Thursday, 16 October 2014

The Virus of Multilateral Trust Funds

Would a stronger core budget funding of the World Health Organisation and less dominance of earmarked funding of global health spending have prevented the current rapid spread of Ebola? This daring thesis was expressed last week at a fascinating workshop, sponsored by the Swiss Network for International Studies (SNIS), at Zurich University, on the “Proliferation of Multilateral Funds”. WHO Director Dr. Margaret Chan recently complained: “My budget [is] highly earmarked, so it is driven by what I call donor interests”.

Earmarked multilateral funding has spread like a virus in recent decades, with earmarking defined in thematic, geographic and/or institutional terms. Bilateral donors channel earmarked funds to multilateral development agencies that implement activities for them, but without allowing these agencies to use the funds at their own discretion. The hybrid funding mechanism is often dubbed multi-bi aid. The Figure[1] below shows that earmarked multilateral funding has really won traction, ironically, since the Paris Declaration 2005 that sought to turn aid delivery less burdensome for the recipients, including by simplifying delivery channels. It is based on 680 (!) distinct multilateral development organisations (much more than the 280 ODA eligible counted by the OECD Report on Multilateral Aid). With a volume of 19 billion USD in 2012, multi-bi aid today amounts to almost 60 percent of the volume of multilateral aid.

Figure: Multi-bi aid relative to multilateral contributions,
                                                        in percent, 1990 - 2012

Source: Reinsberg et al. (2014)

Viewed from a principal-agent perspective (with the donors as principals and the multilaterals as agents), the higher share of non-core budgets in multilateral organizations has gone along the move from collective principals to multiple principals, from member country groupings with largely homogenous preferences to groupings with heterogeneous policy goals. Applied to the UN system, the US called the shots after WWII based on its Western European and Latin American allies in what was then a much smaller country grouping. To the extent that the UN enlarged and raised its member base, the US, the UK and other leading countries lost the majority. This explains the start and rise of voluntary multilateral aid funding from the 1960s. Multiple principals have multiple interests that they see better implemented in earmarked rather than general policy programs.

The recent rapid rise of the share of earmarked contributions to multilateral aid over the past two decades, however, needs further explanation. The rise is closely correlated with the establishment of so-called trust funds, either umbrella funds with many donors or, somehow perversely, single-donor funds. Mid-2013, the overall number of active trust funds at the World Bank was more than 900(!).

So what makes trust funds attractive to donor governments? First, the voluntary nature of multi-bi contributions to trust funds provides more flexibility as they are typically independent of long-term agreements at the international level (such as the three-year IDA replenishment rounds). In contrast, they are released on a short-term basis.  Second, governance and management structures can be designed in ways that may also allow for leverage from private charities, foundations, firms, or charities. Third, and perhaps most importantly, trust funds (especially the rapidly rising single-donor funds) can be tailored to the (ever changing) policy priorities of donor governments. Arguably, the vehicle is well suited to mobilise resources to fund global public goods – under the condition that these resources are complementary (which seems largely the case).

What then has made multilateral trust funds so much more attractive? Two further papers presented at the SNIS workshop provide some convincing hints[2]. First, the simple existence of trust funds influences the aid allocation by donors across available channels: the possibility of earmarking multilateral aid decreases donors' contributions to the multilateral's discretionary core budget and the amount of bilateral aid. Second, the existence of earmarking may also stimulate some active agents (the boss or management team of an international organisation) to enlarge their fiefdoms; I for one have worked in such an organization. Third, policy priorities do change but multilaterals can be very slow to accommodate those changes, partly for bureaucratic inertia but also for perfectly justified governance rules; the more this results in incompatible goals between donors and agencies, the higher will be the incentive to create trust funds and earmark funds. Finally, the same impact will occur to the extent that policy priorities become more heterogeneous across donors.

The demise of the United States as the benevolent Kindleberger-type hegemon in a multipolar world  and the rise of the emerging donors such as China would lead to predict that there will be more, not less, earmarked multilateral funding in the future. The multilateral donor chaos is alive and kicking[3].


[1] Modified from and based on a forthcoming article by Reinsberg, Michaleowa and Eichenbauer (2014), “The rise of multi-bi aid and the proliferation of trust funds”, Handbook of Development Economics 2014.
[2] Eichenbauer and Hug (2014), “The politics of special-purpose trust funds”, unpublished, Heidelberg University and Université de Génève; Reinsberg, Michaelowa, and Knack (2014), “Which donors, which funds? The choice of multilateral funds by bilateral donors at the World Bank”, unpublished, University of Zurich and World Bank.
[3] Reisen (2010), "The multilateral donor non-system: towards accountability and efficient role assignment," Economics - The Open-Access, Open-Assessment E-Journal, vol. 4(5), pages 1-22.

Monday, 22 September 2014

BMZ study ´The Future of Multilateral Concessional Finance´ online

Ready by June 2014, Chris Garroway and myself had to wait a while until the BMZ put our study online. Here is the link:

You can discuss with me (or us) at the following events.

Emboldened by a decade of poor-country convergence and poverty reduction, many are now imagining a world without extreme poverty (often defined as 3 percent of the world population living on USD 1.25 Purchasing Power Parity a day or less) in not too distant a future.  This paper presents the major determinants of the future demand for concessional finance and produces scenarios for multilateral concessional finance eligibility. It then discusses general strategic implications for the future orientation of multilateral concessional finance and presents actionable options for the development of each the International Development Association (IDA), the Asian Development Fund (ADF), the African Development Fund (AfDF) and concessional facilities of the International Monetary Fund (IMF).

 This report considers four major determinants of the future demand for concessional finance: (i) national productivity and welfare, (ii) the extent of poverty and deprivation, (iii) the capacity to mobilize domestic financial resources and (iv) the vulnerability to exogenous shocks and global public “bads”.

 The paper finds that recent studies on poverty and growth projections have been overly optimistic. The 2000s may well have been a special decade, and growth rates in the coming decades may hardly approach the past high levels. This paper projects the number of countries eligible for multilateral soft finance to decline to 26 in 2025, down from 39 in 2012 (based on GNI per capita simulations).

The number of extreme poor population will have been halved globally by 2025, estimated at still more than half a billion according to the basic scenario of the study. Prospective graduates India and Nigeria as well as Democratic Republic Congo are likely to constitute half of global poverty ten years from now. At the same time, the distribution of incomes has blurred the distinction between poor people and poor countries. Projected relative poverty headcount ratios indicate that even if the number of extreme poor population is strongly reduced globally by 2025, most countries will face sizeable problems with social exclusion and relative deprivation. 

Higher domestic resource mobilization may reduce reliance on concessional flows. However, marginal tax rates required to close the poverty gap remain prohibitively high in most developing countries. Moreover, tax effort calculations show little untapped potential for domestic resource mobilization for most Sub-Saharan African countries. In South Asian countries - in particular Bangladesh, India and Pakistan - where a higher domestic tax effort could be envisaged to help combat extreme poverty, challenges for fiscal federalism and cross-state revenue sharing remain nonetheless formidable and require technical support.

Climate change and natural disasters threaten to derail efforts to eradicate poverty over the next decade. In Asia, disaster damage cost in selected IDA recipient countries is substantially higher than concessional IDA flows. In Africa, disaster costs are of about the same magnitude as IDA credits. A major political decision hence relates to the provisioning of global public goods through multilateral concessional finance, especially adaptation to and mitigation of climate change as well as disaster management. Mainstreaming climate change into development cooperation would require multilateral donors to integrate vulnerability to environmental and global risks into their allocation criteria for concessional funds.

Abovementioned uncertainties would suggest a gradualist, precautionary and insurance-oriented approach to the future of multilateral concessional windows. Shrinking multilateral aid would ignore the option value of preserving IFIs and their concessional windows in a world with considerable uncertainty about future poverty outcomes. Realistic options presented in this study are:

  1. redefining eligibility criteria for concessional funds based either on relative or absolute poverty terms, e.g. as per capita income relative to per capita income in a specified grouping, or even as overall levels of extreme poverty; alternatively new and more comprehensive measures such as the UN Human Development Index or the Multidimensional Poverty Index could be considered; 
  2. smoothing transition periods from IDA-only via blend to IBRD-status for upper-middle-income countries (UMICs), with similar graduation paths in the other multilateral windows; such an “IDA+ transitional window” would be available for countries with a per capita income between the current IDA threshold and its double, and funds could be directed towards measures of social inclusion and redistribution; 
  3. strengthening sub-sovereign allocation to take account of the rural-urban duality of inequality and higher disaster risks in certain provinces; 
  4. opening the multilateral-soft windows for regional and global public goods, with climate change mitigation and disaster risk management as tracer sectors; this could take the form of turning MDBs into global/regional public goods facilities or the greening of MDB’s projects.
Finally, this paper presents strategic options for the four soft windows covered in the analysis. It suggests for the IMF to increase its share of blended finance, increase its grant element for Poverty Reduction Growth Trust (PRGT)-only countries and add an insurance-type instrument to the PRGT lending facilities. For IDA, often a lead institution in defining rules for concessional finance, it recommends increasing the grant element of its loans, considering a two-window approach with the second window focusing on transitional support, focusing the performance-based allocation (PBA) on direct poverty reduction outcomes and including vulnerability to environmental and global risks in the allocation criteria. The division of labor between IDA and the AfDF needs to be sharpened in the conceivable case of a largely overlapping client group. The report concludes that the provision of regional public goods in the form of trade development may become an AfDF focus while IDA could concentrate on climate-change related finance.  The AfDF may further consider departing from an IDA-pegged to a specific allocation mechanism tracking the AfDF’s performance; recent changes to its loan policy in its core fields of infrastructure development might be extended to regional integration by the means of structural indicators. The ADF on the other hand, with its largely different client base, is already on a good track with its current effort to merge the ADF with ordinary capital resources (OCR), thereby increasing the institution’s lending and leveraging capacity; the latter would be greatly enhanced by raising China´s and India´s capital shares.

Saturday, 12 July 2014

VI BRICS Summit: Will the BRICS Bank Help Reform Global Financial Architecture?

The VI BRICS Summit, which Brazil will host in Fortaleza and Brasilia from 14th July, will unveil a new multilateral development bank not led by the West. Let´s call it the BRICS bank (it may be called the New Development Bank[1]). I want to reflect on the normative power that would be unleashed in changing global financial governance by the creation of the BRICS bank. What will be the prospects for hastened changes in the global governance structure? Will it be rebalanced away from advanced-country dependence toward the BRICS? The answer comes in three steps:

  • First, the starting point is the current imbalance of emerging powers´ capital shares and voting rights in the existing multilateral banking system; the more imbalanced it is, the higher the pressure to rebalance toward fairer representation.
  • Second, the extent of excess demand (financing gap) for multilateral soft loans will define the demand for concessional flows from the new BRICS bank; joint with relative lending capacity, this will guide how much business – hence political influence- the existing Bretton Woods institutions and Western-led regional development banks might lose in favor of the new competitor bank.
  • Third, how much time it will take for a new BRICS bank to generate the knowledge and ´certification value´ that the existing IFIs have acquired already.

@First: The recalibration of the world economy toward the BRICS is still not reflected in the global financial architecture. The BRICS represent 46% of world population and almost 20% of world GDP in current dollars[2]. At the World Bank the five BRICS together have just 13-14% of shares and votes, according to the IBRD Statement of Subscriptions to Capital Stock and Voting Power (Table 1). By contrast, the G7 group of advanced countries represents only 15% of world population; its share of world GDP corresponds roughly now to its voting shares at the IBRD.

Table 1: IBRD Statement of Subscriptions to Capital Stock and Voting Power, October 2013

Country Group
Capital Stock Shares, %
Voting Power, %

Europe has been a stumbling block toward reform, staying overrepresented in the executive boards of World Bank, IMF and regional development banks. Although overrepresented, Europe´s voice is not united and hence weaker than necessary. Meanwhile, the US retains a blocking minority at the IMF (and informally, joint with allies, at the other international financial institutions). Early 2014, international financial reform and the G20 have suffered a serious blow after the US Congress refused to ratify a capital increase for the International Monetary Fund agreed four years ago. Advanced economies have reneged on their promise to support greater voice and representation for the BRICS and other emerging economies in global governance arrangements. BRICS have thus little incentive to take more responsibility as important stakeholders of the global economy and as financiers of global public goods.

The Asian Development Bank, firmly ruled by Japan and the US, provides an especially stark case of distorted representation. ADB members who are also members of OECD hold 64.6% of total subscribed capital and 58.5% of total voting rights. By contrast, China and India (the other three BRICS are not ADB members…) combine a mere 10.9% of voting rights (Table 2). Japan and the US are by far the biggest shareholders in the ADB with 15.7 per cent and 15.6 per cent, respectively. China, whose economy in dollar terms surpassed Japan’s in 2010, has just 5.5 per cent of voting rights; India, soon to be Asia´s and the world´s most populous country, has 5.4%.

Table 2: ADB Subscribed Capital and Voting Power, end 2013

Country Group
Subscribed Capital Shares, %
Voting Power, %
BRICS (China + India)

Source: Author´s calculation;


An immediate negative consequence of uneven representation is the negative impact on capital resources (to which China could amply provide) and hence lending capacity. Financial constraints on both the concessional window ADF and ordinary capital resources (OCR) are stretching ADB’s capacity to the limit. If ADB is to maintain meaningful levels of involvement in poor ADF countries, it has to find creative ways to enhance its financial capacity[3] – or it has to change representation as Japan´s fiscal resources are limited by rapid ageing, with the risk to turn Japan into a ´middling donor´ (Sawada, 2014)[4]. Although Japan is the largest financial contributor to the ADB, its policy positions are usually framed within the parameters set by the US-Japan relationship, which has effectively limited higher representation of and core funding by China and India in particular.

@ Second: Potential demand for concessional flows from the new BRICS-Bank and its relative lending capacity will determine what share of the business – hence political influence- the existing Bretton-Woods institutions and Western-led regional development banks might lose in favour of the new competitor bank. In a recent UNCTAD paper, Stephany Griffith-Jones has assembled the evidence for the shortage of long-term finance, especially to finance infrastructure, in the developing and emerging countries[5]. Current annual spending on infrastructure in developing and emerging countries has been estimated at $0.8-0.9 trn; existing multilateral development banks contribute merely $40-60bn to that sum while the bulk is being financed from national government budgets ($500-600bn). The annual spending on developing-country infrastructure to finance access to water, electricity, transport and other infrastructure needed to combat poverty, deprivation and climate change have been estimated by various sources (MacQuarie; Estache; MDB working group on infrastructure) at $1.8-2.3trn. The resulting financing gap would be around $1.0-1.4trn.

Figure 1: The Annual Infrastructure Financing Gap

Source: Bhattacharia & Romani (2013)

China – not the other BRICS - has the financial fire power to close a large part of that gap – given the wright incentives and institutions. The politically correct way is to write BRICS and to mean China. The fact that the financing power between China and the other BRICS is very asymmetric, however, is central to future financial governance developments. The new BRICS bank will initially have 50bn subscribed capital, to which each of the five BRICS will contribute $10bn. This sum is negligible by China´s standards, but corresponds to 2.5% of South Africa´s current GDP and almost 9% of her annual tax revenues. Note, however, that just $10bn of capital has to be paid-in ($40bn in guarantees) and that up to 45% of the BRICS bank capital will be allowed to be held by non-BRICS[6]. Nonetheless, the political prior to have equal say at the BRICS bank joint with the limited firepower notably of South Africa provides a certain constraint on the BRICS bank capacity to divert an important part of the multilateral lending business away from the traditional development banks. This may also explain why China has been pursuing recently the creation of yet another multilateral development bank (China-led), the Asian Infrastructure Investment Bank (AIIB), with registered capital planned at $100bn initially double the size of the BRICS bank[7].

@ Third: Stephany Griffith-Jones (op.cit.) provides some estimates on the BRICS bank lending capacity based on the assumption of an eventual total capital endowment of $100 bn. According to her estimates, the level of annual lending could reach, after 20 years, a stock of loans of up to US$350 billion, equivalent to about US$34 billion annually. The latter amount could be used for investment projects worth at least US$68 billion annually, given that there would be co-financing by private and public lenders and investors. Similar to the Latin American CAF development bank, a loan-capital leverage ratio of 2.4 underlies these estimates, among others (profits, rating level). The prospective annual BRICS bank lending sum would be roughly half of the annual flows provided by the major multilateral Western-led IFIs, worth $63bn in 2011, according to latest OECD data. Without taking account of the AIIB and the smaller multilaterals, the BRICS bank might be able to capture roughly a third of mainstream development bank business. Not negligible at all, enough I think to hasten global governance changes. The BRICS bank will be wise to focus on infrastructure finance not merely for the present excess demand but also to deploy China´s knowledge and comparative advantage in that area. A focus on well-defined project finance will help the viability of return to lending for the new BRICS bank.


[1] Christopher Wood (2014), The BRICS New Development Bank and Currency Reserve Arrangement at a glance,, The South African Institute of International Affairs, Pretoria, July 8
[2] VI BRICS Summit (2014),
[3] ADB has recently presented a new proposal to enhance ADB’s financial capacity through a modified management of its capital resources. The proposal entails terminating ADF loan operations and combining ADF loans (and part of ADF liquid assets, projected to be USD 35.3bn in total) with the OCR balance sheet in January 2017. This would increase OCR equity from a projected USD 17.9bn to USD 53.2bn. ADF would henceforth provide only grant assistance, while ADB would continue concessional lending through its OCR window
[4] Yasuyuki Sawada (2014), Japan’s Strategy for Economic Cooperation with Asian Countries, Policy Research Institute, Ministry of Finance, Japan, Public Policy Review, Vol.10, No.1, March 2014.
[5] Stephany Griffith-Jones (2014), „A Brics Development Bank: A Dream Coming True?”, UNCTAD Discussion Papers #215. Her estimates are based on a presentation by Amar Bhattacharya and Mattia Romani (2013), “Meeting the infrastructure challenge; the case for a new development bank”.