Select Committee on International Development Sixth Report


37. DFID spends its money in 145 countries. At first glance, this is a huge portfolio of countries for DFID to be involved in, but this number gives a somewhat false impression of DFID's activities. The Departmental Report includes two tables which outline DFID spending by country. The first, table 4 of annex 1, shows DFID programmes and associated allocations. The second, table 9 of annex 1, shows DFID bilateral expenditure by country in 2000/2001. A comparison of these two tables reveals that whilst DFID spends money in 145 countries, its programmes are limited to 71 countries and territories. The geographical breakdown of DFID's bilateral programmes is: sub-Saharan Africa, 21 countries; Middle East and North Africa, 2 countries; South Asia, 6 countries; East Asia and Pacific, 6 countries; Latin America and the Caribbean, 11 countries; Europe and Central Asia, 18 countries; and, 7 Overseas Territories. As Mark Lowcock, Director for Finance and Development Policy at DFID, explained to us:

    "One of the reasons why we have that very big number you mentioned—140 countries or so—is that we have not taken the step of saying that we want to close down the opportunity say for an NGO to come to us for a grant to work in a particular country even if we are not there, or for a research programme which is part of a broader effort to be excluded from a particular country. So there is a very big tail of countries in which we have a small number of small spending activities which, in many cases, will not involve a DFID presence on the ground but will simply involve a grant, often to an NGO actually".[42]

38. Nevertheless, seventy-one countries is still a lot of countries in which to have programmes. Suma Chakrabarti told us that he shared our concerns about DFID's large portfolio of countries, the danger of "managerial stretch", and the need to consider carefully what value DFID might add by being involved in a country. The trend is to concentrate more resources in a few countries, a move which is being considered as part of DFID's process of setting medium-term strategic priorities for resource allocation.[43] As Mark Lowcock remarked, "In 1997/98, 57 per cent of the bilateral resources were spent in the top 20 countries. That increased to 64 per cent in 2001. Based on our projections for future years we expect that ratio to rise. We also have a list of countries we expect to graduate from programmes. There are about 15 countries over the next few years which we would hope to be able to conclude our direct government to government work in".[44]

39. Historical links and organisational inertia are not good reasons to be involved in a particular country; we are pleased that DFID is thinking about concentrating its efforts in a smaller number of countries. As Suma Chakrabarti put it: "We are part of a collective effort; we do not have to be in every single sector in every single country and we should look at comparative advantage at a country level to try and work out where we can best make the real difference".[45] We expect subsequent Departmental Reports to provide clear information about DFID's assessment of its comparative advantage, its portfolio of countries, and the extent and nature of DFID's involvement in particular countries. We would also welcome further information about how DFID determines which countries will graduate from one form of development assistance to another, and ultimately beyond development assistance.

Country Selection

40. Abstracting from the realities of historical inertia, DFID's two key criteria for allocating resources to countries are: firstly, the level of poverty in the country, and secondly, the commitment of the country's government to poverty reduction and the appropriate policy reforms. DFID's view, in line with that of the World Bank in its publication Assessing Aid,[46] is that development aid is most effective when it is allocated according to these criteria. Convinced of the logic of this argument—as indeed we are broadly—DFID and the Treasury have negotiated PSAs which include targets for concentrating aid resources on low income countries. The 1999-2002 PSA included a target to increase the share of bilateral country programme resources allocated to low income countries from 65 percent to 75 percent by 2002. This target was met in the financial year 2000/2001. The 2001-2004 PSA included a target of increasing the proportion of DFID's bilateral country specific development aid (excluding bilateral humanitarian assistance) going to low income countries from 71 percent in 1998/1999 to 80 percent in 2002/2003. By 2000/2001, a level of 78 percent had been reached. The 2003-2006 PSA targets an increase in the proportion of DFID's bilateral programme going to low income countries from 78 percent to 90 percent.

41. Clearly, progress towards this target must be based on deliberations and decisions about whether or not, and how, DFID ought to be involved in particular countries. The nature of such deliberations was illustrated in a range of cases described to us. A first case concerned DFID involvement in Latin America, a region where it might be thought that DFID does not have a comparative advantage based on language, culture or historical ties. In response to the suggestion that DFID ought to pull out of Latin America, Suma Chakrabarti defended DFID's current involvement there. His defence was four-fold: one, there are seventy-seven million poor people in Latin America and the Caribbean; two, six out of the ten most unequal countries in the world are in Latin America; three, by being involved in the region DFID can help governments to see how to successfully redistribute their resources; and finally, by being involved in the region DFID can engage with other donors such as the Inter-American Development Bank to get them to be more effective and focussed on poverty reduction.[47]

42. A second case was that of Francophone countries in the Sahel. As Richard Manning, Director-General for Policy explained to us:

    "If you look at education in Africa and where the children are who are not in school, one category of country where performance is rather below average is the Francophone countries of the Sahel. If you took an Africa-wide perspective on how to achieve that particular Millennium goal in Africa, you might say there is a strong case for DFID to get seriously engaged in education in Sahel. So far we have decided that would not be in accord with our comparative advantage because, quite frankly, we would have a lot to learn about the system; we are weak in having people with the language skills. We could, of course ... go in in a sort of silent partnership mode and, let us say, give the Dutch some money to spend in Mali, but it is not clear that we are adding much value by doing that, so we have tended to think that we would do better in reaching the Millennium Goals in Africa in concentrating our bilateral aid where we can see prospects of reform—Tanzania for example, Ethiopia, and so on—and then to raise the consciousness internationally of the need to come seriously to grips with the very difficult problems in the Sahel. You could take an alternative stance and say that we could really develop significant new programs here. Indeed, some people have suggested that is exactly what we should do. But we have taken the view so far that the problem is not that there are not enough donors, the problem is that the present system is not delivering adequate results and there is a limit to what we feel we should be doing".[48]  

43. A third case was that of Nigeria, and DFID's selection of which Nigerian states to engage with. We were told that DFID Nigeria has selected four states over the last 3 years, in accordance with its Country Strategy Paper commitment to support State Governments that are committed to reform. Benue was selected in 1999, Jigawa in 2000, and Ekiti and Enugu in 2001. Benue was chosen on the basis of DFID's past experience; Jigawa was selected for a public sector reform programme; Ekiti and Enugu were selected on the basis of criteria concerning the potential both as regards policy reform and potential impact on the MDGs.[49] DFID acknowledged however that "assessing reform potential is highly subjective, particularly in a context of democratic transition when new actors emerge who are strong on rhetoric but have no track record".[50] Our own impression, based on a visit to Nigeria in March 2002, is that an understandable political decision was made—that DFID simply had to engage with the most populous nation in sub-Saharan African—and that from that point on it was a case of selecting the least-worst areas and sectors with which to work.

44. We were also told about an innovative mechanism for country selection, or, more precisely, the ways in which DFID decided which Russian oblasts, or provinces, on which to focus its efforts.[51] As Suma Chakrabarti told us, during the process of producing the 2001 CSP, DFID ran a competition to find those oblasts where DFID might spend its money most effectively, rather than spreading its efforts around eight oblasts as was the case under the 1998 CSP. The criteria against which the oblasts were assessed were:

·   Poverty focus: The degree to which the oblast's development plans address issues of poverty and social exclusion by raising the living standards of ordinary people and reducing inequality.

·   Commitment: The degree to which the oblast wants to involve DFID in key development issues such as the budget, and whether this is backed up with appropriate human and financial resources.

·   Capacity: How realistic are the oblast's development plans, in terms of its needs and resources (human and financial)? Will it be able to deliver on the proposed partnership with DFID?

·   Partnerships: Has the oblast brought in other tiers of government and civil society actors?

·   Other donors: Does the oblast have mechanisms proposed or in place to integrate its work with DFID with its programmes with other donor agencies?

·   Dissemination: Does the oblast have a track record (or potential) of piloting ideas that are replicated elsewhere in the Federation?

45. Two oblasts—Leningrad and Nizhny Novgorod—were the eventual winners of this highly innovative approach to selecting development partners. We were fascinated to hear about the Russian competition for aid. We feel that the wider use of such an approach, with selection criteria made explicit, might help to focus recipients' minds on why the UK is offering development assistance, and what potential partners are expected to deliver in return.

The Nature of DFID Engagement and the Move to Budget Support

46. As we remarked in our report on Financing for Development, decisions about what countries to engage in merge into decisions about how to do so.[52] Given DFID's focus on poor countries with a commitment to reform, we were particularly interested in finding out more about how DFID defines reform and identifies reformers. Richard Manning told us that this is an "impossible question", explaining that: "It is a continuum, is it not, between non-reformers, countries in conflict, to the other end of the spectrum, countries which have adopted full reform packages. Most countries lie somewhere in between. They do not happily sit in categories. You always have to end up reaching a balanced judgement as to where countries are". Explaining further, he told us that it is not Government policy to pull out of non-reforming countries completely. Rather, the Government's policy in such countries is to try and promote change and also to try to get services delivered, even if this is not possible through government channels.[53]


47. For those countries which DFID identifies as "reformers" which have a real commitment to poverty reduction, there is a tentative move towards providing assistance through direct budget support. DFID believes that progress in achieving the MDGs is more likely to be made through helping partner governments to implement long-term poverty reduction programmes than by using development assistance to fund isolated donor-led projects. Budget support minimises the transactions costs associated with aid, and can help to build the capacity of the recipient's systems for financial management and accountability.[54] As the Government wrote in its response to our report on Financing for Development:

    "In suitable conditions, direct budget support is preferable to other forms of aid in several ways: it ensures that funds flow through the central budget increasing transparency, government ownership and accountability for the funds provided; it increases pressure for improvements to central government planning and financial management systems; and it supports effective prioritisation by partner governments."[55]

48. In 2001/02, £290 million of budget support was provided to 17 countries. As a proportion of DFID's programme this represents an increase from 8 to 18 percent in recent years.[56] DFID were however at pains to stress that the provision of non-project assistance was not new and that in the 1980s and early 90s many donors were providing money into government budgets through import programmes.[57] The shift to budget support is neither entirely new, nor at an advanced stage; indeed the number of bilateral projects actually increased by 10 percent between 1997/98 and 2000/01.[58]

49. Budget support can be provided to a partner government's general budget or for a specified sector. DFID sees the provision of support direct to the Finance Ministry as preferable to provision directly to an individual spending department. DFID gave a number of reasons for this; perhaps the most persuasive of which was the need to develop the quality of financial management and systems in developing countries.[59] Mark Lowcock told us that the way in which aid had been managed over recent decades had tended to erode rather than build those systems. Suma Chakrabarti believed that, "providing aid through the budget is a way of [recipient countries] developing their systems rather than imposing the systems that come from a project approach".[60] Paying the money directly to the Finance Ministry has not stopped DFID from notionally earmarking certain sectors, education in Uganda being one of a number of examples. The OECD's Development Assistance Committee has been critical of this practice, which it feels could undermine a common donor approach as donors compete to earmark the few large and verifiable categories of partner governments' expenditure—such as teachers' salaries.[61]

50. DFID's Departmental Report cites Mozambique as an example of the benefits of moving aid to budgetary support.[62] It points out that high aid volumes over a number of years have enabled the country to stabilize and adjust to a market economy . However, much of the aid has been spent outside the government budget and this has imposed a heavy burden on government in coordinating donor and NGO activity. The introduction of budget support funded by DFID and 10 other bilateral donors has brought a more focussed approach to a range of macro-economic and poverty monitoring issues. One of the reasons for moving to direct budget support is to address the problems with other delivery mechanisms. In their oral evidence, DFID officials gave Uganda as an example of budget support enabling the country to put virtually every child through primary school education. Mark Lowcock stated that "Every country wants to have public systems which enable every child to be educated. No-one would make a proposition that you can sensibly try to achieve that through thousands and thousands of NGO programmes. This is a core function of the state."[63]

51. Budget support partnerships are necessarily the result of a long-term process of support for policy reform. This illustrates two related points; firstly, the importance of engaging with a range of countries, rather than simply the best reformers, if progress towards the MDGs is to be achieved; and secondly, the importance of not dropping countries suddenly if they deviate from the reform path. That said, if a country's commitment to reform appears to wane, this is likely to have implications for the nature of DFID engagement, as the case of Kenya over recent years illustrates. As Mark Lowcock recounted to us:

    "In 1999 after many years in which we had a lot of concerns about the commitment of the Government there to reducing poverty and the quality of economic management and administration generally, a reform initiative was launched. A team of people of high integrity and technical competence was appointed by President Moi to run a reform programme. In response to that the leading donors made a commitment that once results started to be delivered they would increase their programmes. So DFID had been spending between £25 million and £30 million a year in Kenya—towards the lower end of that figure actually—in the late 1990's. In response to Kenya instituting this reform process and agreeing a set of measures with the IMF and the World Bank, in 2000/ 2001 we increased our spending to nearly £50 million. At the beginning of 2001 the reform programme went off track; the Government failed to take a number of measures that it had committed itself to take, which we and the World Bank and others thought were very important indicators of their real commitment to reducing poverty. As a result in subsequent years we have reduced our spending plans. That is the way in which in practice we try to take account of changes in what is happening in the real world of the countries we are operating in when we put together our overall budgetary plans".[64]


52. Handing money over to recipient governments carries with it a number of risks. UK taxpayers rightly want to know that money paid to a health department as development assistance will be spent on health care and not on Mercedes cars for health ministers. DFID decides on whether to provide budgetary support against three criteria: the extent of poverty; commitment to poverty reduction; and the likelihood that the resources provided will contribute to poverty reduction. DFID witnesses told us that in most of the countries where they are providing budget support, the governments concerned have shown that they can deliver the outcomes in the areas which DFID cares most about—primary health, primary education and water.[65] There is inevitably a higher level of fiduciary risk associated with providing direct budgetary support and DFID has to balance that risk against the anticipated greater development impact which the use of budgetary support should bring.

53. DFID is seeking to develop its control procedures in order to counter the risks associated with budgetary support and help ensure that UK moneys are properly spent. An important part of this is the Department's work with the National Audit Office to identify key risk areas, to develop a framework for addressing these risks and to ensure that evidence can be provided to prove that budget support moneys have been properly spent. However, it is clearly the Department's responsibility to ensure that appropriate controls and safeguards are implemented properly and operated effectively.

54. The Department recognises the importance of the quality of financial management and government systems in developing countries and believes that budget support offers opportunities to strengthen such systems. For example, the Department is seeking to improve standards of accounting and procurement by strengthening institutions, providing training for civil servants and developing awareness of the role of parliaments. The risks faced by the Department are the same as those faced by other donors and, in this respect, the Department is working with like-minded donors to develop appropriate controls over the use of budget support. Such measures should provide greater assurance over the use of UK taxpayer funds but a clear audit trail for the transfer and use of funds should be maintained. For this reason we welcome the work that the National Audit Office is leading, in co-ordination with its equivalent State Audit Institutions, in developing a common audit approach.

55. The Department has agreed with the National Audit Office that, where partner countries' systems are sufficiently well developed, the audit of budget support may place reliance on records provided by those partner State Audit Institutions. Clearly, State Audit Institutions play a crucial role and the Department should ensure that they are able to meet the demands placed on them. A big issue for the Department is building up the effectiveness and the seriousness with which governments take their equivalent of the Public Accounts Committee. We believe that the National Audit Office is well placed to assist the Department and partner countries in this respect. But validating the use made of budget support also involves much more than this. The Department has recognised the importance of strengthening government institutions and improving financial management systems. A key part of DFID's approach involves building capacity to enable countries to make appropriate assessments and act on the findings.


56. Direct budgetary support has implications for the skills required from DFID's staff. DFID has over many years built up expertise in country teams with highly skilled project managers as well as education and health specialists. On the face of it, a significant increase in the provision of direct budgetary support may suggest that a greater emphasis should be placed on financial management and auditing skills. Suma Chakrabarti was at pains to stress that the need for a policy dialogue with recipient countries meant that sector specialists were still essential but that they had to develop new skills around influencing policy and engaging in a policy dialogue.[66] Mark Lowcock added:

    "Putting money through budget does not mean that we are not interested in the sectoral outcomes. On the contrary. So the nature of the role of our education or health specialists is changing. What we are asking them to do is look at—for their sector or for the country as a whole—whether it is making progress towards the MDGs. What we are not asking them to do so much is to look at a relatively small project and whether that is effective. In that sense what we are doing is scaling up; we are trying to look to the horizon, not just in the little narrow thing that we are financing. So there is a long and productive future role for our sector specialists in this. Those are the areas where we are trying to achieve outcomes."[67]

57. Budgetary support throws up new challenges for DFID staff. There is a danger that DFID offices could become unduly focussed on financial accountability and contacts with recipient governments. DFID is well aware of the danger and of the need to maintain contacts and activities in areas outside government; civil society and the private sector are also vital partners in development. We hope that the development of an effective common audit approach by recipient countries' State Audit Institutions will avoid the need for too great a change in the skills mix of DFID country teams.

58. Provided DFID is able to balance the benefits with the risks, we support the move towards budget support where suitable partners can be found, and outcomes—closely aligned to the MDGs—can be reliably predicted. The willingness to withdraw or temporarily suspend support demonstrates that there is no open-ended commitment to giving budget support[68] It is also reassuring to hear that decisions on where to move to budgetary support are, as with the PRSP itself, part of a collective process of discussion with other donors. There are however some outstanding areas of concern. DFID tries to engage in those countries where it sees itself as having a comparative advantage. It is difficult to see how it is possible to talk about comparative advantage in countries where a number of donors are simply collaborating on pooling contributions to an education department. In a similar vein, where budgetary support is pooled, the ability to attribute outcomes to DFID spending becomes even more difficult. We would like to see more information on the use of budgetary support contained with the annual Departmental Report and suggest that next year's Departmental Report provides a breakdown in the bilateral allocations between money given through direct budget support and allocations for programmes and projects.

59. We were interested to hear about DFID's decision-making processes as regards involvement in all the cases mentioned; Latin America, the Sahel, Nigeria, Russia, Mozambique, Uganda and Kenya. We would welcome the inclusion of similar material in future Departmental Reports. Such information improves DFID's accountability by demonstrating the link between DFID's objectives and spending plans, and by making more explicit DFID's overall strategy. In addition, by facilitating open discussion and learning about what works, a move in this direction might help to improve DFID's effectiveness.

42   Q7 Back

43   Q7 Back

44   Q7 Back

45   Q3 Back

46   Assessing Aid. World Bank, 1998. See Back

47   Q9 Back

48   Q8 Back

49   Ev 28, paragraphs 14-17. Back

50   Ev 28, paragraph 16. Back

51   Q25 and Ev 28 and 29, paragraphs 18-25. Back

52   Fifth Report from the International Development Committee, Session 2001-2002, Financing for development: Finding the money to eliminate world poverty, HC785, paragraphs 76 and 77. See footnote 19 for web-site. Back

53   Q24 Back

54   Q14 Back

55   Third Special Report from the International Development Committee, Session 2001-2002. Government response to the International Development Committee's Fifth Report (2001-02): Financing for Development: Finding the money to eliminate world poverty, HC1269, paragraph 35.

See- Back

56   Q13 Back

57   Q13 Back

58   How effective is DFID? An independent review of DFID's organisational and development effectiveness. DFID Evaluation Report, EV639. Final pre-publication draft. Back

59   Ev 35, paragraph 1. Back

60   Q38 Back

61   Development Cooperation Review of the United Kingdom, OECD Development Assistance Committee, DAC Journal, 2001, volume 2, no. 4, I66. Back

62   DFID Departmental Report 2002, chapter 2, page 18, box 2e. See footnote 1 for web-site. Back

63   Q35 Back

64   Q18 Back

65   Q57 Back

66   Q42 Back

67   Q42 Back

68   Q39 Back

previous page contents next page

House of Commons home page Parliament home page House of Lords home page search page enquiries index

© Parliamentary copyright 2002
Prepared 14 November 2002