Abstract

Transnational companies (TNCs) are becoming increasingly influential in the global governance of climate change. Therefore, it is of paramount importance to understand the factors that explain why some TNCs broadly support policies to tackle climate change, while others oppose them. This study subjects previous findings from small-N case studies to a more systematic fuzzy set qualitative comparative analysis (fsQCA). It investigates previous findings that link exposure to fossil fuels to policy opposition, and transnational operations, exposure to consumers, certain factors in the institutional environment, and pressure from investors to policy support. The study concludes that findings from small-N case literature can explain the necessary conditions for climate policy support in a larger set of TNCs from a wider variety of sectors and geographies beyond GHG-intensive sectors, such as retail, technology, and telecommunication. It concludes by suggesting areas and cases for further research.

During the most recent globalization wave, commonly believed to have started in the second half of the twentieth century, the number of transnational companies (TNCs) has boomed (Zürn 2013). Globalization is characterized by a rise in both multinational markets and TNCs (Strange 1995). This article follows Arts’ (2003, 6) definition of the TNC as a “large-scale, profit-making, commercial organisation with offices and/or production units in many countries around the world.”1 TNCs now account for most of global economic activity. They are increasingly involved in global environmental governance, influencing (international) policy making, researching and developing alternative technologies, shaping public discourse, and even starting private governance initiatives where state-led regulation lacks (Elliott 2004; Falkner 2008; Levy and Newell 2005; Van der Ven 2014; Vormedal 2008).

Globalization has deepened TNCs’ influence over (inter)national policy. It has increased TNCs’ structural power as the central nodes in capitalist economies, being the main drivers of economic growth, investment, and employment upon which governments depend (Fuchs 2013; Newell 2000; Newell and Paterson 1998; Vormedal 2008). For these reasons, TNCs are “key agents” (Falkner 2008, 4). On climate change policy specifically, some TNCs have developed sophisticated strategies to safeguard their interests. Single companies like ExxonMobil have had far-reaching policy influence in key states like the United States (Levy and Newell 2005). This suggests that corporate climate policy strategies should receive more attention (Levy and Newell 2005).

This article examines the drivers behind TNCs’ climate policy strategies through a qualitative comparative analysis (QCA) of twenty-five of the world’s largest TNCs from major sectors and world regions. Its research question is as follows: what factors explain why some TNCs support climate policy, while others do not?

This article first reviews the literature on corporate climate policy positions, identifying six conditions to be tested: Exposure to Fossil Fuels, Low-Carbon Transition, Transnational Nature, Exposure to Consumers, Institutional Environment, and Investor Pressure. Second, it introduces and substantiates the QCA methodology by arguing that the research question is best understood in terms of necessity and sufficiency, and the explanation for TNCs’ climate policy positions is likely to be a causally complex combination of conditions. Third, it describes and discusses the results, arguing that not being exposed to fossil fuels in combination with being in a low-carbon transition is a necessary condition for climate policy support, while the inverse is true for the absence of policy support. The article concludes by identifying avenues for further research.

Literature Review

TNCs seek to influence their regulatory environment because they have to consider how future policies and regulations might influence them on the medium to long term (Vormedal 2010, 2011). International treaties signal the international community’s long-term political objectives by, for example, incentivizing low-carbon development or encouraging and supporting private actors to undertake voluntary efforts (Falkner 2016). Specific “green” policies such as feed-in-tariffs or renewable portfolio standards often precede larger, more general policies like carbon pricing schemes, suggesting that green policy “nurtures a political landscape of interests and coalitions that benefit from a transformation to low-carbon energy use” and coalesce to develop more stringent climate policies (Meckling et al. 2015, 1170). As major users of raw materials and emitters of greenhouse gas (GHG) emissions, TNCs have a major stake in climate policy development (Elliott 2004).

TNCs have long been recognized as important actors with influence over the development of (inter)national environmental policy. The political influence of TNCs is broadly argued to have been increasing. On one side, TNCs have increasing sway over national regulation and state positions in international negotiations (Chasek et al. 2016; Clapp and Dauvergne 2005; Newell 2005; Strange 1992) through expanded traditional political power due to their advantage over other actors in terms of resources, access to decision makers, and technical expertise (Chasek et al. 2016; Kolk and Levy 2001). As Strange (1995) notes, the “downward diffusion” of authority means that states increasingly share authority in the economy and society with nonstate entities, including TNCs. On another side, TNCs are at the end of increasingly complex and global supply chains, with higher sway over suppliers’ environmental practices than governments through requiring compliance with private standards in contracts (Dauvergne and Lister 2010; Fuchs and Kalfagianni 2010; Van der Ven 2014). Increased business involvement in global environmental politics emanates from the maturing relationship between states, business, and civil society and growing recognition that wide-ranging support is needed to solve environmental problems (Falkner 2008). For these reasons, TNCs should be at the heart of global environmental politics research (Andrade and Puppim de Oliveira 2014; Dauvergne and Lister 2010; Downie 2014; Falkner 2008; Levy and Newell 2005; Vormedal 2010).

Global environmental politics has started to investigate the growing political power of corporations—to set agendas; create and enforce rules; and shape norms, regimes, and institutions (Dauvergne and Lister 2010; Downie 2017). In the early 2000s, medium-N comparative studies of TNCs’ role in environmental policy were rare (Skjærseth and Skodvin 2001; Vormedal 2008), especially regarding how corporate political strategies are developed (Levy and Newell 2002, 2005). Studies mostly used qualitative case study methods to compare two or a handful of companies from specific sectors (Betsill 2014), mainly heavy polluters, such as oil (Kolk and Levy 2001; Levy 2005; McAteer and Pulver 2009; Newell 2000; Skjærseth and Skodvin 2001), automobiles (Levy 2005), and chemicals (Falkner 2005; Orsini 2012), or focused on limited geographies (Kamieniecki 2006; McAteer and Pulver 2009). Over the last ten years, quantitative studies examining larger samples have started to appear but are still limited by focusing on certain polluting sectors, such as oil (Van Halderen et al. 2016), coal, and utilities (Downie 2017), or on a specific set of explanatory factors, such as stakeholder pressures (Sprengel and Busch 2011). Outside environmental politics, studies have analyzed the drivers of corporate decisions on mitigating GHG emissions (Abreu et al. 2017; Cadez et al. 2019) and the drivers of general political engagement by corporates (Lux et al. 2011). Multiple authors suggest testing their results across other industries as an important avenue for future research (Abreu et al. 2017; Downie 2017). This article contributes to the literature by analyzing multiple “conditions” that could determine TNCs’ climate policy strategies on a medium-N sample of TNCs across multiple major industries, including and extending beyond heavy polluters. It proceeds by analyzing the literature cited herein to identify the conditions that are argued to determine TNCs’ climate policy positions. For each condition, the directional expectation (DE) is provided.2

Conditions That Determine TNCs’ Climate Policy Strategies

Climate policy, and environmental regulation more broadly, is distributional in nature. Its effects differ across industries and companies, leading to an uneven distribution of benefits and the creation of winners and losers. Critically, corporate preferences on climate policy are shaped according to whether they expect to be a winner or loser (Downie 2017; Falkner 2008; Gasbarro et al. 2017; Meckling 2015; Sprengel and Busch 2011). This general theoretical expectation will underpin the formulation of the conditions that are analyzed.

Exposure to Fossil Fuels (EF)

Industries that either supply or use fossil fuel–based energy on a large scale, such as the coal, oil and gas, heavy industry, automobile, and chemical sectors, could be fundamentally affected by climate change policies (Newell 2000; Skjærseth and Skodvin 2001). These industries have broadly been the most heavily opposed to climate policy (Falkner 2008). On the other hand, companies that market alternative processes or products for environmentally unfriendly activities generally support greater environmental protections (Chasek et al. 2016; Falkner 2008; Gasbarro et al. 2017; Vormedal 2010).

Two factors determine the degree of Exposure to Fossil Fuels.3 First, a larger share of revenue from the production and sale of fossil fuels is expected to negatively affect a firm’s climate policy support. Rio Tinto was the only top-ten US coal producer that did not oppose the Waxman–Markey bill, which attempted to implement an emissions trading system in the United States, in part because coal generated only 8 percent of Rio Tinto’s global revenue (Downie 2017). Furthermore, whether fossil fuels are a core product or a commodity used in manufacturing or product use matters. The oil industry, for example, has been less willing to accept mandatory GHG emission controls than the automobile industry, both in the United States and Europe (Levy 2005). Second, the carbon intensity of TNCs’ activities can be expected to dictate climate policy supportiveness, as higher carbon intensity has been associated with lower supportiveness (Skjærseth and Skodvin 2001). Variation in utility support for the Waxman–Markey bill closely resembled generation portfolios, with the least carbon-intensive utilities being the most supportive (Downie 2017).

DE: A high share of revenue from fossil fuels and high carbon intensity contributes to the absence of climate policy support.

Low-Carbon Transition (LT)

The more firms have invested in transitioning to a low-carbon business model, the more likely they are to support policy. Their compliance costs would be lower than those of their competitors, or a new market would be created for their products or services (Falkner 2005, 2008; Vormedal 2010). For example, the EU Low Fares Airlines Association was more supportive of expanding the European Union Emissions Trading System (EU ETS) to cover air travel than the Association of European Airlines, as the former had already invested in more fuel-efficient fleets for cost reduction purposes (Meckling 2015).

For all companies except car manufacturers, low-carbon transition will be measured through two indicators: renewable energy produced or used as a share of total energy produced or used and the increase in this share over the past three years. For car companies, it will be measured through the fleet average CO2 emissions and the percentage-point change over the past three years, as this is where car manufacturers are most exposed to fossil fuels.

DE: Being in a transition to a low-carbon business model contributes to climate policy support.

Transnational Nature (TN)

For the avoidance of doubt: all companies studied in this article are transnational. However, some are more global than others. The higher a TNC scores on Transnational Nature, the more it is exposed to different regulatory environments. Because TNCs depend on economies of scale, they are unable to benefit from lower regulatory standards in some countries, creating a disadvantage versus domestic producers. TNCs often model production facilities on those already built in home countries and therefore often have higher capabilities to control pollution than domestic competitors in less regulated markets. Therefore, more transnational firms would be more likely to support international rule-setting and harmonization. The transaction costs of operating in various regulatory environments would also be decreased (Chasek et al. 2016; Falkner 2008; Leonard 1988; Lux et al. 2011; Meckling and Hughes 2018; Perkins and Neumayer 2012; Vormedal 2010).

Some authors have conversely argued that profit-maximizing firms externalize their costs by shifting production to states with lower standards.4 No conclusive evidence has, however, been found that environmental regulation specifically has driven such shifts. Leonard (1988, 231) finds that “the costs and logistics of complying with environmental regulations are not a decisive factor in most industrial decisions about desirable plant locations or in the international competitive picture of most major industries.” Other factors include market factors, labor costs, tax incentives, availability of infrastructure and transport, and political stability. Environmental control costs do not override such elements (Leonard 1988).

Several case studies confirm this. German chlorofluorocarbon (CFC) producers that took the lead in the CFC phase-out actively lobbied the German government to harmonize CFC reduction targets across Europe (Falkner 2005). A study of automobile emission standards found that developing countries that export automobiles and related components to markets with more stringent regulations and/or that have a higher influx of foreign direct investment (FDI) into their automobile sectors are more likely to have higher emission standards themselves (Perkins and Neumayer 2012).

The defining characteristics of a TNC are having operations in different countries and being exposed to different regulatory environments (Arts 2003). Therefore, TN will be operationalized through three indicators: number of countries operated in, number of continents operated in, and global distribution of revenue.

DE: A high degree of transnational operations contributes to firms’ climate policy support.

Exposure to Consumers (EC)

Companies that are in direct contact with consumers are more likely to favor stringent climate policy, as it will improve their consumer image. Companies have become regular targets of activism, in part due to growing consumer interest and willingness to “vote with your wallet.” The more a company sells to consumers, the more susceptible it is to consumer pressure (Bartley and Child 2014). As public concern over climate change increases, firms that have experienced heavy public scrutiny are more likely to adopt a proactive climate strategy (Skjærseth and Skodvin 2001). For example, Tesco was pushed by consumer pressure to pledge to reduce GHG emissions by 50 percent between 2008 and 2020 and started experimenting with labeling products according to their GHG footprint (Sprengel and Busch 2011). Upstream companies,5 on the other hand, will more likely oppose it as they face increased production costs without reaping reputational benefits (Falkner 2008).

Nongovernmental organizations (NGOs) have effectively used consumer boycotts to effect changes in companies’ policies (McAteer and Pulver 2009). For example, Greenpeace’s public campaign against Royal Dutch Shell’s proposed disposal of the Brent Spar oil storage buoy at sea caused greater losses than the additional cost of alternatives to Shell’s proposed sea decommissioning. Shell initiated a major reorganization in response to public scrutiny related to the Brent Spar and other incidents (Skjærseth and Skodvin 2001).

Exposure to Consumers will be measured through two indicators: the share of revenue from direct consumer transactions and the number of environmental controversies.

DE: A high share of revenue from consumers and environmental controversies contributes to climate policy support.

Institutional Environment (IE)

The location of a TNC’s headquarters, and thus its senior management responsible for strategy, plays a key role in climate policy supportiveness (Levy 2005; Skjærseth and Skodvin 2001). Less or no policy support in US companies compared to European ones cannot be explained by economic or technological characteristics alone (Levy 2005; Skjærseth and Skodvin 2001; Van Halderen et al. 2016). Previous experience with regulation and a bigger regulatory threat have led European firms to accept the inevitability of climate policies and constructively engage in the policy debate. A lower regulatory threat in the United States has conversely allowed US firms to attempt to entirely block policy (Kolk and Levy 2001; Newell 2000). The fact that Rio Tinto was the only top-ten US coal producer that supported US climate policy, including the Waxman–Markey bill and Clean Power Plan, is partially explained by previous exposure to similar debates in the European Union through its UK headquarters (Downie 2017). Strange (1992) argues that whether a company’s headquarters location helps predict its behavior and interests depends on how geographically dispersed it is, as high dispersion creates political tensions and competing interests within the company. One specific question is hence whether Institutional Environment is only found to influence climate policy support for firms that score low on Transnational Nature.

Second, industry and/or company histories matter. Oil companies that were more supportive of climate policy in the 1990s–2000s, such as BP, Shell, and Texaco, believed in first-mover advantages and preferred acquiring new competencies through gradual, internal, organic growth. Those that were more obstructive, such as ExxonMobil, were expected to follow an acquisition strategy once technology had progressed, risk had declined, and regulatory pressure had increased (Kolk and Levy 2001; Levy 2005).

DE: Institutional Environments with high environmental policy stringency and organic growth strategies contribute to climate policy support.

Investor Pressure (IP)

Investors are argued to hold influence over TNCs’ climate-related activities, as they can directly pressure polluting firms (Cadez et al. 2019; McAteer and Pulver 2009). Publicly traded companies depend on investors for crucial financial resources (Van Halderen et al. 2016), and sustainability practices have been found to affect firms’ creditworthiness (Sprengel and Busch 2011).

Investors are increasingly considering climate-related risks (Mercer 2015). Institutional investors increasingly demand data on GHG reduction strategies and current emissions, evidenced through exponential growth in the Carbon Disclosure Project, Dow Jones Sustainability Index, and other mechanisms (Sprengel and Busch 2011).

Recent investor votes forcing companies to change course on climate change have “sent a shockwave throughout the industry,” signaling “that investors will wait no more for boards who fail to grasp the speed of the energy transition” (Doherty, as cited in Welsh and Passoff 2018, 17). Climate change–related shareholder resolutions almost doubled in the United States from 36 in 2010 to 71 in 2018, reaching well beyond fossil fuel companies to firms such as Apple and Verizon (Welsh and Passoff 2018). Resolutions are increasingly stringent: whereas most asked companies to disclose GHG emissions in 2012, in 2017 the majority concerned climate change strategy and risk reporting (Cook, as cited in Welsh and Passoff 2018).

Even ExxonMobil, widely perceived as one of the most obstructive companies regarding climate policy, has had to change course. The first climate change–related shareholder resolution, in 1998, received 4 percent of the vote. Subsequent resolutions received 10 (2003), 25 (2005), 27.5 (2008), 38 (2016), and 62 percent (2017), prompting Exxon to ramp up its investment in new fuel technologies (Van Halderen et al. 2016).

DE: Investor Pressure will contribute to climate policy support.

Methodology

Qualitative Comparative Analysis

This article asks why some TNCs support climate policy, while others do not. It combines previously separate insights on the drivers of TNCs’ climate policy positions and analyzes them over a medium-N sample of twenty-five cases. QCA is most suitable to address this question for two reasons.

First, this question is best understood in terms of necessity and sufficiency.6 As necessary conditions need to be present for the outcome to occur, a necessary condition is found to hold explanatory power across the study’s cases. Additionally, a sufficiency analysis will uncover which different configurations of conditions are linked to supporting climate policy across the sample. Therefore, QCA enables the study to uncover which findings from small-N case study literature are applicable across a wider set of cases, sectors, and geographies.

Second, QCA is most suitable because it analyzes combinations of conditions to explain how they are linked to a certain outcome (Marx 2008). The reasons why TNCs support climate policy are more likely to be combinations of factors rather than isolated variables. In other words, the study’s interest is not in correlations between variables but in relations between sets (Schneider and Wagemann 2012).

Variation in approaches to climate policy has been studied through alternative methods, such as bi- and multivariate statistics (e.g., Røttereng 2018). Statistical methods are well adapted to provide evidence for certain relationships. However, they treat independent variables as independent causes of an outcome and then assess which variable is most important, while QCA assumes causation is not simple but complex, allowing for “equifinality”—multiple combinations of conditions may generate the same outcome (Berg-Schlosser et al. 2009; Marx 2008; Primc and Čater 2015; Schneider and Wagemann 2012). QCA further broadens causality by relaxing several usual assumptions, including those of additivity, conjunctural causation, uniformity, and symmetry (Schneider and Wagemann 2012).7 Finally, QCA looks at the causes of an outcome rather than the outcomes of a cause, equipping it well for systematically testing existing theories over a medium-N sample of cases (Andreas et al. 2017).

This study employs fuzzy set QCA (fsQCA), which expresses case membership in conditions in degrees between 0 and 1 instead of dichotomizing into either 0 or 1. FsQCA still dichotomizes: case membership of <0.5 denotes “more out than in” and >0.5 denotes “more in than out” of a condition. Therefore, fsQCA enables a more demanding and precise assessment of sufficiency and necessity, minimizes the loss of empirical information and decreases sensitivity to the location of the dichotomization threshold, and is considered best practice by proponents (Ragin 2009; Schneider and Wagemann 2012).

This analysis will use four fuzzy set levels (see Table 1), as it cannot rely on a strong enough body of substantive and theoretical knowledge to justify more detailed set membership scores (Ragin 2009). Appendix B reports the qualitative anchors used to translate case data into fuzzy values.8

Table 1 
Conversion of InfluenceMap Scores to Fuzzy Scores on the Outcome Variable
InfluenceMap ScoreFuzzy Set Score on Climate Policy SupportCase Membership in a Condition/Set of Conditions
>60 Fully in 
50–60 0.67 More in than out 
50 a   
40–50 0.33 More out than in 
<40 Fully out 
InfluenceMap ScoreFuzzy Set Score on Climate Policy SupportCase Membership in a Condition/Set of Conditions
>60 Fully in 
50–60 0.67 More in than out 
50 a   
40–50 0.33 More out than in 
<40 Fully out 
a

Point of maximum ambiguity; not scored.

Data Sources

Appendix9 A lists all data sources. The outcome variable, Climate Policy Support, is based on InfluenceMap’s Lobbying and Corporate Influence database.10 InfluenceMap examines publicly available information, including company websites, advertising campaigns, social media accounts, voluntary disclosures (e.g., CDP, EU Transparency Register), legislative consultations, and reports by respected media sources. InfluenceMap queries these sources for transparency and the content of an organization’s position, examining key policy areas, such as energy policy, emissions trading or taxation, and energy efficiency standards. TNCs are ranked based on their own performance and that of their advocacy groups and trade associations (InfluenceMap 2018).

To my knowledge, InfluenceMap provides the only comparative assessment of the world’s top 100 TNCs’ climate policy support. Particular strengths are the breadth of data sources consulted to inform the overall score and the consideration of both a TNC’s own positions and those of the indirect influencers with which it is affiliated. The overall InfluenceMap score is expressed in a value between 0 and 100. A score of greater than 60 signals that a TNC actively supports policies for a low-carbon future, while a score of less than 40 signals that a TNC actively obstructs this (InfluenceMap 2018).11 These values will serve as the qualitative anchors (see Table 1).

Sample Selection

This study selected cases from the 2018 Forbes Global 2000.12 This list’s focus on size—it ranks companies based on a combination of sales, profits, assets, and market value—means that this study’s universe predominantly contains North American and European, and, to a lesser extent, Asian, companies. The universe also excludes conglomerates like Berkshire Hathaway. Conglomerates own multiple companies that often operate independently, differing on the conditions analyzed in this article. Therefore, these companies would have to be analyzed separately rather than combined under their holding company. It also excludes financial services companies because InfluenceMap excludes these. This limits the analysis, as such companies play an important role in the economy and have supported flagship climate policies like the EU ETS (Meckling 2015; Newell and Paterson 1998). Within this inherently limited universe, a sample was selected to achieve “a maximum of heterogeneity over a minimum number of cases” (Berg-Schlosser and De Meur 2009, 21). For each sector, the three highest-ranking Forbes Global 2000 companies were selected. An exception was made for the technology sector, which was most heavily represented in the Global 2000: a fourth company was added for this sector, Intel, to increase variation on the outcome variable. Heterogeneity was maximized in terms of variation in the world region of their origin and score on the outcome variable. This resulted in a sample of thirty-one TNCs.

Table 2 shows the sample with the abbreviations that will be used to refer to each company in subsequent tables. Six cases13 had to be excluded because key data were unavailable, mainly due to limited public reporting. After exclusions, the sample counted twenty-five TNCs.

Table 2 
Companies Included in Study
CaseSectorRegion of OriginForbes 2018 RankfsQCA Score on OutcomeAbbreviation
Ford Motor Car North America 67 FOR 
Toyota Motor Car Asia 12 0.33 TOY 
Volkswagen Group Car Europe 16 0.33 VW 
BP Energy Europe 36 BP 
ExxonMobil Energy North America 13 EXX 
Royal Dutch Shell Energy Europe 11 0.33 RDS 
Bayer Pharmaceuticals Europe 100 BAY 
Novartis Pharmaceuticals Europe 63 NOV 
Pfizer Pharmaceuticals North America 44 0.67 PFI 
BHP Billiton Raw materials Oceania 108 0.33 BHP 
Glencore International Raw materials Europe 64 GLE 
Rio Tinto Group Raw materials Europe 111 0.33 RIO 
Walmart Stores Retail North America 24 WAL 
Anheuser Busch InBev Retail Europe 41 ANB 
Home Depot Retail North America 121 0.33 HOM 
Apple Technology North America APP 
Samsung Electronics Technology Asia 14 SAM 
Microsoft Technology North America 20 MIC 
Intel Technology North America 49 0.67 INT 
AT&T Telecom North America 15 0.33 ATT 
Verizon Communications Telecom North America 18 VER 
Deutsche Telekom Telecom Europe 79 DET 
Enel Utilities Europe 75 ENE 
EDF Utilities Europe 115 EDF 
Iberdrola Utilities Europe 146 IBE 
CaseSectorRegion of OriginForbes 2018 RankfsQCA Score on OutcomeAbbreviation
Ford Motor Car North America 67 FOR 
Toyota Motor Car Asia 12 0.33 TOY 
Volkswagen Group Car Europe 16 0.33 VW 
BP Energy Europe 36 BP 
ExxonMobil Energy North America 13 EXX 
Royal Dutch Shell Energy Europe 11 0.33 RDS 
Bayer Pharmaceuticals Europe 100 BAY 
Novartis Pharmaceuticals Europe 63 NOV 
Pfizer Pharmaceuticals North America 44 0.67 PFI 
BHP Billiton Raw materials Oceania 108 0.33 BHP 
Glencore International Raw materials Europe 64 GLE 
Rio Tinto Group Raw materials Europe 111 0.33 RIO 
Walmart Stores Retail North America 24 WAL 
Anheuser Busch InBev Retail Europe 41 ANB 
Home Depot Retail North America 121 0.33 HOM 
Apple Technology North America APP 
Samsung Electronics Technology Asia 14 SAM 
Microsoft Technology North America 20 MIC 
Intel Technology North America 49 0.67 INT 
AT&T Telecom North America 15 0.33 ATT 
Verizon Communications Telecom North America 18 VER 
Deutsche Telekom Telecom Europe 79 DET 
Enel Utilities Europe 75 ENE 
EDF Utilities Europe 115 EDF 
Iberdrola Utilities Europe 146 IBE 

Time Frame

Cases were scored on conditions based on annual data for 2015–2017 inclusive, using the average value for each indicator over the three-year period. This period purposefully starts with the year in which the landmark 2015 Paris Agreement was reached. The year 2015 was widely seen as the world’s last chance to get a sufficiently ambitious global climate agreement in time to limit global warming below 2°C. The foundations for Paris were laid at COP15 (2009, Copenhagen), where a compromise was reached that paved the way for voluntary mitigation targets from countries. At COP19 (2013, Warsaw), member states were asked to outline their intended post-2020 climate actions. The United States–China joint announcement14 of November 12, 2014, was “perhaps the biggest breakthrough in the climate negotiations.” Among others, their announcement emphasized the two countries’ personal commitment to a successful outcome in Paris and their determination to decisively implement domestic policies and transition to low-carbon, green economies. This removed “a significant hurdle on the road to Paris” (Chasek et al. 2016, 176–180). TNCs might have changed their climate policy positions as negotiations intensified, especially after the United States–China joint announcement.

In addition, this time frame aligns well with that of the data InfluenceMap uses to determine TNCs’ scores on climate policy supportiveness. It considers data from the last two years while using slightly older sources where no new sources are available and no change in position is apparent. As InfluenceMap scores from 2018 were used, they are based on data for 2016–2018, with occasional slightly older data where necessary.

Results

The QCA was run both for the presence of the outcome, CLIMATE POLICY SUPPORT (CPS), and for the absence, climate policy support (cps),15 as is common in QCA practice (Schneider and Wagemann 2012). The results are presented in that order.16 Analysis was performed using fsQCA 2.0 software.17Table 3 shows the fuzzy set data matrix used for fsQCA.

Table 3 
Data Matrix with Fuzzy Set Scores for All Cases on the Conditions and Outcome CPS
CompanyCondition: SectorEFLTTNECIEIPCPS
APP Technology 0.00 1.00 0.67 0.33 0.67 0.67 1.00 
SAM Technology 0.00 0.33 1.00 0.00 0.67 0.00 1.00 
MIC Technology 0.00 1.00 0.67 0.00 1.00 0.00 1.00 
INT Technology 0.00 1.00 0.33 0.00 1.00 0.00 0.67 
RDS Energy 1.00 0.00 1.00 1.00 0.67 1.00 0.33 
EXX Energy 0.67 0.00 0.67 0.33 0.33 1.00 0.00 
BP Energy 1.00 0.00 1.00 0.67 0.33 1.00 0.00 
TOY Car 0.67 1.00 0.67 0.00 1.00 0.00 0.33 
VW Car 0.67 1.00 0.67 0.67 0.67 0.00 0.33 
FOR Car 0.67 0.00 0.67 0.33 0.67 0.67 0.00 
ATT Telecom 0.00 0.00 0.00 0.00 0.33 0.00 0.33 
VER Telecom 0.00 0.00 0.00 0.00 0.67 0.67 1.00 
DET Telecom 0.00 0.67 0.33 0.00 0.67 0.00 1.00 
WAL Retail 0.00 0.33 0.00 0.00 0.67 0.00 1.00 
ANB Retail 0.00 1.00 1.00 0.00 0.33 0.00 1.00 
HOM Retail 0.00 0.33 0.00 0.33 0.33 1.00 0.33 
PFI Pharmaceuticals 0.00 0.00 0.67 0.00 1.00 0.00 0.67 
NOV Pharmaceuticals 0.00 0.67 0.67 0.00 1.00 0.00 1.00 
BAY Pharmaceuticals 0.67 0.00 1.00 0.00 0.67 0.00 0.00 
GLE Raw materials 1.00 0.00 0.67 0.00 0.00 1.00 0.00 
BHP Raw materials 0.67 0.00 0.00 0.00 1.00 0.33 0.33 
RIO Raw materials 0.67 1.00 0.67 0.00 0.67 1.00 0.33 
ENE Utilities 0.67 0.67 0.67 0.33 0.67 0.00 1.00 
EDF Utilities 0.00 1.00 0.00 0.00 1.00 0.00 1.00 
IBE Utilities 0.67 0.67 0.33 0.00 0.67 0.00 1.00 
CompanyCondition: SectorEFLTTNECIEIPCPS
APP Technology 0.00 1.00 0.67 0.33 0.67 0.67 1.00 
SAM Technology 0.00 0.33 1.00 0.00 0.67 0.00 1.00 
MIC Technology 0.00 1.00 0.67 0.00 1.00 0.00 1.00 
INT Technology 0.00 1.00 0.33 0.00 1.00 0.00 0.67 
RDS Energy 1.00 0.00 1.00 1.00 0.67 1.00 0.33 
EXX Energy 0.67 0.00 0.67 0.33 0.33 1.00 0.00 
BP Energy 1.00 0.00 1.00 0.67 0.33 1.00 0.00 
TOY Car 0.67 1.00 0.67 0.00 1.00 0.00 0.33 
VW Car 0.67 1.00 0.67 0.67 0.67 0.00 0.33 
FOR Car 0.67 0.00 0.67 0.33 0.67 0.67 0.00 
ATT Telecom 0.00 0.00 0.00 0.00 0.33 0.00 0.33 
VER Telecom 0.00 0.00 0.00 0.00 0.67 0.67 1.00 
DET Telecom 0.00 0.67 0.33 0.00 0.67 0.00 1.00 
WAL Retail 0.00 0.33 0.00 0.00 0.67 0.00 1.00 
ANB Retail 0.00 1.00 1.00 0.00 0.33 0.00 1.00 
HOM Retail 0.00 0.33 0.00 0.33 0.33 1.00 0.33 
PFI Pharmaceuticals 0.00 0.00 0.67 0.00 1.00 0.00 0.67 
NOV Pharmaceuticals 0.00 0.67 0.67 0.00 1.00 0.00 1.00 
BAY Pharmaceuticals 0.67 0.00 1.00 0.00 0.67 0.00 0.00 
GLE Raw materials 1.00 0.00 0.67 0.00 0.00 1.00 0.00 
BHP Raw materials 0.67 0.00 0.00 0.00 1.00 0.33 0.33 
RIO Raw materials 0.67 1.00 0.67 0.00 0.67 1.00 0.33 
ENE Utilities 0.67 0.67 0.67 0.33 0.67 0.00 1.00 
EDF Utilities 0.00 1.00 0.00 0.00 1.00 0.00 1.00 
IBE Utilities 0.67 0.67 0.33 0.00 0.67 0.00 1.00 

CLIMATE POLICY SUPPORT (CPS)

No single condition was found to be necessary by the QCA analysis. However, the combination of exposure to fossil fuels + LOW-CARBON TRANSITION was found to be necessary. This means already not being exposed to fossil fuels, or transitioning to a low-carbon business model, leads TNCs to support climate policy. Combinations of conditions can be necessary in QCA if they are both manifestations of a common higher-order concept, which in this case is not being exposed to fossil fuels in the future. In QCA, the coverage parameter indicates the relevance of necessary conditions; a low score would mean that the condition covers many more cases than the outcome (Schneider and Wagemann 2012, 144–145). Coverage values below 0.5 indicate irrelevance. On the other hand, even if a “candidate” necessary condition has very high coverage values, it could still be irrelevant if both the condition and outcome are close to constants, meaning they vary very little across the entire set of cases (Schneider and Wagemann 2012, 146). Table 3 shows that exposure to fossil fuels, LOW-CARBON TRANSITION, and CLIMATE POLICY SUPPORT vary across the cases. In addition, exposure to fossil fuels + LOW-CARBON TRANSITION has a coverage value of 0.73. Therefore, we can conclude that this is a relevant necessary combination that explains variation in the data set on the outcome variable.

To analyze sufficiency, a truth table is constructed (see Table 4). This table sorts cases by their membership of each condition, in other words, whether they are exposed to fossil fuels or not, et cetera. The QCA software’s truth table algorithm delivers various solution terms. The intermediate solution is the core of sufficiency analysis, as it uses only “easy” logical remainders,18 which are both in line with empirical evidence and DEs about whether conditions contribute to the outcome in their presence or absence, in other words, those that are judged to make sense according to the researcher’s theoretical and substantive knowledge (Ragin 2009). The intermediate-solution term delivered by the truth table algorithm is exposure to fossil fuels * LOW-CARBON TRANSITION * TRANSNATIONAL NATURE + LOW-CARBON TRANSITION * transnational nature * INSTITUTIONAL ENVIRONMENT + exposure to fossil fuels * transnational nature * INSTITUTIONAL ENVIRONMENT * investor pressureCLIMATE POLICY SUPPORT. Figure 1 breaks up the intermediate solution into its constitutive paths and shows which cases are covered, uniquely covered by each path, and uncovered by the entire solution term.

Table 4 
Truth Table for CPS
RowEFLTTNECIEIPNo. CasesCases
INT, DET, EDF 
MIC, NOV 
ANB 
IBE 
APP 
WAL 
TOY; ENE 
SAM, PFI 
ATT 
10 VER 
11 VW 
12 BHP 
13 RIO 
14 BAY 
15 HOM 
16 BP 
17 RDS 
18 EXX, GLE 
19 FOR 
RowEFLTTNECIEIPNo. CasesCases
INT, DET, EDF 
MIC, NOV 
ANB 
IBE 
APP 
WAL 
TOY; ENE 
SAM, PFI 
ATT 
10 VER 
11 VW 
12 BHP 
13 RIO 
14 BAY 
15 HOM 
16 BP 
17 RDS 
18 EXX, GLE 
19 FOR 
Figure 1 

Intermediate-Solution “Paths” and Covered Cases for CPS

Figure 1 

Intermediate-Solution “Paths” and Covered Cases for CPS

Climate Policy Support (CPS)

No single necessary condition was found for climate policy support. However, the combination of EXPOSURE TO FOSSIL FUELS + low-carbon transition was found to be necessary for climate policy support. In other words, companies that are exposed to fossil fuels and not taking steps to reduce that exposure are found not to support climate policy. This combination constitutes the higher-order concept of being exposed to fossil fuels for the foreseeable future. With a coverage value of 0.63, this causal combination has a comparatively lower relevance than the finding for CLIMATE POLICY SUPPORT.

Table 5 constitutes the truth table used for the sufficiency analysis for climate policy support. The truth table algorithm produces the following intermediate solution: EXPOSURE TO FOSSIL FUELS * exposure to consumers * INVESTOR PRESSURE + EXPOSURE TO FOSSIL FUELS * low-carbon transition * INVESTOR PRESSUREclimate policy support. Figure 2 again breaks the intermediate solution up into its constitutive paths, providing information on covered and uncovered cases.

Table 5 
Truth Table for cps
RowEFLTTNECIEIPNo. CasesCasesConsistency
EXX, GLE 
BP 
FOR 
RIO 
RDS 
HOM 0.80 
VW 0.67 
VER 0.66 
BHP 0.60 
10 BAY 0.60 
11 TOY; ENE 0.50 
12 IBE 0.40 
13 ATT 0.38 
14 WAL 0.36 
15 APP 0.33 
16 SAM, PFI 0.30 
17 MIC, NOV 0.23 
19 INT, DET, EDF 0.20 
20 ANB 0.12 
RowEFLTTNECIEIPNo. CasesCasesConsistency
EXX, GLE 
BP 
FOR 
RIO 
RDS 
HOM 0.80 
VW 0.67 
VER 0.66 
BHP 0.60 
10 BAY 0.60 
11 TOY; ENE 0.50 
12 IBE 0.40 
13 ATT 0.38 
14 WAL 0.36 
15 APP 0.33 
16 SAM, PFI 0.30 
17 MIC, NOV 0.23 
19 INT, DET, EDF 0.20 
20 ANB 0.12 
Figure 2 

Intermediate-Solution “Paths” and Covered Cases for cps

Figure 2 

Intermediate-Solution “Paths” and Covered Cases for cps

Robustness

Robustness is determined by examining the effect of different methodological choices, where plausible (Schneider and Wagemann 2012). Appendix B presents the performed robustness checks in detail. The coverage of necessary combination of EXPOSURE TO FOSSIL FUELS + low-carbon transition did decrease from 0.63 to 0.55, which indicates that the finding of this necessary condition for climate policy support is less robust than the inverse combination that was found for CLIMATE POLICY SUPPORT. All other findings were robust. Including additional truth table rows in the sufficiency analysis, one of the different choices that could be made, leads to finding one additional sufficient solution for climate policy support: EXPOSURE TO FOSSIL FUELS * low-carbon transition * TRANSNATIONAL NATURE. This confirms the sufficiency analysis for CLIMATE POLICY SUPPORT, which found that both TRANSNATIONAL NATURE and transnational nature can be part of sufficient terms.

Discussion

Key Findings

The finding of exposure to fossil fuels + LOW-CARBON TRANSITION as a necessary condition for CLIMATE POLICY SUPPORT confirms previous inferences that the most fossil fuel–intensive companies are also the least supportive of climate policy. The fact that exposure to fossil fuels by itself is not necessary for CLIMATE POLICY SUPPORT, but is necessary in combination with LOW-CARBON TRANSITION, supports the argument that firms are more concerned with the effects of policy in relation to their competitors rather than in absolute terms (Downie 2017; Falkner 2008; Meckling 2015). Iberdrola and Enel stand out as the only two companies that exhibit EXPOSURE TO FOSSIL FUELS and CLIMATE POLICY SUPPORT. The specific history of European utilities may provide an explanation. European utilities lost more than €100 billion in market value between 2008 and 2013 because they failed to accurately predict the speed of wind and solar photovoltaic cost deflation and coal phase-out (Gray 2015). This forced them to face the reality of the energy transition and may well have provided reasons to constructively engage with climate policy making rather than opposing it (Meckling 2015). The necessary condition that was found for climate policy support, EXPOSURE TO FOSSIL FUELS + low-carbon transition, is the opposite of the necessary condition for CLIMATE POLICY SUPPORT. Although the necessary combination was found to be more relevant and robust for CLIMATE POLICY SUPPORT than for climate policy support, the fact that these conditions—in opposite form—were found to be necessary for both outcomes strengthens the evidence that Exposure to Fossil Fuels and Low-Carbon Transition are key determinants of climate policy strategies. This is an important finding, as it tells us that the necessary conditions that explain the climate policy positions of the heavy carbon emitters, which have been the focus of the literature to date, also apply to less GHG-intensive sectors, such as retail, technology, and telecommunications.

Iberdrola is uniquely covered by path LOW-CARBON TRANSITION * transnational nature * INSTITUTIONAL ENVIRONMENT. This sufficient path draws attention to the presence of a supportive institutional environment. Therefore, this finding confirms the directional expectation that Institutional Environments with high environmental policy stringency and organic growth strategies contribute to CLIMATE POLICY SUPPORT, although the role of Institutional Environment is clearly less important in determining TNCs’ climate policy strategies than the role of conditions Exposure to Fossil Fuels and Low-Carbon Transition.

The finding of TRANSNATIONAL NATURE in one sufficient path combined with finding transnational nature in another sufficient path suggests that there is no systematic evidence that TRANSNATIONAL NATURE contributes to CLIMATE POLICY SUPPORT. Firms that are not exposed to fossil fuels and transitioning to a low-carbon future support climate policy if they are transnational and therefore would benefit from a level international playing field. In other cases, firms that are not transnational can still support climate policy if they have been exposed to an institutional environment with high environmental policy stringency and have pursued organic growth strategies—the commonalities between the second and third solution terms. This suggests that being exposed to different policy strengths in a multitude of markets and being exposed to stringent environmental policy at home are interchangeable in being part of combinations that are sufficient for climate policy support. Furthermore, the finding of transnational nature in both sufficient paths that also contain INSTITUTIONAL ENVIRONMENT supports Strange’s (1992) argument that firms’ headquarters locations only predict their behavior and interests if they are not too geographically dispersed.

Finding INVESTOR PRESSURE in both sufficient paths for climate policy support is surprising. The expectation was that this condition would contribute to CLIMATE POLICY SUPPORT. The effects of recent climate-related shareholder resolutions may need more time to materialize than the short time frame of this study, especially as shareholder engagement is often a multiyear, escalating process. Another explanation would be that investors have tackled the lowest-hanging fruit—the companies that are most exposed to fossil fuels and taking the least action to reduce that exposure.

Furthermore, the finding of exposure to consumers in path EXPOSURE TO FOSSIL FUELS * exposure to consumers * INVESTOR PRESSURE, sufficient for climate policy support, is noteworthy, as EXPOSURE TO CONSUMERS was not found to be part of any sufficient combination for CLIMATE POLICY SUPPORT. Some TNCs have experienced EXPOSURE TO CONSUMERS without changing their climate policy position to CLIMATE POLICY SUPPORT, for example, Volkswagen. Many TNCs that support policy have not been exposed to consumer campaigns, probably because their environmental performance is already relatively strong. In sum, the evidence on Exposure to Consumers does not link it to CLIMATE POLICY SUPPORT.

Limitations and Further Research

This study was limited by incomplete data in two ways. First, the condition Technological Advancement had to be excluded. Previous literature has argued that companies that have already invested or can more easily invest in new, cleaner technology will support international policy, as it would give them a competitive edge (Falkner 2008; Vormedal 2010). A prime example was the electronics industry in the ozone case: in contrast to other CFC users, electronics manufacturers were quick to eliminate ozone-depleting substances because their success had been built on the capacity to innovate and rapidly respond to technological advances and dynamic market conditions (Falkner 2005). This would have been operationalized through past and planned climate-related R&D spend. However, companies do not typically disclose any product- or technology-specific details of their R&D.

Second, a subindicator19 of Institutional Environment was also dropped due to a lack of data. Previous literature has argued that companies’ climate policy positions are path dependent: a long history of opposition to environmental regulation explains more recent opposition to climate policy. The more vehement opposition of the coal industry to the Waxman–Markey bill and Clean Power Plan in comparison to coal-dependent utilities was explained by the former’s long history of opposition to environmental regulation (Downie 2017). Because InfluenceMap only started compiling its data set from 2015, no comparable measure of historical climate policy positions was available. Once comparable data are available, this would be an interesting avenue for future research.

While data were available on other indicators, it was still occasionally limited. For indicator IE1, “environmental policy stringency in HQ location,” the only source of comparable data is the Organisation for Economic Co-operation and Development (OECD; 2018) environmental policy stringency index. The last year for which this index scores all the relevant countries is 2012. As governments and policy priorities change, environmental policy may have become more, or less, stringent since then. In addition, no comparable, coherent data exist on TNCs’ growth strategies. Therefore, whether TNCs seek growth organically or through acquisitions was determined by searching the Factiva database of global news and insight. Because this database includes transcripts of earning calls where senior management regularly explains growth strategy, a judgment on whether a TNC was pursuing internal or external growth was possible.20

QCA can only analyze a limited number of conditions. If the number of possible combinations of conditions far exceeds the number of cases, solution terms would become too complex and only apply to single cases. As the interpretation of QCA’s results depends on comparing similarities and differences across cases, interpreting solution terms in a theoretically meaningful manner would become difficult (Berg-Schlosser and De Meur 2009; Schneider and Wagemann 2012). While this study considered a broad array of literature when formulating the conditions for analysis, and all conditions which were supported widely were included, some more recent, evolving areas of research could not be included. One example constitutes companies’ senior leadership and its degree of “socialization”—exposure to external stakeholders through diverse multistakeholder sustainability networks—as an explanatory factor for green practices in corporates (Van der Ven 2014). Another argues that asset-specific investment into human resources or technological specialization leads companies to adhere to environmental and other standards or voluntarily reintegrate negative externalities (Thauer 2014). Future QCA research could include these factors.

As mentioned previously, the universe of TNCs from which this study’s sample was selected predominantly contains TNCs from North America, Europe, and core Asian economies—China, India, Japan, and South Korea. Therefore, the generalizability of this study’s findings is limited to these regions and does not extend to most developing countries. While it may be complicated to construct a comparable sample from a well-defined universe that would include both the world’s powerhouse TNCs and smaller companies from emerging economies, further studies on the climate policy engagement of such companies would build a more global understanding of the factors that underlie TNCs’ climate policy strategies.

Furthermore, this study’s short time frame may explain the surprising findings regarding Investor Pressure. A study with a longer time frame may be better poised touncover whether those companies that were pressured by investors on climate change become more supportive of climate policy over time than similar companies that are not pressured.

Finally, the two combinations of conditions that were found to be sufficient for climate policy support cover almost all fossil fuel and mining companies, with the exception of BHP Billiton, but almost none of the automobile companies, except for Ford Motors, and none of the telecommunications, retail, or pharmaceutical companies that are not found to support climate policy. This suggests that, while the findings of small-N literature can explain the necessary conditions that determine climate policy positions beyond the most climate-intensive sectors, they cannot explain sufficient conditions for climate policy positions when applied to TNCs from a broader set of sectors. This is probably due to a predominant focus by the literature on heavily polluting sectors, such as fossil fuel and heavy industry, while paying less attention to other sectors. Hence, an important avenue for further research would be to conduct small-N case studies to further explore sufficient conditions for a lack of climate policy support in companies that are not directly linked to fossil fuels. Specific cases that merit further research are Toyota, Volkswagen, AT&T, Home Depot, and Bayer. Further research into why BHP Billiton’s climate policy stance is not explained by this study’s findings, while its peers’ stances are, would also be worthwhile.

Conclusions

This study has aimed to fill a gap in the literature on the climate policy strategies of TNCs by systematically examining the conditions that explain variation in whether and to what extent TNCs support policies to tackle climate change. It shows which explanatory factors, identified in previous literature, are found to drive climate policy support over a broader set of cases, in terms of sector, carbon intensity, and geography of origin. It has found that TNCs that are not exposed to fossil fuels or are transitioning to a carbon-constrained future necessarily support such policy, while TNCs that are exposed to fossil fuels or not transitioning to a low-carbon future necessarily do not support such policy. Furthermore, it has found that an institutional environment where firms have been exposed to stringent environmental policy in the past, pursue organic growth strategies, and/or invest in large R&D budgets is a sufficient condition for policy support in combination with transitioning to a low-carbon future and a high concentration of operations in few countries. These conditions are found to explain variation in climate policy support among TNCs beyond those from heavily polluting sectors, such as energy, utilities, and automobiles, that have traditionally been studied. No evidence, however, was found to confirm theoretical expectations that exposure to consumers, a strong transnational nature, and pressure from investors are linked to policy support. This suggests more research is needed around those explanatory variables to better understand whether and how they drive TNCs’ climate policy positions.

Notes

1. 

Some scholars argue that TNCs should be distinguished from multinational companies (MNCs). MNCs are argued to replicate activities in different world regions, whereas TNCs are truly global, with operations divided across the globe (Arts 2003; Fuchs 2007). What matters for the purpose of this analysis is that a TNC/MNC has operations in different countries and is thus exposed to different regulatory environments.

2. 

Directional expectations state the direction in which a condition is expected to influence the outcome variable (Schneider and Wagemann 2012).

3. 

For an overview of how conditions are operationalized through indicators, see Appendix A, https://www.mitpressjournals.org/doi/suppl/10.1162/glep_a_00560.

4. 

E.g., Herman E. Daly, “The Perils of Free Trade,” Scientific American, November, 1993.

5. 

Upstream companies produce raw materials and manufacture products, while downstream companies distribute and market products. For example, Apple is a downstream company, while Foxconn Technology is an upstream company that manufactures Apple’s products.

6. 

Necessary conditions need to be present for the outcome to occur. Sufficient conditions are always present when the outcome occurs, but the outcome can also occur without their presence (Rihoux and Ragin 2009; Schneider and Wagemann 2012).

7. 

Additivity: several causes can be present simultaneously. Conjunctural causation: effects of single conditions may only unfold in combination with other conditions. Uniformity: a given cause may work toward an outcome in one causal combination, while working against it in other combinations. Symmetry: presence and absence of outcome may require different explanations, respectively. For a more detailed explanation, see Schneider and Wagemann (2012, 76–83).

8. 

Thresholds that determine when a case is fully a member of a set (fuzzy score of 1), fully not a member of a set (fuzzy score of 0), and neither a member nor not a member of a set (fuzzy score of 0.5).

9. 

The Appendices are available here: https://doi.org/10.1162/glep_a_00560.

10. 

Unfortunately, I am not able to include raw data on the outcome variable (Climate Policy Support), as this would require sharing InfluenceMap’s full data set (only a selection of its data are publicly available), which I agreed not to do as a condition for my obtaining the data. However, the fuzzy scores for the outcome variable and conversion rules followed shared in Tables 1 and 2 will give readers an idea of the band in which raw scores for each company fall.

11. 

More information about InfluenceMap can be found at http://www.influencemap.org/, last accessed April 27, 2020. For more detail on InfluenceMap’s methodology, including the full list of data sources and queries used, visit https://influencemap.org/page/Our-Methodology, last accessed April 27, 2020.

12. 

Forbes, “2018 Global 2000 Methodology: How We Crunch the Numbers,” June 6, 2018.

13. 

Foxconn Technology Group, Rosneft, Equinor, Honda Motor, Comcast, Nippon Telegraph and Telephone.

14. 

“US–China Joint Announcement on Climate Change,” White House, Office of the Press Secretary, November 12, 2014, available at: https://obamawhitehouse.archives.gov/the-press-office/2014/11/11/us-china-joint-announcement-climate-change, last accessed April 27, 2020.

15. 

Boolean notations are used: lowercase letters denote the absence of a condition—which corresponds to a fuzzy set value of 0 or 0.33—while uppercase letters denote the presence of a condition, corresponding to a fuzzy set value of 0.67 or 1. A plus means “or,” while an asterisk means “and.”

16. 

See Appendix B for a detailed description of how QCA was applied.

17. 

FsQCA 2.0 is available for download at http://www.u.arizona.edu/~cragin/fsQCA/software.shtml, last accessed April 27, 2020.

18. 

Logical remainders are truth table rows, or in other words, potential combinations of conditions, which are not observed in any of the cases studied.

19. 

Conditions were operationalized through indicators, and some indicators contained subindicators. For example, one of the two indicators through which Institutional Environment was operationalized was “industry/company histories.” This was calculated through two subindicators: “organic/internal growth strategies” and “innovative capacity.” The subindicator that was dropped, “history of environmental policy opposition,” would have been a third subindicator of “industry/company histories.” See Appendix A for a complete overview.

20. 

See Appendix B for more information on how TNCs’ growth strategies were determined.

References

Abreu
,
Mônica Cavalcanti Sá de
,
Aba Rita Pinheiro de
Freitas
, and
Sílvia Maria Dias Pedro
Rebouças
.
2017
.
Conceptual Model for Corporate Climate Change Strategy Development: Empirical Evidence from the Energy Sector
.
Journal of Cleaner Production
165
:
382
392
.
Andrade
,
José Célio Silveira
, and
José Antônio
Puppim de Oliveira
.
2014
.
The Role of the Private Sector in Global Climate and Energy Governance
.
Journal of Business Ethics
130
:
375
387
.
Andreas
,
Jan-Justus
,
Charlotte
Burns
, and
Julia
Touza
.
2017
.
Renewable Energy as a Luxury? A Qualitative Comparative Analysis of the Role of the Economy in the EU’s Renewable Energy Transitions During the “Double Crisis”
.
Ecological Economics
142
:
81
90
.
Arts
,
Bas
.
2003
.
Non-State Actors in Global Governance: Three Faces of Power
.
Gemeinschaftsgüter: Recht, Politik und Ökonomie
2003/4
:
1
53
.
Bartley
,
Tim
, and
Curtis
Child
.
2014
.
Shaming the Corporation: The Social Production of Targets and the Anti-Sweatshop Movement
.
American Sociological Review
79
(
4
):
653
679
.
Berg-Schlosser
,
Dirk
,
Gisèle
De Meur
,
Benoît
Rihoux
, and
Charles C.
Ragin
.
2009
.
Qualitative Comparative Analysis (QCA) as an Approach
. In
Configurational Comparative Methods: Qualitative Comparative Analysis (QCA) and Related Techniques
, edited by
Benoît
Rihoux
and
Charles C.
Ragin
,
1
18
.
Thousand Oaks, CA
:
Sage
.
Berg-Schlosser
,
Dirk
, and
Gisèle
De Meur
.
2009
.
Comparative Research Design: Case and Variable Selection
. In
Configurational Comparative Methods: Qualitative Comparative Analysis (QCA) and Related Techniques
, edited by
Benoît
Rihoux
and
Charles C.
Ragin
,
19
32
.
Thousand Oaks, CA
:
Sage
.
Betsill
,
Michele M.
2014
.
Transnational Actors in Environmental Politics
. In
Advances in International Environmental Politics
, edited by
Michele M.
Betsill
,
Kathryn
Hochstetler
, and
Dimitris
Stevis
,
185
210
.
London, UK
:
Palgrave Macmillan
.
Cadez
,
Simon
,
Albert
Czerny
, and
Peter
Letmathe
.
2019
.
Stakeholder Pressures and Corporate Climate Change Mitigation Strategies
.
Business Strategy and the Environment
28
:
1
14
.
Chasek
,
Pamela S.
,
David Leonard
Downie
, and
Janet Welsh
Brown
.
2016
.
Global Environmental Politics
.
Abingdon, UK
:
Routledge
.
Clapp
,
Jennifer
, and
Peter
Dauvergne
.
2005
.
Paths to a Green World: The Political Economy of the Global Environment
.
Boston, MA
:
MIT Press
.
Dauvergne
,
Peter
, and
Jane
Lister
.
2010
.
The Power of Big Box Retail in Global Environmental Governance: Bringing Commodity Chains Back into IR
.
Millennium: Journal of International Studies
39
(
1
):
145
160
.
Downie
,
Christian
.
2014
.
Transnational Actors in Environmental Politics—Strategies and Influence in Long Negotiations
.
Environmental Politics
23
(
3
):
376
394
.
Downie
,
Christian
.
2017
.
Fighting for King Coal’s Crown—Business Actors in the US and Coal Industries
.
Global Environmental Politics
17
(
1
):
21
39
.
Elliott
,
Lorraine
.
2004
.
The Global Politics of the Environment
. 2nd ed.
Basingstoke, UK
:
Palgrave Macmillan
.
Falkner
,
Robert
.
2005
.
The Business of Ozone Layer Protection: Corporate Power in Regime Evolution
. In
The Business of Global Environmental Governance
, edited by
David L.
Levy
and
Peter J.
Newell
,
105
134
.
Boston, MA
:
MIT Press
.
Falkner
,
Robert
.
2008
.
Business Power and Conflict in International Environmental Politics
.
Basingstoke, UK
:
Palgrave Macmillan
.
Falkner
,
Robert
.
2016
.
The Paris Agreement and the New Logic of International Climate Politics
.
International Affairs
92
(
5
):
1107
1125
.
Fuchs
,
Doris
.
2007
.
Business Power in Global Governance
.
Boulder, CO
:
Lynne Rienner
.
Fuchs
,
Doris
, and
Agni
Kalfagianni
.
2010
.
The Causes and Consequences of Private Food Governance
.
Business and Politics
12
(
3
):
1
34
.
Fuchs
,
Doris
.
2013
.
Theorizing the Power of Global Companies
. In
The Handbook of Global Companies
, edited by
John
Mikler
,
77
95
.
Chichester, UK
:
John Wiley
.
Gasbarro
,
Federica
,
Fabio
Iraldo
, and
Tiberio
Daddi
.
2017
.
The Drivers of Multinational Enterprises’ Climate Change Strategies: A Quantitative Study on Climate-Related Risks and Opportunities
.
Journal of Cleaner Production
160
:
8
26
.
Gray
,
Matthew
.
2015
.
Coal: Caught in the EU Utility Death Spiral
. Edited by
James
Leaton
,
Robert
Schuwerk
, and
Mark
Fulton
.
London, UK
:
Carbon Tracker Initiative
.
InfluenceMap
.
2018
.
Lobbying and Corporate Influence: Methodology
.
Available at: https://influencemap.org/page/Our-Methodology, last accessed April 27, 2020
.
Kamieniecki
,
Sheldon
.
2006
.
Corporate America and Environmental Policy: How Often Does Business Get Its Way?
Stanford, CA
:
Stanford University Press
.
Kolk
,
Ans
, and
David
Levy
.
2001
.
Winds of Change: Corporate Strategy, Climate Change and Oil Multinationals
.
European Management Journal
19
(
5
):
501
509
.
Leonard
,
Hugh Jeffrey
.
1988
.
Pollution and Struggle for the World Product
.
Cambridge, UK
:
Cambridge University Press
.
Levy
,
Daniel L.
, and
Peter J.
Newell
.
2002
.
Business Strategy and International Environmental Governance: Toward a Neo-Gramscian Synthesis
.
Global Environmental Politics
2
(
4
):
84
101
.
Levy
,
Daniel L.
, and
Peter J.
Newell
.
2005
.
Introduction: The Business of Global Environmental Governance
. In
The Business of Global Environmental Governance
, edited by
David L.
Levy
and
Peter J.
Newell
,
1
20
.
Boston, MA
:
MIT Press
.
Levy
,
David L.
2005
.
Business and the Evolution of the Climate Regime: The Dynamics of Corporate Strategies
. In
The Business of Global Environmental Governance
, edited by
David L.
Levy
and
Peter J.
Newell
,
73
104
.
Boston, MA
:
MIT Press
.
Lux
,
Sean
,
T.
Ruseel Crook
, and
David J.
Woehr
.
2011
.
Mixing Business with Politics: A Meta-Analysis of the Antecedents and Outcomes of Corporate Political Activity
.
Journal of Management
37
(
1
):
223
247
.
Marx
,
Axel
.
2008
.
Limits to Non-State Market Regulation: A Qualitative Comparative Analysis of the International Sport Footwear Industry and the Fair Labor Association
.
Regulation and Governance
2
:
253
273
.
McAteer
,
Emily
, and
Simone
Pulver
.
2009
.
The Corporate Boomerang: Shareholder Transnational Advocacy Networks Targeting Oil Companies in the Ecuadorian Amazon
.
Global Environmental Politics
9
(
1
):
1
30
.
Meckling
,
Jonas
.
2015
.
Oppose, Support, or Hedge? Distributional Effects, Regulatory Pressure, and Business Strategy in Environmental Politics
.
Global Environmental Politics
15
(
2
):
19
37
.
Meckling
,
Jonas
,
Nina
Kelsey
,
Eric
Biber
, and
John
Zysman
.
2015
.
Winning Coalitions for Climate Policy: Green Industrial Policy Builds Support for Carbon Regulation
.
Science
349
(
6253
):
1170
1171
.
Meckling
,
Jonas
, and
Llewelyn
Hughes
.
2018
.
Protecting Solar: Global Supply Chains and Business Power
.
New Political Economy
23
(
1
):
88
104
.
Mercer
.
2015
.
Investing in a Time of Climate Change
.
New York, NY
:
Mercer
.
Newell
,
Peter J.
2000
.
Climate for Change: Non-State Actors and the Global Politics of Greenhouse
.
Cambridge, UK
:
Cambridge University Press
.
Newell
,
Peter J.
2005
.
Business and International Environmental Governance: The State of the Art
. In
The Business of Global Environmental Governance
, edited by
David L.
Levy
and
Peter J.
Newell
,
21
46
.
Boston, MA
:
MIT Press
.
Newell
,
Peter J.
, and
Matthew
Paterson
.
1998
.
A Climate for Business: Global Warming, the State and Capital
.
Review of International Political Economy
5
(
4
):
679
703
.
Organisation for Economic Co-operation and Development
.
2018
.
Environmental Policy Stringency Index
.
Orsini
,
Amandine
.
2012
.
Business as a Regulatory Leader for Risk Governance? The Compact Initiative for Liability and Redress Under the Cartagena Protocol on Biosafety
.
Environmental Politics
21
(
6
):
960
979
.
Perkins
,
Richard
, and
Eric
Neumayer
.
2012
.
Does the “California Effect” Operate Across Borders? Trading- and Investing-Up in Automobile Emission Standards
.
Journal of European Public Policy
19
(
2
):
217
237
.
Primc
,
Kaja
, and
Tomaž
Čater
.
2015
.
Environmental Proactivity and Firm Performance: A Fuzzy-Set Analysis
.
Management Decision
53
(
3
):
648
667
.
Ragin
,
Charles C.
2009
.
Qualitative Comparative Analysis Using Fuzzy Sets (fsQCA)
. In
Configurational Comparative Methods: Qualitative Comparative Analysis (QCA) and Related Techniques
, edited by
Benoît
Rihoux
and
Charles C.
Ragin
,
87
122
.
Thousand Oaks, CA
:
Sage
.
Rihoux
,
Benoît
, and
Charles C.
Ragin
.
2009
.
Introduction
. In
Configurational Comparative Methods: Qualitative Comparative Analysis (QCA) and Related Techniques
, edited by
Benoît
Rihoux
and
Charles C.
Ragin
,
xvii
xxv
.
Thousand Oaks, CA
:
Sage
.
Røttereng
,
Jo-Kristian S.
2018
.
The Comparative Politics of Climate Change Mitigation Measures: Who Promotes Carbon Sinks and Why?
Global Environmental Politics
18
(
1
):
52
75
.
Schneider
,
Carsten Q.
, and
Claudius
Wagemann
.
2012
.
Set-Theoretic Methods for the Social Sciences: A Guide to Qualitative Comparative Analysis
.
Cambridge, UK
:
Cambridge University Press
.
Skjærseth
,
Jon Birger
, and
Tora
Skodvin
.
2001
.
Climate Change and the Oil Industry: Common Problems, Different Strategies
.
Global Environmental Politics
1
(
4
):
43
64
.
Sprengel
,
David Christopher
, and
Timo
Busch
.
2011
.
Stakeholder Engagement and Environmental Strategy—the Case of Climate Change
.
Business Strategy and the Environment
20
:
351
364
.
Strange
,
Susan
.
1992
.
States, Firms and Diplomacy
.
International Affairs
68
(
1
):
1
15
.
Strange
,
Susan
.
1995
.
The Defective State
.
Daedalus
124
(
2
):
55
74
.
Thauer
,
Christian
.
2014
.
Goodness Comes from Within: Intra-organizational Dynamics of Corporate Social Responsibility
.
Business and Society
53
(
4
):
483
516
.
Van der Ven
,
Hamish
.
2014
.
Socializing the C-Suite: Why Some Big-Box Retailers Are “Greener” than Others
.
Business and Politics
16
(
1
):
31
63
.
Van Halderen
,
Mignon D.
,
Mamta
Bhatt
,
Guido A. J. M.
Berens
,
Tom J.
Brown
, and
Cees B. M.
van Riel
.
2016
.
Managing Impressions in the Face of Rising Stakeholder Pressures—Examining Oil Companies’ Shifting Stances in the Climate Change Debate
.
Journal of Business Ethics
133
:
567
582
.
Vormedal
,
Irja
.
2008
.
The Influence of Business and Industry NGOs in the Negotiation of the Kyoto Mechanisms: The Case of Carbon Capture and Storage in the CDM
.
Global Environmental Politics
8
(
4
):
36
65
.
Vormedal
,
Irja
.
2010
.
States and Markets in Global Environmental Governance: The Role of Tipping Points in International Regime Formation
.
European Journal of International Relations
18
(
2
):
251
275
.
Vormedal
,
Irja
.
2011
.
From Foe to Friend? Business, the Tipping Point and U.S. Climate Politics
.
Business and Politics
13
(
3
):
1
29
.
Welsh
,
Heidi
, and
Michael
Passoff
.
2018
.
Climate Change
. In
Proxy Preview: Helping Shareholders Vote Their Values
, edited by
Heidi
Welsh
and
Michael
Passoff
,
14
29
.
Oakland, CA
:
As You Sow
.
Zürn
,
Michael
.
2013
.
Globalization and Global Governance
. In
Handbook of International Relations
, edited by
Walter
Carlsnaes
,
Thomas
Risse
, and
Beth A.
Simmons
,
401
425
.
Los Angeles, CA
:
Sage
.

Author notes

*

I express my profound thanks to Mathias Koenig-Archibugi, who gave feedback on both the initial research and various drafts of this article. I am grateful to Thomas O’Neill and Edward Collins from InfluenceMap for sharing their most recent data set and offering valuable suggestions. I also thank Robert Falkner, Jaap Spier, Henk Witte, and Sonia Salamanca, who shared valuable insights and gave feedback on early drafts of this article, and four anonymous reviewers for their constructive feedback during the review process.