Energy system models are mathematical, computer-based models that represent energy systems at different scales. They are used to simulate and explore future pathways for energy systems, demand and supply trajectories, and implications for markets, policy, emissions, and energy security [1–4]. As simplified depictions of reality, they support decision-making processes by informing policymakers about different energy futures and helping them to evaluate different timelines and policy options regarding investment and technology [1, 5, 6].
Following the oil crisis of the 1970s, as well as increasing climate awareness and technological developments in the 1980s and 1990s, many energy models have been developed in and applied to high-income countries (HICs) [7–10], as well as Latin American countries [11–13]. Since the 2015 Paris Agreement, signatories have been required to develop national climate commitments, including Nationally Determined Contributions (NDCs) and long-term strategies. As a result, energy models are increasingly being applied in further low-and-middle-income countries (LMICs) [14, 15]. For example, energy modelling studies have been developed for Ghana [16], Pakistan [17], Indonesia [18] and Ethiopia [19]. However, many models, having been developed in HICs, and transferred to LMICs, are biased towards HIC contexts and do not account for the unique characteristics of LMICs, such as different costs of capital, regulatory structures, low levels of access to electricity, high usage of traditional biomass, political instability, informal economies, and higher climate vulnerability [9, 10, 19–24]. Furthermore, many models focus on reducing emissions in the energy sector, which may not be a priority for LMICs with low baseline emissions and low levels of electricity access [5, 24].
In HICs, energy modelling frameworks[1] such as MESSAGE, TIMES, and LEAP have become an integral part of policymaking processes [25–28]. Some studies have examined and evaluated the connection between energy modelling and policymaking in these contexts [1, 29–31]. The ways in which policy processes draw on models are versatile and differ between jurisdictions [1]. While energy modelling is recognised as a useful input into decision making, it is regularly criticised for falling short of capturing qualitative, socio-economic and policy features which shape energy systems [32][33]. Hence, modelling must not be the only input into policymaking, but is rather to be seen as informing decision-making alongside other factors. Many energy modelling projects are further criticised for a lack of inclusion of non-academic stakeholders, indicating a lack of ‘process democratisation’ (p1) [32].
In LMICs, the connection between energy modelling and policymaking remains largely unexplored. Existing studies on energy modelling in LMICs focus on more general issues, such as data availability and quality, as well as low capacity for energy modelling [5, 34]. Many LMICs lack capacity to draw on energy modelling for decision-making due to limited human and financial resources [6]. A longstanding concern in policymaking has been a lack of institutional capacity to use and engage with energy modelling [20, 35–37]. Thus, it is often the case that modellers from HICs lead or support modelling projects in LMICs, building on their comparatively longer experience. For example, Musonye et al. (2020) found that most energy modelling projects in the Sub-Saharan African (SSA) region were led by European institutions [38]. Likewise, Abrahams et al. (2023) point out that the majority of energy modelling research on Africa was conducted by non-Africa-based authors [15]. This dynamic of external leadership has been criticised for lacking context-specific knowledge and sensitivity and for neglecting the different socio-economic dynamics of countries [5, 20, 39]. For example, global integrated assessment models, such as the ones drawn on by the Intergovernmental Panel on Climate Change, tend to poorly represent African countries, for instance by pooling together Africa and the Middle East as a region [5]. As a result, calls for national leadership on modelling projects and the strengthening of in-country modelling capacity are growing. For Africa in particular, Mulugetta et al. [34] find that capacities for energy system planning differ widely across countries, with some nations being able to base their energy transition decisions on model outputs, while others have only limited capacity to draw on the insights from modelling. The authors call for better and more evidence-based decision making in every African country – an essential requirement given that many nations are making decisions regarding their future energy systems. Generally, the increase and institutionalisation of modelling capacity in LMICs are seen as important for integrating modelling in decision-making and accelerating energy transitions [20, 34, 35].
As the use of energy system models in planning in LMICs grows, the sparsely researched dynamics between energy modelling and policymaking in LMICs warrants further investigation, in particular: the influence of political economy on modelling processes; engagement between modellers, policymakers and other relevant stakeholders; and the ways in which modelling tools are applied to policy questions. Hence, this research sought to address the following questions: 1) what is the status-quo of the use of energy system modelling in policymaking in LMICs? 2) What factors, such as stakeholder interests and engagement, shape modelling processes? 3) And, vice versa, how do modelling results shape policymaking in LMICs?
To answer the research questions, this study drew on 35 semi-structured interviews with energy modellers and policymakers working for institutions such as universities, ministries, utilities, think tanks, international development institutions, and energy companies. The interviews explored participants’ experiences of developing and using energy models in LMICs and, in so doing, contributes to a better understanding of how energy modelling is currently used in policymaking in LMICs, including potential challenges and how they can be addressed. The research was conducted as part of the Climate Compatible Growth (CCG) programme. Funded by the UK Foreign, Commonwealth and Development Office, CCG is a £38 million programme that runs from 2021–2025. CCG aims to support investment in sustainable energy and transport systems to meet development priorities in LMICs. A core focus of CCG is energy modelling, and the views of CCG researchers have substantially influenced the design and objectives of the research programme.
Throughout this research, energy system modellers are referred to as actors who use and work with energy system models on local, national and regional scales, make decisions on the input and assumptions of the model, and perform the modelling exercise [40]. Externals are defined as non-national actors, while nationals are in-country actors. Meanwhile, policymakers are actors involved in the design and implementation of energy-related policies with the opportunity to draw on modelling results to inform their decision-making process. However, sometimes, the two roles cannot be distinguished very clearly [40], for example when modellers work in the government and make or support planning-related decisions. Furthermore, the term LMICs is used while recognising that it is a blanket term describing a huge diversity of country contexts [41]. However, the term is used considering that, in the global context, energy models were firstly applied predominantly in HICs and their more recent use in many LMICs demands the acknowledgment of different starting points and circumstances [24]. While acknowledging this, the research simultaneously aims to carve out and make visible the diversity of contexts regarding capacities, perceptions, interests and influences among LMICs.
The rest of this work is structured as follows. Section 2 describes the materials and methods used to address the research questions, including document analysis and semi-structured interviews with modellers and policymakers. Section 3 presents and discusses the results regarding the role of modelling in policymaking, key factors which shape modelling processes, and ways in which modelling influences policymaking. Finally, Section 4 draws out key conclusions and outlines some areas for further research.