The 9th International Anti-Corruption Conference
DO AID AGENCIES HAVE A COMPARATIVE ADVANTAGE IN FIGHTING CORRUPTION IN
TABLE OF CONTENTS
`Corruption' is integral to the complex patronage networks found in African political systems. To varying degrees, these networks are supported with resources provided by aid, particularly loans from the International Monetary Fund, the World Bank, and the African Development Bank. Both directly and indirectly, aid is heavily implicated in the corruption that it routinely condemns among recipient governments. Consequently, corruption in Africa is not a problem that aid agencies can help solve ... with more aid.
The World Bank has nailed its flag to the global anti-corruption mast by adopting a `zero tolerance' strategy towards corruption among borrowing governments. By now, the logical outcome of zero tolerance should have been a dramatic reduction in lending to countries suffering from `endemic' or `systemic' corruption, including most of sub-Saharan Africa. But apart from Kenya, it appears to be more or less `business as usual' for the Bank and the rest of the aid industry in Africa. Donors, it seems, are not about to undermine their disbursement imperatives because of their (very real) concerns with corruption.
Normally, the World Bank's well known `approval culture' overrides any attempts to make aid to Africa dependent on the level of corruption. Both external and internal assessments of Bank lending show scarcely believable levels of project failure over many years. Projects that deliver significant and sustainable benefits to poor people are the exception, not the rule, especially in sub-Saharan Africa. Corruption, patronage, mismanagement and waste explain this widespread project failure.
There is a growing consensus that structural adjustment lending has failed to turn around African economies, and that the poor have been the major victims.
The widespread failure of project-based lending and structural adjustment programmes to reduce poverty, the central role of corruption in this failure, and sub-Saharan Africa's resultant $200 billion of unpayable external debt constitute the bleak backdrop to my subject.
Current donor-funded anti-corruption initiatives take three main forms:
The paper points out the weaknesses of this approach. Initiatives that would complement those listed above include a much greater degree of transparency in negotiating sectoral and structural adjustment loans, and a full-scale, multi-donor social, economic, and environmental impact assessment programme to determine the benefits derived from poverty-focused aid of all kinds.
The entire aid industry -- the IFIs and the UN system, bilaterals and the EU -- needs to be subjected to the kind of transparency and accountability that it (theoretically) demands of aid recipients.
Such proposals are unlikely to appeal to the aid industry or aid recipients because:
Pressures to increase transparency and accountability among aid donors and recipients will therefore have to come from outside the agencies, from the N GO sector, political pressure groups and grassroots movements, and from democratic assemblies in donor countries unwilling to further subsidise a non-performing industry. The paper looks for ways to operationalise this conclusion.2
Corruption is one of the major factors contributing to the failure of aid to promote African development.3
There is a growing consensus that:
All the above criticisms of aid, and others, apply particularly acutely to the Sub-Saharan African context. Yet one would be forgiven for concluding that it is more or less `business as usual' for donor agencies in Africa. It is true that corruption has become central to donor discourse in their overall concern with improved `governance'. Aid flows have fallen overall, and in some cases, including Kenya, corruption has been cited as one reason for the suspension of aid. 5 Various donor initiatives, discussed below, are underway to help confront corruption in Africa and elsewhere.
Yet these initiatives have had little significant impact on the way donors conduct their affairs. While tinkering at the periphery, donors studiously avoid the deeper issues that undermine the effectiveness of what they do. In the absence of a more persuasive development discourse, aid agencies -- led, of course, by the World Bank -- continue to define the parameters of the debate on aid effectiveness, including corruption, in a narrowly technocratic manner.
This paper considers the nature of corruption in the aid industry, critiques current steps by aid agencies to deal with corruption, and spells out the preconditions for a systematic approach to addressing the question of corruption in aid.
Aid is not homogeneous, and generalisations about `corruption in aid' need careful qualification. Most aid workers are not corrupt but some are, and in some agencies corruption amounts to more than individual initiatives by rogue staff members. Similarly, not all African government officials managing aid are corrupt. Generally, however, they are highly vulnerable to the imperatives of patronage, including pressures from their political bosses and private actors.
Aid money -- a windfall gain similar to a natural resource `rent' -- can be quite easily turned to `unintended' purposes. Low salaries -- often cited as a major cause of petty corruption -- are a relevant, though secondary, consideration.
To identify the comparative advantage of aid agencies in fighting corruption in Africa, the main issues are the following:
Does patronage always imply corruption? Not necessarily. For example, assume a principal secretary in `Azania' uses his influence to have his home district chosen as a pilot for a World Bank project loan to co-finance a local education development fund. This may be used to strengthen relations with local élites, including politicians, bureaucrats and businessmen. There may or may not be a concerted effort subsequently to misuse project funds or opportunities. The important point is that aid, particularly in the form of project loans, opens up many opportunities for patronage, including project location, hiring project staff, training opportunities, procurement and purchases, consultancy, and so on. In the process, covenants and conditions for loans are widely ignored or flounted. Aid agencies are generally aware of this, frequently turning a blind eye.6 Corrupt agencies and staff are intimately involved in the processes of patronage themselves, of course.
Corruption and patronage should also be distinguished from waste, which characterises much project and other aid. Waste is the use of aid resources for legal but unnecessary and/or non-productive purposes. For example, projects are often heavily front-loaded with vehicles, computers and other office equipment and a range of training activities, study tours, and workshops/seminars of dubious (ie non- productive) utility. Technical assistance and consultancy inputs, large components of `aid', often come into the latter category.
Though analytically distinct, corruption, patronage and waste are usually found together, feeding off and reinforcing one other in a systemic fashion. These three, together with normal bureaucratic slowness and inefficiency, which slow down project implementation, help explain the incredibly poor performance of most loan-financed projects in Africa.7
The extent of official misuse of aid is not easily assessed, since much routine corruption is not recorded or publicised. The phenomenal amount of unexplained and irregular government expenditure reported by Auditors General in both recurrent and capital budgets gives an impression of the extent of the problem (Cooksey 1999c). (See Appendix 1). Even where aid money is not directly misappropriated, the conceptual `fungibility' of aid means that any `misuse' of government revenue implicates aid in proportion to aid's overall contribution to government revenues. Thus, if Azania's total budget is donor funded to the tune of 25%, the purchase of one thousand vehicles for the Azanian army (with or without corruption) will effectively be one quarter financed by aid money, including ESAF loans from the IMF. An ESAF loan has helped finance Zimbabwe's involvement in the Democratic Republic of Congo's on-going civil war.8
A less subtle `funging' takes place when IFI loans disappear through direct embezzlement by government officials, as in Indonesia until recently, or the former Soviet Union. The latitude for this practice is quite limited in Africa, where large chunks of new financial aid go straight back to the IFIs to pay interest on previous loans. For a number of years, many African countries have been paying the IFIs more than they receive in new credits. Most corruption is in sectoral projects or `SECALS'.
Other variable amounts of aid money are directly appropriated for unintended purposes, the mechanisms of which are described below.
This is the most crucial area for assessing the impact of corruption in aid, since the IFIs account for a growing proportion of total aid flows to Africa, and a growing proportion of Africa's foreign debt. (See Appendix 2).
World Bank-funded projects are frequently implemented through project implementation units based in central ministries. These units are staffed by officials who hold allegiance to their hierarchical superiors and their political bosses. At the implementation phase, it is relatively easy for the imperatives of corruption, patronage and waste to take over from the formal project objectives, design and management in determining project outcomes.
The Bank admits that project backstopping and supervision from Washington are usually inadequate, and also admits to contributing to the slow implementation of projects. The Bank's frequently criticised `approval culture' means that efforts are concentrated in identifying and appraising new projects, not in assuring the successful completion of those already underway. Although the Bank's Operations Evaluation Department routinely awards poor marks to both Bank staff and national project implementers, the OED's recommendations for improving performance are routinely ignored in the interest of maintaining the Bank's project approval momentum. This process has been much remarked upon and never effectively addressed.
A recent example:
In the course of writing this report, [World] Bank employees have complained that projects are indifferently audited, or, even where audits reveal diversions, that there are no substantial consequences for diversion because the internal institutional incentives of the Bank spur lending. (Thomas and Barkan 1998:27, based on interviews in Africa)
Beneficiaries from corruption in loan-funded projects include the private contractors and procurement agents, both local and expatriate, who are awarded contracts through rigged tendering arrangements. In East Africa, non-indigenous businesses are intimately involved in these processes. Inside contacts in the `system' provide intelligence on opportunities. The best placed businesses are politically well connected, and may help fund the ruling party's electoral coffers from income from aid-related contracts. In this way, aid money helps keep ruling élites in power (Chabal and Daloz 1999: Chapters 7 and 9).
Corruption may be found at any point in the project cycle. A project may be identified with corruption in mind. Corruption may be involved in selecting the consultants who appraise, implement, evaluate or review a project. Corruption may allow projects to be used for illicit activities during implementation. Procurement contracts may be awarded to relatives and cronies. Project funds earmarked for particular purposes may be turned to others. And so on.
World Bank projects have a high failure rate, partly as a result of corruption. Bilateral (grant financed) projects are somewhat different, in that typically the donor retains substantial controls over finance and implementation. This allows for a greater proportion of funds to be used for their intended purpose, helped with a variable degree of modern management and technical competence.
Consequently, some benefits are delivered as intended. Examples would be a road built under tied aid, or an immunisation programme run outside the Ministry of Health, or a rural shallow wells project undertaken by a consultancy company.9
The criticism that such arrangements weaken government, are costly and unsustainable, has led to a new programme orientation among donors. Local élites are usually against such `ring-fenced' projects. Despite the sector programme approach, donors continue to be extremely project-prone in their overall support strategies. They are profoundly sceptical about African governments' ability to use aid for intended purposes, and are nervous about critical commentary at home that could jeopardise future funding. Examples abound where bilaterals have denied or turned a blind eye to evidence of corruption.
After a devastating report on its lending performance in over 1,500 projects, worth $28 billion, the African Development Bank was substantially restructured and its board of directors replaced (Knox 1994). The report states: `The hardest task will be to instil a new culture, which will include fundamental changes in personnel management and in governance.' (page 34). The ADB's African Development Fund gives soft loans to Africa's poorest countries. During 1990-96 the ADF accounted for eight percent of all multilateral loans to Africa.10
The most pernicious form of corruption in aid concerns loans for essentially bogus projects involving collusion between senior agency and government officials. Agriculture, livestock and irrigation are favourite targets for this kind of project, though multisectoral projects are also found. Projects are typically located in remote regions, where the ostensible `target group' of farmers and so on is unlikely to complain that no benefits are forthcoming. Such projects are `ringfenced' in their own way and are by definition impervious to public scrutiny.
Lack of transparency and accountability, particularly in the multilateral agencies, the European Union and the UN system, presents particular obstacles to efforts at corruption control. Over the years, a number of UN agencies have suffered from cronyism and corruption, including UNESCO and FAO. The recent spectacular events in the European Commission climaxing in the resignation of the entire Commission, demonstrate how agencies may fail to practice what they preach in their development ideology, namely good governance, transparency and integrity. 11
In general, bilateral agencies harbour the least internal corruption. As a rule of thumb, the degree of corruption in national politics and the limits on public accountability in national life determine the degree of corruption in bilateral aid. Aid from France, Belgium and Italy has been particularly vulnerable to abuse.
France has a unique profile in this context, given its close ties with its former African colonies. Official financial and political support has helped promote French business interests throughout la Francophonie, and neither Gaullist nor Socialist governments have ever considered official corruption a particular problem that might destabilise existing arrangements. 12
Overfunding (see Appendix 3) encourages corruption in part by providing more aid than governments can effectively absorb. Sectoral overfunding reflects a donor tendency to jump into `flavour of the month' activities. Recent examples include democratisation and governance, small-scale credit and support for NGOs. Competing for relatively small markets forces donors to ignore warning signs of lack of capacity, quality or integrity among recipients. Donors have no incentive to look for proof of corruption when they are themselves force-feeding the beneficiaries, who may receive large grants on the basis of mediocre or negative track records. Despite growing support for the private sector and civil society organisations, central government is still the major target for `overfunding' however. Inadequate donor co-ordination has been blamed for overfunding. The World Bank's OED blames inadequate `absorptive capacity' for much project failure.
Joseph Warioba (1998), chairman of Tanzania's 1996 Presidential anti- corruption commission, made a strong case for tracing the source of corruption in aid to the aid agencies themselves. Warioba said that aid to countries like Tanzania "from bilateral donors and the International Financial Institutions consists of very large amounts of money and the sum which disappears through corruption is also very large." Warioba singled out aid for infrastructural projects and tied aid, giving the example of a road project funded by a bilateral donor in which the contractor had the support of the local development mission despite its poor performance.
The tentacular nature of aid means that any corruption related to implementing particular government policies usually involves aid money. Warioba relates examples of corruption in the privatisation process and in contracting a dubious foreign firm to collect debts from a defunct government bank (the fees paid were more than the debts collected).
The recent example of European Union `reconstruction' aid to Bosnia, where losses are being counted in billions of dollars, does not (yet) have a parallel in Africa.
Emergency relief includes food aid, which is also open to systematic abuse by government officials and the private sector. Given present trends in the incidence of civil disorder in Africa, both internal and international, and `natural' catastrophes, it is likely that these forms of aid will increase in importance in years to come. It is said that corrupt procurement officers in relief organisations self-recruit because of the easy pickings.
The breakdown of civil order is likely to have more impact on aid flows than weak or corrupt government. Violence directed against aid workers will have a growing impact on the industry, including non- emergency/humanitarian aid in countries that are nominally stable.13
Having identified corruption as a central governance issue and a constraint on economic growth and social development, the aid industry proceeded to set up programmes and projects aimed at curbing corruption. What they failed to do was to problematise their own role in facilitating, fuelling, condoning and ignoring evidence of growing corruption resulting from their aid.14
This section critically reviews donor agencies' responses to the challenge of corruption.
Most African governments are implementing donor-funded governance programmes covering a range of activities, including anti-corruption, democratisation, rule of law, and support for "civil society", including NGOs. Economic reforms under structural adjustment programmes are intended to level the economic playing field, eliminate unnecessary regulation and encourage investment and growth. This is not the place to review the success of these projects in detail. Many are too recent to have yielded any concrete results to date. A review by Berg (1998) of a wide range of public sector reforms in Africa showed generally mediocre results.
Formally, the aid industry treats its assistance as largely technical. Technical solutions to policy problems mean... more projects. A rule of thumb: if donor-funded anti-corruption initiatives begin to threaten existing élites in any way, they will be quashed, ignored and/or bypassed.
Kenya is cited as a country where corruption has forced donors to suspend aid. This is indeed the case. However, to put this achievement into perspective, it is worth remembering the following. First, countries should not have to reach the Kenyan level of corruption before donors start to pull the plug. Although corruption in Kenya reflects the worst type of patronage politics, many other African countries are not far behind. Second, there are still fifteen World Bank projects being implemented in Kenya, and most donors still have programmes covering all sectors, albeit on a reduced scale. (A number of bilaterals are planning to pull out for good, however). Third, the IMF has been desperately seeking excuses for getting a suspended ESAF back on track, including `conditionality' that KACA, the newly established anti-corruption body, should fill vacancies for top officials, rather than seeing whether the body had any teeth. (It is, of course, typical for donors to allocate aid on pledges for future changes, not for past achievements). Appendix 4 speculates on why Uganda should be pledged two billion dollars of aid in a context of obviously escalating corruption. Donors to Uganda are effectively undermining local anti-corruption efforts by throwing money at the government.
The concept of conditionality is now challenged in the dominant aid paradigm by the concept of partnership. In this scenario, the national government is responsible for implementing policy reforms, not just doing what donors tell them. The local `ownership' of the policy programme still begs the question of `who's policy?' It is unlikely that donors would fund a reform programme based on the revealed policy preferences of ruling élites, which are frequently diametrically opposed to much that donor-inspired reform prorammes aim to achieve.
All major donors attach (theoretical) conditionalities to their aid, including human rights, gender, environment, democratisation, liberalisation, and so on. Though of largely symbolic value, these conditionalities send the message to African governments that all good things come from abroad, including `development' models, and you will have to change your (essentially bad) ways if you want us to help you develop.
African leaders swallow their pride, take the aid, and then proceed as usual. Political scientists Patrick Chabal and Jean-Pascal Daloz (1999:116) put it like this:
Donors ... seem on the whole to have accepted a degree of non- compliance with the stated aims of aid and a level of failure of targeted aid projects which beggar belief. ... African ruling élites were able to use foreign aid primarily as an additional source of clientelistic revenue.
Although the bilaterals are potentially vulnerable to public disenchantment with `aid', the World Bank suffers from no such constraints. George and Sabelli (1994:239) quote the Financial Times (March 1993) commenting on the World Bank's:
relationships with some of its African clients [which] are profoundly unhealthy: protective, secretive or defensive, sometimes all three... all too often, information that should be in the public domain remains secret. When information about grave mismanagement is withheld, confidentiality overlaps with complicity.
The real, effective `partnership' between donor and recipient is based on the tacit agreement `you pretend to implement our conditionalities and we'll pretend to believe you.' The aid industry's ability to maintain the myth of (at least partial) success in the face of overpowering evidence to the contrary (albeit frequently repressed) bears eloquent testimony to the industry's resiliance and adaptability.
Anyone familiar with the World Bank will tell you that those responsible for supervising, monitoring and evaluating projects are fully aware of much of the corruption that goes on, to which they turn a blind eye. One project manager told me that anything under fifty percent of misused project money would be acceptable to him.
In Azania, the World Bank recently considered closing down its entire programme because of the level of corruption. Instead, it decided `to cut back on financing during the current lending cycle while increasing oversight of existing projects.' According to this unpublished source, `While each World Bank project has an external audit on a regular basis, this has proven to be insufficient in tracking and accounting for funds. ... public procurement is a high priority target for reducing corruption.'
Examples abound where external audits have failed to pick up or follow up evidence of corruption. Most such exercises limit themselves to checking whether the accounts tell a plausible story, not whether they tell the truth. (See Appendix 5). Calling in one of the big five accounting companies to do an independent project audit may or may not be a good way of getting at the truth. Thus the need for `value for money' audits at the very least.
The mindset of the powerful reasons thus: if the medicine doesn't work, increase the dose! Generally, adopting a punitive approach on matters of public policy - for example, concerning crime and punishment, or international drug trafficking - has the opposite effect of that intended. In many cases, anti-corruption exercises have led to mistrust, low morale among staff, the possibility of other forms of corruption developing, the deterioration of work performance and speed, and avoidance mechanisms. Anechiarico describes how anti- corruption initiatives have failed in US public administration.16
The `zero tolerance' option has had all these predictable effects. Competitive bidding has been introduced for contracts worth just a few thousand dollars. Donors introduce procedures to weed out the least suspicion of malpractice, succeeding in slowing down project implementation and frustrating the honest actors along with the rest. Onerous bureaucratic procedures become even more onerous. The costs involved in appearing `sqeaky clean' are not counted, since aid is not organised around principles of efficiency or competition. Both aid workers and recipients resent the assumption of guilt until proven innocent.
The World Bank recently began to publish the names of procurement companies and contractors involved in malpractice, banning them from further bidding on Bank-financed projects. Enhanced internal controls have also unearthed maverick staff, including one highly successful power sector specialist who left the bank to serve as an advisor to the government of Pakistan. Smacking of tokenism, these initiatives are not going to make any substantial inroads on the real issues.
The `islands of integrity' initiative championed by the Global Alliance for Africa and Transparency International, that aims to clean up on procurement and tendering in major projects, addresses a very real and urgent issue. This worthwhile initiative must be linked to a more general view of aid-related corruption along the lines described in this paper. This would allow the inclusion of projects and non- project loans where international procurement is not a central feature.
The formidable capacity of the aid industry to withstand well-founded internal and external criticisms in the interest of `business as usual' makes it incumbent on the critic to look for original solutions to the corruption problem. Before anything can be done to address corruption in aid, the problem has to be properly defined and delineated. For this to happen, corruption must be conceptualised in systemic rather than individualistic terms. I have also argued above that corruption is a problem inherent in the aid relationship, not specifically a problem of one side or the other considered separately.
It is also important to try to delineate the relationship between corruption as a factor undermining aid and other factors working in the same direction. In practice this is a very problematic proposition since so many factors are involved, there is little agreement on the underlying causes of the failure of aid to achieve its objectives, and `proof' will always be contested by alternative interpretations. Consequently, political-ideological interests will continue to play a major role in deciding the fate of aid, in Africa and elsewhere. Put differently, solutions to the problem of corruption in aid are likely to take political rather than policy forms.
I have already commented on the ill-advisedness of adopting punitive approaches to corruption control. Preventive measures are required, with punishment a necessary but not all-pervasive or intrusive adjunct.
For both aid agencies and recipients addressing corruption in aid requires evidence of the nature and scope of the problem, a broad commitment to solve it, and a strategy to address it. The aid agencies must realise that their own survival and legitimacy require a radical revision in their mode of operation. This means exposing their `integrity systems' to the rigours of transparency and accountability.
The major governments funding the IFIs, the European Union and the United Nations have the proven capacity to shake up these organisations. What they generally lack is the political will to do so. Evidence that these organisations are not delivering on their development mandate, that internal or recipient-country corruption is a major cause of this failure, and that the poor country debt burden continues to grow heavier as a result, are powerful forces that could eventually precipitate a melt-down of the entire aid edifice. Donors' repeated commitment to poverty reduction/elimination may force them to question the role of aid if little real progress is made on the ground in coming years.
Possible preventive measures include:
To forward the `mainstreaming' of the issues of corruption and aid, moral leadership should be sought from within the donor community, the major development NGOs and foundations, the research and consultancy industries, the private sector, associations of aid workers, and political parties. Partner organisations are required in Africa. The role of Transparency International is mentioned below.
Huberts (1996) asked a panel of 257 experts about the level of corruption and fraud in sixteen multilateral organisations. The average score was 3 on a scale of 1 (very high) to 5 (little or none) with a range from 2.8 (FAO) to 4.0 (NATO). Huberts' sample consisted of delegates to anti-corruption conferences...
There is no question that debt relief is an overwhelming priority for sub-Saharan Africa. Yet it remains to be seen whether the correct lessons are being learned concerning the origins of these debts and the role of the IFIs in their creation. Unless the irresponsible lending perpetrated by the IFIs and other lending institutions can be brought under control, the next generation of foreign debt will not be long in coming. The only possible solution to this problem is enhanced transparency and accountability for both borrowers and lenders. This message will emerge from the political process in general, from NGO advocacy groups, and from policy research organisations and think tanks.
Many Northern NGOs are still arguing for more rather less bilateral aid on the assumption that aid has an overall positive impact (Eurostep et.al. 1997). The fact that most NNGOs are heavily subsidised by their own governments' aid budgets makes it difficult for them to adopt an openly critical stance on corruption issues.
The `Fifty Years is Enough Network' takes the strategic position that the only good IFI is a dead one. While this position might be intellectually attractive, it is highly quixotic in its ambitions to rid the world of the IFIs.17 A carefully organised and evidence- based build up of pressure from multiple and credible sources should have greater impact in the long-run.
Jubilee 2000 is the last and arguably the most interesting in a string of popular aid lobby groups originating in the North. It has taken the position that debt relief should be linked to poverty reduction programmes to be financed from debt savings. This approach is hardly likely to endear Jubilee 2000 to relief-seeking governments. Recently, Jubilee 2000 has taken up the corruption issue.
Transparency International has achieved a great deal in its first five years in bringing the corruption issue onto the front burner and getting Northern governments to agree to outlaw and punish companies paying bribes for contracts. TI has also influenced the way donors think and act on corruption. But TI does not have the resources to take on the huge issue of corruption in aid. Also, since the World Bank has taken up the anti-corruption initiative, it might appear churlish for TI to begin sniping at the Bank's own record, or that of other donors, in this respect.
`Donor dependency' characterises the majority of consultants and academics involved in aid. The pressures on academic and research organisations to earn their keep from contractual work allows donors to define research/consultancy priorities in ways that primarily serve their own interests. African academics suffer from the same constraints.18
The creeping `privatisation' of the aid industry through the activities of such luminaries as George Sorros, Ted Turner and Bill Gates raises the interesting question of whether the private sector can breath new life into an otiose and sclerotic industry.
African governments naturally resent and resist anti-corruption initiatives by donors. In Tanzania, for example, donor representatives who speak out against corruption have been publicly criticised for exceeding their mandate. The Resident Representative of a UN agency was withdrawn from his post when he tried to publicise evidence of corruption linking his regional office with high-ranking government officials.
In Tanzania, a high-level initiative to deepen the `partnership' concept led to the public criticism of Transparency International for `unfairly' ranking Tanzania as one of the most corrupt countries in the 1998 Corruption Perception Index. Much progress was claimed in improving recipient-donor relations from a low point in the early nineties when a large amount of import support funds disappeared. One significant area in which no progress had been made was in monitoring and assessing the impact of aid projects on the poor. 19
There will be stiff opposition from African governments to attempts to raise the issue of corruption in aid, including aid-funded projects. Outside South Africa, the number of African NGOs and grass-roots CBOs with the capacity to take up the issue of aid transparency is very limited. TI's national chapters in Africa could take up the issue of corruption in aid, should they be convinced that the issue merits their attention. The problem is, TI chapters rely on donors for their core funding and other activities, and are therefore unlikely to want to bite the hand that feeds them.
Research on corruption in aid can be included in `integrity system' and service delivery surveys. These surveys document public opinion and experience of corruption, including its impact on the availability, cost and quality of public services. Such surveys are already part of the battery of public service reform tools aimed at improving the performance of central and local governments. Planned national integrity system research includes the issue of corruption in aid, a possible source of the DCI (Donor Corruption Index).
There is a tension, however, between the TI and the World Bank approach to service delivery surveys. For TI, civil society and the private sector needs to be involved in carrying out the surveys, whereas the Bank sees them as part of civil service reform, and therefore under government control. It is unlikely that surveys controlled by the central and/or local government investigating service quality and the role of aid would be carried out objectively, since the real picture would predictably show up the poor performance of both government and donors.
It is also unlikely that donors would get away with funding independent and potentially shocking research for very long. Finding credible local researchers prepared to risk the consequences of being involved in such activities is problematic. The involvement of outside researchers and consultants could lead to the research being branded as a `foreign plot designed to discredit us.'
For some time the World Bank has been sponsoring investigative journalism workshops, national integrity workshops, and other anti- corruption activities. A round of investigative journalism workshops could be run on the theme of corruption in aid. Corruption in aid should figure in integrity workshops. It is worth remembering that the twin risks to investigative journalists, editors and newspaper editors - being bribed or being killed - will be of major proportions, since so many `big people' have so much to hide. Whistle-blowers also risk their lives if they start telling the truth about certain types of `aid'.
The last two are not `punitive' so much as information-empowerment measures, giving all stakeholders the means to take an intelligent position on the impact of aid.
The issue of the corrupt use of aid should be a central concern in the international anti-corruption struggle.
Corruption is a risk factor for all aid ventures, however well intended. Multi-billion dollar structural adjustment loans from the IFIs (Indonesia, Russia), and European Union post-conflict reconstruction grants (Bosnia), have been subjected to systematic embezzlement and international recycling. 20 Those involved include heads of state and their cronies in the private sector, including the Mafia.
In such company Africa might seem a bit player. Yet the $200 billion owed by sub-Saharan countries to aid agencies and other international creditors is hardly inconsequential. By contrast, the $23 billion so far earmarked for debt relief among the nineteen poorest countries looks like a drop in the ocean, even if it has been touted as a `victory for Jubilee 2000'.
The mechanisms for identifying and dealing with corruption in aid- recipient relations are not difficult to identify. The major precondition for addressing corruption in aid is to define the issue correctly. What is currently lacking is the will to assimilate the lessons of the past in order to clean up the donors' own act. This poses particular challenges for the IFIs. Given the unique position of the World Bank in defining the terms of the development debate, including corruption, I speculate on whether a more fundamental critique of corruption in Africa can be invoked in the interest of good sense and in the search for practical solutions.21
The aid industry, led by the World Bank, tends to define corruption in Africa as a local issue that can be addressed by supporting diverse initiatives of a largely technical nature (judicial reform, investigative journalism, `islands of integrity', support for an anti- corruption agency). In all these and more, it is assumed that corruption is an indigenous problem, to which aid offers an assortment of solutions. It is also assumed that corruption is a largely individual process of `rent-seeking' that can be addressed by changing incentive structures, including wages, accountability, and competition.
If corruption were to be viewed as a systemic political phenomenon, aid agencies would have to ask who their local `anti-corruption' allies might be in any given context. It is not enough for a head of state to adopt a formally anti-corruption stance: he may do this just to placate donors. If aid agencies and their local `partners' have diametrically different definitions of the situation it is difficult to see how the demonstration of largely technical solutions to corruption will have any impact.
Aid is the main vehicle through which external pressures are brought to bear to force the pace of democratisation, governance and anti- corruption initiatives. Yet a growing number of observers now see aid as a major cause of the corruption and governance problems which it is concerned to address. At best, donors have tended to turn a blind eye to the misuse of their aid; at worst, certain aid agencies and agency officials have themselves been actively involved in corrupt practices.
Africa's current crushing foreign debt will eventually be cancelled. Between now and then we must prepare the way for assuring that future aid does not lead to the same disastrous results again.
Endnotes and References
Lack of space prevents a full justification of many of the points made, with appropriate references and qualifications. The sympathetic reader may be prepared to `take my word for it' where full justification for statements made is lacking. Much of what follows is based on personal experience of over twenty years working in and around the aid industry, mostly in East Africa.
See for example Adams (1991), Caufield (1996) George and Sabelli (1994), Hancock (1989), Payer (1982, 1991), Sampson (1981). The World Bank provides its own rich literature on aid effectiveness, coming to often similar conclusions to the above sources, minus the moral indignation. Dollar and Pritchett (1998), for example, recommend that in weak and corrupt policy environments, donors should not give financial assistance but merely policy advice, a recommendation with profound implications for the future of aid to Africa.
The bigger picture for situating aid (in)effectiveness needs to include the globalisation of trade, investment and finance (UNDP 1999) and impending environmental catastrophies of all kinds resulting from economic and population growth, deforestation, loss of bio-diversity, pollution and global warming.
The Wapenhans Report (1992) found that `borrowers respected their agreements in only twenty-two percent of the loans examined.' As a result, "the high incidence of non-compliance undermines the Bank's credibility." George and Sabelli (1994:224).
The Bank's OED evaluations of Bank lending performance make depressing reading, though they generally avoid the `C' word. OED is marginal to Bank lending activities, and staff negotiating new loans are often unaware of the history of loans to the country/sector in question. The Bank is frequently `trying a new approach', making historical concerns irrelevant.
The IMF complains that the government of Zimbabwe provided it with false figures of the cost of its involvement in Congo (actual cost: over $160 million since January). We might ask why Zimbabwe is involved in this (relatively distant war). Other similar examples would include Uganda (Sudan, DRC), Russia (Chechnya), and Indonesia (East Timor). On the criminalisation of the African state, see Bayart et. al. (1998).
Poor project performance has led donors to try an alternative delivery mechanism of basket-funded sectoral programmes. The constraints on sector-wide programmes are even more daunting than those on projects: see Cooksey (1999a). The World Bank claims a fall in the proportion of unsuccessful projects in recent years (Dollar and Pritchett 1998). Why projects fail is discussed in Berg (1993, 1998), Therkildsen (1999), Cooksey (1998, 1999), and Ferguson (1991). Not all observers agree that `projects don't work.' For example, in describing the impact of aid to Africa, van de Walle and Johnston (1996) assert that: `Countless examples of highly successful aid projects exist... the impact of aid is clear, and the results of many individual aid projects are tangible...' The authors do identify some problems with aid: the `Dutch disease' (too much aid leads to currency overvaluation and discourages agricultural exports), `enclave' projects that are unsustainable (mentioned above), lack of donor co-ordination...
The Bank's culture was characterised by `anxiety to meet annual lending targets', `reluctance to delegate authority and an unwillingness to accept responsibility,' `lack of transparency in decision-making', and `suspicion and distrust.' (p26). Other quotable quotes from the report: `Project identification is essentially left to the borrower, even though many African countries are insufficiently equipped to identify the right kind of projects for their needs... ... there are some indications that political pressures may have led to the inclusion of inappropriate projects...' (pp13-14). `Borrower ownership is not an issue; beneficiary and community participation is.' (p13). Project monitoring `is not systematic and so does not allow management control and board oversight.' (p8). Supervision was totally inadequate. A quarter of active projects were considered `problems', concentrated in agriculture and rural development, public utilities and the social sectors. But `there is no assessment of the seriousness of the problems nor their impact on project implementation or sustainability.' (p8).
On the UN system George and Sabelli (1994:122) comment: `Corruption in the UN is considered widespread and many of its agencies are, if anything, less open to scrutiny and amendment than the Bank itself. Some are run as personal baronies, like the Food and Agriculture Organisation in Rome where Edouard Saouma held sway until late 1993, under purely nominal board control, for nearly eighteen years.' Saouma was accused of holding up emergency food for Ethiopia for twenty days because of a personal dislike for the local FAO representative. (Hancock 1989:85). I have heard it said that relief efforts were delayed to allow cronies to get in on the act.
It is claimed that President Mitterand sacked his Minister of Co- operation Jean-Pierre Cot on the insistence of some African presidents who were not happy with his `moral intransigence.' During an earlier period, President Giscard d'Estaign's family was involved in business in Zaire, Gabon, Morocco, Chad, the Central African Republic, the Cameroons, Ivory Coast, Mauritania, Niger and Upper Volta. According to the Wall Street Journal: "In many of these countries, the existing regimes are heavily obligated to President Valery Giscard d'Estaing's government for the French military support that keeps them in power." (quoted by Adams 1991:142-3).
In Tanzania, one bilateral agency recently pulled out project workers from a certain region because the government could not guarantee their safety. Aid vehicles were targeted for theft and sale in neighbouring countries.
This `blame the victim' attitude is frequently noted in relation to the World Bank's critique of the pre-adjustment period, when the Bank wholeheartedly supported interventionist and non-market oriented governments in Africa. The heavy investment in state-owned enterprises during the McNamara epoch (1968-81) helped build up the massive debts that are the centre of attention today.
Wolfensohn 1999. The author `warns governments in developing countries that they will jeopardise their foreign assistance and investment by condoning corruption.' An earlier (1997) Bank source (quoted by Riley (1998:137) states that `the evidence of systemic corruption in itself is not a reason to withdraw the Bank's support.'
A good example is the American government's approach to the Colombian drug trade, where an escalating investment of money and manpower bordering on military intervention coincides with a significant improvement in the quality and fall in the price of New York cocaine.
Failure to confront the corruption issue weakens otherwise authoritative writing on aid to Africa. For example, in `Improving Aid to Africa' van de Walle and Johnston (1996) devote one sentence to corruption (p76) whereas Riley (1998), writing on the `Political Economy of Anti-Corruption Strategies in Africa', including the WB's, fails to consider aid as a potential incentive to corruption.
The author of the report, Professor Gerry Helleiner, claimed the CPI was suspect, and that Tanzania's low ranking would discourage Tanzanians involved in anti-corruption efforts. (See Cooksey 1999e). For the moment, there is a significant World Bank-government alliance that is unlikely to respond positively to any proposals for enhanced transparency of the type made here.
The IFIs claim there is as yet `no evidence' that their loans to Russia were part of the laundering of $7 billion through the Bank of New York. The `fungibility' of aid makes the search for direct evidence of embezzled aid funds a red herring.
The omens are not good. For example, the Wapenhans report (1992) led the Bank to assert that: `Only sound, on-the-ground results - the development impact of projects - are true measures of the Bank's contribution to sustainable development.' Quoted by George and Sabelli (1994:225-6). Consequently, the Bank initiated a process for assessing the impact of social sector loans, covering only thirteen percent of Bank lending. (ibid).
Adams, Patricia (1991) Odious Debts: loose lending, corruption and the third world's environmental legacy, Earthscan, London
Anechiarico, Frank (1999) Alternatives to the Law Enforcement Model of Corruption Control, Paper proposed for the 9th IACC, Durban, October 1999.
Bayart, J-F., S Ellis and B Hibou (1998). The Criminalization of the State in Africa, Oxford, James Currey for the International African Institute
Berg, Elliott (1993)Rethinking Technical Cooperation: Reforms for Capacity Building in Africa. United Nations Development Programme, New York
----- (1998) "Aid and Public Sector Reform", Conference on Aid, University of Copenhagen, October 9-10 (draft)
Caufield, Catherine (1996) Masters of Illusion: The World Bank and the Poverty of Nations, Pan Books, London
Chabal, Patrick and Jean-Pascal Daloz (1999) Africa Works: Disorder as political instrument, Oxford, James Currey
Cooksey, Brian (1997a) Corruption in Aid: How much is acceptable? Governance, Vol.1, No. 1, Economic Development Institute, World Bank, Washington, pp 13-18, January 1997. Reprinted in Africa Analysis No. 267 (March 1997)
----- (1997b) Who is responsible for corruption in aid? Transparency International Newsletter, Berlin, June 1997
----- (1999a) Does Aid Cause Corruption? East African Alternatives, Vol. 1, No. 2, Nairobi, January-February 1999
----- (1999b) Should Aid Survive? East African Alternatives, Vol. 1, No. 3, Nairobi, March-April 1999
----- (1999c) Corruption and Poverty: What are the linkages? Technical Review Workshop on the Socio-economic Consequences of Poverty and Corruption, International Centre for Economic Growth, Nairobi, (26th May)
----- (1999d) Corruption Index: Don't Shoot the Messenger!, East African Nairobi, July 5-11
----- (1999e) Dark deals that cast a long shadow, WORLDAID'99 Directory, Kensington Publications, London, pages 7-8
Eurostep, ICVA (1997) The Reality of Aid 1997/8: An Independent Review of Development Cooperation, Eurostep, ICVA, Actionaid, Earthscan Publications, London
Ferguson, (1991) The Anti-Politics Machine, Berkeley, University of California Press
George, Susan and Fabrizio Sabelli (1994) Faith and Credit: The World Bank's secular empire, London, Penguin
Hancock, Graham (1989) Lords of Poverty, London, Macmillan
Institute of Economic Affairs (1998) Our problems, Our solutions: an economic and public policy agenda for Kenya, Nairobi
Huberts, L (1996) Expert views on public corruption around the globe, Department of Political Science and Public Administration, Free University, Amsterdam
Knox, A David (1994)The quest for quality: Report of the Task Force on Project Quality for the African Development Bank, Oxford, Nuffield College
Payer, Cherly (1982) The World Bank. New York, Monthly Review Press
----- (1991) Lent and Lost, London, Zed Books
Riley, Stephen (1998) "The Political Economy of Anti-Corruption Strategies in Africa", in Mark Robinson (Ed.) Corruption and Development, Frank Cass Publishers, London
Robinson, Mark (1998) (Ed.) Corruption and Development, Frank Cass Publishers, London
Sampson, Anthony (1981) The Money Lenders. New York, Viking Press
Therkildsen, Ole (1999) "Public sector reforms in a poor, aid dependent country", Public Administration and Development Jubilee Conference, St. Anne's College, Oxford University, April 12-145
Thomas, Melissa and Joel Barkan (1998) Corruption and political finance in Africa, World Bank, PREM Seminar Series, December 14
United Nations Development Program (1999) Human Development Report, New York
United Republic of Tanzania (1998). Report of the Controller and Auditor-General for the Financial Year ended 30th June 1997, Dar es Salaam
Toye, John (1991) "Is There a New Political Economy of Development?" in Colclough, Christopher (ed) States or Markets? Neo-liberalism and the Development Policy Debate, Clarendon Press, Oxford pp 321-38
van de Walle, Nicolas and Timothy Johnston (1996). Improving Aid to Africa, Overseas Development Centre, Policy Essay No. 21, Washington D.C.
Warioba, Joseph (1998) Corruption in Aid - The Case of Tanzania, Conference on Strengthening Integrity: The Challenges for Asia - A Global Agenda, Kuala Lumpur, Malaysia, 12-16 September
Wolfensohn, James (1999) A back-to-basics anti-corruption strategy, Economic Perspectives, U.S. Information Agency, Bureau of Information, Washington
World Bank (1996) Tanzania: The Challenge of Reforms: Growth, Income and Welfare (Three Volumes), Country Operations Division, East Africa Department, Africa Region, Washington D.C.
----- (1998a) Kenya: Country Assistance Strategy, Washington D.C.
----- (1998b) Uganda: Recommendations for Strengthening the Government of Uganda's Anti-Corruption Program, Poverty Reduction and Social Development, Africa Region
----- (1998c) Tanzania, Agriculture and the World Bank, OED Review, Report No. 18111, Washington DC