The World Bank JANUARY } E|| !0 _notes NUMBER 35 ECONOMIC POLICY Conditionality revisited: 21 585 A new approach in Burkina Faso A simulation exercise for Burkina Faso alters the usual approach to conditionality in adjustment lending by linking quick-disbursing assistance to outcomes in health, education, and public finance management. This approach enhances govemment ownership of economic reforms. Can focusing on Donorshave typicallylinked quick-disbursing Why a new approach is needed the results of assistance programs to conditions related to Criticism of poor country ownership of policy reforms. But there is growing frus- reforms, the damaging effects of "stop-go" policy reforms tration with how conditionality has worked financing, and the often disappointing in practice. Research has shown that tradi- results ofadjustmentsupportarewell known. make aid tional conditionality usually does not "buy" The weaknesses in the traditional approach reforms. Moreover, because aid funds are to conditional aid described below have also more effective? fungible within government budgets, a been observed in many other borrower coun- results-based approach is needed to enhance tries, and are by no means unique to Burk- aid effectiveness. At the initiative of the Euro- ina Faso. pean Commission, the Strategic Partnership Burkina Faso has been receiving adjust- withAfricaispilotinganewapproachtocon- ment support from the Bank, IMF, and ditionality in Burkina Faso to explore changes other donors since 1991. Yet as late as 1997 that enhance donor coordination, foster only a small group of officials at the Min- country ownership of reforms, and smooth istry of Economy and Finance was famil- aid flows-and so increase aid effectiveness. iar with the reforms, even though most of A driving feature of the approach is a focus the program actually affected other min- on the outcomes of government reforms and istries. Even those aware of the reforms policies. appeared to have little understanding of Beside the Burkina Faso government, the why they were being implemented, beyond pilot involves Austria, Belgium, Canada, Den- being demanded by donors. Disbursement mark, France, Germany,Japan, the Nether- of support was conditional on the legal lands, Switzerland, the World Bank, the aspects of adopting and implementing the International Monetary Fund (IMF), the reforms, not the results. Thus there was lit- European Commission, and the United tle monitoring of actual results with data Nations Development Programme (UNDP). and analysis. Four joint donor missions took place in The thirdjoint donor mission showed 1998-99. Because the exercise is a simula- that even when data were available, tion, it does not have financial implications neither donors nor the government paid for support programs already in place. It has much attention, because support was not constructed links between disbursements contingent on data. The numerous and outcomes and will simulate the effects donors multiplied demands on govern- of such links during the final stage. ment, usually for the same information, FROM THE DEVELOPMENT ECONOMICS VICE PRESIDENCY AND POVERTY REDUCTION AND ECONOMIC MANAGEMENT NETWORK monopolizing scarce domestic capacity Lessons while at times imposing conflicting Thepilotgeneratedfiveimportantlessons. conditionalities. Outcome indicators shift thefocus to Methodology of the pilot results At the outset the donors agreed to use per- The pilot has shown donors and govern- formance results indicators as a key element ments the value of outcome indicators, which of the revised approach and to consider link- shift the focus of reforms to what really mat- ing allocations and disbursements to results. ters: results. This shift changes the nature But it was clear that the design of such a link of the government-donor dialogue and Outcome indicators would need to be based on lessons from the makes for more coherent and realistic objec- early missions. This proved to be a sound tives and interventions. The pilot has demon- shift the focus approach. strated that it is possible to document The government was asked to propose outcome indicators and to measure the of reforms to amatrixofperformance indicators prior to impact of reforms in this manner. the first mission. This matrix consisted For a country that has been consistently what really mainly of input and output indicators but "on track" for the better part of the past was refined over subsequent missions to decade, the use of outcome indicators matters: results focus on performance indicators for bud- revealed some startling facts. For instance, get management and outcome indicators despite 10 years of increasing sectoral bud- for social sectors (box 1). The indicators getallocationsanddonorsupportaswellas chosen relate to outcomes such as vacci- new cost recovery schemes in the health sec- nation rates and books per pupil because tor, attendance rates at health centers have indicators of impact-maternal mortality, steadily decreased. Surveys found that only literacy rates-are subject to factors beyond a small percentage of budget allocations the government's control. reached the decentralized level. In addition, Between missions the government, the private sector complained of grave prob- working with donor representatives, col- lems relating to the judicial system and to lected existing data, carried out surveys governance. Results in the education sector for new data, analyzed results, and in a few were more encouraging yet remained far cases drew policy conclusions from the from what is needed to achieve the gov- analysis. ernment's long-term goals. Box 1 KEY OUTCOME INDICATORS SELECTED FOR THE PILOT Health * Success rate at primary school leaving exam- * Attendance rates at health centers: number of inations new consultations per capita * Number of books per pupil * Number of caesarian sections performed (a * Level of user satisfaction predictor of maternal mortality) * Costs of schooling (publicly and privately * Essential and basic vaccination rates provided) * Number of second prenatal consultations * Level of user satisfaction Publicfinance management * Costs of basic services (publicly and privately * Coherence in budget allocations and sectoral provided) policy objectives * Implementation of the budget Education * Share of the budget going to peripheral * Gross enrollment rates (boys and girls, pub- services lic and private) * Differences between unit costs in public sec- * Gross enrollment rates in first year of primary tor procurement and market prices school (boys and girls) Source: Adapted from government statistics department; rapid surveys. PREMNOTE 35 JANUARY 2000 Partnerships mean shared responsibilities the participants agreed that disbursements The government rightly pointed out that related to budget management and sectoral donors shared responsibility for the poor performance should be based on results and results of policies they had imposed as con- graduated according to the degree of ditions for support. This raised the question achievement. This agreement suggests that ofwhether the choice ofpolicies (and imple- it should be possible to achieve the objec- mentation indicators) is a sovereignty issue. tive of smoothing aid flows. If it is seen as such, a new approach will neces- sitate a shift in donor mentality from Linking disbursements demanding reforms to giving advice. to outcomes Models for linking disbursements to results Donor Ownership is a good start have been constructed, and the final mis- The pilot has also shown that the results- sion (inJune 2000) will simulate how they coordination based approach enhances ownership. There would be applied to the results achieved was increasing participation at the highest in 1999. The purpose behind the simulation has freed scarce political levels as the pilot progressed, and is not to identify a single disbursement link everyone agreed that the exercise changed that should always be used. Rather, the sim- government thinking within ministries as the focus on ulation will highlight issues and principles results grew. The Poverty Reduction Strat- to keep in mind when constructing links. capacity egyPaperapproach, recentlyintroducedby In the chosen models a share of donor the World Bank and the IMF, reaffirms the support would remain linked to a stable pilot's focus on ownership because it asks the macroeconomic framework as determined recipients of concessional lending or debt by the IMF. The rest of the support would relief to devise their own poverty reduction depend on results in budget management strategies using a participatory approach. and in the social sectors (health and edu- cation). For social sector support the gov- Broadly based stakeholder involvement ernment will set public targets for the remains difficult selected indicators. Depending on the indi- The government was initially reluctant to cator, these targets can be a specific num- involve civil society and publicize the out- ber or a quantified or nonquantified come targets. But progress has been made- improvement over a baseline situation. Expe- private sector representatives were active rience has shown that it is important to look participants in the third mission, informa- at trends in indicators rather than absolute tion meetings have been held with a par- levels. The level of allocation and dis- liamentary committee, and so on. The bursement will vary depending on the government is committed to including par- degree to which a target is achieved. liamentarians and other civil society rep- Interpretation of whether a target has resentatives in the final mission. been achieved is ajoint exercise between all the donors and the government. The Donor coordination is worth the logistical assessment is flexible enough to distinguish difficulties between "target not achieved but some Finally, the pilot has shown the logistical dif- progress noted" and "no progress achieved." ficulty but considerable value ofjoint donor An indicator can also be "neutralized" if missions. Donors have been able to draw on the (lack of) results for that indicator depend each other's sector expertise and to arrive on factors beyond the government's control. at a common understanding on the direc- Different indicators or groups of indicators- tion and mechanisms to follow in the pilot. for example, on budget management-can The government has particularly appreci- be given more weight in terms of their impact ated thejoint approach, because it has freed on allocation and disbursement decisions. scarce capacity and ensured coherence in Finally, the simulation will look at the advan- relations with various donors. As noted, all tages of making a direct disbursement link PREMNOTE 35 JANUARY 2000 (affecting disbursements in the same year as * How can stakeholders be involved in the the evaluation) or an indirect one (affecting design and implementation of reforms? allocations for the following budget exer- cise). This approach will influence the pre- Driving feature of the pilot dictability of donor aid flows. The new approach had a defining feature: Surveys have been carried out to mea- it took explicit account of the results of the sure user satisfaction, user costs, the share reforms implemented in disbursement deci- of the budget reaching peripheral levels, sions through the use of outcome indica- and so on. The surveys, which have been tors. Using outcome indicators helps financed by the donors, will continue and increase ownership, improve donor coor- their methodologies will be refined. dination, and smooth aid flows-that is, they If the government The simulation includes few if any condi- enhance the effectiveness of aid. The pilot tions on how the results are to be achieved. shows that these outcomes are possible is to be judged If the government is to be judged on results, despite the methodological difficulties and it should be free to deternine the scope and the mentality shift required of both donors on results, it timing of policy changeswithout microman- and governm ent. agement fromn donors. In areas where the gov- should be free ernment's policymaking capacity requires This note was written by Stefan Emblad (.te- strengtheninig, donors remain committed to fan.emblad@cec.eu.int), Economist, and Gilles to determine providing advice if the government seeks it. Hervio (gilles. herio@cec.eu. int), EconomicAdvi- sor, Directorate GeneralforDevelopment, En ro- policy changes Issues to be addressed pean Commission. Chuck Humphreys (Sector Several issties need to be addressed dur- Managei; Macroeconomics Technical Family ing the final mission: 4, Africa Region) contributed to the note. If * What suppoi-t is needed to ensure reli- you are interested in progress under the pilot, able data for indicators? please contact either the authors of the note at the * How will data be validated by donors? European Commission, or Celestin Mong-a = How will factors beyond the govermment's (x35125) or Marion Eeckhout (x390 75) at the control be taken into account? Bank. * How can allocations and disbursements Ifyou are interested in similar topics, consitder be linked to results and make aid flows joining the Country Strategy and Policy-Based more predictable? Lending Thematic Group. Contact Victoria Elliott * How much freedom are donors ready to (x80535) or Sudhir Shetty (x81939), or click give Burkina Faso in choosing a strategy? on Thematic Groups on PREMnet. This note series is intended to summarize good practice and key policy find- I ings on PREM-related topics. The views expressed in these notes are those of the authors and do not necessarily reflect the views of the World Bank. PREM- rnotes are distributed widely to Bank staff and are also available on the PREM Xwebsite (http://prem). If you are interested in writing a PREMnote, email your P vein dlek in eed Eidea to Sarah Nedolast. For additional copies of this PREMnote please con- ftactthePREMAdvisor Service at x87736. Prepared for World Bank staff