Claire Provost, Fellow, Centre for Investigative Journalism 

Further written evidence on DfID’s use of contractors

In some countries, UK aid is already spent largely through for-profit contractors. For example: the OECD’s latest peer review of UK development cooperation found that in Nigeria 75% was spent through private sector companies.[1] DfID’s use of contractors is also expected to grow. There is an explicit move away from budget support in the government’s new Aid Strategy.[2] Contractors also play a prominent role in UK aid delivery in fragile states, where spending is increasing.[3]

The committee’s decision to examine these issues is thus extremely topical. It could also have significant impact. Greater transparency and proactive efforts to “buy locally” and support developing country capacity could produce significant wins for development. If contracting is going to remain a significant and potentially increasing part of UK aid delivery, it is critical that this process is managed in ways that most benefit local communities in developing countries.

The 2002 International Development Act sets the goal of DfID spending as the reduction of poverty.[4] But it is unclear what the department is doing to ensure that its use of contractors best meets this goal. The Independent Commission on Aid Impact (ICAI) worryingly noted in 2013 that DfID’s “decisions to use contractors are not guided by a strategic plan to deploy the right contractors, including major, niche and innovative new entrant organisations, to best effect.”[5]

Further to oral evidence given on Monday 6 June 2016, I would like to share the following references and additional information with the committee. I would also like to reiterate the suggestion that the committee seek further evidence directly from civil society organisations and businesses in developing countries. The Network of Empowered Aid Response aims to hold donors to account for recent promises to spend more humanitarian aid through local actors.[6] Members of that network could be contacted for their input, for example.

1. DfID’s use of primarily British contractors. DfID’s Annual Report and Accounts for 2014-2015 said: “British companies have continued to be very successful, winning over 90% of contracts awarded.”[7]

1.1 The OECD’s 2014 Peer Review of UK Development Cooperation said:

The UK reports that over 90% of centrally managed contracts – which represent the vast majority of the contract value – go to UK suppliers. This is a concern as it has implications for value for money… the UK should ensure that their procurement regimes are commensurate with the DAC requirements for ODA to be untied de facto and should undertake efforts to ensure there are no unintended or implicit impediments to foreign suppliers winning their contracts.[8]

1.2 It has been suggested that these 90% figures are misleading, as they might include all companies with registered entities in the UK (including subsidiaries of American companies, for example).[9] DfID should clarify how it reports this statistic but also, more importantly, it should report the percentage of contracts awarded to companies in developing countries.  

2. The UK compared to other donors. In 2013 almost 80% of total bilateral aid from OECD donors was formally untied.[10] The UK was a leader on this agenda but today many donors including Australia, Belgium, Denmark, Ireland, the Netherlands and Norway report at least 95% of their aid as untied. (Note: several donors on this list have also met the 0.7% spending target).

2.1 However, the OECD’s Aid Untying: 2015 Progress Report noted:

A very high share (often over 75%) of contracts (volume and/or value) awarded to suppliers in the donor country has been a persistent feature for some donors such as Austria, Canada, Finland, Korea, United Kingdom and the United States... By comparison, Members such as France and Japan report relatively low shares of contracts awarded to their domestic suppliers – 34% for France and 23% for Japan.[11]

It said DAC members including the UK should look at their procurement regimes and ensure they “are commensurate with the requirements for ODA to be de jure and de facto untied.” (Note: for France, 90% of its aid is reported as untied. For Japan that figure is 80%).[12]

2.2 The OECD’s report also reminded donors that contract opportunities (and awards) should be published on a centralised DAC Untied Aid Bulletin Board of Business Opportunities. It said these transparency provisions were “aimed at building confidence that ODA is not only de jure but also de facto untied.” But these “continue to be met poorly in general...For some (e.g. Denmark, United Kingdom) reporting has tapered off, in several cases to zero.”[13]

2.3 The US is the largest OECD donor. In 2013 it reported less than 70% of its aid as untied.[14] But USAID has taken proactive steps to spend more through local institutions and organisations in developing countries. Five years ago, the agency’s then administrator Rajiv Shah said it was “no longer satisfied with writing big checks to big contractors and calling it development.”[15] A target was introduced to spend 30% of programme funds through local partners by 2015. (It did not meet this target, though local spending increased to almost 20% from under 10%).[16]

2.4 It is worth noting that USAID’s procurement reforms were actively opposed by many of the agency’s main for-profit contractors, most of which are based in the US. A group of companies including DAI formed the Council of International Development Companies (CIDC).[17] CIDC in turn employed a professional lobbying firm to represent their interests on Capitol Hill.[18]

3. The argument for spending locally.  Buying goods and services locally can increase the impact of aid by using procurement as a tool to support the local private sector, create jobs, build tax bases, and develop local capacity. It puts local actors in the driver’s seat and can also be cheaper than using international contractors.

3.1 A 2013 report from the Center for American Progress included several examples of how spending through local institutions has saved USAID money. In one case, it said the agency “reduced its cost to build schools in Senegal by 50% by shifting the same contract to a fixed-price reimbursement model with the local government rather than procuring from a traditional international contractor.”[19]

3.2 The Overseas Development Institute (ODI) has also argued that “localising aid” can stimulate the local economy, generate greater local employment, and cause less distortion to the local labour market. It noted that international agencies and contractors compete with local counterparts for recruits, which can “retard the development of the state’s core institutional capacity.” Meanwhile localising aid can also lead to greater government revenues, with local actors subject to local taxes.[20]

3.3 Spending aid through local systems and organisations has been part of the international aid effectiveness agenda for years. The 2008 Accra Agenda for Action urged donors to “increase the value for money of aid by continuing to untie it...and by promoting the use of local and regional procurement.”[21] The 2011 Partnership for Effective Development Cooperation repeated the expectation that untying aid can “present opportunities for local procurement, business development, employment and income generation in developing countries.”[22]

3.4 At the 2016 World Humanitarian Summit, the UK and other donors agreed a new “Grand Bargain” which commits to spending 25% of humanitarian aid through local groups by 2020.[23] It is unclear, however, how exactly the UK will promote, track and disclose progress towards this target. (Moreover, it is unclear why a similar commitment cannot be made for all UK aid). 

4. Competition. The dominance of contractors registered in the UK suggests that there are barriers to entry faced by companies in developing countries. Donors have been repeatedly criticised for not doing enough to level the playing field and support local capacity development. 

4.1 A 2009 report published by the OECD said: “Risk aversion appeared to be part of the explanation for widespread de facto tying.” It said donors’ “often hinted at but not explicitly stated preference” for contracting “within known social networks” is reflected in the high share of contracts awarded to their domestic companies. It added: “Contract requirements, including pre-registration, completion bonds and so forth, make such outcomes more likely.”[24]

4.2 The report included details of a “DfID survey of constraints on local firms” which found that even when aware of opportunities, “developing country consultancies usually did not bid, perceiving the cost of preparing bids was greater than the reward, taking into account the perceived risk of not winning.” It revealed specific challenges including: large contract sizes, “vague and overly legal” bidding documents with “terms not suited to local conditions,” and local organisations’ “lack of reputation outside their own country, even when experienced.”[25]

4.3 A 2011 report from the European Network on Debt and Development (Eurodad) found: “Donors’ procurement practices still favour Northern firms and make it virtually impossible for firms from Least Developed Countries to compete.” It again identified specific issues including “headquarter procurement [that] benefits Northern companies” and “large contracts [that] undermine the chances for small and medium sized enterprises in poor countries.”[26] It said:

“Donors should end all practices that de facto exclude Southern firms from competing. This means creating smaller packages, sized to be manageable by local firms, advertising all tenders locally in local languages, preferably in one joint database per recipient country, and removing unnecessarily restrictive eligibility criteria.”[27]

4.4 The ODI’s “Localising Aid” research similarly identified specific aspects of donor contracting that could be altered to better support the local private sector, including: “specification of technology requirements; size and complexity of bid packages; requirements for a supplier to qualify prior to submitting a valid bid; and efficiency of the payments system, which affects a contractor’s working capital and cash flow.”[28]

4.5 ICAI’s 2013 report on DfID’s use of contractors found that a bid cost more than £40,000 on average to prepare. (The report also noted: “In the long run, DFID pays for contractors’ bidding costs because they are recovered through the fee and overhead mechanisms.”)[29]

4.6 ICAI further noted that it is “hard for new entrants to break into” the framework agreements that DfID is increasingly using.[30] The use of large Payment by Results arrangements has similarly been criticised for its “risk of crowding out smaller players.” Oxfam’s Deputy Head of Programme Strategy & Impact said recently:

“At a meeting in Whitehall two weeks ago for an upcoming £100m plus call it was surprising how few agencies were in the room. The combination of the financial risk transfer, cashflow implications, management costs and risk management itself and the costs of learning and adapting are arguably beyond the risk appetite of many agencies. At a time when DfID is looking to promote southern civil society, and advance small and medium NGOs, this funding model would seem to be working the other way.”[31]

4.7 The “Procurement at DfID” webpage on Gov.uk says there is a branch of UK Trade and Investment (UKTI) called the Aid Funded Business Service, which can help British companies win business with donor agencies. It also has information about DfID’s annual Supplier Conference, in London. It is not clear meanwhile what if any comparable, proactive initiatives exist to support or engage with local contractors or other organisations in developing countries.[32]

5. Subcontracting. DfID’s primary contractors may subcontract significantly with actors in developing countries. But it is unclear how using companies in rich countries as intermediaries is the best strategy for sustainable development and poverty reduction.

5.1 There is insufficient transparency around the use of subcontractors by DfID’s primary suppliers. It is unclear, for example, how much money stays with the primary supplier and how much is spent locally. It is also unclear how UK or other international contractors build up local expertise, capacity and skills. If this is being done effectively, we would expect subcontractors to eventually (and hopefully quickly) become primary contractors themselves.

5.2 Subcontracting with local companies is not the same thing as meeting the promise of untied aid. Developing countries must lead their own development. Their organisations should not remain forever in subordinate relationships to more powerful and more influential actors from the UK and other donor countries. (Note that in the ODI’s reports, they did not count subcontracting through international intermediaries as “localised aid”[33]).

5.3 In 2013 the Center for American Progress said some USAID country offices were taking concrete steps to encourage international organisations to work with local subcontractors “with the intent to build their capacity to manage own grants and contracts.” It said this is an approach with “a direct exit strategy by ensuring local ownership and capacity building in a given project—something that fiscal hawks and development practitioners should both support.”[34]

6. Fragile states. Worryingly, ICAI’s 2015 report on UK Development Assistance for Security and Justice (which reviewed projects in several “fragile states”) said: “the procurement process does not give enough emphasis to local networks and country knowledge.” It found that:

DFID’s procurement of contractors is causing a range of problems, including long delays and rigid or unrealistic programme designs. We saw instances of high quality delivery by non-governmental organisations (NGOs), who may offer greater local knowledge and legitimacy than contractors but often find it difficult to compete in procurement processes.[35]

It further expressed concern that “the tendency towards higher-value programmes with more complex procurement process advantages a limited number of large international firms, at the expense of small firms, NGOs and local partners.”[36]

6.1 The ODI’s research found that even in fragile states, appropriately managed “localised aid” can strengthen both state and non-state institutions:

We found that the risks of localising aid in Afghanistan were not greater, and probably smaller, than not doing so...This implies that this finding may also be true in countries with better-functioning institutions. In our Afghanistan example, localised aid carries a slightly higher fiduciary risk but significantly lower programmatic, contextual and institutional risks.[37]

They stressed: “localising aid does not mean relinquishing control over donor money… a spectrum of options exists for mitigating the fiduciary risks associated.”[38]

7. Transparency. The Devtracker website has made it easier to find information on many UK aid projects. It remains difficult-to-impossible, however, to trace UK aid spending to the ground and understand how exactly it is spent at each stage in the delivery chain.

7.1 For example: The Devtracker website includes a page for the DfID project “Supporting the Government of Sierra Leone to implement its National Water Supply and Sanitation Strategy.”[39] This has three sections 1) a “summary” section with total spending figures; 2) a “documents” section with links to a “Service Provider for Technical Assistance” contract and four contract amendments. (No business case, monitoring report, or any other document is available); 3) a “transactions” section with a list of dated payments to Adam Smith International, the “Receiver Org,” without any further description or detail.

7.2 The Devtracker page for the Sierra Leone project links through to the contract and its amendments published on the government’s separate Contracts Finder website.[40] Here there are 3 documents: Section 1 (Form of Agreement), Section 2 (Terms and Conditions), and Section 3 (Terms of Reference). However, Section 4 (Special Conditions) and Section 5 (Schedule of Prices) of the contract are not published. It is understood that these documents would likely include significant details such as any day rates to be charged by contractors.

7.3 Nowhere on the Devtracker website is it clearly tracked how much money is spent directly (or indirectly) through local organisations in developing countries. It is not clear when contractors sub-contract, or how much money is spent at each stage. A disclaimer says: “DFID aims to be as transparent as possible however, in some cases, we are unable to show all the documents that are associated with a project, for example where there is a potential security risk.”[41] But it is rarely if ever clear exactly when or what information is withheld, for security or any other reason.

7.4 It is understood that DfID holds details such as contractors’ profit margins and day rates. Commercial sensitivity is often invoked as a reason to withhold information. This should be challenged so that the maximum, not the minimum, is released. DfID could require greater disclosure of its contractors. If a company is unwilling to open its books, it could opt-out of bidding for UK aid funded work.

8. Prices and profits. DfID’s Statement of Priorities and Expectations for suppliers includes the request that they seek “fair but not excessive rewards.”[42] But it is unclear how this is defined or monitored, or what the consequences would be for pursuing “excessive rewards.”

8.1 ICAI’s 2013 report said: “DFID has selected contractors that have delivered positive results at competitive fee rates.” But the comparator seems to have been other consultancy companies that work in the UK.[43] It is unclear if comparisons were made with fee rates charged by local companies or other organisations in developing countries.

8.2 A previous 2011 LSE study examined DfID bid and contract data and found “striking price differences in the bids that non- and for-profits make for a same project.” It said:

The overall personnel fees proposed by non-profits were on average 60% cheaper than the equivalent figure proposed by for-profits (that is, holding the tender constant). What is more, for-profit contractors were much more likely to go over-budget ex post. The additional transfers (as a share of initially agreed costs) paid to for-profits ex post were on average 30 to 50% higher than the ex post transfers renegotiated by non-profits.”[44]

3