24 May 2011

by Bernard Rivers

At the Global Fund Board meeting on May 11-12 there were animated conversations in the hallways regarding the recent Associated Press stories on corruption by some Global Fund grant implementers, and regarding some outspoken complaints about the Secretariat that the Fund's Office of the Inspector General had just made in a progress report to the Board.

Underlying these conversations were two core questions: How should the Office of the Inspector General (OIG) fulfil its role in identifying financial management problems and potential corruption by grant implementers? And is the Fund's commitment to almost total transparency always appropriate?

At the meeting, the Board reiterated that it has zero tolerance for corruption and that it is fully committed to transparency. But, unfortunately, the Board was unable to fully resolve the more complex aspects of these issues, and chose instead to wait for the new Chair and Vice Chair and the newly-established High Level Review Panel (both reported on in GFO #146) to recommend a path forward.

If the Fund is to avoid further adverse media coverage and further consequent donor nervousness, it must urgently implement a more effective and fine-tuned approach to the issues of corruption and transparency.

There are two main problem areas where action is required. The first relates to the OIG's working methods; the second relates to tensions between the OIG and the Secretariat.

Problem area 1: The OIG's approach to conducting audits and investigations is overly zealous

The OIG works ferociously hard and has done a terrific job ferreting out the corruption that inevitably exists when an organisation disburses billions of dollars of grant money. This energy is admirable, and GFO has published over 40 articles describing the work.

But as I have written before, the OIG tends to see things in black and white, and seems unable to recognise shades of grey. For instance, OIG audit reports often contain an almost random mix of negative findings, ranging from extremely serious to relatively minor, followed by a statement along the lines of "the OIG is therefore unable to provide assurance that the grant programs managed by this PR were operating fully effectively at the time of the audit."

Of course the OIG can't provide such assurances. First, no organisation operates "fully effectively." But second, the health systems of poor countries all suffer from varying weaknesses that make it very difficult for them to institute gold plated systems of financial control. The OIG's role, in my view, is not to identify all problems; it is to identify the serious ones, particularly those that involve fraud.

One of the results of the OIG's overzealousness is that it gets bogged down in detail and, sometimes, in post-audit disagreements with the parties who have been audited. As a result, the OIG appears to be falling behind its scheduled workplan. In the October 2010 version of its twice-yearly progress report, the OIG stated that it was "on track to deliver" all of the 20 country audits planned for 2010. But of these 20, only four (Haiti, Cambodia, Cameroon and Zambia) had their audit reports published during 2010. One (Rwanda) had its audit report published during 2011; ten (Djibouti, Kenya, Kyrgyzstan, Malawi, Mali, Nigeria, PSI Madagascar, Sri Lanka, Swaziland and Uzbekistan) completed their field work but have not yet had their audit reports published, perhaps because post-audit investigations are taking place; and five (Dominican Republic, Laos, PSI South Sudan, PSI Togo and PSI Headquarters) completed their field work but have not yet had their audit reports published even though there are no known post-audit investigations taking place. (PSI - Populations Services International - is a PR working in several countries.)

Furthermore, four promised 2010 business process reviews have not yet been delivered. (These reviews look at Global Fund-wide programmes and processes such as Voluntary Pooled Procurement.) Compounding the problem, the OIG Investigations Unit had 134 cases in its inventory as of late April, with new referrals coming in at the rate of one every two business days.

As a result of its pursuit of problems both large and small, the OIG has made an average of nearly 70 recommendations per audit report. Dealing with these recommendations has created a substantial burden for the Secretariat and implementers.

A further problem is that the OIG employs or sub-contracts auditors who sometimes have an inappropriately aggressive or arrogant manner. Certainly the OIG cannot and should not be "Mr. Nice Guy" at all times. But the OIG would be more effective if it was smarter in how it handled people. The people who have complained to me about this have so far not been willing to go on the record, usually because their audits have not yet been completed and they don't want to risk making things any worse. But an employee of one PR, regarding which no fraud was identified by the OIG, said with a bitter laugh, "I sometimes wish we had stolen some money, because then at least the pain would have been justified!"

In the absence of more substantive quotes, let me describe one small firsthand example. Last year I offered to introduce the OIG to a witness of possible fraud. When an OIG investigator phoned me to discuss the matter, I started to explain the witness's role and to suggest how best to persuade the witness to be forthcoming. The investigator interrupted with a brusque "I know how to do my job Mr. Rivers, I can assure you of that." I brought the conversation to a rapid close without sharing with the investigator my potentially useful information.

The problem of auditors causing offence may sometimes be the result of the fact that the OIG frequently uses sub-contractors. The OIG in Country A (Switzerland), conducting an audit in Country B, sometimes sub-contracts much of the work to an audit firm in Country C, which sub-sub-contracts to a smaller audit firm in Country D. The sub-contracted staff are sometimes perceived by those being audited as being poorly briefed and trained; and the OIG appears to provide limited oversight concerning how they carry out their work.

In addition, the OIG is at times worryingly lax about respecting the rights of the organisations it audits. For instance, at the conclusion of the OIG's field work for each in-country audit, the OIG holds "exit meetings" not only with the principal recipient (PR) that was being audited, but also with CCM members and, sometimes, with donors. At these meetings the OIG discusses its provisional findings, but provides nothing in writing. The PR then has to wait a considerable time, sometimes over a year, to see a draft report. This long delay places the PR in an intolerable position - allegations about the PR have been made by the OIG in front of others, possibly including the PR's other donors, and the PR's reputation has potentially been harmed; yet all this has happened before the PR has seen or had a chance to respond to anything in writing.

A more appropriate process, more in line with standard audit practice, would be as follows: (a) The OIG meets privately with the PR and verbally presents the OIG's provisional audit findings. (b) Based on the PR's response and its own further research, the OIG determines whether any of the provisional findings need to be modified. (c) The OIG submits to the PR a confidential draft version of its forthcoming audit report. (d) The PR is given a reasonable period of time to submit a written response to the OIG. (e) The OIG makes any changes it deems necessary. (f) If the PR has protested certain findings but the OIG has decided not to modify them, the OIG writes to the PR stating this and explaining why. (g) The OIG then returns to the country to conduct, for the first time, a briefing for the CCM, thereby giving the CCM a chance to prepare for public release of the final audit report.

Another way in which the rights of those being audited have not been adequately respected is that even now, several years after the establishment of the OIG, the Fund has not put in place a process whereby those being audited can lodge with an independent ombudsman or panel their protests or appeals regarding the methods or draft findings of the OIG.

Problem area 2. There are serious tensions between the OIG and the Secretariat

Over the past year and a half, relations between the OIG and the Secretariat have been like a bad marriage, with various rows, some "kiss and make up" sessions, and a mutual attempt to hide the problems from the neighbours except when things get really bad.

The situation was sufficiently serious earlier this year that the Board Chair and Vice-Chair had to organise a series of in-person meetings and teleconferences involving the Secretariat's Executive Director and the OIG's Inspector General. I understand that by the end of a session in Paris in mid-April, working relations appeared to have improved notably. But then, less than two weeks later, the Inspector General sent the Board an OIG progress report which included complaints about the Secretariat whose tone angered every one of the many Board and Secretariat leaders with whom I talked. Indeed, the Executive Director informed the Board that he found this part of the OIG report to be "inflammatory and irresponsible."

My experience with the OIG investigator with whom I talked last year, and the actions of the Inspector General with respect to the latest OIG progress report, illustrate that the OIG, at all levels, seems to have difficulty anticipating how its tone and, sometimes, its expressions of opinion may come across to others. As a result, the OIG has made too many enemies and lost too many friends, including on the Board to which it reports.

Another cause of the tensions is that the OIG sometimes objects to certain forms of PR expenditure that have previously been approved by the Secretariat. Sometimes the OIG not only objects to the expenditure, but also says that the money in question should be refunded. This puts both the PR and the Secretariat in an impossible position, particularly because there is an ambiguity regarding whether the OIG's word on everything is final. If the OIG determines that a PR should not have spent Global Fund money on a specific activity, but the Secretariat believes that the expenditure did not involve corruption and was contractually acceptable, even if unwise, the Secretariat should not be required to force the PR to return the money.

Also contributing to the tensions is the fact that some of the OIG's recommendations are completely outside the OIG's area of expertise. For instance, in one country, the OIG made a "high priority" recommendation that a sub-recipient "should establish a system and procedures to continuously monitor the insecticidal effectiveness of bed-nets available in the country."

A final source of the tension relates to confidentiality. The Fund's policies make it clear that staff from the Secretariat and the OIG are not permitted to publicly disclose information regarding the OIG's work beyond what the Board has received and published. Yet an Associated Press (AP) story on 13 May 2011 said that the AP had "obtained" an "internal chart" showing that across 12 countries, "investigations [by the OIG] reviewed almost $576 million in spending." This internal chart, as reported, appears to contain data that was known only to the OIG and the Secretariat. Both the OIG and the Secretariat have denied to GFO that they provided the AP with such a document, or with any other non-public document. The Secretariat also told GFO that it has never seen a document mentioning this amount (i.e., $576 million).


Transparency has always been a core principle of the Global Fund. Rightly so. But in the specific context of the OIG, to call for transparency regarding everything the OIG writes is simplistic and possibly dangerous.

At its recent meeting, the Board approved a nice-sounding Global Fund Commitment to Transparency (see Decision Point 24). But when, as part of that, the Board reaffirmed "the Global Fund's commitment to a fully independent and transparent Office of the Inspector General," it did not specify what this means in practice.

Certainly, everybody agrees that the Fund should be "transparent" about the OIG's finalised audit and investigation reports (meaning that the Fund should publish these reports). But should the Fund also be transparent about progress reports that contain OIG opinions about how the OIG has been handled by the Secretariat or Board? (Example of such an opinion: The OIG has been "hampered" by the Secretariat in its ability to report fully and completely to the Board.) Should the Fund be transparent about progress reports that contain interim OIG findings that are not yet backed up by publishable evidence? (Example of an interim finding: The OIG is "addressing allegations" about "pervasive" fraud involving fabricated expense claims in Country X.) And should the Fund be transparent about progress reports that contain sweeping but non-specific claims by the OIG, with no backup evidence? (Example of such a claim: The OIG is investigating 17 cases of drug theft in 13 [un-named] countries in Africa; a number of the thefts have been well in excess of $1 million in value.) All of these examples are from OIG progress reports that have been published because of the Fund's transparency rules.

Even with respect to its final reports, the OIG should give the Secretariat and Board an opportunity to comment, before publication, on the report's wording or tone (but not the right to insist on changes), and should give the Secretariat the right to add a cover note providing context. Such actions could save a lot of difficulty. For instance, the first AP article on the Global Fund on 23 January 2011 (discussed in GFO here) achieved an enormous impact partly because of its dramatic headline: "Fraud Plagues the Global Fund." That headline was seriously misleading. The total amount of fraud (as against lesser problems) that the OIG identified in the 15 countries about which it had reported to the Board at that time was $11 million (or 0.28%) of the nearly $4 billion that the Fund had disbursed to those countries. The $11 million was 1.9% of the $576 million in expenditure in those countries for which (according to the AP) the OIG has examined all the paperwork. The AP has twice had to issue corrections about its Global Fund stories. Such errors, and the misleading headline to the first story, might have been less likely if the Secretariat had been permitted to include a cover note providing context.


The Global Fund Board has had the courage to put in place two things that almost no other large funding agency has, namely a vigorous OIG and a strong transparency policy. For this, it deserves congratulations. But the Board has to accept that these two things can quickly turn into an explosive mix if one adds a few more ingredients - such as a Board that is unable or unwilling to provide effective oversight of the OIG; interested media; and, in some countries, political forces that dislike foreign aid and multilateral agencies. Even though this explosive mix is, in part, the perverse result of the Fund's own courageous policies, it may lead to reductions in support from the Fund's more timid donors.

Equally important, the currently over-zealous approach to audit runs the risk of disillusioning thousands of honest and committed implementers and, thereby, of harming vital programmes. While nobody doubts the need to optimise financial management systems and to take a zero tolerance approach to corruption, the results can be counter-productive if this is not handled sensitively and intelligently.

At the recent Board meeting, several speakers made the important point that even though the Fund rightly has zero tolerance for corruption, this does not mean that the Fund should have zero tolerance for risk. As one of them put it, zero risk is only possible if you have zero disbursements which, in turn, means you have zero results.

The Fund cannot eliminate the risk that money will be mis-used; it can merely reduce it. The key question is: how far can the Fund go in trying to reduce this risk before it unduly increases costs or reduces impact? Two examples of perhaps going too far are as follows. First, after the OIG found that some Global Fund money had been used in Mali for "training events" that never took place, the Fund told PRs across the entire portfolio of grants that all training events needed to be pre-approved by the Fund. As a result, the Fund is now sitting on multiple un-approved training plans. Second, the OIG objected to a $200 item being charged by one PR to a Global Fund grant because of the possibility that the item in question had been used not only in connection with a project financed by the Fund but also in connection with a non-Fund project conducted by the same implementer.

In brief: The Fund needs to assess and manage the trade-off between risk and impact. And the OIG should focus primarily on rooting out corruption, not on criticising marginally imperfect management and accounting practices.

The last few months have not been easy ones for the Global Fund. With a new Chair and Vice-Chair in place, this is a good time to take a cool hard look at the issues.

Author's note: I encourage GFO readers, particularly CCM members, grant implementers, and other Global Fund stakeholders, to email me (bernard.rivers@aidspan.org) by 31 May with a short "letter to the editor" responding to this article. We will publish selected letters, edited for length as necessary, in early June. If you don't want your name or country to be revealed in the published letter, please make that clear.

Bernard Rivers (rivers@aidspan.org) is Executive Director of Aidspan and Editor of GFO.

Leave a comment