Engaging firms: The global organisational field for corporate social responsibility and national varieties of capitalism

Abstract This article examines the relationship between national varieties of capitalism and firm engagement with the norms and best practices promoted within the global organisational field for corporate social responsibility (CSR). Using a content analysis of the CSR reports of US and European firms, we show that firms from the coordinated market economies (CME) of Europe engage more substantively with labour and human rights than their US counterparts that operate in a liberal market economy (LME). The environmental commitments of firms in both regions, however, are more developed than practices related to these social issues. These findings support the view that CSR is more developed in CMEs than LMEs, but limit this support to social CSR issues. We posit that firms’ higher levels of engagement with environmental CSR likely reflect the extent to which environmental norms have become embedded in global markets rather than how CSR is promoted by national capitalist systems.


Introduction
Since the 1990s, various private and public actors including the UN, the multi-stakeholder Global Reporting Initiative (GRI) and the NGO-led Forest Stewardship Council (FSC) have developed international codes of conduct in an effort to encourage firms voluntarily to improve and make transparent their environmental and social impacts. Although these codes seek to address a variety of social and environmental problems from child labour to deforestation, increasingly the authors of these schemes have drawn on a common set of norms and best practices such as 'stakeholder engagement', 'transparency' and 'continuous improvement auditing' to bring about change in corporate practice. As such, firms have had to react to what sociologists refer to as a broad organisational field for sustainable and responsible behaviour. Research over the past decade has revealed that firms' engagement with these schemes and the practices contained in them varies considerably according to their size, sector and home country (see Bernhagen & Kollman, 2013). These findings call into question the political and institutional conditions under which large firms are likely to engage with the global norms and management practices promoted within the organisational field for corporate social responsibility (CSR). www.elsevier.com/locate/polsoc Available online at www.sciencedirect.com ScienceDirect Policy and Society 35 (2016) [13][14][15][16][17][18][19][20][21][22][23][24][25][26][27] In this article we focus on the ways in which national styles of capitalism shape firms' engagement with these new CSR norms and practices. This question has gained prominence recently because of the well-known effect that home country has on firms' CSR behaviour and because scholars have developed competing explanations of how different varieties of capitalism contribute to this variation. The debate draws on the insight that a firm's behaviour is shaped by state-society relations and the nature of public authority in its home country. This implies that CSR practices will differ in so-called liberal market economies (LME) and coordinated market economies (CME). The former, which characterises the nature of capitalist relations in Anglo-American countries, relies heavily on the market to regulate key economic relationships such as those between capital and labour. In these systems, economic actors receive only limited support from the state in governing their relationships and public goods provision. By contrast, in the coordinated market economies found in north-western, continental Europe, economic relations largely are managed outside of competitive markets through consensual bargaining between capital, labour, key societal actors and the state.
Although many CSR scholars agree that these different institutional arrangements likely explain some of the crossnational differences in CSR practices, they disagree about how and why. Some scholars argue that CSR institutions and practices will be more developed in LMEs where they serve as a 'substitute' for more extensive public regulation and offer firms a source of legitimacy in countries where the state only minimally compensates those not well served by the market (Jackson & Apostolakou, 2010;Kinderman, 2012). Other scholars posit that firms in CMEs will engage more substantively with CSR because they are embedded in social networks and interact with the stakeholders who promote voluntary environmental and social standards to a greater extent than firms in LMEs (Van Tulder, Van Wijk, & Kolk, 2009).
We join these debates by employing a content analysis of the CSR reports of 40 transnational corporations (TNCs) from the US and the coordinated market economies of Western Europe to examine how firms headquartered in two different varieties of capitalism engage with global CSR norms and practices. We take a somewhat different approach to addressing this question than is typical in the literature. First, unlike many studies of CSR and private governance, we do not focus on the effects of individual codes such as the UN Global Compact (UNGC). Rather we examine the extent to which firms from two different types of capitalism engage with the broader set of corporate social responsibility norms and best practices that underpin many prominent global codes and reporting standards (Tschopp & Nastanski, 2014). The code-based approach has yielded important lessons about how private governance schemes operate and what types of firms are likely to participate in them. But we know far less about what public goods firms are willing to provide or which specific accountability practices they are willing to employ.
Second, we use the concept of engagement to measure firm behaviour rather than CSR code participation or performance as is common in the private governance literature. In addition to offering a more robust indicator of CSR participation, our engagement measure also complements CSR performance studies. The performance measures typically used in the literature such as air emissions or labour violations are important, but tend to be quite narrow and difficult to generalise from broadly. We develop a coding frame that enables us to assess firm engagement across different types of environmental and social issues without conflating them or generalising behaviour from one narrow performance indicator. Further, we analyse the statements by CEOs and senior managers published in these reports to uncover the rationale that these mangers give for undertaking CSR activities. This analysis allows us to gauge which audiences and social actors firms engage with and how important they view the influence of these actors on their CSR practices. The latter is crucial for understanding the effects that different varieties of capitalism have on firms' CSR practices, but is not always well captured by either participation or performance studies.
Our findings partially challenge both major arguments about how varieties of capitalism shape firms' engagement with CSR. First, we find that firms from the coordinated market economies of continental Europe engage more substantively with social CSR issues, such as employee welfare and human rights, than their liberal market counterparts from the US. Further, European firms cite social pressure as a key rationale for CSR engagement and highlight their interaction with external actors more often than US firms. These findings largely support the expectation that firms' CSR practices will be more developed in CME countries because of the emphasis that firms in these countries place on their relations with non-market actors. Our analysis of firms' engagement with environmental CSR, however, tells a different story. Here the variation in levels of engagement between US and European firms is far less pronounced. More importantly, firms in both regions engage more substantively with environmental than social issues. This suggests that the ability of societal actors and the state to persuade firms in CMEs to make greater labour and human rights commitments is real, but limited and contingent on CSR area. It also raises the question of why both European and US firms engage more substantively with environmental issues than other CSR areas promoted by the organisational field, something that varieties of capitalism accounts cannot explain. Our findings suggest that firms' higher levels of engagement with environmental CSR has more to do with the ways in which environmental norms have become embedded in global markets than with how CSR has been promoted by national capitalist systems of any variety.
The article proceeds as follows. In the next section, we locate firms' reporting behaviour within the global CSR organisational field, review the literature on the relationship between varieties of capitalism and private CSR governance, and describe how our empirical analysis adds to this work. In the Section 3 we outline the content analysis methods we use to investigate firms' CSR engagement. Section 4 presents the results before we offer analysis of the findings and consider their implications for future research on private governance in the Section 5.
2. The global CSR organisational field and varieties of capitalism 2.1. Corporate social responsibility and the global organisational field for sustainability Although now often associated with the private codes of conduct that promote environmental and social sustainability, corporate social responsibility continues to be a broad and contested concept. It has remained a moving target that changes over time and according to who is using it. The term itself first gained prominence in the US in the 1970s with the rise of consumer protection movements (Soule, 2009). Corporations also sought to shape the CSR concept during this era by linking it to the community and philanthropic work in which many large corporations, particularly in the US and the UK, engage (Matten & Moon, 2008). The late 1990s saw a resurgence of CSR activism; this time centred on the problems associated with economic globalisation and the perceived difficulties of regulating increasingly powerful corporations that operate across multiple jurisdictions with different regulatory regimes.
To address these problems, transnationally-linked NGOs as well as established intergovernmental organisations (IOs) such as the UN and OECD have created numerous voluntary codes that encourage firms to improve their environmental, labour and human rights practices. These actors increasingly have sought to target firms directly through such codes rather than relying solely on government regulation or binding international treaties to change corporate behaviour (Pattberg, 2005;Soule, 2009). These tactics reflect many NGOs' perceptions that it has become increasingly difficult to lobby governments to pass new regulation or to get states to coordinate such regulatory efforts through international agreements (Vogel, 2006;Kinderman, this issue). Although voluntary in nature, the CSR codes that NGOs and IOs have developed seek to regulate corporate behaviour and many, especially those sponsored by NGOs, have become sophisticated in design. The FSC's sustainable forestry label, for example, requires that an independent, third-party auditor verify a firm's compliance with the code before it can be certified as a participant.
Although the number of voluntary CSR codes has proliferated since the 1990s, these schemes increasingly are structured by common norms, notions of best practice and reporting standards (Tschopp & Nastanski, 2014). Several observers have argued that this convergence has been facilitated by the development of an organisational field of CSR standards bodies (Dingwerth & Pattberg, 2009). An organisational field is defined as ''the organisations, that in aggregate, constitute a recognised area of institutional life'' (DiMaggio & Powell, 1983: 148). Such fields are held together by interactive networks that bring its constituent members into frequent contact. Once established, they tend to lead to organisational isomorphism, or greater similarity, as increased interaction between organisations often produces common ideas of legitimacy (mimetic isomorphism) and professional norms (normative isomorphism) (DiMaggio & Powell, 1983: 150-156).
The CSR organisational field draws on the overarching sustainable development norm that encourages corporations to define success in relation to their 'triple bottom line'. According to this logic, firms should pay equal attention to their environmental, social and economic performance (Loconto & Fouilleux, 2014). The CSR standards bodies that have come into existence since the 1990s have sought to create practices that firms can follow to realise this vaguely defined goal across different environmental and social issues (Dingwerth & Pattberg, 2009). In keeping with their insider strategy, these organisations have developed structures and practices designed, at least in part, to gain the trust of participant market actors. Specifically, the standards bodies in the field are dedicated to promoting: collaborative rulemaking through multi-stakeholder structures, transparency in rulemaking and implementation, and corporate learning through procedures such as target-setting, auditing, third-party certification and reporting (Loconto & Fouilleux, 2014). Thus, the practices endorsed by the organisational field focus more on common procedures than on common performance outcomes, although some standards organisations, such as the FSC, marry these procedures with substantive performance requirements (Dingwerth & Pattberg, 2009).
These notions of best practice now structure many of the prominent CSR codes that seek to improve firms' environmental, labour and/or human rights practices including the Fairtrade logo, the Forest Stewardship Council's sustainable forestry certificate and the Global Reporting Initiative (Dingwerth & Pattberg, 2009). Even the UN's Global Compact, which is not a standards body per se, has adopted many of these practices by creating a multistakeholder body to oversee the programme and by encouraging participants to use the GRI framework to increase the credibility of firms' improvement programmes (UN Global Compact, 2015). Thus, the organisational field has sought to define CSR in terms of broad sustainability goals and a more specific set of best practices. This definition is still evolving and indeed is still contested. Older CSR activities such as community work and philanthropy, for example, remain important to many companies' CSR activities, but have not been fully incorporated in the organisational field.

Explaining firms' (uneven) engagement with the CSR organisational field
Almost two decades of research on private CSR governance has illustrated that the codes promoted by the organisational field have influenced firms' CSR behaviour, but this influence has been limited and uneven (see Bernhagen & Kollman, 2013). More than 8000 firms have joined up to the UN Global Compact, but participation is heavily skewed towards large firms that are located in Europe and to a lesser extent in the US and Japan (UN Global Compact, 2015). Numerous studies have shown that larger firms are far more likely to participate in these schemes than their small or medium-sized counterparts (Bennie, Bernhagen, & Mitchell, 2007;Prakash & Potoski, 2006). Corporations' involvement in CSR codes also is shaped heavily by the nature of their sector activity, with firms in extractive sectors, for example, frequently seeking the reputational gains that CSR participation can bring (Bennie et al., 2007;Dashwood, 2012).
Scholars similarly have sought to explain why firm participation in CSR schemes consistently varies by home country (Kollman & Prakash, 2001;Prakash & Potoski, 2006). Here the findings are less clear, but a number of factors appear relevant. Firms headquartered in democratic countries are more likely to participate than firms from nondemocratic ones (Bernhagen, Mitchell, & Thissen-Smits, 2013;Perkins & Neumayer, 2010). The extent to which a firm's home country is integrated in international markets and/or NGO networks also have proven to be a good predictor of firm participation in many global codes (Berliner & Prakash, 2012;Perkins & Neumayer, 2010). Other scholars have examined how levels of regulation in a firm's home country affect CSR activity (see Tosun et al., this issue). Prakash and Potoski (2013), for example, find that corporations operating in countries with more stringent environmental legislation are more likely to join ISO 14001, a certifiable management system.
More recently, scholars have drawn on the well-established varieties of capitalism literature to explain how home country affects firms' CSR behaviour. Much of this work utilises Peter Hall and David Soskice's framework that compares liberal market with coordinated market economies (2001). These two forms of capitalism are distinguished from one another by the degree to which they depend on market or non-market forms of coordination. In liberal market economies, like the UK and US, 'firms coordinate their activities primarily via hierarchies and competitive market arrangements', while in coordinated market economies, such as Germany, 'firms depend more heavily on non-market relationships to coordinate their relationships with other actors' (Hall & Soskice, 2001: 8). Although many scholars posit that these varieties of capitalism can, at least partially, explain cross-national differences in CSR engagement, they disagree about how and why. Daniel Kinderman has demonstrated that British firms and the business community in the UK, an LME, have used CSR as a complement to the market liberalisation that has occurred since the Thatcher era. In his account, business actors in LMEs strategically employ CSR as a substitute for government regulation, which explains why the UK has been a first-mover on CSR (Kinderman, 2012). Using firm-level data, Jackson and Apostolakou (2010) similarly show that firms from LMEs score better on almost all measures of CSR performanceboth social and environmental-than their counterparts from CMEs. They, like Kinderman, attribute this superior performance to a CSR strategy that firms in LMEs employ to justify light-touch regulation by the state.
These studies have been countered by scholars who conceive of CSR as a mirror of stakeholder relations in different countries. In these accounts firms from CMEs are posited, and often found, to have better quality CSR practices as a result of the consensual relations firms have with the societal stakeholders who promote environmental and social standards. In a recent study, Lim and Tsutsui (2012) analyse the CSR reporting behaviour of firms that participate in the UN Global Compact. They find that firms from countries with liberal economic policy regimes are more likely than their counterparts from other countries to make what they call 'ceremonial commitments' to the UNGC. Similarly, in a study based on a content analysis of corporate occupational health and safety codes, Van Tulder et al. (2009) find that European firms headquartered in CMEs have more substantive CSR commitments and measures than firms in the US or Japan. They argue that these European firms are more 'culturally' attuned to integrating the concerns of societal stakeholders into their business strategies (Van Tulder et al., 2009: 405-410).
These contradictory findings partially stem from the different methods, concepts and measures of CSR utilised in the studies. Kinderman, for example, is interested in how CSR organisations and business associations promote voluntary practices in different countries. The other studies concentrate on firm-level CSR performance, but often use different indicators that measure very different areas of CSR practice. Lim and Tsutsui, as well as Van Tulder, Van Wijk and Kolk, measure performance in relation to specific CSR codes. Jackson and Apostolakou (2010), by contrast, use measures compiled by a business analyst firm. As a result, these studies utilise different definitions of CSR with some focusing simply on labour issues and others covering social, environmental and community activities.
We seek to address these conceptual and measurement problems by taking a different approach. First, rather than studying firms' adherence to the prescriptions of a specific code or their performance according to rating agencies, we examine how firms engage with the environmental and social issues, best practices and norms promoted by the global CSR organisational field. Assessing how firms engage with these global norms and best practices allows us to gauge firms' CSR behaviour across different issue areas-environment, labour and human rights-while keeping each distinct. Studies that focus on specific codes or use established CSR indices generally either focus on a single CSR area or conflate different areas within single measures. The index used by Jackson and Apostolakou, for example, includes both philanthropy and labour issues in their definition of social CSR, despite the clear differences between these areas outlined above.
Second, we measure firm behaviour in terms of engagement with these established CSR norms and best practices rather than by performance per se as much of the literature claims to do. In this way, we avoid the concept stretching that occurs when practices such as adopting a CSR policy or publishing a report are equated with outcome performance. We also avoid the problem of inferring broader CSR performance from narrow measures such as air emissions or violations of specific labour standards. Our strategy allows us to compare firms' CSR behaviour both across countries and across different CSR areas. The latter is important because there are reasons to suspect that firms from LMEs and CMEs engage differently across these areas. The consensual, tripartite stakeholder relations between business, labour and the state fostered in CMEs, for example, should influence labour-related CSR practices, but it is less clear if this would spill over into environmental CSR. Finally, we also examine the rationale that senior managers in the reports give for their firm's CSR engagement. This gives us insight into the extent to which stakeholder relations really drive CSR practices as posited by both sets of varieties of capitalism arguments.

Data and methods
We analyse the CSR reports published by firms from the US and continental Europe to evaluate how firms from two different varieties of capitalism engage with the common sustainability norms and notions of best practice embedded in the CSR organisational field. The data are derived from a content analysis of CSR reports published by 40 firms, 20 headquartered in the US and 20 headquartered in three West European countries classified as coordinated market economies in the literature: Germany, Switzerland and the Netherlands (Hall & Gingerich, 2009). All the reports were published in either 2011 or 2012. The firms in the analysis are among the largest 50 in their region (Western Europe) or country (US) as listed in the Fortune Global 500 index (for a list of the firms see Appendix A). We analysed the most recent stand-alone CSR report for each firm that was available from the CorporateRegister database or from the firm's own website (CorporateRegister, 2013).
We chose to analyse the CSR reports of large TNCs because they are the most likely to publish such reports. This also allows us to hold firm size broadly constant. To this end, we have included firms from three West European countries for comparison with the US, as reliance on a single European country would inevitably result in a sample that is not wellmatched on size. The two sub-samples are also broadly similar with regard to industrial sector. Both contain roughly equal numbers of firms from four sectors: manufacturing, transport, finance and retail. Finally, the two samples are also comparable in terms of presence in different geographical markets (measured by primary sales location), with 60% of companies in each group serving primarily their home-regional markets. By holding these firm-level factors roughly constant, we can better evaluate the effects that different varieties of capitalism have on corporate CSR behaviour.
We analysed three common sections of each report using a self-generated coding frame based on the areas and best practices promoted by the CSR organisational field. The three sections we scrutinised are: the CEO/management statement(s), the environmental section and the social section. The latter examines a firm's commitments to labour/ employee issues and human rights. These two areas represent the issues that the CSR organisational field has promoted most aggressively, unlike CSR issues such as corporate philanthropy. We included the management letter in the analysis to assess the rationale that firms give for their CSR engagement and the extent to which this rationale is shaped by stakeholder relations as the varieties of capitalism literature posits. We identified the environment and social sections of each report by taking the naturally occurring sections so labelled by the firm itself and adding further occurrences to it using a dictionary search. All 40 firms included a management letter in their report. Similarly, a large majority had naturally occurring environmental and social sections, although some have separate sections for employee issues and human rights. The three sections included in our analysis on average represent approximately 50% of a firm's report. The other common sections that firms include are their economic performance/activities, company information and community programmes.
In the management statement we coded the following: rationale for engaging in CSR activity, reference to global CSR norms, reference to CSR codes/standards, and reference to stakeholders. We initially derived our codes for CSR rationale from the business literature and subsequently added codes that emerged from the analysis (Hartman, Rubin, & Dhanda, 2007). We coded for the following areas in the environmental and social sections: management practices promoted by the CSR field, issues for which the firm provides impact measures and issues for which the firm reports improvement goals. Finally, we assess the quality of the data that firms disclose about their impacts and improvement targets using simple indices that measure whether the firm includes quantitative data, a qualitative description and trend data for each issue area included in the social and environmental coding. The two indices-one for impact measures and one for improvement targets-are additive and weigh each quality criteria equally.
There are limitations to our approach. Although we maintain that the strength of our analysis lies in the in-depth examination of firm documents and what these reveal about firms' CSR practices, the reliability of the information contained in these documents is open to question. The reports are strategic documents used for the purpose of enhancing public relations. Although no doubt highly 'spun', we believe there is much to be learned from these documents about how firms engage with CSR and their rationale for doing so. At the very least, the rhetoric contained in the reports reflects what firms' think society and their key stakeholders want to hear. Further, as public documents signed by CEOs, and often scrutinised by NGOs, firms must realise that there is a significant possibility that they can be held to account for any obvious gulfs between their stated commitments and actual behaviour. The fact that we and others in the literature who analyse CSR reports find noticeable patterns of rhetoric and reported behaviour across time, sector and country strongly suggests that these reports represent, if imperfectly, some version of how CSR is actually conceptualised and practiced within firms (Goessling & Vocht, 2007).
Finally, we acknowledge that we are examining only one facet of corporate social responsibility as defined by the firms themselves. Most TNCs include additional areas of CSR engagement in their reports, most obviously their community and corporate giving programmes. As we have emphasised above, these community programmes are distinct from the sustainability codes that have emerged within the organisational field since the 1990s and do not represent the type of regulation conceived of as private governance in the literature. Therefore, we have excluded these areas from the analysis.

CSR engagement in the US and Europe
The results indicate that firms from the US and Europe have been influenced by common notions of best practice. All 40 firms in our analysis publish a stand-alone CSR report indicating that they have committed, at least minimally, to being transparent about their social and environmental impacts and stated obligations. Further, the CSR reports from both regions contain numerous references to key norms of the CSR field such as sustainable development and stakeholder engagement. Many of the firms in the sample also participate in prominent global CSR codes such as the UN Global Compact and the Global Reporting Initiative. These codes are both influential within the CSR organisational field and reflective of its core norms and best practices. Among our 40 firms, 80% have signed up to the GRI Guidelines and 58% are members of the UNGC, indicating at least ceremonial commitment to these codes and the practices contained in them (GRI, 2013;UN Global Compact, 2015). As the content analysis reveals, most firms from both regions flesh out these more general commitments by reporting that they implement a common set of management practices, including having a management system, auditing internal CSR performance, carrying out employee training and monitoring the social and environmental impacts of their supply chain; although both European and US firms are more likely to apply these management practices to their environmental than to their social activities (Fig. 1).
A high percentage of US and European firms further disclose information about their environmental impact and improvement goals in specific areas such as climate change, energy, resource use, pollution, product impact and compliance with environmental regulation. Ninety-five percent of firms from both regions, for example, disclose information about their greenhouse gas emissions (Fig. 2). More than half of the reports from both regions also contain information about their firm's impacts in the other five environmental areas included in the coding. Although firms in the two regions are less forthcoming when reporting on their specific targets for improvement, a majority have identifiable goals for each of the six environmental areas listed above with the exception of compliance and pollution reduction (Fig. 3).
The levels of environmental reporting are thus high among both the European and US firms. The European firms, however, tend to offer somewhat better quality information in the environmental sections of their reports as measured by the index we created based on the type and breadth of data firms disclose about their impact in the six areas examined. The average quality score of the impact data in environmental sections of the European reports was 80 out of a total of 100 points in the index. US firms, by contrast, only score 65 out of 100 points. Similarly, the quality of the improvement goals is also higher in the European reports. Here the latter scored 58 of 100 points, while the US reports  only received 44 of 100 points. 1 Despite these differences in data quality, overall the levels of environmental disclosure by firms in both regions are relatively high, which can be interpreted as a part of a wider environmental strategy, entailing long-term commitments and objectives.
Firms' reporting of their social impacts, i.e. employee issues and broader human rights, is both less precise and more influenced by home country than is the case with their environmental reporting. Few US firms flesh out their general commitment to being a good employer with a discussion of specific areas of impact (Fig. 4) or improvement (Fig. 5).
Although most US firms discuss the diversity of their workforce (90%) and claim to maintain rigorous health and safety standards (75%), only 25% mention their performance in the area of collective bargaining, and fewer than 10% mention their practices on maintaining a living wage for their employees or combating forced labour. Not one US firm reports its impact on child labour. Similarly, US firms publish very little detail about goals for improving their performance in these labour areas. Outside of these core labour areas, US firms disclose even less information about their impacts on non-employee human rights such as those related to indigenous groups or the use of security companies, despite the recent promotion of these issues by the UN and NGOs. Only 12% of US firms publish information on these human rights issues.
European firms flesh out their commitment to labour rights to a greater extent than US firms, but like their US counterparts offer less detail than in their environmental sections. In addition to reporting on the diversity and the  health and safety of their workers, the majority of European firms publish data on collective bargaining. A high percentage also reports their impact on child (55%) and forced labour (45%) as well as their procedures for providing a living wage for their employees (25%). In addition, European firms are better at reporting on their non-employee human rights impacts. Twenty-two percent disclose information about these impacts, most by declaring compliance with the UN International Bill of Human Rights. Further, European firms offer more detailed accounts of their improvement plans in these areas. This again is in contrast to US firms, who offer almost no information about their goals for improving labour standards outside of health and safety (Fig. 5).
Finally, the quality of the impact measures and improvement targets for firms' social performance is again higher in the European reports than in those published by US firms. Using the same quality index as for the environmental sections, the European firms score an average of 56 points out of a total of 100 for the quality of their social impact measures, which is considerably higher than the US firms' average score of 36 out of 100. 2 The difference in the quality of their improvement targets is similarly pronounced, with European firms scoring 34 out of 100 points and US firms only 12 of 100 points. It is notable however, that the quality of the social data reported by firms in both regions is significantly lower than the quality of the environmental data that they disclose.

Rationales for CSR engagement in US and European firms
All 40 firms in the analysis publish statements by the CEO and/or other top managers that summarise how these senior managers define their firms' social and environmental commitments as well as the key principles and rationales that underpin these commitments. As illustrated in Table 1, more than half the firms from each region (55% of European and 95% of US) indicate that their CSR engagement is motivated by pressure from market actors or a desire to be more competitive in the marketplace. In Europe the weaker market motivation is supplemented by a desire to increase firms' social legitimacy as 65% indicate that their CSR activity is influenced by social pressure. By contrast, only 10% of US firms mention social pressure. Senior managers of firms in both regions also frequently state that their firm's CSR commitments are motivated by an internal ethical commitment to a particular CSR area. A high percentage of both European and US firms, 80% and 60% respectively, claim to have an ethical commitment to environmental stewardship. Fifty percent of European and 45% of US firms also maintain that they have an ethical commitment to enhancing employee welfare. European CEOs, however, are much more likely to state they have an ethical commitment to broader human rights (25%) than their US counterparts (5%). The management statements thus indicate that firms' rationales for CSR engagement have distinct regional characteristics, although commonalties exist. CSR activities in both regions are motivated by a mixture of instrumental and ethical factors. The market appears to be a stronger rationale in the US, while social pressure seems to be more important for European firms.
These differences in CSR rationales are in keeping with the stakeholders that US and European firms highlight in their management statements. While more than half the firms in both regions highlight market actors such as customers, business partners, employees and shareholders as important stakeholders, European firms are more likely than their US counterparts to mention the non-market actors associated with concerns about 'social legitimacy' (Table 1). Thirty-five percent of European firms mention NGOs in their management statement; only 15% of US firms do so. Sixty percent of these European firms highlight government regulators as important stakeholders; in the US sample only 25% mention regulators. The most striking difference, however, is how many European and how few US firms mention intergovernmental organisations such as the UN, many of which sponsor prominent CSR codes and norms. Seventy percent of the European firms cite IOs as an important stakeholder, while only 10% of US firms do so. Similarly, while less than half of the US firms indicate that their CSR activities are influenced by a global CSR code such as the UN's Global Compact, 65% of European firms mention the UN code (Table 1). In general, with the exception of their traditional emphasis on the communities in which they operate, US firms appear to be less influenced by, and less enamoured of, non-market actors, particularly non-market political actors that come from outside of the US.
Overall, our analysis of the management statements shows that European and US firms describe their reasons for engaging in CSR activity in distinct ways. In the US, firms emphasise market factors to a greater degree than in Europe. Although US firms also highlight the contributions they make to the communities in which they exist, community generally is not defined in terms of specific societal actors nor is the goal necessarily to increase dialogue with these actors. This stands in contrast to European firms, which seek to engage with key stakeholders, especially national and international regulators, and to link their market success and social responsibilities more closely to stakeholder dialogue. The following two excerpts, the first from Daniel Akerson of GM and the second, a joint letter by senior managers at Daimler, illustrate the differences in these rationales.
We recognize that sustainability feeds our bottom line and that sustaining a profitable business is our ultimate responsibility. Profits enable reinvestment -in R&D to reimagine a car's DNA; in cleaner, more fuel-efficient technologies; in plants that better conserve resources; in improved vehicle safety; in job creation and stability; and in contributions to the communities in which we live and work (Akerson, 2012).
Honorable business people. . .want to not only gain an advantage but also serve the needs of their customers, business partners, employees, and society. . .Our stakeholders today rightly expect that our ''culture of top performance'' should not only apply to our products and technologies but also be reflected in our approach to environmental, social, and ethical responsibility. . .We are counting on you to continue this journey with us as trusted but also, of course, constructively critical partners with a focus on the future (Hohmann-Dennhardt, Zetsche, & Weber, 2012).
The main area in which US and European firms appear to converge in terms of their rationales for CSR engagement is in the environmental area. The latter is also a core focus of CSR activity for firms in both regions and is the area for which they disclose the most and highest quality information about their impacts and plans for improvement.

Discussion and conclusions
What do the results of the content analysis tell us about how different varieties of capitalism shape firms' engagement with the norms and practices of the global CSR organisational field? Our findings provide partial support for the argument that firms in the coordinated market economies of continental Europe tend to engage more substantively with voluntary CSR norms and practices than firms in liberal market economies such as the US. These findings correspond with the results reported by Van Tulder et al. (2009), who argue that firms in continental Europe are more capable of addressing CSR concerns than their US counterparts because they are more culturally attuned to taking stakeholders' demands seriously. In particular this claim is corroborated by our analysis of the social sections of the firms' reports in which European firms disclose significantly more and higher quality information about their labour and human rights commitments than their US counterparts. Indeed, outside of diversity and health and safety, the quality of reporting by US firms on these issues is so poor that it seems fair to conclude that their engagement with these issues is largely ceremonial in nature, as Lim and Tsutsui (2012) maintain (see also Kinderman, this issue).
Further, our results show that European firms' more substantive commitment to social issues appears to be related to the greater value firms in CMEs place on engaging with non-market stakeholders (See Tosun et al., this issue). It is worth noting that European firms also appear to be influenced by market incentives and actors, but to a lesser extent than their US counterparts and, crucially, not to exclusion of other societal actors. The CEO/management statements of the US firms seem to confirm that they have little interest in interacting with overtly political stakeholders or the norms these actors seek to promote. Rather, as is in keeping with firms in liberal market economies, the social practices of US TNCs seem to be largely shaped by market actors. Although US firms appear to have maintained their longstanding attachment to the communities in which they operate, these obligations do not seem to translate into commitments to specific societal stakeholders or lending legitimacy to the social pressure these actors seek to exert. Unions, government regulators and international organisations-the main promoters of social/labour standards-hardly feature in US firms' reports.
If the results of firms' social CSR practices largely support the contention that CSR management will be more developed in CMEs than in LMEs, the results from the environmental sections pose serious challenges to the broader claim that varieties of capitalism shape CSR engagement at all. Although European firms tend to provide more elaborated reporting in their environmental sections than their US counterparts, the differences here are much less pronounced. Perhaps even more importantly, firms from both types of capitalism engage more substantively with environmental norms and management practices than with social norms and practices. This holds for European as well as for US corporations. Indeed the quality of environmental reporting in the US is significantly higher than the social reporting of the European firms.
These findings suggest that the global CSR organisational field has been far more effective in promoting environmental than labour and human rights engagement. A variety of capitalism account cannot explain this finding. First, it is difficult to understand why European firms are more responsive to their stakeholders' environmental demands than social demands. CMEs are supposed to be structured to foster close relations between employees and employers. Further, it is difficult to explain why US firms engage with environmental practices at such a high level when they seem to place so little value on the actors that promote environmental sustainability. Finally, no variant of the varieties of capitalism thesis can explain why firms from LMEs and CMEs appear to be converging around the common norms and practices of environmental care embedded in the global organisational field.
Although our data do not allow us to give a definitive answer to this puzzle, they do provide strong hints about why firms in both the US and Europe engage more substantively with environmental stewardship than labour or human rights. First, there is evidence that the global CSR organisational field itself promotes environmental standards more effectively than labour standards. Dingwerth and Pattberg (2009) argue that the standards bodies that sponsor private labour codes are often reluctant to adopt many of the core principles of the organisational field such as multistakeholder governance and consensual decision-making that business actors find credible. Environmental sustainability standards bodies, by contrast, have embraced these norms. In addition, many of the standardssetting bodies have done a better job of creating specific environmental measures for firms to benchmark themselves against. The GRI reporting standard, for example, has 34 environmental, 16 labour and 12 human rights indicators that it recommends that firms report on (GRI, 2013).
Further, we suspect that environmental norms have simply become more embedded in global markets than the social norms that underpin labour and human rights. This would explain why firms from both liberal and coordinated market economies appear to engage more substantively with environmental management practices. US firms may not need pressure from non-market stakeholders to react. They may simply be responding to what they perceive markets now demand and/or are likely to demand in the future. Similarly, the greater levels of environmental engagement by European firms might suggest that markets, in addition to their non-market stakeholders, reward such behaviour to a greater extent than they reward firms' improved social performance. This interpretation is supported by the fact that market motivation is also relatively strong among European firms. As firms in both regions pay close attention to what markets and market actors demand, it is a common source of influence that could explain their more convergent environmental engagement.
Our findings also imply that environmental norms have been constructed to fit better with market motivations than the norms and practices that make up the social field. Several areas of environmental impact such as resource use, product stewardship and energy use, for example, could easily be perceived to increase firms' economic as well as environmental performance. This perception is echoed in many of the reports published by both European and US firms, which frequently refer to 'eco-efficiencies' generated by their environmental activities and the future opportunities that lie in green product and service markets. The areas highlighted in the social fields of labour and human rights such as wages, collective bargaining and monitoring of non-employee human rights are more likely to be associated with costs-reputational and material-than gains.
Thus there are good reasons to believe that norms of corporate environmentalism, at least as currently conceived, fit better with broader market structures than the norms that underpin the labour and human rights practice of the social field. This is not to argue that environmentalism is inherently more compatible with market activity or the profit motive. Forty years ago, markets did not reward firms for improvements in environmental performance. Today corporate managers appear to perceive markets as doing just that. If markets now reward improved corporate environmental performance and punish poor performers that is because the normative structures of markets have changed. The change may well be one of managers' perceptions, but these perceptions appear to have consequences and implications for the conditions under which private authority CSR regimes are able to influence firm's management practices.
These findings are based on a relatively small sample of firms from just four countries. Generalising from them should be done with care. But we believe the findings and the approach we have taken highlight at least three potentially important avenues of future research. First, our results highlight that future research within the private governance literature should pay more attention to differences across CSR issue areas in addition to studying variation across home country and CSR codes. Our results show that home country, and indeed different varieties of capitalism, shape firms' incentives to provide public goods, but these effects are contingent on issue area and in the case of varieties of capitalism, perhaps more limited than is sometimes implied by the current debate.
Second, more research is needed on what societal issues firms are willing to take responsibility for in the absence of binding mandates. We know far too little about what motivates firms to engage in CSR activities, a question that is both logically prior to issues of CSR code effectiveness and necessary to understand it. This question becomes all the more important if, as our findings indicate, private governance is found to be a more effective tool of corporate environmental than social engagement. A growing number of studies (Bartley, 2010;Dingwerth & Pattberg, 2009) hint at this conclusion. The evidence utilised in this study can only provide hints at why firms engage more substantively with environmental CSR. More systematic research focused on this question is needed to firmly establish this finding and, if substantiated, to explain why it should be the case.
Finally, not enough of the current research on private governance examines the extent to which firms' engagement with CSR is influenced by their activities in host countries. A firm's home country is likely to remain an important determinant of its behaviour given all the reasons discussed in this paper and the literature. But the extent of its international activities and the socio-political environment of the host countries in which it operates-something we hold roughly constant across our two samples-likely also shapes CSR engagement. Significant activity in host countries, for example, influences the calculus of risk that firms face and may also increase managers' exposure to the norms and best practices contained in the global CSR organisational field. More work in the future should examine these host country effects and analyse the ways in which they interact with the influence of firms' home countries.