Demanding Power: Do Protests Empower Citizens to Hold Governments Accountable over Energy?

Energy protests are becoming increasingly common and significant around the world. While in the global North concerns tend to centre around climate issues, in the global South the concerns are more often with affordable energy. Both types of protests, however, have one issue in common: the undemocratic nature of energy policymaking. This paper draws together findings from research conducted in three countries, Mozambique, Nigeria, and Pakistan to ask how and under which conditions do struggles over energy access in fragile and conflict affected settings empower the powerless to hold public authorities to account? In exploring this theme, the study examines what factors support protests developing into significant episodes of contention within fragile settings, and whether these energy struggles promote citizen empowerment and institutional accountability.


Executive summary
Energy protests have been recurrent and significant political events in the twenty-first century. Major episodes have erupted in rich countries and poor; in places with strong and stable states (France, Chile), as well as where states are fragile and conflict common (Mozambique, Haiti); in energy exporting countries, and those dependent on imports. There have been localised protests about electricity, and urban uprisings about fuel prices or transport fares. Some countries have had both these types of protest as well as others, including around new energy deals and infrastructure and transnational campaigns against fossil fuels.
Why do people protest about energy, and what do their protests achieve? Despite their recent prominence, analysis of energy protests is in its infancy. This paper presents findings of research into this phenomenon. The interdisciplinary research synthesised here involved close-grained studies of energy protests in Mozambique, Nigeria, and Pakistan, which were situated within their broader global context through cross-country analysis. The research asked: what political and economic conditions drive such protests? Do major episodes empower people to hold their governments accountable for energy provisioning? How do external and transnational interests shape the conditions in which mass protests unfold? 1 The research was primarily motivated by debates within contentious politics and social movement studies as they applied to governance and citizen participation in developing and fragile states. Energy protests provided a lens through which to understand the political opportunities for citizens to act collectively in the fragile and conflict-affected settings common to many developing countries. They enabled an examination of how popular grievances come to inform policy and politics in settings where political and civic space are captured or closed. The findings also speak to debates about 'energy justice' and how it might be achieved, which underscore the importance of civic participation, transparency and accountability in progress towards 'just transitions'. Finally, understanding energy protests is also important for research, policy, and practice on fossil fuel subsidies and their reform. Mass protests about fuel price rises have frequently blocked or stalled efforts to eradicate subsidies seen as pro-rich and fiscally unsustainable. The research here reads these protests as signals that: a) energy is a basic need for modern life; and b) citizens lack more formal channels through which to voice their energy grievances. 1 The research found that protests were often triggered by grievances where formal channels -political parties, civil society, independent media -were blocked or weak. But while political fragility is linked to energy protests, protest episodes can create or contribute to fragility, depending on how governments respond. Energy provisioning of some kind was seen as part of the social contract in the three countries, and this provided a justification for protests demanding action to protect energy access. Yet there was limited evidence that the raw power of energy protests empowered citizens with respect to energy policy. Meaningful gains in citizen power were short-lived or diffuse, dissipating without institutionalised citizen engagement or lasting organisations. Major protest episodes taught protestors that en masse, they could force the government to hear their complaints. But they rarely elicited more than a heated public hearing, followed by often temporary policy change. In Nigeria, the 2012 episodes reinforced the threat of mass protests against unpopular energy policies, succeeding not in pushing subsidy reform off the table, but in encouraging subsidy reformers to develop more tactical approaches to introducing reforms. In Mozambique, the food and fuel protests of 2008 and 2010 taught the state to deploy extreme violence against protestors, to dampen any nascent struggles.
Nor did energy protests lead to lasting or institutionalised public accountability to citizens, whether for energy or for any other essential public services. In the more disruptive episodes of protest, governments backed down on proposed price rises or introduced short-term fixes for electricity shortages. State tactics were to ward off protests, but rarely involved a visible strategy of strengthening public accountability for energy policy. In Pakistan, hundreds of protests against erratic electricity supply forced it onto the democratic political agenda; but electricity policymaking remains opaque, elitist, and technocratic, and no spaces or institutional practices were created to give citizens voice over their power supply.
Protestors' fleeting sense of power fuelled a self-reinforcing cycle: protestors believed it was all they could do, and governments feared protests but expected resistance to dissipate in time. Political parties in Pakistan and Nigeria took account of protestors' -and the broader public's -grievances about energy to some extent. But no political party in power had strong incentives to abandon either fuel subsidy reforms or energy distribution systems in which powerful interests are invested, and there were no innovations with respect to how to address citizens' energy grievances beyond violence or short-term appeasement. While governments offer little that is new, citizens' repertoires of protest have innovated with the rise of digital and social media, finding new ways of framing narratives and mobilising supporters.
Transnational actors are part of this landscape of energy protests, but their roles can be hidden or obscure. Multilateral aid actors influence subsidy reform, but less often through the 'hard' conditionality of structural adjustment than by shaping the intellectual agenda and policy pathways. Multinationals and foreign governments take a direct hand in national energy policy in the global South; these well-resourced powerful interests help ensure energy policymaking remains closed and secretive. Aid donor support for civil society and good governance has rarely supported popular domestic movements for climate justice or energy security, while transnational climate movements focus on cutting fossil fuel use and play no role in popular protests against energy price rises and scarcity in the global South. While energy price protests in the global South seem at odds with activism against fossil fuels in the North, both share a concern with the need for civic participation, transparency, and accountability in energy policymaking.
There are two main policy and practice implications of this research. First, these protests signal that energy security is shaping up to be a wicked problem: it involves many competing interests, there is no 'correct' solution, and information about the issue is always incomplete (Rittel and Webber 1973). Demands for reliable, affordable access to fuel and energy (the definition of energy security, according to the International Energy Agency (IEA n.d.) will continue to rise, while there is increasing pressure to reduce the use of fossil fuels to minimise the impact on climate change. Efforts to reform fuel subsidies in the interests of reducing carbon emissions and improving equity have also been gathering pace, but subsidy reforms have proven resistant to satisfactory resolution, in part because of policy elite fears of mass protests against higher-priced energy. Such mobilisation expresses deeply felt grievances that lack other civil channels of expression or representation. At the same time, states appear slow or unable to deliver acceptable alternatives to subsidies. Governments in fragile, conflict-and violence-affected settings have not been able to provide legitimate and acceptable social protection buffers that alleviate citizens' anger and protests over the costs of steep fuel prices hikes. The 'negotiations' between citizens and state are costly, often violent, and ultimately unproductive (subsidy cuts are stalled, public spending remains inequitable or inefficient, citizens still do not get the energy they need at prices they can afford). Even 'successful' protest episodes yield little or no mechanisms of accountability to citizens for energy policy. Civic activism around energy security is in its infancy, where found at all. In the three countries we studied closely, energy policy has been taken up by political parties/regimes, but some of the most marginalised and disempowered urban populations have benefited least from the politicisation of energy.
The second key implication is that it is necessary to address institutional failures to hear and respond to citizens' energy concerns. There is no guarantee that ids.ac.uk Working Paper Volume 2021 Number 555 Demanding Power: Do Protests Empower Citizens to Hold Governments Accountable over Energy? 10 improving citizen participation in energy policymaking will resolve this wicked problem. But enabling citizens to voice their energy grievances through nonriotous means is an entry point to date untried in countries with large fossil fuel subsidies and a history of mass protest. While much donor and development policy and many global treaties encourage countries to transition away from fossil fuels, where energy is fossil dependent, the current political bargain around energy prices may make energy transitions contentious. A key part of the solution will be to come up with affordable, accessible, and sustainable energy distribution that moves beyond the current impasse. Governments that have shown willing to deploy armed forces against energy protestors are unlikely to willingly create spaces for citizen participation, nor do transnational actors have innate incentives to do so. Global food policy provides an instructive example of how strong social movements with national and global force have been able to push for and create space in food and nutrition policy processes for citizen voice. As energy becomes more vital to everyday life in modern cities, livelihoods and lives will depend ever more on access to modern forms of energy. Creating global and national policy space for social movements and civic activism is likely to be crucial for progress towards energy justice.

Introduction: Why energy protests matter
This paper summarises research into how and why energy protests emerge and escalate, whether they empower citizens to hold their governments to account for energy provisioning, and the role played by transnational actors in the policy processes associated with these protests. Research into protests against energy price rises and scarcity is of increasing importance because such protests have been frequent and potent in the twenty-first century; they signal the importance of energy as a matter of rights and justice. They may also block fossil fuel subsidy reforms, and so are of particular interest to aid agencies and activists campaigning for cuts. The research on which this synthesis paper is based centres on the perspectives of citizens and protestors, attempting to understand why people take to the streets en masse, and at personal risk, to protest what they see as unjust or unworkable energy pricing or supply policies. The research helps to explain why energy protests have become such an important form of contemporary contention; points towards the weaknesses in energy governance that make such contention increasingly likely; and identifies strategies for addressing energy contention to enable just and sustainable energy transitions that go beyond the technocratic policy interventions currently favoured.

Background and motivations
Energy transitions and green transformations are as political as they are technical and scientific processes (Sovacool 2012;Scoones et al. 2015;Sovacool et al. 2016;Newell 2021). Recent protests illustrate why: mass protests and movements have proven potent opponents of fossil fuel and other energy projects (Temper et al. 2020), and have impeded fossil fuel subsidy reforms (Ross et al. 2017;Lockwood 2015 These protests matter in part because they are part of a 'rich ecology of resistance' that will determine the pathways of energy transitions (Newell 2021: 172). Activism against fossil fuel use, mostly by environmental movements concentrated in the global North, and protests against fuel subsidy cuts or tax rises, mainly in the global South, appear to counter each other, with one side against fossil fuels and the other fighting for their rights to use them. But this apparent opposition overlooks important commonalities: that people across the world are actively demanding to shape their energy supply, and to do so in ways that are just and sustainable (Newell 2021). A growing literature treats popular mobilisation around energy as an important feature of the landscape of climate change policy and energy transitions (Hanna et al. 2016), recognising that 'citizens worldwide are seeking more active political engagement regarding what their energy systems and climate futures will look like' (Temper et al. 2020: 19; see also: Roddis et al. 2018). Energy justice is understood to include not only a fair distribution of energy use, but also procedural justice, or that people are able to participate in decisions about energy policy, with transparency and public accountability for the choices made (Jenkins et al. 2016;Jenkins et al. 2018;Sovacool and Dworkin 2015). The Intergovernmental Panel on Climate Change (IPCC) 2018 special report on global warming of 1.5ºc found that: Identifying and navigating inclusive and socially acceptable pathways towards low-carbon, climate-resilient futures is a challenging yet important endeavour, fraught with moral, practical and political difficulties and inevitable trade-offs [which entail] deliberation and problem-solving processes to negotiate societal values, well-being, risks and resilience and to determine what is desirable and fair, and to whom. (IPCC 2018: 448-49) A critical challenge for climate change policy is that for many people around the world, protest appears to be the best means of engaging in 'deliberation and problem-solving processes' about energy policy, because it is their 'societal values' and 'wellbeing' which are most directly at stake. Activism against fossil fuel use has triggered a series of protests and larger uprisings in the past few years, particularly in the global North, and fossil fuel subsidy reform is now considered an 'emerging global norm' of climate change mitigation efforts (Skovgaard and Asselt 2019: 2). Nevertheless, the policy arguments for fossil fuel subsidy cuts in the global South have emphasised their fiscal -rather than environmental -effects; in contexts where governments have never to date delivered effective social protection, consumer subsidy cuts are perceived as a net loss of welfare  The research findings synthesised here aim to improve our knowledge about energy protests. The starting point was not to inform climate change policies or fossil fuel subsidy reform agendas, but to centre protestors' perspectives (Hossain et al. 2018). The Action for Empowerment and Accountability (A4EA) programme, of which this project was part, focused on countries where political authority was fragile or fragmented, and violence and conflict were endemic and characteristic of citizen-state relations. The prominence of energy contention, in the absence of other important episodes of collective action, suggested that fragility and conflict in some way enabled the escalation of energy protests. In settings where other forms of political action were proscribed, dangerous, or ineffectual, large collectivities had protested energy policies, sometimes facing violence, but occasionally succeeding in reversing policies and influencing political power.
Comparable to food riots in how they dramatise grievances about basic needs, fuel riots draw attention to the relationship between citizens and their state, suggesting that how societies resolve the politics of energy provisioning may be formative of the citizen-state relationship (Tilly 1975). In other words, the research on energy protests was motivated not only by its potential to explain the politics of energy transitions, but also because the politics of energy transitions sheds light on the nature of politics more broadly, drawing attention to bigger questions about the nature of the relationship between citizens and the states mandated to ensure access to the modern forms of energy citizens need for daily life. In the contexts of a global climate crisis and transnational interests in energy and finance, the citizen-state relationship may be structured as much by international as national dynamics.
With these concerns in mind, the research synthesised in this paper asked: How and under which conditions do struggles over energy access in fragile and conflict affected settings empower the powerless to hold public authorities to account (over energy or more broadly, over other issues)? Specifically, we explored: − How and why energy protests emerged and escalated into larger political contention.
− Whether they empowered citizens to hold their governments to account for energy provisioning.
− The roles of transnational actors in these struggles.
This paper summarises findings from the mixed-method, multi-sited, comparative study that was designed to test relevant propositions. The study combined cross-

Key findings
The research found that protests triggered by grievances about energy acquired a particular potency in settings where more institutionalised and formal channels for expressing grievances -programmatic political parties, organised civil society, independent media -were absent, weak, or blocked. Nevertheless, energy price and scarcity protests have occurred and escalated across types of political and energy regimes, levels of development, and regions. While economic and political conditions in fragile states may create opportunities for such protests to escalate, the research also indicates that energy protests can create or contribute to political fragility, depending on how governments respond to them.
Yet there was little evidence from the detailed country case studies that the raw power of energy protests translated into sustained empowerment of citizens with respect to energy policy, or more broadly, in their relations to the state. 3 There is evidence that (political) power was perceived to have been gained by protest movements or by 'the people', broadly speaking. But the evidence also indicated that these gains of power could be subverted, turned to particularistic advantage, or subsumed within party political competition. In Pakistan, for instance, access to electricity was forced onto the democratic political agenda by hundreds of protests over what was seen as an intolerable failure of the state to provide a stable electricity supply; but electricity policymaking remains opaque, elitist, and technocratic, and no new spaces or institutional practices have given citizens voice over the power supply.
Meaningful gains in citizen power after energy protests are often short-lived and diffused, dissipating in the absence of institutionalised mechanisms of citizen empowerment, or of sustained civic organisations or movements. Major protest episodes teach protestors that if they can come together en masse, they can force the government to hear their complaints about unpopular policies. This is important knowledge, closely shaping the likelihood of future episodes; many countries have repeat episodes or long-running campaigns against energy policies. But while energy protests sometimes bequeathed a legacy of movements and activist memories, in the cases examined closely, they rarely elicited more than a heated public hearing and produced little lasting policy change. In Nigeria, the 2012 episodes reinforced the threat of mass protests 3 Unless, that is, they turn into broad revolutionary movements such the Sudanese Revolution of 2019, or the Egyptian Revolution of 2011. These rare and complex moments are the outcome of many factors, in which grievances about energy appeared to have played a triggering, but by no means a decisive, role. against unpopular energy policies, succeeding not in pushing subsidy reform off the table, but in encouraging subsidy reformers to develop more tactical approaches to introducing reforms -timing or policing them so as to avoid or defuse protests. The Mozambique food and fuel protests of 2008 and 2010, for instance, taught the state to deploy extreme violence against protestors, to dampen any nascent struggles.
The research found limited evidence that energy protests led to lasting or institutionalised public accountability to citizens, whether for energy or for any other essential public services. In the more disruptive episodes of protest, governments backed down on proposed price rises or introduced short-term fixes for electricity shortages. State tactics were to ward off protests, but rarely involved a visible strategy of strengthening public accountability for energy policy. One example is Nigeria's Subsidy Reinvestment and Empowerment Program (SURE-P), announced six weeks before the 2012 subsidy cuts (Alleyne and Hussain 2013). This was a social fund designed to reinvest subsidy savings into social protection and investment, the board of which was to include civil society representatives. SURE-P ended after the government of Goodluck Jonathan left office, and while it aimed to provide some services, it did not enable citizens to determine what and how they needed to be compensated for the rise in fuel prices. Notably SURE-P financed infrastructure projects and some public works, maternal and child health programmes, and vocational training for unemployed youth (Alleyne and Hussain 2013); but none of these interventions would have compensated for the rising cost of living associated with the more than 100 per cent increase in fuel prices originally planned.
The fleeting sense of power that protestors gain when they do succeed in mobilising merely creates a self-reinforcing cycle: protestors are persuaded it is all they can and must do, and governments fear the protests, but expect the resistance to dissipate in time. Political parties in Pakistan and Nigeria took on protestors' -and the broader public's -grievances about energy to some extent. But fiscal necessity still frequently compels governments to undertake fuel subsidy reforms, while they have little incentive to reduce corruption or mismanagement in existing energy distribution systems in which powerful interests are invested.
Energy contention thus reveals a 'wicked problem': neither side is satisfied with the outcomes, but the tested strategies of each deepen the impasse and reinforce the conflict. Current strategies for subsidy reform and energy policy formulation offer no evident innovations with respect to how to address citizens' energy grievances beyond violence or short-term appeasement. But while governments offer little that is new, citizens' repertoires of protest have innovated with the rise of digital and social media, finding new ways of framing narratives and mobilising supporters. Transnational actors are part of this landscape of energy protests, but their roles can be hidden or obscure. Multilateral aid actors influence subsidy reform, but less often through the 'hard' conditionality of the earlier structural adjustment period than by shaping the intellectual agenda and policy pathways. Multinationals and foreign governments take a direct hand in national energy policy in many countries in the global South; these well-resourced powerful interests, particularly the major fossil fuel multinationals, help ensure energy policymaking remains closed and secretive. Aid donor support for civil society and good governance appears in general to have stopped short of supporting popular domestic movements or civic action for climate justice or energy security; this is one reason why civil society channels for citizens to articulate and address their energy grievances are blocked. But while transnational policy and business actors are implicated in contention around energy in the global South, transnational (Northern) civil society actors play no such role in popular protests against energy price rises and scarcity (or energy poverty). With its focus on reducing fossil fuel use and carbon emissions, transnational climate justice activism can be thoroughly at odds with the demands for affordable energy by citizens in the global South, where fossil fuels are the primary source available.
This paper provides conceptual, methodological, and empirical analysis of a phenomenon of singularly grave importance, as the world seeks to transition away from fossil fuel dependence. Without understanding why people protest against cuts to energy subsidies, there is no chance of designing lasting reforms that move societies towards fair energy transitions and climate justice. The analysis here shows that while there are no easy answers, the evidence points towards responses that appropriately and effectively centre people's experiences as users of energy. It is clear that citizens need to be engaged in energy policymaking, rather than having solutions imposed upon them which will not stick, or if they do so, will prove neither sustainable nor just.

About this paper
This paper synthesises findings from one of five projects under the A4EA research programme, a consortium of research on fragile and conflict-affected settings based at the Institute of Development Studies (IDS) at the University of Sussex. The A4EA programme explores how social and political action can contribute to empowerment and accountability in fragile, conflict, and violent settings, focused on Egypt, Mozambique, Myanmar, Nigeria, and Pakistan.
The present paper draws on three country case studies, in Mozambique, Nigeria, and Pakistan involving mixed-methods and multi-level research, including the collection of data through process-tracing, literature and document review and key informant interviews (KIIs); construction of events catalogues using news media sources; and focus group discussions (FGD) with communities involved in episodes of energy protest, to explore both their energy needs, and their experiences of energy policies and protests. The paper also integrates findings from a cross-national study of violent protests about energy prices and scarcity (or fuel riots), which explored the relationships between political and economic conditions, including energy production, and the prevalence and distribution of such episodes to date in the twenty-first century.
The paper is organised as follows. The next section sets out our thinking about useful theories for making sense of energy protests and explains our methodological approach. Section 3 sets out findings about whether protests against energy price rises and scarcity are enabled by the nature of politics in settings where political authority is fragile, and conflict and violence characterise the citizen-state relationship. Section 4 explores findings about the extent to which energy protests empower citizens with respect to energy policy, and whether and how protests leave energy policy in a more accountable state.
Section 5 examines findings about the roles of transnational actors in energy protests and the policy regimes that give rise to them. Section 6 concludes with a discussion of the research, policy, and practice implications. To inform the development of researchable questions and appropriate research methodologies, the multi-disciplinary team reviewed four bodies of literature: − Cross-country conflict analysis that uses regression analysis to identify political and economic factors correlated with protest occurrence; within this broad literature, the focus was on analyses of food and fuel riots (Berazneva and Lee 2013;Hendrix and Haggard 2015;Bellemare 2015). − Contentious politics, in particular theoretical and methodological work explaining the mechanisms through which protests emerge, recur, or escalate (McAdam et al. 2001;Tilly 2008), some of which addresses contention around energy specifically (Kitschelt 1986;Plows 2008;North 2011;Hendrix et al. 2009); contentious politics is also informed by the concepts and methods of social history, in particular work on the moral economy justifications for food riots, or popular ideas about the roles of public authorities during periods of dearth or price rises (Thompson 1971(Thompson , 1991Bohstedt 2010Bohstedt , 2014; on the moral economy of fuel price protests, see Guyer and Denzer 2013).
− Empowerment and accountability research, which defines and describes processes, mechanisms and indicators of empowerment and accountability (Batliwala 2007;Ibrahim and Alkire 2007;Kabeer 2005;Gaventa 2006;Gaventa and Barrett 2010;Fox 2016bFox , 2016a. Separately, these bodies of literature offer insights into aspects of energy protests from their different disciplinary and thematic specialisms. Our interdisciplinary research team knit these insights together to build the theoretical and methodological frameworks necessary to support our inquiry. For instance, cross-country conflict analysis describes the distribution and frequency of energy protests, theorising that people are driven to protest as fuel price rises make them worse off than before and because they see the removal of such subsidies as an abrogation of the social contract. There are parallels here with protests against food price increases which appear more likely under authoritarian regimes, where popular discontent is fed by an absence of democratic process or accountability (Arezki and Bruckner 2011). But while popular discontent about governance is common, price changes do not automatically result in protest. For example, while Nigeria has experienced many fuel price protests, the subsidy cuts in 2016 and 2020 occurred without major protests.
Analysis of food riots directs us to pay attention to these events not as the responses of the angry hungry masses, but as political acts of claims making (Hossain and Scott-Villiers 2017). The conditions of relative deprivation matter not as biological needs, but in relation to what citizens expect from their states. In comparable protests, researchers have found that moral economy views that public authorities were responsible for protecting citizens against dearth justified protests against food price rises (Thompson 1971(Thompson , 1991Hossain and Kalita 2014). Modern life increasingly demands modern forms of energy for barest survival, so we could expect to see a comparable moral economy of energy emerging among energy price protests.
Grievances and moral economy ideas alone cannot explain why protests are triggered in some but not other circumstances. The concept of 'political opportunity structures' offers a broad but systematic means of analysing the conditions underpinning protest movements, formalising the recognition that whether a movement emerges and has an impact is closely -and predictablydependent on the incentives created by the political setting (Tarrow 2011;Meyer 2004;Meyer and Minkoff 2004). Political opportunity structures are: consistent -but not necessarily formal, permanent, or nationaldimensions of the political struggle that encourage people to engage in contentious politics … [including] factors -like repression, but also like authorities' capacity to present a solid front to insurgents that discourage contention. (Tarrow 2011: 20) These may be opportunities that signal protests may succeed in shifting political or public opinion, or threats that signal that protests may elicit only state violence. Changes in political opportunity structures -the election of more progressive parties or a political scandal that delegitimates the ruling class, for instance -can make protesting seem more worthwhile.
Political economy analysis of fossil fuel subsidies drew our attention to the reasons why governments raise energy prices sharply, and to how governments respond to protests about energy price rises or scarcity. Governments often raise fuel prices sharply to cut subsidies as part of public spending reforms (Rentschler and Bazilian 2017b;Lockwood 2015 usual object of protests, partly because they are comparatively energy-rich (Cheon et al. 2013;Victor 2009). In energy-rich countries where state capacity to distribute resources is weak, consumer fuel subsidies tend to be common and resilient to reform efforts (Inchauste and Victor 2017). Authoritarian and weakly dominant political regimes are likely to rely on such subsidy regimes as a source of popular legitimacy (Andresen 2008;Rosser 2006).
Nevertheless, not all energy protests are in authoritarian or fossil fuel exporting regimes; state capacities and political regime types, as well as energy production status, vary among the countries that have recently experienced protests (Natalini et al. 2020). Not all countries where fossil fuel subsidies have been sharply cut have experienced such protests (Inchauste and Victor 2017). People appear to be particularly aggrieved about energy subsidy cuts where other forms of social protection are limited, where natural resource wealth is highly concentrated, or where economic performance is poor; in these circumstances, subsidies may come to be seen as part of the social contract (McCulloch et al. 2020;Lockwood 2015). Reflecting on the protests that triggered the Arab uprisings of 2011, El-Katiri and Fattouh note that in the Middle East and North African region, energy subsidies are 'widely perceived as a fundamental economic and social benefit' (2017: 59), such that their withdrawal undermines the legitimacy of state rule.
The localised provision of electricity has enabled research into the non-protest political responses to energy scarcity or price rises within polities. This indicates that under democratic conditions, voters reward providers of affordable electricity; in addition, democratic competition is seen to spur electrification (Ahlborg et al. 2015;Min 2015;Brown and Mobarak 2009). Unsurprisingly, given the choice, citizens tend to vote for power. Absent meaningful political choices, however, and mass protests are likely where space for civic engagement, independent media, organisations capable of mobilising support, and generalised discontent about economic conditions are present (Hossain and Scott-Villiers 2017). Protests often escalate into violence because of state repression; the nature of the state-citizen relationship, and in particular whether the state has the capacity, will, or means to respond to popular demands will also shape the likelihood of fuel riots (Tarrow 2013).

Research questions and propositions
From our reading of these different literatures, and from the learning from an initial scoping study (Hossain et al. 2018), we asked: How and under which conditions do struggles over energy access in fragile and conflict affected settings empower the We further developed a set of research questions underpinned by specific propositions which we explored through the case studies and cross-country research: − Why do struggles around energy specifically appear to have the potential to grow into significant episodes of contention within fragile settings, and which features of fragility help to explain major movements or struggles around energy?
− Did popular struggles over energy strengthen accountability for energy, or for other basic services or rights?
− Did struggles over energy empower individuals or civic or social groups to influence energy policy?
− How did transnational actors in the energy and development sectors impact on energy policies and protests?

Methodology
The study adopted a mix of methods, as suited to answering the different questions it asked, unpacking concepts, and refining approaches to fit the country cases and datasets.

Country case studies
In-depth country cases were developed in Mozambique, Nigeria, and Pakistan, because each had had moments of rapid price change, mass protest, and policy change over the past 15 years. Mozambique had experienced protests around fuel prices and electricity, and the case study examined both. Nigeria had had multiple episodes of fuel price protest, as well as instances where prices rose sharply, and no protest took place. Pakistan had also had hundreds of small and localised electricity protests in the period, and electricity policies shifted in their wake. In each country, research teams comprising scholars with expertise in contentious politics research, popular cultural and media analysis, and the political economy of energy undertook the following: − A catalogue of relevant protest events in the past decade, using digital and physical newspaper archives to develop a database of media coverage to document the timing, location, scale, and immediate effects of protest events.
− FGDs with selected occupation and income groups in areas associated with protests; these aimed to elicit perspectives on energy needs and access, perceptions of the role of government in energy provision, and the motivations and effects of energy price protests. Areas and groups were identified on the basis of findings from the events catalogue; nine FGDs were conducted in Lagos State in Nigeria; four FGDs were conducted in Lahore in Pakistan. In Mozambique, six FGDs were conducted in periurban Maputo and Matola.
− Popular cultural analysis of music, memes, cartoons etc was undertaken to access citizen perspectives on their concerns about energy.
− Process tracing of key moments of energy protests and policy change, using secondary information and key informant interviews (KIIs) with officials, activists, politicians, and transnational actors (where possible). In each country between 15 and 20 key informants were interviewed.
Country research plans and designs were shared with stakeholder groups in each country. Plans were made to ensure ethical and researcher safety standards were set and maintained at the appropriate levels in each context, and that protocols for storing and protecting data, and for anonymising names and places were established and used.

Cross-country analysis
This paper drew on an original database compiled on fuel riots (fuel price protests in which violence took place) described in (Natalini et al. 2020) and extended this to cover the period 2005-18. The aim was to establish more systematically, and across the universe of recent cases, associations between fuel riots and a range of socioeconomic and political factors. The paper combined fuel riot data with monthly data on domestic fuel prices and tested different social and political variables proxying for varying citizen-state relationships to identify the most important underlying factors associated with fuel riots. A total of 59 distinct episodes (encompassing all such events in a single year in a single country) were identified and analysed.

Fragility and the escalation of energy protests
Are energy protests more likely to escalate in fragile settings? Which (if any) features of fragility best help explain major episodes of struggle around energy?
The research studied: a) whether fragile and conflict affected political systems featured more frequent or significant energy protests (compared to more stable or peaceful settings); b) views on the part of energy provisioning within the social contract in fragile settings; c) whether subsidy regimes faced fiscal pressure domestically and/or external pressure from austerity reforms, and; d) whether in settings featuring major energy protests, citizens enjoyed formal political and civic space for popular engagement with energy policies, or if they were blocked from such participation.

Cross-country analysis of fuel riots
We aimed to understand whether fragile and violent political systems were more likely to create political opportunity structures in which protestors had incentives to protest about energy. To explore this, we undertook a cross-country analysis of fuel riots, or price-triggered episodes which featured violence. 4 There were several reasons to believe that major energy protests or fuel riots would be more likely in some kinds of fragile settings, particularly those endowed with exportable energy resources and characterised by a 'natural resource curse' (or governance arrangements intended to protect elite access to such wealth, at the cost of broader development) (Rosser 2006). Authoritarian regimes are more likely than democracies to offer fuel subsidies (Kim and Urpelainen 2016: 552), which are believed to quell or pre-empt social unrest (Victor 2009). Energy-rich countries are particularly likely to have the kind of consumer price subsidies that fuel price protests target (Cheon et al. 2013;Victor 2009); weak state capacity to distribute resources ensures that consumer fuel subsidies are common and resilient to reform efforts (Inchauste and Victor 2017). Finally, authoritarian and weakly dominant political regimes are particularly likely to rely on such subsidy regimes as a source of popular legitimacy (Andresen 2008;Rosser 2006). Major energy protests tend to become more or less common in line with the international oil price (Natalini et al. 2020). We explored relationships between different energy regimes, political systems, and major fuel protest episodes using a purpose-built dataset covering 2005-18. We found that countries with fixed price regimes -and therefore high energy subsidies -were twice as likely as countries with flexible price regimes -i.e., countries without high energy subsidies -to have experienced fuel riots since 2005. This period saw significant fluctuations in fuel prices, so that countries with large fuel subsidies faced fiscal or external pressures to increase prices; these would have been deeply unpopular at moments when other costs of living were also rising. Net fuel exporters were more likely to have fixed price regimes and to experience violent fuel protests than other countries, consistent with the literature showing that energy-rich countries have more consumer fuel subsidies, and often lack the state capacity to distribute resources through other policy instruments.

Figure 3.1 Fuel riots around the world, 2005-19
Source: Authors' own, created with MapChart, CC BY-SA 4.0 While sudden price rises were the single greatest trigger for major fuel protests, the political conditions associated with fuel riots were varied. Fuel riots were not found to be closely associated with regime type. However, fuel riots were more common in countries with higher corruption, and less common in countries that scored well for government effectiveness and the rule of law, as well as control over territory. When controlling for other factors that might influence the likelihood of fuel riots, countries with greater civil society freedom were found to have fewer fuel riots, suggesting that if polities have channels for citizens to express grievances, citizens may be less likely to resort to protest to resist unpopular policy changes. Further, we found that countries where significant anti-system movements exist are much more likely to have fuel riots. This supports the idea that riots are not necessarily spontaneous eruptions of anger but are supported by opposition groups that leverage public anger against price rises as an opportunity to discredit or damage the government.

Mechanisms of escalation: Energy protests in Mozambique, Nigeria, and Pakistan
While aspects of state fragility are associated with the emergence of significant protests triggered by energy contention, we need to understand the mechanisms through which protests about energy escalate from protests into broader struggles. 5 From contentious politics theories, we can draw on the concept of 'scale-shift' or 'change in the number and level of coordinated contentious actions to a different focal point, involving a new range of actors, different objects, and broadened claims' (McAdam et al. 2001: 331).
To identify the mechanisms that turn a protest into an episode of political rupture entails analysing the moments at which specific local grievances are transformed into broader complaints. The shift in scale we are interested in here is vertical and horizontal -a form of 'contentious politics that expand outward and upward from the local level to the state, federal, or transnational level' (Soule 2013: 2). Such shifts would not be possible without mass popular support, and so imply in addition a process of 'diffusion' or that some element of a social movement (e.g., tactic, frame, ideology, protest, repertoire, campaign) is spreading across some set of actors (e.g., organizations, networks, groups, people, communities, states) in a social system either through direct or indirect networks of communication. (Soule 2013: 1) To explore the mechanisms through which energy protests experience scale shift, we sketch the main episodes in each country in which protests about energy prices escalated into more substantial complaints about corruption, collusion, and the failures of economic and political governance. We look at how grievances and collective action diffused across different groups and regions. This overview of the protests and their framing narratives, also identifies the networks through which narratives and tactics were diffused.

5
The cross-country work used a definition of fuel riots that included that the event had to feature violence.
In the country case studies, the focus was not on fuel riots but on the broader category of energy-related protests.

Mozambique
Mozambique has seen explosive episodes of popular mobilisation in 2008, 2010 and 2012, in which protesters have voiced anger not only at the rising cost of urban transport, electricity and food but also at growing inequality and corruption (de Brito et al. 2017). See Annexe 1 Table A1 Protest episodes in Mozambique,  2008-20 and Table A2 Estimated deaths, injuries and arrests during energy protest episodes in Maputo and Matola, 2008-20 for more detail. These revolts spread beyond Maputo, although they were concentrated in the metropolitan area of Greater Maputo, including the satellite city of Matola, comprising together about two million people. Greater Maputo is highly unequal; its visible wealth pockets are surrounded by large concentrations of people living in poverty. In recent years many of these peri-urban zones have been transformed by the growth of bairros de expansão (expansion neighbourhoods) in which upwardlymobile households seeking more living space but priced out of the central areas have moved in alongside poorer traditional residents, bringing with them higher expectations in terms of services and infrastructure. Maputo has higher levels of education and rates of computer, internet or cell phone access than the rest of Mozambique and features a strong perception among many 'of being 'captured' in poverty, and not being able to take part in what the city has to offer, is at the heart of current dissatisfaction among poor people in Maputo's poor bairros' (Tvedten et al. 2013: 4).
In early 2008, the Mozambican government increased the price of fuel to offset higher costs of imported oil. Private passenger transport vehicles (known as chapa) increased fares by up to 50 per cent. The price of bread also rose by 12.5 per cent. Text messages circulated calling for a 'greve' or strike, and on 5 February a wave of violent popular protests erupted, blocking main roads with improvised barricades, and burning tyres, looting shops (particularly of food), and burning or stoning vehicles. The riots paralysed Maputo and Matola for three days. The protests ended when the decision was taken to keep transport fares unchanged, after the government negotiated compensatory benefits with the chapa operators, notably removing the 17 per cent value added tax on diesel.
On 1 and 2 September 2010, text messages again circulated calling for a greve, triggering a second major outbreak of street protests, this time about an increase in the government-controlled price of bread and other basic goods and services, including electricity, after prices of food and fuel had already risen sharply. Similar patterns of violence to February 2008 were seen, this time with ten deaths and hundreds of people injured, and extensive damage. As in 2008, the government reversed its decisions, froze prices, and announced new subsidies, among other measures. In the protests of 2008 and 2010, some two-dozen people were killed, including two children, and over 500 people were injured. On 15 and 16 November 2012, Maputo was paralysed again, this time for a day and a half due to the threat of riots following an announced increase in transport fares. However, the patterns of violence seen in 2008 and 2010 did not recur. Mobile phone communications were restricted, and heavy policing at critical points of the city prevented escalation. There have been no major recurrences of energy protests since 2012. Instead, a pattern has emerged of localised protests both over electricity costs and service failures, and over transport costs (with the latter often led by chapa operators themselves when compensation schemes fail to cover the cost of rising fuel prices), that are stifled before they can spread. In recent years, the government has sought to head off larger-scale protest through a combination of smaller but more frequent price increases, investments in public transport (running cheaper public buses on some crowded chapa routes) and improving the speed with which the electricity utility Electricidade de Moçambique (EDM) responds to local complaints about service failures.
Several features of the scale shift that took place in the 2010 protests in Mozambique stand out. Such moments are rare in a dominant-party state with an over-developed apparatus for repression. The year 2010 marked a moment of acute crisis for low-income urban Mozambicans in particular, when the cost of food, transport, and electricity all increased substantially. The triggers were a combination of factors that together amounted to a subsistence crisis for many: energy prices were one of several grievances, each of individual importance. This broader sense of subsistence crisis shifted into a complaint about political and economic governance. As one interviewee told de Brito et al.: it was a government of intransigent, intolerant individuals who had nothing to do with the cause of the people or with democracy, because one of the pillars of democracy is full respect for human rights. The right to health and the right to food are part of human rights. (de Brito et al. 2017: 166) Messages about protests were diffused through short messaging service (SMS) messages and popular tunes like Azagaia's (2008) 'Povo no poder/Power to the people', which accused the president and his ministers of failing to notice 'that life's not easy here', warning that 'If the police are violent/We'll respond with violence', for which the public prosecutor summoned him to answer questions related to possible incitement to violence. 6 The government introduced mobile phone subscriber identity module (SIM) registration and cancelled prepaid arrangements, to limit the diffusion of protest. Similar price rises and grievances in 2012 did not lead to protests on the same scale, as protestors weighed up the benefits of protesting, against the repression they would likely face. When larger 6 As published in the weekly newspaper Savana, 3 September 2010 (our translation). hikes in electricity and fuel prices were again introduced in 2017 in a context where food prices were also rising sharply, individual transport operators mobilised and calls for a greve circulated on social media, but the chapa owners' organisation disowned the protest and a heavy security presence (including armoured cars parked at major transit points) dissuaded citizens from taking to the streets. Smaller, more localised protests about energy continue -particularly in relation to electricity -but never taking the national stage as in 2008 and 2010.

Nigeria
While Nigeria has had multiple episodes of energy-related protests, the two weeks of protests, (including both violence and disruption as well as peaceful stay-at-home strikes) between 3 and 16 January 2012 remain the single most significant episode. When the government suddenly ended the fuel subsidy, increasing the price of fuel from N65 to N140 per litre (in 2021 prices, US$0.17 to US$0.36), street protests erupted in various parts of the country. The president was accused of unilaterally imposing the hike without consultation, in a clandestine attempt to enrich corrupt officials and market actors. Before the Goodluck Jonathan government scrapped the subsidy, there had been claims of widespread corruption and mismanagement in the oil sector (Okonjo-Iweala 2018), but while the government admitted that a powerful clique had taken advantage of the oil subsidy scheme, it acted to remove the fuel subsidy rather than tackle the corruption.
The labour unions warned that their members would protest if the government decision was not rescinded. Their leadership legitimised and energised ongoing civil society-led protests (see Houeland 2018 for close analysis of Nigerian trade union leadership of fuel price protests over the decades). With the trade unions, social media activists and other media coverage, the protests quickly internationalised by sending solidarity messages to the international labour movement and diaspora Nigerians, some of whom protested outside the World Bank/International Monetary Fund (IMF) premises in Washington DC. At its peak, tens of thousands of protestors were on the streets across the country, including around Lagos, and in Kaduna, Benin City, and Kano. These protests included a nationwide strike by millions of trade union members. The superstar musician Fela Kuti's son, Seun Kuti, came out in support, marching shirtless among protestors and speaking at protest rallies. Riot police beat people with batons before firing into the air as the crowd, angered by the beatings, surged forward. Protestors marched with placards reading: ' "Soon the poor will have nothing left to eat but the rich". T-shirts were emblazoned with "Kill corruption not Nigerians" ' (Hossain et al. 2018: 27). According to one protestor: This government is supposed to be there to protect us, to allow us to make business, to feed and clothe and educate our children, and now they are here shooting us,' said David Amosu, a civil servant who joined one rally in Lagos. (Pflanz 2012) The unions came out in support of the protests, but attempted to curtail the violence: One union leader described the federal government's hugely unpopular move as 'immoral and politically suicidal' and urged Nigerians to resist 'with everything they have'. But … once unleashed, the pent-up anger of the masses could be hard to curtail. Angry crowds vandalised gas stations, intimidated owners into keeping their pumps unused and assaulted a soldier, showing how easily the fragile peace in Africa's most populous nation could spiral into chaos. One young man threw jerrycans of engine oil off the racks at a gas station and tried to damage the station's gas pumps.
After union leader and chairman of the Joint Action Front, Dipo Fashina, asked the young man to stop vandalising the station, he did, but later started one of the first bonfires of the protest in the middle of the highway. (Alamba and Ibukun 2012) Some 15 people in the end lost their lives during the 2012 protests.
Although energy protests are not unusual in Nigeria as they are in Mozambique, the scale of the protests in 2012 was unprecedented. One recent study has explained its scale and significance in relation to the combination of an extended period of democracy and oil-led growth on the one hand, and the absence of broader social participation in the benefits of that wealth and the rise of new social actors on the other (Houeland 2020). Protests around fuel subsidy removal have a long history but were never as widespread or violent before or since 2012. A critical factor was that the 2012 struggle brought together many different actors. Whereas protests and civic action in Nigeria are often divided across religions or communities, the 2012 protests appeared to be spontaneous, widespread, and uniting people across class and other groups, bringing union and civil society activists behind a common grievance; in this respect, these 2012 protests were like other fuel price protests in the 1980s and 1990s (Houeland 2018).
As the fuel price rises became connected to charges of corruption, the opposition party became active, addressing crowds, sometimes supporting protests. The atmosphere was friendly (until the police and army started beating and firing at people), with people playing football in the street and being entertained by famous singers. Popular music and social media memes created and transmitted a common sense of grievance. It was also marked by young people using social media to spread grievances and protests across the country (where in the past it would have been concentrated in the southeast). Since 2012, online activism has been increasingly energetic, but while more frequent, protests have been generally small. 7

Pakistan
Pakistan witnessed some 456 protests between 2007 and 2015 related to electricity supply, access, and pricing. Of these, roughly 10 per cent formed part of a larger protest episode, of which the research identified four major moments. One episode, from 12 to 19 June 2012 in the Punjab, stands out because of the scale of the violence, and the deaths of 10 people. It elevated the electricity crisis to a position of political primacy for citizens and public authorities. By 2013, citizens reported electricity outages as their number one concern, over the wave of Islamist militancy and terrorism that began in 2005.
Some of the most violent protests (between 2010 and 2013) took place in the wealthy and politically competitive regions of North and Central Punjab. The June 2012 protest episode was characterised by extensive economic and social disruption around the country: at least seven properties belonging to electricity sector companies were ransacked in different cities, and police vehicles were torched in Kamalia and Rawalpindi. The main north-south national highway near Toba Tek Singh, Punjab, was blocked for three days, and police and security guard action led to the death of two individuals. Major market and industrial associations carried out 'shutter-down' strikes, ending economic activity during this period.
An important point to note here is the cross-sectional nature of the protesting public, which enhanced its political salience. The protests were not just populated by political activists or workers but were planned and coordinated by sections of the urban business classes, in particular traders, merchants, shopkeepers, and other bazaar-based actors. Data from the events catalogue highlights that traders and associated business segments were the second most frequent segment involved in electricity related contention. Urban political economy accounts of contemporary Pakistan, such as those by Akhtar (2018) and Javed (2017), identify traders and associated business segments as a politically salient demographic, given their role in financing political activity, and their status as local community leaders. Furthermore, opposition parties were quick to capitalise on public anger as it gave them a chance to demonstrate their representative credentials with key urban demographics.
Pakistan's case study differed from protests driven by fuel price contention, but resembled electricity protests in Mozambique in that public authorities reacted 7 Until the 2020 #EndSARS movement, which centred on police brutality. with repression on relatively few occasions. Out of 242 protests where public authority responses were recorded, only 8 per cent saw police using force. These 8 per cent included the protests (such as those in October 2011 and June 2012) where government infrastructure and private property of politicians associated with the ruling party was damaged. The most common outcome (~42 per cent) was protracted negotiations between protestors and local electricity officials, mediated by the police and district administration, to limit unscheduled outages. These were mostly cases where outage problems were contained, and protests did not spread.
An important reason for lower levels of police action is the political dynamic between 2008 and 2013. The parties of the federal government and the Punjab took opposing views over when to clamp down on protests, and policing is a provincial responsibility. Interviews with political representatives and officials hinted that urgency for police action only kicked in with the threat of damage to private property, while road blockages were tacitly allowed to allow protestors to vent. The research uncovered that the police were reluctant to repress protestors due to shared circumstances: police faced the same electricity shortages and empathised with citizen anger. The local administration would often use informal ties with businessmen and political leaders to use their leverage to request protestors to dissipate. Formal action was normally used in the last resort, and only after explicit political approval had been granted by the provincial government.
As with the major episodes in Mozambique and Nigeria, in Pakistan, new alliances and widely shared grievances helped sustain and escalate the protests, which were also supported by political opposition groups. Social media and a relatively free and open mass media helped to diffuse the protests, but a critical factor compared to Mozambique and Nigeria and to other protests in Pakistan, was the relative restraint of the police, partly the result of the dynamics of political competition. The restraint and tacit sympathies of the police enabled the struggles to continue over time.

Affordable energy as part of the social contract
Modern forms of energy are becoming an important part of everyday life, particularly in urban areas in the three case study countries. Provisioning remains uneven across and within countries, and access to electricity and cooking fuels has improved slowly over the past 20 years in the three case study countries (see Figure 3.2 and Figure A1). Pakistan provides a higher proportion of its population with electricity and improved cooking fuels than either Nigeria or Mozambique, but Mozambique has made rapid progress on electricity provision. Energy features in different ways in the social contract between citizens and their states, reflecting in part how those states are situated in energy markets. As Figure 3.3 shows, the countries have widely varying relationships to fuel markets, with Nigeria notable for the dominance of oil in its overall exports. Fuel exports have grown significantly in Mozambique in recent years, while Pakistan continues to rely extensively on fuel imports, as a non-oil producing country.  We know that the possession of natural resource wealth plays some role in the way energy is treated in the social contract. Consumer subsidies serve as primary forms of social protection in countries with natural resource wealth and/or limited state capacity to distribute other resources (see El-Katiri and Fattouh 2017; Inchauste and Victor 2017;Lockwood 2015). By a social contract we do not necessarily mean a constitutional commitment or even a specific political promise to provide affordable energy. Instead, we use the concept to refer to popular political beliefs or commonly held ideas about the basis of rule. For affordable energy to be seen as part of the social contract, people would need to expect actions by their state to ensure energy access as a bare minimum, and the state may lose legitimacy should it fail to meet those expectations.
Energy may feature in social contracts in highly specific ways, as when the nation is itself energy-rich and access to affordable energy is seen as a specific entitlement on the grounds of national energy wealth. This is a kind of resource nationalism, or:  a political discourse, applied to political and economic thinking about how a state and its population should manage and distribute profits derived from natural resources … resource nationalism may also be understood in terms of collective belonging expressed through the idiom of natural resources. It is, in other words, one means by which the imagined community of the nation is constructed. (Koch and Perreault 2019: 611-12) In Nigeria, resource nationalism includes a view that affordable fuel is the principal claim Nigerians make on their state; its value is not only that it keeps living costs lower than without a subsidy, but also that it signals respect for the citizen-state relationship. As one Nigerian scholar of the political economy of fuel explained: It has become basically a cultural and political issue where Nigerians believe since we have petroleum and as government does not give us anything other than cheap fuel, let's hold them to account on the basis that citizens want cheap fuel … emotionally, citizens feel the only thing they benefit from government is petrol.
(KII 9, Nigeria) It bears repeating that fuel subsidy cuts signal not only the material fact that fuel prices and the cost of living will rise, but also that the government is not listening to the people -they are both an economic and a political failure.
In Mozambique, resource nationalism also featured in how people viewed their relationship to energy, notably to the Cahora Bassa hydro-electric dam, owned by the erstwhile imperial power Portugal until 2007, when the government of President Armando Emílio Guebuza (who was subsequently implicated in a massive illegal debt and corruption scandal) bought the Mozambican state a controlling share. The slogan 'Cahora Bassa é Nossa' [Meaning; 'Cahora Bassa is Ours'] is now often used ironically in social media, to point out that national ownership of the dam has not improved electricity access for Mozambicansalthough elites are understood to have benefited. 'Cahora Bassa' is also used as a slang term for the illegal electricity connections to which many residents of Maputo and Matola resort when they have been unable to secure a connection to the grid and/or when prices rise out of their reach.
The country's vast reserves of natural gas have not generated a comparable popular expectation that domestic cooking gas will become more affordable. This is seen as a possibility for the future, contrasting with a current reality in which most of the gas produced by South African energy firm SASOL in Southern Mozambique is exported via pipeline direct to South Africa, while plans for the development of the Cabo Delgado gas fields in Northern Mozambique emphasise exporting liquefied natural gas (LNG) to Asia rather than making it The Pakistan case study offers an alternative perspective on the role of energy in the social contract. Pakistan is not an energy exporter but the extension of grid electricity has historically been a signal of the government's developmental commitment and credentials (Naqvi 2011). In FGDs in working-class Chungi Amer Sidhu in Lahore, respondents said the state was responsible for providing electricity at a 'fair price' and that failure to do so amounted to a failure of governance. One respondent explained: 'if the hakoomat (government) is not ensuring that our fans stay on during the summer months, what good is it?'. Meeting citizens' energy needs involves subsidies, because the cost of generation exceeds the price charged to consumers. The twin dynamic of expectations of the state obligation to supply, and the role of subsidies in ensuring that obligation within existing systems, underlies the mechanisms through which subsidies are connected to escalations of energy struggles in Pakistan.
Our country case studies generally supported the view that subsidies and/or affordable access to energy are seen as citizens' rights, part of the social contract governing citizen-state relations. At the same time, FGDs uncovered a clear and firm sense that some forms of energy were deemed essential for everyday life, so that the denial of energy needs amounted also to a denial of the right to work or eat.
Subsidies were never explicitly framed as a substitute for, or in lieu of, social protection -at least until they were cut or removed -at which point their contribution to keeping everyday life affordable became clear. People in each of these three countries described the painful adjustments they had to make when energy prices rose, indicating that the vital nature of modern forms of energy to everyday life means energy is seen as a right, central to human wellbeing in the same way as food, water, or clean air. Nevertheless, the extent of material hardship people face when subsidies are cut is not the sole determinant of whether they protest, or whether those protests escalate.
While certain kinds of fragility, including the political economy conditions that give rise to fixed price energy regimes, are associated with escalated energy protests, contention over energy also creates fragility in settings otherwise deemed stable; events in France and Chile in 2019 are the clearest examples of this. One implication is that as energy becomes ever more fundamental to daily life, even strong governments that are otherwise viewed as accountable may struggle to meet energy demands in ways citizens view as acceptably fair or affordable.

Subsidy regimes under pressure
We expected subsidy regimes to face pressure to reform, mainly from multilateral agencies. We found that protests tended to escalate at times when subsidies were under fiscal pressure, often arising from the inefficiencies and/or corruption arising from the administration of the subsidies themselves. However, findings about the role of the IMF or other multilateral agencies in subsidy reform efforts were inconclusive; while there may have been more direct influence in the past (see for instance Houeland 2021), contemporary sources of influence appear to be indirect rather than through more overt conditionalities.
In Nigeria, insurgency in the Niger Delta and fluctuating global prices of oil exports and fuel imports created macroeconomic instability and fiscal pressure on a state with weak revenue-raising abilities ( Evidence of corruption has also added to pressure for subsidy reforms in Nigeria: in December 2011, subsidy payments were over 900 per cent more than the appropriated budget (US$7bn as opposed to the budgeted US$642m). More than 20 companies involved in the subsidy scam that was uncovered after a public outcry still face criminal charges for receiving hugely inflated subsidy claims. 8 A total of NGN382bn (US$1bn) was lost to the subsidy scandal (BudgIT 2019: 8). Other sources of pressure on the subsidy regime include devaluation of the Naira and rises in the price of imported fuel, both of which increase the outlay on subsidies (BudgIT 2019). Suspicion of multilateral agencies in Nigeria dates to structural adjustment, and opposition actors and the labour movement have blamed the IMF for austerity measures and subsidy cuts. A visit to Nigeria by the IMF Director Christine Lagarde and the appointment of former World Bank Managing Director Ngozi Okonjo-Iweala as Nigerian finance minister in 2011 raised suspicion about the role of multilateral agencies in pushing for fuel 8 The inquiry into what is widely known as the 'subsidy scam' covered subsidy payments over the period 2009-11, led by the former chair of the parliamentary finance committee. The inquiry found fraud and mismanagement; subsidy claims for products not delivered; overcharging of the government by oil marketers; requisition of foreign exchange for imports of refined products, with the foreign exchange diverted to other uses; unauthorised deductions by the NNPC to itself; and government mismanagement (Okonjo-Iweala 2018). The estimated daily consumption of gasoline was 35m litres, but 59m litres of fuel subsidies were paid, suggesting that potentially as much as 24m litres a day of subsidised gasoline was being smuggled out. Government officials ignored the Petroleum Subsidy Fund (PSF) guidelines, due mainly to corruption and, in some cases, incompetence. A total of NGN1.067tn (US$6.8bn) was estimated to have been misappropriated as subsidy payments. This initial inquiry was derailed by a bribery scandal involving the chair of the inquiry committee, but a subsequent presidential committee arrived at similar conclusions, concluding that NGN382bn (US$2.5bn) was fraudulent or questionable and should be recovered (FOSTER 2020).

ids.ac.uk
subsidy cuts. Supporters of the Occupy Nigeria movement staged protests at the IMF offices in Washington DC in 2012 to highlight this presumed role, indicating that the legacy of the IMF helps political actors resist or even mobilise against reforms, regardless of its actual or direct role.
In Pakistan, the electricity crisis started in 2007: the oil price shock and stagnating economic growth reduced fiscal space to pay the tariff differential subsidy needed to keep generating electricity. Successive governments struggled to increase revenues and exports to pay for energy related imports. These two trends have culminated in spiralling fiscal and current account deficits and repeated bailouts by the IMF. In 2019, the country went for its 11th IMF programme in 30 years on the back of yet another balance of payments crisis. Successive IMF programmes have advocated rationalising electricity related subsidies to bring down fiscal deficits and create alternative mechanisms of social protection in energy markets. The IMF and other donors continue to argue for improved transmission and distribution governance as the primary reform thrust, so that losses can be reduced, recovery rates improved, and overall cost of electricity goes down. They also advocate for the privatisation of distribution companies, which electricity sector labour unions continue to oppose. Anaemic growth and a weak revenue effort coincided with the global recession and oil price shock during 2007-15 to squeeze the state's ability to pay electricity subsidies. The resulting outages contributed to pre-existing fragility by causing a 4 per cent loss to gross domestic product (GDP) growth, and by escalating public anger, leading to widespread mobilisation and periodic outbreaks of violence.
In Mozambique, bilateral and multilateral agencies tend to be cautious in negotiating with government over electricity prices and fuel subsidies because of the possibility of violent repression of protests, but they nonetheless have an influence. This influence had been significantly reduced after many bilateral donors withdrew budget support in the wake of the 2016 'hidden debt' scandal, with the government resisting calls for greater transparency in the expectation that gas revenues (including from deals with energy companies from the donor countries) would reduce its donor dependency. This expectation has been frustrated as declining gas prices and the worsening insurgency in Cabo Delgado have led to the postponement or cancellation of investment decisions, and as a result the government has found itself needing to turn once again to the international financial institutions. The IMF and World Bank are known to push for subsidy reform but are not considered by citizens to be responsible for the rising prices of electricity and transport fuel. Bilateral donors tend to be most active in the off-grid solar sector (through programmes such as the UK aid funded BRILHO programme), with only USAID taking on a high-visibility role in pushing for liberalisation of the electricity sector and regulatory reform.

Space for citizen participation in energy policy
One finding is the elite, technocratic, and untransparent nature of energy policymaking: decisions are made in closed rooms, by powerful actors who often have transnational links. Energy policy had become an important issue for political parties in the wake of mass protests in each of the countries; this had shifted some policy action from technocratic and bureaucratic towards political spaces (see also Houeland 2020).
In Nigeria, we identified no effective institutions through which citizen concerns could be articulated for redress: political parties play little role in aggregating or articulating citizens' energy grievances, although trade unions are often called in to negotiate on behalf of fuel protestors. During the energy crises under the citizen government of Obasanjo in 2003, a committee was set up that eventually metamorphosed into the Petroleum Products Pricing Regulatory Authority (PPPRA). The PPPRA comprised 26 interest groups and stakeholders such as labour, oil, and transport workers unions. However, fuel price rises are not always discussed at PPPRA board meetings before being imposed.
In Pakistan, electricity sector governance remains in the hands of the federal government bureaucracy, international donors, and top elected officials. There is little to no parliamentary or citizen oversight of decision making about the guaranteed returns offered to private electricity producers, tariff setting, subsidy outlay, or other aspects of sector reform. KIIs highlighted the insular nature of policymaking, popular frustration with these processes, and people's reliance on bribery or personal relationships to resolve electricity supply problems. When direct influence fails, communities tend to leverage elected representatives to bargain on their behalf. A 2020 report on financial practices within Pakistan's electricity sector governance uncovered the lucrative contracts between private power producers and the government, and the failure of regulatory oversight to ensure efficiency, suggesting rent-seeking relationships between different business, bureaucratic, and political elites linked to the sector. People reported believing the electricity sector is rife with muk muka [Meaning; 'reciprocal corruption'], while FGDs revealed citizens' sense of disempowerment with respect to energy decision making.
In Mozambique, there are no institutions to channel citizen complaints into the regulation of fuel prices. The Mozambican Federation of Public Transport Owners (FEMATRO -Federação Moçambicana das Associações dos Transportadores Rodoviários), lobbies for tariff increases and negotiates with the government over compensatory subsidies whenever there is a fuel price rise. Accountability for urban transport costs in Maputo is confused by the fact that municipalities regulate the chapa fleets while a provincial-level agency runs government buses and central government sets the fuel price. The parastatal EDM communicates tariff changes to users, chiefly through television advertisements. A team dedicated to community outreach is geared towards ids.ac.uk avoiding sabotage to EDM's infrastructure and electricity theft, communicating through community leaders and Guezi, a theatre group of former EDM workers; however, these are not channels to communicate citizens' grievances back to EDM. Since 2017, the Autoridade Reguladora de Energia (ARENE -Energy Regulatory Authority) is responsible for setting electricity prices but has no capacity for communications or mechanisms for citizen participation.
In the past decade, Mozambican political parties have engaged with popular concerns about energy. The ruling party Frelimo publicised its World Bank funded programme 'Energy for All', with its 2019 election campaign featuring promises to extend electricity access. Opposition parties are also addressing energy concerns: when in October 2020, EDM announced plans to increase electricity tariffs by 10 per cent in 2021, the new urban-based opposition party, Mozambique Democratic Movement (MDM) demanded the plans be reversed. Overall, civil society engagement with the energy sector is embryonic, and there are no platforms for civic engagement with the state. Civil society actors undertake research and advocacy on energy issues but have limited connections with communities. Researchers have suggested that the protests of 2008 and 2010 revealed an 'incapacity of the political system to guarantee a genuine space for negotiating the social interests of the poorest strata of society' (de Brito et al. 2014: 27); civil society similarly lacks a history of mobilisation around energy resources. Citizens increasingly turn to popular culture and social media to express energy grievances.

Energy protests, empowerment, and accountability
Do protests empower protestors to hold public authorities accountable for energy (or other public services)? The findings indicate that despite the noise and fury, and the very public and widespread criticism of the authorities signalled by major protest episodes, there is limited evidence that protesting publics were empowered by their struggles, beyond the fleeting sense of power that collective action can bestow. We found equally limited evidence that the protests increased accountability for energy policy or provisioning, beyond some placatory responses intended to defuse protests rather than to address grievances. Some groups worst affected by unaffordable or inaccessible energy, in particular rural people, women and the poorest, were not visible in protests, nor were they the main beneficiaries of any concessions won beyond short-term gains. We identified neither autonomous citizen organisations capable of advocating on their own behalf, nor institutionalised mechanisms for including citizens or civil society groups in energy policymaking. To the extent that political parties adopted popular energy demands, citizens' protests succeeded in empowering some of their political representatives. Where protests left organisational legacies or coalitions, these played no role in democratising energy policymaking.

A cycle of protest and response without reform
The Nigeria case suggests that protests have empowered citizens to demand public action: the official response has been to either cut the pump price, make new social protection provisions, or both. These responses then became ammunition for the next struggle, as evidence that protests had 'worked'. We agree with Camilla Houeland that '[n]one of the historical fuel subsidy protests altered the system, but they contributed to shifts in popular ideas, civic action and institutional politics' (Houeland 2020(Houeland : 1236. Nevertheless, our research suggests that those shifts were temporary or partial, and amounted to neither an empowered citizenry nor accountable public authorities in any institutionalised or lasting sense. The frequency of protests suggests their inability to effect long-term policy shifts that people can broadly support. The trajectory is predictable: government cuts the subsidy; fuel prices rise; activists and labour unions mobilise crowds; protests erupt; government invites labour union leaders (not activists, not the masses) to negotiate; the subsidy is partially restored, and prices of fuel products revert, although not to the original price. The repetitive, almost ritual, nature of the response to protests do not suggest that citizens gain power from their protests, but that they are trapped into short-term responses with no serious prospects of satisfactory resolution.
Concessions granted to protestors are often little more than political theatre: citizens and government know protests will occur, there will be negotiations, and that government will be forced to reverse cuts to some extent, so they announce a larger initial cut than they expect to make. As a government official with authority over these issues explained: The government would be determined to increase fuel prices and would say okay, they want to increase it to N10 from N8 but they know that people would scream and shout so they'd increase it to N12 so that everybody can scream and shout, and then they bring it to N10, which was their original intention. (KII 16,Nigeria) Such actions are a double win for government: subsidies are reduced, but the government also demonstrates its compassion and responsiveness. Labour leaders also win, because as one interviewee explained: 'Labour will not lose face before their people, and government would have achieved what they want to achieve' (KII 13, Nigeria).

Organised actors and the escalation of protests
The role of organised labour raises further questions about the extent to which protests empower people over energy policymaking: in three major and five minor protests analysed for the Nigeria case study, organised labour under the Nigeria Labour Congress or the Trade Union Congress of Nigeria actively involved civil society groups, and other citizens supported activist groups in mobilising members and other citizens (Houeland 2018). However, governments typically negotiated with union leaders alone, in many cases introducing new benefits for the civil servants represented by unions. Non-unionised citizens saw few benefits.
Although not in direct response to the protests, the Goodluck Jonathan administration created the short-lived SURE-P to mitigate the impacts of the partial petroleum subsidy removal on the population by laying a foundation for developing a national safety net programme; accelerating economic transformation through investment in infrastructure projects; and promoting investment in the petroleum downstream sector (Omafume 2014). Several highprofile infrastructural projects were awarded, such as the Benin-Ore expressway, the reconstruction of the Lagos-Ibadan expressway, and some investments in railway projects. Achievements in social protection and health services were more modest: 120,000 jobs were created for youth, women, and disabled, and 12,000 health care workers were recruited, providing services to 800,000 women (Omafume 2014). No formal evaluation has been undertaken of SURE-P, which appears to have ended without replacement.
Some energy protests did not even achieve such limited wins. In 2016 the Muhammadu Buhari administration cut subsidies and the petrol price jumped from N87 to N145 (US$0.23 to US$0.38) per litre. Unions and federations called strikes, organised marches, and disrupted traffic. However, few citizens joined the protests, and the price of fuel stayed at N145 per litre. One reason was that most people had goodwill towards the new government of Muhammadu Buhari who was personally known for integrity, in contrast to the Jonathan administration of 2012, on whose watch corruption was widely reported. A second reason was a growing understanding of the nature of the subsidy regime in Nigeria. By 2016, people understood the subsidy system was thoroughly corrupt, and that fighting to retain the subsidy was actually fighting for the interests of the 'subsidy cabal'. One focus group discussant from a low-income community put it succinctly: 'Ko si subsidy kankan; sosapo lo wa. [Meaning; 'there was never a subsidy; there was/is only the pocketing of the nation's money'] (Male, FGD 04, low income).
A third reason protests were large and sustained in 2012, but not in 2016 appears to be that the 'big actors' who supported the protests in 2012, providing transport, refreshments, and even rockstar entertainment, were not active in 2016 (in part, because many of them were in government). Protests can be stimulated by unions, political opposition groups, and other influential actors, even though they emanate from mass concerns. But these organised actors, including the opposition Action Congress of Nigeria, are seen by some to 'capture' the protests for their own purposes, subverting the protest goals to pursue political goals apart from citizens' energy concerns.

Pointless protests
Protests also failed to take off in 2016 because rather than developing confidence in their own agency and ability to affect government's energy policy, protesters had protest fatigue as a result of revisiting the same issues without achieving any long-term gains. A common argument was that the concessions sometimes made by the government in response to protests were a 'consolation prize'. One man who had protested in 2012 said: Research participants noted that women and men often had different energy priorities: women's unpaid care work depends particularly on access to cooking fuel, an issue around which energy protests rarely mobilise. However, we uncovered no clear sense of a gender divide with respect to energy grievances. Women were less likely than men to take to the streets to protest price hikes for a range of reasons, but in 2012, respondents noted that women both came onto the streets and struck by staying home. Fear of police and army violence was one reason most protestors were young men, but their grievances were framed as national issues, not as the problem of male urban informal sector workers. Women and men viewed the protests similarly, with age and stage of life the more significant dividing line between those who protest (mostly young urban men) and those who do not. Respondents commonly noted that the state's security agents were more prone to repression than in the previous administration, further reducing their incentives to protest.
In assessing the potency of popular struggles to empower people over policymaking in energy provisioning, lessons from Nigeria show that while protests usually follow increases in fuel pump prices, protesters' demands ranged from the reversal of price rises to improved management of petroleum resources, and an increased minimum wage. However, beyond the occasional reduction of fuel prices, other demands tended to be ignored. The 2012 protesters in Lagos demanded a referendum before the implementation of the controversial fuel sector deregulation policy. Protesters in Abuja during the 2012 episode demanded that the government not only reinstate fuel subsidies, but also unveil an updated budget showing that the subsidy funds had been duly reallocated. Protesters in Benin City wanted the resignation of the finance minister among other demands. These demands met with no response, and do not even appear to have been considered. As in Nigeria, the Pakistan case showed that major energy protests were not the spontaneous eruption of angry citizens but were supported and enabled by preexisting organisations. In the working-class Lahore neighbourhood of Chungi Amer Sidhu, activists linked to local labour unions and a local left-wing political party aiming to compete in local elections, formed a community-based organisation to deal with service delivery complaints of residents. Other members of the protesting public included former women councillors with ties to politicians from the Pakistan Tehreek-e-Insaaf (PTI) party and the Pakistan Muslim League (Nawaz) (PML(N)). These organisers were supported by local shopkeepers and businesses who saw their ability to mobilise young people from the area as a useful way to pressure public authorities.
Similarly, protests in the town of Kamalia, including the major riot in June 2012, would not have taken off without the backing of the politically and socially important market traders association (Anjuman-e-Tajran Kamalia). Their linkages with political parties, particularly the PML(N) in Punjab, helped mobilise large crowds to protest outages. This pattern -of civil society organisations fronting protests with organisational backing from opposition politicians -recurs in many instances where energy protests went to scale, and political party leaders and activists have been central to the escalation of energy contention in Pakistan.

Gender, class, and protest
The Pakistan case found that popular mobilisation around energy mainly featured young urban men and was often inattentive to the needs of other social groups -women and children, rural people, the poorest. Protest organisers were on many occasions better-off urban men, with frontline protestors from the working or lower middle classes. In the protests in Kamalia, for example, trader association representatives called for the protest, held the 'shutter-down' strike, and then ordered employees (sales staff, loaders, and helpers) to block the main highway. The original protestors were then joined by local constituents of the opposition party politician who had subsequently joined the protest as well.
Women may not take to the streets in large numbers around energy issues, but in Chungi Lahore, numerous service delivery complaints -related to water shortages, electricity bill distortions and outages, and natural gas disruptionsoriginated with women residents, who depend on these forms of energy more because of their responsibilities for care work. Women are also the primary users of household electricity, and so are directly affected by outages. Disempowered socio-economic groups (fringe urban communities, rural consumers) are affected more than wealthier and urban people. Even within these groups, there is a gendered dynamic to questions of affordability: women respondents in Chungi explained that pricing remains a pressing concern, because of the trade-offs in household budgets that have to be made to pay for electricity, with impacts typically on women's unpaid care responsibilities. Frequent overbilling, for example, is a persistent complaint, but one that most respondents live with to avoid disconnections or further surcharges. These class and gender dynamics point to areas of concern around energy that do not impact on public discourse and political events in the way that supply disruptions and outages have.
Different groups may experience energy shortages differently, but our interviews found that public authorities prioritised generation enhancement and supply consistency precisely because outages affect groups and communities across social class divides. Home-based solutions like fuel-run generators are costly and inadequate substitutes, especially for commercial and industrial purposes. This is why pressures to ensure electricity provision came from across elite, middle-class, and working-class actors in the form of protests and in extensive backdoor lobbying.
Interviews with protest participants in Kamalia and Chungi and with decision makers at the field and policy levels pointed to mixed legacies of protest from an empowerment perspective. Protest organisers in Chungi felt protests had been effective: that outages were now less common was, they felt, the direct result of people coming out on the streets to hold public authorities to account. This view was echoed by members of the Kamalia traders' association, who said that ex-Prime Minister Nawaz Sharif had paid attention to the demands of the public when he took office in 2013, unlike Zardari (leader of the Pakistan Peoples' Party) who had ignored them.
At the same time, other protest participants were reluctant to connect their presence on the streets to decisions eventually taken by the government to improve generation capacity. Decision making is notoriously insular in Pakistan, and there is little institutionalised contact between voters, politicians, and those governing the energy sector. Several respondents in Chungi complained that even if the supply of electricity had improved, their concerns about costs had been ignored, and electricity distribution company officials ignored their complaints about overbilling. However, officials interviewed during the research appeared wary of prolonged supply outages given the protests witnessed between 2007-15. Officials explained that 'mobs' posed an acute threat at that time, and that the spectre of electricity protests continues to condition how they tackle outages and public anger.

Political responsiveness without public accountability
Politicians and other senior decision makers echoed the energy concern, noting that the electricity supply had become a political non-negotiable since the peak of the protests. Electricity played an important role in shaping electoral outcomes in 2013, and representatives of mainstream parties acknowledge that an outage crisis can prove a political death knell. A senior politician involved with the electricity sector reform process commented that: MNAs [Members of the National Assembly] would visit Islamabad (the capital) and tell us they were afraid to go into their constituencies because of anger. [A politician] said he had a PhD in swear words and even he hadn't heard some of the abuse words he received during his visit to sootar mandi (yarn market in Faisalabad). This is the party's core support base and it had to respond to it someway.
Although electricity outages had become a matter for political competition and political accountability, affordability and associated service delivery aspects were neither prioritised nor addressed in any meaningful way. The architecture of energy governance saw no substantive change during or after the supply crisis. Decision making continues without political or civil society oversight or public participation. The legacy of contention has, so far, been absorbed into broader service delivery politics focused on supply but has not led to the formation of civil society organisations to articulate citizen voice on the issue.
That the issue of energy has been less overtly contentious since 2015 is not a result of repression. It is more closely associated with the fact that the demand of the most influential segments within those protesting has been met. Remaining concerns around affordability mainly affect poorer populations that are underrepresented in the political mainstream, and do not feature in the rent-driven incentive structure of elite actors. Similarly, that no viable organisational form has emerged to represent citizens' concerns relates to the underlying legacy of weak civil society in the country, and that much organising within the civil society domain is closely shaped by international donor agencies. More work is needed to clarify the implications, but it does seem that international donors have not to date given a high priority to empowering citizens with regard to energy.
We did not find that the Pakistani state reduced space for protests, even during the peak of the crisis. Violent police action was witnessed in a minority of cases, with protracted negotiations and political leverage often deployed by municipal administrators to defuse protests. Part of this can be traced to critical features of the political opportunity structure in Pakistan and reflects the prominent socioeconomic and political profile of those organising the contention, as well as tensions between federal and provincial governments which restrained police deployment. The outage crisis affected everyone, and there were strong sympathies between local administrations, including police, and protestors. These conditions help to explain why there was no escalation in state violence or further restriction of civic space.

Mozambique: 'Authoritarian responsiveness'
Mozambican protestors faced the most closed political opportunity structure when it came to protesting food, fuel and electricity price rises in 2008 and 2010. The state's heavy-handed responses to the eruptions of 2008 and, particularly, 2010 narrowed the possibility of a third major protest when grievances around the cost of fuel and food resurfaced in 2012 and again in 2016-17. Compared to Nigeria and Pakistan, Mozambique's energy protests yielded the weakest evidence that citizens gained power to hold authorities accountable.
Mozambicans' concerns about energy have met with 'authoritarian responsiveness' (borrowing from de Brito et al. 2017) and pushed commitments to 'Energy for All' onto electoral agendas, but without establishing citizen participation in policymaking, or political representation of people's energy concerns beyond this vague discursive commitment.

A lack of organisational support
In this single party dominant political system, there is no strong tradition of organised opposition. MDM had been gaining ground through the period of the protests (2007-13) (de Brito et al. 2017), and urban and peri-urban livelihood concerns help explain why. Civil society groups were neither independent from the ruling party, nor closely connected to popular concerns, with limited accountability to grassroots constituencies and agendas that tended to reflect donor funding priorities. Violent protests in 2008 and 2010 grew in a spontaneous way, amplified, and spread through SMS and popular culture.
Organised actors were not visible in the protests, which lacked obvious leaders. The apparent lack of leadership might help explain why the rapper Azagaia faced charges when a ringtone of his song was deemed to have instigated protests. The state took a heavy-handed approach to repressing major protests, and every time that there seemed to be increased circulation of social media messages calling for popular mobilisation, the armoured vehicles brought in to dampen the 2010 protests were deployed to places where crowds had gathered during those protests to remind the population that people had died because they had protested.
Energy has become more politicised in recent years, in particular with popular complaints about the lack of an energy payoff to Mozambicans from national ownership of the Cahora Bassa hydro-electric dam. These complaints appear to be chiefly channelled through social media, and it is not clear what political impact they have beyond lip-service during election campaigns. The political economy of energy in Mozambique has become a high-stakes game, and there are no reasons to believe that citizens will be granted a seat at the energy policy table.

Responses to urban and peri-urban energy grievances
The Mozambican government has moved 'between responsiveness and repression' (de Brito et al. 2017) in its response to energy and related livelihood protests, but learned to keep a tighter control over media, communications, and protesting publics. In order to keep the prices of privately owned public transport affordable the government introduced a compensation package for petrol stations; this ultimately failed because of administrative delays. In 2017 the government ended the generalised fuel compensation scheme and shifted to a flexible pricing regime with a few of the most politically sensitive targeted subsidies (such as those for diesel-fuelled fishing vessels) being maintained.
The deployment of the army and excess violence against protestors has had the desired effect, in that former protestors have had little appetite for further struggle. One male former protestor from the Maputo neighbourhood of Polana Caniço put the general view in these terms: People demonstrated in the past years, as they did in 2008, but now people are afraid. Government action is wrong. The government is aware of citizens' problems but creates barriers for any type of demonstration. The fear of repression makes people withdraw. The excessive use of force prohibits citizens' options.
(KII 34, Mozambique) Other former protestors were of the view that protest was not only dangerous, it was also pointless. As one young woman from the Maputo neighbourhood of Benfica put it: In this context, social media has become an outlet for popular complaint. While memes and WhatsApp messages are often humorous expressions of widely held grievances, they are also critical, and express a broader sense of discontent with the energy -and other -policies of the regime.

Transnational actors and energy protests
Our research on transnational actors was exploratory, and our findings on the roles of transnational actors in energy protests are preliminary: data collection was disrupted by Covid-19, and research teams found some transnational actors in the energy sector inaccessible, a challenge reflected in the exclusive and closed spaces in which energy policymaking takes place. Worldwide, energy policymaking is a high-stakes activity involving national security and high profits and losses (and frequently extensive corruption); it is (as a result) opaque and closed to public oversight or scrutiny. 9 Our starting assumption -that transnational actors played a significant role in protests by insisting on fuel subsidy cuts as part of multilateral austerity packages -was not supported by our findings. The IMF and the World Bank play a role in Mozambique, Nigeria, or Pakistan, but that role is more mediated, less direct, and possibly less important than other factors, including other transnational actors. However, the link between multilateral agency policies and protests warrants further attention: multilateral agencies have a duty to do no harm, and policy prescriptions can alter the political opportunity conditions that make protests likely. Violence frequently occurs during such protests, and such political effects can be predicted. There is, therefore, a question of accountability for policy prescriptions likely to result in riot, even if the policy choices are ultimately those of governments and not the World Bank or the IMF.
The role of the private sector, including multinational energy firms, in energy policy has also proven difficult to study. This is in large part because of the insulated and secretive nature of energy policy decisions, and that in turn relates to the high stakes in terms of geopolitical and national strategic considerations, and of the massive rents to be earned in the sector. Parts of government have, we could propose, relatively strong incentives to keep private sector involvement in energy a closed matter, and to avoid citizen participation or civic oversight.

Transnational actors in Pakistan's energy regime
Since 2007, Pakistan has had three IMF programmes (2008,2013,2019)  removal was a key pillar of each programme, it was particularly prominent in 2008, which coincided with the uptick in electricity related contention. However, the mechanism through which removal and protest are related is not straightforward. Subsidy removal was largely related to the tariff differential subsidy (TDS) offered by the government to bridge the cost of generation and the prices charged to the end-consumer. The IMF, ADB and the World Bank recommended regular increments in tariff to full-cost recovery, to gradually remove the TDS. Successive governments have been reluctant to comply with this fully, given the inflationary and politically sensitive impact of increases in the electricity tariff. However, IMF programme fiscal deficit conditionalities also meant governments were restricted in what they could spend on the TDS. Without subsidies to cover the cost of fuel and generation, private sector electricity suppliers stopped producing, leading to outages.
Overall, then, the IMF's advice for subsidy removal and tariff adjustment was indirectly linked to the outage crisis (and its ensuing protests) because it placed restrictions on the amount of money the government could spend to keep generation online, while keeping tariffs low. In both the 2008 and the 2013 programmes, electricity tariff rises were a key conditionality, that restrained by political considerations, governments complied with only partially. While IMF restrictions on government spending cannot be blamed for the outages, IMF conditionalities shaped the context for significant electricity related contention. Donor agency interviewees downplayed their influence and role in the electricity sector. A World Bank and an IMF staffer both pointed out the gap between what multilateral agencies had proposed and what policies had been adopted.
Transnational actors played a major role in shaping the overall electricity sector in Pakistan. The primary reform process that led to privatisation of generation and a reduced role of the state on one end of the supply chain was conceived and led by the World Bank. Since the start of Pakistan's private electricity market in the late 1990s, the International Finance Corporation (IFC) has made direct investments of over US$850m in 19 projects. The electricity sector is highly centralised with the federal government bureaucracy controlling almost all aspects of purchase, transmission, and distribution. Political oversight is limited to the prime minister's office and the ministers for finance and energy (or water and power as the case was earlier), with preferential access given to powerful domestic business conglomerates (and increasingly Chinese companies) involved in electricity generation. Since 2015, electricity outages declined dramatically, largely as a result of the China Pakistan Economic Corridor (CPEC) collaboration, which is responsible for US$33bn of investment in the power sector through to 2022.
Given the country's bureaucratic structures and history of authoritarian rule, parliament continues to have a negligible role in energy policy formulation, while civil society organisations working in this space are either absent or too weak to influence political discourse; at the same time, fiscal constraints and structural macroeconomic vulnerabilities that lead to periodic crises enhance the role played by lenders and donors in the country's political economy. It is within this environment that transnational actors and technocrats retain a privileged and considerably insular place in framing the debate around energy in Pakistan, one that has not changed much even after a prolonged period of crisis between 2007-15.
External investment in energy infrastructure is likely to be highly contentious, as China's increasing footprint on Pakistan's energy sector from 2013 onwards indicates. Electricity generation projects, sponsored by Chinese loans and built/operated by Chinese companies, have not been subject to substantial political oversight or deliberation. The government's own investigation into the financial health of the electricity sector revealed that in the case of two large coal power projects, Chinese investors asked for an unconditional acceptance of an up-front tariff, which was accepted by the National Electric Power Regulatory Authority (NEPRA). Concerns raised by environmental and food security activists have been ignored, justified by the government with respect to national security and development. Public authorities have repeatedly cited national security and geo-strategic factors as justifications for the lack of transparency and heightened centralised control around Chinese investment in the country (Boni and Adeney 2020).

Transnational actors in Mozambique's energy sector
In the past two decades, energy has moved to the centre of Mozambique's international politics and economics. First, with the reversion of Hidroeléctrica de Cahora Bassa (HCB), the company that runs the Cahora Bassa dam which was built by the Portuguese colonial regime in the 1970s with the explicit intention of earning export revenues by selling power to South Africa and Rhodesia (Isaacman and Isaacman 2013). Second, with the expansion of the exploitation of fossil fuel resources -in particular gas and coal -in Inhambane, Tete and Cabo Delgado provinces. The Tete coalfields have attracted Brazilian, Indian, and Anglo-Australian mining companies, SASOL of South Africa is expanding its natural gas extraction operations in Inhambane, and US, French, Italian and Chinese firms have sought stakes in the huge natural gas basin off the coast of Cabo Delgado, with British Gas emerging as one of the major prospective clients for the region's LNG. International oil firms (including GALP from Portugal and BP from the UK) also dominate the cartel of distributors that negotiates international transport fuel purchases with the Mozambique government and controls distribution through the country's network of petrol stations. Despite their huge influence these external actors are not visible in the country's day-to-day energy politics and as a result residents in Maputo do not seek to hold them accountable to policies related to fuel prices or access to and quality of electricity. ESKOM, the South African energy utility which buys HCB's electricity and then resells it to EDM at a much higher price, is not perceived as a driver of the dysfunctional electricity supply situation, which most simply blame on the corruption of Mozambique's own ruling elite. The external actors that play a more visible role in struggles over energy access, quality and affordability are the donor agencies and international financial institutions (IFIs) who are increasingly active in the sector. The role of the IMF and World Bank in pushing for subsidy reform is acknowledged, though they are not considered responsible for the rising prices of electricity. Bilateral donors, including USAID, the UK Department for International Development (DFID, now FCDO), Deutsche Gesellschaft für Internationale Zusammenarbeit (GIZ -German organisation for international cooperation), the Netherlands and Norway, are most active in the off-grid solar sector (through programmes such as the UK aid funded BRILHO programme), with only USAID taking on a high visibility role in pushing for liberalisation of the electricity sector and regulatory reform (Nhete 2007;Cipriano et al. 2015;Lopes and Bambo 2016;Power et al. 2016). As noted above, development partners are cautious in their negotiations with the government about cutting subsidies as it opens the possibility of violent repression of protests.
In the 2010s, with support from the Brazilian government, Mozambique unsuccessfully attempted to generate biodiesel from jatropha and palm oil and ethanol from sugarcane. Brazil and India have been major players in developing plans for the export of 99 per cent of Mozambique's coal output, mostly for steelmaking in Asia, though Brazilian mining giant Vale recently announced that it would be selling its operations in the Tete coalfields as part of a plan to exit the coal business. International donors and the energy interests of national governments as well as international energy initiatives such as the Power Africa programme and the UN-led climate change convention process have been major influences on energy policy. South Africa reaps most of the benefits of affordable energy from Mozambique's Cahora Bassa dam, and militant groups have emerged in the struggle over the Cabo Delgado gas fields, the development of which have not benefited the local population. Foreign investment in the development of the gas fields has taken place without public deliberation or transparency, and there have been concerns about the terms of contracts and the environmental and social impacts of resettlement programmes. Delgado led by French oil major TOTAL, more than the UK gives for energy transitions or access through its aid programme. The frustrated expectations generated by gas development in Cabo Delgado have helped to drive a local insurgency that has now escalated to the point where TOTAL has been forced to put the project on hold. The violence in Cabo Delgado, which has seen almost 700,000 people forced to flee their homes and numerous reports of atrocities committed by government forces and South African and Russian mercenaries working with them, as well as by the Islamist rebels, has combined with a worsening outlook for global gas prices to trigger a collapse in investor confidence in the gas fields of Northern Mozambique. The failure of promised foreign investment to materialise is likely to push the government back towards greater reliance on the IFIs and bilateral donors.

Transnational actors in Nigeria's energy sector
Nigeria is the largest oil and gas producer in Africa, the ninth largest gas producer in the world, and has the tenth largest crude reserves in the world (BudgIT 2019). Such wealth has attracted transnational involvement since its discovery in 1956. Multinationals including Shell, Exxon, Chevron, and other major oil companies extract and export up to 2m barrels a day of high quality or low sulphur 'Bonny Light' crude from the Niger delta. Very little is refined in Nigeria because its four state-owned refineries are non-functional. International traders export around 900,000 tonnes a year of low-grade, 'dirty' fuel, produced in Dutch, Belgian and other European refineries to Nigeria (Vidal 2020). With proven crude oil reserve estimates of about 37 billion barrels in 2015, Nigeria has the world's second largest crude reserves. However, Nigeria is also one of the largest consumers of refined products in Africa and accounts for over 7 per cent of Africa's refined products consumption. Imports currently account for over 80 per cent of Nigeria's refined product supply (PWC 2017). Efforts on the part of the Nigeria National Petroleum Corporation (NNPC) to overhaul its refineries have to date failed. In 2016, following a plunge in oil prices, sabotage of oil installations and crude theft, Nigeria sought external financing for its refineries, but investors were unconvinced about the viability of the venture (Bala-Gbogbo 2019). It is very difficult for refineries to be viable when the price of their main output -fuel -is heavily subsidised.
Political leaders and champions of energy sector reform have long struggled with the powerful alliances of domestic and foreign interests that ensure the governance of energy remains opaque. As the Nigerian Finance Minister at the time of the 2012 protests, and since 2021 the Director-General of the World Trade Organization, Ngozi Okonjo-Iweala, put it: The story of Nigeria's oil and gas sector is ugly. Although revenues from the sector have, to a substantial extent, helped finance the The World Bank and IMF have consistently supported fuel subsidy elimination in Nigeria as part of broader macro-fiscal and oil sector reforms. Their support has been based on an assessment of the real, opportunity and implicit costs of the vast petrol subsidy, which had amounted to 0.4 per cent of GDP by the time it was, supposedly, removed in 2020. Apart from the price distortions and corruption engendered by the subsidy scheme, both the World Bank and the IMF viewed it as a significant constraint on fiscal space, and therefore on investments in human development (IMF 2019a; World Bank 2021).
Multilateral support for subsidy elimination has also depended on an assumption of the possibility of investments in social protection on a significant scale to alleviate the effects for the poorest; these are assumptions which the majority of Nigerians do not appear to share.
For example, in 2019, the IMF noted that: Analysis based on Nigerian household data shows that a removal of the implicit subsidy (through an increase in the regulated price of Premium Motor Spirit, PMS) would reduce income inequality and, if combined with targeted transfers, could result in a decrease in the poverty gap while leaving significant fiscal space for other productive expenditures.
(IMF 2019b: 23, emphasis added) That 'targeted transfers' could be implemented so as to alleviate the pain of higher fuel costs seems a significant leap of faith, given Nigeria's own recent past experience. An IMF account of the Nigerian experience with fuel subsidy reforms in 2012 noted that the government had been unable to persuade its political allies or the wider population of the need for reforms, despite a sixmonth campaign of advocacy and information (Alleyne and Hussain 2013): This proposal [to cut fuel subsidies drastically] was hotly debated in the press, by business and civil society groups, and it was debated in the National Assembly during the rest of the year, with the government strongly trying to make a convincing case. On January 1, 2012, the price of gasoline was raised to a cost-recovery level-a 117 percent increase. The price of kerosene, a cooking fuel used mainly by poorer households, was not changed. However, in response to intense social unrest, the government scaled back the The state had failed to build trust in its willingness or capacity to reinvest fuel subsidy funds in social policies and programmes, which citizens widely believed would instead be lost to corruption and inefficiency. Indeed, the SURE-P programme, announced just before the 2012 cuts, did not survive the tenure of President Goodluck Jonathan. Nevertheless, up until the removal of the fuel subsidy was announced in 2020, World Bank and IMF official documents continued to profess the hope that social protection instruments would be designed and delivered to mitigate the pain caused.
Yet the popular perception that the Bank and the Fund forced Nigerian governments to cut subsidies, perhaps by withholding finance, is not upheld by the research findings. A representative of the NPEC said that the World Bank and the IMF raised subsidy removal in negotiations over loans, and that diplomatic representatives similarly lost no opportunity to push the government to cut subsidies. Yet several respondents noted that the pressure of international actors like the World Bank, IMF, and donor governments had not been effective, in large part because politicians and government actors had stronger incentives to retain subsidies, including the need to win elections and the material benefits from the leakages in the subsidy system.
Resistance to the recommendations of the IMF and the World Bank is popular with the wider public: the devastating effects of structural adjustment left a legacy of popular antipathy towards the multilaterals, which has enabled opposition political actors to resist or even mobilise against reforms. Nevertheless, the government announced the removal of the mechanism for setting fuel prices in 2020, indicating its intention to leave fuel prices to market forces, an intention that had not (as of April 2021) been enacted. At the time, during the Covid-19 pandemic, protest was relatively muted. However, since then there have been a large number of protests against police brutality, among other issues.
The IMF takes a paradoxical view of what the reduction of fuel and electricity subsidies may mean for living costs and widespread discontent and protest. At the height of the coronavirus pandemic, they noted that: The authorities have undertaken commendable and timely measures to counter the pandemic's impact on lives and livelihood. These encompassed palliative measures, a strict lockdown, removal of fuel and electricity subsidies as well as fiscal and monetary policy support for the health sector and the economy. (IMF 2021: 5, emphasis added) This suggests a view that the removal of subsidies was a 'commendable and timely' way of cushioning the effects of the pandemic on people's lives. This very likely reflects the IMF's view that subsidies are inequitable and a drag on social spending, and that in the medium and long term, the pain to the Nigerian public will be lessened through their removal and the reallocation of these resources for other purposes. The IMF acknowledge the scale of the economic and health crisis facing Nigeria. In the same report they also note that: In October, mass protests against police brutality started under the ENDSARS movement. [10] Social discontent has widened since the beginning of the crisis with rising food inflation, elevated youth unemployment and high social inequality. Headline inflation reached 14.2 percent in October 2020, a 30-month high, while the unemployment rate reached 27 percent in 2020Q2, with youth unemployment increasing to 41 percent. Surveys conducted by the Nigerian authorities also show a significant economic impact on the vulnerable, including a precarious food security condition. (IMF 2021: 6) However, the IMF position ignores two important ways in which subsidy reforms affect mass protests and social discontent. First, while oil prices were low during the pandemic, they have risen rapidly in recent months, causing the subsidy to escalate. The IMF will have been aware of this possibility but does not seem to have fully considered the hardship which would be caused by (and therefore the political infeasibility of) steady rises in domestic fuel prices as the oil price recovers. Second, the multilaterals should be in a position to know that violent responses to justified protests scar the relationship between citizens and their state, reducing trust and making reform more difficult. In situations such as Nigeria in 2020-21, there are strong arguments to be made for reform agendas to adopt pathways that are slower but have a lower risk of violence. The chances of building social consent over the longer term are likely to be greater.
International private actors have been implicated in the subsidy scams. Respondents suggested complex connections between multiple international private actors and fuel marketers exploiting loopholes in tender processes and subsidy schemes. The subsidy scam had three transnational dimensions. First, oil firms had international networks and used loopholes in registrations and licensing to enable their actions to benefit from the schemes. According to one oil and gas consultant respondent, private marketers set up offshore businesses or firms abroad to complicate the chain of actors, with each actor adding a 10 The ENDSARS movement targeted police brutality in Nigeria, and in particular the extrajudicial actions of the Special Antirobbery Squad (SARS premium, in order to present higher bills to the Nigerian government. As Nigerians could set up firms in other countries as part of the chain, the boundary between domestic and transnational was blurred. Second, individual actors involved in subsidy corruption were enabled by transnational financial networks that enable flows and stores of illicit wealth (such as multimillion-dollar properties in Dubai) (Page 2020). Thirdly, not just Nigerian but also courts in other countries have been involved in bringing fraudulent actors to justice. The CEO of Capital Oil was charged by a court in the UK and had assets frozen worldwide. Another UK court jailed the chairman of Ontario Oil and Gas for international fraud and money laundering (Olowookere 2018).
The funds for subsidies cannot be seen in isolation from the revenues generated through oil production, and licenses regularly face allegations of corruption involving international private actors (Oseni 2017;Hildyard 2018). In 2014, accounts of corruption in the Strategic Alliance Agreement (SAA), the scheme through which NNPC allocates oil fields to private actors, illustrated how a web of transnational and Nigerian actors prevent revenues from oil production from reaching federation accounts. In 2013, the Central Bank Governor Lamido Sanusi accused the NNPC of failing to remit billions of dollars from crude oil production, as required by law, to the federation account. He claimed that the SAA was a fraudulent scheme through which local and international oil companies looted the national treasury, and was fired for his troubles (BBC 2014) Nevertheless, in 2014, the IFC committed US$150m to one of the companies implicated in Sanusi's accusations, which watchdog groups considered to be lacking in due diligence (Hildyard 2018).
Subsidies have also leaked through illicit transnational networks of smugglers between Nigeria and neighbouring countries in the region (Mlachila et al. 2016). The court cases resulting from what is widely known as 'the subsidy scam' have brought limited accountability and have not affected a more transparent governance of the fuel subsidy schemes. The complex of political and economic actors, which has transnational dimensions, are a major barrier to enhancing the accountability of the schemes.

Conclusions
Protests about energy are increasingly common and consequential, and on notable occasions in the current century have unseated regimes, unleashed state violence, disrupted economic and social life, derailed economic reforms, and triggered broader movements against corruption and misrule. Most research on energy protests has to date focused on protests against fossil fuel use, mostly in the global North. There has been less analysis to date of protests against energy price rises and scarcity. These are concentrated in the global South and involve the fight for the right to affordable energy, which government and protestors alike often consider most easily delivered via subsidies. These two types of energy protest are at odds in their immediate aims, yet reflect a common concern about energy policymaking: that it is undemocratic, unjust, and unsustainable. A starting assumption of the present research has been that energy price and scarcity protests are significant political events, which need to be understood and responded to if energy transitions are to be just or sustainable.
This paper presented a synthesis of findings of research into protests against energy price rises and scarcity which aimed to centre protestors' concerns about energy. It was motivated by a recognition that these events were politically significant, illuminating the nature of the contemporary state-society relationship in an era when modern forms of energy are increasingly necessary for everyday life in the global South. The research also recognised that such protests frequently block efforts to cut fossil fuel subsidies, which are deemed damaging on fiscal and environmental grounds. Policymakers concerned to advance the cause of fossil fuel subsidy reduction need to understand why energy protests occur, why they escalate, and why the grievances they express can persist, even when protests die down. Through literature review, cross-country analysis of violent fuel-related protests, and three in-depth case studies of major protest episodes in Mozambique, Nigeria and Pakistan, the research was designed to explore why and how energy protests escalated, and whether political fragility and conflict were contributing factors.
The research found that protests triggered by grievances about energy acquired a particular potency in settings where other means of expressing grievances were absent, weak, or blocked. But there have been major energy price and scarcity protests across political and energy regimes, levels of development, and regions. While economic and political conditions in fragile states may create opportunities for such protests to escalate, energy protests can create or contribute to political fragility, depending on how governments respond to them. The research found little evidence that the raw power of energy protests translated into sustained empowerment of citizens with respect to energy policy, or in their relations to the state more generally. Many people perceived protests to have won power, but these gains could be subverted, stolen, or subsumed within other agendas. Meaningful gains in citizen power after energy protests occurred were short-lived and diffuse, dissipating in the absence of institutionalised mechanisms of citizen empowerment, or of sustained civic organisations or movements. Protestors learn that they can force governments to listen, but that protests rarely elicit more than a heated public hearing, followed by often temporary policy change. Governments learned new tactics for cutting subsidies, resorting to repression, or waiting for a more opportune moment.
The limited evidence that energy protests empowered citizens was matched by a lack of public accountability to citizens, for energy or other essential public services. In the more disruptive episodes of protest, governments backed down on proposed price rises or introduced short-term fixes for electricity shortages. State tactics were to ward off protests, but rarely involved a visible strategy of strengthening public accountability for energy policy. The fleeting sense of power that protestors gained when they did succeed created a self-reinforcing cycle: protestors believed these were their only tools, while governments feared the protests, but expected the resistance to dissipate in time.
Energy contention uncovers a 'wicked problem': while neither citizens nor the state are satisfied with the outcomes, the tested strategies of each deepen the impasse and reinforce the conflict. Current strategies for subsidy reform and energy policy formulation offer no solutions to addressing citizens' energy grievances.
Transnational actors play a range of roles in this landscape of energy protests, in ways that are often hidden or obscure. Multilateral aid actors influence subsidy reform, but through indirect means. Multinationals and foreign governments are critically important actors in national energy policy in the global South and are part of the reason why energy policymaking remains closed and secretive.
In this context, it is notable that aid donor support for civil society and good governance has not supported popular domestic movements for climate justice or energy security. In countries in which international aid has been an important resource for civil society, this helps to explain why civil society channels for citizens' energy grievances are blocked. But while transnational policy and business actors are implicated in contention around energy in the global South, transnational (Northern) civil society actors play no such role in popular protests against energy price rises and scarcity (or energy poverty).
A key conclusion of the research is that citizens' protests, in settings where the state has little restraint on its violence, and alternative means of expressing grievances are limited, have been the desperate last resort of people hardpressed to cope with rising costs of living and inadequate energy supplies. These need to be recognised as political claims, which seek to mandate states to ensure people have access to the modern forms of energy they need for daily life. The protests fail to empower citizens to hold their states to account for energy. If anything, states retreat further into insulated and technocratic policy spaces to resolve energy problems. A key lesson is that when the political opportunity structures permit, people will protest against energy policies that prevent them from meeting the living standards they need or expect, in large part because they have no other means of making governments listen to their complaints.
There are no easy answers to these complex matters. But without understanding why people protest against cuts to energy subsidies, there is no chance of designing lasting reforms that move societies towards fair energy transitions and climate justice. The evidence marshalled here points towards responses that appropriately and effectively centre people's experiences as users of energy. Citizens need to be engaged in energy policymaking, rather than having solutions imposed upon them which will not stick, or if they do so, will prove neither sustainable nor just.