Two weeks ago, on Friday 28th October, the Dutch Banking Association and thirteen Dutch banks joined civil society groups, trade unions and the Dutch government in The Hague to sign the long-awaited “Dutch Banking Sector Agreement on international responsible business conduct regarding human rights” , a.k.a. “the covenant”. BankTrack is not a signatory to the agreement, and although we were consulted as a stakeholder at the early stages of the process, the later development of the covenant was shrouded in secrecy.

So when the covenant was published, we were eager to see what was in it. In summary, it’s an ambitious agreement with some far-reaching commitments, but it falls short in some important areas, such that even if fully implemented it is unlikely to lead to banks meeting all their human rights responsibilities. In this post we take a more detailed look at some of the covenant’s strengths and weaknesses and what it might mean for the sector as a whole.

A product of the polder model

First some background on who signed this agreement, and where it came from. There are four delegations of Parties to the agreement – the Dutch Banking Association (NVB), the Dutch Government, two trade union federations, and three civil society organisations (Amnesty, Oxfam Novib and PAX). Thirteen “adhering banks” - including Rabobank, ING, ABN AMRO, SNS Bank and Triodos – also signed the agreement as part of the NVB’s delegation. (For the full list of signatories, see the press release). The agreement remains open for further signatories, including banks and civil society organisations, without restriction to Dutch organisations.

If it seems unusual for such diverse groups to get together around the table and reach a 22-page agreement on human rights, it is – but it also reflects a typically Dutch model of consensus decision-making known as the “polder model”. (The name refers to the notion that in a low-lying country such as the Netherlands, cooperation between all relevant parties, regardless of their denomination, is essential to deal with fundamental issues like dyke maintenance and flood protection). The Social and Economic Council of the Netherlands, the SER, which is the highest advisory body to the government and brings together government, employers, trade unions and civil society groups, epitomises this polder model, and it is they that convened these disparate groups to develop the covenant.

It’s also worth noting that the initial plan of the SER was for a covenant on how banks can meet all their responsibilities under the OECD Guidelines, which includes chapters on environmental impacts, industrial relations, bribery, taxation and other topics alongside human rights. The banks however objected to this ambition, opting instead to first focus on their human rights obligations, to which the other parties reluctantly agreed.

Scope: project finance and corporate loans only, so far

The resulting agreement is a commitment by the 13 adhering banks to meet their human rights responsibilities under the UN Guiding Principles (UNGPs) and OECD Guidelines, through policy commitments, due diligence, tools for information gathering and sharing, transparency and reporting requirements, with further commitments from the other parties to support them in this endeavour where they can.

The first surprise in the agreement is that after narrowing the focus to only human rights, the scope is narrowed even further, to only project finance and corporate loans . Although these are probably the major source of human rights impacts for the adhering banks, this misses out a wide range of other bank activities, from asset management to advisory services and investment banking.

The banks may eventually go further than this – the agreement calls on asset managers and institutional investors to produce their own covenant by the end of 2016 (they had better be quick!) and says the parties to this agreement will take responsibility and develop their own approach in 2017 if this doesn’t happen. But given that bank responsibilities to respect human rights extend to all their activities and business relationships, we would have liked to see the agreement avoid any further limitations in scope.

Consideration of impacts on rights holders and Indigenous Peoples

The covenant goes into welcome detail on human rights due diligence, including focusing on risks to rights holders rather than on risks to the bank and on time-bound improvement plans to address impacts, with ending the relationship as the sanction of last resort. It also emphasises that bank clients must consult with rights holders – although it could go further in ensuring their voices get through to the bank itself.

There is also a focus on Free, Prior and Informed Consent (FPIC) for Indigenous Peoples, which is particularly welcome as it is not explicitly mentioned in the UNGPs. Banks commit to require their project finance clients to ensure FPIC is in place where called for by the IFC Performance Standards or FAO Guidelines on Responsible Governance of Tenure. This is good, although most Dutch banks that do project finance must already follow the IFC’s standards through the Equator Principles. The agreement also goes beyond this by encouraging FPIC, as well as meaningful and effective consultation with other potentially affected groups, for corporate lending clients, with annual reporting to the Steering Committee on efforts and results. We hope this reporting is made publicly accessible so the results can be scrutinised.

FPIC is recognised in UNDRIP as a specific right for Indigenous Peoples, and while the agreement does not ask banks to extend this right to non-Indigenous Peoples affected by their finance, the Parties do agree to promote and explore the broader application of FPIC for other legitimate tenure holders – also a welcome step.

Tools for information sharing

Parties to the agreement further commit to a great deal of information sharing, e.g. on due diligence best practices and any other human rights research they conduct. Parties also agree to work towards a publicly accessible human rights “matrix or database” . There’s also a commitment to jointly carry out between two and four value chain mapping exercises over the three year course of the agreement, on palm oil, cocoa, gold and potentially oil and gas, make these publicly available, and draw conclusions on the results.

All Dutch banks to meet UNGP Reporting Framework

Reporting and transparency are crucial if we are to understand how banks are implementing the covenant, and on this the agreement is quite strong. All adhering banks commit to work towards reporting in line with the UNGPs Reporting Framework, a strong commitment given that ABN AMRO is the only bank globally that is on record as committed to report to this standard. The adhering banks also agree to publish detailed information on their exposure to different economic sectors, which to date few banks have done publicly.

A specific gap that the covenant doesn’t address is the need for banks to publicly account for how they address specific human rights impacts raised by affected stakeholders. Banks are poor at this, often refusing to discuss specific customers outright, as our Human Rights Impact Briefings evidence. But overall the covenant should lead to much improved human rights reporting from banks, and we’ll be paying attention to make sure the results are up to scratch.

A complaints procedure that is NOT a grievance mechanism

BankTrack’s research into bank human rights policies and processes has found that grievance mechanisms – transparent processes for handling complaints about human rights impacts - are one of the biggest gaps in bank practice. Although the UNGPs state very clearly that “business enterprises should establish or participate in effective operational level grievance mechanisms”, there is little sign that banks understand this applies to them, and not just their clients.

Strangely, the agreement includes a commitment for banks to create a complaint procedure that is publicly accessible for employees, clients and third parties – but specifies in a footnote that “this complaint procedure is not a grievance mechanism”. The development of complaint procedures by banks is a welcome and necessary step forward, but this caveat makes clear they won’t meet the effectiveness criteria set out in the UNGPs. It looks like the banks are committing to develop complaints procedures, but only as long as they don’t need to be effective.

Other avenues for remedy - an incomplete patchwork of mechanisms

As well as the adhering banks’ own complaints-channels-that-are-not-grievance-mechanisms, the agreement creates requirements for bank clients to take their responsibility to establish grievance mechanisms that do meet the UNGP effectiveness criteria – requiring them for project finance and encouraging them elsewhere. Again this is positive, but it contrasts with banks’ apparent refusal to take their own responsibility on this score. Banks would benefit more from asking clients to establish grievance mechanisms if they formally participated in these – in line with their responsibility – for example, by being notified of all complaints and taking a role in their remediation. It also gets things the wrong way around, with serious potential impacts for rights holders, by asking corporate clients to establish grievance mechanisms “in the case of severe human rights violations known to the banks” . Companies cannot wait until there is a severe human rights impact to set up channels for grievances – it is too late by that point.

In addition, the NVB will establish a “voluntary advisory expert mechanism” to handle “selected notifications of an alleged breach of the OECD Guidelines", including on human rights . In the first instance this will only handle complaints related to project finance. However, this “expert mechanism” is also not a grievance mechanism – indeed its role is described as being to advise the financial institution on how to handle cases brought, rather than to actually remediate the impact.

Finally, there is an internal dispute mechanism . Although this is only for handling complaints from one Party to the agreement about another, its decision will include recommended time-bound follow-up steps, and may be made public. This could be the strongest avenue for communities or organisations with a human rights grievance against one of the signatory banks to seek remedy – by asking one of the civil society signatories to raise it on their behalf.

Overall, the covenant has a patchy and incomplete approach to providing access to remedy. The UNGPs have only one requirement of multi-stakeholder human rights initiatives – that they should ensure effective and accessible grievance mechanisms are available (UNGP 30). It’s unfortunately far from clear that this agreement lives up to this requirement.

Exploring “cause, contribute and directly linked”.

The covenant also includes an agreement to establish a working group to report on when a bank is deemed to be ‘linked to’, ‘contributing to’ or ‘causing’ specific human rights impacts , by September 2016 - although we understand the working group is not yet in place. As the UNGPs give businesses a responsibility to remedy impacts they cause or contribute to, but not necessarily those to which they are directly linked through a business relationship, clarity on what these terms mean for the finance sector would be helpful.

This is not the first time such a report has been discussed. The ongoing OECD project on Responsible Business Conduct in the Finance Sector is investigating the matter, and the so-called Thun Group of banks, an informal grouping of banks now for years already pondering the implementation of the UN Guiding Principles, indicated in 2014 that it would publish its own view - unilaterally, we assume. However, the OECD project is likely to take another year or so to reach its conclusions, and no word has come out of the Thun Group for quite some time. So the commitment in the covenant to report on this issue as soon as April 2017 seems likely to be the first to see the light of day, and given the involvement of a range of stakeholders in the project including civil society and the OECD, we have high hopes for a fair and thorough assessment of the topic.

Governance

To ensure the agreement gets implemented, there is a Steering Committee made up of two members from each of the four delegations (government, unions, the banking association and civil society), and an independent Monitoring Committee to report to it. The Monitoring Committee will produce an annual report, which may be published, and a final report in 2019, which will be published together with comments by the steering committee. 

What does this mean internationally?

As vitally important as the Netherlands is on the world stage, why should the rest of the world pay any attention to this agreement? For one thing, the world’s large banks will be keeping an eye on their Dutch counterparts, especially those who want to remain at the front of the pack in their approach to human rights issues, to make sure they are not left behind. They even have the option of signing up to the agreement.

For another, the Dutch banks’ work on due diligence best practice and the definition of key terms is expected to feed into the OECD’s Working Party on Responsible Business Conduct in the Financial Sector, thereby having resonance far beyond Dutch borders. And the Dutch ministers that signed the covenant, ministers Ploumen and Dijsselbloem, have said they want to help develop similar agreements at the international level, for example at the EU and the OECD.

Finally, the numerous resources that the signatories have promised to produce - the matrix/database, the reports on increasing leverage and defining cause and contribution, the value chain mapping reports and the reports of the monitoring committee - will be of relevance to banks and many other organisations around the world. They are also likely to keep BankTrack and other bankwatchers busy over the next few years, with the job of keeping a close eye on whether the banks meet their promises over the three year time-frame of this agreement, and above all whether all the paperwork indeed leads to the prevention of human rights abuses by companies financed by Dutch banks.

The official website of the Banking Sector Agreement is here.

Main image: Ministers Dijsselbloem and Ploumen sign the covenant. © Dirk Hol, from Flickr.