Introduction
Multilateral Development Banks (MDB) are international organizations (IO) created with the aim of financing economic development in developing countries. They finance, usually on a reimbursement basis, large scale projects like dams and highways, which often have dramatic impacts on hundred of thousands of inhabitants in affected areas. Responsibility for providing fair solutions to these people belongs to borrowing states, while lenders are usually insulated from accountability.
IOs in general are subjects of international law and thus can have international obligations stemming from general international law, treaties or their constituent instruments. But in the particular case of MDBs, obligations in fields like human rights, labour or the environment are found in neither of these sources of international law. MDBs are not parties to human rights or environmental treaties, and their constituent documents – the Articles of Agreement between Member States – do not make any reference to these topics. As for general international law, it is hard to find applicable rules given the MDB's lack of physical presence at the place of alleged violations, except perhaps for the case of financing notoriously abusive regimes.
Under these circumstances, the voluntary commitments undertaken in MDBs’ environmental, social and governance (ESG) standards are the only concrete social justice standards against which to hold MDBs accountable. With ESGs, development lenders at the same time regulate, by imposing environmental and social standards on borrowers, and self-regulate, by committing, among other things, to monitor borrowers’ compliance. To assess the ESG's potential to enhance MDBs’ accountability, this essay proposes a ccritical reading of these statements, with ‘critical’ used here as ‘showing connections and causes that are hidden’ (Reference FaircloughFairclough, 1992, p. 9) − a meaning employed in various disciplines by authors committed to social change in their fields (see e.g. Reference FreireFreire, 1970 on critical pedagogy). Our central object of research is the ESG entitled ‘Environmental and Social Framework’ (AIIB-ESF) adopted in February 2016 by the Asian Infrastructure Investment Bank (AIIB), initiated by China in 2015 to finance investment in a specific area (Asia) and a specific field (infrastructure). We look at the language in the AIIB-ESF, in a comparative perspective with the language of other MDBs, seeking for the line where AIIB's discourse cease being a MDB discourse and turns into Chinese discourse. In other words, we try to bring to light two different degrees of recontextualization by AIIB, consistent with distinction made in the literature (Reference Chouliaraki and FaircloughChouliaraki & Fairclough, 1999) between appropriation of external discourses, on the one hand – in our case, incorporation of MDB meta-discourse into the AIIB one – and colonization on the other hand; in our case, colonisation of the MDB meta-discourse by the AIIB one – that is, the Chinese one, as China reserved for itself in AIIB powers significantly higher than any other founder in any other MDB.
In the development literature, lenders’ interests, main donors’ especially, were already under scrutiny, and it was demonstrated that for each of them, ‘a different story can be told about the use of aid as an agent of national interest’ (Reference HookHook, 1995, p. 16). But the link between the text as discourse, and the main founders’ cultural traits and rationally pursued interests was not studied. We aim to find manifestations of China's legal culture and national interest in AIIB's ESGs, by assessing intertextuality and reconceptualization in the AIIB policies. The purpose of this endeavour is to evaluate the overall utility of the ESGs to their claimed purposes, and to formulate suggestions for turning them into more useful policy tools.
Critical discourse analysis as methodological tool
Critical Discourse Analysis (CDA) is a methodological tool used to show ‘how discourse is shaped by relations of power and ideologies’ (Reference FaircloughFairclough, 1992, p. 12). It is a moral and political enquiry of the social science researcher, the result of which is intended to have effects in society, e.g. exposing power abuse, or mobilising people to remedy social wrongs (Reference Blommaert and BulcaenBlommaert & Bulcaen, 2000, p. 449). Although not without contestation (see Reference BreezeBreeze, 2011 for a comprehensive review of critiques), CDA has established a firm presence in social science fields, in recent decades. Among the many methodological variants under its umbrella, we use the one proposed in Fairclough (Reference Fairclough, James and Michael2012). The method, devised by the one of the iconic figures of CDA, comprises four stages, the first of which requires ‘a research topic that relates to, or points up, a social wrong and that can productively be approached in a trans-disciplinary way, with a particular focus on dialectical relations between semiotic and other “moments”’ (Reference Fairclough, James and MichaelFairclough, 2012, p. 13).
In our paper, the social ‘wrong’ is the lack of classical – i.e. performed by state actors using legal mechanisms – regulation tying project finance institutions to the environmental and social effects of the megaprojects they finance. Obviously, what constitutes a social wrong is a controversial matter; by identifying self-regulation as a ‘wrong’, we may be seen as already signalling a preference for a regulatory approach. This is not the case, as we show later in the paper; but had it been so, this would not have been inappropriate. As shown by Van Dijk (Reference Van Dijk1993, p. 252), critical discourse analysts should take explicit socio-political stances, spelling out their point of view and speaking on behalf of those who suffer most from dominance and inequality; unlike other types of discourse research, CDA is ultimately political. This feature is at the same time ‘one of the most refreshing things about critical discourse analysis’ (Reference BucholtzBucholtz, 2001, p. 167), and the main target of contestation: critics have pointed to tendencies of bias in text selection and interpretation, to over-obvious analysis and even manipulation of readers (e.g. Reference HammersleyHammersley, 1997; Reference VerschuerenVerschueren, 2001; Reference WiddowsonWiddowson, 1998). Therefore, at a minimum, in order to let their readers critically evaluate the CDA analysis, authors who use CDA should from the onset state and ground their standpoint (Reference HammersleyHammersley, 1997, p. 242). Ours, grounded in the concepts of social justice and distributive justice, is that the millions who stay ‘in the way of development’ are too often ignored by the powerful entities who, with their money, make mega-projects possible in developing countries. If the MDBs’ voluntary commitments towards these people are all we can have to ensure social justice, then they should be at least workable policy tools.
Once identified, the research topic related to the social wrong, in our case the ESGs in general, needs objects of research that can be theorised in a transdisciplinary way. This is stage two, where we put under the microscope the AIIB's ESG, dealt with as Chinese discourse since intuitively, China's interests in having its state owned companies building infrastructure in AsiaFootnote 1 could explain reconceptualization in AIIB ESGs of existing norms in the field of development ESG, in general.
CDA can offer insights on social phenomena only by analysing both individual texts and their relationships with other texts (Reference Leitch and PalmerLeitch & Palmer, 2010, p. 1197) – in our case, the other ESGs. The documents against which the AIIB-ESF, as ‘text’, is discussed are the World Bank's Environmental and Social Framework (WB-ESF, published in 2016), and the similar policies of the regional development banks. In our intertextual analysis, all these become the institutional context of our ‘text’, in the sense that ESGs have achieved, via processes of isomorphic convergence described by institutionalist scholars, a significant degree of uniformity. In other words, the ESGs have become an institution, in the meaning given by sociological institutionalists: ‘formal rules, compliance procedures, and standard operating practices that structure the relationship between individuals in various units of policy and economy’ (Reference HallHall, 1986, p. 19). The institutional is, however, but one of the contexts relevant to our text. As we will see, AIIB-ESF departs in certain areas from the institutionalised model. Institutionalist scholars unconvinced by the theory of mechanical, unwitting convergence towards globally legitimised models explain organisations’ resistance to convergence using the concept of multiple institutional logics (e.g. Reference BinderBinder, 2007; Reference Dunn and JonesDunn & Jones, 2010); in our case, cultural conditionality, or wittingly pursued interests, are ‘logics’ that may reduce the degree to which the organisation (AIIB) is absorbed into universal normative patterns. Accordingly, we have to put the AIIB-produced text in other contexts as well: national (Chinese cultureFootnote 2) and temporal (Chinese contemporary interests) are ‘logics’ that stood in the way of a full immersion in the institutional context. By examining the two additional contexts of our text, we are addressing another important critique to CDA, namely that it has sometimes ‘paid insufficient attention to features of the immediate context, which has led to interpretations which are pragmatically inappropriate or remote from the concerns of the participants’ (Reference BreezeBreeze, 2011, p. 520).
In Fairclough's formulation, step two goes back to ‘asking what it is about the way in which social life is structured and organised that prevents the social wrong from being addressed’ (Reference Fairclough, James and MichaelFairclough, 2012, p. 14). We maintain that what erects barriers to addressing the social wrong is the propaganda hidden in ESGs texts. Stage two is the core of CDA, as it comprises the main operational steps: ‘Select texts, and points of focus and categories for their analysis, in the light of, and appropriately to, the constitution of the object of research; carry out analysis of texts – both interdiscursive and linguistic/semiotic analysis’ (Reference Fairclough, James and MichaelFairclough, 2012, p. 14). Importantly for the case of texts like ESGs, CDA takes into account not only what is said, but also ‘silences, implicatures, ambiguities, and other covert but powerful aspects of discourse’ (Huckin, Andrus, & Clary-Lemon, Reference Huckin, Andrus and Clary-Lemon2012, p. 110).
Our points of focus are developed in the next two sections of this paper, according to the degree of consistency with established discursive patterns in previous texts that inspired the AIIB's: norms where AIIB follows the beaten paths of development CSR, sometimes with subtle adaptation, and areas of normative innovation and discourse divergence. The points we discuss are not meant to aggregate into a coherent and comprehensive comparative analysis of the MDBs’ safeguards; the specialists of the German Ministry Economic Cooperation and Development did this few years ago (see Von Bernstorff & Dann, Reference Von Bernstorff and Dann2013). We only intend to provide, by spotting linguistic details (wording, syntax, emphasis and the like) in texts, enough light on obscure manifestations of national and organisational interest in the language of ESGs. Steps three and four in Fairclough's roadmap – considerations over the necessity, for the social order, of the social ‘wrong’, and possible solutions to the obstacles brought about by the research object – are addressed in the final section.
Appropriation by AIIB of ‘classical’ ESGs
The discourse of the powerful is always characterised by inauthenticity, i.e. a disjunction between words, thoughts and actions (Reference Hodge and CoronadoHodge & Coronado, 2006, p. 532), and this is apparent in all ESGs, which show a stark contrast between grandiloquence in tone and impotence in details. For example, in AIIB-ESF, the legitimatory discourse is visible in the Objectives (para. 5), which are ‘to provide a robust structure’, ‘to ensure the environmental and social soundness’, ‘to provide a framework’ and so on. The fracture between appearance and reality is not only at the text level as a whole, but also inside specific provisions – see for example the requirement, in most of the ESPs is the requirement that the client considers the ‘No project alternative’, in their environmental and social assessments; see e.g. ADB-SPS (p. 16), AIIB-ESF (p. 28) or EBRD-ESP (p. 12). While this legitimisation discourse conveys the message that lenders want to make sure that the project will not do more harm than good, in practice, the analysis of alternatives is often limited to the obviously inconvenient ones, making the project look good by comparison (Reference MacDonaldMacDonald, 2001, p. 1025).
But beyond the legitimatory vocabulary, all ESGs make a clear statement of denial of responsibility, underlining that MDBs are only prepared ‘to assist’ the client in respecting its environmental and social obligations towards the end customer, the people, sometimes by ‘promoting’ specific standards. AIIB for instance will only ‘assist Clients in fulfilling their national and international obligations relating to environmental and social risks and impacts’ (Asian Infrastructure Investment Bank [AIIB], 2016, para. 6). Similarly, EBRD ‘recognizes the responsibility of clients’ (European Bank of Reconstruction and Development [EBRD], 2014, art. 9), while the clearest insulation from responsibility is found in Asian Development Bank [ADB], 2009, para. 16: ‘A basic principle … is that implementation of the provisions of the policies is the responsibility of the borrower/client’. The verb ‘to recognize’ used by EBRD has an anaesthetic effect for the reader: something that only needs recognition is already there and thus its existence cannot be challenged. MDBs convey the message that even if they wanted, they could not take responsibility; their hands are tied by the powerful norm of borrower responsibility, which they can only ‘recognize’. While obviously it is indeed the obligation of the borrower to implement the safeguards, the lender's obligation to refrain from funding certain projects and certain borrowers, and after approving a loan, to oversee compliance throughout the project cycle, should not be downplayed the way ESGs do.
When moving to the operational parts of the documents, we note that words are carefully deprived of their dimension of action – a characteristic of the discourse of powerful, as noted by Freire (Reference Freire1970, p. 60). AIIB-ESF suffers from the same lack of clarity, consistency and comprehensiveness found in the discourse it appropriates (see Reference WrightWright, 2007, p. 10 on the weaknesses of European Investment Bank's relevant policies; on ADB-SPS, see Reference FriedFried, 2008, noting that they are extremely vaguely written). For example, AIIB only ‘believes that the application of the [ESF] … will support Clients in achieving good international practice relating to environmental and social sustainability’ (para. 8). As for biodiversity and sustainable management of living natural resources, AIIB acknowledges them as an objective, but emphasise that it should be balanced with other considerations (para. 16). In cases of highly risky projects, AIIB ‘may require the Client to engage … internationally recognized independent experts’ (para 25); in fact, in the 55 pages of the document, the word ‘may’ is used more than 100 times. It should be however pointed out that vagueness and open doors characterise all the ESGs; for example, the WB-ESF requires compensating for or offsetting significant residual risks or impacts only ‘where technically and financially feasible’ (World Bank [WB], 2016, p. 25). Similarly, the ‘X or at least Y’ strategy, used twice in the AIIB's ESP (‘improve, or at least restore [livelihoods of displaced people]’), is common to the development safeguards, prompting questions on what a commitment to ‘help, or at least do not harm’ someone in need may mean.
A specific manifestation of the use of ambiguity is in the way ESPs categorise the projects as either A (the riskiest ones from an environmental perspective), B (less risky) or C (negligible risks). The AIIB policy, for example, states that
A Project is categorized A if it is likely to have significant adverse environmental and social impacts that are irreversible, cumulative, diverse or unprecedented. These impacts may affect an area larger than the sites or facilities subject to physical works and may be temporary or permanent in nature. The Bank requires the Client to conduct an environmental and social impact assessment (ESIA) or equivalent environmental and social assessment, for each Category A Project … (para. 13).
To be noted that aside from the words carrying an interpretable dimension, like likely and may, AIIB leaves further doors open, by requiring the Client to conduct, for A projects, an environmental and social impact assessment (ESIA) or an ‘equivalent environmental and social assessment’. This is understandable from the bank's managers’ perspective, since a proper ESIA takes time and may result in additional measures and costs, and may thus affect the yearly quantity of loans. By categorising the bulk of the projects as B – where ESIA is at the lender's latitude – the ‘culture of approval’,Footnote 3 well documented in the case of the World Bank (see e.g. Reference Rich, Pincus and WintersRich, 2002, p. 51, speaking of the incongruence between the needs to speed up loans approval and the claims of rooting out the culture of approval; Head, Reference Head2008, p. 191, describing WB's practice of lending for lending's sake), rests undisturbed.
The above examples refer to norms’ emasculation through ambiguity, which serves the purpose of anesthetising critique (by creating the appearance that the principles of social justice are respected) and from here, of dominance.Footnote 4 This is an instance of the text obscuring a norm that does not suit the author, and the keyword here would be manipulation. Another instance of dominance is revealed by the discursive strategy meant to legitimate the controversial change towards using host country and even firm ‘systems’ – a word connoting coherence and the efficiency of a set of rules and institutions, when in fact this is not always the case. True, country systems can and should be strengthened, and the MDBs are in the perfect position to incentivize for that; in fact, AIIB even has a heading saying ‘Strengthening of Country and Corporate Systems’. One may expect to see evaluation and guidance from MDBs in the process of strengthening – but the text only provides that ‘the best way to strengthen these systems is to use them’ (AIIB, 2016, para. 12).
Tens of NGOs have sharply criticised the country systems approach (CSA) in the case of WB.Footnote 5 Park (Reference Park2008, p. 21) sees the shift towards using the CSA as ‘a very strong signal […] that environmental and social safeguard policy norms are no longer norms that should be morally and contractually enforced in all international development loans’. The advancement of most of MDBs towards the CSA, ignoring these concerns, is a reminder that dominance may be reproduced by routine forms of text that appear natural and quite acceptable; after all, there is nothing bad in acknowledging a state's sovereignty, one may say. Here, the keyword is persuasion.
A third manifestation of dominance in the ESGs can be described by the word dissimulation – of inconsistencies between their standards and internationally agreed norms. While in the case of environmental and social matters in general, the ESG drafters benefitted of a lack of universally accepted standards, the case of indigenous populations is different: the world seems to have largely agreed on the substance and boundaries of their rights. The foundations of these rights in the Indigenous and Tribal Peoples Convention (also known as ILO Convention 169) referred to prohibiting non-consensual displacement and were based on self-determination and communal, informal property rights on their lands. Also, as stipulated in article 32 of the United Nations Declaration on the Rights of Indigenous People (UNDRIP),Footnote 6 any administrative decision that affects them is conditioned by the existence of the free, prior, and informed consent (FPIC). Indigenous people now receive significant global support in their quest for self-determination and its particular manifestation of denying consent to an unwanted project, including for environmental concerns. Therefore it is interesting to note the dissimulation of an exit option regarding the FPIC of indigenous peoples, in AIIB-ESF:
When the Bank is unable to ascertain that such broad community support has been obtained from the affected Indigenous Peoples, it excludes from the Project those activities that would affect those Indigenous Peoples. (AIIB, 2016, para. 61)
It is hard to see how the indigenous FPIC can be given for some aspects of an operation, and denied for others; what if the ‘particular operation’ for which the indigenous people deny the FPIC is essential to the whole project?
Another example of dissimulation is the legerity with which all ESGs speak of displacement, as if it was a casual occurrence. According to the UN Committee on Economic, Social and Cultural Rights, ‘instances of forced eviction are prima facie incompatible with the requirements of the [International Covenant on Economic, Social and Cultural Rights] and can only be justified in the most exceptional circumstances’ (UN Committee on Economic, Social and Cultural Rights [UNCESCR], 1991, para. 18, emphasis added). In the field of involuntary displacement, the responsibility of project finance providers should be high, as most mega-projects, which remove hundreds of thousands from their homes and ancestral ways of life, would not be possible without international finance. Instead of leveraging on this position of power vis-à-vis borrowers, and imposing strict conditions, including a verification of the real existence of the public interest, MDBs safeguards trivialise the extraordinary.
Specificities in AIIB's discourse
Very limited provisions on labour rights
While in the previous section we pointed to areas where AIIB followed the discourse of other MDBs in their policies, this section is dedicated to discussing areas where other types of contexts required adaptation of said discourse.
On labour rights, AIIB only ‘ … believes that the following measures contribute to the quality of the Project: […]’ (AIIB, 2016, para. 15, emphasis added). The italicised words show that a labour safeguard is practically non-existent. Moreover, in the operational part of the document, the working conditions are dealt with briefly and under a mixed heading (‘Working Conditions and Community Health and Safety’); the three paragraphs referring to labour insist on the relevance of national laws, put in place some safeguards relating to children employment and the prohibition of involuntary labour, and provide safeguards for labour management relations only for the private sector projects.
Interestingly, labour actually is the chapter in which other MDBs provided clearer – and not weaker, like AIIB – standards of protection. For example, AfDB clearly states its objective to ‘[a]lign Bank requirements with the ILO Core Labor Standards, and the UNICEF Convention on the Rights of the Child, where national laws do not provide equivalent protection’ (African Development Bank Group [AfDB], 2013, p. 49). Similarly, EBRD safeguards deal with labour conditions on 5 pages, again linking the level of protection to the ILO relevant conventions.
But AIIB's environmental and social policy's poorer performance in terms of labour safeguards is less surprising if we look at the International Labor Organization webpage: China ratified only half of the Conventions deemed by ILO as ‘fundamental’.Footnote 7 Four, out of eight fundamental labour conventions, are not ratified by China; for comparison, China is a member of more than 50 multilateral environmental treaties,Footnote 8 which set standards and rules of environmental protection accepted by China. The governmental Chinese-style CSR, with a clear focus on environmental protection and a low gear on other social justice fields, is further emphasised by the adoption of (only) environmental guidelines for its overseas investors (see Ministry of Commerce [MOFCOM], 2013), when for comparison, OECD has in place comprehensive guidelines (see Organization for Economic Cooperation and Development [OECD], 2011), covering environment but also numerous other fields, including labour.
The cleavage between environment and social justice fields, in China's vision now adopted by AIIB, is also visible in its international standing. On the one hand, China advanced substantially in foreign cooperation on environmental protectionFootnote 9; on the other hand, China repeatedly suspended its human rights dialogue with the US. With specific reference to AIIB, it should be mentioned that during his visit to Washington in September 2015, President Xi committed that the new lending institution would abide by ‘the highest international environmental and governance standards’ (Reference DonnanDonnan, 2015) – but the social standards were not mentioned.
No grievance mechanism
In its ESG, AIIB uses the word ‘grievance’ no less than 39 times, although the Bank has not yet designed a grievances mechanism, which makes it the only development bank in this situation, except for the Islamic Development Bank. With only one exception, the word ‘grievance’ in AIIB's ESF refers to the bank's expectancy that clients have such mechanisms in place. The one instance referring to the bank's own mechanism appears in a short paragraph and a footnote, in the context of what the bank announces to be its Oversight Mechanism, ‘currently under development’ (AIIB, 2016, para. 64 and fn13). In April 2017, more than a year after launching its ESF, AIIB made a public announcement that it works on establishing a complaints handling mechanism, the draft of which is expected to be submitted to the Board of Directors by December 2017, after going through public consultation.
Since China is the main drafter of the ESGs of AIIB, this delay in adopting a grievance mechanism reflects the Chinese style of dispute settlement, with preference for settling disputes via bilateral consultation or some other non-compulsory mechanism. China is very reluctant to accept dispute settlement by an independent third party, beyond her control. After the People's Republic of China was founded, it almost never used international tribunals to settle disputes, except for the case of WTO. For example, although there are Chinese Judges in both the International Court of Justice and the International Tribunal for the Law of the Sea, China has never submitted any dispute to these tribunals, nor accepted proposals to settle disputes there. China only started to accept the arbitration mechanism of the International Centre for the Settlement of Investment Disputes (ICSID) in recent years, when the volume of its overseas investment made juridical protection necessary. Therefore, to have an inspection panel to hear complains against projects financed by AIIB simply does not conform to the Chinese culture of dispute settlement.
For those familiar with the Chinese system of petitions (xinfang), it may on the contrary seem strange that a grievance mechanism that in essence is very near to something that China uses for centuries could not be devised faster. Paradoxically, it may be that exactly the experience with xinfang made the Chinese AIIB initiators reluctant in opening a Pandora's box. While petitioning has long been an element in the Chinese repertoire of contention, it is with recent socioeconomic and political reforms that it reached unprecedented dimensions in both quantity and quality (O'Brien & Li, Reference O'Brien and Li2006, pp. 10–11). With regard to quality, what worries the authorities are mainly the ‘skip-level’ petitions, that is, petitioners skipping the local authorities, who are legally the first addressees, and going directly to the central authorities; this is a problem because to avoid embarrassment, local authorities often use unlawful coercion to prevent the petitioners from doing so (Reference ChenChen, 2017, pp. 75–76). Moreover, with the rate of response to petitions being negligible (Reference LingLing, 2014), the aggrieved citizens have realised that ‘troublemaking’ (protests, or at least an implied threat in the way the petition is conceived) is the only way to get attention (Reference Chen, Pan and XuChen, Pan, & Xu, 2016 for instance found that in these cases, the rate of response can go as high as 20%).
In light of the above, it is not unconceivable that delaying at AIIB a grievance mechanism very similar to xinfang, in that it allows injured parties to complain to the Center about wrongdoings by Chinese corporations at local level, is a reflection of the general negative reaction among Chinese officials to the proliferation of petitions.
China's concern with reputation
Among the development banks against which we discuss AIIB's ESF, the only one showing concern for its own reputation is EBRD. Its ESGs state that the project appraisal will ‘include the assessment of the potential financial, legal and reputational risks as well as identify potential environmental or social opportunities’ (EBRD, 2014, p. 4). All other MDBs mention reputation only with reference to contractors, setting forth that the Borrower should work only with reputable ones.
For AIIB instead, reputation is so important, that is listed among the very objectives of the Environmental and Social Framework. The 2nd objective is to ‘[p]rovide a robust structure for managing operational and reputational risks of AIIB and its shareholders’; to be noted the reference to shareholders, which obviously means China, given the distribution of shares among the participating countries. The paramount importance of reputation is better understood if we note that this objective comes before the one of ‘[ensuring] the environmental and social soundness and sustainability of Operations’, which is only the 3rd one.
The phrase is repeated verbatim in the second part of the policies, the operational one, where it is stated in para. 3 that the environmental and social system is designed to ‘[p]rovide a robust structure for managing operational and reputational risks of AIIB and its shareholders’. Finally, reputation is mentioned for a third time in the context of AIIB's assessment of the project.
The firm message sent to Borrowers, but equally to Chinese and other contractors, is in line with general behaviour of Chinese leadership in recent years. Warnings on protecting China's reputation are constantly sent in a variety of domains. In the field of corruption, a report translating the International Chamber of Commerce (ICC) Rules for the Chinese multinational enterprises repeatedly warns them that frequent occurrence of corruption scandals in Chinese overseas investments ‘damages the national image of China’ (Chongyang Institute, 2014, p. 6). In the field of environment, the above mentioned guidelines for Chinese overseas investors, issued jointly by the Ministry of Commerce and the ministry of Environmental Protection in 2013, mention reputation twice, including in the very first article, which sets the document's scope; we find out there that among other things, it was adopted ‘in order to … set up good international images for Chinese enterprises’. Even in the field of tourism, China took steps to curb the embarrassing behaviour of Chinese abroad Pemberton (Reference Pemberton2015).
From a CDA perspective, an analysis of the insertion of three provisions referring to reputation in AIIB's ESG needs addressing two aspects: clarifying the historical and political context that led to such a concern, and understanding the significance of explicit and strong requirements in norms normally intended for other purposes. After all, the other MDBs, as well as any other IOs, are also concerned with their reputation, one of their strategic assets and an important source of legitimacy. All IOs and their member states will normally try to prevent threats to organisation and country reputations – as this is an important side of their international legitimacy – but none felt like stating this concern in its environmental and social policy.
With regard to the first aspect, not only the large proportion of shares reserved by China, but also the intense publicity that preceded the launching of AIIB contributed to a likely connection between China and AIIB in the general perception. To put it simple: if, say, AfDB will finance a megaproject involving outrageous human rights breaches, the public opinion will blame AfDB; if however the same is done by AIIB, the blame will go directly to China, by the same mechanism (conceptualised and analysed in Lopez, Gotsi, & Andriopoulos, Reference Lopez, Gotsi and Andriopoulos2011) governing the transfer of reputational losses from a corporation to its Country of Origin (COO). The already likely association between AIIB and China will be further facilitated by other circumstantial factors, like a negative image acquired in the 1990s and 2000s by China, due to the poor environmental and social performance of its overseas investors in Africa, Asia or Latin America, or the culture of money worshipping in China (see Reference Radavoi and BianRadavoi & Bian, 2014, pp. 25–29 for a list of factors contributing to China's reputational vulnerability).
As for the second aspect, i.e. the explicit and repeated presence of this concern in a document on environmental and social rights, we see it as a laudably strong command from shareholder to management. Too often, things went wrong in development finance because the MDBs’ managementFootnote 10 or the borrowersFootnote 11 ignored the strategic interests of the MDB member states and the organisation itself; the fact that the motivation here is egoistic does not alter the good outcome, i.e. the management and borrowers being sent a warning that damage to China's image will not be tolerated.
China's ambiguous regulatory style
While we have already noted that imprecision and generality characterise all MDBs, it may be the case that the AIIB-ESF's vagueness reflects not the discursive practice of other MDBs’ similar policies, but the Chinese legislative style – one having generality and flexibility as fundamental principles. Drawing on numerous Chinese academics, Keller (Reference Keller1994, pp. 749–751) explains the broad and indeterminate character of Chinese legal texts by factors like historical roots, the necessity of adaptation to local conditions that vary vastly across China, and the CPC need of ambiguous legal documents in order to make recourse to party policies necessary. Reflecting recently on the extent to which rule of law is a reality in nowadays China, Peerenboom (Reference Peerenboom2016) devotes attention to vagueness in legal norms as a hurdle to a genuine rule of law. Again, the poor quality of the law is explained by ideology, geography and culture. Additionally, Peerenboom (Reference Peerenboom2016, n24, p. 348) mentions the need of flexibility in adapting to quickly changing realities. On the other hand, the author admits the possibility of a more objective cause: ‘The fact that the legal system had to be rebuilt virtually from scratch is one of the main explanations for the frequent lack of pertinent laws … and much of the poor drafting and problems in implementation’ (Reference PeerenboomPeerenboom, 2016, p. 344).
While the explanations are multiple and sometimes speculative, ambiguity as a main trait of the Chinese normative style is uncontested. From our paper's perspective, this peculiarity is of high relevance: if the vagueness in AIIB-ESF is not an adaptation to the institutional context, but a reflection of the Chinese legal tradition (the national context), then there is even more room for optimism in respect to the chances of having the AIIB standards improved. Admittedly, ‘[a]mbiguity of the law is at the core of contemporary China's ideological and political ambiguity’ (Reference HanHan, 2016, p. 212), and thus bringing clarity into written norms in general is a political, not a technical, project. But AIIB-ESF are for overseas application, therefore if vagueness in their terms is only a cultural misalignment of regulatory styles between China and the West, then the drafters should be open to improve the text in this regard. This optimism is further cautioned by the fact that the Bank admits the EBRD's policies superiority and even follows them in practice: in a proposed dam project in Georgia, ‘the Bank has decided to apply EBRD's Environmental and Social Policy since EBRD is leading safeguard due diligence in the field’.Footnote 12 If the Bank is willing to follow a better standard, then why not make its own standard better?
Discussion and conclusion
In the two previous sections, we operationalised Stages 1 and 2 in the CDA model proposed in 2012 by Fairclough (Reference Fairclough, James and Michael2012), namely identifying a social wrong (i.e. the fact that MDBs ‘regulate’ in matters related to environmental and social abuses associated with development) and obstacles in addressing said social wrong (i.e. the fact that the MDBs, without which many megaprojects that end in environmental and social devastation would not be possible, construe soft law in a manipulative and dissimulative manner).
As stage 2, since meaning and intelligibility in discourse analysis are dependent on a network of prior and concurrent discourses and texts (Reference AdjeiAdjei, 2013, p. 3), we assessed the ESGs of the recently created AIIB, which were the text we studied, against two types of prior discourses, or contexts; accordingly, we found intertextuality between AIIB's ESGs and those of other MDBs (Section II, which treated the ESGs in general as ‘context’ for the AIIB's one), but also between AIIB's ESGs and other texts emanating from the Chinese government (Section III, which put AIIB's ESG in a historical-political context). The methodological tool we used was critical in the sense that it observed the rules prescribed by the given discourse, but also attempted an explanation of why those rules are shaped in that particular form.
Stage 3, on the necessity of the wrong, requires considering ‘whether the social wrong in focus is inherent to the social order, whether it can be addressed within it, or only by changing it’ (Reference Fairclough, James and MichaelFairclough, 2012, p. 15). The question, in our case, is whether it is possible within the MDBs-as-regulators approach to implement new, more efficient strategies for avoiding the injustices associated with development, or on the contrary, a classic regulatory system is necessary.
The balance in scholarship seems to go towards the first option: ESGs should be preserved and strengthened – let alone that a regulatory system to bind IOs is a rather utopian thought. From a general perspective, it was shown that in a globalised world, the distinction between law and non-law has lost much of its relevance; what matters is not whether norms, rules or standards are labelled as such, but whether their subjects treat them as authoritative (Reference Klabbers, Klabbers, Peters and UlfsteinKlabbers, 2009, p. 98, describing the effect based approaches to international law). In other words, the more and more numerous adepts of the effect, or impact-based approaches to international law contend that we should recognise ‘that law's existence is best measured by the influence it exerts, and not by formal tests of validity rooted in normative hierarchies’ (Brunnée & Toope, Reference Brunnée and Toope2001, p. 65). In this perspective and consistent with one theory adopting this approach, the ESGs, as result of incremental progress and norm diffusion, already belong to what was seen as the Global Administrative Law, a concept proposing that that much of global governance, including self-regulation of institutions like the MDBs, can be understood and as administrative action – thus leading to an emerging coherent quasi-regulation matrix, of the type of national administrative norms (Reference Kingsbury, Krisch and StewartKingsbury, Krisch, & Stewart, 2005, p. 25). If the ESGs are part of the Global Administrative Law, then what is left now is to clarify their formulation and contents.
As far as formulation goes, our article indicated several semiotic dissonances that need eliminated, if there is to make the ESGs a platform to evaluate MDBs’ behaviour. As for contents, it should be reminded that CDA focuses not only on what is said in the texts, but also on silences; in this perspective, the compliance side of the safeguards need strengthened via some additions to the texts. For example, it was suggested that ESPs include guidelines on how to suspend operations in cases of human rights violations (Reference SarfatySarfaty, 2009, p. 648); for the time being, AIIB for instance, in cases of non-compliance by client, only ‘consults with the Client on corrective measures to rectify any failures … ’ (AIIB, 2016, para. 62). If the Client fails to re-establish compliance, AIIB ‘may’ exercise its contractual remedies (AIIB-ESF, para. 56, under the heading ‘Notification of changes’). For comparison, the Equator Principles that govern private project finance, although inspired from the MDBs’ environmental and social safeguards, are stronger in imposing borrowers to obey their relevant obligations: the private financing institutions having adhered to these voluntary code ask their clients to covenant, in the financing documentation, to comply with all relevant host country environmental and social laws, while MDBs only note that the responsibility belongs to borrowers.
But the real problem in need of solution (stage 4 in the CDA model we apply) is not the quality of the texts – solutions to this are known, some being mentioned in the previous paragraphs – but how to persuade MDBs to accept a version of the ESGs more in line with their declared aims, i.e. promoting business that contributes to poverty alleviation without abusing human rights and the environment. In other words, we have a set of standards that in a certain perspective on international law have the potential of becoming ‘law’, if they attain sufficient clarity; the problem is how to convince the issuers to amend them to this end.
Here is where the above discussion on the national/cultural and the temporal/rational contexts of our text of focus becomes important. This paper's aim is to contribute to positive (from our already stated standpoint) change in the AIIB-ESF. Since we have shown that this policy is weaker then those of other MDBs in certain areas, and this weakness can be attributed to Chinese cultural patterns of normative design and rational patterns of economic interest, the question is whether these patterns can be reversed. After all, for almost four decades now, ‘good governance’ has permeated development discourse and especially research agendas, conferences, publications etc. funded by public and private banks and bilateral donors, yet normative progress has not followed (Reference WeissWeiss, 2000, p. 796). Can it follow now in the case of AIIB-ESF?
We suggest as answer a cautious ‘yes’, because a policy window (concept proposed in Kingdon, Reference Kingdon1995) may have been opened by the above mentioned sensitivity of Chinese rulers to reputation in the context of China's recent accession to a world leading position in terms of economic, politic and military might. Chinese policy makers, including those with authority in AIIB, have learned in recent years that reputational damage from overseas failures of Chinese entities, including institutions lending for controversial megaprojects, outweigh the economic advantages of keeping low environmental and social standards. The ‘process sequencing’ model of dealing with the temporal dimension of policy making may explain why the AIIB-ESF's authors could be now willing to give them a shape more acceptable by the civil society, thus breaking away from previous patterns. After comprehensively reviewing the different models of explaining the influence of temporality in policy change, Howlett (Reference Howlett2009, p. 252) argues for the superiority of this model based on the fact that it stresses ‘ … not how outcomes at historical switch points are unpredictable, almost random, accidents, but how they are firmly based or rooted in previous events and thinking as related structural processes of both negative and positive feedback affect actor behaviour’. Under this model, a new policy sequence emerges in response to reactions and counter-reactions, mobilisation and counter-mobilisation, with the negative feedback from relevant actors being especially important (Reference HowlettHowlett, 2009, p. 253).
To show how relevant actors mobilisation and counter-mobilisation can re-sequence established policy patterns, a telling parallel can be made with the radical change that occurred in the 1990s in the way corruption was treated in the World Bank policies. Until then, Bank managers were forbidden from openly discussing corruption matters relevant to Bank projects, because corruption was labelled a political issue (Reference ClarkClark, 2002, p. 221). The groundbreaking reversal of approach is usually associated with the famous characterisation by the WB president of corruption as a ‘cancer’, but the engine for change was the US crusade for imposing global rules against foreign bribery. Normative progress in the field of anticorruption was thus due to a combination between the material, economic US’ interest of reaching a level playing field for American business (which already had to comply with severe regulation of this type) with value-motivated NGO activity, mainly Transparency International. The quest for normative progress in the field of environmental and social safeguards calls for a similar scenario, i.e. the corroboration of normative persuasion from value actors and strategic bargaining from materially motivated actors; the same unexpected alliances that worked in the case of corruption may work here.
For the specific case of AIIB, the interest-motivated action may come from participants to the development lending market. One reason for which MDBs adopt these safeguards is drawing the line for lawful conduct in order to exclude ‘black sheep’ from the developing lending market (Reference Glinski, McBarnet, Voiculescu and CampbellGlinski, 2007, p. 121). A good illustration is the existence, in all ESGs, of areas already at an acceptable standard from a social justice perspective, like the ones on transparency and public participation; here, as the first development bank to emerge after more than two decades, AIIB has a significant contribution to solidifying the norms it chose to appropriate. In fact, the very adoption by the Chinese-led MDB of these standards, which are above domestic standards in some areas,Footnote 13 is generally an illustration of this mechanism of soft-normative progress based on avoidance of the ‘black sheep’ status in a particular industry. In this perspective and given the above discussion on China's reputational interests, China is itself an interested actor, alongside AIIB's competitors in the developing lending market.
As for the value-motivated actors, the particular political context, with many Western democracies having endorsed AIIB and with its ESGs in their infancy, creates favourable conditions for addressing the weaker standards. Here again, the presence of Western democracies like Australia, UK, Germany or Sweden among the AIIB founding members is relevant. The shareholding imbalance ensures that the credit for development assistance will go to China, but as shown above, any environmental and social wrongdoing will be also ‘credited’ to China. Since the Western democracies have little to benefit but also little to lose in terms of reputation, they are suitable value-motivated actors in the quest for the safeguards’ improvement.
In fact, what we propose in stage 4 of our analysis is an application for the specific AIIB context of what was suggested recently by Daugirdas (Reference Daugirdas2014, p. 998) as a general strategy for enhancing IOs responsibility: an intensified transnational discourse – including, as participants, government and IO officials, NGOs, national legislators, courts, political leaders, and private individuals – shaped by the International Law Commission's Draft Articles on the responsibility of international organisations (issued in 2011) and meant to leverage upon the states and IOs need for international reputation as source of legitimacy. The discourse of the non-powerful entities, in our case, should strategically aim at two powerful entities – China, as the main shareholder, and the Western liberal democracies, as important contributors – and make use of what was called ‘consentful contention’, i.e. make the officials ‘prisoners of their own rhetoric’ (Reference ThompsonThompson, 1975, p. 263). That is to say, request policy makers in China to improve the AIIB standards, and those in the UK, New Zealand etc. to live up to the expectations they created when endorsing AIIB, or otherwise point to either of them as hypocritical.
By analysing critically the discourse of AIIB in its institutional, national and historical contexts, this paper purports a dual contribution: to the CDA literature, by adding to the types of discourse susceptible to being scrutinised with this methodological tool, and to the civil society's efforts to make development banks more accountable for the harm that often follows the projects they finance. The authors of course do not have the naivety to imagine that quality safeguards can emerge suddenly just because China will feel ashamed by the critiques of the text drafted by AIIB's personnel, or because New Zealand will feel embarrassed for having its brand, as one of the most environmentally and socially advanced countries in the world, used as umbrella for questionable operations in Asia. This transnational discourse is only the guerrilla part of the strategy for change, an ongoingFootnote 14 one that is necessary but insufficient; what is needed in order to make the ESGs count, to make them more than economic interests dressed up in normative clothes only as much as to cover their shameful parts, is a turning point – one of those moments in history that changed established paths when few would have believed this possible. Together with normative persuasion, including by articles of this type, the action of a strongly (self- or value-) motivated powerful actor is necessary.
Disclosure statement
No potential conflict of interest was reported by the authors.
