The Paradox of Corporate Social Responsibility Standards

“When I obey a rule, I do not choose.

I obey the rule blindly.”

Wittgenstein, Philosophical Investigations

(1958: 219; emphasis added)

Abstract

The purpose of this paper is to provide a constructive criticism of Corporate Social Responsibility (CSR) standards. After pointing out a number of benefits and limitations in the effectiveness of CSR standards, both from a theoretical point of view and in the light of empirical evidence, we formulate and discuss a Paradox of CSR standards: despite being well-intended, CSR standards can favor the emergence of a thoughtless, blind and blinkered mindset which is counterproductive of their aim of enhancing the social responsibility of the organization. We analyze three problems that might underlie the Paradox—namely the problem of deceptive measurements; the problem of responsibility erosion and the problem of blinkered culture. We apply the philosophical tradition of American Pragmatism to reflect on these issues in relation to different types of existing standards, and conclude by suggesting a number of considerations that could help both CSR standards developers and users to address the Paradox.

Introduction

The purpose of this paper is to provide a constructive criticism of Corporate Social Responsibility (CSR) standards—the wide set of national and international standards aimed at advancing the social, ethical and environmental performance of organizations¹—addressing both their theoretical design and their practical implementations, by pointing out three specific problems that are intrinsic to any CSR standard. We argue that these three distinct but intertwined problems can generate what we call the Paradox of CSR standards, that is the risk of the emergence within organizations of a thoughtless, blind and blinkered mindset that is counterproductive with respect to the aim of enhancing the actual CSR of the organization. This risk rises as more standards are implemented.

Our discussion around the Paradox of CSR standards is based on a clear philosophical foundation—namely, the American Pragmatism of the early writings of James (1896, 1912) and Dewey (1927), and the more recent contributions of Rorty (1982, 1990, 1991, 1999) and Putnam (2004).

The structure of the paper is as follows: In the first section we point out a number of potential limitations to the effectiveness of CSR standards, both from a theoretical point of view and in the light of empirical evidence resulting from the existing literature assessing the implementation of CSR standards.

In the second section we formulate the Paradox of CSR Standards and discuss three different problems that may originate it—namely the problem of deceptive measures (P₁); the problem of responsibility erosion (P₂) and the problem of blinkered culture (P₃).

In the third section we apply the core ideas of American Pragmatism to discuss these issues and we suggest a few considerations that should help both CSR standards designers and standards users to achieve more effectively what standards promise.

Although most of our attention will be devoted to highlighting critical issues that arise from our theoretical and empirical examination, we believe that CSR standards are well intentioned and can have positive effects, which might not have been achieved except by following the standard.² It is precisely for this reason that we think it useful to advise users and developers that CSR standards effectiveness may be improved if the intertwined roots of the Paradox are addressed.

The Trouble with CSR Standards—Too Many Different Ingredients in the Melting Pot?

CSR standards include the wide set of national and international standards that—despite relevant differences in their nature, focus, monitoring mechanisms and governance structure—all share a common objective: to advance the social, ethical and environmental performance of organizations by codifying aspects of organizational behavior. CSR, which is sometimes thought of as an overarching concept, is understood and used in widely different ways (from the inspiring principle to adopt anti-corruption initiatives, to a label for community volunteering, for example). The many international initiatives and standards aimed at advancing the CSR of organizations reflect this wide scope of the concept. As a result a number of business ethics and organization scholars have developed frameworks to classify the various CSR standards, providing useful guidance to standard users.³

The number of international CSR standards that have been developed in the last decade is overwhelming, and a number of authors in the business ethics and management literature have already suggested various classification schemes to help managers navigate this intricate jungle. In her detailed book Leipziger (2003) analyzed in depth 32 ‘codes’ and initiatives by distinguishing between performance-oriented and process-oriented codes in the first instance. While the former focus on the definition of outcomes (e.g., avoiding child labor), the latter focus on the process, i.e., the procedures that organizations should follow in order to achieve socially desirable outcomes (e.g., having management systems designed to avoid child labor).

Rasche and Esser (2006) suggested a categorization based on five different criteria to analyze and compare each standard according to its: (1) nature (performance, certification, or process standard); (2) field of application (social, economic, or environmental issues); (3) focus (on accounting, auditing, or reporting processes); (4) geographical reach (global, regional, or local diffusion), and (5) scope (generic, industry, firm level).

Similarly, Gilbert and Rasche (2007, 2008) adopted the term “standardized ethics initiative” to encompass the wide set of “voluntary (usually global) standards defining norms and procedures for organizational behavior with regard to social and/or environmental issues” and point out three particular differences: (1) the type of issues they codify (e.g., social, environmental, or both); (2) the kind of process they standardize (e.g., accounting or reporting, or the assurance process), and (3) the level of specificity of the norms (e.g., ‘aspirational’ general principles or detailed instructions for a specific industrial sector).

An ‘architecture’ of international CSR standards based on the fundamental distinction between specialized CSR standards and overall management systems has been presented by de Colle (2006). According to this classification, specialized CSR standards are types of codification focused primarily on one-issue (e.g., working conditions; environmental emissions, etc.) or one-stakeholder (e.g., employee; contractors, etc.), while overall management systems refer to standards that try to codify “the wide range of social, environmental and ethical issues and look at the relationships between the organization and all its organizational stakeholders, without focusing on one group in particular” (de Colle 2006, p. 336).

The work by Waddock (2008) provides a unifying perspective—what she terms the “new institutional infrastructure for corporate responsibility”—emerging from the interrelationships among the various initiatives, standards, and organizations operating in the CSR field. Waddock’s classification is based on the distinction between “state/government, market/business and civil society categories” (Waddock 2008, p. 87) and her thorough review provides us with up-to-date tables with the most relevant international initiatives in the CSR arena, including normative codes (e.g., The Caux, CERES, and Equator Principles, etc.), reporting initiatives (e.g., GRI, AA1000, etc.), consulting organizations, assurance providers, NGOs, Socially Responsible Investing and business membership organizations, CSR research centers, ratings, CSR publications, watchdogs, and other activists.

We present below (Table 1) a summary based on these various classifications of international CSR standards that highlights four key elements of each initiative—namely its nature, focus, monitoring process and development and governance approach. These four elements assume a particular importance in the light of three specific problems that our Paradox will point out. First of all, standards can be classified according to their nature, i.e., whether they are process or substantive standards. Process standards are concerned with organizational processes (intended to address substantive areas of concern), e.g., the existence of a management system of a specified type to deal with environmental performance. For example, AccountAbility 1000 is a process standard that provides guidance to organizations willing to improve their accountability, by indicating a number of quality principles for the process of social and ethical accounting, auditing and reporting. On the other hand, substantive standards provide organizations with a clear indication for the content of their actions (such as principles, policies, processes, and outcomes) in the area of CSR. For example, The UN Global Compact identifies ten general principles that should guide business behavior at any level; the GRI indicates which areas any organization should report on, to demonstrate its sustainability performance; and SA8000 indicates acceptable policies in dealing with suppliers. Substantive standards can be directly concerned with performance levels in the areas of concern, e.g., the number of properties experiencing foul flooding (Fig. 1).

Table 1 Key elements of international CSR standards

Fig. 1

Positive and negative effects of CSR standards

Second, one can look at the focus of standards, and differentiate between standards that are aimed at helping managing the organization’s overall social, ethical, and environmental responsibilities (‘general standards’), and ‘specialized’ CSR standards that are developed to address in a more specific way a particular issue (e.g., human rights, or the environment) or a particular stakeholder group (e.g., the customers).

Third, CSR standards may or may not include an independent monitoring process, which may be formalized either as a third-party assurance process or as a certification scheme. The inclusion of a monitoring system helps stakeholders evaluate the effectiveness of the implementation of the standards’ requirements by the adopting organizations.

Fourth, we think it is important to look at the process of development and at the governance structure of the standard itself, to assess whether the standard adopts a multi-stakeholder approach, in both these dimensions.

Without purporting to be exhaustive, Table 1 highlights the prevalence of these four different elements in a number of internationally adopted CSR standards. When two dimensions of the same characteristic can be found, we have put in brackets one of the two, to indicate that it is not the most salient characteristic of that standard.⁴ We will refer to this classification in our discussion of the CSR Paradox to point out how different aspects of the Paradox can be expected to have higher or lower relevance according to the key characteristics of CSR standards.

Each framework described above provides standards users with a better understanding of the specific characteristics of the many different CSR standards available, and are therefore valuable contributions to help managers and CSR service providers choose the right ‘ingredients’ to prepare successful CSR recipes. However, they seem to be built on the implicit assumption that the adoption of CSR standards is automatically going to deliver precisely the positive outcomes envisaged by the standard, at least if they are applied properly. As mentioned in the introduction, we think that a more critical approach has to be taken, by analyzing both positive and potentially negative effects that can derive from CSR standards. By assuming such a critical perspective, we will point out that there is an intrinsic paradox in the development of CSR standards. Before formulating the Paradox, in the next section we turn our attention to examining the possible advantages and negative effects generated by CSR standards.

In General, What is Good and What is Bad About CSR Standards?

Looking at the CSR standards from a general point of view, there are a number of observations that we can make about the positive and negative aspects that their adoption within any organization may generate.⁵ There are several dimensions of the beneficial outcomes that are often associated with CSR standards:

• 1. Operationalizing CSR. CSR standards offer practical frameworks to interpret principles and values, i.e., enabling the translation of abstract concepts into manageable tools.

• 2. Avoiding confusion. The development of CSR standards helps to eliminate confusion concerning the use of language and methodologies, by promoting commonly accepted processes and practices.

• 3. Supporting CSR uptake. By definition, the creation of a standard enables any organization (not just the leaders in the field) to adopt the codified practices and processes, thereby promoting a wider uptake of CSR.

• 4. Facilitating stakeholder engagement. By identifying shared ‘rules of the game’, internationally codified norms and principles provide a basis for engagement and partnership between the organization and its stakeholders.

• 5. Promoting continuous improvement. Like standards developed in other fields CSR standards can be designed to be open to experimentation and continuous improvement.

• 6. Enhancing corporate reputation. By adopting CSR standards corporations can enhance their reputation, since standards provide stakeholders with a ‘reference point’ for organizational accountability and enable public judgment about corporate behavior.

• 7. Enabling self-enforcement. Organizations that adopt CSR standards can activate a mechanism of self-enforcement based on the endogenous motivation of the organization’s participants—people are more inclined to behave responsibly if they can rely on being recognized as responsible managers/employees. This reduces external enforcement costs, such as control and sanctions.⁶

On the negative side, there are several possible drawbacks in the use of CSR standards that have been identified by business ethics and management scholars:

• 1. Conceptual inadequacy. Whilst standards have traditionally been developed to deal with technical issues, there are various methodological problems in defining and codifying complex social and ethical issues (Henriques 2010). Moreover, since there are often several standards for any given issue, confusion over terminology may be increased.

• 2. Introducing extra costs. The process of adopting CSR standards and, where appropriate, obtaining external certifications may represent a significant cost, particularly for small and medium enterprises. This is an issue considered, for example, in the development of ISO 26000, the new standard on social responsibility being developed by ISO (ISO 2009). Moreover, Blair et al. (2008) notes that the proliferation of international social responsibility standards raises significant implementation costs for many organizations, despite the potential positive effects that standards can bring in terms of reducing transaction costs.

• 3. Lack of enforcement. CSR standards are adopted on a voluntary basis, therefore they lack the legitimacy and strong compliance mechanisms available to the law. While it is—as noted above at point 7—often intrinsic to the nature of a CSR standard to rely on endogenous motivations rather than on external sanctions, several critics have pointed out that the voluntary nature of CSR standards represents a weakness that can lead to poor effectiveness of the standards (Bondy et al. 2004; Delmas and Keller 2005; Delmas and Montes-Sancho 2010; Simpson et al. 2012). Moreover, often CSR standards include provisions for specific processes (e.g., reporting) or substantive issues (e.g., minimum level of wages) but do not provide an accompanying schema to help stakeholders evaluate whether or how well these provisions have been implemented by adopting organizations (Seidl 2007). This issue can, however, be mitigated by the presence of monitoring mechanisms or independent assurance statements provided by qualified third-parties (as, for example, the AA1000 Assurance Standard requires).

• 4. Obsession with compliance. When organizations adopt a CSR standard, there is an intrinsic tendency to focus on conformity or compliance (since the standards typically identify outcomes and/or processes that can be monitored). If this becomes excessive, it may occur at the expense of building personal relationships and trust among organizational members and their stakeholders: following rules and procedures tends to become a clear organizational focus, at the expenses of values-based decision-making. Moreover, organizations may tend to apply standards without any adaptation to their specific needs and characteristics (lack of flexibility), thereby producing counter-productive effects (Seidl 2007, p. 708). The limitations of compliance-based ethics programmes have been pointed out by scholars in relation to the introduction of the US Federal Sentencing Guidelines for Organization (1991): their influence on corporate ethics has been criticized since they created incentives for companies “to create legal compliance programs” (Izraeli and Schwartz 1998: 1053).⁷

• 5. Over/miscommunication of data. Linked to the issue of a (too) formalistic compliance is the risk that standard users end up communicating a vast amount of data (typically, indicators) that does not represent, however, material information, thereby not allowing stakeholders to make informed judgments on the organization’s social responsibility. This has been described by John Elkington (2002) as the “carpet-bombing syndrome” as the excessive focus on formalistic compliance on reporting requirements might induce organizations to report as much as they can, loosing focus on the core material issues that are meaningful to determine their actual social performance.⁸ As a consequence, organizations fail to activate the reputation mechanism described above (see point 6 in the previous section).

• 6. Stifling innovation. By identifying and codifying ‘socially acceptable’ outcomes or ‘best practice’ processes, CSR standards have the potential to stifle innovation and creativity in dealing with new issues. As Dew and Sarasvathy (2007) point out, innovations carry irreducible uncertainties ex ante with regard to ex-post benefits or harms they may engender. In seeking to estimate a priori which innovations are likely to comply with pre-determined CSR standards and which are not, at least some potentially valuable paths to innovation are likely to be prematurely abandoned.

• 7. Failure of driving systemic change. While CSR standards might be effective in helping a specific organization improve its social, ethical, and environmental performance, it has been argued that there is a need to escalate from the level of corporations to the level of public policy in order to address the scale of problems that affects our societies: “The real challenge is not just to move in the right direction. Rather, the challenge is to move quickly enough to scale in changing direction so as to materially affect large-scale outcomes” (Zadek 2012, p. 214). In other words, CSR standards alone seem unable to drive the systemic (and, sometime, disruptive) change that seems to be needed to tackle issue at the national or global scale.

And in Practice, Are CSR Standards Effective? Some Empirical Findings

Our considerations above suggest that, besides the commonly agreed positive effects associated with the adoption of CSR standards, there are significant difficulties with CSR standards. With a few exceptions (e.g., Gilbert and Rasche 2007; Rasche 2010), these have not been adequately addressed in the literature. But, turning aside from the theoretical considerations for a moment, what is the evidence on the outcomes generated by the adoption of CSR standards? Interestingly, recent studies point out that this ambivalence between the positive and negative effects of CSR standards is not only a theoretical concern: practice confirms that there are issues of counter-efficacy in terms of lack of socially responsible outcomes associated with the use of CSR standards—or, at worst, outcomes generated by CSR standards, that are socially undesirable.

In their study of the effects generated by the self-regulating program “Responsible Care” adopted by US chemical companies, King and Lenox (2000) showed that that program failed to generate better environmental performance for companies adhering to certain a priori standards (these easily fall into the ‘CSR standards’ category according to our broad definition of the term). Their database included over 1,500 firms and the key data observed was the amount of a company’s toxic releases over a 10-year period (1987–1996). The authors concluded that the findings of their analysis provided “no evidence that Responsible Care has positively influenced the rate of improvement among its members.” Disturbingly, King and Lenox also found empirical evidence that members of Responsible Care were “improving their relative environmental performance more slowly than non-members.” These surprising results, concluded the authors, pointed out “the difficulty of creating self-regulation without explicit sanctions”. In other words, the issue of lack of enforcement that we identified above seems in this case to have completely undermined any positive effects consequent upon the introduction of the CSR standard.

Similarly, King et al. (2005) found no evidence that obtaining ISO 14000 certification generates improvements in environmental performance. Analyzing environmental performance data for 7,899 US manufacturing facilities over a 6-year period (1995–2001), the authors concluded that the ISO environmental standard might be seen more as an indicator of an already existing environmental management process, rather than as a sign of superior environmental performance. They also showed an interesting ‘reverse decoupling’ effect: companies that adopted environmental management programs before ISO 14000 was introduced, were then looking for certification to communicate in a credible way their established good performance. These empirical findings on poor or even negative effects of the adoption of environmental and other quality standards seems to be consistent with Saunders (1992) Khazzoom-Brookes postulate (originally the Jevons Paradox), which suggests that increasing environmental (or other resource) efficiency will lead to an absolute rise in the use of that resource. They also account for the disillusionment with which voluntary CSR standards are regarded by significant sections of the NGO community. As a result, few large NGOs participated in the development of ISO 26000, the recent standard on Social Responsibility developed by the International Organization for Standardization (ISO) through a multi-stakeholder process including representatives from different ‘stakeholder’ groups such as NGOs, labor, consumer, industry, Government, and academia.

Anecdotal evidence includes the adoption by Shell of ISO 14001 certification for its oil sands operation in Canada (Shell 2009). As an environmental management system, ISO 14001 is a process standard concerned with “protecting the environment and continually improving its environmental performance”. However, oil sands are one of the most damaging and carbon-inefficient methods of extracting oil. There appears to be a contradiction between the purpose of the standard and its use in this case by Shell.

The Ethical Trading Initiative is concerned with the improvement of outcomes for workers in the supply chain. It has developed a mixed (process and substantive) standard. It is a cross-sectoral initiative, involving companies, NGOs, unions and the UK government. In 2006 it commissioned a study (Barrientos and Smith 2007) which found, for the specific projects studied, that while there were some minor improvements noted in a number of areas major improvements were rare.

SA8000 is another standard for supply chain impacts, which relies on a similar code to that of the ETI. Here again there is little evidence that the standard has made a significant difference, according to Hiscox et al. (2009).

On the other hand, there is also research showing the positive impact of CSR standards. For example, Terlaak and King (2006) analyzed US manufacturing facilities to test whether certification with the ISO 9000 Quality Management Standard generated a competitive advantage. Their 11-year period analysis showed that ISO 9000 had a positive effect, improving the growth of US certified manufacturing companies, as it worked as a signaling mechanism in situations of poor information and fear of opportunism. However, it is worth pointing out that their work did not look at whether the use of ISO 9000 also improved the quality of the manufacturer’s products.

Another positive example is the regulation of the water industry in the UK. Here, due to clear substantive targets together with monitoring, the levels of water service is improving. For example the indicator ‘Properties at risk of sewer flooding incidents (once in ten years)’ has improved from 0.07 % in the 5 years to 2000 to 0.02 % in 2007 (OFWAT 2007).

Blair et al. (2008) point out, after a legal and economic analysis, a number of positive outcomes that are associated with the diffusion of CSR standards, or, in their terms, “widely-applicable social and environmental norms for business process and behavior”. Their perspective shows how standards can help organizations reduce transaction costs in different ways: by reducing the complexity of managing intra-firm relationships (facilitating outsourcing); by lowering costs of monitoring and control (ensuring product or process quality), and by reducing the cost of communication within supply-chains. Therefore, the authors conclude that the diffusion of CSR standards (linked, very importantly, with the provision of third-party assurance) can help the stabilization and growth of business relationships, especially in areas of the world where the law is weak, by providing “the dominant mechanism for regulating business and enforcing contracts”.

How should we interpret this mixed evidence on the positive or negative impacts of CSR standards? Why, despite the growing diffusion of assurance methodologies and standards aimed at improving effectiveness in the adoption of CSR standards (a new version of AA1000 Assurance Standard was released in 2008), CSR standards do not always deliver what is expected?

Recent work by Terlaak (2007) provides an interesting theoretical framework to look at these issues, linking together institutional theory, new institutional economics and corporate social behavior research. Terlaak points out the importance of distinguishing between two central elements of CSR standards—namely the codification and the certification of desired behaviors. While both elements can improve the effectiveness of self-regulation, Terlaak discusses how at the same time, codification and certification can generate counter-effects that undermine the effectiveness of the standardization process by generating a “pattern of compliance that undermines the decentralized enforcement process and, thus, limits the certified management standard’s effectiveness to guide firm behaviors” (Terlaak 2007, p. 969). This perspective points out an intrinsic contradiction within the standards design process itself. It suggests, in other words, that there may be an inherent ‘paradox’ of CSR standards.

The Paradox of CSR Standards

A unique theoretical conceptualization to make sense of the different ‘performances’ of the various CSR standards is not available, and it will probably never be. Standard effectiveness or failure might also depend, as Simpson et al. (2012) argue, to whether the internal capabilities of the organization ‘match’ with the institutional requirements (e.g., setting up a suppliers’ code of ethics) that the standard may include. In other words, we need to acknowledge that it is not possible, ex ante, to predict the pattern of performance of an organization following its adoption of a CSR standard. This perspective represents, in our view, an interesting avenue to be investigated, as it points out an intrinsic contradiction within the project of standardization itself. It suggests, in other words, that there may be an inherent Paradox of CSR Standards—that is, a situation where the consequences of reasoning are “contrary to (para) received belief or opinion (doxa)”, for example, when seemingly reasonable premises yield apparently contradictory conclusions and “the arguments for competing conclusions mirror one other” (Clark 2007, pp. 151–153).⁹ In our case, the Paradox of CSR standards represents the emergence of unintended (counterproductive) consequences that might reduce, or even offset, the positive outcomes of CSR standards on the overall social performance of organizations:

The Paradox of CSR Standards:

The ( thoughtful, responsible and stakeholder-oriented ) mindset

that CSR standards aim to promote among standards users

may be directly counteracted

by the ( thoughtless, blind and blinkered ) mindset

that standards users tend to employ,

the more they focus on implementing CSR standards.

In other words, while CSR standards are designed to promote the organization’s ability to identify, manage and improve its social, environmental, and ethical performance, by focusing on implementation the members of the organization may lose focus on the standards’ genuine purpose, and unconsciously enact or tolerate behaviors that eventually undermine the social responsibility of the organization. Paradoxically, it could be said that the more a CSR standard is successful (in terms of its implementation inside an organization), the more counter-acting behaviors can emerge.

We believe that the CSR Paradox is the result of three intertwined problems:

P₁ :

The problem of deceptive measurements;

P₂ :

The problem of responsibility erosion and

P₃ :

The problem of blinkered culture

Below we discuss more in detail these three roots of the Paradox.

The Problem of Deceptive Measurements: CSR Standards may (Unsuccessfully) Try to Measure the Unmeasurable

How do we measure (CSR) concepts like “supporting human rights”, “refraining from arbitrary discrimination”, “valuing diversity”? How can we measure the quality of a stakeholder engagement process? How do we assess the “adherence to the AA1000 AccountAbility Principles of Inclusivity, Materiality and Responsiveness” or the “reliability of specified information provided”? (both are quotes from AA1000 (2008) Assurance Standard).

From a pragmatic perspective we think that the validity of any measure is given by its “usefulness as an indicator or predictor for some other behavior or characteristic…” (Sellitz et al. 1961, quoted in Petersen 2002, p. 94). The measurement of Human Development or intellectual capital provides an example, where researchers have attempted to “measure the unmeasurable”. While it may be straightforward to identify meaningful (useful) indicators, it may turn out to be impossible to measure such intangible aspects directly. As a result, organizations can try to identify proxies of the unmeasurable aspects, which can be measured. However, the use of proxies can lead to a progressive loss of attention to the important, but unmeasurable aspect. For example, if we measure “human capital” by counting the hours of training per employee provided by an organization, to what extent are we capturing the quality of the training? A second dimension of this problem is given by the (necessary) vagueness of the concepts/dimensions that CSR standards try to capture. As pointed out by Petersen (2002), who critically analyzes and discusses the effect of schema and standards (of quality) on individual behavior within organizations, we might easily resort to “deceptive measures”, and so lose focus on the really important properties that we want to codify and promote within the organization. This might happen as “the unmeasurable qualities may have declined in importance as a result of the focus on the measurable qualities” (Petersen 2002, p. 100).

Looking at the different types of standards described in Table 1, this first dimension of the CSR Paradox seems to be particularly relevant for substantive, specialized and certifiable standards—such as, for example, SA8000. This is because this type of standard deliberately puts a higher emphasis on defining ex ante, specific, measurable performance levels, that can be ex-post verified and certified by an independent external organization. By using proxies, the members of an organization trying hard to increase its CSR performance might lose the focus on the meaningfulness of the standardized measures, and just try to ‘tick the box’ of a given metrics.

An example of the way this problem frustrates the intention behind implementing CSR standards is provided by the following case. In the late 1990s two large oil companies were engaged in a marketing joint venture in Europe. Environmental, health and safety issues were managed through a tool. The application of that tool required an auditor to ask questions that were scored by seeing which of a small number of scored answers most closely applied to the responses given. Managers were assessed according to the total score for their particular unit. Different types of business had different questionnaires (e.g., refineries, fuels distribution, lubrication oil plants, etc.). From a fiduciary angle the problem was that the system did not identify in any meaningful way the most significant risks that were not being properly managed. Obviously, there were certain scores below which it would not be acceptable to fall, but it was possible for the best performing site to have a significant issue that would not have been properly identified to senior management. As far as site managers were concerned, the approach led to behaviors such as maximizing scores where it was easiest to achieve full marks rather than where the highest risks were likely to occur.¹⁰

In general there is considerable anecdotal evidence that standards amplify an unthinking, tick box attitude. From a quality perspective, the tick box approach can be linked to so-called “satisfaction surveys”. Companies draw many false conclusions from levels of satisfaction. This is a particular problem for hotels and cruises. Staff on cruise ships, for example, try to encourage passengers to rate their departments 9 or 10; their salary literally depends on this. Most passengers oblige. Therefore if there is a real issue the cruise lines can say that (other) passengers have not complained. That is because they have not been asked the right question. Also, there is unlikely to be any real correlation between “satisfaction expressed” and future business.

The Problem of Responsibility Erosion: CSR Standards may Erode Individual (and Organizational) Responsibility

By codifying desired behavior, CSR standards can generate a counter-productive tendency as a result of which the members of the organization align their behaviors to the process or outcomes specified by the standard in an algorithm-like mode, rather than in a deliberative, responsible mode. The dangers of such an attitude can be huge, and have been eloquently highlighted by philosophers. Wittgenstein hit the essence of this problem in one sentence: “When I obey a rule I do not choose. I obey a rule blindly” (Wittgenstein 1958, p. 219; emphasis added). The general dimension of this problem is that the more organizational decision-making tends to focus on complying with predetermined logical schemas and rules, the more the decision makers will tend to pay attention to manuals and procedures concerning ‘the right way of making the decision’, possibly losing sight of the importance of ‘making the right decision’. In other words, following rules and standards might develop a tendency towards a quasi-automatic thinking, by which individuals ask themselves “Am I following the rule?” instead of “Am I doing the right thing?”. Similarly, Arendt (1971, p. 7) pointed out that the evil-doing in Nazi-occupied Europe was made possible by the “banality of evil”, that is a situation where individual actors “had not the slightest difficulty in accepting an entire set of rules….as if it were nothing but another language rule”. The resulting absence of individual thinking, of individual responsibility, was therefore made possible by a “morality collapsed to a set of mores—manners, customs, conventions to be changed at will”.

When individuals within an organization model their behavior in order to show compliance with a certain standard, they “somehow lose ownership of the process: it becomes something exterior to them. [..] It is not really something they feel responsible for.” (Petersen 2002, p. 86, emphasis added).

In other words, the second problem underlying the Paradox is that the more CSR standards are successful in shaping individual behavior, the less that behavior tends to be the result of individual, morally responsible and ‘ethically aware’ deliberation. This also holds true for behavior that may be attributed to the organization as a whole: by eroding individual responsibility, organizational responsibility can also be reduced, whether or not this arises directly out of the individuals’ intentions. The paradoxical nature of this problem can be understood as a manifestation of bounded rationality, the concept developed by Herbert Simon (1957). In fact, it is because of the theory of bounded rationality that rational choice theorists and organizational scholars have advocated the use of organizational rules, procedures, and routines to restrict the wide (too wide, for rationally bounded decision-makers) set of choices available to managers. As Elster (1979) explains, sometimes “less (choice) can be more”, i.e., it can be rational to renounce some freedom of choice, when we know that our limited cognitive capacities will fail in front of too ambitious tasks (as in the example of Ulysses, who decided to tie himself up to the mask of his ship, in order to avoid being wooed by the sirens). At the same time, the entrustment of such a procedural rationality (i.e., rule-following) generates within organizations a tendency to reduce the room for individual ethical thinking within decision-making processes. The problem of the responsibility erosion can therefore be reformulated as follows: how can rules and procedures, that are needed to cope with bounded rationality, be designed in such a way that they do not reduce decision-making to mere rule-following behaviors, but allow some flexibility in their interpretation and adaptation to enable responsible decision-making, i.e., the result of a process combining moral reasoning and moral imagination?¹¹

Such a concern seems to be shared by Stansbury and Barry (2007)—not surprisingly, since the title of their article, “Ethics Programs and The Paradox of Control”, already suggests a similar line of reasoning. The authors critically discuss the effectiveness of ethics programs and argues that when organizations place emphasis on coercive control (i.e., implementing a compliance orientation instead of developing a values-based approach) they may undermine the program’s own effectiveness, because “they institutionalize control and thereby risk politicization, indoctrination, and an atrophy of competences” (Stansbury and Barry 2007, p. 256). We will offer a partial solution to this problem from the perspective of American Pragmatism in the next section.

Looking at our classification of CSR standards, we can see how this second dimension of the CSR Paradox is more relevant for process (both general guidance and specialized) standards such as, for example, ISO 9000, ISO 14000, and AA1000. This is because these types of standard focus primarily on the organizational procedures and individual rules of conduct in approaching and managing “CSR issues” (such as setting up an environmental management system or a stakeholder engagement process).

The Problem of Blinkered Culture: CSR Standards may Encourage an Inward-Orientation of the Organization and Its Members, Rather than a Stakeholder-Orientation

The third problem that can lead to the Paradox of CSR standards concerns the effects of the adoption of a standard on the organization’s stakeholder culture, i.e., the “the aspects of organizational culture consisting of the beliefs, values, and practices that have evolved for solving problems and otherwise managing stakeholder relationships.” (Jones et al. 2007).

The tendency to focus on the managerial/procedural aspect of a CSR standard, especially in its implementation process, can be detrimental for the outward-looking stakeholder-orientation of the organization’s members, by shifting the organizational members attention on the implementation process itself: “Rules and regulations […] encourage people to be oriented inwards—towards doing the job as they felt it should be done—rather than outwards, towards the users of public services.” (Davis, quoted in Petersen 2002, p. 87).

Clearly, since a common aim of CSR standards is to put the interests of the organization’s stakeholders before decision-makers, the risk is that this tendency reduces the ability of managers to effectively identify and respond to stakeholder concerns.

This kind of problem emerging in the implementation of CSR standards has been pointed out by Rasche and Esser (2006), who critically discuss—curiously, also using the term ‘paradox’ to indicate the core of the problem—how a managerially centered process of standard implementation produces the counter-productive effect of losing the ability to represent the wide interests of all the organization’s stakeholders: “…such an approach also leads to a paradox: because management defines the scope of the SEAAR [social and ethical accounting, auditing and reporting] process ex ante, it includes only those stakeholders it wishes to communicate with. As a consequence, the very aim of accountability practice, which is to address the information needs and concerns of all relevant stakeholders by providing reliable and relevant information […], is missed due to the flawed decision making process.” (Rasche and Esser 2006, p. 252). This risk might be much more relevant for those CSR standards that are primarily “internally focused”, i.e., aimed at the organization’s managers and employees, with the main objective of creating a strong organizational culture that fosters homogeneous and coherent behaviors at individual and organizational level throughout the company. The focus of these SEEAR approaches is therefore internal, and “the stakeholder dimension—if any—is present in an indirect way, mediated by the primary stakeholder group: the employees.” (de Colle and Gonella 2002, p. 87).

While both substantive and process standards may equally suffer from this dimension of the Paradox, we think that, in light of the above considerations, the key element that might be relevant to reduce (or exacerbate) this potential risk is the presence (or absence) of a multi-stakeholder approach, both in the development phase and in the governance structure of the CSR standard itself.

Making Sense of the Paradox: A Pragmatist Approach

In order to discuss the implications of the Paradox for both standards designers and user, it might be useful to step back and reflect on a good philosophical approach to inform such a discussion.

One of (if not the) key concept of the tradition of American Pragmatism of Peirce, James and Dewey, is that there is no point—and it can actually be useless, or even dangerous—to try to search for the absolute “truth” about the natural and social phenomena we are trying to understand. The search for truth—as accuracy of representation—is replaced by the pragmatist with an approach to scientific inquiry not as a device to represent reality, but rather as a method to make sense of reality: as Rorty suggests “scientific inquiry could be seen as adapting and coping rather than copying” (Rorty 1982, p. 86) and “we should view inquiry as a way of using reality” (Rorty 1999, p. 33)

For the pragmatist, our concepts, beliefs, and theories are the best “tools” that we can develop to understand and make sense of these phenomena. The more useful these tools are, the better they are, as they will enable us to live in a better way. What matters, from a pragmatist’s perspective, is the method through which we develop and refine our beliefs, concepts, theories. Ideas should be based on experience and evolve in a way that resonates with Rawls’s reflective equilibrium: as Rorty points out, “the whole point of Dewey’s experimentalism in moral theory is that you need to keep running back and forth between principles and the results of applying principles” (Rorty 1991, p. 68). John Dewey emphasized the importance of the experimental method in scientific research and pointed out the key elements of this approach: “First, that those concepts, general principles, theories and dialectical developments which are indispensable to any systematic knowledge be shaped and tested as tools of inquiry. Secondly, that policies and proposals for social action be treated as working hypotheses, not as programs to be rigidly adhered and executed” (Dewey 1927, p. 203).

Looking at CSR standards from the perspective of American pragmatism, there is an obvious issue that confronts the nature of any standard: How can a “fixed truth” be useful? How can some ‘ex ante’ defined criteria be valid in light of the experimental method? Here the key issue not only concerns the content of CSR standards, but also their development and implementation processes.

Paradoxically, contrary to its anti-dogmatic attitude, we believe that the pragmatist perspective is not against the use of CSR standards, per se. As we argued in the introduction, our aim is to provide a constructive criticism to CSR standards, not to advocate their dismissal. What the pragmatist perspective enables us to see is that to avoid standard failures (due to the Paradox effects) standard users and designers need to keep in mind that the abstract principles embedded in CSR standards need continuous interpretations and adaptations. These interpretations are necessarily culture-relative, i.e., depend on how meanings are understood in a culture and rules are developed as the results of ‘language games’ within any given community. But this does not mean that no global rules or values can be developed. Contrary to this relativistic drifting, Rorty’s idea of a “cosmopolitan social democratic community” enables us to look at global CSR standards from a pragmatist perspective. Instead of searching for universal moral principles (“meta-narratives”) based on some superior moral law (or on Western standards), Rorty advocates the emergence of a cosmopolitan community sharing some common beliefs and values that its members will have chosen freely, after a process of mutual exchange of ideas (Rorty 1990, p. 212). These shared values (“first-order narratives”, not discovered, but made) should be developed through “richer, more muddled, and more painful syntheses of opposing values” (Rorty 1990, p. 213). In other words, CSR standards can be useful if they are the result of a free conversation among stakeholders of different cultures, based on “tolerant reciprocity” enabling the identification of a common ground “defined by the overlap between their communal beliefs and desires and our own [the Westeners]” (Rorty 1990, p. 213).

In assessing the effectiveness of standards, the first question a pragmatist would ask is “does the standard give rise to the behavior that will deliver what the standard was designed to produce?” On this point, as our literature review has shown, the evidence is mixed. Even for one of the longest established and most widely internationally adopted standards, ISO 14001, the empirical research so far is uncertain. But crucially, there is little strong evidence that adopting the standard results in improved environmental performance.

Standards obviously have many other positive outcomes. These include a crucial aspect that is often overlooked: raising awareness of the issues covered by the standard. This is particularly beneficial for the ultimate goal of any CSR standard, since—unlike standards for much more technical areas—they are typically developed before agreement has been reached on the issues concerned. The result is that the process of standards development itself becomes a process for educating and eliciting agreement from the wide range of participants in the development process. In the light of our personal experience, we can certainly say that this has been true of both the GRI and the ISO 26000 process. The development of ISO 26000 illustrates a large-scale learning process: it involved a working group of more than 400 representatives of six stakeholder groups drawn from over 77 countries (which was an innovative way of working for ISO itself). In addition, the stated purpose of ISO 26000 is explicitly to provide ‘guidance’ to its users, not to provide a set of precise norms of behavior or processes.

From the perspective of American pragmatism the importance of this aspect can be emphasized by looking at standards in the same way as James (1907) looked at beliefs and ideas: “Our beliefs are really rules for action” he wrote, and he added that “to develop a thought’s meaning, we need only determine what conduct it is fitted to produce: that conduct is for us its sole significance” (James 1896, p. 94). Ideas, continued James, are tools; theories are instruments that can help us to move forward, to find better solutions. It is not meaningful to look for the ‘truth’ of ideas, any more than to find out whether they are true instrumentally: “Ideas (which themselves are but parts of our experience) become true just in so far as they help us to get into satisfactory relation with other parts of our experience.” (ibidem: 100). In other words, the pragmatist attitude encourages us to see how CSR standards can be useful tools to promote the kind of behaviors that will deliver socially, ethically, environmentally desirable outcomes. CSR standards are also important as a learning tool, as a device for self-regulation. However, the importance of learning (developing new beliefs, and changing existing ones) in the process of standards development and implementation is often overlooked, by privileging a focus on compliance with the standard’s prescriptions. This issue seems to be more relevant for substantive, certifiable standards, since by their nature they emphasize the precise application and compliance with pre-determined rules and performance measures.

Once standards have been developed, they need to be implemented. This provides a further opportunity for practical learning concerning the relevant issues and their management. A second aspect of this type of benefit is provided by the activities of those who provide assurance of the implementation of standards. The typical role of assurance providers in relation to standards is to provide independent feedback on the implementation of the relevant standard. At least anecdotally, the view of a number of corporate managers is that such feedback actually constitutes valuable advice, which is hard to obtain elsewhere. Again, the pragmatist perspective—particularly, the call for empiricism by James and the idea of experimental method by Dewey—provides us with a philosophical basis for understanding the importance of the practical learning that derives from the process of standards implementation. As James wrote in “The Will To Believe” (1896), we should appreciate the importance of learning from practical experience, and, at the same time, be open to adapt/improve our beliefs in the light of experience: “I am a complete empiricist so far as my theory of human knowledge goes. I live, to be sure, by the practical faith that we must go on experiencing and thinking over our experience, for only thus can our opinion grow more true; but to hold them—I absolutely do not care which—as if it never could be reinterpretable or corrigible, I believe to be a tremendously mistaken attitude…” (James, 1896, p. 79).

In sum, CSR standards, once defined, embody today’s knowledge—both in terms of the ends we aspire to and the means we have available to us in achieving those ends. But knowledge, according to Pragmatists, is not only a tool that can change behavior and direct us toward better consequences, but also a tool that itself changes and needs to be remade through use and experience. It is in this sense that James (James 1912) urges us not only to be empiricists, but also to become radical empiricists. For the radical empiricist, CSR standards can be useful in guiding action toward desirable outcomes, but they may be even more useful in highlighting new problems and identifying new tensions and tradeoffs. In the former case, we may seek to overcome the CSR paradox; in the latter, we can leverage and build upon it to move forward effectively.

Conclusion: A Few Modest Suggestions that CSR Standards Developers and Users Should Keep in Mind

Paradoxes, paradoxically, are not a cause for pessimism. Actually, by bringing to light what we do not know, even by emphasizing the unknowable, they allow us to learn and cope with complex realities in creative ways. Adam Smith, for example, used the paradox of value (namely that water, one of the most valuable things in the world is priced so much lower than diamonds, one of the least) to point out the fact that real economic value derives from human beings: “The real price of every thing, what every thing really costs to the man who wants to acquire it, is the toil and trouble of acquiring it” (Smith 1776).

In such a spirit, we would like to offer a few modest suggestions to limit the negative effects of the Paradox of CSR standards and, more specifically, to address its three underlying problems. Note that we are not proposing a priori solutions, but simply pointing out fruitful avenues for discussion.

In terms of the general significance of the Paradox, we think that its first implication is that the developers of standards should be better aware of both the positive outcomes and the potential negative counter-effects that the application of CSR standards can generate.

Our critical analysis also suggests that users of standards should adopt a more reflexive and learning-based approach. Such an approach would also enable organizations to make a better use of the incremental improvements that derive from the various forms of assurance processes linked with the implementation of CSR standards. The pragmatist’s perspective, emphasizing the importance of an experimental method, is very appropriate in this regard. For example, it would be interesting to explore a two-stage (recursive) model for CSR standards development, in which two different aspects of CSR standards development can be engaged with separately:

Stage 1:

Focusing on defining the desired scope of a CSR standard, by identifying a set of “general principles—not trying to identify ‘performance’ measures yet; and

Stage 2:

Focusing on the definition of the substantive content of a CSR standard, addressing the ‘problem of deceptive measurements’ as result of a stakeholder engagement and experimentation process.¹²

The paradoxical nature of CSR standards suggests that the development of standards should cover both general principles and (more specific) performance aspects if the development process is to be optimally effective, and that stakeholder engagement should be part of the process. Moving back and forth between the abstract (vague) level of CSR principles to a specific, context-based level of contents (CSR meanings), standards users might avoid some of the problems generating the Paradox, particularly the tendency of developing a thoughtless, quasi-automatic mindset. A face-to-face, community-based stakeholder conversation (again, as already suggested by John Dewey in his The Public and its Problems) seems to be a vital mechanism in avoiding this risk and maintaining a thoughtful approach.

It is also important that CSR standards provide guidance on both substantive and process aspects, if their implementation is to be most effective. This can be seen from the unsatisfactory, if not dangerous, nature of standards covering serious issues such as respect for human rights. It seems inappropriate for an organization to rely only on abstract guidance, in the absence of substantive performance requirements for the organization to respect human rights.

To address P₁ (the problem of deceptive measurements), we argue that it would be useful for CSR standards users to acknowledge the fact of measurement problems in many of CSR’s outcome variables. Therefore those who measure should not only question, but also revise and re-test their measurement approaches in an ongoing, pragmatically empirical process. For example, users should be aware of some of the common measurement errors that CSR standards adopters need to worry about. At any rate, it is important for those attempting to measure CSR outcomes to bear in mind, and constantly question, the value and significance of what they are actually measuring in relation to the outcomes they are seeking to achieve through a standard.

One way to address P₂ (the problem of the erosion of individual responsibility) is to acknowledge that one should not (and cannot) build a system that completely does away with human judgment. As Barrett (1979) shows using the overthrow of the earlier Wittgenstein by the later one, there is a pervasive “Illusion of Technique” that often subverts the systems we build to achieve lofty purposes. When we expend all our energies in building fool-proof systems, we end up de-empowering the best amongst us as well. This is consistent with what William James had in mind when he wrote “the trail of the human serpent is over everything.” CSR standards need to create a relevant place for leverage and build upon human judgment, instead of “leaving it all” to rules and guidance formulated a priori. When and how much judgment and when and how much to leave to a set of rules is an enormously interesting problem.

In terms of this specific root of the Paradox of CSR standards, we suggest that organizations adopting a standard need to ask themselves about the mix of human judgment and structural guidance and compliance rules. It is not possible to formulate a complete set of rules that specify what to leave to judgment, since such an approach will fall foul of a similar paradox to that we have set out above. Such a set of rules would either have to accept its own incompleteness leaving room for judgment, or it will seem complete, but fail to lead to the outcomes it was established to create. The writings of Rorty on the cosmopolitan community offers other insights in this regard: despite the problems of imperfect language translations, people in different cultures can talk and learn from each other; “untranslatable” does not mean “unlearnable” and so “the anthropologist and the native agree, after all, on an enormous number of platitudes. They usually share beliefs about, for example, the desirability of finding waterholes, the danger of fondling poisonous snakes….and so on.” (Rorty 1990, p. 215). In others words we do not need a universal ‘meta-language’ (or a unique, global standard) able to perfectly translate every other language, but rather a way to enable people from different communities and speaking different languages to engage in meaningful conversation and learn from each other, and build together shared meanings (or the shared interpretation of a global standard).

To address P₃ (the problem of the emergence of an inward-oriented mindset, rather than stakeholder-oriented), if we put together insights from the measurement problem with the necessity for human judgment, it becomes clear that the judgments we need to think about are not only those of individuals and of the organization as a whole, but also those of stakeholder groups. Just as no system involving “intentional” beings can ignore individual human judgment, so too we cannot leave to individuals the care of multiple stakeholders. All we know about the bounds of human cognition and the evolution of groups argues for an interlocking system of checks and balances rather than an a priori system of rules that works for everyone at all times—even with the added room for individual human judgment. In other words, for the stakeholder voice to have any power or meaning, each stakeholder group has to have a seat at the decision table (i.e., be able to influence decisions). Having those seats available is part of corporate responsibility, but making sure those seats are filled is as much the responsibility of the stakeholder groups concerned as it is of the corporation in question. After the seminal contribution of Freeman (1984), a number of stakeholder theory scholars have pointed out how stakeholder engagement is not to be seen as an optional ‘add on’ activity—such as might entail charity or other philanthropic action—but as an integrated part of strategic management.¹³ At this point in our understanding of CSR standards, this might simply be an idea worth pursuing.

Finally, in the spirit of the Paradox of the CSR standards, we would like to remind both standards designers and users that an organization will only begin to understand itself (and the meaning of CSR for its business) through understanding its stakeholders. CSR standards can be obstacles to this process if they are developed and implemented uncritically and with an excessive emphasis on compliance. However, we believe that they can be a vehicle not only for organizational self-discovery, but also for improved CSR performance—if approached in a Pragmatic spirit.