Report No. 44314- A 0 ANGOLA Investment ClimateAssessment October 2007 Regional Program forEnterpriseDevelopment(RPED) Finance and Private Sector (AFTFP) Afiica Region TABLEOFCONTENTS LISTING OFBOXES. FIGURESAND TABLES ...................................................................m ACKNOWLEDGEMENTS .........................................................................................................V EXECUTIVESUMMARY ...........................................................................................................1 INTRODUCT'ION .........................................................................................................................5 Objective and Rational of the Report........................................................................................... 6 1 SOCIALCONTEXTAND MACROECONOMICBACKGROUND .......................11 1.1 Socio-geographicCharacteristics ................................................................................... 11 1.2 Policy choicesand structural changes............................................................................ 15 1.3 The Economic Outlook:OilIsWellThat EndsWell ..................................................... 17 2 THEBUSINESSENVIRONMENT ..............................................................................22 2.1 Formal Sector: Perceived Constmints ............................................................................22 2.2 Electricity ......................................................................................................................2 6 2.3 Corruption and Crime.................................................................................................. 2 8 2.4 Regulatory Eramework.................................................................................................. 32 2.5 Transportationand OtherConstrai$s............................................................................ 34 3 MICROFIRMS ...............................................................................................................37 3.1 Constraints to Business..................................................................................................37 3.2 Comparisonbetween formal sectorand micro firms.....................................................42 4 FACTOR MARKETS:FINANCIALSECTOR,LABORAND LAND MARKET .........................................................................................................................45 4.1 The Financing of Firms inAngola ................................................................................. 45 4.2 The Formal Labor MarketinAngola............................................................................. 50 4.3 Land Market ................................................................................................................. -57 5 MANUFACTURINGAND FIRM'S PRODUCTIVITY .............................................60 5.1 Labor Productivity and Labor Cost................................................................................60 5.2 Total Factor Productivityand Investment Climate Determinants..................................63 6 SYNTHESISOFRESULTSAND POLICYRECOMMENDATIONS .....................66 6.1 Constraintsto Business.................................................................................................. 66 6.2 FinancialMarket.........A. ................................................................................................ 66 6.3 Productivity ...................................................................................................................67 6.4 Policy Recommendations.............................................................................................. 67 ANNEXA:TECHNICALAPPENDIXFOR PRODUCTIVITYcALcULA~~ONS ...........76 LISTING OF BOXES, FIGURES AND TABLES Boxes: Page 1 Whatisan InvestmentClimateAssasment? 12 Listingof Figures: Oil DependencyforSelectedCountries Evolutionof Angola's Real GDP Per Capita, 1960-2004 Angola: Real GDP growth(1990-2006) Progressin MacroeconomicIndicators curbingInflation Tradable and Nontradable Inflation Rates 2.1 Top 6 Major or Very SevereConstraintsasReportedby All Formal SectorFinnsin Percentages 2.2 Costsof Securityand Theft-International Comparisons 3.1 Percentageof FirmsReporhngMajor orVery SevereConstraints(Top6Constraintsfor Forrid Sector)-Formal SectorversusMicro Firms 3.2 Indmx-2Costs-Formal SectorversusMicroFinns 5.1 Labor Productivity-International Comparison 5.2 Labor Cost per Employee (U.S. Dollars)-ManufacturingSector 5.3 Unit Labor Cost-International Comparison - 5.4 Capital perEmployee-International Comparison 5.5 Total Factor Productivity-InternationalComparison L i g of Tables: SampleDescriphon Basic Povertyand Social Indicators Compositionof GDP by Sector, 1966-2004 Major or Very SevereCons@ahts as Reportedby All Formal SectorFirmsin Percentages Major or Very SevereConstraintsas Reported by ManufacturingSectorFirmsin Percentages Major or Very SevereConstraintsas Reported by Finns-ComparisonAcross Countries IndirectCosts-ManufacturingSector IndirectCosts-Manufacturing Sector-ComparisonacrossCountries I n h c t Costs-All Formal Sectors infrastructureIndicators-All Formal Sectors Inhstructure Indicators-Comparison across Countries Common Percc@on Index - 2007 Perception of Govenunentand Regulations-All Formal Sectors Court System-All Formal Sectors SecurityServicesand SecurityExpenditure-ComparisonacrossCountries-Manufacturing Sector RegulatoryBurden -All Formal Sectors RegulatoryBurden-Comparison acrossCountries LicensingProcess LicensingProcess-Comparison acrossCountries Startinga Business Licenses InventoryHoldings of Most important Input- Manufacturingand All Sectors Percentage of InputsDeliveredby Road -ManufacturingSector Origin of Inputs -Manufacturing Sector Customs-ManufacturingSector-Comparison acrossCountries TradingAcrossBorders Major or Very SevereConstraintsas Reportedby MicroF m s in Percentages Indirect Costs-Micro Firms hhstmcture Indicators-Micro Finns Perception of Govenunent and Regulat~ons-MicroFirms Court System-MicroFirms SecurityServicesand SecurityExpenditure-MicroFirms LicensinglRegistration-MicroFirms Percentage of FirmsReportingMajor or Very Severe Obstaclesto Registeringa Business- Micro Films Regulatory Burden -Micro Firms LicensingProcess-Micro Firms Regulatory Burden -Formal Sectorversus MicroFirms Perception of Govenunent andRegulations-Farmal Sectorversus MicroF h s Mastructure Perceptions-Formal vs MicroFirms Sourcesof Short-term Financein the Formal Sector Sourcesof Long-term Financein the Formal Sector Sourcesof Finance-Formal Sectorversus MicroFinns Sourcesof Short-term Financing -Comparison with other Countries Sourcesof Long-term Financing -Comparisonwith otherCounties Access to Credit in the Formal Sector Access to Credit-Comparisonwith other Countries Collateral -Formal Sectorversus MicroF h s Collateral-Comparisonwith other Countries Cost of debt and duration-Comparisonacross Counties Reasons for Not Applyingfor Loans Formal Sector - Loan ApplicationlRejection Labor Regulations Employment: Full-time PermanentWorkforce-All Formal Sectors Employment: Full-time SeasonalRemporary Workforce-A1I Formal Sectors Description of Workforce- ManufacturingSector UnionizedWorkforce-ManufacturingSector Average Educational Attainment of ProductionWorkers-ManufacturingSector Average Educational Attainmentof ProductionWorkers-Comparison across Countries- ManufactunngSector Firms offeringTraining-ManufacturingSector Firms offeringTraining-ManufacturingSector-Comparisonacross countries Absenteeism -All Formal Sectors HIVPreventionActivities-All FormalSectors Determinants of Earnings-ManufacturingSector Labor Productivity(US Dollars)-Manufacturing Labor Costsper Employee(US Dollars)-Manufacturing ProductionFunction Estimates Total Factor Productivity Extended Production Function Estimates Effects of Accessto Finance Effects of Electricity Effects of Business Licenses Effects of Crime Effects of Cormphon Effects of Transportation ACKNOWLEDGEMENTS Thisreport waspqared by a team leadby GiuseppeIarossi (AFTPS)and comprisingof FranciscoGalraoCarneiro (OPCCE), Ricardo Gonplves (AFTPS), Sofia Silva (AFTPS),and Tania Olivein (AFTPS).David Shiferaw(AFTPS)provided invaluableresearch assistance. OlivierJ. Lambert (AFTFS),ThomasMuller (AFTFS),Abdelmoula M. Ghzala (AFITR),and StephanKL. von Klaudy (FEU) contributedto the policyrecommendations. Commentsand suggestionswere receivedby Maria MargaridaBaessa Mendes (AFMAO),Alberta ChuecaMora (AFMAO),Gilbertode Barros (AFTPS), Vjgayanti Desai (CSMPF), James Habyarirnana (AFTPS), MelanieMbuyi (AFTPS), Joyce Msuya (CAFSC),ChristopherPorter (AFMAO), Manju Shah(AFTPS),DileepWagle (AFTPS), and the participantsto the conceptnote review and ICAreview sessions. Theauthorsare also grateful for commentsreceivedat the workshopsheld in Luandain July 2007 by representatives from the Govenunent ofAngola,The Catholic University, private large companies,SMEs,and the donor community,includingTharaldsenPaul Sverre (Norway Embassy).Officialpeer reviewers were Jose GuilhermeReis (LCSPF), BabatundeOnitri(CAFSS), Joyce Msuya (CAFSC), and SeptimeMartin (AliicanDevelopment Bank). We are grateful to USAID for contributing to the costof data collection. EXECUTIVE SUMMARY Sincethe peace accordsof 2002,Angola has witnessed a surge in grossdomesticproduct. The growth rate of real GDP increased fiom 3.4% in 2003, to 15% in 2006 due in large part to the increasesin oil productionand revenue. Increased oil production fiom new oil fields is expectedto push Angola's GDPgrowth rate to approximately30% in 2007. Angola's macroeconomic stabilizationefforts have been commendablysuccessful.Following the adoption of the September 2003 stabilization program, Angola's annual price inflation has consistently fallen fiom its 2002 level of 100%. In 2005, Angola's 12 month consumer price inflation was 18.5%. In addition, significant increases in oil revenues have allowed for a governmentbudget surplusof about7% of GDPin 2005. The lack of linkages, however, betweenthe capitaland technologyintensiveoil sector and the rest of the economy has meant that non-oil sectors have not experienced equivalent growth. Moreover, Angola's heavy dependence on the oil-sector, which contributed to over 91.92% of exports in 2004 and over 80% of the state budget, and the dramatic increase in Angola's real effectiveexchangerate', has nnpeded the developmentofAngola's manufacturing sector. Despite the favorable outlook in terms of mineral wealth, the Angolan economy will not enter a path of sustainable shared growth without some necessary structural reforms. Because of the long civil war and of the effects of the strong dependence on oil and diamond revenues, the private sector has not evolved outside of the mineral sectors and the quality of the country's institutions remains low. The combination of this state of affairs creates constraints to private sector developmentand forthe diversificationof the economy. Angola's Investment Climate Assessment (ICA) looks in detail at factors constraining the private sector, such as the effective hctioning of the product markets, financial market, infrastructureservices,and the country's legal, regulatory, and institutionalenvironment. The ICA also provides the analytical framework to identi@reform priority by linking constraints to h level costs and.productivity. TheAngola ICA report is largely based on results from a survey of 310 firms in the non-oil private sector, representing a population of 839 firms in the manufacturing and services industry in Angola.' The sample included small medium and large f m s located in Luanda, Benguela and Huambo. The survey also included 115 micro-hs. As benchmarks, the survey utilizes comparator countries and comparator country-groupings. The comparator countries include: Algeria, Democratic Republic of Congo (DRC), South Africa, and Namibia, while comparator country-groups include: low income, Sub-SaharanAfrica, and resource-richcountries. Key Investment ClimateConstraints Access to credit is viewed asoneof the most significantconstraints inAngola with over half of surveyed firms reporting it as a major or very severeconstraint. Smalland medium-sized ' IMFestimates point to an over40% increase in Angola's real effectiveexchangerate forthe twoyears ending2005 Source: INE2003 census. firms,and firmslocatedoutsideofLuandareportedaccessto creditto be particularly constraining. Retainedearningsconstitutethemain sourceof working capital and long-term financeof firms in Angola. Thebanking sectoraccountsfor only 1%and 4% of a typical h ' s total short-termand long-term financing needs respectively.Firms inAngolarely more on internal b d sand less on banks thanfirms in Sub-SaharanAfiica. Similarly,only 2% of firms inAngola have an overdraft, compared toand averageof 35%inthe continent. Cost of finance is not a problem forAngola's firms. Interest rates and collateralrequirements are comparatively attractive in Angola compared to the other countries. Collateral requirements and interest rates on loans are more favorable in Angola than in the DRC, SouthAfiica, Algeria, Namibia, and the country-group comparators. However with about 79% of loan applications in Angola rejected, the problem of accessto credit isrelated to availability and not cost of credit. The complexity of the applicationprocess and unacceptability of collateral are the principal reasonsforlow banking penetrationinAngola. Over 80%of firms in Angola donot apply for a loan. One third of them do no apply because the application process is considered too complex. Similarly of those that applied for a loan over one third was rejected because the collateral provided wasconsidered unacceptableby thebank. Electricity is the main driver of indirect costs and affects the manufacturing sector more than other sectors. Almost half of the firms consider availability of energy a major bottleneck. Urnliability of electricity costs on average 4.6% of sales to a typical Angolan firm. In Angola 84% of all firms experienced power outages, on average 8 times per month. Electricity-related indirect costs affect more heavily large firms (9.8% of total sales), domestic h s (4.8% of total sales) and firms based in Luanda (4.9% of total sales). Generaton, a costly alternative,thus, are owned by almost 70%of h s andprovide approximately31%of electricityneeds. Manufacturingfulns and firmsoutside of ~uandacontend transportationto be a signif~cant constraintto business. Manufacturingfirms in Angola lose 2.1% of their production in transit, a percentage higher than in all comparator countries (except Algeria). In addition, with imports representing40% of manufacturing sector inputs, the average number of days to clear custom is 28 days comparatively the highest than in Sub-Saharan Afiica,. Moreover, for firms outside of Luanda limitedaccesstoroads isproblematic. Corruption is considered a signif~cantbottleneck by fvms In Angola. Close to 40% of them view corruption as a major constraint,particularlylarge h s . Moreover,40% of largefirmsreport informal payments or gifts to be common '%I get things done". Corruption appears to be problematic within the court system as almost 90% of firms believe it to be unfair, partial, and corrupted.Furthermore ICA results indicate that only 23% of firms agree with the statement that laws are consistently and predictably interpreted. In the manufacturing sector, firms in Angola incur indirect costs due to corruption for 3.4% of sales, which is lower than Algeria (7.9%) and DRC (5.1%), but greater than low income (3.1%), Sub-SaharanAfrica (2.5%), and resource-rich countries (1.9%).Angola's low ranking of 147 out of 169 in Transparency International's 2007 CorruptionPerceptionIndex confirmsthe ICA's results. Although demonstrating a positive trend since 2000, Angola's regulatory environment remains problematic for business. On average close to 8% of senior management time is spent with governmentregulations.Although comparablewith other countries in the region, this burden falls disproportionately on large (14.6%) and foreign (10.2%) firms, as well as firms in the retail sector (12.9%).Business licensingand permits are reported as major or very severe constraintsby a greater proportion of firms in Angola than low income, Sub-SaharanAfrica, and resource-rich comparator-groupsof countries. In the Doing Business indicators of 2008,Angola is rankedin the last decile in regulatory quality. Although the establishment of the Guichet Unico has improved Angola's performances on regulatory requirements to begin a business, the costs needed to start a business in Angola are higher than all comparator countries (except the DRC) and country-group averages. Obtaining licenses is still a lengthy process for some firms in Angola. Construction-related permits necessitate more than 3 months for large firms, but only 27 days for small finns. In addition, obtaining an import license takes on average 23 days for small firms,but requires 2 months for large firms.The time and costs needed to build a warehouse in Angola are the highest in the continent(with the exception of DRC regardingcosts). These constraints have a direct impact on costs and productivity of Angolan firms. Value added per worker in manufacturing enterprises in Angola is relatively high at around $5,000. Angola performs better than most comparator countriesexcept SouthAliica, Namibia, Swaziland, and Botswana. On the other hand, Angola has high labor cost, at approximately $3,000 second only to Namibia and SouthAfrica. The shareof value added represented by labor cost inAngola is the highest in all comparator countries.Addressingthe major constraints faced by firmsinAngola will improve their productivity. Electricity, bribes, security and theft account for up to 10-12% of saleslost in 2006. Labor productivityand total factorproductivitywill increase by 6-8% if access to credit is improved, electricity is more reliable, or corruption is reduced. Similarly our survey data shows that sirnpliflmgthe regulatory environment can have an even higher impact on firm productivity. Recommendations: 1. Electricity Improve the monitoring and regulation of the electricity system. Review options for private participation in management contmcts and investments. Separate generation, transmissionand distribution.Ensure operability of independent regulator responsible for price setting. Increaseenergy generation. 2. Credit Enhance credit information infrastructure. Upgrade corporate registries, collateral registries, and public record systems.Facilitate the establishmentof private credit bureaus. Reform collateral law and improve court efficiency. Strengthen accounting framework, enhance disclosure requirement. 3. Corruption & Regulation Declare political will to fight corruption, make resources available and establish an Anti Corruption Agency. Increase effectiveness of GUE by reducing the cost of starting a business. Reduce costs to execute notary deeds. Reduce time required to obtain Commercial Operations Permits and the registration with the Registry of Companies. Build up capacity for the Voluntary Arbitration Law. Shorten time to obtain licenses fiom the Provincial Governor and real estate registry. Reduce sot of inspections. Establish informationsysteminthejudiciary. INTRODUCTION Successivearmed conflicts,which lastedalmostthree decadesafter independence,have devastated Angola and its economy. However, since the peace accord of April 2002, Angolanshave begun a transition toward national reconciliation and lastingpeace. For the Government of Angola (GoA), one of the main challengesaheadis to reconstructthe economyand reunite societyafter a war that has left its most visible marks on the millionsof displacedthat are returningto their areas of origin and demobilized former combatantsthat will need to be reintegrated into society.Peace in Angola has come hand in hand with a surge in GNI~per capita over the past years: per capita GNI rose from USD 470 in 2001 to about USD 1,980 in 2006, primarily as a result of increased oil production and revenue. Even though the national income is currentlyabove the average level in Sub-SaharanAfrica,Angola was nonethelessranked 161st out of 177countries in the UN Human Development Index (HDI)of 2 ~ 6Thisunderscores the magnitude ofAngola's challengesin the . ~ social sphere. Over the period between 1990 and 2001, GDPgrowth exhibitedan irregular pattem that is largely explained by fluctuations in domestic oil production and its international market price. In 2002 growth peaked at 13 percent, thanks largely to the peace agreement and to strong growth in oil production. Growth, however, slowed in 2003, but surged by 11.2 percent in 2004, largely reflecting, once again, developments in the oil sector.Because of few linkages between the capital and technology intensive oil sector and the rest of the economy, non-oil sectors have not shown such growth rates. The economy's dependence on oil is further demonstrated by the fact that it contributed to nearly 50% of all GDP in 2004, whereas services contributed 32% and manufacturing just 4.2%. Additionally, oil accounted for 91.92% of exports in 2004 and contributionsofthe oil sectorto the statebudget exceeded 80percent. Angola is one of Africa's most resource-rich countries. It is Sub-Saharan Africa's second largest producer of oil and the world's fourth largest producer of diamonds,with over 12%market share. In addition,Ahgola is endowed with other minerals,plenty of water for hydroelectricpower and irrigation, vast fertilelands, and abundanttimber and marineresources. The capital-intensive oil sector, mainly located offshore, accounts for about 50 percent of GDP. The formal non-oil economy is dominatedby trade and commerce and non-tradable services. The agricultural sector's contribution to GDProse from 8percent in 2004 to 10percent in 2006, but it is still far from the 24 percent levels of 1991. Much of the explanation for the rise in food production has been the return of displacedpeople after the war. The informal sector has become extremely important mainly due to the disruption caused by the civil war. It is estimated that almost70percent of existingjobs are in the informal sector. 3 GrossNational Income(GNI)pzrcapitausingtheA t h method ( c a n t US$). 4 The UN ednmes that 154out of 1,000infanedie in their fnstyear of life compared with 91 forSub-SaharanAfrican as a whole and 41 percent of all children under five are chronicaUy malnourished The Survey on the FamilyAggregate about Revenues and ExpendihrresO R ) underbkenin 2000-2001,revealed thatthepoverty incidencewas 68percent ofthepopulationwih 28percentofindividualsin extreme poverty situationordestitute. Rural poveltyhas been estimatedat 94percent (ECF',2003). The GoA has been taking stock of the current situation and released in 2004 the final version of The Estratkgia de Combate a Pobreza (ECP), Angola's Poverty Reduction Strategy Program. The ECP is in its final stages of revision and finalization. It states as its major objective the need to consolidate"pace and national unity through the improvement of life conditionsof the Angolan citizens, those that are more vulnerable, by motivating them to actively participate in the social economic development process". For this, the GoA is determined to 'reduce the incidence of poverty by half by 2015 from its 2000-2001 level of 68 percent and acknowledges that although supportfrom other donors isneeded, private sectorparticipation is essential. The World Bank's Interim Strategy Note (ISN) of February 2005 was set to support the government's program for 2005-2006' and emphasizedthe need to encouragethe Private Sector's role through a stronger publiclprivate dialogue h e w o r k and a more propitious operatin % environment for the private sector. The World Bank's ISN of May 2007 reinforced this need , whilst at the same time recognizing that there has been progress: "[the GoA] ... has adopted legislation to streamline the regulatory framework and clan@ land rights and has improved customsprocedures (reducingthe averagepaperworkprocessingtime from 25 days in 2000 to 5in 2006). It has also taken steps to improve access to financial services, including microfinance, by allowingnew entrantsintothe market. Investmentsin infrastructure,includingroads, railways, and electricitygenerationandtransmission will alsoimprovethe investmentclimate." The Regional Program on Enterprise Development (RPED) of the World Bank Africa Private SectorGroup agreedto conductan InvestmentClimateAssessment (ICA)inAngola ObjectiveandRational of the Report Work to improve the investmentclimate is recognized as a key pillar of World Bank Group work to promote economic growth and poverty alleviation in developing count~ies.~The main focus of Investment Climate Assessments (ICAs) is on microeconomic and structural dimensions of a nation's business environment, viewed in an international perspective (See Box 1). To this end, ICAs look in detail at factors constraining the effective functioning of product markets, financial and non-financialfactor markets, and infrastructure services,includingin particularweaknesses in an economy's legal, regulatory and institutional h e w o r k . ICAs also provide the tools and 5 The World Bank's InterimSbategyNote 0of February2005 was setto supprtthe government's program for 2005-2006arcundthreepillars i) enhancingtransparent governance and intensifying capacitydevelop en^ ii) providingbasic senices (especiallyfor retumeg, excombatanis, and other vulnerable groups) and rehabilifation of emergency -, and iii) supporting brondhsed equimble growth, W a U y through impmvemenisintheenvhmmt forprivatesectorgrowth. 6 The World Bank's Interim Strategy Note of May 2007 is organued around three pillars: (a) strengthening public sector management and governmentinstitutional capcity, @) supponh.lgthe rebuildingof criticalinhastructureand h e improvement of servicedeliveryfor povertyreduction, and (c)promotinggrowthof nonmineralsectors. 7'Thecen!nl challengein reaping greaterbenefiis h m globaJizationlies in improvingthe investmentclimate-that is, in providingsoundregulation of indusm, imludimg the promotion of w m p o n ; in overcoming bureaucratic delay and ineficiency; in fightingcormpion; and in improvingthe qualityofhhtnctums. While theinvestmentclimateis clearly important for large,formalsectorfirms,it isjust as important-ifnotmore so-for smallandmedium enterprises(SMEs),the informalsector, agricultumlproductivity,andthegeneration of off-firmemploymentFore thesereasons,the invesrmentclimateitselfis a key issueforpoverty reduction."Nicholas Stem,ChiefEconomisf March 22,2001. analytical b e w o r k to identify reform priorities in a country's investment climate, by linking constraints to firm-level costs and productivity. The main objective of this report is to develop a better understanding of the investment climate constraints that limit the growth and competitivenessofAngolan firms.In particular, the report seeks to measure in a standardized way the investment climate conditions in Angola, to provide comparisons of these conditions with those prevailing in other countries and regions, and to identify the features of the investment climatethat matter most for competitivenessand growth.While we recognize that issues related to macroeconomic and political stability are crucial for constituting a good investment climate, the focus of this report is on microeconomic issues. Indeed, the importance of political economy and macroeconomic issues is now well understood, and they have been addressed at length in other World Bankreports.* BOX1:Whatis anInvestment ClimateAssessment? Investment climate assessments systematically analyze the conditions for private investments and enterprise growth in a country, d r a ~ on g the experience of local firms to pinpoint the areas wherereform is most needed to improve the private sector's productivity and competitiveness. By providing a practical foundation for policy recommendations and involving local partners throughoutthe process,the assessments are designed to give greater impetustopolicyreforms that can speed the private sector's growth. Producedby the World Bank Groupin close partnership with a public or private institution in each country, the investment climate assessmentsare based on a survey of private enterprises designed to capture firms' experience in a range of areas -financing, governance, regulation, tax policy, labor relations, conflictresolution, hbastructure services, technology, and training, among others. All these are areas where difficulties can add substantially to the costs of doing business. The survey attempts to quantify firms' costs related to the investment climate bottlenecks. Using a standard methodology, the assessments then compares the survey findings with those in similar countriesto evaluatehowthe country's privatesector is competing. The findings of the survey, combined with relevant information from other sources, provide a practical basis for identifjmg the most important areas for reform aimed at improving the investmentclimate. The findingsandpolicy recommendationsemerging from the assessmentsare discussed extensively with the private sector and other stakeholders in the country. This broad dissemination of the findings is aimed at engaging not only policymakers but also business leaders, investors, nongovernmental organizations, and the donor community in shaping the national private sector development strategy, forging consensus on the priorities for reform of the investmentclimate,and layingthe groundwork for concreteresponsesto the problems identified. Source:World Bank (2003),Improvingthe Investment Climate in Bangladesh. See World Bank (2006a), "Angola CountryEconomicMemorandum" (CEM);and World Bank (2002b) ''Cowmy RocurementAssessmentR q f l (CPAR) - The InvestmentClimateSurveyQuestionnaire Two survey instrumentswere used. The firstone is the standardRPEDACAquestionnaire for the formal sectorwhich is composed of three parts: The first part is designed for general managers or business owners and deals with the internal structure of businesses and the investment climate within which they operate, including bureaucratic obstaclesand infrastructureconstraints; The secondpart dealswith finance,production and markets and provides information on business performance as well ashuman resourcesand labormarket issues. The last part includes a small questionnaire for a sample of up to 10 workers per business. This data facilitatesan understandingof the interaction between firm peifonnance/businessclimate and labormarket outcomes. A second survey instrument was used for gathering data fiom micro firms (firms with less than 5 full-time employees).Given the characteristicsof the sample,it is a lighterversion of the standard ICAquestionnaire,mainly looking at investment climate data anda basic setof financialdata. TheInvestmentClimateSample The World Bank Enterprise Survey in Angola targeted establishments located in Luanda, Benguela,and Huambointhe followingindustries: 1. Manufacturing:Food and Beverages 2. Manufacturing:Garment 3. -Manufacturing:Other Manufacturing 4. Retail Trade 5. Rest ofthe universe, including: Construction Wholesale trade Hotels,bars andrestaurants Transportation, storageand communications Computerrelated activities The survey also sampled a selection of micro establishments (establishmentswith less than five full-timepermanentpaid employees) fromthe targeted universe, without stratificationby industry. The total sample size was 425 as described in table 1.1. The population fi-ame for establishments with five or more full-time paid permanent employees consisted in a population of 839 establishments. A list of manufacturing establishments in Luanda was obtained from the INE (National StatisticalOffice), but no suitable lists of establishments could be sourced for the other two strata, or for any strata in Benguela and Huambo. Therefore a list of establishments was compiled in the field by walking throughthe industrial and commercialzones of each of the target cities,and identifying establishments likelyto have fiveormore employees. The resultinglistwas combinedwith the manufacturinglist forLuanda toyield the sample frame. Table 1.1:SampleDescription Size Ownership Location Sector Total Small Medium Large Foreign Domestic Luanda Outside* Economy 425 367 53 5 67 358 348 77 Manufacturing 215 178 32 5 18 197 177 38 Micro-firms 115 10 105 98 17 - * Benguela and Huarnbo Finally, the micro establishment stratum covers all establishments of the targeted categories of economic activity with less than 5 employees. For many reasons including the small size of establishments, their expected high rate of turnovers, the high level of "informality" of establishmentsin many activities and consequentlythe difficultyto obtain trustworthy information fromofficial sources, an aerial samplingapproachtoestimatethe population of establishmentsand selectthe samplein this stratumfor allregionsof the survey. Structureof the Report Section 1provides the socialcontext and macroeconomicaspects of Angola. Section 2 deals with formal firms' perception of the economic environment they operate in. We analyze the results of the ICA survey and other countries' ICA surveys (the comparator countries are Democratic Republic of Congo (DRC), Algeria, South Africa and ~amibia)~in order to (i) identify and compare indirect costs of doing business inAngola, (ii) identi@ and compare firms' perception of the main constraints to business, (iii) identify and compare government-driven constraints, (iv) quantifyand compare the impactof corruptionand crime on firms' operations and (v) identifyand compareinfrastructure-drivenconstraints. Section 3 follows the same structure of section 2, but deals mainly with the micro firms. In particular, not only does it focus onpoints (i) to (v) (see above), but it also compares formal sector firms with micro firms inAngola. 9 The selection of Algeria, d~ Democratic Republic of Congo @RC) and Namhia as comparator cwmies is based on the that they are each resource-rich,and at varying levels of development. In addition, DRC and Namibia are neighboring and possible cornpentorcountries. Cornparism with SubSaharanAfrican avengesare madewheredatais available. Section4 analyzes factor markets, namely labor, finance and land. It relies mostly on ICA survey data (Angola and comparator countries) in order to better understand the hctioning of each market and the main constraintsassociated with it. Section 5 looks at labor and capital productivity of manufacturing h s in Angola and quantifies Total Factor Productivity (TFP). It also analyzes the impact of the various constraintsto business on TFPand laborproductivity. Finally,section 6 containsa synthesisof the main results and ourpolicy recommendations. 1 SOCIAL CONTEXTAND MACROECONOMICBACKGROUND 1.1 Socio-geographic Characteristics The Republic of Angola is, after the Democratic Republic of Congo and Sudan, the third largest nation south of the Sahara. It has an area of 1,276,700 sq. km (including the 7,270 sq. km of the oil-rich Cabindaenclave)and is the largestPortuguesespeakingAfrican country. The country's relative climatic diversity represents an advantage and hints at huge potential for agricultural development. Angola's location in the intertropical and subtropical zones of the Southernhemisphere, its proximity to the sea and the cold Benguela stream,and its topographical characteristics are the factors which create two distinct climate regions with two seasons: the dry and cool season (from June to September) and the hot and humid season (from Octoberto May). The northern region from Cabinda to Ambriz has a humid tropical climate with heavy rainfall, while the region from Luanda to Namibe (Mqamedes) has a moderate tropical climate, with the rainfall reduced on the coast by the Benguela wind stream.The southern stripbetween the plateau and Namibia has a desert climate, given the proximity to the Kalahari, with irregular rainfall between 600 and 1000 mm.annually. Temperatures average 23 degrees C in the north and the coastal areas, and 19degrees C in the interior. The relative climatic diversity, due to variations of altitudes across the country, allows for the growth of crops from both tropical and relatively more temperate zones. The vast and diverse territory hosts a large concealed economic potential. Among the abundant natural resources there is plenty of water that provides for hydroelectric power plants and irrigation; amid mineral resources there are abundant oil, diamonds, iron, quartz, ornamental stones and phosphates. In the Cabinda region, very dense forests predominate (Maiombe forests) with economically important timbers such as black wood, ebony, African sandalwood, and ironwood. With a coastline of 1,650 km, Angola's waters are rich in fish, mollusks, and crustaceans.The main petroleum basins under exploration are located near the coast of Cabinda and Zaire provinces. The main diamond producing area is located in the Lunda Norte province. Unfortunately, due to the nonexistence ofproper and comprehensivegeological surveys, the whole mineral potential ofAngola is, to this date, vastly unknown.1° The Angolan population is young and growing rapidly. Recent population figures are difficult to obtain due to the lack of a full national census.A limited censuswas carried out in the province of Luanda in 1983, which was extended to the provinces of Cabinda, Namibe and Zaire in 1984. War-related problems made it impossible to carry out a national census. In 1980, according to official estimates, the population reached 7.7 million. Though data is scanty, the population was projected to grow at an annual rate of 2.9% duringthe 1980sand 1990s,reachingabout 13million in 2003. Estimates by the US Census Bureau suggest that in 2000 some 6.5 million people or about 62% of the Angolan population were under the age of 24 and that by 2025 that segment of the population would be of approximately 10.8millionpeople, or around 60%of the population. -- 10 SeeAmijo and Costa(1997) andAlves da Rocha (2001) foradetaileddescriptionofAngola'snahnal r e s o w endowmenb. 1 1. Populationdensity(8.6 inhabitantsper IGr?) is low and ethno-linguistic groups are geographically separated. The most populous provinces are Huambo, Luanda, Bie, Malange, and Huila, which together account for more than half of the total population.About threequarters of the population come h m three ethno-linguisticgroups: the Ovimbundu (37%) in the Central plateau region, the Kimbundu (25%) living in a belt extending h m Luanda to the East, and the Bakongo (13%) in the Northwest. In addition, mesti~os(Angolans of mixed European and African family origins) amount to about 2%, with a small population of whites, mainly ethnic Portuguese. Portuguese is both the official language and predominant language, spoken in the homes of about two-thirds of the population, and as a secondarylanguageby many more. The existing social indicators are low and represent a huge challenge to the government and development partners alike. According to both the 2001 Household Income and Expenditure survey (IDR) and the 2002 Multiple Industry Cluster Survey (MICS), approximately 70% of the population lives on lessthan 2 dollarsa day and the majority of the Angolans lack access to basic healthcare.About onein fourAngolanchildren die before their fifthbirthday, !N%of whom perish due to malaria, diarrhea or respiratory tract infections, the maternal mortality rate (at 1,800per 100,000births) is one of the highest in SSA, and three in five people do not have access to safe water or sanitation. The HIVIAIDS prevalence rate is, according to official statistics, relatively low, affectingan estimated 3.7% of adults.' ' However, lack of adequate statisticalinformation and a limitednumber of surveillancecenters suggest that the true prevalence rate may be much higher. In terms of education,primary school enrollment is very low at 56%, and suffers from late entries into school and high repetition and drop out rates. Some 33% of the adult population is currently illiterate, though in rural areas this climbsto asmany as 50% (seeTable 1.2). Table 1.2: Basic Poverty and Social Indicators lpo~ulation(million) 1 14.7 1 General living 1 Pouulation S2Ovears 1 60% 1 conditions are far l~oGlationbelow ~overtvline 1 68% 1 from ideal, even for l~ifeexpectancyat birth 1 42.4 1 the middle class,but Under-five mortality (per I000 live births) 250 they are especially HWAIDS prevelance 3.90% dire for the poor. Population who know where to get an HIV test 23% The long period of Population correctlystating 3main ways to avoid HIV infection 17% civil war destroyed - 33% much of the Adult illiteracy rate infrastructure. Most Maternal mortality rate 1800 Angolans, even in Net primary school attendance rate (1-4th grade) 56% urban areas, do not HDI rank (out of 177countries) 161 have reliable access GDVcapitarank (outof 177countries) 128 to safe water - Gini coefficient(income, 1995) 0.54 only 20% in urban Gini coefficient(income,2001) 0.62 Source:IDR (200011); UNICEF (2003); UNAIDS (2004); UNDP (2005) 1 I UNAIDS,Report an Global Aids Epidmic, 2006 Low estimatesare 1.9% andhigh estimates are 9.4%. areas outside of Luanda have access to it, according to the MICS data. Again the inequality is stark12In Luanda, virtually no one in the two lowest asset quintilesreported accessto safe water, while 40% in the highest quintilereported access. For example,in the comuna of Hoji ya Henda, population of 580,000people, only about 15%of the people are connected to piped water while the rest of the population relies on 18public water points, 14of which are functioning.Electricity also is primarily availableto the rich, most of who rely on generatorsgiven the poorly hctioning infrastructureand fkquentpower outages. In Luanda, 82%of the highest quintile reported having electricity,but no onein the bottom 60%reportedhaving any. Public delivery of social servicesis also skewed in favor of the urban rich. For example, in urban areas in 2002, 50% of women reported receiving some form of trained pregnancy assistance, and thispercentage dropped to 40% in the bottom quintile. However, only 24% in rural areas reported receivingthis assistance,withjust 16%in the poorest 20% of the population.An estimated 17%of (surviving)children under the age of 5 had not received any childhoodvaccinations at all in 2002. Accessto education ispoor as well. Only44%of rural childrenof primary school age (grades 14) , are reported to be in school, and Figure1.IOil Dependencyfor Selected 60% of urban Countries children. This is Oil revenues/Totalrevenues partly because about one third of children start Angola school 1-2 years . -A -..A u late, either -- I Tobaaa -. Gabon because the walk Algeria to schools is long, Mexico Norway Venezuela the family cannot afford the fees, 20 Cameroon children are heeded to work at 0 , home, school I I I I 0 20 40 60 80 100 quality is viewed as poor so parents OilexpoMotalexpo* (%) - 1 do not value the child's education very highly, or the parents simply want to keep the children at home for an extra year. Of those who start, only 46%completeprimary school and enroll in fifth grade. Thewar caused the demiseof the rural economyand the subsequentsharprise in urbanization due to the amval of rural refugees. More than 1million lost their lives during the civil war, 3 million fled to the cities and 400,000 crossed the borders into neighboring countries. Upwards of 45% of 12 According to a recent survey conductedby Development Workshop mLuanda, about 30% of the interviewed households did not have access to basic mhtmchm (e.g pipedlsafe water and electricity),as well as to basic services such as health and education in the vicmity.About 56%have access to some level of inlkWmXm and senices. Only 13%have access to a relativelyhigh provision of in6astructureand services(Devebpment Workshop,2003:44). the population became concentrated in urban areas, with more than half of them in Luanda (Adauta de Sousa, 2003). Furthermore, the current population growth at 2.9% per annum has almost doubled the population since 1980, which is now estimated at 14.7 million. Cross- continental transportation links, which served landlocked neighbors as well as the domestic economy, have atrophied. Infrastructure has also deteriorated in the cities, partly because of warfare and partly because inefficienciesin most pamtatal companies and price control policies depress public utility revenues, which hi1 to recover costs in most services. An estimated $4 billion may be required just to restore the road and bridge network, without which little rural activity is feasible. Until 1975, the country was known as an agricultural producer, not an oil exporter. It was the world's fourth-largest exporter of coffee and one of the largest exporters of staple foods in sub- SaharanAfiica-exporting more than400,000 metrictons of maize annually.These grain exports were produced almost exclusively by smallholders using traditional technologies. Oil had not yet achieved the high production levels of the 1980s and thereafter. Today, the economy is heavily dependent on oil, a capital-intensive -tor with few linkages to other parts of the economy and littleimpact on employment.After 1973,the structure of the economychangedsubstantiallyasthe mining and service sectors increased their share of GDP (Table 1.3). To this date, the Angolan economy remains heavily dependent on the oil sector, which represents nearly 92% of total exports and close to 80% of total government revenues-one of the highest dependency rates in Afiicaandelsewhere(seeFigure 1.1). Table 1 3 CompositionofGDPbySector, 1966 2004 - 1%6 1970 1987 19% 2084 Agriculture,Forestryand Fishery 14.2 9.0 12.6 7.0 9.1 Industry 22.2 29.6 57.5 67.8 58.1 Mining 6.3 10.7 51.0 61.2 49.8 Manufacturing 8.7 10.7 3.7 3.4 4.2 Electricityand water 0.9 0.9 0.3 0.0 0.0 Construction 6.3 7.3 2.5 3.1 4.0 Transport and communications 6.3 5.9 2.7 0 0 Commerce 34.0 30.3 7.2 15.0 15.4 Sources:IV Plano de Fornento 1974-1979, Angola;P d Estatistico, 1988-1991;"Angola: "An hhuductory Review." The World Bank,January 1991; data provided by Angolan authorities to IMF and WB. 1.2 Policychoicesand structuralchanges The economyhas experienceda great deal of ups and downsin its growth path during the last four decades. From 1960 to 1973, GDP per capita at 1996 international prices grew steadily, but collapsed by more than 35% after independence (see Figure 1.2).The period between 1974 and 1976andthe eventsassociated with the fight for independencehad a profound impact onAngola's economy insofar as skilled labor fled the country and organhtional capacity diminished. From 1975to 1997,the economy sufferedseveralshocks,the biggest of them beingthe restart ofthe war at the end of 1992which caused anothermajor drop of roughly 39% in GDPper capita in 1993.In addition, changes in oil prices provoked economic contractions during that period and GDP per capita declined at an averagerate of 2% per annum. From 1997to 2004, GDPper capita grew at an average rate of 4.2% per annum with the biggest increase observed in 2002 (about 13%). In mid-2002 gradualist economic policies were adopted and by 2004 the government managed to bring inflation down and to some extent improve ttansparency in the oil and fiscal sectors. Currently,the level of GDPper capita standsat US$ 1,784,which is stillhalf ofthe level observed in 1973. Figure 1.2 EvolutionofAngola's Real GDPPer Capita, 1960-2004 RealGDP Per Capita (19602004) $US 4000 - 3500 - /-", 3000 - #- Cbilwariwiak~[end1992)a a ailpraJlrtbnafsdedbjwar #+- \ -- 2500 - 2060 - 1500 - I000 - 500 - 1986 0 7 ' 1 1 1 " " ' 1 1 " " 1 1 1 1 " ' 1 1 " " ' 1 1 " " " 1 1 r 1 ~ ~ ~ I ,ps9s'pb46$8$7 & 4 p b s o % ~ s e o % b b a @ ~ % b b a o l .$% \4@,.$ 8 ,** ,$P %@ '$++ -GDP percapnaat anert irt'l pnas --GDP percspita at96~rt'lprices After independence, Angola embarked on a system of centralized economic and political management that only in the mid-1980s started to be reviewed. The transition to a market economy took impetus with an ambitious refom program introduced in 1987 that aimed at stabilizing the economy, securing fiscal discipline, encouraging the development of the private sector, and abandoning the system of administered prices. Progress on this agenda has been sluggish and only after the early 2000s, aided by the fortuitous role played by growing oil revenues, the government succeeded in curbing inflation and achieving an incipient macroeconomic stability. In the face of a favorableexternal outlook, the government now has the opportunityto consolidatethe country's transition to a market economy. The reformsranged frombudgetary disciplineto rescheduling of the externaldebtand adjustments to the planning system. At the beginning of 1989, the authorities approved a "Program of Economic Recovery" (Programa de Renrperaqo Econbmica -PRE) oriented to the two main objectives of starting the process of macroeconomic adjustment and of promoting the rapid recovery of production.The PRE initiatedthe implementationof the economicreforms announced in the Program for Economic and FinancialRestructuring(Programade SaneamentoEcon6micoe Financeiro - SEF), which included the following: (1) the reduction of the budget deficit of the statebudget; (2) the adoption of new solutionsto financethe budget deficit; (3)the restructuring of the financial situation of public enterprises; (4) the reform of domestic credit policies; (5) the rescheduling of external debts; (6) adjustments of controlled prices; and (7) adjustments in the exchangerate. On the structural side, the reforms aimed at increasing the role of the private sector and at gradually reducing the importance of the state in the economy, On matters concerned with structural reforms to increase the efficiencyof the productive system, the SEF envisaged a more important role forthe private sector, more autonomyforpublic enterprises,a revision of the law on foreign investment and improvements in the planning system. The SEF explicitly admitted that smaller public enterprises should be transferred to the private sector and that state ownership should remain concentrated largely in key enterprises with strategic roles. As regards improvements in the planning system, the SEF envisaged achieving better coordinationbetween the Annual Plans, the State Budget, and the Foreign Exchange Budget, and more decentralization of planning activitiesh m the Planning Ministry to the planningorganizationsat regional levels. Despite the appropriatefocus, the reformsdid not yield the expectedresults. Governmentefforts to implement the SEF and the PRE proved unsuccessfbl and between 1989 and 2000 some 12 different macroeconomic stabilization programs were introduced with equally hstrating results. On average,there were 1.2programs per year and each of the programs lasted for a period of 10.6 months. Throughout this period, the main obstaclesto the lastingand effective stabilizationof the economy continued to be the lack of fiscal discipline, the excessivecentralization in the planning system and the resulting bureaucratization of the economy, and the inefficieicy of the state in promoting the growth of productivity. Fiscal deficits remained high during the 1990s making it difficult for the authoritiesto reduce inflation, the oil economy remained the main source of state revenues without productive links with the other sectors af'the economy, and the priorities of the war continued to condition government expenditures, which focused primarily on consumption and military expenditures and neglected social and development spending (notably on health, education,and infrastructure). More recently, the government's economic policy has yielded positive results, but sustainability will demand further reforms. With the implementation of a more rigorous monetary policy, the restriction of monetary financing of the budget deficit since 2002, and the implementationof an active exchange rate policy since September 2003, inflation has been significantly reduced. However, the outlook is subject to sigdicant risks, which must be addressed by government actions. Most importantly, in an uncertain environment for oil production and prices, public expenditure growth needs to be set in a medium-term context to avoid the boom and bust cycles that have undermined stability and development in other oil-producing countries. The following paragraphs highlight progress obtained so far and the tensions that will need to be managed to completethe transition to a market economyandto a viable democracy. 1.3 TheEconomic Outlook:OilIsWell That EndsWell Figure1.3:Angola:Real GDPgrowth 1990-2006 The economic outlook in Angola has been transformed 30% by the peace agreement of 2002 and by positive 20% developments in the oil sector. With the end of violent conflict ,ox and the return of more than 4 million IDPs to their original communities since 2002, 0% agricultural production has picked up and the non-oil -lo% economy has shown signs of a vigorous recovery in Angola. -20% Although official and detailed data on the non-mineral economy is scant, the lively -30% A n g o l a I-LOW SSA Income and vibrant informal economy that is now seen in the streets of Luanda is a visible leading indicator of strong economic performance.There have also been encouragingsigns of recovery in public services,construction, and infrastructure rehabilitation. Oil production, which currently accounts for 55% of GDP, is expected to double to 2 million barrels per day by 2007. Largely as a result of increasing oil production combined with rising internationaloil prices, real GDP is estimated to have grown by 20% in 2005, while the economy outside the mineral sectors is estimated to have grown at an annual rate of roughly 10%overthe last 3years. Currentprojections indicatethat GDPis expected to grow by 15%in real terms in 2006 and by 30% in 2007, one of the highest growth rates in the world (see figure 1.3). The macroeconomic framework for 2007 is highly favorable. In our estimates, total government revenues are expected to remain at a level close to 38% of GDP until 2007. On the expenditure side, spending is estimated to decline from 38.5% of GDPin 2004 to 35.7% of GDPin 2005.13In the pursuit of long-term fiscal sustainability,spending should gradually decline in 2006 and 2007 as a share of GDP.]~Such gradual decline in public spending as a share of GDP is not politically unrealistic insofar as real GDP is estimated to have grown by 20% in 2005 and to grow by an average 24% in 2006-2007 supportedby strongperformance in the oil sector and steadyrecovery of the non-oil economy. ,,The figuresare based on information as of March 2006 and reflect the macroeconomic tiameworkagreedbetween the authorities and the Fund duringthe 2006ArticleN cons&ons. 14 In the draft2006 budget recently finalized by the Government,the authorities are projecting a fiscal deficitof 6.9% of GDP in 2006 and an annual inflation rate of 10°'. The fiscal n u m h in our rnaroeconomic fiameworkare different 6om those presented by the Government in the 2006 budget becauseourh a t e s usehigheroil prices for 2006(S56kmel)thanthose used by the Governmentintheir 2006budget ($45hmel). Figure 1.4:Progress in MacroeconomicIndicators &'wok 011Produdlon - M u a l (2001-2041)and Angola: External Debt a8 s Share of GDP ProJadrd(2005-07) lee8 reee 2000 2001 2002 2ooa 1004 Z D O I ~ ~ Q ~ ~ Z O ( Y ~ O O S Z ~ O O ~ ~ Angols: NonQll FI8ul Deflclt a8 a Share of NonQll Angola:Nreke-Month Growth Rates of GDP Monetary Indlutors tW0 2001 2002 2003 2004 ZOO0 ZOO1 2002 ZOO3 2004 Im M3 Reserve Money There have been commendable successes towards macroeconomic stabilization, but there should be a strongeremphasis on the continuingdeficienciesin policy design and implementation. Figure 1.4 shows progress on a number of macroeconomic variables since the year 2000, including oil production. The stabilization obtained so far, however, needs to be strengthened with improved coordination of the fiscal policy with monetary and exchangerate policies. Thesepolicies need to spell out a consistent strategy to absorb the upcoming oil windfall without inhiiiting growth outside the mineral sectors. To avoid the boom and bust cycles that have undermined stability and development in some other oil-producing countries, new public spending in the future should be set in a medium-term context. In addition, the authorities should consider the adoption of a monetary anchor, with the responsibilities for executing monetary policy defined by the Central Bank in order to guarantee a downwardtrend in inflation even in the face of anexternal shock The root cause of past inflationaryepisodes in Angola was the monetization of its fiscal deficits. Angola's main source of fiscal revenue is the taxation of the oil sector, including the state-owned oil company Sonangol. As a result, fiscal revenues have been excessively vulnerable to international crude oil price volatility and have not always been able to keep pace with expenditures. The insufficient control of public spending, including notably large extra budgetary expenditures and the sizeable operational deficit of Banco Nacional de Angola (BNA), have induced large increases in base money. Additionally,in the past, favored interest groups, including Sonangol, have used arbitrage and other tactics to benefit fiom high inflation, for exam le, by delaying payments in domestic currency for oil and other sales received in hard currency.' Until 2002,this combinationof affairshad actually createdpositive incentives for high inflation.I6 More recently, government's efforts to reduce inflation have been successful. Between 1999 and the peace agreement of 2002, annual consumer price inflation fell from around 300% to around 100% (see Figure 1.5). Following the adoption of a stabilization program in Figure 1.5: CurbingInRaUon September 2003, inflation fell sharply again and by December 2004 the 12- An#ob: Yoar on Yoar Inllailon R.ir month inflation rate had declined to 31%. The improvement was largely due to the government's avoidanceof money creation for deficit finance purposes together with smaller fiscal deficits in 2003 and 2004 (that dropped from 6.5% of GDP in 2003 to 1.5%of GDP in 2004, on a commitment basis) and an estimated fiscal surplus of 6.8% of GDP in 2005. Since 2003, f f ~ g # f ~ f $ f ~ g g f ~ g $ L 1% [ f w s 6 s a s 6 k k k 6 k o k I s a government spending has been increasinglyfunded by &sourcesobtained through direct sales of foreignexchange in excess of $2 billion in 2003 and 2004, respectively, which has increased Angola's external liabilities.17The non-oil fiscal deficit as a share of non-oil GDP has also declined substantially since 2000 from around 130%to close to 63% in 2005. In 2005, the cumulativerate of inflation dropped to 18.5% and the projection for 2006 is of an annual rate of 10%. Monetary aggregates have been kept under controlcontributingto lowerinflation. 15 A detailed discwion of public finance management issues can be found m the Bank's PEMFAR report, published m February 2005 (seeWorld Bank,2005). 16From a political economypoint of view, acentralizedeconomic systemtfiat hensb a d on conbullingmarkenenmuragesthedevelopmentofa wealthy elite which tends to create mechanisms to guarantee the appmpriation of profits vrespectrve of exchange rate and price swing so they are largely indifferentto macroeconomicshocksand theneed to stabilizethe economy. In fact, the wealthycan loseh m economicrefm to the extent that comwtive markets and bansparent public financesshrink the scope f a rent extraction. Some commenmrs argue that this was actually one of the reasonsbeyond the war situationthat couldbe usedtoexplain why refoms!alled through the 1990s(Aguilar,2001). 17 A detailed descnpt~onof the kinds and magnitudes of intervention in the foreign exchange market m Angola is available in the 2004 Angola Economic Reportpublished by the Centerof Shdiesand ScientificInvestigationofthe CatholicUniversityofAngola-seeCEIC (2004). There are F'igure 1.6: lkadable andNontradableinflation consequences associated with the current policy to combat inflation. rates First, removal of excess liquidity from circulation reduces the inflationary lnflatlon Rates,January 2003 July 2005 pressures deriving fi0m money (Annualpercentagechange) expansion permitting a decline in the rate of inflation. Second, the use of foreign exchange for the purpose of mopping up liquidity contributes to stabilizing .the exchange rate. Third, keeping the exchange rate stable implies a corresponding constancy in the prices of imported goods, eliminating inflationary pressures fiom this source. Finally, avoiding a policy ' . that requires an immediate fiscal adjusbn&t in favor of one which I , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , yields price stability, while postponing 1 I. 1 { H i f f t ~ fn a g a h cuts in expenditures, generates an O immediate i d visible success in economic management. These effects combined may help the government to build the necessary political capital for the future, when expenditures will invariably haveto be cut down. Achieving sustainable inflationary stability is also essential to harness the growth of the non-oil economy. It is a well established fact that the inflation component of market-oriented reform policies should be expected to be growth-enhancing. In the case of Angola, this is particularly relevant for thenon-oil economy,given the insulation of the oil-economy due to itsenclavenature. This is so because high rates of inflation can be expected to reduce economic growth through a variety of mechanisms which can influence both the rate of capital accumulation and the rate of growth of total factor productivity. One of suchmechanismsis that high inflationmeans unstable _ inflation and volatile relative prices, which reduces the information content of price signals and distorts the efficiency of resource allocation thus harming the growth of total factor productivity over extended periods. Furthermore, governments that tolerate high inflation have lost macroeconomic control, and this circumstance is likely to deter domestic investment in physical capital." Beyond the fiscal sphere, there are important concerns associated with the virtual stabilization of the nominal exchange rate that the current policy has generated. As noted above, exchange rate stability may be seen as a beneficial consequence of this policy by working as a major factor limiting price increases of tradable goods. However, the growth of the monetary aggregates in 18 Barn (1997) presentscross-wlmby evidenceon the negativerelationshipbetween inflation and growth for a sample of 100munbies with annual observationson macmeumomic data for the period 1960-90. His &hding is that, other thing equal, a IO?? increase in the rate of idation reduces long-nm growthby about0.025% peryear. Angola, while slowing, has been faster than would be consistent with the achievement of the inflation objective, with the consequencethat the real exchange rate has appreciated significantly. The implication of this is that the policy has been much less effective in reducing the inflation of the non-tradables (see Figure 1.6).In addition, the scaling up of public spending is likely to exert pressure on the domestic price level. To counter this effect and to keep the declining trend in inflation, the authorities should resort to both sales of foreign currency and new issues of governmentbonds to help mop up excess liquidity. In this context, an appreciation of the nominal exchange rate shouldnot be resisted as it will contribute to reducing inflation. 2 THE BUSINESS ENVIRONMENT 2.1 Formal Sector:PerceivedConstraints Firms were asked to class@ various constraintsof their business environment Table 2.1 contains the percentage of firms perceiving each constraintas major or very severe.Approximately half of the firmsperceive accessto credit and electricityas important constraintsto their operation;crime, corruption and business licensing alsobelong to the group of most importantobstacles to business operation (since over one third of firms identified them as major constraints). Transportation, identifiedby 27% of firms, is alsoa relevant constraint. These perceptions vary across firms. Access to finance appears to affect smaU and medium firms more significantIythanlarge firms.Domestic firmsand firms located outsideLuanda alsoperceive it to be more of a problem than foreign firms or firms located in Luanda. By contrast, electricity appears to be more of a problem for large firmsand for finns in the manufacturing sector. Crime and corruption appear to be more significant constraints for large firms and for firms located in Luanda.Transportationis identifiedasa majorproblem for firmslocated outsideLuanda. Table2.1: MajororVery SevereConstraintsasReportedbyAll Formal SectorFirmsin Percentages Rrmslze m w i p Laation indusby Manuf. - - Rst Canstraint TOTALSmall Med. Large F o r m Dom. Luanda OLndde Food andManUf' LuanB Garments Omer RBtail me universe Acmss to finance(aMihbilityandmsl) 55% 56% 52% 40?4 35% 59% 48% 43% 58% 100% 41% 47% Elecbicity Qime, theffanddisorder 37% 39% 17% 60% 37% 37% 36% 43% 35% CoMpbm 38% 20% 60% Businesslicensingand Pm& 33% 35% 20% 60% 31% 34% Transpatation 27% 26% 29% 80% 25% 28% ~ ~ d inin(am, wad ~ 27%~ 28% 17% 20% ~ m 22% 27% Accessto land 23% 24% 19% 20% 31% 22% Tax rates 23% 23% 20% 60% 13% 25% w - a n i c instability 22Y. 23% 15% 40% 16% 23% CustomsandTrade Regulations 21% 22% 15% 60% 22% 21% Inadequablyeducated wCfkfcrce 21% 23% 6% 20% 15% 22% Tax adrrinisbaiion 18% 19% 11% 20% 13% 19% FunclioningofUte carrts 14% 16% 4% 40% 10% 15% Wlcal instability 14% 15% 4% 0% 7% 15% Labor R6g~laS0ns 12% 14% 2% 0% 6% 13% Telecarmudcations 7% 7% 8% 20% 9% 7% Source:ICA Survey Changing our focusto the manufacturingsector,Table 2.2 containsthe percentage of firms which classified each constraint as major or very severe. A higher share of firms (over 60%) perceive access to credit and electricityas the two most important constraintsto their operation. Similarly, business licensing, crime and corruption are important obstacles identified by over one third of manufacturingfirms. Table 2.2: MajororVery SevereConstraintsas Reportedby ManufacturingSectorFirms in Percentages I F i n size - 1 OwnershipI Location :Ut:: Constraint 1TOTAL Small Med. Large Forelgn Dom Luanda 63% 66% 52% 40% 49% 65% 60% 69% 100% 72% 61% Business licensing and Permits 41% 30% 60% 44% 39% Crime, theft and disorder 43% 19% 60% 23% 41% 37% 26% 60% 28% 36% Practicesof competitorsin inform.sector 36% 16% 20% 17% 34% Transportation 31% 25% 80% 34% 31% l ~ a rates x Macmeconomic instability Inadequatelyeducatedworkforce l~ccessto land Tax administration Customs and Trade Regulations Functioningof the courts Politicalinstability Labor Regulations Source:ICA Survey Figure 2.1 focuses on the perceptions contained in Table 2.1 across locations. Here we can see clearly that the perception of obstacles in Luanda is different fiom that outside Luanda.Access to credit, electricity, and transportation are more important constraints for firms located outside Luanda than for firms in Luanda. By contrast, crime and corruption are a significantconstraintfor firms located in Luanda. Figure 2.1: Top 6 Major or Very Severe ConstraintsasReported byAll Formal SectorFirms in Percentages 60% 50% aOutside Luanda 40% 30% 20% 10% 0% Source:ICA Survey In a comparison acrosscountries (Table2.3), we can see that, as inAngola, accessto creditis even more of a constraint in Algeria, DRC and Namibia - respectively 66%, 64% and 100%of h s report it to be a serious problem.Angola perfoms better thanlow income countries, sub-Saharan Afiica countries or resource rich countries, where over 75% of h s consider it to be a major bottleneck to theirbusiness. Electricity,on the otherhand, appearstobe more of a constraintonly in DRC, where 77%of firms report it to be a seriousproblem. In Algeria, SouthAfrica and Namibia it is perceived to be less of a problem than inAngola. Similarly,crime and licensing seem tobe more of a problem for h s inAngola thanin the rest of the continent.While corruptioninAngola is worse than in SouthAfrica, Namibia, and the DRC, it is within the regional average. Table2.3: MajororVery SevereConstraints asReportedby Firms-ComparisonAcross Countries Low Consbaint income SubSaharan Resource b n n s /Africa ImunMi Access tofinance (availabilityand cost) 55% 66% Electriaty 46% 15% Crime,theftand disorder 37% NIA Cormption 38% 39% Business licensingand Permits 33% 32% Transportation 27% NIA Practices d canpetiton in infam. sedu 27% 62% Access to bnd 23% 41% Tax rates 23% 47% Mauoeconwnicinslability 22% 42% Customs andTrade Regulations 21% 35% I Inadequatelyeducated wakforce 21% 29% l ~ a adminisbation x I1 18% 39% Functioningdthe cwrk 14% NIA Politicalinslability 14% NIA Labor Regulations 12% 17% I~elemmmunications 7% 19% Source:ICA Surveys These perceived constraints have a direct impact on indirect costs. In Table 2.4 we have estimated a number of indirect costs for manufacturingh s in Angola. We can see that the manufacturing sectorhas to bear indirect costs which amount to some 11%of itstotal sales.lgOf these, electricity (production lost due to power outages) is the main component (5.2% of total sales), although bribes (3.4% of sales)arealsorelevant. 19 Care mwt be bewhen intqdng this value as it may double count the& therefore, this d u e should be seen as an uppa limit to indirect costs. Table2.4: Indirect Costs-ManufacturingSector Firm size ( Ownership Location I Indirect costs as % sales TOTAL Small Med. Large Foreign Domestic Electricity 5.2 4.9 6.3 9.6 3.7 5.4 I I Breakage or spoilage in transit 1.4 1.1 2.9 1.1 Theft while in transit 0.7 0.8 0.2 I~heft.robbery or arson 1 0.4 1 0.4 0.3 0.0 1 0.1 0.4 Total indirectcosts 11.2 10.8 12.9 12.6 12.5 11.0 12.7 2.6 Source:ICA Survey In comparison to other countries, we can see fiom Table 2.5 that Angola performs better than Algeria and the Democratic Republic of Congo, where such indirect costs amount to 17.1% and 14% of sales respectively. However South Africa and Namibia report much lower indirect costs (2.7% and 4.7% respectively). Low income and Sub-Saharan African countries report levels of indirect costs higher than Angola, whilst h s in resource rich countries appear to face similar levels of indirectcosts (10.1% of sales). Table2.5: Indirect Costs-ManufacturingSector-ComparisonacrossCountries 1 I Alaria I D. R, Congo SouthAfrica I I Namibia LOW income Sub-Saharan Resource lich1 Indirectcostsas % sales LulgO'a IcnunU. f r i ICouMis Electricrty 5.2 5.3 7.4 0.9 1.2 9.9 8.7 5.8 Bribes 3.4 7.9 5.1 0.3 1.7 3.1 2.5 1.9 I~mductionlostwhile intransit 1 2.1 1 2.4 1 1.0 1 0.8 1 1.0 1 1.5 1 1.4 1 1.3 1 Theft, robberyor arson 0.4 1.6 0.5 0.6 0.8 1.O 1.O 1.1 Totalindirectcosts 11.2 17.1 14.0 2.7 4.7 15.5 13.6 10.1 Source:ICA Surveys In all comparators, electricityis one of the main drivers of indirect costs (and the main driver for all comparators except Algeria and Namibia). In conclusion, the analysis of the indirect costs appearsto confirmthe perception that electricity is a significantproblem forAngolan firms. It is interesting to note that these indirect costs affect different types of firms differently (Table 2.4). Electricityand bribes are more of a problem fordomestic firms (5.4%and 3.7% of total sales respectively) and firms based in Luanda (5.6% and 4.2% of total sales respectively). Electricity affects large firms to a significant extent (9.8% of total sales), whereas bribes and crime affect mostly smallfirms(3.5% and 0.4% of total salesrespectively). When looking at aIl formal sectors, the main conclusion does not change significantly(see Table 2.6)." Total indirect costs amount to 8.3% of total sales, with electricity alone accounting for over half of that total (4.6% of total sales). Large firms (11.4%of sales) face higher indirect costs than smalVmediumfirms. 20 Note thatTable2.6 doesnot includebreakage,spoilageor thefl while intransit Electricity related indirect costs afflictlarge firms (9.8% of total sales), domestic firms (4.8% of total sales) and firms based in Luanda (4.9% of total sales). The manufacturing sector is more affected than the other sectors. Table 2.6:IndirectCosts-All Formal Sectors F l n size Ownership Location lndusby Manuf. - Rest Indirectcosts as %sales T O T NSmall Med. Large Foreign Dom. Luanda Outside Food and - Retail me Luanda ' beYeragesGarments Other univen Elecblcily 4.6 4.5 4.6 9.8 3.1 4.8 4.9 2.2 4.7 5.7 5.5 4.2 3.0 Bribes 3.3 3.4 2.4 1.6 3.3 3.2 4.1 0.0 2.6 4.3 3.8 4.4 1.8 Theft, robberyw arson 0.5 0.5 0.5 0.0 0.8 0.4 0.5 0.3 0.6 0.5 0.3 0.7 0.6 Totalindireclcosts 6.3 8.4 7.5 114 7.3 8.5 9.5 2.5 7.8 10.5 9.5 9.3 5.3 Source: ICA Survey 2.2 Electricity Looking in more detail at electricity,there are several dimensions one can analyse. InAngola 84% of all firms experienced power outages, on average 8 times per month (Table 2.7). Large firms experience power outages more frequently (16 times per month) but with a lower duration (8 hours). Firms based in Luanda are relatively less affected by this problem and the manufacturing sector is more affected than the others. This confirms why electricity is perceived to be less of problem for firmsbased in Luanda and more of a problem for firms in the manufacturing sector - see Table 2.1. To correct for this problem, 68% of firms have their own generators. We would expect a higher percentage of generator ownership where the problem is perceived to be more important: indeed, over 90% of firmsfrom outside Luanda have a generator.These generatorsproduce an average of 31% of the total electricity needs. Large firms rely more on their generators (49% of total electricity) thansmallandmedium firms(28% and43% respectively). Additionally, obtaining an electrical connection is a time consuming process: on average, firms have to wait 60 days-to obtain one. Firms from outsideLuanda have to wait a greatdealmore -on average 182days. Table2.7: InfrastructureIndicators-AU Formal Sectors - I Firmsue Location Industry danuf. Rest lTwe I Gutsidt :ood Manuf. - Manuf. -Retail the Qualitymeasure TOTAL Small Med. Largc Foreign Dan Luanda Luanda ~ n d Garments Other universe - I V % firms experiencedp e r outages 84 85 78 100 89 87 81 87 82 Frequency of power wbges (time 8 7 8 16 8 8 9 6 8 per month) Durationof poweroutages (harm) 21 21 23 8 30 17 26 13 14 I I1% Eledriaty Pmductionlost (% annual sales) 4.6 4.5 4.6 9.8 4.7 5.7 5.5 4.2 3.0 I finns mth agenerata 68 65 84 80 89 71 55 NIA NIA I % eledtiaty coming fmm owl 31 28 43 49 33 23 31 NIA NIA generata Number of days lo obbln electrica 60 46 202 7 99 45 64 25 40 mneclion - Source:ICA Survey In comparison to other countries (Table 2.8), only in the Democratic Republic of Congo do more firms experience power outages (96%) than in Angola. When analysing total outage duration (Erequency of outagesmultiplied by its average duration),Angolan firms report a longer period (80 days) without electricity over a year. Firms in DRC have more frequent power outages (19 times per month)thaninAngola, but they last for less longer(4 hourson average). Overall, the number of days of production lost because of electricity problems is significantly higher in Angola than in comparator countries, althoughthe indirect cost it generates for firms in Angola is lower than in those countries.This may be explained by the higher penetration of own generatorsinAngola (68% of h s in Angola have a generator), which account for a larger share of total electricity needs (compared to other countries) and this may reduce the impact of power outages on the day-to-day operation of firms. However, care must be taken with this line of thought because it is costlier to use a generator to produce electricity, which naturally translates intohigherdirect costsforfirms. Our main conclusion is that even with such a high penetration of generators (compared to other countries),electricityremains a seriousconstrainttobusiness. Table 2.8: InfrastructureIndicators-ComparisonacrossCountries JAW Sub-Saharan Resource TW Qualitymeasure Angola Algeria D. R Congo SouthAiiica Namibia income rich Africa countries countries % firmsexpiend poweroutages 84 68 96 NIA NIA NIA NIA NIA Total outage duration(days) 80 14 36 1.0 0.1 11.5 8.4 4.6 Productionlost(%annualsales) 4.6 5.3 6.9 0.9 1.4 9.1 8.2 5.8 ~l&"ty % firmswith own generator 68 29 42 9 7 41 36 30 % electricitycomingfrom own generator 31 22 19 17 1 14 11 9 Number of days to obtain electricalconnection 60 100 27 6 10 50 42 45 Source:ICA Sweys 2.3 Corruptionand Crime Corruptionisperceived to be a seriousconstraintby firms.Approximately36% of firmsrecognize it to be a major or very severe constraint, and thisp e p t i o n appears to be more common among large firmsand firms located in Luanda (Table2.1) . That corruption is one of the major bottlenecks in Angola is confirmed by other sources. Transparency International's Corruption Perceptions Index (CPI) attempts to quantify the degree of corruption as seen by business people and country analysts, and ranges between 10 @ghly clean)and 0 @ghly corrupt).Table2.9 showsthatAngola is near the bottom of the ranking,with a scoreof2.2 in 2007.Within our comparators,only the Dern.Rep. of Congoperformsworse. Table 2.9: ConuptionPerceptionsindex 2007 - Nevertheless, since the end of the civil war Angola has recorded some improvements on political stability and control of cormption. More recently the World Bank Governance Indicators for 2006 have IDRC 168 1.9 1 shown Angola as one of the top reformers Source:TransparencyInternational in these 2 areas. Whilst positive results have been achieved recently, Table 2.6 shows that corruption still entails a significant indirect cost for firms of approximately 3.3% of sales. Whereas firms located outside Luanda support lower indirect costs because of bribes (and this may help explain why cormption 21This apparent strange resdt is probably due tothe factthat this question is perceived by managersoutside Luanda as more threateningthan those in Luanda. As a matter of fact informal discussions with local expertshave pointed outthat this result is counterintuitive. is perceived to be less of a problem for those firms),large firmsalso appear to face lower i n h t costs. This result implies that the amount of bribe genemlly requested is a fixed amount not dependenton the volume of activityof the firm. From an internationalperspective, corruption is more of a problem in Algeria and DRC, where the cost of bribes is higher than inAngola (Table 2.5). Neverthelesswith an estimated cost of over 3% of sales,corruptionremains a key bottleneckto firmsinAngola. Looking in more detail at corruption, Table 2.10 shows that only 23% of firms believe that governmentofficials have a consistent and predictable interpretationof the law. This uncertainty may be closely linked to corruption.Furthermore some 24% of firmsreport informalpayments or giftsto be commonto "get things done" regardingcustoms,taxes, licenses,regulations,etc, whilst only 7% know in advance the amount of payment needed. Informal payments are perceived to be a more seriousproblem for large and domestic firms,as well as for firmsbased in Luanda. When a government contract is at stake, firms expect to have to pay some 4.5% of its value in such informalgifts orpayments tosecure it. Table2.10:Perceptionof GovernmentandRegulations-All Formal Sectors Firmsize Ownership Location Industry ;: Z:knbs E:f'- Manuf. - Rest of % firms who agree with statement TOTAL Small Med. Large F a e n Dom and Retail h e Wanda :$: universe Consistent and predictable 23.4 23.3 24.4 20.0 30.4 22.2 28.3 0.0 15.5 23.0 26.0 27.3 22.0 interpretationof h e law lnfumal 23.8 24.1 19.4 40.0 16.6 25.0 28.8 0.0 18.3 45.9 30.0 20.7 16.3 commonplace Advance knowledge of infmal 7.0 6.6 6.3 40.0 12.1 6.2 8.5 0.0 8.4 11.5 4.9 7.5 7.7 paymentlgift Percen'ageafannualsalesspent 3.3 3.4 2.4 1.6 3.3 3.2 4.1 0.0 2.6 4.3 3.8 4.4 1.8 on informalpayments/gifts I paidl 1 1 I TI 4 4.3 5.5 10.0 4.7. 4.5 5.6 1 1 to secure contract Of conh3ct "lye 6.1 4.8 6.3 I I I I I I I Source:ICA Survey The court system is another institution where corruption may be a problem. Table 2.11 showsthat firms have a relatively low confidence in it. Almost 90% of firms believe the system to be unfair, partial and corrupted whilst only 50% believe it to be able to enforce its decisions. Clearly, the problem appears to be not so much at the postdecision stage, but at the predecision stage, with 80% of firms consideringthe process slow, 70% expensive,partial and corrupted. This conclusion is reinforced by the fact that whilst 4% of firms had payment disputes in the past two years, only 49% of suchdisputes were taken to court. Table2.11:CourtSystem-All Formal Sectors I I Firmsize I Ownership ( Location I Industry Fz::Lo rJkb- me^ EF- Manuf. - aaraMtica TOTAL Small W. Lam F a i ~ nDm, and Retail Rest Of System Luanda universe Fair,imparUalanduncompted 31.7 31.5 34.620.0 50.7 28.4 38.0 1.6 29.5 17.2 28.4 33.0 41.2 Quick 21.5 21.7 20.4 20.0 32.5 19.8 25.7 1.6 16.9 5.7 21.9 19.8 30.6 Affordable 28.5 29.5 22.2 20.0 38.2 28.8 34.2 1.5 25.3 0.0 29.2 28.5 37.4 Able to enforcedecisions 48.8 49.1 47.8 40.0 52.2 48.2 58.7 1.5 37.9 51.6 58.4 42.6 48.8 Percentage d firms with paymentd'mputesinthepast2 3.9 2.5 9.3 40.0 8.7 3.1 4.4 1.2 1.4 5.7 3.3 4.7 5.8 years seffled by third parties I I I I I I I Source: ICA Survey Crime was alsoreported to be a seriousconstminttobusiness (seeTable2.1).Almost40% of firms complained about this problem. From an internationalperspective, moreAngolan firms complain about crimethan firmsin SouthAfrica, where this is awell known problem. Apart fiom the ICA data, there is little available evidence on the overall crime rates in An ola. According to the Interpol (2000), the incidence of crime is 71.52 per 100,000 inhabitants2'. In comparison, South Africa's is 8,176 per 100,000 inhabitants (Intepl (200l))~~,ranking consistently in the world top crimerankings,especiallyon violent crime. When we look at more objective indicatorsof crime, the picture is quite different.Although many firms complain only 18% of them experienced losses as a result of theft, robbery, vandalism or arson, whilst in DRC and Namibia such losses were experienced by 26% and 35% of firms respectively. South Africa has a higher incidence of crime: 52% of manufacturing firms experiencedlossesdueto it. The impact of crime on the costs of production in Angola is not as high as in other countries. Theft, vandalism or arson generates indirect costs totalling 0.5% of sales, lower than most countries in the region (Table 2.5). Similar results are obtained if we look at the share of f m s that employ security services. Fewer h s in Angola's manufacturing sector pay for security services (Table 2.12). At the same time average security costs (as a percentage of annual sales) are similarto DRC and particularly South Africa (wherecrimeincidenceishigh). 22 Available h m http:llw.justicemitiative.orplregidafridanpla . 23 Availablehh t t p : l l w . j u s t i c e m M v e . ~ d a ~ c a / ~ ~ ~ ~ c a / s a d h a ~ ~ o . Table 2.12: SecurityServicesand SecurityExpenditure-Comparisonacross Countries- ManufacturingSector F i n ske Ownership Location Industry Manuf. - Rest of Outside - Manuf. - TOTAL Small Med. Large Foreign Dom. Food and Manuf. Luanda Luanda Garments Other Retail the bev. universe Percentage Of fins 48 45 68 60 72 44 49 41 56 29 32 57 61 that paidfor securitv Seculity as % 1.7 1.7 1.9 0.8 2.7 1.6 1.8 1.4 1.9 0.8 1.O 3.0 1.7 annual sales Source:ICA Survey Finally, if we sum together the cost due to theft and the costs of security services, we see that Angola perfoms as well as other countries in the region and evei better that the region average (Figure 2.2). Thisleads us to concludethat although crime has some cost implications forAngolan firms,it is not a major constraint. Figure2.2:Costsof SecurityandTheft-International Comparisons --- ---- 4 01- Angola D. R. Congo South Africa Low income SSA SSA Resource rich counbies Source:ICA Survey 2.4 RegulatoryFramework Over the past few years Angola has registered a position and noticeable evolution on regulatory quality. The governance indicators developed by the World Bank clearly show that Angola has moved up fiom the lowest levelssince 2000.Althoughthe evolutionin regulatoryquality has been positive, more needs to be done since Angola still remains among then Sub-Saharan African countries with the lowest score. This confirms h ' s perceptions of business licensing and the overallregulatory burden as a significantconstraint (Table2.1). Table 2.13 shows that on average close to 8% of senior management time is spent dealing with governmentregulations.Thisburden fallsdisproportionatelyon large (14.6%)and foreign(10.2%) firms, as well as h s based in Luanda (8.4%).The retail sector is the most affected (12.9%). On average, close to 68% of all firmswere visited by officials each year, on average 5 times. Large firmsrank high in both coverage (100%) and frequency(7visits),whereas smallh s rank low in both (63% in coverage and 5 visits). Outside Luanda, coverage is high (98%), but frequency is lower(3visits). Table2.13:RegulatoryBurden-All Formal Sectors Firm size Ownership Location lndusby Quality measure TOTAL Small Med Large Foreign O m . ::t:z Lo Ee:ni - Manuf. - and tE: RetaL! Rest Of the Luanda untvene % senior management time 7.7 7.7 7.4 14.6 10.2 7.3 8.4 4.7 5.6 3.9 6.6 12.9 6.9 spentwith regulations % firmsvisiledby officials 67.6 62.9 96.2 100.0 83.4 64.9 61.1 98.3 74.7 42.6 58.6 66.1 83.7 N u m b e r o f i n s ~ v i s ' t s 5.2 4.7 7.1 7.2 4.7 5.3 5.9 3.0 4.9 3.1 5.0 4.7 6.2 Jlast 12 months Source:ICA Survey On the indicator of regulatory burdenAngola performs similarlyto our comparatorcountries,with the exception of Namibia (Table 2.14). Table 2.14: RegulatoryBurden-ComparisonacrossCountries Qualitymeasure countries Africa % senior management time spent 7,7 with regulations Source:ICA Survey Regarding the licensing process, Table 2.15 shows that obtaining licenses is a slow process: it takes 24 days to obtain an import or operating license, whilst a construction-relatedpermit takes some42 days. The processof obtaininglicenses ismuch slowerin Luanda and (withthe exception of importlicenses) it is fasterfor small firms. Table2.15:LicensingProcess Firm size Ownenhip Loqation Industry Outside pz:f. - Manuf. - Manuf. - Number of day. to obtain: TOTAL Small Med. Large Foreign Dom. Luanda Luanda Garments Other Retail the and unlverse Construction-relatedpermit 42.1 27.0 93.4 105.0 39.2 42.8 57.4 7.5 41.3 75.0 81.2 21.3 32.6 An import license 23.9 25.1 19.9 8.0 30.5 21.1 26.7 8.5 15.3 20.0 16.5 23.5 32.3 An operatinglicense 24.1 22.9 24.3 64.0 34.5 21.5 40.1 8.6 18.5 8.0 19.6 23.5 31.8 - Source:ICA Survey An international comparison(Table2.16) showsthat it is common inAngola to wait long to obtain licenses, even though the wait in Angola is a bit better than the regional averages. Nevertheless Angola performs worse than high performance countries such as SouthAfiica and Namibia on all types of license. Table 2.16: LicensingProcess-ComparisonacrossCountries Low income Sub-Saharan Resource Number of days to obtain: Angola D. R. Congo South Ahica Namibia countries Africa rich countries Construction-related permit 42.1 23.6 8.6 21.4 58.7 54.3 51.5 An import license 23.9 12.3 7.0 19.5 28.0 24.7 16.9 An operatinglicense 24.1 23.7 5.0 9.3 15.7 15.2 24.2 Source:ICA Survey Starting a business in Angola was a lengthy (124 days are necessary) and costly process (Table 2.17). Only the DRC fares worse than Angola (155 days), whilst Algeria, South Africa and Namibia appear to have a speedier and less costlybusiness start-upprocess. Both the time needed and the costsof startinga business inAngola arehigher than the Sub-SaharanAfkica average.This estimates however in based on 2005 data. In response to this problem the government has established a new office, the Guichet&co that is supposedto speed up the registration of firms. And the next Doing Business will show that Angola has improved its performance on this indicator. Table 2.17: Startinga Business Angda Algeria D. R. Congo Swth Africa Namibia SubSaharanAfrica Number of procedures 13 14 13 9 10 1 1 1 1 Duration(day.) 124 24 155 1 1 1 1 1 1 1 cost (Xof income per capha) 486.7 21.5 481.1 6.9 18.0 1628 I Min.capital (% of income per capita) ) 74.1 1 46.0 177.3 0.0 0.0 209.9 Source:World Bank Doing BusinessIndicators2006 Nevertheless if we look atprocedures, time, and coststo build a warehouse (Table 2.18), both time (326 days) and costs are much higher in Angola than in the comparator countries (with the exception of DRC regarding costs). The sub-SaharanAiXca average (230 days) is lower,as is the costassociated with the licensingprocess. Table2.18: Licenses Angola Algeria D. R. Congo South Africa Namibia SubSaharanAfrica Numberof procedures I 15 I 25 I 14 1 17 I II I 18 1 Time(days) 326 244 306 174 105 230 Cost (% of incomepercapita) 1239.2 58.9 2281.9 33.5 134.9 1024.5 Source:World Bank DoingBusiness Indicators2006 Thisleadsus to conclude that business licensingremains a problem inAngola. 2.5 'Ransportationand OtherConstraints Transportation emerged fiom Table 2.1 as an important constraint to business, particularly for firms located outside Luanda. From Table 2.4 and Table 2.5, manufacturing firms in Angola lose 2.1% of their sales in transit (breakage, spoilage or theft), more than in all comparators (except Algeria). Strangely, firms located outside Luanda report very low transportation indirect costs, although they perceive transportationto be a significantconstmint. Table2.19: InventoryHoldings of Most Important Input - ManufacturineandAll Sectors - Supply chain problems Finn size Ownership often result in firms outside holding large inventories, Average numberof days TOTAL Small Medium Large Foreign Domestic Luanda Luanda which is inefficient.Table Manufacturing 2.19 shows that on average manufacturing AII Sectors firms hold approximately 14 days of production of theirmost importantinput. Large firms and foreign firms hold on averagemore (20 and 21.6days respectively).The retail sectorholds on average more inputs (20days of production) than the other sectors The smoothoperationalongthe supply chain requires easy transportationof goods between firms. As we can see in Table 2.20, only 31% of manufacturingfirmshave their inputsdeliveredby mad. Whilst not explicitly asked, we presume that other means of transportation are commonly used, possibly railways, airor maritime transport. In particular, we observe that no firms outsideLuanda have their inputs delivered by road. This explains why they perceive transportation to be a significantconstraint(Table 2.l), but report no indirectcosts of transportation(Table2.4). Table 2.20 Percentageof Firmswith Inputs Deliveredby Road -Manufacturing Sector Firm size Ownership Location Industry Manuf. - Rest of TOTAL Small Med. Large Foreign Dom. Luanda ;::$: t:;:ints- !::lf-" Retail the universe Percentage of firms with inputs 3, 35 20 12 33 37 24 40 33 NIA NIA delivered by road Source:ICA Survey Table 2.21 complementsthe results shownin Table 2.20 Some40% of inputsare of foreign origin, with large firms (80%), foreign firms (67%) and firms manufacturing garments (58%) using a higher proportion of foreign inputs. Given the high proportion of foreign inputs used for production, the good functioning of the customs agencies is essential in the supply chain. Only 20% of firms import directly and it takes approximately28 days for incoming inputsto clear customs. Table2.21: Origin ofInputs-ManufacturingSector I I I I I I Fim~size Ownership Location Industry I Manuf. - TOTAL Small Med. Large Fonign Dom. :u\y:: [:tra;;~ Luanda ~ ~ ~ ~ n Retail-universe~the ~ U ~ f ' :Y:- Percentage Of inputs Of 60 66 32 20 33 62 59 64 53 42 66 NIA NIA domestic origin Percentage of inputs o 1 1 1 I 40 ( 1 3 1 68 80 67 38 41 36 47 58 34 NIA NIA Source:ICA Survey In comparisonto other countries,the number of daysneeded to clear customs is clearlyhigh. From Table 2.22 we can see that D.R. Congo (13 days), South Africa (7 days) and Namibia (3 days) have a speedier process for imports to clear customs. All country group comparators perform better than Angola in this respect aswell. Table 2.22: Customs-ManufacturingSector Comparisonacross Countries - - I 1 I 1 1 I ~ n b i ~ ~ m e l ~ ~ ~ i ~ b n n l , " , " , " $ ~ , " Angola Alge"a D. R.Congo South Africa Namibia I ~ v eA. of daysto clear customs (imports)I 28 1 22.9 1 12.8 1 6.7 1 3.4 1 15.4 1 12.4 1 10.2 1 Source:ICA Surveys Whilst time needed to clear customs is one important variable, other aspects are also relevant. Table 2.23 contains some Doing Business indicators for the comparator countries regarding trading across borders. In Angola, importing is costly, time consuming and bureaucracy-ridden when compared to Algeria, South Africa and Namibia. Angola compares well to DRC, but in terms of cost and time for imports it ranks worse thanthe sub-Saharan Afiica average. A similar picture emerges when analysing exports, although the number of documents necessary for exporting is lowerthanin the othercomparatorcountries,with the exceptionof SouthAfiica. Table2.23: 'IkadingacrossBorders Angola Algeria D. R. M n ~ o SouthAfrica Namibia Sub-Saharan Africa Documentsfw export(number) Time for export (dap) Costto export (US$ percontainer) Documentsfw import(number) Time for import(dap) Costto impat (US$ percontainer) . . I I I I Source:World Bank Doing Business Indicators 2006 Therefore, transportation is indeed a problem inAngola, which generates indirect costs for h s . In addition, although customs is not perceived to be a significantconstraint (Table 2.I), objective indicators showthat it may indeed constitutea seriousconstraintfor business, in particular, for the manufacturing sector which imports 40% of its inputs. For h s located outside Luanda, which perceive transportation constraints as more significant, it probably generates lower indirect costs because othermeans of tmnsportationareused. 3 MICRO FIRMS 3.1 Constraints to Business The main constraints identified by micro firmsz4are electricity, access to credit, transportation, access to land, corruption andbusiness licensing(Table 3.1).With the exceptionof access to land, all other constraints are also among the most important for the formal sector (although transportation is particularly relevant for firms located outside Luanda). As in the formal sector, location is a key factor in the perception of constraints: almost all k s outside Luanda identify access to credit as a major or very severe constraint, whilst access to land is a problem only for firmsin Luanda. Table3.1:Majoror Very SevereConstraintsasReportedby Micro Firmsin Percentages Ownership Location Industry Rest ( Constraint TOTAL Foreign Dom. Luanda Lwnda Food and Manuf. I Outside Manuf. - - Manuf. - Retail the Ibev. Garments Other universe Electricity 74% Accessto finance (availabilityand cost) Transportation Accessto land Corruption Businesslicensingand Permits Crime, theft and disorder Customsand Trade Regulations Practicesof competitors in inform. sector Macroeconomicinstability Tax rates Tax administration Inadequatelyeducatedworkforce I (~oliticalinstability l~unctioningof the courts I ILaborRegulations Source:ICA Survey The impact of some of these constraintson micro firms' costs is higher than for formal firms. The breakdown of indirect costs amountsto approximately 10.6%of total sales (Table 3.2). Electricity (5%) and corruption (bribes) (3%) are the two main causesof such costs. Firms located in Luanda are particularly affected (11.6%of sales), whilst firms outside Luanda have comparatively lower indirect costs(2.9% of sales). 24 Micm firmsarefirmswiB less than 5 Ill-timeemplo~es. Table3.2: Indirect Costs-Micro Firms Ownership Location Industry Manuf. - Outsae Food and Manuf. - Manuf. - Rest of Indirectcosts as % sales TOTAL Foreign Dom. Luanda Luanda beverages Garments Other Retail the universe Electricity 4.0 2.4 5.1 5.2 2.4 9.6 1.0 4.0 4.4 5.6 Bribes 3.1 0.2 3.4 3.5 0.2 0.0 0.0 5.0 3.0 3.3 1Breakageor spoilage in transit 1 0.4 1 0.2 0.5 1 0.5 0.2 1 0.0 0.0 1.3 0.5 0.0 1 Thelbile in lansit 0.8 0.2 0.9 0.9 0.1 1.5 0.0 0.4 1.3 0.1 1Theft. robberyor arson 1.4 1.5 1.3 1.5 0.0 0.0 0.0 0.1 1.3 2.2 Totalindirectcosts 10.6 4.5 11.2 11.6 2.9 11.1 1.0 10.7 10.5 11.2 Source:ICA Survey Looking first at electricity, (Table 3.3), 90% of firms experienced power outages, on average 11 times per month, each lasting on average 15 hours. As a result, a significantproportion of firms own their own or sharegenerators(50% of firms),which account for 43% of their total electricity needs. Table3.3: InfrastructureIndicators-Micro Firms Ownership Manuf. - Rest Luanda Outside - Manuf. - Quality measure TOTAL Foreign Dom Food and Manuf. Luanda Garments Other Retailthe universe % firms experiencedpower outages 90 Frequency of power outages (times l, per month) Durationof power outages(hours) 15 1 Electricity Productionlost(% annual sales) 4.8 1 % firmswith own generator ( 50 % electricity coming from own generator Number of days to obtain elecbical 52 connection Source:ICA Survey Regardingcorruption,Table 3.4 showsthat closeto 90% of firmsbelieve that government officials do not interpret consis&ntlyand predictably the law, whilst some 26% of firms report informal payments or gifts to be common to "get things done" regarding customs, taxes, licenses, regulations, etc. The perception of firms in Luanda is more favourablethan that of firms outside Luanda, where almost all firmsbelieve that officials do not interpretthe law consistently and that informalpayments are necessay to get things done. Overall, firms spend on average 3.1% of their annual sales on informal gifts or payments. Also, when a government contract is at stake, firms expect to have to pay some 5.9% of its value in informal giftsorpaymentsto secure it. Table3.4:Perceptionof Governmentand Regulations-MicroFirms -- Ownership Location Industry - - Manuf. - % firms who agree with Outside Manuf. FoodManuf. Rest o TOTAL Foreign Dom. statement Luanda Luanda and bev, Garments Other Retail the 1 universe Consistem and pedimble/ 13.2 12.5 14.3 4.8 0.0 0.0 7.6 18.2 8.9 1 intelpretationof the law p"menwginj 11 20.0 11 11 Informal 25.8 10.0 27.4 28.6 4.8 0.0 0.0 38.1 26.8 23.8 commonplace Advance knowledge of ,4.1 0.0 15.5 15.3 4.8 0.0 0.0 22.9 14.8 11.9 informalpaymentlgift Percentage of annual sales spent on informal 3.1 0.2 3.4 3.5 0.2 0.0 0.0 5.0 3.0 3.3 pvmentslaifts Percentage of contract value I I 1 1 5.1 7.4 5.7 6.6 0.5 1.9 paidto becure contract I I I I I I Source: ICA Survey Similarly to the results of the formal sector, the perceptions of the court system are not very favourable: the great majority (60% to 70%) of micro firms think the court system is unfair, partial, corrupted, and not affordable. Firms in Luanda have the system in higher regard than do firmsoutsideLuanda, although none of them have used the courtsrecently. Table3.5: Court System-MicroFirms Ownership Location Industry z:zLnt; E:kf'- Manuf. - Rest of Characteristics of the court ~~~~~~ TOTAL Foreign Dom. and Retail the system Luanda universe Fair, impartialand uncorrupted 28.5 30.0 28.4 31.6 4.8 61.9 76.4 15.3 23.4 35.6 Quick 24.9 40.0 23.4 27.6 4.8 61.9 76.4 45.8 18.2 20.8 Affordable 42.9 70.0 40.2 48.0 4.8 61.9 76.4 53.4 37.1 44.5 Able to enforce decisions 51.9 80.0 49.2 58.2 4.8 61.9 76.4 61.0 49.1 50.5 Percentage of firms with payment disputes in the past 2 4.5 0.0 5.0 5.1 0.0 0.0 0.0 7.6 3.4 5.9 years settled by third parties Source:ICA Survey Crime doesnot appearas a top 5 constraintformicro firms.Neverthelessit generatesindirect costs totalling 1.4%of sales (Table 3.2). In additiondirect costs (security services)associatedwith crime arein the order of 1.3%of sales (Table 3.6). Table3.6: SecurityServicesand SecurityExpenditure-Micro Firms Ownership Location Industry Rest of Manuf. - Food Manuf. - Manuf. - Retail the TOTAL Foreign Dam, Luanda Outside Luanda and bev. Garments Other univem Percentage of firms that paid 33 50 32 38 0 0 0 23 48 18 for security services Security wst as % annual 2.3 1.2 1.5 0.0 0.0 0.0 0.7 1.8 0.9 sales Source:ICA Survey One interesting result fiom the perceptionsof micro firms is the fact that business licensingis also. identified as a significant constraint. The vast majority of micro h s in Angola have gone through some sort of licensinglregistrationprocess: 87% have an approved or registered company name; 92%have a commercial registration;91% have an operatinggenerallicense and 88% have a tax identification number (Table 3.7). Therefore, it is legitimate for these firms to perceive business licensing as a constraint because most of them have gone through a licensing process. However, this result suggests that defining the informal sector in Angola based on the number of full-timeemployeesmay not be entirelyappropriate (seeHenley et al. (2006)'~). Table3.7: LicensingDtegistration-Micro Firms Mership Location zzLni indusby Percentageoffins with TOTAL Foreign Dorn. ~~~~~ ,","tEv: Food Manuf. - Other Retail Luanda universe Of the Appmvedlregistered companyname 87 90 87 86 100 69 62 54 95 91 Commercial regisbation 92 100 91 91 100 69 62 85 97 91 OperaSng/bade/generaIlicense 91 100 90 90 100 69 62 77 95 94 Tax identificationnumber 88 100 87 87 100 69 62 62 93 94 Source:ICA Survey Indeed, we can see from Table 3.8 that micro h s iden@ time to complete the registration process as the most significantobstacle to obtaining a license, and this is particularly relevant for h s located in Luanda. 25 Henley,A., Arabsheibani,G R andCameim,F. (2006), "Ondefiningandmeasuringthe i n f d sector", World B d PolicyResearch Working Paper3866. Table3.8: PercentageofFirmsReportingMajor orVery SevereObstaclesto Registeringa Business-Micro Firms Ownership Location Obstacle TOTAL Foreign Dom. Luanda Outside Luanda Time to complete regismtion 28% 10% 30% 32% 5% Minimumcapitalrequirements Financialcost of completingregistration I I I I IAdministrativeburdenof complyingwith tax laws 17% 20% 17% 19% 1 1 I O% Other administrative burdens 1 16% 20% 16% 18% 1 1 ::z 1 I Difficultyof gettingnecessary information 14% 0% 16% Financialtax burdenafter regisbation 12% 20% 13% 5% I Strict labor marketrules 5% 10% 4% 5% 0% Source:ICA Survey Looking at more detail at the overall regulatory framework,Table 3.9 shows that on average 8%of seniormanagementtime is spentwith government regulations. Moreover,on averageclose to 83% of all firms were visited by officials an average of 8 times each year. This is similar pattern to formal sectorfirms-see Table 3.11. Table3.9: Regulatory Burden-MicroFirms Ownership Location Industry Rest of Manuf.- Food Manuf. - Manuf. - Qualitymeasure TOTAL Foreign Dam, Luanda Outside Retail the Luanda and bev. Garments Other universe % senior management time 8.2 8.8 8.2 8.9 2.9 1.9 3.1 8.9 9.3 7.3 spent with regulations % firms visited byofficials 82.9 90.0 82.2 80.6 100.0 38.1 23.6 84.7 89.7 79.2 Number of inspection visits 7.5 10.4 7.2 8.2 3.2 5.5 3.0 15.9 7.5 4.2 (last 12 months) Source:ICA Survey Obtaining licenses is a lengher process for micro firms (Table 3.10): it takes some 80 days to obtain a construction-relatedpermit and 64 days to obtain an import license. This is particularlyso for firms in Luanda. Table 3.10:LicensingProcess-MicroFirms Location Industry I Manuf.- Rest of Number of daysto obtain: TOTAL Luanda Outside Foodand - - Luanda Garments Other Retailthe 1 1 I universe1 ~~onst~ctionrelated pt'd 80.1 1 1 11041 7.0 1 NIA NIA NIA 90.0 60.4 1 himportlicense 642 72.9 7.7 7.0 NIA NIA 67.6 64.3 An operatinglicense 34.4 42.1 8.6 8.0 8.0 20.4 28.5 47.8 Source:ICA Survey 3.2 Comparisonbetweenformalsector and microb s As we can see from Figure 3.1, there are minor differencesbetween firmsin the formal sectorand micro h s in the identificationof their main constraints:the top 6 constraints of formal firmsare alsomain constraintsformicro firms,with the only exceptionof accessto land. Figure 3.1: Percentageof FirmsReporting MajororVery Severe Constraints(Top6 ConstraintsforFormal Sector)-Formal Sectorversus MicroFirms ,--- BOY. - --- ---- Source:ICA Survey On the other hand, indirect costs are slightly higher for micro firms, especially because of theft, robbery or arson (Figure 3.2). Figure 3.2: Indirect Costs-Formal SectorversusMicroFirms Eiectllcity Bllbes Wll,mbbgl w amon Tdal indirectccsls Source:ICA Survey Micro finns facea somewhat similarregulatory burden, except for visitsby official (Table 3.11). Table3.11:Regulatory Burden-Formal SectorversusMicroFirms Quality measure Formal Micro I % senior management time spent with regulations 7.7 % firms visited by officials Numberof inspectionvisits (last 12 months) 5.2 7.5 Source:ICA Survey The percentage of annual sales spent on giWinforma1payments is the same for micro firms and firms in the formal sector (Table 3.12). While governmentofficials are held a bit in higher regard .in the formal sector where 23% of firms believe officials to have a consistent and predictable interpretationof the law, compared to only 13%amongmicro firms. Table3.12: Perception of Governmentand Regulations-Formal Sectorversus Micro Firms 1 1 1 % firms who agree with statement Formal Micro Consistent and predictable interpretationof the law 23.4 13.2 Informalpaymentslgiftscommonplace 23.8 25.8 Advance knowledgeof informalpaymentlgift 7.0 14.1 Percentage of annual sales spent on informal 3.3 3.1 paymentslgifts Percentageof contractvalue paidto secure contract 4.5 5.9 Source:ICA Survey Regarding infrastructure (Table 3.13) again micro firms and firmsin the formal sector are equally affected by power outages. The most significantdifference is that more firms in the formal sector have generators (68% versus 50% among micro firms), but these are more important for micro firms (generatorsare responsible for 43% of electricityneeds of micro firms,comparedto 31% in the formal sector). Table3.13: InfrastructurePerceptions-Formal Sectorversus MicroFirms Type Qualitymeasure Formal Miuo I% firms experienced power outages Frequencyof power outages (times per month) Durationof power outages (hours) Electricity Production lost(% annual sales) % firms with own generator 1% electricity coming from om generator Number of days to obtain electricalconnection Source:ICA Survey 4 FACTOR MARKETS: FINANCIAL SECTOR, LABOR AND LAND MARKET 4.1 TheFinancingof FirmsinAngola As we have seen in chapter2, access to finance is a major constraintto business inAngola. Thus, this sectionanalysesfirms' accessto financein detail,alongwith labourmarketand land. In order to operate, firms need short-term and long-term finance. In Angola, the main sources of working capital are firms' own retained earnings (81%), with trade credit (11%) emerging as the second largest sourceof finance (Table 4.1). The banking sectoraccountsfor a meagre 1%of total short-term financingneeds. Firms outsideLuanda rely evenmore on internal funds (94%). Table4.1:.Sourcesof Short-termFinanceintheFormal Sector Firmsize Ownership Location lndustry Rest of Outside - Manuf. - Ynuf. - % short-term financing from TOTAL Small Med. Large Foreign Dom. Luanda Lwndaand bev,Garments Other universe InternalfunddRetainedearning 81.4 83.2 68.8 83.0 75.3 82.5 78.8 93.8 78.1 82.8 80.2 85.9 81.2 Borrowedfmmbanksandatherfinancial 1.2 0.9 3.4 0.0 2.8 1.0 0.8 3.3 1.0 0.0 1.0 0.5 2.8 institutions Purchases On suppliers and 10.7 8.5 26.0 15.0 14.6 10.1 12.4 2.9 14.5 5.5 10.9 6.5 13.2 advancesfrwn customers BorrOWedfrmfami~ylfriendsandother 8.7 7.3 2.5 2.0 7.3 6.5 8.1 0.0 6.4 11.8 8.0 7.2 2.8 informalsources Source: ICA Survey A similar pattern is observed for long-term finance (Table 4.2). Retained earnings are the main source of finance (89%), with a slightly more important role of the banking sector (4%). The firms' relianceon thebanking sector fortheir long-term financedepends sigmficantlyontheir size. Large firms rely on the banking sector for 25% of their long-term finance, while small firmsrely on thebanking sectorfor only4% of theirlong-term finance. Table4.2: SourcesofLong-term Financein theFormal Sector FlBst d atsidSw.-w. -w.- %kf@lmfirmi~fran T O T A s r e l l ~ ~ F a e i ~ h L u a 7 d a ~ F o o d ad bev. uiwse Irhm;l fmWFWAnedeaTirgs 885 90.5 826 68.8 75.3 90.5 87.9 91.1 828 91.7 91.4 E 3 84.9 fran ad 51 4.0 6 1 250 159 34 4.0 9.3 4.8 0.0 1.9 4.7 13.1 i f l m Rnfrases fran alas ad 12 a7 26 6.3 4.4 0.7 1.5 0.0 20 0.0 1.8 0.0 0.0 &atx3franastarers f3mJm-J fran f;rrilyTrimk ad other 80 53 9.9 0.0 5.0 61 7.4 0.0 126 83 5.8 0.0 1.4 i ~ s c u n e s 1 1 1 1 1 I s s l s d ~ ~ d s b t 0.1 0.1 0.0 0.0 0.9 00 0.1 0.0 0.0 0.0 0.0 0.0 0.6 1 Source:ICA Survey A comparison between the sources of financeof the formal sector and of micro firms shows that they have a similarpattern, both for shortand longterm finance (Table4.3). Table4.3: Sourcesof Finance Formal SectorversusMicroFirms - Shwt-term Long-term %financingfrom Formal Micro Formal Micro In comparison 1 1 with other IlntemalhmnddRetainedearnings 81.4 83.0 1 . 5 countries (Table Borrowedfrom banksand other financial institutions 1.2 2.6 5.1 8.5 4.4 and Table Purchases on credit from suppliers and advances from 11.5 1.2 0.0 4.9, firms in customers 1 1 1 Angola have to )Borrowedfrom familyfiriends and other informalsources 6.7 2.8 6.0 2.4 relymoreon their own resources (81% for short- Source:ICA Survey term and 89% for long-term capital) than do finns in Algeria (74% and 73% respectively), SouthAfrica (66% and 58% respectively) and Namibia (60% and 74% respectively). Only DRC shows a similarpattern toAngola. In most countriesthe role of the banking sector as a source of finance ismore important than inAngola.A comparisonwith country groups yields a similar conclusion. Table4.4: Sourcesof Short-term Financing-Comparisonwith OtherCountries Sub-Saharan Resource rich % short-tern financingfrom Angda Algeria D. R. h g o SouthAfrica Namibia countries Africa countries InternalfunddRetainedearnings 81 74 81 66 67 72 70 77 BomMed from banks and other 11 3 17 6 10 11 8 financial institutions Purchases on credit fmm suppliers 11 9 15 12 24 12 14 10 and advancesfrom customers B o m d from familylfriends and 6 1 4 - 3 6 6 5 other informalswrces Issued new equityldebt 0 0- 0 1 0 1 1 2 source:ICA Surveys Table4.5: Sourcesof Long-Term Financing-Comparisonwith OtherCountries LOW income Sub-Saharan Resource rich Algeria D,R, South Africa Namibia countries Africa countries Internalfundrnetained earnings 1 85 58 74 73 71 76 88 Bolrowed from banks and Othe 6 33 20 17 22 18 financialinstitutions 3 1 2 2 2 1 Bormwed from famllylfriends end 5 8 4 7 7 6 other informalsources Issued new equityldebt 0 Source: ICA Surveys Looking at this issue in more detail, Table 4.6 shows that only 1.6%of firms in Angola have an overdraft facility and only 4.1% have loans. F h s outside Luanda report having no overdraft facilities, which may explain why they perceive access to credit as the most significantconstraint to business (seeTable2.1).As onewould normallyexpect, the percentage of large firmswith loans (60%)is larger than that of small (3%) ormedium firms (8%). Table4.6:Access to Creditinthe Formal Sector I 1 I I I I Finn *u Ormership Location Industry Manuf. - Rest of %firmswith TOTAL Small Med. Large Foreign Dom. Luanda Outside Food and - - Retail Me Luanda beveragesGarments Other universe Overdrafts 1.6 1.3 4.2 0.0 4.4 1.2 2.0 0.0 2.8 0.0 0.8 2.8 1.O Lines of credit 1loans 4.1 2.8 7.6 60.0 9.5 3.2 3.1 8.8 2.8 0.0 4.1 2.8 7.9 Source:ICA Survey From an internationalperspectivefirms in allothercountrieshavebetter accessto creditthan firms inAngola (Table 4.7). InAlgeria, 39%have an overdraftfacilityand 53%have loans; in DRC, the closest to Angola, 5% of firms have overdraft facilitiesand 6% loans. Country group comparators highlight amuch greaterrelianceon overdraft facilitieseverywhere other than inAngola. Table4.7:Accessto Credit-Comparisonwith otherCountries LOWincomeSubSaharan Resource rich % firms with Angola D, R, Congo SournAfrica Namibia countries Africa covnhies Overdrafts 1.6 38.6 5.0 68.0 26.8 30.2 35.0 27.4 I Linesof credit1loans 1 4.1 52.9 5.9 39.3 NIA NIA NIA NIA 1 Note: For South Afiica, and countrygroupcomparators,linesof creditare included in overdrafts. Source:ICA Survey-Angola,Algeria, D. R. Congoand SouthAfrica The lack of loans from banks to firmsimplies that firms rely little on banks as sourcesof finance, which confirms that access to finance is indeed an important constraint, especially for small and medium firms. Requiring collateral for loans is a widespread practice in Angola (see Table 4.8). In the formal sector, 93% of h s report it to have been demanded for their loans, whilst all micro f i m had to provide it. The value of the collateral, as a percentage of the amountborrowed, is larger for micro firms(113%of the loan) than forfirmsin formal sectors (99.6%). The most frequently given type of collateral in the formal sector is accounts receivable or inventories (42% of firms have given this type of collateral),whilst for micro firms it is personal assets (36% of firmshave given this typeof collateral). Table4.8: Collateral-Formal Sectorversus Micro Firms I 1 Formal 1% firmswhose loans requiredcollateral 1 93.4 ( 100.0 Value of collateralrequired(% loan) 99.6 113.1 Land, buildings 24.7 28.8 Type of collateralMachineryand equipment including movables 18.3 31.6 required (% of firms with affirmative Accounts receivableand inventories 41.6 11.0 answers) Personalassetsof owner (house, etc.) 30.7 35.8 Other 7.4 0.0 Source:ICA Survey Such collateralrequirementsare very differentinAngola (comparedto othercountries).As we can see fiom Table 4.9, collateral is more frequently demanded in Angola (93% of cases) than in Algeria (82%),the DRC (go%), SouthAfiica (69%) or Namibia (74%). However, its value (as a percentage of the amount borrowed) is lower than in these countries. The same is true when we look at comparatorcountry groups. Table4.9: Collateral-ComparisonwithotherCountries LOWincomeSub-Saharan Resource rich Algeria D,R. A,rica Namibia countries Africa countries % RrmsAose loansrequiredcdlatera 93.4 82.4 90.0 68.9 72.8 81.9 77.9 78.1 Value olmllateralrequired (% loan) 99.6 184.6 134.2 132.8 182.2 171.5 '161.0 173.3 Note: For South Africa,the valuesinthe tablerefer to loanspayablein morethan one year Source:ICA Surveys One reason which may help to explain the low reliance of Angola's firms on the banking sector may be the cost of debt - the interest rate. Avmge interest rates in the formal sector are approximately 6% for overdrafts, 9% for lines of credit and 8% for loans. With inflation (consumerprices) running at around 13%in 2006~~,this results in a negative real interestrate. The cost of debt cannotbe a convincing explanatory factor for the low reliance on banks for financing 26 NF,Dataand Statistics-23%in 2005and 12.9%in2006 (inflation -wnsumerprices). 4s purposes by Angola's firmssince the cost of bank credit is lowerin Angola thanin all comparator countries.(Table4.lo), Table 4.10:Costof Debt and Duration-Comparison acrossCountries Low incomeSub-Saharan Resource rich Line of credit 1 loan Angola Algeria D. R. Congo South Africa Namibia cwnbies Africa countries Averageannual interest rate 8.5 11.3 11.5 11.0 12.4 17.4 15.3 15.0 Duration (months) 48.6 25.3 35.2 70.0 66.1 34.5 43.2 47.4 Note: For SouthAfiica,the values inthe tablerefertoloanspayable inmorethan one year Sowce:ICA Surveys Our conjecture is confirmedby Table4.11: the main reason why Angola's firms (formal sector)do not apply for loansis related to the complexity of the applicationprocess (cited as the main reason by some 30% of firms). Other factors which are cited as reasons for not applying include, in descending order of importance, having sufficient capital and thus not needing a loan (19% of firms), the perception that the loan application would not be approved (18% of firms) and only thereafter high interestrates (16%). Smallfirms areparticularlyaffected by the complexity of loan applications. Table4.11:ReasonsforNotApplying for Loans - Formal Sector Firm size % firms citing as main reason for not Small Med. Large TOTAL applying for loans: No need for a loan - establishment 14.2 33.6 50.0 18.9 has sufficient capital Application procedures for loans or 34.7 18.1 0.0 29.7 line of credit are complex Interest rates are not favorable 16.8 22.8 0.0 16.2 Collateral requirements for loans or 7.1 2.3 0.0 8.4 line of credit are unattainable Size of loan and maturity are 5.4 10.6 0.0 5.4 insufficient Did not think it would be approved 19.2 4.5 0.0 18.4 1Other 2.7 8.1 50.0 3.0 Sowce:ICA Survey Concrete results of loan applications confirm the perception that collateral requirements and the complexityof applicationprocess are the 2 main reasons for a low banking penetrationinAngola. Close to 79% of loan applications in the formal sector are rejected,46% because of unacceptable collateralorco-signers and 23% because of incompleteapplications(Table4.12). Table4.12:LoanApplication/Rejection % firms applying for loansllinesof credit 16.9 15.5 % rejectedapplications 78.8 78.7 Collateralor cosigners unacceptable 46.0 55.9 Insufficient profitability 7.5 14.7 1 Problemswith credithistorylreport Reasons for reiection of loansllines of -&edit Incompleteness of loanapplication I I1 11 I1 Concernsabout levelof debt alreadyincurred O 0 I Other 5.9 0.0 I I I I Source:ICA Survey Overall,access to credit is clearly an importantconstraint forAngolan firms, especially small and medium sized firms. Looking at its two dimensions-availabilityand cost -the ICAdata leads us to conclude that cost of debt is not the main explanatory factor for this finding. The problem is related to the requirements demanded by banks, such as the acceptability of collakraVco-signers and the loan applicationitself. 4.2 The Formal LaborMarket inAngola Laborregulationsappearto be one of the least importantconstraintsfor firms inAngola (seeTable 2.1). In fact, 96%of firms in the formal sectorand 96% of micro firmsreport that laborregulations didnot affecttheir decisionsto hire or fireworkers. In a comparison across countries using Doing Business indicators (Table 4.13), we can see that labor regulations are less stringent in Angola than in the other comparator countries (except Namibia), particularly regarding the hiring of workers. The Difficulty of Hiring indexz7is lower in Angolathan inAlgeria, the DRC and SouthAfrica, and it is alsolowerthan the sub-Saharanf i c a average. The hiring cost (8% of salary) refers to the non-wage labor cost and measures all social securitypayments. It is lower than the sub-SaharanAfica average (13%), but higher than in DRC (6%) and SouthAfrica (2%). 27 Each index varies from 0 to 100,with highervalues indicatingmorerigidity. 50 However, both the rigidity of hours' index and the difficulty of firing index have relatively high values. When combined into the Rigidity of Employment Index (an average of the difficulty of hiring index, the rigidity of hours' index and the difficulty of firing index) we see that Angola has relativelyrigid labor market regulations, with only the DRC having a higher value for this index. Despite this, firms in Angola do not put labour regulations among their top concerns of the business environment(see Table 2.1). Table4.13: LaborRegulations I I Angda I Algeria D. R. Congo SouthAhlca Namibia Sub-Saharan Africa Dniiultyof Hiring Index 33 44 Rigidityof Hwrs Index 80 60 DKiuity of Firing Index 80 30 Rigidty of EmploymentIndex 64 45 Hiringcost (% of salary) 8.0 27.5 Firingcosts (weeksof wages) 58.5 17.0 Source:World Bank DoingBusiness Indicators2006 Small firms account for the majority of total employment in Angola's formal sector firms - approximately 61% of the total. Medium firms account for 27% and large firms for 11%. Since 2002, the majority of h s have added workers - 57% of small firms and 80% of large firms. Thus, an average firm in Angola now has 14.3workers compared with 13.3 in 2002 (see Table 4.14). This growth in the numberof employeesis clearly more felt for firms based in Luanda. Table4.14: Employment: Full-time PermanentWorkforce -AU Formal Sectors I I I 1 I I Flm size Ormershlp Location Industry I I ;ecm;;t Rest c Small Med Large Foreign D m , Luanda :u:yz rnvt p:;n~EE: - Retail the - universe I ( 1 1 1 14.3 IEmployment perfirm - 2006 10.2 31.0 122.01 19 1 13.5 14.5 13.5 17.4 8 6 17.0 10.5 128 Employment perfirm - 2002 1 1 1 1 13.3 8 6 240 108.4 20.2 12.3 13.4 13.1 14.9 6.8 16.5 10.2 112 Percentageof firms addingworkers 511.2 56.5 63.8 80.0 63.3 57.5 69.2 27.5 57 7 68.6 66.4 41 9 56.9 Source:ICA Survey Full-time seasonal or temporary employment is, by comparison,not very common.As we can see from Table 4.15, the average h in Angola employs 2 full-time temporary workers, 13% of which are female. Foreign firms rely more on that type of workers, employing on average 4.4 workers. Firms in the retail sector also rely relatively more on temporary full-time workers (4.6 workers on average). Table4.15: Employment:Full-time SeasonabTemporaryWorkforce All Formal Sectors - Firm size Ownership Location Industry :z: :zgtf'- Manuf. - Rest of TOTAL Small Med Large Foreign D m . Luanda Fttm;;d Retail the universe Employmentperfirm 2.0 2.0 2.4 1.0 4.4 1.6 2.5 0.0 1.4 I.4 1.2 4.6 1.2 % femaleworkers 13 12 33 2 28 10 13 NIA 8 1 5 15 33 Source: ICA Survey The workfo~ein the manufacturing sector consists mostly of production workers (79% of the total), 14%of which are female and 81%of which are skilled.The female share of the production workforce is more significant in the manufacturing of garments (51% of the total). The female share of non-production workers is larger than that of production workers, and this is true for the various firmsizesor sector of activity (see Table4.16). Table4.16:Descriptionof Workforce-ManufacturingSector T Firm sue 1 Ownership I Location I Industry Food and Manuf. - Manuf. TOTAL Small Med. Large Garments Other beverages - % of non-productionworkers % female non-productionworkers % productionworkers % female production workers % skilled productionworkers Source: ICA Survey Most of the workforce in the manufacturing sector is not unionized. From Table 4.17, we can see that only 7%of the workforce is unionized, mainly in medium and large firmsand in firmslocated in Luanda. Table4.17: UnionizedWorkforce-ManufacturingSector Firm size Ownership Location Industry Manuf. - TOTAL Small Med. Large Foreign Dom. Luanda Food and !:zLn,- !E: - beverages Percentageof unionizedworkforce 6.8 2.6 25.8 45.0 6.8 6.8 7.2 4.9 9.8 0.3 6.3 Source:1CA Survey From Table 2.1 (all formal sectors) and Table 2.2 (manufacturing sector) we also fmd that the workforceeducation is not perceived to constitute a significantconstmint.We look deeperintothis issue in Table4.18, which shows the averageeducationalattainment intermsof years of education of a typical production worker in the manufacturing sector. More than half of the firms (81%) report their typical production worker to have between 7 and 12years of education. Roughly this may correspond to complete secondary schooling and some university training. One third of the firms have a workforce with only primary education completed. Only 14%of firms say that their typicalproduction workerhas 3 or feweryearsof education (incompleteprimary schooling). Interestingly, firms outside Luanda clearly report that their typical production worker has fewer years of education than the average: more firms report fewer than 3 years of education (26% compared to the 14% average), 4-6 years (52% compared to the 33% average) and 7-12 years (19% compared to the 48% average). Despite this, firms outside Luanda perceive an inadequately educated workforce to be less of a problem than firms located in Luanda (see Table 2.2). Thismay be related to the type of firm and itsparticular needs in terms of worker's education: firms located outside Luanda may require a less skilled worldbrce, which explains why they perceive education not to be a significant constraint. By contrast, firms in Luanda may require a more skilled worldbrce which, in 31% of Luanda's firms (the percentage of firms which consider an inadequatelyeducatedworkforce to be a major or very severeconstraint), they cannot find. Table 4.18:Average educationalattainmentof productionworkers-ManufacturingSector F i n size I ownership I Location I Industry I Manuf. - TOTAL Small Med. La-rge Foreign Dom. Luanda Outside Eztn& 0":; Luanda Food and - beverages 0-3 years of education 14.1 13.1 22.3 0.0 11.7 14.3 11.8 26.1 15.5 0.0 15.6 4-6 years of education 32.6 33.1 28.4 40.0 33.4 32.5 28.8 52.4 28.8 25.4 35.8 7-12 yearsof education 48.3 48.8 43.0 60.0 49.4 48.2 53.9 19.3 47.6 68.8 45.4 >I3years of education 1.9 2.3 0.0 0.0 0.0 2.1 2.3 0.0 1.4 5.7 1.6 Source: ICA Survey In a comparison with other countries, presented in Table 4.19, we see that the educational level of workers in Angola is considerably lower than that of all comparators (with the exception of Algeria): workers in Angola are more likely to have fewer than 6 years of education than in all country and group comparators.Additionally, the percentage of workers with more than 13years of educationofAngola is much lower than in all countryand group comparators. Table4.19:Average educational attainmentof productionworkers-Comparisonacross Countries -ManufacturingSector Low Sub- Resource Angola Algeria D.R. Congo SouthAfrica Namibia income Saharan rich countries Africa countries <6 years of education 47 58 38 10 NIA 11 10 15 6-12 years of education 48 32 52 78 NIA 78 72 70 >13years of education 2 10 11 12 NIA 13 12 15 Note: For Algeria, the values in the table refer to qualifiedworkers Source: ICA Surveys The question one must ask is why do firms in Angola not complain more vehemently about an inadequately educated workforce (fiom Table 2.1, only 21% believe this to be a major or very severe constraint). One possibility is that the production process of most firms does not require such skills. This would occur if firms have adapted their production technologies to the level of education of the workforce. Another possibility is that firms compensate such low levels of educationwith trainingprogrammes. Training is often a complementfor education, which could help explain why workforce education is not identifiedas a significantconstraint Table4.20 showsthat 19%of manufacturing firmshave offered training to their employees, mostly production workers. Foreign firms are more likely to offer employee trainingthandomestic ones, and the larger the firm is, the more likely it is to offer training to its employees. Although very few firms outside Luanda identify an inadequately educated workforce to be a major or very severe constraint to business (see Table 2.1 and Table 2.2), 25% offer training to their workers (compared to 18%of firmsin Luanda). Such training is particularly targeted atproductionworkers. Table4.20: FirmsOffering'kaining -ManufacturingSector F i n she Ownership ~ocatic-n lndusby Manuf. - TOTAL Small Med. large Foreign Don. Ez:i [;oo&andpzkig$' Luanda ges % of h s offeringtraining 19 14 37 80 50 . 17 18 25 17 17 21 % productionworkers receivingbaining 53 44 74 46 61 50 40 100 86 35 40 % non-productionworkers receivingbaining 12 7 26 11 33 7 13 9 22 0 10 Source: ICA Survey In an international comparison (Table 4.21), we see that firms in Angola offer training less frequently than in all other comparators (with the exception of D.R. Congo). However, when we lookat who receives training,production workers inAngola receive trainingto a largerextent than those comparators. Table4.21:FirmsOffering'kaining-ManufacturingSector-Comparisonacross Countries Sub-Saharan rich Angola Algeria 0.R.Conga SouthAfrlca Namibia cwntries Africa camtries YOof firms offwngtraining 19 32 11 64 44 35 40 31 1 1 1 I I I I 1 1 I % productlmworkers recekingtraining 61 tVA 63 45 48 7 22 19 % nm-productionworkers receiringtraining 12 NIA 37 47 69 21 27 22 Source:ICASurvey This analysislends somesupportto the conjecturethat despite the relatively low level of education of the workforce, firms might perceive this not to constitute a problem if training were indeed a complement of education. Absenteeism is a problem in Angola. As we can see from Table 4.22, 44% of firms report high absenteeism due to sickness and 26% due to HIVIAIDS. The results obtained through the employee questionnaire show that 19% of workers have been sick during the last 30 days, on average missing 4.3 days of work. They also show that 63% of workers see HIVIAIDS as a big concern. Table4.22:Absenteeism-AU Formal Sectors Firm size Ormenhip Location Industry Lcm;i Manuf. - Rest of % reminD high E:!: TOTAL Small Med Large Foreign Drm Luanda absenteeismdue to E$ - Retail lheunivefse Om sickness 44 44 45 40 36 46 33 100 51 48 39 44 47 Familyffriendssickness 12 13 10 20 19 11 15 0 7 0 11 18 17 HIVIAIDS 26 25 28 40 21 27 10 100 35 20 19 26 30 FamilyffriendsHIVIAIDS 4 4 0 0 9 3 5 1 0 0 1 0 4 8 2 Source:ICA Survey Table 4.23 shows that firms are concerned with HTV: 61% of firms have helped spread I-lW prevention messages. However, a much lower percentage has undertaken other important activities, suchas h e condom distributionor anonymousI-lW testing. Table.4.23:HlV PreventionActivities-AU Formal Sectors Firm size Ownership Location Of which Outside TOTAL Small Med. Large Foreign Dom. undertook: Luanda Luanda HIV prevention messages 61 61 62 80 56 62 54 98 Free condom distribution 31 30 39 20 36 30 31 33 Anonymous HIV testing 6 5 10 0 10 5 7 0 Source:ICA Survey In order to ascertain the determinants of wages, we have estimated four basic wage equations using Ordinary Least Squares (OLS) (see Table 4.24). Such regressions allow us to quantify marginal returns to education or experience, as well as other worker (e.g. gender) andlor firm characteristics(e.g.firm size).28 In the first model, we attempt to explain wage levels as a hction of years of education (schooling), years of past experience, gender and the number of hours worked per week. In the second model, we introducean additional explanatoryvariable: whether the worker has undergone - formaltraining in the past. 28 We have excluded 5 observations frmn the sample: 2 because of extremely low reported salaries and 3 because of exmmely high reported salaries. The third model introduces other possible firm-specific explanatory factors, namely firm size and whether the firm is foreign-owned.Finally, in the fourthmodel we control forindustry.29 Table4.24: Determinantsof Earnings-ManufacturingSector I Dependent variable: Log annualwages Std. error I Experience Std. error Female Std. error Log (hourslweek) Std. error 0.159 "* 0.158 "' 0.157 *** 0.157 no Pasttraining 0.276 0.260 Std. error 0.063 "* 0.060 "* 0.059 *** Constant 2.872 3.213 2.743 2.360 0.635 " 0.641 "' 0.645 "' I Std. error 0.064 "* 0.070 Firm size (large) 0.721 0.659 Std. error 0.133 *** 0.133 "* Foreign 0.450 0.438 Std. error 0.083 "' 0.082 "' Sector:manufacturing - other 0.214 Std. error 0.065 "' Observations 484 484 466 466 (***)Statisticallydifferent from 0ata 1% significancelevel (**)Statisticallydifferentfrom 0ata5%significancelevel (*)Statisticallydifferentfrom0at a 10%significancelevel Source:ICA Survey In all models, education appears with clearly positive marginal rates of return: every additional year of education increases a worker's salary by 2.3 - 3.7% (across models). Assuming all other factorsare constant, a female worker typically receives a salary which is 12.2 - 17.8% lower than that of a male worker. Experience appearsto be a significantexplanatory variable in models 1and 2, but once firm-specific characteristicsare controlled for, its coefficient is no longer statistically different from zero. Training yields high returns to workers: those who have undergone formal training courses in the past have salarieswhich are 26 - 27% higherthan those who did not. 29 Severalh e rspecificationswere testedusingunion membership,yearsofwork in thefirm(tenure),Ill-time workersoitus,whetherthe workeris currentlyreceiving trainingand iirm location,whose coefficienaarenot significantlydifferentlium zero, i.e.appearnotto explain salarylevels.Often schoolingor experiencepment decreasingmarginalreturns,i.e. contributeless to salaryincreasesas tfieir level also increases.This was tested by huducing bahvariablessquared in there-ion, but k i r coe5ciena were not statisticallydiffmt fium 0.Othersectordwmnieswerealsotested, bu~theircoe5ciena werenot statisticallydifirentfium 0. Firm size appears to be an importantexplanatory factor of wages, particularly if a firm is large. If this is the case, salaries are some 66 -72% higher than in small firms. The sameresult is true for foreign firms, which (assuming everythmg else to be constant) pay salaries that are 44 - 45% higher than in domesticfirms. Theseresultsmay help explainwhy the level of educationof the workforce isnot perceivedto be a significant constraint. Indeed, training appears to be a good substitute for education: in terms of salary, workers who have undertaken training courses in the past have a salary which is higher than those who have not. In addition,suchtraining coursesare equivalent, in terms of salary, to an additional sevenyearsof education,which is substantial(averageeducationis 8.2years). 4.3 Land Market Overall, access to land is not perceived as a top constraint to business in the ICA survey. In fact, micro firms rank it fourth (afterelectricity,access to finance and transportation),while firms from the formal sectorrank it eighth,as shown in tables 2.1 and 3.1. However, one of the main reasons access to land is currently not regarded as a major problem is that f m s - as well as the general population - tend to consider the land they occupy as their property. Therefore, it is highly probable that access to land will be regarded as a seriousconstraint to business in the near future, as firmsbecome more awareof the potential impactof recent legislationon land rights. The Angolan legal system is derived fiom the Portuguese colonial system, adapted after independence. The Law No. 2030 of 1948was the basis for land law. The new Constitution of 1975, on the other hand, established the overall right of the state to all land, which could be transferredbased on itsuse. The Law 21C of 1992regulated the concessionof land foragricultural use, on the basis of surface land use rights. Also in 1992, the Decree 46A gave the Provincial Governmentthe right to concession,including in urban areas. In 2002, the process of creating a new Land Law began, and a firstdraft was finalized in July that year. The Law (Lei de Terras-Law No. 9 of 2004) was finally published in November2004. It is worth stressing out the difficulties involved and the length of the process of creating legislation regardingan issueas essential toeconomicdevelopmentand stabilityas are land rights. It must be noted, however, that the actual land occupancyand use have littleresemblance to what is established in the legal framework: In fact, the findings of a field research undertaken in the peri-urbanareasof Luanda showedthat 43% of respondents had no awarenessof the legal concept of land rights, and only 13% had a reasonable awareness of land rights issues (Development Workshop,2005). Since independence, Angola's population distribution has changed dramatically. As living and security conditions deteriorated, in particular during the civil war, there was a clear exodus fiom inland towards the main urban areas. In Luanda, in particular, there has been an exponential population growth since independence, not for demographic changes, but due to voluntary and forced migrations from other parts of the country. There were two periods when this trend was most obvious.These were immediately after independence, and when civil war reached its peak. Thereare no official numbers for the 1970sand 1980sbut, between 1995and 2000, it is estimated that the population in Luanda increased from 2,070,000 - with an overall average growth rate estimated at 6.7% per annum - to 3,150,000 - with a growth rate estimated at 8.23% per annum (DevelopmentWorkshop, 2003).And the tendency is for this growth to continue, eventoday. As an emergency measure, duringthe last thlrty years land has been allocated extensivelyin peri- urban areas, as inner areas were already occupied. Whilepen-urban areas grew rapidly, the legal, administrativeand technical capacity of the statewas severelyaffected.In addition,due to war, the national cadastre was taken over by the Ministry of Defence, whose main priority was naturally not urban developmentand, as a result, the cadastrebecame completelyout of date. The Constitution defines that land formally belongs to the state, with the exception of a limited number of entitieswhich had full freehold land rights before independenceand did not lose them since then, namely through abandonment, confiscation or nationalisation. The new Land Law confirmsit, namelyby stressingout in its 6th articlethe point that acquisitionof land by usucapiCo (i.e. the right to occupy land by virtue of a factual occupation fora given period of time) is illegal. This decision was taken despite strongpressure fiomthe various economic agentsinvolved forthe state to recognise that, after thirty years of informal occupation, a legal h e w o r k should be developed to providethe residentstheright to be where they are. In Luanda, some formal market-related activity does exist: after venfylng actual occupation, the cadastre is checked (despite being completely out of date) and, if it contains a previous land registry, there needs to be a public announcementto confirmthat is has been abandoned or if there are still legitimaterights (DevelopmentWorkshop, 2003). Even so, as statesupplyof land has been virtually non-existent, the large majority of urban dwellers have accessed land through informal mechanisms. However, given the conditions in which it happened - often involving local administrativeinstitutions - and the time elapsed since then, most of them considerthey have valid rights to the land. In fact, an active informal market is in place, the most common mechanism being the purchase of land with witnessed purchase documents, and such land occupations and constructions therefore do have some legitimacy. As a result, Development Workshop (2005) observed that, in Luanda, around 80% of those interviewed had occupied their land through informal mechanisms, while 86% felt secure of their land. On the other hand, if we look at the results from the ICA survey, we can see that, in Luanda, around 66% of firms from the formal sector and 46% of micro firmsbelievethey own the land they areoccupying. There are serious limitations on what the government can do in moving to formal rights in the short term.However, there needs to be a distinction between land rights and land titles and, in order to gradually evolve to the latter, it is necessary to emphasise the former (Development Workshop, 2005). At present, firms can only legally register the land if they acquire it from the state or from someone with full freehold land rights. This may explain, for instance, why land is not often used as collateral in loans. On the other hand, if we look at the results from the ICA survey in greater detail, we can see that foreign firms fiom the formal sector already consider access to land to be their fifth constraintto business, after conuption,electricity,crime,theft and disorderand access to finance. This is not surprising, as the Land Law determines, in its 35th article, that only Angolan individuals can buy land from the state. This means that foreign firms can only obtain property rightsby acquiring alreadyprivately-owned land, which is rather scarce,as we have seen. Even today, at a time of peace, it is unlikely that a substantial part of the urban population - and firms, in particular - will move back to rural areas, as they wish to enjoy the benefits of proximity to other productive units, labour market. and commercial opportunity, especially until transportation difficulties, identified as a major constraint for finns located outside Luanda, are overcome. This will create furtherpressure on the land market (both formal and informal), prices will continue to rise, and conflict over occupation rights is likely to increase. Therefore, the land occupationregularisationprocessmust be a priority. 5 MANUFACTURINGAND FIRM'S PRODUCTMTY Before establishing the effects of investment climate constraints on firm's productivity it is important to grasp Angola's position both internationally, by comparing with relevant countries, and nationally, by investigating differences across a number of firm's characteristics. For the international comparisons we use median values for Sub Saharan countries where Investment Climate surveys have been camed out recently. The aggregate values by income levels and resource abundance exclude SouthM c a , because not all variables for comparison are readily available. Nevertheless we also report the closest results for SouthM c a separately so we have a fullerpicture forcomparison in the region. 5.1 LaborProductivity andLabor Cost Labor productivity is measured as value added per em loyee. From Figure 5.1, we can see that Angolan firms perform better than other resource rich3 countries in sub-Saharan Africa, but are r stillwell below otherlower middleincomecountries in the region, particularly Namibia and South Afiica.As it is recognized, productivity in South~ f r i c ais~considerably higher than in the rest of ' Sub-SaharanAfiica. Labor productivity in Angola, however, comparesfavourably with countries such asTanzania,Rwanda, and the DemocraticRepublic of Congo. Figure 5.1: LaborProductivity-International Comparison 1 Value Added per Worker Labor productivity in the Angolan manufacturing sector is higher in foreign owned firms (see Table 5.1)~~.We can also see that firms established in Luanda are more productive than those in the other sampled cities (Benguela and Huambo). 30 The set ofresource rich counmes includes Angola, BOB- DRC, Namibia and Uganda The lower middle income group contains Angola, NamibiaandSwaziland 31 See SouthA6ica: anAssessmt ofthe InvestmatClimate (2006). WorldBank 32 Median value. Table5.1LaborProductivity(US Dollars) Manufacturing - F i n size Ownership Location Industry Value Manuf. - added per TOTAL Small Med. Large Foreign Dom. Luanda Other Food and - Other - worker beverages Garments 8339 8652 6584 8393 9502 8232 9045 5046 5780 5169 10336 2o05 5436 5355 5148 8168 7224 5144 5627 4470 4894 5144 5557 Source: ICA Survey Note:The fmtrow inthetablerefmtothemean andthe secondvaluetothemedian. Next we turn to investigating the labor costs. We measure them as the total costs of workforce divided by the total number of employees (including both full time and temporary workers). Figure 5.2 shows that labor costs inAngola are high in the region, being only below SouthAfrica andNamibia. Figure5.2: LaborCostperEmployee-InternationalComparison Labor Cost perWorker I $ p@8 . 4 . Firms in Luanda a *8 ~ " 9 *a9. (pa +QO +* . * * '@& P a* @ GQ face the highest 5 * c p p @ 6 B @PGo@ @+eq. ** V\+ nV labor costs of the 4 1 4d6$.+ 4p@ #+\6 &5*+B0d*&&* P I * a P 5 + * 1 %s* & %O* regions sampled,but 9 eC 5*9 as we can see in I 8' I G 1 I Table 5.2 the labor in Angolan manufacturing firms does not vary considerably across the different dimensions used in this analysis. 33 Median value Table5.2 LaborCosts perEmployee(USDollars) Manufacturing - F i n size Ownership Location Industry Manuf. - TOTAL Small Med. Large Foreign Dom. Luanda Other Food andManUf' - beverages Garments Other 3377 3470 2806 3720 4267 3296 3553 2561 2928 3018 3701 2005 2740 2747 2687 2740 2687 2751 2762 2627 2653 2922 2751 Source: ICA Survey Note: The firstrow in thetablerefen to themean and the secondvalueto the median. A better way to understandthe weight of labor costsin firms' operationsis to study their unit labor costs, that is, look at total cost of manpower as a Figure5 3 Unit LaborCosts-InternationalComparison proportion of value added. Unit Laborcost This indicator is quite useful for international com~arisonsas it does not depend on national currency units and is, therefore, not subject to exchange rate movements. It is clear from Figure 5.3 that Angolan firms face the highest labor costs in the region. This is expected in countries that rely heavily on labor intensive production methods. As a matter of fact the levels of capital intensity in Angola are abovethosereportedin the DRC,but well below many othercountriesin the region (Figure 5.4). Figure5.4: Capital per Employee-InternationalComparison I Capital perWorker . . 5.2 Total FactorProductivityand Investment ClimateDeterminants We now look at total factor productivity, which is the productivity of h s after both capital and labor have been taken into account. Figure 5.5 shows total factor productivity (TFP) across countries as a proportion of TFP in SouthAfrica. Angolan firms' perform rather well in terms of TFP. They perform at 40% of the productivity of SouthAfrican fms. Angola appears in a better position than most Sub Figure5.5: Total FactorProductivity-International Comparison SaharanAfrican countries. -- Using the methodology o,: described in Annex A we 0.8 estimate a Cobb-Douglas 0.7 0.6 production function for 0,5 Angolan manufacturing firms. 0.4 We proxy capital by the ,, 0.3 product of average capacity 0.1 utilization and the book value O of machinery and equipment, and measure labor by the total number of employees in the establishment. Results are reported in Table5.3. We find that capital elasticity is almost four times lower than labor elasticity. We tested for constant returns to scale in production and we cannot reject the hypothesis that the sum of those elasticitiesof inputs is equal to one. Therefore,there is strongevidenceof constantreturns to scale in productionin the sampled firms.Thismeans that the size of firmsis not immediately a predictor of productivity. We see also see that if a firm is new (lessthan 5 years in operation) it has higher value added. Firms in the food Table5.3Production Function Estimates and beverages sector generate I Coef. SE P-value lower value added. Labor 1 0.778 ("') 0.073 0.000 -Capital 0.193(^**) 0.022 0.000 Firms in Luanda are more Age of Firm 0.234 (***) 0.089 0.008 productivethan firms locatedin Manager's Education -0.136 0.099 0.167 the rest of the country. Foreign Manager's Experience 0.070 0.086 0.413 owned firms are more Food -0.258 ("') 0.085 0.003 Garments -0.145 0.132 0.272 productive. These results Constant 10.0071(***) 0.392 0.000 corroborate the initial findings Observations 159 of labor productivity. (Table R squared 0.486 5.4) The only counterintuitive Test CRS: F stat 0.16 result relates to the influence of Prob>F 0.685 firm size, as we find that (***)Statisticallydifferent 6orn 0at a 1% significancelevel medium sized firms have the Source:ICA Survey highest average TFP. However, this couldbe due to the factthat the number of large firms in the sampleis very low. Table 5.4TotalFactor Pmductivity - - F i nsize Ownership Location industry Manuf. - TOTAL Small Medium Large Faeign Domestitic Luanda Other Food and Garments-Manuf' - Other Average 1.321 1.322 1.404 0.876 1.417 1.312 1.438 0.940 1.185 1.145 1.440 - Median - 0.941 0.942 1.300 0.840 1.300 0.917 1.077 0.859 0.958 1.045 0.921 source: IU\ survey We established in section 2 that fims perceive electricity, access to finance, corruption, business licensing, and transportation as serious constraints affecting Angolan h s . We now attempt to quantifj~the &pact Table55ExtendedProduction FunctionEstimates of each of these Model 1 Model 2 investment climate Coef. SE P-value Coef. SE P-value factors on measures Labor 0.846("') 0.123 0.000 0.825("') 0.123 0.000 Capital 0.196y) 0.024 0.000 0.192('") 0.024 0.000 of productivity. We Accessto Finance estimate an extended Finance -0.143 0.228 0.529 -0.158 0.228 0.489 PV -0.037r) 0.018 0.040 -0.032(.) 0.018 0.078 production hnction Electricitv specification Power0;tagee -0.081("') 0.022 0.000 -0.192("') 0.053 0.000 PV -0.012 0.016 0.438 4.005 0.016 0.758 including both I / I BusinessLicenses objective measures34 ConstrctionPermit 0.227 0.216 0.295 0.266 0.216 0.221 Operation License -0.687("*) 0.264 0.010 -0.633(") 0.263 0.017 of each constraint PV -0.027 0.024 0.256 -0.010 0.023 0.655 and the fims Corruption Informal Payments 0.017 0.012 0.157 0.008 0.012 0.497 perception of such Pavmentsto Officials -0.001 0.010 0.915 0.000 0.010 0.991 constraints (denoted - , PV 1-0.066~) 0.024 0.006 ( - 0 . 0 7 4 ~ ) 0.024 0.002 Crime. Then and Disorder I I pv3'). Two models are estimated: in Crime 0.280(") 0.127 0.028 PV -0.028 0.024 0.245 model 1 we test the Transportation statistical Bmakage 0.010 0.019 0.616 r -0.002 0.021 0.939 significance of objective measures Generator of each constraint as Garments well as firms' Age of Firm ForeignOwnership perceptions; in model MediumSize 2, we use the average LargeSize Luanda of the objective Constant 110.324(***) 0.498 0.000 110.695("') 0.519 0.000 measure of the 159 1 159 I R Observations squared 0.590 1 0.589 1 constraintby location (***)Statisticallydifferent from 0 at a 1%significancelevel and sector of the (**) Statisticallydifferentfrom 0at a 5% significancelevel h s to address (*)Statisticallydifferentfrom0 at a 10%significancelevel Source: ICA Survey endogeneity concerns. Table 5.5 34 ThesemeasuresareexpkinedindetailbelowandinAnm A. 35 PV is an indicator based on the mking of the peneption ofthethreemostserious obstacles. For each finR the most seriousoktaclewas givena weight of3,thesecondmost seriousawight of2 andthethird mostseriow a weight of 1. reportsthe results. Access to finance is measured by a dummy variable that takes the value of one if the firm has an overdraftor bank loan. Here and for the remainder of this section, PVrepresents the correspondent perception of the firm as this being a major or very severe obstacle to business. We see that only the perception variable has a negative impact on productivity. However, it is important to remember at this point that the penetration of access to bank finance is very low in the sampled h s . Overall less than 3% of h s have access to any form of bank finance making it very difficult to find any strong effects. Firms that perceive access to finance as a major constraint are 4% lessproductive. Problems with electricity are measured by the natural logarithm of hours of power outages per month. There is a strongnegative impact of these problems on productivity.We can see that these a decreaseof 1% in the number of hours of production lost due to power outages leads to an 8% increase in TFP. It is apparent that even the high usage of private generators does not mitigate the lossesin productivitycausedby a deficientelectricity supplynetwork Business licenses and permits were also identified as a serious constraint by firms. To measure this, we include binary variables that assume the value of one if it takes over 28 days to obtain a construction pennit and an operational license. We find that the delay in obtaining operational license has a significant andnegative effect on total factorproductivity: h s which have to wait a long time (more than28 days) to obtain an operational license showa level of productivity which ismore than half of the firmsthat donot wait that longto obtainthe licence. Corruptionproblems are approximated by percentage of total sales spent in informal paymentsto get things done and by percentage of sales paid to officials to secure government contracts. The effects of such variables are weak This may reflect the fact that the proxies used do not hlly capturethe corruptionproblem facedby h s . Howeverthe perceptionof corruption as a problem is significantly correlated with TFP in both models labor productivity. Firms that consider corruptionas amajorproblem are 7% lessproductive. Finally problems with transportation are proxied by proportion of value of shipments lost due to breakage or theft while in transit. As with corruption, it is the perception variable that has a consistent negative impact on productivity. Firms complaining about transportation are approximately 10%less productive. The productivity analysis provided above shows that reforms on improving the access to credit, securing a reliable supply of electricity, accelerating the business licensing process, increasing transparency, and improving transport links will have positive effects on firm productivity in Angola. 6 SYNTHESISOF RESULTS AND POLICY RECOMMENDATIONS 6.1 Constraintsto Business Indirect costs inAngola are relativelyhigh: approximately 10%of total salesin the manufacturing sectorand 8% for all sectors.Micro firmsfaceslightlyhigher indirectcosts, totalling 12%of sales. In the formal sector, electricity is the main driver ofsuchcosts. Firms in the formal sector have identified the followingconstraints to business: access to credit, electricity, crime, corruption, and business licensing. Transportation is an important constraint for h s located outside Luanda. Micro firmshave identifiedbroadly the same constraints to business (with different orders of importance), with one exception: access to land appears in their top five concerns.In comparison with other countries, h s in Angola and the DRC have broadly similar main concerns (electricity and accessto finance), but in the latter, macroeconomicconstraints are very important whilst inAngola they arenot. In terms of infrastructure,electricity-relatedproblems aremostly duetopoweroutage duration and frequency. Generators, a costly alternative, become thus necessary for firms to operate, and provide some 31% of total electricityneeds. In comparison with the DRC, Algeria, SouthAfrica, Namibia and comparatorcountry groups, such outagesaremorepronounced inAngola. Business licensing in Angola is a time consuming process: a consfmction permit takes some 42 days and an operating license 24 days. This is both worrying in absolute and relative terms, as firms in the DRC, Algeria, South Africa and Namibia have generally speedier and less costly businesslicensingprocesses. Conuption is also a serious constraint, with firms' perceptionsbeing that government officials do not have a consistentinterpretation of the law, Overall, firmsmake informalpayments worth some 3.3% oftotal yearly sales. 6.2 FinancialMarket Despite strong growth of credit to the private sector in the past three years, the depth of the Angolan financial sector is still very shallow with 11.1%MUGDP (compared to an average 27% in low-income countries).Access to finance as a constraint is mirrored by very low financial sector intermediation with private sector credit accountingfor only 8% of GDP in 2006 (compared to an average of 23% in low-incomecountries). Angolan firms rely essentially on internalfunds to finance themselves,with borrowing (e.g. fiom banks) accounting for only 5% of total long term financing needs. Regarding working capital needs, borrowingis even less common. In comparisonwith othercountries,the role of the banking sectoras a source of finance is less pronounced in Angola. Thisconstraintis confirmedby the fact that a very small percentage (less than 4%) of firms has overdraft facilities,lines of credit or loans from banks, clearly lower than in DRC, Algeria, SouthAfrica, Namibia and compamtor country- groups. Lack of access to finance coincides with high liquidity in the banking sector and the banking systemas a whole fails to intermediategrowingdeposits into credit.Net credit as share of deposits declinedto 38% in 2005.TheAngolan financialsectorremains highly dollarized with 44% of total bankingsector creditand 52%of depositsdenominated in foreigncurrency. Based on sound macroeconomic policies interest rates have been declining significantly (along with inflation) and cost of debt does not appear to be a critical factor limiting firm's access to finance. On the contrary, high access thresholds as documented by complex credit application procedures and high collateral requirements are stated to be the main reasons limiting access to financeby firms. 6.3 Productivity Firmsin Luandahave higher levels of labor productivity than in the rest of the sampled cities even though they face higher labor costs. Labor costs average 54% of value added revealing the prevalence of a labor intensivemanufacturing sector. Angolan firms use low levelsof capital per employee. To complete the picture about productivity in the surveyed firms, estimates of Total Factor Productivity reveal that foreignowned firmsarethe most productive. Investment climate constraints identified as the most serious by the interviewed establishments tend to have a negative impact on productivity,The food and beverages, and garment sectors are lessproductive than thereference one (othermanufacturingfirms). Moreover, foreignowned firms and the ones located in Luanda aremore productive thandomestic firms and those locatedoutside of Luanda respectively. The evidenceon the effects of business climate constraints on productivity shows that reforms on improving the access to finance, improving electricity inliastructure, speeding up the business licensing process, on increasing transparency and improving transport links will have positive effects onproductivity. 6.4 Policy Recommendations 6.4.1 Finance A lot of progress has been made in improvingAngola's payments systems over the last few years. Nevertheless increasing financial depths will require improving the lending environment and encouragingbanks and non-banks to expand financingto the private sector.Rather than providing credit banks in Angola prefer to charge fees fiom transactions, to provide short-term trade financing and to invest in high-yield government bonds. Consequently, the development of Angola's private sector is constrainedby Angola's financialenvironment. - Improvingthe lending environment will be crucial to encourage credit expansionand at the same time avoid undue accumulationof risk in the financial system. Financial institutions inAngola are reluctant to expand credit given their limited ability to identify borrowers with good credit risk (lack of accounting frameworks, market information and credit information) and the ability to enforce contracts and secure collateral (inefficient creditor rights and enforcement, lack of secured lending frameworksand collateral registries).In addition, sustainablefinancial sectordevelopment will require strengthening of bank and corporate governanceframeworks and a robust risk-based supervisory framework. While low financial sector development and lack of access to finance impedes growth in general, low financial sector intermediation capacity poses a particular problem in Angola given the rapid growth of oil revenues in the country. Enhancing the capacity of the financial sector to intermediate funds into productive investment will be crucial for creatingabsorptive capacity and managing the risk of Dutch Disease'. Financial sector development could be a driver on non-oil growth in the private sector. However, credit expansion and investments require 'bankable' projects and financial sector reform should not be seen separate fiom general investment climate reform. The Angolan Governrnent has provided a hmework for future financial sector reform with the passing of the New Law on Financial Institutions and the Securitiesand Exchange Law in 2005. The authorities have identified a broad range of policy areas to address current constraints to accessto finance,including: o HumanResourceCapacityand skiisDevelopmentin the financial sector o Business Registrationandidentificationsystems o CreditRegistriesandCollateralRegistries o CreditorRightsandEnforcement o Long-Term finance,includingreal estateandmortgage finance o DepositInsurance Systems o Microfinance o DiagnosticWorkoninterestrate spreadsand MonetaryPolicy implications A concerted effort by the various Government agencies and development partners to implement reform in the identifiedpriority areaswill key to addressconstraintstoaccessto finance. The BNA initiativeon expandingaccess to creditcontains a significantnumber of relevantpolicy actions targeted in particular at constraints in the lending environment and financial sector capacity. Short-tern recommendations: 1. Enhance credit information inhstructure. The existing public credit registry has very limited coverage and primarily serves supervision puposes. The recent WB/FIRST/IFC review of credit information infrastructurein Angola has found a broad consensusamong banks and regulatory authorities for expanding its functionality under the supervision of the BNA and upgrading the system based on existing international best practice applications.In the short-run,an efficientsolutioncould be to outsource operation,but not ownership,of the registry to an experienced international credit bureau operatoron behalf of the BNA. Later, with a growing credit market, the entry of private credit bureau operators (potentially on shared platforms within SADC) should be enabled and encouraged. 2. Upgrade coporate registries, collateral registries and public record systems. ?he scope of financial information &structure should include efficient access to corporate information,registries of secured lending chargesand courtrecordsetc. 3. Computerize property registrationprocess, simplifytaxes and fees, and make optional the involvement of notaries soas to improveproperty registration.Efficient land registries and the ability to easily perfect and transfer land titles are an important vehicle to provide property owners with access to collateralized financing. On average property registration inAngola requires 369 days comparedto the regional averageof 114days. The lengthmess ofAngola's propertyregistration process,i.e.more than three times the regionalaverage,is principallya result of the 300 daysrequired to receive definitive registrationh m the Real Estate Registry.Backlog andpaper-basedrecords necessitate that all historyof transactions relevantto thepropertymust be checkedeverytime. 4. Clarify the property rights in the new land law in order to reduce codhion and help identify sourcesof collateral. 5. Review standards and codes on creditor rights and insolvency regimes, including the enforcementof property rights and the efficiency of commercial court procedures and the establishmentof alternative disputeresolutionmechanisms. Long-term recommendations: 1. Conduct the planned FSAP. The Financial Sector Assessment Program (FSAP) will provide an appropriate framework for more detailed diagnostics on financial sector constraints with a view to enhance access to finance and increase financial sector intermediation capacity. The FSAP should also provide guidance on addressing issues arising from the linkagesbetween financial sector capacity and the degrees of fixedom of monetary policy. 2. Conduct a detailed household survey on access to finance to identify constraints for householdsandmicro-enterprisesin low-incomesegmentsof thepopulation. 3. Notwithstandingthe objectives to establish a national ID system and expandingthe use of corporate tax IDS, introduce a financial ID system based on biometric. Various experiments based on biometric financial IDSin the region, in particular Uganda, have shown some promise in establishing the basis for credit information and know-your- customer procedures in an environment with high degrees of informality and incomplete public recordsinfr-astructure. 4. Strengthenthe accountingframework, enhance disclosurerequirements and build capacity of the accounting and auditing profession. Good accounting information and reliable auditsare the basis for sound bank risk assessments and the expansion of cash-flow based lending technology. In addition to the planned training and certification program for accountantsand auditors, a review of accounting and auditing standardscould ensure that internationalbest practiceare recognized. 5. Foster asset based lending practices. Review secured lending and leasing frameworks - including the supporting infrastructure like movable collateral registries - and implementation of international best practice in order to promote diversification of bank lending technologiesand emergence of non-bank financialproviders especially in support of SMEsand rural non-farm private sector development. 6. Promote the application of innovative products and technology to expand access to finance. Capacity building for banks and microfinance institutions in the use of different lending technologies secured lending, leasing, mortgage finance and in the longer run the promotion of new products such as warehouse receipts or weather insurance are likely to have high impacton financial depths. 6.4.2 Infrastructure Electricity The Angolan Government's objective of providing sustainable and reliable electricity supply is documented in the "Development Strategy of Angola's Private Sector" approved in September 2002. The strategy aims to increase accessto electricityfrom 20% in 2001 to 36%by 2011,and to increase electricity generation by 7.4% per annum between 2006 and 2011. In addition, the government is committed to the reduction of regional asymmetries in access to electricity. This commitment is congruent with our survey results where 79% of firms outside of Luanda deemed electricityasa constraint comparedto 59%ofthe firmsin Luanda (See Table2.2). In the long-termthe governmentintends on creating a national electricity transmissionsystem that connects all regions to one integrated grid. In addition, a national fund for the electricity sectorto widen access was foreseen in the 1996 electricity law. A range of measures including levies on electricity or petroleum products, state budget, and grants or loans h m international institutions were seen as financing sources for the fund. In October 2002, proposals for activating the fund were submitted to the government. The hnd exists only in concept. Levies on electricity or petroleum products need National Assembly's approval. There is recognition, however, that even when the National Fund is operating it would be wholly inadequate to finance needed investments in the sector,and, therefore,private financingwill be required. The lack of financial stability of ED EL^^ and E N E ~and ~ the considerable accumulated debts among parastatals and between these companies are problematic. Combining losses and non- collection suggests that only 45% of the electricity that EDEL receives from ENE is paid for with the other 55% stolen, lost as heat in the transmission or distribution lines, or sold but not paid for. In addition, the tariff structures are not reflective of costs with costs being historically set below long-run marginal costs.38Technical and non-technical losses totaling 36% in the EDEL system are very high.-Many of the non-technical losses are due to inefficient billing and settlement systems, illegal connections, and the lack of proper metering systems. These financial problems have resulted in EDEL and Em's reliance on direct governmentsubsidies,which are restricted to distribution activities.As a result, the expansion of electricitysupplyhas been limited. Short-term Recommendations: 1. Unbundle the financial structure of ENE into generation, transmission, and distribution activitiesand introducecost-reflectivetariffs, which would necessitatea gradual increasing of tariffs. 36 EDEListhedisbibutivecompanyresponsiblefortheelecbicitysupply in Luandaprovince. 37 ENE is mpomible forelectricitygenedon, mrnission, anddisbibution in themain citicsofthe 15 provinces outside ofLuanda.k is nsponsible forAngola's eldcitygenedon. 38 See World Bank Report "F'rivate Solutions for InkMmchm in Angola" page 43. Authors indicate that long-nm marginal cost is said to be US$O.lIkWh.Theyarehoweverunableto idRltifYthe basis forthiscalculation. 2. Improve the monitoring and regulation to ensure more efficient billing and settlement systems,reduction of illegal connections, andproper metering systems. 3. Amend the legal and institutional framework to allow private operators to play an importantrole in restoring and expandingsupply to consumers in the small or not-so-small isolated networks. 4. Review options for private sector participation in management contracts. With respect to ENE and EDEL activitiesgood rewards to strong private management can be obtained by outsourcing metering andresource collection services.This approachcould be appliedto a range of services including construction, maintenance, and manufacture or treatment of electricitypoles. 5. Reviewoptionsforprivate sectorparticipation in investment. This canbe donethrough the build own operate (BOO), build operate transfer (BOT) contracts for power plants, or a number of variantsofBOO and BOT. Long-termRecommendations: 1. ENE should prepare, publish and regularly update an electrificationplan. Private operator should be contracted to provide support servicesto private and municipal rural electricity operators. A simpler system for licensing of off-grid or small-scale electricity schemes should be introduced. 2. Achieve physical and managerial separation of generation,transmission, and distribution. This will help facilitate the future objective of creating separate and autonomous enterprises for each business fimction. 3. Ensure full operabilityand independence of regulator. The regulator should be responsible in the setting of electricity prices. This is important in reassuring the private sector that prices will be allowed to coverreasonablecostsand earna reasonablerate of return. 4. Move away from uniform tariffs, which imply that a consumer connectedto the ENE grid in isolated areas, pays prices that are a fraction of the production cost in those areas, while those without an ENE connection must pay very high prices for electricityproduced from small diesel generating sets.A policy of nonuniform tariffs would allowprivate developers and ENE to operate supply systems in rural areaswith much reduced orno dependence on the state for operating subsidies. 5. Implement private sector participation in management contracts and investments(through BOO, BOT or variants of the &o). 6.4.3 Regulatory Environment The government of Angola has recently been engaging in a number of legislative and administrative initiativesin an attempt to address obstacles to the establishmentand operation of new businesses in Angola. In August 2003, the Guiche Unico de Empresas (GUE), a public service agency comprising delegations of all services or government bodies involved in authorizing business start-ups was set up. In April 2004, a new Company Law took effect that consolidated rules related to incorporation of commercial companies in Angola, which was previously spread out amongst several laws. Although too early to see the full impact of such measures, Angola has already slightly improved in its ranking in the Doing Business Indicators. There is, however, a need to fUrtherincreasethe efficacyof the GUE, and reduce the related costs. Short-termRecommendations: 1. Increaseeffectiveness of the GUE by reducing the cost to start-up a business. 2. Reduce costs to execute the notary deed of incorporation, provisional registration with registryof companies, and obtainmentof the CommercialOperations Permit. 3. Reduce time required to obtain the Commercial Operations Permit and the two-step registration with the Registry of Companiesfrom thecurrent40 and 30days respectively. Closing a business in Angola is a costly and timely endeavor. The time required to go through insolvency in Angola is approximately6 years, compared to the Sub-SaharanAfiica averageof 3 years. In addition, the recovery rate is 11 cents per dollar, compared to the Sub-Sahara Afiican average of 17 cents per dollar. These obstacles to closing unviable businesses result in the continuation of these loss-making businesses and the subsequent inability to reallocate assets and human capital to more productive uses. In spite of the fact that Angola has bankruptcy laws, they are almost never used. In developing countries, however, it is often the case that debts are settled outside of the legal insolvency procedures. This is one reason why the recently enacted VoluntaryArbitration Law (VAL), which provides a legal Mework for the resolutionof non-judicial disputes, is important. It is, however, necessary to build up the legislativeand administrativeregulatory capacities for foreclosures over the long-term. Short-term Recommendations: 4. Build up the administrative capacity for the recently enacted Voluntary Arbitration Law through the recruitment and trainingof arbitratorsand staff. 5. Encouragethe involvementof creditorsin the foreclosureprocess. Long-termRecommendations: 1. Enact the necessary laws and build up administrative capacities for foreclosures through therecruitmentand training ofjudges and staff. ContractenforcementinAngola necessitates46 procedurescomparedwith the Sub-Sahara African regional average of 40. In addition, the time to enforce a contract is 1,011days compared with the regional average of @I3days. To reach a judgment it takes 1 year and 4 months (485 days) on average,and to enforcethatjudgment it takesalmost the sameamount of time (440 days).Thehigh number of procedures and time required for a contract's enforcementunderscore incapacities and inefficienciesinherentin thelegal andjudicial system. The VoluntaryArbitration Law is useful in providing an alternate mechanism to the court system as a form of dispute settlement.In addition, the establishmentsof informationsystemsfor caseload and judicial statistics will help reduce judgment periods. The US Department of Commerce's Commercial Law Development Program, for example, improved criminal case management in Angola's Luanda Provinceby instituting a computerized casetracking system. In addition,there is a need for the further simplificationof the procedures in commercial disputes. A modificationof the structure of the judiciary to allow for small claims and specialized commercial courts is necessary. Finally, the simplificationand clarification of laws and regulations would result in the freeingup of court-resources for more relevant"dispute settlement"activities. Short-termRecommendations: 6. Establish information systems for caseloadandjudicial statistics. 7. Further simplifythe procedures in commercialdisputes. 8. Simplifyand clarifylawsand regulations Licensinga warehouse in Angola requires 14procedures,takes 337 days, and costs 1110%of per- capita income. The Sub-SaharaAfrican average requires 18procedures, takes 262 days, and costs 2550% of per-capita income. The request for the license from the Provincial Governortakes 180 daysto complete and the registration of the building with the real estate registry takes 90 days. As a result, efforts to hasten the response time of Angola's government bureaucracy and computerization of registry process ought to enable a faster tumaround time. In terms of costs, obtaining power connections and the hiring of the inspection firm comprise 63% and 36% respectively of total "warehouse" licensing costs. Consequently, improvements in electricity infrastructuresought to help diminish the costsof licensinga warehouse. Short-termRecommendations: 9. Shortenthe timeto obtaina license from the Provincial Governor 10. Shorten the time to obtain a license from the real estateregistry 11. Reduce the costs of inspection by increasingcompetitionand reduce the fees forelectrical connections. 12. Computerizeregistryprocess 6.4.4 Corruption Angola remains one of the countriesin the world were corruption is most pervasive.Angola ranks 147 out of 169 countries in the Transparency International's Corruption Perception Index. Objective data from the ICA shows that firms in Angola have to pay close to 4% of the contract value in bribes in order to obtain government contracts. Corruption is most problematic for large firms and h s in Luanda. Tackling corruption is not an easy or a short process. It requires political will, popular support, andnecessary resources. Short-termRecommendations: 1. Clearly and unequivocallydeclarepolitical will to fightcorruptionat the very top level. 2. Allocate necessary resources to the fight: assign 0.5% of national budget permanently to this fight Long-term Recommendations: 1. Establish an Anti Corruption Agency. Recruit investigators and staff and define a clear mandate. 2. Develop an anticorruptioncampaign to build popular support. Foster isset based lendmg prrtdlcg and review l&g andk m g 6ameworks vatepamcrpatlon In environment for ReduceIX& toexecutethenotarydeed of business -4 mmpwon, ~ O V I SlIe~m o n Hlth regdryofcornparuesandobtammentofthe Commercial Operat~cmsPenn~t Reducetunerequired to obtmthe Conunernal wensP m t andthetwostepregisbahon . with theRegistryofCompanies. High costtoclosea business RecentlyenactedVolunkaryArbitrationLaw (VAL) Build up theadministrativecapacity forthe I Enaathenecessarylawsandbuildup Burdensometimeandnumber VOIW& ArbitrationLaw through th adminktdve capacitiesfor foreclosuresthrough of prwdures. recruitmentandtrainingof arbibatmandstaff. themmibnent andbain'ig ofjudges ands M . Encourage the involvement of creditorsin the High costsof licensing Shortenthe time to attain a l i h m the Lengthytimequired RovincialGovmor Shortenthethetoobtainalicensehmthe real estate regisby Compderizeregistrypmcess Reducecosts of inspection by i n m i n g competitionand mimefeesforelectrical connections Lengthy timequired to reach Improved criminalcase- p e n t inAngola's Establishinformationsysm for caseloadand judgmentandenforcea LuandaRovinceinstiaaed judicial statistics. contract. Voluntaq ArbitrationLaw usell in providingan Furthersimplifythe proceduresin commercial alternate mechanismto the courtsystem asa fomlof disputes. disputesettlement SiilifLandclarifylawsand qditions Highcostsondoingbusiness. Someimprovementsin regulatoryqualitynoted, Clearlyandunequivocally declarepoliticalwill Establishan Anti Comption Agency. however,conuptionis stillasignificantpmblem in to fightcomqtionat the verytop level Recruitinvestigatorsandstaffanddefine a Angola. Allocatenecessary resourcestothe fightassign clearmandate. 0.5%ofnationalbudgetpermanentlyto this Developan anticormptioncampaignto fight buildpopular suppon ANNEX A: TECHNICAL APPENDIX FOR PRODUCTIVITY CALCULATIONS We follow the methodsproposed by Escribanoand Guasch (2005)~~'Productivityis estimated by fittingthe followingCobb-Douglasproduction function: where VAiis value added for fumi; Liis the total cost of labor;Ki is the capital stock; 'i is a vector of firm-level control variablesand includes a dummy variable forthe age of the firm (1 if less than 5 years old; 0 otherwise),a dummy variable that assumes the value of one if the manager has some higher education, and a dummy that assumes the unit value if the manager has more than 5 years of experience; F and G are industry dummies for food and garments sectors. Total Factor Productivity (TFP) is constructed as the estimate of ',, the part of value added not explained by labor or capital, after controlling for industry's fixed effects and addressing endogeneiy concerns by the inclusion of averages by location and sector instead of the firm responses4 . The impact of the investment climate variables identifiedin our earlieranalysisonTFPcanbe estimated through the followingequationusing weighted ordinaryleast squares4': Where TFf: is the total factorproductivity(in log) for firm i; ICi is a vector of the investment climatevariables; FCi is a vector of h-level characteristicssuch as size, export orientation, location and ownership status, and the ownahip of a generator.As an alternativeto this two step procedure, we also estimated extended Cobb-Douglas production functionsproposed by Escribano and Guasch, using investment climate constraint perceptions to control for endogeneity of inputs: 39 E s c n h o and Guasch (2005),Assessing the Jnpct of the lnvesbnentClimate on ProductivityUsing Firm-Level Data:Methodologyand the Casesof Guatemala,Honduras, andNicsuagua,World BankPolicy ResearchWorking Paper3621. 40 Due to the homogeneityin the sample it is onlypsible to includeone of these avenge variables at a tim.The choice of miable to include was detemdnedby highest improvementin the gocdnessof fit 41 Unweightedregressionswith robust stmdarderrorswere alsocomputed with resultsthat arequalitativelyidentical to theonesreportedhere. Where IC,. includes investment climate perceptions reported by the firms, whch are considerednot to determine the use of inputs. The variablesused have already been described in the main text. Below we report the full results of the effects of investment climate constraintson TFPand laborproductivity.As alreadymentioned the locationof the firmin the capital has a significant positive effect on productivity. There is some weaker evidence that olderfirms aremoreproductive. TableA1 EffectsofAccess to Finance -1 TFP Model 1 TFP Model2 Labor Productivi Coef. SE P-value Coef. SE P-value Finance* PV I -0.011 0.016 0.5141 -0.011 0.016 0.498 Food I 0.002 0.086 0.9801 -0.024 0.133 0.859 Garments Age of Firm ForeignOwnership MediumSize LargeSize Luanda Constant 1 -0.182 0.106 0.0871 -0.167 0.119 0.161 R squared I 0.0471 0.047 Source: ICA Survey TableA2 Effects of Electricity TFP Model 1 TFP Model2 Labor Productivity Coef. SE P-value Coef. SE P-value Coef. SE P-value Electricity Garments Age of Firm ForeignOwnership Medium Size LargeSize Luanda Constant 1 -0.069 0.127 0.5901 0.430 0.186 0.0211 8.493 0.143 0.000~ R squared 0.0761 0.1031 0.092 Source: ICA Survey TableA3 Effectsof BusinessLicenses TFP Model 1 TFP Model2 Labor Productivity Coef. SE P-value Coef. SE P-value Coef. SE P-value Business Licenses ConstrctionPermit$ 0.240 0.211 0.257 -8.119 4.089 0.048 0.126 0.238 0.597 OperationLicense -0.561 0.266 0.035 -0.517 0.262 0.050 -0.702 0.299 0.019 PV Food Garments Age of Firm ForeignOwnership Medium Size Large Size Luanda Constant ( -0.278 0.096 0.004( -0.393 0.112 0.001( 8.405 0.108 0.000 Rsquared 0.071 1 0.0781 0.087 Source: ICA Survey TableA4 Effectsof Crime TFP Model 1 TFP Model 2 Labor Productivity Coef. SE P-value Coef. SE P-value Coef. SE P-value Crime, Theft and Disorder Security3 0.118 0.083 0.157 2.159 0.653 0.001 0.093 0.093 0.323 Crime 0.334 0.120 0.006 0.338 0.117 0.004 0.215 0.135 0.114 PV -0.038 0.019 0.048 -0.035 0.019 0.064 -0.064 0.021 0.003 Food -0.082 0.088 0.353 -0.766 0.232 0.001 -0.289 0.100 0.004 Garments 0.006 0.125 0.963 0.000 0.123 0.999 -0.245 0.141 0.083 Age of Firm -0.200 0.093 0.032 -0.221 0.092 0.016 0.034 0.105 0.747 ForeignOwnership 0.173 0.145 0.235 0.141 0.143 0.326 0.266 0.164 0.106 Medium Size 0.057 0.131 0.665 0.058 0.127 0.650 0.052 0.148 0.724 Large Size -0.393 0.241 0.104 -0.370 0.237 0.120 -0.442 0.272 0.105 Luanda 0.372 0.113 0.001 -0.243 0.223 0.277 0.311 0.128 0.015 Constant 1 -0.203 0.091 0.0261 -0.293 0.093 0.0021 8.489 0.102 0.000 R squared 0.0861 0.1lOl 0.095 Source:ICA Survey TableA5 Effects of Corruption TFP Model 1 - TFP Model2 I Labor Productivity I Coef. SE P-value 1 Coef. SE P-value I Coef. SE P-value Corruption InformalPayments 0.008 0.012 0.521 0.009 0.008 0.238 0.005 0.013 0.688 Pavmentsto Officials3 0.002 0.009 0.851 -0.271 0.154 0.079 0.004 0.010 0.731 PV ( -0.053 0.019 0.005) -0.051 0.019 0.007) -0.077 0.021 0.000 Food 1 -0.028 0.087 0.7501 0.040 0.094 0.6731 -0.265 0.096 0.006 Garments 0.041 0.127 0.744 0.519 0.299 0.083 -0.194 0.142 0.172 Age of Firm -0.177 0.095 0.062 -0.169 0.093 0.071 0.067 0.106 0.526 Foreign Ownership 0.159 0.146 0.276 0.168 0.145 0.248 0.228 0.162 0.161 Medium Size 0.091 0.130 0.483 0.097 0.129 0.455 0.085 0.145 0.556 Large Size -0.450 0.258 0.082 -0.441 0.242 0.070 -0.522 0.288 0.071 Luanda 0.433 0.113 0.000 1.835 0.804 0.023 0.355 0.126 0.005 Constant -0.198 0.092 0.031 -0.274 0.101 0.007 8.499 0.102 0.000 Rsquared 0.067 0.076 0.099 Source:ICA Survey TableA6 Effects ofTransportation TFP Model 1 TFP Model 2 Labor Productivity Coef. SE P-value Coef. SE P-value Coef. SE P-value Transportation Breakage ITheft PVS Food Garments Age of Firm ForeignOwnership Medium Size Large Size Luanda Constant 1 -0.105 0.097 0.2771 0.711 0.216 0.0011 8.564 0.108 0.000 R squared 0.077) 0.1081 0.102 Source:ICA Survey