Bureau for Policy and Program Coordination August 2004 PN-ACW-878 PPC EVALUATION WORKING PAPER NO. 18 General Budget Support: An Alternative Assistance Approach Mozambique Country Case Study General Budget Support: An Alternative Assistance Approach Mozambique Country Case Study Evaluation Team Members Joseph Lieberson, USAID Economist and Team Leader Diane Ray, Economist, Academy for Educational Development Maxine Lunn, Program Analyst, Academy for Educational Development Bureau for Policy and Program Coordination August 2004 General Budget Support: An Alternative Assistance Approach iii Contents General Budget Support: An Alternative Assistance Approach 1 Summary 1 Background 3 Mozambique’s Political and Economic Overview 7 How General Budget Support Relates to Mozambique’s Development Approach 9 USAID’s Approach to Development 11 Finance and Fiduciary Capabilities 13 Development Management 17 Annex 1. Mozambique’s Sectoral Approaches and the Role of USAID and Other Donors 21 Annex 2. Interview Contact List 27 Annex 3. Bibliography 29 Annex 4. Alternative Assistance Approaches 33 Annex 5. USAID’s Experience with Non-project Assistance 35 Annex 6. General Budget Support Interview Protocols 39 General Budget Support: An Alternative Assistance Approach 1 Donors worldwide employ a wide range of approaches to foreign aid, under the cate￾gories of “project assistance” and “non-proj￾ect assistance.” These donors generally acknowledge that the projects they manage succeed, but once they leave, the projects tend to fail in part owing to a lack of host country ownership. Viewed in this light, donor efforts actively seek ways to encourage host country participation and ownership, and some donors believe that General Budget Support (GBS) offers a workable alternative. GBS is one type of non-project assistance which as a development tool has several applications. Other kinds of non-project assistance or “program aid” include food aid, bal￾ance of payments support, commodity import pro￾grams, sector assistance, and debt relief. While most non-project assistance approaches depend on some degree of earmarking or policy conditionality, GBS relies on broad agreements and an acknowledgment that appropriate development policies are in place. Funds usually flow directly to the finance ministry. If a recipient government demonstrates its commitment to broad development goals and donors agree, GBS donors participate in a planning dialogue with central government officials and pool their aid resources to support the govern￾ment’s development efforts. In applying this approach, the host country finance ministry allo￾cates donor financing and uses its own procurement and accounting systems, which some donors argue is a manifestation of country “ownership” that can lead to sustainable development. In Mozambique, a large group of donors has decided to move away from projects in favor of GBS, provid￾ing a good case study to examine the conditions under which this approach might be recommended. Donors in Mozambique have a long history of coor￾dination, beginning with their humanitarian efforts during and after the civil war. Currently, donors coor￾dinate through a number of mechanisms, including a group comprised of all of the GBS donors plus the United States as an observer and donor sub-commit￾tees. A GBS program in Mozambique has particular appeal to some donors and has garnered their support because of the country’s well-designed development and poverty reduction plan. In Mozambique there are many different donors with a variety of views about GBS. The proportion donor assistance provided using General Budget Support ranges from 10 to 50 percent. Donors also provide sector support as well as through traditional projects, technical assistance and support to the non￾profit and private sectors. While GBS may contribute to sustainable develop￾ment, there are a number of risks inherent in its use. If government institutions and management capabil￾ities are weak, development suffers. Mozambique has established good planning mechanisms, but still has extremely weak accounting and financial control sys￾tems. Public expenditures management is weak, and there is considerable fiduciary risk. These weaknesses contribute to a significant risk of inappropriate spending or outright theft. The shift to GBS is relatively recent and there are some donors who prefer to wait until more experi￾ence is gained in its use. Though it is too soon to know whether improved government performance does result from GBS, this assessment examines the factors necessary for GBS to work successfully in Mozambique. Using the lessons learned from Mozambique, it is possible to identify key factors that should be examined in other countries where donors are considering General Budget Support. These lessons and recommendations are summa￾rized in the Evaluation Brief: “The Conditions That Make General Budget Support Most Effective: Mozambique Country Case Study” (PN-ACU-999). Summary Depending on development constraints and host governments’ institutional capacity to implement development programs, USAID and other donors may choose from a variety of assis￾tance approaches to suit a developing country’s spe￾cial conditions and needs. Brief descriptions follow of different types of interventions that USAID gen￾erally characterizes as non-project assistance. These include General Budget Support (GBS), a recently developed form of non-project assistance. (See Annexes 4 and 5 for a detailed discussion on the pros and cons of different types of assistance.) Balance of Payment Support USAID and other donors may provide resource transfers either in foreign exchange as a cash transfer or as a commodity import program. Balance of pay￾ment support is a modality designed to promote economic and political stability. It is short- and medium-term, and addresses immediate gaps in the country’s balance of payments. Sector Program Assistance Program assistance focuses on sector constraints to sustainable growth and usually refers to a broad area such as agriculture, industry, education, health, exports. It may also refer to a sub-sector such as agricultural marketing, health care financing or child survival. It supports the implementation of reforms and other actions to overcome sector based development constraints. Sector program assistance relies on disbursements based on policy conditional￾ity in separate donor agreements. Sector Wide Approach (SWAp) A SWAp is defined as multi-donor funding for a sector in support of single sector wide policy and expenditure program. A SWAp coordinates all donor assistance in support of a common sector expenditure program that includes a variety of inter￾ventions: budget support, project aid and technical assistance, which may or may not be earmarked to specific expenditures or disbursed through the gov￾ernment’s own budget process. General Budget Support This type of assistance is similar to balance of pay￾ment support since it supports structural adjustment and policy reforms and contains a local currency counterpart to foreign exchange. However, rather than using traditional conditionality, donors agree on an agenda of policy measures which the govern￾ment plans to implement. Once donors are com￾fortable on overall budget priorities, as set out in a medium term budget and expenditure framework, GBS can be provided. With agreement on the budget as a whole, there is no need to earmark aid flows to specific country expenditures. Funds are channeled through the host government’s national budget. Since support is for the budget as a whole, accountability depends on government audited accounts of its total revenues and expenditures. Why the Interest in General Budget Support? Foreign aid is usually provided through donor￾managed projects, though some development practitioners question the success of this approach, noting that if development is to be successful and self-sustaining, the developing country has to have ownership of the development process. They say donors should provide aid funds to host govern￾ments and provide them the space to manage their own development programs. This evaluation identifies the country conditions that led some donors to allocate a portion of their General Budget Support: An Alternative Assistance Approach 3 Background 4 PPC Evaluation Working Paper No. 18 foreign assistance to General Budget Support.1 It examines the benefits and drawbacks of General Budget Support and the country conditions that underlie a successful program. The Case for General Budget Support If a donor runs its own aid project, host govern￾ment ownership tends to be minimal, and that may explain why many donor projects fail. Donor-man￾aged aid projects get the resources they need, and donors have a stake in the success of their projects. But a donor project can be an “island of develop￾ment” that only reaches a small group of people. It often fails to strengthen local institutions or change institutional relationships, is not easily replicated, and benefits are not sustainable. General Budget Support, as some donors argue, is more consistent with the concept of building strong host-country ownership that may lead to a greater chance for sustainable development. Donors, the host govern￾ment, and other country actors work in close part￾nership, as they reach a shared agreement on pro￾gram and budget priorities. Donors then provide GBS to the finance ministry. Funds are not tied to any sector or project but support the overall devel￾opment budget effort. Donors assume that with a change in the process of delivering aid (from projects to budget support), governments will perform effectively in reducing poverty. Since the government is responsible for the program, it will identify the most critical develop￾ment problems and make sure the problems are solved. Donors also assume that a developing coun￾try has the desire and ability to implement success￾ful development programs. They expect that the key benefits of General Budget Support will be improved government efficiency and ownership since GBS will: ■ Improve coordination among donors and har￾monize and align aid with the government’s budget and policy system. ■ Use policy dialogue to reform overall govern￾ment policies and budget priorities, rather than narrowly concentrating on individual projects or sectors. ■ Increase efficiency in public spending. The gov￾ernment directs its own budget allocations, instead of dealing with a large number of donor projects, which often reflect donor priorities rather than those of the developing country. ■ Empower the government to administer its pro￾grams effectively. ■ Make the government accountable to the people of its own country, and not just to the donors. ■ Reduce transaction costs for donors in the long￾term since donors do not have to manage a large number of projects. ■ Reduce transaction costs to the recipient since it does not have to deal with a multitude of donor projects and reporting requirements. The Problems with General Budget Support In some countries where good governance and insti￾tutional capability exist, GBS makes good sense. It is less so in low-income developing countries that tend to have weak institutions and inadequate man￾agement and technical skills. Many donors have serious questions about making large cash transfers to countries that lack the essential capabilities to account adequately for funds. Specific donor con￾cerns center on: ■ Weak management and financial systems and a lack of budget discipline. Many governments typically have weak audit, monitoring and eval￾uation agencies. ■ Corruption and misappropriation of funds. ■ A lack of skilled managers and technicians. ■ A finance ministry that often falls short when it comes to managing an efficient and transparent 1 The U.K. Department for International Development defines General Budget Support as a form of non-project aid where: 1. Funds are provided in support of a government program that focuses on growth and poverty reduction, and transforming institutions, especial￾ly budgetary institutions. 2. The funds are provided to a partner government to spend using its own financial management and accountability systems. http://www.grc-exchange.org/g_themes /cc_budgetsupport.html General Budget Support: An Alternative Assistance Approach 5 budget allocation process. GBS strengthens the hand of the finance ministry. ■ A strong central government that has little inter￾est in empowering civil society and the private sector. Donor funding to the central government could exacerbate this problem. ■ The potential or tendency for General Budget Support to encourage centralization that may harm local, grassroots development—which many view as essential to sustainable develop￾ment. ■ The chances for success at pressing for policy reforms with a line ministry directly (such as agriculture or health). ■ While some transaction costs may decrease, others may increase. For example, GBS requires improved policy and program design and management and months of analytical effort by highly trained staff from both the donor and government. Political History Mozambique is a nation of 17.6 million people, with a land mass twice the size of California. Located on the southeast coast of Africa, and bordering six countries including South Africa, Mozambique is well-placed for trade and commerce. The capital, Maputo, is located in the south. The country became a formal colony of Portugal in 1890. In early 1962 Mozambican nationalist groups united, forming FRELIMO, the Front for the Liberation of Mozambique, and a war for independ￾ence started in 1964. The Portuguese turned over power to FRELIMO in 1974 after a ten-year war. Full independence came on June 25, 1975 with the formation of the People’s Republic of Mozambique. FRELIMO declared the country a Marxist-Leninist state with a centrally planned economy and one￾party rule, spawning opposition led by a group known as the Mozambique National Resistance￾Electoral Union or RENAMO. A devastating civil war resulted. By 1984 real per capita GDP had fall￾en by half compared with pre-independence levels, and half the population was internally displaced or in refugee camps. By October 4, 1992, FRELIMO and RENAMO signed a peace agreement. Country Profile Since the civil war ended in 1992, Mozambique’s growth has averaged 8 percent a year, with recent years even higher at 14.8 percent in 2001 and 10 percent in 2002. Inflation is back down to the sin￾gle digits, 9 percent in 2002, after a spike to 22 per￾cent in 2001. The currency is stable. Success, how￾ever, is fragile. Despite economic gains, floods in 2000 posed threats to sustained growth as does the current HIV/AIDS crisis. Eighty percent of Mozambicans live in rural areas where agriculture employs a large portion of the population. Seventy percent of Mozambique’s popu￾lation lives at or below the absolute poverty level. Approximately 44 percent of children suffer from chronic malnutrition and 6 percent from acute mal￾nutrition. Per capita income is approximately $226 with regional variations. About 48 percent of the population over the age of 15 can read and write, and less than 20 percent of primary school enrollees reach the secondary level. Life expectancy for men is low at 40 years and even lower for women at 38 years. Only 40 percent of Mozambicans have access to healthcare facilities. Consequently, maternal, infant and child mortality rates are high. Mozambique’s largest trading partner is South Africa. Growth in trade is expected to continue due to recent investments in an aluminum smelter and a gas pipeline. However, to promote trade, infrastruc￾ture across the board—road, rail, ports, airports, and telecommunication—needs strengthening. Mozambique is a fledgling democracy with a weak civil society. Democratic processes are in place, but the executive branch has a disproportionate amount of power and FRELIMO dominates the country’s leadership. Economic and democratic freedoms need to be strengthened if the government is to truly represent the needs of the people. There is widespread corruption that potentially threatens Mozambique’s development gains. A lack of legislative or judicial strength along with limited civil society means that there are few ways to con￾trol corruption. Archaic and overly complicated reg￾ulatory systems need reform, to minimize the pen￾chant for corruption. General Budget Support: An Alternative Assistance Approach 7 Mozambique’s Political and Economic Overview 8 PPC Evaluation Working Paper No. 18 Donor Support Donors fund approximately half of the government’s total budget, which includes the entire investment development portion. At the October 2003 Consultative Group meeting donors pledged $790 million. With massive donor funding, the country may soon face “Dutch Disease” whereby a large inflow of foreign currency disrupts prices and mar￾kets, generates inflation, causes the local currency to appreciate and decreases export competitiveness. Donors employ budget support and project inter￾ventions as part of their financing programs. Their budget support follows two main approaches: sector specific assistance such as to agriculture, health, edu￾cation or General Budget Support to the central finance ministry which decides resource allocation to various sectors in accordance with the World Bank’s motivated Poverty Reduction Strategy Paper (known as the “PARPA” in its Portuguese acronym) that states the country’s development priorities. General Budget Support is growing, both in the number of donors participating and in the amount that they contribute. The 2003 mix of donor sup￾port is provided in Table 1 below. In certain instances, donor project financing can achieve host country ownership and avoid duplica￾tion, if the host country is involved in funding deci￾sions. In certain instances, such as with construction projects, the donor’s interests can be preserved with the employment of supervisory engineers. USAID has had success with this approach in Mozambique, working with the Ministry of Public Works and Housing to construct buildings or roads. General Budget Sector Project/NGO/ Donor Support Assistance Private Sector Netherlands 50% 25% 25% European Commission 34% 19% 47% Great Britain 33% 45% 22% World Bank 33% 0% 66% Switzerland 30% n.a. n.a. Norway 25% 25% 50% Denmark 20% n.a. n.a. Sweden 17% 57% 26% Germany 17% 30% 53% Finland 15% 85% 0% France 10% 10% 80% Italy 10% 10% 80% Portugal 0% 10% 90% United States 0% 15% 85% Japan 0% 0% 100% Table 1. Aid Approaches Used in Mozambique by Selected Donors Percentage of Total Aid in 2003 n.a.= data not available Source: Discussion at G-11 Economists Weekly Meeting, September 24, 2003, Maputo, Mozambique, supplemented by interviews with individual donors. Mozambique’s Development Approach I n 1990, Mozambique adopted a poverty reduc￾tion strategy or PARPA as part of its national development plan. It updated the strategy in 1999 and titled it the “2000–2004 Action Plan for the Reduction of Absolute Poverty.” The World Bank and IMF have approved the strategy, as it meets the criteria for Poverty Reduction Strategy Papers (PRSPs) that are geared to host governments’ poverty reduction strategies. The PARPA’s goal is to reduce absolute poverty from 70 percent in 1997 to less than 60 percent by 2005. The PARPA emphasizes economic growth, invest￾ments in health and education and rural infrastruc￾ture in reducing poverty. In the agricultural sector, it focuses on productivity, market access, and expan￾sion of extension services, inputs and credit. In the health sector HIV/AIDS awareness and assistance is given prominence. However, little mention is made of measures to improve governance or to strengthen government service management. For Monitoring and Evaluation, the PARPA is the basis for the Government’s Medium-Term Expenditure Framework. In addition, donors review the annual budget, along with the Annual Economic and Social Plan (PES), within the param￾eters of the Mozambique government’s Performance Assessment Framework (PAF). The PAF contains indicators and targets to measure progress. The Role of Donors and the Role of General Budget Support Donors who provide General Budget Support have formed the G-11 Group2 who meet periodically or weekly, as a whole, in committees or informally. Since USAID does not provide General Budget Support, it has only observer status at the G-11 meetings. The Mozambican government and the G￾11 meet jointly in May each year with 4 days of intensive technical and policy discussions. Through this review process, the G-11 has successfully used General Budget Support as a way to frame policy dialogues that have apparently resulted in better planning, budgeting and country ownership. Donors expect that policy discussions will lead to improve￾ments such as reduced defense spending, elimination of phantom civil service employees and better health and education service delivery to the poor. Donors who provide General Budget Support nego￾tiate policy and budget issues in unison and achieve an impact that no single donor has. The fact that the Mozambique government recognizes donors’ concerns and is willing to negotiate with them makes the policy dialogue process succeed. A check with the G-11 donors reveals that they are very sat￾isfied with a highly effective process in achieving policy reforms. This might explain why Italy, Germany and Portugal are adding budget support to their program as a way to gain a seat at the policy table. USAID has observer status since it does not provide GBS. USAID is a participant owing to its development knowledge and technical skills, but these attributes alone do not translate into influence on the policy debates in the G-11 forum. General Budget Support: An Alternative Assistance Approach 9 How General Budget Support Relates to Mozambique’s Development Approach 2 The G-11 started with 11 donors but now totals 14 and soon will call itself the G-14. It includes: Belgium, The European Commission, Denmark, Ireland, Finland, Netherlands, France, Norway, Sweden, Switzerland and the UK. Recent additions include Germany, Italy and Portugal 10 PPC Evaluation Working Paper No. 18 No single donor provides 100 percent of its assis￾tance to General Budget Support. GBS ranges from a minimum of 10 percent to a maximum of 50 per￾cent of individual members’ total aid. Several donors expect to increase the proportion of their assistance devoted to General Budget Support in the near future, and some strive for what they consider an ideal mix—a 50/50 balance between General Budget Support and other types of aid. Still others, particularly the EU and the British, claim that budget support is their preferred mode, and they hope to move all of their aid to General Budget Support over the long term. Those who argue for a balance say that they are: a) minimizing their risks by not shifting entirely to General Budget Support; b) addressing a need for projects that assist with reforms in the planning and execution of the budget; c) addressing critical needs with the sector min￾istries, particularly in health and education; d) pushing additional funding out to the more remote and poorer provinces; and e) strengthening civil society and the private sector. One unexpected result of GBS has been that gov￾ernment ministries now face a stiff internal debate on their development approach. Previously each ministry dealt directly with the donors and cut its own deal with a donor, often outside the govern￾ment’s budget. Now line ministries are compelled to articulate and defend the logic of their objectives and their approaches to the Ministry of Planning and Finance and other ministries. A healthy internal government debate can build a better overall pro￾gram and promote stronger ownership among all ministries. USAID’s Country Strategic Plan I n 2003, USAID approved a new Country Strategic Plan for assistance to Mozambique. The prior plan (1996–2003) operated in an environment of post-conflict reconstruction and emergency response to catastrophic flooding. The new plan, for fiscal years 2004-2010, was developed in the context of the country’s current high growth, post-drought environment. It also acknowledges the strong degree of coordination within a large donor community and generally positive international assessments of the country’s progress and capacity. Program Structure and Areas of Emphasis USAID’s Country Strategic Plan is projected at about $45 million per year over the next seven years, although additional funding is expected through Presidential Initiatives on hunger, trade and HIV/AIDS. Much of the additional funding will be earmarked for specific expenses, such as anti￾retroviral drugs. As one of the larger bilateral donors, USAID emphasizes local priorities, coordination with other donors, programs with Mozambican partners and impact on poorer and rural populations. USAID’s strategic plans aligns its priorities with those of the Mozambican government, including: increasing rural incomes, promoting labor-intensive manufacturing, expanding child survival and repro￾ductive health services, curbing HIV/AIDS and improving democratic governance at the local level. In addition, gender equity, human capacity develop￾ment and anti-corruption support USAID’s overall objectives. Project assistance is USAID’s primary means for delivering aid, though the mission also participates in a multi-donor, agricultural SWAp known as PROAGRI that provides the means to increase incomes in rural areas. Increasing rural incomes (see chart above) is USAID’s highest development priori￾ty. USAID agricultural funds are allocated for tech￾nical assistance to the agricultural ministry and for private sector and NGO partners and activities. Though USAID has not been formally a member of the G-11, its past experience in developing PROA￾GRI I and II provide useful lessons that inform pro￾gramming decisions for future investments in General Budget Support. Donors agree that plan￾ning and implementation responsibilities for agri￾culture have been successfully transferred to the General Budget Support: An Alternative Assistance Approach 11 USAID’s Approach to Development USAID Country Strategic Plan for Mozambique Proposed Funding by SO (2004–2010) Child survival 18% HIV/AIDS 20% Municipal governance 4% Rural income 35% Labor intensive manufacturing 23% 12 PPC Evaluation Working Paper No. 18 Ministry of Agriculture and Rural Development and agree that the ministry still needs ongoing technical assistance to ensure that priorities are translated into measurable action. Some donors have decided to shift future PROAGRI funding to General Budget Support with the understanding that PROAGRI will receive the funds. Competitiveness Though Mozambique has successfully completed major economic reforms such as liberalizing its for￾eign exchange market and eliminating price and trade controls, it has an excessive system of regula￾tions that harm economic growth and encourage corruption. Areas that need reforms include: ■ The formal sector which faces a gauntlet of diffi￾culties: lengthy and tedious company registra￾tion; restrictive labor laws and arcane tax and customs laws. With excessive and unclear con￾trols, the regulators routinely demand bribes to clear away obstacles. ■ The informal business sector that lacks access to finance, market services and contract enforce￾ment. Firms stay small since they have difficulty expanding their product line or markets. The informal sector is reluctant to move into the for￾mal sector since that means onerous regulations and government officials demanding bribes. ■ Land registration that could benefit from an honest cadastral survey that would give farmers clear land ownership and an incentive to improve productivity. The government would have to change its policies to allow convertibility of land into money. That would mean that land could be sold or used as collateral for invest￾ments to increase productivity. ■ Trade capacity building and trade policy formula￾tion that is hampered by ingrained protection￾ism, excessive government red tape, slow cus￾toms clearance and corruption. The lack of needed infrastructure (roads, bridges, rail, ports, and telecommunications) also harms trade. GBS may not be the way to solve Mozambique’s competitiveness problems, since it requires a joint donor approach. It is always difficult to get all 15 donors to agree on GBS reforms, and donors have different philosophies on the ideal level of govern￾ment regulation for the private sector. Many donors see GBS as a way to support poverty reduction efforts through government health and education programs. They do not focus as much on private sector development. USAID might better deal with private sector issues with its own bilateral program, using technical assistance and policy reform cash transfers. There are also other parts of the USAID Mozambique program that are a poor fit with GBS. USAID’s objectives relating to labor-intensive man￾ufacturing, municipal governance, and the health sector do not lend themselves to GBS. Manufacturing has a private-sector focus and the Mozambique government strongly regulates the pri￾vate sector, particularly private sector labor. This regulation mentality can be better addressed and changed through policy reform, technical assistance and training with performance-based conditionality rather than GBS. Improved municipal governance is also ill suited to GBS. In Mozambique, as in many countries, the government is not eager to encourage a participatory and democratic competing power at the local level which could harm the central govern￾ment. And finally, the U.S. Congress requires health funds to be provided only through projects. That legislation blocks cash transfers to support health sector policy reforms. Public Expenditure Management Areview of Mozambique’s public expenditure management performance was released by the World Bank in December of 2001 and a second one in late 2003. The reviews identified a number of problems and included recommenda￾tions for budget formulation, budget execution, and budget evaluation and auditing. The IMF is working with Mozambique to address issues raised in the review. A major problem for the government and donors is that Mozambique’s budget gives only a partial view of revenue and expenditures and may be a key rea￾son why GBS as currently practiced falls short of achieving broad budgetary reforms. In this context, it is not clear if resource allocations are properly aligned with development needs. Furthermore, the budget preparation process is not fully transparent, and details below sub-headings are not firm. The World Bank reviews call for more information in the budget on the government’s own revenue sources, donor funded activities and tax expendi￾tures. It also suggests using the Medium Term Expenditure Framework as an instrument of budget planning and to make it available to the public. It urges that macroeconomic stability should be an explicit objective of the budget formulation process. Additionally, to make the process more responsive to the needs of the country, the capacity of the National Assembly to analyze the budget and moni￾tor its execution needs to be improved. Budget execution suffers from poor accounting, par￾tial reporting and weak cash management. Many donor projects are “off-budget” and need to be included in the government budget. In public accounting, the single-entry system is being replaced with double-entry bookkeeping; the cash basis accounting system should be replaced with modified accrual accounting; and a new chart of accounts should be implemented. The reports currently used need better data on allocations, adjustments and expenditures and more information on donor-fund￾ed expenditures. Reports on debt, lending and on￾lending, cash flows and tax expenditures need to be created. To improve cash management, the 12,000 government commercial bank accounts need to be consolidated into one treasury account and a new automated treasury payments system should be introduced. Financial planning should be improved, and revenue collection needs to be rationalized. Both internal and external budget evaluations and audits are weak. Internal audits should be used as a management tool and should primarily be the responsibility of the spending units. The Ministry of Planning and Finance should oversee the quality of internal audits. For external audits, the Administrative Tribunal needs the flexibility to build its capacity in financial, human and operational means. Budget evaluations (to assess benefits against costs), are not a current priority. Such evaluations would improve budget management and should be a medium-term objective. Responsibility for the evaluations should fall to the spending agency, under the coordination of the Ministry of Planning and Finance. Assessing Fiduciary Risk There are differences among donors in their approaches and tolerances to fiduciary risk, but all are concerned about possible misuse of funds and corruption. Donors are designing ways to monitor fiduciary risk through an annual joint assessment. They advocate working with the government while systems to minimize risks improve. Some r donors will not provide significant amounts of assistance General Budget Support: An Alternative Assistance Approach 13 Finance and Fiduciary Capabilities for General Budget Support until financial systems and controls have improved. There is no exact for￾mula for making informed judgments about accept￾able fiduciary risk Tolerance for risk can vary accord￾ing to the impact expected and the amount of funds at stake. DFID’s review of 25 Heavily Indebted Poor Countries (HIPC), which includes Mozambique, revealed that none meet international codes and standards for public financial management and accountability. As an initial step, benchmarks can be used to assess risks and to set targets for improve￾ment (see Table 2). These benchmarks are consid￾ered temporary and will be replaced by codes and standards that are being developed by a joint effort of the multilateral development banks and the inter￾national financial management and accounting stan￾dards setting bodies. 14 PPC Evaluation Working Paper No. 18 Good Practice Principles A clear set of rules governs the budget process. The budget is comprehensive. The budget supports pro-poor strategies. The budget is a reliable guide to actual expenditure. Current year expenditure tightly controlled. Government procurement is in line with the principle of obtaining the best value for money spent and transparency in all transactions. Expenditure reporting is timely and accurate. There is effective independent scrutiny of government expenditure. Table 2. Principles and Benchmarks for Fiduciary Risk Assessments Benchmarks for Assessment ■ A budget law specifying fiscal management responsibilities is in operation. ■ Accounting policies and account code classifications are published and applied. ■ All general government activities are included in the budget. ■ Extra budgetary expenditure is minimal. ■ Annual budget allocations are broadly consistent with medium￾term sector expenditure plans or the overall budget. ■ Actual budget results are consistent with the budget. ■ Actual expenditures are reported in-year. ■ Systems are operating to control commitments, accruals, expenditures, and arrears. ■ There is appropriate use of competitive bidding rules. ■ Decisionmaking is recorded and auditable. ■ Effective action taken to identify and eliminate corruption. ■ Reconciliation of fiscal and bank records is carried out on a routine basis. ■ Audited annual accounts are submitted to parliament within the statutory period. ■ Government accounts are independently audited. ■ Government agencies are held to account for mismanagement. Criticisms and recommendations made by the auditors are followed up. Source. Department for International Development, March 2002. “Managing Fiduciary Risk When Providing Direct Budget Support.” Fiduciary risk in Mozambique based on DFID’s 2002 assessment is high, and thus major public expenditure management system reforms are need￾ed. This is in line with an IMF and World Bank study of the financial management of 24 HIPC countries in which Mozambique and 14 other countries require substantial upgrading of their public expenditure management in order to track spending. In Mozambique, the budgets of the line ministries are often a “plus-up” from previous years and not directly linked to the goals of the PARPA. Likewise, actual expenditures do not always reflect the approved budget. The World Bank’s Medium Term Expenditure Framework is designed to link the annual budget and the long-term goals of the PARPA. With no policy or budget discipline, yearly expenditures often have little relationship to the Medium Term Expenditure Framework. Budget breaking expenditures are further com￾pounded by financial leakage via the country’s pro￾curement practices. These practices encourage cor￾ruption owing to the lack of a competitive and transparent culture, trained staff and independence from political pressure. The World Bank estimates a savings to the Mozambique government of 30 to 40 percent if the procurement system were reformed. Quite apart from the foregoing discussion of risk, since funds for General Budget Support are deposit￾ed in a general fund, the money can easily be spent for non development related activities or dispropor￾tionately allocated to a particular region to the detri￾ment of the country’s poverty strategy. Approximately 70 percent of Mozambique’s budget (including central government payroll and procure￾ment) is spent in the Maputo Region which is rela￾tively more prosperous than other provinces. Such unevenness is not surprising in light of strong, cen￾tralized governance and a weak civil society. To address some of the problems above, the newly developed State Integrated Financial Management System (SISTAFE) will track expenditures which can then be linked to the budget. The system has the potential to provide full transparency and accountability in the use of public funds. SISTAFE was designed to be a user-friendly, computerized system that would eliminate paper financial records and improve the timing of reporting among min￾istries and provinces. Most of the weaknesses of the current system can be addressed by SISTAFE, but it needs strong political backing, qualified personnel and adequate equipment to ensure its success. Implementation of SISTAFE is beginning with the Ministry of Education with plans to roll it out to the rest of the ministries and the provinces by 2005. Although donors have much hope for the future success of SISTAFE, some expressed doubt that it could be implemented that quickly or that it would be effectively implemented, especially at the lower levels of the government. Monitoring Fiduciary Risk Denmark and Norway are leading an effort among the G-11 donors to design a joint fiduciary risk monitoring system to address their concerns over the present risks and to track changes that minimize risks. The G-11 economists group is working on a proposal and hopes to have a risk assessment system in place in 2004. One approach to monitoring fiduciary risk is to ask a set of simple questions3: ■ Is reporting accurate and timely? ■ Do actual expenditures match budget allocations? ■ Has value for money and transparency been demonstrated in procurement? ■ Have the objectives for which funds were allocated been achieved? ■ What progress has been made towards improving financial management? The tools that can be used to answer these questions include the government’s own reports, supplement￾ed by other sources such as independent audits. Monitoring should tie into the government’s budget and reporting cycle and its poverty reduction moni￾General Budget Support: An Alternative Assistance Approach 15 3 Department for International Development. March 2002. “Managing Fiduciary Risk When Providing Direct Budget Support.” 16 PPC Evaluation Working Paper No. 18 toring framework. Appropriate government reports may include4: ■ In-year financial reports ■ Annual audited accounts of government expenditure ■ Expenditure tracking surveys ■ National statistical surveys ■ Reports from the apex audit institution ■ Reports from parliamentary scrutiny committees. What happens if the results show weaknesses or sys￾tem failures? Some donors warn against modifying or stopping funding during the current fiscal year, as an interruption of funds would harm develop￾ment momentum by sharply reducing government revenue. Regular monitoring should identify prob￾lems early so they can be addressed. If adequate progress is not made, then funding should be reduced in the following year. The key is to link future funding to the achievement of specific results in the current year. The approach only works if donors and the government agree on specific per￾formance standards and if there are specific and painful penalties for failure to perform. The key to performance-based assistance is to have reliable and up to date financial accounting systems. Presently that information does not exist. Hopefully, it will be available through the integrated financial manage￾ment system—SISTAFE. 4 Department for International Development. March 2002. “Managing Fiduciary Risk When Providing Direct Budget Support.” Country management of donor assistance is complicated by the donors’ relationships with their own headquarters. Some donors, notably the Nordic countries and the EU are very centralized and have limited decision-making power in the field. Other field offices are fairly independ￾ent, or serve as their headquarters’ resource on plan￾ning, such as USAID, DFID and Dutch aid. In implementing General Budget Support, the British have a strict mandate to go forward and the opera￾tional freedom to take the lead among donors. Others, such as the Italians and the Portuguese, are very much in pilot mode, with a range of domestic policy and budget concerns to consider. Thus, despite an agreement in principle among donors to agree on common GBS procedures and assessment frameworks, political realities often require addition￾al layers of accountability and reporting through separate bilateral agreements. Staff Time and Transaction Costs Whether General Budget Support or other non￾project assistance approaches demand more or less host country and donor staff work is still a matter under discussion. Most analysts find that the switch to budget support increases staff time and costs for both the donors and government at least in the planning phases. For example, donors spend a great deal of time analyzing the macro environment, col￾laborating as a group, in discussions with govern￾ment officials and working with outside experts on institutional reforms in procurement, human capac￾ity development, oversight, and measures to reduce corruption. Equally important, government officials devote their time and effort to developing plans and agreements that work with individual donors’ agen￾das, establishing the necessary reporting, monitoring and assessment capacities, and responding to donor concerns. To the extent that donors are able to har￾monize their requirements, perhaps using General Budget Support as a model, the host government’s workload can be reduced. Though the transaction costs to donors of managing multiple projects may be reduced by the switch to General Budget Support or other forms of non￾project assistance, donors still need up-front policy, program analysis and financial analyses. However, unlike other assistance approaches, including project assistance, there is less of a need for donors to do audits, project management, engineering and logis￾tics planning. Country Ownership of the Development Process The Dutch and the British development agencies are advocates of GBS and argue with other donors and with their own governments that General Budget Support will result in more rapid and per￾manent change. They hypothesize that once devel￾opment activities are brought into the government’s budget and planning processes, host governments will have the means to institutionalize their poverty reduction and development approaches. The coun￾try then takes ownership of the process and develops its own planning and implementation abilities. At the very least, budget support forces government ministers to work together to determine priorities and lobby for funding from the Ministry of Planning and Finance, rather than competing for different funding streams directly from the donors. In Mozambique there are a number of tools for pol￾icy dialogue: the PARPA as a guide and economic plan, the Performance Assessment Framework as a monitoring tool, a proposed annual fiduciary risk assessment, an annual Joint Donor review, and ongoing technical assistance to improve capacity, among others. Technical assistance continues to be General Budget Support: An Alternative Assistance Approach 17 Development Management devoted to reforming financial management, linking the budgeting and planning processes, improving procurement and solving ongoing corruption and financial sector problems. Several donors referred to the government’s Performance Assessment Framework matrix as the best tool for assessing the outcomes of General Budget Support. The Joint Donor Review (the G-11’s coordinating mecha￾nism), recommends its adoption as a common framework for donors. As a government product, the donors view the Performance Assessment Framework positively. It was undertaken in close cooperation with them and other stakeholders. Donor Coordination When donors began to coordinate their humanitari￾an assistance during Mozambique’s protracted civil war, foreign aid grew to more than half the coun￾try’s GDP. In the years just after the war, when there were many donors, the country was known collo￾quially as the “Donor’s Republic of Mozambique.” Two of the reasons donor coordination works well in Mozambique may be the leadership capability of specific individuals and strong personal relationships between individual donors’ representatives. Donors consistently cite their coordination and collective impact as one of the principal arguments in favor of participating in General Budget Support. Donors also coordinated on disaster relief, balance of payments support and other non-project mecha￾nisms. They parceled out responsibility for assis￾tance to reach each of the country’s ten provinces and for particular sectors. This division is still per￾ceptible if one examines donor activity by province. Donor activities evolved into several coordinating mechanisms: ■ the Development Partners’ Group at the ambas￾sadorial level; ■ the Paris Consultative Group at the international headquarters level;5 ■ the G-11, which includes GBS donors at the operational and ambassadorial level; and ■ various sectoral and thematic support groups, in agriculture, health, education, water, energy and roads generally centered around SWAp discus￾sions, but also on budget, fiduciary risk, procure￾ment, infrastructure, private sector development and disaster preparedness. The G-11 has a continuing dialogue with the gov￾ernment on its development policies and budget and has successfully pushed for better policy plan￾ning; improved public financial management; estab￾lishment of process and outcome indicators for each sector; better budget execution; tax reform and; and procurement reform. The G-11 has not been as suc￾cessful with financial sector reform or efforts to reduce corruption. Despite their common agreement on the utility of General Budget Support and the progress made in monitoring and evaluation, G-11 members empha￾size their different concerns about its practice and have not yet completely harmonized their coordina￾tion efforts. The Nordic countries and the US (in its “observer” capacity), for example, continue to be wary, focusing on corruption, the need to demon￾strate results and fiduciary risks such as inadequate accounting systems, budget leakage and the ongoing banking scandals. Donors also have different approaches to measuring fiduciary risks and the degree to which they believe they should scrutinize budget and disbursement allocations. Finally, though there is a broad agreement between the gov￾ernment and donors, each donor employs different disbursement mechanisms, conditions and process￾es. Furthermore, despite a belief in harmonization, many donors have separate bilateral GBS agree￾ments with additional issues addressed — usually in response to domestic policy concerns. This compli￾cates the process and seems to run counter to the underlying philosophy of GBS. Technical and Management Skills Government technical capacity is not improving very rapidly, as new entrants into the labor pool pre￾fer to enter the private sector or the NGO commu￾18 PPC Evaluation Working Paper No. 18 5 Consultative group materials for the 2003 annual meeting available at http://www.worldbank.org/afr/mz/ General Budget Support: An Alternative Assistance Approach 19 nity rather than government. In addition, with total university attendees at only about 11,000 nation￾wide, it is difficult to imagine how the pool of those with technical and management skills will be increased. Unfortunately less than 3 percent of Ministry employees have a university degree. In the provinces and districts, the proportion of govern￾ment staff with advanced degrees is much lower. Unlike other African countries, there is no large pool of trained Mozambican expatriates living abroad that can be enticed to return and contribute to their country’s development. In fact, most work￾ers abroad were semi-skilled and unskilled workers in the South African mines or in East German fac￾tories. Their return, due to changes abroad, has put additional strains on the country’s social system. At the same time systems are being put in place that can be operated by relatively unskilled staff, as long as the infrastructure is in place to allow it to oper￾ate. On the expenditure side, the State Integrated Financial Management System (SISTAFE) is expect￾ed to improve accounting and reporting. Training will be organic, with users that adapt easily tapped to train additional users. Where there are insuffi￾cient computers or electrical power, the system can be followed on paper, and data can periodically be entered electronically at technology centers. On the monitoring and evaluation side, the Performance Assessment Framework for the PARPA institutionalizes data collection in support of the development agenda. Ministries have clear responsi￾bilities for data collection and reporting, and con￾tinue to receive some data analysis assistance from organizations such as The International Food Policy Research Institute. Government Institutional Capabilities Donors give the government high marks for doing an excellent job producing its and Performance Assessment Framework. In addition, they applaud the efforts of the Government commission on civil service reform, but some doubt its ability to carry out its ambitious reform agenda. The World Bank agreed to extend a Poverty Reduction Support Credit, in large measure because of improvements in the government’s planning capabilities. This credit will provide $50-$60 mil￾lion annually of General Budget Support and will replace much of the Bank’s sector support. Likewise, the IMF worked with Mozambique to finalize a Poverty Reduction and Growth Facility, a General Budget Support mechanism that emphasizes macro￾economic stability and growth along with financial and public sector reform as tools in fighting poverty. Decentralization The strongest institutional capacity, at least in plan￾ning and budgeting, resides in the Ministry of Finance, but this capacity may be limited to the most senior levels. Anecdotes abound about the Ministry’s difficulties in coordinating with sector ministries, delays in allocating funds to the provinces and budgeting errors that take months to correct. Similar technical capacity, at least for plan￾ning, management and assessment does not appear to extend to the sector ministries (with the excep￾tion of the PROAGRI staff), and becomes even thinner at the provincial and municipal levels. There is also a serious concern about the central govern￾ment’s ability to move funds out to the provinces in a predictable and consistent manner. In addition, government processes, such as procurement and tax collection, are in the early phases of reform. Finally, some donors believe the national statistical agency lacks strong data gathering capacity at the sectoral level, petty and institutional corruption threaten outcomes, and contracting and procurement mecha￾nisms still need improvement system-wide. When an effective accounting system is in place, it will meet donor concerns regarding financial accountability and the predictability of funding. Even more important will be the benefit to govern￾ment operations. As part of this assessment, there was an examination of financial transfers from the Ministry of Planning and Finance to the technical ministries and the transfer of funds from the techni￾cal ministries to provincial and district activities. There are major gaps and delays in funding. Sometimes no funds are received until the last few months of the fiscal year. Agriculture activities fol- low a natural weather cycle and not a budget cycle. If funds are not available on time for planting or harvesting, a whole year may be lost. The disburse￾ment problems for government health and HIV/AIDS lead to delays in health service deliveries increase illness and death rates. As a consequence, people have little confidence in the dependability of government services when they visit a health clinic that has no staff and no drugs. The Government has succeeded in putting a moni￾toring and evaluation system in place at the macro level for the poverty plan’s social indicators (the Performance Assessment Framework), but there is still no agreement on how to effectively track and measure spending as it flows through the whole sys￾tem nor how to take corrective action when prob￾lems arise. SISTAFE may be the first step the gov￾ernment takes towards establishing an effective pub￾lic sector financial accounting system. The Future of General Budget Support in Mozambique USAID/Mozambique provides project and sector support, but not General Budget Support. Some donors suggest that even a small (e.g., 10 percent contribution) toward budget support would give the U.S. a voice with the legitimacy to participate in the G-11’s planning and budget discussions. Participation would also allow the U.S. increased access to ministry documentation and a vote on the development of instruments such as the Joint Donor Agreement or the Performance Assessment Framework. Other donors are not as sanguine, per￾haps because US support at 10 percent might not be viewed as a significant amount to warrant a place at the G-11 table. Even the most ardent of donor supporters of General Budget Support, admit that their own aid ministries and parliaments are concerned about budget leakages and have slowed disbursements based on these concerns. They recognize that their central government can easily cut out budget sup￾port if Mozambique fails to improve its financial management. At the same time, however, it took a few donors willing to brave the risk to begin a process that pro￾vides General Budget Support as a way to influence policies and budget decisions. How that translates to the individual farmer, mother, student or war vet￾eran remains to be seen. If successful, Mozambique will be able to shed its label as the “Donor’s Republic of Mozambique,” reclaim its full sover￾eignty, and gain the ability to improve the status of its most vulnerable residents. 20 PPC Evaluation Working Paper No. 18 General Budget Support: An Alternative Assistance Approach 21 Annex 1: The Government’s Sectoral Approach and the Role of USAID and Other Donors Corruption. Transparency international ranks Mozambique high on its list of most corrupt countries. Local respondents report examples of corruption including high level bank loan corruption and pay￾offs to government workers Decentralization. Government has pledged to decentralize and has put plan￾ning mechanisms in place at the provincial and sector levels that support this process. But the culture is strongly entrenched in centralized government with extensive service delivery responsibilities. Elections at national levels are generally perceived as free and fair, thought with admitted irregularities. Provincial gover￾nors are appointed. Judicial weakness. Weakness are multi￾dimensional and corruption is pervasive. There is no culture of judicial independ￾ence. Individual citizens cannot access lawyers, particularly for defense. There are human rights issues with police and pris￾ons Family law outcomes are generally perceived to be weighted heavily towards men Media. Media are generally perceived to have been relatively free, although some sectors (e.g. police) have been known to take action in response to critical press. There have been two high profile media￾related assassinations and attempts (e.g., the murder of Carlos Cardoso, who was examining a banking scandal touching the highest levels of Mozambican society). Through SO10, “Municipal Governance Increasingly Democratic,” USAID will demonstrate how to repair the disconnect between Mozambican citizens and their governance bodies. This is a small SO for the mission, allowing a coordinating role with civil society. USAID will work in targeted communities to increase par￾ticipation by citizens and civil society and increase democrat￾ic decision making of munici￾pal governments. USAID will bring together municipal communities, including leaders, NGOs, individuals, and private enter￾prises, to identify and address agreed upon priorities. The lessons learned and models of good governance will be shared with other communi￾ties. In the area of corruption, all SOs will work directly with the private sector and NGOs to expose corruption and advocate for adequate and timely government response. SO teams will be able to call on SO10 resources and expertise in the design of anti￾corruption efforts. The World Bank, IMF and other donors are focusing on civil service reform and the SISTAFE financial manage￾ment system. Others are focusing on elimi￾nating corruption in the banking sector. Denmark is active in the legal sector. The Swedish are focusing on government auditing and oversight capability. Others are focusing on human capacity development, procurement reform, strength￾ening and expanding moni￾toring and assessment, the link between budgets and pol￾icy. Strengthening financial man￾agement systems is critical to the future of budget support, as even the staunchest sup￾porters must justify their spending to their own parlia￾ment and there is a percep￾tion that the society is quite corrupt. Support for NGOs is not compatible with budget sup￾port, which is provided to government entities. Corruption is a major barrier to unconditional budget sup￾port. Improving local governance capacity is crucial to the deliv￾ery of grass-roots anti-poverty programs as well as to creating an environment conducive to growth. Table 3. The Government’s Sectoral Approach and the Role of USAID and Other Donors Country Situation and Government Approach USAID Program Other Donor Programs Role of Budget Support Sector: Democracy and Governance 22 PPC Evaluation Working Paper No. 18 In Mozambique, small family farms using traditional technology produce the majori￾ty of agricultural output. There is poor access to improved crop varieties, low use of fertilizers or other yield-enhancing tech￾niques and low use of labor-saving devices or irrigation. To promoted trade, Mozambique needs to work on customs reforms, liberalizing labor laws, improving transport infrastruc￾ture, securing access to land, improving tax law and administration, reducing petty corruption, decreasing red tape and increasing access to financial services. Labor-intensive exports are showing an upward trend, reversing two decades of decline. The Government of Uganda, USAID and other donors are pursuing a two-pronged approach to stimulate economic growth and reduce poverty: (1) increase produc￾tivity in the traditional agricultural sector and facilitate its transformation; and (2) establish an environment where labor￾intensive manufacturing exports can thrive. The first prong reduces absolute poverty in the short- to medium-term and the second sustains high levels of growth and creates employment alternatives to subsistence agriculture over the medium￾to long-term. The PARPA highlights the importance of establishing the basic conditions that will accelerate the creation of non-agricultural employment by encouraging labor-inten￾sive industry and small and medium enterprise. Over the last few years, the Government of Mozambique has promoted exports through implementing the SADC free trade area; preparing a Trade Mainstreaming report; building a strong investment promotion agency; taking sig￾nificant steps in customs reform; imple￾menting tax incentives, supporting export promotion zones; and reducing red tape. USAID supports economic growth, agriculture and trade through two strategic objec￾tives. SO6 – “Rapid Rural Income Growth Sustained in Target Areas” is the mission’s priority SO. USAID will work to increase smallholder sales of agricultural production, expand rural enterprises and improve transport infrastruc￾ture. Through this SO, USAID contributes to the PROAGRI SWAp. SO7 – “Labor-Intensive Exports Increased” is the sec￾ond major prong in the mis￾sion’s strategy. USAID will work to increase exports, cre￾ate jobs and improve econom￾ic freedom in Mozambique. It will do this by expanding markets for exports, improv￾ing the enabling environment for exports, and strengthening the capacity of labor-intensive industries. Contributors to the PROA￾GRI SWAp in descending order by disbursements: USAID, EC, DANIDA, Netherlands, Ireland, Sweden, DFID, IFAD, Italy and the World Bank. Some donors have stopped funding PROAGRI and have switched to General Budget Support. They prefer to see their money going through the Ministry of Planning and Finance, who then sends it to the Ministry of Agriculture and Rural Development, as determined necessary by the government. Sector: Economic Growth, Agriculture, and Trade Table 3. The Government’s Sectoral Approach and the Role of USAID and Other Donors Country Situation and Government Approach USAID Program Other Donor Programs Role of Budget Support General Budget Support: An Alternative Assistance Approach 23 Government spending. Increased almost three-fold since 1999 ($72 million to $196 million) Literacy. Despite low achievements and major setbacks during the civil wars, Mozambique has made incredible progress since independence. In 1975, 3 percent of the population was literate. Currently, literacy stands at 48 percent and is increasing. Basic education. Education is compulsory for 7 years, but not enforced. It is free at primary level; not free at secondary, although even at the primary level, stu￾dents pay a matriculation fee and for books and uniforms and school supplies. The government’s plan focuses on a major school construction program, teacher training and combating corrupt activities such as demanding bribes or sexual favors for school entrance, grades or promotions. Secondary education. Nationwide, there is very low enrollment (around 4,100 12th graders in 2000) due to economic con￾cerns, lack of secondary schools and teach￾ers in rural areas, and significant gender disparity in the higher grades, all issues addressed by the national plan. Tertiary and technical education. Mozambique needs skilled people to build the society and the economy. This is addressed, but not strongly, in the plan. There is one large university. It lacks capacity and also lacks linkages with the technical schools. Nonformal education. Post-war, most students were in first grade — carrying some social stigma for older students. Need to address adult learners through non-formal education and vocational training. This is in the plan, but not elab￾orated upon. USAID currently does not have a basic education pro￾gram. USAID’s program focuses on human capacity development and includes training in its programs as they relate to specific strategic objectives, such as HIV/AIDS education campaigns, exten￾sion worker training, etc. The mission assumes this remains a major part of the government’s strategy. It also recommends that the govern￾ment focus more on non-for￾mal education, given the number of adults who have not been part of the educa￾tion system, the number of repeaters and the need for vocational and life skills. Individual bilateral donors are contributing to the education SWAp that is being devel￾oped, but also have projects in teacher training, school con￾struction and curriculum development. USAID has been involved as an observer. Progress to date has disappointed the donors, who note a lack of quality in the planning and strategies. Some donors have pooled resources for school buildings and are preparing an educa￾tion SWAp. Most donors have considerable concern over slow program progress – cen￾tering on the Ministry of Education’s lack of capacity to absorb and use funds effec￾tively. SISTAFE is being piloted in the Ministry of Education because this is where the majority (52,000) of civil ser￾vants are found. It will not go out to the schools, but at least to the provincial directorates. Bringing education spending on budget and in line with the Poverty Reduction Strategic Plan (PARPA), as measured by the Poverty Assessment Framework, avoids the problem of schools being constructed without access to teachers or materials, an example cited by a World Bank source. Sector: Education Table 3. The Government’s Sectoral Approach and the Role of USAID and Other Donors Country Situation and Government Approach USAID Program Other Donor Programs Role of Budget Support 24 PPC Evaluation Working Paper No. 18 USAID has no separate strate￾gic objective for the environ￾ment. Environmental efforts are related to agriculture and health. The USAID Strategy has identified the environmental linkage with each new strate￾gic objective, and several of the intermediate results tar￾gets. These range from con￾sidering the environmental sustainability of work in a specific export sector, to including environmental impact analysis in rural roads development to ensuring proper disposal of blood and blood products under the HIV/AIDS objective. In addition, USAID partici￾pates in the regional famine early warning system (FEWS), which was successful in deal￾ing with the drought and flooding. At the program level, USAID provides some support for environmental efforts through private sector and agricultural programs. These include sup￾port for eco-tourism such as the linkage of Kruger National Park with Mozambique’s national game park as well as re-stocking other game parks. PROAGRI efforts include a forestry component that aims to increase biological diversity, improve forest management and develop forest products exports. The Nordic and Dutch pro￾grams have projects in fish￾eries, forests and game parks, water and sanitation. Environment is only addressed in budget support as the government addresses these issues in its budget. Other than that, it is addressed by individual donor and NGO projects. There are few to no environment indi￾cators in the Government’s draft Poverty Assessment Framework. Sector: Environment Table 3. The Government’s Sectoral Approach and the Role of USAID and Other Donors Country Situation and Government Approach USAID Program Other Donor Programs Role of Budget Support Natural resources management. Mozambique’s game parks were severely depleted by the war and had limited man￾agement capability. However, there are currently efforts to reintroduce elephants, rhinos, etc. (e.g., Gorongoza). Poaching and illegal logging continue to occur, due to inadequate monitoring and corruption. Many natural resources used by the popu￾lation are not captured in any formal data systems, such as firewood, wild game, fish and medicinal plants. When war or natu￾ral disaster stresses the economy or the political situation, the country’s natural resources provide virtually 100 percent of the rural population’s energy, food, shelter and even clothing. Natural Disaster. El Niño-related drought and flooding in 2002-2002, Mozambique’s plan includes improve￾ments in disaster preparedness and mitiga￾tion. Regulation. The Poverty Reduction Strategy Plan mentions this, but environ￾ment is clearly not a focus in the poverty reduction and economic growth priorities of the government and the donors General Budget Support: An Alternative Assistance Approach 25 USAID’s strategy puts priority on health sector development, primarily maternal and child health and HIV/AIDS pre￾vention, all of which are noted as priorities in the Government’s poverty plan. In addition, the Mission will work to combat petty corrup￾tion among health care work￾ers. USAID’s program concen￾trates on technical assistance and public awareness and education campaigns. With the new Presidential Initiative on HIV/AIDS, pharmaceuti￾cals will be provided. The strategy assumes donor coor￾dination, and expects others to work on the assignment of additional technical personnel to the health sector along with their promotion of “healthy behaviors.” 27 donor agencies or lending institutions providing pooled support of more than $100 million to import drugs. And for provincial-level health efforts. A newly established SWAp provides additional on-budget sector support (in local cur￾rency). The USAID program cannot contribute to this SWAp, but is providing tech￾nical assistance to develop stronger financial manage￾ment in the sector (compati￾ble with SISTAFE) and to maintain a forum with donors involved in the SWAp The Performance Assessment Framework for the PARPA contains multiple indicators to track progress in this sector, along with indicators to track spending down to the provin￾cial level. Many donors con￾tinue to hedge their General Budget Support bets through provincial projects. Sector: Health Table 3. The Government’s Sectoral Approach and the Role of USAID and Other Donors Country Situation and Government Approach USAID Program Other Donor Programs Role of Budget Support Government spending: increased from $59.7 million in 1999 to $89.4 million in 2001 Primary Care. Lack of access to clinics, vaccinations, and basic care providers result in sobering statistics (e.g., 25 per￾cent of children die before age 5). Corruption further hinders the delivery of the services that do exist. AIDS. Mozambique is one of 10 most affected countries, with 13 percent of the population infected and an estimated 1 million AIDS orphans. Overall life expectancy is now reported at 40 for women and 38 for men, and it is drop￾ping. Lack of access to clean water. According to the UNDP, 37 percent of the overall population lacked access to clean water. Additional health sector problems include malaria, tuberculosis, leprosy, women dying in childbirth and malnutrition 26 PPC Evaluation Working Paper No. 18 USAID will emphasize women and gender equity in all of its SOs. Interventions by USAID include: SO6. Women will be benefici￾aries of improvements to fam￾ily farming, and market infor￾mation will be made accessi￾ble to women, SO7. Ensure that labor stan￾dards are applied and policy research conducted to support gender equality. SO8. Expand the concept of health to include the stages of a woman’s life cycle. Include men and youth in reproduc￾tive and maternal health campaigns. SO9. “Community leaders” will include influential women as well as individuals who demonstrate an interest in contributing to behavior change at the community level. Gender issues will be a central part of advocacy cam￾paigns. The disproportionate burden of community and home-based car falling on women and girls will be addressed. SO10. Address some of the reasons why women have less access to information about their rights and fewer oppor￾tunities to participate in deci￾sion making. Ensure that community councils effective￾ly incorporate gender issues into their development plans and promote women’s partici￾pation in the formulation, implementation and monitor￾ing of the plans. Sector: Women and Gender Equity Table 3. The Government’s Sectoral Approach and the Role of USAID and Other Donors Country Situation and Government Approach USAID Program Other Donor Programs Role of Budget Support In spite of attempts to foster greater gen￾der equity in Mozambique, the majority of women, especially rural women, have not experienced significant positive impacts. The Government of Mozambique has demonstrated some commitment to the integration of women into the political and economic processes of the country. The PARPA has a focus on gender and gender working groups were created in both the agriculture and health sectors. USAID Washington Brian Frantz, Economist, AFR/DR Stephen Giddings, Director, Program Office, AFR/DP Janet Kerley, Evaluation Officer, AFR/DP/POSE Mary Alice Klienjan, Office of the General Counsel Deborah Mendelson, AFR/SA, Mozambique Desk Officer Thomas O’Keefe, Deputy Director, AFR/DP Jay Smith, Director, AFR/DP USAID Mission in Mozambique Chris Barrett, Health, HIV/ADIS Tim Born, Economic Growth Christine DeVoest, Agriculture Thomas Johnstone, Controller Jay Knott, Mission Director Julius Schlotthauer, Economist Scott Simons, Agriculture Senior Policy Advisor James Watson, Program Officer International Donors Denmark: Embassy of Denmark Thomas Thomsen, Counselor and DANIDA Representative European Union: Delegation of the European Commission in Mozambique Francisco Carreras, Economics Counselor Erik Von Pistohlkors, Second Secretary, Social Sectors and Humanitarian Assistance Jean Risopoulos, Food Security Officer France: French Development Agency Guillaume Chiron, Economist Francoise Desmazieres, Country Director G-11 Economist Group International Monetary Fund Perry Perone, Resident Representative Italy: Italian Cooperation Andrea Borgarello, Economist Japan: Japan Embassy Ichiro Muto, Head, Economics and Development Netherlands: Royal Netherlands Embassy Weert W. Mostert, Head Development Cooperation Norway: NORAID Lars Ekman, Economist Portugal: Portuguese Cooperation Ines P. Alves, Cooperation Aid for Health Fernando A. Costa, Counselor for Cooperation United Kingdom: Department for International Development Allison Beattie, Health and Education Advisor World Bank Antonio Franco, Senior Country Economist Government of Mozambique Ministry of Agriculture and Rural Development Central Ministry office: Marcelo Chapine, Economic Directorate, Planning Abraham Kahsai, Economic Directorate, Planning Rogeria C. Muianga, Director, Administration and Finance Directorate Manica Province office: Hans Erskog, Financial Advisor (funded by DANIDA) Jose da Graca, Director Cremildo Rungo, Head, Forest and Wildlife Department General Budget Support: An Alternative Assistance Approach 27 Annex 2. Interview Contact List Ministry of Finance and Planning Antonio Laice, National Director of Treasury Emilia Silvestre, National Treasury, Budget Support NGOs World Vision Omo Olupona, Director of Health and Nutrition 28 PPC Evaluation Working Paper No. 18 Africa South of the Sahara. 2003. 32nd ed. NewYork: Europa Publications. Balogun, Paul. 2003. “Issues and Options for Evaluating General Budget Support.” Unpublished report for the U.K. Department for International Development. Batley, Richard. 2002. Mozambique: A Country Case Study. Report for OECD Development Assistance Committee. Birmingham: International Development Department, University of Birmingham. Booth, David, and Andrew Lawson. 2003. “Evaluating General Budget Support: Issues Emerging from DFID Evaluability Study and Proposals for Joint Evaluation Work.” Unpublished aide-memoire, Uganda Ministry of Finance, Planning and Economic Development Workshop, Kampala. Boudreau, Tanya. 2001. Food Economy Baseline Profile of the Lower Limpopo Food Economy Zone: Gaza Province Mozambique. USAID FEWS NET, The Food Economy Group, Ministry of Agriculture, and Ministry of Health. Central Intelligence Agency. 2003. “Mozambique.” In The World Factbook 2003. Clarke, Grayson, Jens Claussen, Rolf Kappel, Jytte Laursen, and Stefan Sjölander. 1999. Programme Assistance to Mozambique; A Joint Donor Review. SIDA Evaluation 99/32. Department for Africa, Swedish International Development Cooperation Agency. DFID. 2002. “Managing Fiduciary Risk When Providing Direct Budget Support.” Draft Policy Paper, March 2002. U.K. Department for International Development Economist Intelligence Unit. 2003. Country Profile: Mozambique. London: Economist Intelligence Unit. EC. n.d. “Variable Tranches in General Budget Support: Implementation Guidelines and Evaluations.” Development Policy and Sectoral Issues, Directorate General for Development. Brussels: European Commission. Falk, Hans, Kåre Landfald, and Pamela Rebelo. 2003. “Mozambique.” Development Policy Review 21 (2): 235–52. Foster, Mick. 1999. “Integrating Aid into Budget Management.” Issue Paper 1. Prepared as background reading for the MEFM/DFID Workshop on Good Practice in Public Expenditure Management, Capetown, June 1999. Foster, Mick. 2002. “Direct Budget Support to Mozambique: Report to the Department for International Development.” Mick Foster Economics, Ltd. Foster, Mick, and Jennifer Leavy. 2001. “The Choice of Financial Aid Instruments.” Working Paper 158. Center for Aid and Public Expenditure. London: Overseas Development Institute. Fozzard, Adrian. 2002. How, When and Why Does Poverty Get Budget Priority? Poverty Reduction Strategy and Public Expenditure in Mozambique. Case Study 5. Working Paper 167. London: Overseas Development Institute. Fox, Justin. 2002. With Both Hands Waving: A Journey Through Mozambique. Cape Town: Kwekla Books. Frantz, Brian. 2002a. “Budget Support SWAps and All That: A Background Note for the DFID Consultations.” Unpublished PPC briefing paper. Washington, D.C.: USAID. Frantz, Brian. 2002b. “Improving Aid Effectiveness: What Role for Sector-Wide Approaches?” PPC briefing paper. Washington, D.C.: USAID. Frantz, Brian, and Carla Komich. 2003. “A Donor Coordination Assessment for USAID/Mozambique.” Washington, D.C.: USAID. Gall, Timothy L., and Susan B. Gall, eds. 1999. Worldmark, Chronology of the Nations. Volume 1. Africa. Woodbridge, Ct: Gale Group, 380–89. Gastow, Peter, and Marcelo Mosse. 2002. “Mozambique: Threats Posed by the Penetration of Criminal Networks.” ISS General Budget Support: An Alternative Assistance Approach 29 Annex 3. Bibliography Regional Seminar, Organized Crime, Corruption and Governance in the SADC Region. Pretoria: Institute for Security Studies (ISS). Hanlon, Joseph. 2002. “Are Donors to Mozambique Promoting Corruption?” Paper submitted to the conference “Towards a New Political Economy of Development: Globalisation and Governance,” University of Sheffield, July 2002. Harding, Alan. 2002. “General Budget Support Evaluability Study: Mozambique Case Study.” Center for the Study of African Economics, University of Oxford. Unpublished revised final report for DFID. Hutton, Guy, et al. “Sector-Wide Approaches in the Swiss Agency for Development and Cooperation (SDC).” Unpublished report for SDC. International Monetary Fund. 2003. Republic of Mozambique: Fifth Review Under the Poverty Reduction and Growth Facility and Request for Waiver of Performance Criterion. IMF Country Report No. 03/288. Washington, D.C. IMF. . Joint Program for Macro-Financial Support, Mozambique. 2003. “Joint Donor Review, 2003.” Redraft No. 3, Unpublished aide-memoire. Jones, Stephen, and Andrew Lawson. 2000. Moving from Projects to Programmatic Aid. OED Working Paper Series, No. 5. Washington, D.C.: World Bank. Lodge, Tom. 2003. “Southern Africa: Angola, Botswana, Lesotho, Madagascar, Malawi, Mauritius, Mozambique, Namibia, South Africa, Swaziland, Zambia, Zimbabwe.” In Global Corruption Report 2003. Berlin: Transparency International: 248–60. Lonstrup, Esther. 2002. “Status of Public Financial Management in Mozambique, Fiduciary Risk Assessment.” Unpublished report for DFID. “Macro-Financial Support to Mozambique: Joint Donor Review, 2002.” Unpublished aide-memoire. Hotel Rovuma Carlton, Maputo, May 6–10, 2002. Mucavele, Firmino Gabriel, and Gilead Isaac Mlay. 1998. Country Note for Mozambique. Kampala: 2020 Vision Network for East Africa and International Food Policy Research Institute. Naschold, Felix, and David Booth. 2002. “General Budget Support Evaluability Study, Literature Review.” London: Overseas Development Institute. Unpublished report for DFID Evaluation Department. Oxford Policy Management and Overseas Development Institute. 2002. General Budget Support Evaluability Study Phase I: Final Synthesis Report. Report to DFID. London: Oxford Policy Management and ODI. Pavignani, Enrico, Stefan Sjölander, and Dag Aarnes. 2002. “Moving On-Budget in the Health Sector of Mozambique: Requirements, Features and Implications of Future Aid Management and Financing Mechanisms.” Report for Swiss Agency for Development and Cooperation. Republic of Mozambique. 2000. “Joint Macro-Financial Aid Program to Mozambique.” November 2000. Republic of Mozambique. 2001a. “Action Plan for the Reduction of Absolute Poverty (2001–2005), Republic of Mozambique. Strategy Document for the Reduction of Poverty and Promotion of Economic Growth.” Final Version Approved by the Council of Ministers, April 2001. Republic of Mozambique. 2001b. “The National Program for Agricultural Development (PROAGRI) Common Flow for Funds Mechanism (CFFM).” Memorandum of Understanding between the Ministry of Planning and Finance, Ministry of Agriculture and Rural Development, World Bank, DFID, USAID, Dutch Embassy, Irish Embassy, Swedish Embassy, Danish Embassy, IFAD and European Commission. Republic of Mozambique. 2003a. “Action Plan for the Reduction of Absolute Poverty (PARPA): Implementation and Perspectives.” Unpublished government document prepared for Consultative Group, September–October 2003. Republic of Mozambique. 2003b. “Matrices PAF/DRSC.” Draft for Consultative Group, August 19, 2003. Republic of Mozambique. Ministry of Agriculture and Fisheries. 1998. “PROAGRI: 1998 to 2003.” Executive Summary. Republic of Mozambique. Ministry of Agriculture and Rural Development. 2003. “PROAGRI Evaluation.” Final Report, Volume I. Republic of Mozambique. Ministry of Health. 2002. “General Health Sector Common Fund Guidelines (Early 30 PPC Evaluation Working Paper No. 18 Draft for Consultation and Consensus Building).” Directorate of Planning and Cooperation (DPC). Republic of Mozambique. Ministry of Planning and Finance. 2003. “PARPA Implementation Evaluation Report: 2001; 1st half 2002 (Sectoral Performance); PROAGRI Evaluation.” Final Report, Volume I. 3rd quarter 2002 (Public Expenditure).” Republic of Mozambique and European Community. “Mozambique-European Community Country Strategy Paper and National Indicative Program for the Period 2001–2007.” Seleti, Yonah. 2000. “The Public in the Exorcism of the Police in Mozambique: Challenges of Institutional Democratization.” Journal of Southern African Studies, 26 (2): 349–64. USAID. 1996. “USAID Policy Paper on Program Assistance.” Bureau for Program and Policy Coordination. USAID/Mozambique. 1998. “Mozambique: 1998 Macro￾Economic Situation Update for the Use of NPA.” R4 Annex. USAID/Mozambique. 2003. “Mozambique Country Strategic Plan FY 2004–2010.” U.S. Department of State. 2003. “Country Reports on Human Rights Practices, 2002: Mozambique.” Bureau of Democracy, Human Rights, and Labor. Washington, D.C. White, Howard.1999. “Reform, Rehabilitation and Recovery: Programme Aide to Mozambique.” Institute of Social Studies. Unpublished evaluation report for SIDA. World Bank. 2001a. “Mozambique Country Economic Memorandum: Growth Prospects and Reform Agenda.” Africa Regional Office, Report No. 20601-MZ. Washington, D.C.: World Bank. World Bank. 2001b. “Mozambique Public Expenditure Management Review.” Africa Regional Office, Report No. 22985-MOZ. Washington, D.C.: World Bank. World Bank. 2002. “Mozambique at a Glance.” General Budget Support: An Alternative Assistance Approach 31 Definitions And Typologies Depending on the type of development prob￾lem being addressed, and the country situa￾tion, there are a number of different ways to deliver assistance. USAID can choose from a vari￾ety of assistance techniques to find the best mix that is suited to a developing country’s special conditions and needs. Before doing that, it is important to define what each type of assistance includes and the benefits and drawbacks of different techniques. Balance of Payment Support6 USAID provides a resource transfer, in the form of foreign exchange (cash transfer) or a commodity import program. It is designed to promote econom￾ic or political stability. It works best when it sup￾ports short- and medium-term economic or political stabilization through measures that address immedi￾ate gaps in the country’s balance of payments or budget. Filling a balance of payments or budget shortfall makes the most sense when the rationale is political, or when the shortfall is temporary. It must be linked to policy reforms that will close the gap by restruc￾turing the national economy. Policy reforms have benefits but they also have costs. USAID resource transfer mitigates the possible short-term drop in consumption and production that sometimes occurs as policy reforms are adopted. While there are eco￾nomic groups that will benefit from reforms, others may be hurt. Potential policy losers often try to sub￾vert reforms. Balance of payments support is rarely successful over a long time period or in countries that are unwilling to change their policies and insti￾tutional environment. Sector Program Assistance7 A sector includes a set of economic activities unified by a common output narrow enough to have an analytical identity and broad enough to encompass significant investment and policy issues. It can refer to a broad area such as agriculture, industry, educa￾tion, health, exports; or it can refer to a sub-sector such as agricultural marketing, health care financing or child survival. This type of assistance focuses on sector constraints to sustainable growth. It supports the implementation of reforms and other actions to break those constraints. Sector assistance enables the government to offset the short-term costs of sectoral policy reform. Disbursement of USAID resources is always linked to fulfillment of conditionality, previ￾ously agreed to by USAID and the host govern￾ment. Sector Wide Approach (SWAp) A SWAp is an unusual acronym, with capital and lower case letters. It is also not the same as sector program assistance. Sector program assistance includes an agreed policy and expenditure plan for a sector. USAID disbursements are dependent upon the government meeting the agreed conditionality. An individual donor usually negotiates its own sec￾tor agreement. In contrast, a SWAp aims to coordi￾nate all donor assistance in support of a common sector expenditure program. SWAp financing typi￾cally includes a range of donor budget support, project aid and technical assistance, which may or may not be earmarked to specific expenditures or disbursed through the government’s own budget process. The defining characteristic of a SWAp is that donor funding for a sector supports a single sector policy and expenditure program. The govern￾General Budget Support: An Alternative Assistance Approach 33 Annex 4. Alternative Assistance Approaches 6 “USAID Policy Paper on Program Assistance.” PN-ACD-317, February 1996. 7 Ibid. ment provides the leadership and donors adopt a common approach to support the government’s effort. General budget support This type of assistance is much like balance of pay￾ment support since it supports structural adjustment and policy reforms. But in contrast to balance of payments support, it focuses attention on the local currency counterpart to the foreign exchange, rather than the foreign exchange itself. The main focus of conditionality is on an agenda of policy measures which the government has agreed to implement. They typically include agreement on overall budget priorities, as set out in a medium term budget and expenditure framework. With agreement on the budget as a whole, there is no need to earmark spe￾cific aid flows to specific country expenditures. Funds are disbursed to the host government’s national budget. Since support is for the budget as a whole, accountability is based on government audit￾ed accounts of its total revenues and expenditures. Project aid A project focuses on a limited set of activities where inputs can generate outputs and achieve a higher order purpose or goal. It identifies a problem such as low crop yields. It then analyzes the problems that are holding back productivity: inadequate tech￾nology, lack of trained manpower, limited markets, high transport costs and lack of needed inputs. It then identifies the most critical problems and devises solutions: development of high-yield seed, training agriculture extension agents, and providing fertilizer. It may decide that other projects will have to deal with markets, transportation, policies and regulatory reform. The project may provide techni￾cal experts, manpower training, fertilizer and a research farm to develop new seed varieties. USAID works closely with the government as it designs and implements the project. USAID usually uses its own procurement, disbursement and accountability pro￾cedures. The project may or may not be included in the government’s budget. The project may be imple￾mented by a U.S. institutional contractor, a U.S. PVO or a local NGO. Projects are closely monitored by USAID officers in the particular country. USAID has a set of tools that protect development assistance funds from financial irregularities and corruption. Projects are imple￾mented through contracts, grants, or credits to recipient government agencies, private voluntary organizations, consulting companies, or other pri￾vate sector institutions. This form of assistance is in contrast to the direct transfer of funds to a country government, whether through block grants, balance of payments assistance, or direct budget support. Which approach to use? “It all depends upon the situation, other things being equal.” “The appropriate mix of aid instruments depends mainly on the degree of consensus on policies and the capacity of the recipient to implement them. Policy is understood to refer to actions rather than merely words, and capacity is understood to encom￾pass not just technical capacity but the ability to adhere to standards of good financial management and good governance.”8 The choice of aid instruments depends upon the confidence USAID has in the developing country government’s management and development policy framework. When the government lacks skilled managers, has weak financial controls, poor accountability, corruption, and weak policies and institutions — then projects make the most sense. 34 PPC Evaluation Working Paper No. 18 8 Foster, Mick and Leavy, Jennifer. “The Choice of Financial AID Instruments.” Overseas Development Institute, London. October 2001. Projects are donor-run and donor-managed. They provide technical experts and training to solve a specific development bottleneck. Projects are fine but if there are problems with insti￾tutions and policies, another approach may be need￾ed. A road can be built but if transport policy fails to stop overloaded trucks, fails to raise funds and does not maintain the road, the road will wash away. If nurses are trained but healthcare financing, drug supplies and clinics are not available, the trained nurses will not solve health problems. When USAID provides project assistance it is not part of the host government’s broader program decisions or budget process. In contrast, with program assistance it gains a seat at the policy table and can work with the government on the many interrelated problems that harm development. Program assistance can solve problems that individual projects can not address. Over the years, USAID’s portfolio has included a substantial share of non-project assistance — cash transfers, budget support, policy reform grants, sec￾tor support, PL-480 Title I and commodity import programs. They have supported both macroeco￾nomic policy reforms (foreign exchange, credit, investment, and privatization) and sectoral policy reforms in agriculture, education and health. In the last 15 years non-project assistance has been used extensively by the Africa Bureau. The Development Fund for Africa (late 1980s) is a good example of balance of payments support. USAID provided cash transfers to African governments that implemented policy reforms. This was a dramatic change from the previous use of USAID managed-projects. There have been a number of policy-based programs in other regions. A review of 60 USAID evaluations covering sector, budget support, and cash transfer programs in several regions generated a number of helpful insights. 1. Staff requirements. Projects rely on project man￾agers and a number of technical experts to design and implement development activities. One of the key assumptions about budget support was that it would reduce staff requirements for both USAID and the host government. Giving a check to the government to implement policy reforms should be simple. But in fact, both the design and implemen￾tation of policy reform programs proved to be very staff-intensive for both USAID and host govern￾ments. Since policy reform is designed to change institu￾tions, it requires extensive technical, economic, social, political and institutional analysis during planning and implementation of the reforms. There is also a need for senior level staff to both analyze the development problems and to design reform programs. As the programs were implemented new questions and problems usually came to the surface, requiring new analysis and new solutions. A 1991 CDIE evaluation of Africa cash transfers supporting policy reform found that: “Policy reform was staff-intensive for both A.I.D. and host govern￾ments. Compared with traditional projects, it required more senior management time and atten￾tion and placed greater demands on professional staff for more sophisticated monitoring of policy reform implementation.” A 1999 world-wide health policy reform assessment found that: “…the analysis needed to design a health policy reform program was much more extensive than project assistance.” A 1997 Africa Bureau evaluation of non-project assis￾tance found that: “Experience with a wide range of sectoral programs indicates that management bur￾dens have been underestimated. There is also a lack of host country ministry institutional capacity to address the managerial and technical requirements of non-project assistance.” General Budget Support: An Alternative Assistance Approach 35 Annex 5. USAID Experience With Non-project Assistance 2. Donor coordination. Coordination was a prob￾lem when donors had different ideas about policy issues. Even with high level strategic agreement, donors often had different agendas that created implementation problems. A 1996 Africa Bureau review of non-project assistance found that: “A key aspect of success or failure in program design is the degree of harmonization among donors and host country officials in setting clear and consistent sec￾toral objectives. Multiple donor agendas and bilater￾al special country interests are often counterproduc￾tive.” The 1991 CDIE evaluation of Africa policy reform found that: “Every donor is in favor of improved donor coordination. However, agreements reached among donors at donor meetings in Paris may not provide the answers to in-country imple￾mentation issues. Donors at times had differing agendas and were pressing for inconsistent and con￾flicting policy changes.” A final problem was identi￾fied in an Africa education sector review. When there are a number of donors and it is difficult to get agreement on a set of policies, there is a tenden￾cy for conditions to become diluted and watered￾down. 3. Donor funding. USAID operates with an annual budget. It usually forward-funds a 5-year project to cover construction, equipment and contract staff. Cash-transfer policy reforms are also funded over a number of years. But it is much easier to reduce or stop a planned series of cash transfers when per￾formance or politics are a problem. A developing country is expecting multi-year support for a series of institutional and manpower changes. A vote in the U.N. that goes against U.S. interests or prob￾lems with a neighboring country may bring a politi￾cal halt to U.S. aid flows. Or, after a year or two, USAID may move on to other priorities or other approaches. That happened in USAID’s Africa pro￾gram. In the late 1980s, USAID planned to switch its whole Africa program from projects to policy￾based cash transfers. However, by the early 1990s the opposite happened — it was moving back to projects. When donors provide their agreed budget support they expect the host government to provide its share. However, an inefficient tax-collection system or economic troubles may mean that it fails to pro￾vide its share of the budget. That puts USAID in a quandary. Does it cut back or stop funding, and lose the momentum and development benefits, or does it try to negotiate a revised approach with the government. USAID had never been able to develop universal guidance to deal with this problem. 4. Fiduciary capabilities. A developing country needs to use funds effectively and be able to account for all of the money provided by the donor. In par￾ticular, donors are concerned about corruption. Large cash transfers to a government budget can be risky if the host government has weak accounting and performance standards. Embezzlement is the most visible. But there can also be a large discrepan￾cy between budgeted amounts and actual expendi￾tures — where funds mysteriously evaporate on the long bureaucratic journey to a rural classroom or health clinic. And finally, there is the quality of expenditure results. What if teachers rarely show up to teach, roads are poorly constructed and health centers lack drugs and trained staff. The next sec￾tion looks at measurement of results. 5. Measuring results and impact. With basic recordkeeping, aid projects can demonstrate bene￾fits. A project can show that so many children were vaccinated; students graduated; or so many miles of roads were built and annually maintained. In con￾trast, even with a good monitoring system, with a policy reform program, it is difficult to measure the overall level of impact let alone beneficiary benefits. The linkage or causality of aid and results is difficult to establish. Did the change in credit policies gener￾ate increased agricultural output or was it the avail￾ability of inputs, weather, or changes in commodity prices? The 1991 CDIE Africa evaluation found that: “Policy reform is a continuing process requiring constant monitoring and adjustment as unexpected effects of each new policy emerge. Developing countries need to develop their own internal capaci￾ty to analyze and set their own policy reform agen￾da. A.I.D. technical assistance can help develop that capacity. The political-economy of policy reform is critical to success. Since reforms are designed to 36 PPC Evaluation Working Paper No. 18 change institutions and structural relationships, it is important to have a monitoring system that will demonstrate the benefits of reform while also help￾ing reformers respond to complaints and arguments that come from those who are hurt.” In most countries monitoring and evaluation capa￾bility was weak, particularly when it had to link pol￾icy changes to beneficiary impact. That meant that both the country and donors failed to fully under￾stand the problems and were often unable to meas￾ure success or make adjustments to correct prob￾lems. There were evaluations of programs that included a successful monitoring effort. The 1997 Zimbabwe grain marketing reform program noted that sustainability was supported by institutionaliz￾ing monitoring capability within the recipient coun￾try. A 1995 evaluation of child survival in the Philippines found that: “Service-delivery indicators need to be selected with care and investments made at the start of the program to obtain accurate base￾line measures that are comparable and reliable.” 6. Demonstrating results to Congress. USAID’s 1980s Development Fund for Africa used cash transfers to support policy and institutional reforms. It was a major departure from the traditional project approach. The first step was convincing Congress that monitoring and evaluation systems at the sector or macro level would be able to provide indicators of performance. Congress accepted the approach. However, not everyone was completely willing to support this approach. Many congressional staffers were used to projects and the types of results that projects produced. Macro and sector level impacts could be measured but it was difficult to provide tracking to the grass￾roots level. In several countries there were many donors and it was impossible to assess a single actor’s contribution. Congress wanted to know how many people were benefiting from the changes and how they benefit￾ed. They wanted a link to U.S. assistance. Initially they asked for a few procurement, implementation and results measures. After a few years, concern about benefits increased. Congress became steadily more demanding, asking for a detailed accounting of results. Within a few years cash transfer programs became more and more like regular USAID proj￾ects. The lesson seems clear. If USAID is to provide cash transfers it needs to develop an agreement early-on with Congress on the types of indicators that can be provided to show results. If Congress wants to see micro-level changes and impact linked to U.S. assistance, projects may be the preferred route. 7. Supporting technical assistance and training. Reform is most successful in countries that already embrace policy reforms and where there is a well￾developed institutional capability for analysis. In countries facing human resource constraints and dysfunctional bureaucratic procedures and systems, program assistance stands little chance of success. Programs failed where governments lacked the insti￾tutional capacity to design and implement pro￾grams. Where the country had the will but lacked the skills, technical assistance often solved the design and implementation problems. However, sustainability could still be a problem if indigenous analytical capabilities were not institutionalized. Evaluations repeatedly found that the amount of technical assistance needed to assure successful poli￾cy and institutional reforms was directly related to the capacity of the host government to design and implement reform programs. Programs that recog￾nized the need for TA and training were much more successful. A 1999 evaluation of USAID’s Zimbabwe Basic Education and Skills Training sec￾tor assistance found that significant long-term train￾ing and staff development (over a decade) helped assure that policy reforms and improved govern￾ment capabilities were successfully adopted. A 1997 sub-Saharan Africa girl’s education assessment urged other programs not to skimp on technical assistance. It also recommended assisting governments to pur￾sue research at the design stage of new programs — rather than waiting until the programs were nearly finished. In a 1996 review of USAID non-project assistance in Benin, Ghana, Malawi, Uganda and Zimbabwe the evaluation stated: “Although the cen￾terpiece of sectoral assistance is a generalized resource transfer, USAID’s programs in Africa typi￾General Budget Support: An Alternative Assistance Approach 37 cally include corresponding technical assistance and training (capacity building) components. The highly effective use of non-project and project instruments addresses the very issue of sustainable development.” There can be a problem of dependency. When tech￾nical assistance ends, it is important to make sure that skills have been transferred. The 1991 CDIE evaluation of policy reform in Africa found that: “Programs that developed the internal policy reform capacity of the host country helped build long-run policy reform sustainability.” Technical assistance by itself is not enough. In the case of Uganda, the Export Policy Analysis and Development Unit became dependent upon a long-term policy advisor. At the end of his assignment he left the country. His skills were not transferred and the unit was not able to carry on. 8. Policy reform performance and conditionality. At the start of the policy reform process the govern￾ment and donors need to mutually agree upon per￾formance conditions that reflect the policy frame￾work and country strategy. An Africa girl’s educa￾tion assessment urged: “…donors to use very specif￾ic language to define performance criteria. Vague or multi-targeted approaches just create problems.” While stressing the need for specific performance measures, it also called for flexibility. It recommend￾ed using a “letter of intent” approach that sets con￾ditions on a rolling basis rather than establishing them at the design stage for a five-year period. Economic and political conditions change and USAID needs to avoid locking in conditions that may be inappropriate in just a few years. A 1996 CDIE evaluation of world-wide investments in agri￾culture recommended against using non-project assis￾tance to force policy change: “USAID should provide non-project assistance to support economic policy reform only in countries where it will be used to facilitate a policy reform processes already under￾way.” A 1995 basic education evaluation found that “Conditioned budgetary support has leveraged sig￾nificant changes in resource distribution and basic education policy. But it is questionable whether the new changes are sustainable without outside assis￾tance and pressure.” A 1995 child survival evalua￾tion in the Philippines found that “Performance￾based non-project assistance can achieve health sec￾tor goals in the Philippines since there is broad donor-host government consensus on needed reforms, sincere commitment to implementing these reforms and a highly capable and trained imple￾menting agency.” 38 PPC Evaluation Working Paper No. 18 Questions for Donors General Questions: Please describe the assistance you give to (country name). What share is GBS/sectoral/project assistance? Why was this method (or combination of methods) of deliv￾ery chosen? General Budget Support Questions: Development Policy and Planning How can general budget support increase country ownership? What is the best way to increase country ownership? How do you measure country commitment? How do you ensure that critical development issues (child survival, HIV/AIDS, education, democracy and governance, civil society, private sector development, trade, investment, etc.) are included in government plans? Are economic growth and private sector development (includ￾ing investment and trade) included in government programs? If yes: Are they effective? Do you press for policy reforms when providing general budget support? If yes: How do you work with other donors to influence policy reform? How do you work with the government to influence policy reform? What policies do you target? How do you select policy reforms? How do you press for these reforms? How do you monitor progress towards the reforms? What happens if the reforms are not achieved? What is the role of technical assistance and training in mak￾ing GBS more effective? How do you and the host government determine what assistance is needed? Is sector or project assistance more appropriate? Why? Under what conditions? Finance How are you assured that your general budget support is additional, and not in place of existing government funding? How timely and predictable is government (and donor fund￾ing) under GBS? What has happened to military and other non-development funding? Does GBS allow the government to divert its own funds to non-development uses? What financial skills do the host government officials need? What minimum level of audit and financial control are neces￾sary? What methods have proven most effective at improving finan￾cial transparency? What indicators of corruption and misappropriation of funds do you use to assess performance? Development Management Does GBS require more or less staff time than project or sec￾tor assistance (from both donor and government perspective)? Does GBS reduce donor and government transactions costs? What are the necessary host country technical skills for staffing development programs in a wide range of technical fields (agriculture, health, HIV/AIDS, education, etc.)? How do you measure and assess capabilities? Beyond individual staff and skills, do government organiza￾tions have other problems that hamper their capacity to effec￾tively deliver services to the people? General Budget Support: An Alternative Assistance Approach 39 Annex 6. General Budget Support Interview Protocols If yes: Please describe the problems. What alternative approaches should donors pursue to address the problems? (For example, use local NGOs or U.S. PVOs and contractors?) How do you measure general budget support performance and beneficiary impact? What monitoring and evaluation systems are required? How do you measure the impact on poverty? How do you demonstrate results to Congress (or parliament)? What performance standards are appropriate for each devel￾opment sector? Do you involve local organizations in monitoring government performance? If a country is not yet ready for general budget support, would sector assistance or other approaches provide an inter￾im step? What direction do you see aid delivery taking in the future? Sector Support or Project Aid Questions: Why do you provide this type of assistance instead of (or in addition to) general budget support? Are you moving towards providing general budget support in (country name)? If yes: Please describe what needs to be accomplished before general budget support is granted. What is your time frame for the transition to general budget support? If no: Why not? What is your opinion of the movement towards providing general budget support in international aid? What are the advantages of general budget support? What are the drawbacks of general budget support? What direction do you see aid delivery taking in the future? What happens to donors that do not provide budget support? Do they really miss out on policy discussions? Do they really miss out on donor coordination? Are there benefits for both the donor and country of not providing budget support? Questions for the Government General Questions: Please describe the assistance you receive from bilateral and multilateral development agencies. What share is GBS/sectoral/project assistance? What kind(s) of assistance do you prefer? Why? General Budget Support Questions: Development Policy and Planning Does general budget support increase your ownership over your country’s development process? If yes: Why/How? Is GBS the best way to increase country ownership? If no: Why not? What is the best way to increase country ownership? Which development issues (such as child survival, HIV/AIDS, education, democracy and governance, civil soci￾ety, private sector development, trade, investment, etc.) do you include in your development plans? If economic growth and private sector development (including investment and trade) are included: Are they effective? Do donors press for policy reforms when providing general budget support? If yes: What policies are targeted? How do you select policy reforms? How do you monitor progress towards the reforms? What happens if the reforms are not achieved? Do you receive technical assistance and training along with general budget support? 40 PPC Evaluation Working Paper No. 18 If yes: What kind? How do you and the donors determine what assistance is needed? Is sector or project assistance more appropriate? Why? Under what conditions? Finance How do you make sure that general budget support increases and does not replace existing government development fund￾ing? What has happened to military and other non-development funding? Is there a problem? How dependable or timely is donor (and government fund￾ing) under GBS? What has happened to military and other non-development funding? What financial skills does your government need to adminis￾ter general budget support? Are there areas where your financial skills need to be strengthened? What level of audit and financial control are required by donors for general budget support? Do you think this is necessary? Should there be more or less auditing and financial con￾trol? What methods have proven most effective at improving your government’s financial transparency? How do you monitor corruption and misappropriation of funds? Do donors have any requirements on monitoring cor￾ruption? Development Management Does GBS require more or less government staff time than project or sector assistance? Does GBS reduce government (and donor) transactions costs? What technical skills do you feel you need to staff develop￾ment programs in a wide range of technical fields (agriculture, health, HIV/AIDS, education, etc.)? How do you measure and assess capabilities? Are there any areas where your technical skills need to be strengthened? Beyond individual staff and skills, are there other problems that hamper your government’s capacity to effectively deliver services to the people? If yes: Please describe the problems. What should your government do to address these prob￾lems? What should donors do to address the problems? How do you measure general budget support performance and beneficiary impact? Do any of your monitoring and evaluation systems need to be strengthened? How do you measure the impact on poverty? What performance standards are appropriate for each devel￾opment sector? Do you involve local organizations in monitoring government performance? Are sector assistance or other assistance approaches, such as project aid, good interim steps before receiving general budget support? Sector Support or Project Aid Questions: Why do you receive this type of assistance instead of (or in addition to) general budget support? Are you moving towards only receiving general budget sup￾port from all donors? If yes: Please describe what needs to be accomplished to achieve this goal. What is your time frame for the transition to general budget support? If none : Why not? What is your opinion of the movement towards providing general budget support in international aid? What are the advantages of general budget support? What are the drawbacks of general budget support? What direction do you see aid delivery taking in the future? General Budget Support: An Alternative Assistance Approach 41 This Evaluation Working Paper can be ordered from USAID’s Development Experience Clearinghouse (DEC). To download or order publications, go to www.dec.org and enter the document identification number in the search box. The DEC may also be contacted at 8403 Colesville Rd, Ste 210, Silver Spring, MD 20910; tel 301-562-0641; fax 301-588-7787; email docorder@dec.cdie.org. For more information, contact U.S. Agency for International Development Washington, D.C. 20523-1000 Telephone: 202-712-4810 Internet: www.usaid.gov PN-ACW-878