Ten months ago, the Global Fund put on hold about $95million in potential disbursements under four grants to the Zambia Ministry ofHealth, because of fraud within the ministry that was first reported by awhistleblower. Seven current or former ministry employees were charged by theZambian government in relation to the fraudulent appropriation of about$350,000 from one of the grants. The funding is still on hold except forpayments for life-saving drugs, which are now bypassing the ministry and beingsent direct to procurement agents or suppliers.

The Fund later decided that the role of principal recipient(PR) for these grants would be taken away from the Ministry of Health andhanded over to UNDP, and the ministry was asked to return $8 million in unspentfunds. Grants with non-governmental PRs were not affected.

The Global Fund Board was informed of these developments.But the public was told nothing; no press release was issued, and noinformation was placed on the Zambia pages of the Fund’s website.

Recently, an Aidspan colleague of mine came across a singlesentence buried deep in a report by the Fund’s Inspector General that revealedthe freezing of disbursements to the ministry. We reported the news in GFO onJune 14. (See Issue 126, at www.aidspan.org/gfo.)We didn’t specify when the action took place, because we didn’t then have thatinformation (though we could see that it was some time during 2009, which weshould have mentioned).

The GFO story was picked up (and some incorrect informationadded) by Reuters. Within three days of the GFO story, there were 14,000 newsstories and related articles about this on the Internet, from New Zealand toSaudi Arabia, and the Global Fund had to issue a press release providingadditional information.

Why was the Global Fund silent on this matter for nearly ayear?

When I put this to the Fund, the response was, “Therewas no suspension of the grants to the Ministry of Health in Zambia. There wasonly a delay in the disbursements.” Thus, according to the Fund, there wasno cause for a press release. But the “delay” in disbursements has thusfar lasted nearly a year, and the Ministry of Health has been removed as a PR.The world’s news media seem to have found that pretty newsworthy.

I’ve been a close observer of the Global Fund since itstarted in 2002. And the Fund’s handling of the Zambia case provides furtherconfirmation of a suspicion that has long been forming in my mind, which isthat the Fund is very reluctant to report, via press release or its website,any news that might worry a donor or that might embarrass the government of acountry that receives Global Fund grants.

But the issue is bigger than that. The Fund is not onlyreluctant to report on its few “tough actions”; it has been reluctant,particularly during the past three years, to take those tough actions in thefirst place.

In my view, the Global Fund can only claim that it is trulytransparent if it gives easy access to both the good news and the bad. And theFund can only claim it is truly using performance-based funding if it iswilling to terminate a grant or to switch to a new PR in cases where the PRpersistently and significantly fails to deliver on its promises, or where thePR reveals clear corruption.

Transparency

In many ways, the Global Fund is remarkable in itstransparency. Impressive amounts of information are publicly availableregarding Board decisions, grant contracts, disbursement data, etc. It would bewonderful if other funding agencies were this open. But the Fund’s transparencyhas some lamentable and persistent gaps.

The obvious place where the Fund should reveal cases whereit has suspended a grant or issued a “No Go” or some similar decisionis on the Fund’s web pages that deal with the grant and country in question.But this almost never happens. In Table 1 below, I list all cases where theFund has taken such actions. Compiling this table took a substantial amount ofresearch, and it would not have been possible without access to non-publiccommunications from the Global Fund Secretariat to Board members and theirdelegations. With the exception of the three Myanmar grants, none of theinformation under “Action taken” in the table is revealed at therelevant grant and country web pages. A reader of those web pages wouldconclude that everything is fine, and always has been, with each of thesegrants.

When I asked the Global Fund why its web pages forindividual countries and grants don’t reveal these bits of bad news, aspokesman agreed that it would be a “good idea” to show suspensions.But he added that No Go grants “will be shown solely as closed grants,”and that this could take some time to show up because a grant is not formally closeduntil all grant activities have ended. But that’s missing the point. The greatmajority of closed grants are ones that simply came to the end of their naturallives.

Performance-based funding

The Fund’s website explains that performance-based fundinglies at the heart of the Global Fund’s operating model; that in order to raiseadditional funding from donors, the Fund must prove that the original fundingled to proven results against performance targets that had originally beenproposed by the country receiving the grant; and that when there arenon-performing grants, the money from those grants is re-assigned to othergrants where results can be achieved. (See

It’s an admirable “tough love” model. But in myview, it has not been adequately implemented during the past three years.

Between mid-2004 and mid-2007, the Global Fund considered,by my calculation, 264 applications from Country Coordinating Mechanisms (CCMs)for “Phase 2 renewal” (that is, applications to have funding extendedfrom the first two years of the grant into years 3 through 5). The Fundapproved 254 of these applications. But for the other 10 (3.6%), it denied theapplication by issuing what it calls a “No Go” decision, which meantthat the grant was being brought to an end three years earlier than originallyanticipated.

I’m not sure that the Fund was being sufficiently tough bytaking such action in only 3.6% of cases. But it was a good start. (TheMinister of Health for one country whose grant received a “No Go”decision told me long afterwards that although he had been upset at the time,he had come to realise that it was the best thing that could have happened. Thedecision led to the CCM being reformed and to new PRs being chosen, and thingsare now going much better.)

But when we examine what has happened between mid-2007 andnow, we see a very different picture. During this period, my calculations show thatthe Fund has received 215 Phase 2 renewal applications – and it has approvedevery one of them. Not a single No Go decision has been issued. (When I asked whetherthis reflects a change of policy, the Fund said that it does not.)

Under performance-based funding, the grants that are themost obvious candidates for a “No Go” decision are the ones that havepreviously received an “unacceptable” rating from the Global Fund.(The Fund issues a rating for each grant once per disbursement – which means,on average, once every six to nine months. The ratings range from A1 – “exceededexpectations” – to C – “unacceptable.”)

Of the CCM applications for Phase 2 renewal that werereceived between mid-2004 and mid-2007, four were for grants that had at someprevious point received an “unacceptable” rating. The Fund did notapprove any of them.

Of the CCM applications for Phase 2 renewal that werereceived between mid-2007 and now, seven were for grants that had at someprevious point received an “unacceptable” rating. All seven wereapproved.

It may well be that some of these “formerly C-ratedgrants” were doing better by the time they were reviewed for Phase 2, anddeserved to be approved. But still, these numbers don’t give the impressionthat performance-based funding is a concept that is being actively pursued bythe Global Fund.

New procedures needed

The Global Fund should establish a committee whose role isto review what action to take regarding each severely underperforming grant. Ifno action is taken, the committee should specify why.

Whenever a grant is issued a “C” rating, or thegrant has had only “B2” or “C” ratings for twelve months,the grant should automatically be referred to the committee. (The grants thatcurrently meet those conditions are shown in Table 2.)

The committee’s decision and report on each such grantshould be sent to each CCM member. A summary of the actions taken, and a linkto the report, should be prominently and permanently placed on the web pagesfor that grant and country. And once a year, the committee should issue areport, and an accompanying press release, documenting all such actions overthe past year.

The purpose of such actions would not be to be vindictive; itwould be to encourage all grant implementers to do their best to ensure thattheir grants continue to be funded, and to show the donors that this is beingdone. I believe that if these actions are taken rapidly, they will increase,rather than decrease, the funding that donors commit to the Fund for the nextthree years at the replenishment meeting this October. Furthermore, if theseactions lead to an increase in the number of grants that are closed down, moneywhich is currently not producing results will be freed up for grants thatotherwise may not be funded in Rounds 10 and 11.

When the Global Fund was established in 2002, it rapidlyobtained substantial and broad-based support. This was partly because the needfor its grants was obviously so great. But it was also because people lookedforward to straight talk and tough logic-based decisions, in contrast – forgiveme – to the UN approach of attempting not to upset anyone. Some of us are stillwaiting.

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

Table 1: GFOcompilation of grants where the Global Fund has taken firm action (in dateorder)

Table 2: GFOcompilation of currently-active grants apparently in trouble, based on GlobalFund ratings as of 17 June 2010

Note: As of 17 June 2010, there were (according to dataat www.theglobalfund.org/en/commitmentsdisbursements)218 currently-active grants that have been rated by the Global Fund. Of these,17 (6%) have a latest rating of “C” or have had a rating of “B2″or “C” since at least one year ago, and thus qualify to feature inthe above table.