Burundi Country Econ Policy

Public Disclosure Authorized
Report No. 51880-BI
Report No. 51880-BI

Republic of Burundi
Country Economic Memorandum (CEM)
The Challenge of Achieving Stable and Shared Growth
March 2011

Poverty Reduction and Economic Management 3
Africa Region
Republic of Burundi

Country Economic Memorandum (CEM)

Document of the World Bank, Co-produced with the Government of Burundi,
the African Development Bank and the Department for International Development (UK)
Currency Equivalents
Exchange Rate Effective as of February 28, 2011
Currency Unit Burundi Franc

USS 1.00 FBu 1,230
Government Fiscal Year

January 1December 31
Weights and Measures

Metric System
Abbreviations and Acronyms
AfDB African Development Bank

BNDE Banque Nationale de Dveloppement conomique
BRB Banque de la Rpublique du Burundi
CEM Country Economic Memorandum
CET Common External Tariff
COMEBU Comptoirs Miniers des Exploitations du Burundi
COMESA Common Market for Eastern and Southern Africa
COMTRADE U.N. Commodity Trade Statistics Database
CPAF Common Performance Assessment Framework
DAC Development Assistance Committee
DDR Disarmament, Demobilization and Reintegration
DfID Department for International Development (UK)
DPAF Donor Performance Assessment Framework
EAC East African Community
EDPRS Economic Development Poverty Reduction Strategy
EITI Extractive Industries Transparency Initiative
EPA Economic Partnership Agreement
EU European Union
FDI Foreign Direct Investment
FSAP Financial Sector Assessment Program
GDP Gross Domestic Product
GWh Gigawatt Hours
HI Herfindahl Index
HIPC Heavily Indebted Poor Countries
HIV/AIDS Human Immune Deficiency Virus/Acquired Immune Deficiency Syndrome
JBSR Joint Budget Support Review
MDG Millennium Development Goal
ii
MEACA Ministry of East African Community Affairs
MFI Microfinance Institution
MTEF Medium-Term Expenditure Framework
NTB Nontariff Barrier
OCIBU Office du Caf du Burundi
OECD Organisation for Economic Co-operation and Development
PAGE Project dAppui la Gestion Economique
PEMFAR Public Expenditure Management and Financial Accountability Review
PPP Public Private Partnership
PRSP Poverty Reduction Strategy Paper
RCA Revealed Comparative Advantage
REGIDESO Rgie de Production et de Distribution d'Eau et d'Electricit
SITC Standard International Trade Classification
SODECO Socit de Dparchage et de Conditionnement
SOGESTAL Socit de Gestion des Stations de Lavage
TFP Total Factor Productivity
VAT Value-Added Tax
Vice President: Obiageli Katryn Ezekwesili

Country Director: John Murray McIntire
Country Manager: Mercy Miyang Tembon
Sector Director: Marcelo Giugale
Sector Manager: Jan Walliser
Task Team Leader: Hannah Nielsen
iii
Table of Contents
ACKNOWLEDGMENTS ................................................................................................................................ vii
EXECUTIVE SUMMARY ............................................................................................................................. viii
INTRODUCTION ............................................................................................................................................ 1
CHAPTER 1 MACROECONOMIC AND FISCAL DEVELOPMENTS........................................................... 2

A. Economic Growth: Historical Perspective ....................................................................................... 2
B. Burundis Public Finances: An Overview ..................................................................................... 11
CHAPTER 2 OBSTACLES TO ECONOMIC GROWTH AND EMERGING OPPORTUNITIES FOR

BROADENING THE BASE FOR GROWTH ......................................................................... 15
A. Obstacles to Economic Growth ..................................................................................................... 15
B. Opportunities in Potential Growth Sectors .................................................................................... 32
CHAPTER 3 THE GOVERNMENTS OPTION TO TACKLE OBSTACLES AND REAP OPPORTUNITIES

FOR ECONOMIC GROWTH ............................................................................................... 52
A. An Infrastructure Action Plan ........................................................................................................ 52

B. Maximizing the Benefits of Burundis EAC Membership ............................................................ 59
C. Reforms Needed to Increase the Private Sectors Contribution to Growth ................................... 70
D. Encouraging Emerging Growth Sectors ........................................................................................ 74
E. Summary of Priority Recommendations ........................................................................................ 81
CHAPTER 4 IMPLICATIONS FOR THE MACRO-FISCAL SITUATION................................................... 84

A. Economic Prospects, Resource Allocation, and Achieving the MDGs ......................................... 85
B. How Can Aid Contribute to Increased Growth? ............................................................................ 91
C. Conclusion ..................................................................................................................................... 96
APPENDIX .......................................................................................................................................... 97
Appendix 1: Action Plan .................................................................................................................... 98
Appendix 2: Export Diversification ................................................................................................. 101
Appendix 3: Regional Integration .................................................................................................... 105
Appendix 4: Infrastructure Action PlanTables and Maps .............................................................. 109
Appendix 5: Infrastructure Action PlanMain Components ............................................................ 113
Appendix 6: Business Environment ................................................................................................. 118
Appendix 7: Map of Burundi ........................................................................................................... 119
REFERENCES ........................................................................................................................................ 120
iv
List of Boxes
Box 1.1: Rwandas Growth Experience and Possible Lessons for Burundi ....................................... 10
Box 2.1: The Transformation of Ugandas ExportsA Success Story.............................................. 21
Box 2.2: Burundis Tax Regime ......................................................................................................... 26
Box 2.3: Land Tenure and Economic Growth .................................................................................... 35
Box 3.1: Proposed Measures to Remove Burundi NTBs within the EAC.......................................... 63
Box 4.1: The Common Performance Assessment Framework in Rwanda ......................................... 95
Box A.1: Potential Benefits, Costs, and Guiding Principles of Regional Integration ....................... 106
List of Figures
Figure ES.1: Real GDP Growth and GDP Per Capita, 19622008 .......................................................... viii
Figure ES.2: Composition of Public Spending, 200108......................................................................... xiv
Figure 1.1: Per Capita GDP in Burundi and Rwanda, 19622008 ........................................................... 2
Figure 1.2: Real GDP Growth and GDP Per Capita, 19622008 ............................................................. 3
Figure 1.3: The Cost Of War: Potential and Actual GDP Per Capita ....................................................... 4
Figure 1.4: Sectoral Contributions to Real GDP Growth, 19972008 ..................................................... 5
Figure 1.5: Private and Public Investment as a Percent of GDP, 19972008 .......................................... 6
Figure 1.6: Annual Growth of GDP, Capital Stock, and Productivity Per Worker, 19902008 .............. 7
Figure 1.7: Real GDP Growth during Post-conflict Periods in Burundi, Rwanda, and Sierra Leone ...... 8
Figure 1.8: Composition of Public Spending, 200108.......................................................................... 12
Figure 2.1: HI for EAC Member-Countries ........................................................................................... 19
Figure 2.2: HI and GDP Per Capita ........................................................................................................ 19
Figure 2.3: Composition of Exports in EAC Countries, 198589 and 200506 .................................... 20
Figure 2.4: EXPY of EAC Countries ..................................................................................................... 22
Figure 2.5: Governance Indicators, 1998, 2000, 2003, and 2008 ........................................................... 28
Figure 2.6: Population Projections, 200550 ......................................................................................... 31
Figure 2.7: Proportion of Individuals Living in Households with a Daily Caloric Intake per Adult
Below 1,900 kcal.................................................................................................................. 33
Figure 2.8: Food Crop Production, 19902007 ...................................................................................... 34
Figure 2.9: Farm-Gate Prices for EAC Members, 19772008 ............................................................... 36
Figure 2.10: Burundi Coffee Production, 19772008 .............................................................................. 37
Figure 2.11: Medium-Contribution ScenarioGovernment Revenue..................................................... 51
Figure 2.12: High-Contribution ScenarioGovernment Revenue .......................................................... 51
Figure 3.1: Funding Arrangements for the Core Infrastructure Program ............................................... 53
Figure 4.1: Selected MDG Trends under the Baseline Scenario, 200615 ............................................ 90
Figure 4.2: Selected MDG Trends under the Alternative Scenario, 200615 ........................................ 90
Figure 4.3: Allocation of Aid by Impact Category, 200107 ................................................................. 92
Figure 4.4: Disbursement Ratio and Commitment-Disbursement Gap, 200107 .................................. 93
Figure 4.5: Exchange Rate and Price Developments, 200108 .............................................................. 94
Figure A.1: Product Space Map of Burundi .......................................................................................... 101
Figure A.2: Transport Corridors for the East African Community ....................................................... 111
Figure A.3: Railway Network of the East African Community ............................................................ 112
List of Tables
Table ES.1: Priority Recommendations ................................................................................................... xv
Table 1.1: Selected Economic Indicators, 19622008 ............................................................................ 3
Table 1.2: Supply-Side Components, 19972008 ................................................................................... 5
Table 1.3: Demand-Side Components, 19972008................................................................................. 6
Table 1.4: Comparison of Selected Indicators Six Years Post-conflict .................................................. 9
v
Table 1.5: Composition of Government Revenue, 200108 ................................................................. 11
Table 1.6: Central Government Fiscal Deficit, 200108 ...................................................................... 12
Table 2.1: Selected Indicators Comparing Burundi to Sub-Saharan Africa and the World Economy . 15
Table 2.2: Classification Summary of Burundis Exports..................................................................... 23
Table 2.3: Sectoral Distribution of Bank Credit.................................................................................... 25
Table 2.4: Annual Projections with One Large-Scale Nickel or Gold Mine ......................................... 47
Table 2.5: Comparison of Annual Official Exports and Survey Production, 2008 ............................... 49
Table 3.1: Selected Indicators for the Base Case Scenario ................................................................... 55
Table 3.2: Key Outcomes for the Base Case and Alternative Scenarios ............................................... 56
Table 3.3: Selected Socioeconomic Indicators for EAC Members, 2008 ............................................. 60
Table 3.4: External Tariffs in the EAC, 200008 ................................................................................. 60
Table 3.5: Similarity and Complementarity Indexes for Burundi, 19952007 Averages ..................... 65
Table 3.6: Specific Measures to Improve the Performance of the Food Crops Sector ......................... 75
Table 4.1: Baseline ScenarioProjections for the Main Economic Variables, 200915 ..................... 85
Table 4.2: Alternative ScenarioProjections for the Main Economic Variables, 200915................. 86
Table 4.3: Baseline ScenarioActual, Estimated, and Projected Government Finance, 200215 ...... 87
Table 4.4: Baseline ScenarioMedium-Term Expenditure Framework, 200715 .............................. 88
Table 4.5: Alternative ScenarioActual, Estimated, and Projected Government Finance, 200215 .. 89
Table 4.6: Alternative ScenarioMedium-Term Expenditure Framework, 200715.......................... 89
Table A.1: Action Plan Priority Recommendations for the Time Period 201015................................ 98
Table A.2: Detailed Classification of Burundis Exports..................................................................... 102
Table A.3: Horticulture: Markets, Constraints, and Opportunities for Burundi .................................. 103
Table A.4: Recommendations for the Artisanal and Small-Scale Mining Sector ................................ 104
Table A.5: Burundis Post-Independence Experience with Regional Integration Agreements .......... 105
Table A.6: Matrix of Strategic Objectives and Indicative Outputs for the MEACA ........................... 107
Table A.7: Policy Matrix for Advancing Regional Integration in Burundi ......................................... 108
Table A.8: Basic Infrastructure Coverage for the East African Community, 2006 ............................. 109
Table A.9: Service Costs and Difficulties in the East African Community ......................................... 109
Table A.10: Description of Various Scenarios Considered in the Report.............................................. 110
Table A.11: Development Expenditures on the Core Program .............................................................. 116
Table A.12: Routine Maintenance Expenditures on the Core Program ................................................. 116
Table A.13: Recommendations for Improving Financial Sector Performance ...................................... 118
vi
ACKNOWLEDGMENTS
This Burundi Country Economic Memorandum (CEM) is the first since 1984. The CEM reviews
developments in Burundi over the past 10 years and identifies the tightest constraints to economic growth.
It draws on studies prepared by the government of Burundi, the World Bank, other partners, and
academics. After synthesizing recommendations from those studies and from original analysis, the CEM
presents a strategy to promote growth, reduce poverty, and improve peoples lives.
The Burundi Country Economic Memorandum (CEM) is a joint report from the World Bank, the
government of Burundi, the African Development Bank, and the U.K. Department for International
Development. The team would like to thank Her Excellency Madame Clotilde Nizigama (Minister of
Economy, Finance, and Development Cooperation, Burundi) and the government steering committee for
the CEM for their excellent collaboration throughout the preparation of the CEM. The report was
prepared under the overall leadership and guidance of Jan Walliser (Sector Manager, AFTP3) and Eric
Bell (Lead Economist, AFTP3), who provided invaluable support, input, and advice during the
preparation of the memorandum. The team also greatly benefited from the support and guidance of John
Murray McIntire (Country Director, AFCE1), Mercy Miyang Tembon (Country Manager, Burundi),
Kathryn Funk (Country Program Coordinator, AFCTZ), and Steffi Stallmeister (Senior Country Officer,
AFCTZ). The former task team leader Dorsati Madani (Senior Economist, CICSA) conceptualized the
CEM, established the core team, launched parts of the analytical work and led early discussions with the
counterparts. The main author and task team leader of the report is Hannah Nielsen (Economist, AFTP3).
The core CEM team includes Eric Mabushi (Economist, AFTP3), Jean-Pascal Nganou (Economist,
AFTP3), Edgardo Favaro (Lead Economist, PRMED), Henry Mooney (Economist, PRMED), Christian
Lim (Economist, African Development Bank), Chiara Selvetti (Economist, DfID), Tania Rajadel
(Consultant, AFTP3), Nicaise Ehoue (Senior Agriculture Economist, AFTAR), Remi Pelon (Mining
Specialist, COCPO), Gilbert Midende (Consultant, AFTP3), John May (Lead Population Specialist,
AFTHE), Caroline Freund (Lead Economist, DECRG), Andre Ryba (Consultant, AFTFW), Aurelien
Beko (Consultant, AFTP3), and Daniel Benitez (Senior Economist, FEUSE). Valuable contributions were
also provided by Kene Ezemenari (Senior Economist, AFTP3), Vandana Chandra (Senior Economist,
DECOS), Israel Osorio-Rodarte (Consultant, DECPG), Susana Carrillo (Senior Governance Specialist,
WBIGV), Craig Andrews (Lead Mining Specialist, COCPO), Charles N'cho-Oguie (Consultant, AFTP3)
and Kalamogo Coulibaly (Consultant, AFTP3). Additionally, the team appreciates very useful comments
from Lev Freinkman (Lead Economist, AFTP3).
The core team worked closely with a steering committee consisting of representatives of various
ministries. The counterpart team included the following members: Domitien Ndihokubwayo, Juvnal
Bumviye, Alexis Bizimungu, Emile Sinzumunsi, Melchior Barantandikiye, Mireille Bizimana, Rose
Kamariza, Adle Mbonankira, Cyriaque Miburo, Willy Ntamagara, Rverien Nivyayo, Lopold
Bizindavyi, Terence Ntabangana, Hilaire Ntakiyica, Fidle Gahungu, Charles Rugema, Bonaventure Sota
and Pierre-Claver Rurakamvye. Additional workshop participants included Lopold Nkunzimana, Gratien
Ninteretse, Pierre Bayihishako, Anaclet Birushabagabo, Balthazar Nakumana, Florence Nshimirimana,
Jean-Bosco Batungwanayo, Jean Mugishawimana, Consolate Musamisomi, Nestor Niyungeko, Cyprien
Gicemzi, and M. Matsebahene.
The CEM benefited immensely from very helpful comments by the peer reviewers for the overall report,
Hinh Dinh (Lead Economist, DECOS) and Peter Moll (Senior Economist, OPCCE); the reviewers for the
regional integration component, Diep Nguyen-Van Houtte (Senior Operations Officer, AFCRI) and Paul
Brenton (Lead Economist, AFTPM). Comments and suggestions from outside the World Bank are also
gratefully acknowledged, in particular, from the Burundi International Monetary Fund team.
The team would also like to extend its gratitude to Mariama Daifour Ba (Program Assistant, AFTP3) and
Aurore Simbananiye (Team Assistant, AFTBI) for logistics support. Resources received from the Belgian
Poverty Reduction Partnership Trust Fund to finance a background study on artisanal and small-scale
mining in Burundi are also greatly appreciated.
vii
EXECUTIVE SUMMARY
1. Burundis growth policy should focus on: (1) closing the infrastructure gap; (2) gaining from
regional integration; (3) improving the business environment; (4) promoting new growth sources; and (5)
strengthening its fiscal position through improved revenue mobilization, more efficient spending, better
public financial management, and more effective aid.
BURUNDIS ECONOMY YESTERDAY AND TODAY
2. Burundi is very poor and has suffered from years of conflict. The 1984 World Bank CEM
said:
Burundi is among the poorest countries in the world; [for a population of 4.2 million] per capita
GNP is estimated at about US$240 (1983); three-fourths of the adult population is illiterate; infant
mortality is still high, despite considerable progress registered in the last twenty years; and access
to potable water and electricity is limited to the urban centers -- 5% of the population.
3. The situation has worsened since 1984. The population is now around 8.4 million. GDP per
capita is about US$110 (constant 2000 prices); a significant fall from the level estimated 25 years ago,
making Burundi the second-poorest country in the world. Starting from an already low developmental
level, Burundi was ravaged by a civil war from 1993 to 2005. War killed some 300,000 Burundians,
displaced many others, destroyed capital, repressed investment, and damaged the States capacity to
provide basic health, education, water and electricity. A simple estimate of the cost of the war indicates
that, without the conflict, Burundis GDP per capita would be about double its current level.
4. A modest recovery has started. FigureES.1:RealGDPGrowthandGDPperCapita,19622008

Real annual GDP growth has stabilized at
3.5 percent since 2005, though this rate is 25% 180
below the average of the East African 20%

160
Community (EAC). Strengthened 140
macroeconomic policies have added 15%
120
Constant2000US$
modestly to growth since 2005, although
10% 100
the global crisis lowered growth in 2009.
Monetary policy is geared toward 5% 80
stabilizing prices and succeeded in 60
keeping inflation on average in the single 0%
40
1962
1964
1966
1968
1970
1972
1974
1976
1978
1980
1982
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
digits between 2005 and 2007. Inflation,

5%
however, increased substantially in 2008 20
because of food and fuel price shocks 10% RealGDPgrowthrate(lefthandside) 0
before returning to estimated 10.5 percent GDPpercapita(constant2000US$,righthandside)

in 2009. Tax revenue has stabilized Source:WorldDevelopmentIndicators.
between 17 and 18 percent of GDP,
despite the narrow tax base and a reduction of the petroleum tax. Another indication of political
stabilization and recovery of the country is the return of more than half a million refugees since 2002.
5. Burundi is not likely to meet many of the Millennium Development Goals by 2015. Burundi
lags behind the world and East Africa on all measures of health and human capital. An estimated 67
percent of Burundians live below the poverty line. The country ranked toward the bottom of the Human
Development Index in 2008. Lack of education among adults and children limits productivity and growth.
viii
Although primary-school enrollment has improved significantly in recent years, completion rates are low,
and secondary and tertiary enrollment rates have not grown apace with primary completion.
6. Compared to other post-conflict countries, however, Burundi has not experienced the
typical growth spurt. As section A3 in chapter 1 shows, Burundis growth in the post-conflict period
does not compare well with other post-conflict countries, such as Rwanda and Sierra Leone. This is
primarily because of the long conflict that continued even after the official peace agreement of 2000.
Investment, output and productivity growth have been less than in comparable post-conflict countries
because Burundians lack confidence in their countrys economic and political stability.
7. The government has begun reforms. The Government has begun some reforms addressing, for
example, coffee liberalization, the business environment, and public financial management. Key reforms
are the ongoing liberalization of coffee, a new investment code, revision of the mining code, the adoption
of a new privatization law, the establishment of the new Investment Promotion Agency (API) and better
public financial management based on a new organic budget law. Critical next steps are adopting the
revised mining code, adopting the revised commercial code, and finalizing the revision of the tax code.
Regarding public finance reforms, the size of the wage bill remains a concern.
WHAT BLOCKS ECONOMIC GROWTH IN BURUNDI?
8. Burundi is mainly rural and subsistence agriculture dominates. Agriculture contributes about
half of GDP and employs approximately 90 percent of the labor force, as detailed in section B1, chapter
2. Poverty is widespread in rural areas. Agricultural productivity is low and has barely improved since
independence. Given high population growth, low productivity gains have led to a decline in production
per capita. Malnutrition is common and Burundi often requires food aid.
9. Burundis exports are undiversified, consisting mainly of low-value, unprocessed primary

products. Export concentration on a few agricultural exports makes Burundi vulnerable to external price
or quantity shocks and to bad weather. Section A4 of chapter 2 demonstrates that, compared to most of
the other EAC members as well as some countries that previously exported mostly coffee, Burundi has
not diversified its exports and this fact has made growth more volatile.
10. The expansion of Burundis main export, coffee, faces severe challenges. Some 600,000
800,000 households (perhaps one-third of the population) grow coffee. Though Burundi has the
agricultural conditions to produce high-quality, high-value coffee, the sector underperforms with
declining production and quality partly because of official prices that do not stimulate production (see
section B2, chapter 2). Unlike Rwanda, Burundi has been slow to match the latest developments, such as
the increased importance of specialty coffees or the marketing of high-quality coffee through direct sales
instead of auctions. Government policy has not always been beneficial to Burundian coffee growers. The
ongoing reform, especially the privatization of the coffee washing stations, is a positive step.
11. Lack of infrastructure raises the costs of Burundis isolation. The poor coverage and state of
infrastructure create costs in time and money that lower the return to work, discourage domestic and
foreign investment, and constrain economic growth, as outlined in section A2, chapter 2. High transport
costs, caused by absence of infrastructure, hinder internal trade and reduce Burundis trade opportunities
with East Africa and the world beyond. The internal obstacles are compounded by the high costs in the
ports of Dar es Salaam (Tanzania) and Mombasa (Kenya), through which Burundi trades. Modern
electricity, which is available only to 2 percent of households, is expensive.
12. The private contribution to growth has been too small. The public sectors share in investment
has been too high, averaging 74 percent of total investment between 2001 and 2004. The main reason is
ix
that Burundi has an unfavorable business environment on top of little infrastructure, one that discourages
investment, activity diversification, and job creation. Domestic private investment has only recently
started to pick up and remains low, averaging about 8 percent of GDP between 2005 and 2008 and almost
half of total investment. Foreign direct investment is less than 1 percent of GDP.
13. Finance is a major obstacle for formal and informal business. Only 2 percent of Burundians
have bank accounts and only 4 percent are members of microfinance institutions (MFIs). Financial
products are undiversified and do not cater adequately to the rural population or small and medium-size
enterprises (see section A5, chapter 2). Given the lack of access to commercial banking, microfinance has
an important role to play; however, nearly 80 percent of members/customers of MFIs are wage earners.
The industry, therefore, excludes a majority of the population, namely non-wage earners, while serving a
market that is uncertain for it in the medium term as it can also easily be served by the banking industry.
14. Burundis population is expected to continue to grow rapidly in the next 40 years, as
illustrated in section A8, chapter 2. Its high population growth rate (2.6 percent in 2005) exacerbates
many of the problems of a country that already has one of the highest population densities in Africa. The
agriculture sector alone cannot support the growing needs of the population, underscoring the need to
create urban and rural non-farm employment.
15. Poor governance and limited institutional capacity imply lower productivity for all sectors
of the economy and inhibit new investment in physical and human capital. Given the importance of
government spending as a share of GDP in the post-conflict economy, increasing the efficiency of that
spending is vital. Section A6 in chapter 2 describes how poor governance is currently an obstacle to more
efficient public spending. In addition, the governments capacity to adopt quick and effective policy
decisions is weak. This fosters a perception that recovery is weak and causes delays in private sector
involvement.
HOW CAN PUBLIC ACTIONS INCREASE ECONOMIC GROWTH?
16. The CEM has identified priorities for public actions to promote growth and to make it
more equitable. Table ES.1 presents the key recommendations which are linked to the relevant text of
the chapters of the main report. Table A.1 in Appendix 1 includes the agency responsible for the
implementation of the recommendation.
17. Burundis growth policy should focus on: (1) closing the infrastructure gap; (2) gaining from
regional integration; (3) improving the business environment; (4) promoting new growth sources; and (5)
strengthening the fiscal position through improved revenue mobilization, more efficient spending, better
public financial management, and more effective aid.
18. Peace and stability are indispensable for growth but will not produce a dividend without
proper management by the Government. The Burundian Government has a larger than usual role in
economic development because it must make up the investment lost during the conflicts and because the
private sector is still weak. The governments main job is to keep the peace and maintain political
stability, without which public and private investment cannot grow. However, the Government should
also realize the peace dividend and benefit from reduced and redirected defense spending. The
Government therefore faces the political and fiscal choice between spending on security and the need to
invest more in infrastructure and in human capital. To the extent possible, and keeping in mind the need
for security and political stability, it is essential to reduce the share of defense and security expenditures
as soon as possible to create fiscal space for increased spending in the priority economic and social
sectors.
x
19. As part of a policy for stability for economic growth, Burundi must lower its population
growth rate. Lower fertility rates would affect most development areas positively; and would particularly
facilitate the formation of human capital, help reduce poverty levels and land shortages, relieve pressure
on the environment, and improve the health of women and children.
Closing the Infrastructure Gap
20. An infrastructure action plan lays out a detailed strategy for improving the infrastructure in
Burundi, taking into account the regional context1. It focuses on the power, transport, and
communications sectors; and proposes actions that need to be taken, their timing, and a possible funding
structure. The action plan could have a significant positive impact on the whole economy, leading to
sustained growth, business, and employment opportunities; more and cheaper infrastructure; and
increased tax revenues. The main features, objectives, and the economic impact of the action plan are
described in chapter 3, section A.
21. Complementary policies to investment in infrastructure are critical to getting the full return
on private investment. Priority interventions for the short and medium term are presented in section A5,
chapter 3. Specific key recommendations are:
Electricity: Electricity supply must be improved by using all possible energy resources, with
more provinces connected to the network and REGIDESO financially restructured.
Transport: Given the geographic isolation of Burundi, the country could greatly benefit from
rehabilitating the national highways and the provincial and community roads. Strengthening the
role of civil aviation could attract international airlines and increase the availability of cargo
space. Rail is also important to Burundi but it can do nothing about it for now.
Gaining from Regional Integration
22. Burundi should benefit from membership in the EAC. As described in section B, chapter 3,
joining the EAC in 2007 opened opportunities for trade in goods and services, for developing physical
and regulatory infrastructure linking Burundi to a larger regional market and to the rest of the world, and
for development of services provided by the regional organization by pooling resources from all member-
countries. Although there is a risk of trade diversion, the larger economic zone would open opportunities
for the production of new goods and services currently not visible in the trade accounts. The ongoing
regional integration process can also create a push for economic and structural reforms and can strengthen
political and regional stability.
23. Regional infrastructure is vital yet Burundi can do little about it. Burundi depends on
Dar es Salaam and Mombasa for most of its external trade. Efficient port management is critical to
improving competitiveness of products coming from the region as is the network of roads, railroads, and
bridges that connects these final destinations to Burundi. EAC membership does not guarantee progress,
but it gives Burundi a forum to advocate for better infrastructure management in regional ports, rails, and
roads that it does not operate directly.
24. A comprehensive approach to regional integration, prioritizing and sequencing actions

needs to be implemented, most importantly addressing institutional constraints and economic reforms.
Key recommendations are the following (see section B4, chapter 3):
1
Developed by the African Development Bank.
xi
Institutional reforms: It is essential that the Ministry of East African Community Affairs
(MEACA) is staffed with competent personnel, and the necessary institutional arrangements need
to be in place, requiring extensive capacity building. It is important to implement a coordination,
monitoring, and evaluation mechanism for all EAC-related issues. Furthermore, a communication
campaign aimed at sensitizing the population to support the integration initiative should be
implemented.
Economic reforms: Eliminating all remaining nontariff barriers is critical. Furthermore, a strong
framework for public-private dialogue is advisable.
Improving the Business Environment
25. The private economy must grow more. The private sector faces constraints in law,
infrastructure, finance, and regulation, plus the tax of bad governance. One critical supply constraint is
finance. A Financial Sector Assessment Program (2009) and an Investment Climate Assessment (World
Bank 2008a) and the Doing Business Indicators show that Burundis business climate discourages
investment. Priority recommendations, as given in chapter 3, section C, are the following:
Access to finance in rural areas must improve. The capacity of the central bank (Banque de la
Rpublique du Burundi BRB) to supervise microfinance institutions needs to be strengthened.
New MFI regulations should allow them to develop new products. Similarly, technical assistance
should be provided to banks to develop new products and lending mechanisms to better meet the
needs of small and medium-size enterprises and of rural areas.
The regulatory environment should be improved to attract domestic and foreign
investment. Critical components are the finalization of the tax code revision, and the finalization
and adoption of the revised mining code. The new investment code includes some of the best
practices in investment legislation, among them investor protection and the freedom to transfer
capital and dividends. The amended tax code includes tax incentives mentioned under the revised
investment code. The revised mining code is meant to prepare the country for potential large-
scale investments and clarify the role of artisanal and small-scale mining.
Increase competition and improved governance. Competition could be increased by finalizing
and adopting the revised Privatization Law. Adopting and implementing the national policy of
good governance and its action plan would demonstrate the recognized importance of improved
governance.
Promoting New Sources of Growth
26. The strategy for new sources of growth has two parts. The first is to address the weak
performance of agriculture, including both food and export crops. It should be the priority to eliminate
widespread food insecurity and malnutrition. Second, in the medium term, diversifying the economy and
strengthening exports could lessen the vulnerability to external shocks and create new opportunities for
economic development and employment (especially off-farm employment). Revenues from current
exports - mainly coffee - can be stabilized and increased through better supply policies, as described in
chapter 3, section D.
27. Given the share of agriculture in employment and land use, addressing the causes of low
agricultural productivity is of utmost importance. As the Sources of Rural Growth Study (Baghdadli,
Harborne, and Rajadel 2008), the National Agricultural Strategy of Burundi (Republic of Burundi 2008b),
and an analysis of the coffee sector show, low agricultural productivity of food and cash crops is the main
obstacles to growth after low domestic investment. As a consequence, Burundi still relies on food aid, has
almost no private external resources (besides aid), and is vulnerable to external shocks. Regarding food
xii
crops, the main constraints are the poor economic incentives for farmers to introduce modern inputs, the
poor physical infrastructure and extension services supporting agriculture, the weak market organization
that is heavily distorted by government intervention, and the low development of farmer associations and
cooperatives. With regard to cash crops, the liberalization of agricultural prices and the definition of a
subsidiary role for the public sector in the production and commercialization of coffee and tea (in
particular, the ongoing privatization of washing stations) are steps in the right direction. However, the
speed of implementation of these programs could be accelerated.
28. Key recommendations include the following:
Food crops: Critical for increasing production and productivity is strengthening input
distribution systems to supply more inputs to small farmers; rehabilitating or modernizing
production tools and infrastructure (including rural roads); and organizing, structuring, and
professionalizing producers.
Coffee: Most important is the complete implementation of the ongoing privatization agenda.
Next steps include strengthening the capacity of the regulatory agency (ARFIC) and the
interprofessional association (INTERCAFE), facilitating access to finance, and improving
processing practices.
29. Diversification out of subsistence agriculture is particularly important. Given demographic

trends and the scarcity of land, Burundi must move away from subsistence agriculture because its income
per unit of land is too low. Although the transformation away from subsistence agriculture will not be
achieved for many years, the move should begin now based on a detailed analysis of potential products
and markets. The analysis presented in chapter 2, section A4, shows that Burundis products are not yet
internationally competitive. Nonetheless, the CEM identifies sectors and products that have great export
potential and are in line with Burundis Poverty Reduction Strategy Paper (PRSP) and National
Agricultural Strategy. Key interventions to facilitate diversification and the increased contribution of
other sectors, such as horticulture and fishing, are the availability of cargo space to export fresh produce
and the compliance with sanitary and phytosanitary standards.
30. Mining can contribute more to Burundis economy. Mining is done by artisanal and small-
scale miners, employing about 20,000 people in low productivity jobs and generating little tax income.
Industrial mining could boost value added, tax receipts, employment, and local supply of intermediate
goods, especially food and construction materials. There is currently no commercial mining because the
war stopped it.
31. To enable the mining sector to play an important role in Burundis economy, several
reforms are necessary. Given the potential contribution of the mining sector, a sectoral strategy
accounting for the characteristics and constraints of the available mineral resources is essential, as
outlined in section D4, chapter 3. The government has only recently started to reform the legal framework
for mining and a few international companies have shown interest. Burundi could greatly benefit from
joining the Extractive Industries Transparency Initiative and from using available facilities for assistance
in negotiations with international companies. To account for the large share of artisanal and small-scale
mining, a unit could be created within the Ministry of Energy and Mines to deal specifically with this
subsector. Support for the formalization of artisanal mining is also advisable because it would raise
productivity and allow small miners to capture more value-added.
xiii
Expanding Fiscal Space
32. The government has to expand fiscal space. As outlined in section A of chapter 4, Burundis
fiscal balance is expected to improve moderately in the next 5 years, but domestic resources will remain
limited and external resources are needed to finance the fiscal deficit. The fiscal position needs to be
strengthened by improving revenue mobilization, implementing prudent spending policies, and reforming
public financial management. The Public Expenditure Management and Financial Accountability Review
(PEMFAR) and an analysis of the medium-term framework demonstrate the need for improved allocation
of government resources. A key recommendation is that defense and security spending be reduced to the
extent possible while ensuring sufficient resources for the security of the country, as noted above. At the
same time, the public sector wage bill needs to be controlled. Resources made available by these
measures must be channeled to priority sectors and pro-poor expenditure.
FigureES.2:CompositionofPublicSpending,200108
33. Burundi could improve its fiscal stance
and direct additional resources toward priority
50 sectors. Although the authorities have made a
40
significant effort to improve the countrys fiscal
stance, increased and inefficient spending as well
percentofGDP
30 as inadequate revenue mobilization could result in

larger, unsustainable fiscal deficits. High defense
20
and security spending and the high wage bill are
10 the main obstacles to reallocating public resources
toward priority economic and social sectors.
0
2001 2002 2003 2004 2005 2006 2007 2008
34. Making aid more effective. To promote
Militaryandsecurityexpenditure Othercurrentexpenditure
Capitalexpenditure growth, external resources could be shifted from
Exceptionalexpenditure
Source:InternationalMonetaryFund.Note:Exceptional relief and humanitarian aid to investments in
expenditureincludesspendingonDDRandonelections. infrastructure and human capital. An analysis of
recent aid disbursements in chapter 4 shows a
declining trend in humanitarian and emergency aid, which is consistent with the improved security in the
country. Resources freed should be shifted to production-oriented sectors with more growth potential,
especially agriculture and infrastructure. This shift does not imply that spending on social sectors should
be reduced; to the contrary, investments in these areas are crucial for the long-term development of the
country.
35. Aid management of aid must change to promote growth. Because of the conflict and post-
conflict status of the country, aid has been primarily allocated to humanitarian and emergency projects;
production-oriented sectors, such as agriculture and infrastructure, have received less support than they
would have. The unpredictability of aid flows has exacerbated the spending bias toward government
consumption and away from investment in production-oriented sectors and in human capital, as shown in
section B, chapter 4. Improved donor coordination will facilitate the allocation of aid; in that regard, a
common performance assessment framework to make aid more effective must be agreed.
xiv
TableES.1:PriorityRecommendations
Area Recommendation Timeframe Approximatecost Costasshareoftotal
(201015) (US$millions) expenditureprojected
for201015
Closingtheinfrastructuregap
Electricity DevelopBurundisdomestichydroelectricpotential(Kaganuzi,Mpanda,Kabu16) Shorttomediumterm 80 1.3%
Provide110kVlineswithnecessarysubstationstoallprovincialcapitals Shorttomediumterm 60 1.0%
FinalizethefinancialrestructuringofREGIDESO Shortterm tobedetermined
Transport Rehabilitate,upgrade,andmaintainnationalhighways Shorttomediumterm 600 9.8%
Rehabilitateprovincialandcommunityroadnetworks Shorttomediumterm 13 0.2%
Buildhumanandinstitutionalcapacityingovernmentagenciesthatare Shorttomediumterm 15 0.2%
responsibleforregulationandmanagementofroadtransportactivities
Prepareabusinessandmasterplanforcivilaviation Shortterm 1 <0.1%
MaximizingthebenefitsofBurundisEACmembership
Institutional RecruitcompetentstaffinMEACA Shortterm 0.2 <0.1%
reforms Makeinstitutionalarrangements(includingcapacitybuildingandotherregional Shorttomediumterm 0.5 <0.1%
integrationissueswithinthenewministry)
Implementthecoordination,monitoring,andevaluationmechanisms Shortterm 0.3 <0.1%
Implementcommunication/sensitizationcampaign Shortterm 0.5 <0.1%
Economic EliminateremainingNTBs(includingequipmentatborderposts,onestop Mediumterm 1 <0.1%
reforms windowattheport,andsoforth)
Strengthenandoperationalizethepublicprivatedialogueframework Shorttomediumterm 0.3 <0.1%
Increasingthecontributionoftheprivatesectortogrowth
Financial ProvidetechnicalassistancetobanksandMFIstodevelopproductsthatmeetthe Shorttomediumterm 1 <0.1%
sector needsofsmallandmediumenterprises
Establishunitswithinbanksthatspecializeinlendingtosmallandmedium Shorttomediumterm Includedinprevious
enterprises,andunitsresponsibleforruralandagriculturalcustomers technicalassistance
RevisetheregulatoryframeworklimitingthedevelopmentofMFIs;andremove Shorttomediumterm 0.1 <0.1%
thebanonMFIsengaginginleasingandissuingmortgages,subjecttoprior
approvalfromthecentralbank
ImprovecapacityoftheBRBtosuperviseMFIs Shorttomediumterm 0.4 <0.1%
Regulatory Finalizeandadopttherevisedminingcode Shortterm 0.6 <0.1%
environment Finalizetherevisionofthetaxcode,includingtheintroductionofthesimplified Shorttomediumterm 0.35 <0.1%
imptsynthtique(synthetictax)
Governance PromulgatetherevisedPrivatizationLaw Shortterm tobedetermined
Adoptandimplementthenationalpolicyofgoodgovernanceanditsactionplan Shorttomediumterm 30 0.5%
xv
Encouragingemerginggrowthsectors
Foodcrops Strengtheninputdistributionsystems(technicalassistance) Shorttomediumterm 0.5 <0.1%
Rehabilitate,create,andstrengthenlocalinfrastructureforstorage,conservation, Shorttomediumterm 10 0.2%
transformation,andcommercializationofagriculturalproducts
Organize,structure,andprofessionalizeproducers Shorttomediumterm 3 <0.1%
Coffee Strengthenthecapacityoftheregulatoryagency(ARFIC) Shortterm 1.5 <0.1%
Launchthesecondroundofwashingstationsanddrymillsprivatizing Shortterm 0.5 <0.1%
Promotemicrocreditandmatchinggrantprogramstohelpfarmersfinanceinputs Shortterm 30 0.5%
orreplanttrees
Strengthenthecapacityoftheinterprofessionalorganization(INTERCAFE) Mediumterm 1.5 <0.1%
Improveprocessingpractices Mediumterm 0.5 <0.1%
Diversification Developanactionplantoestablishfacilitiestocomplywithinternationalsanitary Mediumterm 1.5 <0.1%
andphytosanitarystandards
Improvetheavailabilityofcargospacetoexportfreshproduce Shorttomediumterm 0.5 <0.1%
Mining Developandimplementasectoralstrategythatexplicitlyaccountsforthe Shorttomediumterm 0.15 <0.1%
characteristicsandconstraintsofthedifferenttypesofmineralresources
BecomeacandidatecountryoftheEITI,bevalidatedasacompliantcountry,and Shorttomediumterm 0.15 <0.1%
useavailablefacilitiesforassistancewithnegotiations
RevisetheinstitutionalframeworkbycreatingaunitwithintheMENtodeal Shortterm 0.3 <0.1%
specificallywithartisanalandsmallscalemining
Developaprojecttosupporttheformalizationofartisanalandsmallscalemining Shorttomediumterm 1 <0.1%
Source:Authorscompilation.
Note:Totalexpenditureprojectedfor201015referstothealternativescenariopresentedinsectionA,chapter4,assumingaconstantFBu/US$exchangerate.
xvi
INTRODUCTION
1. This Country Economic Memorandum (CEM) is the first for Burundi since the 1980s. It has been
developed in collaboration with the government of Burundi. The CEM has been prepared in cooperation
with the African Development Bank and the U.K. Department for International Development.
2. Burundi is one of the poorest countries in the world, and has suffered from many years of civil
conflict and its consequences. In the last years, peace has been established and a promising recovery of
the economy has started. Economic growth rates, however, are not in line with what has been projected in
the latest Poverty Reduction Strategy Paper (September 2006). Real GDP growth had been projected to
average almost 7 percent between 2006 and 2009 in that strategy paper, but actual growth will average
just above 4 percent for the same period.
3. The report reviews the economic developments in the past and tries to identify the most binding
constraints to growth. The CEM then sets out a strategy to address these constraints to promote increased
and participatory growth, reduce poverty, and improve the livelihood of the population.
4. The report draws on a number of background studies conducted on various subjects relevant to
the countrys economic development and on existing reports and studies from the government of Burundi,
the World Bank, other donors, and academics. The CEM provides a synthesis of various
recommendations and attempts to prioritize and sequence key actions.
5. The CEM is not meant to offer a comprehensive view of all economic issues in Burundi. It
focuses on the main constraints that have been identified for short- and medium-term growth, and
suggests strategies to address those constraints. Furthermore, the necessity of a growth-supporting
environment is highlighted and the role of government in supporting the economy is addressed. Given the
scope of the CEM, some long-run growth-enhancing measures (such as policy reform priorities in social
sectors) are not directly addressed.
6. The scope of the CEM, as well as preliminary results of its analysis, have been discussed in
Burundi with the members of the CEM steering committee and with a wider audience, including
representatives from various ministries, other institutions, and donor agencies.
1
C HAPTER 1 MACROECONOMIC AND FISCAL D EVELOPMENTS
1.1 Burundi is one of the poorest countries in the world. At US$111.3 in 2008, its per capita
income (in constant 2000 U.S. dollars) ranks second-to-last among countries whose statistics are
compiled by the World Bank, ahead of only the Democratic Republic of Congo. Burundi has not yet
started the transition from a traditional society, with most of the population is employed in subsistence
agriculture, to a modern society where most of the population lives in urban areas and is employed in
manufacturing and services. In most countries that have already undergone such transformation, this
transition has taken decades; along the modernization path, subsistence agriculture was gradually
displaced as more sophisticated and organized forms of agricultural, manufacturing, and services
production increased their relative shares of GDP.
1.2 Several obstacles are critical to this transformation and modernization of the economy.
They are political instability and the menace of violence; poor physical infrastructure, particularly the low
supply of electricity; low productivity in agriculture; the low level of integration of the economy with the
rest of the world and between the rural and urban economies within the country; poor governance and
government decision-making capacity; the low levels of education and health of the population; and high
population growth. Through strong links and interdependencies, these obstacles cannot be seen
independently. The capacity to surmount these obstacles will determine Burundis future economic
growth trends.
1.3 Political instability and the menace of conflict is still a powerful deterrent to economic
growth and poverty reduction. The assassination of the presidents of Burundi and Rwanda in 1993
sparked an ethnic clash and genocide in Rwanda and a civil war in Burundi, which destabilized both
countries for years going forward. The Arusha Accords of 2000 initiated a peace process in Burundi, but
insecurity remained high during most of the
following five years because several rebel Figure1.1:PerCapitaGDPinBurundiandRwanda,19622008
groups did not surrender their weapons and did
not join the peace process until well into the 100 50
decade. 80 40
1.4 The cost of war has been substantial.
LCU'000
LCU'000
60 30
Figure 1.1 illustrates some of the costs of
political instability and conflict in Burundi and 40 20
Rwanda. From 1965 to 1992, per capita GDP in
Burundi increased steadily; but from 1993 to 20 10
2000, as violence erupted, per capita income 0 0
contracted steadily. As of 2000, the cost of
1962
1964
1966
1968
1970
1972
1974
1976
1978
1980
1982
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
political violence amounted to about 62 percent

of the GDP. In contrast, although GDP declined Rwanda(LHS) Burundi(RHS)
by an even wider margin in Rwanda in 1994, the Source: World Development Indicators. Note: LCU = local currency
time of contraction was limited and high growth unit;LHS=lefthandside;RHS=righthandside
rates were experienced in the post-conflict
period.
1.5 The following section analyzes the performance of the economy from a historical perspective,
section B gives an overview of Burundis public finances.
A. ECONOMIC GROWTH: HISTORICAL PERSPECTIVE

1.6 This section starts with an overview of the past five decades, analyzes growth during the period
of political turmoil, and discusses the causes underlying slow growth following the Arusha Accords of
2
2000. Also included is a comparison of Burundis economic performance with those of Rwanda and
Sierra Leonetwo post-conflict countries that share similarities with Burundi. Furthermore, an outlook
of macroeconomic developments is provided.
A1. Economic Growth in the Past Five Decades and the Cost of Burundis Civil War
1.7 Overall, growth has been low in Burundi over the last five decades. From 1962 through
2008, real output grew at an average annual rate of 2.9 percent (table 1.1), while per capita GDP
grew at an average of only 0.8 percent. This 46-year period has three different subperiods: a period
Figure1.2:RealGDPGrowthandGDPperCapita,19622008
of steady, although slow increase in per
capita income from 1962 to 1992; a
25% 180 period of severe economic contraction
20%
160 from 1993 to the early 2000s; and a
140 period of slow growth and stagnation of
15% per capita GDP after 2000 (figure 1.2).2
120
Constant2000US$
10% 1.8100 The outbreak of the civil war in
1993 led to a severe disruption of the
5% 80
economy. The inflation rate reached
0%
60 unprecedented levels, the exchange rate
40 depreciated, and capital stock was
1962
1964
1966
1968
1970
1972
1974
1976
1978
1980
1982
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
5%
20
destroyed. In addition, neighboring
countries imposed an embargo and donor
10% RealGDPgrowthrate(lefthandside) 0
GDPpercapita(constant2000US$,righthandside) support for Burundi was suspended. All
these factors led to a negative average
Source:WorldDevelopmentIndicators.
growth rate of real GDP and real GDP per
capita. The decline of per capita GDP during the 1990s contrasts with the rest of Sub-Saharan Africa,
where it grew steadily. Furthermore, per capita GDP growth in Burundi was considerably more erratic
than for Sub-Saharan Africa as a whole: for the period 19802008, such growth in Burundi displayed a
standard deviation of 4.6 percent, compared with 2.1 percent for Sub-Saharan Africa.
Table1.1:SelectedEconomicIndicators,19622008
19622008 19621992 19932000 20012008
RealGDPgrowthrate(%) 2.9 4.4 3.1 3.0
Populationgrowth(%) 2.1 2.2 1.1 2.8
RealGDPpercapitagrowthrate(%) 0.8 2.2 4.2 0.2
GDPpercapita(2000US$) 124.6 129.7 120.1 109.4
Consumerpriceinflation(%) 10.2 8.2 17.7 9.4
Exportsofgoodsandservices(%ofGDP) 10.1 10.8 8.9 8.7
Importsofgoodsandservices(%ofGDP) 22.5 22.4 21.1 36.6
a
Currentaccountbalance(%ofGDP) 5.0 4.6 2.9 7.6
Grossdomesticsavings(%ofGDP) 1.8 2.0 4.9 15.9
Grossfixedcapitalformation(%ofGDP) 7.1 7.7 5.1 8.6
Officialdevelopmentassistance(%ofGDP) 16.8 14.7 16.7 40.0
Aidpercapita(currentUS$) 23.8 28.1 24.6 41.8
Revenueexcl.grants(%ofGDP) 16.9 19.7
Grants(%ofGDP) 3.4 12.6
Expenditureandnetlending(%ofGDP) 24.5 35.7
Source: World Development Indicators and International Monetary Fund. Note: = not available. a. including
officialtransfers;startingin1962,somedataarenotavailable.
2
Nkurunziza and Ngaruko (2002) further divide the time between 1962 and 1992 into two subperiods: 196272, characterized by
increasing political instability and weakening economic performance, but overall moderate GDP growth; and 197392,
characterized by somewhat lower political tension (with the exception of the domestic conflict in 1988) but very volatile
economic growth.
3
1.9 Peace was only restored slowly after 2000. The signing of the Arusha peace agreement in
August 2000 marked the start of a transitional period and return to democracy, as peace was slowly
restored. A three-year transitional government was formed in November 2001, and democratic elections
were held in August 2005. Reconstruction and rehabilitation efforts began, donors resumed their support,
and the economy started to recover. Real GDP growth returned to positive levels, but Burundi has not yet
recovered from the steady and deep fall in GDP provoked by the war.
1.10 The cost of war was substantial for Burundi. The cost of the war in Burundi may be gauged as
the difference between the actual
GDP and the potential GDP Figure1.3:Thecostofwar:potentialandactualGDPpercapita(ConstantFBu)
(defined as the output that would 31,000
have occurred had Burundi grown ActualGDPpercapita
29,000
between 1992 and 2000 at the PotentialGDPpercapita
median rate of economic growth 27,000
experienced in 196292. Figure 25,000
1.3 contrasts these two concepts. 23,000
The cost of war can be estimated
21,000
as the difference between actual
and potential GDPs. The gap 19,000
amounted to about 62 percent of 17,000
actual GDP as of 2000, and the
15,000
average loss between 1993 and
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2000 was roughly 41 percent. The
potential GDP would have reached
FBu 30,000 in 2008, almost Source:AuthorscalculationsbasedonWorldDevelopmentIndicatorsdata.
double the current level.
1.11 External savings declined as a result of the war. The direct negative impact of the war was
amplified by the sharp reduction in external savings. The current account deficit narrowed from an
average of 10.4 percent of GDP in the years 198092 to 2.6 percent of GDP in the years 19932000.
This reflects a large deceleration of inflows of external savings, which served to amplify the impact of the
conflict on aggregate demand.
1.12 Compared with Rwanda, the conflict has had a more severe impact on growth. In Rwanda,
the conflict was brief but extremely intense, causing large loss of life. Burundi, however, experienced a
prolonged period of political instability and open conflict, which continued to disrupt economic activity
for well over a decade. Hence, the destruction of physical capital and the disruption of economic activity
were appreciably higher in Burundi than in Rwanda. And although the cost of conflict in Burundi was
limited initially, it grew rapidly and accelerated over time. Moreover, the peace accords of 2000 were not
immediately able to fully reestablish law, order, or complete government control over the entire country;
and the economic impact of political instability, uncertainty, and conflict extended for a prolonged period.
Disarmament of the last rebel group did not take place until 2009.
A2. Slow Recovery after 2000

1.13 Per capita GDP stagnated over the past decade. To develop a better understanding of the
specific factors underlying Burundis slow progress toward recovery in the post-conflict period, we first
decompose the rate of growth by (1) sector or origin, (2) component of aggregate demand, and (3) factor
of origin (growth accounting). The analysis focuses on the subperiods 19972000 (last years of the civil
war), 200104 (transitional period), and 200508 (fragile return to democracy).
4
Supply-Side Analysis
1.14 The Burundian economy is rural and mainly agriculture based, relatively unsophisticated,
and undiversified. The primary sector clearly dominates the economy (table 1.2), representing nearly
half of economic output and involving the vast majority of the population. The main driving forces of the
primary sector are food crops, averaging a share of about 81 percent from 2005 to 2008, followed by
livestock and export crops with 13 and 4 percent, respectively.
Table1.2:SupplySideComponents,19972008
19972000 20012004 20052008
Share Growth Contribution Share Growth Contribution Share Growth Contribution
ofGDP rate ofGDP rate ofGDP rate
Primarysector 53.2 0.2 0.1 50.5 1.4 0.7 44.4 0.8 0.3
Secondarysector 12.3 3.0 0.4 13.6 5.5 0.7 15.5 5.9 0.9
Tertiarysector 26.7 0.7 0.1 29.1 4.5 1.3 33.4 6.8 2.2
GDPatmarketprices 100.0 0.8 0.8 100.0 2.5 2.5 100.0 3.5 3.5
Source:AuthorscalculationsbasedonInternationalMonetaryFunddata.
Note:DifferencesbetweenoverallGDPandthetotalofthesectorsaretheresultofindirecttaxes,whicharenotdisplayed.
1.15 The share of the secondary and tertiary sectors has been increasing. The secondary sector
which includes manufacturing, processing, construction, and related industriesrepresents the smallest
overall share of the economy. It averaged about 16 percent from 2005 to 2008, with manufacturing and
construction as its main subcomponents. The
Figure1.4:SectoralContributionstoRealGDPGrowth,19972008
tertiary sectorservices, transport, and
commercerepresents a relatively large and 6
growing share of economic output,
contributing about one third of GDP 4
between 2005 and 2008. The strength of this
percentagepoints
sector, however, primarily reflects an 2

increase in public services.
1.16 The primary sector remains the 0
main determinant of overall growth. In 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008
figure 1.4, it is evident that, despite their 2
improving performance, neither the
secondary nor the tertiary sector has 4
succeeded in eclipsing the influence of the Primarysector Secondarysector
primary sector on overall economic Tertiarysector RealGDPgrowth
performance, particularly given the Source: AuthorscalculationbasedonInternationalMonetaryFunddata.
relatively small proportion of the population
involved in these activities. Also clear is the extreme volatility of primary sector performance and its
contribution to growth.
1.17 The poor performance of the food crop sector is the main reason underlying stagnation of
per capita GDP. Food crops are a source of both livelihood and sustenance for most of the population. In
the near term, any strategy aimed at accelerating growth will have to focus on improving the performance
of this dominant sector.
Demand-Side Analysis
1.18 Despite the recent investment expansion driven by foreign aid, investment and savings rates
are very low in Burundi. Gross fixed capital formation reached an average of 15.9 percent of GDP in
200508, more than double the average in the 19972000 war period (table 1.3). In comparison, gross
fixed capital formation in Rwanda averaged 21.0 percent of GDP; and in Sub-Saharan Africa, the average
5
was 20.0 percent. Public fixed capital formation has increased during the post-conflict period as a result
of the reconstruction effort. This increase in the stock of productive resources has benefited from a
rebound in public investment and a recovery in private investment since 2006. However, this rebound
comes after a prolonged period of underinvestment, which is only now catching up to precrisis levels
(figure 1.5). Information available on remittances indicates that, in contrast to other African countries,
they have been insignificantaveraging less than US$800,000 in the last five years or less than 0.1
percent of GDP.3
Table1.3:DemandSideComponents,19972008
19972000 20012004 20052008
Component Share Contribution Share Contribution Share Contribution
ofGDP ofGDP ofGDP
Consumption 101.0 1.7 102.1 1.4 110.7 6.3
Public 13.6 1.1 23.4 2.2 24.1 0.6
Private 87.4 0.5 79.2 0.3 86.6 5.3
Investment 6.2 1.8 8.8 2.2 15.9 2.2
Public 5.3 1.8 6.6 1.8 8.4 0.3
Private 0.9 0.0 2.2 0.4 7.6 1.9
Resourcegap 7.4 0.2 10.8 1.1 16.6 2.1
Exports 8.1 0.6 8.5 0.4 9.2 0.1
Imports 15.5 0.8 19.3 1.5 25.8 2.2
GDPatmarketprices 100.0 0.8 100.0 2.5 100.0 3.5
Source:AuthorscalculationsbasedonInternationalMonetaryFunddata.
Note:Theindividualsharesandcontributionsmaynotadduptothetotalbecauseofchangesininvestment
stock,especiallyinthe200508period.
1.19 A combination of official flows Figure1.5:PrivateandPublicInvestment,19972008(percentofGDP)

and domestic resources has financed
12
these investments. Foreign investment
flows have yet to show an appreciable 10
acceleration. The very low level of 8
percentofGDP
foreign direct investment (averaging less 6

than 1 percent of GDP over the last 10
years) has been caused by a number of 4
factors, including the protracted war and 2

an unfavorable business environment. 0
The ongoing regional integration process
will help create opportunities for regional
foreign direct investment. Privateinvestment Publicinvestment
1.20 Public expenditure contracted Source:AuthorscalculationsbasedonInternationalMonetaryFunddata.

sharply during the crisis, but has
recovered considerably since 2000. It contributed about half of the GDP growth between 2001 and
2008.4 In line with the expansion of development assistance, this increase in government spending can be
attributed largely to recurrent expenditures (representing about 60 percent of overall expenditures), grant-
financed capital expenditure, exceptional expenditures, and military expenditures. This expansion relies
heavily on external financing and grants; and it has resulted in a significant deterioration in the budget
balance, which calls into question its sustainability over time without a significant improvement in
domestic revenue mobilization.
3
These are World Bank staff estimates based on the International Monetary Funds Balance of Payments Statistics Yearbook
2008. For more information, see Rhata, Mohapatra, and Silwal (2009).
4
General government final consumption expenditure includes all government current expenditures for purchases of goods and
services (including compensation of employees), as well as most expenditures on national defense and security.
6
1.21 The export base is narrow and undiversified. Exports have been low, averaging less than 10
percent of GDP over the last 10 years. The main export crops are coffee and tea, with coffee accounting
for about two thirds of the export value. The performance of exports has been erratic, owing largely to the
volatility of coffee production. Imports, on the other hand, are increasing, mainly financed by the
increasing level of foreign aid. Consequently, the current account deficit has been rising.
1.22 In summary, neither exports nor private investment (except over the past three years) have
contributed significantly to economic growth during the recovery period. This fact, combined with
the analysis of growth by sector of origin, supports the view that the stagnation of agriculture and low
expected return on investment have been at the heart of poor economic performance in Burundi during the
post-conflict period.
Growth Accounting
1.23 Productivity growth has been negative throughout almost the entire past five decades.5 Only
the 196272period displayed positive productivity growth, and the highest growth rate of GDP per
worker (which was attributed about equally to
productivity growth and capital accumulation). Figure1.6:AnnualGrowthofGDP,CapitalStock,and
ProductivityperWorker,19902008
From the onset of political violence in 1993 to
the signing of the Arusha Accords in 2000, the 10
growth of output and physical capital declined Arusha
Accords
markedly, and productivity growth was negative. 5
The decline is attributable to disruption of
percent
domestic economic activity, destruction of 0

physical capital, loss of foreign aid and capital
5
flows, the economic embargo imposed by
neighboring countries from 1996 to 1999,
10
displacement of the population, and suspension
of many government services over the period.
Official aid flows helped finance an increase in GrowthRateofRealGDPperWorker
GrowthRateofRealCapitalStockperWorker
capital stock form 2001 to 2008; even so, the rate TotalFactorProductivity
of productivity growth continued to decline, Source:AuthorscalculationsbasedonWorldDevelopmentIndicators
resulting in negative GDP growth per worker andInternationalMonetaryFunddata.
(figure 1.6).
1.24 Agricultural productivity has been low and declining. Given the importance of the agricultural
sector in terms of its share of overall output and the large proportion of the population involved, the
productivity of the sector is especially crucial. The data show that agricultural productivity has declined in
Burundi since the early 1990s, with value added falling to less than US$65 per worker (in adjusted 2000
dollars) by 2005. This decline stands in sharp contrast to the Sub-Saharan African average of nearly
US$288 that same year.6
A3. Comparison with Rwanda and Sierra Leone

1.25 How does Burundis economic performance compare with that of other post-conflict
countries? Although such comparisons are clearly difficult and imprecise, given the many qualitative
differences across countries (including those with respect to the depth, duration, and impact of crises),
some insights can be drawn from the experiences of other post-conflict countries.
5
A growth accounting decomposition can isolate the origins of growth by factor of origin (labor and capital) and total factor
productivity (TFP). Dividing the post-independence period into several subperiods, as discussed in section A1, the growth
accounting decomposition has been based on a Cobb-Douglas production function. The output elasticity with respect to capital
per worker has been estimated as 0.42.
6
Source: World Development Indicators.
7
1.26 Burundi, Rwanda, and Sierra Leone are at a similar developmental level. The three countries
share similar levels of human development, life expectancy, literacy and educational enrollment, and
poverty. All three compare poorly with most other countries assessed by the United Nations in the context
of the Human Development Index: Rwanda ranked 167th, Burundi 174th, and Sierra Leone 180th, among
182 countries in 2009.
Figure1.7:RealGDPGrowthduringPostconflictPeriodsinBurundi, 1.27 Burundi has not experienced

Rwanda,andSierraLeone the post-conflict boom that Rwanda
%40 and Sierra Leone have seen. Economic
performance during the years following
30
cessation of major hostilities in the three
20
nations2000, 1994, and 2000,
10 respectivelyhas differed widely (table
0 1.4). Burundi underperformed its
percent
10 t2 t1 t t+1 t+2 t+3 t+4 t+5 t+6 contemporaries over this period with
20 respect to overall and per capita growth
30 and in investment, despite displaying the
40 highest levels of foreign assistance and
SierraLeone
50 Rwanda
government expenditure. Rwanda and
Burundi Sierra Leone experienced explosive
60
growth (20 and 18 percent, respectively)
Note:Theyearduringwhichthecrisisendedofficiallyisdenotedastthat
during the first three post-conflict years,

is,2000forBurundiandSierraLeone,and1994forRwanda. Thismethodology whereas Burundi grew at a modest
isalsousedforrelatedcomparisonsinthischapter. average of 2 percent (figure 1.7).
Although both Rwanda and Sierra Leone
experienced more significant reductions in per capita income during the crisis period, per capita income
levels returned to well more than twice the level of Burundi only six years after the crisis.
1.28 This is partially explained by the fact that in Burundi, unlike in Rwanda and Sierra Leone,
the conflict ended neither rapidly nor completely. The protracted peacemaking process also led to
allocation of aid to sectors without a direct impact on growth. A higher share of aid has been disbursed for
humanitarian and emergency operations in Burundi than in the other two countries. Aid with a short-run
impact on growth, to the contrary, received the smallest share in Burundi, whereas (especially in Sierra
Leone) more aid was allocated to short-run growth-enhancing sectors. To determine what other factors
explain the differences in post-conflict performance, the origins of growth in each country by sector,
component of aggregate demand, and factor endowment or TFP are considered.
1.29 Regarding supply-side decomposition, several important differences are apparent:
First, the services sector has outperformed all other sectors in Burundi by a large margin, which
primarily reflects a substantial aid-financed expansion of government services (table 1.4).
Second, the contributions of agriculture to overall output in both Rwanda and Sierra Leone are
higher than that displayed by Burundi, despite the fact that the proportion of the population
involved in agricultural activity is much lower in Sierra Leone and roughly equal to that of
Rwandathat is, the proportion of the total population involved in agriculture in 2005 was 46
percent in Burundi, approximately 22 percent in Sierra Leone, and 47 percent in Rwanda.7
Third, the contribution of agriculture to growth in Burundi was negative, compared with the
significant positive contributions of agriculture in the other two countries. Agricultural
7
These figures are based on the most recent data for agricultural sector employment available from the World Development
Indicators database.
8
productivity has been significantly higher in Rwanda and Sierra Leone, with agricultural value
added per worker being two to five times as high as in Burundi in the post-conflict years.
1.30 The demand-side decomposition shows that the most significant divergences relate to the
growth contributions of investment and exports:
The average level of gross fixed capital formation relative to GDP in Burundi was less than half
of that in Rwanda and about two thirds that in Sierra Leone during the first six post-conflict years.
The relative share and contribution of private investment to output in Burundi was about one
sixth the level in Rwanda and one fifth the level in Sierra Leone over the same period.
Measured as a proportion of GDP, overall investment has generated about nine times the
contribution to growth in Rwanda than in Burundi, and more than five times as much in Sierra
Leone. Those figures underscore the relatively low productivity of capital displayed by Burundi
during the post-conflict period. The strong contributions to growth observed might stem from
higher levels of productive investment undertaken by the private sector in both Rwanda and
Sierra Leone.
The performance and contribution of exports to growth have fluctuated considerably in all three
countries. On average, Burundi has underperformed both Rwanda and Sierra Leone. In Rwanda,
the contribution of exports of goods to growth was limited; however, exports of services excelled.
Exports grew fast in Sierra Leone, but the nature of exports (minerals) implies much lower impact
on creation of jobs and value added outside the sector than the isolated figure may suggest.
1.31 Coulibaly, Duffy, Table1.4:ComparisonofSelectedIndicatorsSixYearsPostconflict(averages)
and Ezemenari (2008)
provide a composition of SelectedIndicator Burundi Rwanda SierraLeone
GDPgrowth(%) 2.7 14.4 12.8
growth for Rwanda during ContributiontorealGDPgrowthof agriculture 0.7 6.5 4.3
the post-conflict period. industry 1.2 3.4 2.7
Although somewhat services 4.7 6.0 7.5
different in its approach Agriculturalvalueaddedperworker(2000US$) 74 171 401
GDPpercapitagrowth(%) 0.7 8.3 8.4
from the one presented ODA(%ofGDP) 38.7 25.9 33.0
above, this study also Grossdomesticsavings(%ofGDP) 13.4 3.1 4.8
estimates TFP growth using Grossfixedcapitalformation(%ofGDP) 10.5 15.8 12.3
Public(%ofGDP) 6.8 7.7 4.9
a Cobb-Douglas framework. Private(%ofGDP) 3.6 8.1 7.4
In contrast to Burundi, Governmentfinalconsumptionexpenditure(%ofGDP) 23.9 10.6 14.9
productivity growth in Export(%ofGDP) 8.9 6.5 18.5
Imports(%ofGDP) 21.3 25.7 30.5
Rwanda was significantly Foreigndirectinvestment,netinflows(%ofGDP) 0.01 0.2 3.3
positive. From 1995 to Source:AuthorscalculationsbasedonWorldDevelopmentIndicatorsandInternational
2003, TFP growth is MonetaryFunddata.Note:BurundiandSierraLeone200106,Rwanda19952000;
estimated to have accounted contributiontoGDPfiguresareconsistentlyavailableforonlyfiveyearspostconflict.
for about two thirds of the
observed increase in output, with growth in employment producing most of the remainder. This finding is
striking, given that the augmentation of physical capital during the period was quite modest in relative
terms, underscoring the tremendous efficiency gains during the period.
1.32 The increase in TFP in Rwanda is explained by several factors, particularly by the development
of human capital via investments in health care and hospitals, education and professional training, and
knowledge infrastructure; and the formation of civil service. Trade and financial liberalization also are
considered to have supported productivity growth by allowing a freer flow of resources and supporting
financial deepening and lower costs of capital, as well as increased government revenues that were then
put toward productivity-enhancing investments. In this context, it seems clear that an important factor in
Rwandas success has been the governments ability to support productivity enhancements via prudent
9
policies, reforms, and public investments in critical areas (see box 1.1). This, along with increasing
private investment, has been at the center of the countrys rebound.
Box 1.1: Rwandas Growth Experience and Possible Lessons for Burundi
In its Vision 2020 document published in 2000, the government of Rwanda established targets for GDP growth and
poverty reduction to be achieved by 2020. Rwanda has a long way to go to achieve some of the targets set under
Vision 2020among them, to raise per capita income from US$230 to US$900 and reduce poverty from 60 to 30
percent of the population. The country will face tremendous challenges in meeting these targets. Nevertheless,
almost a decade after the establishment of the Vision 2020 targets, it has made substantial progress.
A key factor in Rwandas progress has been its commitment to market-based reforms, a limited role for government,
and support for private sector initiatives. Some reforms were under way prior to the crisis in 1994: liberalization of
trade and exchange rate regimes; accession to regional arrangements, including the Common Market for Eastern and
Southern Africa and the East African Community; and participation in the U.S. African Growth and Opportunity
Act. These and related reforms were critical, given the need to look beyond the small domestic market for sources of
growth. These measures also allowed the government to reduce military expenditures and increase spending on
education, health, agriculture, and infrastructure needs. Prudent macroeconomic policies and the adoption of a
medium-term perspective with respect to planning and budgeting have provided a stable economic environment for
investment and reform.
A political economy analysis of the causes of the genocide stressed the need for increased voice, inclusion, and
widespread participation in the political and decision-making processes. These findings led to adoption of a
decentralization program involving legal, institutional, and policy reforms to delineate central and local government
roles, responsibilities, and accountability mechanisms. The decentralization was followed by a refocusing of the
governments efforts on service delivery to communities. These and related reforms have supported both political
stability and growth.
A key dimension of the reform process has been efforts to support the agricultural sector, given its size and
importance for the population and economy as a whole. A national strategy for agriculture (articulated in the first
postcrisis Poverty Reduction Strategy Paper) was implemented, including policies aimed at increasing productivity
through the use of agricultural inputs, improved water management, and soil conservation. Recently, the government
has adopted a program to strengthen the private sectorled distribution of fertilizer, which has been successful.
These reforms have contributed to increases in per capita agricultural income by 5.6 percent from 2006 to 2007 and
by 8.0 percent from 2007 to 2008.
Rwanda has also adopted many reforms aimed at improving the business environment, which led to its being named
the most improved country in the World Bank report Doing Business 2010. Reforms have focused on reducing
administrative costs of doing business, improving transparency, and a implementing a zero-tolerance policy on
corruption.
Those reforms have led to substantial progress in the key indicators of growth and poverty reduction. Between 1995
and 2005, some of the growth experienced was catch-up growth, following the events that occurred in 1994. But
much of the growth was also the product of structural reforms. As a result, GDP growth averaged 10 percent (and
per capita income averaged 5.4 percent), with average inflation below 10 percent.
Looking forward, the government continues to focus on strengthening the institutional framework for public
investment, supporting private sector initiatives, and reducing the cost of doing business. Despite success in
reforming the sector, the government recognizes the need to support a transition from heavy reliance on the
agricultural sector over the medium term, and it has emphasized the development of the services sector and
investment promotion in its current strategies.
In sum, Rwandas success has been the result of a stable and responsible government focused on reform, sound
macroeconomic policies, the progressive liberalization of finance and trade, public financial and expenditure
management reforms, agricultural extension and investments, improvements in the business environment, and
initiatives aimed at increasing social welfare and inclusion.
1.33 In summary, Burundis post-conflict performance does not compare well with that of
Rwanda or Sierra Leone. One of the major factors underlying this divergence is certainly Burundis
slow and sporadic progress from open conflict to peace, relative to the relatively short transitions in
10
Rwanda and Sierra Leone. In this context, the performance of productivity is particularly telling, having
stagnated and fallen despite increased investment during the post-conflict period. Clearly, any recovery
and increase in growth potential will require productivity enhancements, as well as additional private
investment. But these improvements will not occur unless Burundi initiates a comprehensive reform of
the agricultural sector and invests in significant infrastructure improvements. The private sectors in
Rwanda and Sierra Leone perceive that the investment climate (especially in Rwanda) is much more
favorable than in Burundi, and that perception must have had a meaningful impact on the far more
dynamic contribution of private investment to recovery in Rwanda than in Burundi.
B. BURUNDIS PUBLIC FINANCES: AN OVERVIEW

1.34 The structure and management of public finances has an important role for the
development of the country, given the weakness of Burundis private sector and the countrys
dependence on foreign financing. The following section gives an overview of the structure of Burundis
public finances as well as the public finance reforms that have been undertaken or are currently being
implemented.
B1. Structure of Public Expenditure

1.35 Burundi has a narrow tax base, and most revenue stems from domestic activities. More than
90 percent of Burundis domestic revenue is derived from taxes, with an average of 71 percent coming
from taxes on domestic activity (taxes on goods and services and income tax) and approximately 19
percent from taxes on international trade during the period 200108 (table 1.5). Despite its narrow tax
base and the reduction of the petroleum tax, domestic revenue roughly has maintained its 2006 level of
about 19 percent of GDP over 2007 and 2008. Further increases in domestic revenue will depend on
improved revenue administration and the elimination of excessive tax exemptions (see box 2.2 in chapter
2). In this regard, the
Table1.5:CompositionofGovernmentRevenue,200108(percentofGDP) government has plans to revise
Revenuesource Average 2006 2007 2008 Average the general tax code and unify
200105 200608 the local and national revenue
Totalrevenue 28.3 36.8 39.5 43.5 39.9 administrations. As a result of
Domesticrevenue 20.3 18.9 18.6 18.5 18.7
Taxrevenue 18.4 17.3 17.2 16.6 17.0
Burundis accession to the East
Incometax 5.0 4.9 5.0 4.7 4.9 African Community, tax policy
Taxesongoods&services 9.0 8.9 8.7 8.8 8.8 will be reviewed to bring it in
Taxesoninternationaltrade 4.2 3.1 3.2 2.9 3.1 line with other member-
Othertaxrevenue 0.2 0.4 0.3 0.2 0.3
Nontaxrevenue 1.9 1.6 1.4 1.9 1.6 countries. The introduction of a
Grants 8.0 17.9 20.9 25.0 21.3 common external tariff could
Current 4.8 13.5 13.2 16.0 14.3 lead to a reduction in
Capital 3.2 4.4 7.7 9.0 7.0
Source:InternationalMonetaryFund.
international tax revenue.
1.36 Burundi relies heavily

on grants. Grants represent an increasing source of revenue, rising to 57 percent of total revenue and
thereby raising government revenue to 43.5 percent of GDP in 2008. This indicates that Burundi is
heavily reliant on a category of revenue that is subject to significant delays and is less predictable than
other sources of revenue. Grants are primarily disbursed as current grants, including budget support as
well as support for elections and for disarmament, demobilization, and reintegration.
1.37 Total government expenditure has been rising rapidly in recent years. It has increased from
27.2 percent of GDP in 2001 to 44.1 percent of GDP in 2008 (figure 1.8). This is a very large share of
public expenditure, and considerably higher than those in other Sub-Saharan African economies.
Recurrent expenditure accounts for more than 60 percent of total expenditure and net lending. The wage
bill has been rising over time, from 7.3 percent of GDP in 2001 to 11.4 percent in 2008. The rise since
11
2005 reflects the hiring of new teachers8 and a 15 Figure1.8:CompositionofPublicSpending,200108
percent salary increase in 2006the first increase 50
since 2001. Capital expenditure has more than
doubled (from 6.4 percent of GDP in 2001 to 15.1 40
percentofGDP
percent in 2008) as a result of a steep rise in 30
foreign grants. When military expenditures,
exceptional expenditures, and externally financed 20
capital expenditure are excluded, government
10
expenditure has been rising much more slowly.
0
1.38 The government has increased public
2001 2002 2003 2004 2005 2006 2007 2008
expenditure allocations to priority sectors over
Militaryandsecurityexpenditure Othercurrentexpenditure
the past six years. As a result of the availability of Capitalexpenditure Exceptionalexpenditure
funds for heavily indebted poor countries (starting Source:InternationalMonetaryFund.Note:Exceptional
in 2005) and increased donor support, the expenditureincludesspendingonDDRandelections.
government increased priority economic and social
expenditures from 4.8 percent of GDP in 2001 to about 12.4 percent on average during the period 2006
08, mainly reflecting an increase in education expenditures. The increased share of expenditure allocated
to social spending reflects the governments commitment to increase pro-poor spending. Demobilization
has reduced defense expenditures, but the increase in the police force has more than offset the savings. A
reduction in security sector spending is needed to provide more fiscal space for further increases in social
and economic priority sector expenditures.
1.39 The fiscal position improved in recent years, in line with increased external financing. The
fiscal deficit (on a commitment basis and after grants) shifted to an average of 0.7 percent of GDP in
200608 (from an average deficit of 5.0 percent in 200105), mainly because of increased grants
disbursement (table 1.6). The share of the deficit funded by external partners increased from 27 percent in
2001 to 95 percent in 2008. Shortfalls in budget support in 2006, partly reflecting donors governance
concerns, were offset by spending cuts in nonwage spending and increased domestic borrowing.
Following the 2007 governance incident, the government took mitigating measures (increasing some
taxes, cutting expenditures, and postponing until later years increases in benefit payments to the civil
service) to maintain the overall fiscal deficit within the International Monetary Fund program.
Table1.6:CentralGovernmentFiscalDeficit,200108(FBubillion,unlessotherwisestated)
2001 2002 2003 2004 2005 2006 2007 2008
Overallbalance(commitmentbasis)
Includinggrants 28.4 7.9 40.2 43.5 54.1 17.1 5.7 10.6
%ofGDP 5.2 1.4 6.2 5.9 6.3 1.8 0.5 0.8
Excludinggrants 39.6 33.1 88.7 144.3 144.3 182.2 210.4 355.0
%ofGDP 7.2 5.7 13.8 19.7 16.8 19.3 19.8 25.6
Changeinarrears(reduction) 21.3 9.4 2.2 58.5 10.2 13.7 21.9 11.0
External(currentinterest) 18.5 6.6 4.2 49.0 10.1 1.8 0.4 0.0
Domestic 2.8 2.8 6.4 9.5 0.1 11.9 21.5 11.0
Overallbalance(cashbasis)
Includinggrants 7.1 1.5 42.4 101.9 64.3 30.8 16.3 21.6
Excludinggrants 18.3 23.6 90.9 202.7 154.5 195.9 232.3 366.0
Financing 18.3 23.6 90.9 210.3 165.2 199.8 237.7 366.8
External 5.0 41.3 71.8 143.7 156.3 175.4 212.2 347.6
Privatizationproceeds 0.0 0.0 0.0 0.0 0.4 3.1 0.1 0.2
Domestic 13.3 17.7 19.1 66.7 8.5 21.3 25.4 19.0
Bankingsector 23.4 13.0 4.9 60.5 14.5 37.0 0.7 19.0
Nonbanksector 10.1 4.7 14.2 6.2 6.0 1.2 9.2 0.0
Financinggap/errorsandomissions 0.0 0.0 0.0 7.6 10.6 3.9 5.5 0.8
Source:InternationalMonetaryFund;authorscalculations.
8
Primary-school enrollment has doubled since primary school fees were abolished in September 2005.
12
1.40 Given Burundis risk for debt distress, most external financing is provided in the form of
grants. The grant element of external finance has been increasing in recent years, reaching about 90
percent in 2008. The increasing share of grants in external finance is in line with the debt-sustainability
analysis of the Heavily Indebted Poor Countries (HIPC) Initiative Completion Point. It stresses that the
provision of external financial assistance needs to be heavily weighted toward grants, given the countrys
vulnerability to external shocks and its low and volatile export performance.
B2. Public Finance Management Reform

1.41 Since 2005, the government has taken steps to improve public finance management. The
government introduced a double-entry accounting system, a new nomenclature that helps identify pro-
poor expenditures, and an interim computerized financial management information system. In response to
a governance scandal in 2008, the government introduced further changes in the 2008 annual budget law.
In May 2008, a new, organic budget law was adopted, and it was promulgated in December 2008. The
government is also consolidating its cash management by closing off-budgetary accounts. A new
procurement law was promulgated in February 2008, after considerable delays.
1.42 Despite progress, there remain major weaknesses affecting the quality and transparency of
public finance management. The government and the World Bank recently completed a Public
Expenditure Management and Financial Accountability Review (PEMFAR) to prioritize future reforms. It
identifies four main challenges:
1. improving the allocation efficiency of public expenditures by reducing security sector
expenditures and increasing public expenditures to education, health, and infrastructure, in line
with the Poverty Reduction Strategy Paper (PRSP);
2. maintaining control of the wage bill, including by eliminating ghost workers;
3. continuing public finance management reforms to increase transparency and strengthen fiscal
discipline, including by consolidating the budget, streamlining budget execution, reducing the use
of exceptional procedures, and improving internal and external audit systems; and
4. professionalizing the public service to improve performance, including by creating staff
evaluation and grading systems and reducing the number of political appointees.
1.43 Increasing the share of priority sectors and pro-poor expenditures in planned and executed
budgets needs to be supplemented by programs aimed at better tracking the destination and
improving the quality of public expenditures. Although the public expenditure tracking survey exercise
should continue, it should be complemented by identifying ways and means of associating beneficiaries
representatives and civil society in evaluating the performance of public services and the quality of public
expenditures. An International Development Association health sector project approved in June 2009 will
initiate and test mechanisms aimed at involving local communities in monitoring basic health services.
Extending the same approach to other key sectors should be encouraged in the context of future
development policy operations.
1.44 The size of the wage bill is a serious public finance management issue in Burundi.
Representing an unsustainable level of 11.8 percent of GDP in recent years, the size of the wage bill
constrains resources that should have been otherwise allocated to growth-enhancing productive activities
(such as investment).
1.45 External and internal control and audit institutions have been strengthened, but much
remains to be done. Virtually nothing has been done to coordinate their programs so as to eliminate
unnecessary duplications and focus their activities on high-priority objectives. The Inspection Gnrale
de lEtat (State General Inspectorate) is the most appropriate institution to initiate a dialogue between
existing internal oversight departments and institutions, with a view to developing work plan for
consistent, risk-based audits. In parallel, the operational independence and professionalism of the Cour
13
des Comptes (Audit Court) should be further strengthened because it has a major role to play in enhancing
the transparency of public finance management and the accountability of the executive. The state
accountability institutions also should progressively learn to rely on the work of internal oversight bodies
as they gradually strive to comply with international standards.
14
C HAPTER 2 O BSTACLES TO ECONOMIC G ROWTH AND
EMERGING O PPORTUNITIES FOR BROADENING
THE B ASE FOR G ROWTH
A. OBSTACLES TO ECONOMIC GROWTH

2.1 Private investment is low because the expected rate of return on capital is low. During the
period of conflict, instability, and uncertainty, the ability of the economy to attract external savings fell
considerably; and domestically financed investment was constrained by modest levels of income and
domestic savings. However, it is difficult to argue that domestic savings have been a constraint since
2005, with a current account deficit (the supply of external savings facing the country) of about 15
percent of the GDP.
2.2 There is not a unique reason why the return on capital is low. In addition to political
instability and the menace of conflict, these reasons include (1) isolation from the rest of the world; (2)
poor physical infrastructure, particularly the low supply of electricity; (3) poor government policies
regarding the largest sectoragriculture; (4) a narrow and undiversified export base; (5) a weak private
and financial sector; (6) poor governance and government decision making and tax revenue collection
capacity; and (7) the low level of education of the population, especially in the context of the high
population growth rate. The main obstacles to economic growth are discussed next.
A1. Geographic and Economic Isolation

2.3 Low integration within the domestic economy and with the rest of the world has deep
geographic, historical, and political origins. Burundi is landlocked, nestled among the highly unstable
Democratic Republic of
Table2.1:SelectedIndicatorsComparingBurunditoSubSaharanAfricaandthe Congo to the west, Rwanda to
WorldEconomy the north, and Tanzania to the
Indicator Burundi SubSaharanAfrica World south. Poor physical

PercapitaGDP (2000US$) 111 619 6,024 infrastructure, low average
Ruralpopulation(%oftotalpopulation) 90 64 50 educational attainment, and
Exportsofgoodsandservices(%ofGDP) 10 33 28 weak channels of interaction

Agriculturevalueaddedbyworker (2000US$) 64 288 939 with the rest of the world
Agriculturevalueadded(%ofGDP) 45 16 3 imply that Burundi remains
a
Phonesper100people 4 25 71
b
fairly isolated in a highly
Dependencyratio 72 85 55
volatile region.
Sources:Latestavailabledata(2005and2008,respectively)fromtheWorldDevelopment
IndicatorsdatabaseandtheInternationalMonetaryFund. 2.4 The economic
a.Numberoffixedandmobilelines.B.Ratioofpopulationyoungerthan15andolderthan
64toworkingagepopulation(1564years). geography conjoined with
the infrastructure deficit
lead to high transport costs
that affect trade negatively. Poor roads, bridges, and rail infrastructure hinder internal trade and reduce
Burundis opportunities for exchange with the region and the rest of the world. The internal obstacles are
compounded by the high cost of using the ports of Dar es Salaam in Tanzania or Mombasa in Kenya, poor
storage facilities, inadequate freight handling capacity, and some customs and administration issues that
have proved particularly difficult for Burundian firms to navigate. The poor state and coverage of
infrastructure in Burundi imply high transport costs and delays that lower the expected rate of return on
capital, discourage domestic and foreign investment, and constrain economic growth. In the agriculture
sector, transport costs represent, on average, 35 percent of import prices and 40 percent of export prices.
15
High infrastructure costs and lack of access are major obstacles to improved incomes and well-being for
the very large part of the population that depends on agriculture for a livelihood.
2.5 Two economic dimensions of this reality are the small size of the domestic market and the
low export share of GDP. The size of the domestic market (US$1.1 billion in 2008) is very small; and in
the World Development Indicators database, Burundi ranks 146th of 170 countries for which comparable
GDP data are available. Moreover, with the exception of Liberia, countries with GDPs lower than
Burundis are all small states.9 According to International Monetary Fund data, exports were equal to
about 10 percent of GDP in 2008, the lowest in the world after Rwanda10; and significantly below the
average for Sub-Saharan Africa and the world as a whole (table 2.1).
2.6 Exports and imports are critical to the development of any countryeven more so when the
countrys domestic market is small. However, Burundis export base is narrow and has not changed
significantly in the past 20 years. Coffee is responsible for more than 90 percent of foreign exchange
earnings. In turn, Burundis imports have been growing rapidly as a result of aid inflows; in the medium
term, however, they are constrained by the countrys low capacity to export. The development of
Burundis export base in the short and medium terms is directly associated with its capacity to overcome
stagnation in agriculture, particularly in the coffee sector, and with its development of new products
(goods and services). The development in neighboring countries needs to be taken into account.
2.7 In that context, joining the East African Community (EAC11) has been a momentous
decision that has presented both opportunities and challenges. However, becoming a member of the
EAC will require considerable institutional changes and policy reforms, including with respect to
Burundis tax and customs regimes, its legal system, and regulatory structuresall challenges that will
require considerable effort and time.
2.8 Many potential benefits arise from membership in the EAC. Joining the EAC opens
opportunities for trade in goods and services, for development of physical and regulatory infrastructure
linking Burundi to a larger regional market and to the rest of the world, and for development of services
provided by the regional organization through the pooling of resources from all member-countries.
Although there is a risk of trade diversion, the larger economic zone may help open opportunities for the
production of new goods and services currently not visible in the trade accounts.
2.9 The importance of the regional agreement for the development of physical infrastructure
cannot be overstated. As mentioned above, Burundi depends on Dar el Salaam and Mombasa for its
exports and imports. Efficient port management is critical to improve the competitiveness of products
coming from the region; the network of roads, railroads, and bridges that connect these final destinations
with the rest of the region is also crucial. The EAC membership does not guarantee progress in any of
those directions, but it is a clear step toward opening opportunities to integrate the country with the rest of
the world and to broaden possibilities of exploiting economies of scale in many areas.
A2. Infrastructure Deficit

2.10 Burundi suffers from extreme infrastructure gaps in road access, power generation,
communications infrastructure, and access to water and sanitation. On just about any measure of
infrastructure coverageroad density, telephone density, power generation capacity, or service
coverageBurundi and the other EAC countries lag behind most other regional groupings in the world.
Burundi also lags behind other EAC member-countries in access to basic infrastructure services (see table
A.8 in appendix 4). About 90 percent of the population lives in rural areas. Despite the importance of
agriculture, only a relatively small portion of the rural population has access to all-season roads. The road
9
Following the World Bank definition, a small state is a sovereign state with a population of fewer than 1.5 million.
10
However, Rwandas exports of services as opposed to goods and services) have grown significantly in the recent past.
11
The EAC is the regional intergovernmental organization of the Republic of Kenya, Uganda, the United Republic of Tanzania,
and (since 2007) the republics of Burundi and Rwanda.
16
densities in areas of arable land are substantially lower in Burundi than elsewhere in Africa and in other
low-income countries.
2.11 Poor physical infrastructurein particular, the low supply of electricityresults in high
production costs and inhibits new investment in the rest of the economy. About 72 percent of
Burundian firms surveyed recently by the World Bank ranked poor access and reliability of electricity as
the most significant obstacle to new investment. Despite abundant hydroelectric potential, Burundi
displays a severe generation and transmission capacity deficit that began in years of underinvestment and
of deterioration and damage during the conflict period. Only 2 percent of households currently are
thought to have electricity service, compared with 16 percent for Sub-Saharan Africa and 41 percent for
other low-income developing countries. Bujumbura accounts for two thirds of all connections (Benitez,
Estache and Niyungeko 2009). As of 2008, the power utility, REGIDESO, had about 31 megawatts of
total installed electrical generation capacity, although only about 26 megawatts of this capacity was in
use. Domestic electrical production capacity in per capita terms is the lowest in the world. Limited access
to electricity is an important barrier to development. Currently, the electricity shortage is met, in part,
through imports from the Democratic Republic of Congo, but these are not sufficient to satisfy demand at
current prices. A strategy to tackle this is critically needed. It should address both supply-side and
demand-side issues, and take into consideration the extended period that such an extensive investment
and reform program will require. Burundi is also lagging in mainline and mobile telephone densities, as
well as Internet access. Teledensity remains poor at 3 percent of the population, with more than 90
percent of subscribers concentrated in urban areas.
2.12 In general, the following are examples of widespread problems found in the EAC (World
Bank 2008c):
Poor roads and bridgesThe main regional corridors and core road network are mostly paved.
Acute problems are caused by the variable conditions of parts of the core/regional network and
the increasing levels of traffic congestion on roads in and around major urban centers; those
factors, in turn, affect access to the ports of Dar es Salaam and Mombasa. Figure A.2 in appendix
4 gives the locations of the main road transport corridors in East Africa.
Dilapidated railwaysDisrepair of railroad infrastructure and equipment is characterized by
aging tracks, shortage of cars, and inadequate locomotives. Most of the railway systems in Kenya,
Tanzania, and Uganda have now been contracted out to concessionaires. This is expected to lead
eventually to improved operational and financial performance of these rail networks. Figure A.3
in appendix 4 gives the location of the rail transport network in the EAC.
Inadequate ports and inland container freight stationsProblems include inadequate storage and
handling capacity; handling equipment (such as cranes) in poor condition; and lack of adequate
rolling stock at Bujumbura, Dar es Salaam, and Kigoma. In 2007, for example, about 700
containers were being offloaded each day at the port in Dar es Salaam, although Tanzania
International Container Terminal Services was able to deliver no more than 300containers out of
the port daily. Inland container freight stations have similar problems.
Insufficient cargo vessel capacity at the Great LakesAge and disrepair of the merchant fleet on
Lake Tanganyika and Lake Victoria are serious problems. There is no shipyard on Lake
Tanganyika that could maintain or renovate the existing fleet. With the current low levels of
traffic on the lake, however, a shipyard may not be viable.
Inadequate facilities at the border postsProblems at the border posts include understaffing
relative to the volumes of activities and cargo handled; limited parking space for cargo trucks at
most posts; lack of truck scales forcing customs agents to estimate loads; inadequate office
facilities, including lack of computer equipment; and inadequate bonded warehouses.
17
Poor power supplyA number of border posts have no electricity; for others, the power supply is
unreliable because there is a lack of fuel for generators. As a result, bookkeeping must be done
manually.
Expensive cross-border communicationCounterpart revenue authorities located immediately
across a border communicate on landlines charging international rates. Private telecommunication
companies (such as MTN, Safaricom, and Zion) are developing special arrangements for cross-
border calls at local rates, but these do not cover Burundi.
2.13 Access to infrastructure services is limited, and the poor state of infrastructure leads to
substantially higher costs. Prices for services can be two to three times greater than prices in other
countries, further undermining the competitiveness of Burundi business in regional and global markets
(table A.9 in appendix 4). The cost and adequacy of these services affect commercial opportunities for
small farmers, entrepreneurs, and businesses both small and large. Business surveys in Burundi
consistently identify the cost of power and the poor reliability of the service as the single most important
obstacle to increased business investment.
A3. Poor Agricultural Policy

2.14 Low productivity in agriculture is one of the main barriers to development. According to
World Development Indicators data, agricultural output, which is dominated by subsistence farming,
represented about 45 percent of GDP in 2005, and involved about 90 percent of the labor force. Value
added per worker in the sector is US$64 (in constant 2000 dollars)well below the average for Sub-
Saharan Africa and about 15 times smaller than the global average, indicating very low productivity in the
sector.
2.15 Given the size of the agriculture sector and its potential to benefit large segments of the
population, addressing the causes of low productivity is of utmost importance. The immediate causes
of low productivity are low human capital and low use of modern machinery, fertilizers, productive seeds,
and modern methods of production. The underlying reasons are these: poor economic incentives for
farmers to introduce modern inputs; poor physical infrastructure and extension services supporting
agriculture; market organization that is very poor and heavily distorted by government intervention (as in
coffee production); and slight development of farmer associations and cooperatives to provide storage,
commercialization, and financing of crops and other supportive services.
2.16 Increasing government intervention in the coffee sector, the main cash crop, has been a
failure in the past. The government has tried to improve performance in the sector through liberalization
and privatization, but the effort has been partial and there is considerable uncertainty about the roles of
the private and public sectors. Furthermore, prices have been distorted by government intervention and
have stagnated, providing few incentives to producers to improve quality.
2.17 The government is now moving in the right direction. Liberalizing agricultural prices, defining
a subsidiary role for the public sector in the production and commercialization of coffee and tea, and
tabling the privatization of washing stations are positive steps. Unfortunately, these programs are being
implemented very slowly. Even if privatization occurs in coffee and tea, the challenges to increasing
productivity in these crops will not be eradicated. Transitioning from the current situation to one where
private organizations are responsible for services supporting farmers is not straightforward.
A4. Narrow and Undiversified Export Base

2.18 Burundis export base is narrow and undiversified, making the country vulnerable to
external shocks and dependent on foreign financing. This section outlines the composition of
Burundis current exports, also compared with the other EAC members. Furthermore, Burundis potential
to diversify its exports is explored.
18
Burundis Current Exports
2.19 Burundi very narrow export base (averaging only about 10 percent of GDP) has not
changed notably in the last 20 years. In comparison, other EAC members, especially Tanzania and
Uganda, have experienced an increase in exports as a share of GDP since 1985, coupled with a significant
increase in GDP per capita. The same development can be observed, on average, for low-income and
Figure2.1:HIforEACMemberCountries Sub-Saharan African countries. Burundis
exports remain very concentrated. One possible
1.0
0.9
measure of export concentration is the
0.8 Herfindahl index (HI).12 The HI lies between 0
0.7 and 1, where 1 represents exports that are
0.6 extremely concentrated (only one product), and 0
0.5 stands for a completely diversified export basket.
0.4
Economies with an index below a threshold 0.1
0.3
0.2
are considered to have a highly diversified export
0.1 basket. The HI for Burundi has remained high
0.0 throughout the entire period considered here
197079 198084 198589 199094 199599 200004 200507 (figure 2.1). It averaged about 0.5 from 1985 to
Burundi Kenya Rwanda 2004; and it has even risen above 0.8 during
Tanzania Uganda recent years, reflecting the higher share of coffee
Source:COMTRADE,SITC24digits;calculationbytheEconomicPolicy
andDebtDepartmentoftheWorldBank.
in total exports.
2.20 Compared with most of the other EAC
members, Burundi has a relatively high level of export concentration. Rwandas exports are similarly
concentrated, but Kenya and Tanzania have low and decreasing HIs (figure 2.2). The continued decline of
the HI for Uganda demonstrates the countrys Figure2.2:HIandGDPPerCapita
success in diversifying its export base. Its HI 2,500
was as high as 0.9 in the early 1990s, implying ElSalvador
almost fully concentrated exports, but it has 2,000 Colombia
GDPpercapita(2000US$)
decreased significantly to below 0.2 since Guatemala

2001. 1,500
Honduras
2.21 The high concentration of exports is
1,000
correlated with low GDP per capita. Figure
2.2 illustrates the negative relationship 500
Coted'Ivoire
Kenya Rwanda
between a high HI and GDP per capita for a Tanzania
CAR
Uganda SierraLeone
number of coffee-exporting countries. It also 0
Madagascar Burundi
demonstrates that Latin American countries 0 0.2 0.4 0.6

that used to be mainly coffee exporters have HerfindahlIndex
been more successful in diversifying their Sources:WorldDevelopmentIndicatorsandCOMTRADE,SITC24digits;
economies and increasing their income levels. calculationbytheEconomicPolicyandDebtDepartmentoftheWorld
Burundi is clearly at the very bottom of the Bank.
curve with a very low GDP per capita and strong export concentration.
2.22 Burundis economy is highly dependent on primary products, predominantly coffee.
Burundi lags behind the development of all other countries that were exporting primarily coffee in the
1980s. All other coffee exporters have reduced their dependence on coffee and established alternative
export products. This dependence on coffee makes Burundi vulnerable to fluctuating external demand,
price shocks, and unfavorable climatic conditions. In addition, the coffee sector is suffering from
12
The HI is calculated as
, where is the share of the considered variables (in this case, exports) and N is the total
number of products exported. Two factors can lead to a lower HI, a more even distribution of shares, and an increase in the
number of products.
19
declining output and quality. The sector, however, is undergoing a comprehensive reform (as described in
detail in section B2).
2.23 The second most important primary export product is tea, which has remained in the
shadow of coffee in the past. The structure of the tea sector has remained largely unchanged since the
1960s when the first tea plantations were established. Constraints the sector faces include structural
inefficiencies of the tea factories and plantations, poor incentives to smallholders and estate laborers,
limited use of inputs and extension services, and nonexistent research. The tea sector, however, has the
potential to contribute significantly to export earnings; and a reform program of the sector is currently
under way.
2.24 Horticulture exports are only a small share of Burundis total export basket. Horticulture
exports from Burundi began in the mid-1980s and picked up momentum into the 1990s, peaking in 1993.
Burundi has fallen back since that period, while market opportunities and food procurement systems
worldwide have changed radically. The global economic and financial crisis in 2008 has also caused a
decline in horticulture exports. Flower exports especially have suffered in recent years, also because they
were supplied to only one buyer and no diversified customer base had been established.
2.25 Gold used to be one of the main export products. Gold reached as much as a 30 percent share
of exports in the early 1990s, but the civil conflict hampered the development of the mining sector. The
country has large deposits of several minerals (such as nickel, coltan, and cobalt) that have not been
exploited commercially (see section B4).
Figure2.3:CompositionofExportsinEACCountries,198589and200506
19851989 2005 2006

100% 100%
90% 90%
80% 80%
70% 70%
60% 60%
50% 50%
40% 40%
30% 30%
20% 20%
10% 10%
0% 0%
Burundi Kenya Rwanda Tanzania Uganda Burundi Kenya Rwanda Tanzania Uganda
PP PP HV RB LT MT+HT PP PP HV RB LT MT+HT
Source:COMTRADE,SITC24digits;calculationbytheEconomicPolicyandDebtDepartmentoftheWorldBankandtheauthors.
Note:Thesearethetechnologyclassifications:primaryproduct(PP),highvalueprimaryproduct(PPHV),resourcebased(RB),lowtech(LT),
mediumtech(MT),hightech(HT).
2.26 Burundis export structure is characterized by very low use of technology. Following Lall
(2000), exports (excluding primary products and natural resourcebased products) can be classified
according to the technology level needed to produce themnamely low, medium, and high. High-value
primary products represent a subgroup of primary products. The share of primary productsmainly
coffee and teain Burundis export basket has even increased in the recent years, reaching 93 percent in
200506. The importance of resource basedproducts (mainly gold) has decreased since the early 1990s.
The share of low-, medium-, and high-tech product has always been minimal, reaching only about 2
percent of total exports in 200506.
20
Box 2.1: The Transformation of Ugandas ExportsA Success Story
Uganda was in a situation very similar to that of Burundi in the 1980s, recording very low GDP per capita and
exporting almost exclusively coffee after a prolonged period of civil conflict. Like Burundi, the country is
landlocked and has access to one of the Great Lakes. Uganda, however, has been successful in stabilizing the
country and diversifying its export base. It has a natural comparative advantage in unprocessed, primary exports
because of its abundant natural resources, but is disadvantaged by an unskilled workforce and its landlocked
location. During the 1960s and 1970s, its export basket contained mostly coffee and cotton; other commodities
(copper, tea and mate, hides and skins) had shares of less than 10 percent. The cotton sector collapsed in the 1970s
because of heavy taxation; it has never recovered. During the 1980s and early 1990s, the pattern of exports shifted
toward a coffee-only trend.
In the 1990s, the abolition of the export tax and trade liberalization led to a dramatic export recovery. The mid-
1990s marked a structural shift in Ugandas export pattern. The story line transformed from mainly coffee to mainly
coffee, fish, and flowers. Fish was a new natural resourcebased discovery. The share of coffee plummeted from
around 94 percent of total exports in the mid-1980s to 28 percent by 2004, while the share of cotton remained
roughly 45 percent. A few traditional exports (such as tea and mate, animal feed, and oilseed) either lost their
shares or vanished. Several nontraditional exports emerged. Exports of fish started in the late 1990s and their share
accelerated from 4 percent to 26 percent between the early 1990s and 2004 (second only to coffee). The shares of
other high-value exports (such as fruits and vegetables, spices, tobacco, leather products, maize, hides, and skins)
also gained ground. The graduation of fishery productswhich catered informally to the local and neighboring
marketsto export products was not a natural outcome of market forces. After repeated bans by the European
Union because of failure to comply with its phytosanitary standards, concerted efforts by the government of Uganda
in partnership with external donors succeeded in establishing the regulatory standards necessary for the fishery
exports industry to take off.
The government of Uganda has been generally proactive in promoting export growth and diversification. Policy
measures taken by the government include
replacement of the export licensing system with a system of export certificates without value limitations
establishment of the Uganda Revenue Authority, leading to the rationalization of the tax and revenue
system
creation of a new Investment Code, specifying duty and tax exemptions to exporters and offering additional
protection to foreign investors
The creation of the Uganda Investment Authority, a one-stop investment clearing agency that promotes,
facilitates, and supervises investments in Uganda.
To catch up with its comparators outside Sub-Saharan Africa, Uganda will need to scale up high-PRODY exports to
raise their weight, reduce that of coffee, and boost its EXPY. Ugandas EXPY has increased considerably in the
past, but its more recent trend has worrisome implications for its development. Around 2000, the emergence of high-
PRODY fish products augmented Ugandas EXPY significantly; but the absence of similar products since that time
flattened out its EXPY during 200105. To transform into a middle-income country, Uganda has few options but to
diversify into at least some higher-PRODY products. Catch-up experiences indicate that a few high-PRODY
products sufficed in the early stages of diversification, but persistence was necessary for sustained growth.
Sources: Boccardo and Chandra 2008; Dijkstra 2001.
2.27 In contrast, the other EAC members have successfully diversified their export portfolio
(figure 2.3). Particularly notable is Uganda (box 2.1). Twenty years ago, Uganda was almost exclusively
exporting coffee. Now, however, high-value primary products, such as fish fillets or cut flowers, make up
more than one third of its exports. Kenya and Tanzania have expanded their shares of high-value primary
products, as well as low- and medium-tech products. Rwanda is still mainly relying on primary products
and mineral resources, and has not yet been able to establish high-value exports. The case of Uganda
shows that the promotion of high-value primary products is a viable strategy to diversify the economy and
increase the income level.
21
Burundis Export Potential
2.28 Burundi has been only partially successful in introducing high-value primary products,
such as cut flowers. High-value primary products as well as resource-based products can contribute to an
increasing income level to the same extent that manufactured products can contribute. Hausmann, Hwang,
and Rodrik (2007) have defined a measure that ranks products according to their implied
income/productivity leveltheir PRODY.13 PRODY is a good measure to assess the potential of a
specific product to enhance per capita income. The evaluation of a product according to its PRODY
which is product specific, but not country specificcan help Burundi choose products that can contribute
most to a higher income level. Especially high-value primary products (such as cut flowers,14 palm oil, or
fish fillets) may be appropriate products for Figure2.4:EXPYofEACCountries
Burundi to focus on in the medium term. The
4,000
average (unweighted) PRODY of the top 10
3,500
export products from Burundi has more than
constant2000US$
3,000
doubled from US$2,878 in 198589 to
2,500
US$6,705 in 200506. The share of the
2,000
products with a high PRODY, however,
remains very low. 1,500
1,000
2.29 The income potential of Burundis 500
export basket remains low. Based on the 0
PRODY, Hausmann, Hwang, and Rodrik
197079198084198589199094199599200004200506
(2007) developed another measure, called the Burundi Kenya Rwanda
EXPY, which evaluates the income potential Tanzania Uganda
of the entire export basket of a specific Source:COMTRADE,SITC24digits;calculationbytheEconomicPolicy
country.15 They find that there is a strong andDebtDepartmentoftheWorldBank.
positive correlation between the sophistication
of the export basket, measured by the EXPY, and subsequent growth. According to their results, the
relationship runs from higher sophistication to growth. Although Burundis EXPY has increased, it
remains low, thus demonstrating the limited income potential of the countrys export basket. Burundi lags
behind the development in other EAC member-statesespecially Tanzania and Uganda, which also
started at rather low EXPY levels in the 1980s but have been successful in exporting larger shares of
high-PRODY products in recent years (figure 2.4).
Classification of Burundis Exports

2.30 Burundis exports are dominated by a few classics, and there is only a very limited number
of emerging champions. Based on the development of their revealed comparative advantage (RCA)
between 1985 and 2006, products can be classified in four categories: classics, disappearances, emerging
champions, and marginals.16 A summary of Burundis exports, classified according to this approach, is
given in table 2.2.17
13
PRODY is calculated as the weighted average of the GDP per capita of the countries exporting the product.
14
As mentioned above, flower exports have suffered significantly during the crisis. Flowers had only been supplied to one buyer,
who cancelled orders in response to low demand.
15
EXPY is computed as the weighted average of the PRODY, with the export shares of the respective country as weights.
16
A country has an RCA if its share of a certain product in its total exports is greater than the share of that product in global
exports (Balassa 1986). If a product has an RCA less than or equal to 1, it is assigned a value of 0; an RCA greater than 1 is
assigned a value of 1. For a more detailed description of the applied framework, see Chandra and Osorio Rodarte (2009) or
Boccardo and Chandra (2008).
17
A detailed table is given in table A.2 in appendix 2.
22
Table2.2:ClassificationSummaryofBurundisExports
Exports(US$thousands) Shares(%oftotalexports) PRODY RCA
Category 198589 200506 198589 200506 (US$,PPP) 198589 200506
Classics 131,925 58,750 95.53 93.18 3,190 1 1
Disappearances 1,539 22 1.05 0.01 5,036 1 0
Emergingchampions 1,030 915 0.62 1.62 4,304 0 1
Marginals 2,655 1,558 1.07 2.03 5,876 0 0
Source:COMTRADE,SITC24digit;calculationbyauthorsandtheEconomicPolicyandDebtDepartmentoftheWorldBank.
Note:PPP=purchasingpowerparity.
2.31 The classics, products for which Burundi has maintained an RCA, are dominated by coffee,
followed by tea and nonmonetary gold. Smaller shares can be attributed to fresh fish; plants, seeds, fruit
used in perfumery and pharmacy; and goat, kid, sheep, and lamb skins (between 1 percent and 2 percent
of total exports). The classic products account for more than 90 percent of exports and have a comparably
low PRODY, on average. Their performance has been disappointing in recent years, with the value of
total exports dropping to about half of the profit realized in 198589.
2.32 The only Burundi products that have been identified as disappearances are bovine and equine
hides as well as bones, horns, and the like. Those products had an RCA of 1 in the past, but lost their
comparative advantage.
2.33 Products are classified as emerging champions if their RCA was 0 in the past, but turned into 1 in
200506. Their share of total exports is still very small, but increasing. Only very few emerging
champions have been identified, including ores and concentrates of other nonferrous metals; cut flowers
and foliage; bulbs, tubers, and rhizomes of flowering or foliage plants; and leather or other hides and
skins.
2.34 Many products have been classified as marginals. The share of total exports, however, remains
low, at less than 2 percent. A number of marginal products have the potential to develop into emerging
champions and contribute notably to the diversification of the export base. The list contains several high-
PRODY products, a number of them high-value primary products (such as bananas, other fruits, and
vegetables) and resource-based products (such as sawlogs and different alloys).
A5. Weak Private and Financial Sectors

2.35 Burundi has one of the most unfavorable business environments in the world. The World
Banks 2010 Doing Business survey ranked Burundi 176 out of 183 countries18 surveyed in their Ease of
Doing Business Ranking (World Bank 2009a), near the bottom relative to other countries assessed in the
areas of dealing with permits, getting credit, protecting investors, trading across borders, and enforcing
contracts. Another cross-country assessment of business conditions ranked Burundi 133 out of 133
countries surveyed in its Global Competitiveness Index.19 Survey respondents highlighted corruption,
political instability, inefficient government bureaucracy, and crime and theft among their top concerns.
2.36 As a consequence, the country suffers from low productivity growth, thereby impeding
private sector efficiency and productivity gains and ultimately slowing overall growth. The results of
an enterprise survey carried out in 2006 showed that domestic firms clearly have a negative perception of
Burundis business climate (World Bank 2008a). The main constraints cited by managers of formally
registered firms were access to electricity, access to and cost of financing, political and macroeconomic
18
Burundi is only followed by Repblica Bolivariana de Venezuela, Chad, the Republic of Congo, So Tom and Principe,
Guinea-Bissau, the Democratic Republic of Congo, and the Central African Republic.
19
The World Economic Forums 200910 Global Competitiveness Report, which assessed and compared 133 countries based on
both surveys and quantitative measures. The Global Competitiveness Index is a composite measure of individual indicators,
including specific indicators related to the quality of institutions, infrastructure, economic stability, health and education, and the
like. For details, see http://www.weforum.org/pdf/GCR09/GCR20092010fullreport.pdf.
23
instability, tax rates, and business practices of informal competitors. Findings from a survey of the urban
informal sector suggest the lack of infrastructure, access to finance, insufficient and volatile demand, and
lack of professional and technical skills as the main constraints. According to the Heritage Foundations
2011 Index of Economic Freedom, Burundi reaches a score of only 49.6 out of 100.20 Poor governance
also impedes the development of the private sector. Implementing fundamental structural reformswhich
has already begunis necessary for private sector-led growth.
2.37 The Burundian private sector is underdeveloped. It includes a large number of small
enterprises, most of which operate in the informal sector. The formal private sector includes about 3,000
registered enterprises, employs some 37,000 workers, and produces mainly for the local market.
Construction, agricultural processing, brewing, energy, and communications are the main activities of the
very small industrial sector. The service sector is more important, but it is dominated by growing public
services, accounting for about two thirds of the sector.
2.38 The development of the private sector is hampered by several constraints. They can be
broadly grouped into four categories: (1) structural constraints, (2) supply-side constraints, (3) poor
regulatory environment, and (4) poor governance. The government has recognized the situation and
started the reform process, but much remains to be done.21
Structural Constraints
2.39 Structural constraints include the landlocked position of the country, the lack of market
links, and the poor condition of the infrastructure. According to the latest Investment Climate
Assessment (World Bank 2008a), the lack of reliable infrastructure is seen as the main constraint for
enterprises. Eighty-seven percent of manufacturing establishments cite the provision of electricity as a
major or very severe constraint, and transport infrastructure is considered a major or very severe obstacle
by more than 27 percent of respondents. The severity of Burundis infrastructure gap and how it
negatively affects the country are outlined in section B of chapter 3.
Supply-Side Constraints: Finance and Workers

2.40 The main supply-side constraints are the limited access to and high cost of finance, the
limited availability of skilled workers, and low labor productivity of labor
Access to Finance22
2.41 The financial sector of Burundi is rather small and lags behind those of its neighbors in the
EAC. The commercial banks suffer from high numbers of nonperforming loans and internal weaknesses,
although there are prospects for changes with the recent entry of foreign banks; the payment system is
archaic, and there is no card payment system; the largest microfinance network is facing difficulties;
insurance companies operate without a license and without supervision; and access to finance for housing,
small and medium enterprises, and agriculture is quite limited. Consequently, the financial sector is in no
position to play its role efficiently in Burundi.
2.42 Access to financial services is limited. Available data show that only about 1.90 percent of the
total population have bank accounts, 0.42 percent use bank credit services, and 4.00 percent are members
of microfinance institutions. Postal checking services manage about 120,000 accounts, mainly those of
civil servants. Movements on these accounts are rare, except at month-end when wages are paid. There is
20
This score makes Burundi the 148th freest country (out of 179 countries) and the 31st out of 46 countries in Sub-Saharan
Africa (Heritage Foundation 2011).
21
Currently being prepared is a World Bank project that will provide $20 million to support the development of the financial and
private sectors. This section draws on the findings and recommendations outlined in the relevant project appraisal document.
22
This section benefits from the results of the International Monetary Funds and World Banks Financial Sector Assessment
Program.
24
also a high concentration of bank branches in urban areas. Given most peoples lack of access to
commercial bank services, microfinance institutions have an important role to play. However, nearly 80
percent of members/customers are wage earners. Thus, the industry continues to exclude a majority of the
population (nonwage earners) while serving a market that is uncertain for it in the medium term
(uncertain because the banking industry could rather easily serve these customers if it makes that its
objective). The financing of agriculture or production in general remains marginal. Because of difficulties
in obtaining access to credit and the high cost of it, 74 percent of microenterprises and informal firms and
67 percent of enterprises in the manufacturing sector use internal funds or reinvested earnings as their first
and foremost source for financing assets. Sixty-five percent of firms that needed financing in 2006 did not
even request a loan, mainly because of the high interest rates, the guarantee requirements, and the
complexity of the process (World Bank 2008a).
2.43 Burundis performance trails that of countries at a similar level of development.
With 4.0 accounts per 1,000 inhabitants, Burundi is below the average for such countries as the Central
African Republic (5.6 per 1,000), Uganda (5.8), and Cameroon (14.4). This low level reflects the
widespread use of cash. Fiat money represented 38 percent of the money supply as of December 31, 2007.
2.44 The distribution of bank credit is not consistent with the structure of the economy. A
breakdown of credit by
economic sector shows Table2.3:SectoralDistributionofBankCredit
a large concentration
Sector 2006 2007 2008
in the commerce FBu Share FBu Share FBu Share
sector, millions (%) millions (%) millions (%)
representing 60 percent Agricultureandrelatedbusiness 22,143 11 8,587 4 27,544 10
of the outstanding Industry 1,976 2 3,66 2 5,431 2
Construction 9,826 5 6,325 3 10,251 4
credit of all banks at
Commerce 132,261 67 144,118 70 158,295 60
end-2008 (table 2.3), Miscellaneousservices 30,350 15 43,880 21 61,907 24
whereas this sector Total 196,556 100 206,576 100 263,428 100
accounts for less than Source:BRB.
10 percent of GDP.
Financing of agriculture increased from 4 percent to 10 percent of total credit, whereas agricultures share
of GDP was more than 40 percent.
2.45 The financial products supplied by banks are little diversified. Bank credit is generally short
term. Other financial services are marginal. On the deposit side, demand deposits are prevalent, for which
remuneration can be freely set; and within time deposits, those maturing in less than a year seem the most
widespread. In 2007, broad money represented 35 percent of GDP, which is short of the 41 percent
average for Sub-Saharan Africa and well below that of South Africa, (where it represents 56 percent).
Microfinance institutions have limited products, are not allowed by law to offer leasing services or
mortgage loans, and cater mainly to the needs of wage earners.
2.46 The Investment Climate Assessment conducted in 2006 (World Bank 2008a) confirmed that
the lack of access to bank credit is a major obstacle. Cost of and access to bank credit appears to be a
major constraint for 69 percent of manufacturing firms and 43 percent of firms in the trade sector. For
these firms, cost and access rank second as a major concern, behind electricity. Cost and access are the
most important constraint for 68 percent of firms from the informal sector. Compared with other EAC
countries, Burundi is in the worst situation.
2.47 Fully aware of the problems, the authorities already have taken a number of measures. The
country has a new banking law, a new central bank charter, and a new law on microfinance. The central
bank has also put into place the supervision of microfinance institutions. The authorities requested that the
International Monetary Fund and the World Bank undertake a Financial Sector Assessment Program
(FSAP). The program took place in January 2009.
25
Limited Availability of Skilled Workers and Low Productivity of Labor
2.48 Burundis labor market regulations are relatively adaptable, according to the Investment
Climate Assessment from 2006. In fact, issues related to the labor market (difficulty in hiring, rigidity of
hours, redundancy costs, and so forth) are the least important obstacles (World Bank 2008a), and the
labor rigidity index is low compared with other Sub-Saharan African countries. Low wages prevail, and
that could be seen as a comparative advantage in the short run.
2.49 There is, however, an urgent need to improve technical and professional skills in the labor
force. Currently available education does not seem to provide the private sector with enough skilled
labor. Nearly all employees surveyed for the assessment report the need for training (94 percent), mostly
in technical, professional, and production competencies. Furthermore, language skills (French and
English) and information technology skills have been cited as required training.
2.50 There is also a need to improve the flow of information on the labor market, especially
regarding available positions, to better match supply and demand. Currently, there is no formal
system in place, and positions are communicated only through social networks. To ensure equal access to
the labor market and a reliable flow of information, objective channels of dissemination must be
established.
Poor Regulatory Environment

2.51 Administrative and bureaucratic procedures in Burundi are cumbersome. Paying taxes in
Burundi is complicated by the multiplicity of taxes that discourages the formalization of businesses and
leaves room for uncertainty and corruption. Microenterprises see registration formalities as regulatory
deterrents to entering formal markets. Thus, small, informal firms established by such entrepreneurs are
unable to grow and benefit from economies of scale or to access additional benefits of formalization (such
as commercial bank finance and business development services). Investment regulations are burdensome
and discourage foreign direct investment.
Box 2.2: Burundis Tax Regime

Burundis current tax regime is riddled with large tax exemptions that make it overly complex and opaque and lead
to important distortions that negatively affect producers. Tax exemptions are widely perceived as inefficient, leading
to discretionary decisions and delays. The various exemptions (which mainly concern imports) not only reduce
revenue collection, but also make tax administration more complex. According to a recent study, 60 percent of
imports entering Burundi in 2006 had partial or total exemptions, costing FBu 103.7 billion (that is, 10.7 percent of
GDP and 65.6 percent of tax receipts).
The governments decision to introduce a broad VAT, together with membership in the EAC, will lead to broad
changes in the tax regime. Under the assumption of a broad-based VAT introduced at 17 percent and without EAC
adhesion, it has been estimated that it will lead to a reduction in tax revenue of FBu 1.6 billion, equivalent to 0.2
percent of GDP. However, the combined effect of introducing the VAT and joining the EAC is expected to have a
large negative effect on overall receiptsestimated to FBu 15 billion (1.5 percent of GDP). This anticipated
negative effect highlights the importance of introducing a broad-based VAT regime, reducing import exemptions,
and improving tax administration.
2.52 Tax policy is not conducive for the development of the private sector. The tax burden is one
of the highest in the world, and consequently is seen as one of the main obstacles for the private sector.
According to the Doing Business 2010 report (World Bank 2009a), the total statutory tax rate over profits
ratio is 279 percent, which may lead to tax evasion. The tax regime would need to be simplified and tax
exemptions avoided. The value-added tax (VAT) of 18 percent was recently introduced (see box 2.2), but
it needs to be accompanied by information campaigns and taxpayer training.
26
2.53 A revision of legislation and codes is necessary. Burundis legal, regulatory, and procedural
framework is outdated. Many key laws have been revised recently, but are often not accompanied by
appropriate regulations, which causes no or arbitrary implementation.
2.54 The government already has taken important steps in the right direction, but much remains
to be done. In June 2008, the Council of Ministers approved a new investment code that includes some of
the best practices in investment legislationamong them, investor protection, freedom to transfer capital
and dividends, and a streamlined approval process. The investment code was declared law in September
2008 and has been implemented, providing additional incentives to investors. Unfortunately, the revision
of the tax code has been significantly delayed. A competition law is being drafted, and two insolvency
laws were adopted recently. The draft versions of the commercial code and the code for public and private
companies were revised, and (in line with international standards and the Organization for the
Harmonization of Business Law in Africa [Organisation pour lHarmonisation du Droit des Affaires en
Afrique OHADA] recommendations) both have been adopted by the Council of Ministers and submitted
to Parliament. The mining code is currently under revision. It should prepare the country for potential
large-scale investments and clarify the role of artisanal and small-scale mining.
2.55 The capacity to enforce contracts is weak. The economy is characterized by a weak application
of the laws for businesses; and by weak capacity to enforce commercial contracts and have commercial
disputes resolved fairly, predictably, and professionally in the commercial court. Burundi has only one
commercial court in Bujumbura, and it lacks adequate facilities, equipment, and skills despite receiving
training and equipment through the Project dAppui la Gestion Economique (PAGE). The high turnover
of magistrates also contributes to the courts poor performance in handling its caseload. The government
has taken a number of measures to improve the performance of the court and other sector institutions.
These measures include salary increases to discourage corruption, increases in the number of magistrates,
and a reform of the rglement intrieur (internal rules). The percentage of cases whose duration exceeds
60 days decreased from about 42 percent in March 2008 to roughly 22 percent one year later. Moreover, it
fell to 14.3 percent and 10 percent in May and July 2009, respectively.
2.556 In an effort to provide quicker delivery of commercial decisions, an arbitration center was created
in 2002 and the appropriate legislation was adopted in 2008. Almost 30 arbitrators from various sectors
have been trained, and some public education has been initiated. However, although the center has
received some start-up funds from PAGE, it lacks the financial resources to further develop this tool for
dispute resolution.
A6. Poor Governance

2.57 After a prolonged period of conflict and ethnic tension, improving governance, government
decision making, and the capacity to raise taxes are critical to economic recovery. Poor governance
and corruption are also major obstacles to private sector development; and they lead to lower investment,
both domestic and foreign. Poor governance is difficult to measure; even so, survey-based assessments of
perceptions among both citizens and businesses indicate problem areas.
27
2.58 A survey commissioned by the
Figure2.5:GovernanceIndicators,1998,2000,2003,and2008
government of Burundi through the
World Bank in November 2007 found
corruption, access to and abuses of
the judicial system, and security and
criminality to be among the
respondents main concerns (Republic
of Burundi 2008a; World Bank 2008b).
Poor governance has been identified as
one of the main reasons for the
prolonged civil conflict in Burundi. The
same study also found that the majority
of households, as well as many
representatives of firms and
nongovernmental institutions, had
doubts regarding the ability and
willingness of the government and its
political representatives to promote
economic and social development. It is
obvious from figure 2.5 that the Source:Kaufmann,Kraay,andMastruzzi2009.
situation in Burundi has improved since Note:Theconfidencelevelis90percent.
the end of the civil war, but the country
still ranks among the worst performers in terms of governance.23 The Ibrahim Index of African
Governance yields similar results. Burundi ranks 38th among 53 African countries in the 2009 Ibrahim
Index, showing some improvement to previous years (Mo Ibrahim Foundation 2009).
2.59 With respect to corruption, the study cited many specific factors that have exacerbated this
problem (Republic of Burundi 2008a). These factors include extreme poverty, low government salaries,
and limited access to basic and public services, which have worsened incentives with respect to bribes.
The survey also revealed that the police force, customs, tax collection, public procurement, and justice
system were prone to corruption, with the police force singled out as the most frequent recipient of bribes.
Other issues (such as public employment practices) also were identified as areas of concern because
respondents saw political party affiliation or personal connections rather than technical competence as
motivations for government hiring. Transparency Internationals recently released Corruption Perception
Index ranks Burundi 168th among 180 countries, representing a considerable deterioration from rank 131
in 2007 and rank 158 in 2008.24
2.60 Public enterprises dominate the economy. A 2007 study (Republic of Burundi 2007a) found
that the government had shares in 48 entities (40 enterprises and 8 financial institutions), 16 of which
were fully publicly owned. Public enterprises also dominate the export sector (coffee in the process of
being privatized, tea, cotton, sugar). Several public enterprises are faced with a difficult financial
situation. They are burdened by the governance problems typically encountered in many countries,
including political interference from the government. Their indebtedness increases faster than their
turnover, and debt obligations often exceed cash flow. By the end of 2006, the overall debt of the 21
23
The governance indicators presented here aggregate views on the quality of governance provided by a large number of
enterprise, citizen, and expert survey respondents in industrial and developing countries. These data are gathered from a number
of survey institutes, think tanks, nongovernmental organizations, and international organizations. The Worldwide Governance
Indicators do not reflect the official views of the World Bank, its executive directors, or the countries they represent. These
indicators are not used by the World Bank Group to allocate resources.
24
For details. see Transparency International,
http://www.transparency.org/policy_research/surveys_indices/cpi/2009/cpi_2009_table.
28
largest public enterprises amounted to FBu 128 billion (about 14 percent of GDP), including debt to
suppliers, the public sector, foreign lenders, and local banks. The Service Charg des Entreprises
Publiques (SCEP), a special entity created in the 1990s to oversee public enterprises, does not have the
capacity and the resources necessary to bring about substantial improvements in the performance of the
sector and to plan and monitor the privatization program envisaged by the government. The government
recognizes that it needs to retrench from productive sectors, but the existing analytical base does not
provide sufficient information to fully understand the financial situation of many public enterprises, its
impact on government finance, and the best options for reforms.
2.61 Poor governance and government decision-making capacity imply lower productivity for all
sectors of the economy and inhibit new investment in physical and human capital. Given the
importance of government spending in the post-2000 economy, increasing the efficiency of spending is
necessary to obtain the highest results from aid spending programs. Poor governance is currently an
obstacle to this effect. In addition, the governments capacity to make decisions regarding policy,
spending, and tax revenue is weak. This state of affairs fosters a perception that recovery is weak and
causes delays in private sector response. The slow decision-making process regarding the reform of the
coffee sector is a clear example of this problem and a major determinant of the slower response of the
economy of Burundi, relative to Rwanda or Sierra Leone, following the reestablishment of peace.
2.62 Low government decision-making capacity is a critical barrier to development. Low

decision-making capacity is often the result of political dissent within a society and of maneuvering by
interest groups. But it also is the result of low human capital and disorganization. One strategy that
several small states use to circumvent this barrier is to outsource the provision of some government
services to regional organizations. Burundi is moving in this direction by joining the EAC. But there are
areas of government decision making that cannot be outsourced. Meeting the challenge requires
reorganizing current capacity and prioritizing its use.
A7. Low Levels of Education and Health

2.63 To support higher levels of growth and development, one first-order requirement will be to
invest in improvements in health and human capital. By most measures, Burundi lags behind the
world and behind other countries in the region in these and related areas. Deficits in these areas clearly
stem from the countrys low levels of urbanization, income, and development. In this context, Burundi is
not likely to meet many of the Millennium Development Goals (which focus on health and education) by
2015 (World Bank 2008b). An estimated 67 percent of the population lives below the national poverty
line; and the country displayed a Gini coefficient of 0.38 percent based on data available in 2008,
illustrating the fact that poverty is widespread.25 As noted above, Burundi ranked 174th among 182
countries on the United Nations Development Programmes Human Development Index in 2008.
2.64 Low educational indicators reveal severe limitations to increasing productivity and growth
potential. In 2006, only about half of the population was able to read and write in the national language
Kirundi, and only about one third was literate in a foreign language (IMF 2009). Although enrollment
rates for primary schools have improved tremendously in recent years (to more than 70 percent), partly as
a result of the introduction of free primary education in 2005, completion rates have remained low (at
about 40 percent) (World Bank 2008b). Secondary enrollment stood at just 19 percent in 200607,
reflecting high dropout rates, poor infrastructure and access to resources, and high poverty in rural
regions. At the tertiary level, evidence suggests that performance is even worse, with university
enrollment rates estimated at 2.6 percentvery low even by regional standards. Although private
universities have become more prominent in recent years, a review of the PRSP in 2009 emphasized that
25
The Gini coefficient is a widely used measure of income inequality, with a scale ranging from 0 to 1. A score of 0 indicates
perfect equality (every household with equal earnings), and a score of 1 indicates the inverse (a single household earns all of the
national income).
29
poor academic infrastructure at the tertiary level is a serious impediment to improving educational
outcomes and to increasing productivity and economic potential in Burundi.
2.65 Against this background, education is a key means of broadening opportunities for the
population, of strengthening integration of the society as a whole, and of encouraging deeper
integration between Burundi and the rest of the world. The challenge is how to accomplish this in a
country where there are few qualified teachers and the literacy rate in English and/or French is extremely
low. In this context, modern information and communication technology and programs designed to
promote the role of women in the household may be crucial ingredients in addressing the education
challenge.
2.66 This is especially critical in light of the high population growth rate. The fertility rate in
Burundi stands at 5.4 children per woman, and the net population growth rate increased to 2.6 percent
between 1950 and 2005. The population of Burundi is expected to grow rapidly in the next 40 years.
Rapid population growth produces several problems. First, it prevents countries from building their
human capital and achieving their education for all target, a prerequisite for economic development.
Unless fertility levels decline, attaining the Millennium Development Goals will not be possible. Second,
high fertility levels jeopardize poverty reduction efforts, particularly among people in poor quintiles. Poor
households suffer most because high fertility is an additional burden with regard to access to quality
education and health services (World Bank 2007). Third, rapid population growth elevates the pressures
that countries face regarding food security, land tenure, water supply, and environmental degradation.
Slower population growth eases the security problems that often result from conflicts over scarce
resourcesconflicts exacerbated by unsustainably high rates of population growth and widespread youth
unemployment. Fourth, high fertility levels have detrimental consequences for the health of women and
children.
A8. The Challenge of the Demographic Transition

2.67 Today, it is widely recognized that a countrys accelerated demographic transition and
faster fertility decline may positively influence its socioeconomic outcomes (May 2009). The
demographic transition or modernization is defined here as the gradual shift from high crude birth rates
and high crude death rates, in equilibrium, to low levels of fertility and mortality that reach a new
equilibrium. In Sub-Saharan Africa, these demographic changes have not occurred in earnesthence, the
sluggish economic growth experienced in this region, as stressed in the seminal study by Ndulu et al.
(2007). The relationship between rapid population growth and socioeconomic development is most
noticeable in four different areas, namely, economic growth and human capital formation, poverty
reduction, environmental protection, and health status.
2.68 The current population of Burundi is estimated at 8.3 million (mid-2009 estimate) and is
very young, with 41 percent of the Burundians below age 15 (Population Reference Bureau 2009). The
population of Burundi has increased inexorably since 1950, when it was estimated at 2.5 million. It has
almost quadrupled in a period of 50 years. The young age of the population has far-reaching implications
for the formation of human capital, particularly for education and health. Given the small territory of the
country, the already high population density can only be expected to increase. Greater density will add to
the pressure on the agricultural sector and ecosystems unless migratory movements take place to
neighboring and/or more distant countries.
2.69 In absolute terms, the Burundian rate of net population growth has increased since 1950. It
increased from 1.8 percent in 1950 to 2.6 percent in 2005 (United Nations 2009). However, it is important
to note that given Burundis history of conflict (which peaked in the early 1970s and in the 1990s), the
rate of net population growth (taking migration into account) was not regular at all. The periods of
conflict are clearly characterized by much lower rates of population growth as a result of increased
30
mortality rates from direct war and war-related causes, probable postponement of births, and migratory
movements.
2.70 Mortality levels and fertility are very high, but appear to be slowly decreasing despite sharp
variations. The infant mortality, under-five mortality, and maternal mortality rates remain quite high.
Life expectancy for both sexes at birth is low: 49 years, compared with the Sub-Saharan African average
of 51 years. However, the HIV/AIDS epidemic appears to be somewhat under control. Fertility is
estimated at 5.4 children per woman and the slow trend is generally downward, despite some variations
over time. Burundis total fertility rate is close to the Sub-Saharan Africa average rate of 5.3 children per
woman. However, the contraceptive prevalence rate for modern methods was estimated at only 8 percent
of married women.
2.71 As a result of mortality and fertility global trends, the population of Burundi will continue
to grow rapidly during the next 40 years. The current rate of population growth is estimated at 2.1
percent per year, leading to a doubling time of 33 years (Population Reference Bureau 2009). The
medium variant of the United Nations 2008
population projections indicates that the population Figure2.6:PopulationProjections,200550
of Burundi should reach 14.8 million in 2050,
18
assuming that the total fertility rate declines from its lowvariant
mediumvariant
current level of 5.4 children per woman to 2.3 16
highvariant
children per woman in 204550 (figure 2.6). 14
However, the 2050 population would be 13 million if millions
12
fertility declines to 1.8 children per woman in 2045
50, but almost 17 million it declines only to 2.8 10
children per woman during those years. These 8
United Nations projections assume significant
improvements in the mortality levels: the expectancy 6
of life at birth would increase from 50.3 years in 2005 2010 2015 2020 2025 2030 2035 2040 2045 2050
200510 to 65.7 years in 204550. Finally, net Source:UnitedNations2009.
migration is assumed be negative as of 2010 onward.
However, one could also assume that net migration could become positive if more Burundians decide to
return to their country of birth (United Nations 2009).
2.72 Even if fertility declines dramatically and quickly reaches the replacement level, Burundis
population will continue to grow and probably double during the next 70 years or so. This situation
is caused by the youthfulness of the age structure and the phenomenon called the population momentum.
The population momentum occurs when the very large number of people in the reproductive age groups
continues to fuel the population growth, even though fertility levels decline.
2.73 The trauma caused by Burundis checkered political history has put any meaningful
discussion of population issues on hold for almost 20 years. As a consequence, population and
reproductive health issues have been neglected. First, Burundi has not undertaken or completed the major
data collection operations that would have provided an adequate base of knowledge about the
demographic situation in the country. Second, the population and reproductive health issues have not been
addressed forcefully in Burundis development strategies and poverty reduction documents. The Poverty
Reduction Strategy Paper has highlighted the issue of rapid population growth, but has not proposed clear
strategies and programs to deal with it. Third, although Burundi is preparing a national population policy
(with the support of the United Nations Population Fund), there is no guarantee that this document will go
much beyond similar population policies drafted in the past decades in Sub-Saharan Africa. Moreover,
given the difficult political context, the population policys implementation remains a challenge.
2.74 Several steps are necessary to address the challenges related to the demographic transition,
including
31
gathering additional information on demographic trends and gaining a better understanding of the
demographic challenge;
repositioning population and reproductive health issues within development strategies and
strengthening population institutions;
developing a comprehensive national population policy;
revamping Burundis family planning program to accelerate the demographic and fertility
transition;
developing new strategies of family planning distribution; and
strengthening the health system, which could bring substantial gains in the supply of family
planning services.
2.75 The experience of other countries shows that successful population policies can be
implemented in Sub-Saharan Africa too. Necessary, however, is a high level of commitment and a
strong implementation of supply-side family planning programs. In Rwanda, a new national population
policy is currently being prepared, superseding the one enacted in 2003; and political support for family
planning programs is strong. Huge gains in the supply of family planning services have been made
possible by strengthening the health systems. As a result, Rwanda made impressive progress in the
contraceptive prevalence rate between 2005 and 200708, accomplishing more than a third of what is
required to reach the replacement level of fertility.26
2.76 Burundi needs to address its high population growth rate. The attention paid, or unpaid, to
population issues will predicate the development prospects of the country for many decades to come. The
fertility decline will affect most development sectors positively. In particular, it will facilitate formation
of human capital, help reduce poverty levels, mitigate the pressure on the environment, and enhance the
health of women and children. At this juncture, Burundi needs support from its development partners to
design the needed policies and programs to address this fundamental challenge.
A9. Conclusion
2.77 Identifying the most binding constraint to growth by analyzing the economic dynamics of
the country has proved difficult. Given the set of constraints Burundi faces, it is difficult to single out
one specific area that can be considered the most binding constraint. The key problem areas are linked,
creating negative synergies. Low productivity in agriculture, for example, can surely be seen as one of the
main barriers to development. Likewise, the large infrastructure deficit (especially the low supply of
electricity and the high transport costs) prevents the development of almost all sectors. The small size of
the country, coupled with the high population growth rate, necessitates creation of off-farm employment;
development of that employment is hampered, in turn, by the undiversified economy, the infrastructure
deficit, and low educational attainment. Poor governance and the dismal business environment aggravate
the situation by deterring investment, which could create additional employment and income. Burundi
therefore needs to address some of the most binding constraints simultaneously. As a necessary
prerequisite, peace, security and stability must be maintained.
B. OPPORTUNITIES IN POTENTIAL GROWTH SECTORS

2.78 Burundi is a mostly rural economy. It is highly dependent on the mainly subsistence-oriented
agriculture sector. The country relies on small-scale farming and focuses mostly on food crops. Despite
increasing demand from a high population growth rate, productivity is low and has barely improved over
the past four decades. Food insecurity is widespread, and a large part of the population is undernourished.
As a result, there is an urgent need to improve the performance of the agriculture sector to stabilize it and
26
For more details, see May and Kamurase (2009).
32
increase its growth. The same applies to the currently exported crops (mainly coffee and tea), which are
suffering from low production volumes and low quality. The productivity of the food crops subsector
urgently needs to be improved to increase food security and improve the nutrition of the population. That
will lead, in turn, to a healthier and more productive population.
2.79 Burundis economy is not diversified. Burundis product and export base is currently extremely
concentrated. Efforts should be undertaken now to pave the way for the diversification of the economy to
achieve more broadly based growth, to reduce the likelihood of negative growth periods, and to create
employment opportunities. Burundi has the potential to diversify in different areas, most notably in agro-
processing and mining. Although this is an ambitious objective that can only be achieved over the long
term, the course should be set now to transform the economy.
B1. Agricultural Recovery: Food CropsPerformance and Challenges

2.80 The World Bank recently Figure2.7:ProportionofIndividualsLivinginHouseholdswithaDaily
conducted a comprehensive study CaloricIntakeperAdultbelow1,900kcal
addressing the sources of rural growth,
with a focus on agricultural recovery
(Baghdadli, Harborne, and Rajadel
2008). In addition, the government of
Burundi developed a National
Agricultural Strategy in 2008 and a
National Food Security Program. The
following section provides a summary of
the main findings and recommendations
of these documents.
2.81 The contribution of food crops
to GDP is significant. Food crops
contribute between 35 and 40 percent of
GDP and 80 percent of agricultural
output. Food crops occupy about 85
percent of the cultivated land and 28
percent of total land. There is a high
incidence of malnourishment, which
seriously affects the health and
productivity of the population.27
Consumption of own production
accounts for nearly three quarters of the
average rural households income.
Poverty is widespread in rural areas,
where most of the population lives.
Severe land constraints28 have led to an Source:Baghdadli,Harborne,andRajadel2008.
overexploitation of the land, which has
further contributed to land degradation and soil erosion.
2.82 Burundis food supply is not meeting the increasing demand, leading to food insecurity and
the need to rely on food aid. Figure 2.7 shows the proportion of individuals living in households with a
27
According to the 2008 Global Hunger Index, the situation is only worse in Eritrea and the Democratic Republic of Congo. The
severity of the situation in Burundi is considered extremely alarming (von Grebmer et al. 2008).
28
With 257 inhabitants per square kilometer, Burundi exhibits one of Africas highest population density rates.
33
daily caloric intake below the required minimum of 1,900 kilocalories.29 The World Food Program
provides monthly assistance to an average of 635,000 people, mostly in the food-insecure north and
northeastern areas of the country.
Figure2.8:FoodCropProduction,19902007
2.83 Production has stagnated (figure 2.8)
while the population has continued to
4,000
increase, leading to a constant decline in
3,500
production per capita. The main food crops
thousandsoftons
3,000
(ranked by production) are bananas, roots and
2,500
2,000
tubers, legumes, cereals, vegetables and fruits,
1,500
and oilseed. The use of land is already at its
1,000 limits (using traditional technology), and the
500 average size of a plot is only 0.5 hectare. The
0 soil is characterized by low fertility, about a
third of the soil is acidic, and the hilly terrain is
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
prone to soil erosion. The irrigation potential is
Cereal Legumes Rootsandtubers Bananas mostly underexploited, with less than 10
Source:RepublicofBurundi2008b.
percent of the potentially irrigable area under
irrigation.
2.84 The agriculture sector is not well organized. The majority of producers in Burundi are
households, and only a minority of them has joined a producer organization. The production and
distribution of agricultural inputs (seed, fertilizer, crop chemicals, and machinery) are not well organized
and managed and thus contribute to low usage and high input prices. Only limited processing takes place
(mainly at the household level), and there is almost no industrial processing. Most of the food crops are
marketed only at a local level, and food markets are generally underdeveloped. Because most households
engage in subsistence farming, only a modest amount of produce gets to the market. Farmers are poorly
organized and have only very limited access to information on market developments, opportunities, and
prices.
2.85 The lack of infrastructure and extension services hampers the development of the sector.
The insufficient supply of electricity and the poor condition of rural roads are obstacles to an increase of
production as well as marketing of the products. Extension services, including agronomic research, are
weak, as are training, input supply, and distribution.
2.86 Overall, the food crop subsector in Burundi faces a number of challenges that can be grouped
into the following five main categories:
1. Agronomic constraintsThe fertility of the soil is poor, and degradation of the ecosystem is
progressing. There is only limited use of inputs because of a deficient distribution system and
high input prices. Food crops production relies mainly on traditional extensive cultivation
methods, whereas commercial agriculture based on modern technology and inputs remains
underdeveloped.
2. Climatic constraintsClimate change severely affects agriculture in Burundi. Although the
country generally receives more rainfall than most African countries, rainfall shortage has been
observed, leading to volatile production.
3. Technological constraintsThere is weak agricultural research and a poor system to distribute
knowledge about innovative technologies. Specific issues include limited use of improved
management practices (water management, irrigation, transformation, and conservation of
29
This estimate is made using the Questionnaire des Indicateurs de Base du Bien-Etre (QUIBB) 2006 household survey data (see
Baghdadli, Harborne, and Rajadel 2008). The suggested minimum intake is between 1,900 and 2,100 kilocalories.
34
products); limited promotion of alternative and renewable energy sources; and insufficient human
resources and laboratories for quality control of inputs, food and cash crops.
4. Institutional constraintsExtension services (including research, postharvest transformation and
conservation, production and distribution of agricultural inputs) are inefficient. There is a lack of
skills and human capacity. The involvement of the private sector is limited. Weak coordination
and lack of harmonization of interventions persists.
5. Socioeconomic constraintsOne of the main constraints is a lack of resources to rehabilitate
infrastructure and equipment to support agricultural production, reforestation of the damaged
ecosystem, and a reactivation of extension services after the long civil war. Furthermore, the
population increase puts pressure on the use of existing land. Access to finance is limited for rural
households, and the business climate is not conducive to private investment.
Box 2.3: Land Tenure and Economic Growth

Uncertain land tenure rights delay the resolution of land-related conflicts. The legal environment fosters uncertainty
over land tenure rights. The first obstacle to the implementation of an efficient legal land system in Burundi stems
from the coexistence of customary and statute law. For instance, according to customary law, farmers who have
cultivated and managed plots over a certain period of time earn ownership rights to those plots. This system has led
to harsh conflict over the status of some state-owned lands, particularly in swamp areas. Moreover, customary law in
some regions separates ownership of the land from ownership of what stands on the land; statute law does not
recognize such a distinction. This difference has fueled bitter disputes over coffee or oil palm tree plantations.
Womens access to property is another matter of great concern because, according to customary law, family land
cannot be transmitted to daughters. Statutory law, however, completely contradicts customary law in this matter.
Finally, issues pertaining to the resettlement of long-term refugees provide yet another striking illustration of
contradictory legal frameworks. According to the Arusha Accords, long-term refugees should be given the right to
retrieve their property; if such retrieval is impossible, they should receive fair compensation. However, statute law
stipulates that a 30-year prescription applies to land-related claims. The situation is further complicated by the
somewhat unclear answers provided by customary law. It does not recognize the 30-year prescription, but it
acknowledges that long-term occupants have earned some rights over the land on which they live.
In addition, the efficiency of the judiciary and institutional systems is gravely affected by an unclear delineation of
responsibilities and overlapping functions. The judiciary system intervening in land disputes relies on three entities:
the Bashingantahe, local authorities who arbitrate conflicts using statue law as a reference, and local tribunals
(tribunaux de rsidence) who return verdicts based on statute law. The institutional level is as confused because at
least four different ministries are involved in land-related issues (although functions and prerogatives are often
poorly defined and thus generate unnecessary confusion). As a result, local authorities have been implementing ad
hoc solutions to land disputes, especially where refugees are concerned, without prior consultation or coordination
with governmental entities.
Uncertain land tenure rights hamper productivity at the farm level through two channels. First, it reduces incentives
to invest in the property. The economic literature stresses the importance of tenure security and land transferability
rights for long-term investments. Second, land titling is generally thought to foster access to financial services,
although the extent to which and the mechanisms through which it does so are not always well known.
Landownership security thus should be given special attention today as the country considers upgrading its
agricultural production through specialty coffee and tea and horticulture. Moreover, land tenure insecurity reduces
the attractiveness of the economy for both domestic and foreign investors because it involves higher risks and
undermines the reliability and credibility of Burundian economic actors. Finally, the development of sustainable
land management policies to fight the dramatic land depletion in Burundi will include reinforcing landownership
rights. Sustainable land management is indeed a time-consuming, costly, and long-term process that only farmers
who are secure in their property rights will be willing to undertake.
A global discussion of land issues is essential to determine how land tenure could better contribute to shared growth.
Particularly, it will require thinking about ways to enhance wage labor to reduce tensions over access to land
property, specifically by diversifying sources of growth outside the agriculture sector (which could be linked to the
development of urban centers).
Source: Baghdadli, Harborne, and Rajadel 2008.
35
2.87 In addition, uncertain land tenure rights hamper productivity and delay the resolution of land-
related conflicts. A global reflection on land issues is essential to determine how land tenure could better
contribute to shared growth. In particular, it will require thinking about ways to enhance wage labor to
reduce tensions over access to land property, specifically by diversifying sources of growth outside the
agriculture sector that could be linked to the development of urban centers. (See box 2.3 for a discussion
of land tenure and economic growth in Burundi.)
B2. Improving the Performance of the Coffee Sector

2.88 Coffee has been the main cash crop and source of export revenue in Burundi for a long
time. Coffee is produced by 600,000800,000 households, and it usually constitutes their main source of
cash income. However, the coffee sector is suffering from weak performance, with declining production
and quality. The involvement of the government has not always been beneficial to the performance of the
sector. Coffee has the potential to contribute significantly to the countrys economy and its populations
livelihood. Burundi has the agro-ecological conditions to produce high-quality coffee. The current reform
of the sector, including the ongoing privatization process, is a step in the right direction; but more remains
to be done, as outlined in this section.
Past Performance of the Coffee Sector in Burundi

2.89 Despite the sectors potential and the governments strong involvement, performance
remained weak (particularly in the 1980s).
Coffee was first introduced by the Belgians in Figure2.9:FarmGatePricesforEACMembers,19772008
the 1930s. Its production was then mostly
coercive. After the countrys independence in 200
Price(UScentsperpound)
1962, the sector was operated entirely by

private stakeholders until its nationalization in 150
1976. In an effort to boost the industrys
revenues, the government deployed substantial 100
efforts in the early 1980s to increase
50
production; but those efforts were only partially
successful. Whereas the share of fully washed
0
coffee in total output increased steadily, the
overall results were disappointing. Annual
production fell well below the expected output. Burundi Kenya Rwanda
Despite an increase in the production of fully Tanzania Uganda
washed coffee, which should have led to higher Source:InternationalCoffeeOrganization.
premiums, prices paid to growers stagnated Note: Pricesfor2008areforthefirstthreequartersonly.
throughout the 1980s (figure 2.9).
2.90 Prompted by the sectors weak performances and the steep fall in international coffee prices
in the early 1990s, the government launched a reform in 1992 that has shaped the organization of
the coffee sector to this day. The reform strived to
improve incentives throughout the value chain to promote the production of high-quality coffee;
increase efficiency at the washing station and dry mill levels by calling in private actors;
enhance sale revenues by fostering competition among exporters; and
rationalize government intervention by converting the public entity Office du Caf du Burundi
(OCIBU) into a coffee marketing board.
2.91 The 1992 reform entailed a downsizing of OCIBU, as well as an attempt to improve
efficiency by promoting private sector participation in the industry. Ownership of the 133 washing
36
stations and two dry mills was transferred to a public entity, which leased them on 30-year contracts that
remain in effect to this day. All washing stations have been managed since then by five public-private
companies, the Socit de Gestion des Stations de Lavage (SOGESTALs, or Washing Stations
Management Companies); and the dry mills have been managed by the Socit de Dparchage et de
Conditionnement (SODECO, or Hulling and Processing Company). The SOGESTALs became
responsible for providing inputs to growers and financing the purchase of coffee cherries. To improve
sales efficiency and transparency, an auction system was set up to replace the state monopoly on exports
of Burundis green coffee.
2.92 This partial privatization did not provide proper incentives along the value chain. The
remuneration system, controlled by OCIBU, initially was aimed at stabilizing growers revenues by
guaranteeing a floor price to stimulate production. Farmers were paid on delivery at the washing station
and could receive a complementary payment when prices were high enough. But the floor rate usually
acted as a de facto ceiling price for producers because washing stations were not in competition (each of
them operated in an area designated by OCIBU). This had a negative impact on the farmers incentives to
produce higher-quality coffee. Washing stations and dry mills also had little incentive to increase quality
because their percentage fee was set in advance, without regard to actual performance. In addition, the
newly introduced auctions failed to improve the marketing process. The licensing system was set up so
that only exporters with adequate counterpart risk would be authorized to operate. But the granting of
licenses lacked transparency, and the number of bidders was so drastically limited that collusion was
almost inevitablethus capping earnings of upstream actors.
2.93 The outbreak of war in 1993 put an end to the reform process, and the political and
economic collapse that ensued further worsened the coffee sectors performance. The protracted
conflict led to the deterioration of existing trees, which were no longer properly cared for, and to a lower
replacement rate for aging trees. Washing stations and dry mills also were neglected during that period,
leading to the depletion of part of their assets.
2.94 As a result, the reforms, flawed at
Figure2.10:BurundiCoffeeProduction,19772008
conception and interrupted by the conflict,
45 failed to improve the sectors performance.
Production(thousandsoftons)
40 Production levels in the 1990s fell below those

35 displayed during the previous decade (figure
30 2.10). Cyclical variations partly explain these
25 low outputs. But smuggling of coffee cherries
20 to neighboring countries (another testimony to
15 the sectors failure) is also believed to have
10 contributed to artificially decreasing official
5 production figures. Prices paid to growers
0 remained low. The erratic evolution of
international prices for coffee further
contributed to decreased earnings in the 1990s
and early 2000s. However, growers in other
Source:InternationalCoffeeOrganization. East African countries managed to fare better
than those in Burundi. As shown in figure 2.9,
farm-gate prices in Kenya, Ethiopia, Tanzania, and Uganda stood well above those in Burundi. At a time
when other countries were adapting to changes in the global coffee market (notably, modification of
consumer patterns and the rise of specialty coffee), Burundi trailed behind and failed to seize
opportunities arising from these new developments.
2.95 Reforms resumed in 2005 in an attempt to complete the process initiated in 1992; but they
failed because of the lack of a comprehensive approach. In the first half of 2005, the transitional
government issued a series of decrees to liberalize the coffee sector. Prices paid for cherries and washed
37
parchment were deregulated, and private operators were allowed to invest in the industry. In addition, the
government no longer guaranteed loans made by banks to OCIBU to finance the coffee campaigns. That
measure, however, was not accompanied by a transfer of property rights over the washing stations to the
SOGESTALs, putting them in a difficult situation: they were expected to finance the purchase of cherries
and other expenses, but had no access to working capital because they had no collateral to offer banks.
OCIBUs role was also reduced by ministerial decree to that of a regulatory agency. It still remained
responsible for holding the auctions through which green coffee was sold. Other issues (such as the
suspected collusion among exporters, in particular) were not properly addressed. The series of ad hoc
measures signed into law in 2005 thus failed to reform the coffee sector because they did not provide a
comprehensive deregulation framework.
2.96 The lack of a comprehensive and meaningful reform has impeded the coffee sectors revival.
Although production soared in 2008, it stood at only 35,000 tons, the average level displayed by the
sector in the 1980s (figure 2.10). Moreover, prices paid to growers remain among the lowest of the region
(figure 2.9). Aging coffee trees are becoming a major concern, further increasing cyclical variations.
Incentives offered to growers remain insufficient to induce an increase in output and quality. The
SOGESTALs and SODECO are in a dire financial situation. Low production volumes in the past years
have undermined their capacity to repay loans incurred to finance the campaigns. The banking sector is
now reluctant to extend new loans to the industry, and recent campaigns were marked by tough
negotiations in which the government intervened despite the theoretical deregulation initiated in 2005. In
addition, the ill-adapted remuneration grid does not always enable the washing stations and dry mills to
cover their operating costs.
2.97 In December 2008, the government adopted a privatization strategy for the coffee sector,
following lengthy consultations with stakeholders. According to the strategy, washing stations will be
put up for sale in groups of three to six. These relatively small entities are expected to raise
competitiveness through closer collaboration with growers. They are believed to be better positioned than
the SOGESTALs (each of which manage 2030 washing stations) to enhance quality and support the
introduction of differentiation and traceability systems that are indispensable in accessing specialty
markets. By putting washing stations in competition, this new organization is intended to benefit growers.
To some extent, growers should be able to choose the stations to which they will sell their cherries, thanks
to increased price transparency. The sale strategy also would include two dry mills in good condition,
despite their oversized processing capacity and high operating costs. Their production systems are flexible
enough to adapt to the handling of smaller volumes of parched coffee. The privatization strategy includes
some innovative features that promote more inclusiveness in the Burundi coffee sector. For example, it
has opened up 75 percent of the washing stations equity to private sector participation, while allotting the
remaining 25 percent to producer associations.
2.98 The strategy is currently being implemented, envisaging complete privatization. In June
2009, the government of Burundi decided to launch the tender for the sale of 117 washing stations,
without first amending the convention between OCIBU and the SOGESTALs (as had been requested).
The government made this exceptional move to avoid derailing the reform process, which had already
experienced a long delay. At the closing of the bidding process in August 2009, four potential contenders
responded to the tender for the sale; and three of them presented offers that met the technical and financial
requirements to acquire 6 of the 26 lots of washing stations. Following lengthy discussions between the
government and SOGESTALs, the OCIBU-SOGESTALs convention was amended on August 20, 2009.
Based on this amendment, SOGESTALs would continue to manage unsold washing stations. This
arrangement laid the legal ground for bid winners to access the washing stations without contest by
farmers and their organizations. In November, 13 washing stations (out of 117) were transferred to the
bidding winner (leaving 104 unsold stations under SOGESTALs management, based on the terms of the
recently amended convention).
38
2.99 The first phase of the privatization process has ended with mixed results. It could be
expected that an internationally recognized company effectively managing 13 washing stations would
offer an excellent alternative to SOGESTALs, and would force them to improve the remaining poorly
managed washing stations. The new private investor could help address the credit and payment issues and
deliver additional services in the area of intervention to influence the overall way of doing business in the
coffee sector. The new context (with few privatized washing stations) can be viewed as a pilot phase from
which lessons may be drawn to improve future transactions. This first phase also can lay the ground for
an environment conducive to a second bidding round, particularly after the elections when political
pressure on the government is reduced. Even a newly elected government could feel more comfortable to
accelerate the reform if the experience is deemed successful. Conversely, poor first-phase privatization
outcomes may put an end to the overall process and possibly lead to the return of the government to
manage the sector. 30
2.100 The success of privatization will also depend on the governments ability to set up an
efficient regulatory framework. The coffee sector will be able to improve its performances only if rules
are adequately designed, clearly communicated, and fully implemented. Investors will be willing to
participate in the privatization process only if a proper regulatory authority has been established and
market rules have been set. Until today, the industry has been organized primarily by OCIBU. Despite
recent reforms, OCIBU fixed floor prices for coffee and was unable to prevent collusion among exporters
of green coffee. The dissolution of OCIBU and the creation of a regulatory agency within the Ministry of
Agriculture were accomplished in early June 2009. The creation of the Coffee Sector Regulatory Agency
(ARFIC) has strengthened the role of the private sector. The first priority for this agency is to determine
competition rules to govern the industry. Its responsibilities include establishing quality norms and
classification systems, controlling traceability procedures, and delivering certifications.
2.101 Delays in implementing necessary reforms jeopardize the coffee sectors chances of
recovering from decades of weak performance. It is imperative to move forward decisively. Although
there appears to be a consensus on the privatization of the sector, disagreements pertaining to the
modalities of this privatization have slowed the process. Meanwhile, the industry is suffering from the
depletion of its assets (for example, aging trees, outdated processing equipments, and neglected feeder
roads connecting farmers to washing stations). But the biggest threat to the sectors recovery is the risk of
a general fatigue among stakeholders. Privatizing and liberalizing are only the first steps in the direction
of the sectors rebound. Much remains to be done for example, organizing growers into competent
producer associations, promoting market intelligence, branding Burundian coffee on international
markets, and modernizing production and processing procedures to enhance quality. It is more urgent than
ever before to lay the foundations for a thriving coffee sector.
New Opportunities from Recent Global Market Trends

2.102 The world coffee market has changed considerably since the early 1990s. Global
consumption increased steadily over the past 15 years, reaching a total of 128 million bags in 2008. In the
past eight years alone, it grew by 2.5 percent a year, on average. Consumption in emerging economies
and producing countries accounts for part of this upward trend. Demand in major consuming countries
has increasedsometimes substantiallyin recent years. Nevertheless, Prices have fluctuated widely
over the past 15 years because supply did not always meet demand.
2.103 The emergence of specialty markets profoundly modified the global coffee industry. These
niche markets include gourmet, fair-trade, and organic coffee for either in-home or out-of-the-home
consumption. Targeted customers value high-quality coffee and flavors linked to specific origins. They
are usually willing to pay a premium that guarantees that coffee was produced under certain conditions
30
The EAC is the regional intergovernmental organization of the Republic of Kenya, Uganda, the United Republic of Tanzania,
and (since 2007) the republics of Burundi and Rwanda.
39
and that growers were offered a fair priceor, more generally, that labor and environmental norms were
respected. Specialty coffee markets have developed at a fast pace. In the United States, the industry
weighs more than US$20 billion a year and is growing at a 9 percent annual rate. It also has benefited
from the expansion of emerging economies, such as China where specialty coffee sales increased by
90 percent between 1998 and 2003 (Clay, DeLucco, and Ottaway 2007). As an example, Starbucks, the
worlds leading specialty roaster and retailer, has seen its volumes of roasted coffee rise steadily from less
than 100,000 bags in 1995 to 3.5 million bags in 2008. This shift in consumption patterns altered supply
and demand dynamics in the coffee sector. Marketing of high-quality coffee is now more and more
frequently channeled through direct sales from producers to roasters who develop long-term business
partnerships and agree on higher premiums.31
2.104 Burundi has the potential to become a noteworthy player in specialty markets. Commodity
coffee, however, should not be neglected because it will continue to account for a significant share of the
countrys output for some time. Burundi is endowed with an ideal climate and terrain to grow high-
quality Arabica coffee beans. Recent visits from international roasters and retailers, who expressed a
strong interest in Burundis coffee, demonstrate the countrys potential to access specialty markets.
Starbucks offered coffee from Kayanza in very limited release in 2008, after collaborating with the
regions SOGESTAL. In principal, Burundis coffee industry has the necessary experience and assets as a
basis to shift toward specialty markets. To capitalize on the potential, however, the following measures
must be taken:
Quality standards must be raised and stabilized to access specialty markets and increase
earnings from the sale of commodity coffee. First, the pricing system should be revised to
improve incentives so that growers actually gain by an increase in quality and, importantly, so
they fully understand how added value is shared along the value chain. As a second priority,
access to finance for farmers needs to be facilitated. This would promote the use of inputs and
facilitate the replacement of aging trees, which would positively affect both yields and quality.
New market demands will require processing facilities to pay greater attention to tasting
qualities. Burundis coffee sector is currently structured around the selection of coffee according
to grading characteristics, which depend mainly on bean size, color uniformity, and defect count.
Specialty markets assume these criteria are met, but go one step farther to differentiate coffee
according to flavor profiles to match consumer demands. Hence, although selecting high-grade
beans remains fundamental, greater emphasis needs be placed on high-quality flavors.
Burundi will have to adopt traceability and differentiation systems if it is to access specialty
markets. In these two areas, the coffee sector will be starting from scratch. Ideally, output from
small groups of growers will be treated separately from the time it is brought to the washing
station onward. Processing coffee lot by lot enables washing stations to distinguish excellent
batches from poor ones and thus channel the excellent batches to specialty markets and the other
batches to the commodity market. Being able to trace coffee to its origin is also important for
marketing purposes, notably to communicate the products story. Finally, quality control and
coffee differentiation will require an expansion of the countrys cupping capacity.32
To deal with the growing complexity of global coffee markets, Burundi must rethink current
marketing practices. Quality improvements will be fruitful only if they are complemented by
carefully designed marketing strategies aimed at (1) improving revenues from the sale of
commodity coffee, and (2) penetrating specialty markets. Proper gathering of market information
and, more important, its dissemination along the value chain are indispensable to understanding
31
Other elements, such as availability and reliability, affect price differentials applied to green coffee according to its country of
origin. For instance, Burundis coffee is currently offered at a discount of 100 points (that is, minus 1 cent per pound) on the New
York Coffee Exchange.
32
To this day, there is only one cupping laboratory in Burundi, at OCIBU.
40
price trends and supply and demand dynamics. Suboptimal decisions could thus be prevented.33
As for specialty markets, they rely increasingly on the development of long-term partnerships
between producers and roasters or retailers. Countries that have successfully operated a transition
toward specialty markets, such as Rwanda, usually have managed to build a strong image of their
produce and industry to attract such partners. This is all the more important for Burundi because
its competitors enjoy a 15-year head start in producing specialty coffee.
B3. Broadening the Base for Economic Growth

2.105 Recent research confirms that there is a strong correlation between diversification and
long-run economic growth.34 Developing countries that have successfully achieved and sustained high
growth rates have typically been successful in diversifying the structure of their economy. Imbs and
Wacziarg (2003) established a U-shaped relationship between diversification and GDP per capita, finding
that countries first diversify until they reach a certain developmental level; then they start to specialize
again. Countries that have achieved a high level of product sophistication relative to income level have
been found to grow more than the average (for example, China and India) (UNIDO 2009).
2.106 The key findings regarding the diversification of the economy are the following:
It is not necessary to move directly from primary products to manufacturing of high-tech
products. On the contrary, it is advisable to diversify unsophisticated products first and increase
their market share before making a move to more sophisticated products. Hausmann, Hwang, and
Rodrik (2007) have developed several measures to evaluate the income level associated with a
certain product. Their findings show that high-value primary products, such as fresh milk or olive
oil, can be associated with a high income level similar to that for motor cars.
It is also neither necessary nor sensible to jump directly to new products. Instead, the volume of
products already produced and exported (although currently in small amounts) can be expanded.
When considering new products, picking a winner (a product that will survive on international
markets and contribute substantially to Burundis economy) has a chance of success only if there
is sufficient information to understand why that specific sector will be doing well (for example,
see Brenton and Newfarmer [2007]).
An increase and diversification of exports can lead to higher productivity and growth through
learning by exporting (Page 2009). Firms that produce goods for the export market have higher
levels of productivity than those producing goods only for the domestic market. One reason is the
knowledge and skills transfer from international partners and buyers. In addition, a self-selection
takes place, allowing only firms with higher productivity to successfully access international
markets. Although it is an important and valid option in the short run, a focus on the (small)
domestic market is not sufficient for broad-based development over the long run. A strategy that
first targets regional markets to acquire the necessary experience before entering the global
market seems appropriate.
Ultimately, a shift to manufacturing is recognized to have great potential. High GDP growth rates
in low-income countries typically have been supported by high growth rates of manufacturing
value added per worker. Although the direction of causality between high GDP and the growth of
manufacturing value added is not entirely clear, research suggests that manufacturing growth and
structural change generally lead economic growth. Shifting labor and capital from low- to high-
productivity sectors increases the overall productivity in the economy and thereby subsequently
33
One example of a suboptimal decision is the 2007 granting of a monopoly on the sale of Burundis green coffee to a single
New Yorkbased stockbroker.
34
For example, see UNIDO (2009), Hesse (2009), or Lederman and Maloney (2009). It should be noted that correlation does not
necessarily imply causality.
41
increases the income level (for examples, see UNIDO [2009] and Hausmann, Pritchett, and
Rodrik [2005]). Manufactured exports are often labor intensive, thereby creating off-farm
employment opportunities. The labor intensity can also be an equalizing factor, creating
employment for the rural population as well as for women.35
Resource-based products, however, should not be discounted as a potential source of growth and
diversification in the short and medium terms. According to Page (2008), resource-rich countries
have led Africas recent growth turnaround. Although resources can be depleted, they provide an
opportunity to diversify exports and generate government revenues (which, in turn, could be
invested to improve infrastructure in a way that benefits the whole economy). Natural resource
production and exports, however, typically offer less employment opportunities and added value.
2.107 Economic diversification is particularly important for Burundi. Given the currently projected
demographic trends and the scarcity of land, Burundi needs to gradually move away from traditional
agriculture because it adds little or no value as a main source of income in the long run. In the short to
medium term, traditional agriculture will be essential for the development of Burundi; but it will not be
able to support the growing population (see section A of this chapter) and contribute to sustained high
growth over the long term.
2.108 A diversified economy can create much-needed employment, especially off-farm
employment. With a more diversified product base, Burundi could achieve broader and more inclusive
growth and reduce its vulnerability to external shocks. Although the transformation of the economy will
not be achieved in the next few years, implementing policies that facilitate diversification should be
initiated now.
2.109 In this chapter, Burundis potential to diversify its export base is analyzed and recommendations
are given to ease the process of structural reform. Several policies are needed to support a successful
diversification strategy, including addressing weaknesses in infrastructure, business environment, skills,
and institutions; and benefiting from regional integration arrangements (discussed in chapter 3).
Which Products Are within Reach?

2.110 Given Burundis current export structure, what products within reach could significantly
contribute to higher export earnings, increased product diversity, and ultimate development of the
economy? To help answer this question, Hausmann and Klinger (2006) have developed the product space
methodology that maps the distance between products. Hidalgo et al. (2007) visualize this approach in a
network representation. The densest part of the product space is typically made up of manufactured
products, whereas unprocessed agricultural goods tend to be in the sparsely populated part. Being in the
dense part of the product spacethat is, close to a range of other productsmakes it easier for a country
to venture into new products. Hidalgo et al. find that, unlike other regions, Sub-Saharan Africa exports
few product types, all of which are on the periphery of the product space.
2.111 Burundis products are placed on the periphery of the product space, complicating the
move to new products. The product space map for Burundi (highlighting the position of the products
identified in the previous section as classics, disappearances, and emerging champions) is plotted in figure
A.1 of appendix 2. It becomes clear that all of these export products lie on the periphery, far from the
dense part of the product space. That peripheral location makes it difficult for Burundi to move into new
products and develop an RCA in them.
2.112 Several products have been identified as potential export products. Using different measures
developed by Hausmann and Klinger (2006), and the location of Burundis products in the product space,
a number of products can be identified as close to the current classic export products and therefore easier
to reach than products that are far from the current exports. These products can serve as starting points in
35
According to Ross (2008), labor-intensive manufacturing is an important source of wage employment for women.
42
addition to the already identified emerging champions. Furthermore, the existing classics, which represent
only a small share, can be expanded.36 Products that have been identified include
frozen fish (excluding fillets), crustaceans, and mollusks;
beans, peas, lentils, and other leguminous vegetables;
fruit (fresh or dried);
tobacco;
live animals, including zoo animals (mainly ornamental fish);
precious and semiprecious stones; and
sugars.
2.113 These finding are in line with the products that Burundi has identified as diversification
opportunities. In its 2006 PRSP, the following areas were listed as export enhancing: coffee; tea; cotton;
mining; fisheries; tourism; handicrafts; and nontraditional exports, such as fruits and vegetables, flowers,
ornamental plants, aromatic and medicinal plants, and palm oil. This is also in line with the
recommendations of the National Agricultural Strategy from 2008 (Republic of Burundi 2008b),
identifying coffee, tea, cotton, cinchona bark, and horticulture products as the main products to be
promoted as exports. A similar list (with the addition of sugar cane) can be found in a report evaluating
the requirements to meet the Millennium Development Goals (Republic of Burundi 2007b). A strategy for
Burundis commercial and industrial development, drafted by the OTF Group (2008) with the support of
PAGE, suggests that the country should focus in the short term on five main export subsectors: coffee,
tea, tourism, horticulture, and mining. In their study on sources of rural growth in Burundi, Baghdadli,
Harborne, and Rajadel (2008) identify three subsectors with high growth potential: coffee, tea, and
horticulture. Among horticulture options, promising products include specialty vegetables and fruits.
2.114 Detailed information about the prospective sector and product is essential. It is important to
note that government intervention to identify new export products is unlikely to work if it is not based on
good and detailed information about why the sector and the product is likely to be successful. Moreover,
the chance of success is higher if Burundi already has some experience in producing and exporting the
product (as is the case for marginals and emerging champions). A sequenced and gradual transition needs
to take place.
2.115 In summary, the following product groups emerge as the intersection of the product space
analysis with the strategies developed by the government of Burundi, if the main classical exports
(coffee, tea, gold, cotton) are excluded: fishing, horticulture, and mining.
Fishing
2.116 Burundi currently produces fish mainly for domestic consumption, and it exports only small
amounts. The country has favorable natural conditions, and fish can become an important contributor to
exports just as it is a source of nutrition for the domestic population.
2.117 Fishing supports a substantial number of people, but fish production is stagnating. The
fishing sector supports about 300,000 people who live primarily in the poorest communities (Republic of
Burundi 2008b). Fish consumption is currently below that of the sub-region (FAO 2003) and decreasing
as a result of stagnating fish production and population growth. The fishing sector is divided into semi-
36
Based on the concept of proximity, Hausmann and Klinger (2006) further construct a product-specific measure (called path)
that is the distance-weighted number of products around a certain product. In addition, they develop a country-specific measure
(called density) that measures how close a specific product is to the current exports of the country and how easy it is for a country
to develop and maintain an RCA in that product.
43
industrial, artisanal, and traditional fishing (with artisanal fishing accounting for more than 90 percent of
the fish caught).37 The amount of fish produced in fish farms is very small.
2.118 The fishing sector faces several constraints. The main constraints are the weak institutional
capacity of all actors in the sector; degradation of shipping equipment resulting from fishermens lack of
access to finance; noncompliance with sanitary standards by unloading, drying, and smoking facilities;
pollution; overfishing; and generally inadequate management of fishing resources.
Horticulture
2.119 The horticulture sector is an important source of diversification and of income and employment
creation. Products that have been identified include fresh and dried fruits and vegetables, cut flowers,
ornamental plants, and medicinal plants (mainly cinchona bark).
2.120 The horticulture sector has great potential for a gradual transformation from primarily
unprocessed products to agro-processing. During the transition, the necessary know-how, skills, and
equipment can be acquired. Agro-processing includes the transformation, preparation, and preservation of
agricultural products (following harvest). It is one of the main sources of employment and income
creation in the manufacturing sector, especially in developing countries that depend predominantly on the
agriculture sector (UNIDO 2009).
2.121 A move into agro-processing can be especially advantageous for a country like Burundi,
where the majority of poor people live in rural areas. Agro-industries can be decentralized and are
primarily based in rural areas. Therefore, they are an important tool to promote socially inclusive growth
by creating farm and nonfarm employment and income and by raising the living standard of the rural
poor. Creating nonfarm employment is essential for Burundi, given its high population growth rate.
2.122 The potential of the horticulture sector has been recognized. In all strategic documents
produced by the government and in World Bank analyses, horticulture has been uniformly identified as
having the potential to contribute significantly to the diversification of Burundis product base. The
products can be grouped into the two main categories: fruits and vegetables, and flowers and plants. The
following specialty fruits and vegetables have been identified: pineapple, avocado, passion fruit, small
bananas, papaya, mango, cherimoya, green podded peas, baby vegetables, leafy greens, selected roots and
tubers, and macadamia. Flowers and plants include roses, dracaena, and cinchona bark, among others.
2.123 The horticulture sector faces several constraints. The National Agricultural Strategy (Republic
of Burundi 2008b) lists a number of constraints to the development of the horticulture sector, including
the lack of market knowledge, leading to suboptimal strategic positioning of products on
international markets;
insufficient knowledge of international norms and standards, particularly phytosanitary standards;
the lack of storage facilities and availability of air transport;
producers who are not well organized; and
lack of extension services and training for producers.
Mining
2.124 Mining can be an important source of incomeespecially the mining of ores and concentrates of
nonferrous base metals, including nickel, coltan, and cobalt. The mining sector is discussed in more detail
below.
37
On average, only about 13,000 tons of fish have been caught annually between 2001 and 2005, a significant decline from about
21,000 tons in 19901995 (Republic of Burundi 2008b).
44
B4. The Potential of the Mining Sector to Contribute to Growth
2.125 Burundi is believed to have considerable mineral resources. The mining sector, however, is
dominated by some 20,000 artisanal producers, with no industrial-scale production. It performs below its
potential. The development of the sector has been held up for years by the civil war. The government
recently has undertaken to reform the legal framework to make it more conducive to the development of
the sector. A number of international companies have been interested in investing in commercial
exploration of the mineral resources. If these reform and exploration efforts prove successful, one or more
mineral deposits could be developed on an industrial scale. This would provide the economy with a
significant boost in terms of production value, tax receipts, employment, related infrastructure, and local
economic development. However, all of this depends on international market prices and requires that
Burundi overcome infrastructure obstacles. Even if major investments seem unlikely in the short term,
mining deserves special attention from the government because of the potential contribution of large-scale
mining to the countrys future growth, and because current artisanal and small-scale mining needs to be
improved.
Situation of the Mining Sector

2.126 Mineral production accounts for a small part of Burundis GDP.38 Production is limited to
gold, colombo-tantalite (coltan or tantalum ore), and relatively small quantities of cassiterite (tin ore),
kaolin (china clay), limestone for cement, construction stone, sand and gravel, and peat for fuel. The most
important mineral potential is in nickel, gold, and coltan; but the country also has resources of other
metals and industrial minerals like vanadium, phosphate, and rare earth metals.
2.127 The number of nickel deposits known in Burundi suggests good potential. Those deposits are
located all along a strip of land 2040 kilometers wide, stretching from Mugina in the extreme south to
Muremera to the northeast. Deposits have been found from the two main types of nickel deposits that can
be found globallynamely sulphide and laterite. The recent price collapse has had a major impact on
nickel supply, and many projects have been either put on hold or cancelled. However, the global demand
for nickel, closely correlated with global industrial production, is expected to resume and continue to
grow over the long term.
2.128 Musongati is the main laterite nickel deposit in the country. Located in central Burundi, some 85
kilometers east-southeast of Bujumbura, it is the most explored laterite nickel resource. Musongati is
among the 10 largest undeveloped laterite deposits known worldwide. It has an inferred resource of some
180 metric tons. Musongati can be compared in size with other laterite projects recently put forward:
Ravensthrope (Australia), Ambatovy (Madagascar), or Goro (New Caledonia).39 A comprehensive mining
convention, negotiated with an international company in 1999, was cancelled by the government after the
company declared force majeure for security reasons in 2000. The exploration license was granted to
another international mining company at the end of 2008, and several international companies are actively
involved in discussions with the government about the further development of the mine site.
2.129 Muremera is the most well-known nickel sulphide prospect in Burundi; and may be the same ore
body as the Kabanga deposit in Tanzania, a couple of kilometers across the border. A cross-border mining
operation might be possible, even though the feasibility of the Kabanga mine as a stand-alone project is
being evaluated and the ownership of both projects is separate. At the time of its discovery in the 1980s,
Muremera was said to have reserves of as much as 30 million tons of nickel sulphide ore. An international
company secured 100 percent ownership of the Muremera project in March 2009, and continues
exploration for nickel sulphide.
38
Mining and energy accounted for about 1 percent of Burundis GDP in 2008, according to the latest figures from the
International Monetary Fund.
39
This comparison is theoretical because resources/reserves of those projects do not present the same level of certainty. In the
case of Musongati, contrary to the three others, we believe resources are still classified as inferred.
45
2.130 Nickel exploration has been quite dynamic on other prospects. Important lateritic nickel resources
include Nyabikere and Waga, which have combined inferred resources of 81.4 metric tons. The holder of
the Musongati exploration permit also has the permit for those two potential deposits. One of the worlds
largest producers of primary nickel recently has acquired permits to explore the nickel sulphides of
Rutovu, Buhoro, and Bukirasazi, which could be seen as an encouraging sign in the current global
outlook.
2.131 Gold is only exploited on an artisanal and small scale. This has been true since colonial times
in many places across the country. A few small-scale companies, associations, or comptoirs exploit gold
legally. However, most of them work outside the legal framework set by Law 1/015 of 2000, the most
recent law dealing with artisanal mining, mineral trading, and exportation. Gold production represents a
significant but informal contribution to the economy. Most of this gold does not pass through the formal
customs system, and a large part of the exported gold comes from the Democratic Republic of Congo
(Global Witness 2009). If gold prices can be expected to remain at historically high levels, above US$750
an ounce on a nominal basis over the coming 510 years (CPM Group 2009), the contribution of gold to
Burundis economy can be significant.
2.132 No large-scale gold operation is currently being planned. So far, the only gold exploited in
Burundi derives from alluvial deposits found in the riverbeds. These deposits are said to be of relatively
low grade. However, the government has prospected attractive indexes in the northeast. The Burundi
Mining Corporation (BUMICO), a government-private venture, was exploring the possibility of
producing gold on a commercial basis at Muyinga. The governments efforts to promote the countrys
gold potential have attracted several exploration junior companies. One international company has an
exploration permit in the Mabayi-Butara region; but as part of its cost reduction measures, it has put its
Burundi operations on hold.
2.133 Coltan has been exploited by artisans or small companies in Burundi since the 1930s.
Deposits are known in the northeast and north of the country. Informal mining still occurs in rivers or
alluvial areas in several provinces. Comptoirs Miniers des Exploitations du Burundi (COMEBU), which
has a 25-year concession around the Kabarore mine, is the only legal operation. The coltan artisanal
production level is closely related to the tantalite price. Artisanal and small-scale coltan mining has
contributed modestly but constantly to Burundis economy. Development of small-scale mining will be
limited by the bad image of conflict coltan in the Great Lakes region. Indeed, major users of tantalum
ore have committed not to acquire any material containing tantalum from countries in the Great Lakes
region. Development of large-scale tantalum mining is unlikely because there is little room for
newcomers in the supply market.
2.134 Cassiterite (tin ore), and wolframite (tungsten ore) often accompany coltan. Unlike gold
miners, a single company is said to organize the production of artisanal tin miners. COMEBU is the only
company that has a mining concession, and it operates small mechanization. Some unorganized cassiterite
artisanal miners still exist. According to the Ministry of Mines, declared production has been very erratic,
from less than 10,000 tons in 2003 to more than 100,000 tons in 2008.
2.135 Other metals and industrial minerals include vanadium, rare earth metals, phosphate,
platinum-group metals, carbonate materials of the type needed to produce cement, kaolin and peat.
However, further exploration work must be carried out for most of those materials. Hydrocarbon
research operations were conducted in the 1970s and 1980s in Lake Tanganyika and in the Ruzizi plain.
Hydrocarbon indicators were detected in the depths of the basin. Promotion activities are ongoing to
attract companies to pursue exploration in Lake Tanganyika.
Potential Contribution of Mining to the Development of Burundi

2.136 The mining sector has been identified as a potential source of growth. The disparity between
the contribution of the mining industry to the national economy and its larger potential is partly caused by
46
Burundis landlocked location, the inadequate export infrastructure in surrounding countries, and the
protracted war. However, the government has identified the mining sector as a potential source of
diversified growth and income creation. The 2006 PRSP (Republic of Burundi 2006) envisages activities
to (1) reinvigorate research, (2) favor the development of artisanal and semi-industrial activities, (3)
develop a mining code attractive to investors, and (4) attract international investors to exploit some of the
mineral deposits. In 2008, the government launched a revision of the mining legislation and regulation.
This legislative reform should allow the submission of a new mining code with a new model convention
agreement in 2010.
2.137 Substantial benefits can accrue to Burundi from a properly structured and administered
mineral industry. In addition to providing foreign exchange earnings, mining activity will produce
additional revenue through taxes and royalties, improve the professional and technical skills of nationals,
and provide a nucleus for regional and local economic development. Successful large-scale development
of Burundis mineral potential would require investment in energy production and export roads, but
artisanal and small-scale mining can already be improved to provide more benefit to the economy.
Contribution of Industrial Mining

2.138 Mining could become the second most important sector after agriculture. Assuming that a
commercial deposit exists and that mining companies decide to go ahead, mining could have a major
impact on growth and would probably become the second most important sector after agriculture. A few
years after the investment decision, export revenues and government tax revenues would jump to new
levels. A more in-depth projection would have to assess the percentage of growth that this would imply: it
would have to evaluate the added value of each mining projectnot only the production value, but also
all possible economic, social, and environmental benefits and costs related to such an investment.
2.139 Two different scenarios have been considerednamely, development of a large-scale nickel
mine and of a large-scale gold mine. The assumptions and projections are summarized in table 2.4. The
macroeconomic contribution of a gold mine would be significant when operating at full capacity. It would
represent between US$39 million and US$48 million of production value, and between US$7 million and
US$12 million in tax revenues annually. The main part of production would likely be exported, so it
would have a direct effect on export earnings. In terms of direct employment, such a project could create
a limited number of jobs, possibly several hundred. Indirect employment could vary a lot and typically
can be increased by helping local entrepreneurs create businesses and supply the mine. The life span of
such a gold mine is also likely to be limited to 515 years, compared with the minimal 2025 years of a
nickel mine.
Table2.4:AnnualProjectionswithOneLargeScaleNickelorGoldMine
Nickelmine Goldmine
Assumptions
Capacity 50,000metrictons 2metrictons
Operatingcosts US$3.2perpound US$400perounce
Taxation royalty3%,incometax30%,participationofstate15%,dividendtax15%
Price US$4perpound US$7.3perpound US$600perounce US$750perounce
Projectedoutcome(annualcashflow,US$millions)
Sales 441 800 39 48
Operatingprofit 88 447 13 23
Netdistributiontoinvestor 38 214 6 11
Totalgovernmentcashflow 50 233 7 12
Effectivetaxrate(%) 57.0 52.1 54.0 52.7
Sources:AuthorscalculationsbasedonAfDB(2009)andonamodelcreatedbyRobertParsonsinworkpresentedtothegovernment
oftheLaoPeoplesDemocraticRepublicinSeptember2008.
Note:Theprojectionsofthenickelmineexcludethepotentialcontributionofcobaltasabyproduct.
47
2.140 The Musongati nickel deposit has the greatest potential to be exploited on an industrial
scale in the near future. Two possible options for the actual mining operation have been proposed at one
time or another: one is the export of nickel ore; the other is the export of metal from a refinery at the mine
site. The first option would involve transporting some 4 million tons of ore each year from the mine site
to the Port of Dar es Salaam for shipment overseas to a refinery. These large volumes would require
access to the rail transport network of Tanzania. This scenario therefore depends on completion of one of
the current options for the proposed public rail extension from Tanzania into Burundi, and construction of
rail spurs to the mine sites. The other option is to refine the ore at the mine site and transport refined metal
to the coast for shipment abroad. If the ore is refined at the mine site, the quantities of metal to be shipped
are estimated at about 50,000 tons a year. In this scenario, transporting the metal by road to the railhead at
Kigoma is seen by industry analysts as the preferred option. The working assumptions are that the
Musongati ore field will be developed and brought into production by 2017, nickel and cobalt will be
refined at the mine site, and the mining company will ship the refined metal by road to the railhead at
Kigoma.
2.141 The macroeconomic contribution of this nickel mine and processing plant would be huge. It
would represent between US$397 million and US$800 million of production value40; and between US$50
million and US$233 million of tax revenues annually when operating at full capacity, depending on the
nickel price assumption (table 2.4). It is assumed that production starts by 2017. The production would
likely be totally exported, so it would have a direct effect on export earnings. In terms of employment,
such a project could create 1,0002,000 direct jobs, and 2,0005,000 indirect jobs. To those permanent
jobs should be added the numbers of short-term jobs that would be required during the construction
phase. A portion of that workforce inevitably would come from abroad; but if training campaigns can be
organized in advance, the local population could benefit from a large number of those employment offers.
The effective tax rate illustrates the fairness of the deal between the investors and the government. It is
given here only as an indication because the assumptions are too broad to provide a realistic rate in
absolute terms. It is still possible to interpret the decrease of the tax rate when prices increase as a reason
to introduce a possible windfall tax41 to better share exceptional benefits in periods of high commodity
prices.
2.142 In addition to the macroeconomic contribution, mining development can stimulate local
economic development. The construction of one or more large mines and the infrastructure to support
them would produce many positive local and regional effects on employment, the development of supply
industries, and community development around mining. The government must take measures to actually
optimize this contribution to the local economy by encouraging corporate social responsibility or training
and capacity building of the workforce. Without such measures, mining development may have several
negative effects on the local population, including disruption of the indigenous land usage, inflation,
pollution, prevalence of certain diseases (HIV/AIDS), political friction, and the like. Moreover, measures
must be taken to ensure that local communities receive part of the benefitsfor example, channeling part
of the mining taxation revenues to those communities, either through decentralized authorities or through
dedicated instruments like mining foundations, trusts, or development funds.
Contribution of Artisanal and Small-Scale Mining

2.143 Mining employs a substantial part of the population. A recently conducted survey of the entire
artisanal and small-scale mining sector finds that the sector currently employs about 20,000 people;
approximately 10,000 of those workers are mining metals and the other 10,000 are mining construction
material (Midende 2009). Assuming an average family size of 10, this implies that roughly 200,000
40
The value added is not estimated here; therefore, we cannot estimate the percentage of the GDP that such a production would
represent.
41
A windfall tax is a higher tax rate on profits that ensue from a sudden windfall gain to a particular company or industry. Many
governments explored the possibility of including such a tax for the mining sector following the 200708 commodity boom.
48
Burundians are affected by mining. Income from mining usually complements income from agriculture;
and mining is typically carried out in the dry season, when the intensity of activities in the agriculture
sector decreases. Mining also offers possible employment for numerous demobilized soldiers.
2.144 Official and survey production statistics differ widely, making it difficult to assess the
impact of artisanal and small-scale mining on the economy. Table 2.5 contrasts official production
statistics with those determined by the survey of the
industry. It becomes clear that most of the official gold Table2.5:ComparisonofAnnualOfficialExports
exports are not being produced in Burundi, but rather are in andSurveyProduction,2008
transit mainly from the Democratic Republic of Congo. Material Officialexports Surveyproduction
Large amounts of coltan and wolfram, on the other hand, (approximate)
leave the country without being officially recorded as Gold 2,170kilograms 500kilograms
Coltan 84tons 150tons
exports. Prices offered directly at the mining site are also Wolfram 608tons 1,000tons
well below those offered in the parallel market or by small Source:Midende2009.
exporting companies.
2.145 The contribution of mining to the government budget is negligible, and its contribution to
the balance of payments is limited. Most minerals are channeled through comptoirs en transit, foreign
companies that are exempt from paying both the VAT and the mining tax and do not have to repatriate
foreign exchange earnings. Formalization and supervision of the sector as well as revision of the tax
exemption could improve the contribution of mining to the budget and increase the income of miners and
intermediaries.
2.146 If not addressed properly, artisanal and small-scale mining in Burundi can have numerous
negative implications. Those implications include the following:
Use of basic techniques: Extraction techniques are basic, using primarily primitive tools.
Artisanal exploitation of a mining site usually reaches only half of the resources and leaves the
rest behind, leading to quickly accumulating losses.
Health, hygiene, and security problems: Mining sites are not secured, and accidents occur
frequently. The sites are typically far from any infrastructure, such as health facilities and
schools. Temporary housing set up near the site does not provide proper protection, sanitary
conditions, and clean drinking water. In addition, the population mix of the makeshift
communities and the presence of many demobilized soldiers lead to frequent conflicts.
Child labor: Children who work illegally at the mining sites are prone to be injured and to
develop pulmonary infections and eye and skin diseases. Many are also malnourished and not
able to attend school regularly.
Environmental impact: Mining that is not well controlled and organized can lead to deforestation,
destruction of vegetation and soil, erosion, and water pollution resulting from the treatment of
minerals. Abandoned sites are a real danger to people and animals.
Scenarios for Low, Medium, and High Contributions of Mining to the Economy of Burundi
2.147 What follows are three scenarios for the future of mining in Burundi. They are not intended
to be accurate calculations, but rather to illustrate the wide spectrum of impacts in terms of government
revenues. To maximize the chances that Burundi benefits from its mining sector, as illustrated by the
medium- or high-contribution scenario, the country will have to implement significant policy reforms.
2.148 The low-contribution scenario assumes that no policy reforms would be undertaken other than
the ongoing mining legislation and regulation update; and the market outlook does not permit Burundis
potential to be fully realized. The following outcomes would result:
49
Artisanal and small-scale mining would remain primarily informal, providing the government
with erratic revenues of US$0.10.5 million, based on the royalty revenues coming from artisanal
and small-scale mining in recent history.
The contribution of formal mining to GDP would probably remain around 1 percent.
The share of mineral ores in official exports would vary between 1 percent (2001) and 5 percent
(2006).
However, the contribution of informal mining would be a lot more significant (with a stable
number of 70,000 artisanal miners exploiting gold, coltan, and other minerals, including
construction material); but it would offer very little benefit in terms of government revenues.
2.149 The medium-contribution scenario assumes that enough policy reforms are undertaken to
formalize artisanal and small-scale mining, to attract investments, and to prepare the country with
strengthened institutions for industrial mining to make a successful contribution to the economy. It also
assumes that the market outlook is good enough to favor the launch of a medium-scale gold mine with
circumstances corresponding to our low gold price projection (table 2.4). It would yield the following:
Artisanal and small-scale mining would progressively formalize and contribute up to a stabilized
US$2 million per year to government revenues.
The gold mine could be committed around 2013, arrive at full production around 2015, and
operate for five to seven years.
Government revenues would reach US$9 million in 201718, but could quickly decrease
thereafter if no other mining operation is launched (figure 2.11).
2.150 Such a scenario would still need significant policy reforms to improve the geology data
infrastructure, the capacity of institutions to monitor the operations, and the overall governance of the
sector.
2.151 The high-contribution scenario assumes that major policy reforms are undertaken to formalize
artisanal and small-scale mining, attract investments, prepare the country with strengthened institutions,
and properly integrate good management of mining all along the value chain. It also assumes that the
market outlook is good enough to favor the launch not only of a medium-scale gold mine, but also of a
large-scale nickel mine operating for 25 years. Circumstances are supposed to correspond to the
conservative nickel and gold price projections (table 2.4). Results would be as follows:
In addition to the impacts of the medium-contribution scenario, the nickel mine would make a
major contribution to the economy and annual revenues could exceed US$50 million (figure
2.12).
Policy efforts to prepare for such an eventuality should be more consequent as they probably
would have to adopt an integrated approach like EITI++.
Experiences of World Bank assistance in other countries suggest that the related public
investment could range from US$8 million to US$10 million for a technical assistance project
lasting four to five years.
More costly public investment in infrastructure related to nickel mining will be necessary as well
(see section A in chapter 3).
50
Figure2.11:MediumContributionScenario Figure2.12:HighContributionScenario
GovernmentRevenue GovernmentRevenue
10 60
9
50
8
7 40
US$millions
US$millions
6
30
5
4 20
3
2 10
1 0
0
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2040
2041
2042
2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020
ASM Goldmine ASM Goldmine Nickelmine
Source:Authorscalculations. Source:Authorscalculations.
Note:ASM=ArtisanalandSmallScaleMining Note:ASM=ArtisanalandSmallScaleMining
B5. Conclusion
2.152 Burundi has the potential to improve its growth performance considerably. Given the
structure of the economy and the constraints it is facing, the low performance of the agriculture sector
requires immediate attention. In the medium term, more broad-based growth and increased off-farm
employment are necessary to sustainably reduce poverty and improve the populations living conditions.
Sectors that have been identified in this chapter as encouraging new sectors are primarily horticulture,
mining, and fishing.
2.153 To support growth, Burundis policy reform and implementation efforts should focus on
four main areas: (1) closing the infrastructure gap, (2) maximizing the benefits from the ongoing
regional integration process, (3) improving the business environment in which the private sector operates,
and (4) promoting encouraging new growth sectors. The government can further contribute to promoting
growth by strengthening the countrys fiscal position through improved revenue mobilization, prudent
spending policies, and reformed public financial management. These issues will be addressed in detail in
the following chapters.
51
C HAPTER 3 THE G OVERNMENT S O PTION TO TACKLE
O BSTACLES AND R EAP O PPORTUNITIES FOR
ECONOMIC G ROWTH
3.1 As described in chapter 1, Burundis economic recovery has so far been sluggish and below
what can be expected for a post-conflict country. Only in recent years has economic growth improved
slightly, but it has remained volatile. A number of reasons for the current situation and its origin were
outlined in the chapter 2. The challenge now is to turn the economic performance around, consolidate the
recent upward trend, and lay the foundation for high and sustainable growth in the future to significantly
reduce poverty and improve the living conditions of the population.
3.2 How can the government help create an environment that is conducive to accelerated
growth? Although there are many areas where improvements are needed, Burundi should focus on the
following main objectives: (1) closing the infrastructure gap, (2) maximizing the benefits from the
ongoing regional integration process, and (3) improving the environment in which the private sector
operates, and (4) promoting new growth sources. Furthermore, the government needs to strengthen its
fiscal position, which will be discussed in chapter 4.
A. AN INFRASTRUCTURE ACTION PLAN 42

3.3 There is a broad consensus that infrastructure plays an essential role in growth (for
example, see Ndulu et al. [2007], UNIDO [2009], or Foster and Briceo-Garmendia [2009]). Burundi has
a large infrastructure deficit in almost all areas, including transport, electricity, and communications. For
example, only 2 percent of the population has access to electricity, compared with 16 percent for Sub-
Saharan Africa and 41 percent for other low-income developing countries. To make matters worse, the
country is currently plagued by frequent power outages that hamper all areas of economic activity.
3.4 The poor state of the infrastructure also negatively affects Burundis trade. Domestic roads
as well as transport corridors to the main ports in Dar es Salaam and Mombasa, on which Burundi
depends, need to be upgraded and expanded to lower transport costs and improve access to regional and
international markets. According to Naud (2007), one of the main reasons for the poor performance of
many African countries is the proximity syndrome, which he defines as the cumulative result of long
distances to markets, landlocked locations, and suboptimal agglomeration patterns. Because of the cross-
border nature of the required infrastructure projects, improved regional coordination and cooperation are
essential to overcome the proximity gap.
3.5 The infrastructure action plan, developed by the African Development Bank and presented in the
following section, lays out a detailed strategy for improving the infrastructure in Burundi, taking into
account the importance of the regional context. It focuses on the power, transport, and communications
sectors; and proposes actions to be taken, their timing, and a possible funding structure. If realized, the
action plan will have a significant positive impact on the whole economy.
A1. Main Features of the Action Plan

3.6 The focus of the proposed infrastructure action plan is on power, transport, and communications.
Discussions with the government of Burundi have confirmed the key objectives for the program.43
3.7 One of the objectives of the program for the power sector is to ensure, through increased
investments in domestic and regional generation capacity, that the business community and households
42
This action plan has been developed by the African Development Bank; see AfDB (2009) for details of the proposed program.
43
Details of the proposed plan are given in appendix 5.
52
have access to reliable power supply 24 hours a day. Furthermore, the action plan proposes to establish a
national transmission and distribution grid by 2015, with all 15 of the provincial capitals linked to this
grid and supplied with electricity 24 hours a day at reasonable cost. With the grid in place, the
electrification rate is supposed to increase from the current 2 percent of households to 25 percent by 2020
and to at least 40 percent by 2030. By 2020, 85 percent of urban households would have continuous
access to the national distribution network; and by 2030, one third of all rural households would be linked
to the grid.
3.8 The key objectives of the proposed action plan for the transport sector are to lower the costs of
transport for the entire economy and to improve access to local and international markets. The proposed
program focuses on road transport and civil aviation, and provides for further investigation of the
possibility of a rail extension from Tanzania into Burundi. The main component of the transport program
is the upgrading and expansion of the countrys road network over the next decade. The international
airport in Bujumbura would be expanded and modernized. The options for extending the Tanzanian rail
network into Burundi have also been reviewed as part of this study.
3.9 The proposed communications program aims to improve access to the international
communications network substantially and to lay the foundations for a national communications grid that
will provide communities and business through Burundi with low-cost voice and data communications. A
high priority is attached to the immediate development of a national communications grid of fiber-optic
cable and digital microwave that would be linked to the regional network.
3.10 The three main sources of funding for the program are the government of Burundi, including the
electricity utility and the airport authority; the donor community; and the private sector. The government
and donor community each fund major portions of the power program, but the strategy is to look to the
private sector to own and operate the
Figure3.1:FundingArrangementsfortheCoreInfrastructureProgram
new domestic hydroelectric stations that
are proposed. In the case of the roads 3,000
US$millions(2007constantprices)
Government
program, the donor community would
2,500 Donors
fund about 80 percent of the program. 10% Private
The civil aviation program would be 2,000
financed by the private sector, except for 23%
a small amount of donor and 1,500
government funding for the human and 37% 74%
institutional capacity building that will 1,000
be required. For the program as a whole,
500
the government would fund about 27 40%
16%
percent, the donor community about 56 0
percent, and the private sector about 17 Power Transport Communications
percent (figure 3.1). Funding for
investments for the nickel mine project Source: AfDB2009.
would be provided, to a large extent, by
the mining company involved in the project.
A2. Key Policy Objectives

3.11 A concerted program of action will be needed over the next two decades to overcome the
countrys current serious infrastructure deficiencies in power, transport, and communications. The
proposed infrastructure action plan is built around three key policy objectives:
1. Stepping up spending on infrastructure will ensure improved access to services while reducing
the cost and improving the reliability of these services, thereby removing some of the main
obstacles to sustained strong economic growth. The core infrastructure action program to be
53
implemented over the next two decades would cost US$4.6 billion (at 2007 constant prices). In
addition, US$1.2 billion would be spent on maintenance of assets in these three sectors.
2. The US$5.8 billion program will create substantial opportunities for the development of domestic
business activities. To generate a strong domestic supply response to this ambitious program and
ensure sustained economic growth of 67 percent a year, the government will need to design and
implement comprehensive programs for development of small and medium-size businesses and
for skills development in the labor market. Both of these programs are essential complements to
the infrastructure action program.
3. Implementation of the proposed action plan will require increased attention to coordination of the
power, transport, and communications programs within the government. It will also require close
cooperation between the government and the donor community in designing and implementing
these programs. Emphasis on coordination on these two fronts can improve the alignment of
investments with national and regional priorities and the overall efficiency of the development
process in Burundi. It is important to emphasize that, in fact, an expenditure of some US$5.8
billion over the next two decades is required to meet the current national objectives for
infrastructure development. These programs will also lay the foundation for a major investment in
nickel mining in the decade aheadan endeavor that can bring substantial additional benefits to
the country.
A3. Economic Impact of the Action Plan

3.12 A number of scenarios are considered to assess the economic impact of implementing the
action plan.44 In addition to the base case scenario, five alternative scenarios are considered (table A.10
in appendix 4). In all of these scenarios, it is assumed that internal security in Burundi continues to
improve and that there is social and political stability; that the government continues to adhere to sound
macroeconomic policies; and that Burundi, with the help of the international donor community, continues
to make steady progress in reducing its vulnerability to debt distress. It is also assumed that action is
taken to improve the business and investment climate and that there is a strong domestic supply response
to the proposed infrastructure action plan.
Economic Impact in the Near Term

3.13 For the short and medium terms, the recent encouraging recovery in economic growth can
be maintained. Even if no further domestic policy initiatives are taken, there will be further sharp
increases in public investment expenditures until about 2011, because of the ongoing major donor-funded
program of infrastructure rehabilitation. However, these projects come to a close by about 2014; and as a
result, spending under these programs declines sharply from 2012 onward. Analysis of the impact of the
ongoing infrastructure rehabilitation program suggests that GDP growth may be in the range of 4 percent
in real terms in 2009, and that it would increase to about 5 percent a year in 2010 and 2011. As the
ongoing infrastructure program phases down after 2011, economic growth could decline to about 4
percent a year in 201314. Serious delays in implementing the ongoing donor-funded program would
result in slower economic growth in the immediate future.
3.14 The key point that emerges from this assessment of the economic outlook for the near term is that
a strong push to improve the basic infrastructure services in Burundi can sustain economic growth in the
near term and lay the foundations for a period of strong economic growth over the next two decades.
44
These scenarios have been developed by the African Development Bank, as document in AfDB (2009).
54
Impact of the Base Case in the Longer Term
3.15 The long-term economic benefits that accrue to Burundi under the base case scenario are
substantial. The base case scenario includes the core infrastructure program and the nickel mining
project. Benefits include the following:
Sustained growth of the domestic economy that creates new business opportunities and increases
incomesGDP grows in real terms by about 7.2 percent a year over the next two decades in this
scenario. GDP per capita increases by 4.5 percent a year, reaching roughly US$324 by 2030 (at
2007 constant prices). Sustained growth of incomes in this range begins to have a significant
impact on the incidence of poverty in the country, with a substantial number of people at or just
below the poverty line moving out of official poverty (although many of these people would
still be vulnerable to downturns in the economy resulting from droughts or other disruptions).
Increased opportunities for productive employmentEmployment in the nonfarm sector grows at
6 percent a year. Over the next two decades, about 1.3 million jobs are created in the nonfarm
sector, mainly in urban areasequivalent to almost half of the 2.7 million new entrants into the
labor force in that period. The share of employment in agriculture declines steadily to about 70
percent of the workforce by 2030. By 2030, the industrial sector, including mining, is projected to
account for almost 10 percent of employment; and the services sector is projected to account for
about 20 percent.
Improved access to infrastructure services and lower costs for themThe sustained strong
growth in the economy and employment stems from the major investment in basic infrastructure
proposed under the action plan. More
reliable power supplies, improved Table3.1:SelectedIndicatorsfortheBaseCaseScenario
transport and communications Indicator 2010 2015 2020
services, and lower costs for these Totalelectricitysupply(gigawatthours) 196 471 1,981
services enhance the business Exportvolume(thousandsofmetrictons) 37 50 124
Investment(%ofGDP) 31.8 85.3 16.7
environment and investment Public 24.1 22.9 10.7
opportunities for the private sector and Private 7.7 62.4 6.0
raise Burundis international Source: AfDB2009.
competitiveness. (For selected
indicators, see table 3.1.)
Increased tax revenues and expanded public servicesThe combination of strong economic
growth and the start-up of the mining operation have significant implications for the revenue
position of the government.
Alternative Scenarios
3.16 Given that funding for all components is not guaranteed and that private participation in particular
cannot be enforced, a number of alternative scenarios have been developed. Table 3.2 summarizes the
outcomes for each of the five alternative scenarios considered and compares them with the base case.
Several key points emerge from this analysis:
In the event that the nickel mining project does not proceed, but the core infrastructure program is
implemented in full (scenario B), GDP grows at about 6.4 percent a year over the next two
decadessufficient to create a substantial amount of additional productive employment, improve
incomes and productivity in urban and rural areas, and contribute to a significant reduction in the
incidence of poverty in the country. The main economic impact is a substantial drop in export
income and government revenues.
If private investment is not available for the power and aviation sectors (scenario C), there is a
further modest decline in growth performance, with GDP increasing by an average of 6.2 percent
55
a year over the next two decades. The decline in the growth rate is limited because it is assumed
that all the additional power that is needed to meet demand is imported from the East Africa
Power Pool grid. The implication is that by 2030, imported power accounts for 90 percent of total
supply in Burundi.
In scenarios D and E, the level of public investment in the proposed infrastructure program
declines from US$3.8 billion over the next two decades to less than US$1 billion. The growth
performance of the economy declines sharply. In scenario E, the GDP growth rate is more than
two full percentage points below that of the base case. The economy has difficulty in absorbing
new labor force entrants into productive employment opportunities. In addition, much larger
numbers of people remain in low-productivity agricultural pursuits. Although there may be some
reduction in the incidence of poverty in the country in this scenario, the total number of people in
absolute poverty would increase substantially over the next two decades.
Scenario F assumes that one of the rail extensions into Burundi goes ahead, but that the nickel
mining operation is based on refining ore at the mine site. In the absence of the 4 million tons of
ore exports each year, the economic impact of the rail extension is modest and GDP growth is not
much higher than in the base case. Moreover, in this scenario there is a risk that large public
subsidies would be needed for the operation of the rail network in Burundi.
Table3.2:KeyOutcomesfortheBaseCaseandAlternativeScenarios(GDPat2007constantprices)
Scenarios
A B C D E F
Indicator Implement Private Only50%of Only20%of Implement

Base core fundingfor publicfunding publicfunding basecase
case infrastructure actionplan availablefor availablefor andrail
program notavailable actionplan actionplan extension
GDPin2030(US$millions) 4,560 3,895 3,745 3,313 2,868 4,721

GDPgrowthrate(annual%for201030) 7.2 6.4 6.2 5.6 4.8 7.4
GDPpercapitain2030(US$) 324 277 266 236 204 336
Source:AfDB2009.
A4. Key Policy Issues for the Government

3.17 The following important policy issues for consideration by the government and the donor
community have been identified:
Early action on the core infrastructure programIt is important for Burundi to maintain the
current growth momentum generated by the ongoing donor support for infrastructure
rehabilitation. An early launch of the proposed infrastructure action plan requires a series of
decisions by the government and donor community regarding the design of particular programs
and projects and the funding arrangements for these activities. There is a degree of urgency
associated with some of these decisions because implementing a number of key components of
the program must begin in 2010 if the above-mentioned targets for the power and transport
sectors are to be realized and if the current growth momentum in the economy is to be sustained.
Private funding mobilized for infrastructureThe proposed action plan calls for increased use of
public-private partnership (PPP) arrangements to mobilize private sector funding for the program.
In general, accessing private funding under these types of arrangements enables governments to
avoid or defer public spending on infrastructure without forgoing its benefits. To manage the
risks associated with the proposed PPPs and to ensure that they provide high-quality
infrastructure services in an efficient manner, the government will need to give careful attention
to the following three broad sets of concerns: (1) the legal framework governing the PPPs; (2) the
processes for selecting and implementing PPPs, including the roles played by relevant
56
government agencies; and (3) the contractual obligations on which the PPPs are based that
directly determine the fiscal risk incurred by the government.
Viability of a rail extension into BurundiAn important point that emerges is a series of
questions about the extension of the Tanzanian rail network into Burundi. If the rail is unable to
carry freight generated by the nickel mining operation in Burundi, then the volumes of freight
available from regular commerce appear to be too small to justify the investment at any time
within the next two decades. If refining metal at the mine site is the most attractive option
available to potential investors, there may be little appetite for the alternative (perhaps higher-
cost) option of using a rail system in Burundi for the freight services for the mine. More analysis
is needed on these options before final decisions can be made by the government. In any event,
there will be a need for close cooperation and coordination with Tanzanian authorities if the mine
is to go ahead, because the rail system would be used to transport metal from Kigoma to the port
at Dar es Salaam.
Degree of dependence on imported powerThis analysis points to the reemergence of a power
supply deficit by about 2024. The key policy question for the longer term is whether to develop
other potential domestic sites to keep dependence on imported power at prudent levels, or to
allow increased dependence on imported supplies from the East Africa Power Pool network. A
related question concerns the likely cost of new sources of domestic supply, compared with
imports from a low-cost producer such as Ethiopia.
3.18 A 20-year program of this magnitude inevitably faces risks and uncertainties, large and
small, foreseen and unforeseen. Many possibilities can be considered, including, for example, major
political risks such as deterioration in internal security, or civil disturbances in neighboring countries that
affect the overall performance of the Burundi economy and its attractiveness as a destination for private
investment. There are also risks that stem from the international environment, including sharply higher
petroleum or raw material prices that may adversely affect the attractiveness of investing in Burundi.
3.19 The risks and uncertainties of most interest at this stage relate to the design, funding, and
implementation of the proposed program. To manage these risks, the government and donor
community will need to strengthen coordination mechanisms for the infrastructure sectors, starting with
early completion and adoption of the proposed master plans. Regular meetings with donors may then be
required to monitor progress in program implementation. Risks include shortfalls in available funding,
delays in implementation, and macroeconomic difficulties arising from the high investment levels. A
further key risk is the inadequate supply response. Aid-financed projects usually create a strong demand
for imports because the domestic economy is not able to supply the required goods. Action should
therefore be taken to ensure an adequate supply response to the infrastructure plan.
A5. Priority Interventions in the Short and Medium Terms

3.20 Based on the detailed analysis conducted in the context of this study, the following priority
recommendations concerning the power, transport, and communications sectors could be identified for the
short and medium terms (201015).
3.21 Improving access to electric power:
Improve demand management through
installation of prepaid meters;
distribution and promotion of compact fluorescent lights;
utility energy audit; and
promotion of energy efficiency investments by large consumers.
57
Accelerate electricity supply through
development of Burundis domestic hydroelectric potential, commissioning Kaganuzi (5
megawatts), Mpanda (10.4 megawatts), and Kabu 16 (20 megawatts) plants;
development of a small-size run-of-the-river hydropower plant;
participation in regional hydroelectric projects.
Expand the transmission grid through
upgrading and expansion of regional network to 220-kilovolt network to establish regional
links;
provide 110-kilovolt lines with necessary substations to all provincial capitals;
develop system protection and control functions for the network.
Strengthen financial capacity of REGIDESO through
financial restructuring of REGIDESO, including a revision of the tariff structure;
separate REGIDESO into water and electricity companies.
3.22 Improving transport infrastructure:
Develop and maintain road infrastructure through
rehabilitation of the road network, with selection of roads according to economic, social,
and connectivity aspects;
development of a program for regulating and managing road transport;
capacity building and technical studies (such as transport industry surveys, passenger and
freight information, and assessment of road conditions); and
a clear strategy for maintenance, addressing maintenance works, funding of the required
program, and involvement of the private sector and the community.
prepare a civil aviation business plan that addresses
the role of the civil aviation authority (Rgie des Services Aronautiques- RSA), with
particular attention to measures needed for compliance with the International Civil Aviation
Organization (ICAO) and Civil Aviation Safety and Security Oversight Agency
(CASSOA)45;
training and other capacity-building programs to meet these standards; and
the outline of a commercial plan for the development of Bujumbura airport.
3.23 For accelerating the development of communications, in addition to already ongoing projects
financed mainly by the World Bank, the following measures are recommended:
extension of ongoing capacity building to support the regulatory and policy environment; and
additional support for the development of new applications throughout the country, including the
development and/or expansion of e-education, e-health, and e-commerce services for
communities, schools, hospitals, and business entities.
45
This agency was founded by the East African Community (EAC) member-states Kenya, Tanzania, and Uganda in April 2007.
58
B. MAXIMIZING THE BENEFITS OF BURUNDIS EAC MEMBERSHIP
3.24 Regional integration is another critical component of Burundis growth strategy. Being
landlocked, and with long distances to the nearest seaport, the government recognizes that increased
economic integration is key to achieving greater competitiveness in regional and global markets and in
laying the foundations for an extended period of strong growth. In addition to easing the current high
transport costs and improving access to goods and services, regional integration will also provide a larger
market for Burundis exports and help attract foreign investment. True regional integration would help
Burundi achieve returns to scale, especially with increased labor mobility. The EAC agreement can also
create a push for economic and structural reforms. In addition, negotiating trade agreements as a group
may give more incentive for other countries and regions to enter into a free trade agreement with Burundi.
For example, the EAC is negotiating an Economic Partnership Agreement (EPA) with the European
Union (EU). A final gain from the EAC may come from its power serving as a stabilizing factor in the
region.
3.25 Like many other developing countries, Burundi has been a member of several regional
agreements in the past, with an overall unsatisfactory experience.46 There are various reasons for the
poor results, including missing prerequisites, such as stability and political will, and the lack of common
interest. The objective of this section is therefore to provide a coherent strategic framework to guide
Burundis actions toward a successful membership in the EAC, building on lessons learned. Gains and
risks of Burundis accession to the EAC are discussed, and policy recommendations are developed to help
Burundi maximize the benefits of the already ongoing regional integration process.
3.26 One of the biggest challenges of the integration process will be in developing and
implementing a sustainable regional integration strategy and its road map. Because the moratorium
(requested by Burundian authorities in order to be prepared) ended in July 2009, it is critical to set up
such a strategy to ensure Burundis successful participation in the EAC. The strategy should focus on the
priorities and maximize the contribution of development partners and other actors. Suggestions include
the following: (1) prepare an action plan, including sequencing and prioritization of actions; and (2) create
or reinvigorate communication and coordination mechanisms.
3.27 There are a number of potential benefits and costs to joining a regional integration process.
A summary of those benefits and costs and of guiding principles for an effective integration process is
given in box A.1 in appendix 3. A number of actions, however, are necessary to maximize the benefits
from the regional integration process. In the concluding section, recommendations for policy action are
proposed as well as an action plan matrix suggesting the way forward.
3.28 The following section offers recommendations to help Burundi maximize the benefits of its
accession to the EAC. It proposes a road map and highlights critical actions that need to be taken
immediately and in the medium term to move forward with the regional integration agenda. Areas of
focus include institutional reforms, capacity building, increased involvement and facilitation of the private
sector, and a clear coordination and communication mechanism.
B1. Brief History of the EAC: Objectives, Achievements, and Challenges

3.29 The EAC targets a deeper form of regional integration. Two decades after the collapse of the
previous East African Cooperation, the Treaty establishing the EAC was signed in November 1999 by
Kenya, Tanzania, and Uganda. The EAC intends to go beyond trade and economic arrangements,
commonly known as a market-oriented integration process; and to implement a strategy aimed at
46
Since its independence in 1962, Burundi has engaged in various regional agreements and initiatives, including: the Economic
Community of the Countries of the Great Lakes (CEPGL) (1976), the Kagera Basin Development Organization (KBO) (1979),
the Economic Community of Central African States (ECCAS) (1983), the Cross-Border Initiative (CBI) (1992), and the
Preferential Trade Area (PTA) (1984) that became the Common Market for Eastern and Southern Africa (COMESA) (1993). A
synthesis of the performances of the different agreements is presented in table A.5 in appendix 3.
59
developing and improving regional infrastructure, managing regional commons, and producing regional
public goods, thereby moving toward a deeper form of regional integration.
3.30 The characteristics of the EAC member-states vary considerably. Since July 1, 2007, the
EAC has comprised five member-states. Burundi and Rwanda were officially accepted as new members
of the EAC in November 2006, and were effectively admitted by July 1, 2007. The area of the EAC has a
total size of 1,702,000 square kilometers, has a GDP of about US$75 billion,47 and represents a
population of 131 million. There is considerable variation in the economic performance of the member-
states (table 3.3). Kenya is far ahead, with a GDP about 20 times as high as that of Burundi, which has the
smallest economy in terms of GDP.
Table3.3:SelectedSocioeconomicIndicatorsforEACMembers,2008
Indicator Burundi Kenya Rwanda Tanzania Uganda Total
GDP(currentUS$millions) 1,163 23,507 4,457 20,490 14,529 75,146
a
GDPpercapita(2000constantUS$) 111 464 313 362 348 370
a
RealGDPgrowthrate(200208,annual%) 3.2 4.5 7.2 6.9 7.8 6.1
Population(millions) 8.1 38.5 9.7 42.5 31.7 130.5
Landsize(thousandsofsq.km) 25.7 569.1 24.7 885.8 197.1 1,702
Populationdensity(populationpersq.km) 314 67 394 48 161 n.a.
a
Lifeexpectancy(years) 51 54 50 56 53 54
a
Note:n.a.=notapplicable. Weightedaverage.
3.31 The ultimate objective is to establish a political federation built on a single market with free
circulation of goods, services, and factors of production and harmonized economic policies and
regulatory arrangements. So far, the EAC has progressed relatively quickly. It has adopted an approach
of parallel implementation of programs regarding all steps of the market integration process. The
integration process of the EAC benefited notably from the COMESA free trade agreement48 and the
membership of most countries in the Cross-Border Initiative for regional integration.49 Substantial
progress has been made in a relatively short period of time and with limited institutional capacity at the
regional and country levels. This is a signal of strong political commitment to building an institutionally
coherent and economically integrated core group, which could inspire the rest of Africa.
3.32 Encouraged by the good performance so far, negotiations started for creating a free trade
area and a customs union among Kenya, Tanzania, and Uganda. In contrast to a number of other
initiatives in Africa over the past three decades, the EAC has moved quickly to eliminate a large
proportion of tariffs on intra-EAC trade as an integral part of establishing a free trade area among the
founding members. The common
external tariff (CET) was adopted Table3.4:ExternalTariffsintheEAC,200008
on July 1, 2009, and is fully Country 2000 2001 2002 2003 2004 2005 2006 2007 2008
effective. (The development of the Burundi 25.3 22.5 15.7 13.4
CET for the EAC member- Kenya 20.5 21.9 18.5 15.0 15.0 15.3 15.7
Rwanda 11.7 11.0 18.5 15.9 16.8
countries is given in table 3.4.) Tanzania 18.2 15.2 15.1 14.5 14.3 16.4
The schedule of accession allows Uganda 10.0 9.8 9.3 9.7 9.3 9.0 13.8 13.3 16.1
commensurate action on tariffs by Source:UnitedNationsTradeAnalysisandInformationSystem.
Burundi and Rwanda to be Note: =notavailable.
47
That figure is based on 2008 current prices, according to World Development Indicators.
48
Kenya has been a member of the COMESA free trade agreement since its beginning in October 2000, Burundi and Rwanda
became members in January 2004, and Uganda joined in 2005. However, a tariff of 20 percent was still in place for Uganda
relative to the COMESA countries. Tanzania withdrew from COMESA a few years ago to focus on its membership in the South
African Development Community (SADC).
49
This initiative was launched in early 1992, mainly with the support of the World Bank, and aimed at accelerating the
liberalization of economies in the subregion to strengthen regional integration in eastern and southern Africa.
60
undertaken until 2010. Equally, negotiations started in 2006 to establish a common market by 2010
(including allowing for free movement of labor by 2009) and a monetary union by 2012. Furthermore,
efforts are under way to fast-track the establishment of a political federation, the final stage of the
integration process, by 2015.
3.33 The strategic agenda of the EAC, however, is very ambitious and might need to be
reconsidered to avoid undermining both the efficiency and the credibility of the process. The
development strategy of the EAC has many quite ambitious objectives and policy actions covering a very
wide range of issues, without a good sense of priority (World Bank 2008d). Lack of prioritization means
spreading EACs scarce human, institutional, and financial resources extremely thinly over too many
activities. As a result, such a broad and overloaded agenda frequently leads to slow progress, missed
deadlines, and a lowering of credibility of the EAC secretariatwhile it may also be interpreted as weak
political commitment by the member-states.
3.34 The challenge of maintaining the momentum created in the past few years will require
prioritization of objectives and sequencing of planned actions, including securing adequate
resources. To achieve a deeper integration, the EAC secretariat could focus on seeking and setting up
agreed policy priorities. Given the previous progress and the necessity of keeping the common interest in
mind, suggestions for immediate intervention could focus on deepening the trade integration process and
dealing with regional public goods in infrastructure. This includes, in particular, (1) finalizing free trade
area issues within the group, (2) consolidating the customs union, (3) developing a common trade policy,
(4) streamlining overlapping commitments, and (5) establishing and starting the implementation of a
regional infrastructure road map. Definitely, a realistic and gradual agenda could make maximum impact
and build confidence in the EAC secretariat and the member-states.
B2. Burundis EAC Membership: Benefits and Risks

3.35 There are several potential benefits and risks of regional integration. The following section
outlines the challenges of Burundis accession to the EAC, the possible noneconomic gains (especially
those related to the reform agenda), and the trade gains that could result from EAC membership. It further
explores how the industrial and services sectors could benefit from foreign direct investment (FDI), the
role of regional infrastructure, and the possibility of Burundi becoming a regional financial and transport
hub.
Burundis Accession to the EAC: Challenges and Strategy

3.36 Burundi authorities have shown strong commitment to the EAC membership. Since Burundi
was officially admitted as a member of the EAC on July 1, 2007, strong efforts have been made by the
authorities to align with the EAC requirements, especially in political, institutional, and economic areas.
The political will resulted in the signing of the treaty, the ratification of protocols and conventions, and
the quest for development partners support alongside the progressive stabilization of the country and
(although late) a communication campaign aimed at sensitizing the population to support the integration
initiative.
3.37 An institutional structure has been put in place to further the integration agenda. In
November 2007, a new government included (for the first time) a ministry in charge of regional
integration and EAC affairs, including a vice-ministry exclusively in charge of EAC affairs. In early
January 2009, these institutions were restructured into a full ministry in charge of EAC affairs, as
practiced within the other EAC member-states. This constituted a strong signal of political commitment to
enable the dedicated ministry to play its role fully.
3.38 Burundi, however, still faces institutional weaknesses regarding the challenges ahead (both
internal and external). First of all, although the working language in the EAC is English, most Burundians
speak only French. This language misalignment has severe consequences in terms of common
61
understanding and follow-up on agreed decisions. Beyond that barrier, the government experienced
difficulties in paying member contributions and allowing staff to fully attend the frequent regional
meetings. This fact has raised questions about Burundis capacity to engage in a regional program, and it
could undermine its credibility. Furthermore, compared with the other member-states, current staffs
capacities, skills, and experience are limited and do not always meet the technical requirements of the
regional integration process.
3.39 Moreover, Burundi cannot move ahead without addressing the structural constraints facing
its economy. Although Burundi remains the smallest economy within the region (table 3.3), a process of
catching up will have to deal with several constraints, including among others (1) poor physical and
energy infrastructure (including transportation system); (2) lack of dynamism and competitiveness of the
domestic private sector as well as a weak financial system (including the payment system); (3) a
nonconducive business environment (including the lack of an encouraging statutory framework for a
favorable investment climate); and (4) more generally, poor governance and the absence of a holistic
approach to benefit from the regional economic opportunities.
3.40 The biggest challenges will be prioritizing and sequencing the next steps within a
sustainable regional integration strategy and its road map. Given the delay in implementing the
integration process, it is critical to set up a comprehensive strategy to ensure Burundis successful
participation. The necessary steps would be described in a coherent road map for the accession, including
a sequenced action plan reflecting immediate priorities. Such a strategy would enable the government to
plan accordingly, move on quickly with the most urgent measures, and maximize the contribution of
development partners who can focus on specific areas of the road map.
The Need to Push the Reform Agenda

3.41 In addition to offering economic gains (see below), a regional integration process also
provides noneconomic gains through regional commitment mechanisms. These gains can be grouped
into two categories: (1) those addressing political and security issues and (2) those pushing economic and
structural reforms. Regarding the first category, for example, Burundi could benefit considerably from a
regional political stabilization effect. Furthermore, regional integration mechanisms tend to strengthen
an individual countrys bargaining power as well as the credibility of its policies (Baldwin, Franois, and
Portes 1997). In addition, they tend to seriously lower security risks through regional agreements and
other common arrangements. Moreover, there is a hope that the land issues that are a permanent source of
conflict in Burundi could be addressed at least partially through the establishment of the EAC common
market, expected in 2010.
3.42 The potential noneconomic gain the Burundian strategy should focus on is advancing the
ongoing reform agenda. Because the accession is subject to an alignment on some regional prerequisites,
the regional mechanism is an excellent tool to advance the reform agenda. This will be particularly true in
the following five main areas:
1. Implementation and harmonization of fiscal instrumentsAs far as Burundis accession to the
customs union is concerned, there are numerous measures that need to be implemented without
delay. They include the harmonizing the legal framework, including the fiscal and customs codes;
computerizing the tax department; introducing the value-added tax in lieu of the sales tax by July
1, 2009; adopting the common external tariff in 2009, including the negative list (and eventually
the timeline for alignment); aligning with the EAC budget calendar in 2010; and establishing the
Burundi Revenue Authority, which is expected to be in place by January 2010 and fully effective
by March 2010. Many these fiscal reforms are under way or have been completed.
2. Development of an integrated financial marketThe Central Bank of Burundi needs to accelerate
the financial sector reform to comply with EAC standards and norms. A recent study showed that
information on current volumes of cross-border trade in financial products is generally sketchy
62
and incomplete, making further integration problematic by obscuring the extent of cross-border
links.50 Specific actions would include promoting market access for investors from the EAC,
setting up a specific unit responsible for systems and methods of payment, and defining a strategy
and policy modernizing the payment systems and necessary infrastructure in line with the plans of
the EAC (EAC 2006). This can be done in cooperation with the Central Bank of Rwanda, which
already has started such a reform.
3. Reduction of tariff and nontariff barriers (NTBs) and development of a common trade policy
Burundis substantial progress in dismantling its tariff barriers and accessing the COMESA free
trade area since January 2004 will be instrumental for its accession to the EAC free trade area.
However, as in the other EAC member-states, NTBs in place in Burundi remain a serious concern
for intraregional tradeespecially customs and administrative procedures, including the length of
the clearance formalities and the multiplicity of institutions of control at the port in Bujumbura.
Proposed measures to remove Burundis NTBs are described in box 3.1.
Box 3.1: Proposed Measures to Remove Burundi NTBs within the EAC
x Establishing a mechanism for identifying and monitoring the disposal of NTBs. Preferably, this
mechanism would include the public sector and private sector representatives concerned with issues of
trade facilitation.
x Consolidating institutions in charge of customs clearance and administrative control at the port of
Bujumbura in a one-stop window to streamline the transit and customs formalities.
x Improving layout and equipment of border posts.
x Introducing competition in the selection of companies for preshipment inspection.
x Gradually establishing an infrastructure for quality control and certification.
Source: Adapted from World Bank (2008d).
4. Streamlining of overlapping agreementsBurundi is part of multiple arrangements that may

negatively impact the efficiency of its programs. In addition to the fiscal cost of being a member
of several organizations, multiple memberships imply multiple administrative procedures; and
that poses problems for a country with limited human, institutional, and financial capacity. As
mentioned above, as a member of the EAC, Burundi is negotiating an EPA with the EU. In
September 2009, a road map for the finalization and eventual signing of the Framework EPA was
approved by the respective council of the EAC.51
5. Promotion of an environment conducive to businessRegarding the business environment,
Burundi will need to underpin recent considerable progress. Although Burundi recently drafted
several business instruments,52 thereby taking into account the issue of harmonization with the
regional existing tools, the duty exemption, the tax incentives to export procedures, customs
regulations, and transit cargo regulations have not yet been harmonized on the EAC level; and
limited progress has been made in harmonizing standards. In the meantime, it will be instrumental
for Burundi to push forward in terms of implementation to allow regional companies and
investors to operate efficiently and stimulate investment and growth in the near future.
50
World Bank Financial Integration within EAC: Mission Report, May 2009.
51
Currently, however, the lack of consensus between the EAC and the EU on the sequencing of meetings is the main roadblock
to finalizing negotiations.
52
These instruments include revision of the investment code and a presidential decree on the setting up of an investment
promotion agency; revision of the commercial and public-private company codes; drafting of a competition law; and drafting of a
presidential decree on the public-private dialogue framework.
63
Trade Gains from Burundis EAC Membership
3.42 What are the potential gains from trade owing to Burundis accession to the EAC? To
answer this question, necessary to examine how Burundis trade structure has changed over time and to
consider the importance of the regional trade agreements to which it belongs. It is not surprising that
Burundi is likely to reap much larger traditional gains from trade from the Economic Partnership
Agreement with the EU than through the EAC. Burundis export structure is very complementary with the
EU import structure, and vice versa. However, there are a few interesting features of the EAC. In
particular, trade with the EAC has enabled Burundi to move into differentiated goods, which it does not
export to Europe. In addition, the EAC has been a very important market for expanding into new goods.
This offers a new channel through which regional agreements between similar countries can be beneficial
in Africa.
Are Burundi and the Other EAC Members Natural Trade Partners?
3.43 Regional trade agreements are more likely to be welfare improving if they are among
countries that are major trade partners prior to the agreement (see Summers [1991] and Krugman
[1991]).53 The logic is that countries that trade disproportionately more with each other are less likely to
cause trade diversion. In an extreme casefor example, if all of Burundis trade were with EAC
countriesthe agreement would be identical to free trade from a Burundian perspective. Regional trade
agreements among natural trading partners are also more likely to facilitate deeper integration, which
further expands welfare gains. Large trade flows provide bigger incentives for deep integration, and
similar tastes will facilitate institutional harmonization. In addition, proximity is required for integration
of many transport-related systems, such as railroad or trucking.
3.44 The share (measured in value) of Burundis exports that go to members of the EAC has
increased since 2000, but has remained below 15 percent in recent years. This is very low, compared
with the Southern Cone Common Market (Mercosur) and the North American Free Trade Agreement
(NAFTA) in which the share is more than 25 percent and the EU in which it is above 50 percent. In terms
of imports, the share from EAC members has also increased and is now around 20 percent.
3.45 Export similarity and complementarity indexes can be used to analyze Burundis trade
pattern. To understand how similar or complementary trade is between Burundi and other regions or
countries, export similarity and complementarity indexes are calculated between Burundi and other
countries in the EAC, other regions such as COMESA (non-EAC), and the EU. In a first step, the Finger
and Kreinin (1979) export similarity indexes (which describe how similar two trade structures are) are
calculated.54 If the product share distributions of the exports of two countries are identical, the index will
be 100; if the export patterns are completely different, the index will be 0. Another way to gauge the
importance of complementarity versus similarity in trade is to examine whether Burundis main export
products are imported on net by its trade partners. The complementarity index is the total share of
Burundis exports that a trade partner imports more than it exports. For example, Burundis biggest export
is coffee; therefore, net importers of coffee will be natural trading partners.55 A number close to 0 implies
that the country is a net exporter of the products that Burundi exports.
53
Krishna (2003) finds little empirical support for the natural trade partners hypothesis. However, Krishnas results are highly
dependent on the model he is estimating. In addition, he examines only 24 bilateral pairs, and it is possible that this is too limited
a sample to find evidence.
54
The index of export similarity is defined by , where ESI is the similarity between exports of
country A and country B. is the share of product i in js exports to the world.
55
First, we create an indicator that is 1 if a country is a net importer (imports exports > 0) of a given product. Next, we
calculate the share of Burundis exports that is on net imported by other countries and regions as
, where is Burundis share of exports in industry i, and is an indicator for whether country or region A is
a net importer of that product.
64
Table3.5:SimilarityandComplementarityIndexesforBurundi,19952007(Averages)
Index ExportSimilarity ExportImportSimilarity ImportExportSimilarity
Kenya 18.3 11.2 36.8
Rwanda 47.1 10.8 13.9
Tanzania 16.8 12.0 26.1
Uganda 50.2 12.3 20.7
EAC 26.5 11.8 31.4
OtherCOMESAcountries 4.7 15.0 41.1
EU 4.2 91.0 68.9
World 4.2 n.a. n.a.
Source:AuthorscalculationsbasedonCOMTRADEHS19926digitdata.
Note:n.a.=notapplicable.
3.46 The following findings can be derived from an analysis of the export similarity indexes
(table 3.5, Export Similarity column):
Within the EAC, exports are most similar to Uganda, where about half of exports, on average, are in
the same products.
Overall exports are more similar with countries in the EAC than with the other COMESA
countries, with the EU, or with the world.
Burundi and other EAC countries tend to be competing exporters.
Although trade diversion remains a concern, it does not seem likely to have a large impact on
Burundi, and all positive effects of regional integration mentioned above will outweigh any
negative effects.
3.47 The complementarity index (or export-import similarity index) shows clearly that the EUs
trade patterns are very complementary with Burundi (table 3.5). The EU is a net importer of 90
percent of Burundis exports. In contrast, EAC countries and other partners in COMESA are net
importers of only as much as 12 percent and 15 percent of Burundis exports, respectively.
3.48 Finally, the results of an analysis of potential import/consumer gains suggest that Burundi
is far more complementary with the EU than with EAC countries. Burundi is treated as the importer,
and the share of the regions exports of which Burundi is a net importer is examined. In other words, this
index looks at whether there are consumer gains from the various agreements. Again, results show that
there is a lot more complementarity between Burundi and the EU than with the EAC (table 3.5). Burundi
is a net importer of roughly 70 percent of the products that the EU exports, whereas it is a net importer of
more than 30 percent of EAC exports. This suggests that in the absence of differentiation and/or
diversification of the product base, the trend is not likely to change and the country could not expect to
maximize the gains resulting from its commercial policy.
Are There Gains from Trade in Differentiated Goods?

3.49 Although export similarity indexes and export-import complementarity are good ways to
model traditional trade, these relationships break down with differentiated goods. In theory, trade in
differentiated goods is still likely to be abundant, even when both partners export the same goods.
Differentiated goods are goods for which consumers appreciate variety. To measure the extent of
differentiation, data from Rauch (1999) is used. Rauch splits all traded goods into three categories: (1)
those traded on an organized exchange, (2) reference priced goods, and (3) differentiated goods. The first
two categories are meant to capture goods that are roughly homogeneous; the third category covers goods
for which brand and producer are relatively more important.56
56
About 10 percent of goods were not classified and therefore were excluded from the analysis.
65
3.50 Whereas the extent of trade in differentiated goods highlights an important avenue for
growth with the EAC, trade in differentiated goods is quite volatile; and it is too early to determine
how the EAC membership will affect it. Overall, differentiated goods account for less than 10 percent
of Burundis exports, though their share has grown in recent years. In contrast, Burundis exports of
differentiated goods to the EAC make up about two thirds of trade. Most of Burundis imports are in
differentiated goodsthough as a share of total imports from its trade partners, differentiated goods are
less important among the members of the EAC.
How Important Is the Extensive Margin of Trade?

3.51 Finally, a decomposition of trade into the extensive and intensive margins is conducted to
understand their importance for Burundi. The extensive margin is defined as exports of new products
and/or to new partners, and the intensive margin is products that were already exported in the second half
of the 1990s. Specifically, the shares of trade arising from new and existing goods were calculated. The
base period is average trade from 1995 to 1999.57
3.52 The EAC offers a good opportunity for Burundi to export new products. On average, the
extensive margin accounted for about one quarter of exports from 1999 to 2007. The extensive margin is
more important for exports to the EAC, where it accounted for about three quarters of exports. Only
exports of new products were considered to determine whether the EAC is used to test new products.58
New products accounted for about 7 percent of trade in 2007 overall; in the EAC, this amount rises to 28
percent. This implies that the EAC may offer some advantages in terms of getting into new products.
These results also imply that entering new markets with existing products is more common than finding
new export products. On the import side, the extensive margin is important because it may indicate
consumer gains from variety. Whereas about one third of Burundis imports are new country-product
pairs, the figure rises to more than one half for the EAC. Gains are likely to be even greater in completely
new goods; however, the share of completely new imports is very low.
3.53 Burundi, however, might not be able to successfully maintain and grow new exports. The
extensive margin is measured as products/markets that are new since 2000. One issue with the figures,
whether for new product markets or new products, is that they tend to be very volatile from year to year.
This is surprising because, over time, one would expect the new exports to become more important
(assuming they grow) because the base year is not changing. If Burundi enters new markets and grows its
exports in later years, a larger extensive margin over time should be observed. The persistent volatility of
the extensive margin suggests that Burundi may be struggling to maintain and grow these new exports.
Benefits for the Industrial and Services Sectors from Regional Foreign Direct Investment
3.54 The flow and sustainability of FDI, sources of technology transfer and know-how, as well as
job creation will depend on the capacity of Burundi to offer economic externalities. According to the
theory of new economic geography, externalities include (1) the size of the market; and (2) the existence
of intermediary services generated by the presence of public goods that reduce transaction costs
(Krugman 1995; Fujita and Thisse 2002), as well as auxiliary services. Whereas market size refers to the
possibilities of generating economies of scale (through openness and competition), intermediary services
are derived from public goods such as infrastructure for transport and energy, hospitals, and schools.
57
The decomposition of trade is as follows: , where Vit is the value of exports of product i to
country j in year t, I is the set of products that exists in both the first period and period t, and N is the set of new products in year t
that is not in the countrys 199599 export bundle. This is an identity where total trade is decomposed into two parts: (1) products
that were exported (imported) in periods, the intensive margin; and (2) exports (imports) of new products. The share of trade
resulting from the extensive margin is defined as the new goods relative to total trade.
58
In this case, new products are counted as exports of products that were not exported between 1995 and 2000. New markets for
existing products are not counted.
66
There is, however, a minimum threshold below which these externalities do not produce economies of
agglomeration (Hugon 2003).
3.55 Concerning the regulation, overall, the reforms under way have made encouraging, but
insufficient, progress in a short time. It will be unlikely that, in a competitive market, companies are
positioning themselves in areas where the business environment remains difficult, leads to additional
costs, or is simply ineffective.
3.56 Regarding industry and service sectors and companies, the strategic regional partnership
seems to be the best strategy for an effective integration, and the potential to attract regional FDI
exists. Provided that the prerequisites mentioned above are gradually put in place, the interest of this
partnership for regional FDI will be twofold because firms can (1) increase profit margins resulting from
economies of scale generated by the enlargement of the market; and (2) relocate to production centers to
lower production costs, particularly those of labor. In Burundi, wages are low and labor is available
potential incentives to relocate.
3.57 This partnership could take the form of joint ventures with local small and medium
enterprises, equity participation in the shareholder, or subsidiaries from which the economy would
take the maximum benefit. There are several benefits of FDI for the domestic market, including large-
scale production with lower prices, higher standards of quality, technology and know-how transfer in the
industry, expansion of the tax base, and more investment flows and job creation with a multiplier effect
on the national income.
The Role of Regional Infrastructure

3.58 Public infrastructure plays a critical role in promoting economic growth. Here, infrastructure
is understood broadly to include transport (via roads, rail, ports, and airports), energy,
telecommunications, and other externalities (financial markets, commercial and institutional basic
services). In a regional context, infrastructure influences the spatial concentration through two channels:
(1) lowering transport and transaction costs, and (2) reducing fixed and variables costs. The development
of such infrastructure on a regional level accelerates the achievement of a threshold effect necessary to
realize positive externalities, and thus leads to the stimulation of regional growth.
3.59 Burundi faces a substantial infrastructure deficit. An infrastructure action plan that has been
prepared by the African Development Bank reports that, because of the limited access and the poor state
of infrastructure, the cost and adequacy of these services affect commercial opportunities for the whole
business community (farmers, entrepreneurs, and small and large businesses). Such a situation leads to
substantially higher costs, undermining the competitiveness of Burundis business in regional and global
markets. (The infrastructure action plan is described in more detail in section A of this chapter.)
The Possibility of Burundi Becoming a Regional Integration Financial and Transport Hub
3.60 The role of a transshipment hub using the port of Bujumbura could be a viable option for
Burundi. Because of its geostrategic location, one expects that the port of Bujumbura has the potential to
return to its former role as traditional gateway to providing goods and transport services to countries
bordering on Lake Tanganyika, especially to the eastern portion of the Democratic Republic of Congo. In
the context of regional integration, development of intense mining and other economic activities in
Burundi, eastern Democratic Republic of Congo, and Rwanda could efficiently use the services of the
port of Bujumbura. In addition to its location, it is adequately equipped and currently operating at less
than 50 percent of its capacity, according to statistics from the port authority. Mining activities could
include exploitation of the mineral potential in eastern Democratic Republic of Congo, with all related
possible economic activities and economies of agglomeration.
3.61 The future role of the port of Bujumbura, however, primarily depends on developments
regarding the Tanzania Railways Corporation. Until recently, the bulk of cargo movements between
67
Burundi and Dar es Salaam was carried by rail to and from Kigoma (Tanzania) or Mpulungu (Zambia),
and by vessel from these ports to and from the port of Bujumbura. In recent years, lack of suitable rolling
stock and the age of the cargo-handling equipment at the port of Kigoma, and the deterioration of the
Tanzania Railways Corporation have seriously undermined the activity of the port of Bujumbura. The
volume of trade entering and leaving Burundi through the port has declined sharply and is now a third of
what it was in 2000. Possible scenarios range from a disappearance of the need for transshipment from
rail to vessels at Kigoma if a proposed rail extension into Burundi is constructed to carry freight and
passenger traffic direct to Bujumbura, to an improvement in Tanzania Railways Corporations
performance and a move back to rail transport for bulk cargo. Even in the most optimistic case, though,
only modest investments to upgrade equipment would be necessary.
3.62 Concerning the financial sector, recent studies recommend focusing on domestic issues in
the short and medium terms. Priorities are strengthening the existing structure, improving the access to
financial services, and lowering their cost. Certainly, the proximity of an economic area such as the
eastern Democratic Republic of Congo, with intense economic activity operating entirely in the informal
sector, can be seen to have large potential. Burundi, however, first needs to strengthen its domestic
financial, regulatory, and institutional base, and bring them to the level of regional standards in the EAC.
This means that Burundi is facing two constraints: (1) without an efficient and solid financial sector that
serves a large proportion of the population, Burundi will not attain a satisfactory level of growth and
private sector development, and poverty reduction will be slow; and (2) Burundi needs to fill the
regulatory and institutional gap that exists between it and the four other EAC member-countries.
Consequently, it is too early for Burundi to seek a specialized niche within the EAC where the country
may develop a comparative advantage and assume a regional leadership role, including expanding
services to the eastern part of the Democratic Republic of Congo. Over the next five years, Burundi
should focus on these domestic issues, with the EAC as a benchmark that guides regulation, supervision,
and institutional strengthening.
B3. Risks and Risk Mitigation

3.63 In a country that has a large deficit of technical and institutional capacity, the challenge
posed by a regional integration policy is enormous; and there are many risks that need to be
mitigated. These risks include the following:
1. CapacityWeak technical and institutional capacity may delay or compromise the
implementation of the agreed program. In this regard, a capacity-building program is being
developed with assistance from the U.K. Department for International Development (DfID); but
permanent technical assistance will be needed, at least in the first years.
2. Trade diversionAlthough trade diversion remains a risk, the possible negative effect is expected
to be limited, given the small range of products traded in the EAC.
3. Loss of revenue resulting from adoption of the common external tariffA study of the fiscal
impact estimated a loss of 1.7 percent of total revenue. This is not a substantial loss, given
existing alternatives of compensation, mainly through tax reform (expanding the tax base,
eliminating exemptions) and the mechanisms of compensation agreed between COMESA and the
EAC for Burundi and Rwanda.
4. Indirect costsIndirect costs of regional integration are related to the capacity of regional hubs to
attract and combine the best opportunities (research, education, employment, and so forth),
leading to brain and capital migration. Other indirect costs are related to the likelihood of criminal
activity taking advantage of free movement in a fragile area. The first problem is a political
economy issue and needs to be addressed on the community level. The latter poses a security
problem and also needs to be addressed at the community level; it further requires the setting up
of a monitoring mechanism and the establishment of regional security.
68
5. Other risks related to persistent insecurity at the western boarder of BurundiAlthough
establishing technical committees for monitoring and evaluation and donor support should
mitigate such risks, the government should remain focused on its regional commitment to keep
the momentum.
6. Timing of the implementation of priority measuresThe timing is considered very ambitious for
the newest member-statesespecially Burundi. It may undermine the states implementation. In
this regard, the measures are regularly negotiated according to the principle of variable geometry,
recognized within the EAC.
B4. Policy Recommendations

3.64 The adoption of a strategy based on a comprehensive approach to regional integration is the
appropriate way to maximize the chances of success in the accession of Burundi to the EAC. It is
clear that membership in the EAC is a significant challenge for the country. Weak technical and
institutional capacities are major constraints. However, these problems can and should be addressed. The
main recommendations to advance the regional integration process are these:
1. Comprehensive approachLessons from past experiences show that the basic principles of
regional integration are important. The authorities must give them the consideration they deserve.
For Burundi, adopting a strategy based on a comprehensive approach to regional integration
seems to be the appropriate way to a successful accession to the EAC.
2. Improvement of infrastructureThis strategy requires that national and regional infrastructure
(energy, transport, and communications) and regulatory reforms form the backbone of a
successful integration with the EAC.
3. Sequenced processA successful accession strategy will require a sequenced process that gives
high priority to immediate measures while gradually implementing those for the medium and
long terms. To this end, the policy matrix gives a comprehensive and sequenced action plan
where immediate priorities are clearly identified (see table A.7 in appendix 3). The immediate
priorities and objectives of the action plan are the following:
Continue institutional reformsImplementing institutional reforms is important; and the
government needs to keep the momentum, particularly by organizing training sessions and
mobilizing the budget required to fully participate in the regional meetings. Measures
include recruiting qualified staff for the Ministry of East African Community Affairs,
acquiring adequate equipment, training staff in English and in negotiation skills, and
implementing a communications strategy.59
Economic reformsThe deepening of economic reforms and the enhancing of private
sector economic activity will be cornerstones of success during the first stages of accession
to the EAC. This involves the following considerations:
For trade and investment, (1) support for the diversification and discovery of new
markets (Ministry of Trade and the Chamber of Commerce); (2) elimination of
remaining NTBs (including those in the financial system); (3) an effective and
practical support to promotion of the private sector and improvement of its
competitiveness by enhancing the framework for public-private dialogue; and (4)
improvement of the business environment by ensuring compliance with EAC
standards, and the implementation of an attractive tax code.
59
See table A.6 in appendix 2 for a matrix of strategic and indicative outputs for the Ministry of East African Community
Affairs.
69
For the industry and services sector, a strategic partnership is a viable option.
Burundis private sector should try to benefit from FDI, particularly in terms of
technology and know-how transfer.
For services and service delivery, the priority is to reform and develop the financial
system, including the strengthening of banking sector activities, to improve the port
services and develop the information and communication technologies sector.
For infrastructure, special emphasis should be placed on energy, transport, and
communications. Given the required capacities, the government should anticipate the
strengthening/training of its human resources and the request for adequate technical
assistance, when necessary.
4. Projects for the medium and long termsIn the meantime, the government could start or continue
discussions on the medium- and long-term projects to prepare feasibility studies and mobilize
resources from donors and the private sector.
5. Coordination mechanismA coordination mechanism is critical at this stage. It should include
the government, the donor community, the private sector, and civil society. It can clarify the roles
of the potential actors and help maximize their contributions.
6. Communication mechanismThe communication mechanism is essential and needs to be
finalized and implemented as soon as possible. (An action plan is currently being finalized with
technical assistance from the U.K. Department for International Development.)
C. REFORMS NEEDED TO INCREASE THE PRIVATE SECTORS CONTRIBUTION

TO GROWTH
3.65 Burundi has one of the most unfavorable business environments worldwide. The business
environment and, more specifically, the investment climate is considered essential for a successful
diversification of the economy (Page 2008). The private sector in Burundi, however, faces several major
constraints (discussed in section A of chapter 2). The regulatory framework is poor: administrative and
bureaucratic procedures are cumbersome, the tax burden is one of the highest in the world, legislation is
outdated, and the capacity to enforce contracts is weak. The economy is dominated by public enterprises,
several of which face financial problems. Moreover, the financial sector is small and not well developed;
and access to finance is a major obstacle for companies in the formal and informal sectors.
3.66 A number of measures are necessary to enable the private sector to contribute more to
economic growth. Several regulations are already under revision or have been revised, and the
momentum must be sustained. Given below are recommendations to overcome the main obstacles the
private sector in Burundi is facing. They are in line with the areas of activity outlined in the countrys
PRSP (Republic of Burundi 2006) regarding private sector development: (1) short-term emergency
actions, (2) ongoing reforms of the legal and regulatory frameworks, (3) reinforced incentives for growth
in private investment, and (4) government disengagement from the productive sector and privatization of
public enterprises.
3.67 In this section, the way forward is outlined to improve the business environment in Burundi.
Recommendations are given to overcome supply-side constraints, continue the improvement of the
regulatory environment, increase competition, enhance the countrys image as a business and investment
destination, and strengthen the dialogue between the government and the private sector.
C1. Overcoming Structural Constraints

3.68 To provide the private sector with the necessary means to operate its firms and increase
productivity, it is essential to fill the main infrastructure gaps. An infrastructure action plan
specifically addressing the energy, transport, and communications sectorsprovides details on how these
70
obstacles can be overcome (see section A). Burundi can also benefit from the regional integration process
that is currently occurring. Potential gains for the private sector include improved links between the
domestic and regional markets, regional infrastructure projects, and an incentive to accelerate the reform
agenda (as outlined in section A).
C2. Addressing Supply-Side Constraints

3.69 Regarding the financial sector, Burundi is facing two constraints. First, without an efficient and
solid financial sector that serves a large proportion of the population, Burundi will not attain a satisfactory
level of growth and private sector development, and poverty reduction will slow. Second, Burundi needs
to fill the regulatory and institutional gap with the four other member-countries of the EAC. Clearly, as a
matter of priority, Burundi needs to focus on strengthening its financial sector, improving access to
financial services, and lower their cost. Over the next five years, the country must focus on domestic
issues and view the EAC as a benchmark that guides regulation, supervision, and institutional
strengthening. It is too early for Burundi to seek a specialized niche within the EAC, where the country
may develop a comparative advantage and assume a regional leadership role.
3.70 Recommendations for the financial sector are articulated around two main schemes: shoring up
stability and improving access. This means taking actions on both the supply of and the demand for
financial services. A summary of the necessary measures to be taken to improve the performance of the
financial sector is given in table A.13 in appendix 5. Key recommendations include the following:
1. Increased stability
state banksprivatize state-owned banks and improve supervision;
payment systemintroduce a clearing and real-time gross settlement system;
microfinancestrengthen supervision and institutions;
insurance sectorestablish independent regulatory/supervisory agency; and
pension systemchange management structure, internal restructuring, and parametric
reform.
2. Improved access
small and medium enterprises and rural areasdevelop new products and lending
mechanisms, build the capacity of the Banque Nationale de Dveloppement conomique
(BNDE) to lend in rural areas;
microfinancechange regulation to broaden the scope of microfinance institutions to
develop new products and reach more customers;
housing finance: authorize microfinance institutions to offer mortgage loans, decentralize
land registries;
insurance sectordiversify products, develop insurance brokers; and
availability of informationstrengthen existing information centers, offer assistance to
improve financial statements and business plans, strengthen the accounting and auditing
profession.
3.71 To promote technical and professional skills and facilitate transparent procedures in the labor
market, policy recommendation include the following (World Bank 2008a):
1. In the short term,
a needs assessment of required skills for private companies;
71
evaluation of the possibility of joint programs between public institutions and private firms
to improve technical and professional skills;
facilitation of the learning of English;
reexamination of labor market regulations.
2. In the medium term,
improvement of employees skills by making schooling available to a larger part of the
population, providing more opportunities for technical and professional education, and
strengthening existing skills;
revision of certain elements of the labor code.
3.72 In addition, the flow of information on the labor market needs to be improved by supporting
organizations that facilitate the match of supply and demand for labor by circulating relevant information
to a wider audience. It would be possible to create a national employment fund with a focus on improving
the availability and flow of information, setting up programs to improve technical and professional skills
based on market needs, and supporting employment creation in general.
C3. Continued Improvement of the Regulatory Environment

3.73 Continually improving the regulatory environment is necessary. The currently ongoing
reforms need to be completed as soon as possible and the appropriate regulations implemented. Specific
suggestions include the following:
1. Adoption by Parliament of the revised Commercial Code and the Code for Public and Private
Companies.
2. Finalize the revision of the tax code. Furthermore, a study of the reform of the tax code
completed in 2007 suggests replacing the impt forfaitaire (lump-sum tax) and other small taxes
paid by small enterprises with a simplified impt synthtique (synthetic tax) based on verifiable
and mostly quantitative criteria.60 The new tax would then be incorporated in the general tax code
under revision.
3. Finalize the revision of the mining code.
4. Prepare the accompanying regulation for commercial codes.
5. Further improve the efficiency of the commercial court and the arbitration centers.
C4. Improved Governance

3.74 The government recognizes the important role of improvements in governance. A National
Governance Strategy and Action Plan were officially launched in early 2009, and the government
committed to approve it before the end of the year. The design of the plan follows a systematic and
participatory approach aimed at integrating the different dimensions of governance. A Guide des
Consultations was prepared, outlining the context; main recommendation of previous studies on
governance and corruption in Burundi; and the framework, methodology, and organization of thematic
focus groups (Republic of Burundi 2009).
60
This reform would have several benefits: (1) it would help draw into the formal economy those informal entities whose
primary motivation for evading tax is not fraud, but an inability to meet the complex and burdensome rules and regulations; (2)
newly formal businesses would have better access to credit and government contracts, thus increasing their contribution to the
Burundian economy; (3) the reform would expand the tax base and provide the government with a more predictable revenue
stream; and (4) it would reduce the possibility of corruption in the present method of estimating business turnover. The
government is expected to seek technical advice (from the Foreign Investment Advisory Service) on the structure of the new
impt synthtique.
72
3.75 General recommendations include
implementing a policy of good governance and fighting against corruption;
involving civil society and the private sector during the entire elaboration process, and following
up and evaluating policies and projects;
strengthening the capacity (training and equipment) of public institutions, civil society, and
private sector organizations that are involved in promoting good governance and the fight against
corruption;
reforming the government and the public administration by putting in place or improving existing
institutional and legal frameworks; and
disseminating relevant judicial tools for good governance and the fight against corruption.
3.76 Several specific recommendations are given as well, grouped into three categories: (1)
socioeconomic governance; (2) political and administrative governance; and (3) governance related to
justice, peace, and security.
3.77 Burundi could benefit from an increase in competition. A draft Privatization Law was
prepared, approved by Parliament, and promulgated by the government; but the new text is considered
unsatisfactory by the private sector, donors, and many public institutions. The government has agreed to
undertake a revision of the new law. A consultant was identified to complete the revision and enable the
government to approve a new draft and submit it to Parliament by the end of 2009. The objective of the
new law would be to modernize the legislative framework to ensure that the governments withdrawal
from productive sectors will not result in noncompetitive market outcomes (creation of private
monopolies). The Privatization Law needs to be in line with modern international practices.
3.78 The ongoing privatization process of the coffee sector is an encouraging sign (see section B2 of
chapter 2 for details). Whether the economy will benefit, however, depends on the continued
implementation of the privatization agenda. In general, the performance and corporate governance of
public enterprises need to be improved.
C5. Improvement of the Countrys Image as a Business and Investment Destination

3.79 Burundi needs to improve its image and create incentives to attract investment. The
credibility of the country as a business destination for domestic and foreign investors must be
strengthened. Policy recommendations in this regard include the following:
develop a communication strategy to inform the public about the measures taken to improve the
business environment and about progress made regarding political and social stability;
regularly update the investment guide for potential foreign investors;
adopt an action plan aimed at reforming business regulation and the practice of creating and
operating businesses through a substantial reduction of the time and costs associated with these
procedures
C6. Strengthened Dialogue between the Government and the Private Sector
3.80 The dialogue between the public and private sectors in Burundi historically has been very
weak. Limited consultation and partnership with the private sector have tended to result in poorly
informed public sector decision making that often fails to foresee the impact of decisions on the private
sector. A general lack of mutual accountability between the public and private sectors has also resulted in
very weak policy implementation. In addition, many of the private and public enterprises and
organizations (including the Ministry of Commerce, the National Bureau of Standards, the BRB, the
Chambers of Commerce, and the accounting profession) need support to strengthen their capacity to
73
provide policy leadership and create the enabling environment for businesses to grow and be competitive
domestically, regionally, and internationally.
3.81 Private enterprises believe that one of the most important initiatives to improve the business
climate in Burundi is to strengthen the policy and operational dialogue between the government
and private enterprises. In June 2008, a presidential decree established a public-private sector
consultation frameworkincluding a general assembly, technical groups, and a permanent secretary; to
day, however, no meeting has been organized by the government. The private sector also needs to decide
which organizations will represent its interests in this dialogue with the government.
D. ENCOURAGING EMERGING GROWTH SECTORS

3.82 The strategy to encourage emerging growth sectors in Burundi over the short and medium
terms proposed in this Country Economic Memorandum is based on two pillars. First and foremost,
the country must address the low performance of the agriculture sector, including food and export crops.
Priority should be given to eliminating widespread food insecurity and malnutrition. Furthermore,
revenues from current exports (mainly coffee) need to be stabilized and increased. Second, in the medium
term, diversifying the economy and strengthening exports could lessen the nations vulnerability to
external shocks and create new opportunities for economic development and employmentespecially
off-farm employment. The measures described in the previous sections on infrastructure, regional
integration, and the business environment will enable the development of the emerging growth sectors.
D1. Agricultural Recovery: Food CropsProspects and Priority Interventions

3.83 Given the projected increase in population, the land situation is likely to worsen, reinforcing
the need for off-farm employment (see section A of chapter 2). Although additional marshland can be
developed, it will not be sufficient to meet the needs of the growing population. There is, therefore, a need
to shift from extensive to intensive agriculture. Given the current low productivity, the potential for
intensification is real. Improving input delivery and soil and water management is essential.
3.84 Burundi has the potential to produce a wide variety of crops. It has sufficient rainfall in
general and water resources for extended irrigation systems to withstand localized droughts. There are
significant deposits of mineral resources, such as phosphates or limestone that can be used to produce
fertilizer especially for the acidic soil. The country also has an abundant workforce and a number of
experienced technicians who can be trained in the use of new technologies. Strategic investments are
necessary to implement improved technologies (irrigation and drainage systems, terracing, and so forth).
3.85 The recent accession to the EAC offers a range of possibilities for Burundi. Extending import
and export markets, however, requires a lowering of the transport costs to enhance the competitiveness of
Burundis products (see section A of this chapter).
3.86 To use the potential of the country and increase the performance of the agriculture sector, several
priority interventions are necessary. Cross-cutting interventions include rehabilitating rural
infrastructure, improving delivery of financial services to the rural sector, and improving the business
climate. Specific measures are listed in table 3.6. In addition, a mechanism for monitoring and evaluation
must be put in place.
74
Table3.6:SpecificMeasurestoimprovethePerformanceoftheFoodCropsSector
Objective Measure
Sustainablegrowthofagricultural x Intensifyfoodcropproduction,by
productivityandproduction - strengtheninginputdistributionsystems
- improvingsoilandwaterconservationandmanagement
- reinforcingtechnologydevelopmentandtransfer
x Rehabilitatetoolsofproductionandinfrastructurerationallymanagenatural
resources,by
- restoring,improving,andconservingsoilfertilityinparticularand
naturalresourcesingeneral
- rehabilitating,creating,andstrengtheninglocalinfrastructurefor
storage,conservation,transformation,andcommercializationof
agriculturalproducts
- improvinganddevelopingroadinfrastructuretofacilitateaccessto
marketsanddistributionofinputs
x Promotelastingproductionsystemsanddevelopexistingpotential
Professionalizationofproducersand x Organize,structure,andprofessionalizeproducers
developmentofprivateinitiatives x Involvetheprivatesectorinmodernizingthesector
Strengtheningofmanagement x Revive,professionalize,anddecentralizesupportstructuresforproduction
capacitiesanddevelopmentcapacities x Contributetotheprotectionofagriculturalworkersagainstnutritional
ofthesector deficiencies,malaria,andHIV/AIDS
x Improveandstrengthenthecapacitiesofthegovernmentto
project/forecastmarketdevelopmentsandopportunities
Sources:Baghdadli,Harborne,andRajadel2008;RepublicofBurundi2008b.
D2. CoffeeThe Way Forward

3.87 Burundis coffee sector could improve its performances considerably within the next five
years if it undertakes key measures to increase production and enhance quality. According to Clay,
DeLucco, and Ottaway (2007), production could grow by 2535 percent in five years if incentives to
producers are rethought, access to inputs is improved, a greater focus of processing practices on quality is
ensured, and selling strategies are adapted to target specialty markets. Another study (by the OTF Group
[2008]) estimates that Burundis coffee export revenues could more than double within five years, thanks
to targeted interventions. Promoting the use of fertilizers could raise production by 20 percent, while
developing good agricultural practices (such as mulching) could generate an additional increase of
12 percent. Improving washing stations processing methods would enhance quality, which could
translate into a 15 percent price increase. Finally, modifying marketing practices and branding Burundian
coffee could help prices grow by 10 percent. To achieve such results, the industry will need to (1) foster
an enabling environment; and (2) introduce improvements along the value chain, both immediately and in
the medium term.
Immediate Actions
3.88 Setting the market rules and the structure of the industry is indispensable to ensure a proper
development of the coffee sector and attract the type of investors required to enhance its performances. In
addition, improving the industrys practices should start at the production level.
Fostering an Enabling Environment

3.89 To improve the business environment in the short term, the following actions are recommended:
1. Strengthen the capacity of the regulatory agency (ARFIC)A regulatory agency was created by
presidential decree in June 2009. It should swiftly establish an efficient regulatory framework for
the industry, a prerequisite to attract first-class investors. Privatizing and liberalizing of the sector
75
will not guarantee that competition prevails after the reform; it will be up to the agency to
introduce rules fostering competition and to ensure that they are followed.
2. Pursuing the privatization agendaBidding invitations for the sale of the washing stations and
dry mills were published on June 15, 2009. The government lifted some of the obstacles to the
divestiture (for example, amendment of the 30-year lease contract with the Socits de Gestion
des Stations de Lavage [SOGESTAL - the washing stations management companies] and the
Socit de Dparchage et de Conditionnement [SODECO - the hulling and processing company]
on August 21, 2009). But it will need to address further issues in the upcoming months.
Reforming the coffee sector is a sensitive matter in post-conflict Burundi, and trying to build a
consensus is essential. The ministry responsible for governance and privatization launched a
communication campaign to inform the general public about the process and consult with all
stakeholders of the industry, some of whom are voicing strong concerns. The employees of the
washing stations, dry mills, and the Office du Caf du Burundi (OCIBU), for instance, are
understandably preoccupied with how privatizing will affect their jobs. The dissatisfaction of
producer organizations with the proposed ownership structure of the washing stations also will
need to be addressed. The credibility of the process depends on the governments ability to stay
the course and deal with these obstacles. This steady focus will assure stakeholders that the
privatization will proceed and be run smoothly, which is essential to retain serious bidders in the
process. The intricacy of the process and the upcoming elections should not be allowed to
interrupt the ongoing privatization because the coffee sector cannot afford further confusion.
3. Launching the second round of privatizing washing stations and dry millsWith about 10
percent of privatized assets, managing the sector could be more complicated than is anticipated.
The government would have to launch a second round of bidding to enable new owners to be
operational during the 2011/12 crop season. The best period to launch such a tender of sale would
be eight months prior to the beginning of the crop season.
Improving Production Practices

3.90 In the short term, production practices can be improved by the following efforts:
1. Developing and strengthening producer organizationsStrengthening producer organizations
should accompany the privatization process. Existing organizations are believed to represent
merely 20 percent of Burundis coffee growers today. Because the sector relies on smallholders,
enhancing quality and increasing output greatly depend on the development of producer
organizations. Such associations are also an indispensable means through which growers can
participate in the industry, as shareholders of the washing stations or as promoters of their own
coffee.
2. Alleviating growers financial constraints to improve their access to inputs and replace aging
treesTwo of the industrys immediate priorities should be to increase output and improve
quality, starting at the production level. Some major obstacles to raising yields and minimizing
cyclical variations are growers low use of inputs, poor agronomic practices, and their inability to
replace aging trees. Promoting microcredit and matching grant programs would be instrumental
in helping farmers finance the purchase of inputs or the replacement of trees. Such programs
should be set up within a year, and should be designed to foster producer organizations.
Medium-Term Actions
3.91 The regulatory framework and proper market structure should be complemented by other
measures taken within the next two to three years to promote the industrys development. Improvements
at the processing stage should also be introduced within that time frame.
76
Fostering an Enabling Environment
3.92 In the medium term, the following actions can promote an enabling environment in the coffee
sector:
1. Improving financial servicesAs is the case for the rest of the economy, the scope of financial
services offered to the coffee industry is relatively limited. Processing factories, for instance, have
little access to working capital because the banking system in Burundi has little experience (and
interest) in offering such services to small businesses. Incentives should be designed to encourage
banks to develop such products, and technical assistance should be provided to key institutions to
disseminate good practices. In addition to financing issues, the industry suffers from a lack of
financial advisory services. Although these services are virtually nonexistent in Burundi, they
would substantially benefit the coffee sector, notably by helping it optimize revenues from the
sale of green coffee on commodity markets. Building such capacity should be a priority in the
upcoming years.
2. Strengthening the capacity of the interprofessional association (INTERCAFE)The coffee sector
needs an entity devoted to (1) representing it in exchanges with the government and the
regulatory agency, (2) coordinating actions aimed at improving performances along the value
chain, (3) promoting Burundian coffee on global markets, and (4) fostering market intelligence.
An interprofessional association is instrumental for an industry intent on accessing specialty
markets. Financed by a small tax on coffee exports and/or membership fees, for instance, it can
represent all stakeholders in the sector (producers, washing stations, dry mills, roasters, and
exporters) and enable them to voice their concerns and needs. It can be a valuable partner for the
state and the regulatory agency, especially in the phase immediately following privatization. It
can also be a forum through which the industry could coordinate efforts aimed at improving
quality. The government recently has supported the creation of an interprofessional association
whose capacity needs to be reinforced to contribute to the management of service delivery.
Improving Processing Practices

3.93 To improve processing practices in the medium term, the following actions are recommended:
1. Introducing traceability processesImproving washing station and dry mill practices will go
along with the privatization process, which offers stakeholders more visibility. The first priority
in the next two to three years should be to introduce mechanisms to ensure proper traceability
throughout the value chain. It is indispensable to separate high- and lower-quality coffees to
channel produce to the right markets and obtain premiums for quality. In addition, traceability
procedures will help improve incentives along the value chain, enabling the sector to reward
better quality.
2. Initiating coffee differentiationWhen well-functioning traceability mechanisms are in place, a
priority will be to promote coffee differentiation to effectively penetrate specialty markets.
Accomplishing this will require investing in cupping laboratories, initially in regions most likely
to produce specialty coffee; and then, if appropriate, investing in relevant washing stations.
Developing appropriate know-how and expertise, both at cupping facilities and washing stations,
will be crucial and will rely on the sectors ability to bring in expertise from other coffee-
producing countries.
3. Promoting collaboration along the value chain to enhance qualityThe coffee sectors
development is currently impeded by a lack of coordination along the value chain to enhance
quality. This is primarily the result of to ill-designed incentiveswhich deregulation and
privatization are expected to improve. But programs intervening at each key stage of the value
chain could prove invaluable to support Burundi in catching up with its competitors. Such
77
programs are useful in disseminating best practices and can demonstrate to potential investors
Burundis capacity to produce high-quality coffee. These interventions could benefit from a
scaling-up and could be complemented by financial support to replace outdated equipment and
invest in new assets.
D3. Diversification of the Economy

3.94 Burundi currently has a very narrow export base, but has the potential to notably diversify
its economy. A successful diversification of its economy is essential over the long run to be able to better
withstand external shocks, generate income, and create employment (especially off-farm employment).
Research has shown that there is a link between increased diversification and higher GDP per capita,
where diversification usually leads to higher income levels.
3.95 To realize this transformation, a number of steps must be taken by the government and the
private sector. Here are some general suggestions:
Focus on existing products. Burundi should focus on existing products and those well within
reach to start the diversification process. The selection of products and sectors needs to be based
on detailed information and thorough analysis. Possible products are those that have been
identified as emerging champions, as classics with a currently low export share, and as marginals
that have the potential to develop a revealed comparative advantage. Existing strategies, such as
the National Agricultural Strategy (Republic of Burundi 2008b), should be used to guide the
diversification process.
Diversify gradually. Diversifying should be a gradual, sequenced process that starts with the
expansion of the current market share, a diversification of unsophisticated primary products, and
a subsequent move to more sophisticated and eventually manufactured products. Sectors that
have the greatest potential are mining, agro-processing, and fishing.
Benefit from regional integration. The ongoing regional integration process should be used to
maximize Burundis benefits from its accession to the EAC. Reduced tariffs and NTBs will
enable Burundi to have easier access to a larger market, facilitating an increase in exports to the
regional market. This market also can be used to introduce new products and build expertise and
skills before attempting to access international markets. To a large extent, Burundi can also
benefit from regional infrastructure projects that could improve transport and logistics
significantly.
Close the infrastructure gap. The large infrastructure gap needs to be closed. This applies to a
range of areas, including roads, air transport, and electricity. Regarding transport infrastructure,
the insufficient availability of cold storage and a cold chain, in general, needs to be addressed. (A
detailed infrastructure action plan is discussed in section A).
Improve the business environment. The business environment must be improved to attract foreign
investors as well as to provide incentives for domestic investors and enterprises. The business
environment in Burundi is one of the most unfavorable in the world. Reforms need to be
implemented without delay (as discussed in section C).
Comply with sanitary and phytosanitary standards. Burundi needs to develop an action plan and
establish the necessary facilities/laboratories to comply with international sanitary and
phytosanitary standards. Without compliance with these standards, it will be extremely difficult to
export outside the region. Burundi could benefit from the experience of other EAC members in
this area.
3.96 Below is a discussion of specific sectoral recommendations.
78
3.97 Necessary interventions for the fishing sector include the following:
developing a comprehensive strategy for the sector, involving all stakeholders;
improving the capacity to manage the natural resourcesimplying, for example, the
implementation of annual fishing quotas to avoid overfishing and to guarantee a responsible,
profitable, and optimal exploitation of the existing fishing resources without conflict between the
main economic actors;
setting up a regional coordination mechanism to ensure proper implementation of programs and
projects in the sector;
promoting the participatory management of fish and developing comprehensive
legislation/regulation of fishing and fish farming;
promoting commercial fish farming;
improving access to finance; and
complying with sanitary standards.
3.98 A number of interventions can be suggested to increase the performance of the horticulture
sector, including the following61:
Overcome logistical gridlock. Burundi has only limited and inflexible cargo capacity available.
Possible solutions are chartering planes and negotiating with airlines to make additional cargo
space available. The infrastructure action plan (presented in section A), also addresses the
expansion and modernization of the international airport in Bujumbura, with the objective of
obtaining the International Civil Aviation Organizations Certificate of Aerodromes (which
could attract major airlines and freight companies).
Identify potential markets. There is a trade-off between accessing developed markets and
emerging and regional markets. (A summary of opportunities and constraints is presented in table
A.3 in appendix 2.) The ongoing regional integration process has to be taken into account as well.
As pointed out in the section on Burundis benefits risks of membership in the EAC, an analysis
of the extensive margin of trade shows that more new products are exported to the EAC than to
the rest of the world. Products therefore can be tested regionally before they are introduced to the
global market.
Identify products with growth potential. A number of products (listed above) can be identified as
having substantial growth potential.
Develop an action plan. The proposed action plan consists of the following components:
development of a comprehensive strategy for the horticulture sector, involving all
stakeholders;
implementation of a competitiveness assessment of a range of products based on
profitability thresholds in promising markets;
increased production through technical assistance, strengthening sanitary and phytosanitary
management capacity, development of management capacities of producer cooperatives and
export enterprises, and establishment of a small investment fund to address needs for seed
capital; and
development of efficient supply chains, including a substantial improvement of postharvest
logistics and transport infrastructure (cold storage facility, collection/packing facilities that
61
These recommendations are mainly based on Baghdadli, Harborne, and Rajadel (2008).
79
comply with sanitary and phytosanitary standards, and so forth), increased availability of
agricultural inputs and packaging material, and the creation of a favorable business climate.
D4. Development of the Mining Sector

3.99 The geology of the country is overwhelmingly the most important factor for all investors,
but the government has the potential to make the sector more attractive. Surveys of investors
decision parameters indicate that the key determinants for engaging in mining development rank in the
following order: (1) good geological prospects, mining tradition, and potential; (2) clear and secure
mining rights and title (mining legislation); (3) attractive and competitive fiscal conditions (tax
legislation); (4) ownership and control of operations (mining legislation); (5) political stability and
transparency of governance (government institutions); and (6) availability of infrastructure.
3.100 Policy measures include improving the following:
1. Availability of geological informationThis could mean undertaking new geological mapping,
with the support of international organizations such as the World Bank. Burundi actually
benefited in the late1960s from the United Nations Development Programmes assistance in
achieving an inventory of mineral resources with an evaluation of the mining potential. As a
result, most of the deposits, including nickel, were identified. This type of undertaking might
need to be updated through newly available technologies. If enough information exists, access to
current and pertinent information can be improved by setting up modern geologic data
infrastructure.
2. Mining legislationIn 2008, the government launched a reform of the mining code to be
completed in 2009. A key aspect of this reform is ensuring the continuity of tenure, because an
investor will not invest in the necessary capital-intensive exploration phase without certainty that
he or she has the right to exploit the reserves that are found.
3. Mining taxationRelevant questions have been raised worldwide in recent years concerning the
fairness of profit and benefit sharing between the investor and the government. Interestingly
enough, tax considerations only come in third position, which tends to show that the government
should not be tempted to lower its tax rate too much; rather, it should make sure the level of
taxation is fair, predictable, and sustainable.
4. Government participationThe suggestion to have a 15 percent free share for the government
should be explored cautiously. Governments involvement in projects through equity
participation may not result in significant dividend flows under normal circumstances (in the case
of gold in our projection). Such dividends are sometimes more reliably collected through
appropriate royalty and tax regimes, which are consistently applied every year (as opposed to
dividend payments that may be forgone during years of expansion or other discrete corporate
decisions). The merit of dividends is further questioned by the additional risk that the government
must accept by being a shareholder, because the government not only shares the profits of the
good years but also shares the potential losses during difficult timesthus putting scarce
government funds at unnecessary risk.
5. Governance and transparencyWith good governance, the exploitation of mineral resources can
generate large revenues to foster growth and reduce poverty. However, when governance is weak,
it may result in poverty, corruption, and conflict. The Extractive Industries Transparency
Initiative (EITI) supports improved governance in resource-rich countries through the verification
and full publication of company payments and government revenues from oil, gas, and mining.
Such an initiative is now seen as a positive signal by the industry, which has to commit to the
same transparency standards.
80
6. InfrastructureGovernment has three options for the terms on which it could offer to provide
infrastructure for prospective mining projects: (1) no provision, (2) provision in return for an
equity share equal to the value of the infrastructure, and (3) provision in return for user charges.
Circumstances will dictate the choice, but it is desirable that government policy be spelled out
clearly for investors.
3.101 Although the mining sector can make a positive contribution to the economy when it is
properly managed, there also are many countries that do not succeed in turning their mineral
wealth into development. This resource curse that sometimes prevents mineral-rich countries from
reducing their poverty should call for a particular effort in building capacitiesnot only in geology and
mining; but also in mining legislation, taxation, accountability, negotiation, and inspection. This will be
even more relevant in a country like Burundi, where the mining tradition is weak.
3.102 Assistance may be needed in negotiating with private investors. If private investor interest in
the Musongati nickel mining project is to move forward, the government will need to engage in
negotiations with one or more potential investors the terms and conditions for the investment. The
government will need to assemble an experienced legal and technical team for these negotiations, and it
may need advice and assistance from a donor in this regard. Assistance for negotiations is available, for
example, through the recently created Extractive Industries Technical Advisory Facility, funded by the
World Bank.
3.103 Decisions on exploitation of a depletable natural resource and use of the proceeds from
exploitation must take future income into account. International mineral commodity prices, especially
for nickel, have been subject to wide fluctuations during the past 10 or more years. Many countries have
experienced major economic management problems as a result of both precipitous declines and rapid
increases in prices. Proper stabilization policy is needed.
3.104 The World Bank has proposed a new, integrated approach that can help the government
ensure the exploitation of nonrenewable mineral resources is turned into sustainable development.
Taking stock of the EITI,62 and tentatively calling it EITI++ because it goes one step farther upstream
(where do mining revenues come from?) and one step farther downstream (what are mining revenues
used for?), this approach tends to promote sound management of extractive industries all along the value
chain, transforming mineral wealth into poverty reduction and sustainable development.63
3.105 Several measures need to be taken to improve the contribution of the artisanal and small-
scale mining sector, not only to the government budget and balance of payments, but also to the
livelihoods of a large part of the population. The institutional, legal, and regulatory frameworks need to
be revised to formalize artisanal and small-scale mining in Burundi. (A project supporting the
formalization of the sector should be developed, involving all stakeholders. Detailed recommendations
are given in table A.4 in appendix 2.)
E. SUMMARY OF PRIORITY RECOMMENDATIONS

3.106 Based on the analysis carried out in the previous chapters, the following priority
recommendations have been identified for the short and medium terms (see table A.1 in appendix 1
for a detailed action plan indicating the time frames, responsible agencies, and approximate costs):
1. Peace and stability must be maintained. Without a stable political and economic situation, a
successful economic recovery is not possible. Peace and stability enable national reconciliation,
continued democracy, and more effective security expenditure.
62
See details at http://www.eitransparency.org.
63
The EITI++ approach could help the government prepare for mining and for petroleum exploitation.
81
2. Following through on the improvement of infrastructure is critical for development. As
detailed in the infrastructure action plan, different scenarios are possible for its implementation.
Given the importance of improved infrastructure (especially electricity) for the development of all
sectors in the economy, the government should be guided by this action plan and follow through
on its implementation. Specific recommendations include the following:
Electricity: The electricity supply must be improved, more provinces connected to the
network, and REGIDESO financially restructured.
Transport: Given the geographic isolation of Burundi, the country could greatly benefit
from rehabilitating the national highways and the provincial and community roads.
Strengthening of the role of civil aviation could attract international airlines and increase the
availability of cargo space.
3. Burundi should maximize its benefits from the ongoing regional integration process. To do
so, a comprehensive approach that prioritizes and sequences actions needs to be implemented.
Most important, it should address institutional constraints and economic reforms. Key
recommendations are these:
Institutional reforms: It is essential that the Ministry of East African Community Affairs
(MEACA) is to be staffed with competent personnel, and the necessary institutional
arrangements must be in place, requiring extensive capacity building. It is important to
implement a coordination, monitoring, and evaluation mechanism for all EAC-related
issues. Furthermore, a communication campaign aimed at sensitizing the population to
support the integration initiative should be implemented.
Economic reforms: Eliminating all remaining NTBs is critical. Furthermore, a strong
framework for public-private dialogue is advisable.
4. Economic growth has to be led more by the private sector. A recently conducted Financial
Sector Assessment Program, an Investment Climate Assessment (World Bank 2008a), and the
regularly compiled Doing Business Indicators show that Burundis business climate is not
conducive for engaging domestic and foreign investors. Here are the priority recommendations:
Access to finance in the rural areas needs to be improved. Therefore, the regulation should
be changed to broaden the scope of microfinance institutions to enable them to develop new
products. Similarly, technical assistance should be provided to banks to develop new
products and lending mechanisms to better cater to the needs of small and medium-size
enterprises and rural areas. The capacity of BRB to supervise microfinance institutions
needs to be strengthened.
The regulatory environment should be improved to attract domestic and foreign investment.
Critical components are the finalization of the tax code revision, and the finalization and
adoption of the revised mining code.
Increase competition and improve governance. Competition could be increased by
finalizing and adopting the revised Privatization Law. Adopting and implementing the
national policy on good governance and its action plan would demonstrate the recognized
importance of improved governance.
5. Burundi needs to improve agriculture productivity. As shown by Baghdadli, Harborne, and
Rajadel (2008); the National Agricultural Strategy (Republic of Burundi 2008b); and an analysis
of the coffee sector, low agricultural productivity of food and cash crops is one of the main
obstacles to growth. As a consequence, Burundi still relies on regular food aid, has limited
external resources (other than donor funds), and is vulnerable to external shocks. Key
recommendations include the following:
82
Food crops: Critical for increasing production and productivity is strengthening input
distribution systems; rehabilitating production tools and infrastructure (including rural
roads); and organizing, structuring, and professionalizing producers.
Coffee: Most important is the complete implementation of the ongoing privatization agenda.
Next steps include strengthening the regulatory agency (ARFIC), and interprofessional
association (INTERCAFE), facilitating access to finance, and improving processing
practices.
6. Burundi could benefit from diversification of its economy. A product space analysis shows
that Burundis products currently lie on the periphery of the product space, complicating the
move to new products. Nonetheless, the analysis identifies sectors and products that have great
export potential and are in line with Burundis Poverty Reduction Strategy Paper and its National
Agricultural Strategy. Key interventions to facilitate diversification and the increased contribution
of other sectors, such as horticulture and fishing, are availability of cargo space to export fresh
produce and compliance with sanitary and phytosanitary standards.
7. The mining sector could play an important role in Burundis economy. Given the potential
contribution of the mining sector, a sectoral strategy accounting for the characteristics and
constraints of the available mineral resources is essential. Burundi could greatly benefit from
joining the EITI and using available facilities for assistance in negotiating with international
companies. To account for the large share of artisanal and small-scale mining, a unit could be
created within the Ministry of Energy and Mines to deal specifically with this subsector.
Developing a project to support the formalization of artisanal and small-scale mining is advisable.

83
C HAPTER 4 I MPLICATIONS FOR THE MACRO - FISCAL
SITUATION
4.1 The government has an important role in the recovery of the country. Given the weakness of
the private sector and the large needs of a post-conflict country in stabilizing and reconstructing the
economy, the government plays an essential role in promoting growth and reducing poverty. The
government, however, has only limited domestic funds available and relies heavily on foreign assistance
(primarily in the form of grants) financing most of the budget deficit and current account deficit.
4.2 To maintain macroeconomic stability and promote growth, it is critical to improve the
countrys fiscal stance. Reforms, however, are complicated by the challenges of rebuilding the countrys
social and physical infrastructure after years of conflict. The importance of fiscal discipline is well
illustrated in the 2008 Public Expenditure Management and Financial Accountability Review (PEMFAR)
(World Bank 2008e). Although some savings may come from reductions in defense spending, they will
not be sufficient because expenditures for public security (police) are expected to increase. Consequently,
external financing (preferably grants and highly concessional loans) will be needed to finance the fiscal
deficit, at least in the short and medium terms. However, given the volatility of aid to Burundi over recent
years and the unpredictable nature of external capital inflows, the authorities should find ways to rely less
on foreign financing over the long term, particularly by (1) improving the tax administration and (2)
strengthening the effective use of public resources for development goals. To do so, it is necessary to
establish a sound public expenditure management system aimed at promoting fiscal discipline and the
efficient and effective operational performance of the public sector.
4.3 High levels of defense and public security spending are the main obstacle to reallocating
public resources toward priority economic and social sectors. Therefore, to the extent possible and
keeping in mind the need for security and stability, it is essential to reduce the share of defense and
security expenditures over time to create fiscal space for increased spending in the priority economic and
social sectors.
4.4 Although the authorities have made significant efforts to improve the countrys fiscal
stance, increased spending and inadequate revenue mobilization could create larger, unsustainable
fiscal deficits. There is a real risk that a larger deficit may lead to increased borrowing from the domestic
financial market. This problem could raise the cost of funds, potentially crowding out private investment
and dampening growth. Therefore, effective revenue mobilization and efficient expenditure management
are indispensable to maintain growth-supportive macroeconomic policies.
4.5 The allocation and coordination of external aid needs to be examined. In the past, because of
the countrys fragile nature, aid was allocated primarily to humanitarian and emergency projects; whereas
production sectors, such as agriculture and infrastructure, received limited funds. Unpredictable aid flows
have complicated the allocation of aid. When resources previously allocated to humanitarian causes are
freed up by stabilization of the political situation, how those resources are allocated must be carefully
decided.
4.6 This chapter assesses the macroeconomic prospects for Burundi in the medium term and
discusses the medium-term outlook for public finances, considering two different scenarios. The chapter
also evaluates the impact of resource allocation on growth and on efforts to achieve the Millennium
Development Goals (MDGs), and addresses the question of how aid can best contribute to the
development of the economy.
84
A. ECONOMIC PROSPECTS, RESOURCE ALLOCATION, AND ACHIEVING THE
MDGS
4.7 This section presents macroeconomic projections and an assessment of the projected public
finances. The results are presented according to two different scenarios, a baseline scenario and a reform
scenario that assumes the priority recommendations identified in previous chapters are carried out.
A1. Medium-Term Macroeconomic Outlook

4.8 The proposed baseline medium-term macroeconomic framework (201015) reconciles the
macroeconomic framework of the International Monetary Funds Poverty Reduction and Growth
Facility program with the Poverty Reduction Strategy Paper (PRSP) targets of the authorities. It is
based on recent economic trends and assumptions about future developments (anticipating continued
progress of the reform process).64 The baseline scenario reflects the more voluntarist policies of the
authorities, spelled out in other sources of information for assessing the medium-term growth profile.
These sources include sector strategy papers in priority areas, such as education, health, rural sector,
energy and public works, the results of a study on the sources of rural growth undertaken jointly by the
World Bank and the government (Baghdadli, Harborne, and Rajadel 2008), and the recent performance of
Burundis economy presented in this report (chapter 1).
Table4.1:BaselineScenarioProjectionsfortheMainEconomicVariables,200915
Variable Actual Est. Projected Average
2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 201015
Annualpercentagechange
RealGDPgrowth 5.1 3.6 4.5 3.5 4.0 4.4 4.9 5.5 5.9 6.1 5.1
Consumerprices(periodaverage) 2.7 8.3 24.5 8.8 6.9 8.7 8.7 9.4 9.7 9.9 8.9
Exportsofgoodsandnonfactor 67.8 27.2 34.6 15.9 20.9 10.1 8.8 9.4 9.7 9.8 11.5
services(US$)
Importsofgoodsandnonfactor 59.1 3.4 26.2 8.4 5.0 5.8 6.7 8.0 9.0 9.6 7.4
services(US$)
PercentofGDP
Currentaccountdeficit(incl.grants) 14.5 15.7 12.3 13.0 13.6 16.3 14.0 13.0 13.6 12.6 13.8
Currentaccountdeficit(excl.grants) 36.3 37.4 32.8 28.4 28.8 27.7 26.2 25.1 23.7 22.8 25.7
Revenue(excludinggrants) 18.9 18.6 18.5 19.0 19.1 19.2 19.4 19.5 19.7 19.9 19.5
Totalexpenditureandnetlending 38.2 38.5 44.1 50.4 50.8 45.6 44.9 43.8 40.5 39.3 44.1
a
Overallbalance(excl.grants) 19.3 19.8 25.6 31.4 31.7 26.3 25.5 24.3 20.8 19.4 24.7
b
Overallbalance(incl.grants) 1.8 0.5 0.8 58.5 8.4 6.9 5.2 4.6 2.4 2.1 4.9
c
Grossinternationalreserves 3.3 3.8 6.3 7.1 6.0 6.0 6.0 6.0 5.0 4.0 5.5
Sources:InternationalMonetaryFund,WorldBank,andBurundiauthorities.
a.Commitmentbasis.b.IncludesgrantsreceivedundertheHIPCInitiative.c.inmonthsofimports.
4.9 The development of the main economic indicators is given in table 4.1. This outlook assumes
that
after some softening in 2009, the economic growth rate will get stronger;
inflation will remain below 10 percent, owing to lower levels of petroleum prices, improved
liquidity management, and limited financing of government activities by the central bank;
64
Some of the medium-term growth projections are obtained through policy simulations performed with the Burundi
Macroeconomic Model developed for the preparation of the medium-term expenditure framework. The growth path in this model
is determined by macroeconomic and sector policies that affect long-term productive capacity (human capital, infrastructure, and
direct investment flows); short- and medium-term fluctuations in aggregate demand (monetary and fiscal policies); and the
external environment (oil price and terms of trade). The above endogenous growth model was not estimated econometrically for
Burundi. Rather, elasticities estimated through cross-country regressions were used in the model. Other components of the
macroeconomic framework are determined by accounting equations and standard financial programming principles. The
medium-term budget framework allocates the envelope of resources into sector budgets based on government priorities dictated
by exogenous codes.
85
fiscal policy will remain prudent, with stable government revenue and gradually declining public
expenditures;
mainly as a result of the weakening of the world demand following the current global economic
downturn, export sector performance will deteriorate slightly in 2009 and improve in 2010
(whereas import growth will slow down in 2009);
the current account deficit remains fairly constant in the following years; and
external financing will continue to be in the form of grants and highly concessional loans.
4.10 An alternative scenario evaluates the impact of accelerated reforms and increased
infrastructure spending on growth, poverty, and attainment of the MDGs. As the experiences of
other countries have demonstrated (for example, Rwanda and Sierra Leone), especially in a post-conflict
context, Burundi could achieve even higher growth if the commitment of the government to resolutely
accelerate the execution of its overall reform program remained strong. In this alternative scenario, the
authorities are expected to continue and accelerate reforms to improve transparency and governance in the
management of public resources and to create a business environment that is adequate for the
development of the private sector. In addition, the government is expected to carry out the infrastructure
investments proposed under the priority recommendations in the previous chapters. These reforms are
assumed to yield the following results (table 4.2):
There will be a substantially higher growth rate than in the baseline scenario, averaging 8.8
percent for the period 201015.
There will be a similar development of inflation, compared with the baseline scenario.
Greater import volume and value triggered by the large infrastructure projects will be countered
by greater export volume and value resulting from the better business environment, a more
diversified economy, and improved infrastructure. Consequently, the current account balance is
projected to be lower, on average, than in the baseline scenario.
There is prudent fiscal policy with revenue projections similar to those in the baseline scenario,
but higher spending because of increased capital expenditure.
Table4.2:AlternativeScenarioProjectionsfortheMainEconomicVariables,200915
Variable Actual Est. Projected Average
2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 201015
Annualpercentagechange
RealGDPgrowth 5.1 3.6 4.5 3.5 4.0 4.2 8.4 11.1 12.5 12.9 8.8
Consumerprices(periodaverage) 2.7 8.3 24.5 8.8 6.9 8.7 9.0 9.6 9.9 9.9 9.0
Exportsofgoodsandnonfactor 67.8 27.2 34.6 15.9 20.9 9.8 13.7 17.3 19.1 19.5 16.7
services(US$)
Importsofgoodsandnonfactor 59.1 3.4 26.2 8.4 5.0 5.6 9.3 12.2 14.1 14.9 10.2
services(US$)
PercentofGDP
Currentaccountdeficit(incl.grants) 14.5 15.7 12.3 13.0 14.1 15.2 11.0 9.7 11.4 12.2 12.3
Currentaccountdeficit(excl.grants) 36.3 37.4 32.8 28.4 28.8 27.7 25.8 24.2 22.4 21.2 25.0
Revenue(excludinggrants) 18.9 18.6 18.5 19.0 19.1 19.2 19.4 19.6 19.8 20.1 19.5
Totalexpenditureandnetlending 38.2 38.5 44.1 50.4 50.2 46.7 47.1 45.3 40.1 36.6 44.3
a
Overallbalance(excl.grants) 19.3 19.8 25.6 31.4 31.1 27.4 27.7 25.7 20.2 16.5 24.8
b
Overallbalance(incl.grants) 1.8 0.5 0.8 58.5 8.4 6.9 5.0 4.2 2.1 1.7 4.7
Sources:InternationalMonetaryFund,WorldBank,andBurundiauthorities.
a.Commitmentbasis.b.IncludesgrantsreceivedundertheHIPCInitiative.
4.11 A deterioration in the projections of either scenario is possible, as the effects of the global
economic slowdown deepen. First, the overall fiscal deficit (including grants) may be exacerbated if
donor countries cut back on international aid in the face of global economic instability. Finding the
86
resources to finance this budget deficit would be difficult, although Burundi may be able to use savings
that it expects to accrue from reduced debt servicing following the recent debt relief. Second, a potential
stronger decline in migrant remittances could further affect household incomes and the external current
account. In the past few years, the volume of remittances (which are transferred through informal
channels) has increased significantly, although it is difficult to obtain actual figures. Third, a reduction in
projected foreign direct investment is possible, notably in the coffee sector (for example, limited response
of foreign investors to the bidding invitations for the sale of coffee washing stations).
A2. Medium-Term Outlook for Public Finance

4.12 Under the baseline scenario, shortfalls in government revenues, resulting from lower-than-
expected GDP and trade, and increases in government expenditures are likely to result in reduced
fiscal space for economic stimulus and social protection. As a result, the budget deficit (before grants),
estimated at an average of 17.5 percent of GDP in 200208, is expected to widen to about 31.4 percent of
GDP in 2009 (table 4.3). This is partly because of a slight decrease in domestic revenue as a result of the
global economic crisis coupled with an increase in total expenditure arising partly from the need to
implement the recent peace accord with the last rebel group. Because of increased expenditures in light of
the elections and the payment of past government commitments to civil servants, this deficit is projected
to further increase to 31.7 percent of GDP in 2010. In 2012, it is projected to decline to 25.5 percent, and
to decline further to 19.4 percent in 2015 as revenue increases (partly as a result of improved tax
collection performance expected from the newly created Burundian Revenue Office). The ability of the
government to reduce military and security expenditures following the anticipated progress of the peace
process also explains fiscal position improvement. Furthermore, recurrent expenditures will decrease
because of lower interest payments following the January 2009 Heavily Indebted Poor Countries
Initiative completion point. The debt stock is expected to fall substantially for the same reason. Wage and
salary control will also contribute to expenditure discipline.
Table4.3:BaselineScenarioActual,Estimated,andProjectedGovernmentFinance,200215(PercentofGDP)
Actual Estimated Projected Average
200208 2009 2010 2011 2012 2013 2014 2015 201015
Indicator (average)
Domesticrevenue 19.6 19.0 19.1 19.2 19.4 19.5 19.7 19.9 19.5
Taxrevenue 17.8 17.1 17.1 17.2 17.3 17.4 17.5 17.7 17.4
Nontaxrevenue 1.9 1.9 2.0 2.0 2.1 2.1 2.2 2.2 2.1
Totalexpenditureandnetlending 36.9 50.4 50.8 45.6 44.9 43.8 40.5 39.3 44.1
Recurrentexpenditure 23.1 25.1 25.5 25.3 25.2 24.5 23.6 22.4 24.4
Salaries 9.3 12.4 12.2 12.0 11.8 11.5 11.2 10.8 11.6
Goodsandservices 7.0 5.5 6.4 6.6 7.1 7.1 7.0 6.6 6.8
Transfersandsubsidies 3.7 5.7 6.0 5.9 5.6 5.3 5.0 4.6 5.4
Interestpayments 3.1 1.5 0.9 0.8 0.6 0.5 0.4 0.3 0.6
Otherexpenditure(incl.exceptional) 1.8 10.0 8.3 3.3 2.9 2.5 0.7 1.4 3.2
Capitalexpenditure 12.2 15.4 17.0 17.0 16.8 16.8 16.2 15.5 16.6
Domesticallyfinanced 3.2 3.6 3.5 3.4 4.6 5.5 6.3 6.9 5.0
Foreignfinanced 9.0 11.8 13.5 13.6 12.2 11.3 10.0 8.6 11.5
Netlending 0.2 0.1 0.1 0.0 0.0 0.0 0.0 0.0 0.0
Primarybalance 3.4 12.5 10.1 8.6 9.8 10.0 9.8 9.0 9.6
Includinggrants 3.1 58.8 8.4 6.9 5.2 4.6 2.4 2.1 4.9
Excludinggrants 17.2 31.4 31.7 26.3 25.5 24.3 20.8 19.4 24.7
Sources:InternationalMonetaryFundandWorldBankstaff.
4.13 Although financing requirements appear moderate over the entire period (201015) in the
baseline scenario, Burundi will remain dependent on external assistance, particularly in the form of
budget support, to meet projected external financing requirements. Provided Burundi stays on track
with its reform agenda, most of the financing requirements will be covered by budget support; savings
recorded from Heavily Indebted Poor Countries (HIPC) Initiative debt relief, multilateral debt relief
87
initiatives, bilateral donors of the Paris Club; and additional debt forgiveness by bilateral donors. Aid
pledges (mostly grants) are projected to be at least roughly 25 percent of GDP each year, unless some
bilateral donors stop providing budget support. Overall, budget support (including special programs)
accounts for about 5060 percent of total assistance to Burundi.
4.14 The high reliance on external assistance is expected to gradually decline because foreign
grants and loans are projected to decrease slightly until 2015. The gradual decline of the high reliance
on external assistance could be explained by the projected stability of domestic revenue as well as the
commitment of the authorities to rationalize public expenditures and thus create more fiscal space on the
basis of domestic resources.
4.15 Capital spending is expected to increase modestly under the baseline scenario. Total
expenditure (excluding debt service) is expected to increase initially from 36.9 percent of GDP between
2002 and 2008 to 50.8 percent of GDP in 2009 and to 47.1 percent of GDP a year in 201012; however, it
will decline to 41.2 percent in 201315 (table 4.3). Following a significant increase in 2009 to
accommodate the implementation of the peace accord with the last former rebel group, total expenditures
are expected to rise again in 2010 (mostly because of the need for resources to prepare for elections).
Reduced debt services following Burundi debt relief status in the context of the HIPC Initiative also
explains the increase in total spending. Capital expenditure increases to an average of 16.6 percent of
GDP over the projection period (201015), which is consistent with priorities set in the Poverty
Reduction Strategy Paper. Economic and social sectors (education, health, and productive infrastructure)
are expected to be the main beneficiaries of this increase in capital spending (table 4.4). The budget share
of capital expenditure is projected to increase modestly to 40.1 percent, on average, in 201015 (from an
average of 33.7 percent in 200209). Although still below what is required to achieve the MDG targets,
the projected trend is encouraging (figure 4.1).
Table4.4:BaselineScenarioMediumTermExpenditureFramework,200715
Expenditure 2007 2008 2009 2010 2011 2012 2013 2014 2015 201015
(percentoftotalexpenditure)
Recurrentexpenditure(excl.debtservice) 63.2 60.6 60.4 59.1 59.0 59.3 58.8 58.9 58.8 59.0
Capitalexpenditure 36.8 39.4 39.6 40.9 41.0 40.7 41.2 41.1 41.2 41.0
Sectoralallocation(%oftotalexpenditure,excludingdebtservice)
Socialsector 28.9 28.8 33.7 34.3 35.2 36.1 36.9 37.8 38.6 36.5
Education 23.2 20.3 23.8 24.0 24.4 24.9 25.3 25.7 26.1 25.1
Healthandsocialdevelopment 5.7 8.5 9.9 10.3 10.7 11.2 11.6 12.1 12.5 11.4
Productiveinfrastructure 6.0 15.8 12.8 13.1 13.1 13.1 13.2 13.2 13.2 13.1
Miningandenergy 1.0 2.2 2.2 2.2 2.2 2.2 2.2 2.2 2.2 2.2
Transportandtelecommunications 0.4 0.4 2.9 3.0 3.0 3.0 3.1 3.0 3.1 3.0
Water,environment,andurbanism 0.3 0.5 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6
Publicworksandequipment 4.3 12.7 7.0 7.2 7.2 7.2 7.3 7.3 7.3 7.2
Productivesectors 6.4 5.3 8.8 9.6 10.1 10.5 11.1 11.6 12.1 10.8
Agriculture 5.0 4.1 8.3 9.0 9.5 10.0 10.6 11.1 11.6 10.3
Industryandcommerce 1.4 1.2 0.5 0.6 0.6 0.5 0.6 0.6 0.6 0.6
Othersectors 58.7 50.1 44.7 43.0 41.6 40.3 38.8 37.4 36.0 39.5
Sources:WorldBankstaffandBurundiauthorities.
4.16 This increase is justified by the reallocation of nonproductive expenditure toward

socioeconomic priority expenditures following the consolidation of peace, the reduction of military
expenditure, and continued efforts to rationalize the national budget. The budget share of productive
infrastructure increased markedly in recent years, from 6.0 percent in 2007 to 15.8 percent in 2008; the
increase is mostly explained by growth in spending on public works and equipment (construction)
projects. The share of productive infrastructure is projected to be steady at 13 percent of GDP over the
201015 projection period. The budget share of productive sectors (mostly agriculture), which decreased
between 2007 and 2008, is estimated to increase to 8.8 percent in 2009, reflecting the authorities
priorities for rural development. Furthermore, that priority is reflected in the projected budget share of
88
these sectors, which is expected to rise by 2011 above the critical 10 percent level recommended by the
New Partnership for Africas Development (NEPAD).65
Table4.5:AlternativeScenarioActual,Estimated,andProjectedGovernmentFinance,200215(PercentofGDP)
200208 2009 2010 2011 2012 2013 2014 2015 201015
Indicator (average)
Domesticrevenue 19.6 19.0 19.1 19.2 19.4 19.6 19.8 20.1 19.5
Taxrevenue 17.8 17.1 17.1 17.2 17.3 17.4 17.6 17.8 17.4
Nontaxrevenue 1.9 1.9 2.0 2.0 2.1 2.1 2.2 2.3 2.1
Totalexpenditureandnetlending 36.9 50.4 50.2 46.7 47.1 45.3 40.1 36.6 44.3
Recurrentexpenditure 23.1 25.1 25.5 25.3 24.3 22.5 20.4 18.1 22.7
Salaries 9.3 12.4 12.2 12.0 11.4 10.6 9.7 8.8 10.8
Goodsandservices 7.0 5.5 6.4 6.6 6.9 6.5 6.0 5.4 6.3
Transfersandsubsidies 3.7 5.7 6.0 5.9 5.4 4.9 4.3 3.7 5.1
Interestpayments 3.1 1.5 0.9 0.8 0.6 0.4 0.3 0.2 0.6
Otherexpenditure(incl.exceptional) 1.8 10.0 8.3 3.3 2.8 2.3 0.6 1.2 3.1
Capitalexpenditure 12.2 15.4 16.5 18.1 19.9 20.5 19.1 17.3 18.6
Domesticallyfinanced 3.2 3.6 3.0 4.5 8.2 10.2 10.5 10.3 7.8
Foreignfinanced 9.0 11.8 13.5 13.6 11.8 10.3 8.6 7.0 10.8
Netlending 0.2 0.1 0.1 0.1 0.0 0.0 0.0 0.0 0.0
Primarybalance 3.4 12.5 9.5 9.7 12.5 12.6 10.8 8.1 10.5
Includinggrants 3.1 58.8 8.4 6.9 5.0 4.2 2.1 1.7 4.7
Excludinggrants 17.2 31.4 31.1 27.4 27.7 25.7 20.2 16.5 24.8
Sources:InternationalMonetaryFundandWorldBankstaff.
4.17 Under the alternative scenario, capital expenditure will be substantially higher, but
accelerated reforms will prevent an increase in the overall balance. Increased spending, mainly on
infrastructure, will bring capital expenditure to an average of 18.6 percent of GDP in 201015 (table 4.5),
compared with 16.6 percent of GDP in the baseline scenario. Lower recurrent expenditure resulting from
accelerated reform efforts in public finance management and the civil service, however, will offset most
of the increase in capital spending. The resulting overall balance is therefore projected to be similar to that
under the baseline scenario. The domestic revenue pattern under the baseline scenario is maintained,
because it already assumes substantial efforts from tax collection authorities.
Table4.6:AlternativeScenarioMediumTermExpenditureFramework,200715
Expenditure 2007 2008 2009 2010 2011 2012 2013 2014 2015 201015
(percentoftotalexpenditure)
Recurrentexpenditure(excl.debtservice) 63.2 60.6 60.4 59.9 57.5 54.3 51.8 51.2 50.8 54.3
Capitalexpenditure 36.8 39.4 39.6 40.1 42.5 45.7 48.2 48.8 49.2 45.7
Sectoralallocation(%oftotalexpenditure,excludingdebtservice)
Socialsector 28.9 28.8 33.7 33.1 38.0 37.2 37.2 37.9 38.3 36.9
Education 23.2 20.3 23.8 23.1 26.6 25.8 25.6 25.8 26.0 25.5
Healthandsocialdevelopment 5.7 8.5 9.9 10.0 11.3 11.4 11.6 12.0 12.4 11.5
Productiveinfrastructure 6.0 15.8 12.8 21.9 19.9 22.5 23.7 23.4 23.6 22.5
Miningandenergy 1.0 2.2 2.2 4.8 1.9 1.9 2.0 2.1 2.1 2.5
Transportandtelecommunications 0.4 0.4 2.9 0.8 2.5 2.6 2.8 2.8 2.8 2.4
Water,environment,andurbanism 0.3 0.5 0.6 4.0 0.6 0.6 0.6 0.6 0.6 1.2
Publicworksandequipment 4.3 12.7 7.0 12.3 14.9 17.4 18.3 17.9 18.1 16.5
Productivesectors 6.4 5.3 8.8 5.3 7.0 7.6 8.3 9.0 9.4 7.7
Agriculture 5.0 4.1 8.3 4.9 6.5 7.1 7.8 8.4 8.9 7.3
Industryandcommerce 1.4 1.2 0.5 0.4 0.5 0.5 0.5 0.6 0.5 0.5
Othersectors 58.7 50.1 44.7 39.8 35.1 32.6 30.9 29.7 28.7 32.8
Sources:WorldBankstaffandBurundiauthorities.
65
According to the 2003 Maputo Declaration, the Comprehensive Africa Agricultural Development Program from NEPAD is
based on two major principles: (1) pursuit of a 6 percent average annual growth rate at the national level in the agriculture sector,
and (2) allocation of 10 percent of the national budget to agriculture.
89
4.18 The budget share of productive infrastructure will be significantly higher than in the
baseline scenario, without compromising spending on social sectors. Implementing large
infrastructure projects (as recommended in the previous chapters) will yield an average budget share of
productive infrastructure of 22.5 percent for 201015 (table 4.6), compared with an average of 13.1
percent under the baseline scenario. Spending on social sectors, however, will remain similar, as assumed
under the baseline scenario. The share of other sectors, which comprises nonpriority spending, is expected
to decrease significantly.
A3. Prospects for Poverty Reduction and MDG Attainment

4.19 Given the assumptions of the Figure4.1:SelectedMDGTrendsundertheBaselineScenario,200615
baseline scenario, progress will be
7% 120%
made with on social indicators, but it is
not likely that the MDGs will be 6% 100%
achieved. The macroeconomic 5%
80%
framework used for the projections 4%
60%
(Burundi Macroeconomic Model) was 3%
linked to poverty through growth-poverty 2%
40%
elasticities. It is assumed that GDP per 1% 20%
capita, inequality/variance of per capita
0% 0%
income, social/basic infrastructure, and
2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
net enrollment rate are the key factors
Povertyincidence(RHS) Netprimaryenrollment(RHS)
affecting the incidence of poverty. Using GDPgrowth(LHS)
a growth-poverty elasticity of 1.25, the
Sources:BurundiauthoritiesandWorldBankstaff.
baseline macroeconomic framework Note:LHS=lefthandside;RHS=righthandside.
indicates that the incidence of poverty
would be about 38 percent in 2015down from 67 percent in 2006, but short of the MDG target of 18
percent (see figure 4.1). The same applies for primary-school enrollment and child mortality. Progress is
expected to be made, but not enough to meet the MDGs.
4.20 Under the accelerated reforms scenario that includes increased spending on infrastructure,
Figure4.2:SelectedMDGTrendsundertheAlternativeScenario,200615
growth will be notably higher and
the incidence of poverty
14% 140% substantially lower (see figure 4.2).
12% 120% Economic growth will average 9
10% 100% percent over 201015, and its broader
distribution would reduce the poverty
8% 80%
rate by more than 11 percentage
6% 60%
points (to 27 percent in 2015,
4% 40% compared with 38 percent in the
2% 20% baseline scenario). This is still
0% 0% insufficient, however, to attain the
2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
MDG target by 2015. It is evident that
Povertyincidence(RHS) Netprimaryenrollment(RHS) successfully implementing structural
GDPgrowth(LHS) reforms (such as privatization and
Sources:BurundiauthoritiesandWorldBankstaff. regulatory reforms) that are required
Note:LHS=lefthandside;RHS=righthandside.
to improve the overall business
climate, along with public finance management reforms and the necessary investments in infrastructure,
would result in accelerated growth in human capital, infrastructure, and private investment. Consequently,
there are clear indications that a higher pace of poverty reduction is well within reach and primarily
dependent on the authorities commitment, willingness, and their ability to forcefully deepen reforms and
swiftly implement policies.
90
B. HOW CAN AID CONTRIBUTE TO INCREASED GROWTH?
4.21 Burundi has not experienced the expected post-conflict growth spike seen in other post-conflict
countries (see section A3, chapter 1). This is despite the high levels of aid Burundi has been receiving.
The objective of this section, therefore, is to understand why the increased aid has not yet noticeably
enhanced growth.
B1. Allocation of Increased Aid and Impact on Growth

4.22 The effect of aid on growth has been widely discussed in the literature, without a consensus.
A large body of literature on aid and growth typically has examined the impact of aggregate aid on
growth, often using cross-country regressions, and it has found mixed results. Realistically, aid is
recipient specific and is given for a multitude of reasons, not solely to promote growth. Often aid provides
humanitarian relief following war or natural disasters, or it is given to achieve donors political
objectives, with growth as a secondary goal. Other times, aid is conditioned on the specific social
objectives of the donors, such as promoting education, health, or gender equality. These political or social
objectives may or may not fully match the development goals of recipient countries, thus affecting their
ability to successfully implement their growth and development strategies.
4.23 It is important to identify the objectives of aid in the analysis of its impact on growth.
Clemens, Radelet, and Bhavani (2004) identify three categories of aid and their potential impact on
growth. In the first category, aid is destined for immediate consumption in the aftermath of a disaster or
war to stabilize the country and facilitate peace building, thus exhibiting little or even negative correlation
with growth. There is, however, an important long-term effect on growth because peace and stability have
to be established to enable the recovery of the country. A second category of aid provides resources for
infrastructure and production sectors, which can affect growth in the short term. A third category of aid
focuses resources in sectorssuch as health and educationthat have long-term impact on growth, but
no visible short- or medium-term correlation. Using regression analysis over four years, the authors find a
strong, positive, and causal relationship between short-impact aid and growth that is two to three times
larger than that found in studies using aggregate financial aid.
4.24 The absorptive capacity of aid recipients affects the economic returns to aid and its impact
on growth. McGillivray and Feeny (2008) study the well-established U-shaped relationship between aid
and growth as it relates to fragile countries absorptive capacity. They find that many of the highly fragile
states (from a per capita income perspective) are over-aided because they can efficiently absorb only
approximately one third of the amount of aid that other countries can absorb.
4.25 The absorptive capacity for aid is systematically different in post-conflict countries. Collier
and Hoeffler (2004) find that there is no significant difference in absorptive capacity and no bounce-back
of growth in the immediate aftermath of peacemaking (that is, the first three years). After that initial
period, absorptive capacity approximately doubles. Aid is especially effective and critical for growth after
peace has been fully established, but its effectiveness diminishes about seven years following the end of
the conflict.
4.26 Budgetary practices and the low predictability of aid disbursement further affect recipients
capacity to absorb aid and therefore affect the growth impact of aid.66 Many least-developed
countries continue to finance recurrent expenditures with domestic resources while financing investment
expenditures through foreign aid. The combination of budgeting investment expenditures based on aid
receipts and low predictability of aid affects the composition and effectiveness of spending, and can
further aggravate the bias toward government consumption. In fact, as Celasun and Walliser (2008) show,
governments tend to cut investment spending in times of aid shortfalls and expand government
consumption if additional aid becomes available unexpectedly. This is especially true in low-income
66
The predictability of aid flows is defined as the difference between commitments and actual disbursements in a given year.
91
countries that rely heavily on external assistance to fund their budget and have only limited access to
international financial markets to fund budget shortfalls.
Allocation of Aid to Burundi

4.27 The allocation of aid is disaggregated according to its impact on growth, following the
approach of Clemens, Radelet, and Bhavani (2004), modified to capture Burundi-specific
characteristics.67 The three considered categories are (1) emergency and humanitarian aid, (2) aid with a
short-term impact on growth, and (3) aid with a long-term impact on growth. Budget support and debt
relief are allocated similar to the composition of the budget.
4.28 The pattern of aid allocation
Figure4.3:AllocationofAidbyImpactCategory,200107
between 2001 and 2007 shows that
Grossdisbursement,currentUS$millions
there is not necessarily a
contradiction between increased aid 160
Humanitarianaid
flows and low growth performance 140
Shortrunimpact
if a lagged impact of aid is taken into
account (figure 4.3). The highest share 120 Longrunimpact
US$millions(current)
of aidabout 41 percenthas been 100

allocated to emergency and
humanitarian aid. In Burundi, it 80
consists mainly of emergency/distress 60
relief and emergency food aid. This
type of aid is not directly growth 40
enhancing, and might even coincide
20
with a negative growth rate because it
is disbursed in times of crisis. It is, 0
however, a necessary condition for 2001 2002 2003 2004 2005 2006 2007
establishing peace, and it has a long-
Source:AuthorscalculationbasedonOECDDACdata.
term stabilizing effect. The Note:Datafor2001reflectsonlycommitmentsbecausenoinformation
expenditure pattern observed in ondisbursementsisavailable.
Burundi is typical for a country getting
out of conflict: the humanitarian aid remains high immediately after the end of conflict when a fragile
situation prevails. Emergency and humanitarian aid has been declining since 2005, however, with
improving security in Burundi.
4.29 Aid allocated to sectors that contribute to growth in the long run is about 37 percent of total
aid, and it is the second-largest and increasing category. This is aid allocated fully or partially to
education; health; population programs; water supply and sanitation; multisectoral projects; and to
planning, development, and research in other areas. Spending in these areas is considered to build human
capital and improve service deliveryefforts that are expected to increase the productivity of labor in
Burundi in the long run.
4.30 Aid allocation to short-run growth-enhancing sectors68 has been small. Between 2001 and
2007, only about 24 percent of total aid went into these sectors. If disbursements for peace building are
67
This approach is possible by using data from the Creditor Reporting System of the Development Assistance Committee (DAC)
of the Organisation for Economic Co-operation and Development (OECD), providing a breakdown of aid disbursement by 194
purpose codes. Each purpose code has been assigned a category that relates to its impact on aid. Data are available on
commitments starting in 1995, but the low coverage ratio of the data enables the analysis of disbursement data only from 2002
onward. To ensure consistency with the previous section, however, data for 2001 is included on the commitment basis. This also
enables the analysis of the entire period following the official end of the conflict after the Arusha Accords in 2000.
68
These sectors include transport and storage, energy, the financial sector, agriculture, manufacturing, mining, and peace
building.
92
excluded, however, the allocation to sectors with a short-run impact is reduced to merely US$290 million
or 18 percent of total aid in that period. Considering these results, it is not surprising that the increased
amounts of aid have not led to a boost in economic growth rates.
4.31 The effect of the allocation of external resources is compounded by the limited amount of
the domestically financed budget that has been spent on infrastructure and production-oriented
sectors. This makes a significant positive effect on growth unlikely in the short to medium term.
Spending on priority sectors (education, health, infrastructure, and agriculture) has improved, but it is still
considered low because domestic resources are relatively limited.
Predictability of Aid
4.32 In the case of Burundi, the low predictability of aid likely has aggravated the spending
pattern of the government, exacerbating the bias toward government consumption and away from
investment in production-oriented
sectors and human capital. This has not Figure4.4:DisbursementRatioandCommitmentDisbursementGap,
been conducive to promoting growth in the 200107
medium term, especially because the 120 300
country already spends a very limited Disbursement
ratio(LHS)
amount of its domestically financed budget 100 250
on investment. In Burundi, the average aid
80 200
disbursement ratio (that is, the disbursed
US$millions
percent
amount divided by the committed amount) 60 150

was 68 percent between 2002 and 2007,
40 100
but varied considerably from one year to
Commitment
another. The average difference between 20 disbursement 50
commitments and disbursements was gap(RHS)
US$133 million, peaking at US$286 0 0

million in 2006 when only just over 50 2002 2003 2004 2005 2006 2007
percent of the committed amount was Source:AuthorscalculationbasedonOECDDACdata.
disbursed. Figure 4.4 shows the large Note: LHS=lefthandside;RHS=righthandside.
volatility in the disbursement ratio.
4.33 In the next few years, the issues of low predictability and volatility of aid will be especially
important, in light of the financial crisis. Aid commitments and disbursements may decline as donors
address domestic financial issues. The situation is particularly critical because Burundi is now entering
the true post-conflict reconstruction phase. According to Collier and Hoeffler (2004), this is the period
when aid is especially effective and crucial for growth. According to their findings, this aid effectiveness
diminishes after about seven years following the end of the conflict. It is therefore important that the
amount of aid provided by the donor community remains at the current high level for the foreseeable
future.
B2. Impact of Increased Aid on Competitiveness

4.34 The fact that Burundi has received increased aid flows (which it has mostly spent and
absorbed) raises concerns about currency appreciation and its impact on the competitiveness of the
economy and the potential for some Dutch disease effect.69 In fact, the real effective exchange rate
appreciated slightly between 2005 and 2008, when the aid surge was observed: by 1.5 percent annually or
69
In the case of Dutch disease, prices of non-tradables increase more than those of tradables, leading to an appreciation of the
real effective exchange rate and ultimately to the contraction of the tradables sector.
93
5.2 percent in total.70 The slight deterioration of the terms of trade during the same time period is another
indicator for a possible loss of competitiveness.
Figure4.5:ExchangeRateandPriceDevelopments,200108 4.35 This overall slight loss of
120 30 competitiveness, however, has been
REER(2000=100)
CPI(12monthpercentchange) caused primarily by higher inflation
110 (RHS) 25
(LHS)
compared with that of the trading
100 20 partners, not by appreciation of the
nominal exchange rate.71 The
90 15 nominal effective exchange rate
80 10
depreciated by an average of 5.1
percent
percent a year or 20.1 percent overall
Index
70 5 between 2005 and 2008. The higher

60 0
inflation, therefore, has more than
offset the depreciation of the nominal
50
NEER(2000=100)
5 exchange rate. The development of
40
(LHS)
10
these key indicators is shown in figure
May07
4.5. The high inflation rate has been
Jan01
May01
Sep01
Jan02
May02
Sep02
May03
Jan03
Sep03
Jan04
May04
Sep04
Jan05
May05
Sep05
Jan06
May06
Sep06
Jan07
Sep07
Jan08
May08
Sep08
caused partly by natural events and
developments in the commodity
Source:AuthorscalculationbasedonInternationalMonetaryFunddata. markets that are exogenous for the
Note:CPI=consumerpriceindex;LHS=lefthandside;RHS=righthandside government of Burundi, and cannot be
NEER=nominaleffectiveexchangerate;REER=realeffectiveexchangerate. attributed to the increase in aid flows.
For instance, part of the price
increases in 2005 was caused by a drought affecting domestic agricultural production, whereas the high
inflation in 2008 resulted from the global food and petroleum price crises.
4.36 The relative price development of tradables and nontradables confirms that the slight
appreciation of the real effective exchange rate was not caused by the increase in aid. In the case of
Dutch disease, prices of nontradables increase more than those of tradables, leading to an appreciation of
the real effective exchange rate and, ultimately, to the contraction of the tradables sector. In Burundi,
however, the price level of tradables increased by an average of 13 percent between 2005 and 2008, and
the price level of nontradables increased by only 11 percent, on average.72
4.37 In addition, goods exports, dominated by coffee and tea, do not seem to have suffered
probably because of their dependence on nonprice factors (such as infrastructure and supply
constraints) rather than prices and exchange rates (see also Foster and Killick [2006]). In fact,
Burundis export performance improved from an average of 6.1 percent of GDP during 200004 to 6.4
percent of GDP during 200508, indicating that the appreciation of the real effective exchange rate has
not been detrimental to exports.73 Dutch disease does not seem to be a significant problem in Burundi so
far. This is confirmed by the International Monetary Funds assessment of Burundis exchange rate,
which concluded that the real exchange rate has been broadly in line with economic fundamentals (IMF
2008). Despite the absence of an indication of a negative effect on competitiveness via the real effective
exchange rate to this point, the possibility of a Dutch disease effect remains a risk.
70
For a more detailed discussion of absorption and spending of aid in Burundi, see Nielsen and Madani (2010).
71
The real effective exchange rate is calculated as the product of the nominal effective (that is, trade-weighted) exchange rate and
the inflation differential between Burundi and its trade partners.
72
Average year-on-year inflation is based on monthly data; authors calculation is based on International Monetary Fund data.
73
The share of other, nontraditional exports remained fairly constant.
94
Box 4.1: The Common Performance Assessment Framework in Rwanda
The Common Performance Assessment Framework (CPAF) consists of a set of indicators and policy actions
selected from the Economic Development and Poverty Reduction Strategy (EDPRS) Results and Policy Matrix and
the Joint Governance Assessment. It forms the basis for budget support donors joint assessment of the government
of Rwandas performance in implementing the EDPRS. The CPAF is complemented by the Donor Performance
Assessment Framework (DPAF), which consists of a set of indicators and actions drawn from international and
national agreements relating to aid and its effective delivery, and forms the basis for governments assessment of
donor performance. The assessments would consider all partners performance (including donors giving nonbudget
support) at both the aggregate and individual levels.
The Process for Elaboration of the CPAF
The process of developing the CPAF involved extensive consultations and drew from the work of the existing 15
sector working groups. Each group is chaired by a lead government institution and cochaired by a lead donor. Initial
consultations involved the development of sector policy and results matrixes, with technical support and financing
from the U.K. Department for International Development (DfID), the World Bank, and other donors. Indicators for
the CPAF were agreed on the basis of consultations between government and budget support donors, and were
limited to a list of 18 strategic outcomes and 47 indicators. To emphasize the commitment of all budget support
donors and the government, a Memorandum of Understanding was signed. In that document, all the parties agreed to
use the new reporting mechanisms (CPAF and DPAF) as part of a mutual accountability and harmonization process
aligning with the Paris Declaration.
Using the CPAF to Report on and Monitor EDPRS Achievements
Reporting of performance against individual CPAF indicators is led primarily at the sector level, in advance of the
Joint Budget Support Review (JBSR). Sector performance reports are prepared in time for further discussion at the
JBSR. The dialogue in the JBSR is based on the joint assessment of progress in EDPRS implementation, including
governance indicators and public finance management reform, using the CPAF as a single framework. For
additional governance topics not covered in the CPAF, an assessment is to be reviewed on the basis of the Joint
Governance Assessment annual review.
A forward-looking joint review will be held prior to the Cabinet submission of a draft Finance Law or budget for the
following fiscal year. This review enables signatories to discuss proposed budget allocations, forthcoming activities
in relation to public financial management reforms, and the revision of the contents of the CPAF. This joint review
will be held in March of each budget calendar year.
A backward-looking review of performance will be undertaken jointly within the four months that follow the end of
the government fiscal year in question. During this review, signatories will discuss the progress and challenges
encountered in the implementation of the EDPRS in the previous year, reviewing budget execution and audited
government accounts of the year before and value for money.
Process for Revising the CPAF
Policy actions and targets of the CPAF are revised on an annual basis during the forward-looking JBSR. Under
exceptional circumstances, policy actions and indicators may be changed or adjusted as long as they remain
consistent with the EDPRS and with ongoing government policies and programs. The revised CPAF is agreed by all
signatories and takes the form of an annex to the minutes of the JBSR. Joint reviews provide a forum for dialogue
concerning the performance of both the government and its budget support partners, and draw from the DPAF. It is
undertaken following the same annual cycle as the assessment of progress in implementing the EDPRS based on the
CPAF.
Source: Adapted from World Bank (2009b).
95
C. CONCLUSION
4.38 Although the end of Burundis aid dependency is not foreseeable in the near future,
improved allocation and efficiency of spending can enhance the impact of aid on growth in the
short to medium term. As a result of the conflict in Burundi, security and military spending has been
high. As the security situation in the country stabilizes, those funds can be reallocated to production-
oriented sectors or social spending. An increase in the fiscal stance by improving revenue mobilization as
well as public finance management will also lead to more available resources and more efficient
spending.
4.39 A scenario assuming accelerated reforms and increased spending on infrastructure
demonstrates that higher GDP growth and increased poverty alleviation are within reach. If the
government implements the priority measures identified in the previous chaptersparticularly regarding
the management of public resourcesthe improvement of the business environment to foster private
sector development and contribute to the economy, as well as the development of the necessary
infrastructure, the economic and social prospects can be improved substantially.
4.40 Donor coordination could be strengthened. Donor coordination is important for a number of
reasons. First, coordination is needed to ensure that the disbursement of donor funds is in line with the
governments objectives set out in the PRSP and in Vision 2020. Second, it is needed to take capacity
constraints into account and avoid the overburdening of authorities with different administrative
processes. Third, improved donor coordination can ensure consistency among the programs of different
donors. In that regard, a Common Performance Assessment Framework (CPAF) is advisable to align
donors objectives with those of the government and to agree on a set of measures to be taken. For an
example of how the CPAF has been developed and is used in Rwanda, see box 4.1)
4.41 The small amount of domestically financed investment spending has done little to promote
economic growth in the past years. This bias toward recurrent spending, however, has been aggravated
by the low predictability of aid, which is not conducive to a growth-enhancing composition of public
expenditure. Especially in the current economic crisis, firm commitments and timely disbursements by
donors are very important to ensure the allocation of funds in growth-enhancing areas.

96
APPENDIX
97
APPENDIX1:ACTIONPLAN
TableA.1:ActionPlanPriorityRecommendationsfortheTimePeriod201015
Area Recommendation Timeframe Responsibleagency Approximate Costasshareoftotal
(201015) cost(US$ expenditureprojected
million) for201015
1. Closingtheinfrastructuregap
Electricity DevelopBurundisdomestichydroelectricpotential Shortto MEN/REGIDESO 80 1.3%
(Kaganuzi,Mpanda,Kabu16) mediumterm
Provide110kVlineswithnecessarysubstationstoall Shortto MEN/REGIDESO 60 1.0%
provincialcapitals mediumterm
FinalizethefinancialrestructuringofREGIDESO Shortterm MEN/MF tbd
Transport Rehabilitate,upgrade,andmaintainnationalhighways Shortto MPW/MTR/OdR 600 9.8%
mediumterm
Rehabilitateprovincialandcommunityroadnetworks Shortto MPW/MTR/OdR 13 0.2%
mediumterm
Buildhumanandinstitutionalcapacityingovernment Shortto MTR/OdR 15 0.2%
agenciesthatareresponsibleforregulationand mediumterm
managementofroadtransportactivities
Prepareabusinessandmasterplanforcivilaviation Shortterm MTR 1 <0.1%
2. MaximizingthebenefitsofBurundisEACmembership
Institutional RecruitcompetentstaffinMEACA Shortterm MEACA 0.2 <0.1%
reforms Makeinstitutionalarrangements(includingcapacity Shortto MEACA 0.5 <0.1%
buildingandotherregionalintegrationissueswithinthe mediumterm
newministry)
Implementthecoordination,monitoring,andevaluation Shortterm MEACA 0.3 <0.1%
mechanisms
Implementcommunication/sensitizationcampaign Shortterm MEACA 0.5 <0.1%
Economic EliminateremainingNTBs(includingequipmentat Mediumterm MTI/CD 1 <0.1%
reforms borderposts,onestopwindowattheport,andsoforth)
Strengthenandoperationalizethepublicprivate Shortto MTI/MEACA 0.3 <0.1%
dialogueframework mediumterm
98
(201015) cost(US$ expenditureprojected
million) for201015
3. Increasingthecontributionoftheprivatesectortogrowth
Financial ProvidetechnicalassistancetobanksandMFIsto Shortto Banks/BRB/develop 1 <0.1%
sector developproductsthatmeettheneedsofsmalland mediumterm mentpartners/
mediumenterprises administration
responsiblefor
privatesector
Establishunitswithinbanksthatspecializeinlendingto Shortto Financial Includedin
smallandmediumenterprises,andunitsresponsiblefor mediumterm institutions/BRB/ previous
ruralandagriculturalcustomers government technical
assistance
Revisetheregulatoryframeworklimitingthe Shortto BRB 0.1 <0.1%
developmentofMFIs;andremovethebanonMFIs mediumterm
engaginginleasingandissuingmortgages,subjectto
priorapprovalfromthecentralbank
ImprovecapacityoftheBRB tosuperviseMFIs Shortto BRB 0.4 <0.1%
mediumterm
Regulatory Finalizeandadopttherevisedminingcode Shortterm MEN 0.6 <0.1%
environment Finalizetherevisionofthetaxcode,includingthe Shortto MF/MTI 0.35 <0.1%
introductionofthesimplifiedimptsynthtique mediumterm
(synthetictax)
Governance PromulgatetherevisedPrivatizationLaw Shortterm MGGP tbd
Adoptandimplementthenationalpolicyofgood Shortto MGGP/MF 30 0.5%
governanceanditsactionplan mediumterm
4. Encouragingemerginggrowthsectors
Foodcrops Strengtheninputdistributionsystems (technical Shortto MA 0.5 <0.1%
assistance) mediumterm
Rehabilitate,create,andstrengthenlocalinfrastructure Shortto MA 10 0.2%
forstorage,conservation,transformation,and mediumterm
commercializationofagriculturalproducts
Organize,structure,andprofessionalizeproducers Shortto MA 3 <0.1%
mediumterm
99
(201015) cost expenditureprojected
(US$millions) for201015
Coffee Strengthenthecapacityofthe regulatoryagency (ARFIC) Shortterm MA 1.5 <0.1%
Launchthesecondroundofwashingstationsanddry Shortterm MA/MGGP 0.5 <0.1%
millsprivatizing
Promotemicrocreditandmatchinggrantprogramsto Shortterm MA/MF 30 0.5%
helpfarmersfinanceinputsorreplanttrees
Strengthenthecapacityofthe interprofessional Mediumterm MA 1.5 <0.1%
association(INTERCAFE)
Improveprocessingpractices Mediumterm MA 0.5 <0.1%
Diversification Developanactionplantoestablishfacilitiestocomply Mediumterm MTI 1.5 <0.1%
withinternationalsanitaryandphytosanitarystandards
Improvetheavailabilityofcargospacetoexportfresh Shortto MTI/MTR/MA 0.5 <0.1%
produce mediumterm
Mining Developandimplementasectoralstrategythat Shortto MEN 0.15 <0.1%
explicitlyaccountsforthecharacteristicsandconstraints mediumterm
ofthedifferenttypesofmineralresources
BecomeacandidatecountryoftheEITI,bevalidatedas Shortto MEN 0.15 <0.1%
acompliantcountry,anduseavailablefacilitiesfor mediumterm
assistancewithnegotiations
Revisetheinstitutionalframeworkbycreatingaunit Shortterm MEN 0.3 <0.1%
withintheMENtodealspecificallywithartisanaland
smallscalemining
Developaprojecttosupporttheformalizationof Shortto MEN 1 <0.1%
artisanalandsmallscalemining mediumterm
Note:BRB=BanquedelaRpubliqueduBurundi;CD=CustomsDirectorate;EAC=EastAfricanCommunity;EITI=ExtractiveIndustriesTransparencyInitiative;MA=Ministryof
Agriculture; MEACA = Ministry of East African Community Affairs; MEN = Ministry of Mines and Energy; MF = Ministry of Finance; MFI = microfinance institution; MGGP =
Ministry of Good Governance and Privatization, MPlan = Ministry of Planning; MPW = Ministry of Public Works; MTI = Ministry of Trade and Industry; MTR = Ministry of
Transport and Telecommunication; NTB = nontariff barrier; OdR = Office des Routes; REGIDESO = Rgie de Production et de Distribution d'Eau et dElectricit; tbd = to be
determined.Totalexpenditureprojectedfor201015referstothealternativescenariopresentedinsectionA,chapter4,assumingaconstantFBu/US$exchangerate.
100
APPENDIX2:EXPORTDIVERSIFICATION
FigureA.1:ProductSpaceMapofBurundi

Cotton (other than
linters), not Goat and kid Sheep and Ores and
carded or combed skins, raw lamb skins with concentrates of
Fish, fresh wool on, raw other nonferrous
or chilled, base metals
excluding
fillets
Tea
Bulbs,
tubers, and Leather of
rhizomes of other hides
flowering or and skins
of foliage
plants
Cut flowers
and foliage
Coffee,
whether or
not roasted
Gold,
nonmonetary
Bovine and equine

hides (other than
calf), raw
Bones, horns,
ivory, hooves,
claws, and so Plants, seeds, fruit
forth used in perfumery
and pharmacy

Classics Disappearances Emergingchampions
Source:AuthorsestimatesbasedoncalculationsbytheEconomicPolicyandDebtDepartment;mapgeneratedusingsoftware
byHidalgoetal.(2007),availableathttp://www.chidalgo.com/productspace/data.htm.
101
TableA.2:DetailedClassificationofBurundisExports
Shares(%oftotal
Productname Tech Exports(US$thousands) exports)
code PRODY 198589 200506 198589 200506
Classics: RCA198589=1;RCA200506=1
Coffee,whetherornotroastedorfreedofcaffeine PP 1,936 96,524 51,274 71.02 80.41
Tea PP 1,655 5,186 4,480 3.83 8.52
Cotton(otherthanlinters) PP 1,500 1,288 1,287 0.73 2.75
Gold,nonmonetary RB 5,716 27,171 1,273 18.67 0.81
Fish,fresh(live/dead)orchilled,excludingfillets PP 4,919 269 196 0.20 0.36
Goatandkidskins,raw(fresh,salted,dried,pickled) PP 1,217 962 126 0.70 0.14
Plants,seeds,fruitusedinperfumery,pharmacy PP 3,622 303 92 0.24 0.14
Sheepandlambskinswithwoolon,raw(fresh,salted) PP 4,956 222 21 0.15 0.04
Total n.a. 3,190 131,925 58,750 95.53 93.18
Disappearances: RCA198589=1;RCA200506=0
Bovineandequinehides(otherthancalf),raw PP 5,653 708 22 0.52 0.01
Bones,horns,ivory,hooves,claws,coral,shells,andsoforth PP 4,419 832 0.52
Total n.a. 5,036 1,539 22 1.05 0.01
EmergingChampions: RCA198589=0;RCA200506=1
Oresandconcentratesofothernonferrousbasemetals RB 1,982 282 384 0.08 0.70
Cutflowersandfoliage HVPP 4,015 61 375 0.03 0.69
Bulbs,tubersandrhizomesoffloweringoroffoliageplants;cuttings,
slips,livetreesandotherplants HVPP 9,062 342 102 0.25 0.17
Leatherofotherhidesorskins LT 2,156 344 53 0.26 0.06
Total n.a. 4,304 1,030 915 0.62 1.62
Marginals:RCA198589=0;RCA200506=0;averageexportvalue>USD25,000
Sawlogsandveneerlogs,ofconiferousspecies RB 8,841 498.12 0.555
Tobacco,notstripped LT 3,317 342.33 289.31 0.124 0.185
Othertoolsforuseinthehand LT 10,548 1.64 216.80 0.001 0.482
Otheroutergarmentsoftextilefabrics LT 5,408 4.79 177.88 0.002 0.264
Footwear PP 7,765 18.76 91.18 0.002 0.199
Undergarments,knittedofcotton LT 4,975 12.21 64.13 0.003 0.087
Bananas,freshordried HVPP 5,183 8.77 58.05 0.002 0.065
Otheroutergarmentsandclothing,knitted LT 6,020 29.55 30.69 0.004 0.043
Tobacco,whollyorpartlystripped RB 1,531 377.47 27.00 0.145 0.017
Woodofnonconiferousspecies,sawn,planed,tonguedandsoforth LT 3,667 9.78 24.95 0.003 0.041
Tarpaulins,sails,awnings,sunblinds,tentsforboats,andsoforth LT 4,852 39.85 25.94 0.006 0.045
Sawlogsandveneerlogs,ofnonconiferousspecies RB 2,287 335.79 18.29 0.095 0.020
Calfskins,raw(fresh,salted,dried,pickled/limed) RB 4,065 406.17 10.56 0.160 0.007
Worksofart,collectorspiecesandantiques PP 8,542 34.19 8.12 0.024 0.005
Woodofconiferousspecies,sawn,planed,tonguedandsoforth RB 11,578 15.23 6.00 0.008 0.004
Shirts,men's,oftextilefabrics LT 4,936 66.01 5.28 0.018 0.003
Beans,peas,lentilsandotherleguminousvegetables HVPP 2,376 4.70 0.003
Otherpreciousandsemipreciousstones MT 2,846 579.51 0.26 0.323 0.000
Trousers,breechesetc.oftextilefabrics LT 4,789 0.14 0.000
Chemicalproductsandpreparations,n.e.s. RB 19,098 0.69 0.10 0.000 0.000
Otherfreshorchilledvegetables HVPP 5,477 44.88 0.02 0.023 0.000
Animals,live,n.e.s.,incl.zooanimals,dogs,catsandsoforth RB 3,680 41.28 0.029
Fish,frozen(excludingfillets) PP 6,457
Crustaceansandmolluscs,fresh,chilled,frozenandsoforth PP 3,369
Fruit,freshordried,n.e.s. HVPP 5,187 64.23 0.034
Sheepandlambskinswithoutthewool PP 2,349 19.46 0.004
Sheepandlambskinleather LT 2,526 70.22 0.044
Cottonyarn LT 4,262
Cottonfabrics,woven,unbleached,notmercerized LT 4,128
Portlandcement,cimentfondu,slagcementandthelike RB 5,109 127.79 0.017
Containers,ofglass,usedforconveyanceorpacking LT 6,824
Nickelandnickelalloys,unwrought RB 10,397 4.43 0.001
Aluminumandaluminumalloys,unwrought RB 9,077
Zincandzincalloys,unwrought RB 8,334
Total n.a. 5,907 2,615 1,532 1.07 1.98
Source:COMTRADE,SITC24digits;calculationbyauthorsandtheEconomicPolicyandDebtDepartmentoftheWorldBank.
Note:=noexports;n.a.=notapplicable;n.e.s.=notelsewherespecified;RCA=revealedcomparativeadvantage.
102
TableA.3:Horticulture:Markets,Constraints,andOpportunitiesforBurundi
Market Constraints Opportunities
Developedeconomy x Lowtoleranceforfailuretocompeteon x Massivemarketofmillionsofconsumers
markets cost,quality,volume,timing,andflexibility x Largeethnicpopulationswithhigherdemandfor
(particularly x Longertraveldistance;requireshigher exoticgoods
EU) standardsforquality,safety,andpackaging x Organicandenvironmentallyfriendlyproductsin
materials highdemand
x Morestrictenvironmentalregulations x Capacitytocustomizeproductionofsmallorders
regardingpackagingmaterialsinEU (comparedwithotherproducts)ofspecialties
markets x Retailindustryorganizationhighlyfocusedon
x Commoditizationofmainstream categorymanagement;prefersupplierswho
horticultureproducts,suchasmangoes, provideseveralorallproductsunderthesame
papayas category(forexample,tropicalsorleafygreens)
x HighqualitystandardsdifficultforBurundi ratherthanjustoneproduct
tomeetintheshortterm(forexample, x Directaccesstosupermarketsandmajor
EurepGAPregulations) wholesalers
x Shrinkingmarketwindowsforcountries
exportingthesameproducts
Emerging/regional x Costdrivenmarkets x Relativelyshorttraveldistanceswithinthecountry
markets x FeederroadstotrunkroadsinBurundi favortransportofperishableproducts
(surrounding limitedtogather/consolidateproduct x Highertoleranceforlowerquality,lessvolume,
countries) acrossdifferentproductionareas longertiming,andlessflexibility
x Truckingavailabilityalsolimited,especially x Riseofsupermarketsinseveralneighboring
reefercontainersnecessarytoreachsizable countries(suchasKenya)representsanincipient
markets210daysaway(suchasKampala, opportunitytoestablishlongtermsupply
Kigali,Mombasa,Nairobi) contracts
x Productsmaycompetewithlocal x Incipientcategorymanagementcriteria
productionifnoqualitydifferentiationis
attained
Emergingmarkets, x Increasinglydemandingoncost x Highpotentialtoestablishcategorymanagement
theMiddleEast x Salesprogramsmustincludesupplying supplyprogramswithsupermarketsandthe
countriesoriented severalproductsunderonecategory, hospitalitybusinesssector
towardfreetrade ratherthanonlyoneortwoproducts x ShorterflyingtimethantoEUmarkets
(Bahrain,Kuwait, x Directtransportserviceslacking x Foodproductsfacenoquantitativerestrictionson
Oman,Qatar,and x Difficulttonegotiatepricing tariffornontariffbarriers(exceptvegetableoils)
theUnitedArab x Highqualitydemanded x Thesecountriesimport90percentoftheirfood
Emirates) x Buyersmaynotbetrustworthy,requiring needs
veryclosesupervisionandintensive x Onestopredistributionhub(Dubai)toother
managementofaccounts MiddleEasternnations
x Difficulttofindconsistentwholesalers x Moretoleranceonquality,foodsafety,and
x Neighboringcountries(suchasKenya) packagingstandards
alreadyhavestartedtodevelopthese x Paymentsareuponarrivalandacceptanceof
marketsfortemperateandtropical shipments
products.Afghanistan,withUSAIDsupport, x Highdemandfortropicalandsemitropical
hasstartedanaggressiveprogram products
exportingspecialtyvegetablesandleafy x PriceshigherthaninEuropeanmarkets
greens.
Source:Baghdadli,Harborne,andRajadel2008.
Note:EU=EuropeanUnion;USAID=U.S.AgencyforInternationalDevelopment.
103
TableA.4:RecommendationsfortheArtisanalandSmallScaleMiningSector
Area Recommendations
Revisionofthe CreatewithinMENaunitwithintheDGGMthatdealsspecificallywithallaspectsofartisanalandsmallscale
institutional mining,including
framework x grantingofexploitationpermits;
x administrativeandtechnicalsupervision;
x training;
x managementsupervisionofenvironmentalaspectsofmining.
Revisionofthelegal - Allregulationsshouldbeguidedbytheminingcode,implementedcentrallybytheDGGM.The
andregulatory regulatoryframeworkshouldbesummarizedandtranslatedintoKirunditobeaccessibletotheartisanal
framework andsmallscaleminers.
- Therolesandresponsibilitiesofthedifferentstakeholdersneedtobeclarifiedthatis,thegovernment,
concessionariesandotherprivateoperators,provincialandlocalauthorities,miningassociationsand
cooperatives,andtheminers.
- Giventhecurrentlyongoingrevisionoftheminingcode,thefollowingcriticalpointsshouldbe
considered:
x Revisionand/orprecisionofthedefinitionofartisanalandsmallscaleminingactivities,the
supervisionofartisanalminers,theallocationofartisanalexploitationtitles,andthetransferand
leaseoftitles.
x Miningtitleandtheresponsibilitiesofthetitleholdershouldcontainregulationsregardingthe
relationshipbetweenthetitleholder(concessionary)andthelandowner.
x Fairdistributionofrevenueshouldbeguaranteedforallstakeholders;
x newmininglegislationshouldrecommendastrategytodevelopandpromoteartisanalandsmall
scaleminingthatisintegratedwithotherdevelopmentstrategiesinthecountrytoensurethe
contributionofminingtopovertyreduction,thepromotionoftheroleofwomeninthesector,
andthefightagainstchildlabor.
x Environmentalaspectsshouldbeintegratedintherevisedlegislation.
x Roleofcommunitiesshouldbestrengthened.
Implementationand ThesectoralstrategyoftheMENneedstoaccountexplicitlyforthecharacteristicsandconstraintsofthe
rationalizationof differenttypesofmineralresourcesfoundinBurundi,byconsideringcurrentknowledgeofthesubsoil,the
strategiestoexploit timehorizonandrealisticforecastoftheexploitationofthedeposits,thefragilityoftheeconomy,andthe
themineral lackofopportunitiesforemploymentcreationinruralareas.
potential
Formalizationof Proposedprojecttosupporttheformalizationofartisanalandsmallscaleminingshouldincludethe
artisanalandsmall followingactivities:
scalemining x Developanintegratedmanagementplanforartisanalandsmallscalemining,accountingfor
technical,organizational,legal,institutional,social,andenvironmentalaspects.
x WithintheDGGM,createandsupportaunitthatdealsspecificallywiththepromotionofartisanal
andsmallscaleminingandtheprotectionoftheenvironment.Strengthenthecapacitiesofthe
institutionsrelatingtoartisanalandsmallscalemining.
x BegintechnologytransferwithtwopilotprojectsintheprovincesofCibitoke(gold)andKayanza
(coltan),Kirundo(wolfram),andMuyinga(goldandwolfram)throughprivateoperatorswith
supportoftheDGGM.
x Establishand/orencouragetheestablishmentoforganizationalstructureswithinthemining
community.
x Integrateexcombatantsaswellasunemployedyouth.
Furthermore,thefollowingconditionsneedtobefulfilledtoformalizethesector:
x improvedaccesstofinance;
x improvedaccesstominingtechnologyandequipment
x improvedinvestmentclimateforlocalandforeigninvestors;and
x aninstitutionalframeworkthatmanagesthesectorinatransparentandefficientmanner.
Source:Midende2009.
Note:DGGM=DirectorateGeneralforGeologyandMining;MEN=MinistryofMinesandEnergy.

104
APPENDIX3:REGIONALINTEGRATION
TableA.5:BurundisPostIndependenceExperiencewithRegionalIntegrationAgreements
Agreement Members Objectives Instruments Results
CEPGL:Economic Burundi,Democratic Promote Reducingbarriersand Littleprogressin
CommunityoftheGreat RepublicofCongo, cooperationand improvingmobilityof liberalizingtradeand
Lakes,createdin1976 Rwanda economic factorsofproduction, factormobility:some
development jointindustrialprojects industrialprojects
throughacustoms (agriculture,energy,
unionandsectoral health,andconstruction)
cooperation
OBK:Organizationfor Burundi,Rwanda, Promoteprojects Jointindustrialprojects
Developmentand Tanzania,Uganda fordevelopmentof
Managementofthe theKageraRiver
KageraRiverBasin, Basin
createdin1979
CEEAC:Economic Angola,Burundi, Economic Studyofmethodsof
CommunityofCentral Cameroon,Republic development recoveryprocess
AfricanStates, ofCongo,Gabon, throughacustoms
establishedin1983 EquatorialGuinea, unionandsectoral
CentralAfrican cooperation
Republic,Democratic
RepublicofCongo,
Rwanda,SoTom
andPrincipe,Chad
COMESA:Common Burundi,Democratic Economicandsocial Removebarriersto SettingupofFTA(13of
MarketofEasternand RepublicofCongo, development trade;PTABankfor 19members);trade
SouthernAfrica,created Comoros,Djibouti, throughacommon TradeandDevelopment; facilitationby
in1994(formerlyPTA, Egypt,Eritrea, market,monetary ProgramITF/RIFF; harmonizingcustoms
createdin1981) Ethiopia,Kenya, andfinancial clearinghouse;Centerfor proceduresand
Madagascar,Malawi, cooperation,and CommercialArbitration, documentsacross
Mauritius,Namibia, coordinationof CourtofJustice, multiplechannels
Uganda,Rwanda, macroeconomic Interregional (SYDONIA,DTDR,andso
Seychelles,Sudan, policies CoordinationCommittee, forth);freemovementof
Swaziland,Zambia, andInsuranceandRe people(evenpartialfree
Zimbabwe InsuranceRegional movement)
Company
Source:Hugon2003.
Note:DTDR=Dclarationdetransitdouanierroutier;FTA=freetradearea;PTA=PreferentialTradeArea;SYDONIA=System
DouanierAutomatis.

105
BoxA.1:PotentialBenefits,Costs,andGuidingPrinciplesofRegionalIntegration
PotentialBenefitsofRegionalIntegration
x IncreasedreturnsandincreasedcompetitionEnlargingmarketsthroughintegrationofsmalleconomies(1)promote
possibilitiestoachieveeconomiesofscaleandimproveefficiencies;and(2)increasecompetition,leadingtolowerprices
andexpandedsupply.
x TradeandlocationeffectsPreferentialreductionsintariffswithinregionalagreementscaninduceshiftsinbothdemand
andsupply.Neteffectsonnationalincomedependoncostsofalternativesupplyandtradepoliciestowardnonmember
countries.
x InvestmentsRegionalcooperationandestablishedagreementscanattractmoreforeigndirectinvestmentby(1)
enlargingmarkets(particularlyforlumpyinvestmentsviableonlyaboveacertainsize)and(2)reducingdistortionand
loweringmarginalcostofproduction.
x CoordinationandcollectivebargainingpowerRegionalintegrationagreementsmayenablecountriestocoordinate
negotiatingpositionsininternationalforums,thusraisingvisibilityandpossiblystrengtheningbargainingpower.
x ManagementofsharednaturalresourcesManywatersheds,mineraldeposits,fisheries,andsensitivenatural
environmentsaresharedamongcountries.Nocountryactingalonecanensuresustainablemanagement,whichdepends
oncollaborativeefforts.
x ManagementofregionalcommonsEffectiveactiontocombatmigratorydiseases(suchasHIV/AIDSandmalaria)and
vulnerabilitiesarisingfromclimatechangedependsoncollaborativeeffortsamonggroupsofcountries.
x PolicylockinandcommitmentmechanismRegionalagreementscanprovideacommitmentmechanismforcountries
domestictradeandotherpolicyreforms,reducingthelikelihoodofpolicyreversals.Thismechanismbearsuponpolitical
aswellaseconomicreforms.
x InsuranceIntegrationagreementsprovidemembersinsuranceagainstexogenousshocks(termsoftradeshocks,conflict,
protectionismindevelopedcountries,andimpactsfromclimatechange).
x SecurityRisksofconflictareperceivedtobeloweredthroughintegrationagreements,asaresultofimproved
intraregionalconfidenceandtrust,commondefensearrangements,andinterdependenceinkeyaspectsofcountries
nationaldevelopment.
PotentialCostsofRegionalIntegration
x TradediversionDisplacementoflowcostproductsfromnonmembersbyhighercostproductsfromcountrieswithina
regionalintegrationarrangementhasbeenamajorproblemwithseveralregionalintegrationagreements.Welfaregains
arerealizediftradecreationdominatestradediversion,butthisoutcomecannotbeensuredinadvance.
x RevenuelossTradeintegrationagreementsreducegovernmentstariffrevenues,bothdirectlythroughtariffcutsand
indirectlythroughshiftsawayfromimportsfromnonmembersthataresubjecttotariffs.Netcostsdependonhowmuch
newtradeisgeneratedfromtheintegrationagreement,buttheycanbehighwhentradefacilitationisdifficult.
x IndirectcostsThesecostsarisefromthefreemovementofpeopleandcapitalacrossnationalborders(capitalflightand
lossofskilledhumanresources,forexample),andextravigilanceisrequiredtopreventcrossbordercrime.
GuidingPrinciplesforEffectiveIntegration
x Openregionalismiscritical.Regionalandglobalintegrationarecomplements,andneedtobepursuedinparallelandwith
equalvigor.
x Attentionmustbepaidtotheneighborhood.Neighborhoodshaveimportantimplicationsforthefocusandsequencingof
strategiestowardregionalintegration,particularlyforlandlockedcountries.
x Startingsmalliseasier.Thesizeofthegroupisimportant.Thelargerthenumberofparticipants,themorecomplexare
thecoordinationchallenges,makingfailuremorelikely.
x Stayfocusedandproceedincrementally.Experienceshowsthatthisislikelytogivebetterresults.
x Developcrediblemechanismstocompensatelosers.Participantsoftendonotsharebenefitsequally;mechanismsof
compensationarecriticalifmoreambitiousintegrationistotakehold.
x Subsidiarityisimportant.Thesubsidiarityprinciplerespectsnationalsovereigntyandcapacityandavoidsoverloading
scarceregionalcapacity.
x Variablegeometryisnecessary.Variablegeometryisanonuniformmethodofintegrationthatenablessubgroupstomove
fasterthanthewholegrouportomovetoadeeperform.Itinvolvesheterogeneousstrategiesindesignand
implementationofregionalarrangements.
Sources:WorldBank2008c,d.
106
TableA.6:MatrixofStrategicObjectivesandIndicativeOutputsfortheMEACA
Strategicobjectives Indicativeoutputs
1. Tobuilduptheorganizationaland x Staffarepreparedtocontributetotheworkoftheministry
institutionalstructureoftheministry x MEACAisorganizedinanoptimalwaytodeliveritsmandate
sothatitcansuccessfullyplayitsrole x Staffhavetherightskillstoplan,prioritize,monitor,andfollowuponall
ofpromotingtheintegrationof ministryactivities
BurundiintheEAC x Staffhaveadequateequipment(computers,communications
equipment,andsoforth)
x Thenecessaryinformationisavailablewhereandwhenrequired
2. Toadopteffectivecoordination x Strategicandeffectivestructureforcoordinationisinplace
mechanismsforimplementingEAC x Proceduresforcoordinationareinplace
affairs x Communicationmechanismisoperating
x Focalpointshavethenecessaryresourcestobeeffective(information,
skills,mandate,resources,andsoforth)
x Appropriatelinkswithrelevantgovernmentinstitutionsareinplaceand
rolesandresponsibilitiesareclear
3. TonegotiatewithEACpartnersthe x Technicalskillsandinformationneededfornegotiationareacquired
processofBurundisfullintegration x Logisticsforcooperationareinplace(transport,funds,information,
delegatedauthorityallinatimelymanner)
x MEACAisclearaboutcurrentstatusoflegislationcomparedwiththe
acquiscommunautaire
x Progressagainstmilestonesismonitoredandfollowupactionistaken
asrequired
x Thereisassurancethatnegotiationsarebeingcarriedoutinanefficient
andeffectivemanner
4. Tomanageasystemof x Eachstakeholdergrouphastherequiredinformation
communicationinvolvingall
stakeholders
5. Tocoordinatetheprivatesectorand x Legalandinstitutionalframeworksaremadeavailable
civilsocietysotheycancontributeto x Privatesectorandcivilsocietyparticipatefully
thesuccessoftheprocessof x ThereisacompetitivenessreviewofthekeysectorsofBurundis
BurundisintegrationintheEAC economy
6. Toensuremechanismsand x Informationoncurrentstatusandchallengesfacingintegrationis
capacitiesareinplacetomonitorand availablewhereandwhenrequired
evaluatetheintegrationprocess x Appropriateactionistakenonthebasisofmonitoringandevaluation
datagathered
x ActivitiesofBurundiandotherEACmemberstatesinregionaland
multilateralorganizationsaremonitored
Source:StrategyoftheMEACA,draft,March2009.
Note:EAC=EastAfricanCommunity;MEACA=MinistryofEastAfricanCommunityAffairs
107
TableA.7:PolicyMatrixforAdvancingRegionalIntegrationinBurundi
Measure Timeline Responsibleinstitutionand
technical/financialassistance
A.Implementationandharmonizationoffiscalinstruments
Finalizethetaxandcustomscodes ST MF/WB
Computerizethetaxdepartment ST MF/WB/DFID
IntroducetheVAT ST MF/PAGE/WB
AdoptEACfiscalyear ST MF
ImplementtheBRA ST/MT MF/DFID
B.Implementationofprivatesector;financial/bankingandpaymentinstrumentsa
Developamodernpaymentsystem ST/MT BRB/IMF
Developotherfinancialmarketandbankingmeasures MT/LT BRB/IMF
Strengthenthepublicprivatedialogue ST/MT MTI/CCIB/WB
C.Implementationoftradeinstruments
AdopttheCET ST MTI/CD/MF/WB
Eliminateremainingtariffs ST/MT MTI
EliminateremainingNTBs MT MTI/CD
DisseminatetheVATandtheCET ST/MT MTI/WB
Participateindevelopingandadoptingacommontradepolicy MT MTI
Implementtheinvestmentandexportpromotionagency MT MEACA
D.Institutionalarrangements
EnsureasuccessfulaccessiontotheCU ST MEACA
RecruitcompetentstaffinMEACA ST MEACA
ParticipateintheEACmissions ST/MT MEACA
Providetraining(Englishlanguageskills) ST MEACA/DFID
Implementthecoordination,monitoring,andevaluationmechanisms ST MEACA
Implementthecommunicationstrategy ST MEACA/DFID
Strengthenregionalintegrationskills ST/MT MEACA/DFID
ImplementothermeasuresoftheMEACAinstitutionalmatrix(ActionPlan, MT MEACA/DFID
seetableA.6)
E.Infrastructure
Roads,railway,pipeline,port,airport MT/LT MTR
Communications(submarinecablenetwork) ST/MT/LT MTR
Energy ST/MT/LT MEN
F.Mining
Exploitationofmineralresources MT/LT MEN
Note:BRA=BurundiRevenueAuthority;BRB=BanquedelaRpubliqueduBurundi;CCIB=ChamberofCommerceandIndustry
ofBurundi;CD=CustomsDirectorate;CET=commonexternaltariff;CU=CustomsUnion;DFID=DepartmentforInternational
Development;EAC=EastAfricanCommunity;IMF=InternationalMonetaryFund;LT=longterm;MEACA=MinistryofEast
AfricanCommunityAffairs;MEN=MinistryofMinesandEnergy;MF=MinistryofFinance;MT=mediumterm;MTI=Ministry
ofTradeandIndustry;MTR=MinistryofTransportandTelecommunication;NTB=nontariffbarrier;Page=ProjetdAppuila
GestionEconomique;ST=shortterm;VAT=valueaddedtax;WB=WorldBank.
a.SeealsotableA.13.

108
APPENDIX4:INFRASTRUCTUREACTIONPLANTABLESANDMAPS
TableA.8:BasicInfrastructureCoveragefortheEastAfricanCommunity,2006
EastAfricanCommunity OtherCountryGroupings
Sub Otherlow
Indicator Burundi Kenya Rwanda Tanzania Uganda Average
Saharan income
Africa countries
Roads
Percentofroadspaved 10 14 19 9 23 15
Pavedroaddensity 31 134
Roaddensityforarableland 13 12 12 9 14 12 137 211
Totalroaddensity 48 11 57 9 36 12
Electricpower
Generationcapacity 6 37 326
Electricitycoverage 2 13 5 11 8 10 16 41
Electricpowerconsumption 14 138 61
Communications
Mainlinedensity 4 8 2 4 4 5 16 78
Mobiledensity 25 201 33 146 67 126 175 76
Internetdensity 7 76 11 10 50 39 34 30
Waterandsanitation
Accesstoimprovedwater 71 57 65 55 64 60 58 72
Accesstoimprovedsanitation 41 42 23 33 33 35 31 51
Source:AscitedinAfDB(2009):WorldBank2008c;AfricaDevelopmentIndicators,2008/09;andwww.infrastructureafrica.org.
Note:=notavailable.Roaddensityisinkilometersofroadpersquarekilometerofarableland;telephonedensityisinlines
per1,000people;generationcapacityisinmegawattsper1millionpeople;electricity,water,andsanitationcoveragearein
percentageofpopulation.
TableA.9:ServiceCostsandDifficultiesintheEastAfricanCommunity
SubSaharan
Indicator Burundi Kenya Rwanda Tanzania Uganda
Africa
Supplyproblems
Waterfailureforfirmsreceivingwater(daysperyear) 12 85 105
Electricalpoweroutagespermonth(number) 12 14 12 11
Numberofdaysforelectricalconnection 24 51 18 44 33
Firmswithaccesstoprivategenerator(%) 42 71 58 46 29
Fuelprices(US$perliter)
Dieselfuel 1.22 0.98 1.08 0.99 1.01 0.98
Gasoline 1.20 1.12 1.11 1.04 1.17 1.03
Communicationscosts(US$)
Pricebasketforinternet(permonth) 52.00 79.20 29.40 93.60 95.80 42.10
Fixedlinelocalcallfor3minutesinpeakhours 0.07 0.11 0.08 0.16 0.28 0.14
Cellularlocalcallfor3minutesinpeakhours 0.58 0.64 0.79 0.69 0.67 0.77
InternationalcalltoUSfor3min.inpeakhours 0.07 0.13 0.18 0.14 0.28 0.13
Source:AscitedinAfDB(2009):WorldBank2008c,AfricaDevelopmentIndicators,2008/09.
Note:=notavailable.
109
TableA.10:DescriptionofVariousScenariosConsideredintheReport
ProgramorProjectIncludedinScenario
Currentstrategy InfrastructureActionPlan
Nickel Rail
Scenario Descriptionofscenario fornational Withpublic Withprivate
mining extension
development investment investment
project intoBurundi
continues component component
A BaseCase Yes Yes Yes Yes No

CoreInfrastructureAction
B Yes Yes Yes No No
Planimplemented
PrivatefundingforAction
C Yes Yes No No No
Plannotavailable
Only50%ofpublicfunding Implement50%
D Yes No No No
availableforActionPlan ofActionPlan
Only20%ofpublicfunding Implement20%
E Yes No No No
availableforActionPlan ofActionPlan
ImplementationofBase
F Yes Yes Yes Yes Yes
Casewithrailextension
Source:AfDB2009.
110
FigureA.2:TransportCorridorsfortheEastAfricanCommunity

Source:AfDB2009.

111
FigureA.3:RailwayNetworkoftheEastAfricanCommunity

Source:AfDB2009.
112
APPENDIX5:INFRASTRUCTUREACTIONPLANMAINCOMPONENTS74
Main components of the infrastructure action plan
The focus of the proposed infrastructure action plan is on power, transport and communications.
Discussions with the Government of Burundi have confirmed the key objectives for the program.
Power sector
The proposed program for the power sector has six key objectives:
x Through increased investments in domestic and regional generation capacity, ensure that the
business community and households have access to reliable power supply 24 hours a day.
x Establish a national transmission and distribution grid by 2015, with all 15 of the provincial
capitals linked to this grid which would supply electricity 24 hours a day at reasonable cost.
x With the grid in place, increase the electrification rate from the current two percent of households
to 25 percent by 2020 and at least 40 percent by 2030. By 2020, 85 percent of urban households
would have continuous access to the national distribution network, and by 2030, one-third of all
rural households would be linked to the grid.
x Give priority to further development of domestic energy sources to avoid excessive dependence
on imported supplies of electricity. At the present time, about 45 percent of total supply is
imported. The design of the Action Program aims to keep this dependence at less than 50 percent
and close to current levels until 2024, after which there would be a gradual increase in imported
supplies of electricity.
x Improve demand management and reduce system losses.
x Ensure that the national electricity utility is built up into an effective and financially sound entity.
To meet the projected demand for power, the required generation capacity for Burundi would be about
600 MW by 2030, compared with less than 40 MW at present. The domestic and regional power plants
identified and included in the proposed action plan would be sufficient to meet the needs of the country
until the mid 2020s. Provided that no further major mining or power intensive industrial projects are
launched, the supply deficit would grow to about 1,240 GWh by 2030. The implication is that Burundi
would need an additional 200 MW of capacity to meet domestic demand. The working assumption in the
proposed program is that this shortfall is met by the import of additional electricity. The key policy
question for the longer-term is whether to investigate other domestic hydro power sites within Burundi
and develop these in order to keep dependence on imported power to prudent levels, or whether to allow
increased dependence on imported power. If all the additional required capacity was domestic, the share
of domestic power in total consumption would be 75 percent by 2030. A potential issue with these
domestic sites is that the cost of the electricity produced may be significantly higher than that imported
from Ethiopia via the EAPP grid.75
The ongoing operational and financial rehabilitation of the national power utility, REGIDESO, is central
to its role as a major source of funding for the future power program. With continued prudent financial
management the utility would develop into a major corporation by 2030, at which time it would have
assets of about USD 1.6 billion and revenues of about USD 300 million a year (both at 2007 constant
prices). The improved cash flow in the decade ahead would allow the utility to pass on benefits to
consumers with an ongoing reduction in power tariffs from about 2016, which would lower the average
tariff to less than nine U.S. cents a kWh. From about 2020 onwards, the utility would be able to fund the
74
This section draws from AfDB (2009).
75
The trade-off between the degree of self-sufficiency in power supply and the cost of power and its effects on the
competitiveness of Burundi business is discussed in more detail in AfDB (2009).
113
bulk of the power development of the country from its own resources and from prudent access to
commercial sources of debt financing.
Transport sector
The key objectives of the proposed action plan for the transport sector are to lower the costs of transport
for the entire economy and to improve access to local and international markets. The proposed program
focuses on road transport and civil aviation, and also provides for further investigation of the possibility
of a rail extension from Tanzania into Burundi.
The main component of the transport program is the upgrade and expansion of the road network of the
country over the next decade. The proposed program would rehabilitate and pave the entire 1,950 km of
national roads by 2020, and for those national routes where traffic densities are high it would upgrade
roads to enhanced standards that can accommodate the increased traffic. By 2030, the urban road network
would be expanded from the current estimated 650 km of roads to about 1,650 km so that all 2.6 million
urban dwellers at that time are within 500 meters of a road that can carry vehicular traffic. The other key
component is a program that will improve provincial, community and local feeder roads in key
agricultural areas to facilitate access to product markets at home and abroad, and to key inputs, such as
fertilizer, required for production activities. In conjunction with the rehabilitation of the road network, the
program would increase substantially budget allocations for routine maintenance of these facilities.
The international airport in Bujumbura would be expanded and modernized. The intent is to ensure,
within the next five years, full compliance with the ICAO-mandated standards and procedures for
international passenger travel and for freight. This would be done through a program of staff training and
investment in infrastructure. As a result of this program, the international airport would obtain an ICAO
Certification of Aerodromes. These improvements would allow Burundi to attract major international
airlines and air freight companies, thus opening up opportunities for development of tourism and air
freight of high value export products to European and Middle East markets.
The options for the extension of the Tanzanian rail network into Burundi have also been reviewed as part
of this study. What emerges is that if the nickel deposits in the Musongati area of Burundi are developed
for the export of nickel ore, and if the mine sites have access to the rail network, the four million tons of
ore to be carried makes the investment in the rail extension quite attractive. The potential problem is that
international investors interested in developing the site may prefer to refine the ore at the mine and ship
the metal to export markets. In this case the volume of mine-related freight is substantially smaller an
estimated 50,000 Mt of nickel metal a year and a small quantity of cobalt. Based on the assessments made
for the future volumes of international freight in and out of Burundi each year that are unrelated to the
mine activities, it would appear that the rail extension is not economically viable unless it carries a large
volume of mine freight. If one of the two options were to go ahead in the absence of the mine, it is likely
that the rail operator would require substantial public subsidies. The program proposes further evaluation
of the costs and benefits of a possible extension of the Tanzanian rail network into Burundi.
The proposed program would lower the costs of transport for the entire economy and improve access to
local and international markets. The benefits of lower transport costs under the program would be
substantial; for example, at current costs for road freight of 13 U.S. cents per ton/km or more, it costs
about USD 230 to move a ton of fertilizer from the ports in Kenya and Tanzania to Burundi. A reduction
in transport costs to 8 U.S. cents a ton/km, the rate that prevails in these neighboring countries, would
lower the freight cost of the fertilizer by almost USD 100 a ton. These types of cost reductions can have a
significant impact of the profitability of farming and other business activities.
114
Communications network
The proposed communications program aims to improve substantially access to international
communications network and to lay the foundations for a national communications grid that will provide
communities and business through Burundi with low cost voice and data communications.
East Africa is the only heavily populated region of the world that does not have access to the long-
established international system of submarine cables that allow for low cost transmission of data and
voice communications. The Eastern Africa international submarine cable network is currently being laid
off the coast of East Africa with funding from the World Bank and a consortium of private investors. It is
expected to become operational in 2010. Burundi would be linked to this low-cost international
communications network via fiber optic cables that are currently being laid in Kenya and Rwanda. This
cable extension to Bujumbura is expected to be completed by June 2010. With further developments
already underway, Burundi would have four separate access routes to this regional communications
network and the submarine cable.
Against this backdrop, the key elements of the communications program are as follows:
x A high priority is attached to the immediate development of a national communications grid of fiber
optic cable and digital microwave that would be linked to the regional network. This program is
already funded by the World Bank and will begin implementation shortly. On completion in 2012, it
will have laid a fibre optic network of some 400 km throughout Burundi, along with a digital
microwave network that will serve particular communities throughout the country.
x Launch an ambitious program to expand access to this low cost international communications
network for schools, hospitals, universities, the business sector, and local communities throughout the
country.
x Develop a range of applications, including e-government, e-commerce, e-schools, and e-health and
complete the ongoing work on the legislative framework for the communications industry and
corresponding regulatory framework related to e-security, fraud, privacy, data protection, and
intellectual property rights.
x Promote the entry of additional private suppliers of communications services throughout the country
to ensure competition and service quality.
Such a national grid would give communities, businesses, and a wide range of institutions throughout the
country, access to low cost communications within the country and with the region and rest of the world.
Rural as well as urban communities are expected to benefit from this program. According to the World
Bank, for example, incomes of agricultural producers can rise by about nine percent through the use of
mobile telephones.
Building human and institutional capacities

Given the size of the proposed program to be implemented over the next two decades, Burundi will have
to make a large investment in building domestic capacities to manage the program and benefit fully from
the services it will be able to provide. There are four specific components for the proposed capacity
building program:
x Strengthen capacities of individual ministries for project design and implementation, including
for example, arrangements for procurement and site supervision. These programs are of particular
importance for the ministries responsible for power and transport.
x Strengthen and restructure arrangements for oversight and regulation of the power, transport
and communications industries, given the substantial structural changes that would occur under
the proposed Action Program.
x Strengthen capacities for regular collection and analysis of survey data for these three industries.
In the case of road transport, for example, surveys of transport service providers will provide
115
basic information about the evolving amounts of passenger and freight traffic, the costs of service
provision and prices of services offered to the public. Regular surveys of road traffic will be
required for assessments of evolving road maintenance and upgrading requirements.
x Adopt appropriate standards for infrastructure construction and training of skilled workers for
these industries. With specific standards for two lane paved national highways, for example, the
donor community can then ensure that projects they fund do comply with these standards.
Similarly, there is a need for clear standards for accredited institutions that train skilled trades
people such as electricians. In developing these standards, close attention should be given to the
evolving requirements of the EAC.
The proposed program would provide substantial support for the development of these capacities and for
a wide range of technical studies that will be required for informed decision making in the early phases of
the program. There is considerable urgency associated with the start of these capacity building programs.
Required investment for infrastructure
Development expenditures
The core infrastructure action plan calls for development expenditures of USD 4.6 billion (at 2007
constant prices, table A.11) over the next two decades. Successful implementation of this proposed
program will essentially close the large infrastructure gap between Burundi and many other developing
countries.
The power program would involve expenditures of about USD 2 billion, including about USD 465
million of private investment in new domestic generation facilities that would sell power to the national
grid. The roads program would require about USD 2.1 billion. The civil aviation program would involve a
public-private partnership (PPP) arrangement under which the upgrade and operation of the airport and
related services would be handled by one or more private contractors. The total amount of investment
required for the aviation sector is estimated at USD 260 million over the next 20 years. The program
includes about USD 120 million for the further development of the national communications grid and
widespread community access to this grid.

TableA.11:Developmentexpendituresonthecoreprogram Table A.12: Routine maintenance expenditures on the core
(US$millions,2007constantprices) program(US$millions,2007constantprices)
201019 202030 Total 201019 202030 Total
Publicexpenditures Publicsectoroutlays
Powersector 813 764 1,577 Powersector 132 409 540
Transportsector Transportsector
Roads 1,139 989 2,129 Roads 105 160 265
Ports 13 15 28 Ports 4 6 10
Civilaviation 11 6 16 Civilaviation 3 3
Subtotal 1,163 1,009 2,172 Subtotal 112 166 277
Communications 48 28 75 Communications 9 19 29
Total 2,024 1,801 3,825 Total 253 594 847
Associatedprivate Associatedprivate

investment sector
Powersector 458 8 465 Powersector 79 153 232
Civilaviation 190 55 245 Civilaviation 33 83 116
Communications 24 33 57 Communications 4 14 18
Total 672 96 767 Total 115 250 366
Total 2,695 1,896 4,592 Total 368 844 1,212
Source:AfDB2009. Source:AfDB2009.
116
The bulk of these expenditures would be capital outlays on infrastructure assets such as road networks,
airport facilities, power stations and transmission and distribution lines, and communications networks.
About three percent of the outlays (USD 170 million) would be used to meet the cost of the wide-ranging
program of capacity building initiatives and various technical studies included in the Action Plan.
In the event that the railway extension was to proceed, the estimated cost is about USD 600 million (at
2007 constant prices), not including the cost of rail extensions to mine sites in the Musongati area. It is
assumed that these latter costs would be met by the mine operator, in the event that the rail transport
option was to be used. The proposal is to use a PPP-type arrangement to fund and operate the rail service
if the project was to proceed. A small amount of public investment would be required for various studies
and for capacity building within the government for oversight and regulation of the rail services.
Expenditures on routine maintenance

The combination of prolonged civil war and limited public financial resources led to deterioration in the
basic infrastructure of the county. The program of infrastructure rehabilitation that is ongoing and the
proposals to complete this rehabilitation in the decade ahead will restore and upgrade these facilities. The
challenge then will be to step up allocations for routine maintenance of this infrastructure, and thereby
contain the need for further large outlays for rehabilitation.
The proposed action plan for Infrastructure calls for USD 1.2 billion of outlays on routine maintenance
over the next two decades (table A.12). The bulk of these expenditures are for the maintenance of the
power and roads infrastructure. As the accompanying table indicates, these estimates for maintenance
spending include expenditures by the public sector, as well as private sector outlays on power and civil
aviation infrastructure that would be operated under some form of public-private partnership (PPP)
arrangement.
The proposed program represents a major step-up in funding for routine maintenance. In the case of the
government component, outlays for maintenance average USD 25 million a year in the decade ahead.
This compares with maintenance outlays for transport, power and communications infrastructure in the
range of USD 8 million a year at the present time. In the 2020s, maintenance outlays by the public sector
would step up to an average of about USD 85 million a year. These maintenance programs will provide
important business opportunities for local companies and will create substantial additional employment
opportunities. As the subsequent discussion indicates, early action will be needed to develop the requisite
skills in the labor force (electricians, for example) for these programs.
- 117 -
APPENDIX6:BUSINESSENVIRONMENT
TableA.13:RecommendationsforImprovingFinancialSectorPerformance
Constraints Measures Timeline
Stability
Somestatebanksareindifficulty; Privatizetwostateownedbanks 2010
supervisionofbanksisnotupto Improvesupervisionbyintroducingchartofaccounts 2009
par Introduceriskbasedsupervisionofcommercialbanks 2010
ComputerizetheBRB 201011
Paymentsystemisarchaic Introduceelectronicmassclearingandrealtimegrosssettlementsystems 2010
Introducecardsystem 2011
Passnewlawonmeansofpayments 2011
Microfinancepresents Strengthensupervisionthrough chartofaccountsand 2010
weaknesses training 200912
Strengtheninstitutionsthroughinternalrestructuring 200912
Intheinsurancesectorthereisan Establishindependentregulatory/supervisoryagency 2010
absenceofanysupervisionand Restructureinsurancecompaniestomeetprudentialnormsorclosethem 2011
thereareundercapitalized
institutionsthatdonotmeet
prudentialnorms
Deficitsexistinthepension Changethemanagementstructure 2011
systems;therearehighcostsand Introduceparametricreform 2011
lowefficiencyindelivering Conductinternalrestructuring 2011
services Undertakeastudyofasecondpillar 2012
Access
Banksarenotwellpreparedto Providetechnicalassistancetodevelopnewproductsandnewlending 2011
catertosmallandmedium mechanism,andestablishspecializeddepartments
enterprisesandruralareas BuildupthecapacityoftheBNDE(BanqueNationaledeDveloppement 2011
conomique)tolendinruralareas
Microfinanceinstitutionsareill Changeregulationtobroadenthescopeofmicrofinanceinstitutionsand 2010
equippedtodealwithsmalland allowthemtodevelopnewproductsandreachmorecustomers
mediumenterprisesandrural Developfinancingorganismsasprovidedforinthelaw 2012
areas,andarelimitedintheir EstablishrelationshipwiththeBNDE 201112
lendingcapacity
Depthandreachoffinancial Developmobilephonebanking 2012
sectorareverylow Developspecializedinstitutions(leasing,venturecapital,factoring) 201213
Developwarehousing 2012
Littlehousingfinanceisavailable DevelopactivitiesofFPHU(Fondsdepromotiondelhabitaturbain)) 2011
Developnewproducts 2011
Adoptalandcode 2010
Decentralizelandregistriesandincreasetheirbudgetallocation 2011
Authorizemicrofinanceinstitutionstooffermortgageloans 2010
Providetraininginhousingfinance 201013
Insufficientinformationis Strengthentheaccountingandauditingprofession 201011
availableonpotentialborrowers; Assistfirmsinimprovingtheirfinancialstatementsandbusinessplans 201012
thereisalackofstructured Instituteadialogueamongfinancialinstitutions,thenonfinancialprivate 2010
borrowers sector,andgovernment
Strenthenruralplayersthorughtheirstructuringintointegratedvaluechains 201013
Strengthenexistinginformationcenters(forexample,introductionofunique 201011
identifier,computerization,andsoforth)
Insuranceproductsarelimited Developbrokers 200910
Grantentrytoforeigncompanies 2011
Diversifyproductsoftheinsurancecompanies 201213
- 118 -
APPENDIX7:MAPOFBURUNDI
- 119 -
References
AfDB (African Development Bank). 2009. An Infrastructure Action Plan for Burundi: Accelerating
Regional Integration Tunis, Tunisia.
Baghdadli I., B. Harborne, and T. Rajadel, eds. 2008. Breaking the Cycle: A Strategy for Conflict-
sensitive Rural Growth in Burundi. World Bank Working Paper No. 147. World Bank, Washington, DC.
Balassa, B. 1986. Comparative Advantage in Manufactured Goods: A Reappraisal. Review of

Economics and Statistics 68 (2): 31519.
Baldwin, R., J. Franois, and R. Portes. 1997. The Costs and Benefits of Eastern Enlargement: The
Impact on the EU and Central Europe, Economic Policy 12 (24): 12576.
Benitez, D., A. Estache, and D.-M. Niyungeko. 2009. Overcoming Obstacles to Growth in Burundi: The
Role of Infrastructure. Unpublished manuscript, World Bank, Washington, DC.
Boccardo, J., and V. Chandra. 2008. Why Uganda Needs to Diversify Its Exports, Background paper
for the Uganda Country Economic Memorandum, World Bank, Washington, DC.
Brenton, P., and R. Newfarmer. 2007. Watching More Than the Discovery Channel: Export Cycles and
Diversification in Development Policy Research Working Paper 4302, World Bank, Washington, DC.
Celasun, O., and J. Walliser. 2008. Predictability of Aid: Do Fickle Donors Undermine Aid
Effectiveness? Economic Policy 23 (55): 54594.
Chandra, V., and I. Osorio Rodarte. 2009. Structural Transformation for Sustainable Growth in Burkina
Faso. Background paper for the Burkina Faso Country Economic Memorandum, World Bank,
Washington, DC.
Clay, D., P. DeLucco, and A. Ottaway. 2007. Integrative Report in Support of the Burundi Coffee Value
Chain Diagnostic and Strategy Development. Economic Management Support Project, Bujumbura,
Burundi.
Clemens, M., S. Radelet, and R. Bhavani. 2004. Counting Chickens When They Hatch: The Short-Term
Effect of Aid on Growth. Working Paper 44, Center for Global Development, Washington, DC.
http://www.cgdev.org/content/publications/detail/2744
Collier, P., and A. Hoeffler. 2004. Aid, Policy and Growth in Post-Conflict Societies. European
Economic Review 48: 112545.
Coulibaly, K., N. Duffy, and K. Ezemenari. 2008. Productivity Growth and Economic Reform: Evidence
from Rwanda. Policy Research Working Paper 4552, World Bank, Washington, DC.
CPM Group. 2009. The Current Market Environment and Outlook. Presentation at the World Bank,
Washington, DC.
Dijkstra, T. 2001. Export Diversification in Uganda: Developments in Non-Traditional Agricultural

Exports, Working Paper, 47/2001, African Studies Centre, Leiden, The Netherlands.
- 120 -
EAC (East African Community). 2006. Development Strategy 20062010: Deepening and Accelerating
Integration, East African Community Secretariat, Arusha, Tanzania.
FAO (Food and Agriculture Organization). 2003. Burundi : Profils des pches et de laquaculture par
pays. Washington, DC. http://www.fao.org/fishery/countrysector/FI-CP_BI/fr.
Finger, J., and M. Kreinin. 1979. A Measure of Export Similarities and Its Possible Uses. Economic
Journal 89 (356): 90512.
Foster, V., and C. Briceo-Garmendia, eds. 2009. Africas Infrastructure: A Time for Transformation.
Washington, DC: World Bank.
Foster, M., and T. Killick. 2006. What Would Doubling Aid do for Macroeconomic Management in
Africa. Working Paper 264, Overseas Development Institute, London, UK.
Fujita, M., and J.-F. Thisse. 2002. Economics of Agglomeration: Cities, Industrial Locations, and
Regional Growth. Cambridge, UK: Cambridge University Press.
Global Witness. 2009. Faced with a Gun, What Can You Do? War and the Militarisation of Mining in
Eastern Congo. July. http://www.globalwitness.org/fwag/.
Hausmann, R., J. Hwang, and D. Rodrik. 2007. What You Export Matters. Journal of Economic
Growth 12 (1): 125.
Hausmann, R., and B. Klinger. 2006. Structural Transformation and Patterns of Comparative Advantage
in the Product Space. Working Paper 128, Center for International Development, Kennedy School of
Government, Harvard University, Cambridge, MA.
Hausmann, R., L. Pritchett, and D. Rodrik. 2005. Growth Accelerations. Journal of Economic Growth
10 (4): 30329.
Heritage Foundation. 2009. 2009 Index of Economic Freedom. Washington, DC.

http://www.heritage.org/Index/Country/Burundi.
Hesse, H. 2009. Export Diversification and Economic Growth. In Breaking into New Markets:
Emerging Lessons for Export Diversification, ed. R. Newfarmer, W. Shaw, and P. Walkenhorst, 5580.
Washington DC: World Bank.
Hidalgo, C., B. Klinger, A.-L. Barabsi, and R. Hausmann. 2007. The Product Space Conditions the
Development of Nations. Science 317 (5837): 48287.
Hugon, P. 2003. Les conomies en dveloppement lheure de la rgionalisation. Paris: Karthala.
Imbs, J., and R. Wacziarg. 2003. Stages of Diversification. American Economic Review 93 (1): 6386.
IMF (International Monetary Fund). 2008. Burundi Selected Issues: Assessing External Competitiveness
in Burundi. IMF Country Report No. 08/292, Washington, DC.
. 2009. Burundi: Poverty Reduction Strategy PaperAnnual Progress ReportJoint Staff

Advisory Note. IMF Country Report No. 09/91, Washington, DC.
- 121 -
Kaufmann, D., A. Kraay, and M. Mastruzzi. 2009. Governance Matters VIII: Aggregate and Individual
Governance Indicators 19962008. Policy Research Working Paper 4978, World Bank, Washington,
DC.
Krishna, P. 2003. Are Regional Trading Partners Natural? Journal of Political Economy 111 (1): 202
31.
Krugman, P. 1991. The Move toward Free Trade Zones. Federal Reserve Bank of Kansas City.
Economic Review (November/December): 524.
. 1995. Development, Geography, and Economic Theory. Cambridge, MA: MIT Press.
Lall, S. 2000, The Technological Structure and Performance of Developing Country Manufactured
Exports, 198598. Oxford Development Studies 28 (3): 33769.
Lederman, D., and W. Maloney. 2009. Trade Structure and Growth. In Breaking into New Markets:
Emerging Lessons for Export Diversification, ed. R. Newfarmer, W. Shaw, and P. Walkenhorst, 3954.
Washington DC: World Bank.
Lederman, D., M. Olarreaga, and L. Payton. 2009. Export Promotion Agencies: Strategies and Impacts.
In Breaking into New Markets: Emerging Lessons for Export Diversification, ed. R. Newfarmer, W.
Shaw, and P. Walkenhorst, 21122. World Bank, Washington, DC.
May, J. 2009. Rapid Population Growth in Burundi: Implications for the World Bank. Unpublished
manuscript. World Bank, Washington, DC.
May, J., and A. Kamurase. 2009. Demographic Growth and Development Prospects in Rwanda:
Implications for the World Bank. World Bank, Washington, DC.
McGillivray, M., and S. Feeny. 2008. Aid and Growth in Fragile States, Research Paper No. 2008/03,
United Nations UniversityWorld Institute for Development Economics Research, Helsinki, Finland.
Midende, G. 2009. Etude sur les exploitations minires artisanales du Burundi. Bujumbura, Burundi.
Mo Ibrahim Foundation. 2009. 2009 Ibrahim Index: Burundi. London, UK.

http://www.moibrahimfoundation.org/en/media/get/20091004_burundi.pdf.
Naud, W. 2007. Geography and Development in Africa: Overview and Implications for Regional
Cooperation. Discussion Paper No. 2007/03, United Nations UniversityWorld Institute for
Development Economics Research, Helsinki, Finland.
Ndulu, B., L. Chakraborti, L. Lijane, V. Bamachandran, and J. Wolgin. 2007. Challenges of African
Growth: Opportunities, Constraints, and Strategic Directions. World Bank, Washington, DC.
Nielsen, H., and D. Madani. 2010. Potential Benefits and Risks of Increased Aid Flows to Burundi
Research Working Paper 5180, World Bank, Washington, DC.
Nkurunziza, J., and F. Ngaruko. 2002. Explaining growth in Burundi: 19602000. Working Paper No.
162, Centre for the Study of African Economies, Department of Economics, Oxford University, Oxford,
UK.
- 122 -
OTF Group. 2008. Vision et Stratgie Industrielles et Commerciales du Burundi: Rapport Final.
Economic Management Support Project, Bujumbura, Burundi.
Page, J. 2008. Rowing Against the Current: The Diversification Challenge in Africas Resource-Rich
Economies. Global Working Paper No. 28, Brookings Institution, Washington, DC.
. 2009. Africas Growth Turnaround: From Fewer Mistakes to Sustained Growth. Commission
for Growth and Development Working Paper No. 54. World Bank, Washington, DC,
Population Reference Bureau. 2009. 2009 World Population Data Sheet. Population Reference Bureau,
Washington, DC. http://www.prb.org/Publications/Datasheets/2009/2009wpds.aspx.
Rauch, J. 1999. Networks Versus Markets in International Trade Journal of International Economics
48 (1): 735.
Republic of Burundi. 2006. Poverty Reduction Strategy Paper. September, Bujumbura.
. 2007a. Assistance technique au Service Charg des Entreprises Publiques (SCEP) en matire de
rorme, restructuration et privatisation des Socits Participation Publique. Projet dAppui la Gestion
Economique, February, Bujumbura.
. 2007b. Planification base sur les objectifs du millnaire pour le dveloppement : Rapport
gnral dvaluation des besoins pour latteinte des OMD. Ministry of Development Planning and
National Reconstruction and United Nations Development Programme, March, Bujumbura.
. 2008a. Etude diagnostique sur la gouvernance et la corruption au Burundi: Rsum Excutif.

Ministre la Prsidence Charg de la Bonne Gouvernance, de la Privatisation, de lInspection Gnrale
de lEtat et de lAdministration Locale, May, Bujumbura.
. 2008b. Stratgie Agricole Nationale 20082015. Ministre de lAgriculture et de lElevage,

July, Bujumbura.
. 2009. Stratgie Nationale de Gouvernance et de Lutte Contre la Corruption: Guide des

Consultations Sectorielles. Ministre a la Prsidence Charg de la Bonne Gouvernance et de la
Privatisation, in cooperation with the World Bank and the Institut de Dveloppement Economique,
Bujumbura.
Rhata, D., S. Mohapatra, and A. Silwal. 2009. Migration and Remittance Trends 2009. Migration and
Development Brief 11, World Bank, Washington, DC.
Ross, M. 2008. Oil, Islam, and Women. American Political Science Review 102 (1): 10723.
Summers, L. 1991. The Move toward Free Trade Zones: Comment. Federal Reserve Bank of Kansas
City. Economic Review (November/December).
UNIDO (United Nations Industrial Development Organization). 2009. Industrial Development Report
2009. Breaking In and Moving Up: New Industrial Challenges for the Bottom Billion and the Middle-
Income Countries. Vienna, Austria: UNIDO.
United Nations. 2009. World Population Prospects: The 2008 Revision. Department of Economic and
Social Affairs, Population Division, New York. http://www.un.org/esa/population.
- 123 -
von Grebmer, K., H. Fritschel, B. Nestorova, T. Olofinbiyi, R Pandya-Lorch, and Y. Yohannes et al.
2008. Global Hunger Index: The Challenge of Hunger 2008. Welthungerhilfe, International Food
Policy Research Institute, and Concern Worldwide, Bonn, Germany.
World Bank. 2007. Population Issues in the 21st Century. The Role of the World Bank. Health,
Nutrition, and Population Discussion Paper, Washington, DC.
. 2008a. Burundi: Une valuation du climat des investissementsRapport Final. Report

prepared by Etude Economique Conseil Canada, Montreal.
. 2008b. Country Assistance Strategy for the Republic of Burundi for the Period FY09FY12.
Report No. 44193-BI, Washington, DC.
. 2008c. Non-Tariff Measures on Goods Trade in the East African Community: Synthesis
Report. Report No. 45708-AFR, Washington, DC.
. 2008d. Regional Integration Assistance Strategy. Final Report, March 11, Washington, DC.
. 2008e. Republic of BurundiPublic Expenditure Management and Financial Accountability

Review (PEMFAR): Improving Allocative Efficiency and Governance of Public Expenditure and
Investing in Public Capital to Accelerate Growth and Reduce Poverty. Report No. 42160-BI,
Washington, DC.
. 2009a. Doing Business 2010: Burundi. World Bank, Washington, DC.
. 2009b. Republic of Rwanda: Fifth Poverty Reduction Support Grant, Program Document.
Report No. 46455-RW, World Bank, Washington, DC.
- 124 -

Burundi Country Econ Policy

Încărcat de

Informații document

Drepturi de autor

Formate disponibile

Partajați acest document

Partajați sau inserați document

Opțiuni de partajare

Vi se pare util acest document?

Este necorespunzător acest conținut?

Drepturi de autor:

Formate disponibile

Burundi Country Econ Policy

Încărcat de

Drepturi de autor:

Formate disponibile

Public Disclosure Authorized

Report No. 51880-BI

Report No. 51880-BI

Public Disclosure Authorized

Public Disclosure Authorized

Public Disclosure Authorized

Currency Unit Burundi Franc

Government Fiscal Year

Weights and Measures

Abbreviations and Acronyms

AfDB African Development Bank

Vice President: Obiageli Katryn Ezekwesili

ACKNOWLEDGMENTS ................................................................................................................................ vii

EXECUTIVE SUMMARY ............................................................................................................................. viii

CHAPTER 1 MACROECONOMIC AND FISCAL DEVELOPMENTS........................................................... 2

CHAPTER 2 OBSTACLES TO ECONOMIC GROWTH AND EMERGING OPPORTUNITIES FOR

CHAPTER 3 THE GOVERNMENTS OPTION TO TACKLE OBSTACLES AND REAP OPPORTUNITIES

A. An Infrastructure Action Plan ........................................................................................................ 52

CHAPTER 4 IMPLICATIONS FOR THE MACRO-FISCAL SITUATION................................................... 84

REFERENCES ........................................................................................................................................ 120

BURUNDIS ECONOMY YESTERDAY AND TODAY

4. A modest recovery has started. FigureES.1:RealGDPGrowthandGDPperCapita,19622008

below the average of the East African 20%

digits between 2005 and 2007. Inflation,

WHAT BLOCKS ECONOMIC GROWTH IN BURUNDI?

9. Burundis exports are undiversified, consisting mainly of low-value, unprocessed primary

HOW CAN PUBLIC ACTIONS INCREASE ECONOMIC GROWTH?

Closing the Infrastructure Gap

Gaining from Regional Integration

24. A comprehensive approach to regional integration, prioritizing and sequencing actions

Improving the Business Environment

Promoting New Sources of Growth

28. Key recommendations include the following:

29. Diversification out of subsistence agriculture is particularly important. Given demographic

30 as inadequate revenue mobilization could result in

political violence amounted to about 62 percent

A. ECONOMIC GROWTH: HISTORICAL PERSPECTIVE

A2. Slow Recovery after 2000

sector, however, primarily reflects an 2

1.19 A combination of official flows Figure1.5:PrivateandPublicInvestment,19972008(percentofGDP)

foreign direct investment (averaging less 6

factors, including the protracted war and 2

1.20 Public expenditure contracted Source:AuthorscalculationsbasedonInternationalMonetaryFunddata.

domestic economic activity, destruction of 0

A3. Comparison with Rwanda and Sierra Leone

Figure1.7:RealGDPGrowthduringPostconflictPeriodsinBurundi, 1.27 Burundi has not experienced

B. BURUNDIS PUBLIC FINANCES: AN OVERVIEW

B1. Structure of Public Expenditure

1.36 Burundi relies heavily

B2. Public Finance Management Reform

A. OBSTACLES TO ECONOMIC GROWTH

A1. Geographic and Economic Isolation

A2. Infrastructure Deficit

A3. Poor Agricultural Policy

A4. Narrow and Undiversified Export Base

decreased significantly to below 0.2 since Guatemala

demonstrates that Latin American countries 0 0.2 0.4 0.6

19851989 2005 2006

Classification of Burundis Exports

A5. Weak Private and Financial Sectors

Supply-Side Constraints: Finance and Workers

Poor Regulatory Environment

Box 2.2: Burundis Tax Regime

A6. Poor Governance