33698 Investment Climate Assessment Competing in the Global Economy: An Investment Climate Assessment for Uganda World Bank: Linda Cotton, James Habyarimana, Chad Leechor, Jean Michel Marchat, John Paton, Vijaya Ramachandran, Manju Kedia Shah, Michael Wong Uganda Manufacturers Association Consultancy & Information Services (UMACIS): William Kalema, Charles Ntale August 2004 Contents 2 Acknowledgments 4 Infrastructure 47 Executive Summary 5 Electricity 48 Telecommunications 48 1 The Economic Environment 15 Water 49 Investor Perceptions 16 Waste Disposal 49 Policy Initiatives 18 Taxes, Regulation, and Administrative Past Accomplishments 18 Corruption 51 Ongoing Efforts 19 Access to Finance 52 Firm Performance 52 2 Growth and Productivity of Firms 21 Investment 52 Characteristics of Enterprises and Entrepreneurs 22 Exports 53 Human Capital Endowments of Entrepreneurs 23 Employment Growth 53 Characteristics Determining Firm Growth 24 Unit Labor Costs 53 Characteristics and Productivity of Capital 26 Age of Capital Stock 26 4 Factor Markets: Finance and Labor 55 Capacity Utilization 27 The Financial Market 56 Capital Intensity 27 Sources of Finance 56 Productivity of Capital 28 Developments in the Banking Sector 57 Productivity and Cost of Labor 29 Access to Bank Finance 57 Labor Productivity 29 Determinants of Access to Bank Finance 58 Unit Labor Costs 31 Bank Finance Instruments and Collateral Total Factor Productivity and Requirements 59 Technical Efficiency 32 Cost of Finance 61 Determinants of Total Factor Productivity 32 Are Firms Credit Constrained? 61 Performance of Firms Relative to the Trade Credit 63 Efficiency Frontier 33 A Comparison with Kenya 64 3 The Investment Climate 36 Policy Issues 66 Perceived Constraints in the Business The Labor Market 67 Environment 37 Health Status 68 Uncertainty 38 Illness and Workdays Lost 69 Corruption 39 Treatment Sources and Payment Methods 69 Infrastructure Performance 41 Awareness of HIV/AIDS 69 Regulatory Burden 42 Remuneration and Determinants of Wages 71 Effective Protection of Manufacturing 45 Wage Levels across Sectors 71 How Has the Business Environment Changed? 46 Wage Levels in Manufacturing 71 Constraints in the Business Environment 46 Determinants of Manufacturing Wages 73 Contents 3 Institutional Rigidities 75 Labor Contracts 75 Labor Unions 76 Labor Regulations 76 Policy Issues 78 5 Policy Implications 80 Maintaining Macroeconomic Stability 81 Encouraging Private Provision of Social and Infrastructure Services 82 Establishing a Low-Cost Operating Environment 82 Establishing a Competitive Investment Environment 83 Improving Tax Administration 83 Ensuring Sound Financial Market Development 84 Raising Firm Productivity 84 Increasing the Efficiency of Services 85 Addressing Distortions in Trade 86 The Role of the World Bank 86 Annex 5.1 Policy Suggestions for Improving the Investment Climate in Uganda 88 APPENDIXES 92 Appendix 1. The Sample 94 Appendix 2. Labor Market Features 102 Appendix 3. Protection of Manufacturing 119 Appendix 4. Investment Climate Indicators 130 REFERENCES 139 Acknowledgment 4 The authors are grateful to Andrew Stone, Sudarshan Canagarajah, Christiane Kraus, William Steel, Agata Pawlowska, Robert Blake, Ibrahim Elbadawi, Judy O'Connor, Melanie Marlett, Andre Ryba, Jakob Svens- son, and seminar participants in Kampala for comments on various drafts. Executive Summary 5 Uganda's economic growth since the late 1980s The assessment is based on a survey of regis- has resulted largely from restoring and rehabilitating tered enterprises in the Ugandan private sector con- the country's productive capacity. Going forward, ducted in November 2002­April 2003 by the World growth will need to come increasingly from new in- Bank's Regional Program on Enterprise Development vestments or new activities. That will require more in- (RPED) in collaboration with the Uganda Manu- vestment, more intensive acquisition of know-how, facturers Association Consultancy and Information and more complex collaboration between local and Services (UMACIS). The survey covered 392 firms foreign partners. It will also require a far greater role across four sectors (commercial agriculture, construc- for private sector investment. While Uganda has ben- tion, manufacturing, and tourism) and three regions efited enormously from development assistance for al- (central, northeast, and southwest). The questionnaire most two decades, foreign aid may decline in the next contained a range of questions on such issues as the decade. production process, cost of inputs, access to finance, Indeed, Uganda must accelerate private sector types and cost of labor used, and costs incurred in investment and growth if it is to achieve the goals set preventing or treating HIV/AIDS. A separate survey out in its Poverty Eradication Action Plan. To reduce administered to up to 10 workers in each firm allowed the poverty rate to 10 percent by 2017, GDP growth a detailed analysis of the labor market. will have to exceed 7 percent a year, requiring an in- The results of the survey are validated through vestment rate of 30 percent or more of GDP. Attaining comparison with other studies and surveys. The sur- these targets will be feasible only with reforms to pro- vey results are consistent, for example, with those re- mote private investment. This need is recognized in ported in the World Bank's Doing Business database, the pillars guiding the Poverty Eradication Action Plan, which relies on expert assessments to provide inter- the first three of which have a bearing on the private national benchmarks for business regulation (World sector: Bank 2003a). In addition, the findings are compared with those of a similar survey from 1998 to evaluate · Creating a framework for sustainable economic progress in the business environment and in the pri- growth and structural transformation. vate sector's performance. · Ensuring good governance and security. What are the findings of the investment climate as- · Increasing the ability of the poor to raise their sessment? The business environment creates obsta- incomes. cles for the private sector that are clearly slowing its · Improving the quality of life of the poor. growth, particularly by limiting productivity growth. But there is reason for optimism: the comparison of data What are the prospects for the private sector in from the 1998 and 2002/03 surveys shows that Uganda? To find out, this report assesses the per- Uganda's business environment has improved-- formance and productivity of private firms in Uganda especially in aspects relating to regulation and infra- as well as the constraints imposed by the investment structure--and the positive changes are already af- climate on their operation, investment, and growth. fecting firms' performance. Investment rates have The assessment takes into account the perceptions of risen, exports are growing, and firms, especially the private firms and puts its findings into an international largest ones, are operating at higher efficiency. context. It also analyzes the policy implications of its Nonetheless, Ugandan firms have a long way to go to findings and offers recommendations. compete with firms in other parts of the world. And Executive Summary 6 while the business climate in Uganda is improving, it enough to compensate. A common measure for inter- remains much harsher than those in rapidly advanc- national wage comparisons is the monthly pay of un- ing countries like China. For Uganda to become com- skilled production workers, the most homogeneous petitive as a host country for exporters, it must both wage category. In Uganda these workers earn about catch up with other countries' reforms and keep pace $57 a month on average. This compares favorably with their progress. with earnings in such countries as Nigeria and Kenya, where the range is $73­100 a month. Earnings in Performance and Productivity of Firms Uganda also compare favorably with those in India Ugandan firms appear to have high-quality capital and China. and significant excess capacity, suggesting a poten- But while Uganda does well on this simple mea- tial to improve their performance if problems in the sure of labor cost, measuring the output of workers business environment can be addressed. relative to their wages shows that the economy has a To begin with, Uganda's capital stock is excep- long way to go before becoming globally competitive. tionally young. More than 40 percent of its manufac- The median ratio of wages to value added (a proxy for turing firms have capital stock averaging less than 5 unit labor cost) at the firm level in Uganda (0.39) is years old, and another 35 percent have capital stock much higher than the ratio in East Asian countries averaging 5­10 years old. By contrast, in most other when they were at roughly equivalent stages of devel- Sub-Saharan African countries a large share of the opment (0.16­0.35 in the 1960s and 1970s). capital stock is more than 20 years old. The compar- What drives the productivity and growth of firms in atively young age of Uganda's capital stock suggests Uganda? To investigate this question, the assessment that the technology is more recent, of better quality, looks closely at differences in the characteristics of and more productive. entrepreneurs and the firms they control. As is typical Indeed, Uganda's capital stock is remarkably pro- in Africa, most firms in Uganda (almost 70 percent) ductive compared with that of other countries in East are entrepreneur owned--that is, owned and run by Africa. In a year's time every dollar of capital in an individual or family. Indigenous Africans own most Uganda generates twice as much in value added as of these firms (75 percent), and entrepreneurs of a dollar of capital does in neighboring countries. Asian ethnicity most of the rest. But Asian entrepre- Still, Uganda has the lowest capital intensity neurs own most of the larger firms and therefore con- among the sample of countries recently surveyed in trol a large share of the business assets in Uganda. the neighboring region. While Uganda has about Results from regression analysis and estimation of $1,500 of capital per worker, this ratio is several times a production function underscore the importance of greater than in countries such as Kenya, Tanzania, access to education. The learning of entrepreneurs-- and Zambia. whether through advanced education or work experi- In sharp contrast to the relatively high productivity ence in a foreign firm--is among the most important of capital in Uganda is the relatively low productivity of factors in determining a firm's productivity and growth. labor. In Ugandan manufacturing the median annual Compared with indigenous African entrepreneurs, value added per worker ($1,085 in 2002) lies in the ethnic Asian entrepreneurs have more education and general range for East Africa but well below that in much more experience on average before starting India ($3,432) and China ($4,397). their firms. And they are more likely to obtain exter- Even with lower labor productivity, Ugandan firms nal loans for start-up. For these reasons Asian entre- would still be able to compete if wages were low preneurs' firms grow more rapidly. But access to Executive Summary 7 university education for indigenous entrepreneurs to get things done, pay more in bribes than other helps boost the growth of their firms. This result is ro- types of firms--almost 4 percent of their revenue on bust across the countries surveyed in East Africa. average. Large, exporting, and foreign firms find adminis- The Business Environment trative and regulatory problems more of a nuisance Why do Ugandan firms still suffer low productivity and than smaller and domestic firms. Their senior man- slow growth? One set of answers comes from the sur- agement spends more time on average dealing with vey responses of owners and managers of manufac- regulations. Moreover, these firms spend more than turing firms asked to evaluate constraints to invest- twice as much time in inspections with government of- ment and growth in Uganda's business environment. ficials and lose almost 10 times as much money (as a The responses indicate that financing obstacles are share of sales) in fines or seized goods as a result of the greatest constraints--both the cost of finance (in- these inspections. Surprisingly, it also takes large and terest rates) and access to finance (collateral require- foreign companies longer to clear exports through ments). Burdensome tax rates also rank high, a com- customs. mon finding by firm surveys. Also considered a big constraint is unreliable electricity supply. Finance and Investment Around 60 percent of firms reported that the cost Reform of the commercial banking sector has been an of finance is a major or severe constraint, and 45 per- important pillar of the economic reform program that cent reported the same for access to finance.1 When Uganda started in the early 1990s. But while the re- the results are disaggregated by firm size class, it be- form has strengthened the banking system, govern- comes clear that the cost of finance is a constraint felt ment borrowing from banks is crowding out the pri- across the board. Not surprisingly, however, access to vate sector. finance becomes a less important constraint as firm Ugandan manufacturing firms rely substantially on size increases. internal resources. On average, they use internal Unreliable electricity supply has real conse- funds to finance about 80 percent of their working quences. Manufacturing firms in Uganda estimated the capital needs and 71 percent of new investments. share of production lost due to power outages and fluc- Bank credit is the next most important source of tuations at an average 6.3 percent. These losses are finance. Still, only 32 percent of firms in the Ugan- substantially larger than those in China (1.8 percent), dan sample have bank credit (loans or overdrafts), though smaller than those in Kenya (9.3 percent). compared with 80 percent of firms in Kenya. Perceptions of the most important constraints vary Commercial bank finance accounts for 7 percent of among different types of firms. Exporters find some working capital needs and 13.5 percent of new in- aspects of the economy significantly more problem- vestment needs. atic than nonexporters do, such as macroeconomic What matters most in determining whether firms instability (inflation, exchange rates), corruption, mo- have access to finance? The possession of adequate nopolistic behavior of competitors, crime, and trans- collateral, the credibility of information produced by port. Similarly, foreign-owned firms reported greater firms (for example, whether firms have their accounts sensitivity than domestically owned firms to corrup- audited by external agencies), and the size of firms. tion, regulatory policy uncertainty, customs and trade The age of firms also matters. Age is a particularly im- regulations, and crime. Foreign-owned firms, which portant factor in Uganda's manufacturing sector, tend to face the most requests for informal payments which is dominated by young firms. Executive Summary 8 When firms do obtain external finance for invest- rigidities--relating to labor contracts, unions, and reg- ment, most must rely on short-term loans and over- ulations--seem to be relatively limited. drafts. Nearly 40 percent of the loans held by firms in The health of the Ugandan labor force is a major the sample have a maturity of a year or less. Banks' concern, however. Health conditions in the country apparent preference for short maturities may reflect a are poor, and up to 37 days of production per worker very risky environment or a culture of nonrepayment. are lost annually because of health-related issues, Overdraft facilities involve security requirements with obvious implications for firm productivity. similar to those for bank loans. Some 60 percent of Improving the health status of the labor force as firms with overdraft facilities reported being required quickly as possible is therefore critical. to post collateral. That overdraft facilities are backed Despite the obvious health concerns, Ugandan by fixed assets suggests unusually risk-averse be- firms appear to underestimate, ignore, or conceal the havior by banks. One possible explanation is that extent of the HIV/AIDS problem. The private sector overdrafts are typically used to finance medium-term needs to expand measures aimed at HIV/AIDS pre- expenditure. Anticipating this, banks require that firms vention, diagnosis, and awareness raising, targeting post substantial security before extending overdraft these measures to wage earners in lower income facilities. classes. Only about 37 percent of firms reported tak- After bank finance, trade credit is the second most ing actions aimed at preventing and raising aware- important source of external finance for working capi- ness of HIV, mostly advertising or counseling. tal in Ugandan manufacturing, covering 5.3 percent of Surprisingly, almost 60 percent of managers reported needs. Nearly 60 percent of firms in the sample re- that HIV/AIDS has had little or no impact on their work- ported purchasing inputs on credit, compared with force. This response may reflect a simple lack of in- nearly 85 percent of firms in Kenya. Whether firms can formation, an inability to distinguish HIV/AIDS from obtain trade credit depends in part on whether firms other illnesses, or a drastic underestimation of the providing the credit have access to external finance. problem. Where the older, larger firms that tend to provide trade Employees have a much more acute perception of credit do have access to external finance, this can the HIV/AIDS problem. A large share across all sec- mitigate lack of access to bank finance for smaller, tors surveyed are willing to be tested for HIV and to younger firms. pay for the tests as long as the testing is anonymous. These attitudes are a very favorable factor in fighting The Labor Market HIV/AIDS in the work environment, with clear external- The high labor costs in Uganda relative to worker pro- ities for labor productivity and for Ugandan society. ductivity point to a need to better understand the link Wages appear to vary widely across subsectors between labor market dynamics and labor productiv- and regions in Uganda, suggesting that the labor ity, two factors that affect unit labor costs. The overall market may not be competitive or fully integrated. picture of the labor market in Uganda is relatively pos- Indeed, the results of wage equations estimated with itive. The labor force is relatively well trained, com- worker data indicate that a purely competitive model paring favorably with the labor force in other African does not explain wage formation in Uganda. countries for which recent survey data are available. It Increasing labor mobility through better infrastructure is second only to Nigeria's labor force in higher edu- and improving access to education and vocational cation and has the largest share of workers with vo- training in all regions would help reduce the wage dif- cational or technical training. Moreover, institutional ferentials observed in the private sector. Executive Summary 9 In manufacturing workers' earnings appear to be district administrations pose the key challenges to only weakly linked to their performance, as shown by macroeconomic stability. Several strategies could be the small share of bonuses in their pay and the in- pursued to reduce the fiscal deficit: increasing trans- significance of hours worked in explaining the level parency and participation in the setting of sectoral of earnings. Since a higher level of effort does not budget ceilings, improving budget monitoring sys- translate into a noticeable increase in earnings, em- tems, minimizing unbudgeted supplementary expen- ployees have little incentive to work harder. To in- ditures that lead to spending cuts for other depart- crease incentive, earnings should be clearly linked to ments, increasing budget efficiency so as to allow the performance. government to achieve development goals even while reducing or maintaining spending levels, and encour- Policy Recommendations aging stakeholders in the development process to as- The findings of the investment climate assessment sist local governments in decentralizing social ser- confirm those from recent consultations with the pri- vices, budget processes, and financial management. vate sector, underscoring the need for Uganda to con- The public sector has largely withdrawn from solidate and build on core reforms of the past decade ownership and management of purely private opera- if is to transform itself into a competitive economy. tions, but the private sector has not yet exploited in- Uganda needs to maintain a stable macroeconomic vestment opportunities in the provision of social and in- framework, establish a low-cost business environ- frastructure services. By pursuing a policy of divesting ment, strengthen the financial sector, privatize and re- and contracting out services, the government could re- form key utilities, and raise firm productivity by boost- duce the public administration budget while also lever- ing capacity utilization and the efficiency of the labor aging private investment. The Ministry of Public and financial markets. Service, in cooperation with the privatization unit, Now that Uganda has achieved sustained periods needs to continue identifying and divesting services of macroeconomic stability, the challenge is to dem- that can be contracted out to the private sector. And onstrate to investors that its macroeconomic frame- the private sector needs to take a more active role in work is sustainable over the long haul, capable of seeking out investment opportunities and encouraging withstanding not only regional insecurity but also po- policy reforms that would allow greater investment. litical transition. The government must keep its admin- The survey data detail the substantial regulatory istrative budget low, reinforcing the image of a lean, burden with which the private sector must cope. This professional civil service capable of performing with- regulatory burden, and the associated high costs for out increasing donor funding. The Ministry of Finance, the private sector, result from inadequate regulatory Planning, and Economic Development, in collabora- capacity, an unclear regulatory framework, and incon- tion with the Ministry of Public Service, needs to work sistent interpretation of policies and regulations. To on modernizing the civil service, which should be ca- ease the burden, the government needs to accelerate pable of delivering the services the private sector regulatory and institutional reforms aimed at improv- needs to compete successfully. ing and modernizing the business operating environ- In addition, the Ministry of Finance needs to de- ment. It also needs to ensure that these reforms-- velop spending ceilings for key categories in the which include drafting new business laws and budget. Persistent pressure to expand military spend- reforming such key institutions as the Uganda ing and the creation of more public institutions and Revenue Authority and the Business Registrar--are Executive Summary 10 implemented consistently. Beyond strengthening the to proceed with any capacity building program. The regulatory framework, it will be critical to improve the Uganda Development Bank, now a major source of attitude and work habits of the civil service, particu- distortion in the formal incentive structure, needs to be larly at the local government level, where the capacity transformed into an institution that allows broad de- and commitment to address regulatory and institu- velopment of the financial market and provides the fi- tional reform issues are in short supply. nancial coverage financial institutions need to reach To tackle corruption, the government should build high-risk clients. on existing anticorruption policies, such as the re- To improve tax administration, the Ministry of quirement that public officials declare their wealth. Finance, Planning, and Economic Development Other recommended measures include adopting ad- needs to ensure that tax laws are in effect that are ditional anticorruption legislation, reforming public clear, unambiguous, and consistent with the invest- sector pay, providing adequate resources to anticor- ment code. It also needs to ensure that tax policy is ruption agencies, ensuring proper follow-up of find- predictable and consistent. The government needs to ings issued by commissions of inquiry, combating a devise strategies for widening the tax base by gener- culture of impunity, strengthening accountability at all ating more revenue from small businesses and the in- levels of government, and addressing corruption at formal sector. For the Uganda Revenue Authority an lower levels of government. In addition, further efforts urgent priority is to transform itself into an efficient in- need to be made in increasing transparency in fiscal stitution with a reputation for integrity, an institution policy and public procurement, such as in budget ex- that enforces tax laws while remaining cognizant of its ecution and reporting, and in limiting the use of sup- role in creating an enabling environment for the pri- plementary appropriations by the executive branch. vate sector. The agency also needs to focus on run- Arrears need to be properly assessed, and contingent ning the value added tax refund system efficiently for liabilities monitored and controlled. And public pro- the private sector while reducing fraud. curement systems need to be improved. Since cor- To ensure the development of a sound financial ruption is partly rooted in the political system, the pri- market, the government needs to develop additional vate sector has a key role to play as a monitor and reform policies to support financial service providers advocate, demanding minimum standards and assist- and improve their ability to respond to the needs of the ing entrepreneurs in challenging abuses of institu- private sector. While competition in the financial mar- tions, systems, and regulations. ket has improved, this has not yet led to a substantial To establish a competitive investment environ- decline in intermediation costs and thus interest rates ment, the government needs to develop a transparent or to an increase in lending to the private sector. High incentive structure and updated legal framework for interest rates on government treasury bills are having investment. Key priorities are to accelerate the com- a crowding out effect on lending to the private sector. mercial legal reforms started in 1999 and update the Moreover, the high-cost business environment results investment code. In addition, such institutions as the in high administrative costs for loans. And firms have Uganda Investment Authority, Export Promotion limited access to long-term loans. Board, and Tourism Board need to have their roles To help improve access to long-term financing, clarified and be set on a long-term, sustainable path. the government needs to focus on such key areas as The government has proposed merging these institu- pensions, insurance, and capital market develop- tions, but this has not yet been done, making it difficult ment. It also needs to increase funding for commercial Executive Summary 11 courts, to enable them to function more efficiently. At several years--as a matter of urgency. In addition, the the same time the government needs to provide in- government needs to ensure that the three new com- centives for better compliance with accounting stan- panies that have assumed the functions of electricity dards by the private sector. Establishing a credit reg- generation, transmission, and distribution are fully op- istry would give firms greater incentive to provide erational and attracting adequate private participa- high-quality information on their operations and fi- tion. In the water and sanitation sector the responsible nances. Both the government and the private sector ministry needs to fast-track the reforms aimed at in- could play an active role in addressing privacy con- creasing investment and expanding service, also cerns related to such a registry. under way for years. In the transport sector the gov- To help boost firms' productivity and capacity uti- ernment could accelerate the railway privatization and lization, the government needs to support private sec- improve the road network, at least the major economic tor­led skills development and technology transfer ini- routes. The Civil Aviation Authority needs to focus on tiatives. There are many ways to do this, including regulation and on encouraging private investment to providing tax credits to firms that engage in worker improve the efficiency and effectiveness of air trans- training or adopt new technology. The government port services. As suggested in the government's strat- could also support worker training and apprenticeship egy for structural transformation, several measures programs that are designed and implemented by need to be taken immediately, including restructuring the private sector to meet its needs. Support should unsustainable debts in the utilities sector, providing be targeted to micro, small, and medium-size enter- entry points for private participation, and creating a prises, since these make up more than 90 percent of multi-utility regulatory agency. Ugandan firms. The government needs to continue its work in The government should also look for ways to en- trade reform, lowering overall tariffs, promoting effi- courage entrepreneurship and improve access to busi- cient resource allocation, and keeping excise taxes ness education. Access to university education may be and other forms of nontariff protection to a minimum. particularly important in offsetting the advantages of in- While Uganda's effective rates of protection are rela- herited ownership and family-based business knowl- tively low overall, nontariff trade barriers remain high. edge. It may be worthwhile for both the government In particular, ad hoc excise taxes are creating distor- and donors to revisit their priorities in this area. tions in the economy, as indicated by the wide range To help improve the health of workers, the govern- of effective rates of protection. The present structure ment should consider using existing HIV/AIDS aware- of protection biases sectoral incentives, undermining ness programs to increase knowledge on how to con- efficient development of the manufacturing sector. trol the disease. Enterprises should be encouraged to Uganda is discussing new tariff rates with its part- take a more active role in controlling the spread of ners in the East African Community. This is also the HIV/AIDS--for example, by forming partnerships be- right time to change other elements of the trade policy tween business associations and voluntary counsel- regime. To improve allocative efficiency, as well as ing and testing programs. simplify tariff administration and reduce incentives for To address key constraints in the utilities sector, fraud, the government should consider reducing the the government should strengthen the regulatory number of tariff rates from the three now in use to two framework to facilitate private investment and com- or, in the long term, even to one. Decreasing the use plete the restructuring of the sector--under way for of ad hoc excise taxes would increase transparency, Executive Summary 12 lessen distortions, and help further reduce protection. The private sector could play an important role in ensuring that as many distortions as possible are removed. Notes: 1. The finding of poor access to credit is corrobo- rated by the World Bank's Doing Business data- base, which reports a very low index of informa- tion availability for Uganda (World Bank 2003). Similarly, studies conducted by the Bank of Uganda show that, apart from a few large estab- lishments, firms in Uganda lack the creditworthi- ness and collateral to ensure access to external fi- nance (see, for example, Kasekende and Opondo 2003). Executive Summary 13 UGANDA'S INVESTMENT CLIMATE AT-A-GLANCE Uganda China India Indicator 1995a 2002a 1995a 2000a 1995a 2000a Macroeconomic environment Gross national income per capita (PPP U.S. dollars)b 1,000 1,320 2,650 3,940 1,860 2,390 Population (midyear, millions) 19 23 1,205 1,261 929 1,016 Average annual growth of GDP (percent)c 7.0 6.1 12.1 8.2 5.2 6.1 Openness (imports + exports as a percentage of GDP) 32.6 39.8 45.7 47.1 25.7 28.2 Private investment (percentage of GDP) 10.2 15.0 15.8 16.7 16.9 16.6 Public investment (percentage of GDP) 5.4 7.0 18.9 19.2 7.7 7.1 Net inflows of foreign direct investment (percentage of GDP) 2.1 2.5 5.1 3.9 0.6 0.5 Microeconomic environment Inputs Average education of manufacturing workers (years) -- 5 -- 10 -- 10 Excess labor force -Redundant workers as a share of total (percent) -- ­7.7 -- -- -- 17.3 Share of inputs below quality standards (percent) -- 6.1 -- -- -- -- Stock of inputs (days of production) -- 27.9 -- -- -- 28 Research and development spending (percentage of sales) -- -- -- 2 -- -- Governance Informal payments (percentage of revenue) -- 2.44 -- -- -- -- Share of firms lacking confidence in judiciary (percent) -- 70 -- -- -- -- Average annual visits by government officials -- 13.4 -- -- -- 10.5 Share of senior managers' time spent with government officials (percent) -- 0.4 -- 9.2 -- 16.0 Infrastructure Share of firms with own generator (percent) -- 35 -- 16 -- 69 Longest wait to clear imports in previous year (days) -- 11.2 -- 8 -- 11 Telephone lines in largest city (per 1,000 people) -- 37 -- 294 -- 131 Personal computers (per 1,000 people) -- 3 -- 12 -- 3 Paved roads as a share of total (percent) -- 67 -- 88 -- 56 Finance Cost of capital (lending rate, percent) -- 16.7 -- 5.85 -- 12.29 Share of credit from financial institutions (for investment, percent) -- 11.6 -- 25 -- 36 (Table continued on next page) Executive Summary 14 UGANDA'S INVESTMENT CLIMATE AT-A-GLANCE--continued Uganda China India Indicator 1995a 2002a 1995a 2000a 1995a 2000a Credit to private sector (stock, percentage of GDP) -- 7 -- 125 -- 25 Entry and operation Cost of labor (median ratio of average wage to average value added) -- 0.39 -- 0.23 -- 0.21 -- Not available. a. Data are for the year specified or the most recent year available. b. PPP U.S. dollars are adjusted for purchasing power parity. c. The data shown for 1995 refer to 1991­95; those shown for 2000 (or, in Uganda's case, for 2002) refer to 1996­2000. Source: World Bank, World Development Indicators database; Investment Climate surveys, Uganda, 2002/03, China 2000 and India, 1999. 1. The Economic Environment 15 Investor Perceptions 16 Policy Initiatives 18 1. The Economic Environment 16 Uganda's Poverty Reduction Strategy, the Poverty the country today, with annual sales averaging less Eradication Action Plan, sets a goal of reducing the than $30,000. Like such enterprises elsewhere in the country's poverty rate to 10 percent by 2017. To region, most of these firms face severe managerial achieve that goal, Uganda will need to accelerate and financing constraints in scaling up their operation GDP growth to more than 7 percent a year on average and investment. Identifying these impediments as and boost its investment rate to 30 percent or more of well as strategies for overcoming them is a key objec- GDP. Given Uganda's impressive track record, these tive of this report. Large firms, particularly new en- requirements appear to be within reach. But closer trants in banking, power, and telecommunications, analysis suggests that the agenda is more challeng- may be in a position to grow rapidly, but their collec- ing than it seems. tive size is not enough to sustain the robust economic Since the late 1980s economic recovery in growth envisioned by the government. New invest- Uganda has come primarily from aid-financed reha- ments in new businesses, including those by foreign bilitation and reconstruction of the country's produc- investors, will need to play a key role. tive capacity. During that period the country has re- Foreign direct investment has become an in- ceived large development assistance flows relative to creasingly important part of the Ugandan economy. In the size of its economy. But foreign aid is expected to the 1970s and 1980s there was substantial capital decline, and growth will therefore need to come from flight out of the country. In the early 1990s capital out- new investments or new activities. That will require a flows ceased, but capital inflows amounted to little deeper commitment of capital, more intensive acqui- more than a trickle. Foreign investor sentiment gradu- sition of know-how, and more complex collaboration ally improved, however, and foreign direct investment between local and foreign partners. steadily rose. Since 1998 foreign direct investment Particularly important for attaining the poverty has exceeded $200 million a year, accounting for goal is the role of private investment. With severe lim- more than 20 percent of domestic capital formation. its on domestic resource mobilization and significant But investment inflows remain far below what is social spending commitments, government invest- needed to achieve the poverty goal. Meeting the ment is not expected to increase much above the growth and investment targets under the Poverty present level of about 6 percent of GDP. As a result, Eradication Action Plan would have required foreign private investment will need to increase from the level direct investment of $800 million in 2000, but even in recent years of about 10 percent of GDP to an av- though inflows peaked in that year, they amounted to erage of more than 24 percent over the time horizon only $254 million. of the Poverty Eradication Action Plan (ending in 2017). Such a high rate of investment is common in some developing countries, such as those in East Investor Perceptions Asia. But it is rare in Africa and in Uganda. To make it a reality, investors at home and abroad will need to Investor perceptions are generally measured by sov- gain confidence in the economic future of Uganda. ereign credit or country risk ratings issued by such in- Today the private sector in Uganda is a large col- ternationally recognized agencies as Euromoney, lection of small and medium-size firms and a very few Institutional Investor, Moody's, and Standard & Poor's. large firms, mostly in food processing and finance. The rating methodologies have well-known biases According to the Bank of Uganda, about 800,000 (especially against African countries), and not all in- small and medium-size enterprises are operating in vestors rely on these rating services. But the ratings 1. The Economic Environment 17 do have an impact on investor decisions and thus on tries with the best policies, perceptions change slowly. capital flows. Moreover, they can be influenced by regional events The surge of foreign direct investment in Uganda well beyond a country's control. in the 1990s corresponded to a major upgrade in the Globally, the competition for foreign direct invest- country's risk rating. In 1992 Institutional Investor, for ment is intense. Africa has been able to increase the example, gave Uganda a credit rating of 5 on a scale inflows of foreign direct investment in the past of 0­100, ranking it lowest among the 25 African coun- decade, but the volume is still very small on a per tries rated. But by 2000 Institutional Investor's rating of capita basis and compared with those in other re- Uganda had soared to 23--still low on a global basis gions. In 1998, for example, Sub-Saharan Africa re- but among the top ratings in Africa. Since then, how- ceived about $4.4 billion of foreign direct investment, ever, there has been no further improvement. In fact, a substantial increase from the less than $800 million in 2002 Institutional Investor downgraded Uganda's in 1990 but only a small share of the $180 billion going rating to 20 (table 1.1). Inflows of foreign direct invest- to all developing countries that year. The region re- ment have shown a corresponding decline. ceived about $7 per capita, compared with the devel- Given Uganda's need to attract significantly more oping country average of $33. In Uganda today, an- foreign direct investment, the recent decline in its rat- nual foreign direct investment amounts to about $11 ing is a cause for concern. The key to improving in- per capita--considerably higher than the regional vestor perceptions is to improve the conditions that average but still with significant upside potential. influence them. Much of the text in the following chap- To attract more foreign investment, Uganda needs ters deals with those conditions, drawing on a recent, to establish itself as a safe haven for investors in detailed firm-level survey, and the last chapter sug- Africa, as Botswana and Mauritius have done. It also gests policies for improving them. But even for coun- needs to provide global investors with a compelling Table 1.1 Country Credit Ratings and Rankings, Selected African Countries, 2002 Credit rating Regional Global One-year ranking Country ranking Value change 1 Botswana 39 59.0 2.3 9 Ghana 98 25.7 0.3 10 Kenya 103 22.9 1.2 12 Tanzania 107 21.3 0.7 14 Uganda 111 20.0 ­1.4 17 Mozambique 116 19.1 0.1 30 Burundi 143 11.3 0.8 35 Congo, Dem. Rep. of 149 8.7 1.2 Average 114 21.6 1.2 Note: The table includes only a sample of the Sub-Saharan African countries rated. Ratings range from 0 to 100, with 100 being the highest rating possible. Source: Institutional Investor, September 2002, p. 170. 1. The Economic Environment 18 reason to bypass South Asia or East Asia, for exam- Past Accomplishments ple, and come to Africa instead. Uganda faces both Uganda launched major reforms in the late 1980s. regional and global competition. But the first step is to The government began by liberalizing the trade re- reverse the recent setback. Unless the country effec- gime, removing quantitative restrictions and eliminat- tively addresses the causes of the backsliding, the ris- ing onerous export taxes. In 1990 it ended the allo- ing trend in capital inflows over the past decade may cation of foreign exchange, moved to market come to an end, with grave implications for private in- determination of the exchange rate, and made big im- vestment, growth, and, ultimately, the country's ability provements in fiscal discipline. In the early 1990s the to meet its poverty goal under the Poverty Eradication government liberalized the coffee industry, disengag- Action Plan. ing from the marketing, transport, and financing of coffee exports. Finally, Uganda initiated a program of restitution and incentives for returning Asians, who Policy Initiatives provided capital as well as managerial and entrepre- neurial skills during the early economic reconstruc- Uganda, despite its many disadvantages, has out- tion. These measures played a key part in turning performed most countries in Sub-Saharan Africa in around the Ugandan economy after two decades of attracting foreign capital, with per capita inflows of precipitous decline. foreign direct investment almost twice the average Although many of its competitors in the region for the region. According to the United Nations Con- have yet to make these fundamental reforms, Uganda ference on Trade and Development (UNCTAD) and was just getting started. During the 1990s it shifted its the International Chamber of Commerce, Uganda reform agenda toward promoting growth, tackling in- ranks among the 10 African countries making the creasingly difficult challenges. Between 1992 and most progress in upgrading their business environ- 1995 Uganda returned confiscated property to its ment in 2000­03. The key to Uganda's performance rightful owners, mostly Ugandans of Asian origin. It has been its ability to design and implement good liberalized the investment code by eliminating prefer- economic policies--an ability that has eluded most of ential treatment of domestic investors and making in- its competitors in the region. These policies are re- vestment approvals virtually automatic. By 1999 the flected in a record of strong and sustained economic government had successfully privatized about two- growth (box 1.1). This track record has brought the thirds of the 150 public enterprises, primarily through government dividends in credibility, which in turn in- sales of corporate control to strategic investors, both spires confidence among investors. domestic and foreign. Going forward, Uganda will need to rely increas- Even so, not all the planned reforms were suc- ingly on its credibility and good policies to win in- cessfully implemented. And some reforms led to new vestors' confidence. Unlike China or South Africa, for problems. Many large public enterprises and agen- example, Uganda must overcome disadvantages of cies remained engaged in the delivery of commercial market size and logistics. It must deliver attractive re- services, particularly in utilities and infrastructure. The turns on investments despite the high costs inherent effort to privatize these entities met with resistance in its location and its economy. Moreover, after more and technical complexity. Budgetary exigencies led to than a decade of reforms and impressive achieve- a big increase in the taxation of imports, and revenue ments, the agenda remaining for Uganda will become from this source has since accounted for more than increasingly difficult. half the resources that the government has mobilized 1. The Economic Environment 19 Box 1.1 Uganda's Economic Track Record Uganda has turned in an exemplary economic per- rity fronts. In 1997 the government adopted a major formance for almost two decades. Starting from a low poverty reduction initiative involving substantial in- point in 1986, after 15 years of mismanagement and dev- creases in budgetary allocations for health and edu- astation, the country has produced a record of recovery cation. Soon thereafter Uganda achieved universal en- and growth seldom seen in Africa. Between 1990 and rollment in primary education, the first country in 2001 Uganda outperformed neighboring Kenya and Sub-Saharan Africa to do so. Meanwhile, the country Tanzania by a significant margin, and even Mozambique made efforts to contain the spillover of conflicts in neigh- lagged behind for most of the decade (see figure). boring countries (the Democratic Republic of Congo, Uganda's real GDP grew by about 8 percent a year over Rwanda, and Sudan) as well as the threat of domestic this period, while inflation decelerated sharply to single- unrest in the north. This track record was not lost on the digit levels. donor community. Uganda became the first country to During this period of strong economic growth qualify for debt relief under the Heavily Indebted Poor Uganda also made much progress on social and secu- Countries (HIPC) initiative. Real Per Capita Income, Selected East African Countries, 1990­2001 Index (1990 = 100) 250.0 200.0 Kenya Mozambique 150.0 Tanzania Uganda 100.0 50.0 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 Source: World Bank data. domestically. A heavy reliance on import taxes inhibits ity and unreliability in such crucial services as water, trade and, like export taxes, handicaps exporters.1 power, transport, and telecommunications. As a re- sult, private firms needed to provide for costly alter- Ongoing Efforts native or backup services, such as power generators. As the new millennium dawned, the largest and The extra costs were passed from firm to firm and strategically most important public enterprises contin- from sector to sector, creating economywide ineffi- ued to burden the economy. Most provided poor qual- ciency that eroded the competitiveness of local pro- 1. The Economic Environment 20 duction and undermined Uganda's attractiveness as a based in South Africa. And in July 2003 the govern- host for foreign direct investment. Some of these pub- ment reached the final stage of negotiation on the lic enterprises dominated the most promising busi- concession for power distribution with an international nesses, including tourism, agribusiness, and financial consortium consisting of CDC Globeleq and Eskom services, making it difficult for more efficient, dynamic Enterprises, the sole bidder in a competitive tender. firms to succeed or even to enter these sectors. But Meanwhile, a number of independent power produc- the government did not give up on its reform and pri- ers indicated to the government that their prospective vatization efforts. investments in power generation would be contingent Indeed, some of the Ugandan government's re- on the privatization of distribution. cent advances in restructuring the economy have The progress in restructuring the telecommunica- been exceptionally impressive by regional standards. tions and power sectors should boost investor confi- In telecommunications the government achieved dra- dence in Uganda. Many countries across the con- matic growth in access to services and big improve- tinent have targeted these economically sensitive ments in quality in just two years by relying on private industries for reform, but few have succeeded. Many participation and competitive markets. The govern- good intentions have been derailed by the technical ment launched the reform in 1996, restructuring the complexity, the resistance of vested interests, and the public enterprise responsible for telephone and postal apprehension of consumers. By regional standards, services into three new entities: telecommunications, Uganda carried out these reforms fairly quickly (in the post office, and the post office bank. More impor- about two years in each case), largely ending the tant, it opened the telecommunications sector to com- drain on government resources. petition and private participation, awarding two oper- The analysis in the following chapters provides a ating licenses to private service providers. By 1998 strategy for private sector development in Uganda the number of telephone lines in Uganda had almost that builds on these past accomplishments. Using the doubled, rising from fewer than 50,000 to 96,000, with pillars of the Poverty Eradication Action Plan as a most of the gains coming from mobile service. guide, the analysis highlights key constraints and pol- The power sector, often considered the most seri- icy solutions for the government and the private ous bottleneck in the economy, was the next target for sector. restructuring. In 2001, after securing enabling legisla- tion for liberalization, the government established an electricity regulatory authority and divided the Notes Uganda Electricity Board into three separate compa- nies responsible for generation, transmission, and dis- 1. One interesting study argues that managers' inter- tribution, respectively. The restructuring strategy national orientation is inadequate: their attitude to- called for privatizing power generation and distribu- ward exporting is either ambivalent or negative tion while retaining transmission in the public sector. (Bankunda 2004). Through econometric estima- In August 2002 the government awarded the conces- tion, the study shows that managers' attitude to- sion for power generation on a competitive basis to ward exports is a significant determinant of firms' Eskom Enterprises, an established industry leader exports. 2. Growth and Productivity of Firms 21 Characteristics of Enterprises and Entrepreneurs 22 Characteristics and Productivity of Capital 26 Productivity and Cost of Labor 29 Total Factor Productivity and Technical Efficiency 32 2. Growth and Productivity of Firms 22 What are the key drivers of growth and productivity in own in Uganda to see how these differences affect the the Ugandan private sector? This question serves as performance of these firms. Entrepreneurs from ethnic the motivation for this chapter, which analyzes the minorities play a substantial role in Sub-Saharan growth and productivity of Ugandan firms of different African economies, especially in manufacturing. They sizes and ownership types using survey data on man- are sometimes credited with providing valuable skills ufacturing firms. A good understanding of what deter- and stimulating growth in manufacturing. But they also mines the competitiveness of Ugandan firms is crucial are often accused of exploiting local resources to meeting the goals of the Poverty Eradication Action through unfair advantages based on networks and Plan. Growth in the private sector is essential for fuel- personal characteristics. ing overall economic growth. And policies to promote The results show that, on average, Ugandan firms growth will bring high returns if they are targeted to the have lower labor productivity than firms in other Sub- factors most important in determining the perform- Saharan African countries--and much lower labor ance of private firms. productivity than those in China and India. But this The data come from a survey of registered enter- is partly offset by their capital productivity, which is prises in the Ugandan private sector undertaken in higher than that in the other countries in Africa. November 2002­April 2003 by the World Bank's Re- What factor is most important in determining the gional Program on Enterprise Development (RPED) performance of firms in Uganda? The human capital in collaboration with the Uganda Manufacturers of entrepreneurs, as reflected in education, previous Association Consultancy and Information Services experience in other firms, and experience in export- (UMACIS). This survey covered 392 firms across four ing. Government policy directed toward augmenting sectors (commercial agriculture, construction, manu- these learning channels would help improve the com- facturing, and tourism) and three regions (central, petitiveness of private firms in Uganda. northeast, and southwest). (For a detailed description of the sample, see appendix 1.) The survey allows an assessment of the business environment in Uganda Characteristics of Enterprises and as well as a comparison of the performance of its pri- Entrepreneurs vate sector with that in several other countries for which recent survey data are available. Almost 70 percent of firms in the Ugandan sample are How do firms in Uganda compare with those in entrepreneur owned (that is, owned by an individual other Sub-Saharan African countries and elsewhere? or family), a slightly smaller share than in Kenya and How do they differ across size classes? How do en- Mozambique but a much larger share than in Nigeria trepreneurs differ within Uganda and across coun- (table 2.1). Most private firms in Uganda (around 83 tries? And what characteristics of firms and entrepre- percent) were established by entrepreneurs rather neurs matter most in determining the performance of than purchased or inherited. This is again similar to firms? The chapter investigates these issues by com- what is found in other African countries recently sur- paring the firm-level data from Uganda with data from veyed, except Mozambique, where most firms were several Sub-Saharan African countries (Kenya, state owned until being purchased by individuals Mozambique, Nigeria, Tanzania, and Zambia) and, through the government's privatization initiative in the where possible, China and India. It also examines dif- 1990s. Unlike in the other countries, in Uganda a sig- ferences in characteristics between ethnic African nificant share of firms (23 percent) are organized as and ethnic Asian entrepreneurs and the firms they sole proprietorships. These firms are mainly small 2. Growth and Productivity of Firms 23 Table 2.1 Characteristics of Private Manufacturing Firms, Selected Sub-Saharan African Countries (percent) Uganda Kenya Mozambique Nigeria Share of firms owned by entrepreneurs 69.3 74.3 72.4 44.9 Firms by form of acquisition Established 82.9 81.1 43.7 86.5 Bought 14.4 16.9 43.3 3.1 Inherited 1.0 1.2 9.3 7.3 Firms by legal status Sole proprietorship 23.0 5.2 -- -- Partnership 8.3 2.0 -- -- Limited liability 63.0 88.5 -- -- Firms by source of start-up finance Own savings 85.4 65.1 78.5 79.8 Informal loans 8.1 8.6 6.0 1.9 Bank loans 6.6 20.4 9.2 10.7 Firms by ethnicity of entrepreneur African 75.1 14.1 -- -- Asian 20.6 72.8 -- -- European 2.9 10.5 -- -- -- Not available. Note: Percentages do not add to 100 because of the category other (not shown). Source: World Bank, RPED investment climate surveys, Nigeria, 2001, Mozambique, 2002, Uganda, 2002/03, Kenya, 2003. ones that may lack knowledge of the formal registra- larger firms and thus control a large share of business tion process required to become limited liability com- assets in Uganda. panies. The data show that obtaining financing to start a Human Capital Endowments of business in Uganda is difficult. Some 85 percent of Entrepreneurs firms were started with the savings of entrepreneurs. A comparison of human capital in Uganda and other This large share--larger than in the other Sub- countries reveals interesting differences (table 2.2). Saharan countries--probably reflects the poor devel- Most entrepreneurs in Uganda (almost 90 percent) opment of the financial sector and the credit con- have at least a secondary education; 40 percent have straints facing firms (for more on this issue, see the a university degree. Although the share with a univer- section on firms' access to finance in chapter 4). sity degree is smaller than that in Kenya and Nigeria, Around 75 percent of firms in Uganda are owned by it is larger than that in Mozambique. In comparison, indigenous Africans, and most of the rest by Asian en- almost all entrepreneurs in China and India have a trepreneurs. But Asian entrepreneurs own most of the university degree. 2. Growth and Productivity of Firms 24 Table 2.2 Education and Experience of Manufacturing Entrepreneurs, Selected Developing Countries Uganda Kenya Mozambique Nigeria China India Entrepreneurs by highest level of education achieved (percent) None 3.4 0.0 0.0 3.1 0.0 0.4 Primary 8.1 4.0 17.8 5.21 0.1 0.6 Secondary 19.8 23.2 44.1 10.8 15.2 9.8 Vocational 29.2 13.4 23.5 -- -- -- University 39.6 59.4 14.7 70.8 84.7 89.2 Average years of experience 5.0 5.4 -- -- 10.4 9.95 Share of entrepreneurs with experience in foreign firm (percent) 21.6 22.9 -- -- -- -- -- Not available. Source: World Bank, RPED investment climate surveys, Nigeria, 2001, Mozambique, 2002, Uganda, 2002/03, Kenya, 2003; and Investment Climate Unit firm surveys, China 2000 and India 1999. Within Uganda there are differences between ethnic ployees between start-up and the present (table 2.4). groups. On average, Asian entrepreneurs have more By contrast, 60 percent of small firms (10­49 employ- education and much more experience before starting ees) remained in the same size class. This lack of mo- their firms than African entrepreneurs do (table 2.3). In bility may be a result of constraints facing small firms, addition, a much larger share of Asian entrepreneurs such as lack of access to finance, skilled labor, and have experience in a foreign firm before starting their new technology. own enterprise. And they are more likely than their What drives the growth of manufacturing firms in African counterparts to buy a firm and to obtain external Uganda? This question is addressed by first examin- loans for start-up. For all these reasons Asian entre- ing whether initial firm size and firm age matter in de- preneurs start at a larger scale and their firms grow termining firm growth. Then different attributes of en- faster. On average, African-owned firms start with 14 trepreneurs are introduced into the equation, to see employees and end up with 31 employees after almost how such factors as human capital, sources of finance 10 years, while Asian firms start with 39 employees and at start-up, and the nature of a firm's acquisition affect grow to 104 employees after a little more than 10 years. the rate of firm growth. These results are robust across a wide range of countries The results, similar to those found elsewhere in surveyed in Sub-Saharan Africa. Sub-Saharan Africa, show that human capital plays an important part in determining the rate at which firms Characteristics Determining Firm Growth grow. Firms whose owner-managers have secondary The mobility of manufacturing firms between start-up or university degrees grow significantly faster than and the present is greater with size. Medium-size those whose owner-managers have only a primary ed- firms (50­99 employees) are much more mobile than ucation (table 2.5). Firms able to obtain formal loans smaller firms: more than half of those starting out in at start-up are not likely to grow faster, but they do this size class were able to grow to 100 or more em- start bigger. 2. Growth and Productivity of Firms 25 Table 2.3 Characteristics of Manufacturing Entrepreneurs and Their Firms by Ethnicity, Uganda African Asian Entrepreneurs by highest level of education achieved (percent) None 4.4 0.0 Primary 11.5 0.0 Secondary 62.4 33.3 University 21.6 66.7 Entrepreneurs' prior experience Average years of experience 3.8 8.1 Share who worked in foreign firm (percent) 15.9 47.9 Share who established own firm (percent) 87.9 75.0 Share of entrepreneurs obtaining start-up finance (percent) Informal loans 6.4 9.6 Formal loans 7.0 17.3 Firm characteristics Firm size at start-up (employees) 14 39 Average firm age (years) 9.7 10.1 Current firm size (employees) 31 104 Average annual growth (percent) 0.19 0.21 Source: World Bank, RPED investment climate survey, Uganda, 2002/03. Table 2.4 Mobility of Manufacturing Firms across Size Classes, Uganda (percent) Firms by size class (employees) in 2002/03 Firms by size class (employees) at start-up Micro Small Medium-size Large (<10) (10­49) (50­99) (100+) Micro (<10) 31.6 56.6 7.4 4.4 Small (10­49) 5.8 60.5 17.4 16.3 Medium-size (50­99) 0.0 26.3 21.1 52.6 Large (100+) 0.0 7.1 0.0 92.9 Note: Firms are classified by their size at start-up. Of the 14 firms starting in the largest size class (100+ employees), only one downsized between start-up and the time of the survey. Source: World Bank, RPED investment climate survey, Uganda, 2002/03. 2. Growth and Productivity of Firms 26 firms, more likely for African than for Asian entrepre- Table 2.5 Determinants of Growth of neurs, the firms tend to grow more slowly. On the Manufacturing Firms, Uganda whole, Asian firms enjoy the benefits of inherited Variable Model 1 Model 2 ownership and experience. But access to university Constant 0.51* 0.58* education for indigenous entrepreneurs may help off- (0.03) (0.08) set these advantages. Thus it may be worthwhile for Log of employment ­0.05* ­0.08* (0.01) (0.01) both the government and donors to revisit their priori- Log of age ­0.11* ­0.11* ties in this area.1 (0.01) (0.02) Food 0.01 0.04 (0.02) (0.03) Characteristics and Productivity of Textiles and garments 0.002 0.02 Capital (0.05) (0.04) Wood and furniture ­0.03 0.03 Uganda has both low capacity utilization and high (0.03) (0.06) marginal productivity of capital. Low capacity utiliza- Metals 0.004 0.01 (0.03) (0.05) tion can indicate demand constraints or lack of avail- Secondary education 0.07*** ability of inputs on the supply side. For Uganda, a (0.04) small market with low purchasing power is the most University education 0.11* likely reason for the low capacity utilization. But the (0.04) country's capital stock is of relatively high quality, so Log of experience 0.01 Ugandan firms will probably be able to meet in- (0.01) creases in demand with greater efficiency than their Loan at start-up 0.03 neighbors. (0.04) African ­0.06*** (0.03) Age of Capital Stock Informal loan ­0.06 Uganda's capital stock is exceptionally young. More (0.05) than 40 percent of its manufacturing firms have capi- Owner established ­0.07** tal stock averaging less than 5 years old, and another (0.04) 35 percent have capital stock averaging 5­10 years * Significant at the 1 percent level. old (table 2.6). This contrasts sharply with the rest of ** Significant at the 5 percent level. Sub-Saharan Africa: in most countries in the region a *** Significant at the 10 percent level. large share of the capital stock is more than 20 years Notes: Dependent variable is rate of change in employ- ment. Figures in parentheses are standard errors. old. Preliminary estimates for Kenya indicate that half Source: Authors' calculations based on data from World the sampled manufacturing firms have capital stock Bank, RPED investment climate survey, Uganda, 2002/03 averaging 11­20 years old. Uganda's young capital stock is due to two main factors, the higher rates of investment in the 1990s and Indigenous African entrepreneurs have less edu- the large inflows of foreign exchange that led to capi- cation and experience than their Asian and European tal accumulation in the postwar period. The relatively counterparts, and their firms grow significantly more young age of its capital stock may well indicate that slowly. And when entrepreneurs establish their own the technology is more recent and of better quality-- 2. Growth and Productivity of Firms 27 Table 2.6 Manufacturing Firms by Average Age of Equipment, Uganda (percent) Age of equipment (years) <5 5­10 11­20 >20 Full sample 42.3 35.0 14.9 7.7 (42.0) (40.2) (31.7) (24.0) Firm size class (employees) Micro (<10) 60.3 26.3 9.6 3.8 Small (10­49) 40.6 35.9 16.3 7.2 Medium-size (50­99) 30.7 46.2 20.6 2.5 Large (100+) 37.5 34.1 12.5 15.9 Subsector Food 45.8 35.3 10.2 8.8 Textiles and Garments 19.8 41.8 21.3 17.1 Wood and Furniture 61.2 21.5 12.2 5.2 Metals 44.8 44.0 9.0 2.3 Note: Figures in parentheses are standard deviations. Source: World Bank, RPED investment climate survey, Uganda, 2002/03. and thus more productive. If other constraints to the facturing subsectors, with food and metal firms hav- private sector (such as the poor business environ- ing the highest capacity utilization. ment) are addressed, it is quite likely that firms could better utilize their capital and achieve faster growth. Capital Intensity Uganda has the lowest capital intensity in the sample Capacity Utilization (table 2.8). It has only about $1,500 of capital per Firms may have a young capital stock, but how effi- worker; in sharp contrast, this ratio is several times ciently they use it depends on factors driven by prod- as large in recently surveyed neighboring countries. uct demand or bottlenecks in raw material supply. Large firms in Uganda have the most capital per How efficiently is Ugandan capital used? A compari- worker, more than four times the median for the sam- son with other African countries shows that capacity ple. Micro firms have only slightly more than half the utilization in Uganda is about average (table 2.7). median. Not surprisingly, exporters (firms exporting Ugandan firms use close to 60 percent of their ca- 10 percent or more of their annual sales) have sub- pacity, and the dispersion across firm size classes is stantially more capital per worker (a median of about 15 percent, within the typical range for African $3,277) than nonexporters ($1,408), and foreign- countries. Large firms in Uganda use a greater share owned firms substantially more ($3,930) than domes- of their capacity than smaller ones, presumably be- tically owned firms ($1,408). cause they have better-quality capital equipment, Interestingly, the largest firms in Uganda have larger market shares, and better access to labor and higher capital intensity than firms of a similar size in other inputs. There is also dispersion across manu- India, but lower capital intensity than the largest firms 2. Growth and Productivity of Firms 28 Table 2.7 Capacity Utilization in Manufacturing, Selected Sub-Saharan African Countries (percent) Cameroon Côte d'Ivoire Ghana Kenya Tanzania Zambia Uganda Full sample 46.9 70.7 54.3 63.3 51.1 48.4 58.4 (28.5) (25.3) (27.4) (28.2) (27.2) (30.3) (22.6) Firm size class (employees) Micro (<10) 40.5 66.6 52.5 56.3 58.8 50.4 50.6 Small (10­49) 44.3 68.4 55.7 65.6 48.5 50.2 58.1 Medium-size (50­99) 47.0 67.9 48.4 67.3 38.8 42.9 60.8 Large (100+) 60.6 78.5 59.6 69.3 42.3 46.4 65.0 Subsector Food 50.7 70.8 57.4 67.3 46.2 50.1 58.8 Textiles and Garments 38.0 67.9 51.1 59.9 47.3 43.4 54.4 Wood and Furniture 55.0 68.8 52.3 67.1 55.2 53.4 55.7 Metals 41.3 77.3 57.0 59.5 53.0 47.7 61.2 Note: Figures in parentheses are standard deviations. Source: World Bank, RPED investment climate surveys, Ghana, 1994, Cameroon, 1995, Côte d'Ivoire, 1996, Uganda, 2002/03, Kenya, 2003, Tanzania, 2003, Zambia, 2003. Table 2.8 Median Ratio of Capital to Labor in Manufacturing, Selected Developing Countries (U.S. dollars per worker) Firm size class (employees) Tanzania Uganda Kenya Zambia India China Micro (<10) 1,040 845 -- -- 1,859 -- Small (10­49) 7,433 1,408 7,436 15,578 2,000 5,434 Medium-size (50­99) 7,493 2,453 16,816 18,175 2,962 6,070 Large (100+) 19,279 6,667 11,420 8,178 4,158 8,525 All firm size classes 7,757 1,464 11,496 12,161 2,380 7,654 -- Not available. Note: The sample for Kenya included only 3 firms with fewer than 10 employees, and that for China only 10 firms. Source: World Bank, RPED investment climate surveys, Uganda, 2002/03, Kenya, 2003, Tanzania, 2003, Zambia, 2003; and Investment Climate Unit firm surveys, China 2000 and India 1999. in Sub-Saharan Africa. The largest firms in the region Productivity of Capital appear to be substituting capital for labor, perhaps Uganda has remarkably high capital productivity because of labor laws or a lack of skilled labor. compared with that in other Sub-Saharan African countries. In Uganda, despite capacity utilization of 2. Growth and Productivity of Firms 29 Table 2.9 Median Ratio of Annual Value Added to Capital in Manufacturing, Selected Developing Countries Firm size class (employees) Tanzania Uganda Kenya Zambia India China Micro (<10) 1.33 0.80 -- -- 0.80 0.13 Small (10­49) 0.37 0.67 0.30 0.16 1.11 0.59 Medium-size (50­99) 0.61 0.43 0.46 0.24 1.48 0.67 Large (100+) 0.26 0.89 0.34 0.35 1.16 0.47 All firm size classes 0.43 0.70 0.35 0.23 1.10 0.51 -- Not available. Source: World Bank, RPED investment climate surveys, Uganda, 2002/03, Kenya, 2003, Tanzania, 2003, Zambia, 2003; and Investment Climate Unit firm surveys, China 2000 and India 1999. only 60 percent, every dollar of capital generates Ugandan firms, and in Kenyan firms it is just as high twice as much in value added in a year's time as a as in Indian firms. An alternative presentation of the dollar of capital does in neighboring countries. data shows that all the countries in East Africa have far to go to catch up with China in labor productivity (figure 2.2). Productivity and Cost of Labor Within Uganda the productivity of labor varies widely among firms of different sizes. Microenter- In sharp contrast to the relatively high productivity of prises are the least productive, with value added per capital in Uganda is the relatively low productivity of worker only 50 percent that of the median for the sam- labor. A simple comparison of labor productivity sug- ple (table 2.10). Small firms are at 80 percent of the gests that Ugandan labor cannot compete with labor median. Indeed, micro and small enterprises in in other parts of the world. But if lower wages offset Uganda are much less productive even than their the lower productivity, Ugandan firms would still be African counterparts, perhaps reflecting the large dis- able to compete with those elsewhere. How does parities in education levels across firm size classes Uganda compare in this area? Relatively high unit (see appendix table A2.10). Firms with 50 or more labor costs suggest that it has a long way to go before workers show labor productivity above the median. In it can compete globally. large firms labor productivity is three times the me- dian for the sample--comparable to levels for large Labor Productivity firms in other African countries and similar to those for The gap in labor productivity between Ugandan firms medium-size firms in India. and firms in other countries is substantial. Value Labor productivity in Uganda also varies between added per worker in Uganda is 68 percent lower than other groups of firms. Exporters have significantly that in India and 96 percent lower than that in China higher value added per worker ($2,901) than nonex- (figure 2.1). Even the gap between Uganda and its porters ($1,117). And foreign firms produce much neighbors is fairly substantial. In Tanzanian firms more per worker ($2,747) than domestic firms labor productivity is 28 percent higher than in ($1,182). 2. Growth and Productivity of Firms 30 Figure 2.1 Labor Productivity in Manufacturing Relative to That in India, Selected Developing Countries 40 28 worker 20 per 0 added 0 value in ­20 ­22 ­40 difference ­39.9 ­60 Percentage ­68 ­80 China Kenya Zambia Tanzania Uganda Note: The base case is labor productivity in India in 2002­03. Data are based on nominal exchange rates. Source: World Bank, Investment Climate surveys, Uganda 2002/03, Kenya, 2003, Tanzania, 2003, Zambia, 2003, China 2000. Figure 2.2 Median Annual Value Added per Worker in Manufacturing, Selected Developing Countries Productivity of Labor by Country: Median Value Added per Worker in US$ China 4379 Kenya 3457 India 3432 Zambia 2680 Tanzania 2061 Uganda 1085 0 500 1000 1500 2000 2500 3000 3500 4000 4500 Source: World Bank, Investment Climate surveys, Uganda, 2002/03, Kenya, 2003, Tanzania, 2003, Zambia,. 2003, China 2000 and India 1999. 2. Growth and Productivity of Firms 31 Table 2.10 Median Annual Value Added per Worker in Manufacturing by Firm Size Class, Selected Developing Countries (U.S. dollars) Firm size class (employees) Tanzania Uganda Kenya Zambia India China Micro (<10) 989 578 -- -- 3,147 1,920 Small (10­49) 1,526 897 2,439 2,668 2,931 4,595 Medium-size (50­99) 3,288 1,379 4,127 3,836 3,228 4,797 Large (100+) 3,499 3,338 4,138 2,439 5,321 4,193 All firm size classes 2,061 1,085 3,457 2,680 3,432 4,397 -- Not available. Source: World Bank, Investment Climate surveys, Uganda, 2002/03, Kenya, 2003, Tanzania, 2003, Zambia, 2003, China 2000 and India 1999. Unit Labor Costs wage (w) to labor productivity (Q/L). For a country to Uganda's competitiveness is inextricably linked to its have a low (competitive) unit labor cost, it must keep labor cost. Two indicators of labor cost are crucial: the nominal wages low, keep its exchange rate competi- wages paid to workers, and the output of these work- tive, or increase its labor productivity--or do a combi- ers relative to their wages. nation of these things. Let's look first at wage levels. In Uganda unskilled Because physical measures of output compara- production workers in manufacturing earn about $57 ble across countries are difficult to obtain, an approx- a month on average--less than comparable workers imate measure of unit labor cost is used here--the in the Philippines, Kenya, Nigeria, and Thailand but ratio of wages to value added at the firm level, aver- more than in India, where unskilled workers earn aged across the sample of firms (wL/pQ), where p is about $45 a month. In China, where wages have risen the deflator for physical value added. Based on this rapidly in the past decade, unskilled production work- measure, the unit labor cost in Uganda is comparable ers earn about $85 a month on average. to that in several countries in Sub-Saharan Africa but Now let's look at the total cost of labor per unit of higher than that in India and China (table 2.11). output--or the unit labor cost. When the unit labor By definition, the unit labor cost is higher in coun- cost is converted to a common currency, it allows in- tries that have higher wages or lower labor productiv- ternational comparisons of the competitiveness of ity (or both). While overvalued exchange rates have labor. The unit labor cost in U.S. dollars is defined as hampered Africa's competitiveness, data show that high unit labor costs have also played a part. ULC = (w.L/Q)(1/e) Countries in Sub-Saharan Africa, including Uganda, have higher unit labor costs today than East Asian where w is the manufacturing wage, L is the amount economies had at roughly equivalent stages of devel- of labor employed, Q is a physical measure of output, opment (table 2.12). Earnings in Africa today are and e is the exchange rate defined as domestic cur- about two-thirds higher than they were in East Asia in rency units per U.S. dollar. The unit labor cost can the 1960s and 1970s, while productivity in Africa is also be approximated by the ratio of the nominal about one-fourth lower. 2. Growth and Productivity of Firms 32 Table 2.11 Unit Labor Costs in Manufacturing, Selected Developing Countries (median ratio of wages to value added) Firm size class (employees) Tanzania Uganda Kenya Zambia India China Micro (<10) 0.45 0.33 -- -- 0.29 -- Small (10­49) 0.56 0.41 0.38 0.41 0.30 0.38 Medium-size (50­99) 0.42 0.41 0.41 0.47 0.25 0.34 Large (100+) 0.25 0.35 0.34 0.39 0.24 0.29 All firm size classes 0.39 0.39 0.36 0.41 0.27 0.32 -- Not available. Source: World Bank, Investment Climate surveys, Uganda, 2002/03, Kenya, 2003, Tanzania, 2003, Zambia, 2003, China 2000 and India 1999. Disaggregating the data for Uganda by firm size Foreign firms have lower unit labor costs than domes- class shows that the unit labor cost is highest for small tic firms. And firms owned by nonindigenous entre- and medium-size firms, and lowest for micro and preneurs have lower unit labor costs than those large firms. But large firms in Uganda have much owned by indigenous entrepreneurs.2 higher unit labor costs than those in Tanzania, India, and China. Indeed, large firms in Uganda lag farther behind their counterparts in China than do medium- Total Factor Productivity and Technical size firms (in percentage terms). Efficiency The unit labor cost also ranges widely across manufacturing subsectors: it is lowest for the textile The measures of partial factor productivity in the pre- and garment industry, and highest for the wood and ceding sections provide some insight into firm per- furniture subsector. The unit labor cost for exporters formance in Uganda. But considered in isolation, they (0.24) is less than half that for nonexporters (0.55). provide a misleading indication of overall productivity. As shown, labor productivity is very low in Uganda compared with other countries, while capital produc- Table 2.12 Historical Unit Labor Costs, Selected East Asian Economies tivity is high. This section investigates the net impact of labor and capital on a firm's value added by look- Median ratio of wages to ing at total factor productivity. Economy Year value added Indonesia 1981 0.21 Determinants of Total Factor Productivity Korea, Rep. of 1963 0.26 Total factor productivity is first examined by estimat- Malaysia 1970 0.27 ing a Cobb-Douglas production function using the Singapore 1963 0.35 standard ordinary least squares technique. A log- Taiwan (China) 1961 0.16 linear specification is used, with the log of value Thailand 1970 0.24 added as the dependent variable. The results of these Source: Lindauer and Velenchik 1994. regressions are in table 2.13. The first model exam- 2. Growth and Productivity of Firms 33 The second model augments the production func- Table 2.13 Determinants of Firm Productivity tion by adding some characteristics of firms and en- in Manufacturing, Uganda: Ordinary Least trepreneurs. The results show that the education of Squares Regression Results entrepreneurs and managers is significant in deter- Variable Model 1 Model 2 mining a firm's value added. Firms whose manager Constant 3.02*** 3.31*** (0.55) (0.52) has a university education outperform all others, while Log of capital 0.35*** 0.33*** those whose manager has a secondary education are (0.06) (0.05) significantly more productive than those whose man- Log of labor 0.78*** 0.71*** ager has a primary education or none.3 A manager's (0.11) (0.10) years of experience at a foreign firm are also signifi- Capacity utilization 0.02*** 0.01*** cant in determining the productivity of a firm. This (0.004) (0.003) result shows the importance of general training re- Food 0.11 0.11 ceived by employees at foreign enterprises; this train- (0.23) (0.23) ing helps entrepreneurs start and manage more pro- Textiles and garments 0.27 0.02 (0.38) (0.34) ductive firms. But being an exporter in Uganda does Wood and furniture ­0.48 ­0.40 not seem to matter in determining firm performance.4 (0.28) (0.26) Metals ­0.10 ­0.04 Performance of Firms Relative to the (0.37) (0.33) Efficiency Frontier Secondary education 0.49* In addition to the traditional approach to estimating (0.27) productivity, the stochastic frontier approach is used to University education 0.58* shed light on which types of firms in Uganda perform (0.32) better than others. This approach relaxes the assump- Years worked in foreign 0.06*** country (0.02) tions of a traditional production function by attributing Exporting 0.001 a technical inefficiency component to the error term (0.004) rather than assuming it to be purely random. This tech- Adjusted R2 0.77 0.81 nique is best suited for countries with noisy data, like * Significant at the 10 percent level. those in Sub-Saharan Africa. The frontier itself is ran- ** Significant at the 5 percent level. dom, determined by the "best practice" firm in the *** Significant at the 1 percent level. country. There is a dispersion of firms below the fron- Note: The dependent variable is the log of value added. tier based on their technical inefficiency. Figures in parentheses are standard errors. Source: Authors' calculations based on data from World A preliminary stochastic frontier is estimated for Bank, Investment Climate survey, Uganda, 2002/03. Uganda, including only subsector controls along with capital, labor, and capacity utilization. Individual firm ines the basic production function specification and efficiencies are calculated from the total factor pro- controls for subsectors. The coefficients of labor and ductivity frontier. The results show that the average capital are positive and significant, as expected, and technical efficiency in Uganda is 0.51, indicating so is the coefficient of capacity utilization. The co- that, on average, firms are only 50 percent as effi- efficients of labor and capital are close to 1 (1.03), cient as the best practice firm (table 2.14). Low aver- indicating constant returns to scale in Ugandan age efficiency is typically associated with uncompet- manufacturing. itive, segmented markets. In competitive economies 2. Growth and Productivity of Firms 34 efficient than those started by an entrepreneur without Table 2.14 Results of Stochastic Frontier Estimations such experience. Overall, the frontier results are similar to the pro- Variable Coefficient duction function results: learning channels are the Intercept 1.36* (0.72) most significant driver of firm performance. Log of capital 0.34*** Learning--whether through export experience, work (0.046) in a foreign firm, or advanced education--helps boost Log of labor 0.79*** firm performance. Government policies that help de- (0.094) velop these learning channels will thus foster private Capacity utilization 0.021*** sector growth in Uganda. (0.005) Food 0.34 Table 2.15 Average Efficiency of (0.23) Manufacturing Firms Grouped by Various Textiles and garments 0.18 Characteristics, Uganda (0.39) Wood and furniture ­0.38 Characteristic Average efficiency (0.28) Firm size class (employees) Metals 0.17 Micro (<10) 0.50 (0.35) Small (10­49) 0.52 Average technical efficiency 0.51 Medium-size (50­99) 0.48 * Significant at the 10 percent level. Large (100+) 0.54 ** Significant at the 5 percent level. *** Significant at the 1 percent level. Market orientation Note: Figures in parentheses are standard errors. Source: Authors' calculations based on data from World Exporter 0.56 Bank, Investment Climate survey, Uganda, 2002/03. Nonexporter 0.51 firms' average efficiency tends to be close to 0.75 or Ownership 0.80. Foreign 0.53 When average efficiency is estimated for firms Domestic 0.51 grouped by different characteristics, no significant differences emerge across size classes (table 2.15). Entrepreneurs' highest level of education achieved Nor is there a significant difference between foreign- Primary 0.47 owned and domestic firms. Whether a firm exports Secondary 0.51 does matter, however: exporters are more efficient Technical 0.53 than nonexporters. University 0.60 But the most important factor in explaining differ- ences in firm efficiency is education. Firms whose Entrepreneurs' prior experience manager has a university degree are 13 percent more Foreign firm experience 0.56 efficient than firms whose manager has only a primary No foreign firm experience 0.51 education. Similarly, firms started by an entrepreneur Source: Authors' calculations based on data from World who had previously worked in a foreign firm are more Bank, Investment Climate survey, Uganda, 2002/03. 2. Growth and Productivity of Firms 35 Notes 4. The percentage of foreign ownership was also tested as an explanatory variable. It was not sig- 1. Entrepreneurs with high ability are likely to choose nificant. In addition, several investment climate and achieve a higher level of schooling while at variables were included in the production function the same time managing their firms better. This to evaluate their impact on productivity. These might create an upward bias in the estimates of specifications also were not significant. Several the effect of schooling. factors probably account for this: the sample of 2. Worker earnings data show that foreign firms do firms includes relatively little variance; firms com- not exploit Ugandan workers, however; in fact, pensate for the poor investment climate (for ex- they often pay higher wages. The lower unit labor ample, electricity users facing high power costs costs result from higher productivity, presum- purchase generators) and select activities that ably due to higher-quality capital equipment and minimize its adverse impact (for example, avoid- workers. ing continuous process manufacturing); and the 3. University education may be correlated with characteristics of the investment climate may well greater ability, and the coefficient on this variable be embodied in (and correlated with) the attri- may consequently be biased upward. butes of labor and capital. 3. The Investment Climate 36 Perceived Constraints in the Business Environment 37 Effective Protection of Manufacturing 45 How Has the Business Environment Changed? 46 3. The Investment Climate 37 Uganda's continued growth over the past decade has qualitative rankings are difficult to compare across ranked the country among the success stories in eco- countries, given probable differences in firms' experi- nomic development. But the government's ambition is ence and expectations. to achieve investment, export, and economic growth performance similar to that of the East Asian "tigers" before their financial crisis of the late 1990s. To inform Perceived Constraints in the Business its development strategy, it seeks to understand what Environment factors continue to constrain investment and growth in the private sector. Manufacturing firms in Uganda largely consider fi- This chapter draws on data from the 2002/03 firm- nancing and tax obstacles the greatest constraints to level survey as well as other sources to identify the their operation and growth (table 3.1). Finance ranks most important constraints in the investment climate. high among hindrances in the business environment, On the whole, the available evidence supports the whether it is the cost of finance (high interest rates) or finding that Uganda has a difficult investment climate, access to finance (difficult collateral requirements). with significant improvements needed to allow the Some 45 percent of firms reported that access to fi- private sector to flourish. Yet a comparison of the nance is a major or severe constraint. 2002/03 survey data with results from a similar survey The top constraints perceived by foreign firms in 1998 provides hopeful news: Uganda's business (those with 10 percent or more foreign ownership) di- environment has improved, and the gains are already verge from those perceived by domestic firms. having positive effects on firms, particularly the larger Foreign firms identify corruption as their second ones. Large firms (with 100 or more employees) are biggest constraint after macroeconomic instability more likely to invest and export today than they were and feel more constrained by regulatory policy uncer- in 1998. They are also operating at higher levels of ef- tainty, customs and trade regulations, and crime than ficiency. Continued improvements in the business en- do domestic firms. Similarly, exporters (firms export- vironment, particularly a better electricity supply and ing 10 percent or more of their annual sales) and lower interest rates, would further boost efficiency, en- nonexporters have differing perceptions of macroeco- abling Ugandan firms exposed to international com- nomic instability (inflation, exchange rates), corrup- petition to compete with their foreign counterparts. tion, monopolistic behavior of competitors, crime, and The analysis of the investment climate begins with transport.1 Both foreign-owned firms and exporters a ranking of constraints in the business environment reported being less constrained by lack of access to as perceived by private manufacturing firms in finance than domestic firms. But on almost half the is- Uganda. The manufacturing sector's perceptions of sues in table 3.1, foreign-owned firms and exporting the investment climate are not meant to be a definitive firms reported being significantly more adversely af- tool for setting priorities. But these subjective rankings fected than their counterparts. of the importance of different factors in the business When the qualitative rankings are disaggregated by environment can make a useful contribution to the dis- firm size class, it becomes clear that the cost of finance course. And quantitative data, such as the number of is a constraint felt across the board (table 3.2). Not electricity outages or the amount paid in bribes, can surprisingly, access to finance becomes less restrictive add significant weight to arguments about which fac- as firm size increases, while the effects of corruption are tors are most critical. While perceptions of constraints felt more. Microenterprises perceive corruption as a less are shown for other countries as well as Uganda, important constraint--standably, since their low profiles 3. The Investment Climate 38 Table 3.1 Manufacturing Firms' Evaluation of General Constraints to Operation, Uganda (percentage of respondents evaluating constraint as major or very severe) Full Foreign Domestic Non- Constraint sample firms firms Exporters Exporters Cost of finance (interest rates) 60.3 54.1 62.0 62.5 60.2 Tax rates 48.3 43.3 49.6 48.9 48.4 Macroeconomic instability 45.4 57.6 41.3 64.3 41.7 Access to finance (collateral requirements) 45.0 36.5 47.7 37.2 46.6 Electricity 44.5 48.5 43.1 52.4 42.9 Corruption 38.2 55.0 33.3 56.4 35.0 Tax administration 36.1 42.2 34.5 42.9 35.1 Anticompetitive or informal practices 31.1 34.4 30.2 41.5 29.4 Skills and education of available workers 30.8 25.4 32.0 36.6 30.0 Regulatory policy uncertainty 27.6 38.1 23.7 42.9 24.6 Customs and trade regulations 27.4 38.1 23.2 33.3 26.3 Crime, theft, and disorder 26.9 37.3 23.5 36.4 25.3 Transport 22.9 28.8 20.9 36.4 20.2 Access to land 17.4 24.6 15.6 17.1 17.4 Labor regulations 10.8 12.3 10.4 14.6 10.1 Business licensing and operating permits 10.1 13.4 9.2 8.9 10.4 Telecommunications 5.2 6.2 4.9 7.0 4.5 Note: Results showing differences of more than 10 percentage points between foreign and domestic firms, or between exporters and nonexporters, are highlighted. Source: World Bank, Investment Climate survey, Uganda, 2002/03. and small revenues mean that they are less likely to be do Ugandan firms rate such issues? Large firms seem targets. There is convergence among size classes on to have more confidence in the judiciary than small the perception that telecommunications, licensing, labor firms, making greater use of the courts in settling pay- regulations, and access to land limit firm growth less ment disputes (table 3.4). Small, domestic, and non- than the other factors. exporting firms are more likely than large, foreign, and An international comparison shows that issues of exporting firms to perceive inconsistency in the inter- concern in Uganda--cost of finance, tax rates, macro- pretation of regulations. economic instability, and tax administration--are also Firms' perceptions of predictability in the deci- issues of concern in its East African neighbors (table sions by outside bodies shape their assessments of 3.3). Interestingly, a much smaller share of firms in certainty in the economy in general. Thus another way China perceive these issues as important constraints. to evaluate these issues is to look at firms' perceptions of their ability to make investments profitable, as re- Uncertainty flected in the share of profits they reinvest. In Uganda Unpredictability in the decisions made by government this share varies little among different types of firms, bodies, such as the judiciary or regulatory agencies, averaging around 42 percent. Of course, this appar- can create an unstable business environment. How ent similarity may reflect differences in the factors de- 3. The Investment Climate 39 Table 3.2 Manufacturing Firms' Evaluation of General Constraints to Operation by Size Class, Uganda (percentage of respondents evaluating constraint as major or very severe) Medium- All firm Micro Small size Large size (<10 (10­49 (50­99 (100+ Constraint classes employees) employees) employees) employees) Cost of finance (interest rates) 60.3 61.5 59.6 60.0 59.4 Tax rates 48.3 53.3 47.6 33.3 47.2 Macroeconomic instability 45.4 35.7 49.6 48.0 57.1 Access to finance (collateral requirements) 45.0 51.9 43.3 30.8 40.6 Electricity 44.5 45.7 36.3 69.2 51.4 Corruption 38.2 24.8 46.3 56.5 38.7 Tax administration 36.1 52.3 30.3 48.0 37.1 Anticompetitive or informal practices 31.1 22.1 32.8 43.5 41.7 Skills and education of available workers 30.8 32.1 27.2 41.7 32.4 Regulatory policy uncertainty 27.6 25.6 24.8 37.5 34.3 Customs and trade regulations 27.4 27.4 26.3 24.0 33.3 Crime, theft, and disorder 26.9 22.9 26.8 38.5 30.6 Transport 22.9 20.2 18.1 48.0 30.6 Access to land 17.4 14.8 20.2 13.0 19.2 Labor regulations 10.8 14.3 5.6 12.0 17.1 Business licensing and operating permits 10.1 11.1 8.7 15.4 8.3 Telecommunications 5.2 4.7 3.3 12.0 8.3 Note: The top five complaints in each category of firms are highlighted. Source: World Bank, Investment Climate survey, Uganda, 2002/03. termining how much a company reinvests, such as to rank among the most corrupt countries in the working capital or credit available. international indices. The Second National Integrity Survey . . . carried with it a clear Corruption message: large-scale corruption and embezzle- Corruption was highlighted as a serious problem in ment at the top, which is carried out with Uganda in a survey of manufacturing firms conducted impunity, has worked to encourage the by the World Bank in 1998. It continues to be an im- proliferation of administrative corruption at the portant source of concern. A participant in the April grassroots. (Consultative Group Meeting 2003) 2003 Consultative Group meeting of donors in Kampala highlighted the need for major reform: The Consultative Group urged the passing of anti- corruption legislation, public sector pay reform, ade- Notwithstanding the Government's many laud- quate resources for anticorruption agencies, proper able policy and technical achievements, there follow-up on the findings of commissions of inquiry, is a widely held perception in Ugandan society and actions to combat a culture of impunity, that corruption is pervasive, institutionalized and strengthen accountability at all levels, and address on the increase. In addition, Uganda continues corruption at lower levels of government. 3. The Investment Climate 40 Table 3.3 Manufacturing Firms' Evaluation of General Constraints to Operation, Selected Developing Countries (percentage of respondents evaluating constraint as major or very severe) Constraint Uganda Kenya Zambia Tanzania China Turkey Cost of finance (interest rates) 60.3 73.3 82.1 56.2 21.6 28.2 Tax rates 48.3 68.2 57.5 72.1 34.1 38.1 Macroeconomic instability 45.4 51.3 73.9 42.0 26.0 53.7 Electricity 44.5 48.1 39.6 57.6 28.1 17.3 Corruption 38.2 73.8 44.6 50.0 22.4 23.7 Tax administration 36.1 50.9 27.5 54.7 23.7 33.1 Anticompetitive or informal practices 31.1 65.3 38.7 23.9 17.6 22.7 Skills and education of available workers 30.8 27.6 35.8 24.6 26.7 12.8 Regulatory policy uncertainty 27.6 51.5 57.0 30.8 28.0 53.8 Customs and trade regulations 27.4 39.9 32.4 30.8 21.1 8.9 Transport 22.9 37.4 30.4 22.5 19.4 8.4 Access to land 17.4 24.6 17.4 24.3 16.3 6.0 Labor regulations 10.8 22.5 16.9 11.9 19.4 8.7 Telecommunications 5.2 44.1 32.9 11.6 16.5 10.9 Source: World Bank, Investment Climate surveys, Uganda, 2002/03, Kenya, 2003, Tanzania, 2003, Zambia, 2003, China 2000, Turkey, 2002. Table 3.4 Manufacturing Firms' Perceptions of Uncertainty in the Business Environment, Uganda (percent) Large Small Full firms (100+ firms (<100 Foreign Domestic Non- Indicator sample employees) employees firms firms Exporters exporters Share of firms disagreeing that interpretations of regulations are consistent and predictable 40.0 33.3 41.0 36.4 41.0 31.1 41.4 Share of profits reinvested in firm 41.9 42.7 41.8 45.5 41.1 39.9 42.0 Share of firms disagreeing that they have confidence in the judiciary 69.9 61.1 71.1 71.2 69.8 70.5 70.0 Share of payment disputes settled by third parties or resolved in court 50.0 53.3 48.7 43.5 53.6 57.1 48.7 Source: World Bank, Investment Climate survey, Uganda, 2002/03. 3. The Investment Climate 41 A related concern is transparency in fiscal policy firm the widely held view that corruption is a serious and public procurement. The International Monetary problem. Not surprisingly, foreign-owned and export- Fund (IMF), in its 2003 Report on the Observance of ing firms seem to be the biggest targets of govern- Standards and Codes in Uganda (2003b), argued that ment officials requesting bribes. Foreign-owned firms while fiscal transparency had improved since the reported forfeiting the largest share of revenue-- 1999 report, many recommendations in the earlier re- almost 4 percent--for "informal payments" to get port had yet to be implemented. The IMF emphasized things done (table 3.5). the need for transparency in budget execution and re- Obtaining an electricity connection is the service porting and strongly recommended limiting the use of most likely to require a bribe, followed by getting a supplementary appropriations by the executive telephone connection. Large companies seem to bear branch. It also called for properly assessing arrears the largest burden; almost a third reported having to and monitoring and controlling contingent liabilities. In pay a bribe to acquire a telephone connection. addition, the Organisation for Economic Co-operation and Development recently argued for strengthening Infrastructure Performance accountability and institutional capacity in public pro- The quality of infrastructure plays a crucial part in firm curement (OECD 2003). productivity. The manufacturing firms surveyed in Uganda has made efforts to reduce corruption, Uganda reported about 39 power outages in the pre- but progress has been slow. Corruption remains per- vious year on average, though large firms reported vasive throughout the government and can cost firms considerably more, at 54 (table 3.6). Firms estimated a great deal in time and money. The survey data con- the resulting production losses in the 4­7 percent Table 3.5 Indicators of Corruption as Reported by Manufacturing Firms, Uganda (percent) Large Small Full firms (100+ firms (<100 Foreign Domestic Non- Indicator sample employees) employees firms firms Exporters exporters Informal payments required as a share of revenues 2.4 1.1 2.6 3.9 1.9 3.0 2.3 Share of firms reporting requirement for gift or payment For a mainline telephone connection 18.3 28.6 16.4 18.4 18.5 18.2 17.6 For an electricity connection 21.5 21.4 21.5 18.2 22.3 24.0 21.2 For a construction permit 12.3 20.0 11.1 10.0 13.2 5.6 14.8 For an import license 3.6 0.0 4.3 5.3 2.8 0.0 4.5 For a trading license 4.2 3.4 4.3 3.8 4.3 2.7 4.5 Share of revenue typically reported for tax purposes 76.7 87.3 75.2 81.3 75.3 86.1 74.7 Source: World Bank, Investment Climate survey, Uganda, 2002/03. 3. The Investment Climate 42 Table 3.6 Infrastructure Performance as Reported by Manufacturing Firms, Uganda Large Small Full firms (100+ firms (<100 Foreign Domestic Non- Indicator sample employees) employees firms firms Exporters exporters Frequency of power outages (times in previous year) 38.6 54.2 36.5 40.3 38.0 38.3 37.7 Share of production lost due to power outages (percent) 6.3 4.5 6.5 7.4 5.9 3.7 6.7 Share of firms with own generator (percent) 35.3 69.4 30.7 67.7 26.0 53.3 31.9 Share of firms that have built own well (percent) 13.0 30.6 10.6 27.9 8.7 24.4 10.6 Days to obtain a telephone connection 33.2 23.1 35.1 17.7 39.3 35.1 32.8 Days to obtain an electricity connection 38.3 39.1 38.3 46.8 36.2 48.7 36.6 Source: World Bank, Investment Climate survey, Uganda, 2002/03. range. Slightly more than a third of all firms own gen- as Ugandan firms to obtain telephone and electricity erators, while about two-thirds of large and foreign connections. firms do. The Ugandan government has undertaken efforts Despite complying with demands for informal pay- to improve the electricity supply. But the proposed ments, firms face long delays in obtaining connec- $550 million Bujagali hydropower project, intended to tions for telephone service (33 days on average) and extend electricity to rural areas and export energy to electricity (38 days). Large and foreign firms are able neighboring countries, was halted in 2002 because of to obtain telephone connections faster than others, investigations of corrupt dealings with the U.S. com- however. But exporting firms wait longer than others pany AES (Financial Times 2002). According to the for an electricity connection. This result is particularly Economist Intelligence Unit, several new investors are troublesome, since exporters are more sensitive to being sought and alternative deals, perhaps less ex- electricity constraints and since the growth of the pensive, are being considered (EIU 2004). Mean- Ugandan economy depends on the growth of exports. while, privatization of the state-owned electricity distri- International comparisons show that while bution and transmission companies has stalled Uganda may do better than Kenya on some indicators recently and investor interest is lower than desired. It of infrastructure performance, it falls far short of the seems that the goal of increasing access to electricity performance in China (table 3.7). Ugandan firms need to10 percent of the rural population (from one percent) wait only about a fourth as many days as Kenyan firms by 2010 will be difficult to meet. to obtain a telephone connection, for example. But while Ugandan firms lose an average 6.3 percent of Regulatory Burden production because of power outages, Chinese firms The private sector in Uganda must cope with a sub- lose only 1.8 percent. The share of firms owning gen- stantial regulatory burden (table 3.8). Large enter- erators in China is only about half that in Uganda. And prises and foreign-owned companies find administra- Chinese firms must wait less than half as many days tive and regulatory problems a greater nuisance than 3. The Investment Climate 43 Table 3.7 Infrastructure Performance as Reported by Manufacturing Firms, Selected Developing Countries Indicator Uganda Kenya Tanzania Zambia China India Frequency of power outages (times in previous year) 38.6 33.1 67.2a 37.2a -- -- Share of production lost due to power outages (percent) 6.3 9.3 9.2 4.5 2.0 -- Share of firms with own generator (percent) 35.3 70.0 55.0 38.2 16.2 68.9 Share of firms that have built own well (percent) 13.0 33.5 34.7 59.9 15.6b 50.8b Days to obtain a telephone connection 33.2 123.7 23.1 132.5 12.0 -- Days to obtain an electricity connection 38.3 65.6 23.1 120.7 19.0 -- -- Not available. a. Frequency of power outages and surges. b. Percentage of firms that own a well. Source: World Bank, Investment Climate survey, Uganda, 2002/03, Kenya, 2003, Tanzania, 2003, Zambia, 2003, China 2000,Turkey, 2002. small and domestic firms. The same can be said of survey by the International Finance Corporation's exporting businesses, since most are also foreign Foreign Investment Advisory Service found that about firms. The senior management of large, foreign, and a third of respondents were generally dissatisfied exporting firms spends slightly more time on average with the quality of regulation (FIAS 2003). Roughly 62 dealing with regulations than does the management of percent were dissatisfied with the bureaucracy. The small or domestic enterprises. These firms also spend most problematic areas of regulation identified by more than twice as much time in inspections and firms were taxation administration, customs, and ac- meetings with government officials and lose almost 10 cess to land. Some agencies are addressing con- times as much money (as a share of sales) in fines cerns by introducing new procedures, but they are or seized goods as a result of these inspections. doing so before the changes are reflected in laws Surprisingly, it also takes large and foreign companies and regulations. The result is varying interpretation by longer to clear exports through customs. Not surpris- officials, confusion in the business community, and ingly, firms exporting 10 percent or more of their sales concern that the procedures could be summarily wait less time than firms exporting less for exported changed. goods to clear customs.2 Backing the perceptions of Uganda-based in- The main regulatory obstacles for small and do- vestors are the views of international rating sources mestic enterprises tend to differ from those for large such as the Wall Street Journal­Heritage Foundation and foreign firms. A larger share of these businesses Index of Economic Freedom, which provide an impor- perceive regulations as being inconsistently or unpre- tant gauge of how well Uganda is doing compared dictably applied, and not in their favor. Interestingly, with other developing countries. Uganda scores small and domestic firms also face a much larger poorly in ratings by international agencies for reasons share of inspections by local authorities--almost three highlighted in this report--bribes, high fees, compli- times as large--as do large or foreign firms. cated licensing procedures, regulations that burden Other results are consistent with the findings of businesses, and significant barriers to opening a busi- the firm survey. An administrative and regulatory cost ness. Reinforcing these perceptions are an outdated 3. The Investment Climate 44 Table 3.8 Regulatory Burden and Administrative Delays as Reported by Manufacturing Firms, Uganda Large Small Full firms (100+ firms (<100 Foreign Domestic Non- Indicator sample employees) employees firms firms Exporters exporters Regulation (percent) Share of firms disagreeing that interpretations of regulations are consistent and predictable 40.0 33.3 41.0 36.4 41.0 31.1 41.4 Share of senior manage- ment's time spent dealing with regulations 0.4 0.07 0.04 0.06 0.04 0.05 0.04 Share of revenues typically paid to officials to get things done 2.4 1.1 2.6 3.9 1.9 3.0 2.3 Share of firm revenues typically reported for tax purposes 76.7 87.3 75.2 81.3 75.3 86.1 74.7 Inspections Days last year spent in inspections or required meetings with officials 13.4 25.6 11.8 26.0 9.7 18.2 12.6 Share of meetings or inspections by local authorities (percent) 19.4 7.6 21.4 8.9 23.0 16.3 20.1 Cost of fines or seized goods (percentage of sales) 0.1 0.3 0.0 0.2 0.0 0.2 0.0 Share of interactions in which informal payment requested (percent) 6.7 9.4 6.3 6.8 6.7 7.3 6.6 Value of informal payments (percentage of sales) 0.3 0.6 0.3 0.4 0.3 0.4 0.3 Import delays (days) Average wait to clear customs 5.8 5.5 5.9 5.6 6.1 5.3 6.0 Longest wait to clear customs 11.2 10.1 11.5 12.5 9.8 10.9 11.3 Export delays (days) Average wait to clear customs 3.5 4.2 3.2 3.5 3.6 3.3 3.7 Longest wait to clear customs 6.0 6.3 5.9 6.5 4.8 5.8 6.2 Source: World Bank, Investment Climate survey, Uganda, 2002/03. 3. The Investment Climate 45 investment code describing incentives that no longer lowered average tariffs from 17.8 percent in 1994 to exist, budget declarations on incentive packages that 10.6 percent in 2002. The average tariff for all imports are not implemented, and nontransparent incentives is now around 9 percent. In 2002 average tariffs in for selected investments. Uganda were well below those in other Sub-Saharan Other evidence comes from the World Bank's African countries. Doing Business database, which shows that the cost Thus on the basis of the tariff schedule Uganda's of starting a business in Uganda is relatively high trade policy regime seems fairly liberal. But tariffs are (World Bank 2003a). Entrepreneurs in Uganda can ex- not the only policy tools affecting the domestic price pect to undertake 17 different steps to set up a busi- of tradable goods. Some nontariff, ad valorem restric- ness, a process that takes 36 days on average and tions are still in effect, including an import license involves a cost equal to 135 percent of gross national regime, a withholding tax, and an excise tax. These income per capita. taxes, especially the excise tax, are nonneutral in their effects on the economy: Effective Protection of Manufacturing · The nontariff protection increases nominal protec- tion well beyond what is suggested by tariffs. Another element of the business environment is the · The additional trade taxes increase the dispersion extent of protection of Ugandan manufacturing firms in nominal protection compared with the tariff- from foreign competition. The survey data show that induced dispersion. Although excise taxes cover while the tariff regime in Uganda is quite liberal, im- only 8.6 percent of all tariff lines, their coverage portant distortions remain in nontariff protection. varies greatly at the category level. These distortions have implications for economic effi- · The nontariff protection allows authorities to pro- ciency and need to be addressed. An analysis sug- tect specific sectors while still complying with in- gests that Uganda needs to persevere in removing ad ternational trade rules and demonstrating a re- hoc excise taxes. It also needs to move its dialogue duction in tariff protection. with the East African Community forward with a view to minimizing distortions. All this translates into a structure of effective pro- In the 1990s Ugandan authorities, viewing foreign tection that is uneven and much higher than the tariff trade as a major engine of economic growth, under- levels might suggest. The estimated rate of effective took reforms aimed at fostering trade (see chapter 1 protection varies widely across manufacturing sub- and appendix 3). In addition, Uganda has promoted sectors, ranging from roughly 28 percent in the wood trade relations with other developing countries subsector to almost 80 percent in the textile and through regional integration agreements. In (East leather products subsector (see appendix table A3.5). Africa a proposed customs union would entail a com- The dispersion of effective rates of protection is also mon external tariff. A World Bank study (2003) sug- large within subsectors. These patterns suggest that gests that the common tariff that has been agreed on the structure of protection is nonneutral in its impact. implies a decline in tariff rates for Kenya and Tanzania Indeed, firm-level measures of nominal protection and an increase for Uganda.3 show that raw materials tend to be less protected than As part of the trade reforms, Uganda has drasti- final products, that differences exist between domes- cally reduced tariff protection over the past decade tic and foreign firms, and that distortions in protection (see appendix table A3.1). For manufactured goods it levels are significant. 3. The Investment Climate 46 So the picture of manufacturing protection in Table 3.9 Samples in 1998 and 2002/03 Uganda is mixed. Tariff protection has declined, but Surveys by Size and Location of Firms, the true level of protection remains quite high and dis- Uganda tortions persist. Most of the protection is due to the (percent) use of excise taxes. An important consequence of the 1998 2002/03 structure of protection is to bias sectoral incentives, Firm size class (employees) which is detrimental to the efficient development of the Micro (<10) 17.7 18.0 manufacturing sector. (See appendix 3 for a detailed Small (10­49) 39.9 51.0 discussion of the protection of manufacturing in Medium-size (50­99) 13.6 11.3 Uganda.) Large (100+) 28.8 19.7 Location How Has the Business Environment Central region 64.0 75.0 Changed? Northeast region 25.0 12.5 Southwest region 11.0 12.5 In 1998 the World Bank conducted a survey of Source: World Bank, RPED survey, Uganda, 1998, and Ugandan manufacturing firms quite similar the one Investment Climate survey 2002/03. undertaken in 2002/03, allowing a comparison over time of firms' perceptions of constraints to their in- northeast region: 25 percent were in the Jinja and vestment, operation, and growth. How has the busi- Mbale-Tororo areas, compared with 12.5 percent in ness environment changed, and what has been the 2002/03. The sectors represented in the two samples impact on investment and growth? differ somewhat, limiting comparability to some Overall, the business environment has improved extent. since 1998, especially in aspects relating to regulation and infrastructure. The improvements have benefited Constraints in the Business Environment firms' performance: investment rates have risen, ex- In 1998, as in 2002/03, Ugandan firms were asked to ports are growing, and firms, especially the largest rank key constraints to their operation and growth. ones, are operating more efficiently. Nonetheless, as Ranking highest among their concerns were the price this chapter has shown, the business climate in and quality of utility services (such as water, tele- Uganda remains much harsher than those in rapidly phone, and electricity services), high taxes, and inter- advancing countries like China. est rates (figure 3.1). Government corruption, access The samples for the two World Bank surveys are to finance, tax administration, and the cost of raw ma- roughly comparable in the size distribution of firms, al- terials and supplies formed a second tier of con- though the 1998 sample contains slightly fewer small straints. Rounding out the group of at least moderate firms and proportionally more large enterprises (table constraints were local competition, lack of demand, 3.9). But the average size of firms is almost identical: lack of business support services, crime and security, in 1998 the sampled firms had 123 employees on av- lack of skilled labor, uncertainty about government erage, while in 2002/03 they had 130 on average. The policies, inflation, and exchange rate issues. vast majority of firms in both samples were selected A comparison of the results from 1998 and from the Kampala-Entebbe area in the central region. 2002/03 shows that constraints generally lessened In 1998 firms were slightly oversampled from the during the period. Firms still perceive problems with 3. The Investment Climate 47 Figure 3.1 Ranking of Constraints Perceived by Manufacturiung Firms, Uganda, 1998 Constraints to Uganda Firms, 1998 High Utility Prices High Taxes Poor Utilities (electric, phones...) Interest Rates Corruption Access to Finance Tax Administration Cost of Raw Materials/Supplies Competition from Local Firms Insufficient Demand Lack Business Support Services Crime/Security Lack of Skilled Labor Uncertainty about Gov't policies Inflation Exchange Rate: level/fluctuations Government's debt burden Access to raw materials/supplies Competition from Imports Other Regulations Access to Land Unclear Property Rights Import/Export Regulations Political Instability 1.00 1.50 2.00 2.50 3.00 3.50 4.00 (1 = No Obstacle, 2 = Minor, 3 = Moderate, 4 = Major, 5 = Very severe) Note: Constraints were rated on a scale of 1 (no obstacle) to 5 (very severe) Source: World Bank, RPED survey, Uganda, 1998. high tax rates and the cost of finance, but the average vices in Uganda have already improved significantly score has fallen somewhat for such factors as corrup- in recent years, especially with the introduction of mo- tion, access to finance, and tax administration. A ma- bile phones. jority of firms no longer consider access to land, busi- ness licensing, and telecommunications to be Infrastructure problems. While the survey data show tremendous improve- Questions about infrastructure differed between ments in the quality of electricity service and the use the two surveys, so not all responses are directly com- of mobile telephony, firms were more likely to report parable. But some conclusions are nonetheless clear. water and fixed line telephone services as major or Although power supply has improved over the years, severe constraints to operation in 2002/03 than in its reliability and adequacy remain the leading infra- 1998. And while there have been improvements in the structure constraint for Ugandan enterprises. Current public provision of waste disposal services, slightly deregulation efforts should improve electricity supply more firms considered the quality of these services to significantly in the coming years, doing much to im- be a major or severe constraint in 2002/03 than in prove firm-level productivity. Telecommunications ser- 1998. 3. The Investment Climate 48 Electricity. Manufacturing firms saw big improve- 2002/03. Again, the gains were biggest for micro firms ments in the number of days of production lost be- in agro-industry: among these firms the share cause of power outages between 1998 and 2002/03 dropped from 88 percent to only 22 percent. In other (figure 3.2). While firms found themselves without manufacturing the share declined from 56 percent to power for 84 days on average in 1998, they reported 34 percent. 39 outages in 2002/03. The share of firms owning gen- Electricity connection times have also improved. erators consequently fell, from 41 percent in 1998 to While firms had to wait 12 weeks on average for an 35.3 percent in 2002/03. electricity connection in 1998, they had to wait only 38 Particularly dramatic were the improvements for days in 2002/03. All these improvements suggest that micro firms in agro-industry. While in 1998 such firms recent policy initiatives to boost the quality of electric- reported a staggering average of 130 days of pro- ity service have been successful. duction lost to electricity outages, in 2002/03 they re- ported a loss of just 35 days. While the improvements Telecommunications. Cellular phone service in were greatest for agro-industry, there were also per- Uganda has grown tremendously since the first cellu- ceptible gains for other manufacturing subsectors, lar operating license was issued in 1997. The share of particularly among small firms. firms reporting use of mobile phones rose from 41 per- The share of firms reporting electricity as a major cent in agro-industry and 50 percent in other manu- or severe problem declined sharply. Agro-industry facturing in 1998 to 93 percent in each sector in firms reported the largest improvements. While 78 2002/03. The largest increases in mobile phone use percent of these firms cited electricity as a major or occurred among micro and small firms (figure 3.3). In severe problem in 1998, only 22 percent did in 1998 no microenterprises reported using mobile Figure 3.2 Average Number of Days Lost Due to Power Outage by Firm Size Class, Uganda 140.0 129.9 120.0 114.8 101.7 100.0 88.2 80.0 76.8 69.1 66.5 Days 59.5 60.0 55.4 42.4 40.9 40.0 34.8 36.4 36.8 31.9 24.9 20.0 0.0 Micro (1­9) Small (10­49) Medium (50­99) Large (100+) 1998Ag 2002Ag 1998Om 2002Om Source: World Bank, RPED survey, Uganda, 1998, and Investment Climate survey 2002/03. 3. The Investment Climate 49 Figure 3.3 Share of Firms Using Mobile Phones for Business by Firm Size, Uganda 120.0 100.0 100.0 100.0 96.3 97.0 91.7 92.3 82.6 80.0 78.3 80.0 77.4 60.0 50.0 40.0 37.0 31.6 33.3 20.0 0.0 0.0 0.0 Micro<10 Small 10-49 Medium 50-99 Large 100+ Mobile-1998Ag Mobile-2002Ag Mobile-1998Om Mobile-2002Om Source: World Bank, RPED survey, Uganda, 1998, and Investment Climate survey 2002/03. phones, but in 2002/03 nearly three-quarters did. tion technology and call center services will be con- Among large firms almost 100 percent reported using strained by the quality of fixed line telephony sug- mobile phones in 2002/03. gested by the survey data. While mobile telephony appears to have brought significant benefits to enterprises, fixed line telephony Water. One measure of the quality of water service is remains a constraint to firm operation. One indicator is the number of days in the previous year firms report the wait to obtain a telephone connection. For manu- having an inadequate water supply. This measure im- facturing firms as a group, the average wait fell from proved between 1998 and 2002/03, falling from an av- 11 weeks in 1998 to 33 days in 2002/03. Because of erage 31 days to 6.1 days. The median number of substitute mobile phone service, however, most firms days with insufficient water also fell, from four days to do not consider this wait an important constraint. zero. A larger share of firms in agro-industry reported Nonetheless, more firms reported water service as problems with fixed line telephony as major or severe. a major or severe constraint to doing business in Less than 5 percent of firms reported major or severe 2002/03 than in 1998 (figure 3.4). This result suggests problems with cellular phone service, underlining the that while water service has become more reliable, difference in the quality of service by fixed line phone more needs to be done to improve its quality. providers. The quality differences between fixed line and mo- Waste Disposal. The share of firms providing their bile service are a particular concern, since fixed line own waste disposal service gives a good indica- telephony is likely to be cheaper in a small economy tion of the quality of the service available. This share such as Uganda. Moreover, recent policy initiatives to fell in each size class and sector between 1998 establish Uganda as a potential supplier of informa- and 2002/03 (figure 3.5). Firms in agro-industry 3. The Investment Climate 50 Figure 3.4 Share of Firms Reporting Water Services as Major or Severe Problem by Firm Size, Uganda 70.0 61.3 60.0 52.2 50.0 40.9 40.0 38.9 36.4 33.3 33.3 34.6 32.0 Percent 30.0 26.3 25.0 25.0 26.1 20.0 16.7 17.4 10.0 0.0 0.0 Micro (1-9) Small (10-49) Medium (50-99) Large (100+) 1998Ag 2002Ag 1998Om 2002Om Source: World Bank, RPED survey, Uganda, 1998, and Investment Climate survey 2002/03. Figure 3.5 Share of Firms Doing Own Waste Disposal by Firm Size, Uganda 120.0 100.0 100.0 94.7 96.0 87.5 83.3 82.6 80.0 78.3 60.6 60.0 57.7 52.2 52.5 54.5 Percent 50.0 45.2 42.6 40.0 33.3 20.0 0.0 Micro<10 Small 10-49 Medium 50-99 Large 100+ 1998Ag 2002Ag 1998Om 2002Om Source: World Bank, RPED survey, Uganda, 1998, and Investment Climate survey 2002/03. 3. The Investment Climate 51 experienced the largest improvement, with the share countries like the United States firms perceive taxes providing their own waste disposal falling from more as the leading constraint. than 85 percent in 1998 to less than 50 percent in By contrast, the regulatory burden imposed on 2002/03. That this sector should be most affected is Ugandan firms appears to have lessened signifi- unsurprising, since it is more sensitive than others to cantly. There are several ways to measure this burden, poor waste disposal. but among the most important is how much time en- Despite the improvements, firms remain dissatis- trepreneurs and senior managers must spend in deal- fied with the quality of the waste disposal service ing with regulation--time taken away from running available. The share of firms reporting major or severe their firm's core business. The 1998 survey found that problems with waste disposal service rose slightly be- senior managers devoted 14 percent of their time to tween 1998 and 2002/03, suggesting that while the regulatory compliance (such activities as filling out quantity of service has increased, the quality still falls paperwork and ensuring compliance with regulations short of firms' expectations (figure 3.6). relating to waste disposal, product standards, worker benefits, environmental safety, occupational safety, Taxes, Regulation, and Administrative and the like). This average was heavily influenced by Corruption a few firms (9 percent of respondents) whose man- Ugandan firms identified high taxes as the second agers reported spending more than 40 percent of biggest constraint to doing business in 2002/03, just their time on regulatory compliance. For the other as they had in 1998. But firms throughout the world firms the average was 10 percent. By 2002/03 the av- dislike taxes, and even in relatively low-tax OECD erage had fallen sharply to 4 percent. The managers Figure 3.6 Share of Firms Reporting Waste Disposal as Major or Severe Problem by Firm Size, Uganda 70.0 61.3 60.0 52.2 50.0 40.9 40.0 38.9 36.4 33.3 33.3 34.6 32.0 Percent30.0 26.3 25.0 25.0 26.1 20.0 16.7 17.4 10.0 0.0 0.0 Micro (1-9) Small (10-49) Medium (50-99) Large (100+) 1998Ag 2002Ag 1998Om 2002Om Source: World Bank, RPED survey, Uganda, 1998, and Investment Climate survey 2002/03. 3. The Investment Climate 52 of larger firms reported spending more of their time Firm Performance dealing with regulations (7 percent) than those of The comparison of the 1998 and 2002/03 survey data small firms (3 percent). shows that Uganda has seen an improvement in its Survey data suggest that corruption also may business environment.4 Much remains to be done: have lessened. The 1994 RPED manufacturing firm firms in Uganda still face a much more difficult busi- survey that asked firms how often corruption occurred ness environment than do firms in East Asia. But the among firms in their line of work, 40 percent acknowl- positive changes between 1998 and 2002/03 are al- edged occasional bribery and 28 percent reported ready affecting firms' performance, as reflected in that bribery occurred often or always, while 31 per- their investment patterns, exports, growth rates, and cent reported no bribery. In the 2002/03 survey 20 unit labor costs. percent of firms reported paying some bribes. This share differs significantly across size classes: while Investment. More firms in Uganda reported making only 12 percent of small firms reported paying bribes, recent investments in 2002/03 than in 1998. In 1998 28 percent of larger firms did. In both 1998 and about 55 percent of firms reported making any invest- 2002/03 firms reported most often paying bribes to ment in the previous three years, while in 2002/03, 63 speed bureaucratic processes (such as license and percent did. Moreover, the average ratio of investment customs approvals); to obtain government services to capital stock rose from 13.3 percent in 1998 to 15.6 (such as electricity or telephone); to reduce taxes, percent in 2002/03. fees, or fines; or to be considered for or to win gov- In the smallest size classes almost 50 percent of ernment contracts. firms reported making recent investments in 2002/03. The largest firms were the most likely to invest, how- Access to Finance ever (table 3.10). For the smaller firms the ratio of in- In 2002/03, as in 1998, the top financial concern of vestment to capital stock remained much the same be- Ugandan firms was the cost of bank finance. Access tween 1998 and 2002/03. But for large firms this ratio to finance is also a large problem, and microenter- jumped from 0.13 to 0.25, showing that these firms prises consider it one of the top five constraints along have been making bigger investments. This growth in with interest rates. In Uganda interest rates have fallen investment is corroborated by capital vintage: as since 1998, but firms surveyed in 2002/03 still com- noted elsewhere in this report, Ugandan firms have the plained that the cost of finance is high. youngest capital stock in Sub-Saharan Africa. Table 3.10 Investment Rates of Manufacturing Firms, Uganda Share of firms investing in previous three years Ratio of investment (percent) to capital stock Firm size class (employees) 1998 2002/03 1998 2002/03 Micro (<10) 44.2 44.4 0.18 0.16 Small (10­49) 49.5 50.9 0.14 0.14 Medium-size (50­99) 57.6 55.8 0.08 0.13 Large (100+) 62.9 67.8 0.13 0.25 Source: World Bank, RPED survey, Uganda, 1998, and RPED investment climate survey 2002/03. 3. The Investment Climate 53 Exports. The share of firms exporting remained al- first period and 4 percent in the second (table 3.11). most the same over the period, at 14 percent. The The picture is also similar across size classes, except share of total manufacturing output exported was also for the largest firms. These firms grew faster in similar, at around 10 percent. But differences emerge 1995­97 than in 2000­02. But the standard deviation across firm size classes. Larger firms reported ex- is much smaller for the later period, indicating greater porting a significantly larger share of their output in stability in employment in 2000­02. 2002/03 (47.5 percent) than in 1998 (31.4 percent). Data for casual employees are missing for 2000, This change indicates the growing competitiveness so total employment cannot be compared. But a quick and efficiency of large firms in Uganda compared with comparison of casual employees as a share of the their international counterparts. Smaller firms, operat- total reveals a surprising fact: in both periods the av- ing in segmented local markets, are more shielded erage share is 29 percent, a large share compared from international competition and cater mostly to with that in other countries. This share is even larger local customers. for medium-size and large firms and does not differ significantly between the two periods. The incentive to Employment Growth. Both surveys asked firms hire casual workers is usually their lower cost. In coun- about their current employment and their employment tries like Uganda, where the HIV/AIDS crisis is severe, in the previous three years. While the intervals are rel- hiring casual workers might be a way for firms to avoid atively short, the data can be used to observe the high health benefit costs (for a more detailed discus- growth of existing firms in each period. For the full sion of this issue, see chapter 4). sample the growth rates are quite close, with perma- nent full-time employment growing at 5 percent in the Unit Labor Costs. Unit labor costs increased be- tween 1998 and 2002/03 (table 3.12). But the differ- ence is greatest for the smallest firms, where the ratio Table 3.11 Average Growth in Permanent of wages to value added rose dramatically.5 For the Full-Time Employment in Manufacturing largest firms unit labor costs declined sharply, be- Firms over Three-Year Intervals, Uganda coming increasingly competitive with those of firms in Firm size class China and India. (employees) 1995­97 2000­02 Micro (<10) 0.00 ­0.01 (0.23) (0.19) Table 3.12 Unit Labor Costs in Small (10­49) 0.04 0.04 Manufacturing, Uganda (0.22) (0.19) (average ratio of wages to value added) Medium-size (50­99) 0.09 0.12 Firm size class (0.19) (0.25) (employees) 1998 2000­03 Large (100+) 0.12 0.04 Micro (<10) 0.39 0.69 (0.26) (0.16) Small (10­49) 0.40 0.57 All firm size classes 0.05 0.04 Medium-size (50­99) 0.51 0.59 (0.24) (0.20) Large (100+) 0.48 0.30 All firm size classes 0.43 0.56 Note: Figures in parentheses are standard deviations. Source: World Bank, RPED survey, Uganda, 1998, and Source: World Bank, RPED survey, Uganda, 1998, and RPED investment climate survey 2002/03. RPED investment climate survey 2002/03. 3. The Investment Climate 54 Notes Uganda, it also suggests that this loss can be off- set by revamping the customs administration to 1. There is some overlap between the categories of increase efficiency and reduce leakage. firms, which may explain some of the similar re- 4. This section draws on a note by Andrew Stone sponses. Among exporting firms, 56.8 percent are summarizing the key results from the 1998 survey foreign owned and 43.2 percent are domestic. (Stone, 1998). 2. Border crossings in landlocked Uganda are also 5. This result should be interpreted with caution, slow. Exports take 15.4 days on average, and up since it may be driven by differences in the sam- to 24.2 days, to cross the border. While few com- ples between the two periods, differences in data parative data are available, these delays probably collection, or even differences in accounting prac- also impose a burden on firms trying to export tices between small and large firms. their goods. 3. While the World Bank study (2003) predicts that the tariff changes will lead to a revenue loss for 4. Factor Markets: Finance and Labor 55 The Financial Market 56 Sources of Finance 56 Developments in the Banking Sector 57 Access to Bank Finance 57 Cost of Finance 61 Are Firms Credit Constrained? 62 Trade Credit 63 A Comparison with Kenya 64 Policy Issues 66 The Labor Market 67 Health Status 68 Remuneration and Determinants of Wages 71 Institutional Rigidities 75 Policy Issues 78 4. Factor Markets: Finance and Labor 56 Factor Markets as they are discussed in this chapter ations of deposit-taking institutions, the performance refer to financial and labor factor markets. Constraints of insurance and contractual savings institutions, to the efficient functioning of these markets impose money management by the Bank of Uganda, clear- hardships on firms including those related to legal, ance processes for checks, and the provision of sup- regulatory or institutional impediments. port for capital market development. The following section examines the survey results on the sample firms' access to bank finance, their use The Financial Market of trade credit, and the costs, correlates, and require- ments for each of these sources of external finance Firms rely on a range of sources to finance working points. The recommendations that emerge from this capital and new investment needs. In a world of per- analysis is very much in keeping with those of the ear- fect capital markets the source of finance would be ir- lier assessments of the Ugandan financial sector. relevant to firms' financing decisions. But since mar- Together, the ICA firm survey results and past assess- kets are imperfect, chiefly because of information ments of the Ugandan financial sector point to the asymmetries, the cost of finance differs across recognition that much progress has been made but a sources. great deal remains to be done. Most important, the This section looks at what determines the cost and government needs to address the risks posed by availability of the two most important sources of exter- small banks, implement anti­money laundering legis- nal finance for manufacturing firms in Uganda: bank lation and a credible monitoring system to enforce it, finance and trade credit. These determinants can be phase out direct government involvement in microfi- divided into two broad categories: supply and de- nance, and continue to improve banking supervision. mand. On the supply side are such macroeconomic factors as overall financial depth, macroeconomic stability, fiscal discipline, the capacity to manage Sources of Finance shocks, and the effectiveness of the legal system. Factors on the demand side include the quality of fea- The 2002/03 survey data for manufacturing firms in sible projects, the ability to produce credible informa- Uganda show that their demand for external financing tion relevant to the lending decision, and ownership of is determined by the extent to which they can meet collateral. Policy interventions to improve the opera- their working capital and investment needs through in- tion of the financial sector will require closely examin- ternal resources (retained earnings) and by the cost of ing all these. But here the focus is on demand-side external financing. Internally generated funds are the policy recommendations. cheapest source of funding for any firm, and Ugandan In assessments of the Ugandan financial sector firms show a clear preference for relying on internal by the World Bank (1999) and the International resources. On average, firms in the sample rely on re- Monetary Fund (2003a), the major policy recommen- tained earnings to finance about 80 percent of their dations focus on boosting savings mobilization, in- working capital needs and 71 percent of their new in- creasing access of the rural poor to financial services, vestment needs (table 4.1). expanding the availability of term finance, as well as After internal funds, banks are the most important improving the implementation following activities: the source of finance for manufacturing firms in Uganda. pension system, the insurance sector, the enforce- Firms use commercial bank financing to meet 7 per- ment of financial contracts and supervision, the oper- cent of their working capital needs and 13.5 percent 4. Factor Markets: Finance and Labor 57 Uganda started in the early 1990s. The financial sec- Table 4.1 Sources of Finance for Manufacturing Firms' Working Capital and tor reform has focused on lifting interest rate ceilings, Investment Needs, Uganda liberalizing the foreign exchange market, and intro- (percent) ducing higher capital requirements for financial insti- Working tutions. New operating licenses were issued, with the Source of finance Capital Investmentb expectation that they would encourage competition in the banking sector and eventually lead to lower bor- Retained earnings 79.95 71.06 rowing costs and broader savings mobilization. By Local commercial banks 5.65 11.64 1995 the number of private domestic banks in opera- Foreign-owned commercial tion had more than doubled from prereform levels. In banks 1.32 1.83 addition, the government privatized the largest com- Leasing arrangements 0.09 2.36 mercial bank in 2002 after an initial attempt in 1997.1 Development financea 1.45 2.20 Uganda suffered its first banking crisis in the post Trade credit 5.31 0.48 reform era in 1998­99. Four banks, three domestic Credit cards 0.00 0.00 and one foreign, were closed between September Equity, sale of stock 1.81 1.95 1998 and May 1999 as a result of imprudent banking Family and friends 1.35 2.02 practices. In addition, a bank and a nonbank financial Informal sources 0.36 1.46 institution were closed in 2002. Earlier, a small bank Other 2.71 4.51 had been closed in 1993, followed by the restructur- Total 100.00 100.00 ing of two banks in 1995. a. Includes all financing from the Uganda Development The closure of weak banks strengthened the Bank and East African Development Bank and donor re- sources managed by the Bank of Uganda's Development banking system. But the costs of the banking crisis, Finance Department. along with those of defense and macroeconomic b. The sample used to calculate the figure for new invest- management, have had clear implications for lending ment includes only firms with positive investments in 2002. activity in the postcrisis period. Government borrow- These firms represent just over half of all surveyed firms (52 percent). ing from commercial banks has overtaken private sec- Source: World Bank, Investment Climate survey, Uganda, tor borrowing--a classic example of government bor- 2002/03. rowing crowding out the private sector (figure 4.1). With commercial banks the predominant source of of their new investment needs. Trade credit is the sec- external finance, the shift in their asset portfolio away ond most important external source of working capital from credit to the private sector is likely to have signif- finance, accounting for 5.3 percent. Leasing finance, icant implications for the performance of firms. development finance, sale of stock, and borrowing Businesses that are riskier or less profitable are likely from family and friends each finance about 2 percent to be crowded out of the market for external finance, of new investments. and small and medium-size firms will be affected most. Developments in the Banking Sector Access to Bank Finance Reform of the commercial banking sector has been an Only a third of the sample firms have access to bank important pillar of the economic reform program that finance--that is, have either a currently active loan or 4. Factor Markets: Finance and Labor 58 Figure 4.1 Commercial Bank Activity, Uganda, 1997­2003 900 800 700 shillings 600 500 Uganda 400 of 300 200 Billions 100 0 1997 1998 1999 2000 2001 2002 2003 Government securities Credit to private sector Source: Bank of Uganda 2003. access to overdraft facilities (table 4.2). The size of contrast, among firms that do not produce externally firms is strongly associated with bank financing. The audited accounts, only 11 percent have access to share of firms reporting access to bank finance bank finance. steadily increases with size: while only 7.4 percent of The age of firms also matters. Older firms are microenterprises have access to bank finance, 73 more likely than younger ones to obtain bank finance. percent of large firms do. This finding is evidence of the importance of relation- Foreign-owned firms (those with 50 percent or ships in credit markets characterized by information more equity owned by non-Ugandan nationals or enti- asymmetries, costly monitoring, and weak enforce- ties) are more likely than domestically owned firms to ment of property rights. Longer relationships with a have access to bank finance. Around 64 percent of bank allow firms to use their track record to obtain these firms have access to bank finance, compared finance. with only 24 percent of domestically owned firms. One reason for the difference may be size: foreign-owned Determinants of Access to Bank Finance firms tend to be larger than domestically owned To determine the importance of each of the correlates ones.2 But it is impossible to determine whether the of access to bank finance, a probit model is run, con- disparity in access to finance captures differences be- trolling for firm size and performance, firm age, use of yond size. Foreign ownership may bring with it a cul- external auditors, foreign ownership, and subsector ture of keeping accurate accounting information that and firm location. Since current firm size and perform- facilitates access to bank finance. ance are likely to be endogenous to access to fi- The credibility of information produced by firms is nance, start-up size and firm growth in 2000­01 are strongly associated with access to bank finance. used as instruments. Among the firms that have their accounts audited by The multivariate regression results show that firm external agencies (about 59 percent of those sur- age, foreign ownership, and firm assets are all pos- veyed), 47 percent have access to bank finance. By itively associated with access to bank finance 4. Factor Markets: Finance and Labor 59 light the importance of the ownership of plant and Table 4.2 Share of Manufacturing Firms with buildings and the production of credible accounts in Access to Bank Finance, Uganda determining access to finance. Firm characteristic Percent Full sample 32.3 Bank Finance Instruments and Collateral Firm size class (employees) Requirements Micro (<10) 7.4 Loans and overdraft facilities are the predominant Small (10­49) 22.2 forms of bank debt. Loans typically finance the acqui- Medium-size (50­99) 47.1 sition of new plants and machinery, while overdraft fa- Large (100+) 72.9 cilities finance short-term liquidity requirements, en- Ownershipa abling firms to meet short-term obligations when cash Domestic 24.1 flows are temporarily low. Foreign 63.5 Most loans held by firms in the sample are of short duration, confirming the widely held view in the private Accounts externally audited sector that there is a paucity of long-term financing in- Yes 47.0 struments. More than 93 percent of the loans are No 11.0 backed by collateral. The average collateral to loan Decade in which established ratio is 116 percent (the median is 100 percent). Large <1960 52.9 firms and foreign firms post significantly less collateral 1960s 66.7 per shilling borrowed than smaller firms and domestic 1970s 36.4 firms. On average, 75 percent of the collateral takes 1980s 40.0 the form of immovable property (land, buildings, or 1990s 28.4 both). This confirms the finding that the ownership 2000s 21.1 of valuable fixed assets is an important determinant of a. A firm is considered foreign owned if 30 percent or more access to bank finance. Such collateral requirements of its equity is owned by non-Ugandan nationals or entities generally limit access to bank finance for small and Source: World Bank, Investment Climate survey, Uganda, medium-size enterprises. 2002/03. The overdraft facilities held by firms in the sample (table 4.3).3 The strong correlation between firm as- involve a similar set of security requirements. Some 60 sets and use of external auditors and start-up size percent of firms with overdraft facilities were required makes it difficult to identify the independent effect of to post collateral, mainly land, buildings, and other the capacity to produce reliable accounts. But firms' fixed property. That overdraft facilities are backed by use of external auditors is strongly and positively cor- fixed assets suggests unusually risk-averse behavior related with access to bank finance when the log of by banks or thin markets for finished manufactured assets is dropped in specifications 4 and 5. In speci- goods. One possible explanation, suggested by the fication 6 the log of assets is replaced with a dummy substitutability between overdrafts and loans, is that variable that takes the value 1 if the owner is a non- overdrafts are typically used to finance medium-term indigenous Ugandan or if the firm is a business group expenditure. Anticipating this, banks require that firms affiliate or is a corporation. While the results are not post substantial security before extending overdraft conclusive about the channels through which firm as- facilities. The average and median number of days sets and use of external auditors operate, they high- that firms with overdrafts reached the limits provides 4. Factor Markets: Finance and Labor 60 Table 4.3 Determinants of Access to Bank Finance for Manufacturing Firms, Uganda Variable (1) (2) (3) (4) (5) (6) Log of start-up size ­0.032 ­0.023 ­0.021 0.039 0.031 ­0.013 (0.025) (0.026) (0.026) (0.024) (0.025) (0.056) Log of assets 0.101 0.100 0.097 0.079 (0.018)*** (0.018)*** (0.018)*** (0.088) Firm uses external auditors 0.088 0.077 0.085 0.280 0.267 0.018 (0.090) (0.092) (0.092) (0.071)*** (0.073)*** (0.204) Foreign owned 0.002 0.002 0.002 0.002 0.002 0.001 (0.001)** (0.001)** (0.001)** (0.001)** (0.001)* (0.001) Employment growth, 2000­01 0.014 0.014 0.012 0.040 0.046 0.003 (0.095) (0.098) (0.096) (0.095) (0.094) (0.078) Firm age 0.012 0.011 0.011 0.018 0.018 0.009 (0.005)** (0.005)* (0.005)** (0.006)*** (0.006)*** (0.004)** Firm age squared ­0.000 ­0.000 ­0.000 ­0.000 ­0.000 ­0.000 (0.000)** (0.000)** (0.000)** (0.000)*** (0.000)*** (0.000)*** Uganda ­0.323 ­0.316 ­0.326 ­0.395 ­0.373 ­0.279 (0.067)*** (0.069)*** (0.070)*** (0.072)*** (0.075)*** (0.137)** Chemicals 0.110 0.091 0.087 0.085 0.045 (0.098) (0.098) (0.101) (0.103) (0.088) Metals&construction materials 0.089 0.080 0.075 0.066 0.026 (0.080) (0.081) (0.079) (0.080) (0.054) Furniture/wood 0.118 0.117 ­0.010 ­0.002 0.042 (0.101) (0.102) (0.101) (0.103) (0.078) Textiles/publishing 0.161 0.152 0.107 0.127 0.078 (0.086)* (0.086)* (0.087) (0.087) (0.074) Firm located outside capital 0.090 0.106 0.118 0.036 (0.067) (0.063)* (0.063)* (0.064) Nonindigenous owner 0.027 0.082 (0.073) (0.078) Firm not owned by individual or family 0.202 (0.107)* Business group affiliate 0.021 (0.070) Observations 398 398 398 391 388 366 Log likelihood ­180.47 ­178.36 ­177.46 ­189.09 ­186.39 Pseudo R2 0.35 0.35 0.36 0.30 0.31 R2 0.42 * Significant at the 10 percent level. ** Significant at the 5 percent level. *** Significant at the 1 percent level. Note: The dependent variable is access to bank finance (0/1). Reported coefficients are marginal effects for probit specifica- tions 1­5; specification 6 is a two-stage least squares estimation with the log of assets instrumented by nonindigenous owner, firm not owned by an individual or family, and business group affiliation. Figures in parentheses are robust standard errors. Source: Authors' calculations based on data from World Bank, Investment Climate survey, Uganda, 2002/03. 4. Factor Markets: Finance and Labor 61 further evidence of the intensive use of overdrafts trade finance accounts for the largest share of credit (table 4.4). Larger firms and foreign-owned firms ap- to the private sector.5 pear to use overdraft facilities more intensively than Uganda's monetary policy plays a part in the high others. cost of borrowing for the private sector. The public deficit in Uganda is financed by donor inflows, and for- eign exchange is converted to Uganda shillings to fi- Cost of Finance nance government expenditures. That expands the money supply, necessitating the mopping up of ex- Ugandan firms face high costs of borrowing from cess liquidity. The government has two instruments at banks. Firms reported paying an average of nearly 20 its disposal for doing so--the sale of treasury bills and percent in real terms for every shilling borrowed in the the sale of foreign exchange. With donor inflows being previous three years (figure 4.2). Interest rates on a significant proportion of the money supply (M2), in- overdrafts and loans are high across all categories of terventions by the Bank of Uganda in the treasury bill firms (table 4.5). With the real costs of borrowing so and foreign exchange markets are substantial. The high, it is unsurprising that trade finance accounts for Bank of Uganda appears to alternate between the two the largest share of credit to the private sector.4 instruments; consequently there are times when inter- Figure 4.2 shows the trend in borrowing rates over est rates are high and the value of the shilling is rela- the last five years. As the figure illustrates, borrowers tively low, and times when the opposite is true. During have been paying an average of nearly 20 percent in the survey period interventions in the treasury bill mar- real terms for every shilling borrowed in the last three ket were high and the private sector thus faced high in- years. Table 4.5 shows interest rates across firm size terest rates. Lowering the cost of finance would require and firm ownership; rates of interest on overdrafts and that the government reduce the public deficit and thus loans are high across all types of firms. With such its need to sterilize inflows of foreign exchange. high real costs of borrowing it is not surprising that Are Firms Credit Constrained? Table 4.4 Days in Previous Year on Which Manufacturing Firms with Overdrafts The findings on the access to and cost of finance sug- Reached Limit, Uganda gest that Ugandan manufacturing firms might be Firm characteristic Average Median credit constrained. To find out if this is so, firms were Full sample 131.37 90.00 asked whether they would like to borrow more at the Firm size class (employees) current interest rate. While this is the standard defini- Small (10­49) 112.85 90.00 tion of being credit constrained, it does not include Medium-size (50­99) 62.60 8.50 firms whose applications for credit have been rejected Large (100+) 169.74 150.00 or firms that self-select out of the credit market. Under a broader definition a firm is also said to be credit con- Ownership strained if it has not applied for a loan because of Domestic 100.12 45.00 inadequate collateral, a cumbersome application Foreign 166.93 150.00 process, or an expectation that its application will fail, Source: World Bank, Investment Climate survey, Uganda, or if it has applied for a loan and been rejected. 2002/03. Whether a firm is credit constrained under this 4. Factor Markets: Finance and Labor 62 Figure 4.2 Lending, Borrowing Rates and Inflation Lending, Borrowing rates, Inflation 25 20 15 % 10 5 0 1996 1997 1998 1999 2000 2001 2002 Year Lending rates Deposit rates Bank rates Inflation, av Source: World Bank, Investment Climate survey, Uganda, 2002/03. Table 4.5 Annual Interest Rate Charged Manufacturing Firms on Overdraft Facilities and Loans, Uganda (percent) Firm characteristic Overdrafts Loans Full sample 17.3 16.7 Firm size class (employees) Micro (<10) 16.0 21.8 Small (10­49) 18.3 16.8 Medium-size (50­99) 19.6 17.3 Large (100+) 16.0 15.9 Ownership Domestic 18.4 18.7 Foreign 16.1 14.0 Decade in which established <1960 15.3 18.2 1960s 19.7 18.9 1970s 23.5 24.0 1980s 16.0 14.7 1990s 17.6 17.0 2000s 17.4 13.4 Source: World Bank, Investment Climate survey, Uganda, 2002/03. 4. Factor Markets: Finance and Labor 63 definition is of course self-reported. Firms whose loan Trade Credit applications are turned down because their projects are bad are not actually credit constrained but would Trade credit is an important source of working capital be included under this definition. finance. By making it possible to purchase inputs on Some 30 percent of the firms with bank loans or credit, it allows firms to continue to operate even when overdraft facilities reported that they would like to bor- their cash flows are low. The volume of trade credit in row more. Among the firms that had never applied for an economy depends on supply and demand factors. a bank loan, 31 percent had not done so because In Uganda manufacturing firms face a number of fac- they believed they would not get the loan or found the tors likely to increase the demand for trade credit. For process too cumbersome. Slightly more than a quar- importing and exporting firms the country's land- ter of the firms in the sample reported being credit locked position creates transport-related risks. The constrained (table 4.6). The share was highest among supply of trade credit depends on three important micro firms: 41 percent of these firms reported failing factors: to get bank debt at the prevailing interest rate, com- pared with 19 percent of large firms. · Access to finance (for firms that supply trade credit). · Flows of information between contracting parties. · Quality of the legal system. Table 4.6 Share of Manufacturing Firms Reporting Being Credit Constrained, Uganda Information flows influence a firm's willingness to Firm characteristic Percent extend credit. Repeated interactions, whether through Full sample 26.3 the extension of credit or other transactions, enable a firm to acquire information about the reliability of its Firm size class (employees) contracting partner. Information suggesting that the Micro (<10) 40.7 partner has a high probability of survival is likely to in- Small (10­49) 26.8 crease the supply of trade credit. The concentration of Medium-size (50­99) 14.7 sectors is likely to affect the frequency of interaction Large (100+) 18.6 and therefore the amount of information transmitted to Ownership potential contracting parties. Domestic 28.7 Where the quality of the legal system is high, the Foreign 17.5 supply of trade credit is also likely to be high. Where the legal system is efficient, and the cost of legal ad- Decade in which established judication therefore low, firms will have a greater will- <1960 17.6 ingness to extend credit. 1960s 33.3 Nearly 60 percent of the firms in the sample re- 1970s 27.3 ported purchasing inputs on credit, with 85 percent of 1980s 20.0 large firms but only about 50 percent of small and 1990s 28.4 micro firms doing so (table 4.7). Similarly, while 84 2000s 26.3 percent of foreign-owned firms reported purchasing Source: World Bank, Investment Climate survey, Uganda, inputs on credit, only 53 percent of domestic firms did. 2002/03. Older firms use trade credit more intensively than 4. Factor Markets: Finance and Labor 64 A Comparison with Kenya Table 4.7 Share of Manufacturing Firms Purchasing Inputs on Credit, Uganda As the Ugandan private sector debates the merits of Firm characteristic Percent joining the East African Union, one important concern Full sample 59.4 it has voiced is the competitive edge of the Kenyan Firm size class (employees) manufacturing sector. Do Kenyan firms have better Micro (<10) 45.3 access to finance than their Ugandan counterparts? Small (10­49) 51.6 This section looks at that question with the aim of con- Medium-size (50­99) 73.5 tributing to policy to address any differences where Large (100+) 84.7 possible. A comparison of the share of firms with external fi- Ownership nance and those with bank loans in Kenya and Domestic 53.0 Uganda suggest that the Ugandan banking sector ra- Foreign 84.1 tions credit more intensively, forcing firms to rely on Decade in which established both loans and overdraft facilities? (figure 4.5). Also, <1960 82.4 banks' preference for short maturities may reflect a 1960s 83.3 very risky environment or concerns about nonpayment 1970s 45.5 There are no significant differences in collateral 1980s 55.6 requirements between Kenya and Uganda, however 1990s 56.6 (figure 4.6). Nor are there significant differences 2000s 63.2 across categories of firms within each country. The Source: World Bank, Investment Climate survey, Uganda, low collateral requirements reported by microenter- 2002/03. prises reflect noncommercial financing through devel- opment banks. younger firms, confirming the role of relationships and But a comparison of the share of firms reporting reputation in the provision of external finance. being credit constrained points to the comparative Firms accepting trade credit reported using it to disadvantage that manufacturing firms in Uganda purchase slightly more than half their inputs (figure face in obtaining finance. Across virtually all cate- 4.3). Large firms receive more trade credit than small gories of firms, a smaller share of manufacturing firms ones, and foreign firms more than domestic ones. This reported being credit constrained in Kenya than in implies that the volume of trade credit taken does not Uganda (figure 4.7). compensate for lack of access to bank finance, but in- Taken together, the data show a lower level of fi- stead enhances any inequality in access to external nancial intermediation in Uganda than in its neighbor. finance. On average, Kenyan manufacturing firms rely on in- The firms in the sample reported making an aver- ternal resources to finance less than half their invest- age of nearly a third of their sales on credit (figure ment and working capital needs. They are able to ob- 4.4). The firms most likely to have access to external tain commercial bank finance for 25 percent of those finance are also those most likely to provide trade needs. Moreover, Kenyan firms use trade credit much credit. This suggests that the provision of trade credit more intensively than Ugandan firms, relying on it to to some extent mitigates the lack of access to bank fi- fund 15 percent of their working capital needs. A com- nance among smaller, younger firms. parison of the share of firms using trade credit sug- 4. Factor Markets: Finance and Labor 65 Figure 4.3 Share of Inputs Purchased on Credit by Manufacturing Firms with Trade Credit, by Size Class and Type of Ownership, Uganda 80 70 60 50 40 Percent 30 20 10 0 Very Small Medium Large Domestic Foreign Total small Source: World Bank, Investment Climate survey, Uganda, 2002/03. Figure 4.4 Share of Sales Made on Credit by Figure 4.5 Share of Manufacturing Firms Manufacturing Firms, by Size Class and with Bank Loans and with External Finance, Type of Ownership, Uganda Kenya and Uganda (Percent) 8 Total Foreign 6 Domestic ms fir of 4 Large tion opor Pr Medium 2 Small 0 Very small Kenya Uganda 0 10 20 30 40 50 60 Loans External finance Percent Source: World Bank, RPED survey, Uganda, 1998, and Investment Source: World Bank, RPED investment climate survey, Climate survey 2002/03. Uganda, 2002/03 and Kenya, 2003. 4. Factor Markets: Finance and Labor 66 Figure 4.6 Share of Loans to Manufacturing Firms Requiring Collateral, Kenya and Uganda Firm size class (employees) Ownership 100 100 80 80 60 60 Percent 40 Percent 40 20 20 0 Micro Small Medium- Large 0 (<10) (10­49) sized (100+) Domestic Foreign (50­99) Decade in which firm established Use of External Auditors 100 100 80 80 60 60 40 Percent Percent 40 20 0 20 No external audits External audits 0 Kenya Uganda <1960 1960s 1970s 1980s 1990s 2000s Source: World Bank, RPED investment climate survey, Uganda (2002/03), Kenya (2003). gests another glaring source of comparative disad- decisions) needs to be improved. And small and vantage--nearly 85 percent of manufacturing firms in medium-size firms need access to credible and Kenya purchase inputs on credit (figure 4.8). affordable auditing services to improve their ac- cess to external finance. This should be a central focus of business development services providing Policy Issues support to small and medium-sized firms. · Establishing a credit registry should cement firms' This analysis of firms' access to bank finance, their incentives to maintain reliable records and reduce use of trade credit, and the costs, correlates, and re- the transaction costs associated with lending to them. quirements for each of these sources of external fi- · Improving laws relating to foreclosure might in- nance points to recommendations very much in keep- crease access to finance for firms whose asset ing with those of the earlier assessments of the holdings are small and of uncertain value. In addi- Ugandan financial sector: tion, providing the commercial courts with suffi- cient resources to operate effectively should in- · The capacity of firms to produce credible ac- crease the availability of both bank finance and counts (or other information relevant to lending trade credit for the private sector. 4. Factor Markets: Finance and Labor 67 Figure 4.7 Share of Manufacturing Firms Reporting Being Credit Constrained, Kenya and Uganda Firm size class (employees) Ownership 40 04 30 03 20 02 Percent Percent 10 0 0 Micro Small Medium Large 01 (<10) (10-49) (50-99) (100+) Domestic Foreign Decade in which firm established Use of External Auditors 40 40 30 30 20 20 Percent Percent 10 10 0 No external audits External audits 0 Kenya Uganda <1960 1960s 1970s 1980s 1990s 2000s Source: World Bank, RPED investment climate surveys, Uganda (2002/03), Kenya (2003). · Reducing public deficits would help reduce in- rience, and demographics. (See appendix 2 for a de- flows of foreign exchange and the need to sterilize tailed discussion of the Ugandan labor market.) those inflows through the sale of treasury bills. The overall picture of the labor market in Uganda That in turn would reduce the cost of finance for is relatively positive, health issues aside. The labor force the private sector. is relatively well trained, and wage levels are relatively competitive. But when labor productivity is taken into account, the picture becomes bleaker. The median The Labor Market annual value added per manufacturing worker in Uganda is low: $1,085 in 2002, compared with $3,432 A well-functioning labor market is critical both for in India and $4,397 in China in around the same period. boosting macroeconomic performance in Uganda and The low labor productivity prevents Uganda from taking for achieving sustainable improvements in the living advantage of its relatively low wages. standards of the poor. With this in mind, a large section As seen in chapter 2, labor costs in Uganda are of the firm survey in Uganda was devoted to collecting high relative to worker productivity. Thus there is a worker data. A sample of up to 10 workers in each firm need to better understand the link between labor mar- were interviewed, to obtain information on wages, oc- ket outcomes and labor productivity, two factors that cupation, union status, education, tenure, layoff expe- affect unit labor costs. This section focuses on three 4. Factor Markets: Finance and Labor 68 Figure 4.8 Share of Manufacturing Firms Using Trade Credit, Kenya and Uganda Firm size class (employees) Ownership 100 100 80 80 60 60 Percent 40 Percent 40 20 20 0 Micro Small Medium Large 0 (<10) (10-49) (50-99) (100+) Domestic Foreign Decade in which firm established Use of External Auditors 100 100 80 80 60 60 Percent 40 Percent 40 20 0 20 No external audits External audits 0 Kenya Uganda <1960 1960s 1970s 1980s 1990s 2000s Source: World Bank, RPED investment climate surveys, Uganda (2002/03), Kenya (2003). key areas--the health status of workers, the determi- faces big challenges in health. Its health indicators nants of wage levels, and the relative importance of are among the lowest in the world. Life expectancy at institutional rigidities in the labor market. Its findings birth, for example, was only 42.1 years in 1999.6 suggest that the health of the labor force and the rel- Moreover, Uganda has been severely affected by the atively weak link between wages and the performance HIV/AIDS pandemic, now the leading cause of death of workers need to be addressed to improve the com- among adults, followed by tuberculosis and malaria petitiveness of Ugandan labor. Increasing labor mo- (EIU 2002). Public policies have succeeded in slow- bility and eliminating wage differentials for the same ing the annual growth in HIV infections, a rate that types of jobs across regions and sectors might also had reached 6.3 percent in 1996. But the growth rate help. in 1998­99 was still around 3.1 percent (UBOS 2001). Studies assessing the impact of HIV/AIDS on economic productivity in Africa have estimated it to Health Status be in the neighborhood of 1 percent of GDP (Sachs and Bloom 1998). Given this situation, it is particu- The health status of Uganda's labor force has impor- larly important to investigate the responses of firms tant consequences for worker productivity. Uganda and employees to HIV/AIDS. 4. Factor Markets: Finance and Labor 69 Illness and Workdays Lost vate health care and public facilities. These choices While no comparative data across Africa are available are probably linked to the level of earnings for each on illness and workdays lost, the survey data on these category of employee. Firms' own health care facilities indicators for Uganda suggest that the health status of and those of nongovernmental organizations (NGOs) its labor force needs to be addressed. Across all sec- do not appear to be workers' first choice; only 12.5 tors surveyed, about 24.7 percent of workers reported percent and 0.8 percent of employees seek treatment having been ill within the previous 30 days. This share in these facilities. This probably reflects limited avail- ranged from 17.4 percent in construction to 21.2 per- ability of such facilities and uncertainty about the cent in tourism, 23.2 percent in commercial agricul- quality of the treatment they provide. ture, and 25.6 percent in manufacturing. Illness led to When it comes to paying for health care services, an average loss of 3.2 workdays over a 30-day period most workers reported that the costs are borne by across all sectors. In case of illness a majority of work- their household (44.4 percent) or directly reimbursed ers said that they would obtain treatment from health by their employer (31.5 percent). The share of workers care providers (55.2 percent), and in a large share of who can depend on reimbursements from their em- these cases the worker's own household would bear ployer varies by rank. Unskilled workers tend to rely the expense (43.4 percent). heavily on their own funds, with 49.9 percent reporting In manufacturing, where about a quarter of the that their household bears the cost of treatment. Only workforce reported having been ill within the previous 19.3 percent of these workers reported being reim- 30 days, illness led to an average loss of about 3.1 bursed by their employer, compared with 43.4 percent workdays over the 30-day period--or about 15 per- of professionals. cent of available workdays (table 4.8). Under the as- sumption of a constant rate of illness over a year, that Awareness of HIV/AIDS translates into 37.2 days of production lost per worker About 37 percent of the firms surveyed reported un- on average. Based on the total factor productivity es- dertaking actions aimed at preventing and raising timation described in chapter 2, a 15 percent loss of awareness of HIV (table 4.10). Firms in commercial labor translates into a production loss of 11 percent. agriculture and tourism lead the effort. Firms in con- The share of workers who reported having been ill struction and manufacturing are less active: only 26 in the previous 30 days grows with the size of the firm. percent of construction and 32 percent of manufac- But the number of workdays lost tends to decline as turing firms undertake preventive activities. When the size of the firm increases. This suggests that firms do undertake activities to address HIV/AIDS, larger firms take appropriate steps to ensure that their they limit their efforts mostly to advertising (77.4 per- workers receive proper treatment or that employees of cent) and counseling (42.5 percent). larger firms, which usually pay higher wages, are able Surprisingly, almost 60 percent of managers re- to pay for better treatment. ported that HIV/AIDS has had little or no impact on their workforce. This response may reflect ignorance, Treatment Sources and Payment Methods an inability to distinguish HIV/AIDS from other ill- Most workers in manufacturing (almost 57 percent) nesses, or a drastic underestimation of the problem. rely on private health care providers to obtain treat- By contrast, employees have a much more acute ment (table 4.9). Managers and professionals rely perception of the HIV/AIDS issue, perhaps as a result more heavily on private health care (more than 60 of the public campaign of recent years. Between 80 percent), while production workers use a mix of pri- percent and 93 percent of the sample of workers know 4. Factor Markets: Finance and Labor 70 Table 4.8 Illness and Workdays Lost in Previous 30 Days as Reported by Manufacturing Workers, Uganda Share of workers Average Workdays Firm size class reporting being missed because (employees) ill (percent) of illness Micro (<10) 23.31 4.28 Small (10­49) 24.75 3.69 Medium-size (50­99) 25.50 2.96 Large (100+) 28.19 1.80 All firm size classes 25.59 3.07 Source: World Bank, Investment Climate survey, Uganda, 2002/03. Table 4.9 Health Facilities and Payment Arrangements for Treatment Used by Manufacturing Workers in Different Job Categories, Uganda (percent) Skilled Unskilled Non- production production production All job Management Professionals workers workers workers categories Preferred source of treatment Health care facilities of firm 11.50 20.93 15.48 8.31 12.88 12.56 Private health providers 67.26 62.02 49.23 55.50 52.36 56.94 Public facilities 19.76 14.73 34.37 33.74 33.48 28.82 Facilities of NGOs or charities 0.59 0.78 0.31 1.96 0.00 0.84 Other 0.88 1.55 0.62 0.49 1.29 0.84 Payment arrangements used for treatment No significant expenses Received free or low-cost treatment 17.99 9.30 7.43 20.54 9.40 14.16 High expenses Reimbursed by employer 37.17 43.41 31.58 19.32 37.61 31.45 Reimbursed by insurance company 4.42 4.65 1.55 0.98 2.14 2.44 Covered by financial support from friends and family 2.65 6.20 12.07 8.31 5.13 7.11 Borne by household 37.46 35.66 47.37 49.88 45.73 44.42 Other 0.29 0.78 0.00 0.98 0.00 0.42 Source: World Bank, Investment Climate survey, Uganda, 2002/03. 4. Factor Markets: Finance and Labor 71 correlated with income. Managers are ready to pay Table 4.10 Firms and HIV/AIDS in Selected more (9,353 USh, or $5.30) than unskilled workers Sectors, Uganda (3,235 USh, or $1.80). Share of firms (percent) Firms undertaking preventive activities by sector Remuneration and Determinants of Commercial agriculture 57.78 Wages Construction 26.32 Manufacturing 32.00 Wage levels are a critical determinant of unit labor Tourism 67.86 costs. As shown in chapter 2, unit labor costs in Full sample 37.24 Ugandan manufacturing are high, and a better under- Activity undertaken by active firms standing of what drives manufacturing wages is Prevention messages 77.40 needed to understand why this is so. This section an- Free condom distribution 28.70 alyzes the determinants of wage rates in manufactur- Counseling 42.50 ing using data from the survey of workers. While the Anonymous HIV testing 4.70 survey covered four sectors, the vast majority of the Financial support of infected dependents 8.20 1,803 workers interviewed were in manufacturing Perceived impact of HIV/AIDS on workforce (1,436), while only 160 were in tourism, 138 in com- High absenteeism among infected workers 15.10 mercial agriculture, and 69 in construction. High absenteeism among workers needing to care for infected family members or friends 15.40 Wage Levels across Sectors High staff turnover 6.70 About 1,522 of the workers surveyed provided usable, No effect 59.00 detailed data on their earnings, including their base Note: Data are for the commercial agriculture, construction, wage, allowances, and bonuses. Monthly earnings manufacturing, and tourism sectors. average about $197 over the sample of workers (table Source: World Bank, Investment Climate survey, Uganda, 4.11). But earnings differ significantly across sectors, 2002/03. ranging from $135 a month in commercial agriculture to $892 in construction. In addition, managers' where to be tested. Moreover, the level of awareness salaries in construction are much higher than those in tends to increase with employees' position in their firm the other sectors, reflecting the greater scarcity of (figure 4.9). Overall, employees rank HIV/AIDS as a skills and higher level of technical knowledge in this significant concern (giving it an average ranking of sector. Indeed, the earnings for almost all job cate- 4.07 on a scale of 1 to 5, with 5 being the highest; see gories are highest in construction (interviews of con- appendix table A2.12). struction workers were limited in number, however). Another indication of employees' perception of the Manufacturing ranks second, with average monthly importance of the issue is that 72 percent of workers earnings of $180. are ready to pay to be tested at their firm, as long as the tests are anonymous and voluntary (see appendix Wage Levels in Manufacturing table A2.12). On average, employees would be willing In manufacturing, wages account for about 88 per- to pay about 5,914 USh ($3.40) to be tested. The will- cent of cash earnings on average, cash allowances ingness to pay differs across job categories and is for about 10 percent, and performance bonuses for 4. Factor Markets: Finance and Labor 72 Figure 4.9 Share of Employees Aware of HIV and Willing to Pay for Testing in Selected Sectors by Job Category, Uganda (percent) 100 91.31 92.90 90 84.33 83.18 80.41 80 77.25 75.72 71.60 72.19 70 67.35 60 50 40 30 20 10 0 Management Professionals Skilled Unskilled Nonproduction production production workers workers workers Aware of HIV Willing to pay Note: Data are for the commercial agriculture, construction, manufacturing, and tourism sectors. Source: World Bank, Investment Climate survey, Uganda, 2002/03. about 1.3 percent. Bonuses account for a relatively than the sample mean, about $57.50 a month com- small part of earnings compared with the share in pared with the average of $180.40. some other African countries; they were around 6 per- Manufacturing earnings also vary a great deal cent of cash earnings in Nigeria in 2000 and about 7.5 across subsectors. Earnings are highest in the paper, percent in Eritrea in 2001. printing, and publishing industry and the metals sub- Wages are usually paid monthly. This is the case for sector, and lowest in the furniture subsector (see ap- almost 81 percent of workers in the sample and 72 per- pendix table A2.14). Cash earnings tend to increase cent of production workers (see appendix table A2.13). with firm size and differ greatly depending on geo- Interestingly, a significant share of production workers graphic location. Firms in the central region pay much (about 15.5 percent) are paid by the piece, which al- higher wages than those in other regions. lows firms to link workers' pay to their productivity. Particularly useful is to examine the pattern of Male workers earn about 24 percent more than fe- earnings for unskilled production workers, the most male workers on average, suggesting a possibility of homogeneous wage category and thus fairly compa- gender discrimination in the labor market (table 4.12). rable across firms and countries. The wage level of In addition, there is substantial dispersion of earnings unskilled production workers also serves as a barom- within job categories, as shown by the large standard eter for foreign firms deciding whether to invest in a deviations. Earnings also vary sharply across job cat- country. In Uganda unskilled production workers earn egories. Unskilled production workers earn much less about $57 a month on average, similar to levels in 4. Factor Markets: Finance and Labor 73 Table 4.11 Monthly Cash Earnings in Selected Sectors by Job Category, Uganda (U.S. dollars) Commercial Job category Item agriculture Construction Manufacturing Tourism Full sample Management Average earnings 143.61 2,982.81 409.49 347.35 464.68 (148.6) (5,006.8) (751.8) (391.5) (1,168.9) Observations 24 10 289 13 336 Professionals Average earnings 85.71 419.14 286.54 474.99 300.41 (121.2) (144.3) (330.1) (881.8) (384.1) Observations 5 8 117 10 140 Skilled workers Average earnings 183.41 230.32 125.52 144.27 132.56 (120.9) (118.1) (146.2) (83.1) (143.9) Observations 12 14 303 13 342 Unskilled workers Average earnings 119.58 137.67 57.48 80.18 65.62 (321.3) (142.1) (58.4) (59.2) (120.3) Observations 46 3 360 16 425 Others Average earnings 231.43 193.93 102.91 69.38 101.63 (157.6) (165.6) (111.9) (53.1) (109.4) Observations 2 8 229 40 279 All job categories Average earnings 135.28 892.33 180.40 165.21 196.96 (249.3) (2,597.7) (402.6) (347.1) (591.2) Observations 89 43 1,298 92 1,522 Note: Computed on the basis of the earnings reported by workers in early 2003 and converted into U.S. dollars using the official exchange rate of $1 = 1,750 USh. Figures in parentheses are standard deviations. Source: World Bank, Investment Climate survey, Uganda, 2002/03. such countries as Nigeria or Kenya, where the range log of individual earnings as the dependent variable. is $73­100 a month. Earnings in Uganda are lower The results are shown in table 4.13. than those in India and China. But as the discussion The starting point is a basic wage equation related of unit labor costs in chapter 2 shows, when the pro- to the individual characteristics of workers (Mincer ductivity of workers is taken into account, Uganda is 1974). This first equation assumes that employers are not competitive with the fast-growing economies of able to discern differences in productivity among Asia. workers depending on their education, gender, and experience and compensate them accordingly. The Determinants of Manufacturing Wages second equation differs from the first in its inclusion of The apparently large variation in manufacturing sector and subsector dummy variables and firm- wages across subsectors and regions suggests that specific variables. (For ease of reading, these dummy the labor market may not be competitive or fully inte- variables are not shown in the table.) grated. To understand the wage differentials, wage In both specifications the variables relating to equations were estimated with worker data, with the human capital--years of education, years of experi- 4. Factor Markets: Finance and Labor 74 Table 4.12 Monthly Cash Earnings in Manufacturing by Job Category, Uganda (U.S. dollars) Male Female All Job category Item employees employees employees Management Average earnings 403.23 472.79 409.49 (764.6) (617.0) (751.8) Observations 263 26 289 Professionals Average earnings 292.03 269.90 286.54 (343.8) (289.5) (330.1) Observations 88 29 117 Skilled production workers Average earnings 128.71 94.24 125.52 (151.7) (65.4) (146.2) Observations 275 28 303 Unskilled production workers Average earnings 54.51 72.04 57.48 (50.3) (87.2) (58.4) Observations 299 61 360 Nonproduction workers Average earnings 105.27 100.16 102.91 (134.1) (79.4) (111.9) Observations 123 106 229 All job categories Average earnings 187.40 151.08 180.40 (429.3) (260.9) (402.6) Observations 1,048 250 1,298 Note: Computed on the basis of the earnings reported by workers in early 2003 and converted into U.S. dollars using the official exchange rate of $1 = 1,750 USh. Figures in parentheses are standard deviations. Source: World Bank, Investment Climate survey, Uganda, 2002/03. ence with the firm, other work experience--all have a gender discrimination becomes an issue only when positive and statistically significant effect on wages. workers are employed by firms in the central region or Thus the greater a worker's endowment of human firms with foreign ownership. capital, the higher his or her wages are. Formal train- Subsector dummy variables are mostly significant. ing also has a positive and significant effect. The vari- In addition, in the second equation some firm-level able "weekly hours worked" is significant in the first variables are significant. Working for firms located in but not the second equation. This result suggests that the central region or with foreign ownership translates earnings in manufacturing are weakly correlated with into higher earnings. The age of the firm is insignifi- performance. One possible explanation is that cant, however, suggesting that if firms benefit from bonuses are such a small share of manufacturing any reputational effects, they do not pass the benefits earnings in Uganda. on to workers. The gender dummy variable is insignificant in the The results suggest that wages are not competi- first equation but highly significant in the augmented tively determined in Uganda. If the labor market were equation with firm-specific effects. This suggests that competitive, none of the firm-specific variables and 4. Factor Markets: Finance and Labor 75 subsector dummy variables would be significant in Table 4.13 Estimates of Wage Determinants explaining earnings. Increasing labor mobility through in Manufacturing, Uganda better infrastructure and expanding access to educa- Variable (1) (2) tion and vocational training in all regions would pre- Intercept 2.813*** 2.633*** sumably help to remove the distortions in the labor (21.07) (11.69) market and reduce the wage differentials observed in Worker characteristics the private sector. Education in years 0.082*** 0.076*** (11.10) (11.40) Experience with firm in years 0.045*** 0.053*** (3.77) (4.58) Institutional Rigidities Experience in years squared ­0.001*** ­0.001*** (­2.68) (­3.52) Beyond the distribution of firms and workers, three Other professional experience 0.049*** 0.042*** other factors may affect the structure of the labor mar- in years (7.28) (6.73) ket: the type of labor contracts in use, the role of labor Gender dummy variable 0.024 0.127** unions, and the constraints imposed by labor regula- (1 if male, 0 otherwise) (0.38) (2.11) tions. This section examines the role of institutional Weekly hours worked 0.005*** 0.002 (3.77) (1.55) rigidities in determining wage and employment Training dummy variable 0.384*** 0.351*** levels. (1 if postschool training, 0 otherwise) (5.59) (5.55) Labor Contracts Most workers in the sample--about 59.5 percent-- Subsector dummy variables No Yesa hold permanent full-time jobs (table 4.14). Permanent employment is especially prevalent in tourism (80 per- Firm characteristics cent of workers) and commercial agriculture (76 per- Age in years 0.000 cent). By contrast, permanent employment is much (­0.27) less developed in manufacturing (56 percent of work- Central region dummy variable 0.455*** ers) and construction (34 percent). Thus firms in these (1 if in central region, 0 otherwise) (8.28) two sectors have some flexibility in employment levels. Foreign ownership dummy variable 0.344*** (1 if 50 percent or more of capital Manufacturing firms in Uganda rely more on tem- is foreign) (0.03) porary (casual and part-time) labor contracts than do Observations 1,232 1,232 firms in other recent manufacturing surveys by the F-statistic 65.594 42.544 Regional Program on Enterprise Development R2 0.273 0.387 (RPED).7 But a few subsectors are exceptions (table 4.15). In the wood subsector, for example, 71 percent ** Significant at the 5 percent level. *** Significant at the 1 percent level. of workers are permanent employees, and in the con- Note: Dependent variable is the log of individual earnings. struction materials industry 67 percent are. Also Equations have been estimated using ordinary least noteworthy is the inverse relationship between the squares. Figures in parentheses are White's consistent size of firms and the share of permanent employees. t-ratios, used to correct for heteroskedasticity in the data. a. Partially significant. In microenterprises about 64 percent of workers are Source: Authors' calculations based on data from World permanent, while in large firms around 55 percent Bank, Investment Climate survey, Uganda, 2002/03. are. 4. Factor Markets: Finance and Labor 76 Table 4.14 Employment Structure in Selected Sectors by Firm Size Class, Uganda (percent) Micro Small Medium- Large All Sector and employment (<10 (10­49 size (50­99 (100+ firm size category employees) employees) employees) employees) classes Commercial agriculture Permanent full time 76.19 44.30 3.33 78.87 76.38 Casual full time 9.52 13.16 0.00 9.11 9.14 Part time 14.29 42.53 96.67 12.02 14.49 Construction Permanent full time 100.00 66.67 16.67 31.04 33.97 Casual full time 0.00 3.70 83.33 61.04 56.39 Part time 0.00 29.63 0.00 7.92 9.64 Manufacturing Permanent full time 63.82 62.04 56.81 55.42 56.10 Casual full time 25.66 31.46 31.16 26.41 27.10 Part time 10.53 6.51 12.03 18.17 16.80 Tourism Permanent full time 83.33 85.87 85.61 76.27 79.53 Casual full time 13.33 9.78 13.16 16.94 15.34 Part time 3.33 4.35 1.23 6.79 5.13 Full sample Permanent full time 67.19 61.70 58.74 59.35 59.54 Casual full time 22.57 26.82 27.35 24.68 24.97 Part time 10.24 11.49 13.90 15.97 15.48 Source: World Bank, Investment Climate survey, Uganda, 2002/03. Labor Unions erage to strikes in the previous year?. The larger the The survey data suggest that union membership is firm, however, the larger the number of days lost. low in Uganda. Only around 10 percent of firms re- Further confirming the low profile of unions, only ported having employees who belong to a union, one 6.5 percent of firms reported that union-negotiated of the smallest shares found in RPED surveys (table wages affect nonunion workers. By contrast, prelimi- 4.16). In firms where there is union membership, nary estimates for Kenya in 2002 indicate that in about 45 percent of employees on average belong to 56 percent of firms nonunion workers share union- a national union. As a result of the low membership negotiated wages and benefits. levels, unions have a negligible impact on wage and employment levels. Labor Regulations Moreover, few days are lost to strikes. Firms re- The regulatory framework seems to have little impact ported losing less than half a day of production on av- on the labor market in Uganda. Asked to rate 4. Factor Markets: Finance and Labor 77 Table 4.15 Employment Structure in Manufacturing by Firm Size Class, Uganda (percent) Micro Small Medium- Large All Subsector and (<10 (10­49 size (50­99 (100+ firm size employment category employees) employees) employees) employees) classes Agro-industry Permanent full time 66.67 54.32 56.30 59.32 59.00 Casual full time 30.89 38.54 34.56 14.71 16.72 Part time 2.44 7.14 9.14 25.97 24.28 Chemicals and paints Permanent full time 100.00 54.61 62.57 52.30 54.58 Casual full time 0.00 30.50 37.43 42.34 39.73 Part time 0.00 14.89 0.00 5.35 5.69 Construction materials Permanent full time 50.00 53.06 55.97 75.20 67.27 Casual full time 0.00 38.65 30.59 24.80 28.47 Part time 50.00 8.30 13.45 0.00 4.26 Furniture Permanent full time 53.23 69.02 43.58 18.37 49.27 Casual full time 27.42 25.85 0.56 81.63 35.59 Part time 19.35 5.12 55.87 0.00 15.14 Metals Permanent full time 100.00 71.10 0.00 40.51 41.35 Casual full time 0.00 25.10 100.00 57.10 56.23 Part time 0.00 3.80 0.00 2.39 2.42 Paper, printing, and publishing Permanent full time 92.86 72.22 59.51 56.20 60.58 Casual full time 7.14 24.65 29.75 31.14 29.27 Part time 0.00 3.13 10.74 12.66 10.16 Plastics Permanent full time -- 62.07 69.23 25.00 50.53 Casual full time -- 37.93 30.77 40.00 37.01 Part time -- 0.00 0.00 35.00 12.46 Textiles and leather products Permanent full time 16.67 83.41 100.00 57.89 65.78 Casual full time 83.33 13.82 0.00 35.25 28.71 Part time 0.00 2.76 0.00 6.86 5.51 Wood Permanent full time 33.33 81.18 76.74 66.67 70.89 Casual full time 0.00 0.00 23.26 33.33 25.32 Part time 66.67 18.82 0.00 0.00 3.80 Full sample Permanent full time 63.82 62.04 56.81 55.42 56.10 Casual full time 25.66 31.46 31.16 26.41 27.10 Part time 10.53 6.51 12.03 18.17 16.80 -- Not available. Source: World Bank, Investment Climate survey, Uganda, 2002/03. 4. Factor Markets: Finance and Labor 78 Table 4.16 Union Membership and Days of Production Lost to Strikes in Selected Sectors, Uganda Commercial Full agriculture Construction Manufacturing Tourism sample Union membership Share of firms with employees belonging to a union (percent) 11.11 0.00 10.00 12.00 9.8 Share of employees belonging to a national union (percent)a 43.75 0.00 44.12 58.33 45.2 Days of production lost to strikes in previous year by firm size class (employees) Micro (<10) 0.0 0.0 0.0 0.5 0.1 Small (10­49) 0.0 0.0 0.5 0.0 0.4 Medium-size (50­99) 0.0 0.0 0.1 0.3 0.1 Large (100+) 0.3 0.1 1.4 0.0 1.0 All firm size classes 0.1 0.1 0.5 0.2 0.4 a. Data refer only to firms in which any employees belong to a union. Source: World Bank, Investment Climate survey, Uganda, 2002/03. constraints relating to labor regulations on a scale of 0 verse consequences for firm productivity. Im- (no problem) to 4 (high obstacle), firms gave re- proving the health of workers is therefore critical. sponses suggesting that such regulations impose a · A specific area of action should be HIV/AIDS. The minimal burden (table 4.17). Only 5 percent of firms survey data show that firms tend to underestimate, reported experiencing significant problems with layoff ignore, or conceal the extent of the HIV/AIDS procedures and the cost of retrenchment.8 problem. Even though the Ugandan government has taken steps to combat HIV/AIDS, the private sector remains relatively uninformed. Yet a large Policy Issues share of employees across all sectors surveyed are willing to be tested and to pay for the tests as This brief review of some of the main characteristics of long as testing is anonymous. These attitudes are Uganda's labor market points to some issues that a very favorable factor in the fight against need to be addressed to improve the competitiveness HIV/AIDS in the work environment, with obvious of its labor: externalities for labor productivity and for Ugandan society. · The health of the labor force is a major concern. · A continued focus on education and professional Health conditions in the country are poor, and up training remains vital for the long term. to 37 days of production per worker are lost annu- · Workers' earnings in manufacturing appear to ally because of health-related issues, with ad- be only weakly linked to their performance, as 4. Factor Markets: Finance and Labor 79 Table 4.17 Firms' Evaluation of Labor Regulations as Constraints in Selected Sectors, Uganda Share of firms citing as a significant Average constraintb Constraint ratinga (percent) Hiring procedures for local workers 0.1 2.81 Hiring procedures for foreign workers 0.1 4.59 Layoff procedures and cost of retrenchment 0.3 5.10 Limits on temporary hiring 0.1 3.06 Dealing with the Inspectorate of Labor 0.3 3.92 Note: Data are for the commercial agriculture, construction, manufacturing, and tourism sectors. a. Ratings are on a scale of 0 (no problem) to 4 (high obstacle). b. Refers to ratings in the 3­4 range on the scale of 0­4. Source: World Bank, RPED investment climate survey, Uganda, 2002/03. shown by the small share of bonuses in their pay more than 95 percent of the firms in the sample and the insignificance of hours worked in explain- are less than 50 years old. ing the level of earnings. Thus workers have little 4. Another reason for the large share of trade finance incentive to increase their level of effort. To in- is the lower uncertainty about the value of collat- crease incentives, earnings should be clearly eral posted by the trading enterprises. linked to performance. 5. Lower uncertainty in the value of collateral posted · Institutional rigidities do not appear to be severe. by the trading enterprises is another reason for the Ugandan firms are able to hire workers on flexible larger share of private sector credit. contracts and have few problems with unions. 6. Data are from the World Bank's World Develop- ment Indicators database. 7. In Nigeria, for example, about 89 percent of manu- Notes facturing workers had permanent full-time labor contracts in 2001, and in Eritrea about 85 percent 1. The initial sale of this bank in 1997 to a Malaysian did in 2002. But in Kenya only about 58 percent of firm was invalidated when the firm transferred manufacturing workers had permanent labor con- ownership to a consortium of Ugandan investors tracts in 2002­03. well ahead of the stipulated schedule. 8. Other data, however, suggest that some institu- 2. Nearly 80 percent of domestically controlled firms tional rigidities prevail. Most notably, the World have fewer than 50 employees, compared with Bank's Doing Business database indicates that only 30 percent of foreign-controlled firms. Ugandan firms have greater difficulty in laying off 3. The quadratic relationship between age and ac- workers than do firms in other countries in the cess to bank finance is maximized at age 50. But region or in OECD countries (World Bank 2003). 5. Policy Implications 80 Maintaining Macroeconomic Stability 81 Encouraging Private Provision of Social and Infrastructure Services 82 Establishing a Low-Cost Operating Environment 82 Establishing a Competitive Investment Environment 83 Improving Tax Administration 83 Ensuring Sound Financial Market Development 84 Raising Firm Productivity 84 Increasing the Efficiency of Services 85 Addressing Distortions in Trade 86 The Role of the World Bank 86 Annex 5.1 Policy Suggestions for Improving the Investment Climate in Uganda 88 5. Policy Implications 81 The findings of the investment climate assessment Recommendations. Now that Uganda has confirm those from recent consultations with the pri- achieved sustained periods of macroeconomic sta- vate sector in Uganda. These findings include a rela- bility, the challenge is to demonstrate to investors that tively high-cost business environment, low investor its macroeconomic framework is sustainable over the confidence (particularly with respect to the macroeco- long haul. Uganda's economic and political institu- nomic situation), low productivity, and low capacity tions must be capable of withstanding not only re- utilization. Although core reforms in the past decade gional insecurity but also political transition. The gov- have improved the investment climate, Uganda needs ernment must keep its administrative budget low, to consolidate and build on those reforms if it is to reinforcing the image of a lean, professional civil ser- transform itself into a competitive economy. The coun- vice capable of performing without increasing donor try needs to maintain a stable macroeconomic frame- funding. The Ministry of Finance, Planning, and work, establish a low-cost business environment, Economic Development, in collaboration with the strengthen the financial sector, privatize and reform Ministry of Public Service, needs to work on modern- key utilities, and raise firm productivity by boosting izing the civil service, which should be capable of de- capacity utilization and the efficiency of the labor and livering the services the private sector needs to com- financial markets. pete successfully. Persistent pressure to increase military spending and the creation of more public in- stitutions and district administrations pose the key Maintaining Macroeconomic Stability challenges to macroeconomic stability. The growing budget deficit and continued decline in the exchange Macroeconomic instability is among the most impor- rate, reflecting the persistent trade imbalance, under- tant constraints identified by Ugandan firms--and the mine confidence in the sustainability of growth. most important one in the view of foreign-owned com- Private firms surveyed for this report share these panies and exporters. That investor perceptions are concerns. so negative is interesting, since Uganda has sus- As discussed in the World Bank's Public tained high GDP growth, low inflation, and low ex- Expenditure Review of Uganda in 2003, several change rate volatility for the past 10 years or more. strategies could be pursued to reduce the fiscal These perceptions suggest that investors are no deficit: increasing transparency and participation in longer comparing Uganda's economic performance the setting of sectoral budget ceilings, improving today with that of the past decade, but instead focus- budget monitoring systems, minimizing unbudgeted ing on recent indicators of the economy's perform- supplementary expenditures that lead to spending ance, which point to an overall slowdown. Declining cuts for other departments, increasing budget effi- economic growth, rising public spending (especially ciency so as to allow the government to achieve de- for public services), inconsistent tax policies, and velopment goals even while reducing or maintaining poor tax administration are the key concerns for in- spending levels, and encouraging stakeholders in the vestors. Continued high military spending may also development process to assist local governments in be shaping investors' expectations about inflation and decentralizing social services, budget processes, and macroeconomic stability. financial management (World Bank, 2003b). 5. Policy Implications 82 Encouraging Private Provision of Social days a year on average--dealing with government of- and Infrastructure Services ficials. This regulatory burden, and the associated high costs for the private sector, are a result of inade- The government's policies and actions are consistent quate regulatory capacity, an unclear regulatory with an overall policy of establishing the private framework, and inconsistent interpretation of policies (rather than the public) sector as the provider of jobs and regulations. Forty percent of the manufacturing and incomes. But the provision of large-scale funding firms surveyed complained that regulations are not in- by the international community, with its emphasis on terpreted consistently. In addition, nearly 40 percent social spending, has helped put the public sector of the firms rated corruption as a major obstacle to back into the driver's seat in investment and has also doing business. led to the crowding out of lending to the private sec- tor. While the public sector has largely withdrawn from Recommendations. The government needs ownership and management of purely private opera- to accelerate regulatory and institutional reforms tions, the private sector's interest in investing in the aimed at improving and modernizing the business op- provision of social and infrastructure services has not erating environment. It also needs to ensure that these been fully exploited. reforms--which include drafting new business laws and reforming such key institutions as the Uganda Recommendations. The government needs Revenue Authority and the Business Registrar--are to pursue a policy of divesting and contracting out implemented consistently. The Ministry of Justice has services--a policy that would help reduce the public identified 44 key laws that have a significant impact administration budget while also leveraging private in- on the commercial legal system. These laws, which vestment. The Ministry of Public Service, in coopera- are only partially updated, need to be reviewed, tion with the privatization unit, needs to continue iden- processed, and implemented. The Ministry of Justice tifying and divesting services that can be contracted is also responsible for implementing the act governing out to the private sector. And sector ministries need to the Business Registrar (the Business Registrar's Act), establish an appropriate legal and regulatory frame- which still lacks the ministry's approval. While the work for private investment or public-private partner- Business Registrar is clearly in need of capacity build- ships in the provision of social services. The private ing, such efforts will be effective only if this act is fully sector, though already a major provider of services, approved and implemented. Beyond strengthening needs to evaluate the potential of private, for-profit the regulatory framework, it will be critical to improve operations and be more vocal about expanding its the attitude and work habits of the civil service, par- role. ticularly at the local government level, where capacity is more limited. Recently Uganda has introduced laws allowing Establishing a Low-Cost Operating local governments to collect fees, prompting local of- Environment ficials to develop a series of fees for the private sec- tor. Local administrations, applying these fees at their The survey data underscore the substantial regulatory discretion, have singled out larger investors for higher burden with which the private sector in Uganda must fees, regardless of exemptions granted under the in- cope. As noted in chapter 3, senior managers of man- vestment code. To help reduce the cost of doing busi- ufacturing firms spend substantial time--more than 13 ness in Uganda, regulatory agencies need to adopt a 5. Policy Implications 83 new role of enhancing the quality of services and ex- sound, and adequately serviced. In addition, there are panding outreach. a number of proposals to establish industrial zones In addition, the government should build on exist- near Kampala and Entebbe, such as the 980-hectare ing anticorruption policies, such as the requirement Kampala Industrial and Business Park at Namanve. that public officials declare their wealth. Other mea- While there has been long-standing interest in estab- sures to address corruption could include adopting lishing an export processing zone in Uganda, imple- additional anticorruption legislation, reforming public mentation has been delayed by a lack of consensus sector pay, providing adequate resources to anticor- on the development model, despite the Special ruption agencies, ensuring proper follow-up of find- Economic Zones Bill drafted in 2002. This draft bill is ings issued by commissions of inquiry, combating a being revised to take into account the needs and culture of impunity, strengthening accountability at all views of prospective investors. levels of government, and addressing corruption in lower levels of government. Further efforts need to be Recommendations. The government needs made in increasing transparency in fiscal policy and to establish a competitive investment environment that public procurement, such as in budget execution and is based on a transparent incentive structure and up- reporting, and in limiting the use of supplementary ap- dated legal framework for investment. Key priorities propriations by the executive branch. Arrears need to are to accelerate the commercial legal reforms started be properly assessed, and contingent liabilities moni- in 1999 and update the investment code. In addition, tored and controlled. And public procurement sys- such institutions as the Uganda Investment Authority, tems need to be improved. Since corruption is partly Export Promotion Board, and Tourism Board need to rooted in the political system, the private sector has a have their roles clarified and be set on a long-term, key role to play as a monitor and advocate, demand- sustainable path. The government has proposed ing minimum standards and assisting entrepreneurs merging these institutions, but this has not yet been in challenging abuses of institutions, systems, and done, making it difficult to proceed with any capacity regulations. building program. The Uganda Development Bank, now a major source of distortion in the formal incentive structure, needs to be transformed into an institution Establishing a Competitive Investment that allows broad development of the financial market. Environment And the export processing zone needs to be made fully operational to attract high-quality investment. Uganda competes with other locations around the world for private investment. Meanwhile, countries in Africa, Asia, and Europe are continuously improving Improving Tax Administration their investment environment with the aim of attracting domestic and foreign investment. Among the features Investors in Uganda consider the level of taxation the that tend to draw international investors are export second biggest constraint to doing business and also processing zones and related investment incentives. rank tax administration high among the burdens they To help attract investment to Uganda, the govern- face. The most common complaint is that the income ment has decided to set up an export processing tax law is inconsistently interpreted even within the zone, acknowledging the need for industrial land out- Uganda Revenue Authority, leading to contradictory side Kampala that is well planned, environmentally rulings, unpredictable value added tax refunds, and 5. Policy Implications 84 corporate tax bills that are inconsistent with the guide- business lending and trade finance. But the greater lines. Adding to the problem, the staff of the Uganda competition still has not led to a substantial decline in Revenue Authority have limited technical skills, and intermediation costs and thus interest rates or to an appealing tax rulings before the Tax Administration increase in lending to the private sector. High interest Tribunal is time consuming and ineffective. rates on treasury bills are having a crowding out ef- fect on lending to the private sector. The high-cost Recommendations. The Ministry of Finance, business environment results in high administrative Planning, and Economic Development needs to en- costs for loans. And firms have limited access to sure that tax laws are in effect that are clear, unam- long-term loans. Compared with firms in Kenya, those biguous, and consistent with the investment code. It in Uganda have less access to financial products also needs to ensure that tax policy is predictable, across the board. keeping to a minimum new tax measures that depart from the policy. While most entrepreneurs recognize Recommendations. The government needs the need for a sustainable revenue base, they believe to develop additional reform policies to support finan- that tax revenue is generated from a very small part of cial service providers and improve their ability to re- the private sector--larger firms in the formal sector. spond to the needs of the private sector. The reforms Thus the government needs to devise strategies for need to focus on such key areas as pensions, insur- widening the tax base by generating more revenue ance, and capital market development, to improve from small businesses and perhaps even the informal firms' access to long-term finance. Options to reform sector. In addition, the Uganda Revenue Authority the pension system also need to be considered. Also needs to transform itself into an efficient institution important is greater funding for commercial courts, with a reputation for integrity, an institution that en- which would enable these courts to function more ef- forces tax laws while remaining cognizant of the need ficiently and thus increase lenders' willingness to ex- to foster a productive private sector. The agency also tend credit. At the same time the government needs to needs to focus on running the value added tax refund provide incentives for better compliance with ac- system efficiently for the private sector while reducing counting standards by the private sector. Establishing fraud. a credit registry would give firms greater incentive to provide high-quality information on their operations and finances. Legislation to address related privacy Ensuring Sound Financial Market concerns would also need to be finalized. Development Firms in the survey ranked the cost of finance and Raising Firm Productivity access to finance among the top five constraints to their operation and growth. The government has The productivity of labor and the efficiency of wage completed key reforms in the financial sector, trans- determination both warrant attention from policymak- forming it from a vulnerable system with collapsing ers. Labor markets function inefficiently in part be- banks to a relatively sound one able to compensate cause of the low labor skills and low labor mobility. for banking losses without government intervention. The poor labor mobility may be due to difficulties Competition in the financial market has improved, workers have in demonstrating informally learned with commercial banks showing more interest in small skills to potential employers in different regions and to 5. Policy Implications 85 the close relationship between labor and land, which both the government and donors to revisit their priori- is often owned informally and without title. ties in this area. As the survey data show, Ugandan manufacturing To help improve the health of workers, the govern- firms use close to 60 percent of their capacity, and the ment and the private sector should consider using ex- value added per worker is just over $1,000 a year, in isting HIV/AIDS awareness programs to increase the same range as in other small Sub-Saharan African knowledge on how to control the disease. Enterprises economies. But Ugandan firms compare poorly in should be encouraged to take a more active role in productivity with firms in countries outside Africa, controlling the spread of HIV/AIDS--for example, by such as China and India. In addition, small firms in forming partnerships between business associations Uganda have lower capacity utilization and productiv- and voluntary counseling and testing programs. ity than large firms. Support aimed at increasing firm productivity should be focused especially on micro, small, and medium-size enterprises, since these Increasing the Efficiency of Services make up more than 90 percent of Ugandan firms and their growth is crucial to the creation of a broad- According to the survey results, difficulties in access based private sector and thus to long-term economic to electricity and to land are major constraints for development. Ugandan businesses. Limited and unreliable electric- A critical factor in firm productivity is the health of ity supply hampers the operation of businesses. workers. Firms lose up to 37 days of production per Transport costs are also a key constraint to doing worker annually because of health-related issues. business in Uganda. Survey respondents did not Moreover, firms appear to underestimate, ignore, or mention this constraint often, however, largely be- conceal the extent of the HIV/AIDS problem. cause many businesses have taken steps to over- come it by developing in-house transport services. Recommendations. The Ugandan govern- ment needs to support private sector­led skills devel- Recommendations. The government's re- opment and technology transfer initiatives aimed at in- structuring of the utilities sector, under way for several creasing productivity and capacity utilization. There years, needs to be completed as a matter of urgency are many ways to do this, including providing tax to address critical constraints to doing business. Also credits to firms that engage in worker training or adopt recommended is that the government strengthen the new technology. The government could also support regulatory framework, to facilitate private investment worker training and apprenticeship programs de- in the utilities sector. signed and implemented by the private sector to meet The government restructured the Uganda its needs. Electricity Board in March 2001 and transferred the The government should look for ways to encour- generation, transmission, and distribution functions to age entrepreneurship and improve access to busi- three separate companies. The government needs to ness education. For indigenous entrepreneurs, edu- ensure that these companies are fully operational and cation is among the most important factors in are attracting adequate private participation. Other determining firm performance. Access to university work is also under way in the utilities sector and in in- education may be particularly important in offsetting frastructure. The ministry responsible for water and the advantages of inherited ownership and family- sanitation is developing a strategy for private partici- based business knowledge. It may be worthwhile for pation based on appropriate regulation. The reforms, 5. Policy Implications 86 which have been under way for years, need to be fast- fraud, the government should consider decreas- tracked to allow further investment and expansion of ing the number of tariff rates from the three now in service. In air transport some infrastructure has been use to two or, in the long term, even to one. improved, but many issues still need to be addressed · The selective use of the excise tax for such to ensure adequate services meeting the needs of the products as tobacco and alcoholic beverages private sector. The Civil Aviation Authority needs to is probably justified, but there is little rationale focus on regulation and on encouraging private in- for using the tax for other products. Ending its vestment to improve the efficiency and effectiveness indiscriminate use would increase transpar- of air transport services. As suggested in the govern- ency, lessen distortions, and help further reduce ment's strategy for structural transformation, pre- protection. sented at the Consultative Group meeting of donors in · The government needs to continue its work in Kampala in April 2003, several measures need to be trade reform, lowering overall tariffs, promoting ef- taken immediately, including restructuring unsustain- ficient resource allocation, and keeping excise able debts in the utilities sector, providing entry points taxes and other forms of nontariff protection to a for private participation, and creating a multi-utility minimum. regulatory agency. The Role of the World Bank Addressing Distortions in Trade The World Bank is supporting many of the recommen- On the basis of the tariff schedule, Uganda's trade dations described above. In addition, on the basis of policy regime appears to be fairly liberal. But tariffs the results of the investment climate assessment, the are not the only tool used to protect domestic produc- Bank is proposing a $57 million International Devel- ers. Several nontariff ad valorem restrictions are still in opment Association credit to help promote private force, and the true level of protection is thus much sector development. Key components of the project higher than the tariff schedule suggests. As the wide include the following: range of effective rates of protection show, distortions persist. The biggest source of the distortions is the ex- · A matching grants program to address low pro- cise tax, used indiscriminately and showing disper- ductivity through a $13 million loan for skills de- sion across tariff lines. While the excise tax provides velopment and in-house training. needed revenue for financing public spending, the · A range of business development services to help government needs to be careful about distorting in- improve financial management and audits and centives in manufacturing. lower the risk of lending to the private sector. · Infrastructure provision through the development Recommendations. Uganda is now dis- of an export processing zone outside Kampala cussing new tariff rates with its partners in the East and a proper institutional and regulatory frame- African Community. It is also the right time to change work to support infrastructure services. other elements of the trade policy regime: · Mobilization of private participation in infrastruc- ture through a $1 million loan for the provision of · To improve allocative efficiency, as well as simplify technical assistance and assistance in building a tariff administration and reduce incentives for regulatory framework. 5. Policy Implications 87 · Provision of technical assistance to support insti- identified in this assessment. These grants, to be pro- tutional development aimed at increasing the vided on the basis of clear and transparent criteria, availability of term finance and other products will develop the capacity of Private Sector Foundation suitable for micro, small, and medium-size firms. Uganda and other business associations as well as the dialogue conducted by the Investors Round Table. Funding will also go to capacity building grants to Implementation of this World Bank project will be facilitate a sustainable dialogue between the World closely coordinated with other donor activities in Bank and the private sector on the key constraints Uganda. 5. Policy Implications 88 Annex 5.1 Policy Suggestions for Improving the Investment Climate in Uganda Policy objective Policy issues Observations Policy suggestions Maintaining Declining economic growth, Macroeconomic instability is the · Maintain a low administrative macro- rising public spending third most important constraint budget. economic (especially for public services), identified by Ugandan investors-- · Have the Ministry of Finance, stability inconsistent tax policies, and and the most important one Planning, and Economic poor tax administration are key identified by foreign-owned Development and the Ministry of investor concerns. companies and exporters. Public Service modernize the civil service. · Implement recommendations made in the 2003 Public Expenditure Review, including improving the process for setting sectoral budget ceilings, improving budget monitoring systems, and decentralizing certain services. Encouraging Private sector investment is There is a need to promote private · Have the Ministry of Public Service private provi- being crowded out. investment in the provision of and the privatization unit continue sion of social social and infrastructure services identifying and divesting services and by pursuing a policy of divesting that can be contracted out to the infrastructure and contracting out services. private sector. services · Have the sector ministries establish an appropriate legal and regulatory framework for private investment or public-private partnerships. Establishing The private sector must cope Uganda has a high-cost business · Complete the review and a low-cost with corruption and a sub- environment as a result of implementation of 44 commercial operating stantial regulatory burden. inadequate regulatory capacity, laws identified as key by the environment an unclear regulatory framework, Ministry of Justice. and inconsistent interpretation of · Complete the approval and policies and regulations. Nearly 40 implementation of the Business percent of manufacturing firms Registrar's Act. rated corruption as a major · Inculcate in the civil service, obstacle to doing business. particularly at the local government level, an attitude of valuing and serving the private sector. · Build on existing anticorruption policies by increasing the transparency of public procurement, limiting the use of supplementary appropriations by the executive branch, and monitoring contingent liabilities. Annex continued on next page 5. Policy Implications 89 Annex 5.1 Policy Suggestions for Improving the Investment Climate in Uganda (continued) Policy objective Policy issues Observations Policy suggestions Establishing a In competing with other The government needs to establish · Update the investment code. competitive locations around the world for an attractive investment environ- · Clarify the roles of such institutions investment private investment, Uganda is ment based on a transparent in- as the Uganda Investment Authority, environment handicapped by the poor centive structure and updated Export Promotion Board, and ratings of its investment legal framework for investment. Tourism Board and set them on a climate. long-term, sustainable path. · Transform the Uganda Development Bank into an institution that allows broad development of the financial market. · Fully implement the proposed export processing zone with assistance from the World Bank. Improving tax The income tax law is Investors in Uganda consider the · Have the Ministry of Finance, administration inconsistently interpreted even level of taxation the second big- Planning, and Economic within the Uganda Revenue gest constraint to doing business Development ensure that tax laws Authority, leading to contra- and also rank tax administration are clear, unambiguous, and dictory rulings, unpredictable high. consistent with the investment code. value added tax refunds, and · Keep to a minimum new tax corporate tax bills that are measures that are inconsistent with inconsistent with the guidelines. previous policies. Moreover, the staff of the · Widen the tax base by generating Uganda Revenue Authority more revenue from smaller have limited technical skills, businesses and perhaps the and appealing tax rulings before informal sector. the Tax Administration Tribunal · Transform the Uganda Revenue is time consuming and Authority into an efficient institution ineffective. with a reputation for integrity, an institution that both enforces tax laws and creates an enabling environment for the private sector. · Have the Uganda Revenue Autho- rity focus on running the value added tax refund system efficiently while reducing fraud. Ensuring sound High interest rates on treasury Key financial sector reforms have · Focus on such key areas as financial bills are having a crowding out led to a relatively sound banking pensions, insurance, and capital market effect on lending to the private system, but the improvements have market development. development sector, the high-cost business not yet resulted in a substantial de- · Increase funding for commercial environment results in high cline in costs, reduction in interest courts to enable them to function administrative costs for loans, rates, or increase in lending to the more efficiently. and access to long-term loans private sector. · Create incentives to encourage is limited. better compliance with accounting Annex continued on next page 5. Policy Implications 90 Annex 5.1 Policy Suggestions for Improving the Investment Climate in Uganda (continued) Policy objective Policy issues Observations Policy suggestions standards by firms and thus reduce information asymmetries in lending--such as by establishing a credit bureau. Raising firm Ugandan firms have poor The labor force has low skill levels · Support private sector­led skills productivity productivity compared with and poor health, and informal development and technology firms in countries outside (untitled) land ownership may transfer initiatives. Africa, such as China and restrict labor mobility. · Target support to micro, small, and India, and their capacity medium-size enterprises. utilization is low. · Formalize worker training and apprenticeship programs. · Encourage entrepreneurship and improve access to business education. · Use existing HIV/AIDS awareness programs to increase the knowledge on how to control the disease. · Encourage enterprises to take a more active role in controlling the spread of HIV/AIDS. · Encourage partnerships between business associations and voluntary counseling and testing programs. · Create property rights that are legally recognized and tradable. Increasing the The provision of utility and Electricity is a particularly limiting · Restructure unsustainable debts in efficiency of transport services is limited factor, with the limited and the utilities sector, provide entry services and unreliable. unreliable supply hampering the points for private participation, and operation of businesses. Many create a multi-utility regulatory businesses have taken steps to agency. overcome the transport constraint, · Establish industrial parks that are however, by developing in-house easily accessible, provide adequate transport services. utility services, and meet environ- mental standards. · Ensure that the restructured electricity companies are fully operational and attracting adequate private participation. · Have the ministry responsible for water and sanitation fast-track regulatory reforms to facilitate private participation in the sector. Annex continued on next page 5. Policy Implications 91 Annex 5.1 Policy Suggestions for Improving the Investment Climate in Uganda (continued) Policy objective Policy issues Observations Policy suggestions · Accelerate the railway privatization and improve the road network (at least the major economic routes). · Have the Civil Aviation Authority focus on regulation and on encouraging private participation. Addressing Nontariff trade barriers remain Excise taxes are creating · Continue the work in trade reform, distortions high. distortions in the economy. lowering overall tariffs, promoting in trade efficient resource allocation, and keeping excise taxes and other forms of nontariff protection to a minimum. Appendix 92 Appendix 1: The Sample 94 Table A2.6 Manufacturing Employees by Job Table A1.1 Universe and Sample Frame 94 Category, Firm Location, and Table A1.2 Sample Frame by Firm Size Class, Geographic Origin, Uganda 109 Sector, and Subsector 95 Table A2.7 Average Age of Employees in Table A1.3 Sample Frame by Sector and Selected Sectors by Firm Size Region 97 Class and Gender, Uganda 110 Table A1.4 Surveyed Sample by Firm Size Table A2.8 Employees' Experience in Class, Sector, and Region 98 Selected Sectors by Gender, Uganda 110 Table A1.5 Manufacturing Sample by Firm Size Class and Subsector 100 Table A2.9 Employees in Selected Sectors by Gender and Highest Level of Table A1.6 Manufacturing Sample by Region Education Achieved, Uganda 112 and Subsector 101 Table A2.10 Employees in Selected Sectors Appendix 2. Labor Market Features 102 by Highest Level of Education Figure A2.1 Geographic Distribution of Achieved, Uganda 112 Employment by Sector, Table A2.11 Estimates of Wage Determinants Uganda, 2001/02 105 in Selected Sectors, Uganda 114 Figure A2.2 Average Age of Manufacturing Table A2.12 Employees and HIV in Selected Employees by Gender and Job Sectors, Uganda 116 Category, Uganda 111 Table A2.13 Manufacturing Employees by Figure A2.3 Manufacturing Employees by Highest Method of Payment, Uganda 117 Level of Education Achieved, Selected African Countries 113 Table A2.14 Monthly Earnings of Unskilled Production Workers 118 Table A2.1 Structure of Formal Employment, Uganda, 2001/02 102 Appendix 3. Protection of Manufacturing 119 Table A2.2 Employment in Selected Sectors Figure A3.1 Simple Average Tariffs for by Firm Size Class and Subsector, Manufactured Commodities, Uganda, 2001/02 104 Uganda and Selected Country Groups, Selected Years, 1994­2002 122 Table A2.3 White-Collar Workers as a Share of Permanent Employment in Table A3.1 Applied Tariffs by Major Selected Sectors by Firm Size Commodity Group, Uganda, Class, Uganda 106 Selected Years, 1994­2002 120 Table A2.4 Change in Permanent Employment Table A3.2 Weighted Average Tariffs by in Selected Sectors, Uganda, Major Commodity Group, Uganda 1998­2002 107 and Selected Country Groups, Selected Years, 1994­2002 121 Table A2.5 Employees in All Sectors by Job Category and Geographic Table A3.3 Characteristics of the Most Distorted Origin, Uganda 108 Nominal Protection Coefficients by Product Category, Uganda, 2002 124 Appendix 93 Table A3.4 Firm-Level Nominal Protection Table A4.5 Sources of Finance for Coefficients for Outputs and Manufacturing Firms in Uganda, Raw Materials in Manufacturing, in International Comparison and Uganda, 2001/02 125 by Firm Characteristic 133 Table A3.5 Firm-Level Effective Rates of Table A4.6 Manufacturing Firms' Credit, Protection in Manufacturing by Loans, and Liabilities in Uganda, Subsector, Uganda, 2001/02 127 in International Comparison and Table A3.6 Firm-Level Effective Rates of by Firm Characteristic 134 Protection in Manufacturing by Table A4.7 Financial Sector and Property Type of Ownership and Subsector, Rights Indicators as Reported by Uganda, 2001/02 127 Manufacturing Firms in Uganda, in Appendix 4. Investment Climate Indicators 130 International Comparison and by Table A4.1 Structure of Manufacturing Firm Characteristic 135 Sample for Uganda Investment Table A4.8 Labor and Training in Manufacturing Climate Survey 130 in Uganda, in International Table A4.2 Manufacturing Firms' Competitors, Comparison and by Firm Suppliers, and Customers in Characteristic 136 International Comparison Uganda, Table A4.9 Regulatory Burden and by Firm Type 130 Administrative Delays as Reported Table A4.3 Manufacturing Firms' Evaluation by Manufacturing Firms in Uganda, of General Constraints to Operation in International Comparison and in International Comparison by Firm Characteristic 137 Uganda, by Firm Type 131 Table A4.10 Indicators of Uncertainty and Table A4.4 Infrastructure Performance as Corruption as Reported by Reported by Manufacturing Manufacturing Firms in Uganda, Firms in Uganda, in International in International Comparison and Comparison and by Firm by Firm Characteristic 138 Characteristic 132 Appendix 1: The Sample 94 The sample of Ugandan firms was designed as a Third, the sample frame was further refined by re- stratified random sample on the basis of recent cen- moving a few enterprises in relatively inaccessible sus data from the Uganda Bureau of Statistics. areas in northern Uganda. The final sample frame in- cluded 1,439 enterprises employing 80,589 people, or 96.9 percent of firms and 98.8 percent of employ- Devising the Sample Frame ment in the formal sector (table A1.1). For sampling purposes, the sample frame was The sample frame of the survey was devised in three stratified by location, manufacturing subsector, and steps. First, a list containing 4,698 officially registered size. To adequately reflect the geographic distribution companies with 10 or more employees was obtained of firms, three regions were defined on the basis of ex- from the Uganda Bureau of Statistics, which compiled isting districts: the central region, which includes the the list during a nationwide census in 2001/02. The list Kampala district; the northeast region; and the south- includes all activities in the formal sector, which had a west region. Similarly, to obtain the correct distribution total of 182,687 permanent employees. It also pro- of activity in manufacturing, nine broad subsectors vides the name, address, sector, International were defined using the ISIC: agro-industry; chemicals Standard Industrial Classification (ISIC) code, and and paints; construction materials; furniture; metals; number of employees for every formal manufacturing paper, printing, and publishing; plastics; textiles and firm existing in 2001/02. leather products; and wood. Finally, to allow com- Second, only firms in four sectors--commercial parisons with data sets compiled for other African agriculture, construction, manufacturing, and countries by the Regional Program on Enterprise tourism--were retained from this list. This selection of Development (RPED), three standard size classes sectors defined the universe of formal sector firms to were used: small (10­49 employees), medium-size be considered for the survey. This new list contained (50­99), and large (100 or more). Clusters were then 1,485 firms employing 81,566 workers. defined on the basis of the location, size, and sector Table A1.1 Universe and Sample Frame Universe Sample frame Sector Firms Employment Firms Employment Manufacturing 797 52,628 751 51,651 Percentage of universe 94.2 98.1 Construction 95 6,653 95 6,653 Percentage of universe 100.0 100.0 Tourism 479 10,453 479 10,453 Percentage of universe 100.0 100.0 Commercial agriculture 114 11,832 114 11,832 Percentage of universe 100.0 100.0 Total 1,485 81,566 1,439 80,589 Source: Uganda Bureau of Statistics census of firms, 2001/02. Appendix 1: The Sample 95 of firms. This three-level stratification defined 82 clus- sectors except tourism, where small firms make up ters, including 56 in manufacturing alone. about 94 percent of the total but provide about 69.5 The size distribution of firms in the sample frame is percent of employment. The tourism industry in uneven (table A1.2). Although small enterprises make Uganda is thus made up largely of small firms. up most of the firms in the sample frame (about 82 The importance of the central region is immedi- percent of the firms employ between 10 and 49 work- ately apparent (table A1.3). This region accounts for ers), they account for only 25.2 percent of the employ- about 64.6 percent of the firms and 62.4 percent of ment. The sample frame's 146 large firms (with 100 or the employment. It dominates in all sectors ex- more employees) provide most of the employment cept commercial agriculture, located mainly in the (66.4 percent). This characteristic is shared across all southwest. Table A1.2 Sample Frame by Firm Size Class, Sector, and Subsector Small Medium-size Large (10­49 (50­99 (100+ Sector and subsector employees) employees) employees) Total Manufacturing Agro-industry Firms 227 22 51 300 Employment 4,017 1,403 29,716 35,136 Chemicals and paints Firms 27 7 4 38 Employment 562 445 797 1,804 Construction materials Firms 45 4 3 52 Employment 721 306 743 1,770 Furniture Firms 118 3 1 122 Employment 1,708 175 135 2,018 Metals Firms 67 10 8 85 Employment 1,128 670 1,494 3,292 Paper, printing, and publishing Firms 42 2 4 48 Employment 917 150 1,026 2,093 Plastics Firms 13 6 2 21 Employment 261 367 330 958 Textiles and leather products Firms 44 5 10 59 Employment 844 299 2,647 3,790 Wood Firms 23 2 1 26 Employment 444 108 238 790 All subsectors Firms 606 61 84 751 Employment 10,602 3,923 37,126 51,651 (Table continues on next page) Appendix 1: The Sample 96 Table A1.2 Sample Frame by Firm Size Class, Sector, and Subsector (continued) Small Medium-size Large (10­49 (50­99 (100+ Sector and subsector employees) employees) employees) Total Tourism Hotels Firms 436 16 11 463 Employment 7,050 1,073 1,995 10,118 Transport Firms 14 2 0 16 Employment 217 118 0 335 All subsectors Firms 450 18 11 479 Employment 7,267 1,191 1,995 10,453 Construction Building completion Firms 6 0 0 6 Employment 145 0 0 145 Building installation Firms 8 0 0 8 Employment 117 0 0 117 Complete construction Firms 53 14 14 81 Employment 1,017 909 4,465 6,391 All subsectors Firms 67 14 14 95 Employment 1,279 909 4,465 6,653 Commercial agriculture Cereals Firms 8 0 2 10 Employment 200 0 255 455 Crops, animal husbandry Firms 40 0 0 40 Employment 605 0 0 605 Vegetables, nursery products Firms 4 1 17 22 Employment 75 60 3,857 3,992 Other animal farming Firms 8 0 0 8 Employment 133 0 0 133 Fruits, nuts, beverages Firms 7 9 18 34 Employment 174 692 5,781 6,647 All subsectors Firms 67 10 37 114 Employment 1,187 752 9,893 11,832 Full sample frame Firms 1,190 103 146 1,439 Employment 20,335 6,775 53,479 80,589 Source: Uganda Bureau of Statistics census of firms, 2001/02. Appendix 1: The Sample 97 Table A1.3 Sample Frame by Sector and Region (percent) Commercial Full sample Region Manufacturing agriculture Tourism Construction frame Firms Central 72.0 34.9 69.0 85.3 64.6 Northeast 12.9 2.6 15.7 8.0 15.3 Southwest 15.0 62.5 15.2 6.6 20.1 Employment Central 64.4 34.9 69.0 85.3 62.4 Northeast 4.6 2.6 15.7 8.0 6.0 Southwest 31.0 62.5 15.2 6.6 31.5 Source: Uganda Bureau of Statistics census of firms, 2001/02. Selecting the Sample placements from the sample frame were required to be similar in size and to operate in the same sector Following the stratification of the sample frame, firms and the same region. were selected randomly in each cluster. A total sam- The manufacturing subsample is the largest com- ple of 410 firms was drawn, with an overall sampling ponent of the overall sample. It includes 300 firms, or rate of 28.5 percent. about 40 percent of the existing manufacturing firms The surveyed sample differs slightly from the in the sample frame, and accounts for about 75 per- drawn sample because of some replacement sam- cent of the manufacturing employment. This subsam- pling, but it retains the key characteristics: the impor- ple reflects the dominance of the central region and tance of firms with 100 or more employees and the large firms (tables A1.5 and A1.6). But it differs slightly prominence of the central region (table A1.4). The sur- from the sample frame in its structure. In the subsam- veyed sample includes 392 firms (a sampling rate of ple 68 percent of firms are in the central region (com- 27.2 percent). Firms with 100 or more workers ac- pared with 72 percent in the frame) and 19.7 percent count for about 86 percent of the employment in the of firms have 100 or more employees (compared with sample, and the central region accounts for 66 per- 10.6 percent in the frame). cent of the enterprises and about 66.5 percent of the In short, the overall distribution of employment in employment. the surveyed sample is correct with respect to the lo- The difference between the theoretical and the cation of firms, but the "fit" is imperfect for the size dis- surveyed sample is explained by several factors: tribution because it overstates the importance of firms some firms refused to be interviewed, several firms with 100 or more employees compared with the frame. changed size class (the survey was undertaken about Nonetheless, the surveyed sample takes into account a year after the completion of the census), and some the dominance of large firms and the importance of firms either did not exist or had changed their activity. the central region. And although the sampling rate is Some of these "missing" firms were replaced with only about 27 percent, it still accounts for about 67 firms with similar or identical characteristics. These re- percent of officially recorded employment. Appendix 1: The Sample 98 Table A1.4 Surveyed Sample by Firm Size Class, Sector, and Region Medium-size Sector Micro (<10 Small (10­49 (50­99 Large (100+ and region employees) employees) employees) employees) Total Manufacturing Central Firms 35 98 25 46 204 Employment 199 2,152 1,678 24,060 28,089 Northeast Firms 5 30 3 7 45 Employment 26 607 208 1,503 2,344 Southwest Firms 14 25 6 6 51 Employment 79 423 434 7,819 8,755 All regions Firms 54 153 34 59 300 Employment 304 3,182 2,320 33,382 39,188 Tourism Central Firms 2 7 6 7 22 Employment 12 142 465 1,458 2,077 Northeast Firms 1 0 1 0 2 Employment 9 0 55 0 64 Southwest Firms 1 2 1 0 4 Employment 9 42 50 0 101 All regions Firms 4 9 8 7 28 Employment 30 184 570 1,458 2,242 Construction Central Firms 2 7 0 7 16 Employment 5 227 0 2,777 3,009 Southwest Firms 0 2 1 0 3 Employment 0 43 60 0 103 All regions Firms 2 9 1 7 19 Employment 5 270 60 2,777 3,112 (Table continues on next page) Appendix 1: The Sample 99 Table A1.4 Surveyed Sample by Firm Size Class, Sector, and Region (continued) Medium-size Sector Micro (<10 Small (10­49 (50­99 Large (100+ and region employees) employees) employees) employees) Total Commercial agriculture Central Firms 1 8 0 8 17 Employment 6 173 0 2,931 3,110 Northeast Firms 1 0 0 1 2 Employment 5 0 0 765 770 Southwest Firms 5 10 2 9 26 Employment 31 222 150 5,796 6,199 All regions Firms 7 18 2 18 45 Employment 42 395 150 9,492 10,079 Full sample Firms 67 189 45 91 392 Employment 381 4,031 3,100 47,109 54,621 Source: Uganda Bureau of Statistics census of firms, 2001/02. Appendix 1: The Sample 100 Table A1.5 Manufacturing Sample by Firm Size Class and Subsector Micro Small Medium-size Large (<10 (10­49 (50­99 (100+ Subsector employees) employees) employees) employees) Total Agro-industry Firms 23 54 12 33 122 Employment 123 1,204 897 21,821 24,045 Average employees/firm 5.3 22.3 74.8 661.2 197.1 (2.2) (9.7) (12.3) (1,219.6) (688.7) Chemicals and paints Firms 2 8 3 5 18 Employment 10 141 171 803 1,125 Average employees/firm 5.0 17.6 57.0 160.6 62.5 (1.4) (7.9) (5.0) (80.7) (75.7) Construction materials Firms 3 24 7 6 40 Employment 6 458 461 1,492 2,417 Average employees/firm 2.0 19.1 65.9 248.7 60.4 (1.7) (8.9) (14.8) (158.7) (100.4) Furniture Firms 19 24 3 1 47 Employment 124 410 179 245 958 Average employees/firm 6.5 17.1 59.7 245.0 20.4 (1.3) (9.4) (7.6) (36.5) Metals Firms 3 12 1 5 21 Employment 18 263 70 7,102 7,453 Average employees/firm 6.0 21.9 70.0 1,420.4 354.9 (1.0) (7.8) (2,839.8) (1,409.1) Paper, printing, and publishing Firms 2 12 5 4 23 Employment 14 288 326 790 1,418 Average employees/firm 7.0 24.0 65.2 197.5 61.7 (2.8) (12.2) (9.7) (122.9) (81.0) Plastics Firms 0 5 1 1 7 Employment 0 116 65 100 281 Average employees/firm n.a. 23.2 65.0 100.0 40.1 (13.7) (32.6) Textiles and leather products Firms 1 10 1 3 15 Employment 6 217 65 729 1,017 Average employees/firm 6.0 21.7 65.0 243.0 67.8 (12.2) (27.9) (92.6) Wood Firms 1 4 1 1 7 Employment 3 85 86 300 474 Average employees/firm 3.0 21.3 86.0 300.0 67.7 (10.3) (106.0) Full sample Firms 54 153 34 59 300 Employment 304 3,182 2,320 33,382 39,188 Average employees/firm 5.6 20.8 68.2 565.8 130.6 (2.0) (9.9) (12.4) (1,222.5) (580.3) n.a. Not applicable. Note: Figures in parentheses are standard deviations. Source: World Bank, Investment Climate survey, Uganda, 2002/03; authors' calculations. Appendix 1: The Sample 101 Table A1.6 Manufacturing Sample by Region and Subsector Central Northeast Southwest Subsector region region region Total Agro-industry Firms 82 19 21 122 Employment 14,727 1,493 7,825 24,045 Average employees/firm 179.6 78.6 372.6 197.1 (578.2) (103.2) (1,207.9) (688.7) Chemicals and paints Firms 16 1 1 18 Employment 984 10 131 1,125 Average employees/firm 61.5 10.0 131.0 62.5 (77.5) (75.7) Construction materials Firms 21 10 9 40 Employment 1,722 240 455 2,417 Average employees/firm 82.0 24.0 50.6 60.4 (129.7) (13.8) (62.1) (100.4) Furniture Firms 29 7 11 47 Employment 711 101 146 958 Average employees/firm 24.5 14.4 13.3 20.4 (44.7) (9.3) (18.5) (36.5) Metals Firms 16 3 2 21 Employment 7,300 131 22 7,453 Average employees/firm 456.3 43.7 11.0 354.9 (1,612.8) (49.0) (5.7) (1,409.1) Paper, printing, and publishing Firms 21 0 2 23 Employment 1,395 0 23 1,418 Average employees/firm 66.4 n.a. 11.5 61.7 (83.3) (3.5) (81.0) Plastics Firms 7 0 0 7 Employment 281 0 0 281 Average employees/firm 40.1 n.a. n.a. 40.1 (32.6) (32.6) Textiles and leather products Firms 8 4 3 15 Employment 626 359 32 1,017 Average employees/firm 78.3 89.8 10.7 67.8 (93.4) (124.3) (4.2) (92.6) Wood Firms 4 1 2 7 Employment 343 10 121 474 Average employees/firm 85.8 10.0 60.5 67.7 (143.1) (36.1) (106.0) Full sample Firms 204 45 51 300 Employment 28,089 2,344 8,755 39,188 Average employees/firm 137.7 52.1 171.7 130.6 (584.5) (80.6) (783.3) (580.3) n.a. Not applicable. Note: Figures in parentheses are standard deviations. Source: World Bank, Investment Climate survey, Uganda, 2002/03; authors' calculations. Appendix 2: Labor Market Features 102 The Ugandan society is overwhelmingly rural. Only mains uneven across regions and sectors. In addi- 11 percent of Ugandans live in urban areas, and 40 tion, while overall employment increased slightly be- percent of these live in Kampala (EIU 2002). The tween 1998 and 2002, employment in manufacturing Ugandan labor force consisted of roughly 10.9 million declined. people in 1999, or about 49.1 percent of the total The formal labor market employs almost 183,000 population (World Bank 2001). Kampala is the coun- people in different sectors of activity (table A2.1).1 try's main economic and industrial center, and this Manufacturing, the largest sector, accounts for about area and its immediate surroundings account for 28.8 percent of employment, followed by education about two-thirds of firms. The formal labor market is (15.1 percent) and business services (14.4 percent). therefore only a small fraction of the total. Using both All other sectors each employ less than 10 percent of the detailed firm-level data from the 2002/03 Regional workers. Women account for almost a third of total em- Program on Enterprise Development (RPED) survey ployment, but this share varies greatly across sectors. and the data from the Uganda Bureau of Statistics While women account for only 15.7 percent of em- firm census in 2001/02, this appendix examines re- ployment in transport, they account for 61.5 percent in cent patterns of and changes in employment. The the health sector, where female employment is more data suggest that the distribution of employment re- traditional. Table A2.1 Structure of Formal Employment, Uganda, 2001/02 Share of total Female employment employment as a share Sector Firms Employment (percent) of total (percent) Manufacturing 797 52,628 28.81 20.1 Education 1,214 27,590 15.10 47.1 Business services 491 26,236 14.36 20.1 Trade 951 18,092 9.90 21.5 Commercial agriculture 114 11,832 6.48 53.0 Tourism (hotels, tour operators) 479 10,453 5.72 47.9 Construction 95 6,653 3.64 8.2 Health 154 6,481 3.55 61.5 Communications 17 5,338 2.92 72.0 Finance 100 4,527 2.48 43.8 Transport 105 4,337 2.37 15.7 Utilities 14 3,424 1.87 17.7 Other agriculture 85 1,698 0.93 48.0 Personal services 33 988 0.54 24.1 Mining 13 814 0.45 16.3 Insurance 18 799 0.44 57.6 Nongovernmental organizations 18 797 0.44 26.7 Total 4,698 182,687 100.00 31.0 Source: Uganda Bureau of Statistics census of firms, 2001/02. Appendix 2: Labor Market Features 103 Structure of Employment in Selected 69 percent of employment in the sector. But unlike Sectors other sectors, tourism is dominated by small firms, which account for about 93 percent of the sector's en- Agro-industry dominates manufacturing employment terprises and about 69 percent of its employment. in Uganda, accounting for about 67 percent of the Overall, employment in these four sectors exhibits total. The share of other subsectors ranges between the usual concentration patterns with two exceptions: 1.5 percent and 7.5 percent (table A2.2). Some 72 the subsectoral concentration in manufacturing is un- percent of manufacturing firms are located in central usually large, and the employment pattern by firm size Uganda (including the Kampala district) and account class in tourism differs from the usual linear pattern. for about 63 percent of manufacturing employment Another interesting feature of the distribution of (figure A2.1). Small firms (employing 10­49 workers) employment in the sample is the average share of are substantial in number, but account for only about management and professionals (white-collar workers) 21 percent of manufacturing employment. By con- in firms' workforces. It is often argued that the com- trast, the 85 large enterprises (100 or more workers), paratively high cost of labor in Africa may be attribut- though only 10.6 percent of all manufacturing firms, able to an excess of white-collar workers. RPED man- account for 70.7 percent of employment. ufacturing surveys in Africa in the 1990s measured the Manufacturing firms in Uganda have a geo- share of white-collar workers at 20­30 percent. graphic and size concentration similar to that in the Uganda is a bit of an outlier: white-collar workers rest of Africa. But the extent of the dominance of agro- account for 34.3 percent of the employees in the en- industry is unusual. The largest subsectors in other tire sample (table A2.3). But this large share encom- countries surveyed typically account for up to 40­45 passes wide variation among sectors. The smallest percent of manufacturing employment. shares are in tourism (15.7 percent) and commercial As in manufacturing, the distribution of employ- agriculture (18.8 percent). The shares are much larger ment by firm size class is almost linear in other sec- in construction (31.8 percent) and manufacturing tors, such as construction and commercial agricul- (38.1 percent). Indeed, the share in manufacturing is ture: the larger the firm, the larger the share of sector the largest in the RPED manufacturing surveys employment. In construction, for example, census (Nigeria has the next largest share, with 37.1 percent data suggest that the many small firms employ 19 per- in 2001). But as in other African countries, the share of cent of the workers in the sector, while the few large white-collar workers in manufacturing tends to de- ones employ 67 percent. crease as the size of the firm increases. Although employment patterns by firm size class The share of white-collar workers also varies are similar in construction and commercial agriculture, among manufacturing subsectors--ranging from 25.5 the geographic distribution of employment differs. percent in the textile and leather products subsector Firms in construction are overwhelmingly located in to 49.5 percent in the furniture subsector. There are central Uganda, which accounts for 85.3 percent of several competing explanations for this pattern, in- employment in this sector. But most firms in commer- cluding structural differences and variations in pro- cial agriculture are in the southwest, which accounts duction technologies among subsectors. In addition, for about 62.5 percent of employment in that sector. white-collar workers are a kind of overhead cost that Employment in tourism is also highly concentrated can be spread over the greater number of employees geographically, with the central region accounting for in larger firms. Appendix 2: Labor Market Features 104 Table A2.2 Employment in Selected Sectors by Firm Size Class and Subsector, Uganda, 2001/02 Small Medium-size Large (10­49 (50­99 (100+ Sector and subsector employees) employees) employees) Total Manufacturing Agro-industry 4,223 1,521 29,716 35,460 Chemicals and paints 562 445 797 1,804 Construction materials 736 369 743 1,848 Furniture 1,902 175 135 2,212 Metals 1,249 730 1,494 3,473 Paper, printing, and publishing 927 150 1,026 2,103 Plastics 261 367 330 958 Textiles and leather products 914 299 2,767 3,980 Wood 444 108 238 790 Total 11,218 4,164 37,246 52,628 Tourism Hotels 7,050 1,073 1,995 10,118 Transport 217 118 0 335 Total 7,267 1,191 1,995 10,453 Construction Building completion 145 0 0 145 Building installation 117 0 0 117 Complete construction 1,017 909 4,465 6,391 Total 1,279 909 4,465 6,653 Commercial agriculture Cereals 200 0 255 455 Crops, animal husbandry 605 0 0 605 Vegetables, nursery products 75 60 3,857 3,992 Other animal farming 133 0 0 133 Fruits, nuts, beverages 174 692 5,781 6,647 Total 1,187 752 9,893 11,832 All sectors 20,951 7,016 53,599 81,566 Source: Uganda Bureau of Statistics census of firms, 2001/02. Appendix 2: Labor Market Features 105 Figure A2.1 Geographic Distribution of Employment by Sector, Uganda, 2001/02 (percent) 100 6.6 15.2 90 8.0 30.7 80 15.7 70 62.5 6.0 60 50 85.3 40 2.6 69.0 30 63.2 20 34.9 10 0 Manufacturing Commercial Tourism Construction agriculture Central region Northeast region Southwest region Source: Uganda Bureau of Statistics census of firms, 2001/02. Change in Employment, 1998­2002 25.3 percent. By contrast, other subsectors experi- enced moderate or even large growth in permanent Employment data for 1998, 2000, and 2002 are avail- employment. (The 100 percent increase in employ- able only for firms that existed throughout the entire ment in the wood subsector in 1998­2002 should be period and thus exclude firms entering or exiting dur- viewed with caution, since the sample includes only ing this time. In the overall sample permanent em- seven firms in this subsector.) Small and medium-size ployment grew from 30,104 in 1998 to 31,056 in 2002 firms saw permanent employment increase, while for firms that existed over the entire period (table micro and large firms experienced significant de- A2.4). But as performance varied among sectors, a clines. significant reallocation of labor occurred. While per- manent employment boomed in commercial agricul- ture (increasing by 23.4 percent in 1998­2000 and by Worker Characteristics 19.1 percent in 2000­02) and grew significantly in construction and tourism in 1998­2002, it declined in Labor productivity depends largely on the character- manufacturing. istics of workers, such as their age, education level, In manufacturing as a whole permanent employ- and years of experience. The firm-level survey data ment declined by 8.6 percent in 1998­2002. Firms in suggest that the Ugandan labor force is fairly experi- agro-industry and metals were hit hardest, with de- enced and reasonably well educated but faces health clines in permanent employment of 13.3 percent and issues (including HIV/AIDS) that will need to be Appendix 2: Labor Market Features 106 Table A2.3 White-Collar Workers as a Share of Permanent Employment in Selected Sectors by Firm Size Class, Uganda (percent) Micro Small Medium-size Large All (<10 (10­49 (50­99 (100+ firm size Sector and subsector employees) employees) employees) employees) classes Commercial agriculture 35.0 25.4 100.0 5.6 18.8 Construction 80.0 20.6 30.0 38.9 31.8 Manufacturing 59.2 37.2 29.4 26.3 38.1 Agro-industry 51.0 41.3 38.8 28.2 39.2 Chemicals and paints 33.3 54.3 19.7 13.1 34.8 Construction materials 100.0 38.6 23.5 15.1 36.9 Furniture 68.1 35.6 49.2 35.6 49.5 Metals 34.8 33.7 -- 32.2 33.5 Paper, printing, and publishing 22.5 31.7 16.8 35.6 28.3 Plastics -- 34.0 11.1 32.0 30.4 Textiles and leather products 100.0 18.8 23.3 22.9 25.5 Wood 100.0 30.2 13.6 -- 39.1 Tourism 22.2 13.1 18.6 13.0 15.7 Full sample 56.6 34.5 29.0 21.9 34.3 -- Not available. Source: World Bank, Investment Climate survey, Uganda, 2002/03. addressed to improve productivity. (For data on em- This is hardly surprising, because most formal enter- ployees and HIV in all sectors, see table A2.12 at the prises are also located in this region. This suggests end of this appendix.) This section draws on data from that internal migration to find employment is not wide- interviews in early 2003 of a subsample of 1,803 em- spread; workers tend to find work close to home. ployees, mainly in manufacturing but also in commer- Within each region, the majority of workers in manu- cial agriculture, construction, and tourism.2 facturing also originate from that region (table A2.6). For manufacturing firms in the central region 52.4 per- Origin cent of workers originate from that region. For firms in Most of the labor force in manufacturing comes from the northeast region, the share is about 70 percent, two regions in Uganda--central (36.4 percent) and and for those in the southwest region, it rises to 84 eastern (20.9 percent; table A2.5). Workers originat- percent. The smaller share of manufacturing workers ing from the central region account for about 34­40 with local origins in the central region suggests that percent of total employment in the firms surveyed. (continued on page 108) Appendix 2: Labor Market Features 107 cent 3.16 47.00 13.83 ­8.63 51.30 34.81 27.30 16.95 14.56 56.40 18.30 66.89 30.91 ­13.32 ­25.34 100.00 ­17.83 ­12.92 Per 1998­2002 83 15 otal 126 198 401 118 163 168 ­23 235 491 340 952 T 2,459 ­1,973 ­2,094 ­1,025 ­2,676 Change 584 387 814 118 452 336 106 2002 7,691 1,037 1,553 3,021 1,517 1,225 1,440 20,888 13,623 18,040 31,056 manent Per employment cent 2.88 5.71 6.98 19.13 20.66 22.38 13.82 15.69 37.39 21.74 12.49 ­2.96 ­5.62 ­12.81 ­19.20 ­14.32 ­10.17 ­15.43 1998­2002 2000­02 Per 29 47 44 16 60 99 otal 100 284 123 ­12 136 ­44 Change T 1,235 ­505 ­3,069 ­3,238 ­3,292 ­1,849 Uganda, 484 340 770 102 329 276 118 2000 6,456 1,008 1,269 3,526 1,418 1,089 1,484 Sectors, 23,957 16,861 21,332 32,905 manent Per employment Selected in cent 4.79 7.28 2.97 9.30 23.39 10.65 25.39 10.16 11.84 10.63 ­0.97 13.84 64.29 ­8.53 10.58 48.37 34.91 ­12.86 2002/03. Per 1998­2000 97 98 36 74 ­1 40 otal 117 108 ­11 136 355 616 384 T 1,224 1,096 1,144 ­520 2,801 Uganda,, Change Employment survey 911 386 304 696 103 289 168 129 734 1998 5,232 1,152 4,046 1,282 1,100 22,861 15,717 20,716 30,104 manent manent Climate Per Per employment e in oducts publishing Investment pr , Change (employees) class agricultur paints and Bank, y materials leather (50­99) and size cial class orld A2.4 subsector , m e and W printing,, (<10) (10­49) (100+) fir size o-industr o ce: nitur ableT ood m Sector and Commer Construction Manufacturing Subsector Agr Chemicals Construction Fur Metals Paper Plastics extilesT ourism W Fir Micr Small Medium-size Large T otalT Sour Appendix 2: Labor Market Features 108 Table A2.5 Employees in All Sectors by Job Category and Geographic Origin, Uganda (percent) Skilled Unskilled Non- production production production All job Geographic origin Management Professionals workers workers workers categories Northern Uganda 4.67 5.33 9.11 4.49 11.21 6.81 Eastern Uganda 15.42 12.43 27.08 19.39 27.41 20.87 Western Uganda 17.52 18.93 14.06 17.35 17.76 16.91 Southwestern Uganda 6.78 7.69 10.68 16.94 8.41 10.77 Central Uganda 35.75 36.69 34.38 40.00 33.96 36.38 Europe 1.87 2.37 0.78 0.41 0.31 1.00 Asia 10.98 14.20 1.56 1.22 0.31 4.69 Other East Africa 3.04 1.18 1.04 0.00 0.62 1.17 Other Africa 2.34 1.18 1.30 0.20 0.00 1.00 Other 1.64 0.00 0.00 0.00 0.00 0.39 Source: World Bank, Investment Climate survey, Uganda, 2002/03. internal migration, when it occurs, is directed mainly firms. Small firms appear to serve as an entry point for toward that region. newcomers who cannot obtain jobs in larger firms with more stringent skill requirements. Age Workers in the firm sample are on average about 32.5 Tenure years old (table A2.7). Workers in tourism tend to be The tenure of workers (the length of time they have the youngest (about 29.6 years), and those in con- been employed at their firm) averages about 5.2 struction the oldest (37.3 years). Women in the sam- years across all sectors (table A2.8). Average tenure ple are slightly younger than their male colleagues is longest in construction (7.7 years) and shortest in (about 30 years old, compared with 33 years for men). tourism (3.7 years). Female workers have a shorter The labor force in manufacturing (with an average average tenure (4.7 years) than their male colleagues age of 32.5 years) is slightly younger than that in (5.3 years). Kenya (36 years). Not surprisingly, the oldest employ- In manufacturing the average tenure is about 5.2 ees in manufacturing are managers (with an average years. Managers stayed with their firm the longest (7 age of 37.8 years), while most staff in other positions years), and unskilled workers the shortest time (4.3 are younger (figure A2.2). Thus age and employment years). Firms in the plastics and wood subsectors rank appear to be correlated in the Ugandan sample, tend to keep their employees longer than do those in similar to the findings of other RPED manufacturing other subsectors, with average tenures of 7.2 and 6.3 surveys in Africa. years. Prior experience in other manufacturing firms is Younger workers tend to be concentrated in micro also significant. Employees had an average of about firms, and older ones in larger firms (see table A2.7). 3.9 years' experience in self-employment or employ- Moreover, there is an almost perfect linear relationship ment at another company before joining their present between the age of workers and the size class of firm. Appendix 2: Labor Market Features 109 Table A2.6 Manufacturing Employees by Job Category, Firm Location, and Geographic Origin, Uganda (percent) Employee's Skilled Unskilled Non- Geographic production production production All job Firm location origin Management Professionals workers workers workers categories Central region Northern Uganda 0.86 2.70 8.53 3.73 13.50 5.64 Eastern Uganda 9.91 9.01 23.22 17.43 24.54 17.12 Western Uganda 10.34 18.02 9.95 8.71 14.11 11.38 Southwestern Uganda 1.29 8.11 6.16 4.15 1.23 3.86 Central Uganda 52.59 42.34 47.87 65.56 45.40 52.40 Europe 1.72 1.80 0.95 0.41 0.61 1.04 Asia 14.22 16.22 1.42 0.00 0.00 5.64 Other East Africa 4.74 0.90 0.95 0.00 0.61 1.57 Other Africa 2.59 0.90 0.95 0.00 0.00 0.94 Other 1.72 0.00 0.00 0.00 0.00 0.42 Northeast region Northern Uganda 5.26 11.11 10.94 8.06 21.21 10.22 Eastern Uganda 57.89 33.33 60.94 70.97 51.52 60.44 Western Uganda 5.26 11.11 1.56 1.61 6.06 3.56 Southwestern Uganda 0.00 0.00 3.13 0.00 3.03 1.33 Central Uganda 17.54 33.33 18.75 19.35 18.18 19.11 Asia 12.28 11.11 3.13 0.00 0.00 4.44 Other Africa 1.75 0.00 1.56 0.00 0.00 0.89 Southwest region Northern Uganda 6.00 0.00 2.04 4.72 2.56 3.95 Eastern Uganda 6.00 11.11 4.08 1.89 5.13 3.95 Western Uganda 38.00 33.33 42.86 45.28 46.15 43.08 Southwestern Uganda 40.00 33.33 44.90 40.57 41.03 41.11 Central Uganda 6.00 11.11 0.00 6.60 5.13 5.14 Europe 0.00 0.00 2.04 0.00 0.00 0.40 Asia 2.00 11.11 2.04 0.00 0.00 1.19 Other East Africa 0.00 0.00 2.04 0.00 0.00 0.40 Other Africa 0.00 0.00 0.00 0.94 0.00 0.40 Other 2.00 0.00 0.00 0.00 0.00 0.40 Source: World Bank, Investment Climate survey, Uganda, 2002/03. Appendix 2: Labor Market Features 110 Table A2.7 Average Age of Employees in Selected Sectors by Firm Size Class and Gender, Uganda (years) Small Medium-size Large All firm Micro (<10 (10­49 (50­99 (100+ size Sector and gender employees) employees) employees) employees) classes Commercial agriculture Male 37.6 36.8 31.3 33.1 34.7 Female 30.0 27.7 28.0 33.1 30.8 All employees 35.9 34.6 30.9 33.1 33.8 Construction Male 42.0 37.5 -- 39.3 38.5 Female -- 32.1 -- 34.7 33.6 All employees 42.0 36.4 -- 38.1 37.3 Manufacturing Male 29.3 31.8 34.5 36.5 33.0 Female 28.3 29.6 30.6 31.3 30.3 All employees 29.2 31.4 33.5 35.3 32.5 Tourism Male 34.5 29.9 32.0 32.6 31.9 Female 26.5 25.7 24.2 28.2 26.0 All employees 31.5 28.2 29.1 31.0 29.6 Full sample 30.1 31.6 32.5 34.8 32.5 -- Not available. Source: World Bank, Investment Climate survey, Uganda, 2002/03. Table A2.8 Employees' Experience in Selected Sectors by Gender, Uganda (years) Tenure Prior experience All All Sector Male Female employees Male Female employees Commercial agriculture 5.5 5.4 5.5 5.4 1.2 4.5 Construction 7.8 7.6 7.7 5.6 2.5 4.9 Manufacturing 5.3 4.9 5.2 4.1 2.6 3.9 Tourism 4.2 2.8 3.7 4.1 1.8 3.3 Full sample 5.3 4.7 5.2 4.3 2.3 3.9 Source: World Bank, Investment Climate survey, Uganda, 2002/03. Appendix 2: Labor Market Features 111 Figure A2.2 Average Age of Manufacturing Employees by Gender and Job Category, Uganda (years) 40.0 35.0 30.0 25.0 20.0 15.0 10.0 5.0 0.0 Management Professionals Skilled Unskilled Nonproduction production production workers workers workers Male Female Source: World Bank, Investment Climate survey, Uganda, 2002/03. Education and Training ers interviewed reported having no education (table The links between economic growth and education A2.9). For about 19 percent of the workers the highest level are well understood. The externalities generated level achieved is primary education, and for 31 per- by education (greater adaptability, easier learning-by- cent, secondary education. The predominance of sec- doing) increase the stock of human capital, resulting ondary education in the labor force is similar to that in in higher productivity and growth. At independence other Sub-Saharan African countries surveyed. About Uganda had one of the best education systems in 16 percent of the workers have completed university Sub-Saharan Africa. Because of neglect during the education. While gender differences exist, they are Amin era and subsequent economic difficulties, many smaller than those in other Sub-Saharan African coun- parts of the system are now in poor condition. tries surveyed?. Interestingly, many more female work- Nonetheless, education is still highly valued, and stan- ers have completed vocational or technical training dards have not slipped as badly as might be ex- (34.4 percent) than male workers have (28.6 percent). pected (EIU 2002). In general, the government's pol- The education level of workers in Uganda varies icy is to shift funds from secondary and tertiary to widely across sectors and firm size classes. The con- primary education. The authorities have also empha- struction industry employs the most educated work- sized the need to achieve gender equality. The survey ers. Almost 45 percent of the workers in this sector data reflect these trends. have a university or professional degree (table A2.10). Overall, the Ugandan labor force is fairly well edu- For other sectors this share is between 14.8 percent cated. Across all sectors only 3.8 percent of the work- and 16.1 percent. Appendix 2: Labor Market Features 112 Table A2.9 Employees in Selected Sectors by Gender and Highest Level of Education Achieved, Uganda (percent) Education level achieved Male Female All employees None 3.72 4.10 3.80 Primary 20.80 12.02 19.01 Secondary 29.44 38.25 31.25 Technical or vocational 28.60 34.43 29.79 Bachelor's degree 11.88 7.92 11.07 Master's degree or higher 3.44 1.91 3.13 Professional 2.11 1.37 1.96 Note: Data are for the commercial agriculture, construction, manufacturing, and tourism sectors. Source: World Bank, Investment Climate survey, Uganda, 2002/03. Table A2.10 Employees in Selected Sectors by Highest Level of Education Achieved, Uganda (percent) Technical Master's Sector, subsector, or Bachelor's degree or and firm size class None Primary Secondary vocational degree higher Professional Commercial agriculture 8.70 24.64 31.88 18.84 12.32 2.17 1.45 Construction 0.00 2.90 17.39 34.78 23.19 18.84 2.90 Manufacturing 3.85 20.32 31.46 29.57 10.30 2.59 1.89 Subsector Agro-industry 5.06 23.95 29.34 25.13 11.80 2.53 2.19 Chemicals and paints 1.14 14.77 34.09 27.27 14.77 5.68 2.27 Construction materials 3.94 16.75 29.06 37.44 7.88 2.46 2.46 Furniture 3.68 29.47 37.89 25.26 1.58 0.53 1.58 Metals 8.16 18.37 35.71 27.55 7.14 2.04 1.02 Paper, printing, and publishing 0.00 7.89 28.07 41.23 16.67 3.51 2.63 Plastics 3.85 3.85 30.77 46.15 3.85 11.54 0.00 Textiles and leather products 0.00 8.43 34.94 32.53 21.69 2.41 0.00 Wood 0.00 31.25 31.25 37.50 0.00 0.00 0.00 Firm size class (employees) Micro (<10) 10.13 29.75 36.08 18.35 5.06 0.00 0.63 Small (10­49) 3.18 24.89 34.44 28.51 7.38 1.45 0.14 Medium-size (50­99) 4.95 16.83 30.20 31.68 9.90 3.47 2.97 Large (100+) 1.86 9.84 24.73 35.11 18.09 5.32 5.05 Tourism 0.65 9.03 34.84 39.35 11.61 1.94 2.58 Source: World Bank, Investment Climate survey, Uganda, 2002/03.. Appendix 2: Labor Market Features 113 While 14.8 percent of workers in manufacturing Uganda compares quite favorably with other have a higher education, this share differs widely African countries for which recent survey data are among subsectors. The largest shares of workers with available (figure A2.3). It is second only to Nigeria in a university or professional degree are found in sub- higher education and has the largest share of em- sectors with significant technological or educational ployees with vocational or technical training. requirements: textiles and leather products (24.1 per- cent), chemicals and paints (22.7 percent), and paper, printing, and publishing (22.8 percent). Estimates of Earnings Functions Workers with secondary education account for 28­38 percent of the labor force in manufacturing subsec- Data on workers' earnings in Uganda show differ- tors. ences across regions and subsectors. To identify the Not surprisingly, those in manufacturing with uni- causes of these earnings differentials, a few wage versity or professional degrees tend to work for large equations were estimated, with the log of individual firms. University graduates, postgraduates, and pro- earnings as the dependent variable. The results are fessionally trained workers make up about 28.5 per- shown in table A2.11. cent of the workforce of large firms, but only 6­16 per- Two pairs of equations were estimated, both for cent of the workforce in other size classes. all four sectors surveyed (commercial agriculture, Figure A2.3 Manufacturing Employees by Highest Level of Education Achieved, Selected African Countries (percent) 100 4.3 14.8 11.8 90 11.5 19.9 80 24.5 70 29.6 25.1 60 49.6 50 42.5 40 31.5 44.7 30 22.7 20 20.3 10 20.0 11.9 8.9 0 3.9 1.1 1.4 Uganda (2003) Kenya (2003) Eritrea (2002) Nigeria (2001) None Primary Secondary Technical or vocational University or professional Source: World Bank, Investment Climate surveys, Uganda, 2002/03, Eritrea (2002), Kenya (2003), and Nigeria (2001). Appendix 2: Labor Market Features 114 Table A2.11 Estimates of Wage Determinants in Selected Sectors, Uganda All sectorsa Manufacturing Variable (1) (2) (3) (4) Intercept 2.663*** 3.228*** 2.813*** 2.633*** (20.11) (16.29) (21.07) (11.69) Worker characteristics Education in years 0.090*** 0.083*** 0.082*** 0.076*** (12.26) (12.38) (11.10) (11.40) Experience with firm in years 0.034** 0.038*** 0.045*** 0.053*** (2.40) (3.02) (3.77) (4.58) Experience in years squared 0.000 ­0.001 ­0.001*** ­0.001*** (­0.47) (­1.07) (­2.68) (­3.52) Other professional experience in years 0.045*** 0.041*** 0.049*** 0.042*** (7.06) (7.12) (7.28) (6.73) Gender dummy variable 0.085 0.122** 0.024 0.127** (1 if male, 0 otherwise) (1.41) (2.11) (0.38) (2.11) Weekly hours worked 0.006*** 0.004*** 0.005*** 0.002 (4.65) (3.23) (3.77) (1.55) Training dummy variable 0.371*** 0.367*** 0.384*** 0.351*** (1 if postschool training, 0 otherwise) (5.91) (6.13) (5.59) (5.55) Sector or subsector dummy variables No Yes*** No Yesb Firm characteristics Age in years 0.001 0.000 (0.55) (­0.27) Central region dummy variable 0.420*** 0.455*** (1 if in central region, 0 otherwise) (7.65) (8.28) Foreign ownership dummy variable 0.350*** 0.344*** (1 if 50 percent or more of capital is foreign) (6.41) (6.09) Observations 1,442 1,442 1,232 1,232 F-statistic 75.680 58.583 65.594 42.544 R2 0.270 0.348 0.273 0.387 ** Significant at the 5 percent level. *** Significant at the 1 percent level. Note: Dependent variable is the log of individual earnings. Equations have been estimated using ordinary least squares. Figures in parentheses are White's consistent t-ratios, used to correct for heteroskedasticity in the data. a. Data are for the commercial agriculture, construction, manufacturing, and tourism sectors. b. Partially significant. Source: Authors' calculations based on data from World Bank, Investment Climate survey, Uganda, 2002/03. Appendix 2: Labor Market Features 115 construction, manufacturing, and tourism) and for with firm-specific effects. This result suggests that manufacturing alone. The first pair (shown as 1 and 3 gender discrimination becomes an issue only for in the table), basic wage equations related to the indi- workers employed by firms in the central region or by vidual characteristics of workers (Mincer 1974), as- non-Ugandan firms. sume that employers are able to discern differences in In both samples sector or subsector dummy vari- productivity among workers depending on their edu- ables are often significant, as are some firm-specific cation level, gender, and experience. The second pair variables. Working for firms located in the central re- of equations (2 and 4) differ from the first in their in- gion or for firms with foreign ownership entails higher clusion of sector or subsector dummy variables and earnings in both samples. The age of the firm is in- firm-specific variables (for ease of reading, the significant in both samples, however, suggesting that dummy variables are not reported in the table). if firms benefit from any reputational effects, they do Equation 2, estimated over the entire sample of work- not pass the benefits on to workers. ers, includes sector dummy variables referring to the The results suggest that a purely competitive four sectors surveyed, while equation 4, estimated model is not relevant for explaining the process of only for the subsample of workers employed in manu- wage formation in Uganda, whether in the full sample facturing, includes subsector dummy variables refer- of workers or in the manufacturing sample. In a purely ring to subsectors such as agro-industry, chemicals competitive model of wage formation none of the and paints, and so on. firm-specific and sector dummy variables would be In both samples of workers, whether for the basic significant. wage equation or the augmented variant with firm- specific effects, the variables relating to human capi- tal--years of education, years of experience with the Notes firm, other work experience--all have the expected positive effect and are statistically significant. Thus 1. The formal labor market, as defined here, does the greater a worker's endowment of human capital, not include employment with the central adminis- the higher his or her wages are. Formal training also tration. The data are for firms with 10 or more has a positive and significant effect. The variable employees. "weekly hours worked" is significant except in equa- 2. The interviews covered 1,436 workers in manu- tion 4, suggesting that earnings have less correlation facturing, 160 in tourism, 138 in commercial agri- with hours worked in manufacturing than in other sec- culture, and 69 in construction. Because of the tors. One possible explanation is that for Ugandan small number of interviews in sectors other than manufacturing workers, the share of bonuses in earn- manufacturing, results for these sectors are only ings is among the smallest found in RPED industry indicative. surveys in Sub-Saharan Africa. In both samples the gender dummy variable is highly significant only in the augmented equations Appendix 2: Labor Market Features 116 Additional Tables on the Labor Market Table A2.12 Employees and HIV in Selected Sectors, Uganda Share willing to pay to be tested Rating of at their firm Amount employees would HIV/AIDS Share if testing is be willing to pay to be as a concern knowing where voluntary and tested at their firm (1 = lowest, to be tested anonymous (Uganda shillings) Job category 5 = highest) (percent) (percent) Average Maximum Management 4.04 91.31 71.60 9,353.04 100,000.00 Professionals 4.28 92.90 77.25 8,359.52 100,000.00 Skilled production workers 4.23 84.33 75.72 4,516.96 50,000.00 Unskilled production workers 3.90 80.41 67.35 3,235.42 50,000.00 Nonproduction workers 4.08 83.18 72.19 5,567.26 100,000.00 All job categories 4.07 85.52 72.00 5,913.15 100,000.00 Note: Data are for the commercial agriculture, construction, manufacturing, and tourism sectors. Source: World Bank, Investment Climate survey, Uganda, 2002/03. Appendix 2: Labor Market Features 117 Table A2.13 Manufacturing Employees by Method of Payment, Uganda (percent) Hourly Daily Weekly Monthly By the piece Job category Management 0.31 4.04 1.55 88.20 5.90 Professionals 0.00 0.79 0.79 96.85 1.57 Skilled production workers 0.00 3.42 4.66 82.61 9.32 Unskilled production workers 0.24 11.98 3.91 63.57 20.29 Nonproduction workers 0.00 3.40 5.53 89.36 1.70 All job categories 0.14 5.80 3.53 80.78 9.75 Production workers (skilled and unskilled) Subsector Agro-industry 0.35 7.96 3.81 79.58 8.30 Chemicals and paints 0.00 5.13 7.69 82.05 5.13 Construction materials 0.00 1.10 3.30 82.42 13.19 Furniture 0.00 8.59 4.69 55.47 31.25 Metals 0.00 22.22 9.52 41.27 26.98 Paper, printing, and publishing 0.00 1.92 0.00 98.08 0.00 Plastics 0.00 0.00 8.33 91.67 0.00 Textiles and leather products 0.00 22.22 2.78 58.33 16.67 Wood 0.00 0.00 0.00 42.86 57.14 Firm size class (employees) Micro (<10) 0.00 14.71 0.00 62.75 22.55 Small (10­49) 0.00 10.43 4.01 62.30 23.26 Medium-size (50­99) 0.00 5.26 5.26 89.47 0.00 Large (100+) 0.63 0.63 6.88 90.00 1.88 All production workers 0.14 8.21 4.24 71.96 15.46 Source: World Bank, Investment Climate survey, Uganda, 2002/03. Appendix 2: Labor Market Features 118 Table A2.14 Monthly Earnings of Unskilled Production Workers in Manufacturing, Uganda U.S. dollars Uganda shillings U.S. dollars Uganda shillings Subsector Firm size class (employees) Agro-industry Micro (<10) Average value 48.71 85,250.03 Average value 55.76 97,578.70 (55.5) (97,104.9) (69.9) (122,305.9) Observations 140 140 Observations 72 72 Chemicals and paints Small (10­49) Average value 89.74 157,053.62 Average value 49.47 86,567.01 (68.1) (119,230.4) (40.8) (71,431.2) Observations 23 23 Observations 182 182 Construction materials Medium-size (50­99) Average value 57.91 101,343.05 Average value 55.52 97,168.40 (54.3) (95,002.8) (29.6) (51,832.8) Observations 39 39 Observations 48 48 Furniture Large (100+) Average value 36.29 63,510.05 Average value 86.39 151,190.88 (26.9) (47,150.8) (91.3) (159,715.0) Observations 73 73 Observations 58 58 Metals Average value 90.53 158,431.37 Firm location (82.0) (143,556.3) Central region Observations 34 34 Average value 67.03 117,308.13 Paper, printing, and publishing (66.4) (116,219.8) Average value 109.25 191,188.41 Observations 222 222 (75.3) (131,724.8) Northeast region Observations 23 23 Average value 46.93 82,133.33 Plastics (60.2) (105,393.3) Average value 66.12 115,714.29 Observations 45 45 (28.0) (49,078.7) Southwest region Observations 7 7 Average value 39.79 69,630.24 Textiles and leather products (19.9) (34,747.8) Average value 38.80 67,892.16 Observations 93 93 (18.5) (32,310.2) Observations 17 17 Wood All unskilled production workers Average value 47.14 82,500.00 Average value 57.48 100,594.49 (13.5) (23,629.1) (58.4) (102,223.2) Observations 4 4 Observations 360 360 Note: Computed on the basis of the earnings reported by workers in early 2003 and converted into U.S. dollars using the official exchange rate of $1 = 1,750 USh. Figures in parentheses are standard deviations. Source: World Bank, Investment Climate survey, Uganda, 2002/03. Appendix 3: Protection of Manufacturing 119 The succession of extreme events in Uganda since its Moreover, coffee export earnings influenced the independence in 1962 has been detrimental to the exchange rate and the administered coffee producer development of its manufacturing sector, which re- prices. The dominance of coffee in export earnings mains small. At the end of the 1990s manufacturing and government revenue explains the persistent over- accounted for about 8.7 percent of GDP in Uganda, valuation of the exchange rate, which was fixed until compared with an average of about 12 percent in the mid-1980s. In addition, tight foreign exchange al- other low- and middle-income countries in Sub- location procedures discouraged alternative interna- Saharan Africa (World Bank 2001). In 1994­99 formal tional activities (Morrissey and Rudaheranwa 1998). manufacturing provided about 5.4 percent of export Faced with a dramatic economic situation, the earnings and employed about 50,000 people, around government launched a comprehensive economic re- 0.5 percent of the available labor force (World Bank covery program in 1987 with support from the World 2001; WTO 2001). Bank, the International Monetary Fund, and other The ability to trade competitively is crucial to donors. The focus of the recovery program shifted in Uganda's economic development, as is implementing the early 1990s from economic growth and macro- an appropriate set of policies to promote competitive- economic stabilization to structural reforms. Several ness (Bonaglia and Fukasaku 2002). This appendix measures were taken to foster international trade, analyzes the level of protection granted to the which had come to be viewed as the major engine of Ugandan manufacturing sector based on the firm- economic growth. level data collected in the 2002/03 survey by the Among the most important measures was to liberal- World Bank's Regional Program on Enterprise ize the foreign exchange market, which had become a Development (RPED). major impediment to growth. The government made rapid progress in deregulating this market. Initially the exchange rate policy involved repeated devaluation Overview of Trade Policy Reforms and rationing of the available foreign exchange under various mechanisms. But by 1993­94 the exchange rate During the period following independence, Uganda's was determined entirely by the market and exporters export earnings came mainly from agricultural com- were no longer forced to surrender their export receipts. modities. In the 1970s, as in many other African coun- In addition, the government has pursued a per- tries, trade policy in Uganda was characterized by sistent policy of trade liberalization since 1987: heavy taxation of exports and significant quantitative restrictions on imports (Collier 1997). · In 1991 the government removed most nontariff Coffee was among the few exports that survived barriers. the economic mismanagement of the 1970s because · The government has gradually alleviated the bu- coffee production requires few inputs and coffee reaucracy surrounding imports, though there re- could be largely smuggled abroad (Henstridge 1996). mains room for improvement (Collier 1997). The As a result, coffee exports accounted for about 94 RPED survey data show that in 2003 customs percent of merchandise export revenue in 1982 (EIU clearance for manufacturers still took an average 2002).1 This dominance of coffee helps explain why of 5.8 days and as long as 11.2 days Uganda's trade policies in the 1970s and early 1980s · The government shifted the structure of trade were heavily influenced by conditions in the world cof- taxes from export taxation to import taxation. It fee market (Morrissey and Rudaheranwa 1998). abolished the coffee tax (or stabilization tax), Appendix 3: Protection of Manufacturing 120 which once provided a large part of government Structure of Nominal Protection revenues, in 1996. Needing to maintain and even increase revenues, it then shifted the burden of The dramatic changes in Uganda's trade policy taxation toward imports.2 Import taxes thus be- regime in the past decade and a half call for an as- came the main instrument of trade policy, but their sessment of the present situation. This section ana- rates gradually declined as import liberalization lyzes the structure of nominal protection based on the proceeded. The first round of liberalization oc- existing duty scheme and a subsample of 228 manu- curred in 1992, with ad valorem rates ranging from facturing firms from the 2002/03 RPED survey.3 It first 10 percent to 60 percent. This range was reduced estimates the overall level of nominal protection and to 10­50 percent in 1994. In 1996 the highest tar- compares it with that in other countries. Because this iff was 30 percent, and in 2002­03 it was 15 per- analysis is based on official schedules and does not cent. The simple average for the manufacturing take into account the fact that firms have multiple out- sector was 8.2 percent in 2001­02. The maximum puts and inputs, often subject to different duties, it is rate of 15 percent applies mostly to imports of also necessary to compute firm-level indexes of nom- manufactured tobacco, clothing, furniture (except inal protection. metal), and plastic products. Tariff and Nontariff Nominal Protection Uganda has also promoted trade relations with Tariff protection in Uganda has been drastically re- other developing countries through regional integra- duced over time (table A3.1). Agriculture faced the tion, such as through the Common Market of Eastern largest reduction in tariff protection, with the average and Southern Africa (COMESA) and the East African tariff decreasing from 21.7 percent in 1994 to 12.32 Community. Tariff rates on imports from COMESA percent in 2002. Manufacturing also had a big decline countries, notably Kenya and Tanzania, were 0, 4, and in protection, with tariffs decreasing from 17.8 percent 6 percent in 2001/02. in 1994 to 10.6 percent in 2002. Table A3.1 Applied Tariffs by Major Commodity Group, Uganda, Selected Years, 1994­2002 (percent) 1994 2000 2002 Simple Weighted Simple Weighted Simple Weighted Commodity group average average average average average average Agricultural, forestry, and fishery products 21.70 17.25 12.34 2.49 12.32 2.81 Mineral commodities 11.48 9.25 7.69 6.41 7.65 7.16 Manufactured commodities 17.80 18.48 10.85 8.11 10.62 8.16 Other commodities 15.61 29.78 10.18 14.95 10.66 14.33 Source: UNCTAD, TRAINS database and Standard Industrial Classification (SIC) system. Appendix 3: Protection of Manufacturing 121 With the strong emphasis on trade liberalization, Comparisons of tariff data with two country groups including the elimination of all quantitative restrictions, confirm that Uganda made deep reductions.6 In 2002 tariffs have become the main trade policy instrument tariff protection in Uganda was lower than the average in Uganda. By 2001­02 tariff protection in Uganda rates applied in both least developed countries and was based primarily on three ad valorem rates--0, 7, Sub-Saharan African countries, whatever the sector, and 15 percent--applied on the CIF (cost, insurance, including manufacturing (table A3.2; figure A3.1). and freight) value of imports. The average most- Indeed, tariff protection in Uganda was closer to that favored-nation tariff was around 9 percent. About 16.1 in industrial countries, and for agricultural goods it percent of tariff lines were duty free, and 83.9 percent was even lower. dutiable.4 There were no recorded international or na- On the basis of the tariff schedule, Uganda's tional tariff peaks (WTO 2003).5 trade policy regime now seems fairly liberal. But Table A3.2 Weighted Average Tariffs by Major Commodity Group, Uganda and Selected Country Groups, Selected Years, 1994­2002 (percent) Country or country group and commodity group 1994 2000 2002 Uganda Agricultural, forestry, and fishery products 17.25 2.49 2.81 Mineral commodities 9.25 6.41 7.16 Manufactured commodities 18.48 8.11 8.16 Other commodities 29.78 14.95 14.33 Industrial countries Agricultural, forestry, and fishery products 5.57 3.52 3.62 Mineral commodities 0.33 1.36 0.01 Manufactured commodities 8.05 5.38 4.86 Other commodities 0.34 0.08 0.10 Least developed countries Agricultural, forestry, and fishery products 35.40 8.25 7.10 Mineral commodities 51.65 15.91 15.40 Manufactured commodities 69.81 20.03 17.74 Other commodities 44.12 15.79 14.96 Sub-Saharan Africa Agricultural, forestry, and fishery products 7.79 14.70 6.58 Mineral commodities 15.92 20.96 8.42 Manufactured commodities 15.59 18.10 14.22 Other commodities 27.14 22.39 18.00 Source: UNCTAD, TRAINS database and Standard Industrial Classification (SIC) system. Appendix 3: Protection of Manufacturing 122 Figure A3.1 Simple Average Tariffs for Manufactured Commodities, Uganda and Selected Country Groups, Selected Years, 1994­2002 (percent) 40 35 30 25 20 15 10 5 0 1994 2000 2002 Uganda Industrial countries Least developed countries Sub-Saharan Africa Source: UNCTAD, TRAINS database and Standard Industrial Classification (SIC) system. tariffs are not the only policy tool affecting the do- Because of the multiple taxes affecting the do- mestic price of tradable goods in Uganda. Some mestic price of imports, restricting the analysis to tar- nontariff restrictions are still in force, mainly for moral, iffs could be misleading. A proper picture of nominal health, security, or similar reasons (WTO 2001). protection in Uganda can be obtained by using nom- These restrictions, all computed on an ad valorem inal protection coefficients, which allow an analysis of basis, include an import license commission of 2 per- the pattern and level of protection, based on the latest cent and a withholding tax of 4 percent collected on available official data. The nominal protection coeffi- all imports (though the withholding tax has been cient is usually defined as the ratio of the appropri- waived on all imported raw materials; WTO 201). ately adjusted domestic price to a comparable world Internal taxes include an excise tax and a value price. When a protection regime is based entirely on added tax. An excise duty in the 10­15 percent range ad valorem duties and there are no quantitative re- is levied on many goods, imported or locally pro- strictions, as is the case now in Uganda, nominal pro- duced.7 And a value added tax of 17 percent applies tection coefficients equal one plus the duty rate. As to most imported or domestically produced goods, defined here, this duty rate includes the tariff rate, the with the important exception of unprocessed agricul- import license commission, the withholding tax, and tural products (WTO 2001). the excise tax.8 Appendix 3: Protection of Manufacturing 123 The nontariff protections increase nominal protec- Nominal Protection at the Firm Level in tion. Although the average most-favored-nation tariff is Manufacturing around 9 percent, the average nominal protection co- Most manufacturing enterprises in Uganda produce a efficient equals 1.1521, the equivalent of a 15.21 per- variety of goods and use a range of raw materials that cent duty. By definition, the maximum tariff is 15 per- do not always fit into a single tariff category. Thus as- cent, but the maximum nominal protection coefficient sessing the true level of nominal protection granted to is now 1.481 (for tobacco products), the equivalent of a firm requires computing a weighted average nomi- a 48.1 percent tariff. The impact is nonnegligible, but nal protection coefficient at the firm level, with the the overall level of protection, with a few exceptions, weight being the share of each product or raw mate- remains within reasonable limits. rial in the company's total sales or purchases. This is More important, the excise tax increases the dis- done here for 228 manufacturing firms using 2001/02 persion of nominal protection coefficients relative to data relating to he firms' five most important outputs the tariff-induced dispersion.9 Although excise taxes and five most important raw materials as reported in cover only 8.65 percent of all tariff lines, their cover- the RPED survey (table A3.4). age varies greatly at the category level. For some In 2001/02, firms in the metals industry had the products, such as those in the milling industry (cate- lowest nominal protection (a weighted average nomi- gory 11), excise taxes cover only 3.4 percent of the nal protection coefficient of 1.141), and those in the lines, while for others, such as meat products (cate- wood, plastics, and textile and leather products sub- gory 2) or beverages (category 22), they cover all tar- sectors the highest (1.243, 1.212, and 1.207). Most iff lines. In addition, because there is some variation in firms in the last three subsectors produce combina- the excise tax rate, the overall variability is increased. tions of goods that face maximum import tariffs of 15 The tariff-induced standard deviation in nominal pro- percent and are often subject to excise taxes. Firms in tection coefficients is 0.039, but when excise taxes the chemical and paint industry and plastics subsec- are taken into account, the standard deviation is tor seem to face the lowest protection on inputs (1.095 0.057. and 1.088). The dispersion in weighted nominal pro- Thus despite the use of just three tariff rates and a tection coefficients remains significant for both out- large decline in tariff protection, some products still puts and inputs. It is equal to roughly a third of the face significant dispersion at the two-digit level (table mean value of the nominal protection coefficients. A3.3). This probably leads to inefficiencies in the allo- Foreign firms have a slightly lower weighted aver- cation of resources across sectors.10 Such products age nominal protection coefficient on inputs than do as ships (position 89), electrical machinery and domestic firms--while domestic firms have a slightly equipment (position 85), vehicles (position 87), and lower coefficient on outputs. Only in the construction mineral fuels (position 27) face the largest internal dis- materials subsector do locally owned firms have a tortions. In addition, some of the products shown in higher nominal protection coefficient on outputs than table A3.3 are inputs or raw materials for manufactur- their foreign counterparts. The dispersion of weighted ing. Changes in duties on inputs are usually difficult to nominal protection coefficients on outputs is higher for cope with. This may provide an incentive for some foreign-owned firms. For inputs there is no clear rela- firms to try to reduce import taxes by misidentifying tionship with firm ownership. goods while keeping them within the same two-digit These differences may be explained in part by the classification. bargaining power of a few local firms able to obtain Appendix 3: Protection of Manufacturing 124 1.48 1.23 1.23 1.23 1.23 1.23 1.23 1.23 1.23 1.23 1.21 1.36 1.21 1.23 1.23 1.23 1.21 1.23 1.21 1.23 1.23 1.23 1.23 1.21 value 2002 Maximum ficient d nominal 6.59 6.31 6.01 5.59 5.40 coef 5.17 5.09 4.99 4.68 4.66 4.56 4.52 4.48 4.46 4.43 4.40 4.39 4.30 4.26 4.21 4.17 4.13 4.04 12.60 Uganda, ,y Standar deviation otection Unweighted pr 1.40 1.09 1.14 1.16 1.15 1.15 1.17 1.18 1.18 1.17 1.09 1.34 1.19 1.18 1.18 1.18 1.18 1.15 1.19 1.20 1.20 1.22 1.18 1.17 value Categor verage A oduct 0.00 7.00 Pr cent) 15.00 15.00 15.00 15.00 15.00 15.00 15.00 15.00 15.00 15.00 15.00 15.00 15.00 15.00 15.00 15.00 15.00 15.00 15.00 15.00 15.00 15.00 by Maximum (per d 0.00 0.00 6.15 5.80 5.50 4.85 0.00 4.00 4.00 3.97 4.65 0.00 4.39 3.99 4.32 3.94 4.37 4.02 3.85 4.14 3.40 3.41 3.99 4.00 ficients f Standar deviation Coef arif T 0.00 6.41 7.26 4.32 7.00 7.16 1.51 0.49 6.94 8.99 cent) 15.00 10.46 14.53 11.87 15.00 11.38 11.63 13.57 10.28 10.38 10.63 12.74 12.50 10.96 eighted average otection W (per Pr 0.00 7.60 9.38 8.40 8.63 7.00 2.53 8.61 cent) 15.00 11.12 11.00 10.48 15.00 12.71 11.33 12.00 11.70 12.38 13.69 13.64 13.11 13.69 10.78 11.08 Simple average (per Nominal ts, oducts wheat ted par pr ticles ar ts; oducts measuring, inulin; their cosmetics, Distor substitutes par pr coal stock, washing similar and ches; es and goods char ollingr y oils star and fumes, metal ticles Most agents, ar tobacco distillation wood per structur and malt; ed aluminum ed esses, base the their eparations instruments vinegar tramway pr fats cinematographic, of of of y floating equipment,,y wood; e mattr and confectioner and origin, of equipment and ecision face-active oducts; esinoids;r ticles cinematographic ticles pr pr ar manufactur and oils, edible sur y or ar manufactur and ding filaments sugar vegetable spirits, animal ticles categor glasswar ar and machiner and railway and of bedding, oils boats, ecorr fuels, and photographic, and organic industr e; and and database. accessories toiletries eals oduct nitur Pr obaccoT Ships, Electrical sound ehicles,V oducts eparations ood and Mineral Cer Miscellaneous Man-made Sugars Animal Optical, checking, Beverages, Pr Soap, pr Milling Cotton Glass Aluminum Photographic Fur W Essential and Miscellaneous Miscellaneous Characteristics TRAINS AD, A3.3 code UNCT monized ce: ableT Har system 24 89 85 87 27 10 21 54 17 15 90 22 05 34 11 52 70 76 37 94 44 33 83 96 Sour Appendix 3: Protection of Manufacturing 125 6 5 5 6 2 3 4 3 2 6 3 5 90 11 32 39 17 16 12 20 45 70 26 37 14 12 10 228 183 vations Uganda, Obser Raw 1.020 1.078 1.084 1.103 1.014 1.089 1.008 1.072 1.020 1.008 1.020 1.080 1.090 1.190 1.014 1.090 1.008 1.090 1.008 1.020 1.078 1.084 1.103 1.087 1.089 1.090 1.072 1.020 1.020 materials ficient Manufacturing, Minimum coef 1.031 1.060 1.080 1.169 1.035 1.105 1.130 1.130 1.127 1.031 1.113 1.130 1.130 1.210 1.036 1.166 1.130 1.210 1.036 1.031 1.060 1.080 1.169 1.035 1.105 1.130 1.130 1.127 1.031 in Outputs Raw 1.270 1.120 1.271 1.273 1.193 1.170 1.158 1.270 1.261 1.273 1.270 1.120 1.182 1.270 1.116 1.170 1.158 1.138 1.270 1.270 1.117 1.271 1.273 1.193 1.170 1.091 1.270 1.261 1.273 Materials materials Raw ficient Maximum coef and 1.643 1.310 1.267 1.310 1.308 1.297 1.310 1.305 1.310 1.643 1.643 1.289 1.222 1.249 1.308 1.297 1.310 1.210 1.643 1.310 1.310 1.267 1.310 1.206 1.224 1.210 1.305 1.310 1.310 Outputs Outputs Raw 0.065 0.014 0.055 0.044 0.046 0.032 0.048 0.076 0.104 0.068 0.078 0.016 0.038 0.057 0.052 0.039 0.075 0.034 0.066 0.060 0.013 0.059 0.044 0.040 0.031 0.000 0.079 0.104 0.067 for d materials. materials raw tant Standar deviation ficients 0.071 0.076 0.044 0.029 0.063 0.051 0.074 0.040 0.076 0.061 0.115 0.067 0.043 0.027 0.137 0.054 0.096 0.000 0.092 0.051 0.089 0.043 0.030 0.043 0.043 0.046 0.044 0.076 0.051 impor Outputs Coef most five 2002/03. Raw 1.146 1.095 1.126 1.227 1.119 1.111 1.088 1.167 1.168 1.150 1.127 1.096 1.137 1.230 1.070 1.111 1.086 1.114 1.121 1.151 1.095 1.124 1.226 1.130 1.111 1.090 1.177 1.168 1.157 and average otection materials ficient Pr Uganda,, outputs coef eighted 1.199 1.190 1.194 1.217 1.141 1.180 1.212 1.207 1.243 1.197 1.221 1.195 1.170 1.229 1.165 1.223 1.240 1.210 1.209 1.193 1.186 1.199 1.216 1.136 1.166 1.183 1.206 1.243 1.194 tant survey W Outputs Nominal impor Climate most five m-Level oducts oducts ms oducts publishing pr publishing pr fir publishing ms' Investment pr Fir fir ms paints and fir paints and paints and to Bank, y materials leather y materials leather owned y materials leather and and and eferr A3.4 orld e and printing,, e and printing,, e and W printing,, Data sample ce: ableT o-industr nitur subsectors eign-owned o-industr nitur subsectors o-industr nitur subsectors 2001/02 ood ood Subsector Full Agr Chemicals Construction Fur Metals Paper Plastics extilesT W All For Agr Chemicals Construction Fur Metals Paper Plastics extilesT All Domestically Agr Chemicals Construction Fur Metals Paper Plastics extilesT W All Note: Sour Appendix 3: Protection of Manufacturing 126 better protection or exemptions from the state. Where vestors might shift resources. The firm-level effective foreign-owned firms have lower nominal protection on rate of protection is defined as inputs or higher protection on outputs, this can be (VA D k - VAk W) seen as compensation granted for the extra costs W (such as deficient infrastructure, high transport costs, ERPk = VAk and low labor skills and productivity) and perceived for a firm k where VADk is the value added at domestic high risks of operating in Uganda. In the early 1990s, prices or the tariff-distorted value added and VAWk is as many authors have noted, investors perceived the value added expressed in world prices or simu- Uganda as the riskiest country in Africa--and Africa lated for the same sector in the absence of trade re- as the riskiest region. Despite the dramatic improve- strictions. ments in the economic situation, investors still per- Thus the effective rate of protection indicates the ceive Uganda as a risky country (see table 1.1 in extent to which the value added changes as a conse- chapter 1). quence of the entire tariff structure under the assump- tion that there are few or no nontariff barriers that may cause further distortions. Other things being equal, Effective Protection of Manufacturing the higher the nominal tariffs on output, the higher the effective rate of protection is--and the higher the tar- The data in the previous section suggest that the iffs on inputs, the lower it is. overall structure of nominal protection in Uganda is Effective rates of protection estimated on the appropriate but that some deficiencies remain, basis of a restricted subsample of 71 firms for which mainly in the allocative efficiency of the duty structure the required output and input data are available at the (given some significant dispersion) and the eventual subsectoral level show wide variation across subsec- tax advantages granted to specific firms.11 But nomi- tors, just as for nominal protection coefficients. The nal protection has an impact not only on the value estimated rate of effective protection ranges from of a firm's output but also on the cost of its in- roughly 28 percent in the wood subsector to almost 80 puts, and thus affects the value added generated by percent in the textile and leather products subsector enterprises. (table A3.5). This wide variation suggests that the This effect is usually captured by the effective rate structure of protection is not neutral in its impact of protection, which, in a partial equilibrium setting, across subsectors. measures the proportional change in an industry's Effective protection also varies widely within sub- value added attributable to the tariff structure and rel- sectors, as shown by the standard deviation and other ative to a free trade benchmark that is usually proxied statistics. In the textile and leather products industry, using world prices or prices from a freer but otherwise for example, the effective rate of protection ranges comparable competitor country. The effective rate of from a low 24 percent to a very high 168.6 percent. protection depends not only on the tariffs on a firm's Part of the reason for this wide disparity may be the final product but also on tariffs on inputs and input co- fact that in the least developed countries firms within efficients in production. Effective rates of protection at the same subsector tend to use very different produc- the sectoral level and the ranking of sectors by these tion technologies (and, as a result, the technical coef- values synthesize the impact of the overall structure of ficients vary widely from firm to firm). The dispersion in protection and indicate the directions in which in- weighted nominal protection coefficients, largely Appendix 3: Protection of Manufacturing 127 Table A3.5 Firm-Level Effective Rates of Protection in Manufacturing by Subsector, Uganda, 2001/02 (percent) Standard Subsector Average Maximum Minimum deviation Agro-industry 47.70 178.42 1.05 0.40 Chemicals and paints 75.50 137.48 13.53 0.88 Construction materials 46.37 105.61 3.05 0.29 Furniture 67.12 173.32 -0.89 0.50 Metals 52.91 111.12 15.91 0.37 Paper, printing, and publishing 33.19 53.87 14.07 0.20 Plastics 30.78 34.20 27.37 0.05 Textiles and leather products 79.69 168.62 24.03 0.66 Wood 27.97 27.97 27.97 All subsectors 52.26 0.42 Source: World Bank, Investment Climate survey, Uganda, 2002/03. Table A3.6 Firm-Level Effective Rates of Protection in Manufacturing by Type of Ownership and Subsector, Uganda, 2001/02 (percent) Type of ownership Standard and subsector Average Maximum Minimum deviation Foreign Agro-industry 46.21 96.23 18.06 0.29 Chemicals and paints 13.53 13.53 13.53 Construction materials 50.28 74.58 25.99 0.34 Textiles and leather products 168.62 168.62 168.62 All subsectors 56.00 0.48 Domestic Agro-industry 48.05 178.42 1.05 0.43 Chemicals and paints 137.48 137.48 137.48 Construction materials 45.50 105.61 3.05 0.29 Furniture 67.12 173.32 ­0.89 0.50 Metals 52.91 111.12 15.91 0.37 Paper, printing, and publishing 33.19 53.87 14.07 0.20 Plastics 30.78 34.20 27.37 0.05 Textiles and leather products 57.46 132.48 24.03 0.51 Wood 27.97 27.97 27.97 All subsectors 51.65 0.41 Source: World Bank, Investment Climate survey, Uganda, 2002/03. Appendix 3: Protection of Manufacturing 128 induced by the use of excise taxes, probably also ex- defined as the number of harmonized schedule plains a significant share of the variability within sub- six-digit duties at least three times the country's sectors. overall simple average, divided by the total num- Just as for nominal protection, the effective pro- ber of harmonized schedule subheadings. tection granted to firms is marginally higher for foreign 6. International comparisons are restricted to tariffs than for local firms (largely because of the high effec- because not all countries report data on quantita- tive protection in the textile and leather products in- tive restrictions, nontariff barriers, and other non- dustry) and there is wide dispersion (table A3.6). tariff protection to international trade bodies. 7. Higher excise duties are levied on cigarettes (130 percent), alcoholic beverages (70 percent), and Notes soft drinks (15 percent). Only oil products under the 27.10 harmonized schedule heading are sub- 1. After the collapse of the International Coffee ject to a non­ad valorem excise duty, ranging from Organization's export quota agreement in 1989, 200 to 580 USh per liter. The ad valorem equiva- however, the share of export earnings from coffee lents have been computed using the UNCTAD fell sharply, to 53 percent in 1993. Since then the methodology (Stawowy 2001). share has fluctuated largely in line with world 8. The nominal protection coefficient equals [Pjk /Pjk ] D w prices, averaging about 56 percent in 1994­2000. for a firm k producing a good j, with Pjk being the D 2. Although the rationale at the sectoral level for re- domestic price and Pjk the relevant world price. w moving or reducing the coffee export tax was When quantitative restrictions or other nontariff strong, it is unclear whether the general equilib- trade barriers are in use, the domestic price re- rium effects of this measure were properly as- sults from various other factors (such as the insti- sessed at the time (Collier 1997). Removing ex- tutional framework, the degree of competition in port taxes and replacing them with import taxes domestic industry, and the supply-demand bal- was often equivalent to taxing exports (the Lerner ance generated by regulatory policy). In this case Equivalence Theorem). That is, taxing imports was a tariff-based nominal protection coefficient does another way to tax exports. not fully capture the extent of distortions. The nom- 3. Firms with missing or incomplete data were re- inal protection coefficient is then better proxied by moved from the computations. computing the ratio of the ex-factory price to the 4. The share of duty-free tariff lines is the number of CIF (cost, insurance, and freight) import price. harmonized schedule subheadings for which all In Uganda a good approximation is to assume tariff line duties are equal to zero, divided by the that the dominant distortion is induced by ad val- total number of subheadings. The share of du- orem duties, since all quantitative restrictions and tiable tariff lines is the number of harmonized various other non tariff barriers have been. In this schedule subheadings for which not all tariff line case the total duty (tj) is the sum of the relevant duties are equal to zero, divided by the total num- tariff rate, the import license commission, the with- ber of subheadings. holding tax, and the excise tax. The domestic 5. An international peak is defined as the number of price is then Pjk = (1 + tj) Pjk , which simplifies the D w harmonized schedule six-digit duties higher than nominal protection coefficient to (1 + tj). 15 percent, divided by the total number of harmo- 9. The import license commission and the withhold- nized schedule subheadings. A national peak is ing tax have a fixed rate and apply equally to all Appendix 3: Protection of Manufacturing 129 products. Thus they affect only the level of nomi- 11. This concern is also expressed by the Inter- nal protection, not its dispersion. national Monetary Fund (2002a), which underlines 10. Although tobacco products appear in table A3.3, the limits of the selective approach to incentives they are not considered in the subsequent discus- that discriminates among firms and favors rent sion because their high level of taxation is appro- seeking and eventually corruption. priate for health policy reasons. Appendix 4: Investment Climate Indicators 130 Table A4.1 Structure of Manufacturing Sample for Uganda Investment Climate Survey (percent) Share of sample Share of sample Firm size Firm activity Small (<100 employees) 12.00 Agro-industry 40.67 Large (100+ employees) 88.00 Chemicals and paints 6.00 Construction materials 13.33 Market orientation Furniture 15.67 Exportera 12.00 Metals 7.00 Nonexporter 88.00 Paper, printing, and publishing 7.67 Plastics 2.33 Firm ownership Textiles and leather products 5.00 Publicly listed company 1.33 Wood 2.33 Publicly held limited company 2.67 Privately held limited company 60.00 Partnership 8.33 Firm location Sole proprietorship 23.00 Central region 68.00 Cooperative 2.00 Northeast region 15.00 Other 2.67 Southwest region 17.00 a. Exports 10 percent or more of sales Source: World Bank, Investment Climate survey, Uganda, 2002/03. Table A4.2 Manufacturing Firms' Competitors, Suppliers, and Customers in International Comparison Uganda, by Firm Type Low High Uganda Eritrea Pakistan China India Morocco Smalla Large capacityb capacity Average number of competitors Domestic private firms 30.65 14.69 85.90 -- -- -- 33.17 12.37 34.57 15.16 State-owned firms 0.14 0.29 0.70 -- -- -- 0.16 0.03 0.15 0.16 Foreign-owned firms 9.87 0.14 2.90 -- -- -- 10.83 2.91 7.72 19.82 Average number of suppliers Domestic private firms 15.89 -- 93.20 -- -- -- 15.92 15.68 15.74 17.65 State-owned firms 0.05 -- 1.30 -- -- -- 0.02 0.24 0.04 0.09 Foreign-owned firms 1.13 -- 7.90 -- -- -- 0.97 2.38 0.95 1.89 Average number of customers Domestic private firms 87.37 -- 133.10 -- -- -- 73.09 190.33 91.81 80.17 State-owned firms 1.14 -- 1.30 -- -- -- 1.19 0.73 0.85 2.05 Foreign-owned firms 4.37 -- 14.30 -- -- -- 2.83 15.70 4.34 5.25 -- Not available. a. Small = below 100 employees, Large = larger than 100 employees. b. Low Capacity = below 75 percent capacity, High = higher than 75 percent capacity and above Source: World Bank, Investment Climate surveys, Uganda, 2002/03, Eritrea, Pakistan, 2002, China, 2000, India, 1999, Morocco, 2000. Appendix 4: Investment Climate Indicators 131 Table A4.3 Manufacturing Firms' Evaluation of General Constraints to Operation in International Comparison (percentage of respondents evaluating constraint as major or very severe) Uganda, by Firm Type Low High Constraint Uganda Eritrea Pakistan China India Morocco Smalla Large capacityb capacity Telecommunications 5.19 13.92 9.2 23.5 -- -- 4.7 8.3 6.2 4.6 Electricity 44.48 36.71 39.2 29.7 -- -- 43.5 51.4 18.8 44.9 Transport 22.9 17.95 10 19.1 -- -- 21.8 30.6 19.7 30.6 Access to land 17.39 21.52 20.4 14.7 -- -- 17.2 19.2 44.4 17.3 Tax rates 48.31 29.11 45.6 36.8 -- -- 48.5 47.2 9.9 49.5 Tax administration 36.11 15.19 46.1 26.7 -- -- 35.9 37.1 37.1 38.7 Customs and trade regulations 27.44 8.97 24.6 19.3 -- -- 26.4 33.3 21.4 28.9 Labor regulations 10.81 5.06 15 20.7 -- -- 9.8 17.1 24.1 10.1 Skills and education of available workers 30.82 40.51 12.8 30.7 -- -- 30.6 32.4 25.0 33.2 Business licensing and operating permits 10.1 2.53 14.5 21.3 -- -- 10.3 8.3 18.2 9.6 Access to finance (collateral requirements) 45.04 -- 37.6 22.8 -- -- 45.6 40.6 9.6 43.0 Cost of finance (interest rates) 60.3 36.71 42.6 21.8 -- -- 60.4 59.4 29.8 61.3 Regulatory policy uncertainty 27.56 29.11 40.1 32.9 -- -- 26.5 34.3 42.7 29.4 Macroeconomic instability (inflation, exchange rate) 45.42 79.75 34.4 30.2 -- -- 43.7 57.1 51.3 48.7 Corruption 38.21 2.53 40.4 27.3 -- -- 38.2 38.7 28.2 43.3 Crime, theft, and disorder 26.85 1.27 21.5 20 -- -- 26.3 30.6 27.5 25.0 Anticompetitive or informal practices 31.11 7.59 21.3 23.7 -- -- 29.5 41.7 58.6 31.9 -- Not available. a. Small = below 100 employees, Large = larger than 100 employees. b. Low Capacity = below 75 percent capacity, High = higher than 75 percent capacity and above Source: World Bank, Investment Climate surveys, Uganda, 2002/03, Eritrea, Pakistan, 2002, China, 2000, India, 1999, Morocco, 2000. Appendix 4: Investment Climate Indicators 132 Table A4.4 Infrastructure Performance as Reported by Manufacturing Firms in Uganda, in International Comparison and by Firm Characteristic Low High Indicator Uganda Eritrea Pakistan China India Morocco Smalla Large capacityb capacity Frequency of power outages (times in previous year) 38.59 105.44 14.50 -- -- 15.90 36.49 54.20 36.29 38.81 Share of production lost due to power outages (percent) 6.25 5.47 5.40 2.00 -- -- 6.52 4.54 4.49 6.65 Share of firms with own generator (percent) 35.33 43.04 41.80 16.20 68.94 16.73 30.68 69.44 38.98 33.63 Share of firms with own well (percent) 13.00s 48.10 43.80 15.60 50.77 29.14 10.61 30.56 20.34 11.06 Share of production lost in shipment (percent) -- 0.36 -- 1.20 -- -- -- -- -- -- Days to obtain a telephone connection 33.16 256.33 47.30 12.00 -- -- 35.07 23.14 45.86 27.94 Days to obtain an electricity connection 38.33 98.68 46.80 19.00 -- -- 38.33 39.14 23.18 37.15 -- Not available. s. Share of firms that have built own well. a. Small = below 100 employees, Large = larger than 100 employees. b. Low Capacity = below 75 percent capacity, High = higher than 75 percent capacity and above Source: World Bank, Investment Climate surveys, Uganda, 2002/03, Eritrea, Pakistan, 2002, China, 2000, India, 1999, Morocco, 2000. Appendix 4: Investment Climate Indicators 133 Table A4.5 Sources of Finance for Manufacturing Firms in Uganda, in International Comparison and by Firm Characteristic (percent) Low High Uganda Eritrea Pakistan Chinaa Indiaa Moroccoa Smallb Large capacityc capacity Sources for working capital Retained earnings 79.95 73.96 65.40 51.5 30.37 61.95 82.49 61.36 82.21 76.25 Banks and other financial institutions 5.65 23.45 7.40 20.6 36.10 19.55 13.36 4.59 4.83 6.69 Trade credit 5.30 0.26 4.60 4.1 -- 7.58 4.90 8.22 5.48 4.11 Equity 1.80 0.01 12.70 0.6 13.04 2.56 2.05 0.00 1.94 1.69 Informal sources 0.35 1.28 1.30 8.6 -- -- 0.38 0.13 0.01 0.08 Other 2.71 1.04 8.20 6.3 20.48 8.31 2.47 4.47 1.99 4.40 Sources for new investments Retained earnings 71.05 63.08 55.60 -- -- -- 74.44 50.48 72.27 65.86 Banks and other financial institutions 11.64 31.17 8.20 -- -- -- 8.91 28.20 11.10 12.47 Trade credit 0.47 0.00 1.70 -- -- -- 0.27 1.72 0.16 0.71 Equity 1.95 2.67 14.10 -- -- -- 2.27 0.00 2.70 0.00 Informal sources 1.46 0.58 2.60 -- -- -- 1.47 1.37 0.81 0.86 Other 4.51 2.50 11.00 -- -- -- 4.14 6.72 2.83 9.67 -- Not available. a. Data refer to shares of total capital (working capital and new investment). b. Small = below 100 employees, Large = larger than 100 employees. c. Low Capacity = below 75 percent capacity, High = higher than 75 percent capacity and above Source: World Bank, Investment Climate surveys, Uganda, 2002/03, Eritrea, Pakistan, 2002, China, 2000, India, 1999, Morocco, 2000. Appendix 4: Investment Climate Indicators 134 Table A4.6 Manufacturing Firms' Credit, Loans, and Liabilities in Uganda, in International Comparison and by Firm Characteristic (percent, except where otherwise specified) Low High Indicator Uganda Eritrea Pakistan China India Morocco Smallb Large capacityc capacity Share of firms with overdraft or line of credit 22.66 47.44 22.80 21.80 -- 77.38 19.31 47.22 19.46 30.50 Share of credit currently unused -- 29.53 43.50 29.70 -- 25.46 -- -- -- -- Share of firms with a loan from a bank or other financial institution 20.00 44.87 19.50 45.80 -- -- 16.66 44.44 17.25 30.50 For the most recent loan or overdraft Share requiring collateral 92.00a 85.71 80.30 82.80 -- -- 92.50 83.30 93.02 92.30 Average value of collateral required as a share of loan 116.04 168.06 70.90 86.80 -- -- 116.12 115.76 121.39 117.52 Average interest rate 16.70 9.81 14.80 5.40 -- -- 16.89 16.20 16.88 15.99 Average duration (months) 43.80 46.53 8.30 15.10 -- -- 39.00 55.92 39.84 47.16 Share of total borrowing denominated in foreign currency 8.70 2.86 0.50 7.70 9.49 3.89 6.21 25.91 6.64 17.18 Long-term (one year or more) liabilities as a share of total liabilities -- 29.50 7.10 9.40 28.69 8.96 -- -- -- -- Short-term liabilities as a share of total liabilities -- 66.53 13.50 47.00 22.82 46.63 -- -- -- -- Equity earnings (or share capital) and retained earnings as a share of total liabilities -- 52.80 63.40 43.70 48.48 30.73 -- -- -- -- -- Not available. a. Refers only to the most recent loan. b. Small = below 100 employees, Large = larger than 100 employees. c. Low Capacity = below 75 percent capacity, High = higher than 75 percent capacity and above Source: World Bank, Investment Climate surveys, Uganda, 2002/03, Eritrea, Pakistan, 2002, China, 2000, India, 1999, Morocco, 2000. Appendix 4: Investment Climate Indicators 135 Table A4.7 Financial Sector and Property Rights Indicators as Reported by Manufacturing Firms in Uganda, in International Comparison and by Firm Characteristic Low High Indicator Uganda Eritrea Pakistan China India Morocco Smalla Large capacityb capacity Financial sector Share of firms whose financial statements are audited by outside auditors (percent) 59.00 89.74 41.60 -- -- -- 54.90 88.80 56.63 70.68 Clearance time through firms' financial institution (days) For a check -- 1.44 1.90 4.50 -- -- 1.43 1.47 1.46 1.33 For a domestic currency wire -- 1.43 2.40 4.80 -- -- 1.45 1.40 1.94 0.80 For a foreign currency wire -- 6.33 3.20 3.00 -- -- 7.67 5.00 7.00 3.00 Property rights Land Share owned (percent) 62.69 -- 87.90 44.00 -- -- 62.64 63.05 64.19 66.66 Share leased or rented (percent) 34.30 -- 11.60 22.20 -- -- 33.94 36.94 32.11 42.45 Average length of lease or rental contract (years) 24.34 -- 1.25 6.28 -- -- 18.41 55.33 17.21 1.00 Buildings Share owned (percent) 73.06 -- 90.50 56.90 -- -- 70.95 88.75 73.84 71.79 Share leased or rented (percent) 26.51 -- 9.00 43.20 -- -- 28.61 11.25 25.93 26.95 Average length of lease or rental contract (years) 2.29 -- 7.63 4.43 -- -- 2.25 3.03 2.08 3.73 -- Not available. a. Small = below 100 employees, Large = larger than 100 employees. b. Low Capacity = below 75 percent capacity, High = higher than 75 percent capacity and above Source: World Bank, Investment Climate surveys, Uganda, 2002/03, Eritrea, Pakistan, 2002, China, 2000, India, 1999, Morocco, 2000. Appendix 4: Investment Climate Indicators 136 Table A4.8 Labor and Training in Manufacturing in Uganda, in International Comparison and by Firm Characteristic (percent, except where otherwise specified) Low High Indicator Uganda Eritrea Pakistan China India Morocco Smalla Large capacityb capacity Labor composition Share of workers who are permanent 56.09 91.13 86.60 85.30 -- -- 56.15 56.02 56.10 56.27 Share of permanent workers who are female 13.29 41.35 3.10 42.50 -- -- 7.75 54.54 11.21 22.01 Share of temporary workers who are female 29.46 29.46 -- 19.40 -- -- 27.11 32.78 32.76 33.03 Share of permanent skilled workers who are foreign nationals 2.94 2.94 1.95 -- -- -- 3.38 1.74 2.66 1.79 Labor turnover New employees in previous year as a share of total 4.40 14.92 8.30 -- -- -- 5.75 2.47 3.99 4.20 Employees who left in previous year as a share of total 1.60 17.32 5.50 10.20 -- -- 1.50 1.70 1.92 1.26 Average time to fill a skilled technician vacancy (weeks) -- -- 1.50 5.20 -- -- -- -- -- -- Average time to fill a production or service worker vacancy (weeks) -- -- 1.30 13.90 -- -- -- -- -- -- Excess workforce due to regulatory restrictions (as a share of total workforce?) -- -- 3.60 16.70 -- -- -- -- -- -- Training and education 0.00 Share of workforce with less than 6 years' schooling 12.76 -- 59.30 1.10 -- -- 17.21 6.08 7.31 21.39 Share of workforce with more than 12 years' schooling 8.48 -- 9.50 3.00 -- -- 8.23 8.87 9.74 6.10 Share of firms offering formal training 29.66 9.59 11.10 71.70 -- -- 78.65 21.34 74.69 25.30 Share of skilled workers receiving training 3.04 -- 36.00 47.70 -- -- 2.52 3.76 4.11 1.31 Labor unrest Days lost in previous year to labor disputes or civil unrest 0.52 -- 1.30 0.30 -- -- 0.42 1.25 0.61 0.29 -- Not available. a. Small = below 100 employees, Large = larger than 100 employees. b. Low Capacity = below 75 percent capacity, High = higher than 75 percent capacity and above Source: World Bank, Investment Climate surveys, Uganda, 2002/03, Eritrea, Pakistan, 2002, China, 2000, India, 1999, Morocco, 2000. Appendix 4: Investment Climate Indicators 137 Table A4.9 Regulatory Burden and Administrative Delays as Reported by Manufacturing Firms in Uganda, in International Comparison and by Firm Characteristic Low High Indicator Uganda Eritrea Pakistan China India Morocco Smalla Large capacityb capacity Regulation Share of firms disagreeing that interpretations of regulations are consistent and predictable (percent) 40.00 -- 64.80 14.10 -- -- 40.96 33.33 40.09 35.59 Share of senior management's time spent dealing with regulations (percent) 0.4 4.96 10.10 11.80 -- -- 0.04 0.07 0.04 0.06 Share of revenues typically paid to officials to get things done (percent) 2.44 1.55 2.00 1.40 -- -- 2.64 1.10 2.91 1.19 Share of firm revenues typically reported for tax purposes (percent) 76.71 84.20 -- 95.90 -- -- 75.17 87.29 75.70 79.13 Length of wait for business registration (days) -- -- -- 22.00 -- -- -- -- -- -- Inspections Days in previous year spent in inspections or required meetings with officials 13.43 6.32 41.10 29.10 10.70 -- 11.76 25.56 11.52 19.91 Share of meetings or inspections by local authorities (percent) 19.40 -- -- -- 68.00 -- 21.36 7.63 21.04 13.56 Cost of fines or seized goods (percentage of sales) 0.05 0.02 0.23 0.60 -- -- 0.02 0.27 0.06 0.00 Share of interactions in which informal payment requested (percent) 6.69 -- -- 2.20 -- -- 6.31 9.38 5.61 11.76 Value of informal payments (percentage of sales) 0.32 -- 0.15 -- -- -- 0.27 0.57 0.35 0.25 Import delays (days) Average wait to clear customs 5.79 11.79 17.20 7.90 10.60 2.70 5.88 5.53 5.23 6.88 Longest wait to clear customs 11.15 22.62 30.20 12.50 21.20 5.40 11.53 10.06 10.48 11.59 Export delays (days) Average wait to clear customs 3.53 3.20 9.60 5.40 5.00 1.70 3.22 4.18 3.32 3.00 Longest wait to clear customs 6.03 3.90 17.10 8.00 9.20 2.70 5.91 6.30 6.21 3.75 -- Not available. a. Small = below 100 employees, Large = larger than 100 employees. b. Low Capacity = below 75 percent capacity, High = higher than 75 percent capacity and above Source: World Bank, Investment Climate surveys, Uganda, 2002/03, Eritrea, Pakistan, 2002, China, 2000, India, 1999, Morocco, 2000. Appendix 4: Investment Climate Indicators 138 Table A4.10 Indicators of Uncertainty and Corruption as Reported by Manufacturing Firms in Uganda, in International Comparison and by Firm Characteristic (percent, except where otherwise specified) Low High Indicator Uganda Eritrea Pakistan China India Morocco Smalla Large capacityb capacity Uncertainty Share of firms disagreeing that interpretations of regulations are consistent and predictable 40.00 -- 64.80 14.10 -- -- 40.96 33.33 40.09 35.59 Share of profits reinvested in firm 41.87 -- -- -- -- -- 41.76 42.71 39.31 46.83 Share of firms disagreeing that they have confidence in the judiciary 69.86 -- 4.10 2.40 -- -- 71.14 61.11 75.36 56.90 Share of payment disputes settled by third parties or resolved in court 50.00 -- 30.20 5.30 -- -- 48.65 53.33 44.12 64.29 Planning horizon for investments (months) -- -- -- 2.30 -- -- -- -- -- -- Corruption Informal payments required as a share of revenues 2.44 1.55 -- 1.40 -- -- 2.64 1.10 2.91 1.19 Share of firms reporting requirement for gift or payment For a mainline telephone connection 18.32 -- -- -- -- -- 16.36 28.57 16.16 23.08 For an electricity connection 21.47 -- -- -- -- -- 21.48 21.43 22.95 19.44 For a construction permit 12.33 -- -- -- -- -- 11.11 20.00 8.93 25.00 For an import license 3.64 -- -- -- -- -- 4.26 0.00 4.76 0.00 For a trading license 4.21 -- -- -- -- -- 4.31 3.45 3.52 7.69 Share of revenue typically reported for tax purposes 76.71 84.20 -- 95.9 -- -- 75.17 87.29 75.70 79.13 -- Not available. a. Small = below 100 employees, Large = larger than 100 employees. b. Low Capacity = below 75 percent capacity, High = higher than 75 percent capacity and above Source: World Bank, Investment Climate surveys, Uganda, 2002/03, Eritrea, Pakistan, 2002, China, 2000, India, 1999, Morocco, 2000. References 139 Anderson, James E. 1998. "Effective Protection Re- International Monetary Fund. 1999. Uganda: Selected dux." Journal of International Economics 44: Issues and Statistical Appendix. Country Report 22­44. 99/116. Washington, D.C. Bankunda, Geoffrey. 2004. "Corporate Managers' ------. 2002a. Uganda: Selected Issues and Statis- International Orientation and the Export Per- tical Appendix. Country Report 03/84. Washing- formance of Firms in Uganda." Eastern Africa ton, D.C. Social Science Research Review 20 (1): 27­50. ------. 2002b. Uganda: Staff Report for the Article IV Biggs, Tyler, and Manju Shah. 1997. "Trade Reforms, Consultation. Country Report 03/83. Washington, Incentives on the Ground, and Firm Performance D.C. in Ghana." RPED Working Paper 105. World Bank, ------. 2003a. Uganda: Financial Stability Assess- Regional Program on Enterprise Development, ment. Country Report 03/97. Washington, D.C. Washington, D.C. ------. 2003b. Uganda: Report on Observance of Bonaglia, Federico, and Kiichiro Fukasaku. 2002. Standards and Codes. Country Report 03/85. Trading Competitively: A Study of Trade Capacity Washington, D.C. Building in Sub-Saharan Africa. Paris: OECD Kasekende, Louis, and Henry Opondo. 2003. Development Centre. "Financing Small and Medium-Scale Enterprises: Collier, Paul. 1997. "The Trade Policy Review of Uganda's Experience." Bank of Uganda Working Uganda." World Economy 20: 649­64. Paper WP/03/01. Kampala. Consultative Group Meeting. 2003. "Statement of Lindauer, David L., and Ann D. Velenchik. 1994. "Can Uganda's Development Partners on Governance African Labor Compete?" In David L. Lindauer and Anti-Corruption." Kampala. April. and Michael Roemer, eds., Asia and Africa: EIU (Economist Intelligence Unit). 2002. Uganda Legacies and Opportunities in Development. San Country Report. London. Francisco: ICS Press. ------. 2003. Uganda Country Report. London. Michalopoulos, Constantine. 1999. "Trade Policy and ------. 2004. Uganda Country Report. London. Market Access Issues for Developing Countries: Ettori, François M. 1992. "Measure and Interpretation Implications for the Millennium Round." Policy of Effective Protection in the Presence of High Research Working Paper 2214. World Bank, Capital Cost: Evidence from India." Policy Development Research Group, Washington, D.C. Research Working Paper 873. World Bank, Asia Mincer, Jacob. 1974. Unemployment Effect of Minimum Region, Country Department IV, Washington, D.C. Wages. NBER Working Paper 39. Cambridge, FIAS (Financial Investment Advisory Services, Mass.: National Bureau of Economic Research. International Finance Corporation). 2002. Morrissey, Oliver, and Nicodemus Rudaheranwa. 1998. Uganda: Administrative Barriers to Investment "Ugandan Trade Policy and Export Performance in Update. Washington, D.C. the 1990s." CREDIT Discussion Paper 98/12. Financial Times. 2002. "Ugandan Project Faces University of Nottingham, Centre for Research in Attack over Cost of Power." November 21. Economic Development and International Trade. ------. 2003. "Power behind the Throne." Novem- Newbery, David, and Nicholas Stern, eds. 1987. The ber 8. Theory of Taxation for Developing Countries. New Henstridge, Mark. 1996. Coffee and Money in York: Oxford University Press. Uganda: An Econometric Analysis. Oxford: Oxford OECD (Organisation for Economic Co-operation and University Press. Development). 2003. Public Procurement: Les- References 140 sons from Kenya, Tanzania, and Uganda. OECD World Bank. 1999. "Uganda: Project Appraisal DEV/DOC (2003)06. Paris. Document for Financial Markets Assistance Private Sector Foundation Uganda. 2002. Annual Project." Report 19264UG. Washington, D.C. Report. Kampala. ------. 2001. World Development Indicators 2001. Sachs, Jeff and David Bloom. 1998."Geography, Washington, D.C. Demography, and Economic Growth in Africa." ------. 2003a. Doing Business Database. Washing- Brookings Papers on Economic Activity. Volume 2. ton, D.C. Washington, D.C. ------. 2003b. The Republic of Uganda Public Ex- Stawowy, Wojciech. 2001. "Calculation of Ad Valorem penditure Review, 2003. September. Washington, Equivalents of Non­Ad Valorem Tariffs-- DC. Methodology Notes." DITGSC-UNCTAD Working WTO (World Trade Organization). 1995. Trade Policy Paper. United Nations Conference on Trade and Review: Uganda, 1995. Geneva. Development, Geneva. ------. 2001. Trade Policy Review: Uganda, 2001. UBOS (Uganda Bureau of Statistics). 2001. website: Geneva. http://www.ubos.org/. ------. 2003. World Trade Report 2003. Geneva. UNCTAD (United Nations Conference on Trade and Development). 2002. World Investment Report. Geneva.