To Stop Project Proliferation, U.N. Consolidates Local Office Funding

You've probably heard of UNDP, the U.N. Development Programme, because it publishes the Human Development Report each year, which includes the influential Human Development Index. You must recognize the name UNICEF (the U.N. Children's Fund), and you might know about the FAO (Food and Agriculture Organization) and WHO (the World Health Organization). But can you tell me what ILO, UNAIDS, UNESCO, UNFPA, UNHCR, UNIDO, UNODC, and UNV are? All these acronyms represent U.N. agencies that deliver aid in Vietnam, each with its own budget, staff, and offices. (Though when I lived there, some shared buildings.) The same agencies operate in parallel in many other countries too, often with overlapping mandates and independent projects.The Associated Press quoted UNDP Administrator Kemal Dervis on this inefficiency last week:

"In the business of public policy and foreign aid, excessive competition is actually very inefficient by encouraging the proliferation of small projects ... and actually hurts the effectiveness of these resources," Dervis said, citing a recent analysis by the Center for Global Development, an independent think tank.

Ten years ago, to deal with the problems created by having so many U.N. aid agencies, Secretary General Kofi Annan created a U.N. agency. The U.N. Development Group (UNDG) has no less than 28 member bodies, whose activities it is supposed to coordinate. As head of UNDP, Dervis chairs it. (Before assuming both posts 18 months ago, Dervis was a fellow at CGD, and before that, finance minister of Turkey.)The UNDG evidently disappointed--unsurprising since it apparently had no control over budgets. But reform is in the air at the United Nations today, and Dervis just announced an interesting pilot initiative to pool funding for local U.N. organizations, starting with six of the twelve in Vietnam. Also slated for mergers are Albania, Cape Verde, Mozambique, Pakistan, Rwanda, Tanzania, and Uruguay. Time will tell whether the reforms will live merely on paper or will become real in the sense of strengthening incentives for U.N. offices to work together, combine projects, and minimize their impositions on overworked officials in recipient governments.In justifying the consolidation program, Dervis appeared to refer to the CGD working paper, Competitive Proliferation of Aid Projects. There, I build a theoretical model that generates some intuitive prescriptions. The focus is on the adminstrative burden that aid projects put on strapped recipient governments--the meetings they must attend, the reports they must file, the "missions" to the "field" of visiting aid officials that they must host. The burden of this oversight activity is assumed to be proportionally larger for small projects. But it comes with a benefit too: closely watched projects run better, with the potential gains being largest where national governance is poorest. The model suggests that aid projects ought to be bigger where national governance is better (out of trust), where recipient resources such as tax revenue are tight (to avoid overloading the government), and where total aid is high (for the same reason).Taking inspiration from the work of Stephen Knack and Aminur Rahman, the paper also shows that if donors are "selfish," caring more about their own projects than other donors' projects, they have an incentive to fund smaller, more numerous aid activities in order to attract the recipient's resources, whether matching funds or officials' time, to donor's own projects and away from other donors. Since aid agencies do not pay for these resources--economists would say a market is missing--they can ignore the costs the diversion imposes on other donors. The result is competitive proliferation, an arms race in which each donor divides its aid pie more finely in a fight for the recipient government's attention. The smallest donors, like many of those U.N. agencies, can come out ahead in the game, but overall development goes down.Since this is a model, I quivered slightly at Dervis's apparent certainty that "excessive competition is actually very inefficient" in the real world. But good models systematize common sense. And it is a real cause for concern that Mozambique hosts as many aid projects as China, some 2,000 by my count. The worry that "donor darlings" are being killed with kindness is voiced in several CGD products, including the Commitment to Development Index aid component, which penalizes project proliferation; a paper by Sheila Herrling and Steven Radelet on the Millennium Challenge Account; and Nancy Birdsall's Seven Deadly Sins, which puts proliferation under the heading of "Envy." An important point in Birdsall's paper is that as easy as it is to poke fun at the inefficiencies of the U.N. system, the many bilateral donors--national government agencies--contribute to proliferation through their mere existence. One solution is for bilaterals to spend less through their own bureacracies and more through multilaterals such as the U.N., World Bank, and the Global Fund to fight AIDS, Tuberculosis and Malaria. And maybe small donor agencies, from Italy to Finland, should just close their doors.