I doubt that Tim Costello, Director of World Vision, and Aid Watch, the left-wing aid watchdog, agree very often. But they both, along with many others, think that too much government funding is given to private contractors to run Australian aid projects.
The Aid Watch website warns that “A major concern of Australia’s aid program is that it favours commercial interests in aid delivery.” Quoting 2006 figures, it notes that 45% of the aid program goes to private companies.
And then as recently as April this year, on ABC’s Lateline, Tim Costello said that “Australians would be surprised to know that some 40 plus per cent of the aid budget goes to private contractors, to Australian companies.”
Well, they should be surprised, because it isn’t true. The Aid Watch figures are completely out of date. This graph from the recent Independent Aid Review shows how AusAID actually spends its money, and how that has changed. (Disclaimer: I was one of the five panel members commissioned to write the Review.) AusAID funds spent through private contractors were indeed 45% of the total in 2005-06, but fell to only 21% in 2009-10.
This dramatic change in the way Australia gives aid has gone largely unnoticed, though credit to Bernard Keane of Crikey for highlighting the Review’s findings in this area. AusAID rarely if ever reports its aid in terms of delivery mechanisms. Perhaps the first time the dramatic fall in the share of the aid budget channeled through the private sector was made public was in AusAID Director-General Peter Baxter’s opening address to the ANU Doubling Aid conference in February of this year.
What lies behind the numbers? The move away from ‘managing contractors’, as they are often called, has little to do with the arguments traditionally made against them. Any number of objections have been made over the years against contractors: the aid they deliver benefits only Australia (the ‘boomerang aid’ argument); aid shouldn’t be managed by for-profit companies; the particular companies which Australian aid funds are non-transparent; and they are too few in number to provide genuine competition in aid tenders.
However valid these arguments (or not), the reasons for the change in Australian aid lie elsewhere. The limitations of the contractor-led model of aid delivery were exposed in the first half of the last decade, when the Australian government put together a number of interventions, most prominently the RAMSI stabilization intervention for the Solomon Islands and the post-Tsunami package for Indonesia, whose size, urgency and complexity demanded a much more hands-on approach. More generally, there was a sense that development progress in PNG and the Pacific – always the acid test for AusAID – was tenuous at best, negative at worst, and that it was therefore time to try a different approach.
AusAID has since delivered no more than what it promised in the 2006 White Paper. This policy document stressed the need to work in partnership with recipient governments, other bilaterals, and multilaterals, and promised “greater direct, field-based, policy engagement with partner governments and hands-on management of the programs … rather than simply contracting out all services to private contractors.”
The biggest winner from this bipartisan reorientation of the aid program – started by the Coalition, continued by Labour – has been the multilaterals, such as the World Bank and the UN agencies, whose share of the aid program has increased from 32% to 40. In the past, the fact that – to simplify – multilaterals focus on Africa and we on Asia and the Pacific limited our funding to them. But a new strategy was developed to increase funding to multilaterals for activities in our region. And more Australian aid now passes through multilaterals than any other type of aid delivery channel.
Another winner has been the Australian public sector. AusAID staff numbers have certainly risen sharply, and more aid also goes through other government departments. The graph above only shows spending by other government departments of funds they receive from AusAID (5% of the AusAID spend), but they receive another 10% of the aid budget directly as well.
Perhaps most surprisingly, Australia now gives much more of its aid directly to the recipient or partner governments: the share of this type of aid has skyrocketed from 2% to 8. Australia had become very reluctant to actually hand over aid funds to partner governments after the bad experience with budget support to PNG in the 1970s and 1980s. But again, a rethinking has occured. Budget support to PNG was given with no strings attached. Funding to recipient governments now has lots of strings attached, and works well. The Indonesia schools program is the best example.
NGOs have also gained, with their share increasing from 8% to 13. This seems to be the only change which has been partisan in nature. Funding to Australian NGOs has increased sharply since Labour came to power, with core funding alone rising since then from $30 million to $98 million in the most recent budget.
And then finally universities have doubled their share of the aid budget from 3% to 7. As summarized here, governments from both sides have come to love scholarships in recent years.
These changes put together add up to a quiet revolution in Australian aid. Contractors haven’t gone out of business. The overall aid budget has roughly doubled in the last five years so funding to them has stayed roughly constant in nominal terms. But they are no longer the dominant force they used to be in the aid program.
I think that’s a good thing, and no doubt both Tim Costello and Aid Watch would agree. A more diversified approach to aid delivery makes sense. But it is not without its risks. Contractors are hardly ever given projects which they don’t have to compete for. Multilaterals never bid for projects. Nor do government departments. Universities and NGOs often don’t either. More generally, projects managed by contractors tend to come under greater scrutiny, both before and after the intervention, than those managed in other ways. Go to the AusAID website, and you’ll see far more evaluations of projects managed by contractors than of those managed by multilaterals or NGOs.
The point of this blog is not, however, to argue the case for or against what has happened, but simply to draw attention to the change. It’s time for the debate over the aid program to move on.
Stephen Howes is Director of the Development Policy Centre at the Crawford School of Economics and Government, ANU.
My apologies for not returning to this sooner as I have been undertaking fieldwork in Indonesian schools and districts. There are three different issues that should worry us about this aid modality in Indonesian education, both in relation to the completed AIBEP and what is now planned by Australia.
First, is the question of whether school construction support is justifiable in Indonesia at all at its present stage of development. Is Indonesia really unable to fund school construction from its own budget when, for example, it allocates so much of its budget to subsidising the sale of fuel? A convincing case that school construction is an effective way to assist in educational development has not been demonstrated. Neither has the case for the significant amount of funding to be provided. Even if such case can be made, should Australia be spending money on an activity that is not sustainable? The Independent Completion Report (ICR) you refer to provides a clear warning that school construction may not be sustainable: “Maintenance of new school facilities and provision of budgets that can cover this together with teaching materials and books remains a concern.” (p.iii)
The second issue is one of how donors and the government of Indonesia can reliably know where these new school assets are actually needed. The poor quality of Indonesian data for educational planning is widely understood to be an impediment to good planning. Only now are some districts and provinces beginning to get a reasonably accurate understanding of the educational assets they have, their school demographic data and therefore their needs. This understanding is occurring through aid-supported data management tools and analytical skills development. So, how can AusAID have reliably known where to build new schools to meet actual needs in the past?
The third issue of concern is the ICR itself. I am not convinced that the ICR on the Indonesia Basic Education Program is ‘compelling evidence’ at all. Evidence, yes, but is it compelling? The publicly available ICR documents on the AusAID website show why this is an issue; the documents paint a clear picture of the tight time frames, over-management of the evaluation process by AusAID, complex terms of reference that run to 11 pages, and firm editorial control. The documents raise serious questions about true ‘independence’. This is not to question, in any sense at all, the integrity and hard work of the independent evaluators who were so pressed for time in the evaluation that they were only able to visit 11 schools (out of a construction total of 2014), but rather the bureaurcratic environment in which they operated that lead to the finished ICR.
That the claimed evidence for implementation through govermment working ‘very well’ may not be the good evidence we would wish, is flagged by these remarks in the ICR:
“An issue raised by MoNE and some development partners for future education programs was whether more emphasis should be given to refurbishment of existing schools rather than constructing new schools. Indonesia has approximately one million classrooms, with 30-50% of these in need of repairs or refurbishment. The cost to refurbish schools is approximately 55% of the cost to construct new schools. AusAID and other development partners need to examine whether rebuilding of schools nationwide, and extending facilities by adding laboratories and libraries, would have a greater impact than expanding the number of schools (p.iii)”
“It is understood that there has been no formal review of the structural design to confirm extent of earthquake resistance however, the design appears to be more resistant than comparable regional buildings. The use of a standard structural design solution across all regions irrespective of site conditions remains a concern… (p.4)”
“School site selection was an issue that led to challenges beyond the anticipated expertise … With respect to construction criteria only, it appears that sites were often chosen based simply on availability, e.g. the site may have been ‘less desirable’ vacant public land or unproductive land donated by a benefactor. Other factors such as proximity of rivers and main roads need to be considered – for example, at one school site visited it was found that some nearby students did not attend due to an extremely busy highway between. At another site, the new school was located approximately 100 metres off the access road – sites fronting the access road were above flood level, but not the school site; this particular site is flooded with up to two metres depth for seven months of the year. (p.7)”
“As noted, some sites were not ideal for school construction. Some exceed minimum allowable slopes; some are adjacent to natural (e.g. abutments) and other manmade hazards (e.g. busy roads). The District was responsible for site preparation prior to construction, but this did not always occur due to limited District financial and human resource capacity. Typically, drainage and essential landscaping to school sites was incomplete at the time the school becomes operational. (p.7)”
I do not believe that donors building schools in Indonesia can be justified now, except in disaster recovery situations. It might be good for publicity and it does make schools available for some children, but there is clear evidence from the ICR itself to cast serious doubt on this working “very well”: doubts about sustainability are raised; there are doubts whether construction is better than refurbishment; and serious concerns about design and site selection are raised in at least two of the 11 schools visited – a very high proportion indeed.
If this is the “best example” of funding to recipient governments then I think we have much to be concerned about.
Thank you for the comments. Robert, have a look (if you haven’t already) at the Independent Completion Report on the Indonesia Basic Education Program, which can be found here. I think that document provides compelling evidence that the modality chosen (implementation through the govermment) worked very well in this case.
Stephen,
I think any confusion in this area stems largely from the lack of clear and comprehensive information from AusAID. Hopefully this will change under the new Transparency Charter recommended by the Aid Review.
However I’m not sure that the issue of “boomerang aid” or the larger issue of aid getting to the poor has been solved, despite the share of aid to private contractors decreasing.
Thank you for the post for and clarifying several issues.
I am interested in your assertion that “Funding to recipient governments now has lots of strings attached, and works well. The Indonesia schools program is the best example.”
Can you clarify why you believe the Indonesia program is the best example and how the ‘strings’ ensure this is so?
With thanks.
Thank you for setting out these significant shifts in aid delivery. Your blog suggests that AusAID is increasingly giving consideration to the most appropriate channel for aid delivery based on the particular program objectives, rather than simply defaulting to use contractors. This is definitely a good thing.
However, I think your discussion of the greater scrutiny of contractors isn’t correct. Firstly, evaluations of multilateral activities aren’t on AusAID’s website, but they are the websites of the multilaterals. While it varies across organisations, many have strong evaluation policies and practices and there is little point in AusAID duplicating this function.
Similarly, many NGOS have strong approaches to evaluation and increasingly these are available publically. See for example http://www.careevaluations.org ; CARE’s Electronic Evaluations Library which contains over 400 evaluation reports from CARE projects from all over the world.
You suggest that the fact that contractors bid for projects increases the level of scrutiny on their work. Of the new Australian Government grants CARE received last year, 72% were won through competitive processes. In addition, Australian NGOs go through a rigorous accreditation process with AusAID, renewed every 5 years, to ensure that we have the capacity to effectively and efficiently deliver aid. There is a lot of scrutiny on NGO work, although some of it operates differently to the scrutiny on contractors.
Thanks for an interesting post.
Julia