Do Political Institutions Influence the Dismantling of Fossil Fuel Subsidies? Lessons from the OECD Countries and a Comparative Analysis of Canadian and German Production Subsidies

Abstract

Despite a global consensus that fossil fuel subsidies should be reformed, limited progress has been made. The study assesses whether domestic political institutions insulating politicians from backlash and compensating those affected by reforms make subsidies easier to dismantle. It was found that proportional representation and corporatism were correlated with lower levels of fossil fuel subsidies in OECD countries. A comparative case study of coal production subsidies in Germany and gas production subsidies in Canada suggests that political insulation and compensation contributed towards the dismantling of fossil fuel subsidies. The findings provide an understanding of the impact of corporatism and electoral systems on reform.

Keywords:

• Fossil fuels
• subsidy reform
• proportional representation
• corporatism
• policy dismantling
• comparative climate policy

Introduction

Fossil fuel subsidies are government measures that reduce the cost of consuming or producing fossil fuels, and are provided through a diverse array of support mechanisms (Wooders et al. 2019). Despite international commitments to their reform and increasing attention to their environmental and economic costs, fossil fuel subsidies in Organisation for Economic Co-operation and Development (OECD) countries have remained stable around US$100 billion each year in the period 2010–2020 (IISD and OECD 2022). Subsidies in OECD countries consist of direct transfers of public money to the production of fossil fuels (such as coal mines and gas fields) and tax rebates for investment in fossil fuel production and consumption. They constitute anti-climate policies (Compston and Bailey 2013) which need to be dismantled to achieve net-zero targets.

Studies on fossil fuel subsidy reform have often focused on the exogenous and economic factors that pressure lower income countries to dismantle fossil fuel subsidies, for example, international oil prices or fiscal deficits (see e.g. Rentschler and Bazilian 2017a, 2017b; Mahdavi et al. 2022). Such exogenous and economic factors are less likely to play a role among OECD countries than in lower income countries, as the former are less vulnerable to fluctuations in international fossil fuel prices and have more stable political institutions.

The differences between these countries in phasing out their respective subsidies point to the importance of domestic political institutions modulating pathways to reform (see online Appendices 1 and 2). While German hard coal production subsidies have been fully dismantled, Canadian fossil gas production subsidies have persisted (for Germany, see Furnaro 2022, for Canada, see Corkal et al. 2020). Comparative climate policy research suggests that domestic political institutions, such as proportional representation (PR) and corporatism, can contribute to lasting reform by insulating policymakers and compensating organised opposition to the reforms (Jacobs 2011; Finnegan 2022; Meckling et al. 2022). Countries with more proportional electoral systems tend to adopt more stringent climate policies than those with majoritarian systems as proportionality affords governing parties greater political insulation from electoral backlash when enacting unpopular reforms. Similarly, the literature highlights the influence of corporatism on climate policymaking. Corporatist arrangements institutionalise the access of carbon-intensive actors to policymaking (Martin and Swank 2012; Mildenberger 2020), facilitating the negotiation of compensation for the costs associated with decarbonisation policies (Wood et al. 2019).

The growing literature on the role of political institutions regarding fossil fuel subsidies have neglected electoral systems and corporatism in favour of institutional factors less relevant to OECD countries, such as authoritarianism and corruption (Cheon et al. 2013; Krane 2018; Kyle 2018; Fails 2022). We draw on the literature on the role of corporatism and electoral systems for climate policies (e.g. Böhmelt et al. 2016; Gronow et al. 2019; Mildenberger 2020; Finnegan 2022; Meckling et al. 2022). This raises the questions that this article seeks to answer, namely whether proportional representation and corporatism (1) are associated with lower levels of fossil fuel subsidies in the OECD countries, and (2) influence fossil fuel subsidy reform through the causal mechanisms of insulation and compensation.

To answer the first question, we study the statistical relationship of PR and corporatism to the levels of fossil fuel subsidies in OECD countries. We address the second question by exploring the mechanisms of insulation and compensation, associated with respectively PR and corporatism, vis-à-vis coal production subsidies in Germany and fossil gas production subsidies in Canada. Since the regression analysis is limited to data on subsidy levels and largely time-invariant institutional factors, the comparative case studies explore the roles of these two political institutions and the causal mechanisms through which they may influence reform.

Our paper contributes to the literature on fossil fuel subsidies (Rentschler and Bazilian 2017a, 2017b; Skovgaard and van Asselt 2018) theoretically by studying the effect of corporatism and electoral systems on reform, conceptually by studying fossil fuel subsidy reform as policy dismantling, and empirically by studying the under-researched fossil fuel subsidies in higher income countries. Furthermore, it contributes to the literature on comparative climate politics (Steinberg and Vandeveer 2012; Meckling et al. 2022) by studying the role of domestic political institutions on the dismantling of anti-climate policies.

The paper proceeds with our framework for studying the reform of fossil fuel subsidies. We then apply the framework first to the regression analysis and then to comparative case studies of Germany and Canada before offering a discussion of our results and concluding remarks.

A Framework for Studying the Dismantling of Fossil Fuel Subsidies

To study fossil fuel subsidy reform, we first disentangle the often unspecified concept of fossil fuel subsidy reform. Drawing on the policy dismantling literature, we examine policy instrument changes in terms of fossil fuel subsidy intensity and density (see inter alia Jordan et al. 2013; Gravey and Jordan 2016; Gürtler et al. 2019; Pollex and Lenschow 2020; Schaub et al. 2022). Policy instrument density refers to the number of policy instruments whereas policy instrument intensity refers to the size and scope of policy instruments. The size of a subsidy can range from thousands to billions of US dollars, whereas the scope can extend from individual corporate entities like state-owned oil companies to entire economic sectors like agriculture and transportation. Hence, we interpret a reduction in policy instrument density or intensity as constituting dismantling, whereas an increase is considered to be an expansion.

Our focus on PR and corporatism is motivated by existing literature on climate policy, particularly Meckling et al. (2022) and Finnegan (2022), emphasising the importance of domestic political institutions on climate reforms. Within the scope of these reforms and specific to our empirical focus, Mahdavi et al. (2022) argue that individual fossil fuel subsidy reforms are driven by idiosyncratic combinations of domestic and international factors, such as fiscal crises, low fuel prices, and fossil fuel reserves. Our research delves into the non-ideosyncratic contribution of PR and corporatism on subsidy reform and levels.

Against this backdrop, we posit that pathways towards dismantling fossil fuel subsidies are shaped by institutions that insulate policymakers from political backlash and compensate businesses and consumers for the costs of reform. The costs of subsidy reform must be distributed by politicians across economic actors in a way that is politically tenable while limiting their exposure to political risks. This is particularly relevant to the success of fossil fuel subsidy dismantling as the public has limited enthusiasm for fossil fuel subsidy reform and attempts have often provoked backlash (McCulloch et al. 2022). Increased costs of subsidy reform for voters (consumers) can trigger electoral backlash against governing parties, whereas increased costs for producers can trigger counter-mobilisation efforts that block policy change (Finnegan 2022).

Political institutions are capable of insulating policymakers from the political backlash of disgruntled voters and facilitating compensation measures to overcome organised opposition from incumbent fossil fuel interests (Lindvall 2017). The two sets of political institutions expected to influence these mechanisms are: (a) electoral systems for insulation, and (b) corporatism for compensation.

Proportional representation may insulate policymakers from backlash by decreasing electoral competitiveness and blurring politicians’ accountability towards voters. They offer incumbent governments increased electoral safety, making it easier for them to impose costs on consumers (Finnegan 2022). Specifically, PR decreases the marginal expected loss in a party’s seat share for a given decrease in the party’s national vote total. With coalition governments being the norm in PR systems, voters will have a harder time attributing unpopular policies to specific parties or politicians. Consequently, PR systems have been associated with imposing higher costs on consumers as part of higher climate policy ambition.

We therefore expect PR (and mixed) systems to have lower subsidy levels than majoritarian systems as policymakers enjoy greater political insulation. Accountability is obfuscated as voters are unsure which political party within a coalition government is responsible for dismantling, which restricts their ability to sanction them by withdrawing their vote (Plescia and Kritzinger 2022).¹ We submit hypothesis 1 (H1) and argue that the association between PR and fossil fuel subsidy levels may operate through the following mechanism: Insulation from political backlash in more proportional systems make fossil fuel subsidies easier to dismantle than in majoritarian systems.

(H1) Proportional representation is associated with lower levels of fossil fuel subsidies than majoritarian systems

Corporatism, understood as the degree of centralisation of collective bargaining within a country, has been shown to influence policymaking (e.g. Martin and Swank 2012) and institutionalise the policy access of carbon-intensive actors (Mildenberger 2020). Corporatist institutions lower the costs of negotiating long-term distributive agreements by establishing stable and credible bargaining between industry, trade unions, and the government. Institutionalised concertation between the state and carbon-intensive actors allows for negotiating compensation for losers of fossil fuel subsidy reform. Thus, phasing out fossil fuel subsidies might be facilitated in corporatist systems as compensation through long-term agreements (e.g. retraining or early retirement for workers) is easier to attain (see the German case study below). Previous research has shown that corporatist institutions may channel resistance to, for example, environmental policy change, away from public opposition and into closed policy forums, and thereby increase the possibility for ambitious policy change by decreasing audience costs (Finnegan 2022). In countries with low levels of corporatism, incumbent opposition groups are more likely to oppose climate policies, including fossil fuel subsidy reforms, in public fora such as the media or the streets (Andersen 2019; Mildenberger 2020; Driscoll 2021; Finnegan 2022). Public relations campaigns by organised opposition groups, persuading the public that fossil fuel subsidy reforms are too costly or ineffective, can increase the likelihood and severity of political backlash.

Relevant actors losing out from fossil fuel subsidy dismantling include individual companies and their industry associations, as well as labour unions representing employees within the affected sectors. In corporatist countries such actors enjoy privileged access to policymaking, allowing them to directly influence policy design and negotiate compensation for transition costs from decarbonisation policies (for example, decommissioning coal mines and power plants) (Wood et al. 2019). Gaining political support from industry through compensation can enable governments to increase costs for voters.

We put forward a second hypothesis (H2) about fossil fuel subsidy levels and argue that the association between corporatism and fossil fuel subsidy levels may operate through the following mechanism: Compensation agreed through corporatist structures makes fossil fuel subsidies easier to dismantle.

(H2) Corporatism is associated with lower levels of fossil fuel subsidies

Methods and Data

Using a mixed methods explanatory-sequential approach we first determine whether the two explanatory variables, the degree of corporatism and the type of electoral system in 34 OECD countries between 2010 and 2020 (see online Appendices 3 and 4), have a statistically significant effect on the level of fossil fuel subsidies. Data limitations, inconsistent government classification, and varying degrees of disclosure limit us to examining subsidy levels within this timeframe as a proxy for density and intensity. Subsequently, we undertake a comparative case study of production subsidies in Germany and Canada to explore the role of insulation and compensation mechanisms without necessarily providing a full account of all causal factors explaining the differences in subsidy reform between the two countries.

We utilise data from the online database Fossil Fuel Subsidy Tracker (IISD and OECD 2022) to capture our response variable – the total subsidy level expressed in nominal millions of US dollars and normalised on a per capita basis. In addition to total subsidy levels, we report results for production subsidy levels to reflect the focus of our case studies. A decrease in country subsidy levels indicates whether we can expect to find policy dismantling in terms of policy instrument density, intensity, or both. In some instances, there may be an increase in density, but an accompanying decrease in intensity, or vice versa. To assess whether there has been an overall shift towards dismantling or expansion, we treat price movements in country levels of fossil fuel subsidies to infer net changes.

Our statistical analysis focuses on two main explanatory variables. The first is the electoral system used in the primary national legislative chamber, categorised as majoritarian, mixed, or proportional.² The second is a continuous variable measuring the level of centralisation in collective bargaining across OECD countries. In addition to the two main explanatory variables, we also include several control variables in our study (see online Appendix 4 for details):

1. Government effectiveness: This accounts for a government’s capacity to enact effective subsidy reforms. This captures country-level perceptions on the quality of public services, civil service, and policy formulation and implementation.

2. GDP and population: We control for a country’s overall wealth and population size, as wealthier and more populous economies tend to either supply or demand more subsidies. We transformed these variables by taking their natural logarithm for a better approximation of a normal distribution and to improve interpretability.

3. Oil price: To address external variations in oil prices, we control for the international market price of a barrel of oil. This helps account for internationally driven fluctuations in subsidy levels.

4. Interest rates: Short-term (three-month) government bond interest rates serve as an indicator of the fiscal pressure governments might face to reform fossil fuel subsidies.

5. Energy intensity: Countries with more energy-intensive economies tend to consume more fossil fuels. It follows that actors in these countries would be more sensitive to price movements with greater associated risks of political backlash.

6. Energy mix: Countries that are less reliant on fossil fuels are generally less likely to have interest groups opposing fossil fuel subsidy reform.

7. Fossil fuel production: The per capita production of coal, oil, and fossil gas is an indicator of the relative power and influence of the fossil fuel industry over policy design.

We included one-year lags for (5), (6), and (7) to mitigate endogeneity issues given the possibility of reverse causality. Lagging these control variables ensures that they do not contemporaneously influence the dependent variable, reducing the risk of bias.

The coal production subsidies in Germany and the gas production subsidies in Canada constituted two of the largest production subsidies among the OECD countries. These cases differ on the response variable (dismantling of German subsidies but not of Canadian subsidies) and explanatory variables (Germany having a mixed-member proportional system with high levels of corporatism and Canada a majoritarian system with low levels of corporatism). The subsidies were similar in that coal and fossil gas production constituted politically important sectors respectively in Germany and Canada, with large and politically influential extraction companies and highly concentrated in particular regions. We argue that these similarities between the political roles of the sectors and the subsidies outweigh the differences between coal and gas production, such as coal being a more polluting fuel and German coal production facing more competition from low-cost producers than Canadian gas production.

The case study analyses were conducted using secondary materials such as academic texts, international and non-governmental organisation (NGO) reports, and newspaper articles. These materials have been studied to discern the roles of insulation and compensation in policy processes concerning fossil fuel subsidy reform, as well as the relationship between these mechanisms and our explanatory variables. In the case of Germany, this concerns particularly the processes leading to the 2007 decision to phase out hard coal production subsidies. Conversely, in the case of Canada, where no comparable reform occurred, we direct our focus to the period from 2010 to 2023, following the G20 commitment in 2009 to phase out fossil fuel subsidies.

Statistical Analysis

We ran separate models for our two explanatory variables to avoid multicollinearity as countries with PR often have corporatist institutions which would confound our results (see Figure 1). For corporatism, we used two-way fixed effects estimators as we were interested in within-country variation and can address all unobserved country- and time-specific heterogeneity (Models 1 and 2). For electoral systems, as we were primarily interested in variation between countries, we first used pooled OLS estimators (Models 3 and 4) and then ran within-between “hybrid” models (Models 5 and 6) following the approach set out in Bell and Jones (2015) and Schunck (2013). For Models 1–4 we utilised Driscoll–Kraay standard errors as they are heteroscedastic and autocorrelation consistent, and are robust to general forms of cross-sectional and temporal dependence (Hoechle 2007). Pesaran’s (2021) test of cross-sectional independence and Wooldridge’s (2002) test of serial correlation both verified the presence of cross-sectional dependence and serial correlation in the data.

Figure 1. Regression results for both corporatism and electoral system explanatory variables with total and producer subsidies (per capita)

Figure 1. (Continued).

For electoral systems, we found that countries with PR appear to have on average US$24.15 per capita lower total fossil fuel subsidies and US$23.32 per capita lower production subsidies than countries with majoritarian systems. Mixed systems were also correlated with lower subsidies, with US$62.09 and US$20.56 per capita less in respective total and production subsidies (Models 3 and 4). This is considerable given that the mean value for total subsidies per capita is US$120 across the 34 OECD countries. In Model 5, proportional and mixed systems were correlated with lower levels of total fossil fuel subsidies, but these results were not statistically significant. However, in Model 6, proportional systems were negatively correlated with production subsidies (−US$31.55), as were mixed systems (−US$30.02). These results were in line with H1, except for the non-significant results in Model 5.

For corporatism, we found that a unit increase in the explanatory variable was negatively correlated with total subsidies (Model 1), but slightly positively correlated with production subsidies (Model 2). In line with H2, the observed correlation showed a marginal decrease in total subsidies of approximately US$17 per capita. However, the marginal decrease in total subsidies does not appear to be driven by production subsidies, which saw a slight increase of US$2.182 per capita (Model 2), suggesting that corporatist countries are more likely to impose costs on consumers than producers. These results were statistically significant at the 0.10 level and in line with H2.

To make better sense of these results, we now turn to our comparative case studies. We examine how political insulation and corporatist negotiated compensation led to subsidy dismantling in Germany and limited policy change in Canada.

Comparative Case Studies

Germany: Dismantling Coal Production Subsidies

Coal production subsidies in Germany amounted to about €2–3 billion per year in the years before 2007, making Germany the largest provider of production subsidies among the OECD countries. Despite receiving subsidies, coal production and employment in Germany had declined since the 1950s, when it exceeded 700,000 in 1958, reaching only a few thousand by 2020 (Brauers et al. 2020). In 2007, the German parliament voted to end such subsidies by 2018. Given the high costs of hard coal production in Germany, this de facto ended German hard coal production (Oei et al. 2020). From 2007 to 2018, annual direct transfers to hard coal production amounted to €1–1.8 billion (G20 2017). State transition aid for coal-mining regions and workers will continue until 2027 (Gencsü 2019). Direct transfers to lignite production, the other kind of coal in Germany and concentrated in eastern Germany, had ended in the 1990s, although some policy instruments persist, such as tax exemptions for lignite production (especially at the Länder level) and direct subsidies for lignite-based electricity (G20 2017; Gençsü 2019). Yet after the reduction in policy instrument density the producer subsidies are at the much lower level of millions of euros (OECD 2022).

The German mixed-member proportional system has historically led to coalition governments, and in 2005 it led to the formation of a grand coalition government of the centre-right Christian Democratic Union (CDU), its Bavarian sister party the Christian Social Union (CSU), and the Social Democratic Party (SPD). The previous SPD–Green government had not adopted a phase-out due to the strong ties between parts of the SPD (especially in its historical stronghold of North Rhine-Westphalia) to the coal industry and associated labour unions (Leipprand and Flachsland 2018; Furnaro 2022). Yet in 2007 the grand coalition could after a year of intense negotiations among the coalition parties – the Länder of North Rhine-Westphalia and Saarland, the German Hard-Coal Corporation (owning most hard coal mines), and the labour unions for the mining, chemical, and energy sectors – adopt the agreement to phase out the subsidies by 2018. The agreement contained a review in 2012 of whether subsidies could continue beyond 2018 for some mines. The law was amended in 2011 under the then CDU–FDP coalition government to delete the review clause in light of recent EU legislation phasing out coal-mining subsidies (Federal Government of Germany 2011). The 2018 end date was a compromise between the CDU, which wanted a quicker phase-out, and the SPD (especially the parts of the party with ties to North Rhine-Westphalia), which together with the labour unions wanted to retain subsidies for a few mines (Spiegel 2007).

Thus, on the federal level, the SPD was more insulated from voter backlash than during the previous government since responsibility was shared with the CDU, which could not capitalise on such backlash. The parties outside of government also had limited possibilities for utilising voter frustration, as they were either pro-reform (Greens, the FDP) or abstained (the Left). In North Rhine-Westphalia, led by a CDU–FDP coalition, the CDU and the SPD adopted more opposing stances.

Phasing out hard coal subsidies had been on the political agenda for decades before the 2007 agreement, pointing to the role of the grand coalition in dismantling the subsidies. Yet the insulation offered by grand coalitions does not allow for fossil fuel subsidy reform on its own, as is evident in the process leading to the 2019 decision to phase out coal. Here, the governing CDU/CSU–SPD coalition set up a national Coal Commission with representatives from industry, labour unions, lignite-producing Länder and regions, experts, and environmental NGOs. This strategy, which also involved compensation, aimed to insulate the governing parties from losing voters to the far-right Alternative für Deutschland in the lignite-producing eastern Länder (Furnaro 2022, pp. 9–10).

Corporatism played an important role for the dismantling of hard coal production subsidies, including by allowing for the compensation that played an explicit role in the reform. Not only have corporatist structures been very strong within West German hard coal production, but the coal sector has been a core component of German corporatism since World War II (Furnaro 2022, p. 5). The corporatist structures involved coal and steel trade associations, labour unions, and the federal and Länder governments in North Rhine-Westphalia and Saarland. They constitute an incumbent regime that has tried to fend off challenges from globalisation (especially cheap coal, gas, and oil imports), EU policies promoting free markets, and calls for environmental regulation (Rentier et al. 2019; Oei et al. 2020). Subsidies have been an integral part of the corporatist structures since 1957, and have limited major layoffs (Storchmann 2005).

Given the incumbency of the German hard coal sector, it was not a foregone conclusion that dismantling would have been agreed in 2007. It is also unlikely that such an agreement would have been acceptable to coal sector actors without the possibility to negotiate vast amounts (billions of euros) in compensation to affected workers and regions. This compensation ensured that workers older than 42 years would not become unemployed and younger workers had ample opportunities for re-training (Federal Government of Germany 2007; Brauers et al. 2020). The coal sector, especially the labour unions, and the SPD also ensured that the end date would be delayed from 2012 to 2018 (Brauers et al. 2020). While the most important labour union, IG BCE (representing mining, chemical, and energy sector workers), would have preferred the subsidies to continue, they and the SPD ensured that no workers would be made redundant by the agreement, which was a key reason for accepting it (BCE 2006, 2007).

Direct subsidies to lignite mining in eastern Germany, where there were fewer corporatist structures, ended more abruptly in the early 1990s, and led to the firing of 100,000 workers or two-thirds of the workforce (Herpich et al. 2018). Yet the 2019 agreement to phase out coal in Germany, and thus also the remaining lignite production subsidies, involved corporatist structures in the stakeholder-driven Coal Commission and ensured compensation of up to €2 billion per year to the affected regions, workers, and companies (Reitzenstein and Popp 2019; Furnaro 2022).

Canada: Dismantling Gas Production Subsidies by Symbolic Action

Fossil gas production in Canada has been on the rise since 2010, accounting for 38 per cent of total primary energy supply in 2019 (Canada. Natural Resources Canada 2021). Much of the expansion in gas production was made possible through increased government support of the fossil gas sector via federal and provincial subsidies. Fossil fuel subsidies from the federal government in 2020 shifted from exploration to the development of fossil fuel production and export infrastructure (primarily to the US), including for liquified natural gas (Corkal et al. 2020). Canada’s subsidy levels for fossil gas production in 2020 are almost at the same level as they were in 2010, with a significant increase between 2019 and 2020 from US$494 million to $1,226 million (IISD and OECD 2022).

In 2015, Prime Minister Trudeau mandated the Department of Environment and Climate Change and the Department of Finance to work together to “phase out and rationalize over the medium-term inefficient fossil fuel subsidies while providing targeted support for the poorest”, a promise that has been reaffirmed regularly since the 2009 G20 commitment (G20 Heads of State and Government 2009). Although the Liberal government has maintained its position of dismantling fossil fuel subsidies, its track record does not align with the political declarations. Increasing criticism of the Liberal government’s slow fulfilment of the reform declarations has come from the National Democratic Party, Green Party, and Bloc Quebecois between 2010 and 2020. The Conservatives, in government from 2006 to 2015, and the official opposition party since 2015, have been keen to expand fossil fuel infrastructure and support the oil and gas industries.

Beyond party politics, the Office of the Auditor General of Canada issued a report in 2017 on fossil fuel subsidies stating that “without a clear understanding of the fossil fuel subsidies covered by the G20 commitment and without an implementation plan with timelines, the departments cannot ensure that they are providing the support needed for Canada to meet the commitments by 2025”. In 2022, the Liberal government pointed to its reform of nine tax measures that provided preferential tax treatment to the fossil fuel sector, with five directly resulting in the reduction of policy instrument density (HC Deb 17 May 2022). However, with fossil gas production subsidies at the same level or higher in 2020 than they were in 2010, the intensity of individual policy instruments are likely to have expanded and thus offset the reduction in instrument density.

Canada has been behind schedule in its joint G20 peer review process with Argentina to identify “inefficient” fossil fuel subsidies within their respective countries. Where other peer reviews took 12–18 months, Canada and Argentina were well into their fifth year (Corkal and Gass 2020). In 2023, 14 years since Canada first committed to phasing out inefficient fossil fuel subsidies, the country published a framework for identifying which subsidy instruments will be included in the reform efforts to dismantle the subsidies in 2024. Though the federal government has developed a comprehensive definition of inefficiency, it contains exemptions for fossil fuel production projects that include emission reduction measures, such as carbon capture and storage, as well as the public financing of fossil fuel projects through publicly owned companies, such as loan guarantees for pipeline expansion (Williams 2023).

Drawing on the policy dismantling literature, we argue that the Liberal government has been pursuing what Bauer and Knill (2012) term a strategy of dismantling by symbolic action whereby one can expect announcements about the dismantling of policies with limited follow-through. The lack of follow-through may stem from institutional constraints or the government responding to pressure from certain groups to dismantle subsidies, even if they are not fully convinced that such dismantling is politically advantageous (Bauer and Knill 2012). Pairing limited dismantling with highly visible political declarations has been the strategy pursued by successive loss-sensitive Liberal governments in a precarious electoral position. Under a majoritarian system, which disincentivises unpopular reforms, the Liberals are not insulated from electoral competition. In 2019, the Liberals lost their parliamentary majority, losing their only four seats in Alberta to the Conservatives, who also took the plurality of seats in British Colombia (the two provinces account for 98 per cent of fossil gas production in Canada). In the hopes of reclaiming his majority, Trudeau called for an early election in 2021 and the Liberals were able to reclaim their plurality in British Colombia and two seats in Alberta, but again only managed to form a minority government. Given how competitive elections are as a result of majoritarian rules, politicians will have been concerned about voter backlash and less likely to increase consumer prices for fossil fuels which would also have affected major industries (Finnegan 2023).

The Canadian economy is to a large part made up of several carbon-intensive industries such as steel production, intensive agriculture, and automotives. A Canadian transition towards decarbonisation would take place without the presence of corporatist structures to negotiate compensation for subsidy dismantling. Though not directly part of subsidy reform, the phasing out of coal production has been coupled with a “just transition” initiative which has included labour movements and affected communities. Some coal miners were offered early retirement, while others transitioned to mine reclamation or other sectors, including fossil fuel ones. Contrary to our expectations, it shows that the Canadian government is capable of providing compensation, but the decision to dismantle might hinge on the smaller size and political influence of the coal sector when compared to oil and gas.

Pluralist, rather than corporatist, state–business relations in the Canadian case have resulted in considerable alignment between the fossil fuel industry and policy decisions made by provincial governments. Interest groups such as the oil and gas lobby have since the 1990s embedded themselves in key policy networks that hinder climate action, particularly in the fossil-fuel-rich and conservative stronghold of Alberta (Lecours and Béland 2022; Viens 2022). Pluralist systems can face challenges in adequately representing the interests of less powerful or marginalised groups (Mildenberger 2020). This can lead to policy outcomes that disproportionately favour the interests of powerful fossil fuel interest groups at the expense of broader societal interests. On balance, pluralist institutions have been viewed as complicating climate reforms (Karapin 2016). Limited insulation for policymakers and the absence of compensation through centralised collective bargaining – increasing the transaction costs of negotiating a longstanding agreement – along with heavily embedded carbon-intensive interests at the provincial level generate few incentives and substantial costs to dismantle these subsidies.

Discussion

The regression analysis demonstrated a negative significant association between proportionality and total and production fossil fuel subsidies per capita. Thus H1 was fully supported. The case studies show that the German grand coalition was insulated from electoral backlash stemming from production fossil fuel subsidy reform in a way the Canadian government was not. This insulation allowed for a fuller policy dismantling in terms of reduced policy intensity and density, unlike in Canada, where gas production subsidies were only subject to dismantling by symbolic action. Yet, regarding insulation, what might matter more than just the electoral system per se is the government type. Government coalitions come in all shapes, with grand coalitions being more of an exception than the norm among PR and mixed systems, so it may be difficult to generalise from the German case to other coalition governments. At the same time, proportionality may matter for fossil fuel subsidy reform in other ways than insulation, such as by allowing for more Green party seats in parliament, although the German Green party was not an important actor in the German reform.

The relationship between corporatism and fossil fuel subsidies is also rather clear. The regression analysis showed that corporatism was associated with lower levels of fossil fuel subsidies per capita, confirming H2. Yet corporatism was positively correlated with production subsidies, indicating that governments in corporatist countries may prefer to shift costs away from industry producers and towards consumers. The case studies showed that the mechanism of compensation agreed through corporatist structures made it easier to reach an agreement among business organisations, labour unions, and federal and regional governments in Germany. In Canada, there are no such structures that could allow for similar negotiations of compensation, although compensation has been utilised in the coal production phase-out. The use of compensation in the Canadian coal production phase-out highlights that compensation may be dependent on a different set of factors, not all of them including corporatism.

We argue that insulation and compensation by way of electoral proportionality and corporatism constitute permissive conditions within a constellation of factors shaping reform. Insulation and compensation are not on their own sufficient conditions for reform of fossil fuel subsidies, since such reform requires actors that promote them. It is worth noting that pro-reform NGOs were active in both Canada and Germany and do not explain the differences between the two countries. In Canada, environmental NGOs including the International Institute for Sustainable Development have put fossil fuel subsidies on the agenda, whereas in Germany economic rather than environmental concerns drove the opposition to subsidies, and the Green party and NGOs were more concerned about nuclear power than coal production subsidies (Renn and Marshall 2016). A more prominent alternative factor explaining the difference between the two cases may be the EU; it had continuously pushed the German government to phase out coal-mining subsidies, and was important for the decision to abandon the 2012 review of the phase-out.

We contend that countries with higher levels of insulation and compensation can act as important first movers to break the fossil fuel subsidy lock-in (Skovgaard and van Asselt 2018), particularly when those savings are reinvested into driving down the costs of renewable energy alternatives (Meckling et al. 2022). This in turn can allow for late adopters to introduce reforms more rapidly due to lowered barriers of entry and facilitate a global trajectory towards phasing out fossil fuel subsidies and disrupting the wider carbon lock-in (Bernstein and Hoffmann 2019).

Conclusions

Our analyses showed that domestic political institutions, namely PR and corporatism, are important to fossil fuel subsidy levels and reforms. Our comparative case study showed that two mechanisms, insulation due to electoral proportionality and compensation agreed within corporatist structures, provide pathways to reducing the density and intensity of fossil fuel subsidies. We contribute to the literature on fossil fuel subsidies, policy dismantling, and comparative climate politics by identifying how PR and corporatism are relevant for the dismantling of fossil fuel subsidies, and not just explicit climate policies.

These findings have implications for future research, which should pay more attention to domestic political institutions and their role in influencing the viability of pathways to sustainable reforms. This may involve investigating how the influence of exogenous and international factors, such as oil price levels or commitments to fossil fuel subsidy reform, is moderated by these institutions.

Disclosure Statement

No potential conflict of interest was reported by the authors.

Supplementary Material

Supplementary data for this article can be accessed at https://doi.org/10.1080/13876988.2024.2328605.

Additional information

Funding

This work was supported by the Swedish Energy Agency under grant no. 48761-1, “Pathways to Breaking the Fossil Fuel Lock-In: Assessing the Influence of the Paris Agreement and the Sustainable Development Goals on Fossil Fuel Subsidy Reform”.

Notes

1. It is important to note that the relationship between government accountability and higher subsidies is not limited to OECD countries. Matthew D. Fails (2022) found that consumption subsidies tend to rise when accountability increases when a country transitions from being a closed autocracy to an electoral autocracy.

2. V-Dem, our source for electoral systems, classifies Germany as a “mixed” system even though it is mixed-member proportional system, but “mixed-majoritarian” systems have been classified as majoritarian. Hence, the mixed category is closer to the PR category in terms of proportionality than to the majoritarian category.