Climate Change, Energy Transition, and Constitutional Identity

Abstract

Through its potential to contribute to mass suffering, economic disruption, and social unrest, climate change poses a security threat to the constitutional identities of states (as democratic, autocratic, or hybrid regimes). This paper proposes a conceptual framework of mediated causality for climatic impacts on constitutional identity and engages in novel theory-building for one mediating vector of change: the post-fossil energy transition. Theories of the “oil curse” and of “carbon democracy” are compared and critiqued for their contributions to understanding the potential impacts of decarbonized energy systems on democracy. Two counterintuitive conclusions emerge. First, transitioning away from petroleum may not result in increased democratization, as the oil curse implies. Second, post-fossil energy systems are unlikely to become structurally decentralized, as advocates of “energy democracy” suppose, and may even need to remain centralized in order for popular mobilization around energy to help maintain or reinvigorate democratic rights.

El cambio climático supone, por su potencial para contribuir al sufrimiento en masa, a los disturbios económicos y al malestar social, una amenaza para la seguridad de las identidades constitucionales de los Estados (como regímenes democráticos, autocráticos o híbridos). Este artículo propone un marco conceptual de causalidad mediada para los impactos climáticos sobre la identidad constitucional y se adentra en una novedosa construcción teórica de un vector mediador del cambio: la transición energética post-fósil. Se comparan y se someten a crítica las teorías de la «maldición del petróleo» y de la «democracia del carbono» por su contribución a la comprensión de las posibles repercusiones de los sistemas energéticos descarbonizados en la democracia. Se desprenden dos conclusiones contraintuitivas. En primer lugar, es posible que el abandono del petróleo no se traduzca en una mayor democratización, como sugiere la maldición del petróleo. En segundo lugar, es poco probable que los sistemas energéticos post-fósiles se descentralicen estructuralmente, como suponen los defensores de la «democracia energética», e incluso es posible que deban seguir centralizados para que la movilización popular en torno a la energía contribuya a mantener o revitalizar los derechos democráticos.

Par son potentiel de contribution à la souffrance de masse, aux perturbations économiques et aux troubles sociaux, le changement climatique menace la sécurité des identités constitutionnelles des États (en tant que régimes démocratiques, autocratiques ou hybrides). Cet article propose un cadre conceptuel de causalité médiatrice pour les répercussions climatiques sur l'identité constitutionnelle et élabore une théorie inédite d'un vecteur de changement médiateur : la transition énergétique après les combustibles fossiles. Les théories de la « malédiction du pétrole » et de la « démocratie du carbone » sont comparées et critiquées par rapport à leurs contributions à la compréhension des conséquences potentielles des systèmes d’énergie décarbonée sur la démocratie. Deux conclusions qui défient toute logique apparaissent. D'abord, la transition pour se libérer du pétrole pourrait ne pas engendrer une plus grande démocratisation, comme l'implique la malédiction du pétrole. Ensuite, les systèmes d’énergie après les combustibles fossiles ont peu de chances de devenir structurellement décentralisés, comme l'entendent les partisans de la « démocratie énergétique ». Ils pourraient même être contraints de rester centralisés afin que la mobilisation populaire autour de l’énergie permette de maintenir et de revigorer les droits démocratiques.

Introduction

Early in 2022, the flares atop oil and gas installations around Kazakhstan were eclipsed by other flames from burning buildings and automobiles in nearly a dozen Kazakh cities. Mass protests briefly endangered the incumbent government after it implemented policy changes that resulted in the doubling of fuel prices for consumers (Auyezov 2022). A similar dynamic of popular unrest that rose in tandem with energy prices had previously featured in the Arab Spring of 2011, when civil wars and changes of regime were ignited in a few states in the Middle East and North Africa (MENA).

Less than 2 months after the Kazakh fuel protests, Russian military forces entered Ukrainian territory in an apparent effort to replace their neighbor's incumbent government. This conflict also implicated the economics of fossil-fuel energy. In addition to the consequences for western Europe of denial of Russian supplies of natural gas in 2022 and 2023, European states’ previous efforts to decarbonize their economies threatened to disarm the fossil-fuel weapon and thereby diminish Russia's geopolitical leverage as a major supplier, making Russian military action against Ukraine seem more advantageous to Moscow in the near term (Harvey 2022).

Events of this type illustrate the concerns that motivate scholars to investigate the intersection of energy security and geopolitics. However, they also implicate a larger field of relationships involving climate change and regime change. In Kazakhstan, the rise in energy costs was reversed to pacify violent protests in 2022, but future changes in climate could impose intolerable costs on a wide range of places and for a wide range of reasons, including but not limited to the price of obtaining energy. Also because of climate change, but outside the field of energy politics strictly speaking, electoral processes and civil liberties in Europe and the United States have been affected by flows of economic and political refugees from MENA and Central America, respectively—flows that have been swelled by severe drought or intensifying storms, or both (Goldstone and Diamond 2020, 867–68; ODNI 2021, 18). At some threshold of pressure, institutions either crumble, adapt, or get redesigned. It should therefore be a central concern in the study of environmental security to investigate the potential impacts of various kinds of ecological disruption on political institutions.

Academic political research lacks a well-developed theory to guide such an investigation, but related mid-range theories and empirical literatures abound. The goal of this article is to advance the theory-building process for the interaction of biophysical (independent) variables, socioeconomic (intervening) variables, and political (dependent) variables.

The argument below proceeds in three stages. In the second section, I defend the premise that change in the constitutional identities of states, and specifically the threat of de-democratization, constitutes an important but overlooked impact related to ecological degradation in general and climatic instability in particular. In the third section, I suggest that the problem of climatic impacts on political institutions should be conceptualized in terms of mediated causality, before outlining a conceptual scheme in which ecological factors of change are distinguished from the social and economic vectors through which they exert pressure on political institutions. In the fourth section, I focus on energy sources and systems as one of the socioeconomic vectors, comparing and critiquing two social-science theories (the “oil curse” and “carbon democracy”) for their utility in understanding the potential political impacts of a post-fossil energy transition. Whereas the literature on the oil curse has been empirically rich but theoretically uncertain, the literature on carbon democracy has been theoretically provocative but empirically underexplored. I conclude by suggesting how to combine the strongest traits of these two research agendas in future studies of energy transitions and their regime-level impacts.

Environmental Security and Constitutional Identity

This section explains how change in the constitutional identities of states could be regarded as a problem of international environmental security and articulates the basic conceptual parameters of constitutional identity, including democratic, autocratic, and hybrid regime types.

Two broadly contrasting conceptions of security foster different approaches to identifying threats from ecological change, with a third, bridging approach also available. For any of these three understandings, change in the constitutional identity of a state—its most basic, regime-level institutions—presents a plausible threat of insecurity.

A traditional conception of “national” security refers to the threat of organized violence from an external source, thereby prioritizing military preparations against such threats (Haftendorn 1991; Walt 1991). A second, broader notion of “human” security includes nonmilitary factors such as social, economic, and ecological preconditions of decent human existence (King and Murray 2002; Mitchell and Carpenter 2019). The second approach has the potential to bring problems of ecological change into a closer, more constitutive relation to the concept of security (Goldstein 2016). A middle path is indicated by Levy's focus on the “drastic[] degrad[ation]” of “some of the nation's most important values” (Levy 1995, 40), a definition used to identify decreasing atmospheric ozone and increasing atmospheric carbon as “direct physical threats” (Levy 1995, 46–54).

For any of these three approaches, the potential for change in the constitutional identities of states represents a major threat. Levy's focus on national values could be broadened beyond physical welfare, leading us to regard the democratic, autocratic, or hybrid character of a national political regime as a value in need of securing in and of itself. Indeed, Busby's recent conception of climatic security, as a third way between the poles of “national” and “human” security, explicitly characterizes regime stability and national “way of life” as security concerns (Busby 2021, 19–25). Even on the narrowest possible approach, however, the capacity of states to provide military, economic, or other services to human populations would likely be affected by alterations in their basic constitutional structures. Those interested in either national or human security, or something bridging the two, should therefore be equally concerned about the possibility of regime change induced by climate change.

But what sorts of change, exactly, might be involved in the constitutional identity of states? Since political regimes are ensembles of institutions, particular institutional differences determine how we apply labels such as “democracy,” “autocracy,” and “hybrid regime.” A substantial body of opinion in modern social science holds that the institutional hallmark of modern democratic regimes is the presence of periodic competitive elections (e.g., Dahl 1998). Autocracies do not have them, and hybrid regimes have them only in partial or semifunctional forms.

Elections are central to regime types to the extent that they assist popular power as against elite power (Bollen 1990). Voting contests in which larger numbers trump smaller numbers are supposed to mitigate the non-electoral advantages of small but cohesive groups (i.e., of elites). Of course, elections can be tilted back in elites’ favor in various ways; hence the importance of the modifiers “periodic” and “competitive.” Democratic theorists believe that elections must occur at regular intervals, in an iterative process, because holding one election and then stopping all future elections could result in an entrenched elite. They also tend to think that elections must be competitive, so that popular pressure could in principle result in substantial changes of leaders or governing mandates. If an electoral process were not truly competitive, the result would be similar to stopping elections altogether, from the perspective of the balance between popular pressure and elite prerogative.

Moreover, making elections both periodic and competitive is often taken to be impossible without certain formal rules and privileges, principally freedoms of expression and association (Dahl 1998, 96–98). This assumption adds a second institutional layer of rights and liberties to the minimalist vision of electoral democracy.

It may be a Western conceit, fostered under contingent cultural and economic circumstances, that elections cannot as a practical matter be meaningfully open or reliably competitive without effective bills of rights to protect all citizens from restrictions on expression and association. I therefore distinguish three layers of meaning in the phrase “democratic regime.” At the conceptual level, basic democracy is any political regime in which popular pressure prevails over elite prerogative. This is a regulative ideal more than a regime type per se. Next, at the first operational level, electoral democracy is a regime based on periodic competitive elections. At the second operational level, liberal democracy includes periodic competitive elections while adding effective civil liberties.

Even where joined together, competitive elections and civil liberties are insufficient to guarantee popular control over political elites. Empirical scholarship has demonstrated the partial or mixed success of electoral accountability in established democratic regimes (for reviews, see Anderson 2007; Maloy 2014), and historical scholarship has revealed that the early modern originators of the idea of electing representatives viewed the mildness with which the practice actually constrains political elites as one of its virtues (Manin 1997; Maloy 2008). However, elections are still widely regarded as necessary to institutionalize popular power in large-scale modern states, even if not sufficient to do so. They create possibilities for democracies that autocracies lack.

Ultimately, if we are interested in identifying the potential for change in constitutional identity, we need to examine specific institutional impacts related to elections or liberties, or both. Before attempting to follow the causal chain to its end, however, we must first recognize the conceptual complexities associated with placing ecological conditions at the beginning of that chain.

Climate Change and Mediated Causality

This section describes a factor-vector model of mediated causality for investigating the relation between climatic change and political change. I isolate six factors of climatic disruption that appear capable of operating through six vectors of social change; these six vectors, in turn, channel climatic dynamics into political outcomes. (For an alternative scheme of mediated causality that links climate change and democratization in particular, rather than institutional change in general, see Burnell 2012, 818–20.)

I am posing here a different sort of question from the ones usually debated in academic studies of ecological conditions and political institutions. The kind of theory we need is a causal theory more than a normative theory. The latter type investigates questions relating to the priority of one norm over another—for example, when the benefits of environmental protection should take precedence over the benefits of liberal–democratic rights (Mittiga 2022)—whereas causal theory is concerned with the priority of cause over effect. Within this latter type of inquiry, it is common to analyze political institutions as causes and ecological conditions as effects, yielding a literature in which democracy's theoretically superior performance over autocracy has been met with only “mixed” or highly conditional empirical results (e.g., Bernauer 2013; Lagreid and Povitkina 2018; Kammerlander and Schulze 2020). Here, however, we have to reverse the causal arrow: What impacts could climate change or related ecological transformations have on political institutions?

In order to place such questions into a coherent theoretic framework, we must be talking about mediated causality. Ecological change induces certain first-order changes in human activity that, in turn, exert second-order effects on political institutions.

For ecological factors, consider a number of the most direct effects that humans are experiencing or will experience as a result of climate change. (What follows is based on Brooke 2014, 564–68, Wallace-Wells 2019, and Human Rights Council 2019, §§ 3–15.)

1. Direct heat. Rising average temperatures are expected in most parts of the globe, yielding more heat-related deaths and fewer productive labor hours, especially for people with limited access to air-conditioning. Material infrastructure such as railroads and pavement will be subject to more rapid deterioration and higher costs of maintenance.

2. Storms and disasters. More frequent and more destructive weather events result from higher average temperatures. Natural disasters increase anxiety and stress among victims. Flooding may cause the abandonment of some inhabited areas, and public spending on flood prevention may increase to keep others viable. Insurance markets may collapse in some areas, probably preceding residential or commercial abandonment. Public spending may rise to bail out property owners and insurance companies.

3. Sea-level rise. Coastal industries and residential communities may face significant economic costs of adaptation or relocation. Ports and refineries tend to be concentrated in such regions. The importance of seaborne trade to the energy sector in particular and globalized commerce in general heightens the potential economic impact.

4. Sea-temperature rise. Fisheries have already started migrating to new waters to find ambient temperatures that enable them to survive and reproduce. Coral bleaching results from higher temperatures (and lower oxygen content) of seawater, and healthy coral is critical to supporting a wide range of marine life used as seafood.

5. Rainfall patterns. More frequent and longer-lasting droughts are already observed in some regions, but higher annual rainfall may be expected in others.

6. Species relocation and extinction. All of the above factors may contribute to significant geographic shifts in the habitat of various nonhuman species, affecting multiple conditions of human life such as food supplies and disease-carrying agents. Extinction of some species may have health-related effects on humans.

Turning from these factors of climate change to vectors of human activity gets us closer to political change along the causal chain. These social and economic vectors, and the “indirect pathways” that causal processes take through them (Busby 2021, 32–35, 64–67), are already well known to scholars of international security: war, agriculture, revenue, migration, energy, and disease (WARMED).

• War. Organized violence is historically one of the most important triggers of revolutionary change (Skocpol 1979), as well as lesser changes in political institutions. A substantial empirical literature over the last three decades has investigated the environmental aspects of both internal and cross-border violence (Diehl and Gleditsch 2001; von Uexkull and Buhaug 2021), with mixed results for the salience of climate change as a traditional (military) security concern. Plausible arguments have been raised that certain dynamics of future climate change, including geoengineering as a mitigative tactic, could generate new threats to security within the state system (Sears 2021, 16). In this sense, studying previous historical drivers of climate-induced conflict may understate the true threat.

• Agriculture. Production of food may rise for certain regions and crops (especially in higher latitudes) while declining for others (especially in the subtropics and tropics). Even in higher latitudes, more frequent harvests and higher yields could be nullified by the destruction of whole crops through extreme heat and reduced rainfall. Impacts on fisheries should be considered as part of the general vector of food production. (Compare Nordhaus 2013, 82–90, with Wallace-Wells 2019, 49–58.)

• Revenue. Adaptation to and recovery from the adverse impacts of global warming will be costly. A conservative estimate gives a range of −1 to −5 percent for the likely annual impact on total global output, implying net contraction for most countries (Nordhaus 2013, 139–40); other estimates include an impact on average incomes as severe as −23 percent (Human Rights Council 2019, § 40). (Such estimates generally take the year 2100 as a benchmark, implying less extreme but still negative impacts in prior years.) For public finances, climate change promises to increase demand for direct aid and infrastructure improvements while decreasing tax revenues. International and central banks may or may not be able to supply adequate liquidity in a context of declining or negative real growth.

• Migration. As disease-carrying organisms and food supplies relocate, residential communities themselves will be pushed and pulled within and across existing states’ territories. Some political volatility is bound to result from internal migration alone, as subnational jurisdictions are forced to incorporate new residents (Fischer 2017, 208). Given that nearly every warming forecast includes disproportionately severe impacts on poor states, increased migration into rich states is widely expected. More nationalist and autocratic styles of politics may therefore gain further electoral purchase (Ghosh 2016, 194–95; Human Rights Council 2019, §§ 65–66).

• Energy. The use of energy is essential to all social processes, including production of greenhouse gases and adaptation to their impacts. Policy demands for reducing carbon pollution imply major changes to energy-producing and -consuming practices, while the potential for price volatility (as with agricultural commodities) has significant implications for social and political unrest.

• Disease. Modern medicine and nutrition reduced human mortality under relatively stable climatic conditions after around 1800, but future warming trends may disrupt long-established social and technical infrastructures for basic bodily security. Strains on supplies of fresh water under higher average temperatures portend physical weakness and psychological desperation (Wallace-Wells 2019, 86–93). Rapid spread of disease in the wake of significant deviations from average temperatures has historically been a source of political instability in contexts ranging from ancient Rome to nineteenth-century China (McMichael, Woodward, and Muir 2017, 150–59, 218).

Why are these six vectors the relevant ones for investigating political–institutional change as the outcome of interest? War, agriculture, and disease can motivate people to reconsider established institutions because they directly implicate individual and collective survival. Migration can become similarly salient in contexts that raise questions of communal identity: who exactly is included in or excluded from the collective entities whose survival is deemed politically paramount. Energy and revenue pervade all aspects of state capacity to promote security: they are the biophysical and human currencies (respectively) without which collective action cannot be organized and executed. Later, in the concluding section, I will address more specific institutional impacts on democracy's two main components, competitive elections and civil liberties.

Two critical points emerge from the factor-vector model of mediated causality between climatic change and political change. First, multiple climatic factors may feed into a single social vector. For example, intense heat waves could raise policy demand for low-carbon energy sources while sea-level rise could impose high costs of adaptation on existing high-carbon energy infrastructure. Second, one of the six social vectors may interact with another before making its impact on political institutions. We will now compare two theoretic approaches to the E in WARMED. The first approach conceptualizes the energy vector's role in political change by passing it through the revenue vector; the second formulates a more direct kind of impact for energy on politics.

Energy Regimes and Political Regimes

This section probes what two theories—the “oil curse” and “carbon democracy”—can tell us about likely impacts on political institutions if and when the fossil-fuel era wanes substantially while some sort of post-fossil regime takes its place.

How likely is a global post-fossil energy transition? There is now significant political and diplomatic pressure for states and firms around the world to leave fossil fuels “in the ground” while increasing reliance on decarbonized energy-producing processes involving solar cells, wind turbines, and nuclear reactors. Even without voluntary, politically driven change, increasing costs and diminishing returns of future extraction of available fossil fuels (Ahmed 2017, 22; Hall and Klitgaard 2018, 388–99, 494–95) could incentivize a large-scale energy transition. The extent and pace of this kind of shift are far from certain, and it would likely be a matter of generations rather than years or decades (Grubler 2012; Smil 2016).

The Oil Curse

The most developed academic research agenda on the relationship between energy systems and political institutions involves the “oil curse.” This literature has been empirically rich but theoretically uncertain. Here, I attempt to identify what it can contribute toward anticipating the political–institutional impacts of future climate change and energy transitions. I do so by distinguishing three distinct versions of the theory: the resource curse, the petroleum curse, and the NOC (national oil company) curse.

The notion of a resource curse originally arose to explain various social, economic, and political outcomes in states that gain significant wealth from exploiting natural resources (e.g., Karl 1997). These outcomes include heightened levels of autocratic rule, corruption, and civil war. For the particular nexus of energy systems and political institutions, the most robust finding of over two decades of research is that “one type of resource wealth—petroleum—… tends to make authoritarian regimes more durable” (Ross 2015, 240). The curse, then, lies in the power of petroleum to restrict democratic possibilities in producing states. (“Oil” and “petroleum,” including natural gas, are often used interchangeably in the literature.)

If oil's curse on democracy is real, a tantalizing prospect looms in the future. Could a global shift away from fossil fuels effect a corresponding shift in favor of democratic regimes? Much of the academic literature on such relationships is either wholly atheoretical (e.g., Andersen and Aslaksen 2008, 2013; Aslaksen 2010) or easily criticized for insufficient attention to the impacts of international actors and dynamics on domestic actors and dynamics (see Rajan 2011; Hendrix 2018). Notable efforts have been made to support conditional or even null findings (e.g., Dunning 2008; Haber and Menaldo 2011) on the basic theory of the oil curse, largely through alternate empirical strategies. Here, we focus instead on the main lines of theoretical formulation and evolution.

Through what causal mechanisms could the use of oil and gas as energy sources hinder democracy? Theorists have offered several causal logics in answer to this question (e.g., Ross 2012, 4–6), all of which rely on some sort of fiscal mechanism. The revenues from oil production are what impact the shape that political institutions take. More specifically, the massive scale of petroleum sales in the global economy offers a fiscal windfall to incumbent elites in petro-states that may be used to secure (“buy off”) popular consent to oligarchic rule without recourse to more democratic forms of political representation. When Ross concludes that the scale and secrecy of oil revenues are what make oil-rich states less democratic (Ross 2012, 105), he is highlighting two aspects of a fiscal strategy for autocrats to preclude democratic opening.

The problem is that, if revenue is the mechanism, any other resource could in principle yield fiscal windfalls large enough to have anti-democratic consequences, under the right circumstances. Yet, according to Ross, oil and gas are the only resources that show the predicted anti-democratic impacts after empirical testing (Ross 2015, 253). This conclusion initially seems to betray an incoherence in the literature: a narrow petroleum curse cannot be explained by a broader, resource–curse logic. If oil and gas, why not also coal? For that matter, why not also non-energy resources such as ores and gems?

The resolution of this incoherence, though, comes from the theoretic implications of a widely embraced empirical result. Most scholars accept Ross's restriction of the chronological scope of the theory: the curse was only inaugurated at some point between 1970 and 1980, after many oil-producing states in the developing world expropriated the assets of American and European oil companies and vested them in NOCs (Ross 2012, 1–4, 63, 75). This version of the curse is therefore doubly restricted: first, it is not a resource curse but a “petroleum curse”; second, it is not strictly a petroleum curse but rather an “NOC curse.” Private firms based in Europe and the United States have gone from controlling 90 percent of the world's oil supply after the Second World War to less than 20 percent in recent years, with NOCs now producing the majority and 1970 widely recognized as the key point of inflection (Hughes 2016, 25–29). The effects of the oil curse, in other words, are less a function of oil than of the management of its production by a particular kind of agency. Renaming the oil curse as the NOC curse conveys an important insight because it reconciles the restricted empirical findings with the original fiscal logic of the theory.

However, even the restricted form of the NOC curse remains open to challenge. One important strand of the literature has stayed true to the original resource curse after reaching contrasting conclusions about the range of commodities with anti-democratic political impacts. These studies do so by supplementing the fiscal logic with a logic of concentration. They consider (1) the spatial concentration of a resource and (2) the structural concentration of its production as relevant sources of variation within the independent variable. For example:

Point resources are extracted from a narrow geographic or economic base and include oil, minerals, and plantations. The fact that these resources are spatially concentrated implies that they can be protected and controlled at a relatively modest cost. An abundance of these resources is typically associated with inequality in terms of power and the division of the surplus, and often is accompanied by vertical relationships between agents (shareholders, managers, laborers). Diffuse resources, on the other hand, are spread thinly in space, and harvested or utilized by agents characterized by horizontal relationships of rough equality. (Bulte, Damania, and Deacon 2005, 1031–32)

This analysis concludes that corruption and other weaknesses in the rule of law are significantly related to minerals and ores in general, including but not limited to petroleum (Bulte, Damania, and Deacon 2005, 1033–35). Still, these outcome variables do not directly bear on questions of electoral or liberal democracy versus autocracy.

Another notable analysis does focus on electoral (not liberal) democracy as the dependent variable. At the same time, it similarly operationalizes the independent variable in broad terms, as “income from oil, natural gas, coal, precious metal, and industrial metals,” with the result that a significant negative effect is found on a state's likelihood of transitioning from autocracy to electoral democracy (Wiens, Poast, and Clark 2014, 786). Yet another important study has found that a curse-like dynamic driven by oil and coal results in less competitive elections and higher levels of corruption in subnational governments within the United States (Goldberg, Wibbels, and Mvukiyehe 2008).

This type of approach represents a broadening of the petroleum curse, back to a genuine resource curse. These studies thereby achieve “concept-measure consistency” (Goertz 2006, 95–96) between the theorized mechanism (the fiscal logic) and the independent variable. They manage this by accounting for variation in the spatial and structural concentration of resources and of their production. However, two concerns remain. First, these studies may not be taking seriously enough the potential for variation in ownership and control (two parameters that are conceptually and empirically distinct; see Jones Luong and Weinthal 2006) of domestic mineral assets. Second, they are still working with theories that treat resources as marketable commodities: a reasonable but incomplete approach. The functioning of an energy source as an engine of revenue may be one facet of its political significance, but other characteristics pertaining to its functioning as energy per se will merit further consideration below.

A final limitation of all three versions of the oil curse is noteworthy: they offer theories about political impacts on producing states but not on consuming states. The distinction between net importers of petroleum and net exporters is crucial but often insufficiently operationalized in empirical studies (Colgan 2021, 3–4). Moreover, almost every country in the world relies heavily on the consumption of oil. A decarbonized energy regime would affect consuming states as much as, and possibly in different ways from, producing states. Even if a post-fossil energy regime released pro-democratic fiscal pressures inside producing states, these might in theory be overwhelmed in two respects. Not only might non-fiscal pressures in producing states turn out to be anti-democratic after the demise of fossil fuels, but either fiscal or non-fiscal pressures from the energy transition in consuming states might turn out to be anti-democratic.

What is clear from this brief review is that the petroleum curse per se has not survived two decades of empirical examination as a coherent theory. Only the (broader) resource curse and the (narrower) NOC curse remain standing as the viable alternatives. The implications for a post-fossil energy transition that might have been drawn from the oil curse—that reduced reliance on petroleum in the global energy system should mean increased chances for democracy—are therefore diminished. For example, if dramatic declines in the market value of petroleum reserves created a new type of “post-rentier state,” the loss of a habitual fiscal crutch could result in quasi-colonial status for such states vis-à-vis larger neighbors, as has arguably already happened to Nauru in relation to Australia since the exhaustion of the former's phosphate reserves (Kirkpatrick 2018).

We are left to draw lessons from the other two versions of the curse. The logic of the resource curse implies a long-run phenomenon, and one of the better empirical studies has confirmed it as far back as 1901 (Wiens, Poast, and Clark 2014, 788). If the resource curse is true, future studies of the effects of decarbonization on constitutional identity can learn nothing special from it about energy per se, but they should still be attentive to the regime-level implications of fiscal windfalls. It may be that producing states under the fossil-fuel regime would be forced to diversify economically under the next energy regime, and that economic diversification would have relatively pro-democratic impacts in its own right. This sort of process, though, would depend a great deal on variable national responses and capacities.

The logic of the NOC curse, for its part, entails even greater ambiguity and caution about the post-fossil future of democracy. Where NOCs currently dominate their home states’ energy production, not all of them will necessarily wither away after the transition to decarbonized fuels. Many NOCs have broad mandates over welfare policies, not just the supply or regulation of energy (Victor, Hults, and Thurber 2012). Existing stocks of managerial talent, technical infrastructure, investible assets, and political connections could enable some of them to survive the transition. Even where NOCs did go into decline, other state-owned enterprises specializing in different aspects of the energy system might take over a similar role, preserving the same fiscal logic based on revenues captured for the state. Instead of a renaissance of democratization after the demise of NOCs, a future generation of researchers might therefore be studying a new kind of fiscal curse based on different vehicles of state control over energy.

Carbon Democracy

The theory of “carbon democracy” addresses similar themes to the “curse” literature but challenges it in key respects. In its current state, this research tradition is theoretically provocative but empirically underexplored. Here, I assess the nature of that theoretic challenge and how successful it is, with an eye for how it might be used to investigate potential political–institutional impacts of future climate change and energy transitions.

Mitchell's Carbon Democracy (Mitchell 2011) is our primary source for the theory of the same name, supplemented by a small number of precursors and responses. Mitchell's book richly combines diplomatic, economic, and intellectual history to explain certain features of modern democracy and international relations. For energy-system impacts on political regimes, Mitchell valuably calls attention to two types of factor: the structure of energy-production processes that are distinctively associated with different energy sources and the domestic effects of complex international interactions of states and firms.

Particularly relevant are the first two chapters of Carbon Democracy, which analyze the discrepant political impacts of an energy regime based on coal and of one based on oil. On Mitchell's analysis, coal uniquely favored democratization. Industrial economies’ heavy reliance on this mineral in the late nineteenth century facilitated the development (and defense) of universal suffrage and the welfare state.

Between the 1880s and the interwar decades, workers in the industrialized countries of Europe and North America used their new powers over energy flows to acquire or extend the right to vote and, more importantly, the right to form labor unions, to create political organizations, and to take collective action including strikes. In most cases, these changes enabled mass-based parties to win power for the first time. (Mitchell 2011, 26)

The key mechanism linking coal and democracy was industrial sabotage. Organized groups of miners, railworkers, and dockworkers—small in numbers but strategically placed in networks for producing and distributing coal—possessed means of shutting down the urban economies of countries such as the United States, the United Kingdom, Germany, and France by cutting off their chief source of energy. Wealthy industrialists and their political defenders therefore felt compelled to make concessions to organized labor as the price of keeping coal-fired economies in continuous operation. The French word sabotage (meaning obstruction or slow-down) entered the English lexicon just prior to the First World War precisely to describe the tactics of the more militant labor unions around Europe when they used strikes to exploit their economic position for political leverage (Mitchell 2011, 21–26). This leverage is “what was missing [before the age of coal,] … not consciousness, not a repertoire of demands, but an effective way of forcing the powerful to listen to those demands” (Mitchell 2011, 21). The result was what Mitchell has elsewhere called “welfare democracy” (Mitchell 2009, 406).

The age of oil was different, on Mitchell's account. As oil was ascending in importance in the early decades of the twentieth century, this new source of fossilized energy could be produced with smaller numbers of workers and a higher proportion of skilled engineers and managers to the overall labor force. Moreover, because oil is liquid rather than solid, it can be distributed more quickly and with less labor than coal, through pipelines running from multiple points of production to major nodes of distribution (Mitchell 2011, 36–38).

Whereas the movement of coal tended to follow dendritic networks, with branches at each end but a single main channel, creating potential choke points at several junctures, oil flowed along networks … where there is more than one possible path and the flow of energy can switch to avoid blockages or overcome breakdowns. These changes in the way forms of fossil energy were extracted, transported, and used made energy networks less vulnerable to the political claims of those whose labor kept them running. (Mitchell 2011, 38–39)

Perhaps the most striking aspect of Mitchell's analysis of the political dimensions of the coal-to-oil transition comes from his examination of Winston Churchill's service as a cabinet minister during and after the First World War. Mitchell shows that Churchill deliberately moved the British navy away from coal-powered and toward oil-powered engines. Among the various possible motives for such a policy, Churchill's papers show a clear awareness (and fear) of the power of British labor unions in the mining and railway sectors to bring coal production to a halt, thereby endangering the navy's prime source of fuel (Mitchell 2011, 61–63).

The distinctive contribution of Mitchell's theory of carbon democracy is a logic of leverage. Energy regimes affect political regimes to the extent that their processes of production and distribution are susceptible to sabotage by ordinary citizens. A credible threat of sabotage can be used as leverage in political struggles, but the degree of credibility depends on the material infrastructure of energy. Without a credible threat over energy, ordinary citizens are vulnerable to technocracy and plutocracy: rule by managers, financiers, and their legal and political fixers.

Mitchell's logic of leverage helpfully extends earlier efforts to explain the impacts of energy sources on political institutions in terms of a logic of concentration. Two aspects of concentration have previously been highlighted as politically relevant features of fossil fuels, with Mitchell now having added a third.

First, the physiochemical concentration of stored solar energy in masses of coal, oil, and gas has often been considered a natural support for concentrated (autocratic) political power. Social and physical scientists alike often endorse the basic intuition that “the use of fossil fuels, which are concentrated energy, tends to concentrate both economic and political power in less area” (Hall and Klitgaard 2018, 212). Since all humans stand to benefit from the use of stored solar energy, the ability of a small number of humans in circumscribed locations to control massive quantities of energy should confer on them disproportionate social power. However, this relation should depend on a second kind of concentration as well: the spatial concentration of deposits of a given source of energy. With fossil fuels, as Perelman explained, “if you have a fairly small piece of ground which happens to have a big seam of coal or pool of oil underneath it, you can be very wealthy without having much land … Even very small holes in the ground may yield power (literally, energy available per unit of time) vastly greater and more reliable than that available to the holders of the historically largest estates, kingdoms, and empires.” For this reason, “the transition from a society based on non-renewable resources (area-independent) to one based on renewable resources (area-dependent) must result in major changes in social, political, and economic relationships” (Perelman 1980, 396).

As we have already seen, one strand of the literature on the resource curse takes spatial concentration seriously. Point resources are spatially concentrated and include minerals besides oil and gas; they also include non-energy resources such as gems and ores (Bulte, Damania, and Deacon 2005). Yet “spatial [centralization or] decentralization does not automatically imply structural [centralization or] decentralization” (Fotion and Elfstrom 1982, 240). The political question is about authority and control over the operations of an energy regime, however spatially centralized or decentralized.

Mitchell's logic of leverage can therefore be viewed as an elaboration of a third, structural type of concentration in energy systems. It suggests that different sources of energy vary in their ability to be used by different parties in political competition. What counts is not the concentration of the energy source as much as its social and technical embeddedness in larger processes of production, distribution, and use. Even between oil and natural gas, physiochemical and technical differences in storage and shipping have respective impacts on the oil and gas markets (Bashir 2017, 303), and Mitchell's approach directs us to consider such technical systems for their political impacts as well. It is thus an attempt to offer a more genuinely political account of the evolution of political regimes in response to energy regimes.

The theory of carbon democracy offers other improvements on the curse literature in particular. Whereas the resource curse describes the political impacts of commodities capable of yielding fiscal windfalls, carbon democracy is genuinely about energy sources per se. Because international dynamics between states and firms are central for Mitchell, his approach does not necessarily support the NOC curse. British and American geopolitical maneuvers from the early twentieth century onward (Mitchell 2011, 43–108), not the actions of oil-producing states around 1970, mark the key moments in Mitchell's story. In short, the curse of oil on democracy may be viewed as an effect of imperial elites in the oldest democracies, instead of (or in addition to) viewing it as an effect of kleptocratic elites in peripheral states.

Mitchell's analysis has theoretical and empirical weaknesses, nonetheless. Distinctions between coal (highly prone to sabotage) and oil (less prone) are carefully drawn, but alternative explanations for diminished democratic openings in the middle and late twentieth century cannot be ruled out. New technologies of automation and mechanization were used across a wide range of industries to curtail the leverage available to organized labor, not just in the oil industry (Morrison 2013, 1121). We need more fine-grained empirical analyses of the coal-to-oil transition that are organized in more state- or firm-specific case studies than Mitchell's book provides. Even for the earlier transition from wood, wind, and wave to a coal-powered society, the relationship between technical, social, and economic developments and political institutions is conducted at too high a level of generality to be empirically compelling. For example, the extension of formal voting rights to poor white men preceded the coal era in the United States, while subsequent extensions to black men and to women appear to have been unrelated to the political leverage of coal-miners or railway workers.

Also important are limitations in Mitchell's conceptualization of political regimes. His narrative tends to isolate universal suffrage and workers’ rights as the principal components of democratization, but democratic theory tends to favor greater institutional specification. Dahl broke down the two components of modern democracy (elections and liberties) into six more specific institutions (Dahl 1998, 83–99), which were in turn said to embody five core values. Even further political–institutional impacts could be distinguished in areas such as federalism and the division of executive, legislative, and judicial powers.

Crucially, though, the theory of carbon democracy teaches us that the relevant dynamics would be not only fiscal, as in the resource curse, but also more broadly structural. Mitchell perceives no political determinism in the physiochemical or spatial concentration of a fuel stock; what counts is the structural concentration of the sociotechnical systems organized around it. The logic of leverage forces us to ask how far the organized masses have the practical wherewithal to interrupt the production of wind turbines and solar cells, of the new generation of “modular” nuclear reactors, or of the transmission of electricity on the grid.

Conclusion: Post-Fossil Energy and Democratic Prospects

Academic researchers on climate change and environmental security have raised provocative questions about the geopolitical implications of a post-fossil energy transition. In contrast to traditional military and diplomatic approaches to American foreign policy that focus on state action to secure foreign supplies of oil for domestic consumption (Gholz and Press 2010), some believe that increasing reliance on renewable and low-carbon sources could fundamentally alter regional priorities and “energy for security” deals (Sivaram and Saha 2018). Moreover, transnational rather than national management of electrical grids may be called for (Reusswig, Komendantova, and Battaglini 2018, 252–53). The prospect of a post-fossil transition, in short, has some observers looking for energy security beyond the state.

The implications of decarbonization may also be profound for states themselves. The first of this paper's theoretic contributions to the study of the politics of climate change is to argue for the inclusion of change in the constitutional identities of states as an aspect of environmental security. Second, I have proposed a conceptual scheme of mediated causality for investigating climate change's potential impacts on political institutions, featuring six ecological factors that operate through six social and economic vectors (the WARMED framework). Third, with an in-depth critical analysis of theories of the oil curse and of carbon democracy, I have called into question conventional expectations about the possibility of a post-fossil democratic dividend.

This third issue now requires more granular analyses of the independent variable of energy-system change and the dependent variable of political-regime change.

Decentralized Energy and Democracy

Knowing that the oil curse implied a post-oil future that should be friendly to democracy, Mitchell offered an alternative theory of the energetic–political nexus. He isolated energy per se as a vector of causality, as opposed to the doubly mediated mechanism of the oil curse, running from energy through revenue before impacting political institutions. However, Mitchell was noncommittal in addressing the political possibilities associated with a future post-carbon transition (Mitchell 2011, 231–54). Even if we accept his argument that the age of coal was more friendly to democracy than the age of oil, the next thing after oil could take us in either a pro- or an anti-democratic direction.

One reaction to Mitchell's theory has held that decarbonized energy does in fact harbor democratizing tendencies. The underlying assumption behind recent theories of “energy democracy” is that, if both the generation and the transmission of energy were highly decentralized, political power would tend to devolve away from central states toward local communities (Fischer 2017, 228). After all, Mitchell's logic of leverage suggests that political power comes from power over energy flows. Turning his theory of carbon democracy into a companion theory of post-carbon democracy could therefore imply that masses of “prosumers” (simultaneously producers and consumers) of energy could wrest power away from the concentrated elites that manage, finance, and guard (judicially and militarily) the current fossil-energy regime (Szulecki 2018, 31–33).

This sort of argument, however, rests on the uncertain proposition that post-fossil resources—typically assumed to include wind, solar, nuclear, and hydroelectric energy—could function on a large scale with radically diffusive structures of both production and transmission. On the contrary, nuclear and hydropower plants have traditionally required massive investment in construction and maintenance projects under centralized management (Burke and Stephens 2018, 82–83). Wind turbines and solar cells can be located in a much more spatially dispersed fashion, but the electrical grids over which their product travels have tended to be highly integrated over long distances and therefore also in need of centralized management (Reusswig, Komendantova, and Battaglini 2018, 236).

A further weakness in causal arguments for post-fossil democracy is that, although wind and sunshine are relatively abundant and widely distributed resources, other ingredients needed for production and distribution of energy are not. The material inputs for wind turbines and solar cells include rare metals whose scarcity and price make them inaccessible to small-scale buyers on international commodity markets (Burke and Stephens 2018, 85–86). Additionally, the “rare earth” minerals used in small- and large-scale battery storage (for transportation and electricity, respectively) tend to be spatially concentrated in a few regions of the planet, raising the likelihood of nationalization under state-owned enterprises (Mahdavi 2020, 223–25). Even if future technology could enable the manufacture of effective batteries with more accessible materials, such as salt, the research, fabrication, and distribution of these new products would seem unlikely to be accomplished in highly decentralized fashion. The legal constraints of intellectual property, for example, would have to change before centralized control over technology could be ruled out. In short, even a hypothetical wind- and sun-driven energy system promises to remain reliant on “big government” or “big business,” or both.

Comparative empirical results also sit uneasily with the expectation that post-fossil energy fosters pro-democratic decentralization. In two middle-income democracies, Brazil and South Africa, transitioning to wind and solar generation in the electricity sector does not appear to be straightforwardly empowering for ordinary citizens (Hochstetler 2021, 132–74). Efforts to provide access to electricity for underserved citizens through distributed (i.e., decentralized or “off the grid”) generation, especially through solar cells, have confronted the reality that relatively small numbers of citizens and industries are able to provide themselves with the necessary infrastructure. Most people in Brazil and South Africa still rely on national grids, publicly financed and regulated. In such contexts, post-fossil technologies offer opportunities for the defection of those most able to pay the costs of maintenance of traditional centralized grids. While private grids flourish, the public grid deteriorates. Lower-class citizens are unlikely to find their political power enhanced by such a situation.

In short, current assumptions that energetic decentralization implies political democratization may not be taking seriously enough Mitchell's emphasis on structural over spatial concentration. If spatially dispersed wind- or sun-based generation systems currently remain dependent on centrally managed systems of distribution, Mitchell's logic of leverage would consider the latter fact more politically relevant than the former. Genuine structural decentralization, in which the many points of generation were the same as, or very close to, the many points of use, would have to overcome considerable technical, material, and infrastructural challenges to eliminate or to bypass centralized grids. How would millions or billions of ordinary citizens acquire their own rooftop solar cells or backyard wind turbines without resorting to established, oligopolistic networks of public and private bureaucracies for permission and provision? If we assume that the necessary materials and technologies for producing renewable energy are controlled by elites, the politically relevant kind of structural concentration stretches back to the very beginning of the sociotechnical process.

Even so, energy democracy's prospects might remain favorable, by means other than structural decentralization. After all, Mitchell's logic of leverage does not require broad popular ownership or control over materials and technologies. In order to democratize their countries in the nineteenth century, workers only needed access to points of potential sabotage within the coal-based energy system. If a wind-and-solar economy offered similar points of leverage, popular mobilizations would only need to interrupt the systems that elites build and maintain in their own interests. In this sense, it might be better for democracy if post-fossil energy regimes remain highly centralized—and therefore less resilient or secure.

Energetic Impacts on Elections and Liberties

If centralized control of the raw materials, production, and distribution of energy remains intact even after decarbonization, the mechanisms of the oil curse and carbon democracy remain relevant. However, the challenge of directing scholarly attention to more granular detail on the side of the dependent variable would then come to the fore.

Theories of the oil curse and of carbon democracy both assume that political elites are self-serving agents who prefer the work of excluding others from power to that of including them. Other theoretic traditions might challenge this assumption, but the oil curse and carbon democracy happen to share it. Where they differ is about when and how exclusion may be sustained. The logic of a revenue-based curse implies that some energy sources make it easier than others for oligarchs to monopolize fiscal resources for sustaining an exclusive regime, whereas the logic of leverage (in the theory of carbon democracy) explicitly contrasts different energy sources’ technical properties and processes for their vulnerability to sabotage by non-elites.

These theoretic logics tend to operate at high levels of generality. If we are interested in constitutional identity and institutional change, however, we must descend on the ladder of abstraction. Discerning the energetic impacts on democracies, autocracies, and hybrid regimes requires distinguishing the two institutions of competitive elections and civil liberties.

Elections and liberties are supposed to embody institutionally what the Dahlian conception of liberal democracy (or “polyarchy”) considers its two major components: contestation and inclusion (cf. “opposition” and “participation”; Dahl 1971, 2–9), respectively. Periodic elections embody consent and accountability, although this is often an imperfect partnership; liberties related to political communication and organization keep elections on a broad platform of meaningful participation. What do the oil curse and carbon democracy have to say about these disaggregated components of liberal democracy?

The revenue-based logic of the oil curse may suggest that the decline of petroleum should open up more space for elections than for liberties. If political elites have to resort to more diversified sources of revenue, they will need to broaden their basis of political consent. However, civil liberties are not as essential to the formalization of consent as they are to enabling unforeseen changes of policy and personnel. If autocratic or hybrid regimes wanted to preserve relatively closed political orders but had to do so with more diversified (post-oil) fiscal structures (Van de Graaf 2018, 113–14), they could try to incorporate new non-state interests (and sources of revenue) into minimally inclusive electoral processes. However, they would not necessarily have to make such processes fully liberal, in the Western sense. This kind of scenario might project a post-fossil future with fewer autocracies and more hybrid regimes, but not necessarily more democracies.

This would be a most-change scenario, of course. In a least-change scenario—in which autocratic elites manage to capture new energy sectors (or other revenue-producing facilities) through other state-owned enterprises—the pressure on autocracies to become hybrid regimes should not materialize, on the revenue-based logic.

The perspective of carbon democracy, in contrast, may suggest stronger pressures toward democracy, at least under certain assumptions. The interaction of the energy and revenue vectors is not operative in the logic of leverage, but a particular energy regime's own technical requirements and material dimensions are paramount. If organized labor can credibly threaten energy networks, Mitchell assumes that labor's demands tend to spread to ordinary citizens more generally. A kind of democratic solidarity operates throughout the polity, despite the fact that strategically placed energy workers may be small in number. On these assumptions, a post-fossil energy regime that reversed the diffusive and security-enhancing features of oil networks might sustain demands for not only elections but also liberties (such as rights to organize economically and politically). The goal of energy sabotage would then be a level of inclusion that hybrid regimes fail to offer. This kind of scenario might project a post-oil future with fewer autocracies, fewer hybrid regimes, and more democracies.

Like the logic of energy-based revenue, however, the logic of leverage holds the potential for authorizing more pessimistic forecasts for post-fossil democratization. The distinct implication of Mitchell's theory is that an energy regime that is more diffusely networked, with more planned redundancy, should be less friendly to democracy. The irony of carbon democracy is that more “power for the people” (i.e., universal access to secure energy) implies less “power to the people” (i.e., universal access to secure democratic rights).

Broadening the empirical referents of the theory beyond Mitchell's case studies of wood-to-coal and coal-to-oil transitions may help to resolve this paradox. Much depends on the precise “energy mix” of the decarbonized era, as well as the pacing of the transition from fossil to post-fossil energy. Researchers should use insights from past transitions while probing emerging characteristics of post-fossil energy in greater detail. They should also consider further descending on the ladder of institutional generality, from elections and liberties to more specific institutional features within the dependent variable, such as separation of powers (among executive, legislative, and judicial bodies) and geographic structure (i.e., federal versus unitary states).

References