Über die Autorin bzw. den Autor

Clifford F. Zinnes

Auszug. © Genehmigter Nachdruck. Alle Rechte vorbehalten.

Tournament Approaches to Policy Reform

Brookings Institution Press

Chapter One

Why is it that there is almost universal agreement that foreign aid-sometimes called donor aid or development assistance-has not been particularly effective (Collier 2002; Espina and Zinnes 2003; Easterly 2006a and 2006b; World Bank 1998)? Slowly the donor community has come to realize that the problem is not primarily one of insufficient funding. Rather, it appears to be related to the incentives created by the relations among donor country voters, donor organizations, technical assistance implementers, intermediary recipient governments, and final aid beneficiaries, among others (Murrell 2002). Clearly, some of these relationships are horizontal and some are vertical. In some there may be "teamwork" or competition; in others there is hierarchy, which may be well or poorly managed (or supervised).

In short, the root of the problem of aid effectiveness is institutional. By institutions I mean the set of rules, strategies, payoffs, and players, as well as player beliefs about all of these. Thus "institution" may refer to culture, norms, markets, firms, government, organizations, and legislation. It also includes donor-recipient assistance contracts. Associated with this insight is the ever greater attention paid to governance, monitoring and evaluation, indicator design, public participation, and participatory development (Williams and Kushnarova 2004). Likewise, these are not issues limited to donor aid but are also applicable to local-level government services and regulation as well as private sponsors of diverse initiatives. Part of the reason for the emerging deeper understanding of these challenges comes from advances made over the last two decades in what is now called the new institutional economics (NIE).

1.1 Challenges to Effecting Change

Consider briefly the sponsors' conundrum. They wish to provide development assistance to recipients (or, in the case of a central government, revenues to decentralized local governments) in an environment foreign to the sponsors and in such a way that the sponsors' explicit and implicit objectives are met. These tend to be highly multidimensional (Alesina and Dollar 2000) and include a desire for the consequences of sponsor intervention to be sustainable in the long run. Likewise and at least officially, sponsors would like their funds to be applied in a cost-effective fashion. Unfortunately, this is more of a challenge than first meets the eye.

To understand why one must examine the path or steps through which this process might typically pass. First, the sponsor must manage and empower its own staff to identify an appropriate area for recipient country improvement and then determine the requisite intervention to address the problem. It must then find and contract an implementer, which it must monitor. The implementer will generally have to interact with agencies of the recipient government, which may in turn need to delegate to their subordinate territorial units. When presented this way, the opportunities for mistakes as well as malfeasance or shirking are formidable. Given the limited local knowledge as well as operative control the sponsor has in a typical situation, it is no wonder things do not always go the sponsor's way, ignoring whether its own objectives were appropriate in the first place.

These issues of control and monitoring are aspects of what economists refer to as principal-agent problems. They encompass some of the central challenges for the provision of development assistance (Murrell 2002) and thus are dealt with repeatedly in this discussion. Principal-agent issues often may manifest themselves hierarchically in a "chain of command." Clearly, the longer the chain, the more susceptible an intervention is to unanticipated breakdowns or even failure. Such considerations should encourage a modicum of modesty in application designs.

But there are still further challenges. Sponsor interventions such as reform activities, capacity building, or other local improvements generally require that local stakeholders coordinate-or at least cooperate-among themselves. Lack of trust and a "zero-sum" attitude lead to reluctance to work together across political groups, economic sectors, and jurisdictions. Likewise, how can a sponsor separate serious from frivolous local requests for assistance? In other words, how can a potential aid recipient signal to a sponsor its seriousness to engage in high-level efforts, that is, prove that its incentives are aligned with sponsor objectives? What credible commitment mechanism can the recipients employ?

Sponsors have historically, albeit unintentionally, responded to these dilemmas in several ways. One is to assess their own impact by measuring project inputs, for example, the number of entrepreneurs trained, the value of loans placed, amount of seed distributed, or whether an environmental law was promulgated. This is easier to document to donor constituencies back home (and less risky to staff member careers) than proving outcome effectiveness associated with each of these aforementioned inputs, namely, an increase (due to training) in small- or medium-scale enterprise (SME) value added, the sustainable impact of the project loan, the impact on rural household caloric intake, or the degree of firm compliance, respectively.

Donors have also engaged in "conditionality." A country promises to do X in the future and the donor promises to give Y in the present. A more sophisticated version of this is for the donor to require the recipient to "go first," though this is often difficult because Y is often an input to X. There has been a growing literature as to why conditionality often fails. This makes the so-called merit good (X) appear to the recipient (or rather the "agent" of the recipient or principal) as a "price" it must pay to get what it wants (Y). It is hardly likely that the recipient will implement Y with much enthusiasm under such an arrangement. Moreover, where the government is not fully accountable to its population's interests, conditionality tends to be "time inconsistent," namely, once the donor has fulfilled its side of the bargain, it is not in the donor's interest to penalize the country if the government defaults on its side-especially if the government changes over the course of the agreement period. Kenya and the donor community have played this game so many times that The Economist in 1995 called it a "ritual" (as quoted in Svensson 2003). Why would donors tolerate this behavior? Svensson (2003) and others point out that part of the reason is the way in which donors allocate assistance. First, the donor's country teams that develop the projects rarely experience much of an opportunity cost to the funds they program. They operate under the implicit rule of "use it or lose it." Second, within departments of the donor organization, there tend to be positive bureaucratic incentives to maximize their budget size rather than their program effectiveness, an observation made many years ago by Niskanen (1971) in conjunction with government in general.

Now sponsors have come full circle and are beginning to focus on meeting their own objectives more effectively. For example, participants at the 2005 Paris High-Level Forum issued the Paris Declaration on Aid Effectiveness, in...