The problem with facilities, and aid

The increasing reliance on facilities – large, multi-faceted programs run by consulting or so-called “contracting” companies – has become controversial in the Australian aid program. It is not surprising therefore that facilities have started to feature on the Devpolicy Blog. Since November 2017, we have run three articles about facilities (here, here and here). They all provide interesting insights, arguments and contributions. That these three articles are all (overall) positive about facilities is understandable given that they are all written by people working for companies who run them. What is odd is that no one has bothered to write us a critical article on facilities.

What we see in relation to facilities is a good example of Bill Easterly’s characterisation of aid as a “cartel of good intentions.” Everyone wants to advertise how good their aid project is; no one wants to criticise the aid projects of others. It might give aid a bad name, or give you a bad name, and make your own funding less secure. That’s one reason why we see many more articles on our blog promoting particular aid projects than ones criticising them.

This is also why we run the aid stakeholder survey every three years. It provides those implementing the aid program – both private sector contractors and NGO executives – with a safe space to make known their views on the aid program. In our most recent 2018 survey, we asked a number of questions about facilities. They did not come out looking good.

As the graph below shows, a majority of both private sector contractors and NGO executives agreed that the emphasis on facilities in the management of Australian aid has had a negative effect on the quality of Australian aid.

Are aid facilities having a positive or negative impact on aid effectiveness?

This unhappiness might be dismissed as sour grapes. A small group of large contracting firms run DFAT’s facilities. Perhaps others are simply unhappy about the resulting squeeze. To get around this problem, we also sought the views of those directly managing aid projects at least some of whose funding comes via facilities (just under half of those we surveyed). Contractors were equally divided between those who thought the facility improved their project’s effectiveness, and those who thought it made things worse. Among NGOs, the great majority was in the latter (negative) group. We also asked the same respondents about transaction costs. A solid majority of both groups thought facilities increased their transaction costs; very few thought they reduced them.

The impact of facilities on projects (according to those who are funded by them)

Colin Adams, the author of the most recent Devpolicy piece on facilities, mentions these findings in passing, but summarises them as “mixed feelings,” which to me is an understatement of their negativity. Adams goes on to provide two examples of good facilities, but does acknowledge that the facility model can be improved, since the current approach “constrains flexibility and innovation.” Jacqui de Lacy also has a good discussion of the risks facilities can give rise to.

Over recent years, as a minor aid implementor myself, I have observed a mix of pros and cons of being funded via a facility. Some of my experiences have been quite positive, but when I think about the negative response of most stakeholder survey participants, two cons come to mind. One is that having to work through an intermediary can delay decision making. That’s not surprising. Unless there is a high level of trust, getting agreement among three parties (the implementor, DFAT and the facility) is probably going to be more difficult than getting agreement among two. The other is that the expiry date of the facility can become a major complicating factor in the life of the smaller implementing partner. Sub-contracts are shortened to fit in with that expiry date. Or, after that point in time, the implementing partner might be transferred to the care of another contractor – if the latter wins the facility contract this time. The new contractor might have quite different conventions, understandings and requirements. The result is an increase in both uncertainty and transaction costs.

What are the solutions? There will always be some reliance on large, contractor-managed programs, but the question is whether the pendulum has swung too far. DFAT could hire more staff. It could then shift back to more bilateral arrangements and to smaller facilities. But will the government allow DFAT to hire more staff, and will DFAT, a generalist organisation, be comfortable with a larger number of aid specialists?

Adams puts forward a different solution. He says a partnership model is needed, and to get there he wants the performance of both parties – DFAT and the contractor – to be assessed, not just the performance of the contractor. I don’t see that happening. DFAT is the principal. The contractor is the agent. As long as DFAT is the government agency paying the bills, and the contractor is a private sector entity that has to bid for DFAT contracts, then there is not going to be a relationship of equals in which the performance of both parties is assessed.

A radical but perhaps workable alternative to both of the above would be to establish a state-owned implementing aid agency. Germany implements its aid program in large part through GIZ, a state-owned company with a global reputation for excellence. That might seem unimaginable in Australia, but remember that we ask ACIAR to manage aid investments in agricultural research. We don’t require it to bid for contracts or to work through intermediaries. Bidding for contracts sounds like a way to promote efficiency, but, as I mentioned earlier, the DFAT-contractor relationship is inherently bound by the inflexibility of a contract: basically, a fixed amount of money for a fixed number of years, with all the stop-start and uncertainty that that involves. A within-government DFAT-implementor relationship may be the only way to deliver the longer-term, more flexible, more trusting approach that Adams is right to acknowledge we need. (It might also avoid the problem of contracting out policy dialogue which facilities can give rise to.) At the very least, the idea of emulating the ACIAR approach more broadly across the aid program is worthy of consideration, especially if DFAT doesn’t want to go on an aid-specialist hiring spree.

Whatever the solutions, the latest stakeholder survey suggests that are some major problems to be addressed. The six-page review of facilities put out by DFAT tells us that the Department is thinking about the issue, but is hardly an adequate response.

In the meantime, here at the Devpolicy Blog we will continue to welcome well-informed articles about both aid successes and aid problems. And we will keep running the Australian aid stakeholder survey as our modest contribution to disrupting Easterly’s cartel of good intentions.

Notes: The graphs don’t add to 100%, because “no effect” responses are omitted. Find out more about the 2018 and earlier stakeholder surveys, including the detailed 2018 responses which this blog summarises. 

image_pdfDownload PDF

Stephen Howes

Stephen Howes is Director of the Development Policy Centre and Professor of Economics at the Crawford School of Public Policy at The Australian National University.


  • Thank you Stephen and others for this interesting discussion thread, which I have followed with interest as a former long-term AusAID and DFAT senior aid program manager.

    There seems to be a consensus that facilities can offer gains in terms of efficiency, flexibility/responsiveness and innovation but DFAT’s own independent review acknowledges that facilities are complex to manage:

    “Far from enabling DFAT staff to adopt a less ‘hands-on’ approach, facilities require intensive and ongoing DFAT oversight, engagement and management of both the development content and the delivery process.High quality aid policy development, strategic programming and effective aid delivery depend on fostering and retaining staff who can: engage in deep, content-oriented policy dialogue with partner governments; establish and manage contracts with delivery partners that enable rather than constrain effective aid delivery; and make quality choices about activity focus (and how/when these need to change) that optimise results (effectiveness and value for money). DFAT has some of this, but not enough – in either breadth or depth.”

    The integration of AusAID and DFAT resulted in the departure of the majority of AusAID’s internationally recruited sector specialist staff, who previously provided high level technical guidance and oversight especially of more complex programs. A lot of experienced program managers also left and many others understandably moved fully into the diplomatic stream to optimise their career options. Those that are left or have been newly inducted increasingly find their ability to devote time to program management squeezed by the core public service duty of providing advice and briefing to Ministers and senior officers.

    DFAT has tried to remedy this situation by insourcing external technical expertise through panels or standing offers and by expanding the use of mechanisms such as facilities. Whilst this may have benefits in drawing in fresh expertise and new ideas, it has also major risks – most important of all, in my view, the erosion long-term of DFAT’s ability to engage in that “deep, content-oriented policy dialogue with partner governments” the review says is crucial.

    As Erin Anderson noted in her post, it is critical that DFAT provide a good ‘brain’ to steer programs in dialogue with partner governments and other stakeholders- whether those programs are delivered through old fashioned ‘body shops’, facilities, individual projects, sector-based programs, multilateral or NGO partnerships or otherwise. DFAT can do this either through its own staff, through facility/contractor managers, through independent external technical advice, or a combination of these. However there is a ‘moral hazard’ risk of contractor capture, in DFAT staff relying too heavily on advice from a contractor’s own managers or technical specialists to help steer large and complex programs, without independent technical advice and support. This can increase the potential for proliferation of activities and poor coherence and fragmentation over time (the ‘Christmas tree’ syndrome), as arguably contractors are commercially incentivised to increase activities and inputs. Over-reliance on contractor advice by inexperienced and over-stretched government program managers can also lead to contractors influencing positive performance assessments for programs they manage (the ‘dog judging the dog show’ syndrome). The recommendation of the DFAT review for independent quality assurance for all complex facilities seems to be a good way to mitigate these risks.

    Another major risk of facilities, particularly multi-sector facilities, is that they can be established relatively rapidly in pursuit of ill-defined higher level outcomes, without sufficient up-front investment in understanding sector context and in developing some kind of agreed road-map or trajectory towards the higher level outcomes (eg at least a set of notional benchmarks or ‘stepping stones’). This takes time and resource investment in joint analysis and design with partners, to enable joint progress monitoring, discussion and problem-solving along the way. That is not to deny the benefit of adaptation, seizing political opportunities and exploratory and innovative approaches during implementation, but to better enable such decisions to be made consciously and explicitly, rather than by ‘finding the way as we go’. The example cited by Jacqui de Lacy of Australian aid training midwives and providing medicines to birthing facilities in West Timor without being able to address the lack of water in these facilities is a good case in point: rather than something to be ‘fixed’ retrospectively through a facility, I would suggest that is something that good old-fashioned sector analysis and donor dialogue should have picked up and addressed even before any project design.

    In arguing for facilities, Colin Adams in his post rather glibly states that ‘transformational development outcomes are not achieved through simple, sector-specific programs’. Yet, like our own, most partner governments structure their own development plans, budgets and institutions around sectors, as do most major donors including the multilateral banks. So a strong understanding of sector context remains critical to ‘transformational development’…and neither are sector-specific programs simple!

    In Tonga, where Australia has supported a long-standing health sector support program, agreed benchmarks and outcomes for the program have included among others: a continuation of increased spending on health as a proportion of the overall budget (and within that an increased allocation to public health); development of a national non-communicable disease strategy and media campaign; development and implementation of increased excise duty on tobacco and unhealthy food; infrastructure investments to improve environmental health (eg water supplies); and expansion of rheumatic disease screening and treatment to all children in Tonga. These are hardly minor achievements and were based on a joint program design with the Tongan government based on extensive analysis, guided by DFAT and independently-contracted health sector experts and supported by high-level annual performance reviews and discussions between the two governments.

    As it happens (based on reading DFAT’s latest annual performance report on the Tonga aid program), it seems that due to disasters, changes in senior personnel and other factors, the Tongan government’s implementation on this program is faltering – and may well need supplementation through a facility or similar contractor-led modality!

    My argument is not with facilities (or any other particular modality), but with the need for DFAT to be able to devote the specialised resources needed to manage these and other complex programs well, in order for Australia’s aid to be effective and to have impact at scale and at the level of policy.

    From discussions with colleagues in other parts of government, DFAT is also far from alone in having lost or downgraded program management expertise. This seems to have been a broader trend caused by the ongoing shrinking of the federal public service and the prioritisation of ‘policy'(more often than not briefing needed by ministers or senior officials to address the growing demands of the media cycle) over implementation in government over the last decade or more. I would argue the need to address the challenge of ensuring the “high quality policy development, strategic programming and effective delivery” and to “engage in deep, content-oriented policy dialogue with partners” cited by DFAT’s review extends more broadly across government- witness the state of service delivery to our own indigenous people, in addressing homelessness, in environmental management and in a range of other policy areas.

    How the professional management expertise needed to manage complex facilities and other programs is best re-established in DFAT (and arguably elsewhere) seems to me to be a crucial policy issue for the future effectiveness of Australia’s aid program…and more broadly for effective service delivery across government. Options seem to include establishing specialist technical, research and implementing agencies in government (a la ACIAR and GTZ); through private sector and NGO outsourcing (delivery) and in-sourcing (management and technical advice); through partnerships with research institutions, NGOs, international agencies or a combination of these.

    In relation to DFAT and the management of facilities specifically, the recommendations in the independent review seem to be a good place to start.

  • A very intetesting post, many thanks Stephen. I’d be interested to hear from those familiar with the GIZ model. My understanding is that it still implements projects with defined timeframes and budgets so it would not necessarily be more efficient approach than the managing contractor approach. I’d also be interested in any evidence of efficiency or effectiveness gains from facilities versus other models. The aid stakeholder survey gives us informed opinions but triangulating these findings with some cost-benefit analyses would be useful. Given the changes within DFAT over the last 5 years, an evidence-informed discussion about how best to deliver the aid program is very timely.

    • Thanks Elisabeth. I should have credited Richard Moore with the GIZ idea. He recommended it in a speech to ACFID back in 2016. However, in his more recent review of DFAT-AusAID integration, Richard has backed away from the idea. I’m still attracted to it. I do feel that the rapid turnover (individual and institutional) in our aid programs undermines efficiency and effectiveness. Moving away from bidding to dedicated providers seems to me a positive way forward, and a state-owned provider could help with that transition. But I certainly agree that there is a lot more to be analysed and written on this subject.

Leave a Comment