modelling social aspects of the energy transition: What is the current representation of social factors in energy models? en. \u003cem\u003eEnergy \u003c/em\u003e239, 121706. issn: 0360-5442. https://www.sciencedirect.com/science/article/pii/ S036054422101954X (Jan. 2022).\u003c/li\u003e\n\u003cli\u003ePeng, W. \u003cem\u003eet al. \u003c/em\u003eClimate policy models need to get real about people \u0026mdash; here\u0026rsquo;s how. en. \u003cem\u003eNature \u003c/em\u003e594. Bandiera_abtest: a Cg_type: Comment Number: 7862 Publisher: Nature Publishing Group Subject_term: Climate change, Economics, Policy, Renewable energy, 174\u0026ndash;176. https://www.nature.com/articles/d41586-021-01500-2 (June 2021).\u003c/li\u003e\n\u003cli\u003eKeppo, I. \u003cem\u003eet al. \u003c/em\u003eExploring the possibility space: taking stock of the diverse capabilities and gaps in integrated assessment models. en. 16. Publisher: IOP Publishing, 053006. issn: 1748-9326. https://doi.org/10.1088/1748-9326/abe5d8 (Apr. 2021).\u003c/li\u003e\n\u003cli\u003eTrutnevyte, E. \u003cem\u003eet al. \u003c/em\u003eSocietal Transformations in Models for Energy and Climate Policy: The Ambitious Next Step. en. \u003cem\u003eOne Earth \u003c/em\u003e1, \u003csup\u003e423\u0026ndash;433. \u003c/sup\u003eissn: 2590-3322. https://www. sciencedirect.com/science/article/pii/S2590332219302246 (Dec. 2019).\u003c/li\u003e\n\u003cli\u003eHirt, L. F., Schell, G., Sahakian, M. \u0026amp; Trutnevyte, E. A review of linking models and socio-technical transitions theories for energy and climate solutions. en. \u003cem\u003eEnvironmental Innovation and Societal Transitions \u003c/em\u003e35, \u003csup\u003e162\u0026ndash;179. \u003c/sup\u003eissn: 2210-4224. https://www.sciencedirect.com/science/article/pii/S2210422420300368 (June 2020).\u003c/li\u003e\n\u003cli\u003eVan Sluisveld, M. A. E. \u003cem\u003eet al. \u003c/em\u003eAligning integrated assessment modelling with sociotechnical transition insights: An application to low-carbon energy scenario analysis in Europe. en. \u003cem\u003eTechnological Forecasting and Social Change \u003c/em\u003e151, 119177. issn: 0040-1625. https://www.sciencedirect.com/science/article/pii/S0040162518300593 (Feb. 2020).\u003c/li\u003e\n\u003cli\u003eBrutschin, E. \u003cem\u003eet al. \u003c/em\u003eA multidimensional feasibility evaluation of low-carbon scenarios. en. \u003cem\u003eEnvironmental Research Letters \u003c/em\u003e16. Publisher: IOP Publishing, 064069. issn: 1748-9326. https://doi.org/10.1088/1748-9326/abf0ce (June 2021).\u003c/li\u003e\n\u003cli\u003eSteffen, B. Estimating the cost of capital for renewable energy projects. en. \u003cem\u003eEnergy Economics \u003c/em\u003e88, 104783. issn: 0140-9883. https://www.sciencedirect.com/science/ article/pii/S0140988320301237 (May 2020).\u003c/li\u003e\n\u003cli\u003eVerrier, B., Li, P.-H., Pye, S. \u0026amp; Strachan, N. Incorporating social mechanisms in energy decarbonisation modelling. en. \u003cem\u003eEnvironmental Innovation and Societal Transitions \u003c/em\u003e45, 154\u0026ndash;169. issn: 2210-4224. https://www.sciencedirect.com/science/article/pii/ S2210422422000934 (Dec. 2022).\u003c/li\u003e\n\u003cli\u003eLi, F. G. N. Actors behaving badly: Exploring the modelling of non-optimal behaviour in energy transitions. en. \u003cem\u003eEnergy Strategy Reviews \u003c/em\u003e15, 57\u0026ndash;71. issn: 2211-467X. https://www.sciencedirect.com/science/article/pii/S2211467X17300020 (2021) (Mar. 2017).\u003c/li\u003e\n\u003cli\u003eKoecklin, M. T., Longoria, G., Fitiwi, D. Z., DeCarolis, J. F. \u0026amp; Curtis, J. Public acceptance of renewable electricity generation and transmission network developments: Insights from Ireland. en. \u003cem\u003eEnergy Policy \u003c/em\u003e151, 112185. issn: 0301-4215. https://www.sciencedirect. com/science/article/pii/S0301421521000549 (Apr. 2021).\u003c/li\u003e\n\u003cli\u003eStern, P. C., Dietz, T., Nielsen, K. S., Peng, W. \u0026amp; Vandenbergh, M. P. Feasible climate mitigation. en. \u003cem\u003eNature Climate Change \u003c/em\u003e13. Number: 1 Publisher: Nature Publishing Group, 6\u0026ndash;8. issn: 1758-6798. https://www.nature.com/articles/s41558-022-01563-7 (Jan. 2023).\u003c/li\u003e\n\u003cli\u003eBraunreiter, L., van Beek, L., Hajer, M. \u0026amp; van Vuuren, D. Transformative pathways \u0026ndash; Using integrated assessment models more effectively to open up plausible and desirable low-carbon futures. en. \u003cem\u003eEnergy Research \u0026amp; Social Science \u003c/em\u003e80, 102220. issn: 2214-6296. https://www.sciencedirect.com/science/article/pii/S2214629621003133 (Oct. 2021).\u003c/li\u003e\n\u003cli\u003eMoore, F. C. \u003cem\u003eet al. \u003c/em\u003eDeterminants of emissions pathways in the coupled climate\u0026ndash;social system. en. \u003cem\u003eNature \u003c/em\u003e603. Number: 7899 Publisher: Nature Publishing Group, 103\u0026ndash;111. issn: 1476-4687. https://www.nature.com/articles/s41586-022-04423-8 (Mar. 2022).\u003c/li\u003e\n\u003cli\u003eFreeman, R. Modelling the socio-political feasibility of energy transition with system dynamics. en. \u003cem\u003eEnvironmental Innovation and Societal Transitions \u003c/em\u003e40, 486\u0026ndash;500. issn: 2210-4224. https://www.sciencedirect.com/science/article/pii/S2210422421000800 (Sept. 2021).\u003c/li\u003e\n\u003cli\u003eLi, F. G. N. \u0026amp; Strachan, N. Take me to your leader: Using socio-technical energy transitions (STET) modelling to explore the role of actors in decarbonisation pathways. en. \u003cem\u003eEnergy Research \u0026amp; Social Science \u003c/em\u003e51, 67\u0026ndash;81. issn: 2214-6296. https://www.sciencedirect. com/science/article/pii/S2214629618306029 (May 2019).\u003c/li\u003e\n\u003cli\u003eLonergan, K. E. \u003cem\u003eet al. \u003c/em\u003eImproving the representation of cost of capital in energy system models. English. \u003cem\u003eJoule \u003c/em\u003e0. \u003csup\u003ePublisher: Elsevier. \u003c/sup\u003eissn: 2542-4785, 2542-4351. https://www. cell.com/joule/abstract/S2542-4351(23)00073-9 (Mar. 2023).\u003c/li\u003e\n\u003cli\u003eCreutzig, F. \u003cem\u003eet al. \u003c/em\u003eThe underestimated potential of solar energy to mitigate climate change. en. \u003cem\u003eNature Energy \u003c/em\u003e2. Number: 9 Publisher: Nature Publishing Group, 1\u0026ndash;9. issn: 20587546. https://www.nature.com/articles/nenergy2017140 (Aug. 2017).\u003c/li\u003e\n\u003cli\u003eGilbert, A. Q. \u0026amp; Sovacool, B. K. Looking the wrong way: Bias, renewable electricity, and energy modelling in the United States. en. \u003cem\u003eEnergy \u003c/em\u003e94, 533\u0026ndash;541. issn: 0360-5442. https://www.sciencedirect.com/science/article/pii/S0360544215015133 (Jan. 2016).\u003c/li\u003e\n\u003cli\u003eWilson, C. \u003cem\u003eet al. \u003c/em\u003eEvaluating process-based integrated assessment models of climate change mitigation. en. \u003cem\u003eClimatic Change \u003c/em\u003e166, 3. issn: 1573-1480. https://doi.org/10.1007/ s10584-021-03099-9 (May 2021).\u003c/li\u003e\n\u003cli\u003eWen, X., Jaxa-Rozen, M. \u0026amp; Trutnevyte, E. Hindcasting to inform the development of bottom-up electricity system models: The cases of endogenous demand and technology learning. en. \u003cem\u003eApplied Energy \u003c/em\u003e340, 121035. issn: 0306-2619. https://www.sciencedirect. com/science/article/pii/S0306261923003999 (June 2023).\u003c/li\u003e\n\u003cli\u003eGlotin, D., Bourgeois, C., Giraudet, L.-G. \u0026amp; Quirion, P. Prediction is difficult, even when it\u0026rsquo;s about the past: A hindcast experiment using Res-IRF, an integrated energy-economy model. en. \u003cem\u003eEnergy Economics. Eighth Atlantic Workshop on Energy and Environmental Economics \u003c/em\u003e84, 104452. issn: 0140-9883. https://www.sciencedirect.com/science/ article/pii/S0140988319302336 (Oct. 2019).\u003c/li\u003e\n\u003cli\u003eTrutnevyte, E. Does cost optimization approximate the real-world energy transition? en. \u003cem\u003eEnergy \u003c/em\u003e106, 182\u0026ndash;193. issn: 0360-5442. https://www.sciencedirect.com/science/ article/pii/S0360544216302821 (July 2016).\u003c/li\u003e\n\u003cli\u003ePolzin, F. \u003cem\u003eet al. \u003c/em\u003eThe effect of differentiating costs of capital by country and technology on the European energy transition. en. \u003cem\u003eClimatic Change \u003c/em\u003e167, 26. issn: 1573-1480. https: //doi.org/10.1007/s10584-021-03163-4 (2021) (July 2021).\u003c/li\u003e\n\u003cli\u003eCotterman, T., Small, M. J., Wilson, S., Abdulla, A. \u0026amp; Wong-Parodi, G. Applying risk tolerance and socio-technical dynamics for more realistic energy transition pathways. en. \u003cem\u003eAp\u003c/em\u003e\u003cem\u003eplied Energy \u003c/em\u003e291, 116751. issn: 0306-2619. https://www.sciencedirect.com/science/ article/pii/S0306261921002609 (June 2021).\u003c/li\u003e\n\u003cli\u003eNicolli, F. \u0026amp; Vona, F. Energy market liberalization and renewable energy policies in OECD countries. en. \u003cem\u003eEnergy Policy \u003c/em\u003e128, \u003csup\u003e853\u0026ndash;867. \u003c/sup\u003eissn: 0301-4215. https://www. sciencedirect.com/science/article/pii/S0301421519300187 (May 2019).\u003c/li\u003e\n\u003cli\u003eSteffen, B., Karplus, V. \u0026amp; Schmidt, T. S. State ownership and technology adoption: The case of electric utilities and renewable energy. en. \u003cem\u003eResearch Policy \u003c/em\u003e51, 104534. issn: 0048-7333. https://www.sciencedirect.com/science/article/pii/S0048733322000610 (July 2022).\u003c/li\u003e\n\u003cli\u003ePriesmann, J., Nolting, L. \u0026amp; Praktiknjo, A. Are complex energy system models more accurate? An intra-model comparison of power system optimization models. en. \u003cem\u003eApplied Energy \u003c/em\u003e255, 113783. issn: 0306-2619. https://www.sciencedirect.com/science/ article/pii/S0306261919314709 (Dec. 2019).\u003c/li\u003e\n\u003cli\u003eMoret, S., Babonneau, F., Bierlaire, M. \u0026amp; Mar\u0026eacute;chal, F. Overcapacity in European power systems: Analysis and robust optimization approach. en. \u003cem\u003eApplied Energy \u003c/em\u003e259, 113970. issn: 0306-2619. https://www.sciencedirect.com/science/article/pii/S0306261919316575 (Feb. 2020).\u003c/li\u003e\n\u003cli\u003eSchwanitz, V. J. Evaluating integrated assessment models of global climate change. en. \u003cem\u003eEnvironmental Modelling \u0026amp; Software \u003c/em\u003e50, \u003csup\u003e120\u0026ndash;131. \u003c/sup\u003eissn: 1364-8152. https://www. sciencedirect.com/science/article/pii/S1364815213001965 (Dec. 2013).\u003c/li\u003e\n\u003cli\u003eOreskes, N. Evaluation (Not Validation) of Quantitative Models. \u003cem\u003eEnvironmental Health \u003c/em\u003e\u003cem\u003ePerspectives \u003c/em\u003e106. Publisher: [National Institute of Environmental Health Sciences, Brogan \u0026amp; Partners], 1453\u0026ndash;1460. issn: 0091-6765. https://www.jstor.org/stable/4641214 (1998).\u003c/li\u003e\n\u003cli\u003eMuttitt, G., Price, J., Pye, S. \u0026amp; Welsby, D. Socio-political feasibility of coal power phaseout and its role in mitigation pathways. en. \u003cem\u003eNature Climate Change \u003c/em\u003e13. Number: 2 Publisher: Nature Publishing Group, 140\u0026ndash;147. issn: 1758-6798. https://www.nature.com/ articles/s41558-022-01576-2 (Feb. 2023).\u003c/li\u003e\n\u003cli\u003eHarmsen, M. \u003cem\u003eet al. \u003c/em\u003eIntegrated assessment model diagnostics: key indicators and model evolution. en. \u003cem\u003eEnvironmental Research Letters \u003c/em\u003e16. Publisher: IOP Publishing, 054046. issn: 1748-9326. https://dx.doi.org/10.1088/1748-9326/abf964 (May 2021).\u003c/li\u003e\n\u003cli\u003eKrey, V. \u003cem\u003eet al. \u003c/em\u003eLooking under the hood: A comparison of techno-economic assumptions across national and global integrated assessment models. en. \u003cem\u003eEnergy \u003c/em\u003e172, 1254\u0026ndash;1267. issn: 0360-5442. https://www.sciencedirect.com/science/article/pii/S0360544218325039 (Apr. 2019).\u003c/li\u003e\n\u003cli\u003eDeCarolis, J. F., Hunter, K. \u0026amp; Sreepathi, S. The case for repeatable analysis with energy economy optimization models. en. \u003cem\u003eEnergy Economics \u003c/em\u003e34, 1845\u0026ndash;1853. issn: 0140-9883. https://www.sciencedirect.com/science/article/pii/S0140988312001405 (Nov. 2012).\u003c/li\u003e\n\u003cli\u003eVinichenko, V., Vetier, M., Jewell, J., Nacke, L. \u0026amp; Cherp, A. Phasing out coal for 2 C target requires worldwide replication of most ambitious national plans despite security and fairness concerns. \u003cem\u003eEnvironmental Research Letters \u003c/em\u003e(2023).\u003c/li\u003e\n\u003cli\u003eCherp, A., Vinichenko, V., Tosun, J., Gordon, J. A. \u0026amp; Jewell, J. National growth dynamics of wind and solar power compared to the growth required for global climate targets. \u003cem\u003eNature Energy \u003c/em\u003e6. Publisher: Nature Publishing Group UK London, 742\u0026ndash;754 (2021).\u003c/li\u003e\n\u003cli\u003eGambhir, A. Powering past coal is not enough. en. \u003cem\u003eNature Climate Change \u003c/em\u003e13. Number: 2 Publisher: Nature Publishing Group, \u003csup\u003e117\u0026ndash;118. \u003c/sup\u003eissn: 1758-6798. https://www.nature. com/articles/s41558-022-01574-4 (Feb. 2023).\u003c/li\u003e\n\u003cli\u003eLi, F. G. N., Trutnevyte, E. \u0026amp; Strachan, N. A review of socio-technical energy transition (STET) models. en. \u003cem\u003eTechnological Forecasting and Social Change \u003c/em\u003e100, 290\u0026ndash;305. issn: 0040-1625. https://www.sciencedirect.com/science/article/pii/S0040162515002267 (Nov. 2015).\u003c/li\u003e\n\u003cli\u003eWise, M. \u003cem\u003eet al. \u003c/em\u003eRepresenting power sector detail and flexibility in a multi-sector model. en. \u003cem\u003eEnergy Strategy Reviews \u003c/em\u003e26, 100411. issn: 2211-467X. https://www.sciencedirect. com/science/article/pii/S2211467X1930104X (Nov. 2019).\u003c/li\u003e\n\u003cli\u003eDe Boer, H. S. ( S. ) \u0026amp; van Vuuren, D. ( P. ) Representation of variable renewable energy sources in TIMER, an aggregated energy system simulation model. en. \u003cem\u003eEnergy Economics \u003c/em\u003e64, 600\u0026ndash;611. issn: 0140-9883. https://www.sciencedirect.com/science/article/ pii/S0140988316303528 (May 2017).\u003c/li\u003e\n\u003cli\u003eDespr\u0026eacute;s, J. \u003cem\u003eet al. \u003c/em\u003eStorage as a flexibility option in power systems with high shares of variable renewable energy sources: a POLES-based analysis. en. \u003cem\u003eEnergy Economics \u003c/em\u003e64, 638\u0026ndash;650. issn: 0140-9883. https://www.sciencedirect.com/science/article/pii/ S0140988316300445 (May 2017).\u003c/li\u003e\n\u003cli\u003eFisch-Romito, V., Guivarch, C., Creutzig, F., Minx, J. C. \u0026amp; Callaghan, M. W. Systematic map of the literature on carbon lock-in induced by long-lived capital. en. \u003cem\u003eEnvironmental Research Letters \u003c/em\u003e16. \u003csup\u003ePublisher: IOP Publishing, 053004. \u003c/sup\u003eissn: 1748-9326. https://doi. org/10.1088/1748-9326/aba660 (Apr. 2021).\u003c/li\u003e\n\u003cli\u003eFouquet, R. Path dependence in energy systems and economic development. en. \u003cem\u003eNature Energy \u003c/em\u003e1, 16098. issn: 2058-7546. https://www.nature.com/articles/nenergy201698 (July 2016).\u003c/li\u003e\n\u003cli\u003eHirth, L. \u0026amp; Steckel, J. C. The role of capital costs in decarbonizing the electricity sector. en. \u003cem\u003eEnvironmental Research Letters \u003c/em\u003e11. Publisher: IOP Publishing, 114010. issn: 17489326. https://doi.org/10.1088/1748-9326/11/11/114010 (Nov. 2016).\u003c/li\u003e\n\u003cli\u003eEgli, F., Steffen, B. \u0026amp; Schmidt, T. S. A dynamic analysis of financing conditions for renewable energy technologies. en. \u003cem\u003eNature Energy \u003c/em\u003e3. Number: 12 Publisher: Nature Publishing Group, 1084\u0026ndash;1092. issn: 2058-7546. https://www.nature.com/articles/s41560-0180277-y (Dec. 2018).\u003c/li\u003e\n\u003cli\u003eCousse, J., W\u0026uuml;stenhagen, R. \u0026amp; Schneider, N. Mixed feelings on wind energy: Affective imagery and local concern driving social acceptance in Switzerland. en. \u003cem\u003eEnergy Research \u0026amp; Social Science \u003c/em\u003e70, 101676. issn: 2214-6296. https://www.sciencedirect.com/science/ article/pii/S2214629620302516 (Dec. 2020).\u003c/li\u003e\n\u003cli\u003eRuhnau, O., Eicke, A., Sgarlato, R., Tr\u0026ouml;ndle, T. \u0026amp; Hirth, L. Cost-Potential Curves of Onshore Wind Energy: the Role of Disamenity Costs. en. \u003cem\u003eEnvironmental and Resource Economics. \u003c/em\u003eissn: 1573-1502. https://doi.org/10.1007/s10640-022-00746-2 (Nov. 2022).\u003c/li\u003e\n\u003cli\u003eHarold, J., Bertsch, V., Lawrence, T. \u0026amp; Hall, M. Drivers of People\u0026rsquo;s Preferences for Spatial Proximity to Energy Infrastructure Technologies: A Cross-country Analysis. \u003cem\u003eThe Energy Journal \u003c/em\u003e42. issn: 01956574. http://www.iaee.org/en/publications/ejarticle. aspx?id=3703 (Oct. 2021).\u003c/li\u003e\n\u003cli\u003eStokes, L. C. \u0026amp; Warshaw, C. Renewable energy policy design and framing influence public support in the United States. en. \u003cem\u003eNature Energy \u003c/em\u003e2. Number: 8 Publisher: Nature Publishing Group, 1\u0026ndash;6. issn: 2058-7546. https://www.nature.com/articles/nenergy2017107 (June 2017).\u003c/li\u003e\n\u003cli\u003eOECD. Product Market Regulation Network Sectors indicators: Methodology for calculating the 1975-2018 time series. en (2018).\u003c/li\u003e\n\u003cli\u003eJaxa-Rozen, M., Wen, X. \u0026amp; Trutnevyte, E. Historic data of the national electricity system transitions in Europe in 1990\u0026ndash;2019 for retrospective evaluation of models. en. \u003cem\u003eData in Brief \u003c/em\u003e43, 108459. issn: 2352-3409. https://www.sciencedirect.com/science/article/pii/ S2352340922006540 (Aug. 2022).\u003c/li\u003e\n\u003cli\u003eEurostat. Eurostat Dissemination Database. Publications Office of the European Union. https://ec.europa.eu/eurostat/data/database (2023).\u003c/li\u003e\n\u003cli\u003eWen, X., Jaxa-Rozen, M. \u0026amp; Trutnevyte, E. Accuracy indicators for evaluating retrospective performance of energy system models. en. \u003cem\u003eApplied Energy \u003c/em\u003e325, 119906. issn: 0306-2619. https://www.sciencedirect.com/science/article/pii/S0306261922011667 (Nov. 2022).\u003c/li\u003e\n\u003cli\u003eGuivarch, C. \u0026amp; Monjon, S. Identifying the main uncertainty drivers of energy security in a low-carbon world: The case of Europe. en. \u003cem\u003eEnergy Economics \u003c/em\u003e64, 530\u0026ndash;541. issn: 0140-9883. https://www.sciencedirect.com/science/article/pii/S014098831630086X (May 2017).\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"University of Geneva","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"nature-portfolio","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"","title":"Nature Portfolio","twitterHandle":"","acdcEnabled":false,"dfaEnabled":false,"editorialSystem":"ejp","reportingPortfolio":"","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"societal transformation dynamics, energy systems modelling, integrated assessment modelling, socio-technical transitions, hindcasting","lastPublishedDoi":"10.21203/rs.3.rs-4312891/v3","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-4312891/v3","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eIntegrated assessment and energy system models are challenged to account for societal transformation dynamics to produce feasible low-carbon pathways. Yet, empirical evidence is lacking on which factors should be incorporated, how and to what extent this would improve the quality and relevance of modeled pathways. Here, we include six societal factors related to (i) infrastructure dynamics, (ii) actors and decision making and (iii) societal and institutional context into an open-source simulation model of the national power system transition. We apply this model for 31 European countries and, using hindcasting (1990\u0026ndash;2019), quantify which societal factors improved the modeled pathways. We find that, if well-chosen and in most cases, incorporating societal factors can improve the hindcasting performance by up to 24% in terms of modelled installed capacity of individual technologies, but there are also situations where hindcasting performance can become worse. The combinations of most relevant societal factors differ among countries and model outputs, but infrastructure lock\u003cem\u003e-\u003c/em\u003ein, public acceptance and investment risks contribute more strongly and frequently to model performance improvement. Our study hence paves the road to evidence-based choice of societal factors to be included in energy transition modeling in a systematic and transparent way.\u003c/p\u003e","manuscriptTitle":"Multi-country evidence on societal factors to include in energy transition modeling","msid":"","msnumber":"","nonDraftVersions":[{"code":3,"date":"2024-05-16 17:46:17","doi":"10.21203/rs.3.rs-4312891/v3","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"nature-energy","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"nenergy","sideBox":"Learn more about [Nature Energy](http://www.nature.com/nenergy/)","snPcode":"","submissionUrl":"","title":"Nature Energy","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature Research","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"20e8de00-2d50-4b93-8eca-3e0b91741c08","owner":[],"postedDate":"May 16th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":31960617,"name":"Scientific community and society/Energy and society/Energy supply and demand"},{"id":31960618,"name":"Earth and environmental sciences/Environmental social sciences/Socioeconomic scenarios"},{"id":31960619,"name":"Earth and environmental sciences/Environmental social sciences/Climate-change mitigation"},{"id":31960620,"name":"Scientific community and society/Energy and society/Energy economics"}],"tags":[],"updatedAt":"2025-02-22T08:05:21+00:00","versionOfRecord":{"articleIdentity":"rs-4312891","link":"https://doi.org/10.1038/s41560-025-01719-7","journal":{"identity":"nature-energy","isVorOnly":false,"title":"Nature Energy"},"publishedOn":"2025-02-21 05:00:00","publishedOnDateReadable":"February 21st, 2025"},"versionCreatedAt":"2024-05-16 17:46:17","video":"","vorDoi":"10.1038/s41560-025-01719-7","vorDoiUrl":"https://doi.org/10.1038/s41560-025-01719-7","workflowStages":[]},"version":"v3","identity":"rs-4312891","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-4312891","identity":"rs-4312891","version":["v3"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}