Towards balanced aid relationships

Toronto, 19 Oct (Prof Gerry Helleiner*) - “Partnership” between the rich countries and the poor, as the latter struggle for development and poverty reduction, has been part of approved rhetoric in the “development community” for a very long time, at least since 1969 when the Pearson Commission published its report and titled it “Partners in Development”.

It has rarely been effectively practised. Some practitioners, have long doubted whether ‘partnership in development’ was possible.

By the mid-1990s, the donor-driven character of aid programmes and the limited local “ownership” that inevitably accompanied them had brought many analysts and policymakers, at last, to the realization that a new kind of “partnership” between rich nations and poor was required in aid relationships - though not, it must hastily be added, in global economic governance.

To make such new partnerships work and to achieve real developing country “ownership”, there has to be a shift away from the previous relatively passive mind-set, common among aid recipients, towards active leadership in the development of “home-grown” development programmes. Developing countries, particularly their governments, have to develop clearer views as to precisely what forms of external support they require. In one of the relatively few recent statements of African views on these matters, this point is made explicitly and clearly.

Of course, aid donors must mean what they say about rethinking and reforming current aid modalities. However, there is still a curious “disconnect” between donors’ general rhetoric on these issues and actual practice on the ground.

The current donor consensus that “aid works” when domestic policies are of the character that the World Bank perceives as “right”, and when these policies are truly domestically “owned”, is based on World Bank analysis that has been subject to such serious methodological challenge that it cannot be sustained.

In the mid-1990s, in relation to the changing aid relationships in an African country, donor agencies were asked about their understanding of ‘ownership’ issues, and it brought some remarkable responses:

·        “Ownership exists when they (recipients) do what we want them to do but they do so voluntarily;”

·        “We want them to take ownership. Of course they must do what we want. If not they should get their money elsewhere;”

·        “We have to pressure the local government to take ownership;”

·        “We have to be realistic. Our taxpayers want to be sure their money is being used well. They want to know there is someone they can trust, a national of their own country, in charge.”

·        “I routinely instruct my staff to draft terms of reference for technical cooperation projects and then spend half an hour with a local government official on it.”

Yet there is intuitive and obvious sense to the proposition that if overall policies are grotesquely inappropriate, aid is unlikely to have much effect; and that unless sound policies are domestically supported they are unlikely to be sustainable.

Argument as to the details of appropriate policy, sequencing, threshold effects, and the role of initial conditions is bound to continue. Now that domestic ownership is so much emphasized one would expect that, when push comes to shove, such argument between donors and recipients would now more frequently grant the benefit of the doubt to recipients.

But it is still difficult to find hard supporting evidence of any such change in donor behaviour (as opposed to donor rhetoric). Both the international financial institutions and the bilateral donors continue to seek detailed policy influence, even if it is now ostensibly within a recipient-led “comprehensive development framework”.

In any case, the actual role of official development assistance (ODA) is only likely to be comprehensible, and analysis of its effects of use for policymaking purposes, at the level of specific individual countries.

In the extensive experience and literature of structural adjustment and development in low-income countries there has been no shortage of policy prescriptions and performance indicators for the adjusting countries. From early emphasis on macro-economic policies and indicators, to later more microeconomic measures, e.g. privatization and liberalization, to still later emphasis on governance and institutions, and now to poverty reduction, the international financial community has kept the pressure on for policy change and quantifiable measures of their extent.

At the same time, concern has grown over the effects of aid dependence, for which appropriate measures also have to be devised.

As the emphasis has changed, measurement of aid recipients’ “performance” has frequently become more difficult. Measures of “good governance” have been devised—incorporating such elements as the extent of the rule of law, assessments of governmental effectiveness, and the frequency of corrupt and illegitimate payments to officials. So have measures of local “ownership”.

But how to weight and aggregate the disparate components of concepts like these remains subject to argument; in the end it is a matter of arbitrary judgment.

As concepts of poverty expand to incorporate dimensions other than sheer income, together with education, health and the like, and/or anthropometric measures, e.g. weight and height for age, similar problems arise. Vulnerability, powerlessness and voicelessness, emphasized, for example, in the World Development Report, 2000 on poverty are not easy to quantify; power and voice also raise issues of the distribution of income and assets, which has its own huge literature on alternative measurement approaches.

Yet poverty reduction is now proclaimed to be the principal objective of IMF/World Bank programmes and international development assistance. Evidently performance, of the currently approved sort, will be more difficult to measure than it was in the “old days” of IMF credit ceilings, inflation and growth rates.

Of one thing, however, one can be sure: as quickly as new concepts of appropriate policies and performance appear, legions of (primarily Northern) research professionals will embark upon fresh efforts to clarify and quantify them.

One can perhaps understand, and even rationalize, all of this continuing effort to measure policy change and “performance” in the low-income countries which are, after all, the object of global development effort. But there can be no doubt that the effort has been essentially driven by the “needs” of the aid donor community, rather than those of the developing countries themselves.

One cannot help wondering whether equivalent expenditure on the research priorities of policymakers and researchers based within developing countries would not have been far more effective use of “development” funding. I do not propose to enter here into a debate as to what these local research priorities might be; they are bound to be highly area- and country-specific. Rather, I want to call attention to the enormous imbalance in measurement and monitoring effort within the so-called “aid relationship”.

What is most striking in the widely-shared aspiration towards a new form of aid partnership is the failure to follow it up with a more balanced approach to performance monitoring. Although the details have changed, nothing essential has changed in the degree of reporting required of the aid-receiving countries or the intensity of monitoring of their performance by the IMF, World Bank and individual bilateral donors. Indeed, with the introduction of the Poverty Reduction Strategy Paper (PRSP), external demands upon already overstretched authorities in the low-income countries have probably risen.

Nothing has been done, however, to increase the (extremely limited) transparency or accountability of any of the bilateral aid donors or international institutions as they interact with the low-income countries in a purportedly “new” form of aid partnership.

What information, and in what form, would be most useful to the low-income partners in the aid relationship? What performance indicators should be measured and reported on the side of individual aid donors?

At present, the only major official source of aid performance data and performance evaluation is the Development Assistance Committee (DAC) of the OECD. Its published data are the product of information supplied by donors themselves. It uses its own (highly arbitrary) definition of official development assistance (ODA), and what it reports (and evaluates) is only at a highly aggregated level, the level of total performance by each individual donor country. Donor performance evaluations are undertaken via “peer review” by other DAC members.

Aid recipients have not been involved in any DAC decision-making as to the definition of development assistance, the determination of which data to request and report, the nature of its reports and evaluations, etc. Nor have they been involved in performance evaluations. The DAC is very much a donor organization and it is designed to serve the needs of its members. Recipients are not members and have no voice. If its data reporting systems and performance evaluations are of limited usefulness to aid recipients, this should therefore not occasion much surprise.

Unfortunately, there is a significant (typically two-year) lag in the availability even of these data. Valuable as all of these data may be for general and ex post analysis, they are of no use to developing country policymakers who require current, country-specific and detailed information for budget preparation and planning. Nor are the performance indicators and peer evaluations usable in the building of partnerships between the donor community and individual recipient countries.

Northern NGOs have made valiant efforts to provide more independent assessments of aid efforts, and even to publish valuable information on developing country debt to OECD official agencies, but they, like everyone else, are seriously hampered by the lack of transparency in aid and official lending.

It is worth asking what the recipients would really like to have reported and evaluated, if they were in charge of the monitoring and evaluation systems:

·        Obviously the most important consideration for aid recipients is that data and evaluation systems relate to their own budgeting and planning needs—and their own country-specific statistical categories and decision-making timetables. To be useful to them, donor performance monitoring and evaluation must take place at the level of activities within their own countries, the activities over which they, at least in principle, have jurisdiction and can exercise their sovereignty. Strange as it would seem to any visiting Martian seeking to understand how “aid partnerships” work, such recipient-country-level systems, those most likely to be useful to recipients, do not exist.

·        Aid donors evidently feel no compunction to report to the governments of the countries in which they conduct their activities as to what exactly they are doing there, what they have done in the past or what they intend to do in the future, let alone to do so in harmonized categories or according to timetables (or, in some cases, even in a language) that might be most useful to the local authorities. In the relatively infrequent instances when national governments have asked donors to supply such information, they have typically pleaded inability to do so or have complained of the inordinate cost of attempting it.  In consequence the economic decision-making in the more aid-dependent of the low-income countries is severely constrained in terms of critical data.

According to DAC data, official development assistance (ODA) amounts to significant proportions of many recipient countries’ GNP (in 1996-97) -- ranging from about 12% for Uganda, 13.4% for Tanzania, 39% for Nicaragua and Rwanda, and 41% for Mozaambique.

The degree of donor compliance with recipient governmental requests for standardized and timely aid data should therefore be an important performance indicator for donors. Such compliance may depend upon the nature of the data request, but donor recipient dialogue should be able to engender agreement as to what is most useful and feasible to supply. The performance indicator may have to be fairly crude, e.g. a dichotomous (yes/no) measure for each donor.

·        A common popular misconception about ODA is that it is all passed through a recipient government system, even through its budget. For better or for worse, however, this is typically not the case. High proportions of ODA expenditures are made directly to the suppliers of goods and services to aid agencies private firms, NGOs, individuals.

Some of these direct expenditures are made to nationals (firms, NGOs, individuals, sometimes even local rather than national governments) of the recipient country; traditionally, more have gone to foreigners, notably from the donor country. In the latter case, these funds do not register in the balance of payments statistics of either the donor or recipient country, except indirectly when/if the recipients spend some of them in the recipient country. Needless to say, decisions as to the uses and recipients of such “direct funds” are made exclusively by the donors. In Tanzania, where strenuous efforts have purportedly been made to transfer “ownership” of development programmes from aid donors to the government, only 30% of ODA was estimated to flow through the government budget in fiscal year 1999.

The proportion of each donor’s ODA expenditures that finds its way into the national budget system is therefore another reasonable performance indicator for donors; this should be inclusive of debt forgiveness and contributions to debt servicing funds.

·        A related issue is the degree to which donor projects and expenditures are coordinated and integrated into national and sectoral plans and/or recognize the declared priorities of recipient governments. The clearest and simplest manifestation of donor willingness to coordinate their support and follow national leadership is through contributions to sectoral or cross-sector “basket funds”, administered by recipient governments in accordance with objectives and priorities agreed with the contributing donors. Donor support of this kind should be reflected in the data on the share of ODA making its way through the recipients’ budgetary systems. But donors may also consciously tailor their activities and projects to recipient priorities, whether national or sectoral, and/or attempt to coordinate their support, standardize their accounting and reporting systems, reduce transactions costs for recipients, etc. without going all the way to “basket fund” contributions (which some donors are constrained, by their own national legislation, from making). On the other hand, they may continue, as they have so often done in the past, to set their own agendas and “push” projects that are not high in the recipients’ order of priorities.

Some attempt should be made to assess donor coordination and willingness to accept local priorities in a systematic way. To some degree, what transpires in this respect is the product of the recipient government’s determination to take leadership. In this respect, the assessment might be considered as among the most important indicators of the success of the aspired-for partnership, transfer of leadership and achievement of local “ownership”. Perhaps a quantitative (negative) indicator of this, if it is feasible, is the percentage of ODA commitments or expenditures which appear to “stand alone”, outside of agreed priorities or coordination systems.

·        Short-falls from ODA promises

Aid donor announcements and even formal commitments often bear little relationship to subsequent actual disbursements. There are many reasons for this: administrative delays; recipient failure to meet pre-stipulated donor conditions, e.g. on local cost co-financing; changed political or economic circumstances in either donor or recipient countries, etc. By no means all the fault for donor shortfalls (overspending does not often occur) rests with the donors. For effective policymaking, however, one must have reasonably accurate resource projections, on a year-by-year basis, and preferably for longer periods such as are covered by a medium-term expenditure framework (MTEF). It may be more important to have predictable and reliable resource inflows than to have large flows that are highly erratic and uncertain.

There must be a presumption that, where general macroeconomic management remains sound, and particularly in the case of general or sectoral budget support, the primary responsibility for exceptionally large shortfalls rests with the relevant donors. Their actual disbursements should therefore be monitored in the context of their own prior commitments. Their shortfalls, individual and collective, should constitute another performance indicator. It would also be useful to calculate shortfalls in different kinds of ODA, not least those identified as especially valuable in the aid that falls within recipient’s budgetary system and its integration and coordination within national plans and priorities.

·        It is important to recognize the exceptional need for liquidity and contingency finance in the poorest and least developed countries. Their structures and size make them peculiarly vulnerable to “shocks” from weather, terms of trade and even (though this is less widely recognized) private capital flows (Helleiner, 2000b). At the same time, their access to commercial bank finance is limited (and/or costly) and the opportunity cost of the holding of foreign exchange reserves is always high in poor countries. IMF funding availability falls far short of the amounts required fully to offset these countries’ shocks. It is, in any case, even in the case of its so-called “Compensatory and Contingency Finance Facility (CCFF)”, not available without new conditions, and hence delays and heavy transactions costs at a time of already increased pressures on policymakers’ time and energies; the IMF thus can no longer be described as a source of increased “liquidity”, even with respect to the limited funds it can provide.

Bilateral donors, who routinely disburse (collectively) far greater amounts in support of poor countries than the IMF or World Bank, could if they chose purposefully alter the time profile of their disbursements for budget or balance of payments support in response to individual recipient countries’ shock-generated needs for liquidity. Such “compensatory” variability of donor flows would help to impart greater predictability to entire country programmes rather than merely to donor flows; and this could be extremely helpful to recipient countries. Donors might well devote greater attention to this potential stabilization role. Those able to perform such a role should obviously be favourably recognized for doing so rather than recorded as offering unstable and unpredictable finance.

·        The tying of aid has long been recognized as costly to recipients, particularly when it relates both to its use and to its procurement source. It is particularly costly to the poorest countries who are least likely to be able to respond to its potential costs by taking maximum advantage of fungibility.  Despite years of effort, OECD DAC members have still not been able collectively to agree to untie all aid to the least developed countries.

Another obvious donor performance indicator, then, is the percentage of ODA which is provided, whether in project or programme form, on an untied basis with respect to country of procurement. Since some donors have been willing to permit local sourcing or sourcing in other poor countries, while retaining the tying requirement on any “external” expenditures, it would probably be best also to record the percentage of ODA for which such partial sourcing freedom exists.  Technical assistance/cooperation raises so many further issues (see below) that these measures of aid donor tying should be calculated exclusive of technical assistance/cooperation expenditures as well as in total.

·        Technical assistance/cooperation expenditures have played a major role in overall aid to the poorest countries. That role has been controversial and is highly politically sensitive. The emerging consensus among aid analysts is that, great as the need for technical expertise may be in most of the poorest countries, traditional technical assistance/cooperation activities have been signally ineffective in sheer cost-benefit terms. Expatriate expertise is frequently ill-informed and/or insensitive to local realities; typically generates little domestic learning, memory or capacity-building; sometimes serves donor rather than development interests (including donor monitoring and control objectives); and is always extremely costly.

As both developing countries and donors have shifted their emphasis (at least at the level of their rhetoric) to long-term capacity-building, the limitations of the traditional model of expatriate technical assistance have been increasingly recognized.

The latest World Bank research report on African prospects states: “... on balance, it is likely that [these] aid programs have weakened rather than strengthened capacity in Africa. Technical assistance has served to displace local expertise and even substitute for civil servants pulled away to administer aid-funded programs precisely the opposite of the capacity building intentions of both donors and recipients.”

Technical cooperation expenditures in Sub-Saharan Africa still amount to about $4 billions per year and about one-quarter of all bilateral assistance to the region. In some countries these expenditures account for 40% of total ODA. Under the traditional modalities, these numbers are simply too high; and recipients resent their perceived opportunity costs.

Another suitable (negative) donor performance indicator could be the percentage of its aid which is spent upon donor-country tied technical assistance/cooperation. Although there are plenty of “useful” expatriates working in poor countries, the presumption must be that this is not generally now a wise use of limited aid funds, particularly when it has not been requested, and that recipient freedom from procurement tying increases overall cost-effectiveness. Hence good donor performance means a low percentage devoted to tied technical assistance. One could imagine some positive indicator of contributions to long-term capacity-building as a complement to this somewhat “negative” indicator; but this would have to be somewhat subjective and hence more difficult to devise.

·        Fundamental to the credibility and effectiveness of any performance monitoring is the independence of the evaluator(s). Neither the DAC (OECD) nor the Bretton Woods institutions can be trusted to be neutral and apolitical in their assessments of donor performance. (There is room for doubt as to their record of neutrality as to the performance of recipients as well.) Political influences may also bedevil the potential UN role in such activities. Although the UNDP has not as yet shown much interest in issues as potentially sensitive to its own major contributors, it (or possibly UNCTAD) could nevertheless serve as an appropriate financier and organizer of independent donor performance assessments via contracting with private individuals, teams of individuals (“panels”?), or consulting firms to provide these services.

The production of the UNDP’s annual Human Development Report is handled in this manner. So are many of the other research and technical cooperation activities of both UNDP and UNCTAD. Alternatively, the work could be funded and contracted by groups of “like-minded” donors. Whoever the financiers/organizers, it must be clear to all that the assessors retain absolute independence and that the contractors/donors carry zero responsibility for their conclusions.

·        Since change in aid relationships is likely to take some time and since, in any case, every effort should be made by donors to reduce recipient transactions costs and take a longer view, the current one-year cycle for donor consultations and Consultative Group (CG) meetings is too short. The more balanced assessments of donor and recipient performance recommended above, and probably CG meetings themselves, need not take place so frequently. A two-year cycle might be most appropriate for a start.

Aid relationships have been difficult to change in low-income countries. Despite much donor rhetoric on the need for recipient ownership of development programmes and the building of new forms of donor recipient partnership, aid-supported programmes are still basically donor-driven. The continuing imbalance in aid relationships is manifest in many ways. An important and previously neglected dimension of the problem is the imbalance in performance monitoring as between donors and recipients. Whereas the behaviour and performance of low-income developing countries is measured and assessed in ever-increasing detail within the international community, the behaviour and performance of their donor “partners” receives only cursory attention, except at an aggregate level which is of little operational usefulness to individual recipients. When it comes to performance monitoring, as in so many other spheres, the powerful (the donors and the international financial institutions) still call all the shots.

Genuine partnership in development requires the monitoring, by independent assessors, of individual and collective donor performance at the level of individual aid-recipient developing countries. Do donors live up to their rhetoric and their promises? In what measurable ways? It is not difficult to devise measures of donor performance at the recipient country level; and some have been suggested above. Instituting systems of donor performance monitoring at the recipient country level can assist in improving understanding of aid effectiveness; promote the new forms of partnership of which there is so much talk; and, most important, assist policymakers in low-income countries in their difficult task of promoting poverty reduction and development. It is long overdue. It is time for it to be done.

(* Prof Gerry Helleiner is Emeritus Professor of Economics at the Toronto University in Canada. The above, based on some of his recent presentations and writings on this issue, is reproduced with Prof Helleiner’s permission) .

The article by Prof Gerry Helleiner in SUNS #4765, "Development: Towards balanced aid relationships", was abstracted from Prof Helleiner's writings on aid conditionality and ownership, compiled and edited by Mr. Brian Tomlinson of the Canadian Council on International Cooperation (an umbrella organization of Canada's development-oriented NGOs) for the NGO-organized September meeting in San Jose, Costa Rica of the Reality of Aid Project. A fuller version of the article will appear in the September-October edition of the UNDP journal, "Cooperation South".

[c] 2000, SUNS - All rights reserved. May not be reproduced, reprinted or posted to any system or service without specific permission from SUNS. This limitation includes incorporation into a database, distribution via Usenet News, bulletin board systems, mailing lists, print media or broadcast. For information about reproduction or multi-user subscriptions please contact: