In October 2015, the Washington Post ran a story that compared the World Bank’s performance to that of other bilateral and multilateral development finance institutions. It identified the Bank as a leader among its peers in the value-for-money that it provides to its shareholders (and their taxpayers).
The story was based upon the findings from a report entitled Listening to Leaders: Who Do Developing World Leaders Listen To and Why? The report was produced by AidData, a research lab at the College of William and Mary that fielded a survey of nearly 7,000 development policymakers and practitioners in the summer of 2014. This survey generated comparative data on the development policy influence of 100+ bilateral and multilateral institutions. Government officials from 126 low-income and middle-income countries and semi-autonomous territories were asked to evaluate the agenda-setting influence of development institutions—more specifically, the level of influence that a given bilateral or multilateral development institution had on their government’s decision to pursue specific reforms in a specific policy domain. Participants in the survey were only given the opportunity to evaluate institutions that were actively engaged in their country and in their area of policy expertise. They were also only given the opportunity to evaluate institutions that they had firsthand experience working with between 2004 and 2013.
Once the 2014 Reform Efforts Survey closed, AidData aggregated and weighted the micro-data to identify which aid agencies and development banks have the most influence on the development policy priorities of low-income and middle-income governments. Then, by comparing each institution’s actual agenda-setting influence against its predicted agenda-setting influence (on the basis of their average annual grant and loan volumes), AidData constructed an index that identifies which bilateral and multilateral development partners punch at, above, or below their financial weight. They found the World Bank to be a standout performer that punches above its weight, efficiently converting its loans into greater-than-expected development policy influence. By comparison, they found that bilateral heavyweights like the United States, the United Kingdom, Japan, and Germany punch at or below their financial weight. This pattern suggests that taxpayers in wealthy countries may obtain better value-for-money by channeling their money through the World Bank than through their own bilateral institutions.
AidData’s report also explored several potential reasons why some donors exert outsized influence on the policy priorities of partner countries. It found that alignment of sector aid allocations with partner country priorities (identified via citizen surveys) is a particularly strong predictor of development policy influence. This finding is relevant to the World Bank because it ranks much higher than most other multilateral and bilateral development institutions on alignment with country priorities (see Figure 1).
Figure 1: Alignment with Partner Country Priorities and Agenda-Setting Influence (Reproduced from Custer et al. 2015)
Note: Agenda-setting influence is measured on a scale of 0 to 5, where 0 indicates "No influence at all" and 5 indicates "Maximum influence." Alignment is equal to a z-score based on the share of ODA allocated to partner countries’ top development priorities (Birdsall and Kharas 2014).
With new research funding we recently began to delve deeper into these rich micro-data to better understand the Bank’s sectoral areas of comparative advantage and disadvantage vis-à-vis other bilateral and multilateral development partners. We found that, compared to other bilateral and multilateral development partners as a group, the Bank exerts significantly greater agenda-setting influence in 16 out of 23 policy domains evaluated in the 2014 Reform Efforts Survey. Figure 2 depicts the Bank’s relative influence in each policy area, ranked from most to least influential. Energy and mining, infrastructure, public expenditure management, and macroeconomic management are among the various policy domains where the Bank receives uniquely high marks from in-country decision-makers—at the priority-setting stage of the reform process. In the other 7 policy domains (justice and security, customs, land, tax, decentralization, democracy, and family and gender) that were examined in the survey, the Bank's agenda-setting influence is roughly on par with that of other bilateral and multilateral development institutions.
Notice that, according to government officials in the Bank’s client countries, the policy influence of the Bank varies substantially not only across all 23 policy domains, but also within areas that fall within the mandates of the Bank's individual networks or Global Practices (GPs). Nearly one-third of these policy domains pertain to the Governance GP’s work, with public expenditure management ranked very highly (4th of 23), public administration in the middle (11th), and justice and security at the bottom.
Figure 2: Average Agenda-Setting Influence by Policy Domain, the World Bank Compared to All Other Multilateral and Bilateral Development Partners (as a Group)
Going forward, AidData and World Bank DEC researchers hope to leverage the rich micro-data from the 2014 Reform Efforts Survey to better account for these empirical patterns and thus learn how the Bank can most effectively assist and advise public sector decision-makers. Potential channels of policy influence at the country level include investment lending, development policy operations (and associated prior actions), economic and sector work (ESW), technical assitance (TA), and informal policy dialogue. To the extent these and other activities are measurable, analysis of the data may tell us something about the relative payoffs to different interventions—in terms of policy influence. We also anticipate that a new set of opportunities for organizational learning at the Bank will arise as AidData completes future rounds of its omnibus survey. The second round of its survey closed early this month, so new insights (including those that leverage the panel dimension of this survey) are already within reach.