United in disharmony

Development Policy13 Apr 2010Amy Pollard

Harmonizing aid seems like a straightforward, sensible thing to do. But how easy has it been in practice? Amy Pollard tells of the challenges involved in establishing a multi-donor office in Indonesia, while Rosalind Eyben assesses the implications for agencies trying to make aid more effective.

Harmonizing aid seems like a straightforward, sensible thing to do. If aid programmes overlap, or if several donors are attempting to do the same thing in the same country, surely they should coordinate their efforts and work together? This self-evidently reasonable notion was a key element in the Paris Declaration on Aid Effectiveness (2005). But how easy has harmonization been in practice?

In 2005, just in time for the launch of the Paris Declaration, the UK’s Department for International Development (DFID) established the Decentralization Support Facility (DSF) in Jakarta, Indonesia. This was to be an innovative experiment that would, for the first time, synthesize the decentralization programmes of multiple donors, and establish a joint office in which they would work.

From the start, the tensions between individuals and organizations ran high. For many staff it was the most difficult professional environment they had ever experienced. ‘I’ve never seen anything like it’, said one, ‘people getting so angry, crying, and completely breaking down’. German participants described the office as a haifischbecken – a ‘pool of sharks’ – while World Bank staff called it a ‘pit of snakes’.

This story describes some of the struggles in the first few years in the life of the facility.

The idea that development aid will be more effective if donors work together is now well established. In the Paris Declaration on Aid Effectiveness (2005), increasing harmonization was seen a key element in the effort to improve the impact of aid. The argument for harmonization is that there are too many overlaps between donors, and they would achieve better development results if they worked together to coordinate their work. Rather than donors being in competition, or simply ignoring each other, harmonized aid should ‘reduce transaction costs’ for recipient countries, because governments will no longer have to deal with many, often conflicting donor programmes. At least, that’s the theory.

One ambitious attempt to realize the potential of harmonization is the Decentralization Support Facility (DSF), established by the UK’s Department for International Development (DFID) in Jakarta, Indonesia. The facility would eventually bring together, under one roof, staff from no less than three Indonesian ministries and nine donor agencies, and integrate their support for decentralization.

Decentralization has been a dramatic, tumultuous process in Indonesia. Rolled out after the fall of Suharto in 1998, it provoked a radical transformation of almost every major institution in Indonesia – and was seen by donors as a major opportunity to promote pro-poor reform. The DSF was to be no ordinary donor office. It was not ‘an organization’, or a legal entity of any kind, but a radical experiment in aid design. The attempt to develop joint programmes on decentralization, which would have a real impact on poverty in Indonesia, would sorely test all concerned.

Interestingly, although the initiative was intended to promote collaboration between donors by breaking down the institutional barriers between agencies, the pressure to try to harmonize their work intensified the differences between them. The Paris Declaration was founded on the hope that harmonization would make the aid system simpler. Why, then, did relationships at the DSF become so complex?

Conflicting interests

The most common explanation of this apparent paradox was that it was all about ‘interests’. Each organization and individual was seen to have their own incentives and disincentives, which structured their motivations and actions. The way these interests were understood – or misunderstood – by others had a critical impact on how the relationships within the DSF developed.

One important set of interests concerned the different financial positions of the various agencies. The bilateral donor partners (Germany, the Netherlands and the UK) had access to grant funds, which were more flexible than the loan funds of multilaterals like the Asian Development Bank (ADB), the UN Development Programme (UNDP) and the World Bank. There was competition between UNDP and the World Bank to access the DSF trust fund set up by DFID, and this led to increasing mistrust. When UNDP staff complained that the World Bank had too much control in the DSF’s governance structure, for example, this was seen as motivated by their own interest in accessing the fund. When money was allocated to the World Bank, UNDP raised concerns that the system for allocating funds had set up a conflict of interests in favour of the Bank. While the UNDP complaint was ostensibly about ‘process’, it was interpreted by others as financially motivated sour grapes.

Interests were seen not only as financial, but also bound up with the different strategies the agencies used to solve problems. UNDP was often perceived to have a particular interest in ‘process’. One afternoon I interviewed the head of a World Bank programme in Indonesia. I asked him about how he went about solving problems in multi-donor work, and he offered to show me. He called a senior colleague at UNDP, and put the call on speaker-phone. It was a convivial call with someone he knew well, but the two agencies were in the midst of a dispute over a joint project. Each subtly defended their own staff and discussed how to resolve the issue. As he spoke, he switched between documents on his screen, working a carefully formulated text on his position into the flow of conversation. As he listened, he even answered the odd email.

When he hung up, he asked me what I thought was interesting about what I had heard. I said it was his slick combination of reading, emailing, speaking and listening – he said I wasn’t supposed to notice that. The interesting thing, he thought, was that his UNDP colleague had said they should decide who would send out invitations for an upcoming meeting. ‘You see – process – that was her solution’. His previous experience with UNDP had sensitized him to any action that could be interpreted as springing from this ‘UNDP approach’. He was able to speak fluently and precisely because he had anticipated the issues and had key prompting documents open and ready. He had expected a ‘UN approach’, and perhaps interpreted her responses in a different way than he would have done had she worked for another agency. In this, we can see how the kind of interests that actors anticipate in others may influence the kinds of interests they perceive.

Meanwhile, World Bank staff were seen to have an interest in turning the DSF into ‘another World Bank office’, and to skew the institutional arrangements in their favour. The relationship between the ADB and the Bank was particularly tense. At the opening of the DSF office, for example, one senior ADB manager was surprised to see posters bearing the logo of the ADB as a ‘DSF partner’, when such a status had not yet – he believed – been agreed. ‘We were just being co-opted … it was never done in any formal way’. He decided not to complain, as he did not in fact want ADB to withdraw from the DSF. But he did make a lot of noise behind the scenes – not about the World Bank trying to dominate the DSF, but about the fact that the invitations to the opening were issued by the World Bank and not the DSF. His initial discontent about the assumption of ADB involvement was thus transferred into a complaint ‘about’ the invitations to the opening.

Some ADB staff felt that even the DSF office furniture betrayed the World Bank’s interest in dominance. ‘You can tell this is a World Bank office – they use the same furniture all over the world’. Indeed, the DSF furniture was the same as that in the Bank’s office in Aceh, ordered by the same officer from the same supplier, so it seems likely that it was chosen for convenience rather than to show dominance. Early evaluations of the office noted that disputes over issues that would have been trivial in other circumstances, became points of contention in the context of the tensions developing within DSF.

One day, a UNDP manager alluded to the invisibility of the World Bank’s interest in domination. She advised me conspiratorially, ‘You’ve read Gramsci, haven’t you, darling, this is a hegemony!’ For her, the Bank’s interest in dominating the DSF could be read from myriad issues in the office, from the vocabulary that Bank staff used in meetings, to clerical errors in the printing of business cards.

Given such sensitivities, staff went to great pains to anticipate how the documents and reports they produced might be perceived by others. There was a consensus that ‘managing perceptions’ was key – both to prevent agitation among agencies and to ensure that the donors did not pull the plug. But it soon became clear that perceptions were not going to be manageable, and anticipating the reactions of others was enormously stressful.

For those leading the design of DSF, the challenge was to write the facility’s planning documents in ways that would weather any eventuality in an uncertain future. They needed to have a robust and flexible interpretive scope. It was not for the author to be direct about whether or not a text resolved tensions between different agencies. That was left to the readers to judge for themselves, on behalf of their respective organizations. Thus, in many of the email discussions on drafting documents for DSF, there was a lot of innuendo. It was expected that all readers within this politicized bureaucracy would be able to read between the lines.

Homeopathic solutions

DSF design documents were written in such a way that every agency could see hints of things they agreed with. The manager of the DSF design process described how ‘You can’t pander to the agendas of particular agencies, because then you will alienate the others. And you can’t just throw in buzzwords, because then the whole thing stops making sense.’

Designing the DSF, he believed, was rather like concocting a homeopathic remedy. ‘Homeopathy is based on the idea that water molecules contain the “memory” of substances they once contained. But homeopathic solutions are so dilute that there is no molecular evidence … of those substances. Homeopathy relies on the placebo effect and faith. The DSF was established to reduce poverty through harmonization. But these core concepts have become so diluted through the design and redesign process that no evidence of them remains. So, to achieve its goals, the DSF [will have to] resort to placebos and faith.’

In other words, all donors would have to accept that it was in their own interest to ensure that DSF documents recognized the interests of others. Indeed, placebos can be effective, whether they are based on reason, conjecture or delusion. The interests intimated in these documents carried with them a sense of something missing, to be filled in by each reader as they read between the lines. Thus, the sense of something missing, which is at the heart of perceiving the interests of others, is an engine of power. Just as a child learning to walk must lose its balance in order to move forward, when institutions struggle, power has to remain an issue of doubt.

What happened next?

I finished my fieldwork in June 2007, but my informants report that DSF has remained ‘a rollercoaster ride’, with decision making an arduous, fraught and time-consuming process. There has been a heavy change-over of staff. The scale of the facility’s ambition has been tempered and, with more conventional management arrangements in place, the Indonesian government is now leading the initiative.

As I completed my fieldwork before the DSF began implementing its programmes, I am in no position to evaluate whether DSF succeeded in its aims, or to assess its impact on poverty. For further information on the facility’s progress, visit


Unfortunately, due to the age of this contribution and several migrations to online content management systems, the footnotes in the text may have been lost. The footnotes below are listed in its original order of appearance in text.
  1. The story is based on Amy Pollard’s PhD dissertation, Power in Doubt: Aid, Effectiveness and Harmonisation amongst Donors in Indonesia (University of Cambridge, UK, 2009). Her research involved in-depth ethnographic fieldwork over a period of 14 months between 2005 and 2007, including 140 formal interviews, discussions with more than 80 informal participants, and (under the Freedom of Information Act) accessing hundreds of emails and documents in DFID’s archive, ‘Quest’. Under the terms of her contract with DFID Pollard has the right to publish and reproduce her fieldwork material freely. All the individuals represented here have had the opportunity to comment and give feedback on this article.
    The events described here occurred between 2004 and 2007, when the DSF was becoming established and designing its joint programmes, and practically no work was being implemented. For this reason, there is no possibility of using this story to assess the overall success of DSF as an initiative – it simply reports some of the difficulties in its early years. All the individuals represented here have had the opportunity to comment and give feedback on this article.
  2. The DSF is currently led by the government of Indonesia – the National Development Planning Agency (BAPPENAS), the Ministry of Finance and the Ministry of Home Affairs – and nine donors: the Asian Development Bank, AusAID, CIDA, DFID, the Government of Federal Republic of Germany, the Government of the Netherlands, UNDP, USAID and the World Bank. In 2005, only five donors were members, and there was no formal position for the Indonesian government in its management structures.