Banks Do Not Assess Project Risks Accurately. People and our planet pay the price.

Do higher risk projects generate more complaints?

When development financial institutions (DFIs) review proposed investments, they often assign a risk categorization to the project to reflect the level of risk and harm to local communities and the environment that might result from the project.  Every institution has its own unique way of categorizing risk, but the general pattern followed includes assigning projects to one of four broad categories: (1) Category A, or high risk projects, representing the highest level of potential risks and harm; (2) Category B, or moderate risk projects, representing limited and often reversible risks or harm; (3) Category C, or low risk projects, with minimal or no negative impacts; and (4) Category FI, or financial intermediary projects, representing investments flowing through other financial institutions, whose risks have only more recently begun to be subjected to more scrutiny and nuanced risk assessments. 

The categorization of risk assigned to a project matters a great deal to communities who bear the brunt of project impacts.  The project risk category dictates which of the DFI’s safeguard policies are triggered: the higher the risk, the more safeguards are put in place to ensure adverse impacts are properly identified and mitigated. As an example, under the IFC’s current policy from 2012, certain project information has to be disclosed to affected communities regardless of the risk. But, if there are moderate risks identified, consultations must also be undertaken to allow communities to express their views of the project risks and to allow the Bank’s client to consider and respond to those views. If there are significant risks identified, i.e., a project is considered Category A, there must be a robust informed consultation and participation process where clients are required to take into account the views of affected communities in their decision-making processes.

Given the very low numbers of projects with category A ratings (4% of IFC Investment Services projects; 9% of World Bank projects; 1% of Inter-American Development Bank projects;  and 12% of Asian Development Bank’s sovereign projects)*, we wanted to understand if complaints submitted reflected these same proportions.  

To explore this, we focused on four institutions where we had the most robust datasets: International Finance Corporation (IFC), World Bank (WB), Inter-American Development Bank (IDB) and Asian Development Bank (ADB).  We had hoped to also include the European Investment Bank (EIB) in our analysis, as they are one of the largest DFIs in the world, however we were unable to find data either through the EIB Services website or through the EIB Complaints Mechanism website to identify the risk category of projects at the assessment phase.

We learned that the independent accountability mechanisms of these institutions consistently have more than 25% of their complaints coming from Category A projects, and for the WB, IDB, and ADB, more than 40% of complaints were about category B projects.** For the IFC, 29% of its projects were Category B but because of the CAO’s data gaps, we were unable to match a risk category for a third of complaints submitted.  It’s important to note that complaints have been raised across all risk categories, including 1 to 8% of complaints for Category C projects across all four banks and 0 to 8% of complaints for FI projects across all four banks. 

Note: In addition to the categories A, B, C, FI mentioned above, we include a few more categories in this chart. Mixed refers to complaints implicating multiple investments/projects with different risk ratings. Uncategorized projects are often advisory services or technical assistance projects that are considered by the institution to be out of scope for their safeguards policies and therefore are not assessed for their environmental and social risks. Unknown refers to complaints where insufficient information is provided to identify the project and its risk category.

We draw two conclusions from this.  First, category A projects seem to be leading to a disproportionately high number of complaints. Second, category B projects, though purportedly having less adverse impacts, continue to generate a significant number of complaints to IAMs.  

The drivers for complaints

To better understand the drivers for these conclusions, we looked across all the data we have where we were able to match a complaint to a project risk category and pulled out the top issues identified across complaints.  

Common issues raised in complaints

The same five issues rose to the top for both Category A and Category B projects (the only difference being that "unknown" issues from undisclosed complaints pushed "livelihoods” to sixth place for Category B complaints).  Top issues raised in complaints—displacement, livelihoods, and community health and safety—point to significant adverse harms being felt by communities, regardless of whether the project was classified as Category A or B.

Consultation and disclosure

One notable difference is that inadequate consultation and disclosure was by far the most complained about issue for Category B projects.  This makes sense when you consider that Category B projects typically have lower requirements for consultation and disclosure than Category A projects.  This lower standard is keenly felt by communities impacted by Category B projects. 

Take for example the community members in Ukraine affected by an EBRD, IFC, and DFC-funded mega poultry farm run by MHP, who have been waiting years to get remedy for the pollution and road impacts, health and safety risks, and retaliation generated by the project.  As described in a recent opinion piece, the problems stemmed from an under-categorization of the project by its lenders (it was assessed as category B), which resulted in inappropriate board  approval, lack of access to information and participation, insufficient  guarantees and inadequate oversight over a mega-agribusiness project.  Both the EBRD and IFC’s independent accountability offices are expected to release investigative reports on whether these institutions should have done more to acknowledge the risks of the projects in order to prevent harm to communities and the environment.

Misclassification of projects

Next, we narrowed down and looked only at complaints that were found eligible for a compliance investigation or dispute resolution process.  For complaints that went through a compliance investigation process where there was a finding of non-compliance, we see very similar proportions of category A and lower risk categories represented.  

Delving deeper into the data we found ten examples, at least one from each of the four institutions we reviewed, where the institution’s IAM made a finding that a category B project was incorrectly categorized or that the assessment to categorize the project was inadequate.*** This is a shocking prevalence, given how few complaints make it to this substantive stage (only 10% across these four institutions).  (We have previously written about the various (unjustifiable) reasons given for rejecting complaints and barriers to accessing these processes.)  

The impacts in such cases can be severe, as seen in São José Dos Campos in Brazil, where community members of Banhado were forced out of their homes without adequate consultation or adequate resettlement offers.  The IDB’s mechanism found that that this category B project should have been categorized as a high risk project and wrote, “the great uncertainty and insecurity felt by the Banhado residents regarding their future over the nine years the IDB was involved in the resettlement plans, during which they were not attended to adequately in accordance with the Relevant Operational Policies, constitutes moral harm linked to the Bank’s noncompliance.” 

Dispute resolution and confidentiality

For complaints that went through a dispute resolution process where an agreement was reached, we see far more Category B projects represented, with the exception of the IFC CAO, where Category A and B projects are represented in similar proportions. Unfortunately, many of these agreements are confidential, so it’s difficult to understand the drivers for these differences. 

Conclusion

Given the significant adverse impacts to communities raised across projects with different risk categorizations, DFIs need to significantly improve their risk categorization in the first place and the application of the safeguards to all projects, regardless of a project’s categorization.  

Given the few numbers of complaints that manage to get an investigation report, we think it’s significant that a sizable portion of those reports make a finding of ‘inadequate classifications’ or ‘inadequate assessments to support the classification’.  DFIs must heed these findings to re-classify projects, activate proper safeguards for the project, and ensure future investments avoid these mistakes.

DFIs must start with adequate consultation and disclosure. This is a cornerstone safeguard to ensuring that proper mitigation measures are in place for other issues.  Without robust consultation and disclosure processes, DFIs and their clients will miss important adverse impacts that eventually cause harm and cost the Bank and their client more to address after-the-fact.  This is especially true for projects classified as Category B, where less robust consultation and disclosure standards are applied. A key part of this consultation and disclosure is ensuring that communities know about the grievance mechanisms available to them, including the instituiton’s IAM.  This helps ensure that communities can raise concerns about all projects, even projects that may not initially appear to be risky. 

Finally, DFIs should also revisit their safeguards around displacement, livelihoods, and community health and safety, and they should adhere to them regardless of risk classification.  It is striking that these issues consistently rise to the top whether a project is categorized as high risk or moderate risk.  Clearly more work needs to be done around improving safeguard measures for these issues and/or ensuring the highest safeguards are appropriately applied when these issues are raised.

ENDNOTES

* These percentages exclude projects for which no risk rating was included.  

** A note on methodology.  We recognize that every bank engages in its own unique risk analysis, which might not directly map onto or align with other institutions.  For the purposes of this analysis, we attempted to standardize across the different institutions’ risk categorizations, mapping each institution’s risk rating into one of the following categories: A (high risk); B (moderate risk); C (low risk); and FI (unrated financial intermediary investment). 

*** The ten complaints include: Amaggi Expansion-01/IFC Executive Vice President Request (IFC CAO complaint), Dinant-01/CAO Vice President Request (IFC CAO Complaint), Eco Oro-01/Bucaramanga (IFC CAO complaint), Forest Concession Management and Control Pilot Project (WB IP case), Program to Improve Highway Corridors in Paraguay (IDB MICI complaint), Quellaveco Mining-01/Moquegua (IFC CAO complaint), Third Power Project, Fourth Power Project, and proposed Bujagali Hydropower Project (IFC CAO complaint), Transitional Support for Economic Recovery Credit and Emergency Economic and Social Reunification Support Project (WB IP complaint), Sustainable Urban Transport Investment Program - Tranche 3 (ADB SPF CRP complaint), Sao Jose dos Campos Urban Structuring Program (IDB MICI complaint).