Theoretical assumptions and framework of analysis

The argument for internal drivers of corporate social responsibility is based on transaction cost economics (Coase Reference Coase1937; North Reference North1990; Williamson Reference Williamson1975) and bargaining theory (Iklé Reference Iklé1964; Jönsson Reference Jönsson, Carlsnaes, Risse and Simmons2002; Lake and Powell Reference Lake, Powell, Lake and Powell1999). It assumes bounded rationality: actors are “intendedly rational, but only limitedly so” (Simon Reference Simon1961: xxxiv). They make purposive choices in light of the information and knowledge they have and, to the best of their ability, choose a strategy which meets their exogenously given, transitive interests. However, the information and knowledge they have, as well as their cognitive capacities, are limited. In an environment that is complex and/or uncertain, this implies that “approximation must replace exactness in reaching a decision” (Williamson Reference Williamson1975: 21). Uncertainty exists when decision problems are not deterministic. Neither a set of alternative paths, nor a rule for generating them is available. Complexity refers to the inability to ascertain the structure of the environment due to cognitive limitations. Approximation implies the possibility that actors make erroneous decisions (Ostrom Reference Ostrom and Sabatier1999: 46) and try to exploit the lack of information of others by behaving opportunistically (Williamson Reference Williamson1975: 9). Opportunism is the making of false promises and the faking or misrepresentation of information in the attempt to influence the approximation of others. The acquisition of information reduces the risk of opportunism. The costs this involves are the costs specific to a transaction or interaction. The degree of complexity and uncertainty actors confront (hence the transaction costs) depend on the level of information of the involved actors and the institutional environment. Institutions are, from this perspective, condensed information. They are safeguards against opportunism and reduce transaction costs.

Based on these assumptions, this study explains corporate social responsibility as an outcome of implicit or explicit bargaining.² Bargaining is a mode of collective decision-making to be distinguished from choices that are made by numerical aggregation such as voting procedures (see Jönsson Reference Jönsson, Carlsnaes, Risse and Simmons2002: 217; Zartman Reference Zartman1977: 621). It is also different from hierarchical steering by means of a judge who aggregates conflicting interests into a single decision. In bargaining, the parties are left to themselves to solve their conflicts of interest and reach a single decision. Bargaining situations are characterized by three elements: interdependence, common interest, and conflict (see Iklé Reference Iklé1964: 2; Jönsson Reference Jönsson, Carlsnaes, Risse and Simmons2002: 218). Interdependence implies that the outcomes resulting from a given set of alternative actions depend on how others choose to behave.³ An interdependently structured situation is therefore similar to a game in which each actor must take into account the choices of others when assessing his or her choice. Common interests are present when there is a need for cooperation, that is, when mutuality leads to superior outcomes than unilateralism. For example, a transaction is in the common interest of a fruit trader and a consumer intending to buy fruit. Conflict refers to diverging preferences over outcomes. The fruit trader may prefer to sell fruit at a high price, while the customer’s preference is that of a low price. Processes of implicit or explicit bargaining set out to find a solution to the conflict, which is suboptimal for at least one of the parties – but still within the limits of the common interest. For example, if both parties to the transaction agreed on a medium price, this would constitute a suboptimal outcome for both, but nonetheless considered preferable to no transaction at all.

Bargaining is a mode of joint decision-making that can be found in different institutional settings. In (neo-)corporatist arrangements, bargaining is formally institutionalized within tripartite forums (Schmitter and Lehmbruch Reference Schmitter and Lehmbruch1979), sometimes even officially named “bargaining councils,” as in South Africa (Müller-Debus et al. Reference Müller-Debus, Thauer and Börzel2009a). Bargaining also takes place in other institutional contexts, but often informally. In hierarchical organizations, management is often confronted with “managerial dilemmas” (Miller Reference Miller1992) that necessitate joint decision-making through bargaining, as will be discussed in detail below. In a market context, bargaining over price takes place when market participants do not know the exact match of supply and demand and the costs of production. Also in hybrid forms of social order located between markets and hierarchies, as in a semi-integrated supply chain, bargaining is a mode of decision-making that actors rely on. The relation between a buyer and a supplier may be formalized by a contract. Conflicts over issues that are contractually regulated are therefore subject to hierarchical decision-making by means of a judge (i.e. court action). However, inevitably incomplete contracts, the possibility of unforeseen events, opportunism, and the costs of litigation often necessitate additional bargaining over governance (Héritier et al. Reference Héritier, Müller-Debus and Thauer2009). Finally, in institutional contexts that prescribe decision-making through numerical aggregation and voting procedures, actors draw on bargaining when, for example, parliaments or executive councils have come to a deadlock regarding important decisions. “Issue linkage” or the suggestion of “package deals” are bargaining attempts that can effectively render voting behavior.

Potential outcomes of bargaining processes are threefold. First, bargaining can fail to reach an agreement. Bargaining failure implies a collectively suboptimal outcome.⁴ Second and third, bargaining can be successful, but can result in either asymmetric payoffs or in a true compromise. Either outcome depends on the bargaining position and power of the actors. Bargaining positions and power, as well as the success and failure of bargaining processes, depend on the interests and level of information actors have and the strategic environment, i.e. the set of available alternative actions, the way events may unfold, the institutional context, and the information environment (see Lake and Powell Reference Lake, Powell, Lake and Powell1999). Power, from a bargaining perspective, is thus the result of a complex social setting and not an absolute measure as, for example, “capabilities” in realism (Waltz Reference Waltz1954).⁵

In processes of bargaining, actors optimize their strategy on the understanding that others are equally bounded rational actors. Bargaining processes can be implicit and explicit. Explicit bargaining involves verbal communication and other signals informing the opponent of one’s own preferences and bargaining position. Explicit bargaining increases the level of information on which an agreement is reached; this can prevent bargaining failure or a collectively suboptimal outcome. At the same time, opportunism is all-pervasive (Williamson Reference Williamson1975). However, opportunism can also be costly. Demanding too much can lead to bargaining failure (Miller Reference Miller1992: 48). Lying can damage the reputation of a player when detected and thus may weaken his future bargaining position. Bargaining can also take on the form of non-verbal signals, such as “tit for tat” (Axelrod Reference Axelrod1981) – or be entirely implicit. Non-communicative, single-shot game theory formalizes implicit bargaining processes. Thereby the “Nash equilibrium” is found “by calculating each player’s best replies to each of the strategy combinations that might be played by others, and identifying those strategy combinations that are best replies to each other” (Hargreaves Heap and Hollis Reference Hargreaves Heap and Hollis1998: 101). This process of mutual adaptation, explicit or implicit, is the causal mechanism unfolded by bargaining situations (see Miller Reference Miller1992: 48).

Asset specificity: concept and theory

Asset-specificity theory is a branch of transaction cost economics developed by Oliver Williamson and his followers.⁶ The central concept of the theory, “asset specificity,” refers to non-transferable investments in an exchange relationship. If market demand for goods, services, or know-how created by an investment in support of a transaction is absent, it cannot be transferred to another transaction and is therefore unique to the task (Williamson Reference Williamson2002: 175). However, non-transferability is rarely absolute. Another way to describe asset specificity is therefore that the investments made to support a particular transaction have a higher value in relation to that transaction than they would if they were redeployed for another purpose (McGuinness Reference McGuinness, Thompson, Frances and Mitchell1994). Asset specificity can take on various forms. For example, it can be in human skills (“human asset specificity”), specialized machine tools (“physical asset specificity”) or natural resources that are linked to a certain location (“site specificity”).⁷

Asset specificity can be one-sided or two-sided (Joskow Reference Joskow1988; Klein Reference Klein1988; Williamson Reference Williamson1975). One-sided asset specificity implies that only one party to a transaction makes an investment so that it can produce something that is otherwise not attainable on the market or only hardly so. An example in this respect is a supplier’s investment in tooling instruments specifically for the making of a component that one particular customer has ordered. Two-sided asset specificity refers to non-transferable investments made by both parties to the transaction, such as when buyer and supplier share the purchasing costs for the new tooling instruments (examples in Héritier et al. Reference Héritier, Müller-Debus and Thauer2009).

Asset specificity transforms the bargaining situation between transaction partners. This transformation is “fundamental” (Williamson Reference Williamson1996: 16). The theory assumes an initial “large-n” situation. Large numbers of buyers (who demand goods) and a large number of providers (who supply the demand) transact frequently and interchangeably. Immediate substitutability of all exchanges renders potential information asymmetries between the transaction partners and unforeseen events unimportant on account of its disciplining effect on the parties. Hence, information deficits are absent and the level of certainty is high.

Non-transferable investments change this. Buyers and suppliers “can redeploy the specialized assets to their next best use … only at a loss of productive value” (Williamson Reference Williamson2002: 176). Transaction partners become dependent on each other. A situation emerges in which small numbers of buyers confront a small number of providers. In the extreme case, buyer and supplier enter an exclusive exchange relation. Both are vulnerable in this situation, as the disciplining effect of the market is absent and because exchange contracts are inevitably incomplete. Unforeseen events have to be dealt with jointly, thereby information asymmetries persist and bear the risk of opportunism (Williamson Reference Williamson1975: 26). Both fear being cheated.⁸ A collectively suboptimal outcome could derive from the exchange. Consequently, both parties either refrain from asset-specific transactions altogether or agree on devices that help mitigate risks, discourage opportunism, and reduce uncertainty and complexity. “Governance” – institutional provisions sidelining the exchange contract – is the means by which the bargaining partners can achieve this. Asset-specificity theory therefore posits that, in bargaining situations characterized by non-transferable investments, a governance order will be the outcome of bargaining processes between the transacting parties. In extreme cases, vertical integration is predicted.

Governance rules define, for example, under which conditions the exchange contract is fulfilled. They may also provide transaction partners with positive incentives for the fulfillment of the contract or impose penalties for premature termination. In addition, governance can stipulate mechanisms for information disclosure and verification as well as sanctioning mechanisms, how disputes are to be settled and what procedures are to be followed in the likely case of unforeseen events. A fundamental distinction with respect to the content of governance rules pertains to “product” versus “process regulation” (Vogel Reference Vogel1995, Reference Vogel2005).

Product regulation defines the physical attributes of the goods under exchange, such as quantity and size as well as quality-related aspects (see Vogel Reference Vogel1995: 18). Product regulation can be utilized for a number of purposes. First, in exchange relations, product regulation helps to establish whether an exchange contract is fulfilled or breached. Second, through high product standards, firms can keep foreign competitors operating on lower standards out of the market – if they succeed in lobbying government to raise the legal standards that regulate the product under exchange to their level (Börzel et al. Reference Börzel, Héritier, Kranz, Thauer and Risse2011; Greenstein and Stango Reference Greenstein and Stango2007). This strategy option exists, because product standards are exempt under the General Agreement on Tariffs and Trade (GATT)/World Trade Organization (WTO) regime. Strong regulatory states or markets, such as the European Union (EU) or the US, can therefore make use of high product standards, in effect turning them into a hidden form of import barrier. Most firm activities in the realm of product standards are consequently oriented towards anticipated or actual legislation in order to raise the legal standards and to keep external competitors out. Third, enhancing product quality standards allows firms to target a higher end of the market, promising higher margins. For these three reasons product regulation often follows “race to the top” dynamics, i.e. competition for higher standards (Abbott and Snidal Reference Abbott, Snidal, Mattli and Woods2009; Börzel and Thauer Reference Börzel and Thauer2013; Vogel Reference Vogel1995, Reference Vogel2005).

This study concentrates exclusively on process regulation, however.⁹ Process regulation formulates requirements for the production of goods. Management systems concerned with corporate social responsibility are just one example among many standards – such as minimum wage, energy consumption reduction in the production process, or recycling. Usually, process regulation is considered to be more prone to “race to the bottom” dynamics – i.e. regulatory downsizing for competitive cost advantages (Vogel Reference Vogel1995: 18). The reasons for this are twofold: first, process regulation cannot be used as an import barrier under the WTO/GATT regime and is thus not a playing field for anti-competitive strategies of firms – i.e. for keeping competitors out of the market. Second, only in rare cases does process regulation add measureable value to the goods that are being produced in the eyes of end consumers. In fact, process regulation in most cases just increases production costs and has no direct marketable value. Hence, process standards are the unlikely case of corporate social responsibility.

However, why would transacting parties insist on process standards if their transactions involve asset-specific investments? It will be argued here that process regulation reduces uncertainty and information deficits and discourages defection and opportunism in the critical period between contract agreement and contract execution. Also, in the case of technologically complex products, the verification of product regulation can be very difficult and costly. It is for this reason that transaction partners will rely on process regulation. Therefore, it will be argued that, wherever hazards have to be mitigated as a result of asset specificity, process regulation occurs.

Intra-organizational asset specificity, managerial dilemmas, and corporate social responsibility

The “new” turn this study takes now is to apply asset specificity intra-organizationally, and to relate it to the emergence of corporate social responsibility standards. It is necessary to address certain assumptions about the nature of firms before the application of asset specificity to the intra-organizational setting. The following arguments for internal drivers of corporate social responsibility assume a context of imperfect vertical integration: central management is in control over the flow of organizational resources (see Demsetz Reference Demsetz, Williamson and Winter1991: 161–2) and therefore in a “take it or leave it” (Sappington Reference Sappington1991: 47) position vis-à-vis organizational subunits and individual subordinates. Analogous to markets, however, hierarchies are imperfect in that the information on the basis of which actors make decisions is incomplete (Hart Reference Hart, Wilson, Grant and Coen2010). “Hidden information” (Miller Reference Miller1992: 138) and expert knowledge create information asymmetries. Consequently, decision-making in hierarchies is confronted with similar problems as in a market context. Opportunism, shirking, and “hidden action” (Miller Reference Miller1992: 120) are pervasive. Hierarchy failure is a possible outcome as organizational steering is a constant challenge. Management deals with it via intra-organizational bargaining, negotiation, and coalition politics.¹⁰

So when management assigns a critical mass of organizational resources to an organizational subunit or to subordinates in order to support the fulfillment of a task, but subsequently fails to transfer this investment without a loss, the decision situation is transformed. Management becomes vulnerable and dependent on the subunit and loses its strong, authoritative “take it or leave it” position. Unforeseen events can no longer be dealt with through a reallocation of resources; they require a joint reaction. Opportunism and shirking become threats, while the disciplining effect of hierarchy is weakened. Central management thus finds itself in a situation in which it will have to bargain with the respective subunit over an intra-organizational governance order, sidelining asset-specific allocations of resources to safeguard its authority. This study treats situations in which the mode of intra-organizational social coordination is transformed from a hierarchy to one in which management becomes dependent on and vulnerable to the behavior of subordinates, as “managerial dilemmas” (Miller Reference Miller1992).

Corporate social responsibility is key to resolving managerial dilemmas in two ways: first, it reduces information asymmetries. In situations of managerial dilemmas, corporate social responsibility is implemented to structure, document, and monitor work processes, identify areas of improvement, provide for performance measures, facilitate resource use efficiency, and define escalation procedures. Hence, a valuable side effect of the adoption of standards defining corporate social responsibility is that, in these situations, it enables management to collect information about what subordinates do and how processes can be improved.

Second, corporate social responsibility reduces uncertainty. Since the 1990s, environmental, social, health, and labor standards proliferate on the global level, addressing not only states but also firms. On the local level, however, and, by assumption, in a context of limited statehood, these standards are not enforced. This discrepancy between the global and the local is a source of uncertainty for companies. Firms that invest in an area of regulatory void in an asset-specific way worry about whether their investment decision will still be tenable in the future. They ask themselves, for instance, whether labor costs and energy prices will increase, or whether the firm will be forced to reintegrate negative externalities of production (caused, for example, by NGO campaigns or increasing prices due to resource shortages). Yet firms that from the very start adhere to the highest international standards in the asset-specific allocation of resources reduce this kind of vulnerability to changes in the environment: they have already prepared for such changes. Therefore, a management that has an asset-specific relationship with one or more subunits and so cannot close down a unit easily in response to a shift in its strategic environment will insist on implementing corporate social responsibility standards and policies.

It must be noted that these arguments, while they take into account the existence of external drivers for corporate social responsibility, are essentially about internal drivers. More precisely, the argument developed here is that two firms, exposed to the same external pressure factors and risk environment, will nonetheless show very different levels of corporate social responsibility in their behavior on account of the distinct internal managerial dilemmas they face.

Internal driver 1: the human resources dilemma

The first particular dilemma situation analyzed is the human resources dilemma. It refers to ongoing investments in skills of employees that are hard to attain on the labor market. Such investments create asset specificity in the employment relation in the sense that the required skills level can be achieved only subject to these investments.¹¹ In comparison, ongoing investments in skills that are openly available on the labor market are not asset specific, or only less so. These investments could be easily reassigned to the task of headhunting for new employees on the labor market who can already boast of these required skills. Hence, in this case, management has strategy options for the procurement of the requisite skills level of production and thus remains in a “take it or leave it” position vis-à-vis employees.

The goal, asset-specific investments in skills – skills that are not attainable from the labor market or only with difficulty – pursues a higher productivity and/or specific production techniques for specialized niche-market products. Since management is in a “take it or leave it” position, employees can be forced to attend training courses that teach such skills. However, once employees have attained the required specific skills, the bargaining situation is transformed. For management, “turnover is costly, since a similarly qualified but inexperienced employee would have to acquire the requisite task-specific skills before he would reach a level of productivity equivalent to that of an incumbent” (Williamson Reference Williamson1975: 59). Hence, management cannot easily “hire and fire” employees anymore. The mode of social coordination has changed from a hierarchy to a situation in which management becomes dependent on employees, and vulnerable to their behavior. The dilemma situation emerges. More precisely, asset-specific investments in skills create expert knowledge and information asymmetries, in addition to related problems of opportunism, cheating, and shirking. Also, unforeseen events in the environment may affect the productivity of the investment, yet lie beyond the direct control of management and thus necessitate a joint response. Social conflict, disease, and drug abuse are examples of general societal developments with a potential negative impact on the productive value of the investment in skills. Hence, management has to find a way to cope with the uncertainty that persists with respect to the future productivity of the resource allocation and has to find a way to reduce emerging information asymmetries. This is how dilemma situations turn into a driver. They put management under pressure to act and find ways to deal with them.

This study argues that, in this specific situation, labor-related corporate social responsibility can resolve the dilemma and will, therefore, be the preferred strategy option of management. Labor-related corporate social responsibility can entail services which are so beneficial to employees as to discourage turnover, absenteeism, and low productivity. In addition, they also mitigate problems in the investment environment that could permeate the workforce and affect its productivity.¹² Corporate social responsibility may thus envisage premiums for workers upon satisfactory fulfillment of the tasks in question. Also, special pension schemes and insurances, extensive health care services, and services promoting physical wellbeing, regeneration, and a healthy lifestyle, can remedy potential threats in the environment that risk affecting the firm. In substance, the labor-related corporate social responsibility that emerges in situations of asset-specific allocation of resources depends largely on the features of the investment environment. More specifically, it depends on where public service provision and regulation are lacking, and needed. Thus, the question is: in which areas does management confront uncertainties and complexity as a consequence of limited statehood and thus engage in corporate social responsibility with the aim of reducing uncertainties and information asymmetries that may negatively affect the asset-specific investment?

In a context in which labor conditions are entirely unregulated or where the implementation of existing regulation is absent, the logically derived outcome, as argued here, is that firms will issue employees with steady and formal contracts, offer them insurance packages, and guarantee the payment of a minimum wage for limited working hours. If social problems, such as drug abuse, are widespread in the investment environment and impinge on the firm, the human resources dilemma will motivate management to set up and run anti-drug campaigns. The empirical assessment conducted further below analyzes businesses in South Africa in terms of their management’s behavior regarding a major problem in the investment environment, HIV/AIDS, and the huge uncertainties and complexities this illness entails. Firms that confront the human resources dilemma, in this context, will set up HIV/AIDS workplace programs.

While it can be assumed that employees have a strong interest in such labor-related services, the mentioned governance provisions allow management to safeguard its power position during and after the allocation of organizational resources. However, in contrast to employees, such governance provisions are costly for management. Hence, it is in the management’s interest to offer only the bare minimum in services necessary to maintain its authority, while employees will drive for a maximum in services. The extent of services resulting from the process of explicit (for example, in labor councils between representatives of management and of employees) and implicit (finding the best reply to the opponent’s strategy) bargaining will therefore critically depend on the level of asset specificity created through the investment.¹³ The more management becomes dependent on employees and vulnerable to their behavior, the greater will the uncertainty in the environment be perceived. The same is true for information asymmetries between management and employees. Hence, the broader and more sophisticated will be the labor-related corporate social responsibility services which management agrees to offer employees.

In summary, with respect to internal driver 1, I argue that the human resources dilemma causes the engagement of firms in health, social, and other labor-related corporate social responsibility.¹⁴

Internal driver 2: the technological specialization dilemma

The second internal driver for corporate social responsibility is the technological specialization dilemma. It elaborates in detail a factor that has already been considered in literature, though mainly as a control variable in econometric analyses of corporate social responsibility, which is the “level of technological advancement” of a firm (for example, Khanna et al. Reference Khanna, Koss, Jones and Ervin2007). Technologically complex products often require a considerable amount of pre-commercial planning and allocation of resources. Once research and development (R&D) has a marketable product ready, central management allocates resources to start up production. It is in this phase that the technological specializationdilemma occurs on account of production-specific assets, i.e. intra-organizational asset-specific investments in the production unit and technology. Production-specific assets refer to the commitment that is made by management to support the task delegated to a subunit to start up and run production. The longer the period before envisioned return of investment and the more resources dedicated to the production unit, the greater is management’s commitment. The production-specificity of the company’s assets increases accordingly – in line with the ever growing managerial dilemma.

In comparison, if the period before envisioned return of investment is short, production-specific assets are absent.¹⁵ In this case, if management is dissatisfied with the output of production, it can reallocate resources swiftly. The unit is closed or restructured. A new production site is opened or production is shifted to another site. Thus, management remains in a “take it or leave it” position vis-à-vis the subunit.

If a long period before return on investments is envisioned – that is, if production-specific assets are created – the situation changes and turns into a managerial dilemma. Management becomes dependent on and vulnerable to the behavior of the production unit. It will have to deal with unforeseen events that have an impact on production costs. In addition, the technological specialization dilemma implies the creation of expert knowledge about processes, specialized machinery, and production techniques within the production unit. Information asymmetries are a consequence and bear the risk of opportunism. Hence, management fears that the subunit will not fulfill the task as envisioned. Return of investment and the making of profits are at risk.

Standard economic instruments of organizational management, such as the creation of an internal price mechanism, according to which management incentivizes the subunit on a certain price per unit of the production output, does not provide for a way out of the dilemma. The technological specialization dilemma occurs in the period before production begins, the period during which the production site is being constructed. In this period, instruments of organizational steering that are related to the production output, such as a price per unit, are obviously not applicable.

On some occasions an incentivization on a price per unit is possible, but bears risks and may cause reverse effects. Where the technological specialization dilemma persists throughout long periods of production or over a whole production cycle – which is often the case in high technology production – price per unit incentives may even create a “moral hazard.”¹⁶ Management incentivizes the subunit on a low price per unit to inspire efficient production processes. However, price per unit is a rather short-term efficiency measure. Being fully aware that management will bail out the subunit in the event of a crisis in the future, this unit may take high risks and use up assets and resources to meet management’s expectation of a low price per unit in the short run, but may thereby undermine its long-term efficiency and diminish the profitability over the whole production cycle. Insisting on the delivery of a certain price per unit in an early stage of the production cycle alone is therefore not a strategy by which management can overcome the technological specialization dilemma.¹⁷ An additional governance order sidelining the asset-specific investment is instead necessary to reduce information asymmetries, uncertainties, and the risk of “moral hazard” in the relationship between management and the production unit.

Therefore, in order to reduce uncertainty and information asymmetries and to mitigate “moral hazard,” a management that confronts the technological specialization dilemma will insist on strict production process standards. Such standards will structure, document, and monitor work processes, identify areas of improvement, provide for performance measures, facilitate resource use efficiency, and define escalation procedures. Hence, management will insist on the implementation of standards that facilitate information about what subordinates do and about how processes can be improved. Such standards, which will often take on the form of management systems, can be quality-oriented, such as in case of ISO 9001 quality management systems; not every process standard is thus an instance of corporate social responsibility.¹⁸ However, in a context of regulatory void, and depending on the degree to which allocated resources are asset-specific, management will, in addition, insist on standards that extend to environmental aspects, as this guarantees resource-use efficiency and mitigates potential legal and reputational risks. Examples of such process-oriented environmental corporate social responsibility are ISO 14001 environmental management systems (Potoski and Prakash Reference Prakash and Potoski2006), VDA 6.1, or TS 16949 integrated quality and environmental management standards in the automotive industry.

In the upcoming bargaining situation, it is common knowledge that management’s authority vis-à-vis the subunit is diminished once resources have been allocated. Anticipating this transformation, the representatives of the unit will behave opportunistically and promise to fulfill the task of running production efficiently as envisioned by management, knowing that they can renegotiate the terms and conditions of the agreement from a better bargaining position after production-specific assets have been created – or simply pursue their own interests ignoring central management. Central management, however, anticipates this loss of authority as well. It will therefore insist on negotiations with the subunit that go beyond agreements based on “mere promise” (Williamson Reference Williamson, Foss and Mahnke2000: 601) and press for the installment of governance rules as a backup to the commitment through corporate social responsibility standards with strict information disclosure, monitoring, and sanctioning mechanisms. For the representatives of the subunit this is a suboptimal outcome, but since central management can still negotiate from a “take it or leave it” position it can impose corporate social responsibility unequivocally at this point.

This study therefore argues with respect to internal driver 2 that the technological specialization dilemma causes production process-oriented (environmental) corporate social responsibility.

Internal driver 3: the foreign direct investment dilemma

The third internal driver is the foreign direct investment dilemma. This is a specific instance of internal driver number 2, the technological specialization dilemma. More specifically, it involves central management, located in a highly regulating country, making a substantial investment with a long duration before returns are generated in a branch located in a foreign country with weak or limited regulatory capacities. The consequences of such investments are thus analogous to the ones that give rise to the technological specialization dilemma. The difference is, however, that the foreign direct investment dilemma also features a trans-border element in the relation between management and the subunit in which it invests, as well as diverging regulatory contexts, which make the dilemma appear more severe. These elements increase uncertainty and information asymmetries.

Managers who confront the foreign direct investment dilemma will consequently ask themselves whether the branch really organizes processes efficiently or whether it exploits the emergent information asymmetries and acts opportunistically. The question is whether the branch will make use of assets and resources, or of lax environmental, social, and health regulation in order to boost its short-term revenues and thereby risk undermining the long-term profitability and efficiency of its operations knowing that headquarters will come to the rescue in the event of a crisis (“moral hazard” problem). Analogous to the technological specialization dilemma, in situations characterized by the foreign direct investment dilemma, headquarters’ management will insist on strict process standards to reduce information asymmetries and uncertainty with respect to changes in the environment of the branch and the mitigation of “moral hazard.” In a context of limited statehood, just as in the case of the technological specialization dilemma, depending on the severity of the dilemma, management will insist on standards that extend from quality to environmental aspects as well, as this guarantees resource-use efficiency and mitigates potential legal and reputational risks.

However, unlike the technological specialization dilemma, management will, in its effort to hold the branch accountable to the highest attainable standards, insist on a transfer of standards from “home” to its operations “abroad” in the case of a strong foreign direct investment dilemma. Hence, the foreign direct investment dilemma is a driver of the diffusion of regulatory standards from highly regulating to weakly regulating countries. This is the main difference between the foreign direct investment dilemma and the technological specialization dilemma in terms of the impact they have on corporate social responsibility. As it features a cross-border element and a management located in a country with particularly strict and concise regulation, the foreign direct investment dilemma is a source of “race to the top” (Börzel and Thauer Reference Börzel and Thauer2013; Vogel and Kagan Reference Vogel and Kagan2004) dynamics, i.e. the convergence of international standards at the highest level. This is especially so when competitors confront similar foreign direct investment dilemmas. The efficiency and long-term profitability of the branches abroad then become a decisive competitive advantage.¹⁹

In this respect, the foreign direct investment dilemma links up with two arguments in literature, which both concern or derive from diffusion theory and research (DiMaggio and Powell Reference DiMaggio and Powell1991; Shipan and Volden Reference Shipan and Volden2008; Simmons et al. Reference Simmons, Dobbin and Garrett2006). First, it specifies the “home country” hypothesis and closes a gap between quantitative and qualitative approaches. According to this hypothesis, firms originating from a highly regulating country will transfer the high standards from “home” to their operations in weakly regulating countries “abroad” (Greenhill et al. Reference Greenhill, Mosley and Prakash2010; Prakash and Potoski Reference Prakash and Potoski2007). While this claim finds confirmation in some econometric studies, qualitative analyses observe that subsidies of such firms often do not transfer any standards (Börzel et al. Reference Börzel, Héritier, Kranz, Thauer and Risse2011; Héritier et al. Reference Héritier, Müller-Debus and Thauer2009). This study specifies the hypothesis, thereby extending its validity to qualitative approaches. More specifically, it will be argued that while the hypothesis holds, it does so only under the condition that the operations abroad constitute an asset-specific investment for management at “home.” It is therefore the foreign direct investment dilemma, and the degree to which it is present, that drives the transfer of standards from “home” to operations “abroad.”

Second, it adds a “vertical” perspective to the analysis of diffusion dynamics, which so far have been analyzed from a strictly “horizontal” perspective. By a horizontal perspective, diffusion theorists understand a focus on interaction effects in large populations of cases: “what theorists of diffusion explicitly reject is the notion that processes of policy change can adequately be understood by conceiving of [actors] as making decisions independently of each other” (Simmons et al. Reference Simmons, Dobbin and Garrett2006: 787). Accordingly, the phenomenon of diffusion refers to “any process where prior adoption of a trait or practice in a population alters the probability of adoption for remaining non-adopters” (Strang Reference Strang1991: 325). The argument on the foreign direct investment dilemma contributes to the analysis of the diffusion of business standards as it defines (vertical) organizational characteristics as a precondition for (horizontal) diffusion effects. More precisely, it helps to define how and why an individual firm may develop the idea to transfer “home” standards “abroad” in the first place, which is a necessary condition for diffusion dynamics to take place at a later stage. It may also contribute to the analysis of diffusion dynamics themselves as it helps us to identify which firms of an industry sector will be affected by diffusion dynamics, and when. The general argument, with respect to the former question, is that firms are affected by diffusion dynamics on account of the degree to which they confront the foreign direct investment dilemma. With respect to the latter, in the very moment a firm receives investments with long durations before returns are generated, it is affected by diffusion dynamics. Taken together, both arguments may help future research to identify which industry sectors are more and less prone to the diffusion of business standards. They may also contribute to a better understanding of the evolution and specific pathways of standard diffusion over time.

Note that while this argument relates to external drivers, it is essentially about internal drivers. The argument here is that two firms of an industry sector experiencing diffusion dynamics will show a different reaction to and participation in the diffusion process. This is due to the variant foreign direct investment dilemmas they are confronted with. Likewise, a single firm of an industry sector in an area of regulatory void may remain unaffected by diffusion dynamics up to the point it receives long-term investments.

To summarize, regarding the third internal driver of corporate social responsibility, this book argues that the foreign direct investment dilemma leads to a transfer of high standards if the branch in the weak state “abroad” is a specific asset to management “at home” in the highly regulating country.

Internal driver 4: the brand reputation dilemma

The fourth internal driver is the brand reputation dilemma. The argument with respect to this dilemma bears strong similarities to a hypothesis in literature which maintains that branded firms will engage in corporate social responsibility for reputational reasons (Börzel et al. Reference Börzel, Hönke, Thauer, Börzel and Thauer2013; Haufler Reference Haufler2001; Marx Reference Marx2008; Mol Reference Mol2001). The brand reputation dilemma emerges as a consequence of asset-specific investments in marketing to support alternative means of product differentiation to the price mechanism. Price-driven firms rely – in their interaction with consumers – on anonymous market forces.²⁰ Intra-organizationally, this strategy has no repercussions except that it implies decision-making by means of hierarchical steering and economic instruments such as an internal market, which gives subunits incentives to produce at a low price per unit. Short-term cost cutting is the strategy, production-specific assets are minimized and negative externalities are maximized while management remains in a “take it or leave it” position throughout the production process. However, once management decides to offer alternative means of product differentiation to consumers in order to support their decision to buy, intra-organizational repercussions will occur. Two investments in this respect can be differentiated.

First, resources can be allocated by management to support a differentiation of products according to their degree of sophistication, technical complexity and other quality-related features (Anton et al. Reference Anton, Deltas and Khanna2004; Parker Reference Parker2002). Such investments are made to target a high-end niche market in which consumers are willing to pay a premium for quality.²¹ Management becomes thereby dependent and vulnerable in two ways: on the one hand, the allocation creates the technological specialization dilemma and therefore renders management dependent on the production unit; on the other hand, production-specific assets also imply one-sided asset specificity in relation to consumers.

More precisely, production-specific assets are investments made by management that lose their productive value when redeployed to another consumer segment, in which no premiums are paid (i.e. the mass segment). Hence, management becomes dependent on the high-end market segment. Asset specificity is one-sided in this relation, as consumers do not make any non-transferable investments when buying the products from a high-end market firm. Consumers do not commit themselves to purchasing a particular good from a specific brand or seller. They do not even commit to purchasing a certain type of product from the high-end market segment, since they can always choose the option to buy a cheaper product from a mass segment. Hence, should the firm not offer the expected price–quality ratio consumers will opt for another seller. These considerations reinforce the need for management to impose strict corporate social responsibility rules on the production unit as a means of cost and quality control. The difference to the argument for the second internal driver, the technological specialization dilemma, is that this fourth driver goes beyond the explanation of intra-firm corporate social responsibility. It applies to the exchange between organizationally different buyers and suppliers as well. Consumers hold the brand, from which they purchase a product, responsible for its quality and the way it was produced. Even if much of the value added to a product was actually generated within a supply chain, they hold the brand responsible for what they have purchased nonetheless.

Second, management can allocate resources to the marketing unit for the creation of a brand name, which can be a powerful product differentiation mechanism.²² Such an investment in the reputation of a firm is not necessarily competing, but rather complementary to a high-end market strategy. Strong branding can go hand in hand with a strong quality orientation. Indeed, branding can provide the end-consumer with information about the superior quality of a product. However, branding can also add value to products that have no material differences. To be more precise, investments in marketing can support the creation of a social status-value of the brand. As in the case of a high-end market orientation, an intra-organizational allocation of resources to support the creation of a brand name creates one-sided asset specificity. The resources lose their productive value if they are transferred to a mass or low-end consumer segment. Intra-organizationally, the asset “brand name” is also not redeployable, for example to support the creation of a more sophisticated production site, to upgrade R&D, or the level of skills of employees. High-end market consumers, however, are not dependent on the firm. They do not invest in their buying decision and are therefore free to choose between the full range of sellers and between different market segments. Just as investments in product quality, investments in a brand name have intra-organizational repercussions. Public outcry over bad business practices devalues the investments made to support a positive image. To safeguard its investment, management will therefore impose strict corporate social responsibility policies on the production unit and on its suppliers, for which firms are held responsible.²³

In summary, the fourth internal driver of corporate social responsibility is the brand reputation dilemma, which causes the engagement of firms in in-house and supply-chain corporate social responsibility.

Scope conditions

The empirical analysis in the subsequent chapters will show that these arguments for internal drivers of corporate social responsibility are widely applicable and gain explanatory power in highly diverse contexts. However, three scope conditions apply. The first scope condition pertains to the institutional-political context. The arguments in this book apply to contexts of regulatory void, so-called areas of “limited statehood” (Krasner and Risse Reference Krasner and Risse2014; Risse Reference Risse2011) – environments where the state is unwilling or incapable of setting high standards, providing essential services, or enforcing legal obligations. Everything else being equal, variation of internal drivers should, under these conditions, lead to variant levels of corporate social responsibility. In areas of consolidated statehood, by contrast, the theory of internal drivers does not apply in the same way. As the literature on the “varieties of capitalism” points out, consolidated statehood is explained historically, at least to a certain degree, by the emergence of regulation that mitigates the risks of predominant patterns of asset-specific investments in an economy (Busemeyer Reference Busemeyer2009; Hall and Soskice Reference Hall, Soskice, Hall and Soskice2001; Iversen Reference Iversen2005). Interlocking webs of firmly established legal obligations and the provision of services by way of public institutions are therefore (usually) already present here. These webs reflect the specific needs of the prevailing type of capitalist order and its comparative advantages (Hancké Reference Hancké, Wilson, Grant and Coen2010; Thelen Reference Thelen2004). In addition, social movements and organizations such as the environmental movement and NGOs or unions feed into this web of obligations and public services (Esping-Anderson Reference Esping-Anderson1990; Korpi Reference Korpi2006). In consequence firms often deem themselves to be over-regulated in contexts of consolidated statehood: they do not see the need for additional voluntary standards.²⁴ In Germany, for instance, there is a strong emphasis on skilled labor, a sophisticated web of obligations, and collective institutions that organize health care services including occupational health care for workers (Jackson and Deeg Reference Jackson and Deeg2006; Streeck and Yamamura Reference Streeck and Yamamura2001). Firms usually do not exceed these levels, even if their workforce is highly skilled. There is simply no need for it: high levels of regulation and service provision are already in place. In areas of limited statehood, by contrast, interlocking systems of firm obligations and public institutions do not exist in the same way and the provision of services and governance is generally insufficient. Instead, the “shadow of anarchy” (Börzel and Risse Reference Börzel and Risse2010; Börzel and Thauer Reference Börzel and Thauer2013; Scharpf Reference Scharpf1997) looms over firms: the knowledge that the risks associated with asset-specific investments will remain unmitigated, if they do not adopt high standards voluntarily. Regulatory void – limited statehood – is thus a scope condition for internal drivers to be effective in the way predicted in this study.

Note, however, that areas of regulatory void (and, accordingly, the shadow of anarchy) are, according to the way the concept is understood here, not confined to failing, failed, or weak states – or to the “global south” (Draude et al. Reference Draude, Risse and Schmelzle2012: 11). “Area” here pertains to functionally, socially, historically, or territorially defined political spaces. Spaces of regulatory void can thus occupy parts of state territories or transcend state borders and be transnational or global in nature – or concern a specific policy field or historical period. In other words: there may be areas of regulatory void to be found in the midst of Germany, for example – known to be a rather “strong” state – or Denmark or the US.²⁵ The theory of internal drivers applies to all areas of limited statehood – within Europe or the US in the same way as in the emerging markets context of South Africa and China.²⁶

The second scope condition pertains to the content of governance rules: more specifically to the common distinction in literature between standards in the area of “product regulation” and “process regulation” (Vogel Reference Vogel1995, Reference Vogel2005). Standards regulating products define the physical attributes of goods, such as quantity, size, and quality-related aspects (Vogel Reference Vogel1995: 18). Standards regulating the process of production formulate requirements in regard to the way a product is manufactured or otherwise made. Both follow different logics: product standards tend towards “race to the top” dynamics. The real issue, however, is process standards, which are particularly prone to “race to the bottom” dynamics (see above). The arguments presented in this study merely explain process standards, the more unlikely case of firm behavior in relation to regulatory standards. They do not explain product standards. The causal logic of the arguments set this scope condition: process standards reduce vulnerabilities, uncertainties, and information asymmetries in the period between an agreement among two exchanging parties and the execution of this agreement (that is, the delivery of a good in exchange for a price), which is the critical period in situations of asset specificity. Product standards, by contrast, are of no use here. They can be ascertained only with or after the execution of the agreement (i.e. upon delivery of goods in exchange for the agreed price). This is, from the perspective of the transacting parties, too late should the agreement involve asset-specific investments. In consequence, product standards do not help decision-makers in firms overcome the managerial dilemmas inherent in asset-specific investments, and are therefore not an outcome of the internal drivers theorized here.

Third, the argument for internal drivers applies to market, rather than politically driven firms. The rationale underlying internal drivers presupposes that the management of main assets and resources is a key variable for a firm’s commercial success. That is to say that if assets are mismanaged, there is a good chance that the firm will endure a loss in efficiency and profitability and encounter debts or bankruptcy caused by a slump in competitiveness. It is these negative yet anticipated consequences of asset mismanagement that motivate managers to engage in the governance of asset-specific investments and to make use of corporate social responsibility standards, management systems, and policies. Politically driven firms, in comparison, find their key factor for success in public affairs management (Coen et al. Reference Coen, Grant, Wilson, Wilson, Grant and Coen2010: 16–17). This is not to say that internal drivers are entirely irrelevant for firms that are more politically driven than market-driven. However, the game they play is a different one and it is more often than not the case that political pressure on firms regarding their behavior will trump the causal force of internal drivers. Examples of politically driven firms include most state-owned companies, such as Petrobras in Brazil, the energy provider Eskom in South Africa, or firms that rely disproportionally on public spending or subsidies such as the aviation industry or energy suppliers in general (Chick Reference Chick, Wilson, Grant and Coen2011). Likewise, firms whose business model relies on licensing by the state, as in large parts of the mining and military industries, are predominantly politically rather than market-driven (Baldwin and Cave Reference Baldwin and Cave1999; Besley Reference Besley2006). Moreover, many former communist countries, such as China or Russia, where considerable parts of industry sectors are still state-owned or partly state-owned, require foreign investors in strategic key industries to obtain licenses from the government to enter the market and/or do so in the form of joint ventures with local, often state-owned enterprises. Also in these cases – an example of which would be the automotive industry in China – firms are in all likelihood more politically driven than market-oriented.²⁷

Within these limitations, the arguments addressed in this study can be applied to a whole range of cases. In fact they decisively contribute to a better understanding of corporate social responsibility for a significant part of global production. In cases where the state takes a distant stance towards industry regulation and does not own, license, or otherwise dominate the logic of an industry sector, the arguments of this book apply and may be tested with respect to their explanatory power. The empirical analysis will assess the arguments in the context of the South African and Chinese textile industry and its various subsectors and segments, as well as the South African automotive industry with its various subsectors and segments. The arguments developed here are applicable beyond that, though, as, for example, to textile and lower electronic production in South-East Asia, automotive production in Latin America, household items production in India, furniture manufacturing, high-tech electronics, and processor production in various countries around the world. Table 2.1 summarizes the main theoretical arguments for internal drivers and the scope conditions within which they apply.

Table 2.1 Internal drivers of corporate social responsibility

Before turning to an empirical evaluation of the theory, the next chapter will discuss in detail what corporate social responsibility is and how we can detect it. It will also lay out the organization and design of the inquiry, the case selection, and the institutional and policy context of the analysis.