Corporate citizenship or corporate social responsibility (CSR) may be defined in terms of proactive efforts by business decision-makers to contribute to sustainable development (WBCSD 2002). Many hope or even expect that such efforts will be decisive. For instance, the United Nations (UN) Global Compact argues, ‘Through responsible business practices, business is making a unique and significant contribution to implementing the Millennium Development Goals’ (MDGs) (www.unglobalcompact.org, accessed 16 September 2005).

Prominent business leaders are no less ambitious and optimistic (WBCSD 2005).

However, in a context of slow progress towards achieving the MDGs, particularly in sub-Saharan Africa (UN 2007), we have little evidence that corporate citizenship efforts are indeed fulfilling their promise. Current development trajectories are clearly unsustainable (UN 2007; see also MEA 2005; IPCC 2007) and a more committed transition is required, characterised by a co-evolutionary process involving many diverse role-players – with a special responsibility for governments – and demanding particular attention to innovation and fairness. ‘Many companies seem more involved in this journey than many governments are’ (Kemp et al. 2005: 23), and companies and the markets they operate in play a crucial role in innovation and the efficient allocation of resources.

Nevertheless, big business, in particular, has been a key beneficiary and protagonist of globalisation, responding to critics by developing the concept and practice of corporate citizenship. Current uncertainties as to whether corporate citizenship efforts are making a difference in the sustainability transition thus strike at the heart of broader debates surrounding globalisation and our quest for a sustainable future. To ‘make a difference’, any effort and its impacts need to be commensurate to the scale of the challenge that is to be addressed (see OECD-DAC 2002: 24 for a definition of ‘impacts’ in this context).

More particularly, corporate citizenship efforts need to:

• Live up to the promises made by their corporate protagonists
• Contribute to a net positive impact associated with corporate activity, at various scales of analysis
• Stand up to scrutiny against agreed benchmarks and frameworks for sustainable development, such as the MDGs and the Johannesburg Plan of Implementation

It is thus not so much the discrete impacts of particular CSR initiatives that are in question, but the degree to which CSR improves the broader business contributions to sustainable development (see also Blowfield 2007: 693).

Developing a better understanding of whether corporate citizenship efforts contribute to broader sustainability objectives is important from both an instrumental perspective and a political economy perspective. Regarding the former, it would help corporate managers and policy-makers to make better decisions regarding strategic emphases, resource allocation and the creation of an enabling framework for more effective efforts (Margolis and Walsh 2003; Chatterji and Levine 2006). Porter and Kramer (2006: 1) argue that existing ‘efforts have not been nearly as productive as they could be’ and that a stronger link between CSR and corporate strategy needs to be based, inter alia, on a more systematic measurement of social impact.

The issue of impacts is also significant considering the controversial nature of corporate citizenship. Critics see it as a measure by big business to emphasise voluntary approaches and to pre-empt stricter government regulation (e.g. Christian Aid 2004). The discourse of corporate citizenship, in this analysis, involves an opportunity cost, because corporate citizenship efforts by business and others, including government and multilateral organisations, may preclude other ways of addressing the negative consequences of economic activity (see also Nelson et al. this edition). Hence, if current corporate citizenship efforts cannot show their promised impact, this increases the stakes not only for increased effort by their protagonists, but also for a renewed consideration of alternative responses and mechanisms. A better, shared understanding of business impacts is thus part of the broader struggle for increased accountability of both the public and the private sectors (see Newell and Wheeler 2006).

Despite these imperatives, our current approaches to assessing corporate citizenship impacts are crucially constrained. For a start, case studies are one of the most common means of assessing or even demonstrating CSR’s developmental impact (Blowfield 2007), but they are limited by their parochial nature in that their findings are difficult to generalise and their focus on a particular scale of analysis generally precludes an assessment of broader, cumulative impacts. In addition, particularly the ‘best practice’ examples in business publications (e.g. WBCSD 2005) often give only a partial picture of the overall sustainability performance of the project or initiative in question (for an illustrative critique of such an example, see Baue 2005).

Over and above case studies, of course, there are a wide variety of codes, ratings and rankings that purport to assess companies’ corporate citizenship efforts, and these are undertaken both within and outside of firms, for diverse audiences (see Table 1). However, rather than providing clear indications regarding the impact of corporate citizenship, ‘the proliferation of non-financial performance metrics and associated surveys’ (Chatterji and Levine 2006: 30) contributes to continued uncertainty, because many of them are plagued by low reliability of data (due to non-response bias and other factors), low validity of measurement constructs (because they do not relate directly to performance issues that are important to society) and lack of comparability between methodologies.

Poor performers have incentives to invent and adopt unreliable, invalid, and non-comparable standards because stakeholders will find it difficult to differentiate which standards are valid . . . each additional certification and corresponding acronym can actually decrease overall welfare, even while increasing the amount of measurement (and resulting costs) (Chatterji and Levine 2006: 31).

The academic literature has also not given systematic attention to the impacts of corporate citizenship. In particular, organisational scholarship has been predominantly occupied with the so-called business case, or the impacts of such efforts on the firm’s financial performance. Margolis and Walsh (2003) review 127 such studies focused on the relationship between corporate social performance (CSP) and corporate financial performance. The majority of these studies evaluate CSP in terms of reputation ratings and companies’ disclosure practices, while some use evaluations by ranking agencies such as KLD, which face the methodological problems considered above with regard to reliability, validity and comparability (for instance, ‘KLD uses largely qualitative and subjective measures, which make it difficult to produce comparable and reliable metrics’ [Chatterji and Levine 2006: 44]).

In terms of Wood’s (1991) categories of the outcomes of corporate behaviour, it is apparent that both the predominant ratings and the management literature focus more on policies and programmes than on social impacts (see also Margolis and Walsh 2003). A similar assessment is possible using the categories used in the international development literature (see Rieth et al. this edition): there is an emphasis on interventions’ outcomes (such as organisational commitments) and outputs (e.g. changes in behaviour), instead of impacts (or goal attainment). It has been left largely to scholars in other fields, especially development studies, to focus on actual impacts on stakeholders, particularly the poor. ‘What CSR seeks to measure is significantly different from the focus on changes in people’s lives that international development is concerned with’ (Blowfield 2007: 684; see also Prieto-Carron et al. 2006).

This special edition is an initial attempt to respond to this gap in the literature, in management practice and in the policy debate. Most of its articles are drawn from an eponymous symposium hosted by the UN Global Compact, the Unisa Centre for Corporate Citizenship and the Lindenberg Center for Global Citizenship, which took place at the Ghana Institute of Management and Public Administration in Accra, Ghana, 21-22 November 2006. The next section of this introductory article presents an analytical framework to consider the impacts of corporate citizenship at different, interrelated scales of analysis. This is followed by an overview of some existing efforts at impact assessment – including the contributions in this special edition – and resulting conclusions.

Table 1: Examples of different approaches to impact assessments and reporting, adapted from Gray et al (1997: 331)

Reports compiled by…

A simplified model of business impact on sustainable development

One of the likely reasons why we know relatively little about the impact of corporate citizenship and business operations on sustainable development is the inherent complexity of this relationship. This is due to the wide range and different scales of business activities and the diverse interrelationships between these and multi-layered socio-economic and natural systems, which themselves are interrelated in complex ways. In the language of complex systems theory, these various interrelationships involve context-specific and path-dependent feedback loops, ‘unpredictability, irreversibility and co-evolution’ (Urry 2005: 11). Figure 1 is a schematic illustration of a simple model representing these relationships, though the underlying complexity cannot be ignored. It is based on the initial, tentative distinction between initiatives that have an impact (‘impact of what?’) – on the left-hand side of the diagram – and those aspects of the social and environmental systems that are impacted upon (‘impact on what?’) – on the right-hand side (see also Gitsham this edition for an elaboration on this distinction).

A premise of this model is that the ‘coalface’ of business impacts is at the local level.

This is where people’s lives – as workers, neighbouring community members or consumers – are affected most directly by business activity, in particular companies’ local operations and related CSR initiatives. These local business activities are nested within broader scales of analysis, the first being the relevant companies’ policies and practices. Using the examples provided in Figure 1, for instance, the interrelationship between the Mtubatuba plantation in KwaZulu-Natal, South Africa, and its workers and surrounding communities is influenced by the parent company’s policies and management systems, which in this scenario are those of the pulp and paper company Mondi. The company’s policies and practices are informed by national initiatives, such as the South African National Business Initiative (NBI), as well as sector-specific initiatives, such as the Forest Stewardship Council (FSC). These, in turn, operate in the context (though possibly quite independently) of global, universal efforts, such as the United Nations Global Compact (UNGC).

In terms of the right-hand side of the diagram, there are of course a multitude of stakeholders and processes that are impacted upon by business activity. One of the most prominent frameworks for making sense of these impacts has been the stakeholder model (Clarkson 1995). However, based on the above-mentioned definition of corporate citizenship in terms of the business contribution to sustainable development, Figure 1 suggests a complementary alternative to the stakeholder model. It emphasises three different scales of analysis premised on existing efforts to develop indicators assessing progress in a sustainability transition (for an overview, see Pinter et al. 2005). At the local level, a prominent framework for assessment has been the sustainable livelihoods approach (Bury 2001, 2004; Barney 2003, WBCSD 2005), which emphasises the resources – commonly categorised in terms of productive, human, natural and social capitals – available to households to construct livelihoods for their members.

Local livelihoods and the local governance systems in which they are constructed
(Hamann et al. 2005) exist within, influence, and are in turn influenced by, the National economic and political context, for which numerous high-profile indicator sets exist. Examples include the MDGs, the Human Development Index (UNDP 2006), governance indicators (Kaufmann et al. 2007) and environmental indicators (Esty et al. 2005). Local impacts of business activity are not only accumulated to give rise to the conditions assessed by such national indicators; business activity also has crucial impacts directly on national economic, social and governance systems through, for instance, its influence on consumption patterns, the payment of wages and taxes, and political lobbying. These local and national scales of impact are nested within the global dimension, which, again, encompasses cumulative impacts from lower scales, as well as more direct impacts of business activity in the form of, say, greenhouse gas emissions and lobbying at the World Trade Organisation.

It may be argued that the model in Figure 1 – especially its emphasis on local-level
Impacts – is less relevant for companies in the tertiary sector, such as banks or retailers, than those in the primary and secondary sectors, such as mining or manufacturing firms. Particularly with the increasing emphasis on companies’ supply chains, however, even companies in the services sector need to pay attention to their indirect impacts at the local level. Furthermore, this model is just one of diverse ways to simplify a very complex set of interrelationships; alternative examples include WBCSD 2007, which focuses on a number of identified elements of impact such as ‘jobs’ or ‘human rights’.

The model is nevertheless a useful point of departure for the following overview of illustrative examples of existing efforts at assessing the development impact of corporate citizenship related initiatives. The overview suggests three overarching categories of existing efforts: those focusing on local operations or projects, those using the company as the unit of analysis, and those assessing codes of conduct and multi-stakeholder partnerships. That is, they are represented most clearly by the left-hand side of the model in Figure 1. In most cases, existing evaluation efforts focus on policies and programmes, rather than impacts. In instances where impacts are considered, this is predominantly at the local level; there is a marked absence of efforts to assess actual impacts on aggregated indicators at the national or international level.

Figure 1: Schematic illustration of the nested nature of different scales of business impacts on sustainable development (see text for abbreviations).

An overview of some current efforts at assessing impacts

The local ‘coalface’
The process of ex ante identification of business impacts at the local level is relatively well developed on the basis of environmental and social impact assessments (ESIAs), particularly in the case of large-scale infrastructure or resource extraction projects. But, because the relevant legislative standards and especially their enforcement are weak in many countries, many multinational companies commit to ‘international best practice’, represented, for instance, by the guidelines and requirements of the International Finance Corporation (IFC) and the associated Equator Principles (www.equator-principles.com, accessed 26 September 2007). These require companies to comply with and implement a number of social and environmental performance standards, including dedicated monitoring and reporting (IFC 2006a).

The IFC’s efforts are also noteworthy with regard to those of its Independent Evaluation Group (IEG), which – in line with development agencies’ emphasis on results-based management’ (OECD-DAC 2000) – assesses IFC-supported projects in terms of financial, economic, social and environmental criteria, as well as so-called private sector development impacts, all of which are synthesised in a ‘development outcome’ rating (IFC 2006b). The IEG found, for instance, that for a random sample of 210 operations approved between 1997 and 1999 there was a 59% success rate in terms of development outcome (compared with a 44% success rate in terms of financial performance). Methodologically, the IEG’s evaluations are limited in that they rely on the IFC’s data and include little participatory research (see below).

Private sector activities that are outside the purview of organisations such as the IFC, of course, are likely to face fewer monitoring requirements. Particularly in developing countries where state monitoring of social and environmental standards is often limited, there is often a lack of will or capacity among companies to devote sufficient attention to continuous or ex post monitoring and evaluation of development impacts. A lot depends on the particular proclivity, skill and influence of the relevant personnel on site. Furthermore, ESIAs and their corresponding management and monitoring systems still face important limitations, despite their adaptation of frameworks such as sustainable livelihoods (see, for instance, IFC 2006b).

Methodological challenges are evident also in independent, third-party evaluations at the local level. For instance, Bury (2004) applies the sustainable livelihoods model to assess the impacts of Newmont’s Minera Yanacocha operations in Peru. These have increased households’ access to produced and human capitals, but decreased their access to natural and social capitals. However, the livelihood approach ‘says little, at present, about how …to synthesise these findings into a more parsimonious conclusion that explains whether household livelihoods are better or worse off since [the company] began operations in the region’ (Bury 2004: 89). This challenge relates to a broader, fundamental debate with regard to the substitutability of different forms of capital, particularly in the ecological economics literature (Pearce and Atkinson 1993).

In this context, recent progress in human rights impact assessment is particularly promising. BP was the first to publish such an assessment by independent analysts (Smith and Freeman 2002; for the company’s response see BP 2003), though John Ruggie, Special Representative to the UN Secretary-General on business and human rights, bemoans that ‘relatively few firms conduct these assessments routinely’ (quoted in Baue 2007). While the BP example was an ex ante assessment, the value of human rights impact assessments from an ex post perspective is illustrated by case studies profiled in a report by Rights & Democracy (2007), which argues that, in each of the four ex post evaluations, foreign investment projects have had negative impacts on human rights. Over and above methodological challenges, problems with impact evaluations conducted by companies themselves are often compounded by the company management’s emphasis on those impacts that are of particular value to its perceived priority interests. Instead of measuring impacts on beneficiaries in a broader sense, company-contrived indicators are emphasised, often inspired by tangible project timetables. In this vein, a local manager in one of South Africa’s platinum mining regions noted (Personal communication with Bandile Gcabo, Rustenburg, South Africa, 17 April 2003 (Bandile Gcabo was community relations manager for Lonmin Platinum in the Rustenburg area).

It’s one thing to set targets and hope all goes well, but involving all stakeholders for a sustainable solution takes more time than we are often prepared to give …The easy way is the wrong way. If you want to build a road you can just bulldoze your way to building a road, and then you can show your road for it [in terms of measurable deliverables]; but involving government and other stakeholders in building the road is more difficult, but more sustainable in the long term.

In this edition, Ricarda McFalls’s case study of a Hewlett Packard (HP) project testing the notion of the ‘bottom of the pyramid’ similarly attests to a ‘dark side’ of impact assessment: ‘The project team was under pressure to deliver fast, visible results at the risk of speed overriding substance’.

Impact evaluations are particularly problematic in conflict zones, in which the impact of companies is often the subject – and sometimes even the cause – of significant acrimony. In areas such as the Niger Delta in Nigeria, the impact of some of the most sophisticated and well-resourced corporate citizenship initiatives remains open to question (for opposing assessments in the Nigerian setting, see, on the one hand, Shell 2007 and, on the other, FOE 2004 and Boele et al. 2001). Such circumstances also aggravate the possibility of unintended negative consequences, whereby, for instance, CSR activities can contribute to further conflict (Akpan 2006).

Gitsham’s contribution to this edition shows how different observers and analysts arrive at very different appraisals of BP’s impacts in the area of its Colombian operations.  In such circumstances, the unilateral definition of success measures by the company is unlikely to contribute to conflict resolution, the building of trust and the establishment of collaborative responses to local challenges. Rather, it is likely that the identification of impact indicators may require participative, collaborative approaches with the involvement of all relevant role-players, especially the local poor, though of course this is especially difficult in conflict situations.

There have been important lessons learned with regard to the development of local sustainability indicators by means of participatory processes. Nurick and Johnson (1998) discuss the development of quality of life and corresponding industry impact indicators in the case of south Durban in South Africa, illustrating the depth and context-specific nature such indicators adopt on the basis of careful community participation. Based on subsequent research in the same area, Acutt et al. (2004) show how companies’ failure to meaningfully engage local communities in their impact appraisals has led to a breakdown of trust and thus a failure of the resulting mitigation plans.

Company-level evaluations
The company is the most common level of analysis for organisational scholars and the growing socially responsible investment industry, including a wide array of ratings and indices, such as the Dow Jones Sustainability Index. Most of these approaches, as noted above, are characterised by an emphasis on company policies and practices, rather than impacts on particular stakeholders or objectives. Referring to Chatterji and Levine’s (2006) overview of codes and ratings, Porter and Kramer (2006) bemoan the lack of consistency and rigour in these ratings’ criteria and their weightings, as well as their emphasis on ‘measures for which data are readily available, even though they may not be good proxies for the social or environmental effects they are intended to reflect’ (Porter and Kramer 2006: 3). Coupled to a reliance on unreliable data, this leads to ‘a jumble of largely meaningless rankings, allowing almost any company to boast that it meets some measure of social responsibility – and most do’ (Porter and Kramer 2006: 3). Attempts to construct more rigorous ‘composite sustainability indicators’ at the company level include Krajnc and Glavic 2005.

From a company perspective, the most prominent, systematic mechanism for publicly reporting on sustainable development impacts is, of course, sustainability reporting. Though this practice has come a long way from the initial discussions in the 1970s on social accounting (Estes 1976) and social audits (Medawar 1976) to the current prominence of the Global Reporting Initiative indicators (GRI 2006), most companies’ reports are still not very effective in explaining actual impacts on stakeholders, especially in holistic terms for the company as a whole.

There are three overarching concerns in this regard. The first challenge for the GRI and reporting companies has been the development and application of indicators that are relevant universally, while at the same time pertaining to local contexts that are very diverse and complex. Important recent efforts, therefore, attempt to combine the use of ‘top-down’ indicator systems, foremost among which is that of the GRI, with ‘bottom-up’ systems developed on the basis of participatory methods involving, in particular, local stakeholders that are most directly affected (Chamaret et al. 2007). As far as company initiatives in this regard are concerned, one of the most prominent efforts is the Tata Index for Sustainable Human Communities, which, inter alia, encourages all Tata employees to interact with [local] communities and to jointly create measurable markers’ for the company’s social investments (Branzei and Rao 2007: 11).

The limitation of corporate sustainability reports also relates to their focus on the corporate level of analysis, whereby cumulative impacts of different companies and industries at the local, national or international level are not subject to systematic analysis (see also Blowfield 2007). It remains to be seen whether companies will respond comprehensively to the third-generation GRI principle of reporting within the sustainability context (GRI 2006), but this is unlikely to be successful without greater commitment to regional and sectoral sustainability assessments (e.g. Pope et al. 2004), in which governments will need to play a leading role.

More fundamental critiques of corporate sustainability reports involve claims of ‘greenwashing’ (Bruno and Karliner 2002; Laufer 2003; FOE 2004), or the intentional provision of disinformation for the purpose of image management. Despite the advances made through the GRI, many NGOs continue to view corporate sustainability reports negatively with regard to their ‘credibility and sufficiency’ (O’Dwyer et al. 2005). When companies’ motivations for sustainability reporting are more to do with maintaining legitimacy and controlling stakeholders than discharging accountability to stakeholders and curbing unsustainable trends, this does not only have detrimental effects on the credibility of the corporate citizenship movement. It may also have perverse side effects: ‘The purchase of the “commodity of compliance” sufficient to shift the risk of liability and loss, in certain firms, may result in decreased levels of care by senior managers’ (Laufer 2003: 257).

To respond to these challenges, Laufer (2003: 259) argues for ‘Tripartism, or the integration of a third party into the regulatory arena occupied by an organisation and a regulator’. To some extent, the increasing scrutiny of corporate activity by a global network of NGOs, linked to local activists and making effective use of the media and the Internet, arguably approaches the introduction of a third party in the regulatory arena (see Spar 1998). Yet there is significant scope to enhance this involvement of civil society in company-level impact appraisals, not only – as noted above – to increase local legitimacy of the indicators used, but also to increase accountability and trust. One of the most striking examples of such an approach was a study on the impacts of Unilever on poverty in Indonesia conducted jointly by the company and Oxfam (Clay 2005). This study is also remarkable for its comprehensive analysis of impacts ranging from poor consumers to the country’s macro-economy, illustrating the wide-ranging impacts in the company’s value chain. In his foreword, the company’s chief executive declared this to be a ‘learning partnership which demonstrated how much insight can be gained by working together’ (Cescau foreword in Clay 2005: 11).

In this edition, Scott Marshall and his colleagues argue on the basis of a quantitative study that negotiation and dialogue, rather than confrontational strategies, between firms and stakeholder groups (in this instance, NGOs and institutional investors) has led to notable increases in the quality of disclosure of environmental information. This underscores Blowfield’s (2007) view that the most significant effect of the corporate citizenship movement has been widespread organisational change, manifested in this instance by improved disclosure.

A less confrontational approach by NGOs is also advocated by Klaus Leisinger in his invited contribution to this edition. Arguing from a corporate perspective, he notes that engaging stakeholders is a good investment ‘even if the process is sometimes a difficult one’, but bemoans some NGOs’ ‘campaign strategies that seek to undermine corporate integrity in the interest of sensationalism…[and which may result] in companies falling back to a legalistic, compliance-oriented approach’. Indeed, one of Oxfam’s lessons from the above-mentioned Unilever study was that ‘our analysis needs to be more alert to the differences between multinational companies’ (Clay 2005: 21). At the same time, however, the significant impact and widespread appreciation for this study was probably due in large part to what Covey and Brown (2001) would characterise as Oxfam’s ‘critical cooperation’ approach.

Codes of conduct and cross-sector collaboration initiatives
The crucial role of participatory and partnership approaches in assessing corporate citizenship efforts has also been documented in the implementation of codes of conduct, especially those related to labour practices and impacts on workers. Barrientos (2005) emphasises the manner in which participation by workers, trade unions and NGOs in the implementation and monitoring of the code of conduct can facilitate a continuous ‘learning approach to impact assessment’, which iteratively improves workers’ quality of life. She stresses the need to go beyond commonly emphasised indicators of compliance to ‘examine whether a particular change in employment conditions due to a code has had a positive or negative effect on these broader dimensions of workers’ well-being and poverty’ (Barrientos 2005: 266; see also Prieto-Carron et al. 2006: 982).

In this edition, Nelson and colleagues report on a longitudinal, comparative study of the livelihood impact of codes of conduct in the Kenyan cut flower and South African wine industries. They find that workers are generally better off in companies that had adopted codes in comparison with companies that did not. However, attribution remains a vexing challenge in that it is difficult to assess whether these differences are indeed due to the codes. They also argue that the relatively modest economic benefits that may be accruing to workers are fundamentally limited due to these workers’ lack of power in the broader supply chain, and that development agencies’ focus should perhaps shift to alternative strategies for improving local livelihoods.

Some sort of cross-sector collaboration is a feature of most codes and evaluation processes, and indeed it plays a role in most corporate citizenship strategies and initiatives at all scales of analysis. Such partnerships take a variety of forms and have diverse objectives and governance structures. During the World Summit on Sustainable Development held in Johannesburg in 2002, so-called Type 2 initiatives were touted as a significant contributor to sustainable development, despite vocal concerns. The most recent official UN overview of these initiatives focuses on ‘progress in implementation within the broad categories of partnership building/coordination, capacity-building activities, information sharing and pilot projects’ (UN 2006: 1), but it provides hardly any information on performance as measured against the wider sustainability objectives they are meant to achieve. Hence, five years later, the impact of such initiatives remains uncertain and arguably the expectations raised during the summit have not been realised (for empirical studies of such initiatives that now require updating, see Hale and Mauzerall 2004; AICC 2004).

This is not to say that collaboration initiatives cannot have important benefits that are greater than unilateral efforts. Such benefits are especially tangible in the case of locally driven partnerships’, though these ‘should focus more on measuring results’ (Steets 2006: 6). In this edition, Muthuri highlights how a collaborative approach allows the case study firm to motivate and enable community members to take a more proactive, self-reliant approach to local development. On the other hand, McFalls’s case study provides a more critical view of a public–private partnership, noting that HP’s approach to this particular initiative did not allow for meaningful participation by local beneficiaries in its planning or implementation, and even the government partners were not involved effectively. This propensity for companies to dominate ostensible partnerships is underscored by Ashman (2001), who – like Nelson et al. in this edition – emphasises that the limited economic gains that local citizens in poor countries may get from partnerships should not detract from the necessary political empowerment that would ensure more sustainable livelihoods.

The complexity of assessing the impact of collaboration initiatives is of course heightened at greater scales of analysis; this often leads to an even greater emphasis on interventions’ outputs and outcomes, rather than impacts. This is perhaps most evident in the opposing assessments of one of the most prominent corporate citizenship initiatives, the UN Global Compact, a key criticism of which has been that ‘there is no assessment mechanism (just corporate self-reports)’ (Chatterji and Levine 2006: 32). In this edition, Leisinger and Rieth et al. provide a positive assessment of the Compact’s potential.

The latter use case studies from southern Africa to emphasise the role that a proactive decentralisation strategy can play in embedding its principles and making use of existing social capital in local networks of companies and other stakeholders in developing countries. They note that the Global Compact can play a particularly important role in developing countries such as Malawi and Mozambique, where it is often the first systematic introduction to CSR issues for local role-players and where it can provide a crucial convening platform for cross-sector collaboration. But it is also in these countries where the limit of voluntary initiatives is apparent, as company representatives themselves emphasise the need for greater government regulation, enforcement and monitoring on social and environmental issues.

Conclusions

Despite the prominent expectations that business can make significant contributions to sustainable development, we do not know whether current corporate citizenship initiatives are, indeed, making a difference. Selective examples to demonstrate development impact (WBCSD 2005) or even case studies by independent observers do little to convince sceptics of the broader benefit of corporate citizenship for the poor or the environment, especially when critical counter-examples or interpretations abound. Even the surveys, ratings and rankings undertaken by organisational scholars (see Margolis and Walsh 2003) or investment rating agencies (see Chatterji and Levine 2006) are prone to being constrained by unreliable data, invalid metrics or incomparable methodologies. Furthermore, most of these assessment efforts focus on corporate policies and practices, rather than impacts. So, while the best indication we have is that the corporate citizenship movement has led to widespread organisational change (Blowfield 2007), its impact on sustainable development remains unclear.

This article has suggested that this uncertainty is not primarily due to a lack of effort. Nor is it only due to a lack of suitable metrics or other technical challenges, though some methodological advancement is no doubt necessary to deal with the complexity discussed in this article. Such advancements are particularly important in developing a better understanding of business impacts at regional, national and international scales of analysis, involving accumulated local impacts, as well as direct higher-order impacts resulting from, for instance, tax payments and political lobbying.

Over and above such methodological challenges, a crucial yet rarely considered role is being played by the underlying motives, power relations and value systems informing current approaches to impact assessment, perpetuating conflicting appraisals of the impacts of business activity. Put differently, the broader political economy not only plays a role in underpinning and mediating the socioeconomic and environmental impacts of business activity, but it also constrains our knowledge of these impacts. This is also illustrated by the various instances of the ‘dark side’ of impact assessment discussed in this article, whereby legitimacy- or image-driven approaches to impact assessment and standard setting have perverse, negative consequences for organisational performance and social welfare (Laufer 2003; Chatterji and Levine 2006; McFalls this edition). At a more basic level, this requires greater attention to the issue of who pays for assessing business impacts.

Taking into account both the methodological and political economy aspects considered in this article, the following two considerations may help companies and other roleplayers – with special responsibilities for governments – develop a better, shared understanding of business impacts on sustainable development.

Greater emphasis on assessing business impacts on the poor and their human rights, based on the fairness principle. This is an important response to the complexity of business–society interrelationships discussed in this article, because it focuses the objectives and methods of impact assessments. It is also a crucial application of Rawls’s fairness principle that ‘an inequality is allowed only if…it will work for the advantage of every party engaging in it’ (Rawls 1958: 167, original emphasis). By emphasising that the worst-off should not experience any further deterioration of their livelihoods, it helps navigate the difficult trade-offs between different kinds of business impact. It can also help to bridge some of the different value systems that are currently contributing to such contradictory appraisals of the role of business.

• An emphasis on participatory appraisal and research partnerships:  One of the overarching conclusions of the above discussion is that unilateral efforts by companies to assess their impacts on development are likely to lack crucial information and legitimacy. Hence participatory methods that involve, say, local household members in a livelihoods assessment or the development and monitoring of community-based indicators ought to play a more prominent role. Partnerships between companies and NGOs that are known to be independent and capable of informed criticism (as in Clay 2005) provide not only an important learning opportunity arising from the exchange between different perspectives, but also significant legitimacy in a context in which corporate reports are distrusted by many civil society organisations
• Finally, the research community will also need to be involved more systematically in impact assessment efforts:  Considering the multifaceted complexity of the business-society interrelationship, such research partnerships ought to involve multi-disciplinary research teams applying both quantitative and qualitative methods. The researchers will need to be able to maintain and demonstrate their independence, contributing to the assessment not only methodological research skills and techniques but also an attitude of ‘critical thinking’ (Utting and Zammit 2006: 38). For such partnerships between business and independent researchers to be authentic, the company or companies concerned will need to be able to surrender some power over the process and its outcome.

This willingness currently seems in short supply.

This special edition of the Journal of Corporate Citizenship is the result of a symposium I helped organise in 2006 as head of research at the Centre for Corporate Citizenship at the University of South Africa. I am grateful to those individuals and organisations that made the edition possible and the symposium on which it is based. This includes the conveners of the symposium, the Unisa Centre for Corporate Citizenship, the UN Global Compact, and the Lindenberg Centre for Global Citizenship; the hosts, the Ghana Institute of Management and Public Administration; the symposium committee members, Manuel Escudero, Sanjeev Khagram, Franklyn Manu, Jeremy Moon, Maggie Opondo and Dante Pesce; and – most importantly – the 40 participants from 16 countries and four continents. Crucial support in compiling the special edition was provided by the co-editors, Maggie Opondo and Sanjeev Khagram; the anonymous referees; as well as Mara Brain and Catrina Lucero. This introductory article has benefited from comments on an earlier draft by Nicky Black, Malcolm McIntosh and Fleur Boulogne, and it was prepared with support from the University of Cape Town Environmental Evaluation Unit and TrustAfrica.)

– Picture courtesy of The Co-operative

Corporate citizenship or corporate social responsibility (CSR) may be defined in terms of proactive efforts by business decision-makers to contribute to sustainable development (WBCSD 2002). Many hope or even expect that such efforts will be decisive. For instance, the United Nations (UN) Global Compact argues, ‘Through responsible business practices, business is making a unique and significant contribution to implementing the Millennium Development Goals’ (MDGs) (www.unglobalcompact.org, accessed 16 September 2005).

Prominent business leaders are no less ambitious and optimistic (WBCSD 2005).

However, in a context of slow progress towards achieving the MDGs, particularly in sub-Saharan Africa (UN 2007), we have little evidence that corporate citizenship efforts are indeed fulfilling their promise. Current development trajectories are clearly unsustainable (UN 2007; see also MEA 2005; IPCC 2007) and a more committed transition is required, characterised by a co-evolutionary process involving many diverse role-players – with a special responsibility for governments – and demanding particular attention to innovation and fairness. ‘Many companies seem more involved in this journey than many governments are’ (Kemp et al. 2005: 23), and companies and the markets they operate in play a crucial role in innovation and the efficient allocation of resources.

Nevertheless, big business, in particular, has been a key beneficiary and protagonist of globalisation, responding to critics by developing the concept and practice of corporate citizenship. Current uncertainties as to whether corporate citizenship efforts are making a difference in the sustainability transition thus strike at the heart of broader debates surrounding globalisation and our quest for a sustainable future. To ‘make a difference’, any effort and its impacts need to be commensurate to the scale of the challenge that is to be addressed (see OECD-DAC 2002: 24 for a definition of ‘impacts’ in this context).

More particularly, corporate citizenship efforts need to:

• Live up to the promises made by their corporate protagonists
• Contribute to a net positive impact associated with corporate activity, at various scales of analysis
• Stand up to scrutiny against agreed benchmarks and frameworks for sustainable development, such as the MDGs and the Johannesburg Plan of Implementation

It is thus not so much the discrete impacts of particular CSR initiatives that are in question, but the degree to which CSR improves the broader business contributions to sustainable development (see also Blowfield 2007: 693).

Developing a better understanding of whether corporate citizenship efforts contribute to broader sustainability objectives is important from both an instrumental perspective and a political economy perspective. Regarding the former, it would help corporate managers and policy-makers to make better decisions regarding strategic emphases, resource allocation and the creation of an enabling framework for more effective efforts (Margolis and Walsh 2003; Chatterji and Levine 2006). Porter and Kramer (2006: 1) argue that existing ‘efforts have not been nearly as productive as they could be’ and that a stronger link between CSR and corporate strategy needs to be based, inter alia, on a more systematic measurement of social impact.

The issue of impacts is also significant considering the controversial nature of corporate citizenship. Critics see it as a measure by big business to emphasise voluntary approaches and to pre-empt stricter government regulation (e.g. Christian Aid 2004). The discourse of corporate citizenship, in this analysis, involves an opportunity cost, because corporate citizenship efforts by business and others, including government and multilateral organisations, may preclude other ways of addressing the negative consequences of economic activity (see also Nelson et al. this edition). Hence, if current corporate citizenship efforts cannot show their promised impact, this increases the stakes not only for increased effort by their protagonists, but also for a renewed consideration of alternative responses and mechanisms. A better, shared understanding of business impacts is thus part of the broader struggle for increased accountability of both the public and the private sectors (see Newell and Wheeler 2006).

Despite these imperatives, our current approaches to assessing corporate citizenship impacts are crucially constrained. For a start, case studies are one of the most common means of assessing or even demonstrating CSR’s developmental impact (Blowfield 2007), but they are limited by their parochial nature in that their findings are difficult to generalise and their focus on a particular scale of analysis generally precludes an assessment of broader, cumulative impacts. In addition, particularly the ‘best practice’ examples in business publications (e.g. WBCSD 2005) often give only a partial picture of the overall sustainability performance of the project or initiative in question (for an illustrative critique of such an example, see Baue 2005).

Over and above case studies, of course, there are a wide variety of codes, ratings and rankings that purport to assess companies’ corporate citizenship efforts, and these are undertaken both within and outside of firms, for diverse audiences (see Table 1). However, rather than providing clear indications regarding the impact of corporate citizenship, ‘the proliferation of non-financial performance metrics and associated surveys’ (Chatterji and Levine 2006: 30) contributes to continued uncertainty, because many of them are plagued by low reliability of data (due to non-response bias and other factors), low validity of measurement constructs (because they do not relate directly to performance issues that are important to society) and lack of comparability between methodologies.

Poor performers have incentives to invent and adopt unreliable, invalid, and non-comparable standards because stakeholders will find it difficult to differentiate which standards are valid . . . each additional certification and corresponding acronym can actually decrease overall welfare, even while increasing the amount of measurement (and resulting costs) (Chatterji and Levine 2006: 31).

The academic literature has also not given systematic attention to the impacts of corporate citizenship. In particular, organisational scholarship has been predominantly occupied with the so-called business case, or the impacts of such efforts on the firm’s financial performance. Margolis and Walsh (2003) review 127 such studies focused on the relationship between corporate social performance (CSP) and corporate financial performance. The majority of these studies evaluate CSP in terms of reputation ratings and companies’ disclosure practices, while some use evaluations by ranking agencies such as KLD, which face the methodological problems considered above with regard to reliability, validity and comparability (for instance, ‘KLD uses largely qualitative and subjective measures, which make it difficult to produce comparable and reliable metrics’ [Chatterji and Levine 2006: 44]).

In terms of Wood’s (1991) categories of the outcomes of corporate behaviour, it is apparent that both the predominant ratings and the management literature focus more on policies and programmes than on social impacts (see also Margolis and Walsh 2003). A similar assessment is possible using the categories used in the international development literature (see Rieth et al. this edition): there is an emphasis on interventions’ outcomes (such as organisational commitments) and outputs (e.g. changes in behaviour), instead of impacts (or goal attainment). It has been left largely to scholars in other fields, especially development studies, to focus on actual impacts on stakeholders, particularly the poor. ‘What CSR seeks to measure is significantly different from the focus on changes in people’s lives that international development is concerned with’ (Blowfield 2007: 684; see also Prieto-Carron et al. 2006).

This special edition is an initial attempt to respond to this gap in the literature, in management practice and in the policy debate. Most of its articles are drawn from an eponymous symposium hosted by the UN Global Compact, the Unisa Centre for Corporate Citizenship and the Lindenberg Center for Global Citizenship, which took place at the Ghana Institute of Management and Public Administration in Accra, Ghana, 21-22 November 2006. The next section of this introductory article presents an analytical framework to consider the impacts of corporate citizenship at different, interrelated scales of analysis. This is followed by an overview of some existing efforts at impact assessment – including the contributions in this special edition – and resulting conclusions.

Table 1: Examples of different approaches to impact assessments and reporting, adapted from Gray et al (1997: 331)

Reports compiled by…

A simplified model of business impact on sustainable development

One of the likely reasons why we know relatively little about the impact of corporate citizenship and business operations on sustainable development is the inherent complexity of this relationship. This is due to the wide range and different scales of business activities and the diverse interrelationships between these and multi-layered socio-economic and natural systems, which themselves are interrelated in complex ways. In the language of complex systems theory, these various interrelationships involve context-specific and path-dependent feedback loops, ‘unpredictability, irreversibility and co-evolution’ (Urry 2005: 11). Figure 1 is a schematic illustration of a simple model representing these relationships, though the underlying complexity cannot be ignored. It is based on the initial, tentative distinction between initiatives that have an impact (‘impact of what?’) – on the left-hand side of the diagram – and those aspects of the social and environmental systems that are impacted upon (‘impact on what?’) – on the right-hand side (see also Gitsham this edition for an elaboration on this distinction).

A premise of this model is that the ‘coalface’ of business impacts is at the local level.

This is where people’s lives – as workers, neighbouring community members or consumers – are affected most directly by business activity, in particular companies’ local operations and related CSR initiatives. These local business activities are nested within broader scales of analysis, the first being the relevant companies’ policies and practices. Using the examples provided in Figure 1, for instance, the interrelationship between the Mtubatuba plantation in KwaZulu-Natal, South Africa, and its workers and surrounding communities is influenced by the parent company’s policies and management systems, which in this scenario are those of the pulp and paper company Mondi. The company’s policies and practices are informed by national initiatives, such as the South African National Business Initiative (NBI), as well as sector-specific initiatives, such as the Forest Stewardship Council (FSC). These, in turn, operate in the context (though possibly quite independently) of global, universal efforts, such as the United Nations Global Compact (UNGC).

In terms of the right-hand side of the diagram, there are of course a multitude of stakeholders and processes that are impacted upon by business activity. One of the most prominent frameworks for making sense of these impacts has been the stakeholder model (Clarkson 1995). However, based on the above-mentioned definition of corporate citizenship in terms of the business contribution to sustainable development, Figure 1 suggests a complementary alternative to the stakeholder model. It emphasises three different scales of analysis premised on existing efforts to develop indicators assessing progress in a sustainability transition (for an overview, see Pinter et al. 2005). At the local level, a prominent framework for assessment has been the sustainable livelihoods approach (Bury 2001, 2004; Barney 2003, WBCSD 2005), which emphasises the resources – commonly categorised in terms of productive, human, natural and social capitals – available to households to construct livelihoods for their members.

Local livelihoods and the local governance systems in which they are constructed
(Hamann et al. 2005) exist within, influence, and are in turn influenced by, the National economic and political context, for which numerous high-profile indicator sets exist. Examples include the MDGs, the Human Development Index (UNDP 2006), governance indicators (Kaufmann et al. 2007) and environmental indicators (Esty et al. 2005). Local impacts of business activity are not only accumulated to give rise to the conditions assessed by such national indicators; business activity also has crucial impacts directly on national economic, social and governance systems through, for instance, its influence on consumption patterns, the payment of wages and taxes, and political lobbying. These local and national scales of impact are nested within the global dimension, which, again, encompasses cumulative impacts from lower scales, as well as more direct impacts of business activity in the form of, say, greenhouse gas emissions and lobbying at the World Trade Organisation.

It may be argued that the model in Figure 1 – especially its emphasis on local-level
Impacts – is less relevant for companies in the tertiary sector, such as banks or retailers, than those in the primary and secondary sectors, such as mining or manufacturing firms. Particularly with the increasing emphasis on companies’ supply chains, however, even companies in the services sector need to pay attention to their indirect impacts at the local level. Furthermore, this model is just one of diverse ways to simplify a very complex set of interrelationships; alternative examples include WBCSD 2007, which focuses on a number of identified elements of impact such as ‘jobs’ or ‘human rights’.

The model is nevertheless a useful point of departure for the following overview of illustrative examples of existing efforts at assessing the development impact of corporate citizenship related initiatives. The overview suggests three overarching categories of existing efforts: those focusing on local operations or projects, those using the company as the unit of analysis, and those assessing codes of conduct and multi-stakeholder partnerships. That is, they are represented most clearly by the left-hand side of the model in Figure 1. In most cases, existing evaluation efforts focus on policies and programmes, rather than impacts. In instances where impacts are considered, this is predominantly at the local level; there is a marked absence of efforts to assess actual impacts on aggregated indicators at the national or international level.

Figure 1: Schematic illustration of the nested nature of different scales of business impacts on sustainable development (see text for abbreviations).

An overview of some current efforts at assessing impacts

The local ‘coalface’
The process of ex ante identification of business impacts at the local level is relatively well developed on the basis of environmental and social impact assessments (ESIAs), particularly in the case of large-scale infrastructure or resource extraction projects. But, because the relevant legislative standards and especially their enforcement are weak in many countries, many multinational companies commit to ‘international best practice’, represented, for instance, by the guidelines and requirements of the International Finance Corporation (IFC) and the associated Equator Principles (www.equator-principles.com, accessed 26 September 2007). These require companies to comply with and implement a number of social and environmental performance standards, including dedicated monitoring and reporting (IFC 2006a).

The IFC’s efforts are also noteworthy with regard to those of its Independent Evaluation Group (IEG), which – in line with development agencies’ emphasis on results-based management’ (OECD-DAC 2000) – assesses IFC-supported projects in terms of financial, economic, social and environmental criteria, as well as so-called private sector development impacts, all of which are synthesised in a ‘development outcome’ rating (IFC 2006b). The IEG found, for instance, that for a random sample of 210 operations approved between 1997 and 1999 there was a 59% success rate in terms of development outcome (compared with a 44% success rate in terms of financial performance). Methodologically, the IEG’s evaluations are limited in that they rely on the IFC’s data and include little participatory research (see below).

Private sector activities that are outside the purview of organisations such as the IFC, of course, are likely to face fewer monitoring requirements. Particularly in developing countries where state monitoring of social and environmental standards is often limited, there is often a lack of will or capacity among companies to devote sufficient attention to continuous or ex post monitoring and evaluation of development impacts. A lot depends on the particular proclivity, skill and influence of the relevant personnel on site. Furthermore, ESIAs and their corresponding management and monitoring systems still face important limitations, despite their adaptation of frameworks such as sustainable livelihoods (see, for instance, IFC 2006b).

Methodological challenges are evident also in independent, third-party evaluations at the local level. For instance, Bury (2004) applies the sustainable livelihoods model to assess the impacts of Newmont’s Minera Yanacocha operations in Peru. These have increased households’ access to produced and human capitals, but decreased their access to natural and social capitals. However, the livelihood approach ‘says little, at present, about how …to synthesise these findings into a more parsimonious conclusion that explains whether household livelihoods are better or worse off since [the company] began operations in the region’ (Bury 2004: 89). This challenge relates to a broader, fundamental debate with regard to the substitutability of different forms of capital, particularly in the ecological economics literature (Pearce and Atkinson 1993).

In this context, recent progress in human rights impact assessment is particularly promising. BP was the first to publish such an assessment by independent analysts (Smith and Freeman 2002; for the company’s response see BP 2003), though John Ruggie, Special Representative to the UN Secretary-General on business and human rights, bemoans that ‘relatively few firms conduct these assessments routinely’ (quoted in Baue 2007). While the BP example was an ex ante assessment, the value of human rights impact assessments from an ex post perspective is illustrated by case studies profiled in a report by Rights & Democracy (2007), which argues that, in each of the four ex post evaluations, foreign investment projects have had negative impacts on human rights. Over and above methodological challenges, problems with impact evaluations conducted by companies themselves are often compounded by the company management’s emphasis on those impacts that are of particular value to its perceived priority interests. Instead of measuring impacts on beneficiaries in a broader sense, company-contrived indicators are emphasised, often inspired by tangible project timetables. In this vein, a local manager in one of South Africa’s platinum mining regions noted (Personal communication with Bandile Gcabo, Rustenburg, South Africa, 17 April 2003 (Bandile Gcabo was community relations manager for Lonmin Platinum in the Rustenburg area).

It’s one thing to set targets and hope all goes well, but involving all stakeholders for a sustainable solution takes more time than we are often prepared to give …The easy way is the wrong way. If you want to build a road you can just bulldoze your way to building a road, and then you can show your road for it [in terms of measurable deliverables]; but involving government and other stakeholders in building the road is more difficult, but more sustainable in the long term.

In this edition, Ricarda McFalls’s case study of a Hewlett Packard (HP) project testing the notion of the ‘bottom of the pyramid’ similarly attests to a ‘dark side’ of impact assessment: ‘The project team was under pressure to deliver fast, visible results at the risk of speed overriding substance’.

Impact evaluations are particularly problematic in conflict zones, in which the impact of companies is often the subject – and sometimes even the cause – of significant acrimony. In areas such as the Niger Delta in Nigeria, the impact of some of the most sophisticated and well-resourced corporate citizenship initiatives remains open to question (for opposing assessments in the Nigerian setting, see, on the one hand, Shell 2007 and, on the other, FOE 2004 and Boele et al. 2001). Such circumstances also aggravate the possibility of unintended negative consequences, whereby, for instance, CSR activities can contribute to further conflict (Akpan 2006).

Gitsham’s contribution to this edition shows how different observers and analysts arrive at very different appraisals of BP’s impacts in the area of its Colombian operations.  In such circumstances, the unilateral definition of success measures by the company is unlikely to contribute to conflict resolution, the building of trust and the establishment of collaborative responses to local challenges. Rather, it is likely that the identification of impact indicators may require participative, collaborative approaches with the involvement of all relevant role-players, especially the local poor, though of course this is especially difficult in conflict situations.

There have been important lessons learned with regard to the development of local sustainability indicators by means of participatory processes. Nurick and Johnson (1998) discuss the development of quality of life and corresponding industry impact indicators in the case of south Durban in South Africa, illustrating the depth and context-specific nature such indicators adopt on the basis of careful community participation. Based on subsequent research in the same area, Acutt et al. (2004) show how companies’ failure to meaningfully engage local communities in their impact appraisals has led to a breakdown of trust and thus a failure of the resulting mitigation plans.

Company-level evaluations
The company is the most common level of analysis for organisational scholars and the growing socially responsible investment industry, including a wide array of ratings and indices, such as the Dow Jones Sustainability Index. Most of these approaches, as noted above, are characterised by an emphasis on company policies and practices, rather than impacts on particular stakeholders or objectives. Referring to Chatterji and Levine’s (2006) overview of codes and ratings, Porter and Kramer (2006) bemoan the lack of consistency and rigour in these ratings’ criteria and their weightings, as well as their emphasis on ‘measures for which data are readily available, even though they may not be good proxies for the social or environmental effects they are intended to reflect’ (Porter and Kramer 2006: 3). Coupled to a reliance on unreliable data, this leads to ‘a jumble of largely meaningless rankings, allowing almost any company to boast that it meets some measure of social responsibility – and most do’ (Porter and Kramer 2006: 3). Attempts to construct more rigorous ‘composite sustainability indicators’ at the company level include Krajnc and Glavic 2005.

From a company perspective, the most prominent, systematic mechanism for publicly reporting on sustainable development impacts is, of course, sustainability reporting. Though this practice has come a long way from the initial discussions in the 1970s on social accounting (Estes 1976) and social audits (Medawar 1976) to the current prominence of the Global Reporting Initiative indicators (GRI 2006), most companies’ reports are still not very effective in explaining actual impacts on stakeholders, especially in holistic terms for the company as a whole.

There are three overarching concerns in this regard. The first challenge for the GRI and reporting companies has been the development and application of indicators that are relevant universally, while at the same time pertaining to local contexts that are very diverse and complex. Important recent efforts, therefore, attempt to combine the use of ‘top-down’ indicator systems, foremost among which is that of the GRI, with ‘bottom-up’ systems developed on the basis of participatory methods involving, in particular, local stakeholders that are most directly affected (Chamaret et al. 2007). As far as company initiatives in this regard are concerned, one of the most prominent efforts is the Tata Index for Sustainable Human Communities, which, inter alia, encourages all Tata employees to interact with [local] communities and to jointly create measurable markers’ for the company’s social investments (Branzei and Rao 2007: 11).

The limitation of corporate sustainability reports also relates to their focus on the corporate level of analysis, whereby cumulative impacts of different companies and industries at the local, national or international level are not subject to systematic analysis (see also Blowfield 2007). It remains to be seen whether companies will respond comprehensively to the third-generation GRI principle of reporting within the sustainability context (GRI 2006), but this is unlikely to be successful without greater commitment to regional and sectoral sustainability assessments (e.g. Pope et al. 2004), in which governments will need to play a leading role.

More fundamental critiques of corporate sustainability reports involve claims of ‘greenwashing’ (Bruno and Karliner 2002; Laufer 2003; FOE 2004), or the intentional provision of disinformation for the purpose of image management. Despite the advances made through the GRI, many NGOs continue to view corporate sustainability reports negatively with regard to their ‘credibility and sufficiency’ (O’Dwyer et al. 2005). When companies’ motivations for sustainability reporting are more to do with maintaining legitimacy and controlling stakeholders than discharging accountability to stakeholders and curbing unsustainable trends, this does not only have detrimental effects on the credibility of the corporate citizenship movement. It may also have perverse side effects: ‘The purchase of the “commodity of compliance” sufficient to shift the risk of liability and loss, in certain firms, may result in decreased levels of care by senior managers’ (Laufer 2003: 257).

To respond to these challenges, Laufer (2003: 259) argues for ‘Tripartism, or the integration of a third party into the regulatory arena occupied by an organisation and a regulator’. To some extent, the increasing scrutiny of corporate activity by a global network of NGOs, linked to local activists and making effective use of the media and the Internet, arguably approaches the introduction of a third party in the regulatory arena (see Spar 1998). Yet there is significant scope to enhance this involvement of civil society in company-level impact appraisals, not only – as noted above – to increase local legitimacy of the indicators used, but also to increase accountability and trust. One of the most striking examples of such an approach was a study on the impacts of Unilever on poverty in Indonesia conducted jointly by the company and Oxfam (Clay 2005). This study is also remarkable for its comprehensive analysis of impacts ranging from poor consumers to the country’s macro-economy, illustrating the wide-ranging impacts in the company’s value chain. In his foreword, the company’s chief executive declared this to be a ‘learning partnership which demonstrated how much insight can be gained by working together’ (Cescau foreword in Clay 2005: 11).

In this edition, Scott Marshall and his colleagues argue on the basis of a quantitative study that negotiation and dialogue, rather than confrontational strategies, between firms and stakeholder groups (in this instance, NGOs and institutional investors) has led to notable increases in the quality of disclosure of environmental information. This underscores Blowfield’s (2007) view that the most significant effect of the corporate citizenship movement has been widespread organisational change, manifested in this instance by improved disclosure.

A less confrontational approach by NGOs is also advocated by Klaus Leisinger in his invited contribution to this edition. Arguing from a corporate perspective, he notes that engaging stakeholders is a good investment ‘even if the process is sometimes a difficult one’, but bemoans some NGOs’ ‘campaign strategies that seek to undermine corporate integrity in the interest of sensationalism…[and which may result] in companies falling back to a legalistic, compliance-oriented approach’. Indeed, one of Oxfam’s lessons from the above-mentioned Unilever study was that ‘our analysis needs to be more alert to the differences between multinational companies’ (Clay 2005: 21). At the same time, however, the significant impact and widespread appreciation for this study was probably due in large part to what Covey and Brown (2001) would characterise as Oxfam’s ‘critical cooperation’ approach.

Codes of conduct and cross-sector collaboration initiatives
The crucial role of participatory and partnership approaches in assessing corporate citizenship efforts has also been documented in the implementation of codes of conduct, especially those related to labour practices and impacts on workers. Barrientos (2005) emphasises the manner in which participation by workers, trade unions and NGOs in the implementation and monitoring of the code of conduct can facilitate a continuous ‘learning approach to impact assessment’, which iteratively improves workers’ quality of life. She stresses the need to go beyond commonly emphasised indicators of compliance to ‘examine whether a particular change in employment conditions due to a code has had a positive or negative effect on these broader dimensions of workers’ well-being and poverty’ (Barrientos 2005: 266; see also Prieto-Carron et al. 2006: 982).

In this edition, Nelson and colleagues report on a longitudinal, comparative study of the livelihood impact of codes of conduct in the Kenyan cut flower and South African wine industries. They find that workers are generally better off in companies that had adopted codes in comparison with companies that did not. However, attribution remains a vexing challenge in that it is difficult to assess whether these differences are indeed due to the codes. They also argue that the relatively modest economic benefits that may be accruing to workers are fundamentally limited due to these workers’ lack of power in the broader supply chain, and that development agencies’ focus should perhaps shift to alternative strategies for improving local livelihoods.

Some sort of cross-sector collaboration is a feature of most codes and evaluation processes, and indeed it plays a role in most corporate citizenship strategies and initiatives at all scales of analysis. Such partnerships take a variety of forms and have diverse objectives and governance structures. During the World Summit on Sustainable Development held in Johannesburg in 2002, so-called Type 2 initiatives were touted as a significant contributor to sustainable development, despite vocal concerns. The most recent official UN overview of these initiatives focuses on ‘progress in implementation within the broad categories of partnership building/coordination, capacity-building activities, information sharing and pilot projects’ (UN 2006: 1), but it provides hardly any information on performance as measured against the wider sustainability objectives they are meant to achieve. Hence, five years later, the impact of such initiatives remains uncertain and arguably the expectations raised during the summit have not been realised (for empirical studies of such initiatives that now require updating, see Hale and Mauzerall 2004; AICC 2004).

This is not to say that collaboration initiatives cannot have important benefits that are greater than unilateral efforts. Such benefits are especially tangible in the case of locally driven partnerships’, though these ‘should focus more on measuring results’ (Steets 2006: 6). In this edition, Muthuri highlights how a collaborative approach allows the case study firm to motivate and enable community members to take a more proactive, self-reliant approach to local development. On the other hand, McFalls’s case study provides a more critical view of a public–private partnership, noting that HP’s approach to this particular initiative did not allow for meaningful participation by local beneficiaries in its planning or implementation, and even the government partners were not involved effectively. This propensity for companies to dominate ostensible partnerships is underscored by Ashman (2001), who – like Nelson et al. in this edition – emphasises that the limited economic gains that local citizens in poor countries may get from partnerships should not detract from the necessary political empowerment that would ensure more sustainable livelihoods.

The complexity of assessing the impact of collaboration initiatives is of course heightened at greater scales of analysis; this often leads to an even greater emphasis on interventions’ outputs and outcomes, rather than impacts. This is perhaps most evident in the opposing assessments of one of the most prominent corporate citizenship initiatives, the UN Global Compact, a key criticism of which has been that ‘there is no assessment mechanism (just corporate self-reports)’ (Chatterji and Levine 2006: 32). In this edition, Leisinger and Rieth et al. provide a positive assessment of the Compact’s potential.

The latter use case studies from southern Africa to emphasise the role that a proactive decentralisation strategy can play in embedding its principles and making use of existing social capital in local networks of companies and other stakeholders in developing countries. They note that the Global Compact can play a particularly important role in developing countries such as Malawi and Mozambique, where it is often the first systematic introduction to CSR issues for local role-players and where it can provide a crucial convening platform for cross-sector collaboration. But it is also in these countries where the limit of voluntary initiatives is apparent, as company representatives themselves emphasise the need for greater government regulation, enforcement and monitoring on social and environmental issues.

Conclusions

Despite the prominent expectations that business can make significant contributions to sustainable development, we do not know whether current corporate citizenship initiatives are, indeed, making a difference. Selective examples to demonstrate development impact (WBCSD 2005) or even case studies by independent observers do little to convince sceptics of the broader benefit of corporate citizenship for the poor or the environment, especially when critical counter-examples or interpretations abound. Even the surveys, ratings and rankings undertaken by organisational scholars (see Margolis and Walsh 2003) or investment rating agencies (see Chatterji and Levine 2006) are prone to being constrained by unreliable data, invalid metrics or incomparable methodologies. Furthermore, most of these assessment efforts focus on corporate policies and practices, rather than impacts. So, while the best indication we have is that the corporate citizenship movement has led to widespread organisational change (Blowfield 2007), its impact on sustainable development remains unclear.

This article has suggested that this uncertainty is not primarily due to a lack of effort. Nor is it only due to a lack of suitable metrics or other technical challenges, though some methodological advancement is no doubt necessary to deal with the complexity discussed in this article. Such advancements are particularly important in developing a better understanding of business impacts at regional, national and international scales of analysis, involving accumulated local impacts, as well as direct higher-order impacts resulting from, for instance, tax payments and political lobbying.

Over and above such methodological challenges, a crucial yet rarely considered role is being played by the underlying motives, power relations and value systems informing current approaches to impact assessment, perpetuating conflicting appraisals of the impacts of business activity. Put differently, the broader political economy not only plays a role in underpinning and mediating the socioeconomic and environmental impacts of business activity, but it also constrains our knowledge of these impacts. This is also illustrated by the various instances of the ‘dark side’ of impact assessment discussed in this article, whereby legitimacy- or image-driven approaches to impact assessment and standard setting have perverse, negative consequences for organisational performance and social welfare (Laufer 2003; Chatterji and Levine 2006; McFalls this edition). At a more basic level, this requires greater attention to the issue of who pays for assessing business impacts.

Taking into account both the methodological and political economy aspects considered in this article, the following two considerations may help companies and other roleplayers – with special responsibilities for governments – develop a better, shared understanding of business impacts on sustainable development.

Greater emphasis on assessing business impacts on the poor and their human rights, based on the fairness principle. This is an important response to the complexity of business–society interrelationships discussed in this article, because it focuses the objectives and methods of impact assessments. It is also a crucial application of Rawls’s fairness principle that ‘an inequality is allowed only if…it will work for the advantage of every party engaging in it’ (Rawls 1958: 167, original emphasis). By emphasising that the worst-off should not experience any further deterioration of their livelihoods, it helps navigate the difficult trade-offs between different kinds of business impact. It can also help to bridge some of the different value systems that are currently contributing to such contradictory appraisals of the role of business.

• An emphasis on participatory appraisal and research partnerships:  One of the overarching conclusions of the above discussion is that unilateral efforts by companies to assess their impacts on development are likely to lack crucial information and legitimacy. Hence participatory methods that involve, say, local household members in a livelihoods assessment or the development and monitoring of community-based indicators ought to play a more prominent role. Partnerships between companies and NGOs that are known to be independent and capable of informed criticism (as in Clay 2005) provide not only an important learning opportunity arising from the exchange between different perspectives, but also significant legitimacy in a context in which corporate reports are distrusted by many civil society organisations
• Finally, the research community will also need to be involved more systematically in impact assessment efforts:  Considering the multifaceted complexity of the business-society interrelationship, such research partnerships ought to involve multi-disciplinary research teams applying both quantitative and qualitative methods. The researchers will need to be able to maintain and demonstrate their independence, contributing to the assessment not only methodological research skills and techniques but also an attitude of ‘critical thinking’ (Utting and Zammit 2006: 38). For such partnerships between business and independent researchers to be authentic, the company or companies concerned will need to be able to surrender some power over the process and its outcome.

This willingness currently seems in short supply.

This special edition of the Journal of Corporate Citizenship is the result of a symposium I helped organise in 2006 as head of research at the Centre for Corporate Citizenship at the University of South Africa. I am grateful to those individuals and organisations that made the edition possible and the symposium on which it is based. This includes the conveners of the symposium, the Unisa Centre for Corporate Citizenship, the UN Global Compact, and the Lindenberg Centre for Global Citizenship; the hosts, the Ghana Institute of Management and Public Administration; the symposium committee members, Manuel Escudero, Sanjeev Khagram, Franklyn Manu, Jeremy Moon, Maggie Opondo and Dante Pesce; and – most importantly – the 40 participants from 16 countries and four continents. Crucial support in compiling the special edition was provided by the co-editors, Maggie Opondo and Sanjeev Khagram; the anonymous referees; as well as Mara Brain and Catrina Lucero. This introductory article has benefited from comments on an earlier draft by Nicky Black, Malcolm McIntosh and Fleur Boulogne, and it was prepared with support from the University of Cape Town Environmental Evaluation Unit and TrustAfrica.)

– Picture courtesy of The Co-operative