VoxEU Column Development

Crushed aid: Why is fragmentation a problem for international aid?

The rapid growth in the fragmentation of aid donors is seen by many to be a burden for recipient countries. This column argues that too much fragmentation is not the issue; the problem is that there is too little competition between the suppliers of aid.

The world of official development assistance is rapidly evolving. Four decades ago, few donors used to grant money to few countries. But aid has since expanded tremendously. New donors emerged, and developed countries created increasing numbers of aid partnerships. Today this trend is even reinforced by the transition of some developing countries from being aid recipients to being aid donors (Brazil, China, Russia, Saudi Arabia or Venezuela, to name but a few).

The multiplication of actors on the aid stage has profoundly shaped the way aid is disbursed. It is now fragmented – received in many small pieces from many donors (Deutscher and Fyson 2009). Aid fragmentation is now high on the donors’ policy agenda. They have pledged to decrease fragmentation by improving coordination and by implementing labour division. The 2005 Paris Declaration and the 2008 Accra Agenda for Action explicitly address these issues. The Development Assistance Committee of the OECD actively participates in monitoring progress in fragmentation reduction.

Fragmentation is considered to be a pressing issue because its costs have been shown to be very large for recipients, to the point that it significantly reduces aid efficiency. Having to deal with a plethora of donor missions, requirements, and consultants considerably reduces the value of aid for recipients. It mobilises a great deal of administrative resources in countries where these are often scarce and would be better employed elsewhere. Our recent research (Frot and Santiso 2008, Frot 2009, Frot and Santiso 2010) contributes to the debate on tackling aid fragmentation by giving an overview of its extent and evolution.

How fragmented is aid?

The numbers related to aid fragmentation are quite stunning. To give an idea of the sheer change in the characteristics of aid disbursements over the last forty years, we present two simple graphs.

Figure 1 plots the average number of partnerships across aid donors. The upper line is the number of developing countries in our dataset, and so is the upper bound of the number of partnerships. In the 1960s, donors disbursed aid to, on average, fewer than 50 countries. In 2006, they did so to more than 100. This average masks the fact that the largest donors disbursed aid to virtually every single developing country in recent years.

A direct consequence of donor portfolio expansion is that developing countries receive aid from a rising number of donors. Figure 2 shows that in 1960 each developing country received aid from, on average, two donors. In 2006, it was from more than 28.

Figure 1. Average number of partnerships per donor, 1960–2007


Figure 2. Average number of donors disbursing aid to a country, 1960–2006


How did this happen?

There is more to fragmentation than a simple increase in the number of actors. A large number of these partnerships actually carry little money. Developing countries often bear the administrative costs of aid with few benefits attached. Figure 3 shows the average donor portfolio size and the average number of significant partnerships in a portfolio, according to the Development Assistance Committee definition of fragmentation that classifies a partnership as insignificant if it receives little aid compared with the donor contribution at the global level.

Figure 3. Widening gap


The growing wedge between portfolio size and the number of significant partnerships is fragmentation at work. From the upper curve we learn that aid donors have chosen to disburse aid to more and more countries. The lower curve tells us that this expansion actually created few significant partnerships.

Significant partnerships typically receive more than 80% of a donor's budget, despite only representing a minority within a portfolio. Similarly, more than 80% of a recipient's aid allocation comes from its significant partnerships. Donors have chosen to have very large portfolios, even if that entails relatively small disbursements in some countries.

Going deeper: Sectoral fragmentation

Fragmentation measured at the country level only gives an approximate representation. If, as advised by Development Assistance Committee, fragmentation should be solved by labour division then we need to go deeper to understand exactly where complementarities can be exploited to reduce fragmentation levels.

In Frot and Santiso (2010), a sectoral analysis reveals that fragmentation has become more pronounced in all sectors. The social sector is the most fragmented, and follows the most pronounced trend towards more fragmentation. Two factors explain why social sectors stand out.

  • First, there was a large allocation shift from the agriculture, industry, transport and energy sectors to social sectors (education, population, government), such that these are today overcrowded.
  • Second, social sectors, often with little capital investment required, are also more subject to fragmentation.

Fragmentation is also the outcome of these allocation shifts towards more micro-projects that can easily proliferate.

The sectoral study also shows that fragmentation even in a single country is a complex notion. Social sectors are often very fragmented while others are not. A single figure for each country is therefore often misleading.

For this reason we propose to use radar plots that indicate at a glance fragmentation in each sector and make comparisons between countries easier. This will help to identify sectors and countries most affected by fragmentation.

The other aspect of fragmentation

The aid community often debates about too much fragmentation, and so usually too many donors. But in many countries there are very few donors. Too little fragmentation, or more precisely too little competition among donors, is also an issue.

For instance in 2007, sectors in some countries attracted more than 2000 simultaneous aid projects. On the other hand, the median number of projects in a sector was 19; the average 44. This observation led us to develop an index that identifies countries where a donor enjoys a kind of monopoly power. We say that this is the case when it disburses a large share of aid in many sectors of a developing country.

The literature on fragmentation usually considers that a single dominant donor is beneficial (see for instance Knack and Rahman 2007), and empirical studies, if taken literally, support this view. On the other hand lack of competition should increase the price of aid, as it would on any traditional market. This price is ill-measured, but the literature on aid lists various "anti-competitive" procedures: tied aid, extraneous conditions, hiring of costly external consultants, etc.

It is peculiar that an abundance of suppliers is criticised in the "aid market", when economics underline the virtue of competition almost everywhere. However in the world of aid, the presence of many donors does not imply competition among them, but more often superposition of costs and administrative procedures. Aid monopolies therefore appear desirable if they cut these costs while barely raising the already overinflated price of aid.

What to do about a lack of competition?

This simple remark suggests that the real issue at the heart of fragmentation is too little competition. Numerous donors only multiply monopoly costs, without bringing the benefits expected from competition.

This has implications for how the donor community tackles fragmentation. The current approach is institution-based. Donors and recipients meet in international meetings, and pledge to act. Progress is monitored by a multilateral institution (OECD’s Development Assessment Committee) that cannot constrain donors to implement their pledges, except through a delicate game of naming and shaming.

We wonder about the efficiency of this approach. To deal with a too heavy administrative weight by creating new administrations is somehow ironic. It remains to be proven that these new institutions will lower transaction costs and manage to implement a labour division that donors are often reluctant to effectively achieve. The problem with this approach is that it basically ignores why aid is fragmented. It does not attempt to change the incentives donors and recipients face, and so is unlikely to radically change their behaviours. In particular, it disregards the lack of competition that creates fragmentation.

In another "decentralised" perspective, fragmentation would be reduced as a result of innovation and evolution in a market for aid where donors would face incentives to be more efficient and accountable. Recipients would be able to opt for the most skilled, efficient suppliers. In this world, insignificant partnerships would be driven away because of their inefficiency. Numerous donors would increase competition, instead of only increasing costs. Barder (2009) describes how market mechanisms could be applied to aid.

This decentralised approach argues that fragmentation is a consequence of the current institutional setting where competition is absent. It directly tries to make fragmentation an unsustainable outcome instead of ruling it out by assumption. Its difficulty lies into designing the set of rules that provide the right incentives, and to make donors accept these rules. This is by no means an easy task, but it is more ambitious and promising.

The donor community has decided to adopt the centralised approach, despite its past failures. An additional failure, though hardly surprising, would nevertheless be regrettable for poor countries and would delay once again the structural reforms aid desperately needs.


Barder, Owen, (2009), "Beyond Planning: Markets and Networks for Better Aid", Centre for Global Development, Working Paper 185.

Frot, Emmanuel and Javier Santiso (2008), “Development Aid and Portfolio Funds: Trends, Volatility and Fragmentation”, OECD Development Centre, Working Paper No. 275.

Frot, Emmanuel (2009), “Early vs. Late in Aid Partnerships and Implications for Tackling Aid Fragmentation”, Working Paper, 2009.

Frot, Emmanuel and Javier Santiso (2010), “Crushed Aid: Fragmentation in Sectoral Aid”, OECD Development Centre, Working Paper, No. 284.

Deutscher, Eckhard and Sara Fyson (2008). “Improving the Effectiveness of Aid”, Finance and Development, September, Volume 45, Number 3.

3,570 Reads