AIIB's proposed accountability mechanism needs improvement

The letter below was sent by more than a dozen international NGOs to the AIIB on March 26, 2018, expressing concerns over the draft policy for a new Independent Accountability Mechanism (IAM) at the AIIB, and making recommendations for its improvement.

If the policy is adopted without significant reforms, we would have strong reservations that it can truly function as an IAM and remedy the harms to project-affected people. Far from adopting best practice from other IAMs, the draft policy would create a mechanism that is markedly less accessible and transparent than most other IAMs.

Dear Mr. Sharif:

We appreciate the opportunity to comment on the draft Project-Affected People’s Mechanism (PPM) policy. As you know, our organizations have extensive experience supporting project-affected people and communities in using Independent Accountability Mechanisms (IAMs) of development finance institutions to prevent harms and seek redress. Our recommendations are based on the unique insights we have gained from this experience. As such, we trust that our recommendations will be carefully considered and incorporated into the final version.

We have closely reviewed the draft policy and, while we are pleased to see the contours of a mechanism with a dispute resolution, compliance, and advisory function, regrettably, we have serious concerns about fundamental aspects of the proposed mechanism and each of these functions. If the policy is adopted without significant reforms, we would have strong reservations that it can truly function as an IAM and remedy the harms to project-affected people. Far from adopting best practice from other IAMs, the draft policy would create a mechanism that is markedly less accessible and transparent than most other IAMs. Despite its proposed name, and repeated assertions in the draft, we do not agree that the mechanism will be “people-centered.” We are of the firm view that its structural deficiencies, coupled with the procedural hurdles it poses to potential users, will result in an IAM that fails to achieve its core purpose of preventing and resolving harms experienced by project-affected communities, and enhancing institutional accountability and continuous learning within AIIB. We urge you to substantially revise the draft policy to ensure that the mechanism is empowered and equipped to meet these important goals.

This submission is composed of three parts.

  • This cover letter sets out our most significant concerns, organized according to the following three principles: accessibility, transparency, and legitimacy. We note with concern that these are not included among the policy’s own five overarching principles.
  • Annex 1 assesses the draft policy against the initial best practice recommendations contained in our first submission dated 23 June 2017. We find that the draft policy only fully incorporates 12 of the 60 recommendations.
  • Annex 2 is a marked-up version of the draft policy, identifying the specific paragraphs and text that we find problematic, with cross-references to our recommendations where relevant.

Accessibility

The draft PPM policy makes a prediction that the mechanism will not receive many complaints in its first few years of operation,1 basing this conclusion on the caseloads at other IAMs. Our analysis of the draft leads us to a similar conclusion, but for different reasons. The draft policy describes a mechanism that would be inaccessible to many project-affected communities, creating barriers that would prevent or dissuade its use.

Exclusion of complaints concerning co-financed projects: The draft excludes an entire class of complaints—those relating to projects co-financed by other international financial institutions (IFIs) when the policies of that institution, and not of the AIIB, apply. Complainants should be able to choose the mechanism most appropriate to their case and that will best serve their needs. As is the case at other mechanisms, when complaints are filed at more than one, the IAMs can coordinate to maximize positive outcomes, both for the case at hand and institutional learning and accountability. Accountability must follow the investment, regardless of which policies apply. As long as the AIIB is invested in a project, it should be accountable for that project’s impacts and learn from its shortcomings. As the draft stands, however, not only are co-financed complaints barred from the PPM, but the AIIB would have no obligation to cooperate with other IAM processes or respond to the findings of another institution’s IAM, effectively absolving the AIIB from all responsibility for the impacts of these investments. Nor is any process established for the AIIB to learn from such cases – an important purpose of IAMs, especially at a new institution. As a result, institutional accountability would be absent. This restriction on accessibility is particularly concerning in light of the fact that the majority of the AIIB’s current portfolio consists of projects co-financed with other IFIs.

Unrealistic requirements to demonstrate “substantial” harm: The eligibility requirements of the draft policy set the bar unrealistically high for project-affected communities. Submissions must “make a credible case of potential or actual adverse impact or harm.” A footnote to that paragraph further specifies that the “substantial” adverse impact or harm must be demonstrated, to the satisfaction of the PPM.”2 There are two problems with this. First, there is a risk that the PPM will unduly deny complaints by applying arbitrary and unreasonable definitions of “substantial adverse impact.” Second, the implication that complaints must include evidence to demonstrate harms creates an impractical barrier for complainants. Affected communities are almost always poor and marginalized, and sometimes illiterate, and may not have access to legal representation or support (which is indeed restricted by the PPM policy). At other mechanisms, an initial assessment process allows the IAM to conduct preliminary research, including through interviews of the parties to establish whether there is a valid basis for the complaint, and that it is not frivolous or vexatious. This is an important initial part of the IAM’s function, and the burden of providing evidence up front should not be shifted to affected communities. Relatedly, the draft policy requires that for complaints on projects that use country or client systems, the complainants must establish that AIIB has been “grossly negligent” in determining that the country/client environmental and social protection standards are materially consistent with the AIIB’s own. This legalistic requirement is wholly unrealistic and inappropriate for a “people-centered” complaints mechanism.

Unreasonable preconditions to filing submissions: The draft policy establishes two pre-conditions for filing a complaint: using the project-level grievance mechanism and approaching AIIB staff. No other IAM requires a complainant to first access a project-level grievance mechanism, in part because the project-level grievance mechanism is often designed and operated by the very actor causing the harm. The evidence shows that project-level grievance mechanisms are not effective at addressing the types of complex issues that are usually presented to IAMs. We understand from our engagement with the MD-CEIU that this pre-condition is intended, in part, to strengthen the acknowledged short-comings of project-level grievance mechanisms. If they are accessible, independent, and effective, project-affected communities are likely to use project-level grievance mechanisms; however simply requiring their use will not make them more fair and effective, and instead exposes already vulnerable people to the risk of retaliation without the protection of an external legitimate complaints handling process. We would instead encourage the AIIB to work with or urge its clients to improve project-level grievance mechanisms, which, if effective, would likely lead to fewer cases of harm and thus a reduced need for complaints to the PPM.

Exclusion of nonlocal assistance: The draft would restrict complainants from receiving assistance from “nonlocal” civil society organizations unless that assistance is “adequately justified” by complainants and “endorsed by the PPM.”3 The term “assistance” is so broad that it could be interpreted to prohibit all forms of advice, funding, or other support from regional or international CSOs. If adopted, this would be an astonishing attempt to curtail CSO cooperation and activity. We understand from the MD-CEIU that this provision is intended to help build the capacity of local civil society organizations. While we agree with the notion that local CSOs, which play a crucial a role in supporting communities, should have every opportunity to increase their experience and expertise, we do not think that the PPM is well-placed to build CSO capacity, or that excluding the involvement of international CSOs is the most effective way to do so. In fact, our experience consistently shows that the best outcomes – both for cases and capacity building – are reached when experienced international CSOs work side-by-side with local CSOs to support communities using IAMs. Multiple factors, such as capacity levels, political environment, and security risks, determine whether affected communities request local and/or international CSO representation and other assistance in navigating IAM processes, and the nature and degree of that support. Thus, even if the language in the policy were revised to apply only to nonlocal representation of complainants,  we would still be concerned. If the policy were truly people-centered, it would allow project-affected communities to work with or be represented by anyone of their choosing –just as the AIIB and its clients may choose to involve their legal counsel in IAM cases, when they wish to
do so.

Confusing complaint nomenclature: Instead of simply establishing procedures for submission of “complaints,” the draft PPM creates three types of “submissions” – a “concern,” a “request for dispute resolution,” and a “request for compliance review.”

There are multiple problems with this complex typology and submission filing structure. For one, the terminology is confusing. A “concern” is vaguely defined as a project-related issue that “has not yet crystallized into an identified dispute.”4 Additionally, it is unclear whether there is any practical difference between a “concern” and a “request for dispute resolution,” as both seemingly result in a dialogue process. Further, the preconditions and requirements for filing a “concern” and also a “request for dispute resolution” are identical to those for filing a request for a compliance review: in order to file a “concern,” project-affected people must have already approached the AIIB and been dissatisfied with the response received.5 They must also meet the high bar of making a credible case of potential or actual substantial harm.6 It is clear, therefore, that all three submission types are the same in substance.

Second, given the complexity of investment projects, in reality many cases involve both “concerns” about future impacts –as for example, a project expands or evolves into subsequent phases of development or operation– and complaints about actual harms that have already occurred –for example, during the early stages of development. Therefore, separating these two categories is artificial and impractical, and will ultimately cause the PPM logistical difficulties in handling complaints in an efficient and effective manner.

Thirdly, since all three submission types are true “complaints” as understood by all other grievance mechanisms, the complex typology is unhelpful in achieving the AIIB’s stated goal of “destigmatizing” the complaints process. We could not agree more that “problems are part of the difficult business of development and to satisfactorily solve them is, in fact, an integral part of good development practice.”7 “Destigmatizing” the complaints process requires a cultural shift within the institution. That shift will not be achieved by calling a complaint by another name. Instead, characterizing a complaint as a “concern” or a “request for dispute resolution” serves to marginalize and minimize complainants’ grievances – and therefore ironically to stigmatize these genuine complaints.

Complex and rigid filing system: The complex submission structure acts as a barrier to access. Although the policy claims to take a people-centered approach, it proposes a system that would be difficult for project-affected communities to navigate, requiring them to identify which type of complaint to file at what stage in the project and overcoming obstacles in order to transition from one function to another. There should be no hierarchy in the functions that a mechanism performs. Dispute resolution and compliance review are two different roads that should lead to the same destination, remedy for the complainant and accountability of the institution. The policy itself stigmatizes compliance review by making it more difficult for complainants to request it in the first instance, or to switch from dispute resolution to compliance review, referring to the latter as “escalation.”8 Further, if the dialogue process undertaken for a concern and/or a request for resolution fails before the project has been approved, the complainant must wait until project approval to file a complaint requesting compliance review. There is no reason compliance review could not occur prior to project approval since the bank has standards that apply, and against which compliance can be assessed, during the due diligence phase of the project. Indeed, the mechanisms of the World Bank Group, the AfDB, and the ADB, among others allow this compliance review at pre-approval stages.

A more people-centered approach would allow project-affected people to file a complaint as early in the project cycle as possible and choose whatever function or functions best suits their needs and the sequence of those functions. Confusing and impractical roles of various parties: The draft policy is unclear and internally inconsistent about the role of complainants, clients, and the AIIB in the different functions. For concerns, the policy seems to suggest that the PPM and Bank Management will identify “timely and cost-effective ways in which the issues raised by the concern can be reasonably addressed”9 without input from the complainant on the adequacy of those measures to address the issues raised. Elsewhere, the policy says that concerns would result in a dialogue between “AIIB staff, the Project-affected people and or Client concerned.”10 That appears to allow for the possibility that the client but not the complainant would be consulted on the outcome. In dispute resolution, it is the client whose role is unclear. The text seems to suggest that the dialogue would take place between AIIB staff and the complainant, but not the client. The bank would “ensure that the Client is kept fully informed during the resolution of eligible concerns and disputes raised by Project-affected people, and as needed work with the Client to ensure appropriate implementation of the agreed measures.” This is contrary to the way dispute resolution is conducted by all other IAMs in which the dialogue takes place between the complainants and the client, and often without the participation of bank staff. The proposal is unworkable since any effective dispute resolution process must involve the primary parties to that dispute, which will almost always be the project implementer/owner, whose activities are directly causing the harms and the community experiencing those harms. While project financiers, such as the AIIB, play an important role in dispute resolution and remediation, the primary actor cannot be feasibly removed from the process if it is to have a
chance of success.

Transparency

It is standard practice for IAMs to publish all relevant documentation regarding a complaint, including the complaint, the eligibility determination, the final report or agreement, and monitoring reports (subject to complainant confidentiality requests). The draft appears to commit to disclosing only summaries of PPM findings and assessments.11 That will prevent complainants and other stakeholders from monitoring the PPM to ensure it is fulfilling its mandate. Similarly, access to the compliance review reports and the bank’s response to them
are necessary to ensure that the bank is meeting its environmental and social obligations and its responsibilities under the complaint process. For the mechanism to be regarded as credible, it is essential that there is disclosure of these documents.

Legitimacy

Legitimacy refers to a mechanism that, through its structure and procedures, is capable of achieving its mission, and which engenders trust from the people who would use it. The PPM policy as drafted creates serious problems of legitimacy.

Conflicts of interest: The policy claims that combining oversight functions into one unit – the CEIU – is a “governance innovation among international financial institutions,”12 without explaining how this approach actually improves on previous models. There is a reason why other IFIs have created separate, independent offices for these functions: such separation and independence are critical for ensuring the success of the mechanisms’ work. Although the MD-CEIU has a reporting line to the Board, the mechanism itself is not independent from the other functions of the CEIU unit. The policy is explicit that the staff in the unit will multi-task, presumably working on PPM and the other functions within the unit. Conflicts of interest may arise between the evaluation function and the PPM if a complaint is received on a project that the evaluation team has already reviewed. Any potential or actual conflicts of interest would, it asserts, be adequately and effectively addressed. Even if it were possible to ensure a separation at the staff level, it is difficult to envision measures that would adequately address the conflict, much less the appearance of one. After all, the MD-CEIU directs all of the functions within the unit.

Impractical vesting of multiple responsibilities in the MD-CEIU: The policy is explicit that the staff in the unit will multi-task. The MD-CEIU is tasked with directing all functions of the PPM, even chairing every task-force created to conduct a compliance review. The structure yields serious risk that staff will be too burdened to give each complaint the care it requires, and that the MD-CEIU him or herself will end up being a bottle-neck in the PPM’s process.

Lack of independence from board: The mechanism also lacks additional hallmarks of independence found at other IAMs. Key among these is the mandate to determine whether a compliance review is warranted without approval from the Board. The PPM policy should contain clear eligibility criteria for the mechanism to apply in making such a determination. Allowing the Board to overrule that determination would undermine the independence of the mechanism.

Undue discretion on remedial action plans: The draft policy leaves it to the discretion of the PPM and the Board to request that Management prepare a remedial action plan to respond to findings of non-compliance. Should the PPM or the Board allow the bank to ignore the PPM’s findings of non-compliance, the legitimacy and the independence of the mechanism would be in serious jeopardy. Management should be required to prepare an action plan any time an investigation finds non-compliance.

No external stakeholders in selection of key staff: The draft policy also lacks provisions to include external stakeholders on selection of key staff and strong pre- and post-employment restrictions—all of which further promote the independence of the mechanism.

Conclusion
Given the above significant concerns, in addition to those detailed in the annexes, the undersigned organizations call for a substantial revision of the policy before it is considered by the Board of Directors for its adoption. If you have any questions about this submission, please contact Kris Genovese (k.genovese@somo.nl), SOMO, who will coordinate our organizations’ responses and input.

Sincerely,
Accountability Counsel
Bank Information Center
Bank Information Centre Europe
Both ENDS
CEE Bankwatch Network
Center for International Environmental Law
Centre for Research on Multinational Corporations (SOMO)
Forest Peoples Programme
Green Alternative
Inclusive Development International
International Accountability Project
NGO Forum on ADB
Oxfam
Urgewald