Report No. 38627-ZA South Africa Enhancing the Effectiveness of Government in Promoting Micro, Small and Medium Enterprise February 14, 2007 Private Sector and Finance Africa Region Document of the World Bank MICA Micro-enterprise Investment Climate Assessment MIDP The Motor Industry Development Program MSMEs Micro, Small and Medium Enterprises NAMAC National Office of the Manufacturing Advisory Centers NEF National Empowerment Fund NFKTA National Fabric Knitters' Trade Association OEM Original Equipment Manufacturers RFI Retail Financing Intermediary SACTMA South African Cotton Textile Manufacturers' Association SAWMTA South Africa Cotton Worsted Manufacturers' Trade Association SARS South African Revenue Services SBDCs Small Business Development Centers SEDA Small Enterprise Development Agency SETA Sector Education and Training Authority SIP Strategic Industrial Projects SME Small and Medium Enterprises SMEDP Small and Medium Enterprise Development Program SMMEs Small, micro and medium enterprises SP Sector Partnership SPF Sector Partnership Fund SPII Support Program for Industrial Innovation TEO The Enterprise Organization THRIP Technical and Human Resources for Industry Program TIPS Trade and Industry Policy Strategies TISA Trade and Investment South Africa TWIB Technology For Women VC Value Chain WTO World Trade Organization Acting Vice President: Hartwig Schafer Country Director: Ritva Reinikka Sector Manager: Demba Ba Task Team Leader: Dileep Wagle ___________________________________ Acknowledgments This report was prepared by a team led by Dileep Wagle (AFTPS), and comprising Michaela Weber (AFTPS), George Clarke (AFTPS), David Phillips (GBRW, consultant) and Yasuo Konishi (GDS, consultant). Shirley Faragher (AFCS1) provided assistance in preparation of the report. The team benefited from comments and contributions from Vijaya Ramachandran (Georgetown University) and Michael Ingram (CSMPF). Field work associated with the report was conducted during 2005 and 2006. The survey of micro-enterprises was conducted with the assistance of the Bureau of Market Research, University of South Africa. Data used for the value chain analysis was collected in the field by Global Development Solutions, LLC. The team is grateful to DTI for facilitating the work on the assessment of the incentive framework, through access to program evaluation reports and relevant data on the programs selected. The report was prepared under the general supervision of Ritva Reinikka (Country Director for South Africa) and Demba Ba (Sector Manager, AFTPS). Peer reviewers were Alan Gelb (DECVP), Arvind Gupta (EASFP) and Roy Pepper (CICFI). Advice and comments at different stages of the work were also provided by Zeljiko Bogetic (AFTP4), Bernard Drum (AFTPS), Tonia Overmeyer (University of Cape Town) and Robert Cull (DECRG). The report was discussed, prior to finalization, with DTI's Executive Board, and at the technical level with senior staff at the departmental level, in July, 2006. The report was also discussed with National Treasury, as well as with SEDA, and representatives from other support agencies including the Khula Fund, SMEDP, IDC and the Gauteng Enterprise Propeller. The team benefited greatly from these discussions, as well as with Prof. Christian Rogerson, University of Witwatersrand. TABLE OF CONTENTS Page Executive Summary............................................................................. i 1. THE ECONOMIC AND POLICY BACKGROUND 1 A. Economic Trends Since Independence ..................................................... 1 B. The Evolving Role of SMMEs .................................................................. 4 C. The Evolution of Industrial Support ....................................................... 7 D. The Industrial Policy Rationale ........................................................ 10 E. The Purpose of This Study ............................................................... 14 - Approach .................................................................................... 15 - Opportunity ................................................................................. 15 - Structure of the Report ..................................................................... 16 2. WHAT ARE THE CONSTRAINTS ON AND THE PUBLIC SUPPORT NEEDS OF THE SMME SECTOR? 17 A. The Business Climate for Formal Sector Enterprises ................. ............. 17 - Background: The Investment Climate Assessment ..................................... 17 - Worker Skills .............................................................................. 18 - Labor Regulations ......................................................................... 19 - Macroeconomic Instability ............................................................... 19 - Crime ........................................................................................ 20 - Access to Finance .......................................................................... 20 - Taxation ..................................................................................... 20 - Other Areas of the Formal Sector Business Climate ................................. 20 - Summary of ICA Findings ............................................................... 21 B. The Micro-Enterprises (MICA) Survey................................................ 21 - Micro-enterprise Performance ............................................................ 24 -The Investment Climate and Constraints on MEs ...................................... 26 - Earlier (1999) Johannesburg Survey ................................................ 26 - MICA Findings on Business Constraints .......................................... 28 - Access To and Cost of Finance .................................................... 30 C. Incentive Programs........................................................................ 33 D. Conclusions: the Informal and Formal Sector Environment ...................... 36 3. HOW FAR ARE GOVERNMENT SUPPORT MEASURES CONFIGURED TO ADDRESS 37 SMME NEEDS? A. Introduction: Surveys of Institutions and Beneficiaries ...................... ... 37 - Government Support Programs ......................................................... 37 - Program Survey ........................................................................... 38 - Beneficiary Survey ........................................................................ 39 B. Costs and Benefits from the Providers' View ­ Review of Government Support Program ......................................................................... 40 - International Benchmarks ................................................................. 40 - Performance Scores........................................................................ 42 - Main Findings on Performance .......................................................... 43 - Summary Performance rating ......................................................... 44 - Economic Rationale ................................................................... 44 - Eligibility and Selection of Firms ................................................... 45 - Operating Efficiency .................................................................. 45 - Economic impact/cost effectiveness ................................................ 46 - Inspection, verification and M & E .................................................. 47 - What are the Lessons? ...................................................................... 47 C. The Performance of Four Critical Programs ......................................... 49 (a) The SMEDP .............................................................................. 50 (b) Khula Enterprise Finance RFI Program ............................................... 54 (c) Ntsika Enterprise Promotion Agency ­ LBSCs ..................................... 57 (d) The Industrial Competitiveness and Job Creation Project .......................... 60 - The Competitiveness Fund ........................................................... 60 - The Sector Partnership Fund ......................................................... 63 - The Black Business Suppliers Development Program ........................... 67 D. Costs and Benefits from the Users' View - Results of Beneficiary Surveys .... 68 - Is the Government's strategy helping small firms? ..................................... 69 - How far do firms have any knowledge of Government/DTI support? .............. 69 - How did firms find out about Government/DTI programs? .......................... 70 - What further assistance would the firms apply for? .................................... 71 - How did firms interact with their facilitator/consultant? .............................. 72 - How well did the firms interact with DTI? ............................................. 73 - Is the publicity/promotion of the Government's programs adequate? ............... 74 - How effective were the projects undertaken? .......................................... 75 - How good was program management? .................................................. 76 - Firm's willingness to pay for services? ................................................. 78 E. Summary of the Program Effectiveness Issues ....................................... 79 - Lack of Knowledge of Government Small Firm Strategy ........................... 79 - Ineffective Promotion and Publicity .................................................... 79 - Operational Inefficiencies.................................................................. 80 - Project Outcome/Effectiveness .......................................................... 80 - Program Impact and Selection Criteria .................................................. 80 - Pricing/Public Benefit Targeting .......................................................... 80 - Reapplications and Future Needs ......................................................... 81 F. The Pattern of Government Assistance to SMMEs .................................. 81 G. Improving the SMME Support Infrastructure ...................................... 83 - The Key Issues ............................................................................. 83 - Consolidation and Rationalization ...................................................... 84 - Independent Fund Management ......................................................... 85 - The Role of DTI and TEO if Programs are Spun Off ................................. 85 - Public-Private Initiatives .................................................................. 86 - Using Large Scale Business to Support Small Business .............................. 87 4. INDUSTRY CASES: USING VALUE-CHAIN ANALYSIS TO IDENTIFY OPPORTUNITIES FOR SMME INTEGRATION AND GOVERNMENT SUPPORT 89 A. The Role of Value-Chain Analysis ..................................................... 89 B. Cotton-to-Garments Supply Chain .................................................... 91 - Industry Overview ........................................................................ 91 - Cotton Production ......................................................................... 93 - Cotton Ginning ............................................................................ 99 - Textile Fabrics Production ............................................................... 101 - Garment and Apparel Production ....................................................... 104 - Industry Structure ....................................................................... 104 - The garment value chain ............................................................... 106 - Production economics .................................................................. 107 - Government Incentives: the DCCS ..................................................... 110 - The international garment market .................................................... 112 - The Garment Industry: Adjustment to International Competition ................ 113 - Key Market Drivers and Strategic Options for the Industry and the SMMEs ..... 115 - Implications for future assistance programs .......................................... 116 C. Automotive Components Value Chain ................................................. 119 - Sub-sector profile .......................................................................... 119 - The Impact of the MIDP .................................................................. 120 - Leather Seat Covers: Market Structure and the Supply Chain ...................... 123 - Integrated Value Chain Analysis for Seat Covers ................................... 124 - Seat Covers for export ............................................................... 125 - Seat covers for the domestic market ................................................ 128 - Leather seat covers: implications for SMME entry and future assistance programs ................................................................................. 130 - Forged Steel Towing Ball: The Value Chain ........................................... 133 - Steel Towing Balls: implications for SMME entry and future assistance programs .................................................................................. 136 5. IMPROVING THE ECONOMIC IMPACT AND EFFICIENCY OF PUBLIC SUPPORT TO SMMES : THE WAY FORWARD 138 A. Reviewing the Constraints on SMMEs and the Economic Case for Intervention ................................................................................................................ 138 B. Reviewing the Implementation of Assistance Programs ............................ 141 - Key Effectiveness lessons ................................................................ 142 - Good Promotion of Assistance Programs is essential to their Effectiveness .... 142 C. Reviewing Entry Points for Micro and Small Firms ­ The Case of Textiles, Auto Components and Construction Equipment Industries .......................... 144 - The DCCS and the MIDP ................................................................ 145 - SMME Entry Points ....................................................................... 145 - Cotton and Cotton Textiles ............................................................ 146 - Engineered products (construction equipment and automotive components) ... 147 - Intermediate processed products (Leather seat covers) ............................ 147 D. Reviewing Areas of General State Support to SMMEs ............................ 148 - Skills Development ........................................................................ 148 - Entrepreneurship ........................................................................... 150 - Investment Finance ........................................................................ 150 - Overall Productivity and Competitiveness ............................................. 150 E. The Pattern of Government Assistance ................................................ 151 F. Improving the SMME Support Infrastructure ....................................... 152 - The Key Issues ............................................................................. 153 - Recommendations ......................................................................... 153 - Consolidation and Rationalization .................................................... 154 - Spin-Off and Independent Fund Management ....................................... 154 - Centralization and Decentralization ................................................... 154 - The Role of DTI ......................................................................... 155 - Public-Private Initiatives .................................................................. 155 - Using Large Scale Business to Support Small Business .............................. 155 - Final thoughts on program design ....................................................... 156 - Some Final Comments on the Objective of BEE ..................................... 156 Annexes: Page 1 KEY LABOR LEGISLATION ............................................................... 158 2 CHARACTERISTICS OF ICA FIRM SURVEY SAMPLE........................................ 159 3 SOUTH AFRICAN EXCHANGE RATE TRENDS.......................................... 160 4 MICA SAMPLE CHARACTERISTICS...................................................... 161 5 PRODUCTIVITY MEASURES BY FIRM TYPE, MES IN SOUTH AFRICA .................... 163 6 BENEFICIARY ASSESSMENT SURVEY METHOD ................................. 164 7 QUALITATIVE PERFORMANCE SCORES ............................................ 166 8 ERODING LABOR SKILLS, DECLINE OF APPRENTICESHIPS AND 170 LACK OF INVESTMENT IN RESEARCH AND DEVELOPMENT .............. 9 INCENTIVE PROGRAMS ................................................................. 177 10 TARIFF ON TEXTILES AND TEXTILE GOODS ................................................... 182 11 COTTON PRODUCTION AND MARKET FLOW IN SOUTH AFRICA .................... 183 12(A) COTTON GINNING VALUE CHAIN ...................................................... 184 12(B) COTTON TEXTILE SPINNING AND WEAVING VALUE CHAIN ........................... 185 13 COTTON TEXTILE SPINNING AND WEAVING VALUE CHAIN : BREAKDOWN OF COSTS ........................................................................................... 186 14 TRADE TRANSPORT LOGISTICS ............................................................... 187 15 A NOTE ON THE STEEL SECTOR IN SOUTH AFRICA .................................. 189 16 KEY LABOR LEGISLATION ............................................................... 199 17 REFERENCES AND SOURCES USED ............................................. 200 Tables: Page Table 1.1: Distribution of formal and informal businesses ................................... 6 Table 1.2: Number and % of informal workers by sector, 2001 ............................. 6 Table 1.3: Percentage of Workers Receiving Firm-Based Training, by Country ........ 12 Table 2.1: Productivity Measures by Firm Type, Micro-enterprises in South Africa . 25 Table 2.2: Access to finance by enterprise characteristic ................................ 32 Table 2.3: Micro-enterprises (are far less likely to say that they don't have a loan because they don't want one) ................................................... 33 Table 3.1: Selected programs and their objectives ........................................... 39 Table 3.2: Comparative Matching Grant Fund Performance Benchmarks .................. 41 Table 3.3: Rating descriptions .................................................................... 43 Table 3.4: Summary Performance Results for Selected Support Programs ............. 43 Table 3.5: Approvals and Disbursements .................................................... 51 Table 3.6: Summary of some SMEDP Performance Indicators ............................ 53 Table 3.7: Government Recurrent Funding of Khula ........................................ 54 Table 3.8: Khula RFI Program Output ...................................................... 55 Table 3.9: Khula-funded RFI Loan Activity ................................................. 55 Table 3.10: Aggregate Ntsika and LBSC Network Reported Outcome in 2004 ........... 59 Table 3.11: CF Grants Approved and Cancelled ............................................. 60 Table 3.12: Change in performance of CF Beneficiaries 1999-2004 ...................... 61 Table 3.13: Difference in Performance of Beneficiary and Control Group Firms ......... 62 Table 3.14: Benefits versus Costs .............................................................. 63 Table 3.15: SPF partnership sample ............................................................. 64 Table 3.16: Assisted Firm vs. Control Group Performance ................................. 65 Table 3.17: Main Management Efficiency Measures ......................................... 66 Table 3.18: Processing Time for Grants ....................................................... 68 Table 3.19: Beneficiary Survey ­ Selected Responses ......................................... 69 Table 3.20: Knowledge of Government/DTI Support ........................................ 70 Table 3.21: How Effective were Projects Undertaken? ...................................... 76 Table 3.22: How Good was Program Management? ......................................... 76 Table 3.23: Processing Efficiency ............................................................. 77 Table 3.24: Summary of Assessed Public Support Schemes ............................... 82 Table 4.1: Historical Production of Lint Cotton in South Africa ......................... 94 Table 4.2: South Africa Cotton Production Sub-Sector Profile............................ 95 Table 4.3: South Africa Cotton Yield Rates ................................................. 95 Table 4.4: Breakdown of Irrigated Cotton Production Costs ............................... 97 Table 4.5: Average Profits for Irrigated Cotton ............................................. 97 Table 4.6: Benchmarking Cotton Farming Cost and Yield ................................... 98 Table 4.7: South Africa's Ginning Production Sub-sector Profile......................... 99 Table 4.8: Benchmarking Ginning Costs ..................................................... 100 Table 4.9: South Africa's Textile Production Sub-sector Profile .......................... 102 Table 4.10: Employment and Wages in the Textile Industry 2004 ........................ 102 Table 4.11: Comparison of the Cost of Lint Cotton in Selected Countries ................ 103 Table 4.12: Benchmarking Textile Fabric Production Cost ($/Kg) .......................... 104 Table 4.13: South Africa's Garment Production Sub-Sector Profile ....................... 105 Table 4.14: Sewing Assembly Production Cost Breakdown ................................ 106 Table 4.15: Cotton T-Shirt Assembly Cost Benchmark ..................................... 107 Table 4.16: FOB Price Comparison for 180g T-shirt ($/unit) ............................... 107 Table 4.17: Comparison of Western Cape and KwaZulu Natal Clothing Sectors ....... 107 Table 4.18: Characterization of Two Categories of Apparel Manufacturers in South Africa ........................................................................................................ 108 Table 4.19: Major Characteristics of the South African Apparel Industry................. 109 109 Table 4.21: Comparative Features Of Garment Industry in South Africa and China ... 110 Table 4.22: DCCS Certificate Value 2005 .................................................... 111 Table 4.23: Main African exporters of garments to the U.S. Under AGOA ............. 113 Table 4.24: Characteristics of European and American Brand Name Holders .............. 113 Table 4.25: DTI and Associated Support Schemes applicable to CMT unit production of Garments .......................................................... 117 Table 4.26: Employment and Turnover per Size, Auto Industry, South Africa, 2004 .. 120 Table 4.27: South African Automotive Import Tariff Schedule ........................... 122 Table 4.28: Effective Protection to Exports by the IRCC Facility ........................ 122 Table 4.29: Benchmarking leather Price in EU and South Africa (Rand/mē) ............. 126 Table 4.30: Benchmarking Labor Productivity for Leather Seat Cover Production in India, China, EU and South Africa (units/hour) ..................................... 127 Table 4.31: Benchmarking Defect Rates. Seat Covers China, India, South Africa (ppm) ............................................................................... 127 Table 4.32: Leather Seat Cover Market Potential .......................................... 128 Table 4.33: DTI and associated Support schemes applicable to production of leather seat covers ......................................................................... 131 Table 4.34: Tow Ball Production in South Africa: Sample Manufacturer Profile.............................................................................. 135 Table 4.35: Benchmarking South Africa's Production of Tow Balls with International Industry Norms ................................................... 135 Table 4.36: DTI and associated Support schemes applicable to production of steel tow-balls ........................................................................... 136 Boxes: Page Box 3.1 Respondent: A bedding firm in Rosslyn ........................................... 73 Box 3.2 Respondent: A firm producing fire extinguisher medium in Walkerville ... 74 Box 3.3 Respondent: A software development firm in Edenvale ...................... 75 Box 3.4 Respondent: An engineering firm in Pretoria ................................... 77 Box 3.5 Respondent: A pharmaceutical firm in Johannesburg .......................... 78 Figures: Page Figure 1.1: Growth and Investment have been lower in South Africa than in the best performing middle-income economies ........................................ 2 Figure 1.2: Labor productivity is higher in South Africa than elsewhere in Africa and higher than, or comparable to, other middle-income economies and the most productive areas of China .......................................... 3 Figure 1.3: Investment has been modest, especially compared to the fast growing Asian economies.................................................................. 13 Figure 2.1: Worker skills, macroeconomic instability, labor regulation and crime are firms' biggest concerns in South Africa .................................. 18 Figure 2.2: Labor Regulations are stricter in South Africa than in the best performing comparator countries or in the OECD ........................... 19 Figure 2.3: About half of the firms in the micro-enterprise survey reported that they were registered with one or more government agency ............ .......... 23 Figure 2.4: Labor productivity is higher for MEs in the manufacturing sector in South Africa than for similar firms elsewhere in Africa and other middle-income economies ...................................................... 24 Figure 2.5: MEs in the manufacturing sector have very different concerns from the larger firms in the investment climate survey ................................. 30 Figure 2.6: MEs are consistently more concerned about access to finance throughout the world ........................................................................ 32 Figure 2.7: Few firms reported knowing about any of the incentive programs-- especially smaller firms and firms in Stellenbosch .......................... 34 Figure 2.8: Over 75% of managers were interested in incentive programs ­ mostly related to new investment, training in management skills, and loan guarantees ........................................................................ 35 Figure 3.1 Information Sources ............................................................... 71 Figure 3.2 Type of Program demanded ........................................................ 72 Figure 3.3 Firms Interaction with Facilitator .................................................. 72 Figure 3.4 Publicity and Promotion of Programs............................................. 74 Diagrams: Page Diagram 4.1 Value Chain for Irrigated Cotton Production in South Africa .......... 96 Diagram 4.2 Value Chain for Garment Production ­ T-shirts ............................. 106 Diagram 4.3 Duty Credit Certification Scheme (DCCS)* ................................. 111 Diagram 4.4 MIDP Import Rebate System ................................................... 121 Diagram 4.5 Value Chain for Leather Seat Cover Exported to the EU under the name of the OEM ..................................................................... 125 Diagram 4.6 Value Chain ­ Leather Seat Cover Supplied to Domestic OEMs ......... 129 Diagram 4.7 Value Chain ­ Forged Steel Towing Ball .................................... 133 EXECUTIVE SUMMARY Background and Context 1. Since the early 1990s, South Africa has achieved much in terms of macroeconomic stability, improvement in the business climate, and development of industrial support measures, especially within the `first economy'1 which is operating in an increasingly open trade environment. Growth, which was steady at 2.9% over the past decade, has picked up more recently to 4.5% in 2004, 4.9% in 2005, and a projected 4.7% in 20062. Per capita incomes (at $4960) are among the highest in Sub-Saharan Africa. However, a high degree of economic inequality, the legacy of historical times, still persists, reflected in an unemployment rate variously measured at between 26% to over 40%, making the creation of jobs a critical priority. 2. Given the increasingly capital-intensive nature of the `first economy' in South Africa, there is a much greater potential role for small firms to play in job-creation compared to other economies. Growth in employment in SMMEs in South Africa has, in fact, increased much more rapidly than their contribution to GDP, changing from a ratio of 1.35 in 1995 to 1.5 in 2001. According to recent estimates by DTI3, small businesses represent 98 % of the total number of firms and employ 55% of the country's labor force, contributing approximately 42% to the total wage-bill. Small firms, overall, account for 35% of GDP. 3. The Government has been aware of the importance of the SMME sector, and has actively tried to support its growth through establishment of a comprehensive incentive framework incorporating a number of programs and institutions. A new SMME strategy was presented by DTI in 2003, articulating the Government's strategic directions, creating a new Small Enterprise Development Agency and promoting the objective of broad-based Black Economic Empowerment (BEE). Growth and performance of the sector has, however, not been commensurate so far with the efforts and investments involved. Though there has been a large increase in small and micro-business start-ups over the past decade, survival rates have also been low and a critical mass of small enterprises has yet to be integrated into a number of sub-sectors. In addition, the impact of the various incentive programs, in terms of economic returns and additionality, has been less than consistent. 1First and second economy. The second economy in South Africa is usually defined as the informal and micro/small businesses which are largely owned by historically disadvantaged social groups; the first economy is defined as the other small and medium firms, and the large firms. 2EIU Country Report, April 2006. 3Department of Trade and Industry 2003: "National Strategy for the Development and Promotion of Small Business in South Africa", Pretoria, South Africa. i Focus and Approach of the Study 4. This focus of the study flows from this background. In particular, the question arises of whether incentives and support programs have: (a) been correctly targeted to address the diverse and specific needs of small, especially micro, enterprises; (b) been implemented efficiently by the responsible agencies in terms of their delivery and impact, and (c) have been effective in helping smaller firms access a wider market for their products and services. 5. To assess this, the study adopts the following approach of conducting: · A survey of SMME firms, using the same core questionnaire of the Investment Climate Assessment ­ which would provide consistency for benchmarking ­ but with additional questions included, specifically on the incentives framework. · Analysis, through directed interviews and impact assessment reviews, of how government agencies have been implementing the incentive schemes. This analysis is supplemented, on the demand side, through a small survey of incentive program beneficiaries. The analysis is intended to look more carefully at the economic rationale of the menu of programs, their actual performance and economic impact, the efficiency of their implementation and their effectiveness in reaching out to prospective applicants. · Assessment, through analysis of the value chains of some representative sub- sectors, of the potential for integration of SMMEs into larger product supply chains. Such an exercise would be expected to (a) throw light on the tension between the competitiveness and social agendas, and (b) help identify the entry points for better SMME integration, and measures necessary to take advantage of them. In selecting sub-sectors for this analysis, the choice for the value chain analysis was influenced by considerations of employment potential and the size of incentives being provided for this purpose. Public Support needs of SMME firms 6. Findings of earlier ICA Survey: An investment climate assessment for formal sector business had been conducted in 2004, involving a survey of about 800 firms, predominantly European-owned, and a mix of small, medium and large (plus very large) sizes. Relatively few of these firms reported facing any severe obstacles, but their biggest constraints were: worker skills, macroeconomic instability, labor regulations and crime. Exporters were more concerned about macroeconomic stability than other firms and African-owned firms were slightly more concerned about access to and cost of financing than other firms. Foreign corporations operating in South Africa tended to see the skills and economic stability issue as more serious than other firm categories. Crime was relatively important for Asia-owned firms because they located more in urban areas. Business licensing, communication facilities and access to land were generally not regarded as important. ii 7. Micro-enterprise Survey: Against this background, the findings of the survey of 240 small and micro-enterprise firms of less than 10 employees in 3 locations, based on a similar (but enhanced) core questionnaire, revealed somewhat different results. Respondent firms reported being relatively more concerned about access to finance and infrastructure (access to transportation, electricity, business space and land). There was far less concern about other areas of the investment climate, including regulation, taxation, or official corruption. 8. The survey also revealed that micro-enterprises were not very well integrated into larger product supply chains or, for that matter, into the formal sector in general. Only 2% of their sales overall went to larger firms or to government agencies, and only 8% to other small firms and trade intermediaries. Firms also reported being largely unaware of the various support programs being offered by the Government, despite the fact that 75% of managers reported being very interested in assistance for development of skills and training. Implementation of Support Schemes 9. A review was conducted of 11 representative government programs aimed at supporting small businesses through either management and technical assistance or financial assistance. Eight of these programs were funded largely by the Government, and three by the World Bank. The amount of public funding of these schemes ranged widely from less than R30 million per annum (US$4 million) in the case of the Ntsika LBSC program, to as much as R2.0 billion per annum (US$300 million) for the SMEDP. 10. The eleven assistance programs selected represented a cross section of public sector assistance types. Initial criteria for program selection were that the programs (a) be directed to SMMEs, (b) be representative of types of support, and (c) have, if possible, both a national and provincial focus. In the event, most programs chosen were at the national level. Only two were local, and one of these was a local project of a national organization. Further work would be needed to investigate the performance of the provincial, and non-Government sectors, but the assessment of national level programs is per se important as it is the source of main thrust of SMME support. The following table lists the selected programs and their specific objectives: iii Institution/program Main Public Interest Objective National level DTI/TEO ­ Small Medium Enterprise Development Increase general investment, especially for labor-using small-medium Program (SMEDP) enterprises DTI/EMIA(TISA) National Pavilions Facilitate access to export markets through collective initiatives DTI/ Khula Finance RFI program Strengthen microfinance capacity and micro-enterprise growth DTI/ EIDD (IDC) SPII Accelerate technological advance and growth by strengthening research, development and innovation DTI/ National Empowerment Strengthen black business investment/ market entry through diverse Fund investment finance DTI/Ntsika (SEDA) local General business development and technical advisory support for business service centers SMMEs Provincial level GODISA Trust ­ Softstart Promoting startup businesses through business incubator (targeting ICT (Gauteng) sub-sector) Gauteng Enterprise Propeller Supporting entry and survival by advisory/financial support services to provincial, small, black-owned business (Under GEDA) 11. The review was conducted on the basis of structured interviews with the management of the programs, audited accounts and prior evaluation studies. Programs were evaluated on the basis of a common set of criteria (output, outcome, impact, cost- effectiveness and performance), using international benchmarks where appropriate (e.g. for matching grants funds). To provide comparability across programs reviewed, a scoring system was used to rate performance on each of the criteria evaluated. 12. Results of the evaluation were mixed. The best performers were found to be the EMIA National Pavilions and the Support Program for Industrial Innovation (SPII). Of the three World Bank-supported funds, the Competitiveness Fund came out best, and comparably with the EMIA and SPII programs. All of these exhibited a combination of appropriate economic rationale, good assisted-firm performance, the likelihood of spillover benefits, and reasonable operating efficiency. Weaker programs in terms of rationale and operational capacity were firstly the SMEDP, which implemented excessively loose selection criteria, with consequential impact on economic outcomes; the NEF; and Khula Enterprise Finance's Retail Financial Intermediary (RFI) program, which has appeared to be weak on selection criteria and economic impact. 13. Key lessons: (a) A general problem was found to be that of overambitious or over-complex program goals, lack of targeting and duplication of effort. In several cases it would have been better to focus on a few successful projects to use as models for dissemination rather than to try and scale up in haste; (b) There was a continual problem of excessive administrative burden, derived partly from the need to install safeguards against misuse of public funds, which also led to the use of considerable management time in grant applications and claims ­ and in turn to the widespread use of consultants and facilitators to help firms make applications in return for a fee. (c) In terms of economic impact, overall results were not very positive. Programs often tended to iv become self-justifying over time ­ their outputs being confused with their impact. This made it difficult to monitor them effectively and make adjustments as appropriate. (d) Similarly, there was not enough time for adequate learning to take place. Since 1995, programs were set up very rapidly, before the lessons of earlier programs could properly be taken into account. As a result, many programs failed to avoid some of the pitfalls that they were specifically designed to avoid ­ such as the tendency to become supply- driven, to use a one-size-fits-all approach to business assistance, or to become isolated from other parallel agencies instead of collaborating with them. In retrospect it may have been better to have focused on a few good models and a reduction in the number of programs. 14. Costs and Benefits from Users' Points of View: Results of a small survey of (23) beneficiary firms, provides a bottom-up perspective on these programs. Key findings were that: (a) Most firms expressed concerns about accessibility and bureaucracy (of government programs); (b) Very few firms had any detailed knowledge of these programs prior to application; (c) When firms found out about these programs it was usually through other private business sources (i.e. word-of-mouth), not via any public radio or TV source ­ indicating that publicity and promotion of these schemes could undoubtedly be improved; (d) On a positive note, when firms did access programs they were generally happy with the outcomes in terms of impact on their performance; (e) Most firms expressed willingness to pay for services provided on a heavily subsidized basis under the programs ­ indicating that levels of subsidy could probably be reduced without negatively impacting outcomes. Value Chain Analysis 15. Value chain analysis was conducted of two sub-sectors which had been targeted for support by government incentive programs (in particular the IRCC and MIDP) and/or had prima facie significant employment potential. The sub-sectors selected were the Cotton-to-Garments chain and Automotive Components (leather seat covers and metal tow-balls). 16. Cotton-to-Garments Chain: The cotton textiles industry, which has developed in South Africa since the end of the Second World War, accounts for more 200,000 jobs, direct and indirect. The industry grew under tariff protection, with production historically being driven by domestic demand, and with exports constituting only around 6% of domestic production during the 1970's. Asian investment during the 1980s helped to reinvigorate the sector and nearly two-thirds of investment in the textile industry is currently from foreign sources. The industry underwent a structural transformation after the trade liberalization of the 1990s, with a decline in spinning and fabric production, and a reorientation of garments production towards the export market (exports as a share of production rising from 4% to 21% between 1990 and 2001). South Africa's garments and apparel sector is very widely diversified, producing 38 categories for export to the US under AGOA. However, intense competition from China, post-MFA, in basic (high- volume) commodity type products (such as T-shirts)has eroded much of that market. v 17. All of the basic stages of the sub-sector value chain, from cotton cultivation to apparel production, are present in the country. Viable entry points for small and micro- enterprises are found to exist most strongly at the two ends of the chain ­ i.e. cotton cultivation and garment production, particularly at the high quality end of both production stages. To some extent this reflects the direction in which the industry is being forced to restructure in the face of intense competition from global suppliers, moving up-market into high-end niches where it has greatest comparative advantage. However, ability to take advantage of these entry points will depend on ability to address competitiveness constraints that could otherwise undermine sustainability of these production stages. 18. The value chain analysis brought out that, at the bottom end of the supply chain, the economics of cotton production are highly differentiated between irrigated and non- irrigated lands, yields on the latter being a fraction of those on the former, and qualities of lint cotton grown differing significantly. The quality of cotton on irrigated lands is among the best in the world, commanding premium prices and providing opportunities for expanding the industry. Non-irrigated lands are mostly occupied by smallholder, black farmers who do, in many cases, have the potential to improve yields and production margins by connecting up to irrigation distribution lines that in many cases already extend to nearby farms, but which are constrained in many areas on account of communal structures of land ownership. 19. The analysis also identified some specific inefficiencies impacting the competitiveness of each processing stage. In particular, the ginning, textiles and garments processing stages are affected by high overhead and administrative costs, which reflected relatively high management salaries and benefits that were a legacy of earlier times. At the same time, the structure of wage costs is also high relative to costs and productivity in competing countries, this too being a legacy of the unionized structure of the formal sector labor market. 20. The ability of the industry to survive in an increasingly competitive global market is dependent on its ability to significantly ramp up productivity and contain costs, whilst moving to more specialized segments of the market. Some of these changes are already in place, and also have a bearing on the greater integration of SMME producers. One consequence of this is the need to move to more capital-intensive production techniques, to be able to upgrade product quality. On the other hand, to contain costs, apparel production processes are also increasingly moving away from integrated functions to outsourcing to smaller CMT units. The more dynamic CMT units have been taking the opportunity to build up market access, upgrade machinery and improve product specialization. 21. There will, at the same time, be a need to improve productivity of workers, through vocational training, including increased use of apprenticeships programs. In this context, the DCCS export incentive scheme, providing subsidies to exporters against training obligations was meant to assist in this. In practice, the implementation of DCCS vi was somewhat distortionary and beneficiary employers often did not comply with their training obligations. 22. Auto Components (Leather Seat Covers and Metal Tow-Balls): The automotive sector in South Africa underwent a transformation by the early 1990s, moving from the shelter of tariff protection to greater integration into the global marketplace. Under the impetus of the MIDP the sector became significantly export-oriented, deriving greater economies of scale and production efficiencies, and employing some 263,000 workers directly and indirectly by 2004. The component industry experienced an expansion in parallel with the auto industry, contributing 60% of the exports of that industry ($1.56 billion by 2004). 23. The main impetus to the growth of the industry was provided by the MIDP which was meant to encourage competitiveness through rebate offsetting the impact of import duties via three mechanisms (an import rebate complementation certificate, a duty free allowance and a productive asset allowance). While the schemes produced the desired supply response, the mechanisms themselves generated some anomalies, whereby all players in the product or component supply chains did not benefit equally; in particular, players with greater market power (e.g. the OEMs and the tanneries in the leather seat covers chain) were able to appropriate a greater share of the benefits (at the expense of seat manufacturers). Secondly, the subsidies implicit in the scheme were not only high but failed to generate genuine competitiveness, as evidenced by relatively high domestic resource cost estimates of export or import-substituting output. 24. In leather seat covers, the market structure suggested that the component manufacturers had been unable to create their own market positioning outside of the OEM's supply chains. OEMs maintained tight control over their supply chains via creation of specific product platforms that made it difficult for small producers to easily switch clients. The tiering of the supply chain has been evolving from the late 1990s, as OEMs have exercised greater control over all of the processing stages (wet blue, tanneries and sewing/assembly), and is moving towards still greater integration within the Global supply chain. 25. Opportunities for SMME producers exist at the sewing/assembly stage, mostly in sub-contractual relationships with the larger seat manufacturers (most of whom are also supplying export markets via the OEMs). At the bottom end of the chain, the market is dominated by six major tanneries which, being large enough to supply the entire demand for tanned leather of sufficient quality for seat cover manufacture, are also able to exercise some degree of market power. 26. The analysis revealed some inefficiencies in production at the assembly stage, where the use of only partially-motorized sewing machines results in relatively high defect rates in comparison to competitors from China and India. This results in higher costs of production which are compounded by relatively high costs of labor and lower productivity, signifying the need for investment in training. When adjusted for labor efficiencies, the real labor cost for South Africa's component manufacturers works out to vii four times higher than in China. The incidence of transport costs is also high, as finished product has to be dispatched to European markets (mostly Germany) by air freight. MIDP rebates to a large extent offset South Africa's geographical disadvantage on this account, but the situation is not sustainable indefinitely. 27. In forged steel towing balls for automobiles, the value chain involves relatively simple processing and, hence, a prima facie greater scope for participation of smaller players. South Africa has a forging capacity of 39,000 metric tons per year, which is a fraction of India's (800,000 tons), but adequate to meet domestic consumption needs. Raw materials (hot rolled steel) constitute the main cost driver, and a combination of market structure (where the largest steel producer is able to exercise considerable market power) and implementation of MIDP rebates has resulted in relatively high costs of steel. A consequence of the fact that margins are squeezed is that manufacturers have little incentive to invest in more capital intensive production methods. The use of labor intensive methods results in relatively high reject rates, but there is a robust aftermarket in South Africa capable of absorbing lower quality output for this product. Conclusions 28. The findings of the micro-enterprise survey, the review of the various incentive programs and the value chain analyses indicate that: (a) among specific constraints faced by the SMME sector, the skills gap and the issue of access to finance are of particular relevance; and (b) while the economic rationale that existed in 1995 for SMME support remains valid, there is a need to find cost-effective and well-targeted programs that meet that rationale. 29. The issue of skills development, in particular, is central to the medium-term agenda as a means of raising productivity and, hence, employment in segments of industry - both in the formal and informal sectors. The Government has already taken steps to implement a skills development strategy; however, it has still to make major inroads into the problem. SMME firms still have relatively low participation in the SETA (Sector Education and Training Authority) levy-grant training programs, partly on account of inadequate knowledge of the procedures involved and partly because of difficulties of access. In finalizing the second phase of its national skills development strategy, key measures for the Government to consider should include a significant expansion of vocational training, including revamping of the apprenticeship/internship system; a re-launching of the training levy/rebate system within the SETAs to provide greater incentives for small firm take-up; and an expansion of the skills support and training programs offered by DTI, including grants for large enterprises to design and implement training and skills transfer to small enterprises. 30. As regards the DTI programs, there is a need to improve the effectiveness of promotion, strengthening selection criteria, and modulate the process of scaling up of individual programs. In particular, promotional programs need to be designed so as to ensure maximum coverage per rand expended; targeting of beneficiaries to ensure that viii grants disbursed actually achieve additionality; and the scaling up of programs to be at a pace that would not result in a lack of focus. 31. As regards other incentives, implementation of the DCCS incentives has not been highly effective in ensuring the compliance of beneficiaries with the training and skills development requirements of the scheme; and this will need to be tightened up in the future. Similarly, there is a need for some rethinking of the program of rebates under the MIDP to ensure that they also have a sustainable impact on firm competitiveness. 32. Main recommendations of the report for strengthening the pattern of Government assistance to smaller enterprises include measures for: (a) Consolidation and rationalization of programs: to reduce problems arising from duplication of efforts across different programs and compartmentalization of assistance efforts into too many different areas. (b) Spinning off of fund management: through use of independent fund managers, either through competitive tender or an accreditation process (as with GODISA and Ntsika). The implication for DTI itself would be for it to divest itself of most operational program management, in favor of a much greater oversight, monitoring and evaluation role. (c) Decentralization of assistance program management under various alternative models: e.g. to provincial offices, to local development agencies, to semi-private agencies, etc, as appropriate. Examples of such approaches include the Ntsika LBSC model, the NAMAC model, and the Business Partners SA model. (d) Strengthening of public-private partnerships: as in the case of Khula Finance's private partnerships for micro-institutions. (e) Strengthening of Linkages between large and small/micro enterprises, using the capacity of the first economy to support the second economy. ix x CHAPTER ONE THE ECONOMIC AND POLICY BACKGROUND A. ECONOMIC TRENDS SINCE INDEPENDENCE4 1.1 Since the early 1990s, South Africa has achieved much in terms of macroeconomic stability, improvement in the business climate, and development of industrial support measures, especially within the `first economy'5 which is operating in an increasingly open trade environment. South Africa's relative macro-economic stability over the past decade has resulted from solid macro-economic management and fiscal control, resulting in long run exchange rate strength and a vote of confidence from the major international rating agencies. Exchange rate volatility over 2001 to 2003 has dissipated and the real exchange rate has stabilized since 2004. Growth has been steady, if moderate, over the past decade, averaging 2.9%, picking up more recently to 4.5% in 2004, 4.9% in 2005, and a projected 4.7% in 2006.6 Annual per capita GDP growth averaged just less than 1 percent. Employment has also picked up since 2003 after many years of decline. 1.2 Although South Africa's economy has grown, its growth until 2003 was relatively modest compared to other middle income countries (see Figure ). Over the past decade, per capita growth has been over three times faster in Thailand and Malaysia than in South Africa--despite the Asian crisis in the mid-1990s, and has been close to ten times faster in China. Given the challenges that South Africa will face in the coming years, and the high level of unemployment in the country, sustaining a more rapid rate of growth will be essential. 1.3 Investment has remained at around a low 16% of GDP, with an up-tick since 2003 to 17%. This is considerably lower than the government's target of 25%, and lower than in middle- income countries, especially the fast growing Asian economies (see Figure 1.1). In part this reflects low public sector investment but private investment has been at about 12 percent of GDP. 1.4 Despite an upturn since 2003, the outlook for employment remains unpromising. The measured unemployment rate more than doubled from the already high estimate of 4The following section draws partly on the Investment Climate Study (ICA) for South Africa, World Bank, 2004. 5First and second economy. The second economy in South Africa is usually defined as the informal and micro/small businesses which are largely owned by historically disadvantaged social groups; the first economy is defined as the other small and medium firms, and the large firms. 6EIU Country Report, April 2006. 1 23% in 1991, to an alarming 47.8% of the potential labor force in 2002.7 Unemployment among the white population rose (by 15%), while among mixed race and black South Africans, constituting the majority of the population, it more than doubled. This is consistent with the fact that while South Africa has one of the highest per capita incomes ($49608) in Sub-Saharan Africa, it also has one of the most unequal income distributions in the world. Even with the slight turnaround in the last two years, absolute numbers of employed in industry have declined since 1990, especially in industries such as textiles. Figure 1.1: Growth and Investment have been lower in South Africa than in the best performing middle-income economies. China China Korea, Rep. Korea, Rep. Malaysia Malaysia Thailand Thailand Poland Poland Brazil Brazil South Africa South Africa 0 2 4 6 8 10 0 10 20 30 40 Gross Fixed Capital Formation (as percent of GDP, Per capita GDP growth, 1994-2003 average 1994-2003) Source: World Bank (2005b). 1.5 Exports have diversified out of primary products into manufacturing and have shown reasonable growth but there is concern about South Africa's international competitiveness and the technological level of its manufactured exports, especially compared to other middle-income economies of Asia. 1.6 Despite South Africa's moderate growth performance, the Investment Climate Assessment found that labor productivity was fairly high. In 2002, value added per worker was far higher than elsewhere in Sub-Saharan Africa9 and compared well even with other middle-income countries such as Lithuania, Brazil, Poland and Malaysia--all of which, except Brazil, have higher per capita income than South Africa. 7According to the Bureau of Market Research (BMR). Part of the increase may result from the full accounting for the `homelands' labor force. 8 World Bank: "World Development Report, 2007". 9 Source: ICA. 2 Figure 1.2: Labor productivity is higher in South Africa than elsewhere in Africa and higher than, or comparable to, other middle-income economies and the most productive areas of China. $20,000 s $15,000 Dollar $10,000 US $5,000 $0 htu aci ayn en So Afr Ke enzhS na)hiC( liza Br iasyla Ma Value-Added per Worker Value-Added per worker in Garments Source: Investment Climate Surveys Note: All values are medians for enterprises with available data. Value added is calculated by subtracting intermediate inputs and energy costs from sales from manufacturing. Workers include both permanent and temporary workers. Values are converted to US$ using average exchange rates for 2002 from World Development Indicators. Data is for 2002, except for Lithuania (2003). 1.7 Part of the reason for South Africa's apparent high productivity appears to be that South African enterprises are concentrated in relatively capital intensive industries. Over the 1980s and 1990s, manufacturing became increasingly capital intensive due to both capital intensive investment and an expansion of relatively capital intensive sectors. Firms have about twice as much capital per worker (about $3,500 per worker) as firms in Lithuania, Brazil, and in the most productive areas of China.10 1.8 The ICA findings on capital intensity and high labor productivity, particularly within the textile sector, differ in some ways from those of the value chain analysis in chapter five, which indicates relatively low technology and productivity. The findings are consistent with the high concentration and sharp segmentation of the industry, with a small group of larger-scale and foreign investment enclave firms and numerous small scale domestic origin producers, as revealed in chapter two by the micro enterprise survey. 1.9 Even in the more highly capital-intensive firms, however, because of high formal wage costs in South Africa compared to these comparator countries, unit labor costs (labor productivity in relation to wage-bill) are also higher in South Africa than in most of the comparator countries, except for those in Eastern Europe. High unit labor cost is an indicator of low international competitiveness. 10Although the weak Rand in 2002 might have led to underestimation of the amount of capital per worker in South Africa. 3 B. THE EVOLVING ROLE OF SMMES 1.10 The importance of a strong small, micro and medium enterprise (SMME) sector lies largely in its effects on mobilizing entrepreneurship and introducing flexibility and competition in an economy.11 Within a particular economy there is an efficient size profile of enterprises based on market size, consumption patterns, resources, technology, institutions, and level of competition. Economies with an undeveloped SME sector in relation to these parameters may thus be at a disadvantage. This is especially so in economies that have faced significant structural constraints on SME development, such as former Soviet republics where large-scale State enterprises were dominant, and also in economies such as South Africa where State policies created artificial constraints through restriction on Black and Asian enterprise ownership, which would otherwise have been the main source of small firm growth. 1.11 Apartheid both limited job opportunities in the formal economy for black South Africans and restricted the right of non-white entrepreneurs to establish and own businesses.12 The impact of Apartheid business legislation resulted in low skills, lack of an entrepreneurial tradition and weak SME network institutions such as business associations. Today, in the `townships', the effects of the former restrictions on black- owned businesses are evident, with poor access to infrastructure and undeveloped commercial retail activity, despite their proximity to the major urban centers. Local business development programs are beginning in the more prosperous cities, such as the "Red Door Offices" in the Cape Province. The Soweto Executive Business Council outside Johannesburg is also taking the initiative. 1.12 It is in dual economies that have experienced such structural distortions that the `missing middle' (i.e. the lack of small formal-sector firms) is likely to have a particularly adverse effect on economic growth. The legacy of entrenched economic dualism in South Africa, with an increasingly capital-intensive `first economy,' however, also gives a much greater potential role for small firms in job-creation compared to other economies (see note ). In fact, the growth in employment in SMMEs in South Africa has increased 12 considerably more rapidly than their contribution to GDP. In 1995, this contribution was 1.35 times greater than the contribution of the SMME sector to GDP; in 2001 this factor was 1.5. This high labor absorption capacity, and job-creation potential has to be netted against firm death rates but seems to show a large employment growth contribution. 11The other often-claimed benefit of SME in developing countries, high job creation, has been challenged. This is because while small firms create significant employment they also have high failure rates and the net job creation (and quality and stability of those jobs) may not be as high as that of the medium-large scale enterprise. See Kris Hallberg "A market-oriented strategy for Small and Medium Scale Enterprise, IFC, 1999, and Tyler Biggs "Is Small Beautiful and Worthy of Subsidies?" Draft. World Bank 2003. 12Lund, Frances, and Caroline Skinner. 2004. "The Investment Climate for the Informal Economy: A Case of Durban, South Africa." Background paper for the WDR 2005. 4 1.13 In 1995 the small and medium enterprise sector in South Africa was considered to be relatively small, and in 1997 a report by Ntsika13 found that self-employment rates among the black population in South Africa were very low by international standards, reflecting the structural constraint on firm ownership by historically disadvantaged groups, and implying considerable potential for a catching-up process. The situation seems to have improved over time from this low starting position. For instance, the share of manufacturing SMMEs in employment is estimated to have risen from 39% in 1988 to 44% in 1999. At the end of the 1990s, out of 106,000 production units, about 95% were at SMME scale14 close to international levels while, overall, SMMEs contributed about 35% of GDP and 54% to formal private sector employment.15 1.14 A more recent estimate by the Department of Trade and Industry (DTI)16 is that small businesses represent 98 % of the total number of firms. They employ 55% of the country's labor force and contribute approximately 42% to the total wage-bill. Small firms, overall, account for 35% of GDP. 87% of `survivalist' micro-entrepreneurs are black, mostly unregistered and not eligible for Government programs. Wholesale/retail, construction and transport are the three largest and fastest growing small business sectors. 41% of SMMEs are owned by women.17 1.15 In correspondence with the rising small firm population share, there has been a steep rise in the number of new registrations of private companies, from 6,369 to 32,178 per year between 1990 and 2002, and for close corporations (CCs) from 28,008 to 107,302, especially in urban areas such as Gauteng, Western Cape and Durban. The majority of new CC registrations have taken place in the trade and financial & business services sectors, followed by community/social services and the construction industry. Rural small- and micro-enterprise numbers are not growing as fast due, possibly, to the continued predominance of large agro-producers. The annual change in number of registrations is, in some sense, a barometer of a country's economic and entrepreneurial activity. 1.16 The informal sector in most developing countries is, increasingly, the dominant form of work18 and one that operates beyond the scope of registration, tax and social security obligations. The informal economy in South Africa is less prominent than elsewhere, though recent analysis suggests that it could account for 25% to 30% of South Africans who are presently working. Alternative estimates put the size of the informal sector at between 1.0 million and 2.2 million enterprises, a wide range which highlights the difficulties that DTI and other institutions have in designing assistance. Part of the problem lies in the fact that definitions of formal and informal sector activity have not 13 "The state of Small Business in South Africa" Ntsika Annual Review 1997. 14`The state of Small Business in South Africa' Ntsika Annual Review 1999. 15`Integrated Manufacturing Strategy' DTI, 2002 p18-19. 16 Department of Trade and Industry 2003: "National Strategy for the Development and Promotion of Small Business in South Africa", Pretoria, South Africa. 17Ntsika Enterprise Promotion Agency. "State of Small Business Development in South Africa ­ Annual Review 2002". Pretoria, South Africa, 2002. 18Statistics cited in the ICA (Charmes) indicate that informal work composes 25%, 61% and 40-60% of urban employment in Latin America, Africa and Asia respectively. Further between 80% and 90% of new jobs created in Latin America and Africa were in the informal economy. (Lund, Skinner. 2003). 5 always been precise.19 Ntsika defines the sector as including all survivalist and single- person micro-enterprises, plus micro-enterprises employing up to 4 people, very small enterprises and small enterprises. The following table, from the Standard Bank, shows the upper estimate of formal and informal firms in 1998 and 1999: Table 1.1: Distribution of formal and informal businesses Estimated number of SMMEs in South Africa in 1998 and 1999 1998 1999 % change Formal private 349,720 357,780 2.3 Public service 83,430 88,100 5.5 Total Formal Firms 433,150 445,880 2.9 Informal micro enterprises 1,329,570 1,823,320 3.7 GRAND TOTAL 1,762,720 2,269,200 3.3 Source: Standard Bank 2000, in ILO employment section, working paper No. 34 1.17 The following table, drawn from the September 2001 Labor Force Survey, demonstrates a slightly higher total than for 1999, with 1.87 million informal workers. Informal employment is concentrated in the retail and wholesale trade, with just over half of all informal workers. Further, there are significant numbers working in construction, manufacturing and services. Just over 10% are in manufacturing, a low proportion in comparison to other developing countries. Table 1.2: Number and % of informal workers by sector, 2001 Economic Sector Number % Wholesale/retail 936,977 50.10 Construction 258,442 13.80 Manufacturing 199,108 10.60 Community, social, personal service 163,340 8.70 Private Households 127,277 6.80 Transport, storage, communication 103,541 5.50 Financial 76,851 4.10 Mining and quarrying 2,695 0.10 Electricity, gas, water 1,796 0.10 TOTAL 1,870,027 100 Source: StatsSA (2001) in Lund, Skinner (2003): The Investment Climate for the Informal Economy: A Case of Durban, South Africa, Background paper for the 2005 WDR; 2003. 1.18 While there has been an acceleration of micro-business startups and activity, the constraints on growth remain considerable, as discussed in the subsequent chapter. The survival rate of small enterprises has remained low by international standards and a `vibrant critical mass of small enterprises' has yet to be integrated into a number of sub- sectors.20 At the same time there is some evidence that the `dual' nature of the economy may have intensified with the concentration of formal enterprises in the first economy 19The informal sector has been defined as that which encompasses any value adding economic activity not recorded in official economic statistical data bases See: "State of Small Business development in South Africa", Annual Review, 2002; Ch.2. 20Ntsika Enterprise promotion Agency; Annual Review, 2002, cited earlier. 6 increasing. Recent studies21 have shown that concentration and associated levels of protection have continued to apply across a wide range of industries. C. THE EVOLUTION OF INDUSTRIAL SUPPORT 1.19 Following independence, a major rethink of South African industrial policy took place, pushed both by the new possibilities and opportunities facing the economy and by external factors such as the requirements of WTO accession. The aim was to shift from a policy characterized by import-substitution and limited export promotion measures, largely relying on taxes, tariffs and quotas, to supply-side interventions such as Government-industry partnerships in strategic industries, innovation support, skills development, and SMME advisory programs directed at increasing competitiveness. 1.20 In 1995, when the `National Strategy for the Development and Promotion of Small Business' was adopted, South Africa had been facing stagnant output and declining employment for a number of years. The rate of unemployment among the black population was as much as 40%, with a rising rate among whites. 400,000 people were entering the job market each year while the formal economy was, on average, only absorbing about 10 % of this number. 1.21 The Government adopted a program of policies that aimed at broad macro- economic stabilization and market liberalization, legal and regulatory change and improvement in the business climate. As part of the overall program it also launched a series of direct initiatives to promote micro, small and medium business.22 The legacy of the previous system, that small business was contributing a relatively low proportion of GDP by international standards, seemed to offer scope for expansion and job creation through a drive to increase entrepreneurship. 1.22 During the 1990s the Government phased out some agencies and created a National Small Business Support Network including a number of new SMME support agencies. The 1995 White Paper on Small Business and the National Small Business Development Act of 1996 paved the way for creating new institutions. The two principal initiatives were the Ntsika Enterprise Promotion Agency, to provide advice and assistance to new and small businesses, and Khula Enterprise Finance, to promote the provision of finance, especially microfinance, and guarantees to small businesses. Each umbrella organization in turn instituted a series of programs, including the Local Business Services Centers (LBSC), supported by Manufacturing Advice Centers (MACs), and Tender Advice Centers under Ntsika, and the Retail Financial Intermediaries and the Guarantees programs under Khula. 21Johannes Fedderke and Dabor Szalontai, School of Economics, University of Cape Town. "Industry Concentration in South African Manufacturing: Trends and Consequences, 1972-96" and Johannes Federke "South Africa: Sources and Constraints of Long-Term Growth 1970-2000". 22Under the National Small Business Act five categories of smaller enterprise are defined: `survivalist' (self- employed), micro (up to 5 workers) very small (up to 10 to 20 workers), small (up to 50 workers) and medium scale (up to 100 to 200 workers). 7 1.23 A number of `legacy' programs were retained. These include the Export Marketing and Investment Assistance program (EMIA) and the Support Program for Industrial Innovation (SPII), managed by the State-owned Industrial Development Corporation (IDC), which was, itself, established as far back as 1940. 1.24 The National Empowerment Fund was set up in 1998 to focus on Black-owned business, and the World Bank/DTI supported Industrial Competitiveness and Job Creation project (ICJC project) was launched in the same year. The SME promotion activities of the DTI were from 1999 channeled through the `The Enterprise Organization' (TEO). Of these, notably, the SMEDP was started in 2001 to provide investment grants, focusing ostensibly on smaller, labor intensive firms. A range of other initiatives were started under DTI auspices, such as the Work Place Challenge Fund, designed to promote better workplace relationships. Other initiatives included the GODISA small business incubation centers (partly under the Ministry of Labor), and the Industrial Development Zones (IDZ) program. The DTI's largest program is currently the Motor Industry Development Program (MIDP). 1.25 Increasingly the Government has tried to rationalize the support network by centralizing direction through the DTI. In summary, the DTI oversees a network of 21 affiliated agencies (grouped also under the Council of Trade and Industry Institutions or CTII) plus its own directly managed programs. These may be categorized as follows: · Regulatory agencies - overseeing agencies concerned with regulating competition, microfinance (MFRC), registration of companies and intellectual property (CIPRO), international trade administration (ITAC), and gambling; · Enterprise financing institutions including the Industrial Development Corporation (IDC), Khula Ltd. and the National Empowerment Fund (NEF); · Support service agencies, including Ntsika Enterprise Promotion Agency and National Office of the Manufacturing Advisory Centers NAMAC (now merged into SEDA (Small Enterprise Development Agency); the CSIR (Council for Scientific and Industrial Research); National Research Foundation, National Bureau of Standards, and SA Quality Institute; and · Directly managed SMME support programs: day-to day management of these programs is carried out within DTI by TEO (The Enterprise Organization), EIDD (Employment and Investment Department), and TISA (Trade and Investment South Africa). These have included Bank-supported funds, the SMEDP, EMIA and several others. 1.26 Apart from these functions the DTI is responsible for (a) policy-making and legislative activities; (b) procurement programs; (c) promoting partnerships with Government and Industry organizations; and (d) information and publicity 8 1.27 Since 2004, DTI has continued to expand its menu of incentives. It has launched the Business Process Outsourcing program (which includes development of call centers); a film industry incentive program, and the Micro Apex Fund. The latter fund will be set up as an autonomous organization, with DTI as major shareholder, to provide group loans to the smallest micro-enterprises, taking over some business from Khula. 1.28 In an effort to reinvigorate the industrial promotion process the Government introduced the Integrated Manufacturing Strategy23 in 2002. The central aim of this strategy was to retain export competitiveness by increasing domestic value-added within an efficient export sector encompassing high quality product design, production, distribution, sales and marketing. The Strategy prioritized agro-processing, textiles and clothing, minerals and metals, chemicals, automobiles, and ICT. 1.29 Based on the review and the experiences of eight years of SMME policy, in 2003 DTI issued a new SMME Strategy: "Unlocking potential in an enterprising nation: The Integrated Small Business Development Strategy in South Africa 2004-2014". The Integrated Small Business Development Strategy (ISBDS) sets out three strategic directions: (1) promoting entrepreneurship, (2) unlocking potential through better business environment, and (3) promoting more competitive small businesses. It defines as its targets micro-enterprises, small business in high-growth sectors, and black owned and managed small and medium enterprises. 1.30 The 2003 DTI review and widespread consultations identified a number of weaknesses and suggested that strategies needed to be redesigned, revised and aligned with new developments. The most significant were that: (a) the generic approach towards small business development proved ineffective, suggesting a need for differentiated services, instruments and delivery concepts adapted to the different segments of SMEs; (b) support institutions and programs that had proliferated needed consolidation; (c) wide-spread regional and local difference in policy absorption called for a more bottom-up approach; (d) the promotion of entrepreneurship in the context of basic formal and technical education needed special attention; and (e) business associations needed to play a more significant role in the support process. 1.31 As a result of this re-think, in December 2004 as part of a rationalization process, the Ntsika and NAMAC organizations, along with the CPPP,24 were merged under the Small Enterprise Development Agency (SEDA) and discussions are under way on further rationalizations. An Enterprise Development Bill to strengthen the regulatory regime underpinning enterprise incentives is currently being prepared and is expected to be presented to the Cabinet in 2006. 1.32 The key issue underlying the strategy is the objective of broad-based Black Economic Empowerment (BEE). As outlined in the Microeconomic Reform (MER) agenda and the Integrated Action Plan (IAP) launched by President Mbeki in 2001, BEE is meant to influence every sector of the economy, impacting on all types and sizes of 23Accelerating Growth and Development: the contribution of an integrated manufacturing strategy". (DTI 2002). 24Community Public Private Partnership Program. 9 enterprises. It is intended to bring about significant increases in the numbers of black people who manage, own and control the country's economy, as well as significant decreases in income inequalities. A core component of the BEE strategy is the creation and nurturing of new enterprises undertaking new forms of economic and value-adding services, attracting investment and employing more people in productive activities. Towards this end, the BEE agenda incorporates a number of instruments, including preferential procurement, finance for skills development, and use of a `balanced scorecard' for measuring progress in achieving BEE objectives at the enterprise level. Where appropriate, especially at the sectoral level, the Government actively seeks partnership with the private sector, where this will help accelerate the BEE process. 1.33 Though various government programs are in place to support firms in overcoming these barriers, there have as yet been few sustained successes of MSME growth. BEE was intended to help transform market structures, ownership, employment structures and the integration of small firms into the supply chain, however, national support mechanisms to support the BEE effort are still weak and not well coordinated. It remains to be seen whether the centralization of support mechanisms has been the best approach, or whether a greater emphasis on the role of provincial organizations would have been more effective. Critics have also raised the question of whether the strategy pursued has not been too supply-side oriented, and whether a greater emphasis on a demand-driven approach would not have yielded better results.25 D. THE INDUSTRIAL POLICY RATIONALE 1.34 In summary, South Africa's industrial policy in the 1990s developed a menu of interventions with the following objectives: · reducing barriers on access to finance for micro and small firms; · reducing information barriers to entry for small businesses; · reducing barriers to entry for youth, women and black-owned businesses to offset systemic disadvantages and improve equity; · supporting youth, women and black-owned businesses operation to improve equity; · overcoming systemic constraints on networking/contracting between firms to improve information sharing and productivity; · reducing the cost of/increasing private returns to R & D and innovation; · reducing the monopolistic price of technology acquisition; · easing the lack of information and experience in national and export marketing; and · inducing firms to invest more than they would otherwise do on skills training. 25Christian Rogerson: SMME Infrastructure and Policy in South Africa. 10 1.35 The basic rationale for this range of public interventions was the presence of non- competitive or `failing' markets, in the following areas: 1.36 The labor market, which was segmented and partly non-competitive as a legacy of apartheid and migration control. There was a narrow urban formal labor market (a `labor aristocracy') and an extensive rural informal market characterized by high unemployment. Median wages in the formal sector in South Africa were high and continue to be so. An unskilled formal sector worker earns $246 per month compared to a mere $76 in China. In 1995 segmentation was still acute due to the legacy of the previous market structure, mismatched skills and a union-controlled formal labor market, and by a tendency of formal sector firms to invest in capital-intensive technology to avoid labor issues. The public objective was to increase the employment of black and other previously disadvantaged workers, by lowering the cost of labor to offset these market impediments, through empowerment schemes, and by directly encouraging labor- intensive production through micro and small enterprise. 1.37 A particular feature of the labor market is the burden of regulations. According to the World Bank's Doing Business database the difficulty of both hiring and firing in South Africa is higher than in all of the comparator countries except Brazil and is also higher than the average for the OECD. Overall, South Africa performs worse on most measures than the comparator countries. This is reinforced by the results of a survey on the cost of red tape in South Africa, which showed that entrepreneurs considered labor laws and regulations as one of the most important constraints to expansion and increase in employment.26 Four main pieces of legislation (see Annex 1), governing the labor market contribute to this: (a) The Labor Relations Act, which covers employee rights, collective bargaining and union rights, strikes and lockouts, workplace forums, dispute resolution, dismissals, etc., and which generally places the burden of proof with the employer; (b) The Basic Conditions of Employment Act of 1997, which covers such areas as: working time, leave, remuneration, termination, child and forced labor, sectoral determinations, as well as monitoring and enforcement; (c) The Employment Equity Act of 1998, which covers the basic ideas of unfair discrimination and affirmative action policies of South Africa, as well as the institutions that govern these policies; and (d) The Skills Development Act of 1998, which set up the training levy-grant scheme designed to reinvigorate training (but where the difficulty in making claims has led many small businesses to view this program as more of a tax27 than a benefit). A further piece of relevant legislation is the Black Economic Empowerment Act of 2003, which goes beyond the affirmative action of the Employment Equity Act, and sets goals of changing the racial composition of ownership and management structures of existing and new enterprises. 1.38 Labor Skills are in short supply because of the generic training externalities facing all economies, but particularly because of past segmentation of education and labor markets and economic isolation. Despite an improvement after 1997, levels of education 26See: "Counting the Cost of Red Tape ­ for business in South Africa", SBP, June 2005. 27Hudson, Judi. "RIAS and private sector development: Some thoughts from the South African context," Small Business Project, (Johannesburg, South Africa, 2003): 5. 11 among the workforce remained very low, with the vast majority of workers possessing less than a matriculation certificate.28 There was also a major decline in apprenticeship programs, as well as formal sector enterprise training through the 1990s which particularly impacted black employment and black-owned business startups. Though the public objective is to reduce the price of obtaining skills and to boost their supply through training, according to the results of the ICA, training capacity in South Africa is still particularly weak by international standards. By way of comparison, more than 80% of all workers receive firm-based training in Poland, 68% of unskilled workers in Brazil and 63% in China, whereas less than half of skilled and unskilled workers in South Africa have access to firm-based training (Table 1.3). With the exception of unskilled workers in India, this is the lowest level of training provided among the comparator countries. Table 1.3: Percentage of Workers Receiving Firm-Based Training, by Country Country % of Skilled workers % of Unskilled workers Brazil 77.3 68.3 China 69.1 63.0 India 55.0 33.0 Poland 79.9 86.2 South Africa 44.6 45.8 1.39 In the capital market, uncertainties about economic stability and country risk in the early 1990s raised real interest rates. Investment rates were lower than desirable and supply of/access to capital was uneven. Smaller and non-white-owned firms in particular faced barriers to entry and growth in this respect. It was, therefore, considered to be in the public interest to try to reduce the price of capital by lowering risk, directly increase capital availability to disadvantaged firms, and encourage investment demand. 1.40 In the decade before 1994, investment declined by 3% per year. In contrast, during the last decade, investment has risen by some 5% per year, but overall has remained in the narrow band between 15% and 16% of GDP, with a recent boost to 17%. Although this is comparable to capital formation in Kenya and Senegal, the comparator countries in Sub-Saharan Africa, it is lower than in the middle-income comparator countries (Brazil and Poland) and is far lower than in the fast growing Asian economies. Gross Fixed Capital Formation increased in 2003-2004, but remains lower than in the fast growing Asian economies. 28See: State of Skills in South Africa, 2005; Dept. of Labour, RSA, Ch.4. 12 Figure 1.3: Investment has been modest, especially compared to the fast growing Asian economies. 40 30 20 10 0 South Africa nya Ke negal il Braz land Se Po Thailand Malaysia ea, Rep. China Kor Gross Fixed Capital Formation (as percent of GDP, average 1994-2003) Source: World Bank (2005b). 1.41 Within the 15% to 16% range, compared with the government's "unofficial" target of 25%, investment since the early 1990s, and previously, has been below the levels required to lift South Africa into a high growth bracket. There thus remains an investment gap despite the more recent deepening of the capital market and apparently good access to finance for larger, formal sector, firms. The principal hope is that higher projected levels of government spending will induce private investment through growing confidence in government economic policies and a steady decline in perceptions of political risk and lower local costs for imported plant and equipment. 1.42 The supply of entrepreneurship was constrained by historic factors including lack of opportunities for the black population, lack of linkages between established (white- owned) enterprises and non-white owned businesses, and barriers to entry for the latter. As mentioned earlier, Ntsika reported that self-employment rates among the black population in South Africa were very low by international standards, reflecting the structural constraint on firm ownership by historically disadvantaged groups, and implying considerable potential for a catching-up process. The situation did improve from its low starting position so that, by the end of the 1990s, 95% were at SMME scale,29 close to international levels. Nevertheless, in 1995, it was considered to be a justifiable public objective to reduce entry costs and promote entrepreneurship, especially among the previously disadvantaged, black community. 1.43 Product market- Exports. Past economic isolation had tended to constrain external trade experience and reduce international competitiveness, and the dominance of metals and minerals exports meant that manufactured goods export know-how and information 29`The state of Small Business in South Africa' Ntsika Annual Review 1999. 13 was relatively scarce. During the 1990s considerable trade liberalization occurred with major reductions in export subsidies, import tariffs and levels of protection, in accordance with, and also beyond, the WTO accession requirements. The share of manufactures increased over the years so that, by 2000, manufactures were the major export product group and the export-intensity of manufacturing output was found to have doubled between 1993 and 2000. One of the main sources of this growth was the automotive industry under the MIDP, along with sectors such as iron and steel and nonferrous metals, and certain equipment categories such as TV, communications, professional and scientific equipment. 1.44 Despite its apparent diversification, South Africa was not doing well enough in the 1990s compared to other middle income country exporters, losing ground overall in manufacturing exports, and not moving sufficiently into high growth high-technology exports, such as telecoms and electronics, to the extent of comparable economies.30 In those circumstances the public objective was to promote exports, not only by tariff reform and removal of anti-export biases, but also through direct intervention by lowering barriers to entry such as exporting know-how. E. THE PURPOSE OF THIS STUDY 1.45 The purpose of this study flows from the previous discussion. The creation of jobs is an important economic priority, towards the achievement of which the SMME sector has a key role to play. As described above, the Government has made a concerted effort to encourage the growth and sustainability of this sector, through establishment of an incentive framework that is probably the most comprehensive in Sub- Saharan Africa. However, it is not clear that performance of the sector, as reflected in growth of new entrants and survival of existing incumbents, has been commensurate with the efforts and investments involved. Questions on the efficacy and impact of individual incentive programs continue to arise, giving rise to the questions of whether: (a) Incentive programs have been correctly targeted to address the real and diverse needs of SMMEs; (b) The programs have been implemented efficiently enough by the responsible governmental agencies, in terms of their delivery and impact; and (c) They are effective in helping SMMEs access a wider market for their output and services, from larger enterprises downstream in the relevant product supply chains. 30Source: `South Africa Export Performance: Determinants of Export Supply', L. Edwards and P. Alves Univ Cape Town March 2005. 14 Approach 1.46 To assess this, the study adopts the following approach, of conducting: · A survey of SMME firms, using the same core questionnaire of the Investment Climate Assessment ­ so as to provide consistency for benchmarking ­ but with additional questions included, specifically on the incentives framework. · Analysis, through directed interviews and impact assessment reviews, of how government agencies have been implementing the incentive schemes. This analysis to be supplemented, on the demand side, through a small survey of incentive recipients. Such an analysis would need to look more carefully at the economic rationale of the menu of programs, their actual performance and economic impact, the efficiency of their implementation and their effectiveness in reaching out to prospective applicants. · Assessment, through analysis of the value chains for some representative sub- sectors, of the potential for and effectiveness of integration of SMMEs into larger product supply chains. Such an exercise would be expected to (a) throw light on the tension between the competitiveness agenda, dictated by the need for downstream manufacturing industries to survive in an increasingly globalized world, and the government's social agenda, being promoted through the incentive schemes, and (b) help identify the entry points for better SMME integration, and the measures necessary to take advantage of them. · While a number of sub-sectors could be looked at for this purpose, the choice ­ in this study ­ of sub-sectors for the value chain analysis is influenced by considerations of employment potential and quantum of targeted incentives already being provided for this purpose. On this basis, the cotton textiles and apparels sector and the auto components sector (with greater focus on Tier 2/Tier 3 suppliers, where smaller suppliers would be expected to dominate), most readily suggest themselves. Opportunity 1.47 Among countries in Sub-Saharan Africa, South Africa's experience in enterprise promotional programs is probably the most rich and diverse, notwithstanding that most have been introduced relatively recently, since 1995. Ten years down the road, the time is opportune to revisit the programs, and assess their performance and their alignment to South Africa's priority development needs, and the Government has already begun the process of rationalization, to try and improve their impact and cost-effectiveness. This study can be considered a part of that process and will, hopefully, provide a platform for further, more detailed, analysis in due course. 15 Structure of the Report 1.48 The report groups together the three main strands of analysis to arrive at recommendations for the future of enterprise promotion in South Africa. After setting the framework in chapter one, the next three chapters discuss respectively: (a) the findings of the survey conducted to identify the nature of the constraints facing MSME firms, (b) the efficacy of the institutional structure for enterprise support, from the perspective of both the providers and the users, and (c) the results of the value-chain analysis of three relevant sub-sectors, aimed at identifying the entry points for small and micro enterprises, as well as the competitiveness constraints imposed by inefficiencies at different stages of production. The results are pulled together in chapter five to try and draw conclusions, from a policy perspective, on the rationale for enterprise support, its configuration going forward and the organization of support infrastructure that will be necessary. 16 CHAPTER TWO WHAT ARE THE CONSTRAINTS ON, AND THE PUBLIC SUPPORT NEEDS OF, THE SMME SECTOR? 2.1 To understand whether incentives have been correctly targeted it is useful to understand what enterprises see as their principal needs and constraints to doing business. This allows us to move from general measures into the more specific areas of concern faced by enterprises, and the ways they can be addressed. 2.2 The constraints on enterprises are framed by the investment climate. The investment climate consists of the set of economic incentives shaped by macroeconomic and regulatory policies, public administrative procedures, institutional arrangements, (e.g. property rights, rule of law and governance), and the availability and cost of physical infrastructure. Institutional arrangements influence private investment decisions through their effect on uncertainty and risk, and investment safeguards (protection of property rights, enforcement of contracts, and maintenance of integrity of monetary stands). A. THE BUSINESS CLIMATE FOR FORMAL SECTOR ENTERPRISES Background: The Investment Climate Assessment 2.3 A starting point is the Investment Climate Assessment (ICA) for South Africa that was conducted in 2004 through a survey of about 800 formal sector firms (Annex 2 for details of the sample), distributed between manufacturing, construction and trade. Sample firms were located in four locations, and covered a mix of small, medium and large (plus some very large) enterprises. The sample contained few micro-enterprises (less than 10 employees), especially in the manufacturing sector, and those covered were predominantly owned by European individuals and families, with only 5% owned by Africans or colored individuals/families. 2.4 Looking to the findings of the survey, relatively few firms rated any constraints on their businesses as severe obstacles. Only about 35% of firms rated the biggest constraint, worker skills, as a serious problem and most obstacles were rated as a serious concern by less than one in five enterprises. This is far lower than in most countries, where for example, close to 85 % of firms in Brazil said that tax rates were a major problem and 74% of firms in Kenya considered corruption to be a major problem. 2.5 Bearing in mind this caveat, worker skills, macroeconomic instability, labor regulations and crime, were cited as the main problems, with 29 to 35% of enterprises rating each of these areas as major problems. There were few noticeable differences between firms by ownership, export status, foreign participation, province or sector. The main differences were that exporters were more concerned about macroeconomic 17 stability than other firms and African-owned firms were slightly more concerned about access to and cost of financing than other firms. Foreign corporations operating in South Africa tended to see the skills and economic stability issue as more serious than other firm categories. Crime was relatively important for Asia-owned firms, probably because they located more in urban areas. Business licensing, communication facilities and access to land were perceived as generally not important factors. Figure 2.1: Worker skills, macroeconomic instability, labor regulation and crime are firms' biggest concerns in South Africa. 50% 40% 30% 20% 10% 0% croeconomi Worker Skilc InstabiRegulations ls lity Crimex Ra tes ty ion g ption s ion y ystemto LandnicationLicen s ricit sing rru Ta Labor Policy UnTrade certainRegulatf FinancinCo E st o to Tra Co Ma Anti-competitiAc ve cess practiceFinancingistratnsportationlectLegal SAccess x Admin mmusiness Ta leco Te Bu Percent of firms saying area is a major obstacle to growth Source: Investment Climate Survey Worker skills 2.6 Enterprise managers were more concerned about lack of worker skills than any other issue. Firms found it difficult to attract skilled workers, and were paying a sizeable premium for skilled and educated workers. The differential between management/skilled earnings and unskilled wages (about 7.5:1) is considerably higher than in comparator countries. Notwithstanding this, relatively few enterprises had training programs, far less than in say China, Poland and Brazil.31 Wage differentials from training were also comparatively low, suggesting either undervaluation of or ineffective training, or that managers selected only the most productive workers to receive training. Overall these results suggested that government programs designed to encourage training (e.g., SETA) have not been successful. 31Interviews with employees suggested that more than 80 % of workers in South Africa had not received any formal training. 18 Labor Regulations 2.7 The latest Doing Business database, which measures the rigidity of hiring and firing regulations and the cost of firing,32 shows that labor regulation in South Africa is more rigid than in most of the comparator countries. It is more difficult to hire and fire workers in South Africa than in most of the comparator countries or than in most OECD economies, (although on the cost of firing an individual in weeks of wages, South Africa performs better than the OECD average and for several comparator countries). Figure 2.2: Labor Regulations are stricter in South Africa than in the best performing comparator countries or in the OECD. OECD OECD OECD Brazil Brazil Brazil Lithuania China China Poland Poland Malaysia China Lithuania Lithuania Malaysia Malaysia Poland South Africa South Africa South Africa 0 50 100 0 50 100 0 100 200 Difficulty of Hiring Index Difficulty of Firing Index Firing Costs (weeks) Source: Investment Climate Survey Macroeconomic Instability 2.8 Macroeconomic instability was rated as a serious obstacle to enterprise operations and growth by about 33% of South African firms, and seemed to reflect the fact that despite positive growth and modest inflation, exchange rates have been unstable since 2000 (See Annex 3). Not surprisingly, exporters showed far more concern about economic instability than non-exporters (44% vs. 28 %). About 75% of exporters billing in US dollars, against which the rand has been most unstable, saw macro instability as a major concern. 32World Bank (2004d). Note that calculation errors in an earlier version of the Doing Business data base and a change in methodology resulted in South Africa being misclassified in early versions of the database. These errors have been corrected in the most recent publications. 19 Crime 2.9 Though violent crime rates have declined modestly over the past decade, property crimes have been increasing. The median firm reported losses due to crime and the cost of security higher than in many middle-income countries including China, Poland, Brazil and Russia. Security costs accounted for about two-thirds of the cost of crime, and direct losses accounted for the additional third. In general, manufacturing firms faced fewer losses than firms involved in retail and wholesale trade or construction, and large firms tended to face higher losses than smaller firms. Access to Finance 2.10 Firms in South Africa were generally less concerned about access to and cost of financing (fewer than 20 % rated either as a major obstacle), than in other countries, placing it at eighth and eleventh of eighteen constraints. One reason is that real interest rates have been lower than in most comparator countries. On the other hand, firms financed less investment through banks than in comparator countries and fewer firms had overdraft facilities. South African firms were found to rely heavily on retained earnings to finance both investment and working capital. 2.11 South African firms in general seem to have relatively little demand for external credit. Most firms that did not have loans reported that they did not want or need a loan and very few firms had been rejected for a loan. However, black-owned firms were more concerned about access to credit and cost of financing than others, regardless of their size, sector or age. Black-owned firms were also less likely to have applied for or to hold a loan or overdraft facility, were more likely to have been rejected for a loan, and tended to pay higher interest rates. Taxation 2.12 Most enterprises in South Africa did not see taxation, either tax rates or tax administration, as a major problem. Though tax rates ranked as the fifth greatest obstacle overall, there was a large gap between the fourth and fifth rated obstacles. In contrast, in the large majority of comparator countries (49 out of 52), tax rates are typically seen as a fairly serious constraint.33 Other Areas of the Formal Sector Business Climate 2.13 The International Monetary Fund identifies HIV/AIDS as one of the two main potential constraints on growth in South Africa. In general, however, firms were only moderately concerned about its immediate impact, only about 6% reported that it was having a strong impact on labor productivity and 4% reporting a strong impact on profits. The greatest immediate impact appeared to be on employee absenteeism. Firms did, however, appear to be more concerned about the medium-term impact of the epidemic, due to uncertainty about its impact on market size and profitability. 33Data from rapid response website at World Bank (http://rru.worldbank.org/). 20 2.14 Firms had few complaints about most other areas of the investment climate. Few firms rated infrastructure, regulation, corruption or the court system as serious obstacles. Most firms believed that courts were able to enforce property rights and that court cases are resolved relatively quickly. Losses due to power outages were modest and the cost of power low by international standards. The burden of regulation, while not low, was found to be lower than in most countries in Africa and comparable to most middle income countries. Few firms reported paying bribes to obtains services or win government contracts. Firms in South Africa faced a less predatory regulatory regime than a number of middle-income countries, e.g. in terms of the average number of inspections they had to undergo. In summary, the objective indicators seemed consistent with firm perceptions -- i.e. on most other areas of the investment climate, South Africa appeared to perform relatively well. SUMMARY OF ICA FINDINGS 2.15 Given the relatively favorable investment climate, the question is, why has private investment been so modest? The ICA suggests several possible answers to this question. Exchange rate instability has been especially problematic. Unit labor costs are high, especially for skilled workers. Labor regulation is still somewhat burdensome and the cost of crime is high. All these factors discourage investment, despite the generally favorable climate in other areas. The investment climate has improved quite recently along a number of dimensions ­ interest rates and inflation have fallen, corporate tax rates have been reduced, while there have been generally increasing profits. Meanwhile, investment rates have started to rise and this might reflect the improving climate. Finally, other areas not captured well in the firm survey, such as market concentration and systemic risk, may be important. B. THE MICRO-ENTERPRISE (MICA) SURVEY 2.16 Given the focus of the ICA on formal enterprises, the bulk of which consisted of larger firms of over 50 employees and only a tiny proportion of which was African owned, it was clear that there was a strong need for an independent survey of the smallest firms, to complement the ICA findings. Such a survey would help provide a clearer picture of the constraints faced by these firms, and the support they would need to improve their competitiveness and sustainability. Towards this end, we decided to conduct a firm-level survey of enterprises under 10 employees, on key investment climate issues such as the cost of doing business, the regulatory environment, the labor market and the financial sector, and on their awareness and participation of the incentive framework for SMMEs. The resultant survey of micro-enterprises, in effect a Micro- enterprise Investment Climate Assessment (MICA), provides a detailed cross-sectional snapshot of the investment climate constraints facing small and mostly informal firms, which ­ as it was based on the original Investment Climate Survey (ICS), can be 21 benchmarked with the investment climates for micro-enterprises in other countries around the world.34 2.17 The MICA survey was conducted in the context of this study covering about 240 enterprises in Tshwane, Ekhuruleni and Stellenbosch,35 (fuller details of the survey are provided in Annex 4). Over three-quarters of the surveyed firms were black-owned and, of the remainder, 10% were owned by Europeans and 13% by Asians and others. Asian- owned MEs were concentrated in the retail trade sector while firms owned by `coloreds and others' in construction and manufacturing. About three quarters had between 1 to 4 employees, with European-owned firms tending to be slightly larger. The median size of sample firms was 3 employees. 2.18 Most firms were young, 75% starting since 1994, with a median of 5 years. Most enterprises (60%) were in retailing and service sectors, 23% in light manufacturing, and 17% in construction, a profile similar to the ICA sample. About 38% of firms reported being registered for the value added tax (VAT), in some cases voluntarily (registering would often be helpful in allowing enterprises to recover taxes on their inputs). Fewer firms reported being registered with other government agencies (See Figure 1). Of these, more were likely to be registered with DTI (15%) or the municipal government (16). In total, about 70% of firms not registered with a government agency, were identified as `informal'.36 Retail firms had higher registration rates, possibly because they were relatively more visible to the authorities. Three quarters of business owners were male, more so in manufacturing and construction, and especially in white and Asian owned firms, while women owners were more prevalent in Black-owned businesses such as hairdressing. 34It should be mentioned that as background work for the 2005 World Development Report the World Bank undertook surveys of MEs with fewer than 10 employees in 13 countries, including Tanzania, Senegal, Kenya, Bangladesh, Indonesia, Brazil and Guatemala,. The ICA surveys generally covered registered manufacturing enterprises with 5 or more employees, and also services, construction and commercial agriculture. 35The survey was conducted in 2005 by the Bureau of Market Research at the University of South Africa, on behalf of the World Bank. Registered or unregistered (informal) MEs were randomly selected in retail, manufacturing (e.g. wood-working, welding, dressmaking), construction and services (e.g. hairdressers, automobile mechanics), in Pretoria (central business district, Soshanguve and Mamelodi), one third from Ekurhuleni in Gauteng (Alberton/Germiston, Tembisa and Katlehong/Tokoza), and one sixth from Stellenbosch in Western Cape. 36Reported registration with the tax authorities was regarded as unreliable and therefore excluded as an indicator of `informality'. 22 Figure 2.3: About half of the firms in the micro-enterprise survey reported that they were registered with one or more government agency SETA Retail Trade Other Govt Agency Industry Board Services DTI Municipal Govt Construction Light Any except VAT Manufacturing VAT All Any including VAT 0% 20% 40% 60% 80% 0% 20% 40% 60% 80% % of firms registered % of firms registered with any agency Source: Micro-enterprise Investment Climate Survey 2.19 Only in about 21% of firms did the manager have post-secondary university or vocational training. Managers in manufacturing were less well educated, and managers in black-owned businesses were predominantly only primary-school educated. In the larger enterprises (surveyed by the ICA) the level of education was found to be significantly higher across all sectors, with 80% of managers being university educated.37 2.20 The MEs across all sectors covered in the MICA were found to sell almost exclusively to consumers and individuals within South Africa. Only 2% or less of ME sales overall went to large firms (with more than 100 employees), suggesting poor integration into supply chains. Firms also reported very few sales to Government agencies (1 to 2%). None of the ME firms reported any direct or indirect exports. This lack of integration into larger manufacturing, or to Government, conforms to patterns in other countries. However, sales to small firms and wholesalers were also found to be less important for manufacturing MEs in South Africa than in most other countries. For example they accounted for 30% of ME output in Bangladesh, 24% in Indonesia, 17% in Guatemala and 12% in Brazil, whereas in South Africa the proportion of ME sales to 37Managers of MEs in South African were more likely to be educated at university level than the managers of MEs in other surveyed countries (e.g. Tanzania, Senegal, and Kenya). On the other hand enterprises in Kenya and Senegal were less likely to have a manager with only a primary education than enterprises in South Africa, while enterprises in Tanzania, Bangladesh and Guatemala were more likely to have a manager with only a primary education. The large difference in management education between informal and formal enterprises is general across countries. 23 other small firms and trade intermediaries was less than 8%. Accordingly, there appears to be a lack of ME integration in South Africa both into the first economy and into the formal sector in general. Micro-enterprise Performance 2.21 Interestingly, labor productivity turned out to be higher for MEs in South Africa than in other African countries (Tanzania, Senegal) by a ratio of about 3:1,38 and in other middle income countries (Brazil, Indonesia, and Kenya) by about 1.5:1. Figure 2.4: Labor productivity is higher for MEs in the manufacturing sector in South Africa than for similar firms elsewhere in Africa and other middle-income economies. S$)U( $2,000 erk $1,500 wor $1,000 per $500 added $0 value- uth Africa nzania negal nya la zil Ke desh Bra Ta Se So GuatemaBangla Indonesia Source: ME Investment Climate Surveys Note: All values are medians for manufacturing enterprises with available data. Value added is calculated by subtracting intermediate inputs and energy costs from sales from manufacturing. Workers include both permanent and temporary workers. Values are converted to US$ using average exchange rates for the relevant years from World Development Indicators. Data is for 2002, except for South Africa (2004). 2.22 This suggests that South African MEs are relatively productive. The productivity differential was found to be somewhat higher however for larger enterprises. There were also significant differences in productivity by enterprise ownership in South Africa. The median I in black-owned manufacturing was found to be slightly lower than the median firms in Kenya, Indonesia or Brazil. 2.23 In the following table, among the most notable features of micro-enterprise performance were that registered (i.e. formal sector), European-owned firms were far ahead of informal (African) firms in terms of labor and capital productivity, capital per worker and wages, and fairly well ahead of Asian firms in all these areas except for capital per worker. The larger micro-enterprises (6 to 10 employees) were also well 38Labor productivity only was computed because measurements of capital productivity tend to be of doubtful reliability, and were not available for the comparator countries. 24 ahead of the smaller ones. Wage rates and unit labor costs were also, however, higher in European-owned formal enterprises, suggesting that within similar types of enterprise the larger, European-owned firms were not producing at lower prices or higher profits, which were paid out in high salaries. The most productive enterprise would be a larger, registered, European owned service enterprise in Stellenbosch, and the least productive a smaller, unregistered, black-owned retail enterprise in Ekhuruleni. Table 2.1: Productivity Measures by Firm Type, Micro-enterprises in South Africa. Value Added Wages Unit Capital Capital Employment Sales per per Labor per Produc- Growth Growth Worker Worker Costs Worker tivity All R 9,455 R 2,820 28% R 1,343 4.0 0.0 0.0 Services R 10,999 R 3,453 39% R 2,015 4.1 0.0 0.0 Retail R 7,010 R 2,417 22% R 806 2.7 0.0 0.0 Manufacturing R 8,595 R 1,934 25% R 1,612 2.1 0.0 0.0 Construction R 17,123 R 2,753 33% R 1,220 6.7 0.0 0.0 Registered R 22,160 R 5,802 29% R 3,022 5.3 0.0 0.0 Not Registered R 3,143 R 1,450 26% R 672 2.5 0.0 0.0 African R 6,446 R 1,934 25% R 806 2.8 0.0 0.0 European R 58,958 R 18,466 36% R 8,058 9.6 0.0 10.0 Asian R 18,936 R 7,252 30% R 6,715 2.8 0.0 0.0 Other R 34,534 R 11,886 35% R 3,022 10.4 0.0 0.0 Tshwane R 14,827 R 3,868 19% R 1,007 2.5 0.0 0.0 Ekurhuleni R 2,820 R 967 27% R 1,007 2.4 0.0 0.0 Stellenbosch R 46,082 R 16,188 36% R 4,029 11.3 0.0 0.0 0-5 employees R 8,259 R 2,417 25% R 1,249 3.0 0.0 0.0 6-10 employees R 24,174 R 7,655 31% R 2,015 7.6 8.3 0.0 Source: Micro-enterprise Investment Climate Assessment Note: All numbers are medians 2.24 Comparing MEs to larger firms (from the ICA survey), the overall picture is that large firms were found to be very productive compared to firms elsewhere in Africa and to firms in other middle-income countries. For MEs, however, the differential was considerably less, especially for black-owned firms (Annex 5 shows details of productivity measures by type of firm in South Africa). 2.25 Median wage costs were far higher for larger firms in manufacturing than for MEs, by a ratio of as much as 30:1 (R69,900 in 2000 prices per large firm worker in wages and other benefits, compared to only about R2,000 in MEs (bearing in mind that 25 owner-workers might pay themselves higher wages out of the net income). Registered MEs, however, reported wage costs about five times those of unregistered firms, and black-owned firms reported significantly lower median wages costs than others. These patterns appear to partially reflect differences in labor productivity. Black-owned firms tend to produce less value-added per worker than other firms, as do unregistered firms and firms in Ekurhuleni and Tshwane. Unit labor costs (measured in wages as % of value-added) are, as would be expected, more comparable. 2.26 Within MEs in terms of enterprise growth, sales and employment saw modest increases during the study period.39 The median firm reported no sales or employment growth between 2001 and 2004. This pattern held across all sectors, in all regions where the survey was completed, and for registered and unregistered firms. The only exceptions were in the range 6-to-10 employees who reported 8% employment growth and European-owned firms which reported 10% sales growth between 2001 and 2004, at the median. About 32% of firms reported employment growth and about 38% reported some sales growth. Only about 7% of firms reported lower employment, while 34% reported lower sales. 2.27 Among larger firms (from the ICA survey) the median firm in retail trade, construction and manufacturing reported sales growth of 8% and employment growth of 4%. The median retailer reported 4% sales and 10% employment growth and the median construction firm reported sales growth of 17% and employment growth of 6%. The Investment Climate and Constraints on MEs Earlier (1999) Johannesburg Survey 2.28 By way of additional background, the constraints faced by the small and micro enterprises sector were highlighted in an earlier World Bank survey of 499 informal firms that had been conducted in 1999 in Johannesburg.40 97% of respondents were found to be Black-owned with, on average, three employees. They had little capital and almost no vocational skills. 63% were found to be new enterprises, established since 1995. Women constituted 38% of business owners, mostly in retailing, including hawking, and clothing and food. During 1997-99, while formal job losses continued, informal firms expanded and the number of full time jobs per firm rose from 2 to 3. Similarly, about double the number of informal compared to formal firms invested during this period. Profits averaged only about $250 per month (half the average formal sector employee income in Johannesburg), but low overheads helped businesses to provide a survival resource for the household. 39Even so there are well-known problems associated with using retrospective data--including the potential for survivor bias (i.e., the best performing firms might have grown fast enough that they had over 10 employees by 2004 and the worst performing firms might have exited entirely). 40Chandra, Vandana. 2002. "Constraints to Growth in Johannesburg's Black Informal Sector: Evidence from the 1999 Informal Sector Survey." World Bank Report, 24449-ZA. Washington DC, 2002. 26 The main constraints that firms cited in that survey were as follows: 2.29 Lack of credit was a problem for 80% of informal firms, caused by poor access and high interest rates. Owners financed investment from personal savings. Only 5% relied on private money lenders. Less than 10% used banks, and only 12% had ever applied (and 4% obtained) formal credit, deterred mainly by complex application procedures, lack of collateral and credit history and high costs. Lack of credit access reflected not only lack of formal status and track record, but the effects of past discrimination. 2.30 Infrastructure, services, transport and business space were reported to be inadequate, especially for businesses outside the home. Some 70% of firms had access to water, sanitation and electricity but only 40% were linked to postal services, telephones or public transport. As such, while access to services was a key issue for those not connected, the high cost of services was a constraint for many who had access. Public transport was a critical constraint: 40% of the business owners walked to work or operated from home. Over 70% of employees walked or used taxis. The geographical separation of black townships from the city areas created problems which added to poor transport, unaffordable business spaces and lack of storage facilities. 2.31 Low skills and training. 80% of firm owners had education above grade 4 but only a third had some type of informal vocational training, and no more than 20% received some training in business opportunity identification and machinery/equipment e.g. through the Small Business Development Centers (SBDCs), private firms, and Industrial Training Boards. Informal business owners had a latent demand for training as shown by reported levels of willingness to pay. 2.32 Linkages between informal and formal firms. On the input side most MEs were found to depend on formal large, small and government firms. Most informal firm owners considered the formal sector important for their business and 26% would have liked to formalize after reaching a certain size, to take advantage of its higher and more stable income opportunities. However, only 5-6% of MSMEs obtained business from government sources and there was a low prevalence of sub-contracting (less than 7%) from the large firm tier. Informal firms were found to be weakly linked with formal firms through the sale of their products and services. Roughly 88% of the informal firm's sold their products direct to private individuals or households. Only about 20% of the informal firm's sold their output to small formal firms (SMMEs) and less than 5% to larger formal firms. Aggregating small and large formal firms, middlemen and contractors, as well as government, shows that no more than 36% of the informal firms could rely on the formal sector for sales. 27 MICA Findings on Business Constraints 2.33 The current, wider geographically based MICA survey collected information on access to finance, transportation costs, access to land and quality of the enterprises' facilities, regulation, and taxation. This included both objective indicators and perception-based measures that asked managers what they saw as the major obstacles their firm faced.41 2.34 The MICA confirmed the Johannesburg study finding that access to finance and infrastructure are the main problems facing MEs. 2.35 As in countries like Brazil and Indonesia, managers (of manufacturing MEs) were most concerned about the access to and the cost of financing. Over 40% of managers in the MICA rated the access to financing as a serious obstacle, while about 30% were concerned about its cost. Managers also complained about access to transportation, power, business space and land. There was far less concern about other areas of the investment climate, including regulation or taxation. Fewer than 10% saw corruption as a problem, although given that many of the MEs in the survey were unregistered, this might be because their interactions with government officials were relatively limited. 2.36 There were important differences between the pattern of responses of MEs and those of formal (larger) enterprise in the ICA, especially with regard to the issue of finance. The difference in perception likely reflects the formal/informal divide, rather than the size of the enterprise as such. It is significant that none of the top five obstacles for manufacturing MEs (access to finance, cost of financing, transportation, access to land, and electricity) ranked among the top five constraints for the larger manufacturing enterprises in the ICA (worker skills, macroeconomic instability, labor regulations, crime and tax rates). In general, MEs were far more concerned about financing, infrastructure (transportation, power and telecommunications) and access to land, while large enterprises were more concerned about macroeconomic instability, worker skills, crime, aspects of regulation, especially labor, and taxation. Neither MEs nor larger enterprises were particularly concerned about anti-competitive practices and corruption, and policy uncertainty was only a moderate concern for both. 2.37 MEs were also much more concerned about infrastructure and access to land than larger enterprises were. Nearly 25% of the firms in the MICA sample reported that access to land was a serious constraint on their enterprise's operations and growth, especially in urban areas, whereas for larger firms in the ICA sample; only 3% reported this. Those that neither owned nor rented premises were significantly more concerned about access to land than other firms but, even so, relatively few firms reported that their land situation was highly unstable, and only 4% of firms had had to move due to a lack of secure title to their land in the last year, comparable with other middle income countries where data were available and slightly lower than some African comparators. Most firms had access to water and electricity. In general, firms in locations with better services 41Perception-based measures of the investment climate do however have several drawbacks, being based on individual business interests which might be a poor reflection of the overall needs of the sector. 28 were less likely to see access to land as a serious constraint. For example, whereas 33 % of firms in locations without lighting saw access to land as a concern, only 16 % of firms in locations with lighting did the same. 2.38 Also as mentioned, MEs in the survey tended to be less concerned than large enterprises about macroeconomic instability, regulation and taxation,42 worker education and skills and crime. MEs in the manufacturing sector were less concerned about macroeconomic instability than larger enterprises, except in construction and retailing, where the level of concern was more similar. 2.39 Larger manufacturers still think that instability is a significant concern, although inflation spiked in 2003, reaching 9%, but had been moderate over 1996-2005, while growth had been steady if moderate. The Rand, however, has been volatile, especially over 2001-2003. Despite relative currency stability since 2004, entrepreneurs seem to have retained a perception of significant exchange risk, especially in the case of exporters. Exporters in dollars, against which the Rand was most unstable, were the most concerned. Since none of the MEs exported, it is logical that they did not rate instability as a serious concern. 2.40 The larger firms in the ICA tended to be more concerned about regulation and taxation than the MEs. Whereas 18% of large manufacturing enterprises rated tax rates as a serious obstacle to enterprise operations and growth, only 8% of MEs did so. Large differences could be observed for labor regulation (37% of large compared to only 6% of MEs), trade regulations (17% compared to 6%) and tax administration (11% compared to 6%). One explanation may be that the MEs avoid the full burden of taxation and regulation. Consistent with this, registered MEs tended to be more concerned about tax rates and tax administration, and less about regulation, than unregistered MEs.43 42The finding of the study on regulations is not entirely consistent with other studies such as "Counting the Cost of Red Tape - for Business in South Africa" SPB Jul 2005. In this study firms of 11-49 employees were relatively concerned about regulations, in particular VAT and labor law. There was a strongly inverse relationship between turnover size and the unit costs of regulatory compliance. Informal firms (most of the MICA survey sample) by definition come outside the regulatory net, so this relationship applies only to smaller formal firms. 43These differences remain statistically significant after controlling for other differences between MEs in terms of size, ownership, location and sector of operations (see Appendix). 29 Figure 2.5: MEs in the manufacturing sector have very different concerns from the larger firms in the investment climate survey. el 60% act obs ousriessi 40% ngiyass 20% rmif of % 0% Finanof FinTranspAccess cing ancingortationto ty y s s on atesltions ation ation Policyeconomic Elec nc U Co nis Regul Access to Land tricityertainInstabilitker Skillpracticenications Cost Wor tive Labor cro Tax Ma Anti-competiTelecommu CrimerruptiTax RReguAadmiTradtre ICA firms Micro- and Very Small Source: ICA Rankings are for manufacturing enterprises only. 2.41 Few MEs in the manufacturing and construction sectors saw crime as a serious obstacle (only 11% of MEs in construction and 13% in manufacturing), compared to the 28% of larger manufacturing firms and 29% of larger construction firms that saw crime as a problem. These perceptions were consistent with information on security costs and losses from crime. MEs in retail and services were more concerned about crime (25% of retailers and 31% of services thought it a serious problem), and these sectors also faced greater losses. The subjective and objective data tell a consistent story: MEs are generally less subject to crime than larger enterprises in the same sectors. MEs in the service and retail trade sectors are more likely to see crime as a serious concern than other enterprises and also seem more vulnerable to crime. The cost of crime can be high for a minority of firms in these sectors and firms are more likely to be victims of crime in these sectors. 2.42 Despite its prevalence, very few MEs (2%) were seriously concerned about the economic impact of HIV/AIDS. The low level of concern across different types of MEs contrasts with the responses of larger enterprises in the ICAS, where - in manufacturing - 33 % of managers reported that they had been adversely impacted, about 12% in larger retail trade and 26% in larger construction firms, largely through absenteeism. Access To and Cost of Finance 2.43 The most significant difference in reported constraints on MEs and small/medium and large (formal) enterprises was in access to finance. Since this has clear implications for the design of public intervention it is worth examining it more closely. 30 2.44 Whereas only 13% of the larger manufacturing firms rated access to finance as a serious concern, 43% of the MEs did so. This pattern is common to the comparator countries, where in five out of seven cases, over 50% of MEs rated it as a problem. While cost of finance is a problem, access has been more of an acute hurdle to MEs, especially compared to larger enterprises in South Africa. This is also consistent with the patterns observed in other countries. The difference in perceptions between the MEs and the larger enterprises in South Africa is, however, greater than in most of the comparator countries, suggesting that South Africa's financial sector is particularly focused on large enterprises. 2.45 MEs were far less likely to use bank finance than the larger enterprises. Only 4% of MEs in manufacturing had a bank loan compared to 38% of larger enterprises. However, small (formal) enterprises were more likely to have loans than medium-sized or very large enterprises and financed more of their working capital needs through the banking sector. Thus, the divide is more between formal and informal than between large and small. MEs do not rely heavily upon the banking sector to finance their working capital needs anywhere in the world, and the difference between formal and informal enterprises in this respect is general. 2.46 In fact, despite South Africa's relatively developed financial markets, neither MEs nor formal firms appear to rely very heavily upon the banking sector for financing, compared to comparator countries such as Brazil or Kenya. For large firms this appears to be a matter of business strategy rather than constraints of access. 2.47 In contrast to the larger formal enterprises in the ICA, only about half of the MEs in the MICA said that the main reason that they did not have a loan was because they either did not want or need a loan. Rather, very few had applied and most cited problems of collateral, lack of information, or lack of formal registration, and lack of track record as preventing access. However, very few had actually had a loan application rejected. 2.48 There are also differences between different types of MEs. Larger MEs, especially in manufacturing and construction are more concerned than smaller firms, retailers and service providers, although access does not differ significantly. Firms in the construction sector and, to a lesser extent in the manufacturing sector, did appear to be credit constrained in the sense that (I) they did not have a loan and (ii) they would have liked to borrow at current interest rates, but were unable to do so. 31 Figure 2.6: MEs are consistently more concerned about access to finance throughout the world. Informal firms Formal firms Kenya Kenya Senegal Senegal Tanzania Tanzania Brazil Brazil Bangladesh Bangaldesh Guatemala Guatemala Indonesia Indonesia South Africa South Africa 0% 25% 50% 75% 100% 0% 25% 50% 75% 100% % of firms % of firms Access Cost Access Cost Source: ME Investment Climate Surveys Note: International comparisons are for manufacturing firms only Table 2.2: Access to finance by enterprise characteristic. % saying % of short- access % saying % with % with term assets % credit is serious cost is serious bank loan overdraft financed constrained problem problem facility through banks All 39% 30% 5% 8% 5% 31% Black 43% 32% 4% 6% 6% 36% European 17% 17% 17% 22% 9% 5% Asian 21% 31% 0% 21% 1% 14% Other 35% 35% 0% 0% 0% 25% Manufacturing 43% 28% 4% 5% 3% 38% Construction 51% 41% 3% 8% 8% 54% Retail 32% 29% 6% 10% 5% 16% Services 35% 27% 5% 10% 7% 30% 2-5 employees 38% 30% 4% 6% 4% 33% 5-10 employees 44% 33% 9% 20% 9% 23% University education 48% 38% 11% 19% 7% 17% Vocational/secondary 38% 29% 4% 7% 5% 33% Contracting 48% 28% 2% 7% 6% 31% Growing 41% 32% 6% 18% 10% 22% Source: ME Investment Climate Survey 2.49 Better educated ME managers were more likely to have accessed bank finance, and were also more likely to consider themselves credit constrained. In general, MEs that were growing were more likely to have bank loans, overdraft facilities and working capital finance, and were also less likely to say that they were credit constrained, suggesting that MEs without credit are unable to grow due to lack of financing. 32 Table 2.3: Micro-enterprises are far less likely to say that they don't have a loan because they don't want one. Manufacturing Retail Construction Services Informal Formal Informal Formal Informal Formal Informal Do not want 48% 95% 70% 89% 36% 89% 65% No Collateral 18% 3% 6% 5% 19% 0% 14% Other 16% 2% 8% 5% 28% 11% 11% Interest rates too high 12% 0% 13% 0% 8% 0% 4% Process too difficult 6% 0% 3% 0% 8% 0% 7% Source: Micro-enterprise Investment Climate Survey 2.50 MEs with primary owners of European and Asian descent were far less likely to see access to finance as a serious problem than other firms. European owners were also less likely to see the cost of financing as a serious problem than other MEs. Objective indicators of access to finance are also generally consistent with the view that access to finance is a less serious problem for the European-owned than it is for African-owned MEs in the sample. European-owned MEs were also far less likely to report that they were credit constrained. Overall, the results suggest that access to and cost of finance is a concern principally for black-owned MEs. 2.51 Why is this the case? One possible answer is that banks might discriminate against black-owned businesses. Other answers include unobserved differences in firm performance, registration status, and audited accounts. Other factors that might also affect firm performance are education and business experience of the management team. Many previous studies have noted that, because black people were actively discriminated under the previous regime, the quality of education that they received was often relatively substandard.44 Another possible explanation for the difference is the availability of collateral, which usually depends on the owner's assets and wealth, which was not surveyed. In general, better performing MEs will find it easier to get loans, and higher productivity among European and Asian owned firm might indicate the basis for such lending behavior. Registration is usually necessary for access to finance45 and black- owned MEs were far less likely to be registered than other firms. Black-owned firms also tended not to have audited accounts, which is a general requirement for access to bank finance. C. INCENTIVE PROGRAMS 2.52 To what extent were micro enterprises in the sample aware of the various incentive programs offered by DTI, other national government agencies, municipal governments, etc., to help them grow and enter new markets? Survey results indicate that only about 29% of MEs knew about any Government programs. Managers showed most knowledge (24%) of the Small and Medium Enterprise Development Program (SMEDP). 44See, for example, Fedderke, de Kalt, and Luiz (2000), Case and Deaton (1999) and Selod and Zenou (2003) for discussions of this issue. Selod and Zenou (2003) note, for example, that per capita spending was four times higher for white children than for black children and that the student/teacher ration was twice as high. 45For example, World Bank (2003, p. 22) states "it is impossible for [informal] businesses to obtain bank credit or use courts to resolve disputes'. 33 Only about 13% knew about the Khula guarantee program. Less than 12% of firms, and less than 13% of African-owned firms, had heard of the BBDSP. Few firms could name any other specific program, with only 3 firms mentioning Ntsika's Local Business Service Centers. In addition, one firm mentioned the Get Ahead Foundation, an NGO- financed group lending scheme. Retail firms had heard more about the programs than other firms, and larger MEs (6-10 employees) were slightly more knowledgeable than the smallest (1­5 employees). Managers of firms in Stellenbosch had significantly less information than in Tshwane or Ekurhuleni, suggesting that outreach is worse outside the big city zones. Figure 2.7: Few firms reported knowing about any of the incentive programs--especially smaller firms and firms in Stellenbosch. Retail 6-10 workers 2-5 workers Manufacturing Services Ekurhuleni Tshwane Construction Stellenbosch All All 0% 25% 50% 0% 25% 50% % of firms that have heard of program %of firms that have heardof program Source: ME Investment Climate Surveys 2.53 When firms that had heard of any of the programs were asked about how they had heard of them, half indicated that they had obtained information primarily from family and friends. The next most important source was the electronic and print media, radio and television. However, as in the earlier case, very few firms obtained information from official sources (13%) and publications (19%), with only 1% using the DTI website. Only 3 firms reported that they (4 % of the total who had heard of the programs) had been visited by a government facilitator for any of the programs. Of the 240 firms in the survey, only 20 were found to have obtained information from official sources, indicating that few firms had looked into applying for these programs seriously enough to investigate the options. 34 2.54 Looking to the nature of services of specific interest to microenterprises, firms were asked about their interest in seven specific types of services. 75% were found to be interested in further assistance, of which 49% wanted matching grants for new investment, 37% financial management training, and 36% loan guarantees. More than 25% of managers were interested in the least cited programs - improving management systems and management training. Around half of the MEs were interested in more than one type of incentive, some were interested in all of them. This interest was differentiated across sectors with (e.g. while manufacturers were most interested in investment grants, retailers had a relatively high interest in financial management). 2.55 Geographically, interest was lower in Stellenbosch, partly because MEs in Stellenbosch were less likely to be black-owned (black-owned firms having relatively greater interest in support programs per se). Figure 2.8: Over 75% of managers were interested in incentive programs ­ mostly related to new investment, training in management skills, and loan guarantees. Finance new investment Construction Financial management training Retail Trade Loan guarantees Marketing Manufacturing Employee training Services Improve management skills Improve management systems All Any Program 0% 30% 60% 90% 0% 30% 60% 90% % of firms interested in program % of firms interested in program Source: ME Investment Climate Surveys 2.56 Despite displaying wide interest, few micro-enterprises surveyed had actually applied to any program. The most common reason given was lack of necessary connections (34% of firms that had heard of at least one program), lack of response to initial enquiry (32%), and difficulty with the application process (31%). Only about one- fifth said that they did not need support in areas offered by programs, 18% said co- payments were too high and 10% said that they did not qualify. Firms that got information from official sources had the same responses, but the access problems seemed less severe. 35 D. CONCLUSIONS: THE INFORMAL AND FORMAL SECTOR ENVIRONMENT 2.57 Promotion of the SMME sector requires understanding of both formal and informal sector needs. The principal differences in the business environment, affecting on the one hand formal SMMEs and, on the other hand, MEs and informal enterprises, are reflected in the main obstacles they face. 2.58 The top five obstacles for MEs were · access to finance · cost of financing · transportation · access to land · access to transport/power/telecoms 2.59 The top five obstacles for the formal SMMES were: · worker skills · macroeconomic instability · labor regulations · crime · tax rates 2.60 Neither MEs, nor larger enterprises, were particularly concerned about regulatory enforcement (judicial capacity), anti-competitive practices or corruption. Policy uncertainty was a moderate concern for both. On access to, and cost of, finance there was an important distinction between (informal) micro/small enterprises and formal small firms. Furthermore, there was a lower level of concern overall about business obstacles than in many comparator countries, reflecting the fact that South Africa's business climate as a whole is probably relatively unobstructed relative to comparator countries where market reforms are less advanced. In the rest of this study we draw out of these findings and direct evidence of the impact of Government support measures some conclusions for the future design of supply-side industry support. 36 CHAPTER THREE HOW FAR ARE GOVERNMENT SUPPORT MEASURES CONFIGURED TO ADDRESS SMME NEEDS? A. INTRODUCTION: SURVEYS OF INSTITUTIONS AND BENEFICIARIES Government Support Programs 3.1 From the mid-1990s, the South African Government's support programs were aimed to address specific objectives as discussed in chapter one.46 Two key agencies established were Ntsika Enterprise Promotion, and Khula Enterprise Finance. These reflected a twin-pillar approach to enterprise support ­ supporting businesses on the one hand through management and technical advice, and on the other through financial assistance. These agencies, in turn- managed a range of programs, notably the Local Business Support Centers, the Tender Advice Centers in the case of Ntsika, and the Retail Financial Intermediaries and the Loan Guarantee programs in the case of Khula. The Manufacturing Advisory Centers (MACs) were initiated shortly thereafter, playing a role that somewhat overlapped that of Ntsika but concentrated more on specific technical assistance, while Ntsika provided a range of informational and advisory services. A range of incentive programs followed these initiatives. 3.2 How effective have these programs been? Have they achieved their social and developmental objectives? To assess this we undertook an evaluation,47 based on structured interviews and analysis of data, of a sample of the most relevant Government support schemes, firstly, from the provider's viewpoint, to get an idea of overall institutional effectiveness and, secondly, from the perspective of firm recipients, actual and potential. Apart from looking at the performance of individual assistance programs we aimed also to derive indications of overall institutional effectiveness through the lens of the individual programs. This was in the knowledge that one program is only a partial indicator of the performance of an institution, while with programs that differ in size, design, objectives and performance criteria it is difficult to come up with common performance standards.48 46For a useful review of the rationale for support also see Rogerson, Christian `The Impact of the South African government's SMME programs: a ten year review' (1994-2003) in Development South Africa Vol 21, No 5, December 2004. 47Conducted by Phillips, David A. for the World Bank and summarized in the reports: `South Africa: Industrial Incentives: A Review of Selected SMME Support Programs' (Draft), World Bank, October 2005, and David Phillips: `South Africa; SMME Support Programs - Beneficiary Assessment' (Draft) World Bank, January 2006. 48The issues include inter alia: multiple objectives, lack of clear agreed indicators, and lack of adequate systematic evidence for assessment. See C.Roche and L.Kelly `Evaluating the Performance of Development Agencies' in "Evaluating Development Effectiveness' World Bank Series on Evaluation and Development Vol 7, Transaction Publishers, 2005. 37 3.3 A common approach was taken to each program by assessing: (a) output, (b) outcome, (c) impact, (d) cost-effectiveness and (e) performance issues, using a scoring system over five dimensions of performance. We reviewed a total of eleven assistance programs ­ eight largely funded by the Government, and three by the World Bank. The amount of public funding of these schemes ranged widely from less than R30 million per annum (US$4 million) in the case of the Ntsika LBSC program, to as much as R2.0 billion per annum (US$300 million) for the SMEDP. 3.4 The organizations involved specifically in SMME support in South Africa number in the hundreds. They can be categorized into projects and programs of: · Central Government (e.g. DTI, DOL etc). · Central agencies (e.g. SEDA, NEP, Khula), which are in turn funded by Government, including the DTI. · Provincial and municipal public agencies (e.g. LEDAs, MACs, GEP). · NGOs and non-profit agencies. · Public-private programs (e.g. `Kickstart' with SA Breweries; Semele, Anglo- Khula mining trust fund; Eskom fund). · SME financing institutions (commercial and microfinance banks, venture capital funds ­ e.g. Business Partners SA). · Business consultants specializing in facilitating access to support. 3.5 As detailed in chapter one, the DTI is the principal central source of SMME support, overseeing a network of 21 affiliated agencies and directly managing a range of activities and programs through individual departments, including a range of incentive and support programs. It is mainly the latter group of DTI support programs, largely managed within DTI by TEO ("The Enterprise Organization"), EIDD (Employment and Investment Department), and TISA (Trade and Investment South Africa), that are our focus. Program Survey 3.6 The eleven assistance programs selected represent a cross section of public sector assistance types. The focus on the public sector reflects the priority need to rethink the rationale for public programs. The principal criteria for program selection were: (a) programs directed to SMMEs; (b) representative types of support; and (c) programs having both a national and a provincial focus. In the event, most programs chosen were at the national level. Only two were local, and one of these was a local project of a national organization. Further work would be needed to investigate the performance of the provincial and non-Government sectors, but the assessment of national level programs is per se important as it is the source of main thrust of SMME support. The following table lists the selected programs and their objectives. 38 Table 3.1: Selected programs and their objectives: Institution/program Main Public Interest Objective National level DTI/TEO ­ Small Medium Enterprise Increase general investment, especially for labor-using small-medium Development Program enterprises (SMEDP) DTI/EMIA(TISA) National Pavilions Facilitate access to export markets through collective initiatives DTI/ Khula Finance RFI program Strengthen microfinance capacity and micro-enterprise growth DTI/ EIDD (IDC) SPII Accelerate technological advance and growth by strengthening research, development and innovation DTI/ National Empowerment Strengthen black business investment/ market entry through diverse Fund investment finance DTI/Ntsika (SEDA) local business service centers General business development and technical advisory support for SMMEs Provincial level GODISA Trust ­ Softstart Promoting startup businesses through business incubator (targeting ICT (Gauteng) sub-sector) Gauteng Enterprise Propeller Supporting entry and survival by advisory/financial support services to provincial, small, black-owned business (Under GEDA) 3.7 Structured interviews were held with the management of each of these programs, and in almost all cases, follow up visits were made with management and with staff. Requests for further information were also made where clarifications and a more detailed picture was required. The overall picture in each case was built up on the basis of three main information sources: interviews; published data (e.g. audited annual accounts); and previous evaluation reports. A special effort was made to collect previous evaluations in order to build up as comprehensive a picture as possible. Beneficiary Survey 3.8 To complement the top-down assessment of programs from the providers' (i.e. the Government) side an in-depth, bottom-up, survey was also conducted of 23 firms - 16 program beneficiaries and 7 non-beneficiaries. The assessment also references the MICA survey of chapter two.49 The purpose was to assess how specific enterprises have, in practice, accessed, responded to, experienced and gained from Government support programs. 3.9 The survey group consisted of a sample of 23 SMMEs, ranging from the smallest firm with two employees to the largest employed with 160 employees across several sub- sectors, but confined to Gauteng province. (See Annex 6 for more details on the sample). Thus, the overall sample of enterprises was not fully representative in terms of 49The MICA survey included some questions aimed at determining respondents' awareness of and participation in these incentive programs. 39 geographical area but, nonetheless, taken with the previous MICA survey results, provides useful insights into the response of a diverse group of firm types to the Government programs that have been operating for the past few years. B. COSTS AND BENEFITS FROM THE PROVIDERS' VIEW - REVIEW OF GOVERNMENT SUPPORT PROGRAMS International Benchmarks 3.10 Comparable international performance benchmarks are not readily available for the diverse types of public assistance programs that we are looking at, which include business advisory services, support to financial infrastructure, business incubator programs, investment grants, investment finance, and matching grant facilities. Some benchmarks for enterprise support matching grant facilities are, however, available50 which might provide a starting point. 3.11 Enterprise support matching grant funds approved by the Bank between 1986 and 199951 ranged in size from $2 million to $30 million. The average grant per subproject ranged from $2,000 to $50,000 at current prices. The grants funded advisory or skills development services to between 400 and 1200 small firms. Total administration costs over project life ranged from $0.5 million to $6 million. Two cost-output benchmarks are available - administrative cost percent of grant value disbursed, and administrative costs per sub-project funded, as follows. 50See: Phillips, David A. `Implementing the Market Approach to Enterprise Support ­ An evaluation of Ten Matching Grant Schemes' World Bank Policy Research Working Paper #2589, April 2001. The grants were part of enterprise support projects in Kenya, Uganda, South Africa, India, Indonesia, Mauritius, Argentine, Bangladesh and Tunisia. The evaluation was updated to include more recent results information. 51The grants were all rated at least satisfactory by the Bank, and in one case, Mauritius, highly satisfactory. 40 Table 3.2: Comparative Matching Grant Fund Performance Benchmarks % admin cost/ grant % admin cost Admin cost per Project and date value (appraised) /grant value subproject ($) approved (actuals) (actuals) Indonesia Export 50.0 50.0 3532 Development 1986 Kenya Export 57.0 57.0 1326 Development 1991 Mauritius 15.3 16.4 1842 Competitiveness 1994 Argentine Export 18.1 36 0 6047 Development 1995 Uganda 33.0 a/ 40.0 a/ 1228 Competitiveness 1996 South Africa 6.2 22.7 3507 Industrial Competitiveness 1997 Bangladesh Export 37.0 40.0 4219 Diversification 1999 Tunisia Export 50.0 44.0 6431 Development 1999 3.12 In all cases but one (Tunisia) actual administrative costs were higher than the expected (appraised) level as a percent of funds disbursed. In one case (South Africa) the actual costs were three times the planned level, and in another, double (Argentine). With regard to South Africa the original appraisal report very much underestimated likely administrative costs. 3.13 Taking the actuals, the lowest administrative cost ratio was 16% (Mauritius) while the highest was 57% (Kenya) partly because this was a very small fund. The mean percent is 38%. Some of these values however were biased upwards.52 The mean administrative cost per sub-project was $3516, with a range from $1228 ((Uganda) to $6431 (Tunisia). 3.14 The relation between this benchmark and suitable benchmarks for the other assistance programs is not clear cut. The lowest comparable ratios are those of profit earning financial institutions. In the case of venture funds, management costs for smaller funds in developing countries are typically in the range 3% to 5% of fund value, plus a share of profits. In development banking operations the overhead cost range is similar. In the case of a development agency administrative costs would tend to be higher, from 52For example, some were covering some costs that were not strictly administrative (e.g. some advisory services provided by the management unit), or in another case the project numbers were distorted by partial subprojects. 41 perhaps 8% to 10% of funds disbursed upwards (in the case of IFIs including the World Bank), depending on the nature of the service provided. As shown below, the Pavilions program had overhead costs of only 6% and the SPII program 15% of funds provided. Advisory and finance services such as the GEP are aiming for about 20% to 25% administrative overheads. Non-revenue earning advisory services may have to be measured in terms of physical output per dollar cost rather than by this method. 3.15 The cost ratios for matching grants are widely varying owing to factors such as the size of the fund, the type of fund management, and the project life including extensions. Comparative overhead costs in other assistance programs have tended to be lower per dollar of assistance, probably because of less intensive management. The costs for selected programs in the following impact evaluations can be roughly benchmarked against this range of comparators. Performance Scores 3.16 We take a common approach to assessing each assistance program by analyzing (a) output, (b) outcome, (c) impact, (d) a preliminary cost/benefit assessment, (e) cost- effectiveness and performance issues, and (f) performance evaluation. To provide some standardization to the assessment between projects, a scoring system is used for five key variables.53 We developed, for this purpose, a set of subjective performance scores for the assistance programs on a scale of 0 to 5, with 5 representing the best possible outcome, and zero representing the worst possible outcome. The scores are not additive because, in some cases, they are contingent upon each other. In particular, adequate `economic rationale' is an essential requirement without which other scores may be redundant. Consequently, there is no `sum total' or `average'. A score of 2.5 represents the just- satisfactory level (approximately equivalent to the level at which the economic return would just cover the economic cost of the program). The following rating descriptions apply to the intermediate scores between the minimum (0) which is the worst possible case and the maximum (5) which denotes a result that is the best possible. 53Scoring approaches are widely used but inevitably remain limited in scope and partially subjective. E.g. the UK Government (DFID) has used a scoring system to evaluate its operations, known as the `Value for Money Indicator'. See D. Poate and C.Barnett ITAD April 2003, report for DFID Evaluation Department. 42 Table 3.3: Rating descriptions Score 1 The program has little or no economic rationale in terms of addressing market distortions; OR Beneficiary selection is largely inconsistent with its economic rationale resulting in minimal additionality; OR Operating efficiency is so poor that the program itself likely costs close to the value of the market distortions it is addressing; OR Its economic impact is close to zero (resulting from one or more of the above); OR safeguard and monitoring systems are non-existent. Overall, prospects for adequate improvement are very low. It should be closed down in the near future. Score 2 All, or most of, the above areas are unacceptably poor and earning the equivalent of a sub-marginal economic return. The program has a degree of merit/it may be susceptible to improvement. Alternatively, if there are no current signs of improvement potential, the score signifies that it is likely to merit closing down in the short run. Score 3 All the above areas are satisfactory, or there is sufficient strength in some to offset relative weakness in others. The project is overall worthwhile, would earn a slightly better than minimal required return. Barring an unexpected setback, the program is justified for continuation. Score 4 The economic rationale is very sound; beneficiary selection is well targeted and provides a high level of additionality; operating costs are under control and below benchmark; economic impact or cost-effectiveness is high and there are good controls and monitoring systems in place. Main Findings on Performance 3.17 Using the performance rating system explained above the programs that were evaluated in South Africa were assessed as follows: Table 3.4: Summary Performance Results for Selected Support Programs Program Economic Target Operating Impact/cost- Inspection, Rationale beneficiary Efficiency effectiveness verification Selection M&E Government funded programs SMEDP 2.5 2.0 2.0 2.0 1.5 EMIA Pavilions 3.0 2.5 3.5 3.0 3.0 Khula RFIs 3.0 2.0 3.0 2.0 2.5 NEF 2.5 2.5 2.5 2.0 2.0 SPII 3.5 3.0 3.0 2.5 3.0 Ntsika LBSC 3.0 2.0 2.0 2.0 2.5 GODISASoftstart 3.5 2.5 3.0 2.5 2.5 Gauteng GEP 2.5 2.5 3.0 2.5 2.5 Total 2.9 2.4 2.8 2.3 2.4 Bank-supported programs Competitiveness Fund 3.5 2.5 3.0 3.0 2.0 Sector Partnership Fund 3.5 2.5 2.0 2.2 2.2 BBSDP 3.5 2.8 3.0 Na Na Total 3.5 2.6 2.7 2.6 2.1 43 3.18 Summary Performance rating. In sum, the best performers assessed were the National Pavilions, a program under the Export Marketing and Investment Assistance Scheme (EMIA/TISA) and the Support Program for Industrial Innovation (SPII). Both these are relatively old-established programs with a widely acceptable economic rationale (in terms of diffusing export know-how and supporting potentially socially beneficial R & D) and relatively good operational and follow-up effectiveness, although not without problems. Of the three World Bank-supported funds, the Competitiveness Fund came out best, and comparably, with the two best Government-funded programs. This is because of a combination of an appropriate economic rationale, good assisted firm performance compared to the counterfactual, possible spillover benefits, and reasonable operating efficiency. The BBSDP impact study showed acceptable performance on some dimensions. The SPF seems to have been less effective, largely because of the difficulty in finding strong, self-managed partnerships. The weaker programs, in terms of rationale and operational capacity, are firstly the SMEDP, which appears to have excessively loose selection criteria. The NEF is a newly started program, designed to address directly BEE firms and may face operating problems of cost-recovery although it is early to make judgments. Khula Enterprise Finance's Retail Financial Intermediary (RFI) program has also proved weak on selection and economic impact, considering the large number of RFI failures. The LBSC program proved disappointing for reasons of uneven geographical spread, inadequate funding and management problems. 3.19 Economic Rationale. Appropriate economic rationale is key to program justification, and was relatively strong for most programs. In the case of the World Bank assisted programs, the rationale remains valid of (i) the mobilization of local BDS capacity to assist small firm development, especially by black-owned businesses, and (ii) the development of industrial clusters through partnerships to build technological information and improve productivity. The economic rationale is less clear in two programs. For the SMEDP, investment grants to medium scale enterprises have become less justifiable since the capital market has evolved and medium scale firms themselves are not reporting that they are credit constrained. For the NEF, while the essential rationale to target finance to black-owned business is acceptable, the method of subsidized lending and potentially favorable equity funding is not easily supported, and there are other avenues of financial support to black enterprise, such as loan guarantees against public contract awards. 3.20 Even where the economic rationale is acceptable, this does not per se justify a particular program. Each program has to meets its objectives in a cost-effective way, the most important of which would be the creation of appropriate economic impact. Such impact requires additionality ­ i.e. investment, employment and output activity over and above what the private sector would have provided without assistance. Additionality is a necessary but not a sufficient condition for economic impact which requires public benefits through externalities, usually technological and transaction spillover effects: e.g. technology transfer, skills development, new market creation, and significant profitability-creation within existing markets. Finally, the public benefits have to be provided at acceptable public cost. 44 3.21 Against this background, the scores for all the SMME assistance programs in areas other than `economic rationale' are relatively lower. This means that they do not seem, on the whole, to have exhibited strong design or operational effectiveness. To the extent that these results are typical of the capacity of the agencies managing the programs, these results would also be representative of the capability of the whole public support system. In summary the results on the other four areas were as follows. Eligibility and selection of firms 3.22 The principal reason for relatively low scores on this dimension was weak targeting, and the lack of selection criteria addressing additionality and economic impact and difficulties in selecting providers able to meet norms of efficiency. 54 3.23 The best score applied to the SPII program, where criteria were clear and applied carefully. The worst score applied to the SMEDP, where there is evidence that many (especially medium scale and European-owned) firms did not actually need grants to carry out investments (some firms even applied for grants ex-post) and some failed to meet objectives such as employment creation. 3.24 Apart from the economic impact issue, other concerns include the level of compliance with selection criteria. In this context the NEF, which has rejected over 95% of applications, has applied criteria aggressively compared to the SMEDP which has rejected about 20%. The selection criteria of RFIs by Khula has, in principle, been appropriate but, in practice, too loosely controlled, with about 50% of RFI support loans having to be written off later. The same applied to the Ntsika LBSCs where, despite appropriate selection criteria, many of the accredited centers apparently ceased to function effectively. Selection under the Bank supported funds was based on clearly specified criteria but did not include an a priori judgment about economic impact. The BBSDP had a very high rejection rate, suggesting aggressive application of selection criteria. The CF had a relatively low rejection rate combined with a high cancellation rate after approval, which suggested less efficient selection. Operating efficiency 3.25 The best performers were programs with more well-established procedures and simple, clear operating rules, such as EMIA Pavilions and SPII. The Pavilions program has a low management cost per approval value of 6%. The SPII, which has operated for 25 years, retains a reasonable 15% management cost ratio. Nevertheless there are problems because the throughput of eligible projects for the SPII scheme has been falling while management costs have increased. In the case of the EMIA National Pavilions 54One indicator of weak additionality is if ex-ante willingness to pay for service inputs at the market price is high, overall. In such cases the probability is high that a number of firms would go ahead anyway with improvements, training or investments even without a subsidy; hence, the public rationale for helping them would be relatively weak. Better targeting and selection of grant levels can ensure that firms which do not need assistance are excluded from programs. On the design of matching grant schemes, lessons of experience are discussed in D. Phillips: `Implementing the Market Approach to Enterprise Support' World Bank Policy Research Working Paper, April 2001. 45 program there has been some criticism regarding lack of adequate scope and responsiveness. 3.26 The lowest scores on this criterion went to two programs, SMEDP and Ntsika LBSC, both of which seemed to be overloaded and with inadequate capacity to deliver a large workflow or a complicated agenda. This is reflected in SMEDP's very low management cost per value of approvals of 2%, which signifies not so much operating efficiency as lack of capacity to manage the program at the required level. The large SMEDP workflow has resulted from application overload partly due to funding of medium scale enterprises with no clear need of support, and lack of capacity for follow- up and verification. Ntsika's LBSC management cost ratio was an apparently reasonable 12% but it still suffered from capacity problems and other issues such as inability to coordinate and fund the LBSCs adequately. 3.27 Khula, Softstart and GEP appear to have experienced management who are on top of the issues, despite earlier problems in Khula over 1997-99. Both Khula and GEP have a relatively high (expected) management cost per amount of funds disbursed compared to other similar programs, although lower than the benchmark range for matching grant funds. Khula's own finances are also highly reliant on DTI support but will improve as the weaker part of the RFI portfolio is to be transferred to the new Micro Apex fund. (See below). NEF operational effectiveness is too early to assess, but there are indications of difficulties because of the assumption of heavy obligations in a new financial institution, combined with equity and partially subsidized loans finance which will make it reliant on outside subventions. 3.28 For the CF project, overall management costs per approval were within reasonable limits, both with the external manager by year three, and later on with local management. However, approval rates fell greatly after the departure of the external manager. Processing delays were relatively low while, in the case of the SPF, management costs were low but processing efficiency was also low, with inadequate management capacity. Economic impact/cost effectiveness 3.29 Achieved economic impact differed from potential economic impact. Several programs do not seem to have achieved their potential, leading to a weak overall effectiveness/impact score. For example, the failure rate among RFIs has been high, resulting in a sub-marginal economic return to the RFI program. The LBSCs have had management, program and funding problems, and lack of adequate local participation which have limited their economic benefits. The SMEDP has not significantly improved investment performance compared to firms that did not receive funding, while its employment impact seems to be relatively small (mostly located in the urban areas). Softstart and the GEP have some reasonably good areas of performance but appear marginal on operating features. 46 3.30 For the CF, economic returns were acceptable, explained by a combination of reasonable net additional productivity impact and some spillover benefits, in conjunction with reasonable management costs and grant level (50% of total cost). For the SPF, a higher grant level (65% of total) and fairly high management costs, combined with low, if any, net productivity gains and weak spillover effects, were reflected in negative economic returns. The SPF also seems to have been less effective overall, largely because of the difficulty in finding strong, self-managed partnerships capable of generating profitable additional collective investments and linkages. The BBSDP had the highest grant level (80%) but reasonable additional productivity impact, and may have achieved an acceptable economic return. Inspection, verification, M & E. 3.31 This performance area was marginal or poor for all programs. The best programs in this respect were, again, the EMIA Pavilion program and the SPII program. The former has a well-established monitoring system designed to assess export achievement, but might be able to strengthen it by assessing the performance of assisted vis-ā-vis non- assisted firms in generating export profits. The SPII program has strong verification and quite good follow-up because the ex-post performance of some assisted enterprises is the basis for levies on revenue. However, no programs have fully working M & E systems. The SMEDP, which collects a considerable amount of firm data, does not appear to have the staff capacity to handle the high flow of verifications, follow-ups or evaluations and, therefore, cannot use the information it has available. The NEF has not as yet set up an M & E process, although its position is better because throughput is, so far, relatively low. The Bank-funded programs did not have systematic M & E systems. Plans to introduce a system were not followed through in the CF. What are the Lessons? 3.32 The overall conclusion from this review is not very positive. Achieving economic impact through State programs is difficult, and progress seems to need to be phased so that a learning process takes place. Since 1995 there has been a large number of programs set up, either simultaneously or before the lessons of the earlier programs can be taken into account by the later programs. What other reviews have said is that there should be a focus on a few good models and a reduction in the number of programs. It is difficult to accurately monitor the value of State support programs once they are in place because they tend to become self-justifying - their output tends to be confused with their impact. 3.33 Regarding overall institutional effectiveness the impact of the programs reviewed is probably representative of the effectiveness of the institutions responsible for them. For DTI this implies, not surprisingly, that some are reasonably good models while others are not so good. A general problem is that of overambitious or over-complex program goals, lack of targeting and duplication of effort. In several cases it would have been better to focus on a few successful projects to use as models for dissemination. The desire to scale up assistance as quickly as possible has tended to be at the cost of reduced effectiveness of each operation. This is probably particularly true with the RFIs where a narrower concentration was needed to build sustainable institutions. It seems to have been true 47 also of the LBSCs, where an emphasis on numbers apparently resulted in neglect of existing LBSCs. The NEF is charged with both supporting potentially marginal businesses, achieving financial sustainability, and managing a portfolio of mature investments at the start of its life, which may create similar difficulties. On the other hand, the Pavilions Program and SPII have kept to fairly modest objectives (some may argue too modest). The GEP similarly has built on MAC activities but in a fairly measured and realistic way (e.g. introducing just one financial product).55 3.34 There is a continual problem of excessive administrative burden. This derives partly from the need to install safeguards against misuse of public funds, but it also leads to the use of considerable management time in grant applications and claims. This has, in turn, led to the widespread use of consultants and facilitators to make applications in return for a fee. Such intermediaries can add value by improving grant allocation and speeding up the process, but equally it is important for firms to take more responsibility for their own projects. Furthermore, intermediation has at times been collusive or has worsened grant allocation. In addition, though the administrative cost of these programs has been significant there has in general been inadequate emphasis on many of the programs on promotional expenditure (or such expenditure has not been cost-effective). As a consequence (discussed further in section E and Chapter 5), public knowledge of the programs has often been very limited. 3.35 Despite clear mission statements, many programs (for example the LBSCs) have failed to avoid some of the pitfalls that they were specifically designed to avoid ­ such as the tendency to become supply-driven, to use a one-size-fits-all approach to business assistance, or to become isolated from other parallel agencies instead of collaborating with them. In some cases funding was overstretched and, as a result, not sufficiently reliable to allow good program development. Thus, even where the objectives were justified in terms of public benefits the methods of achieving them were not necessarily cost-effective and in some cases seemed to involve public costs that exceeded the value of the potential public benefits. 3.36 Since 1995, a large number of programs have been set up, either simultaneously or before the lessons of the earlier programs could be taken into account by the later programs. What other reviews have indicated, and which we would concur with, is that there should be a focus on a few good models and a reduction in the number of programs. It is difficult to accurately monitor the value of State support programs once they are in place because they tend to become self-justifying - their output tends to be confused with their impact. 55Another program is the MAC network. This was not in the sample evaluated here but other reviews have suggested that it has been relatively effective compared for example to the LBSC program due to competent management, effective delivery of services, and stronger core skills. See Rogerson; op cit. 48 3.37 To summarize, the main lessons for program design are: At the institutional level, to: · Identify a clear economic rationale for programs of support; · Simplify the institutional structure supporting that rationale by rationalizing, cutting or merging agencies; · Separate service provider agencies from (central) policy, regulatory and oversight agencies, or Government departments; and · Focus on, and test, good program management models using, where possible, specialized local management resources outside the public sector. At the program level, to: · Develop program model and ensure effective performance before scaling up ­ i.e. learn by doing; · Develop a cost-effective promotional effort; · Maintain relatively simple program objectives and reasonable workload; · Select sub-projects according to additionality and public benefit criteria; · Select a subsidy level (grant component) that reflects the level of willingness to pay of the appropriate target group of firms; 56 · Negotiate at time of project preparation between Government departments (such as Treasury and SARS), to anticipate operational problems arising from Government regulations; and · Install unified databases and simple but informative monitoring and follow-up procedures with, if possible, comparative data on assisted and non-assisted firms. C. THE PERFORMANCE OF FOUR CRITICAL PROGRAMS 3.38 The following sections provide an outline of the evaluation of four programs. These are the SMEDP, Khula Enterprise Finance RFI program, Ntsika LBSC program, and the Bank supported Industrial Competitiveness and Job Creation Project, which consisted of the CF (Competitiveness Fund), the SPF (Sector Partnership Fund) and the BBSDP (Black Business Suppliers Development Program). Fuller evaluations are available in the background paper for this study.57 56 In practice, most matching grant schemes have selected a 50% grant level on grounds that this is a simple `matching amount', and the assumption is that target firms would be willing and able to bear the cost of 50% of the activity. If firms are in fact willing to bear substantially more than 50% then part of the subsidy would be redundant. Conversely, if they are not able to bear as much as a 50% share, they will not enter the scheme and to bring them in would require a larger element of subsidy. Either way, it may be justified to phase down the subsidy for repeat applications. 57 Phillips, David A.; op cit. 49 (a) The SMEDP 3.39 The SMEDP is the largest DTI support program outside the MIDP58 and, with advance commitments of over US$1.0 billion, is one of the largest SME programs in the World. As such, it is a particularly important candidate for a cost-benefit assessment of public expenditure. Its intended public objectives are: to increase investment, to promote smaller firms, and to increase employment. The objectives presuppose market failures in the cost of or access to finance, market entry in general and for black business in particular, and as a result of a segmented labor market penalizing black workers.59 3.40 The program started in 2001 and is due to close in September 2006, and its successor is under consideration. A 2004 impact review60 and DTI data showed about 75% of funding going to manufacturing and 25% to tourism, with 75% also in urbanized Gauteng, Western Cape and KZN provinces. The grants are payable over two years for eligible investment in land, buildings, equipment and vehicles, and for a third year dependent on achievement of a labor cost target (more than 30% of operating costs). A firm with the maximum eligible `qualifying assets' (R100 million) could receive up to R3.05 million per annum, or R9.15 million in grants over three years, i.e. about 9% of qualifying investment. The small firm category (up to R5 million in qualifying assets) can receive up to 10% of its investment per annum which, over three years, amounts to up to R1.5 million, a maximum of 30% of investment. Thus, the grant is structured in favor of smaller firms, although the largest grants in absolute terms go to the larger firms. 3.41 The grant and the reimbursement applications require considerable control documentation including audited financial statements under GAAP, owner equity conditions, checking of assets in place, and setting of performance targets for capital structure, turnover, employment, labor costs etc. The approvals are made by an evaluation committee on the recommendation of grant officers, however, eligibility rules are leniently applied initially (e.g. 25% of targeted achievement in first year), with scope for waivers by the Board. 3.42 Total commitments over 2005-2008 are about R7.0 billion (over US$1.0 billion). As of mid-2005, about 6,500 projects had been approved, with disbursements amounting to R1.1 billion (abut US$180 million). The `additional' (private) investment mobilized is, in principle, supposed to be about R10 to 12 billion. According to the management, the application rejection rate is about 20%, including rejections prior to approval. Project cancellations are running at the rate of 20 to 25% of approvals. Rejections are usually on account of failure either to meet selection criteria, special conditions imposed or, in some cases, after approval. 58The Government's MIDP support through the IRCC is running at about R15 billion a year (over $2 billion) in the form of tax and duty exemptions. 59A seminal analysis of the labor market in South Africa is: R. Lucas and D. Schydlowsky: `Evaluation of the RIDP of South Africa ­ Comments on Draft summary of findings and recommendations of Ernst and Young' April 1996. 60`Impact Review of the Small Medium Enterprise Development Program" DTI July 2004. 50 3.43 The throughput of grants (average 30 a month) is considered by the Board once a month for about 2-3 hours thus, only a cursory examination has usually been possible of each application, heavily reliant on staff work. There is, however, a capacity shortage in DTI. Forty staff under a senior DTI director are not sufficient to do required checks, e.g. for assets-in-place. Applications tend to be rejected because of lack of capacity to handle queries and there is little capacity to follow up after the grant period; hence, useful initial data, available from financial statements, are not used for monitoring or reporting. A disbursement lag of up to 12 months has built up with a major increase in disbursements projected from 2005 onwards. Against commitments of as much as R2.0 billion per annum, annual budgetary appropriations have been running at only about R400 million per annum. 3.44 Approvals and disbursements up to date have been as follows: Table 3.5: Approvals and Disbursements 2000-01 2001-2 2002-3 2003-4 2004-5 Total Approvals (projects) 36 596 1933 1711 2252 6533 Disbursements Rm. 1.20 8.58 161.9 354.0 562.1 1087.8 Source: DTI 3.45 Has the program created additional activity? While, in principle, the grants are supposed to leverage ten or twenty times their value in additional investment (according to the size of qualifying assets) interviews suggested that this was not happening. Some firms stated that the grants had been catalytic, but others stated that the grants had been received after investment had taken place and were used simply as a `windfall' rebate or subsidy.61 The 2004 impact study found that assisted firms performed slightly better than non-assisted firms, so there was some additional impact. 30% of assisted firms reported that they had introduced new products; 30% reported market development, and 20% reported that they had restructured. 22% invested sooner than they would have done (the small firms did best on this dimension). But, only 34% of manufacturing firms had actually increased investment as a result of the investment grant (43% in the smallest category but only 17% in the next group, and only 14% of the largest group). At the same time, for smaller firms, the grant was partly replacing inaccessible bank finance.62 3.46 This finding implies that investment decisions were relatively little affected by the grants. This is partly because, in the case of the larger enterprises, it was a small percent of investment (maximum 9%)63, while larger enterprises have longer term investment plans and, as we s, the SMEDP does not seem to have been critical to their growth, since 61According to the rules it is understood that the actual investment has to be made in the same year as the grant award, but could in fact be made after the award. While there are cases where this is legitimate but often not, and it is essential to set the rules so that complementary investment is truly additional. 62Very few (22%) small firms applied for loans and only 17% were successful. In tourism projects only 15% overall stated that the grant led to new investment decisions, more so in the larger tourism enterprises. 63That is the equivalent of a one percentage point reduction in the prevailing applicable cost of finance which is in mid-2005 at prime (10.5%) plus about three percent. 51 many of those interviewed seem to have actually decided not to apply for grant funding while still carrying out some investment. 3.47 Does the SMEDP focus on smaller enterprises? In terms of size of assisted firms, as of 2005, 75% (4,900) of those assisted have been in the category of less than R5 million in qualifying assets, a further 19% (1,240) below R20 million, and the remaining 6% (about 400) have more than R20 million in assets. In terms of value, however, only about 55% of funds have actually been going to the smallest category of firms. Of the 45% going to larger firms, the largest eligible category of 400 firms received 15% (over R 1.0 billion) in grants. It is also important to note that the size of the participating firms is larger than implied by the above asset categories because `qualifying' assets are only a part of the firms' total assets. This means that probably less than 50% of funds actually went to the smaller category of firms. 3.48 What about employment creation? The 2004 report estimated that the average increase in employment in beneficiary firms was about 22%. Investment in smaller firms was more labor-intensive, as data showed that the investment cost per job rising with the scale of production. The job creation potential had however been only about half of the outcome expected in the initial proposals for grants. The 2005 DTI Annual Budget report is less optimistic still, projecting only 27,000 jobs from R1.4 billion of SMEDP funding, plus a claimed R8.2 billion of leveraged investment, or R350,000 (US$55,000) per job created. Interviews in the field supported the low employment impact because the third year employment target was not being met by some firms who merely took advantage of the first two years' grants. 3.49 Does it meet BEE objectives? There are clear indications that the grants have not been going to BEs, since: (a) the average grant actually disbursed was R166,000 and commitments per firm over three years were considerably higher, while, if complementary investment is included, then average asset size per firm must be several million Rand;64 (b), the relatively burdensome demands for documentation (including audited financial statements) would rule out informal or micro-enterprise participation; and (c) the field interviews suggested that black-owned businesses were a very small proportion of those assisted. 3.50 Are the assisted firms generating spillover (social) benefits? While not an explicit goal of the project, catalytic linkage-inducing (or spillover) effects are an important public benefit of assisted investments. The 2004 study suggested that in this respect the effect of the grants was weak. For example, only 11% of manufacturing firms (24% of tourism firms) reported that the grant enabled them to integrate upstream and downstream. 64This size would also be implied since the small firm category qualified with assets of up to R 5 million. 52 3.51 A summary of some indicators of SMEDP performance is as follows. Table 3.6: Summary of some SMEDP Performance Indicators SMEDP operating costs per approval R3,500 Operating costs/ disbursements 2% Disbursement delay Large and increasing Total investment per job created R350,000 plus Total SMEDP funding per job created R30,000 to 50,000 3.52 As mentioned earlier, SMEDP operating costs per million Rand disbursed are extremely low by the benchmarks outlined above, at only 2%. This is despite a disbursement backlog. While this could suggest high operating efficiency, in this case it would appear to be on account of a lack of capacity, resulting inter alia in inadequate screening. The high investment cost per job created is suggestive of this because it indicates that selection criteria were not rigorously applied to weed out firms that were unlikely to meet additional job targets, despite the objectives of the program. Based on the above, qualitative performance scores, using the system discussed earlier, are shown in Annex 7. 3.53 Overall, the impact of the SMEDP does not seem to have been promising. This is born out by this assessment and the 2004 impact study using a control group for comparison. While assisted enterprises reported progress on a number of dimensions, additionality was limited, investment and employment creation below expectations, and spillover effects were not in evidence. In terms of business strategy the grant has been used more as a cash flow supplement than for reducing investment cost. It has helped to mitigate risk and lower hurdles but has not stimulated changes in investment behavior. 3.54 The main conclusions from the assessment were: · Future SMEDP-type programs should focus only on smaller firms; · Selection criteria should include simple indicators of the projects additionality and likely net economic impact;65 · Applying tighter selection criteria will help to alleviate capacity constraints; · The process should be simplified so that applicants do not necessarily have to work through consultants; and · Data from financial statements should be updated and used for monitoring or reporting based on follow-up of grantees after the end of the grant period. 65Selection indicators should answer questions about the public benefit ­ i.e. potential catalytic effect on labor-using investment that can be attributed to the grant. It must be established that without the grant there would be significant delays or lower investment in an opportunity, and also whether such investments create social benefits ­ black employment, entrepreneurship, linkage or technology-spreading effects etc. 53 (b) Khula Enterprise Finance RFI Program 3.55 What are Khula's objectives? Khula Enterprise Finance aims to mobilize finance, improve smaller firm access to finance, encourage commercial banks to provide loans to black-owned business, and promote equity finance for black-owned businesses. It is a limited liability company governed by a Board, with DTI as its major shareholder along with a group of donors. It operates the RFI loans scheme, KhulaStart (micro loans), Emerging Entrepreneur Scheme, Equity Fund, Individual Guarantee Scheme, Empowerment Scheme, Portfolio Guarantee Scheme, Institutional Guarantee Scheme, Land Reform Facility, and a new ICT loan guarantee fund. It has a R40 million joint venture with Anglo-American, and with Shoprite to enable small business to own retail premises. 3.56 DTI funded Khula with initial equity of R320 million in 1996 and has provided additional funds through `shareholder loans', for a cumulative R540 million.66 With DTI help, Khula made a net surplus since 2001 following large write-offs in 1999-2000 of RFI loans, however, asset value adjustments have weakened its financial position. Despite its corporate status, it is vulnerable to political control, and currently is expected to support MFI assistance programs and to subsidize its own lending. DTI's recurrent funding for Khula operations has been as follows. Table 3.7: Government Recurrent Funding of Khula R Million. 2001-02 2002-03 2003-04 2004-05 2005-06 Funding from 64.0 20.0 25.0 25.0 26.5 DTI Source: DTI budget and Khula Annual Reports 3.57 What is the Aim and Scope of the RFI Program? The Retail Financial Intermediaries program (RFI) aims to increase access of very small firms to finance by building capacity in the microfinance institutions. It has provided wholesale funding to Retail Finance Intermediaries, subject to eligibility criteria, including: number of clients, average repayment rates, and operational efficiency. Khula also accredits service providers to provide training and mentoring to RFI managements. The RFI program was initiated with the purchase for R26 million of a loan portfolio from the DBSA, which included some problem loans that Khula later had to write off. 3.58 Khula finances the RFI through one semi-commercial fund (business loans) and three low interest loan or grant funds, known as Pioneer funds, Seed funds, and Capitalization funds. About 10 RFIs out of 50 have been rejected for funding mainly because of weak or untrustworthy management, inadequate business plans or inadequate rollout, while some RFI applicants withdrew largely because of unwillingness to disclose information. The RFI program has now assisted a total of 42 RFIs, of which 22 were startups and 20 were existing funds. 66See published annual reports, 2002, 2003, 2004. 54 3.59 Total cumulative funding of RFIs since inception is close to R600 million. The largest single funding was R120 million (Ithala fund), followed by R58 million (Marang fund) and R53 million (Khethani fund). The top five account for about 50% of total RFI funding. The current (2004) RFI business loan portfolio is R123 million and the capitalization loan portfolio is R24 million, after repayments, write-offs, and conversions to grants. In 2004-05 the business loans were in good condition with only 6% due and payable from 16 RFIs but the more concessional capitalization loans were in poor shape with 40% arrears and 8 out of 16 RFI's outstanding `capitalization' debt being written off, while several other debts are in difficulties or are being converted to grants. Write- offs in 2004-05 were about R10 million, the first since 1999. Disbursement by the RFI program over 2001 to 2005 was as follows: Table 3.8: Khula RFI Program Output 2001-2 2002-3 2003-4 2004-5 Number of borrowing 16 16 14 12 RFIs Business loans Rm. 18.5 38.7 58.2 66.3 Capitalization loans 11.1 16.4 10.3 15.3 Rm. Seed loans Rm. 13.9 15.2 13.0 13.2 Pioneer loans Rm. 0.2 0.1 0.5 2.3 Capacity building 2.2 4.9 3.2 1.8 grants Total disbursements 45.9 75.3 85.2 98.9 Disbursement per RFI 2.9 4.7 6.1 8.2 Source: Khula Enterprise Finance 3.60 The RFIs' own loans funded by Khula have risen steadily over the past three years. Over 2004-5, 13 RFIs funded by the Khula's RFI program disbursed R258 million, over two thirds to businesses owned by black women. R157.0 million (60% of loan value) was on-lent by two RFIs (Marang and SEF) while at the low end, two RFIs on-lent less than R1.0 million each. Table 3.9: Khula-funded RFI Loan Activity 2001-2 2002-3 2003-4 2004-5 Estimated loans by RFIs to clients of 104431 81238 93687 115440 which: first time borrowers 45% 20% 26% Disbursements by RFIs (Rm.) 157.2 230.2 258.0 Of which: group loans 50% 53% 62% individual loans 50% 47% 38% Average loan size (R) 1935.0 2457.0 2232.0 Estimated jobs created a/ 244,431 190,146 215,315 267,686 Investment per job (R) 827 1070 964 Black business beneficiaries % 95 97 97 89 Retailers/service/contractors % 80 80 a/ Khula provides two different estimates of jobs created. These numbers are regarded as optimistic. However the alternative series does not have consistent assumptions. 55 3.61 How far did the RFI Program Assist the Microfinance sector? The program's end-date is indefinite, largely because few of the RFIs funded are yet sustainable. Initial goals were overambitious, with a five year target for RFI sustainability. Considerable turnover of experienced staff also occurred and portfolio management suffered. Despite efforts to weed out poorly managed and unviable RFIs, at least 20 of those assisted have subsequently closed, including 13 of the startups. RFI eligibility was based on its track record, including appropriate activity, adequate management, existence of an adequate business plan, and disadvantaged group ownership. 25% of applicants were rejected on the basis of weak management, inadequate business plans or inadequate rollout. Although a number of problem loans to RFIs were inherited from the DBSA and were not selected by Khula, the rejection rate seems to have been unduly low and selection criteria needed to have been applied more aggressively67. The mortality rate of startup RFIs was nearly double that of existing RFIs, although a few have managed to be successful. Currently 12 RFIs are actively supported, of which five are to be transferred to a new Micro Apex Fund, allowing Khula to concentrate on the more commercial segment of the market. 3.62 The average RFI is smaller than the minimum size for sustainability based on a norm of R30 million (US$5 million equivalent). Similarly the level of on-lending by the RFIs has been too low in most cases to be sustainable68. Only about five are considered to be actually or potentially financially sustainable. Initial goals were overambitious, turnover of experienced staff occurred and portfolio management suffered69. The RFI program is now being shifted out of the more concessional loans, and Khula is looking for successful private companies to `adopt' microfinance programs. Joint funds have been established with the Harmony Mining Group, Shoprite, and SAB-Miller's `kickstart' program. 3.63 Neither Khula as a whole, nor the RFI program, can operate on a commercial basis under present circumstances where it is expected to provide subsidized services. A commercial approach has also been made difficult, according to Khula's management, because donor funds destined for the RFI program were controlled by DTI without the specific performance conditions on the funds (e.g. five-year RFI transition periods) being made known adequately to Khula. It is clearly important for the integrity of the microfinance sector that quasi-market operating rules and agreements are established and adhered to, and therefore to the extent possible protocols should be adopted, whereby Khula can operate within a clear semi-commercial framework70. 67On this basis the targeting/ selectivity performance score was judged to be sub-marginal (see annex). 68While some RFIs are relatively large (for instance, Marang and SEF each disbursed over R50 million (US$ 8 million), in 2004-05 to a client base of 25,000), others, such as Ikusassa and Nicro, with disbursements of about R500,000 to only around 500 clients were less likely to be sustainable. 69Another difficulty faced in jumpstarting a viable microfinance sector has emerged in the `turnkey' RFIs set up under the Khulastart program (on the example of the UN Microstart). These often suffer from a lack of buy-in from owner- managers who do not get the experience of launching and building up enterprises. 70While Khula compensated the DBSA when it assumed DBSA's RFI portfolio in 1996, it is not certain (despite a formal agreement) that it will itself be compensated by the new Apex Fund for the adoption of part of Khula's own RFI portfolio. 56 3.64 Costs and Benefits of the RFI program. The public costs of the RFI program include part of the DTI capital injection and annual subvention to Khula, plus program management costs of about R3 million per annum. Against these public costs, the program needs to generate public benefits in terms of additional microfinance capacity and micro-entrepreneurship. Job creation from micro-loans is strong compared with the cost per job in the SMEDP program. Some limited spillover benefits from microfinance training have occurred, but some assisted RFIs have not shared know-how. Spillover benefits may accrue from the high number of individual loans on-lent by the RFIs. 3.65 An evaluation report in 1999, when the program was in its third year,71 listed a series of problems, some of which appear to be still applicable. These included lack of internal capacity and complex organizational structure; lack of flexibility in loan products (all had uniform interest rates and terms); delays in loan approval and excessive reporting obligations (especially in remote areas); processing delays of up to nine months; need to improve loan tracking, approval and disbursement processes; need to improve monitoring and evaluation. 3.66 For the RFI program, the key issue is to focus on a limited number of model RFIs, develop more flexible products and larger intermediary loan programs per selected RFI. This, in effect, is what the RFI program has moved towards, through reduction in number of RFIs funded, and transfer of its loans to the new Micro Apex Fund. The RFI program would then be left with about ten RFI clients who it would focus its efforts on. A summary table of qualitative performance scores, based on the system discussed above, is in Annex 7. (c) Ntsika Enterprise Promotion Agency - LBSCs 3.67 How did the design of the LBSC program meet its objectives? Ntsika was set up as one of the SMME `flagship' programs and as such is an important candidate for assessment, though it has now been merged with NAMAC72 into SEDA. Up to 2004, Ntsika provided non-financial and business development support through the Local Business Service Center Program (LBSC) and tender advice centers (TACs). Its assistance to SMMEs (under 200 workers) consisted of: (a) business development services (e.g. local business service centers); (b) tender procurement and business linkages support; (c) specialized support services (e.g. service provider network); (d) targeted support (e.g. award schemes) , and (e) program design, research and information (e.g. database compilation). In 2004, 61 staff were employed in the HQ including 16 management. 71`Supporting South Africa's Small, Micro and Medium enterprise strategy. Khula finance ltd and Ntsika Enterprise promotion Agency ­ Discussion points on the way forward". Report to the World Bank by ICC. 1999. 72 NAMAC was the umbrella body for 17 manufacturing advisory centers (MACs). It also operated a series of programs including: BRAIN, a SME referral and information network; FRAIN, a business franchise information, training and advice service; PROJECT, an advisory service for youth micro business; GAIN, assistance to local government in business development, and SEHD, empowering black business ownership. 57 3.68 The objective of the LBSC program was to support small firms through a network of accredited service providers, with special emphasis on disadvantaged groups. Each LSBC was supposed to provide a client-centered service appropriate to a local area including: business information, counseling, linkages and training services; integration of local support services (both financial and non-financial); promotion of community involvement in designing and managing SMME services; and promotion of equity funding. The program was supposed to be locally oriented, demand driven and participatory with local business, Government and non-profit organizations as stakeholders. 3.69 Each accredited service provider would get benefits such as publicity, access to information networks, training and funding, in return for providing quarterly reports to Ntsika and participating in workshops and conferences. The LBSCs, once accredited, would offer services based on 11 areas, or modules.73 To become accredited, an organization would need to have capacity to provide training, information and advice; be legally incorporated; have required staff skills and qualifications; be financial sound and potentially sustainable; and have a Board of Trustees and staff with adequate disadvantaged group representation and experience. 3.70 What was the LBSC's output? In principle, the design of the program, locally focused with strong local partnership, accredited independent support centers, and capacity to provide a range of key services, was consistent with good practice. There were several rationalizations of Ntsika activities in pursuit of improved effectiveness, including changes to accreditation requirements. In 1998, Ntsika took over 19 LBSCs from a program in Cape Town. The target from 1999 was to accredit 72 or more LBSCs a year over three years (265 in total), with an emphasis on community organizations and NGOs. According to an evaluation report,74 by 2000, Ntsika had accredited 10675 LBSCs, up from only 32 in 1999, mainly run by NGOs or community-based organizations and located around the four major cities. After 2000, however, a number of LBSCs ceased business or lost accreditation due to lack of capacity and funding, and the number of LBSCs declined to 81 by 2002.76 3.71 The aggregate Ntsika and LBSC network reported outcome in 2004 for main indicators was as follows. The LBSC program accounted for a good part of total Ntsika output. 73Business development, tender advice, technology transfer (`technopreneur'), local economic planning, rural extension, assistance to disadvantaged groups, pilot projects, ICT provision, and support to education institutions. 74Op cit. `Supporting South Africa's Small, Micro and Medium enterprise strategy. Khula Finance Ltd and Ntsika Enterprise Promotion Agency ­ Discussion points on the way forward". Report to the World Bank by ICC. 1999. 75Some of the data in this section is based on `A Review of the South African Local Business Service Center program' Robin Bloch and Stephen Daze. IDRC June 2000. 76Source: `Synthesis report on LBSC evaluation' Ntsika October 2004. 58 Table 3.10: Aggregate Ntsika and LBSC Network Reported Outcome in 2004 Indicator LBSC 2004 Ntsika 2004 Service providers supported 42 322 Service provider staff trained 152 191 SMEs referred/ counseled/ advised 5725 32307 Entrepreneurs trained 1896 8261 Jobs created/sustained 10836 27018 Source: Annual Report 2004 3.72 The total number of LBSCs in 2002 was considerably lower than the target of 265 in the 1999 LBSC program strategy. Other targets of the strategy, for new SMMEs assisted, jobs created, and entrepreneurs trained were also missed by a considerable margin. The effectiveness of the LBSCs was variable, with a number unable to provide the services expected of them. Funding was unreliable, especially from the Ntsika,77 and local funds had to be pursued, while some in urban neighborhoods had better funding and more capacity. A very small revenue was generated from training, consultations, assistance with tax and loan applications, and rentals. 3.73 Did Benefits Meet Costs? The overall expected public benefits of the LBSC program were (a) skills development and (b) increased potential for business development through enhanced entrepreneurship. Actual outcome numbers in terms of new businesses, entrepreneurs trained and jobs created seem to have been disappointing, both in terms of targets set, and amounts of funding available for mobilizing the expected additional activity. 3.74 The LBSC program cost about 55% of the total Ntsika budget over 2002-2004 and shared an overhead burden of about 25% of total program expenditure. The LBSCs had considerable difficulty in generating fee income especially from the smallest enterprises, although some existing firms yielded fee income. Funding for the Ntsika program as a whole, including the LBSCs, was almost entirely based on DTI and EU grants. 3.75 To justify the LBSC program, the entire public cost (up to R29.2 million in 2004) would need to be covered by the above-mentioned public benefits. The partial evidence on net impact does not suggest that this was achieved. Previous impact assessments addressed many aspects of the program at various stages. Many of the criticisms (e.g. a supply-driven service design and lack of coordination with other local players) were exactly the same as those that the original Ntsika mandate had specifically been expected to avoid. The main criticisms, applicable both to the Ntsika program as a whole and the LBSCs in particular, were: the need to avoid combining policy support and program management; reduction in the number of schemes; more focus on service delivery, supply-driven and generic; more even geographical distribution; reduction in reporting requirements and tightening up on timeliness of those to be provided; speed up in accreditation procedures; greater focus on issues on the ground; more stable and 77Ntsika provided on average only about 10% of funding. 59 predictable funding; the need to firm up links with local private business and local government; more appropriate management and marketing tools and materials, in local languages. 3.76 A table of the qualitative performance scores for the LBSC program based on the system discussed earlier is provided in Annex 7. (d) The Industrial Competitiveness and Job Creation Project 3.77 The Industrial Competitiveness project, supported by the Bank, consisted of the Competitiveness Fund (CF), the Sector Partnership Fund (SPF), and the Black Business Suppliers Development Program (BBSDP) which replaced an unsuccessful micro- enterprise fund called `bumblebee'. The original fund component comprised US$ 24.4 million in external funds with US$19.0 million equivalent in Government funds. The project was approved in 1997 and closed in 2004. The funds were designed to expand the role of SMMEs in the industrial sector by: (a) catalyzing the use of technological, marketing and productivity enhancing business development services, and (b) supporting a process of information sharing and fostering of networking among partnerships at the sectoral level. The Competitiveness Fund 3.78 How did output meet objectives? The purpose of the CF was to accelerate private enterprise development by stimulating the use of foreign and domestic business development services by private firms, accelerating the formation of skills, knowledge, information and contacts. Firms received a grant up to a maximum of R600, 000 for up to 50% of their eligible costs. The firm competitively selected local or foreign business service providers. The CF disbursed R140 million (US$ 17.3 million equivalent) to 1,066 firms over the period, from 1999 to 2004. Table 3.11: CF Grants Approved and Cancelled Year Number approved Number cancelled78 1999 254 - 2000 440 - 2001 266 - 2002 112 34 2003 162 166 2004 13 63 Totals 1247 263 78Cancellation refers both to projects which were completely cancelled and to projects where only partial funding was claimed from the DTI. Information on cancellations is only available from 2002. 60 3.79 A total of 984 CF grants (net of cancellations) were fully paid out. The cancellation rate was 21%, due mainly to underestimation of costs, failure to complete a project or withdrawal. About 8% of applications were rejected by the DTI, mainly due to unacceptability of quoted prices for services, poor business plans or ineligible activities. `Lack of additionality' accounted for only 2% of the rejections. 3.80 About half of grants were below R100,000 (US$15,000 approx). 25% were at the larger scale - R300,000 to R600,000 (US$90,000 approx). The grants were concentrated in the four largest metropolitan areas, namely Johannesburg and Pretoria (Gauteng), Cape Town (Western Cape) and Durban (KwaZulu Natal). The post-approval project cancellation rate was 21%, due mainly to underestimation of costs, failure to complete a project or withdrawal. About 8% of applications were rejected by the DTI, mainly due to unacceptability of quoted prices for services, poor business plans or ineligible activities. There was a comprehensive set of selection criteria, but without a systematic indicator of economic impact. `Lack of additionality' accounted for only 2% of the rejections. Approval time for applications and disbursement were reasonable. Interaction with project management was overall positive, with less than one in five negative opinions on any question. 3.81 Was there a net impact? The majority of the beneficiaries surveyed considered that the CF did benefit them in terms of productivity and competitiveness. A majority reported improved efficiency which they attributed to the CF program, increased market access and a range of other benefits; 82% perceived the program to have assisted them in becoming more profitable, while others saw quality improvements, time savings, and market expansion. However, only a minority saw a specific impact of the CF program on the share of exports in sales. The main changes could be summarized as follows: Table 3.12: Change in Performance of CF Beneficiaries 1999-2004 Capital Productivity 9.5% Capacity Utilization 1.2% Labor Productivity 5.7% Sales 2.9% Employment 0.1% Export Proportion of Turnover 3.3% 3.82 Quantitative results showed that, compared to a control group, beneficiary firms had modest sales growth but increasing capacity utilization, suggesting that capital assets were being utilized more efficiently, consistent with the high rate of growth of reported capital productivity. Employment, however, remained flat with quite strong labor productivity growth. There was reasonable growth in the export share of turnover, notwithstanding the less favorable qualitative responses. BDS service providers saw increased demand for their services, increase in their product/service range and improvements in effectiveness. They did not, however, think that the CF itself had a 61 significant effect on their market, attributing improvements instead largely to a general upswing within the sector. Consistent with this view, beneficiary firms were unwilling to pay full market value for BDS services ex post. 3.83 The spillover effects of the program were reported as good. Fifty seven percent of beneficiaries thought that the CF program led to horizontal spillovers, and three out of four that the CF program led to vertical spillovers. An indirect indicator of spillover effects is the extent to which the CF program served to create/strengthen networks within and between industries. In this context, almost 60% of beneficiaries reported that the CF program had strengthened or helped create such networks.79 3.84 Despite the positive outcome, the extent to which the project created additional activity was more limited. Many of the activities implemented by beneficiary firms with CF funding would have been carried out without CF assistance, but it did allow many firms to implement projects sooner or on a larger scale than they would have done independently. In this way, there was a clear, if modest, net impact. Table 3.13: Difference in Performance of Beneficiary and Control Group Firms Beneficiary Control Group Variables Growth Rate 1999 Growth Rate Difference - 2004 1999 ­ 2004 1999 ­ 2004 Avg. Sales Growth 2.9% 7.6 -4.7% Avg. Employment Growth 0.1% 4.3 -4.2% Capacity Utilization Growth 1.2% 1.5 -0.3% Labor Productivity Growth 5.7% -1.0 6.7% Capital Productivity Growth 9.5% -4.0% 13.5% Growth in export % of sales 3.3% -1.6 4.9% 3.85 The beneficiary companies performed better than a control group in labor productivity, capital productivity and growth in export shares, while the control group performed better in terms of turnover growth and employment growth, and capacity utilization.80 Beneficiary companies tended to invest more in labor-saving but also, apparently, used their investments in plant and equipment quite efficiently, increasing both labor and capital productivity. 79This result should be treated with caution as it is not clear how far the strict definition of spillover effects was applied in practical field interviews. 80Employment and labor productivity growth may be inversely related, especially in the short run. In the longer run labor productivity is the key objective since it is central to competitiveness and long term employment. The apparently anomalous result of falling labor productivity with rising gross sales per worker in the control group is explained by falling percent of value added in sales. This is also the explanation for the falling capital productivity. 62 3.86 The increase in productivity compared to the control group suggested that there was a positive net impact. Non-beneficiary companies, while showing better short term performance in job creation and sales, did not increase their productivity, implying longer-term problems for competitiveness. 3.87 Did benefits meet costs? The total cost of the CF program was about US$ 21 million over 1999 to 2004. It required a three-year extension for full disbursement, in part due to delays when management was transferred to the DTI after the end of the three year contract with the outside managers. The higher cost of the external managers in 1999-2002 was offset by relatively high output, with 77% of approvals being made during the first three years. After the DTI took over the program, costs per approval did fall, however, given that the external managers had the responsibility for launching the program and setting up systems and procedures, their output performance seems satisfactory. Management cost per approved grant value was 11% overall, and per disbursement value was 18%, within the good range according to international benchmarks discussed earlier. Table 3.14: Benefits versus Costs Efficiency measures Value Fund utilization rate 61% Project cancellation rate 38% Management cost per approval R 20,660.39 Management cost % of total approval value 11% Management cost % of disbursed value 18% Days from application to approval 30 days Days from receipt of claim to disbursement 45 days 3.88 The grant proportion of 50% was possibly higher than required, given that 66% of firms would have upgraded even without the CF. However, better selection criteria would eliminate firms (e.g. larger scale manufacturers) for whom the CF funding made little or no impact, and may justify retaining the 50% subsidy, which is on par with projects internationally. 3.89 Overall, on the basis of both qualitative and quantitative data, the project appeared to generate net benefits that reasonably justified the private costs of the beneficiary firms and the investments in it by society. The following summarizes the project's performance ratings: The Sector Partnership Fund 3.90 The objectives and output of the program. The SPF was a pilot vehicle to `kick- start' partnerships capable of providing network opportunities to member firms which would, over time, reinforce productivity gains through information sharing and 63 transactions. The aim was that successful sector partnerships (SPs) would promote commitment to build additional partnerships. Networks, or clusters, of enterprises could make collective investments that would benefit a group or an industry, but which might not be attractive enough for individual firms to invest in on their own account. About 870 individual firms received some degree of direct assistance through 100 upgrading programs in 96 SPs, plus indirectly the member firms of a number of industry associations. The maximum grant was 65% of the total cost for a partnership program, subject to a general limit of R1.0 million, the remainder being paid by the partnership member firms. The average grant approved was R867, 000, and the programs supported included a range of market promotion, cost reduction and productivity upgrading measures. 3.91 What was the impact of the SPF? The impact evaluation assessed a 10% sample of SPs: Table 3.15: SPF partnership sample Partnership and Location No of partner Grants (R Status firms million) Citi Bandwidth Barn (WC) Up to 30 2.0 Completed Deciduous fruits (WC) Association 0.70 Completed ECMAC Capacity building (EC) Up to 37 0.94 Completed Efficient Resource use (GAU) 7 0.11 Completed KZNMAC ISO (KZN) Up to 47 1.22 Completed Landsdowne Clothing (WC) 5 0.46 Cancelled Network Relationship Mgmt (GAU) 5 0.61 Completed Timber loading accreditation (KZN) Up to 30 0.70 Completed Ukukhula Refrigeration (GAU) 6 1.0 Cancelled Wine-of-the month club (WC) 5 0.94 Completed TOTAL 170 + assocn 8.68 (6.70 members disbursed) 3.92 The networks operated quite effectively; and those that did not did so mainly because of their limited size, relatively weak content, or lower importance to their members. The three strongest SPs gained from their relatively high importance to their members and the frequency and content of meetings, or because they were large, open and well organized. Economic impact criteria were not used for selection but otherwise eligibility compliance was high because firms adhered to the requirements of partnership size, openness, type of activity and ownership, and financial contribution from members. Qualitative ratings suggested that the SPF overall had a satisfactory productivity improvement outcome. Five out of ten SPs did satisfactorily or better on this dimension. Of the remaining five, two were cancelled, two others were less than satisfactory, and one has yet to generate an outcome but may well be successful. 3.93 With regard to economic benefits, vertical and horizontal spillover effects were relatively weak. The main reason for this is that partner firms were often not highly interlinked, but operated largely as independent recipients of the upgrading programs. The best result seemed to be from a Wine producing partnership because the network 64 acted jointly to integrate growers, packers and exporters. Horizontal spillovers were present, to some extent, with partnerships in timber transport and ICT. An increase in the ex-post willingness to pay suggests that barriers to the business service market might have been lowered but, with some of the larger firms, a subsidy was probably not needed and no market growth impact resulted. 3.94 The additional network impact was rather weak because, on balance, the project did not build new networks nor significantly strengthen those that were in place. About half of the partnerships have closed since the end of the project. An important reason may be that genuinely self-managed partnerships were a minority while several were initiated by outsiders, including service providers. Another reason is that the SPF funds were often not very significant in relation to the investment and turnover of the firms assisted. The networks would only add value if they identified special catalytic opportunities, and these were difficult to find. 3.95 The economic impact, in terms of changes in productivity of partner firms, was assessed in relation to a control group. While the achieved productivity outcome was reasonable (five partnerships performed above average and four below), the quantitative results for economic impact were poor. Sales and employment in SPF firms increased substantially but less than in the control group. Employment increased significantly faster than sales in both groups leading to falling labor productivity, however, this effect was worse for the SPF firms for which labor productivity fell more than in the control group. While sample size and possible bias must be taken into account, the poor results implied a lack of situations providing a clear collective investment opportunity that could not have been undertaken without the SPF. Table 3.16: Assisted Firm vs. Control Group Performance Assisted firms CF control group firms Sales 7% 12.0% Employment 14% 21.0% Value-added 4.2% 4.9% l Labor productivity -6.6% -1.0% Capital productivity -4.2% Na 3.96 The management costs per amount of funds disbursed were reasonable by international standards. Nevertheless, there were considerable delays in the approval and reimbursement processes, worse than for the CF. There was also a significant cancellation rate (though lower than the CF). The table indicates some main management efficiency measures: 65 Table 3.17: Main Management Efficiency Measures Efficiency Measures Value Comment Program Cancellation rate 18% 18 cancelled out of 100 programs 64% Utilization net of funds Utilization rate of approved grants cancelled R91,200 100 programs in 96 Management cost per program approved partnerships 22% Total disbursement R42 Management cost/disbursement value million Average time lag: approval to MOA 7.4 months For 55 programs in the database Average time lag: claim to reimbursement 90 days For 148 claims since April 2003 3.97 Although management costs were not high compared to some other similar projects, DTI's performance in implementing the program was evaluated at a little below satisfactory, mainly because of problems with applications and claims processing (partly resulting from rules imposed by the Treasury).81 3.98 Some key conclusions from the SPF evaluation were that: (a) it should re- integrate into a broader fund open to individual firms, which would allow greater flexibility in fund utilization; (b) there is need for increased initial management effort and strengthening of the governance; (c) improved project promotion is required, clarifying the role of the service providers and network facilitators; (d) selection criteria should include a simple economic impact assessment which assesses additionality; (e) the provision of greater management resources at the start may ultimately reduce overall costs because it is likely to reduce the life of the project; (f) it is essential to streamline applications, unify databases and records, improve follow up and verification, speed up reimbursement, and install a monitoring and evaluation system; and (g) prior agreements should be made with control agencies such as Treasury to anticipate operational hurdles. 3.99 Qualitative ratings on economic impact were better than the poor quantitative ratings, but the overall impact of the SPF was nevertheless sub-marginal. Public net benefits were lower than would be required to justify the public costs of the program, because of the high 65% subsidy, accompanied by not insignificant management costs. To justify replication there is a need to redesign procedures to improve the effectiveness of such project designs. The overall qualitative performance scores based on the rating system previously described are provided in Annex 7. 81One partnership asked the question "what is the point of an incentive program if it is so difficult to access that it ends up being a disincentive?". 66 The Black Business Suppliers Development Program Objectives and output 3.100 The BBSDP program was launched in 2003 to upgrade black-owned businesses and increase their access to business development services. The rationale was to compensate for their lower than expected achievement in procurement, the skills gap, inability to finance their own training costs and their under-representation in the CF program. The program supported enterprises of up to 20 employees with grants up to a maximum of R100,000 (about US$16,000). The program financed 80% and paid the service provider directly to avoid the need for pre-payment by the beneficiaries. 3.101 Over 2003-5, 672 out of 1,542 applications, and 60% of funds requested, were approved (a high rejection rate compared to the CF or SPF). Average disbursement per firm was R19,000 (US$3,200) with total about R12.8 million (US$2.1 million equivalent). About 86% of the firms applying for the program had 20 employees or fewer, and about 57%, 10 or fewer employees. One quarter had 5 employees or less. 52% of firms applying were in the services sector, 33% in manufacturing and construction, and 15% in retail trade. About 70% of beneficiaries were located in Gauteng, with Western Cape the next most frequent. The grant level averaged 79% of total costs, a relatively high public share of program costs by international experience. 3.102 Did the BBSDP have an economic impact? Those firms that received funding were more likely to engage in new business activities. 63% reported a business expansion in the timeframe after having received the grants. About 40% of participants reported that the activities implemented with BBSDP funding enabled them to participate in public and private tenders. The two most commonly cited activities were improved marketing strategy (63%) and expansion of business, including the hiring of new workers (55%). About half of the participants reported that they had upgraded an existing product line. One-third had introduced a new product line with BBSDP funding, while 16% had obtained an internationally recognized quality certification. Improved marketing and business expansion through hiring jumped by 7 to 8%. 3.103 The impact assessment was based on analysis of an assisted group (44) and a control group (25).82 Over 40% reported improvement in management systems, certification, planning, and control, and about 35% availed themselves of skills training for employees; roughly the same fraction pursued management training. Those that received funding were more likely to re-orient their business than those that did not. Some 40% of participants reported that the activities implemented with BBSDP funding enabled them to participate in public and private tenders. 82The samples were small and bias was partly allowed for by adjusting for size of firm. However, it is likely that biases remained since the assisted group was more formally established than the control group. 67 3.104 Performance indicators used in this evaluation were different from those for the CF and SPF.83 Taking account of small-sample bias, employment growth rates in the assisted firms showed the clearest evidence that BBSDP participants outperformed the control group. The median assisted firm experienced employment growth of 0%, 20%, and 47% respectively in 2002, 2003, and 2004. The median firm in the control group experienced no employment growth in any year. Sales growth in assisted enterprises was slightly more than that of the control group, while profitability was not better (in 2004, firms in both groups reported net profit rates near 20%). Recipient firms invested more than firms in the control group, especially in machinery and equipment per worker. On the whole, the investment results were better for the beneficiary firms and improvement on multiple indicators was steady. 3.105 The processing time for grants was fairly long. Nevertheless there was generally high reported satisfaction with the program and with the project facilitators. Less than one fifth of program participants rated the program or facilitators poor, and 91% of the recipients indicated that they would re-apply. Table 3.18: Processing Time for Grants Processing days Required for Mean High Application to Approval 76.8 366 Approval to First Payment 67.5 232 First payment to Final 34.6 228 3.106 Available evidence shows that there was a reasonable degree of additional output activity generated by the program, compared to what would have happened without the grant. Evidence is not available for possible spillover benefits, BDS market creation nor explicitly on entrepreneurship, which would justify public outlays on such a program. The approval of nearly 700 programs serviced by BDS providers over a two year period may however have been a good `platform' for spillover effects. Overall the BBSDP probably generated an acceptable economic rate of return. Qualitative performance scores based on the ratings given previously are provided in Annex 7. D. COSTS AND BENEFITS FROM THE USERS' VIEW ­ RESULTS OF BENEFICIARY SURVEYS 3.107 The beneficiary survey, as described in section A, was a bottom-up look at Government support programs for SMMEs from the point of view of the firms themselves. The summary findings (on those survey questions susceptible to yes/no responses) were as follows in percentage of responses. 83This study did not measure productivity indicators but used profitability (margin as % of sales). 68 Table 3.19: Beneficiary Survey ­ Selected Responses Yes/good Acceptable No/poor Zero/not known % % % % Does the Government's strategy 13 43 17 27 help small firms? How far do firms have knowledge 30 18 52 0 of DTI support? How did firms interact with their 25 54 21 0 facilitator/consultant? How well did the firms interact 12 18 41 29 (processed via with DTI facilitator) Is the publicity /promotion of the 0 23 64 13 Government's programs adequate? How effective were the projects 57 29 14 0 undertaken (by project)? How good was program 43 57 0 0 management? (by program) Is the Government's strategy helping small firms? 3.108 Only three of the 23 firms had a clear positive opinion. The most emphatic support came from a firm that had participated in the greatest number of projects (including eight SPII grants). Four firms had a negative opinion. Among specific negative replies were one firm that thought the programs should be better targeted, another that felt that Government programs have inherent problems of inaccessibility and delay because of the need to ensure control, and a third that pointed to excessive fragmentation of programs, each imposing different rules. The majority of firms had only a rather vague idea of the Government's strategy and whether it was going in the right direction. In general, concerns expressed were about accessibility and bureaucracy. Two firms, however, considered that the Government should be focusing on wider issues, another that assistance should be better targeted at small to medium firms to generate employment (rather than focus on micro-businesses) and a third that interest rate subsidies were justified because they offset excessive risk premiums placed by the market. How far do firms have any knowledge of Government/DTI support? 3.109 Overall response was poor. Just over half the firms had zero to slight knowledge of Government programs prior to application. Less than one third had good knowledge. Even among applicants, only 6 out of 16 had good knowledge, often through intermediaries. There was no clear pattern whereby smaller firms had less knowledge. In some cases, small firms were more isolated but, on the other hand, larger firms were sometimes less interested in knowing about assistance. Of the non-participants, six out of seven (85%) had zero or minimal knowledge of schemes. One service firm had good knowledge but did not need to participate. 69 Table 3.20: Knowledge of Government/DTI support Total Firms Size:<50 Size:>50 Participants Non- workers workers participants Zero/minimal 4 2 2 2 2 Slight/general 8 6 2 4 4 Reasonable 4 3 1 4 0 Good/extensive 7 3 4 6 1 Total 23 14 9 16 7 3.110 Those that had the best knowledge had made a special effort to find out and make use of Government grants, including converting themselves into BEE firms to have a better chance. There were diverse experiences, ranging from a firm based in Germiston, manufacturing teflon and polypropylene material with 160 workers which had, since its foundation in 1969, participated in the SPII eight times (filing two patents), the CF two times, and the EMIA two times, and had considered applying for the SMEDP; another, based in Johannesburg, which had been rejected for the CF and the NEF and had had a poor experience with the BBSDP, and a third firm, started in 1996, which had no knowledge of any Government programs at all but was interested in applying. 3.111 The MICA survey of chapter two also found a low level of knowledge about any Government programs, the SMEDP being the best known, but even then only by 24 % of managers. Larger micro-enterprises were slightly more knowledgeable than the smallest and urban firms were better informed, implying that outreach is worse outside the big city zones. How did firms find out about Government/DTI programs? 3.112 Several firms cited more than one source of information; the most common being through a business contact, client, competing company or supplier. 19 out of 32 sources cited were private business sources. No respondents cited TV or radio as a principal source, and only seven in all cited the press, official publication and mailing, or advertisement. Of the six times Government was cited, two firms themselves took the initiative by visiting the DTI to get information, and in one case a firm had a personal contact within DTI. The clear implication is that the Government is relying on word-of- mouth between private firms to promote programs. This would not necessarily be an issue per se, but for the fact that it is not doing an adequate job of making its programs accessible. 3.113 Firms were, in general, not pressured to join programs, which implies that collusion was low. Of the 23 respondents, only two suggested any type of pressure. Such behavior would work against market determination of demand, and might give a distorted idea of the appropriate level of subsidy. The responses from the MICA survey (chapter two) were slightly different, reflecting the fact that the MICA group only consisted of micro-enterprises, whereas this survey looked at SMMEs. 70 Figure 3.1: InformationSources 12 10 etiC 8 se 6 m Ti 4 2 0 ,Radio vertising t/article sh ultan ntac TV blication/brochu iling/emailing Ad Officialpu Ma Presscommenvernmentvisit/roadFacilitator/consGe nBusinessCo Go What further assistance would the firms apply for? 3.114 Respondent firms were asked whether they would apply or reapply for a program. All but one of 23 firms said that they were in principle interested in applying or re- applying, despite the problems faced. The most commonly cited program of interest (firms prioritized up to three) was export marketing (26%). Quality certification is also often directly linked to the objective of increased domestic and exports sales. Overall, marketing and sales accounted for 40% of programs in demand, but training did not figure as prominently as expected, accounting for only 21% of citations. The MICA survey (chapter two) also found that a large proportion were interested in further assistance. 71 Figure 3.2: TypeofProgramDemanded 12 s 10 poner 8 6 of.oN 4 2 0 training ttrainingevelopm e eting ribution port me system ification Employeeanagemenmarketd od M Export GeneralmarkSales/distTec hnologysupuctdevelopounting Innovation/prBu siness/acc Qu alitycertcontractualservice Legal/ 3.115 Despite displaying interest, few micro-enterprises surveyed had actually applied to any program, usually because of lack of necessary connections, lack of response to initial enquiry and difficulty with the application process. Only about one-fifth said that they did not need support in areas offered by programs, 18% said co-payments were too high and 10% said that they did not qualify. Firms that got information from official sources had the same responses, but the access problems seemed less severe. How did firms interact with their facilitator/consultant? 3.116 Of 13 firms that used facilitators (11 participants and two non-participants who were rejected), there was about an even balance between those who thought that the facilitators had done a good or essential job and those that did not. Figure 3.3: Firms interaction with facilitator 6 5 4 3 2 1 0 Good Acceptable Poor 72 3.117 Several firms remarked that without the facilitator they would not have been able to participate, usually due to lack of knowledge or lack of time to complete an adequate application. Of those that did not use facilitators two wanted to do it themselves because they felt that it was an appropriate `discipline'. Those that thought the facilitators had done a poor job included two firms where applications had been unsuccessful. Box 3.1 A bedding firm in Rosslyn started in 1995 and employs 12 workers. It is a black-owned business. The firm thinks the Government can help small firms but it needs to take closer control to protect firms from abuse. In 2003 as a result of an approach by a facilitator it applied to the BBSDP for a grant of R100,000 for the development of a business plan for plant expansion to be used to raise finance through the banking system, IDC and elsewhere. The firm applied through the facilitator and funding was approved. However, the firm's proprietor had no knowledge of the process as it was being handled by a `trusted staff member'. The staff member signed off that it was completed, but it was then discovered that the work was unsatisfactory, by which time the firm was no longer contactable and the staff member had left. On contacting the DTI the proprietor was told that since the claim had been signed off there was nothing that the DTI could. The proprietor described the experience as like taking a `second MBA'. The quality of the business plan was poor and it had to be redone. Management thinks it was `taken for a ride' as the cost of the services provided were also well above the market rate as a result of collusion with the service provider. 3.118 An important issue is: why do so many firms need facilitators, and should application processes not be designed so that the beneficiary firms can take more responsibility for their applications? The problem with using intermediaries is that it potentially leads to conflicts of interest or confusion of objectives. How well did the firms interact with DTI? 3.119 The question about interaction with DTI complements the one on knowledge of Government programs. Of the 17 firms that responded (two non-participants and 15 participants) only two were satisfied with the interaction with DTI. One, a chemical products firm, had participated in many programs since its formation in 1969, and another, a mining equipment firm, had a generally satisfactory experience with DTI programs. In two cases firms made a special effort to visit the DTI to find out what programs were available. One firm was very unhappy, and faced financing problems, as a result of being rejected ten months after applying, because of a rules change for the SMEDP. Another heating equipment firm had its claims largely disallowed on what it regarded as a technicality. Yet another reported that most program officers `never' return calls, and that it found only one person in DTI who was responsive to questions. Intermediaries also complained, especially about the application process. 73 Box 3.2 A firm producing fire extinguisher medium in Walkerville started in 1999 and employs 28 workers. It received funds from the CF, SMEDP, SPII and Khula Guarantee. It applied to the NEF but withdrew because it could not find suitable outside shareholders. It applied several times to EMIA but did not receive a response and to Gauteng Enterprise Propeller, but was not eligible. The firm's extensive knowledge of Government programs resulted from spending several weeks at DTI offices to identify officials who could explain Government programs. Apart from these meetings the firm has not received any DTI information, and had difficulty accessing though the DTI call center. The manager handled all applications including visits to DTI to sort out problems such as mislaying of documents. In the case of the CF the firm felt that the critical need was for independent fund management who would market programs. The time required from claims to reimbursement was fastest in the SPII (one month), two months in the case of the CF, and an average of three months for the SMEDP (at the start of the program). The Khula application was delayed by the applicant's own bank. Internal controls and information systems seemed to be good for the CF, but deteriorated. For the SMEDP they were acceptable at the start but also deteriorated. For the SPII they appeared satisfactory, while Khula had recovered from a bad patch. Is the publicity /promotion of the Government's programs adequate? 3.120 The low level of knowledge of Government programs would suggest that promotion is not adequate. The response to specific questions about promotion and publicity were as follows: Figure 3.4: PublicityandPromotionofPrograms 16 14 12 10 8 6 4 2 0 Yes Acceptable Poor Very poor 3.121 The view of DTI's promotional efforts was reasonable in only 5 cases, with all other respondents recording poor or very poor performance. Not one thought that promotional effort was good. This confirms the previous findings on knowledge of Government programs and sources of information on Government programs. The problems cited included poor access to DTI officials; inadequate processing capacity causing long queues; instability of rules; non-targeted promotional material (i.e. using generic DTI ads); and, far too small coverage (e.g. from the often single `road show'). 74 Box 3.3 A software development firm in Edenvale started in 1998 and employs 15. It is seeking BEE status and has brought in a black-owned firm to buy up a share of equity. The firm received funding of R600,000 out of a total upgrading program of R1.2 million from the CF in 1999 (at the start of the program). Management considered the technical service provision to have been highly effective, as a catalyst for a major Government contract to design and manage the national tax e-filing system, in collaboration with a major international consultant. Regarding program promotion much more could be done in the area of procurement preferment on Government contracts, tighter disclosure rules to force larger enterprises to set up genuine BEE companies. Promotion could be done through Registrar of Companies, Chambers of Commerce, and CFA offices. Success stories should be publicized. Regarding program management, the detail requirements were good discipline for a small firm. Documentation was satisfactory; the application format was a template for the required business plan. Time from application to approval was four to five months but the firm thought this was acceptable. Time from claim to reimbursement was two to three months in the first case, and about one month thereafter. The firm wishes to access the SPII program with a further software innovation for which SARS would require `proof of concept'. It also wants further export market support. Finance is however not a binding constraint and the firm has access to bank overdraft funds. 3.122 The key issue is promotional cost-effectiveness, or maximum coverage per Rand expended. Most firms had suggestions on improvement, for example that the DTI make use of the extensive company databases available through the Registry of Companies, or possibly through the Tax Service, to send out newsletters, mail-shots or email distributions to inform potential applicants about DTI programs on a regular basis. Other proposals included advertising in trade magazines (e.g. Engineering News). Much more could also be done with the better publication of preferred procurement information. One respondent considered that independent fund management was critical to good marketing of programs. How effective were the projects undertaken? 3.123 The quality of service provided was rated in 13 out of 23 projects as good to very good (scoring 4 and 5). Firms may have been motivated to be positive simply because they received a grant that increased their cash flow, and because they wanted further assistance. However, in a few cases (in the CF and SPII projects) the firms were emphatic that the program was `essential' to their business. In one case the services provided were catalytic in the winning of a major Government contract to supply systems software. 3.124 At the other end of the scale, there were two firms (producers of PVC products, and bedding) who were very unhappy with the services provided (under the BBSDP), and thought that they had been `taken for a ride'. The untypical score in the table of `1' for the CF was the result of a claim being largely disallowed because of confusion over 75 timing. Three firms stated (with respect to the SMEDP) that the cash did not change their investment plans but was a `windfall'.84 Table 3.21: How Effective were the Projects Undertaken? Ratings (0-5) for service quality of programs in which firms participated. (Total sums to more than 16 because four firms participated in more than one program. Average participation rate per participating firm is 1.45) CF BBSDP SMEDP SPII (3) Khula (1) SSP (1) EMIA (4) (6) (4) (4) First rating 5 3 4 4 3 4 4 Second 5 4 2 5 3 rating Third rating 4 0 3 4 2 Fourth rating 1 4 3 Fifth rating 5 Sixth rating 4 Mean rating 4 2.3 3.3 4.3 3.0 4.0 3.0 3.125 However these results are interpreted, if 13 out of 23 firms regard the services provided as good to very good then this clearly does reflect the fact that the Government's programs have elicited good performance from the local BDS sector, whether or not they created an economic impact. How good was program management? 3.126 The following are the ratings on efficacy of management for the programs participated in by 16 firms: Table 3.22: How Good was Program Management? Ratings (0-5) for management of programs in which firms participated. (Sums to more than 16 participants as four firms participated in two or more programs (Avg. 1.45) CF (6) BBSDP (4) SMEDP SPII (3) Khula (1) SSP (1) EMIA (4) (4) First rating 4 3 3 4 3 3 4 Second rating 4 3 2 4 3 Third rating 5 3 3 3 3 Fourth rating 2 2 2 3 Fifth rating 4 Sixth rating 4 Mean rating 3.8 2.8 2.5 3.7 3 3 3.3 3.127 Bearing in mind as usual the very small samples, the program management ratings overall are about acceptable to good, a result that seems better in some cases than those implied in responses to previous questions. The weakest rating was for the SMEDP, and the strongest the CF. Weakest areas generally were response times, publicity and 84The additionality issue with respect to the SMEDP seems to be particularly critical. In the 2004 impact study (see below), only 34% of firms stated that the grants had `led to an investment', and a far lower percent had said so in the case of larger enterprises. 76 promotion, and application procedures. The highest score (5 for the CF) was given by a firm who approved of the lengthy processing of a project because they thought that it was educational for small firms. Box 3.4 An engineering firm in Pretoria making catalytic converters started up in 1997 and employs 45 workers. The firm applied to the SMEDP for a grant of R540,000 over three years, which was 30% of the total value of the firm's investment program. Following the application and after a wait of about 10 months including follow-up requests the application was denied. This was as a result of the change in the eligibility rules to require evidence of operating profits. An objection was lodged but was unsuccessful. This was a case when the firm relied on the grant to complete an investment package and faced serious cash flow problems as a result of denial of the subsidy. However a bank agreed to accommodate the cash shortfall after lengthy negotiation. The firm thinks the Government should be more transparent about program details, streamline procedures, reduce waiting times, and maintain stable rules. 3.128 Processing efficiency is reflected in response times. The average times for each of the programs accessed by the sixteen participating firms were: Table 3.23: Processing Efficiency Program CF BBSDP SMEDP SPII Khula SSP EMIA (6) (4) 4) (3) (1) (1) (4) Application to 2.8 3.3 5.4 3.7 3.0 2.0 1.3 approval (months) Claim to 2.5 2.5 9.3 2.3 - 2.0 1.3 reimbursement (months) 3.129 The worst performance on processing time was under the SMEDP where, as reported earlier, there has been little control of project applications relative to capacity (one application was rejected after a ten- month wait). With SMEDP claims, the first took the longest and follow-on claims (once the firm's data were in the information system) were usually faster. The best performance among programs with three or more projects reported was EMIA. The next best was the CF. The SPII applications take longer than others because of the relatively rigorous inspection and evaluation process to verify the innovative nature of project proposals. Nevertheless, the SPII processing time was similar to the BBSDP in these few examples. Beyond the SMEDP, application and claims times are overall moderately slow. General response times for DTI enquiries were poor (5), acceptable ((7) and good (2). 77 Firm's willingness to pay for services. 3.130 One of the most important issues relating to the impact of these programs is how far firms are willing to pay for services at market rates, either before or after a project is conducted. If a firm is willing and able to pay market price for the services it requires then there is no public purpose served by subsidizing it. As has been mentioned in the case of the SMEDP, in some cases the support was provided ex post such that, by definition, firms did not actually include it as part of their investment programs. They had already paid market price for financing and the grant was just a windfall cash injection coming in later. This is the clearest case of lack of public purpose (lack of externality). Box 3.5 A pharmaceutical firm in Johannesburg producing mosquito repellent started up in 1996 and employs 20 workers. The management thought Government programs could potentially help small firms but were too slow (especially on reimbursement) and needed to be better organized and promoted (e.g. through mail shots). The firm received from the SMEDP R250,000 per annum (10% of its total investment requirement) or R0.75 million over three years. It has submitted a further application for funding in 2005. However the funding was applied for after the new equipment had already been purchased and installed. In the view of management the funds did not assist startup but constituted a refund on their completed investment which increased their working capital. They may also not meet the SMEDP requirement for a minimum labor cost ratio by year three, in which case they will also have received two years of funding without achieving the employment objective of the program. The assistance was `welcome' but not critical and they did not think that it properly met its public interest goals. Management would like to obtain export marketing assistance but it did not have a financial constraint and could probably pay market price for such programs. 3.131 Other cases are less clear. In two cases an SMEDP grant was an essential component of an investment plan which had to be abandoned or delayed when the grant was not forthcoming. Sometimes an outside intervention can be catalytic even for a well- financed firm. For example, in the case of technical assistance, the issue may not be the financing as much as information and confidence-building. Thus, two firms spoke of the importance of the State sharing the risk when a firm enters a new area (typically an export market) just to lower the information cost barrier to entry. Others spoke of the technical assistance coming at a critical point and jump-starting an upgrading process which otherwise might not have been started, or would have been delayed. 3.132 For the 23 firms (including the firms that had not participated in Government schemes), about five stated that they would have paid, or would pay, 100% of the market price for similar technical assistance. About 13 would generally be prepared to pay 80 to 100% of market price of similar technical assistance. Five would be prepared to pay above 50%. In some cases, a firm would prioritize its investment and be ready to pay for critical inputs while it looked for some support for less-critical inputs. 3.133 These responses suggests that for second round and following rounds of technical assistance support, the percent subsidy could be reduced to line up better with the value of the public benefit ­ e.g. from 50% (CF) and 80% (BBSDP) to around 20 to 30 % after the first injection of assistance. Many grant schemes do, in fact, phase out the subsidy through successive rounds. Alternatively, it might be justified to maintain the level of 78 subsidy but tighten up the selection criteria to focus on those firms that pass eligibility tests but are least able to pay (i.e. more precise targeting of the public benefit). In the case of cash investment grants (SMEDP) the latter approach seems to be important. 3.134 The firms were asked whether their principal constraint in the short to medium term was (a) financing or (b) technical capacity. 16 firms stated that it was technical capacity, four needed primarily finance, and three needed both. Thus, the issue was not so much one of inability to find financial resources, but rather lack of knowledge of, or confidence in, technical support services. E. SUMMARY OF THE PROGRAM EFFECTIVENESS ISSUES 3.135 Overall, the firm's responses indicated clearly that there are major areas in which the functioning of Government support programs needs to be improved, and these responses correspond generally with the `top-down' assessments of individual programs in section. The principal issues that need addressing from the top and the bottom are summarized as follows. Lack of Knowledge of Government Small Firm Strategy 3.136 Only a small proportion of firms were aware of the Government's policies for small firm support. While a few firms supported current initiatives, the majority either could not comment substantively or were critical. Among critical views were that Government programs inherently could not overcome the problems of accessibility and delay caused by need for control of public funds. Another was that there was too much fragmentation which risked delaying rather than accelerating development. Some considered that the Government should be focusing on wider issues such as redefining small firms, increasing income and capital gains tax incentives and easing labor regulations. Ineffective Promotion and Publicity 3.137 Far too few firms are aware of the support programs. Most learn about them from business contacts rather than from the DTI or Government sources. Thus, intermediaries and facilitators often take a key role in introducing firms to the opportunities available, with potential for collusion. The DTI needs to maximize coverage per Rand invested in publicity. One approach suggested is to use the DTI databases such as the Company Registry to do mail-shots and e-mail-shots in the form of newsletters setting out the incentives available. 3.138 In some cases, such as the SMEDP, despite inadequate promotion, the programs have received applications well above their processing capacity. In the SMEDP this is because of the attractiveness of the grant and the number of inappropriate applications. In other cases, such as the CF, disbursements lagged several years behind the original project timetable, and better publicity could probably have remedied this. This is discussed further in Chapter 5. 79 Operational Inefficiencies 3.139 Only a minority of firms expressed general satisfaction with the way programs are run. In some programs the selection procedure is either, on the one hand, over-complex or, on the other, not sufficiently rigorous. On average the application and claims process takes an excessively long time in most of the programs studied. The old established programs such as EMIA tend to be better, while at the other extreme, the SMEDP accelerated approvals but then faced severe processing backlogs. The complication of the application process results in intermediary firms taking charge, which is justifiable where they add value but also risks conflict of interest. Operational information systems and databases are of variable efficacy. In some cases filing is largely manual. Where electronic filing is used there are some problems of lack of harmonization and lack of ability to produce required reports in a timely way. Program evaluations differ but, overall, there is a sense of lack of responsiveness and difficulty accessing program officials, while, in turn, other officials are overloaded. Project Outcome/Effectiveness 3.140 Project output, in terms of the quality and output of technical services provided by suppliers through the grants, was a strong point of the programs studied. Most (though not all) responding firms were satisfied with these services. Some said that they had been indispensable. In one case the technical input provided a catalyst to winning a series of large contracts. In a number of other cases the services provided were rated as `essential'. In a minority of projects recipients were dissatisfied with services provision. Program Impact and Selection Criteria 3.141 In many cases programs did not create additionality. This was the case with some SMEDP grants which were received as `windfalls'. In the case of technical advisory services many firms that received export support were already engaged in exporting and might have been able and willing to pay the market price. Where a public subsidy is used merely as a cash source it is not necessarily serving a public purpose. Selection criteria in most programs need to be fine-tuned to target firms that need subsidies to deliver public benefits. Examples are the firms needing a `jump-start' for risky activities (e.g. opening new export lines), lowering information cost barriers to entry into new activities (e.g. developing innovative software products), offsetting costs that might be unrelated to business activity or risk (e.g. the effect of country risk on interest rates), or providing additional employment to marginalized black workers. Pricing/Public Benefit Targeting 3.142 Most firms seemed to be prepared to pay close to market price for many services, which suggests that subsidy rates (percents) could be reduced from current levels, especially in follow-up rounds of support. Alternatively, subsidies should be better targeted to firms that are clearly eligible for grants (on the appropriate criteria) but least able to pay ­ thereby more effectively generating a public benefit, such as in the case of 80 the SMEDP where some grants have created no additional output, employment, investment or spin-off benefits. Reapplications and Future Needs 3.143 Nearly all the firms surveyed would, in principle, apply for programs in the future, on the assumption that the processing problems can be sorted out. Only a small minority of firms stated that they only needed cash subsidies for investment finance; most responded that their need was for risk-sharing and support in business upgrading. The main requirement is for assistance in the export marketing area. The second most common requirement is for employee training. A third area of need, that is linked to exports, is quality certification. F. THE PATTERN OF GOVERNMENT ASSISTANCE TO SMMES 3.144 The above sections have looked at the effectiveness of assistance programs, individually and in comparison with one another. In this section we look at the overall program assistance strategy of the Government. 3.145 A difficulty in designing the overall program is the conflicting triple objectives of economic efficiency (competitiveness), welfare (poverty alleviation) and political (black) empowerment, and of the appropriate allocation of funding and effort between these objectives. The objectives of simultaneously pursuing economic efficiency and say unskilled labor-intensive investment and micro-enterprise development, along with the dilution of European ownership and management, are subject to considerable tension, especially for economic activities under severe competitive pressure (for example the textile industry as discussed in chapter four). Clear guidelines for resource allocation and a hierarchy of importance are absent, and the SMME strategy has struggled to meet this range of different and sometimes conflicting objectives. 3.146 The efficiency objective was the main concern of several of the selected programs ­ the CF, SPF, SPII, and EMIA. There are a number of other DTI programs (listed in Annex 9) such as the SIP, FIG, IDZ, CIP and CSP, which are designed to address competitiveness and productivity in general. The MAC network also falls into this category. The SMEDP would also qualify under this heading. 3.147 The welfare objective is addressed by programs not covered in our review, such as the Usubomvu Youth Fund, or other heavily supported programs such as the upcoming Micro Apex fund which will deal with the smallest, survivalist, enterprises. Other examples which have a welfare orientation are Rural Community Support under the NEF, technology for women (TWIB) and rural projects under the CPPP (now under SEDA). Within the MAC programs (now managed by SEDA), there is a Youth Pilot and the SEHD for disadvantaged community programs which comes under NAMAC. 81 3.148 The political empowerment objective overlaps with the welfare objective. The BBDSP fell directly within this category. In the case of Khula Finance, the RFI and other programs have funded predominantly black-owned business (over 90%), albeit in urban areas. All the NEF's programs (generator, accelerator and transformer finance) are aimed specifically at black-owned businesses, as are the local initiatives such as GEP. The GODISA program has targeted mainly survivalist micro enterprises in mining and floriculture, which are almost all black-owned. More recently the SPII has been revamped to include an innovation grant program aimed at BEE firms. Several other programs not assessed here (listed in Annex 9) have also addressed black empowerment, including in the formal sector the Workplace Challenge Fund (DTI/NPI) and Skills Support Programs, and in the informal sector the Usubomvu Youth Fund which is directed at black youth, set up by the Ministry of Labor. 3.149 The following table shows two principal public costs of the programs, as another indicator of revealed assistance strategy. Table 3.24: Summary of Assessed Public Support Schemes Program Avg. Annual Coverage of 2003-5 Avg. Annual Commitment Expenditure Management Cost per Funds 2003-5 (U$mn) Committed % SMEDP 220.0 Operations, grants 1% to 2% EMIA Pavilions 7.0 Operations, 6.5% programs Khula (RFIs) 8.0 a/ Operations, 20% program subsidy NEF 6.0 Operating subsidy Approx 50% only (at startup) SPII 8.5 Operations, grants 14% Ntsika (LBSC) 4.5 Operations, Na program subsidy GODISA Softstart 0.8 Operations Na GEP 7.0 Operations, 21% (projected) programs CF 6.0 operations and 21% grants SPF 1.5 operations and 22% grants BBSDP 1.2 operations and 14% grants a/ Includes both prorated annual subvention and a prorated portion of equity inputs 3.150 The first issue is the large disproportion in funding shown in the table. While the LBSC program was initially regarded as a `flagship' assistance program, supporting the key micro and small (black-owned) business sector, it was poorly funded. In contrast, the SMEDP, which has fifty times the annual State funding of the LBSC, was in fact directed at the formal sector, largely European-owned businesses. An amount equivalent to 50% of projected SMEDP commitments (about R1.0 billion (US$150 million) a year) which is currently not going to small firms (i.e. those with under R5 million in qualifying assets) 82 could in principle be diverted into a broad-based micro/small scale firm promotional effort, especially for skill generation. It is noteworthy also that the World Bank- supported programs, designed to introduce innovative instruments to generate business know-how and skills, were among the smallest in terms of their annual commitment of funds. 3.151 The programs also show very variable management costs, ranging from 1or 2% up to over 20% as a percent of program expenditure (excluding NEF which was just starting up). The implications of this are not simple since as the benchmarking discussion shows, different schemes regularly incur different levels of management cost. Furthermore, without evaluating economic benefits it is not possible to make definitive judgments about cost levels. However, the wide range of management costs per amount committed needs to be thought through and might involve some misallocation of effort. 3.152 SMME support at the Central Government level has been concentrated through the DTI in order to try to rationalize and simplify the structure. Central support has had to reach down into the local level, with limited success, as in the case of the Ntsika LBSCs and GODISA incubators. Locally controlled and resourced support programs like the Gauteng GEP might be better able to structure targeted assistance than centrally directed programs. Provincial and regional governments have initiated development programs, as in Gauteng, which through its Blue IQ initiative created the Gauteng automotive cluster, the Automotive Industry Development Centre (AIDC) and the Automotive Supplier Park, or the ICT initiative in the Western Cape Province, with incubators assisting micro and small companies. Some of these initiatives were supported through the Bank's SPF. Their success is still to be established. G. IMPROVING THE SMME SUPPORT INFRASTRUCTURE The Key Issues 3.153 The Government is revamping support institutions, e.g. through the recent merger of Ntsika and NAMAC into the Small Enterprise Development Agency (SEDA). The organization of the institutional structure is extremely important to improve the alignment of capacity with objectives, sharpen focus, ensure stable funding and prevent wasteful duplication. Enhanced monitoring and evaluation capacity in the DTI and other MSME support institutions (e.g. SEDA) is also needed, failing which service design and resource allocation decisions are likely to be sub-optimal. The central issue is how to organize the overall support framework and how to design and operate public support programs to achieve their objectives cost-effectively. 3.154 On the design side this means adequate management capacity in relation to workflow, appropriate selection criteria, efficient application and verification processes, good record keeping, effective disbursement procedures, and good monitoring and evaluation systems to determine impact. 83 3.155 There are a number of key questions. · Is a solution to the improvement of program management to spin it off from Government to independent fund managers? · What role should Government take if such a spinning off were to occur? · Should Government focus on leveraging the capacity of large private firms to assist small firms, or, private initiatives to help micro-businesses ­ in effect using the first economy to boost the second economy? · If so, should government incentives aim at larger firms to compensate them for the risks (e.g. poor quality or late delivery) of accepting a small firm into their supply chain, or at smaller firms to make the necessary adjustments to strengthen their skills and competitive efficiency? Consolidation and Rationalization 3.156 It is not easy to identify the appropriate program structure without a clear understanding of the relative cost-effectiveness of each type of program. It is equally not obvious that programs are too numerous or overlapping without detailed knowledge of each. While there are transaction cost savings from merging programs equally it is important that specialized assistance capacity is not diluted within a general program. Indeed, one of the common complaints from field interviews about DTI promotional efforts is precisely that they are not sufficiently targeted at the requirements of specific programs. 3.157 However, the surveys done for this report and the conclusions of previous work by South African agencies do strongly suggest that duplication is a problem, and that compartmentalization of assistance efforts into numerous program areas creates serious bureaucratic challenges both for the providers and for the recipients, which reduce the value of the assistance. 3.158 The Government has taken rationalization steps with the creation of SEDA in 2004, incorporating NAMAC, Ntsika and CPPP, but at the same time it is also developing new programs, for business process outsourcing, film development, and the micro apex fund. The DTI is increasingly being treated as the main, single-channel agency for assistance from the center. 3.159 One of the recommendations from evaluations of the three Bank­assisted funds under the Industrial Competitiveness project was to also re-integrate such funds in order to reap economies and reduce management transaction costs. The CF and SPF programs would be better re-integrated into a larger more flexible fund with goals clearly differentiated from those of other Government programs, emphasizing the technological spillover and BDS market creation objectives rather than investment or employment per 84 se. The more specialized cluster-building aims of programs like the SPF could be retained within the larger program through some kind of `firewall' allowing amended selection procedure, sub-project objectives and M & E. Independent Fund Management 3.160 Spinning off program management from Government departments and possibly also from the public sector is an approach that appears to have been effective. It has been done through an independent fund manager, as in the case of the Bank-supported Competitiveness Fund. This is an instructive example because management was originally in an independent agency and then was transferred to the DTI midway through the project's life (partly because its life was extended to complete disbursement). Another project (the SPII) is managed independently by a State agency (the IDC), but with the level of autonomy that would allow it to behave like a private agency. The organization of the SPII program seems to have been relatively strong, despite recent increases in the management cost ratio.85 The GODISA and the Ntsika programs also both used non-Government agencies for managing specific incubators and advisory centers, although their record was, as we have explained, less positive. 3.161 The TEO was established within DTI in 1999 with a view to potential spin-off but this did not take place. The reason appears to be that the DTI management was uncertain that outside management would be much more efficient and that it would require potentially difficult monitoring and oversight by DTI. Over time vested interests in maintaining the staff within the DTI also evolved. 3.162 The reason for spinning off management and implementation from Government is one of alignment of interests and objectives, as well as of effectiveness. This is because the more cautious audit and oversight mentality of Government is generally not appropriate to the moderately risk-taking and entrepreneurial approach required to operate a support program successfully. The inadequate promotion of Government programs may also be directly traced to this, whereas independent managers would assume program promotion as a primary task. Hence it is recommended that in general the Government should spin off program operation. While independent management is not a `magic bullet' and needs to be structured properly, the examples of the CF and the SPII (managed by DTI) suggest that independent management has been relatively efficient in practice. The Role of DTI and TEO if Programs are Spun off 3.163 The aim of aligning capacity and objectives suggests that the DTI and other Government departments should no longer be responsible for operational program management. This is not necessarily because of inherent management inefficiencies but because of the complications and conflicts which create problems for the retention of 85The SPII has faced a declining rate of grant applications and has had to hire a promotion officer which has increased management costs, so costs per approval have risen. See Phillips; op cit 85 both management and oversight in the same place. There seem to be two principal issues: · the division of responsibilities between the private and public sector; and · the division of responsibilities within the public sector. 3.164 The appropriate principle is to divide public sector responsibilities into: · policy, regulation, and oversight functions, which would be the responsibility of Government (i.e. DTI); · implementation and management of programs, which would be the responsibility of semi-independent Government, or private, agencies. 3.165 DTI or other Government staff would no longer manage support programs but would provide: (a) policy analysis and development, (b) formulation and development of regulations, and (c) build and maintain internal capacity for monitoring and impact evaluation of programs managed outside. That is, in the latter case staff would take responsibilities for monitoring and evaluating specific programs, not managing them. Public-Private Initiatives 3.166 Government-industry partnerships are justified in particular cases to promote industrial activities with long-term productivity potential, provided this does not imply `picking winners' but rather a process of coordination to identify opportunities on an ongoing basis.86 (The public interest in `supporting losers' is however difficult to sustain). 3.167 As stated, the GODISA incubators and the Ntsika LBSCs were both implemented by private/NGO groups, in the first case chosen through a competitive tender process and in the second through an accreditation process. Khula Finance is currently seeking private sector sponsors for micro-finance institutions. While the LBSC and GODISA programs have faced problems it may be that the model used for choosing project management (competitive tender or accreditation) is still appropriate because if it is carefully handled it is relatively likely to ensure independent and relatively committed on-the-ground management. It is worth drawing lessons from the experience of GODISA and the LBSC program to identify the reasons why it faced problems and to see if it can be applicable to future programs. 86In this context Rodrik writes: "the right way of thinking of industrial policy is as a discovery process - one where firms and the government learn about underlying costs and opportunities and engage in strategic coordination. The traditional arguments against industrial policy lose much of their force when we view industrial policy in these terms. For example, the typical riposte about governments' inability to pick winners becomes irrelevant. Yes, the government has imperfect information, (but) so does the private sector. It is the information externalities generated by ignorance in the private sector that creates a useful public role'. `Industrial policy for the 21st century' Dani Rodrik. Harvard, Kennedy School, Sept 2004. 86 3.168 Another type of link is the establishment of financing schemes linking large private firms and public agencies to provide finance to micro and small (black-owned) firms. Financing vehicles have been put in place to promote black entrepreneurship through corporate responsibility ventures, such as a Khula-Anglo-American mining fund, a Shoprite venture for setting up retail shops, Kickstart' with SA Breweries; the Eskom fund, Semele, and the ACSA fund. Using Large Scale Business to Support Small Business 3.169 Formal-to-informal links are fairly strong because small and informal firms are heavily dependent upon formal suppliers. According to the Johannesburg ME study 80% of the informal businesses purchase inputs/supplies from large formal firms, 64% purchase them from small formal or informal firms, 30% from government agencies and 25% percent from customers. Only about 10% of the inputs are self produced in the informal sector, and only 20% of training services which otherwise is privately provided through large formal providers. 3.170 On the other hand informal-to-formal sector links are fairly weak by international standards. This is revealed in both the MICA and the Johannesburg surveys. The MICA showed only 8% of sales going to other small firms and distributors, 2% to large firms and 2% to Government. The other study found that a large majority (88%) of MEs sell products directly to consumer. Only about 20% of the informal firms sell products to small formal firms (SMMEs) and only 5% sell to larger formal firms. 3.171 The relatively strong formal-to-informal link could be better exploited to promote business development in the informal sector. As mentioned, this could occur through intensified training programs funded by the State but organized through large supplier firms who would provide pre- and post-sales support to ME purchasers of equipment, materials and services. 3.172 The current weak informal-to-formal link is less easy to take advantage of. Given the large percent of goods going straight to consumption it would need solid growth in incomes and demand to stimulate ME production. However there is scope for increasing the forward links from MEs to larger formal firms through promoting the development of subcontracting. One way this is happening is through outsourcing to small firms, such as the `putting out' work of the garment CMT units to informal outsourced producers. Technology and know-how partnerships have also been initiated through spinning off small startup firms from motor vehicle OEMs who have in some interviewed cases supplied equipment at low or zero cost to the startup in return for provision of specified auto components (e.g. springs). Some such outsourcing clusters have been supported through for example the SPF and this model may be developed, provided cost­ effectiveness criteria are carefully adhered to in design. 3.173 The Johannesburg survey showed that about 44% of informal firms had received at least one contract, and 11% had received between two and ten contracts from large formal firms in 1998, while 15-20% reported having received two-to-ten contracts from 87 other similar sized businesses, largely for the supply of goods such as drinks, clothing, furniture, cement, sand, and bricks. The median value of such contracts was about R2,000 from firms and R7,500 from individual contractors, and the average duration of the contracts, two to three weeks. About 56% of MEs reported that they were happy with such contracts as they could bring more stable income, while about 18% were dissatisfied mainly because they received too small a share of the profits. 3.174 Subject to avoiding placing undue cost burdens on larger enterprises that might already be facing severe competition from imports,87 there is an apparently viable development approach in South Africa of large-to-small firm subcontracting through business transaction links, technology and know-how transfer links. This opportunity arises from the legacy of a dual economy with a `first sector' producing to World standards for a range of products, and with productivity levels far ahead of the second sector. 87For an account of competitive pressures in textiles, engineering and automotive parts see chapter 4. 88 CHAPTER FOUR INDUSTRY CASES: USING VALUE-CHAIN ANALYSIS TO IDENTIFY OPPORTUNITIES FOR SMME INTEGRATION AND GOVERNMENT SUPPORT A. THE ROLE OF VALUE-CHAIN ANALYSIS 4.1 Given their limitations of size and scale, integration into product supply chains is an important way that SMMEs can have access to expanding markets for their own products. Integration is, in this way, key to their long term sustainability, and clearly important to the objective of employment expansion and BEE. In recognition of this the Government of South Africa identified textiles and auto components as priority sectors and established major incentive programs to promote local sourcing down their supply chains. 4.2 As found under the MICA survey, the level of integration of micro-enterprises across all sectors into larger supply chains is, in fact, still relatively low, indicating considerable room for progress. Only 2% or less of ME sales in the sample were found to go to larger firms (with more than 100 employees), or to Government agencies (1 to 2%), the bulk of firms appearing to sell directly to consumers. None of the ME firms surveyed reported any direct or indirect exports. The earlier Johannesburg micro- enterprise survey (section 2.2.3) found a slightly higher level of informal-to-formal linkages, but its conclusions similarly indicated that there was little forward integration.88 In other words, as commented on by other researchers,89 South Africa's overall SMME sector appears to compete with larger enterprises in the same product markets rather than complement them. 4.3 This low level of integration into larger-scale manufacturing is not out of line with patterns in other Sub-Saharan African countries. However, in countries outside the region, sales to formal SME firms and wholesalers are relatively higher than in South Africa, accounting for 30% of ME output in Bangladesh, 24% in Indonesia, 17% in Guatemala and 12% in Brazil, against under 8% in South Africa. Thus, by international comparison, there appears to be a lack of ME integration in South Africa not only into the first economy but, also, into the formal sector in general. This has a negative implication, particularly for black-owned businesses, because they are predominantly very small and because of the historic obstacles to black-owned business startup. 88The differences between the Johannesburg study and the MICA could reflect differing ethnic composition of ownership and less representative location, i.e. within the largest conurbation in South Africa. 89Qualmann R. 2000 Economic Development and employment promotion in South Africa: analysis with special reference to SMME promotion and strategy options for German Development Cooperation" cited in Rogerson: op cit. 89 4.4 To address this question, a more detailed analysis of the value-chains (VC) of relevant product lines is needed, to be able to identify where the main entry points for integration of SMMEs exist and, equally importantly, what needs to be done to access them. By analyzing the bottlenecks, efficiencies and competitiveness of specific stages of production and supply and their dynamics, VC analysis can pinpoint areas for improvement and ways in which SMMEs could increase their contribution, as well as where Government support might be required and justified. A fair amount of work on value-chain competitiveness has already been done within South African academic and professional organizations since the late 1990s,90 notably in the auto-parts and textiles sub-sectors, and also in furniture, leather, wine, fruit and other areas. These findings would be utilized, where appropriate, to strengthen our own conclusions. 4.5 The choice of sectors for this analysis (intended to be illustrative rather than exhaustive) is based on: (a) where employment potential is high, and (b) where the major incentives have already been provided. On the basis of these criteria, two sectors, the cotton-to-garments supply chain and the auto components sub-sector, were selected. 4.6 The cotton-garments chain, in particular, accounts for more than 200,000 jobs, direct and indirect, in an industry that accounts for nearly 10% of total manufacturing employment. The auto components chain accounts for over 75,000 direct jobs, plus another 190,000 indirect jobs in the motor trade. Both product lines operate in markets exposed to increasing global competition, so that integration means that the competitive position of SMME suppliers cannot be divorced from the fate of the large-firm segment of end-product producers. Both are recipients of generous government incentives (in effect, public subsidies), aimed at promoting local sourcing of inputs, giving rise to the question of whether these have been successful in their objective and whether the benefits have been transferred to the appropriate recipients. 4.7 The impact of global competition has not been uniformly detrimental to the SMME sector. The gradual restructuring of the garments industry, for instance, has seen an expansion of sub-contracting by larger firms to smaller, more flexible and lower cost CMT units, some of which have been operating quite successfully. The lessons of their success, which may relate to broader issues of skills and productivity, need to be identified and, if possible, replicated. 4.8 Against this background, the VC analysis would need to look at: (a) the dynamics of the product and service supply chain in each sub-sector; (b) the long-term competitiveness of the sub-sector, the degree of strategic re-orientation it must undergo and the implications of this on jobs; (c) what effect incentives have had on short-term and long-term competitiveness; (d) whether the organization of the supply chains for each product creates distortions affecting small-firm participants; and (e) the main options for entry and growth of smaller firms, and how smaller firms can increase productivity and competitiveness so as to be able to participate. 90Value-chain and associated research in South Africa has been done notably by The Industrial Restructuring Project at the University of Kwazulu Natal School of Development Studies, and also by the Universities of Witswatersrand, Stellenbosch, Cape Town, and others, on behalf of the DTI and other organizations. 90 B. COTTON-TO-GARMENTS SUPPLY CHAIN Industry Overview 4.9 Following the end of World War II, the South African textile industry expanded to furnishings, industrial textiles and clothing and then into synthetic fibers in the 1960's.91 The industry grew under trade and tariff protection, receiving direct support for investment from the state-owned Industrial Development Corporation. Historically, production was driven by domestic demand and exports constituted only around 6% of domestic production during the 1970's. Asian investment during the 1980s helped to reinvigorate the sector and nearly two-thirds of investment in the textile industry is currently from foreign sources. 4.10 In the early 1990s, with the ending of the economic sanctions, South Africa embarked on a general liberalization of its trade regime. Anxious to put pressure on its local industries to become internationally competitive, the Government reduced duties on clothing and textiles to levels even below its 1994 WTO commitments. In addition, it kept in place bilateral trading arrangements with Malawi, Mozambique and Zimbabwe, which allowed a range of clothing products to enter South Africa at duty levels below its MFN rates. 4.11 Following trade liberalization, textile output stagnated during the mid-1990s and spinning output declined. Relative prices for both fabrics and garments fell as a result of world competition. On the export side, however, while exports of unprocessed products such as raw wool and cotton fell, those for processed products (yarn, fabrics and garments) saw an increase. Exports, as a proportion of production, rose from about 25% to 36% over 1990 to 2001, but far more rapidly for garments (from 4% to 21%). The overall share of imports of textiles and garments in total consumption also rose respectively from 27% to 37%, and from 16% to 23%, over 1990 to 2001, and further increases have occurred since then. There are also thought to be considerable unrecorded or under-invoiced imports. 4.12 Tariffs will be reduced to zero for all SADC-origin textile and garment products by 2007. For EU products rates will be reduced to 10% in 2007 for woven and knitted fabrics (from 22% in 2004), and 20% for garments (from 40% in 2004). (See Annex 10). 4.13 The industry is one of South Africa's larger manufacturing employers and exporters of manufactured goods. Production at all stages is fairly highly concentrated in a handful of large firms and a large number of smaller firms. In most textile production activities, the top four firms account for over 50% of production, while there are some 900 firms overall in spinning, weaving, finishing and in textile goods and knitted 91For further details on the development of the textile industry se Robert, S. and J. Thoburn `Globalization and the South African Textile Industry' Globalization and Poverty Discussion Paper No 9, Universities of Witswatersrand and East Anglia. 91 products, without including garments. Thus, despite high concentration there is a considerable potential for small firm participation. 4.14 Though production has stagnated over the past few years in the face of intensifying international competition, and employment has steadily declined, the textile sector remains the sixth largest employer in the manufacturing sector and eleventh largest exporter of manufactured goods. It currently employs about 51,000 persons in spinning, weaving, and finishing of all textile goods, 11,000 in knitting mills producing both fabrics and garments, and another 101,000 in the apparel production industry. Taken together, as many as 162,000 people work in the textile/apparel industry; representing 14.8% of all industrial manufacturing employment. With approximately 60,000 further workers in cotton cultivation and ginning, total employment in the sector of approximately 223,000 is about 11% of manufacturing and agricultural employment. 4.15 Despite growing some of the world's best quality cotton on irrigated land, South African textiles are struggling, caught between intense foreign competition and relatively high domestic labor costs. The latter stems from unionization and high management remuneration, both of which are legacies of labor market segmentation and unequal pay during the Apartheid era. 4.16 Because of its size and strategic importance within South Africa, the textile sector is significant in terms of the forming of public policy. In particular, it is a key sector for the whole issue of State support, being an important vehicle for the delivery of skilled and unskilled jobs, and black ownership and employment, over the whole value chain from agriculture through manufacture to retailing. 4.17 All the basic stages in the sub-sector value-chain are present within the country. These are the following: · cotton cultivation · ginners and oil pressers · spinners · weavers · knitters · clothing manufacture · textile goods 4.18 The industry production and market flow in South Africa is represented in Annex 11. 4.19 The cotton textile industry is integrated through an umbrella organization Cotton SA. Ginners, weavers, spinners, textile manufacturers also have their own representative organizations that provide information; promotion; quality standards and training; co- ordination of research; technical support; marketing; lobbying and similar support to their members. A large proportion of the textile workers are represented by the South African Textile workers union. However, despite the inter-relationships, an integrated policy 92 towards growth of the entire industry seems to be lacking. Cotton SA appears focused on making South Africa compete at the global market level in cotton production, whereas South African cotton production is very small-scale and unstable compared with the large global producers, and dry land cultivation is uneconomic. 4.20 As we will see, out of the several stages of cotton textile production, viable entry points for integration of micro and small scale, including black owned business, exist most strongly at the two ends of the value chain, that is cotton cultivation and garment and apparel production. In the following sections we will, therefore, focus on these two stages while simply summarizing findings in other areas. Cotton Production 4.21 Ninety-eight percent of cotton production is grown by approximately 500 larger scale commercial producers, with a combination of irrigation and rain fed cultivation, and a larger group of small scale farmers92 who also farm a combination of dry land and irrigated cotton. Small-scale has fallen steadily from a recent peak of 12% in 1997-98, to 4% in the 2000-01 season, and only 2% in 2004-05. The number of smallholder farmers has decreased from about 4,000 in 2001 to 1,700 today. Generally, this decline has paralleled decline in overall cotton production in South Africa caused by declining cotton prices, more attractive returns from competing crops, and drought conditions. 4.22 South Africa produces some of the best quality cotton available on the world market, meeting or exceeding recommended ICAC minimum requirements and, in some cases, exceeding the quality of standard USA upland cotton. In fact, the only current producer of cotton T-shirts in South Africa also exports to the USA for such high-end retailers as Brooks Brothers. This is an example of a niche player who is able to take advantage of high quality fiber available in South Africa, but stay away from high volume, lower quality commodity level production markets. 4.23 Despite its high quality, total cotton output has fluctuated widely around 25,000 tons per annum during the past ten years, declining significantly between 1999 and 2005. This production level is tiny in comparison with major World producers, especially the largest producers like the USA, which produces over 4.5 million tons and China which produces nearly 6 million tons per annum. South Africa's output amounts to less than 0.1% of the world production, dominated by the USA, China, India and Pakistan. Currently, South Africa imports approximately 50% of its total cotton consumption, largely from lower cost suppliers, Zambia and Zimbabwe. 92Defined by Cotton SA as farming less than 10 Hectares. 93 4.24 The stagnant output of cotton has occurred, despite increasing yields per hectare of irrigated cotton, due to the number of serious droughts in the main production regions and unfavorable prices, thus making cotton a less attractive cash crop compared to the production of for example maize or paprika. Many farms (both irrigated and dry land) have switched to other crops in recent years. Cotton income for the coming season is expected to decline by about 38% and ginning income by about 33%. Table 4.1: Historical Production of Lint Cotton in South Africa Season Production in Metric Tons 1994-1995 22,317 1995-1996 21,472 1996-1997 37,699 1997-1998 24,180 1998-1999 34,507 1999-2000 44,926 2000-2001 25,757 2001-2002 33,893 2002-2003 17,384 2003-2004 15,285 2004-2005 25,798 2005-2006* 19,814 * estimate Source: Cotton SA 4.25 Increasing yields apply only to irrigated cotton. Typical dry land farming provides an average yield rate of 473 kg/ha, far lower than irrigated land yielding on average 3,747 kg, which is high by World standards (e.g. far exceeding Australia, Israel and China). Dry land farmers also face greater unpredictability. 94 Table 4.2: South Africa Cotton Production Sub-Sector Profile 1.0 Land available for cotton production 1.1 Available for dry land cultivation 80,000 ha 1.2 Available for irrigated cultivation 110,000 ha 1.3 Actual land currently under dry land cultivation 10,374 ha 1.4 Actual land currently under irrigated cultivation 13,000 ha 1.5 Total under cotton cultivation 23,374 ha 1.6 10 year average (1994-2005) 58,372 ha 1.7 Utilization Rate 12.4% 2.0 Production level 2.1 Actual production (2005/2006 Estimate) 19,814 MT 2.2 Peak Production level (1999/2000) 44,926 MT 3.0 Farmers in cotton production (100% Private) 2,200 farms 3.1 Less than 10 ha of land (2% of total production) 1,700 farms 3.4 More than 10 ha of land (98% of total production) 500 farms 4.0 Yield rates (average) 4.1 Irrigated cotton yield 3,747 kg/ha 4.2 Dry land cotton yield 473 kg/ha 5.0 Seed Cotton Production cost (average) 5.1 Irrigated production cost $0.26/kg 5.2 Dry land production cost $1.52/kg 7.0 Average market price of seed cotton $0.49/kg 8.0 Cotton import (estimate 2004 - 2005) 29,437 MT (lint) 9.0 Cotton export (2003 - 2004) 0 MT (lint) Source: Cotton SA 4.26 In 2004-05, approximately 60% of cotton was under irrigation and total production represented only about 12% of the total land available and suitable for cotton. Table 4.3: South Africa Cotton Yield Rates Season Yield Irrigated Land Yield Dry Land Seed Cotton (Kg/Ha) Seed Cotton (Kg/Ha) 1994-1995 2,362 318 1995-1996 2.742 778 1996-1997 2,189 403 1997-1998 2,724 580 1998-1999 2,680 545 1999-2000 3,107 777 2000-2001 3,455 593 2001-2002 3,538 515 2002-2003 3,482 475 2003-2004 3,455 492 2004-2005 3,747 473 Source: Cotton SA 95 4.27 For an average dry land cotton farm yield rate of 473 kg/ha, the average production cost per hectare of seed cotton is estimated to be approximately $718.77, which translates to a production cost of $1.52/kg for seed cotton. While this yield is on par with cotton production in many developing countries, dry land production is extremely uncompetitive, with losses of close to $450/ha at current market prices. However, at yields of 3,747 kg/ha, irrigated cotton production cost is only $0.26/kg which is internationally competitive, especially given the quality of cotton. 4.28 One of the reasons why smallholder cotton farms are not achieving higher yield levels is related to the communal ownership of land, a legacy of the communal land tenure that was imposed on the indigenous population by the earlier colonial and apartheid governments. Following the end of Apartheid, a number of land reform measures were initiated in an effort to return lands to black Africans; however progress has been very slow. Communal landownership has often led to a more diffused system of decision-making, reliant on the ability to achieve consensus, in the absence of which productive lands have sometimes had to lie fallow. Also as a result, in many cases, irrigation lines extending all the way to neighboring fields have not been utilized for lack of operating funds to support the operation of pump houses and the upkeep and maintenance. A solution to this issue will be needed if smallholder, black-owned farms are to increase their integration into irrigated cotton cultivation. 4.29 South Africa also has opportunities to produce organic cotton, which currently sells at a premium price, allowing higher profits. Organic cotton is also much more labor intensive than conventional cotton. While still a minute share of world demand, large local retailers such as Woolworths are increasingly looking for supplies of organics. This could be part of a specialist product strategy with high employment possibilities. 4.30 The value chain for cotton production can be divided into nine key stages: Land preparation >Planting >Seeding >Thinning >Stamping >Weeding (3 times/season) > Chemical spraying (5 times/season) >Fertilizing (2 times/season) >Harvesting. 4.31 The diagram shows the value chain cost structure for a typical irrigated farm. DIAGRAM 4.1:Value Chain for Irrigated Cotton Production in South Africa Labor Agro- chemicals 9.6% 90.4% Land Planting Seeding Thinning Stamping Weeding Spraying Fertilizer Harvesting Irrigation Preparation 4.3% 3.6% 6.1% 5.6% 0% 7.5% 23.8% 16.3% 6.8% 26.0% Labor Agro- Labor Operating chemicals Costs 14.0% 86.0% 32.9% 67.1% 96 Table 4.4: Breakdown of Irrigated Cotton Production Costs ZAR Total $ Total % Total Land preparation 272 41.85 4.3% Planting 226 34.77 3.6% Seeding 386 59.38 6.1% Thinning 352 54.15 5.6% Stamping - 0.0% Weeding (3 times/season) 472 72.62 7.5% Chemical spraying (4 times/season) 1,501 230.92 23.8% Fertilizing (2 times/season) 1,031 158.62 16.3% Harvesting 432 66.46 6.8% Irrigation 1,640 252.31 26.0% Total 6,312 971.08 100% 4.32 The table shows the level of typical income expenditure and profits for irrigated cotton. By comparison, typical dry land farmers are producing at considerable losses. Table 4.5: Average Profits for Irrigated Cotton Profit/Loss $/Ha Total Revenue 2,062 Total Cost 971 TOTAL Profit 1,091 4.33 The value chain coefficients for non-irrigated land differ from those of irrigated land entirely because of the absence of irrigation costs, otherwise, input ratios are similar. This is because extension officers provide standardized help to farmers (large and small alike) so that input costs are largely identical between small and large farmers and irrigated and dry land cultivation, apart from the cost of the irrigation itself. Pesticide application is the main cost as application is time-intensive and requires skills and experience. Agro-chemicals (for pesticide and fertilizer), amount to about 50% of total costs, and labor, which accounts for most of the residual. 4.34 As in all chains, the key cost areas are also likely to be the chain leader stages ­ i.e. these stages tend to have control over the chain as a whole. In the case of irrigated farms, agro-chemicals comprise about 38% (in weeding, spraying and fertilizing) and irrigation operating costs about 18% of total costs. Labor again constitutes most of the residual. 97 4.35 The comparison of South African costs and yields in comparison with main World producers, based on representative data, is as follows. In setting out these costs it should be born in mind that one of the pressures on all countries is the presence of rich country subsidies to their own farmers.93 Table 4.6: Benchmarking Cotton Farming Cost and Yield Seed Cotton YieldProduction Cost Production Cost (kg/ha) (US$/ha) (US$/kg seed cotton) Kyrgyzstan 2,450 393.63 $ 0.16 China 3,500 752.00 $ 0.21 Pakistan 1,680 387.34 $ 0.23 Kenya 570 184.00 $ 0.32 Cambodia 1,200 415.00 $ 0.35 South Africa-Dry Land 473 718.77 $ 1.52 South Africa-Irrigated 3,747 971.08 $ 0.26 4.36 South Africa's reasonably competitive irrigated cotton costs suggest that the focus of attention must continue to be placed on irrigation, which also allows predictable yields and profits, which in turn facilitate the encouragement of more cotton production on an individual level. 4.37 To bring a greater number of smallholder and black-owned farms back into cotton cultivation will necessitate improvements in management of land usage, as well as the provision of new irrigation facilities, or connections to existing irrigation lines that may be serving nearby fields. This may mean addressing the issue of communally owned land, discussed earlier. 4.38 A possible model for the introduction of black-owned farms into the irrigated cotton sub-sector might be that of the outgrower schemes in Mozambique.94 On account of various factors, the large cotton companies have been retreating from cotton production in Mozambique, and outsourcing cotton growing to smallholders who now account for 85% of total seed cotton production. Many smallholders have resorted to growing cotton on blocks, sometimes provided by the concessionaires: enabling more intensive technological packages to be applied, and crop yields to be maximized. Under this model, individual smallholders are obliged to sell all cotton produced to the company that is operating in their area, in return for inputs on credit and extension services, plus exclusive marketing rights. In areas where companies are absent, the Cotton Institute of Mozambique has been promoting smallholder production. Many producers are forming associations with the support of some NGOs and cotton companies, as a means of 93Low prices on world cotton markets, partly due to rich country subsidies, have provided little incentive to produce. Prices started falling in the mid-1990s, reaching an all-time low by October 2001. They have risen recently, but are expected to fall again once China has recovered from poor harvests. Major efforts have been made to persuade industrialized countries to reduce subsidies but no significant change has occurred to date. 94Described in GDS-LLC: "Value Chain Analysis for Strategic Sectors in Mozambique", prepared for the Enterprise Development Project, Mozambique, April 2005. 98 obtaining higher than government minimum prices, and to improve overall the efficiency of their activities. Cotton Ginning 4.39 The ginning industry in South Africa is made up of eight firms with one currently under construction.95 Three ginners are also fabric producers who have integrated backward into ginning. They use their spare capacity to provide ginning services to neighboring farmers. In addition to these integrated textile producers, Da Gama Textiles is currently in the process of establishing a gin in the Eastern Cape, although it is facing cotton supply problems. The large ginners traditionally purchased seed cotton outright from farmers, ginned it, and sold the lint themselves but, recently, they have moved to contract ginning, providing only ginning services while the farmer retains ownership over his cotton and sells it to spinners and oil presses. This has created greater uncertainty and further discouraged smallholder production. Because of the relatively high cost of transporting seed cotton, ginners generally serve specific locations. Some contract ginning has been started by local growers associations as an integration activity. 4.40 Ginning outturn (GOT), or the ratio of lint to seed cotton produced by the ginning process, is a critical factor in the competitiveness of the ginning sector. The average GOT achieved by most ginneries in South Africa is approximately 37%, which is considered fairly efficient in production of lint from seed cotton. Table 4.7: South Africa's Ginning Production Sub-sector Profile 1.0 Number of ginneries 9 2.0 Ginning capacity 2.1 Number of saw gins ~200 2.2 Total installed capacity 53,000 MT 3.0 Technology in use Saw gin technology 4.0 Average c 34% 5.0 Capacity utilization 1.0 Ginning outturn (GOT) 5.1 Potential 40% 5.2 Actual 37%96 2.0 Cotton purchasing method 6.1 Direct purchase from farmers 60% 6.2 Ginning Services on Contract Basis 40% 7.0 Cotton purchasing schedule Seasonal 8.0 Average price for seed cotton paid by ginneries $0.49/kg 95The key players in the sub-sector include Clark Cotton, NSK, Makhatini Cotton, Vaalharts Cotton, Oranje Pluismeule, Weipe Cotton Gin, Loskop Cotton Gin and Noordkaap Pluismeule. 96Generally, ginning equipment was well maintained and in good operating order, which is reflected in the high ginning outturn ratio. 99 4.41 The current average cost per kg of lint cotton is approximately $1.68, which includes the cost of the initial seed cotton input at 3.18 ZAR/kg and a GOT of 37%. The current ginning cost per kg of lint cotton is $0.36. Textile mills are purchasing lint cotton at a price of about ZAR11.09/kg. This represents a very small 1.05% profit margin for ginneries on average, subject to price fluctuations. Ginning costs are relatively high in South Africa and the return on investment to improve ginning facilities and quality is generally low to inadequate. 4.42 Low profitability discourages the additional cost of planning, coordination, collection and transport of seed cotton from a wide range of small holders. Although such services are available, they are ad-hoc and, thus, small farmers do have an added burden of transferring seed cotton to the processing facilities. 4.43 In Asia, cotton ginning has often been done by government organizations. In the Chinese market, the Government (Board of Cotton and Jute) is slowly retreating from full ownership of such activities, but it finances operations for procurement agencies, which allows for lower ginning costs. Table 4.8: Benchmarking Ginning Costs Country Ginning Cost (US$/kg lint) China 0.050 Brazil 0.050 Pakistan 0.076 Kyrgyzstan 0.130 Syria 0.230 South Africa 0.367 4.44 Of the total administrative cost (9.2% of total), ginning management comprises 47.5%. Vehicles and housing for gin managers in South Africa account for about a further 8% of the total value of lint produced. Such high management costs are a legacy of the past when labor costs were controlled at low levels. Excessive management costs seem to apply all the way down the textile value chain. 4.45 While there may be an opportunity, through local producers associations, to set up small ginneries with lower overhead costs, it does not appear that this segment of the cotton textile value chain is a promising one for micro/small scale entry. This is because firstly, the industry is dominated by the textile mill purchasers who are tending to set up their own ginneries; secondly, economies of ginning preclude very small ginneries; thirdly, transport costs from scattered outgrowers are higher than for nearby integrated cultivation and, finally, because the potential employment and profitability is low. 100 Textile Fabrics Production 4.46 At present, the manufacturing base is not large overall, with 3,000 looms and 400,000 spindles, but it is quite comprehensive, covering industrial, domestic, household and apparel textiles. There are in excess of 300 companies engaged in some aspect of cotton textile production. At least four of these are vertical mills, with many of these also involved in downstream textile garment production under AGOA. At least one (Da Gama Textiles) is "reverse integrated" into cotton ginning and, thus, has a vested interest in cotton production as well. 4.47 There are 16 spinners active in South Africa, some of whom are independent, while others are part of larger integrated mills. The largest of these are Frame Textiles and Da Gama Textiles. The spinners generally buy lint cotton from ginners (or farmers directly), process it and deliver product to spinners and weavers. The vertically integrated mills produce yarn and woven, knitted, and finished fabrics which are then sold, within their own production structure, at cost. The remaining output from the spinning activities that is not consumed internally is sold directly to local markets or to independent weavers and knitters. 4.48 Currently, spinning, weaving and knitting employs approximately 61,000 workers, while thousands of others are in supporting and subsidiary industries such as transport and packaging. This is in addition to employment in cotton cultivation and ginning but, the employment generating capability of the textile and garment sectors is increasingly under threat. Labor cost pressures have also resulted in shift in production from South Africa to Lesotho where the wage rate is about 60% of South African levels. Textile employment has steadily declined over time. NPI figures show employment in textiles and garments declining 3.5% in 2004, even during an economic upturn, and further losses have occurred in 2005. 4.49 Imports grew at around 2.5% per annum between 1995 and 2004, mainly from Asia, but also from the U.S. and EU. Imports are expected to continue gaining market share. Imports from neighboring countries are small, but some local manufacturers do outsource to Lesotho and Mozambique, to take advantage of lower labor costs. During the last two years, several local companies have closed, leaving a number of factories and warehouses empty. Asian companies and management, particularly in the Eastern Cape region, have bought and set up textile manufacturing concerns. 101 4.50 Textile industry data are summarized in the following table. Table 4.9: South Africa's Textile Production Sub-sector Profile 1. Number of companies currently operating in the sector 32 2. Production and Consumption of Cotton Base Textiles (2004) 2.1 Spinning (in thousands) Installed capacity 120 tons Actual production 72 tons Utilization Factor 60% Consumption 254 tons 2.2 Weaving (in thousands) Installed capacity 578 mē Actual production 319 mē Utilization Factor 55% Consumption 690 mē 2.3 Knitting (in thousands) Installed capacity 32 tons Actual production 21 tons Utilization Factor 66% Consumption 37 tons Source: Compiled by Global Development Solutions, LLCTM from data from Stats SA and TexFed 4.51 All players in the market are represented by the Textile Federation, which consists of four sub-sector representative organizations including South African Cotton Textile Manufacturers' Association (SACTMA), South Africa Cotton Worsted Manufacturers' Trade Association (SAWMTA), National Fabric Knitters' Trade Association (NFKTA), and the Fibre Group. The following table provides data on employment in the industry. Table 4.10: Employment and Wages in the Textile Industry 2004 Sector Workers in Total Worker Average Worker total Remuneration (R) Remuneration (R) Total Manufacturing 1,279,000 70,424,000,000 55,061 Textiles-Spin. 20,226 1,014,000,000 51,830 Weaving, finish Textile - Other 31,230 1,653,000,000 Knitting Mills 10,882 624,000,000 57,342 Clothing Production 101,149 3,327,000,000 32,892 Avg. Wage for unskilled labor = R36/day Seasonal R55/day Permanent Source: Texfed 4.52 Following years of protectionist policies, meeting global market requirements continues to be a challenge for South Africa's textiles sector. Poor performance is reflected in long lead times ­ e.g. 90-150 days from a local textile mill, compared to sourcing from China in less than 60 days. Despite high quality cotton, fabric quality is also seen as a problem, especially erratic shrinkage and fabric bleeding. Local fabric and garment manufacturers suggest that they can source the same quality yarn from Asia for 102 20% - 25% lower cost. As a consequence, it is estimated that only 15% of the fabric used for export products are sourced from local suppliers. Manufacturers are willing to pay for imported fabric from Asia and incur duty to export under AGOA. 4.53 In general, the textile industry has been hampered by low productivity that has greatly contributed to the lack of competitiveness. The relatively high average age of plant and machinery and high labor cost in relation to productivity has exacerbated these problems. The investment ratio has been low (at 3.6-3.8% of revenue) even relative to the low economy-wide rates. Low investment and low levels of training of unskilled workers are linked to poor performance, including high return rates and slow delivery times, compared to international norms. Because of outdated capital equipment, and weak management and production organization, many firms closed down when import tariffs were reduced at a rate faster than required by GATT after 1994. 4.54 The industry began to show a slight improvement from 1998 onwards, following restructuring and the establishment of Cotton SA. The previous local content requirement for ginned cotton was phased out. The companies that survived the trade liberalization attempted to increase exports, and reduce their product range, to achieve economies of scale, by increasing capital- intensity and focusing on niche markets requiring higher quality. But, improved competitiveness has also led to reduced employment. 4.55 The value chain is depicted in Annex 12(A) and Annex 12(B). The main activities are: Combing >Twisting >Weaving >Dyeing >Finishing 4.56 Raw material is the highest single cost item (28% at average cost $1.83/Kg ). 97 Combing is the next highest cost, of which the major component is administrative costs. Dyeing represents the next largest cost, of which a disproportionate amount is also for overheads.98 Average finished cost is $6.42. The table compares costs of lint cotton in five countries. South Africa's lint cotton is relatively expensive but higher quality. Table 4.11: Comparison of the Cost of Lint Cotton in Selected Countries Country Cost per kg (US$) Source of Lint Cotton Kyrgyzstan $1.10 Mostly domestic, some imported Mali $1.25 100% domestic Kenya $1.41 Blended domestic and imported South Africa $1.83 60% imported / 40% domestic Bangladesh $2.14 100% imported 4.57 The cotton cost coefficient of 28% creates a disadvantage at the textile production stage as it is higher than in competitors, including Kenya. However, while the cost of lint cotton is high, the production process itself appears to be competitive, despite poor 97$1.83/kg is the delivered price of lint cotton from the ginnery to the textile. 98This includes automobile allowances and educational benefits for management. 103 delivery times and fairly high overheads. Further textile cost breakdowns are in Annex 13. Table 4.12: Benchmarking Textile Fabric Production Cost ($/kg) Lint as % of Lint Cotton Production Cost Total Cost Total Cost South Africa $ 1.83 $ 4.59 $ 6.42 28.5% Cambodia $ 1.12 $ 5.55 $ 6.67 17.4% Kenya $ 1.41 $ 5.43 $ 6.84 22.0% Mali $ 1.25 $ 4.52 $ 5.77 19.5% Global Development Solutions, LLC 4.58 South African textile economics are, on balance, marginal due to high production costs. Although 300 firms are operating at this stage of production, with regard to entry points for micro/small scale, including black enterprise, a main consideration at the textile production stage is economies of scale, especially given the keen cost competitiveness of the international market and the pressures on South African industry to automate or relocate to low labor cost areas within or outside the country. Consequently, the opportunities for small firms are limited and in terms of SMME promotion there is considerably more scope at the garments stage which we now examine. The one possible exception is in the area of knitted fabrics, where demand from garment exporting firms in Lesotho could considerably surge, with the expiration of the 3rd Party LDC Concession under AGOA. A potential demand for up to 10,000 tons of fabric per annum could exist for South African product if this Concession is not extended beyond 2007.99 Garment and Apparel Production Industry structure 4.59 The garment industry currently has 1,300 apparel manufacturing companies operating in South Africa, plus nearly 2,000 small emerging companies based in the Decentralized Zones100 for access to low cost labor. There is a fair diversity of garment manufacturers, some producing high quality, low-volume products, such as men's suits and dress shirts and others that continue to manufacture basic commodity-type products such as low-end T-shirts and trousers, in a rapidly disappearing market. Total apparel production was officially estimated at 250 million units in 2003 by Stats SA. 4.60 The garment firms can be categorized into four groups: (a) medium-sized, urban- based companies with organized labor, (b) larger-scale urban-based with organized labor for large-scale production; (c) small entrepreneurial companies working outside formal 99FIAS: "Lesotho ­ The Competitiveness of Vertical Integration in Lesotho's Garment Industry", Discussion Draft, April 2006. 100Decentralized Zones were established in 1971 to slow migration of black Africans to the large towns. They allowed reduced wages, were subsidized by the government, and by 1990 had generated significant employment. Since 1994 as subsidies were removed many of the zones have disappeared but some labor resources remain. Unionization has been slow and wages remain below the average. 104 labor and tax structures; and (d) transnational companies (largely Hong-Kong and Taiwanese owned) decentralized to take advantage of low-cost labor. Firms in the first two categories are technologically relatively modern, while those in the final category tend to make most use of payment on piecework basis. 4.61 While there are a large number of textile and garment manufacturing firms, production is dominated by a relatively small number of companies. According to the most recent manufacturing census data, in 2001, four of the largest firms accounted for more than half of all output in eight of the fourteen textile sub-groupings. Exports accounted for R1.4 billion for apparel and R2.5 billion for textiles, mostly to USA and European markets. 4.62 More than 90% of the garments exported from South Africa are made by factories owned or operated by foreign investors, where purchases are organized by buyers in Asia. To cut labor costs, producers are seeking out areas where the labor force is not unionized and is less costly than Durban and Cape Town, the same pressure that has driven production to relocate to Lesotho where high volume-low cost products dominate exports to the U.S. market under AGOA. Table 4.13: South Africa's Garment Production Sub-Sector Profile 1.0 Garment Manufacturers o Large scale (market share of more than 10%) ~40 o Small scale (market share of less than 1%) ~1,200 o Export oriented < 50 2.0 Export Destinations 5.1 USA 42% 5.2 Europe 48% 5.3 SADC 7% 5.4 Others 3% 105 The garment value chain 4.63 In addition to cotton fabric costs, the value chain for a typical cotton T-shirt can be divided into at least six cost stages of value adding activities, namely: Cutting/layering>Sewing/assembly>Finishing>Packing>in-factory inspection DIAGRAM 4.2 Value Chain for Garment Production ­ T Shirts : Materials Labor Electricity Maintenance 22.2% 50.0% 8.3% 19.4% Raw Cutting/ Sewing/ Finishing Packing/ Inspection Overhead Material Layering Assembly 6.8% Loading 1.4% 20.0% 45.9% 5.1% 16.3% 4.4% Salaries Rent Transport Marketing Financing Deprec Other 55.8% 3.5% 9.3% 10.5% 3.5% 3.5% 14.0% 4.64 Apart from bought-in raw materials, once again overheads and, in particular, administration are the most significant contributors to the production cost of the T-shirt. South African garment producers use the same inputs, regardless of quality and cost, in both domestic and export markets. As a result, garments are expensive for the local market where low quality, low price imports continue to gain market share. As an approximate average, it costs $4.29 to produce a cotton T-shirt of high quality for export or high end domestic consumption. This includes $1.97 in fabric (45.9% of total). Sewing and assembly constitute 16.3% of the overall cost, with labor input accounting for half of the cost associated with that stage of production. Overhead runs at 20%. Table 4.14: Sewing Assembly Production Cost Breakdown $/unit % of Total Material 0.16 22.2% Labor 0.36 50.0% Electricity 0.06 8.3% Maintenance 0.14 19.4% 0.72 100.0% 4.65 CMT (cut, make, trim) costs for T-shirts are substantially higher in South Africa than in other countries. Cutting, sewing and finishing costs are higher than those of competitors in Africa. 106 Table 4.15: Cotton T-Shirt Assembly Cost Benchmark CMT Packing Inspection Admin TOTAL South Africa $ 1.21 $ 0.19 $ 0.06 $ 0.86 $ 2.32 Lesotho $ 0.39 $ 0.11 $ 0.05 $ 0.22 $ 0.77 Kenya $ 0.67 $ 0.05 $ 0.06 $ 0.65 $ 1.43 Mozambique $ 0.43 $ 0.02 $ 0.04 $ 0.14 $ 0.63 Mali $ 0.60 $ 0.05 $ 0.06 $ 0.82 $ 1.53 CMT: Cut, make and trim 4.66 In general, while the garment sector is well diversified compared to other African countries, the cost structure at the low-end, high volume segment is significantly out of alignment with Asian manufacturers, as seen from the following cost estimates for T-shirt production: Table 4.16: FOB Price Comparison for 180g T-shirt ($/unit) China India Bangladesh Kenya Nepal South Africa 180g $0.85 $1.15- $1.28 $1.85 $2.00 $2.30 - T-shirt $1.45 $2.85 ($/unit) Production economics 4.67 The average size of apparel operation in South Africa is too low to achieve economies of scale for "commodity" garments. There are two distinct supplier bases for apparel production. Generally, given the high wage and cost structure of garment production in the Western Cape, manufacturers located in this area tend to specialize in high-end niche market products for both local and export markets. In contrast, manufacturers located in KwaZulu Natal (KZN), facing a relatively lower cost structure, tend to target low-end, high volume `commodity' type products (Table 4.17, below): Table 4.17: Comparison of Western Cape and KwaZulu Natal Clothing Sectors Western Cape KwaZulu Natal Production Largely CMTs, but offer full-line Mainly CMTs Cost structure High Low Product niche High-end, high-quality Low-end, low -to-moderate quality Ownership South African Chinese, Taiwanese, Indonesian, and Singaporean Markets High-end domestic and export niche Uncertain ­ displaced by very low price markets, low volume products from China Reject rate <1% High (no reliable data currently available at the factory level, but is considered relatively high by most accounts) 107 4.68 There are clear distinctions between the two categories of producers with respect to their target market, ability to weather competition in the market against high-volume producers like China, use of unskilled labor, etc. The key distinctions are summarized below: Table 4.18: Characterization of Two Categories of Apparel Manufacturers in South Africa Company A: High-End Niche Company B: Low-end Commodity Market Producer Type Producer Product Specifications High quality, specialty fabric or Low commodity type, low price, easy material or high tech fabrics (Gortex construction, straight line stitch type), unique items Production Volume Low Low Product Quality High Low Product Selling Price High Low Market Type Niche market, specialty applications, Commodity type, general public high end sellers consumption, government contracts for bulk Capital or Labor Capital Labor Intensive Production Types of Equipment Automated feed, cutting, assembly, Manual equipment, little or no sewing and packing automated processes Ownership Profile Small family owned or shareholders. Small family owned or in some cases Either is possible but all are small limited shareholders companies BEE Compliant Generally not and not interested in Yes, deemed important and take credit becoming so for unskilled labor production Productivity High Low Unskilled labor 30% 75% Semi-skilled labor 60% 20% Technical/Professional 7% 3% labor Admin 3% 2% Quality of Production High Medium to Low Defect/Rework Rate Less than 1 % 5 % or more 108 4.69 The following are the main advantages and disadvantages faced by the industry. Table 4.19: Major Characteristics of the South African Apparel Industry Advantages Disadvantages Diversified product range; High relative wage, enforced by unions in urban areas; Responsible apparel Low productivity, especially compared to its Asian rivals - (22-24 production is widespread; minutes to make a pair of 5-pocket western jeans, compared to 11- Unskilled, but abundant 12 minutes in Asia); labor force; Lack of productivity incentive for workers. Wages are on a flat rate Sophisticated and with minimum wage base or negotiated contract rates at least above established infrastructure; legislated minimum wage rates (unions prevent introduction of Relatively strong piecework pay scales); infrastructure. Lack of marketing skills, technical, financial and labor management within the management structure and thus those with such skills command a premium salary 4.70 The labor cost disadvantage faced by South African producers, vis-ā-vis principal competitors, is shown in the following table. Table 4.20: Labor Cost Comparison for the Garment Sector Countries US$/hour Bangladesh $0.21 Pakistan $0.26 India $0.27 Nepal $0.31 Sri Lanka $0.35 China $0.68 South Africa $1.38 - $2.22 4.71 South African garment manufacturers additionally suffer from relatively high absenteeism, which generally ranges from 7.9% to 10%, against an international average for the garment sector of around 4.7% (absenteeism in the textile sector is over 4%). 4.72 South Africa also faces some transport cost disadvantage, considering the internal distnces (e.g. Cape Town to Johannesburg 1,500km). While transport costs act as a natural barrier to imports of bulky goods, especially into Gauteng, they also add to the costs of South Africa's exports. a 4.73 The garment industry faces not only severe legitimate competition but, also, some alleged under-invoicing and dumping of imports. The most prominent competition in the garment sector comes from China. This is true, not only for South African manufacturers, but also for all garment producers in Africa and parts of Asia. Taking into consideration China's investments in capital equipment, inexpensive but productive labor force, and other factors that impact on the operating environment, it is becoming increasingly clear that garment manufacturers, particularly those with products targeted towards the low-end of the market, must quickly reconsider their product and market strategy (refer to the table below). 109 Table 4.21: Comparative Features Of Garment Industry in South Africa and China South Africa China Wage ($/hour) $1.38 - $2.22 $0.68 Subsidies DCCS (15% - 25%) 13% Unions Unionized (>60%) Non-Union Input material Imported (duty rates - yarn: Locally available 15%; fabric: 22%) Equipment Lifecycle 10.5 years 4 years Utility pricing Competitive (R0.0234/Kwh) Free (State/ municipal companies) Interest rate 12% - 14% 1.5% Currency Floating Fixed ­ undervalued * Exchange rate: $1 = R6.1 4.74 As mentioned, administrative overhead costs are surprisingly high at 20% of the entire value chain for garment production. A closer look suggests that 55.8% of the overhead costs are accounted for by salaries, while an additional 10% is applied to marketing. Even in Lesotho, where there is a high cost foreign supervisor-to-line worker ratio (1:4), the salary and benefit burden on administrative overhead is only 16.8% or equal to $0.02/unit of output, as compared to $0.48/unit of output in South Africa. Excessive management remuneration emerges as a factor contributing to competitive disadvantage.101 4.75 Some garment manufacturers have accepted that they cannot compete on a commodity level and have targeted small, upscale, or otherwise niche markets. However, the industry is not yet fully targeting the right markets to take advantage of local high quality yarn. There is still a significant mismatch between cotton production and textile/garment production and a need to define a strategy that takes account of the comparative advantage of South African producers. Government Incentives: the DCCS 4.76 Government incentives were introduced to stabilize the industry in the face of competition and promote exports. The main incentive introduced was the DCCS (Duty Credit Certificate Scheme), providing duty exemption permits on up to 35% of export value,102 principally as a means of stimulating exports by reducing the cost of inputs. The following diagram depicts the working of the scheme as it would apply to an apparel manufacturer. Participation in the DCCS was linked to a spending on human resource development and improvements in operational efficiency, to be certified by independent auditors.103 In the original formulation, certificates were not meant to be tradable. 101Management salaries range upwards from $US30,000 a year. Benefits include regular bonuses, entertainment and vehicle allowances, housing, medical and other benefits. 102The rebate provision is contained in the Customs and Excise Act 91 of 1964, rebate item 460.11/00.00/06.00 and a retrospective provision for the refund of customs duties paid on date of importation. 103An enterprise submitted a request for a DCC upon exporting qualified goods outside of SACU. The value of the certificate was a percent of the export value. Depending on the classification the beneficiary was also required to allocate a percentage of export revenue toward training. 110 DIAGRAM 4.3 Duty Credit Certification Scheme (DCCS)* * Department of Trade and Industry (DTI)/ International Trade Administration Commission (ITAC) DCC Duty Request Rebate Local Sell Apparel Purchase CTFL SETA Retailers DCCs Manufacturer Training Import cheap Export Outside garments ofSACU *DCCS ended in 31 March 2005, but an extension has been granted through the South African Interim Clothing and Textile Scheme Market CTFL: Clothing, Textile, Footwear, Leather Global Development Solutions ,LLC 4.77 The DCCS has been modified under the Interim Clothing and Textile Scheme to avoid some unintended effects. In practice, exporters using the DCCS were able to sell their entitlements to retailers, which encouraged exports whilst also promoting imports of competing finished garments, so that the link between exporting and lowering the cost of inputs (for export) was broken. Thus, exporters could simply receive a subsidy whilst other firms could import finished goods, including low end items that were at times supplied at below cost prices. The following were DCC certificate values in 2005: Table 4.22: DCCS Certificate Value 2005 Duty Rebate Rate Category Category B A Clothing 15% 25% Household textiles 12% 17.5% Fabric 8% 12.5% Yarn 5% 8% Category A: Companies with export turnover <15% of total sales Category B: Companies with export turnover >15% of total sales 111 4.78 The percent of export value allowable for duty exemption has fallen from the initial 35%, as have the duty rates applicable. The highest rate of support is for clothing (garments), followed by fabric and, thirdly, yarn. For firms with more than 15% of their turnover in exports, the value of the subsidy through the DCCS comes to the [imported input % of exports] times [% rate of exempted duty]/ total export revenue. For category B clothing firms with a current 40% import tariff, this would be equivalent to a subsidy of about 10% of export revenue. The effect on exports would depend on how far the difference of the import parity price and domestic wholesale price was less or more than the effective export subsidy. A 10% subsidy would probably only be effective for products in which South Africa was not far from being competitive. Nevertheless, surveys have shown that many firms are dependent on it for survival. 4.79 The intended training and performance requirements of the DCCS have not been complied with, while the Sector Education and Training Authorities (SETAs), whose mandate is to advance workplace security and productivity, appear to have had limited success, lacking credibility with manufacturers. Interviews with garment manufacturers suggest that what little in-house training that does take place, particularly for unskilled workers, is limited to instructions on how to fill out job cards, simple arithmetic, and rudimentary dexterity training.104 Employers are also reluctant to invest in training because of high labor turnover among unskilled and semi-skilled workers, partly due to high transport costs.105 The international garment market 4.80 South Africa enjoyed a steady growth of apparel exports to the U.S. (from 5.47 to 28.1 million square meters equivalent) between 2001 and 2003, although the volume of exports is only one-fourth that of Lesotho. Since its manufacturing costs are high, a partial explanation for its export performance could be the implicit subsidy provided by the DCCS scheme, as well as AGOA. 4.81 Contrasting South Africa against three other successful AGOA based apparel exporters, Kenya, Lesotho and Mauritius, there is marked difference in the product strategy. Whereas, South Africa has focused on a diversified range of products, Lesotho has concentrated on 11 product categories. South African apparel exports to the U.S. cover as many as 38 product categories, in which they are dominant in 24 product categories among AGOA apparel exporting countries. Furthermore, South Africa's principal export category is non-woven fabrics (14.4%). 104Manufacturers interviewed tended to attribute this to relatively high labor turnover rates. 105Bezuidenhoust, A., G. Khunou, S. Mosoetsa, K. Sutherland and J Thoburn, 2003, `The Impact of Restructuring in the Textiles Sector on Households in South Africa', Globalization of Poverty Discussion Paper No. 13., Norwich: Overseas Development Group, University of East Anglia. Employers tend to lay off the least skilled first. Over 70% of those laid off were unable to find work again for sometime, which reflects the absence of transferable skills acquired during the period of employment 112 Table 4.23: Main African exporters of garments to the U.S. Under AGOA (million sq mt equivalent) 2001 2002 2003 Ghana na 0.548 6.108 Kenya 14.143 33.901 48.85 Mauritius 5.696 17.795 19.952 Madagascar 16.455 18.348 42.856 South Africa 5.472 16.538 28.064 Swaziland 2.104 22.555 45.278 Malawi 1.01 3.354 6.853 Lesotho 29.716 83.143 98.574 Namibia na 0.269 7.986 Source: US Department of Commerce 4.82 More recent data are more troubling. In 2004-05 exports to the USA market fell in nearly all product categories, with the exception of special fabric and playsuits/sun suits. In some of the smaller lines (e.g. cotton coats, nightwear and bedding) exports to the USA were all but eliminated. The ending of the MFA and the appreciation of the Rand from 2003 have ratcheted up import pressure, while adjustment has been slow, particularly in moving out of "commodity" type garments where buyers and consumers are extremely price sensitive to higher value added products. 4.83 Apart from the USA market, the garment sector is also driven by the EU market. EU buyers have tended to demand smaller volumes of product than U.S. buyers. While markets in the U.S. garment sector are driven by price, EU buyers, while seeking low prices, have tended to be more loyal to their suppliers. Table 4.24: Characteristics of European and American Brand Name Holders Europe America · Small volume · Large volume · Low prices · Lowest price · Loyal to suppliers · Purchasing relationship driven by price · Designate fabric supplier The Garment Industry: Adjustment to International Competition 4.84 Textile wages are well above minimum wage, as a result of union collective bargaining agreements. Working conditions are closely monitored by the unions, mostly in the urban areas. This has led to some small enterprises and foreign investors establishing operations in decentralized zones to take advantage of the lack of union activity in these areas, in order to negotiate lower wage levels. Despite relocation to low labor cost areas, competition with imports of large quantities of low price commodity garments, nevertheless, remains extremely difficult. This applies both to imports from Asia and from Lesotho and Mauritius where companies have similar capacities and productivity while maintaining lower cost of labor in non-unionized mills. 113 4.85 One result of this has been the development of a `putting out' (subcontracting) system to `cut, make and trim' units (CMTs), ranging from micro units to medium sized workshops, employing 3 to 100 workers.106 There are thought to be over a thousand of these units employing perhaps 20,000 or more workers The central factory retains only purchase, design, marketing and sales functions. There are at least three categories of CMT producers: garment manufacturers operating in the Western Cape, who tend to fall into the specialized high-end product market, manufacturers in KZN who tend to produce non-specialized `commodity' type products, and small, medium and micro-enterprises that subcontract to local `commodity' type producers or work on ad hoc jobs. 4.86 The CMTs outsource units are often informal, unregistered and with casual non- unionized and low-paid labor. The implication is that the garment industry has, to an extent, adapted to competition through labor cost flexibility, rather than management restructuring or equipment upgrading/ product innovation. The CMTs, being outside the regulatory system, are not, in themselves, capable of upgrading or innovating, which depends on the contractor units reinvesting whatever profits they can earn. There is also a danger of a `race to the bottom' as the outsourcing units become simply cost centers completely dependent on, and under pressure from, a single customer. Nevertheless, such mode of production has the advantage of (a) high flexibility, and (b) low cost labor. 4.87 Not all CMT units share the cost advantages normally associated with this form of subcontracting. However, as Rogerson has shown107 from a survey of small entrepreneurs in the Witwatersrand, there are also many examples of success. The majority of enterprises in his survey, one third working primarily as CMT units, were found to have achieved huge expansion of their asset bases (in many cases from less than R10,000 at start-up to over R1 million over 5 to 10 years), with commensurate increase in numbers of workers and machinery. The key to this success appears to have been the level of education among the clothing entrepreneurs surveyed, all of whom undertook various business and technical training programs and made a conscious move to relocate to formal inner-city premises that helped provide greater access to markets, and hence, the resources with which to upgrade machinery. As seen from another survey, some of the more `elite' CMT units have been able to produce for the export market, their capacity being sold as part of the regular capacity of larger exporting companies. In a number of other cases108, where regular CMT units have found it difficult to cope with production peaks or awkward styles and run lengths, larger exporters have been able to make use of them to handle instead their entire domestic market business109. In this way, the CMT outsourcing units have shown that they can provide an entry point for 106For a description see e.g. Cristi van der Westhuizen: "Trade and Poverty: A Case Study of the SA Clothing Industry". 107Christian M Rogerson: "Successful SMEs in South Africa: the case of clothing producers in the Witwatersrand", Development Southern Africa, Vol.17, No.5, December 2000. 108Especially in the Durban Metro. 109See Peter Gibbon: "South Africa and the Global Commodity Chain for Clothing: Export Performance and Constraints", in `Clothing and Footwear in African Industrialization' (Ed: Dorothy McCormick and Christian Rogerson), Africa Institute of South Africa, 2004. 114 micro/small firms, including black owned businesses, if efforts are made to derive the benefits of agglomeration economies. 4.88 With regard to the garment industry as a whole, seasonal mass market fashion is the segment where most competition takes place. It is characterized by very short lead times, low demand predictability, and sourcing generally by brand name holders and retailers through purchasing agents and intermediaries. Foreign, mainly U.S, buyers control the value chain by designating approved fabric suppliers. As a result, relationships between retailers and producers have become far more restrictive. Even accepting the possibility of increasing flexibility and lowering labor costs through the CMT units, the competitive advantage of South African suppliers is generally not in this segment, but in the higher value traditional segment. Traditional items such as suits and high value added shirts tend to be characterized by much longer lead times, higher demand predictability and direct sourcing by retailers. Key Market Drivers and Strategic Options for the Industry and the SMMEs 4.89 From the above analysis of each stage of the cotton textile sector, certain key options emerge for the increase of SMME output, including that of black-owned businesses. These are to: · Increase the irrigated area and production of high quality cotton, both overall and for black-owned farms; · Improve fabric production quality and delivery times throughout the industry to facilitate higher local content in economic garment production; · Encourage efforts to reduce costs and increase flexibility through outsourcing of garment production to viable small CMT units; · Address the skill needs of the sector through revamped training programs; · Improve overall business prospects by repositioning garments for high-end markets using high quality cotton fabric; · Explore opportunities to meet medium term demand for knitted fabrics from garment exporters in Lesotho. 4.90 The increase in irrigated area depends on identifying land areas that are both suitable for irrigation, or already having irrigation facilities, and not subject to communal land titles unless such land can be well managed. The land areas also need to be located in adequate proximity to centralized ginneries. 4.91 The garment industry must focus on smaller niche markets for such high quality products, up-market fashion products, etc, with short production runs and longer lead times where South African manufacturers do have some competitive advantage, and where progress has already been made. 4.92 The low level of skills reflects a loss of skilled workers through natural attrition and failure to maintain a strong base of skilled workers. A significant increase in workplace apprenticeship programs and other skill development programs is necessary to 115 improve the current supply of skills but, clearly, what is needed is a comprehensive strategy that deals with ongoing unemployment. Employers contributing to the training levy and registered with the Government have cited the complicated procedures required to access the fund, in addition to the time lag in trying to recoup their investments in training from the funds. 4.93 Access to finance is a further issue for the textile sub-sector as a whole. As with other industries, `at risk' finance is relatively unavailable for textile producers. Banks regard the textile industry as protected and of doubtful long run viability. This results in premium interest rates, often marked up for both industry and individual business risk, and the requirement for onerous performance guarantees and collateral, out of proportion with real risk levels. Typically, banks require order contracts prior to approving a loan so that manufacturers are unable to borrow in anticipation of or to facilitate the winning of contracts. This problem is most acute for SMME businesses. 4.94 A final consideration is that, in some areas of the textile sector facing enormous competitive pressures, the objective of efficiency may not be consistent in the short run with the objectives of maximizing employment or black empowerment. In spinning and weaving, economies of scale are important and manufacturers have invested in capital equipment and less labor-intensive technology to help improve the quality and consistency of their products to target the high-end niche market. As competitive pressures intensify, these trends are likely to strengthen, putting an increasing premium on highly skilled inputs and production methods, and increasing the difficulty for BEE firms to integrate into the product supply chains. The most feasible entry points, therefore, are in the stages where economies of scale are less important. Implications for future assistance programs 4.95 A number of existing or recent assistance programs could be used to support particular entry points into cotton and garments production. In terms of DTI's own programs, this assistance would be principally for garments. From the value chain analysis and the more generic survey results the type of assistance that is needed could be categorized as: a) business startup support (to help new CMT and other units get into production); b) export market promotion; c) skills development (training in technical and management processes); d) production assistance (to increase productivity), and e) general business linkage and marketing support (to diversify from subcontracting). In addition, access to small-scale or micro finance may be necessary in many cases. 4.96 The following table identifies current and recent DTI-associated assistance schemes and those among them which could in principle provide support to garment production. The most applicable schemes are flagged. 116 Table 4.25: DTI and Associated Support Schemes applicable to CMT unit production of Garments Support Scheme Sector Description Yes/No Investment grant for building, SMEDP General (SME) equipment Grant for workplace projects to Workplace General SME and improve labor /management operating ChallengeFund LSE practices. Critical CIP Infrastructure Grant for infrastructure improvement Duty free operations and simplified Export -oriented customs procedures (Coega, E. IDZ industries London, Richards Bay) General foreign Grant to international firms FIG investment establishing projects in SA Selected strategic SIP industries Investment- based tax allowance Manufacturing and Advisory services to producers and CF services business service providers Manufacturing and Matching Fund for cluster/ network SPF services building programs Cost sharing grant for training and BBSDP Black owned firms mentoring Grant for skills development for work General (recipients force training and promotion of long SSP of SMEDP funding) term employment TISA/Customized Government/industry participatory Sector Programs Priority Sectors action groups and export councils National Pavilions; trade mission Export/ inward support; outward and inward TISA/ EMIA investment investment support; exporter training Funding of RFIs and MFIs, institution, Khula Enterprise group and individual bank, guarantees, Finance SME sector equity finance Direct small-loans to very small firms Micro Apex Fund Microenterprises focusing on black-ownership Business development centers, linkage program, information and SEDA small manufacturing advisory centers, tender business support SME sector advice centers GODISA SMEs Multisector Intensive startup support to small Incubation firms in Centers six incubators (mining, biotech, ICT, flowers) plus two technical centers Support SMEs and LSEs Reducing costs of R & D and program for multisector product innovation. Financing for Industrial entrepreneurs in competitive 117 Table 4.25: DTI and Associated Support Schemes applicable to CMT unit production of Garments Support Scheme Sector Description Yes/No Investment grant for building, SMEDP General (SME) equipment Grant for workplace projects to Workplace General SME and improve labor /management operating ChallengeFund LSE practices. Critical CIP Infrastructure Grant for infrastructure improvement Duty free operations and simplified Export -oriented customs procedures (Coega, E. IDZ industries London, Richards Bay) General foreign Grant to international firms FIG investment establishing projects in SA Selected strategic SIP industries Investment- based tax allowance Manufacturing and Advisory services to producers and CF services business service providers Manufacturing and Matching Fund for cluster/ network SPF services building programs Cost sharing grant for training and BBSDP Black owned firms mentoring Grant for skills development for work General (recipients force training and promotion of long SSP of SMEDP funding) term employment TISA/Customized Government/industry participatory Sector Programs Priority Sectors action groups and export councils National Pavilions; trade mission Export/ inward support; outward and inward TISA/ EMIA investment investment support; exporter training Funding of RFIs and MFIs, institution, Khula Enterprise group and individual bank, guarantees, Finance SME sector equity finance Direct small-loans to very small firms Micro Apex Fund Microenterprises focusing on black-ownership Business development centers, linkage program, information and SEDA small manufacturing advisory centers, tender business support SME sector advice centers Innovation industries (SPII) 4.97 It is notable that this wide range of available programs does not include more than one dedicated to skills development/training support. Three of the above programs which did provide training grants have since been discontinued (CF, SPF, and BBSDP). Up to now the training initiative has been implemented largely through the SETAs (see chapter 5), but innovative and accessible specific skill formation grant schemes targeted to the 118 needs of industry may well be necessary as part of the overall strategy to upgrade South Africa's labor force. C. AUTOMOTIVE COMPONENTS VALUE CHAIN 4.98 Since 1995, the automotive sector, like textiles, has been the recipient of sector specific incentives aimed at improving competitive efficiency and promoting employment down the value chain. At one level, this effort has been perceived as a significant success, with exports of motor vehicles, parts and accessories growing some 29% between 1997 and 2001, more rapidly than any other sector in the economy.110 At another level, bearing in mind the size of the incentives provided and the extent of employment actually generated, there is some question as to the cost effectiveness of the effort and the whether its objectives have fully been realized. From the perspective of the potential integration of SMME suppliers, the economics of Tier 2 and Tier 3 manufacturers (producing relatively less sophisticated components and facing lower entry barriers for smaller firms) is of specific relevance. For this reason, the analysis below focuses on two products: leather seat covers (normally Tier 2), which is one of the largest market drivers in the auto components sector,111 and forged steel towing balls (Tier 3), which is one of the few products that utilize locally available raw material (steel), being targeted for both local and international markets.112 Sub-sector profile 4.99 The automotive sector in South Africa had its beginnings in the 1920s. Starting in the 1960s there were a series of programs to increase local content and in 1989 the policy shifted from import substitution to export promotion. In 1995, the Motor Industry Development Program (MIDP) was launched as part of this effort. In 2002, the automotive manufacturing industry (including Original Equipment Manufacturers and component manufacturers) contributed 6.3% of GDP and was the largest manufacturing sub-sector. Its contribution to exports almost doubled in the period 1997-2001 to 9.75% and the sector receives high inflows of investment, including about US$R1.0 billion anticipated in 2005113. The component industry is responsible for over 60% of the exports of the industry, increasing from virtually zero in 1995 to US$1.56 billion in 2004.114 110See: David Kaplan: "Manufacturing Performance and Policy in South Africa ­ A Review" (Paper produced for the TIPS/DPRU Forum 2003); University of Cape Town, September, 2003. 111Catalytic converters and leather seat covers are the two dominant products in component exports. Production of seat covers has reached 300 units/day for the domestic market and 3,800 units/day for the export market, with an annual output of approximately 1.1 million units. 112The selection of these two products was also recommended by NAACAM, specifically because the production and supply chain for these two products embodied many of the problems and challenges faced by in South Africa's automotive component sector as a whole. 113Tralac News, `The Automotive Sector Needs a Plan to Replace MIDP', May 2005. 114UN Trade Statistics. 119 Table 4.26: Employment and Turnover per Size, Auto Industry, South Africa, 2004 Type of Company Average Average Employees Average Turnover (Number of Employees per category per category employees) per Category excluding OEMs excluding OEMs (Total and manufacturers and manufacturers Population) of trucks and buses of trucks and buses (R million) (R million) Micro (up to 10) 6.3 6 4.5 Small (10-50) 27.6 28 8.9 Medium (51-250) 117.1 117 70.2 Large (251-999) 436.7 440 291.7 Extra Large (>1,000) 2,585.9 1,715 1,204.0 Total Number 111,063.0* 76,911** 48,990.0*** Source: DTI. * Total direct employment for the sector ** Total employment for the component sector excluding OEMs, truck and bus manufacturers *** Total component sector turnover 4.100 The expansion of the component industry as well as the creation of new service stations has had a beneficial impact on job creation. An estimated 111,000 workers were directly employed in the auto sector in 2004, out of around 263,000 employed directly and indirectly.115 4.101 South Africa is the 18th largest manufacturer of vehicles in the world and represents 80% of Sub-Saharan Africa's vehicle output. All major motor manufacturers are represented: BMW, Delta, Fiat, Ford, Nissan, Toyota, VW and DaimlerChrysler, in addition to major component manufacturers such as Lear, Valeo, Faurecia and Arvin Exhaust, which have also established production bases in the country. 4.102 The auto industry, which had received heavy protection from import tariffs and local content requirements prior to trade liberalization in the 1990s, had to undergo an adjustment with South Africa's reintegration into the global economy. The Motor Industry Development Program (MIDP) was designed to help the industry adjust and increase competitiveness through a program that combined a phased reduction in import tariffs with an export-import complementation scheme of Import Rebate Credit Certificates (IRCCs), duty drawbacks for exporters and duty-free allowances on imported components, so as to promote higher levels of specialization and exports. The idea was that this would allow for increased economies of scale and hence to sustainable gains in efficiency and productivity. The Impact of the MIDP 4.103 The MIDP support scheme, introduced in 1995 and managed by the DTI, has been critical to the growth of the auto sector through: (a) its direct effect on imports and exports of vehicles and their component parts, and (b) its possible effect up the supply 115Municipality of East London and South Africa Department of Trade and Industry (DTI). 120 chain on raw material prices.116 The MIDP encouraged competitiveness in the sector through gradual reduction of import duties from levels of over 100%, down to levels accepted by the General Agreement on Trade Tariffs (GATT).117 Under the MIDP, firms may reduce duty payable through three mechanisms - the Duty Free Allowance (DFA), an import/export complementation scheme and the Productive Asset Allowance (PAA). 4.104 Mechanism of MIDP: Imports of passenger cars and light commercial vehicles have been subject to the tariffs shown below. The idea is that the tariffs may be reduced, and even nullified, using any or all of the three methods: DFA, IRCC and PAA. Under the DFA, 27% of the wholesale value of the vehicle may be imported free of duty. Under the IRCC scheme (import/export complementation), for every Rand of CBU or components exported, imports worth an equivalent percentage of the value of the local content may be brought in duty free. The certificates may be sold by the exporter to a registered importer. There is also a small fiscal duty which may be reduced using the IRCC system. The PAA is a scheme whereby investors in new plant and equipment can qualify for a duty credit certificate up to 20% of the value of their investment over a five year period. DIAGRAM 4.4: MIDP Import Rebate System End Auto component Import Manufacturer Price - R83 Apply or Leather Trade Rebate Seat Cover Rebate Sales Price- - R100 Duty Inclusive Price - R100 Duties Rebate - MIDP Local (R17) Import Content Rebate And Value Credit Added Automotive Certificate Export Imports Market 116The MIDP has been under scrutiny for compliance with WTO rules. Australia has challenged it under WTO as unfair practice, and it is possible that WTO will eventually rule adversely on it unless it is appropriately reformed. 117There is a schedule of reducing import tariffs for complete built up (CBU) and complete knocked down (CKD) components for cars and light commercial vehicles. CBU medium and heavy vehicles (above 3.5 tons) may be imported at a duty of 20% and components for these vehicles may be imported duty free, except for tires, which attract duty of 6% from the EU, or 10% from elsewhere. Semi-knock down (SKD) vehicles may be imported at the CBU rate, but imported second hand vehicles are not allowed. 121 Table 4.27: South African Automotive Import Tariff Schedule Year CBU (%) CKD (%) 2004 36 28 2005 34 27 2006 32 26 2007 30 25 2008 29 24 2009 28 23 2010 27 22 2011 26 21 2012 25 20 Source: NAACAM, 2004 4.105 Because the import duty rebates apply only to the local content contained in the exports, OEMs have the incentive to maximize local content in the exported products. The verification process118 can, however, be very onerous when there are sub-assemblies and large bills of materials. 4.106 The value of the IRCC to the exporter has been calculated as follows: Table 4.28: Effective Protection to Exports by the IRCC Facility IRCC used for: Vehicles Components Vehicle imports 40% 26% Component Imports 30% 30% Source: Flatters. F, (2002) `From Import Substitution to Export Promotion: Driving the South African Motor Industry' (Quoted in Kaplan; op cit) 4.107 In its working, the IRCCs seem to have resulted in some anomalies. This is because the rebate is based on local content and creates incentives for chain leaders, i.e. suppliers (such as leather tanneries in the case of seat covers) and purchasers of local content (mostly OEMs), to fix prices at levels that maximize the rebates within the delivered price of goods. This does not benefit all players in the chain equally. In the case of seat covers, for example, the component manufacturer is a passive recipient of whatever price is arranged between the OEM and the tanneries. This non-transparent situation affects negatively, not only the manufacturers dependent on MIDP rebates for export competitiveness, but also other downstream manufacturers that rely on domestic raw materials. Steel pricing has similarly been subject to these criticisms. 118To verify the local content, all suppliers must provide their customers down the chain with a DA190 form, itemizing each piece of a component down to the original local raw material or imported piece. A quarterly audit is required by external auditors. 122 4.108 Assessment of the impact of MIDP119 as a whole, suggests that the benefits realized have not necessarily justified the high costs involved. While the MIDP has encouraged the restructuring of the South African auto industry, by encouraging a shift in production patterns towards exports and greater product specialization, it is not clear that the industry has achieved international competitiveness. This is equally true of vehicle assembly and component production. Subsidies provided to investment and exports have been substantial over the years, resulting in a high domestic resource cost of foreign exchange120 and, hence, encouragement of uncompetitive investments. Against this background, employment creation has been relatively modest in the components sector (about 6% between 2000 and 2004), and negative in vehicle assembly (a decline of 17% between 1995 and 2000, and stable thereafter). 4.109 The MIDP has gone through two reviews to fine tune and extend the program through to 2012. Earlier this year, the Government initiated a third formal consultative review of the program, the impact of which is currently being assessed by an external audit firm. The outcome of the review will allow for policy changes to fine tune MIDP's incentive package for 2006 and beyond. 4.110 Against this background, we take a look at the economics of producing leather seat covers and forged tow-balls through analysis of their value chains. Leather Seat Covers: Market Structure and the Supply Chain 4.111 A sample supply chain for auto leather seat cover from leather supplies to the market is provided in the above diagram. The seat cover manufacturer is in the middle of the supply chain but, in fact, the OEMs control the chain, negotiating most transactions from tanning services for a fee to provision of import materials. The OEM is the sole contact point within the marketplace and, in cases where a seat cover manufacturer supplies to export markets outside the OEMs supply chain, it is under the OEM's name. Furthermore, what little control the component manufacturer has over the marketplace is diminished by the fact that its domestic suppliers of leather are aware that, under the IRCC system, component manufacturers are under pressure to maximize local content, and they can use this leverage to increase the price of raw material supplies. 4.112 This market structure suggests that South African component manufacturers have been unable to create their own market positioning in the global marketplace outside the OEM's supply chains. This is partly evident from the fact that 75% of the exports of leather seat covers are concentrated in a single country, Germany. The OEM automakers maintain tight control over their supply chains via creation of specific production platforms that make it difficult for smaller suppliers to switch client. The ones that do have such capabilities to be relatively platform-flexible are usually multibillion, 119See for instance: F. Flatters "The Economics of MIDP and the South African Motor Industry", 5 November 2005; F. Flatters and Nnzeni Netshitomboni: "Trade and Poverty in South Africa: Motor Industry Case Study", (Draft) 20 March 2006; and TSG: "The Potential for Regional Integration of the SADC Motor Industry", August 2004. 120Declining somewhat from 1.60 earlier to 1.29 currently. See: F. Flatters, Nov. 2005; op.cit. 123 multinational auto component manufacturers that have emerged from long term supplier relationships with particular auto manufacturers. 4.113 Dependence on OEMs is not necessarily an adverse situation, as there is much to be gained from being part of the OEMs global purchasing process. Graduating to global OEM supplier status brings with it externalities such as global reputation and acquaintance with the workings of the global market that would, otherwise, be difficult to obtain. Suppliers deemed `appropriate' for access to global supply chains are, however, those who survive the cost squeeze exerted by OEMs. In this sense, becoming a part of an OEM global supply chain will, in all likelihood, be achieved on OEM terms and conditions. Nonetheless, the value-chain enhancement strategy from the industry and government perspective should be to support local component manufacturers able to access global supply chains of OEMs and to become globally competitive in the longer run. 4.114 The evolution of the supply chain for leather seat covers from the recent past is indicative of the potential evolution of the component manufacturers' supply chain in South Africa. The tiering of the supply chain has been evolving from the late 1990s when tanning, finishing, and cutting of leather seat covers were done within the component manufacturing company, while sewing and assembly of seat covers was done by the OEM (See Diagram 1). 4.115 The evolution of component manufacturers from local to global suppliers would face considerable competitive challenges, given the trend towards consolidation in the vehicle industry.121 This consolidation casts OEMs as brand owners, rather than brand manufacturers, so they will be looking for even more systems and modules from suppliers which, in turn, will shift consolidation pressures towards suppliers. It is predicted that only 25 to 30 global Tier 1 suppliers will exist in 2010, compared with over 600 operating today. Tier 2 suppliers also face consolidation in the next 10 years. Their universe is predicted to shrink from 10,000 to 800 in ten years to 2010. INTEGRATED VALUE CHAIN ANALYSIS FOR SEAT COVERS 4.116 Leather seat covers come from bovine hides, after passing through a tanning process in local tanneries. Tanned leather is then either directly exported (50% is exported to non-automotive industries such as furniture), or is supplied to local manufacturers who, in turn, supply OEM's with the final product under OEM specifications for design, coloring, and quality.122 The tanneries are multinational companies that produce to international quality levels, so the following analysis will focus on competitiveness beyond the tanning stage. 121According to a survey from Price Waterhouse Coopers in Automotive Industries Journal, in 1999, 23 car companies were major players. By late 2001 there were only nine major players in the global market, and that number is projected to drop to five by 2010. 122Leather sewing industry for seat covers employs over 3,000 workers, and an additional 3,749 workers in the tanning industry. 124 4.117 Tanned leather is purchased in a pre-cut form from tanneries, after which it goes through: Sewing and assembly > Inspection > Finishing/Packaging. 4.118 Value chain analysis was conducted first for a product supplying export markets, and second for a product to supply domestic OEMs. The first case focuses on a local supplier which exports to Europe. Seat Covers for Export 4.119 South Africa exported an estimated $491 million worth of leather seat covers in 2004, the bulk ($368 million) to Germany. The typical medium size local company, which employs 300 people, is dedicated to exports for the automotive sector and is a result of an OEM spin-off. The supplier exports under the name of the OEM, which makes it possible for the OEM to claim IRCC benefits under the MIDP. The value chain shows that the highest value adding component is material inputs (76.0%), followed by overhead (11.6%) and transportation (8.5%). Locally purchased tanned leather constitutes the bulk of value added (87.2%) at the material stage. Imported materials such as resistors, foams, and plastics are imported at zero rated duties. The manufacturer of seat covers, in fact, has no control over the price and flow of local leather and imported inputs which is within the OEM's supply chain and is tightly controlled. The manufacturers are positioned in the supply chain as sub-contractors and, as a consequence, they are generally driven by price and quality, with little scope for innovation. DIAGRAM 4.5: Value Chain for Leather Seat Cover Exported to the EU under the name of the OEM Imported materials include electrical resistors, foams, interlayer, and extruded plastics. Local Imported Fixed Profits Waste Leather Inputs Costs 87.2% 12.8% 43% 49% 8% Raw Material Sewing/ Finishing/ Overhead Transport Assembly Packaging/ 76.0% 3.7% Inspection 11% 8.5% 0.8% Labor Maintenance Electricity Product air freighted to customers in 94.7% 5.2% 0.1% Europe to optimize delivery times of high-value items Highly labor intensive. Average of 220 man-minutes per piece. No capital invested except the building and manual sewing machines Source: Global Development Solutions, LLCTM 125 4.120 The tanneries are multinational companies which negotiate the price for leather tanning with the OEMs. This is the case in almost all automotive leather supply chains except that, in South Africa, the MIDP benefits that arise from sourcing local inputs may have led to increased leverage on the part of local tanned leather suppliers. This leverage is increased because only six tanneries exist with the capacity to supply tanned leather of sufficient quality for automotive applications, of which four appear to have a cartel-like control over pricing. This is reflected in the fact that, when compared to EU suppliers of tanned leather who sell at R195/m2, South African tanneries sell leather at an estimated 15% premium, at a price range of R221 - R229/m2, part of which is the value of import tariffs and transport costs. Table 4.29: Benchmarking leather Price in EU and South Africa (Rand/mē) EU 195 South Africa 221 ­ 229 4.121 At the same time, since exports are undertaken under the OEM's umbrella, OEMs can, and do, claim IRCC benefits under the MIDP, in this case 12.7% of the selling price which, in effect, reduces the component price and makes the leather seat covers more competitive within the OEM global supply chain. The tanneries benefit from the tariff and transport protection margin, while the OEMs claim the MIDP rebate, but it appears that, in addition, direct negotiations between tanners and OEMs allows part of the MIDP rebate to be passed back through a leather price premium. The automotive component supplier in the middle of the chain sees no direct advantage from the MIDP (apart from the fact that it is keeping the supplier's principal client in business) and has little control over its product and growth. The MIDP also creates a significant administrative burden, which is likely to have some impact on the component suppliers' overhead costs.123 4.122 Overhead cost analysis illustrates well the operating environment of component manufacturers. Overheads constitute 11% of value added, of which the bulk is profits, equal to around 5.5% of the total value added, followed by fixed costs at 43% or around 5% of total value added, and waste, which is equal to approximately 1% of total value added. The relatively low profit margin is indicative of the price squeeze on the component manufacturer. The OEM client dictates the price and source of raw material and also determines the seat cover price. The resulting profit squeeze on the component supplier constrains major equipment investment, which is reflected in the fact that 43% of overhead consists of fixed investments, such as building and facilities. As with many component manufacturers in South Africa, seat suppliers minimize capital investment to reduce risk while, at the same time, employing labor intensive production methods.124 Thus, while the MIDP has been successful in attracting foreign companies to South Africa, the level of investments in technology and know-how transfer and training is limited. 123KPMG South Africa, `MIDP must be state-of-the-art to avoid increasing risk in the motor industry". 124It was noted that the seat cover production facility was using manual sewing machines, which requires minimal amount of investments, rather than employing electric or automated sewing equipment, which could improve production efficiency, but would require substantial capital investments. 126 4.123 Labor constitutes 4.4 % of value added (3.6% for sewing and assembly, and 0.8% for finishing, packing and inspection) and its productivity is low by international standards. Table 4.30: Benchmarking Labor Productivity for Leather Seat Cover Production in India, China, EU and South Africa (units/hour) High Low Cost/Hour (Rand) India 2.7 0.75 n/a China 1.4 0.4 10 EU 0.4 n/a 12 South Africa 0.3 n/a 24 ­ 31 4.124 The sewing machines used at this stage of the process are mostly manual,125 with the aid of a small electric motor, thus the percentage value added is only 5.3% of the total sewing and assembly.126 Given the limited labor skills and manual techniques, the defect rate, while within international norms, could be reduced. Table 4.31: Benchmarking Defect Rates. Seat Covers China, India, South Africa (ppm) High Low China/India >102 10 South Africa >100 25 4.125 Chinese and Eastern European labor produce an average of three seat covers per working day, compared to two in South Africa. This, in effect, increases labor cost per unit to approximately R31/hour, significantly higher than in China and Eastern Europe. This weak performance is related to the fact that component manufacturers, tightly squeezed by their suppliers and customers, do not invest in labor training. Interventions in labor training would improve South Africa's competitiveness, especially in labor intensive manufacturing. There are also general labor productivity issues, resulting from the impact of: (a) Absenteeism which, at 8% to 10%, is significantly higher than in any globally competitive labor force caused, in part, by disturbances in the family environment due to crime and HIV incidence; (b) Labor Unions ­ labor unions are powerful and frequently paralyze transport interconnections, causing disturbances in workplace accessibility; and (c) Difficulties in Black management accession ­ Black middle management is only slowly adapting to a formerly European-dominated management culture. 4.126 Transport costs for export are significant, amounting to 8.5% of value added to Europe (Annex 14). Due to the high value of the final product and the savings on transit stock financing, air freight is the preferred mode of transportation, but this creates a significant cost disadvantage for South African component manufacturers compared to Eastern European suppliers. This geographic disadvantage, along with the other 125 Low investments in improving productivity and output is a reflection of the lack of long-term commitment by foreign component manufacturers currently operating in South Africa. 126With an average price of industrial electricity at about R0.0234/Kwh, the estimated electricity and maintenance cost is approximately R0.686/seat cover. In this context, poor output is linked to poor labor skills and productivity. 127 inefficiencies, is subject to compensation by MIDP rebates. In this particular case, the transport price gap of R220 (or 27) per seat cover between Eastern European competitors and South African manufacturers, is almost equal to the rebate available of approximately R252. From a medium term point of view, however, this off-setting of costs is unlikely to be sustainable. 4.127 There is need for more market-based initiatives to reduce transportation costs to foreign markets. One possible option could be to consolidate resources to offer OEMs warehousing in Europe, and to develop direct customer representation in Europe for South African manufacturers. This would make it possible to (a) reduce transportation costs ­ with a sufficient stock in Europe, vendors could switch from air to sea transportation; (b) Implement Just-in-Time (JIT) ­ increased proximity to customers enabling JIT to be implemented, which can contribute to significant time and cost savings; and (c) Reduce supplier delivery defaults ­ warehousing in the buyers' market would create security stocks that can mitigate any transportation failure risks. 4.128 These measures would increase transit/traveling stock financing costs but the benefits from reduced transport costs would, most probably, outweigh the cost of warehousing. Seat Covers for the domestic market 4.129 The value chain analysis illustrated below is based on a Tier 1 seat cover manufacturer selling into the domestic OEM market in South Africa. The manufacturer employs 720 people and has a presence in 34 countries. It is a multinational corporation, and, as such, already complies with all international quality standards. 4.130 The principal difference between this example, and the export supplier presented previously, lies in higher sewing/assembly costs (labor), lower overheads and lower transport cost shares. Labor constitutes approximately 12% of the value added compared to 4% above. This high labor composition in the value chain is related to both the multinational structure of the company with high overheads as well high labor costs per product manufactured. When labor productivity is factored in, this manufacturer's labor cost are significantly higher than Eastern European and Chinese manufacturers. Table 4.32: Leather Seat Cover Market Potential Pieces/day Total Local market 300 Total Export Market 3,800 Source: Global Development Solutions, LLCTM 128 DIAGRAM 4.6: Value Chain ­ Leather Seat Cover Supplied to Domestic OEMs Local Imported Leather Inputs 87.5% 12.5% Raw Material Sewing/ Finishing/ Overhead Transport Assembly Packaging/ 80% 11.8% Inspection 7% 1% 0.2% Highly quality targets achieved (at 25ppm) at a high labor cost relative to competitors Source: Global Development Solutions, LLCTM 4.131 When adjusted for labor efficiencies, the real labor cost for South Africa's component manufacturers works out to four times higher than in China, and at least three times higher than in Eastern European countries such as Romania and Bosnia. 4.132 The principal attraction for multinationals to continue to produce in South Africa, despite such low labor productivity, is clearly the MIDP rebate scheme which, in this sense, is financing inherent manufacturing labor inefficiencies. The reliance on the MIDP is reflected in a recent case where the rebates were eliminated on leather seat cover exports to Australia (as Australia had threatened to take up the MIDP rebates with the WTO), as a result of which seat cover exports to Australia came to a halt. 4.133 As the value chain in the diagram above shows, raw material supplies constitute the bulk of the value chain of the component manufacturer (70% of the total value-added structure). Multinational seat cover manufacturers are, thus, potentially in no better position than local manufacturers, vis-ā-vis raw material suppliers, if there are distortions in leather prices (e.g. through collusion with the OEMs). Interviews with the company confirmed that the tanneries engaged in cartel-like price setting, and supply leather at higher than normal market prices. 4.134 The MIDP creates room for price hikes by tanneries which increase the price of raw material, and introduces a distortion of 66% - 70% of the value-added structure of seat cover manufacturers. The domination of the tanneries sector by four MNCs also prevents SMEs and, in particular, BEE enterprises, from participating in the supply chain. 129 This situation is also partly the result of the MIDP rebate scheme which obliges OEMs to seek local sourcing through collusion with the tanneries. Such deals are most likely between a small group of larger firms. Leather seat covers ­ implications for SMME entry and future assistance programs 4.135 South Africa's automotive sector has seen significant investment and export growth through the MIDP. At the same time, the application of the MIDP has also given rise to distortions which may have undermined the competitiveness of the industry and raised barriers to entry for smaller component manufacturing units. While quality craftsmanship and technical skills are required for stitching leather seat covers, this is an area where ­ as with cotton garments ­ small production units could continue to enter the market as part of a sub-contracting supply chain. While this would, in principle also be possible at the primary end of the chain, the fact that the market is dominated by a few major tanneries, which are able to supply the entire demand for tanned leather of sufficient quality for sear cover manufacture, suggests that opportunities at this end may be somewhat restricted. 4.136 In light of this, labor productivity issues are central, and access to technical training/skills development will need to be augmented. Fine-tuning of MIDP incentives is also necessary to reduce the opportunity for raw material costs to be inflated and to allow component manufacturers to exercise greater control over their own operations in a supply chain in which material suppliers and purchasing OEMs have the market power and incentive to squeeze their margins. 4.137 The value chain analysis shows potential entry points for small and micro businesses in manufacture of leather seat covers. To the extent that competitiveness can be enhanced down the value chain, there may be potential in the long run for independent leather seat exports outside the automobile sector and protection of the MIDP, as part of the wider leather products export sector. 4.138 Assistance measures should be limited to supporting the development of subcontracting relationships between existing automotive component manufacturers and smaller production units, and also the longer run potential for direct export. The principal types of support needed would therefore be the following: a) productivity increase, b) linkage support/cluster- building and c) export marketing and promotion. 4.139 The following table shows the fit between selected current and recent DTI and associated assistance schemes and the development of leather seat cover production. The most likely applicable schemes are flagged. 130 Table 4.33: DTI and associated Support schemes applicable to production of leather seat covers Support Scheme Sector Description Yes/No Investment grant for building, SMEDP General (SME) equipment Grant for workplace projects to Workplace General SME and improve labor /management operating ChallengeFund LSE practices. Critical CIP Infrastructure Grant for infrastructure improvement Duty free operations and simplified Export -oriented customs procedures (Coega, E. IDZ industries London, Richards Bay) General foreign Grant to international firms FIG investment establishing projects in SA Selected strategic SIP industries Investment- based tax allowance manufacturing and Advisory services to producers and CF services business service providers manufacturing and Matching Fund for cluster/ network SPF services building programs Cost sharing grant for training and BBSDP Black owned firms mentoring Grant for skills development for work SME - recipients of force training and promotion of SSP SMEDP fund employment TISA/Customized Government/industry Action groups Sector Programs Priority Sectors and export councils National Pavilions; trade mission Export/ inward support; outward and inward TISA/ EMIA investment investment support; exporter training Funding of RFIs and MFIs, Khula Enterprise institution, group and individual bank, Finance SME sector guarantees, equity finance Direct small-land to very small firms Micro Apex Fund Microenterprises with focus on black ownership Business development centers, linkage program, information and SEDA small manufacturing advisory centers, business support SME sector tender advice centers GODISA SMEs Multisector Intensive startup support to small Incubation firms in Centers six incubators (mining, biotech, ICT, flowers) plus two technical centers Support program SMEs and LSEs Reducing costs of R & D and for Industrial multisector product innovation. Financing for 131 Table 4.33: DTI and associated Support schemes applicable to production of leather seat covers Support Scheme Sector Description Yes/No Investment grant for building, SMEDP General (SME) equipment Grant for workplace projects to Workplace General SME and improve labor /management operating ChallengeFund LSE practices. Critical CIP Infrastructure Grant for infrastructure improvement Duty free operations and simplified Export -oriented customs procedures (Coega, E. IDZ industries London, Richards Bay) General foreign Grant to international firms FIG investment establishing projects in SA Selected strategic SIP industries Investment- based tax allowance manufacturing and Advisory services to producers and CF services business service providers manufacturing and Matching Fund for cluster/ network SPF services building programs Cost sharing grant for training and BBSDP Black owned firms mentoring Grant for skills development for work SME - recipients of force training and promotion of SSP SMEDP fund employment TISA/Customized Government/industry Action groups Sector Programs Priority Sectors and export councils National Pavilions; trade mission Export/ inward support; outward and inward TISA/ EMIA investment investment support; exporter training Funding of RFIs and MFIs, Khula Enterprise institution, group and individual bank, Finance SME sector guarantees, equity finance Direct small-land to very small firms Micro Apex Fund Microenterprises with focus on black ownership Business development centers, linkage program, information and SEDA small manufacturing advisory centers, business support SME sector tender advice centers Innovation (SPII) entrepreneurs in competitive industries 132 Forged Steel Towing Ball: The Value Chain 4.140 As in India and China, the growth of the forging sector is closely tied to the auto and auto component sector. What is different is the scale of growth of the automotive sector in the three countries. Given the substantial scale of the automotive sector in India and China, the growth projection for these two countries dwarfs the capacity available in South Africa. The scale of production capacity available, and the narrowness of the customer base of South African manufacturers of forged products, will hinder exports into the Asian automotive industry. 4.141 The steel towing ball is produced by local component manufacturers that source steel from local suppliers, after which the steel goes through the following stages of production: Heating, forging, trimming >Punching >Machining>Electroplating>Batch marking; 4.142 South Africa was ranked the 20th largest crude steel producing country in the world by the International Iron and Steel Institute in 2004 and, as a net exporter of primary steel, it was ranked 8th in the world during 2002. South Africa is the largest steel producer in Africa, producing 57% of the total crude steel production of the continent during 2004, amounting to 9.4 million tons. 4.143 The value chain for a forged steel ball is shown below. DIAGRAM 4.7 Value Chain ­ Forged Steel Towing Ball Hot rolling steel purchased at R4,860/ton ­ steel priced at input parity prices Raw Heating Punching Coin Machining Electro- Batch Overhead Material Forging Operation plating Marking Trimming 52.0% 9.6% 3.0% 2.0% 6.0% 8.0% 1.8% 17.6% Depreciation Electricity Labor Profit Maintenan Other ce 40% 33% 27% 45% 28% 27% Source: Global Development Solutions, LLCTM 133 Analysis of competitiveness issues 4.144 Raw material constitutes 52% of value added, that is hot rolling steel sourced locally. This proportion of raw material input could be reduced, but data for the steel sector (see Annex 15) suggests that, particularly during periods of exchange rate fluctuation, there has been some `stickiness' of the pricing mechanism, whereby domestic prices have been significantly out of line with import parity prices. To a large extent, this appears to reflect the market power of the dominant domestic manufacturer. What is noticeable is the fact that the volume and scale of production continues to be low, and there is very little evidence to suggest that manufacturers are taking action or have the capacity to diversify their client and product base. 4.145 The tow ball operation has a relatively high overhead cost. According to the manufacturer, the minimum threshold profit on sales for them would be 8%, which is two to three times higher than the norm in other countries producing a similar product. This is, in part, a reflection of the manufacturer's strategy to maximize short-term benefits based on the MIDP rebate scheme. According to the manager of the plant, if profit margins fell below 8%, they would simply close up shop (as opposed to moving to another location). 4.146 This short time horizon is further reflected in the fact that the maintenance costs are high. In a competitive and efficient automated plant producing a similar product, maintenance costs would run about 1% of the entire value chain. In this particular example in South Africa, maintenance costs are 4.9%, which reflects the age of the equipment (>20 years old), and the constant need for repair. What the manufacturer classified as `other' costs reflects the tooling requirement for this type of operation. While somewhat high, given the various dies required for impression die forging, the figure stated here is not out of the ordinary. 4.147 The depreciation rate is 40% of the heating, forging, and trimming cost, which is approximately 3.8% of the total value chain. This, again, reflects the relative age of the equipment and the labor intensive production method utilized by manufacturers in South Africa. A factory manufacturing a comparable product that utilizes an efficient automated production system would expect to have a depreciation rate of 20% or more, which reflects the investments in capital intensive production equipment. 4.148 Manufacturers' reluctance to invest in more capital intensive production may partly reflect the protections accorded by MIDP but, also, the lack of a skilled labor force available to effectively operate an automated production line. Given the fact that tow balls are a low value added product, with relatively low profit margins, there is little incentive to invest, particularly if this calls for further investment in training to upgrade the technical skills required by shop floor workers to operate automated equipment. 134 4.149 In this context, interviews suggest that low value added `commodity' type products in the automotive component industry, particularly those with high material input cost, will tend to under-invest in capital/technology intensive production methods, and continue to rely heavily on labor intensive production methods until such time that benefits from MIDP incentives no longer outweigh inefficiencies in the current production system. 4.150 The manufacturer profiled for this analysis produces approximately 20,000 units/month, which is reflective of the labor intensive production method which continues to be used by the factory. In contrast, an efficient automated factory producing a similar product would be able to achieve an output of over 100,000 units/month. Table 4.34: Tow Ball Production in South Africa: Sample Manufacturer Profile Production Volume 20,000 units/month Defect Rate >1.5% Markets Domestic: 20%, Exports: 80% Production Method Mechanical impression die forging, oven, induction heating Equipment Age >20 years Minimum Profit Threshold 8% Market Price (EU) R26 4.151 In addition, to the low productivity rate, continued reliance on labor intensive (temporary and unskilled labor) production methods are also associated with high defect rates. In this case, the defect rate is estimated to be greater than 1.5%. In an industry where defect rates are measured in parts per million, the defect rates of production facility in South Africa is extremely high. At the same time, however, there is a robust aftermarket in South Africa which is capable of absorbing some of the sub-standard products. Specifically, it is estimated that over 20,000 enterprises today are selling or providing services in the aftermarket sector in South Africa. Table 4.35: Benchmarking South Africa's Production of Tow Balls With International Industry Norms South Africa Industry Norm Production method Labor intensive Automated Productivity (units/month) 20,000 100,000 Age of Equipment (years) >20 <7 Defect rate 1.5% 100 ­ 250 ppm Profit margin 8% 2% - 3% Maintenance cost (% of 4.9% 1% total) Depreciation 3.8% 20% Global Development Solutions, LLC 135 Steel Towing Balls: implications for SMME entry and future assistance programs 4.152 The OEMs and the steel producers exercise dominance over the value chain which allows them to dictate how rebates under the MIDP are shared. The largest producer and importer of steel (Mittal) also appears to be in a position to exercise a degree of monopoly power over steel prices. Consequently, it is important as an underpinning of the business environment for towballs (and other steel auto-components) that an assessment be made of the incentive effects of the MIDP and the supply conditions for ferrous metals. One way of dealing with the steel pricing issue would be to formalize the agreement between DTI and MITTAL regarding the `developmental pricing model'. Tow-ball production is a relatively low/medium technology process which could continue to be an appropriate entry point for metal-products SMMEs, provided adequate skills are available to encourage investing in the metal forming product lines. 4.153 The following current or recent assistance schemes have provided (or could provide) assistance relevant to developing production of steel towing balls. The most relevant programs are flagged. Table 4.36: DTI and associated Support schemes applicable to production of steel tow- balls Support Yes/no Scheme Sector Description Investment grant for building, SMEDP General (SME) equipment Workplace General SME and Grant for workplace projects to improve ChallengeFund LSE labor /management operating practices. Critical CIP Infrastructure Grant for infrastructure improvement Duty free operations and simplified Export ­oriented customs procedures (Coega, E. London, IDZ industries Richards Bay) General foreign Grant to international firms establishing FIG investment projects in SA Selected strategic SIP industries Investment- based tax allowance manufacturing and Advisory services to producers and CF services business service providers manufacturing and Matching Fund for cluster/ network SPF services building programs Cost sharing grant for training and BBSDP Black owned firms mentoring Grant for skills development for work General (recipients force training and promotion of SSP of SMEDP fund) employment Government/industry Action groups and TISA/CSP Customized Sectors export councils 136 National Pavilions; trade mission Export/ inward support; outward and inward investment TISA/ EMIA investment support; exporter training Funding of RFIs and MFIs, institution, Khula Enterprise group and individual bank guarantees, Finance SME sector equity finance Business development centers, business linkage program, information and SEDA small manufacturing advisory centers, tender business support SME sector advice centers GODISA SMEs Multisector Intensive startup support to small firms in six incubators (mining, biotech, ICT, flowers) plus two technical centers SPII - Support SMEs and LSEs Reducing costs of R & D and program for multisector product innovation. Financing for Industrial entrepreneurs in competitive Innovation industries 4.154 As in the case of garments production, the current DTI menu of programs does not seem to place much emphasis on specific training support and there appears to be a need for some more innovative incentive schemes to complement the efforts being made under the SETAs (discussed in Chapter 5). 137 CHAPTER FIVE IMPROVING THE ECONOMIC IMPACT AND EFFICIENCY OF PUBLIC SUPPORT TO SMMEs: THE WAY FORWARD 5.1 The considerable analytical work that has been conducted on the industrial sector in South Africa, and the SMME sector in particular, to which this study has tried to contribute, indicates that the impact of industrial support policies has been fairly mixed. Industrial support measures have been successful on some fronts, such as in promoting the restructuring and export orientation of the automotive industry, but less so on others, including the growth and sustainability of the micro, small and medium enterprise sector. Given the current high rate of unemployment in the economy, the successful integration of this sector into the domestic economy is critical to achieving the objectives of the BEE agenda; however, it is not clear that the incentive framework is as yet fully aligned to supporting its needs, or that the institutional capacity to implement its programs is completely in place. 5.2 The objective of this study was to look at some of the key factors impacting on the SMME sector, to be able to assess whether incentives have, in fact, been correctly targeted to addressing the main constraints they face, whether they have been efficiently dispensed, whether they have had the desired impact, and the nature of the barriers actually faced by small firms in their efforts to integrate into the wider economy. For this reason, the study focused on identifying investment climate constraints, the economic impact of implementation, and the key competitiveness challenges that would have to be overcome to achieve wider integration. This chapter pulls together some of the evidence gathered and looks at potential next steps to further these objectives. A. REVIEWING THE CONSTRAINTS ON SMMES AND THE ECONOMIC CASE FOR INTERVENTION 5.3 Does the overall rationale for Government intervention in 1995 still hold in 2006? To what extent have economic circumstances changed since the initial assistance programs were introduced in the mid-1990s and, if so, does this imply the need to rethink the policy rationale? 5.4 The combined MICA and ICA studies, and the value chain analyses of the textiles, auto components and construction equipment sub-sectors, pinpointed specific constraints that applied to the broad SMME sector. Among these, principally, were the skills gap and access to finance. 5.5 The skills gap is primarily responsible for the low labor productivity in certain segments of industry. The ICA survey showed high relative labor productivity in some first economy segments, especially foreign owned, because of capital intensive 138 investment over recent years. However, the value chain analysis showed that within some product lines, for example light engineering products such as pumps and towballs, scale was relatively small, production techniques obsolete and labor productivity accordingly low. The future competitiveness of South African industry depends largely on two factors, increasing labor productivity and limiting wage costs. 5.6 Employment and skills Manufacturing employment in the formal sector has declined as a result of labor-saving investment and regulatory constraints on labor market restructuring (reflected in the cost and time required to retrench workers) which has discouraged hiring. Due to a recovery in output from 1999,127 formal manufacturing employment finally started to rise in 2002, moving upward by 2.7% according to the NPI.128 Several industries, however, still showed declines, especially textile and garments.129 5.7 In the longer run, the issues of stagnation or slow employment growth, skills mismatch, continued labor market segmentation, and high formal wage levels for a protected first economy labor force, still remain. As noted in the value chain analysis and in Annex 8, labor training has declined and ICA interviews with firm employees suggested that a large proportion (perhaps more than 80%) of workers in South Africa have not received any formal training in the workplace.130 While the Government has taken concrete steps to develop a skills development strategy, it has yet to make major inroads into the problem. The SETA (Sector Education and Training Authority) levy- grant training program, for instance, has a number of weaknesses. As we have seen, SMMEs have both relatively low participation in it compared to larger firms, and very low grant take-up, due mainly to lack of knowledge of the procedures involved.131 According to the ICA, wage differentials as a result of training are also comparatively low, suggesting either undervaluation of or ineffective training, or that managers select only the most productive workers to receive training. 5.8 In international terms, relative unit labor costs (productivity related to wage levels) fell in the 1990s but still exceeded those of other developing countries. Since 1999 money wages have increased faster than productivity, and the 1990s relative cost improvement may have been reversed. The pressures to raise money wages have been inconsistent with increasing employment in South Africa as they encourage labor-saving investment and further raise the barriers to entry into the formal labor market. Thus, overall, there is not a significant change from 1995 and a range of policy measures to support skilled labor supply through expanded training opportunities, reduce labor cost to 127It is in fact the longest cyclical upswing since 1946, as monitored by the South African Reserve Bank (See Reserve Bank Quarterly Bulletin). 128NPI Productivity Statistics 2004, 2005. The gains were made mainly by paper products, coal/petroleum, nonmetallic minerals and metal products. 129NPI figures show employment in textiles and garments declining a further 3.5%, shedding another 6,000 jobs in 2004. 130This does not appear consistent with the South Africa State of Skills Report, 2005 which states (p43) that 40% of small and micro firm workers underwent training during one year (2002/3). However the latter report included informal training broadly defined. 131 See SA State of Skills Report 2005, p37-39. 139 formal business and boost demand, especially to bring informal (black) workers into employment, continue to be necessary. 5.9 Finance and Investment. The rate of gross capital formation to GDP has risen to 17%,132 but is still well short of the international benchmark of 25% for industrializing economies. However, the medium and large formal enterprise sector is not, in general, credit-constrained according to ICA findings. Long term real interest rates in South Africa have declined significantly since 1998,133 thereby lowering one of the important barriers to investment. A number of industries have announced large capital expenditure plans, including mining (notably platinum), automotive and chemicals, retail, real estate and tourism. 5.10 Nevertheless, a lack of competition in the banking sector appears to have impeded the creation of the smaller specialized and more flexible financial intermediaries that, in many countries, have helped fuel and sustain the growth of small firms. The five largest banks in South Africa account for more than 85% of bank deposits. A recent study134 reported a lack of diversity through second and third tier banking institutions, and that the major banks were undermining new entry in the sector. This has created access problems, confirmed in the MICA study and the value chain analyses, especially for low and middle income clients who face high bank charges, low or negative returns on savings accounts and limited access to banking services. South African banks are slow to move away from collateral-based lending, have a conservative attitude to financing and, as stated in the textile value-chain analysis, tend to `profile' certain industries, especially small firms, as high-risk without necessarily examining specific business risk. MEs are in the worst position, with very limited access and very high costs, and a substantial segment of the population has no access to banking services at all. 5.11 Entrepreneurship. There has been a steep rise in new registrations of private companies, from 6,369 to 32,178 per year, between 1990 and 2002, and for close corporations from 28,008 to 107,302, especially in urban areas such as Gauteng, Western Cape and Durban. This trend is confirmed by the number of companies on the tax register which increased by almost 20% over the two year period 2001/02­2003/04.135 The share of MEs in employment and GDP has also increased,136 suggesting that the incubation of new entrepreneurs is not as critical an objective as it was in 1994. However, the numbers of new businesses do not say anything about quality or survival rates, and there is not yet evidence to show that adequate gains in entrepreneurship have been made. As reported in the value chain studies, lack of education, technical training and management experience impacted negatively on entrepreneurship. 132SA Reserve Bank Quarterly Bulletin March 2005. 133Due to increasing confidence in Rand stability and reduced country risk. See 2005 Budget Review. National Treasury. 134Task Group Report for the National Treasury and the South African Reserve Bank: "Competition in South African Banking" April 2004. 135Source: South African Revenue Services, Annual Report 2004:24. 136Source: Johannesburg microenterprise study. 140 5.12 Technology Acquisition. The ICA shows that first economy firms have increased labor and total factor productivity steadily over the past ten years, implying that technical innovation has been a major component of the growth rate. While we do not have direct information on technological change, the relatively high capital intensity and high labor productivity of formal sector enterprises would suggest that, in the first economy, technology is relatively up to date.137 On the whole, technology acquisition, which has faced barriers in a number of countries, has been relatively open in South Africa. However, according to the value chain studies and the surveys, small firms and, especially, enterprises owned by black individuals have much less capital per worker than firms owned by Europeans or Asian individuals and are often using obsolete equipment. Another study has found that, even if process technology is new in the formal first economy, South Africa is not moving sufficiently quickly into the high product technology to keep pace with other industrializing economies especially in Asia.138 There are indications that the rate of innovation has fallen; one of these is the slowdown in application for grants to the SPII (see chapter three). There, therefore, appears to remain a public interest in supporting and accelerating R & D and high-technology investment both in the formal and, to the extent possible, the informal sector. 5.13 In major policy areas, therefore, the economic rationale that existed for SMME support in 1995 is still valid, and the issue is one of finding cost-effective programs that meet that rationale. B. REVIEWING THE IMPLEMENTATION OF ASSISTANCE PROGRAMS 5.14 As a basic principle, the public benefits of support measures need to exceed the cost of their introduction, or else the intervention might worsen the cost of the economic distortions involved. Considerable emphasis is now placed on the ongoing process of Government-private sector cooperation in identifying appropriate measures and agreeing on what measures can be effective. Government support must be flexible, adaptive and demand-driven, and the process of dialogue with the private sector should, in itself, be an objective of State support. 5.15 The extensive program of State support to SMMEs that has been under way since 1995-6 has, by and large, not been able to meet these criteria for success, nor achieved the quite ambitious goals set for it. Both the MICA survey and the targeted program beneficiary interviews found that program promotion was very weak, there was inadequate knowledge of the programs, and they were difficult for small firms to access due to the procedural complexities at the application and the claims stage. Most firms had heard of the Government programs though word-of-mouth and no advantage was taken of available enterprise databases to enhance promotion and dissemination. The picture was not wholly adverse since interviews did find that that the quality of assistance 137The ICA also found that smaller formal enterprises in South Africa were more capital intensive than larger firms, partly because of the labor intensity of some large firms such as garment manufacturers. 138Source: `South Africa Export Performance: Determinants of Export Supply', L. Edwards and P. Alves Univ. Cape Town March 2005. 141 given, and of program management, was acceptable. There were also quite a high percentage of beneficiary firms that were willing to re-apply for programs, especially in export marketing. But, there are many areas where there is clearly room for improvement overall, as set out in the following section. Key Effectiveness lessons 5.16 The study looked at the implementation of assistance from the point of view of a sample of individual support programs. These divided up into three programs providing financial assistance (SMEDP, Khula and the NEF), and seven providing assistance in skills and know-how development (Ntsika LBSC, Godisa business incubators, EMIA pavilions, SPII innovation support, and three individual Bank-supported matching grant funds, CF, SPF, and BBSDP). One program, the GEP, local to Gauteng, planned to provide both skills support and financial products. The following are the principal lessons. Good Promotion of Assistance Programs Is Essential to their Effectiveness 5.17 Among the programs, the investment grants under the SMEDP have been sufficiently attractive and easy to access that they have tended to face an extremely high flow of applications in relation to processing capacity. For most other programs, however, the surveys of beneficiary firms reported in Chapters 2 and 3 have revealed that program promotion was usually very weak. This could be partly remedied if designs could be simplified in all programs so as to allow easier access (e.g. so that consultants are not needed to fill out applications on behalf of applicant firms). In addition, however, promotional programs themselves need to be designed so as to maximize coverage per rand expended. Road shows, which are carried out in most cases, tend to be relatively expensive in terms of the amount of outreach, and their coverage has in fact been very limited. Use of mass media such as TV and radio spots has been too generic and ineffective in providing information about specific programs and their use. 5.18 One relatively simple approach would be for DTI to make use of the extensive company databases available through the Registry of Companies, or possibly through the Tax Service, to send out newsletters, mail-shots or e-mail distributions to inform potential applicants about specific DTI programs on a regular basis. Such information would include clear messages about the type and amount of assistance, the manner in which it would work, the eligibility criteria and the application procedures. 5.19 Support agencies such as SEDA could similarly utilize their own extensive databases for a similar purpose. Other approaches proposed include advertising in trade magazines (e.g. Engineering News). Much more could also be done with the better publication of preferred procurement information. It is likely that the spinning-off of programs under independent fund managers would enable considerably more innovative promotional methods, which would be an important reason to go this route. 142 5.20 Phased Implementation Allows Improved Learning. The overall conclusion from this review was not very positive. Achieving economic impact through State programs is difficult, and implementation needs to be phased so that a learning process takes place. Since 1995 there has been a large number of programs set up, either simultaneously with, or before, the lessons of the earlier programs can be taken into account by the later programs. It is difficult to accurately monitor the value of State support programs once they are in place because they tend to become self-justifying - their output tends to be confused with their impact. 5.21 Inappropriate selection criteria and procedures negate the economic rationale of a program. Failure to develop and follow appropriate selection criteria seems to have created problems, especially for the SMEDP, which faces an excessive throughput of grant applications in relation to its objectives and to its staff capacity. Medium scale enterprises that are not credit constrained have received grants that are not essential to their investment plans and, thus, provide little or no additionality. If selection is re- focused on small firms and opportunities, where publicly beneficial incremental investment and output are likely to result from the grant, then the workload could be substantially reduced and staff capacity would be more able to cope with the approvals, verifications, monitoring and claims handling. The additionality issue also arose in other programs, including the World Bank assisted funds. In the case of the SPF, appropriate selection was difficult because it required a judgment of whether the groups were likely to enhance productivity through sustainable new linkages and networks. In the Khula RFI program, at least half of the RFIs selected for funding turned out to be financially unsustainable and their obligations had to be written off. In the case of the Ntsika LBSC, the selected center managers also, in many cases, had to be de-accredited. These latter cases were more difficult to deal with than the SMEDP, however, since they involved up- front judgments about the viability of selected beneficiaries rather than problems with the criteria for selection themselves. 5.22 Selection criteria need to be designed to focus, simply, on the additionality and potential public benefits of projects. The programs must also appraise the potential beneficiaries' ability to meet those criteria. In the case of Khula, the decision to consolidate the RFI clients into a small group of stronger institutions, which can be more closely monitored and nurtured to financial sustainability, appears to us to be an example of the most appropriate way forward. 5.23 Too Rapid Scaling Up Results In Lack of Focus. Regarding overall institutional effectiveness, the impact of the programs reviewed is probably representative of the effectiveness of the institutions responsible for them. For DTI, this implies, not surprisingly, that some are reasonably good models while others are not so good. A general problem is overambitious or over-complex program goals, lack of targeting and duplication of effort. In several cases, it would have been better to focus on a few successful projects to use as models for dissemination. The desire to scale up assistance as quickly as possible has tended to be at the cost of reduced effectiveness of each operation. This is probably particularly true with the RFIs, where a narrower concentration was needed to build sustainable institutions. It seems to have been true 143 also of the LBSCs, where an emphasis on numbers apparently resulted in neglect of existing LBSCs. The NEF is charged with supporting potentially marginal businesses, achieving financial sustainability, and managing a portfolio of mature investments at the start of its life, which may create similar difficulties. On the other hand, the Pavilions Program and SPII have kept to fairly modest objectives (some argue too modest). The GEP similarly has built on MAC activities but in a fairly measured and realistic way (e.g. introducing just one financial product). The SMEDP is a prime example of an overambitious program that appears to have resulted in over-commitment of managerial resources and inability to adequately approve, monitor and manage grants. 5.24 Over complexity and Administrative Burden Decreases Firm Involvement and Program Effectiveness. There is a continual problem of excessive administrative burden, leading to relatively high management cost ratios compared to other types of assistance projects. Most firms interviewed for this study complained about this problem, which derives partly from the need to install safeguards against misuse of public funds, but it also leads to the use of considerable management time in grant applications and claims. This has, in turn, led to the widespread use of consultants and facilitators to make applications in return for a fee. Such intermediaries can add value by improving grant allocation and speeding up the process but, equally it is important for firms to take more responsibility for their own projects. Furthermore, intermediation has at times been collusive or worsened grant allocation. 5.25 Failure to Fit Programs to Needs Reduces Impact. Despite clear mission statements, many programs (for example the LBSCs) have failed to avoid some of the pitfalls that they were specifically designed to avoid ­ such as the tendency to become supply-driven, to use a one-size-fits-all approach to business assistance, or to become isolated from other parallel agencies instead of collaborating with them. In some cases, funding was overstretched and, as a result, not sufficiently reliable to allow good locally appropriate program development. C. REVIEWING ENTRY POINTS FOR MICRO AND SMALL FIRMS ­ THE CASE OF TEXTILES, AUTO COMPONENTS AND CONSTRUCTION EQUIPMENT INDUSTRIES 5.26 The value chain analyses allowed for a stage-by-stage look at the dynamics of the supply chains of the selected sub-sectors, with a view to facilitating the identification of entry points for smaller forms that the range of promotional services could target. It also allowed for a disaggregation of cost structures that was helpful in identifying economic efficiencies and inefficiencies at each stage, to be able to determine their impact and implications for the entry of SMME businesses. The analysis also took account of the impact of specific incentive programs and possible distortions arising from their application. 5.27 An important finding of the analyses is that, especially in industries under acute international competitive pressure, such as automobiles and garments, there is little alternative to raising productivity through capital intensive investment, which means that opportunities for employment growth in the unskilled and semiskilled labor category will be limited. A possible alternative strategy is to reorganize through outsourcing/ 144 subcontracting, at lower wage levels. However, this may well only be a viable strategy in the medium term. Thus, there is inevitably some tension between the goals of international competitiveness and the goals of employment expansion and black enterprise and worker empowerment in the longer run. The DCCS and the MIDP 5.28 In all cases, within the cotton textiles, auto components, and construction equipment subsectors, efficiencies turn out to be heavily influenced by the presence of two major incentive programs, the DCCS, and the MIDP. Thus, these two incentive programs have a determining effect on the short and the long term profitability and competitiveness of the respective industries, and for the entry potential of new firms into those industries. The MIDP appears to have a significant impact on the financial viability of production all the way down the chain from raw material producer (steel and leather) to vehicle manufacturer (OEM), such that the ending of the incentive program could cause the closure of many firms within the sub-sector. While the Government has tried to design the program such that it would phase out over time through tariff reduction, in practice, as with all protective barriers, the necessary increase in competitiveness is very difficult to achieve. Yet, it must be achieved and so there is an urgent need to design into it automatic and foolproof efficiency building measures. 139 5.29 With respect to the protection of the DCCS, the considerations are similar insofar as many firms report that they would not survive without protection (mainly against Asian imports). The way in which the DCCS has operated, in practice, has discouraged skills development and encouraged the import of competing products such as garments through transfer (sale) of duty credits at a discount to importers. SMME Entry Points 5.30 The value chain analysis identified a number of strategic entry points where SMMEs and BEs may have opportunities to engage in productive economic activities. However, it also highlighted where the intersection of the competitive agenda (i.e. the need for downstream industries to restructure themselves so as to survive in an increasingly competitive global market) calls for complementary actions necessary in order to capitalize on these entry points. For instance, given the low skill base generally found among workers in SMMEs and BEs, the Government will need to focus on delivering targeted support services to improve labor productivity and the quality of their output. 5.31 While pockets of opportunity exist for SMMEs and BEs along the various value chains, these are limited. For example, in the cotton-to-garment value chain, substantial capital investments are required in the ginning and textile spinning and weaving areas, which limits, if not eliminates, opportunities for SMMEs and BEs unless the work is 139The IMF staff report for 2004 Article IV consultation, August 12, 2004 notes that the Government has argued that an important indirect benefit of the MIDP is investor knowledge and a track record of foreign investment in South Africa. This would to some degree offset inefficiencies created by the incentive. 145 limited to contract labor. In the case of products like pumps and towballs, relatively simple production techniques and limited capital investments are required. On the other hand, the analysis suggests that with sufficient skills improvement activities, opportunities may exist for SMMEs and BEs even with entry level skills. Similar observations can be made for the automotive components sector. While quality craftsmanship and technical skills are required for stitching seat covers, it is an area where small production units can become a part of a subcontracting supply chain. Cotton and Cotton Textiles 5.32 There is an overarching need to review the DCCS incentive mechanism to ensure that it meets the objectives of transitional protection of jobs, restructuring, adjustment, phasing and sun-setting. There is also a need for c 5.33 Cotton cultivation. Irrigation of cotton cultivation is essential to raise yields and meet international competitiveness benchmarks but black owned farms are generally un- irrigated and there has been a switch into other crops. Much of the land operating under communal property arrangements has not, on account of the incentive structures, benefited from especially efficient land management. Provided this hurdle can be overcome, black-owned farms may be able to participate in cotton outgrower schemes in partnership with large commercial cotton farms. A model for this is Mozambique where, increasingly, large firms are outsourcing to smallholders under Government-supervised supply contracts (85% of seed cotton is now produced either on smallholdings or on land made available by the purchasing factory). This will require identifying appropriate areas of land close to ginneries, possibly with existing irrigation facilities, capable of good management not restricted by communal title. Public investment to improve access to existing irrigation infrastructure may be justified in the selected areas. 5.34 Garment production. In an effort to reduce labor costs and increase the flexibility of production, CMT outsourcing has been expanding rapidly, with over 1000 units now in existence, some successful as independent firms while others are in a more dependent subcontracting relationship with large manufacturers. This form of `flexible specialization' is being applied to garments at both low high ends. It is recommended that the Government takes steps to support the development of subcontracting relationships and skills development to integrate viable CMT units capable of developing independent business strategies into the formal garment supply chain, allowing some regulatory flexibility to permit costs to be kept down. Also recommended is support to actively promote trading up into high value cotton garment exports using the available high quality South African cotton, to ensure the longer term viability of the sector. General productivity assistance measures should support technical upgrading, especially of fabric quality, marketing, branding and delivery, and upgrading within the value chain to extend the CMT units over time into new value-added areas of quality control, design, branding and marketing. 146 Engineered products (construction equipment and automotive components) 5.35 There are potential entry points for small and micro businesses in basic casting and forged products, and in some machined products. Based on the value chain analysis there are initial issues that need to be resolved. Firstly, the Government needs to review the impact of the MIDP incentives upstream on steel producers and steel auto parts manufacture, to ensure that no businesses are being hurt by the efforts to maximize domestic value for rebate purposes, and also to assess the long term viability of steel auto parts within the automobile sector. Secondly, it needs to continue urgently to develop mechanisms for ensuring controlled phase-out of the incentives in the long run. Thirdly, the Government needs to ensure that the main supplier/importer is adhering to competitive practices and, in this respect, to review the developmental pricing model currently being promoted. Intermediate processed products (Leather seat covers) 5.36 There are potential entry points for small and micro businesses in manufacture of leather seat covers as part of the leather products sector. As in the case of engineered components, we recommend that the Government reviews the impact of the MIDP on leather seat cover manufacture which is one of the largest auto component producing lines. There is also a need to review the pricing of tanned leather to ensure adherence by the tanneries to competitive practices and avoidance of collusive deals between chain leaders. Since leather products are likely to be within South Africa's areas of comparative advantage there may be potential in the long run for independent leather seat exports outside the automobile sector and protection of the MIDP. 5.37 Assistance measures should support the development of subcontracting relationships between existing automotive component manufacturers and smaller production units, and also the longer run potential for direct export. 5.38 Ch. 4 looked at existing and recent DTI support schemes, some of which could potentially provide support to the garments, leather seat cover and metal tow-balls product lines. In general, the suite of currently--available DTI and associated programs does not seem to be too strongly focused on skills development ­ which suggests the need to introduce a broad and innovative demand-based skills grant program that can replace the earlier CF, SPF and BBSDP programs and complement current SETA programs. 5.39 Another area of support is that of small-scale finance. The Khula programs are focused at the wholesale level, while the micro-apex fund which has only recently started up, focuses only on the smallest firms. If the access to finance constraint reported by small and informal firms in the survey is to be adequately addressed, a redesigned SMEDP investment grants program (or its successor) would need to focus more specifically on this segment of the market. 147 D. REVIEWING AREAS OF GENERAL STATE SUPPORT TO SMMES Skills Development 5.40 The Government implemented a first National Skills Development Strategy during 2002-5. This strategy made reasonable progress in achieving its goals for training and utilization of training grants. The target for National Qualifications Framework (NQF) Level 1 qualifications was only half fulfilled, but better progress was made elsewhere, and the report cites clear evidence of a significant increase in the capacity of the SETA system. For example, 37% of 181,000 levy-paying small firms were supported through skills development initiatives (well in excess of the initial target) and the target of 80,000 `learnerships' appeared to have been achieved. Despite this progress, however, the second phase of the Skills Development Strategy announced in 2005 continues to face poor labor market trends and an adverse training legacy, including a decline in technician graduates and apprenticeships, and an aging skilled workforce that is not being adequately replenished.140 5.41 While a significant proportion of small firms participated in training activities, only 10% actually claimed grants through the levy program on account of the difficulties of access. In 2003, training rates remained low for technicians and for craft and skilled trades workers (20 to 23%) compared with service workers (33%), and training expenditure was lowest in small enterprises (at 1% of payroll) compared with 2.8% in large enterprises. 5.42 The performance of the 25 SETAs established in 2000 was found to be variable. Registration with the SETAs for participation in the levy-grant system was lowest for small enterprises, at 56%, of which only 29% had claimed grants, while for large enterprises it was relatively high. The micro-enterprise sector was the least active participant in training, with only 9% of the registered firms with less-than-6 workers claiming grants. The rate of grant claims varied from as low as 20% in the ETDP SETA and 23% in the Construction and Education SETA, to 78% in the Financial Services SETA. The level of satisfaction with SETA services was generally low. Less than 10% overall rated these services as "good" or "very good". The SETAs and the National Skills Fund also provided support outside the levy system for SMMEs, especially for first-time labor market entrants. 5.43 Among the programs that the State of Skills Report rated highly was the THRIP program which funds R&D and innovation through domestic or foreign academic/ technology institutions through an industry/higher education partnership to produce top- level skills, and the SPII innovation fund, both of which are under DTI. Training initiatives have also been taken in the corporate sector, e.g. by Telkom, which has instituted a `Centers of Excellence' initiative for top level ICT professionals. However, at the craft and technical training level, a look at the available support programs seems to suggest that there is considerable room for expansion of targeted training support under DTI. 140State of Skills Report Op cit 2005. 148 5.44 With the Phase 2 strategy, the opportunity now exists to take more aggressive steps. The objective should be to reverse the observed serious decline in skills, especially for small-to-medium scale formal sector enterprises, through a well funded and broad- based initiative, and an expansion of DTI's contribution in this specific area would be worth considering. 5.45 The fundamentals of the Government's second phase strategy are as follows.141 a) Alignment of all skills development activity with the economic growth, job creation and poverty alleviation policies of government; b) Alleviation of critical skills at three levels: advanced, intermediate and entry-level skills; c) The deepening, consolidation and expansion of enterprise training in the large, medium and small enterprises of the formal economy; d) The expansion of social development initiatives particularly with regard to the training of the unemployed, the provision of Adult Basic Education and Training (ABET) to adults requiring such upskilling, support for the Expanded Public Works Programme (EPWP), and the development of SMMEs; and e) The promotion of quality training across all sectors and institutions (public and private). 5.46 Among the measures that the Government could take within this strategic framework, and which we would support, would be: a) To greatly expand vocational training, including appropriate revamping of the apprenticeship/internship system and the SETAs; b) To re-launch the Government training levy/rebate system within the SETA organizations levy so that it provides better accessibility and greater monetary incentives for firms to make use of it; c) To expand other skills support and training programs offered by DTI including both general training grants and special incentives for large enterprises to design and implement training and skill transfer in small enterprises; d) To ensure that skills development responds directly to demand, largely on-the-job, through apprenticeships and similar schemes; and 141State of Skills Report. Op cit, 2005, p63. 149 e) Following d), to introduce funding mechanisms and incentives for larger enterprises to prepare and execute training programs for linked small firms in priority products. Entrepreneurship 5.47 The objective should be to strengthen black entrepreneurship, maintaining the rate of new micro/small business entry and reducing the rate of closures by strengthening operations. Among the measures would be the following: a) Refine instruments (entrepreneur training vehicles) and modified grant funds (State and private/public) to support business entry and upgrading; b) Design viable financial products (e.g. loan guarantees against subcontract orders and Government procurement contracts, and leasing products) to improve access to finance especially for small and informal enterprises; and c) Ease business entry by expanding industrial zoning and improving utilities and transport infrastructure for micro/small firms. Investment Finance 5.48 Financial sector issues, such as improving access, require a separate in-depth specialist study to create a strategy to deepen the capital market and provide financial products to small and informal firms. Among measures to be taken would be the following: a) Introduce innovative financial products through commercial banks, specialized micro-finance institutions, or through non-bank intermediaries such as venture capital funds, provided viability and sustainability can be assured; b) Revise the SMEDP to target small firms and ensure additionality; c) Revise Khula-type programs to fund a selected small group of viable or potentially viable MFIs, and provide loans or guarantees against contracts and cash flow for small formal and informal firms; and d) Ensure NEF-type financing is sustainable and subject to transparent and clearly justified subsidies. Overall Productivity and Competitiveness 5.49 The objective would be to improve overall productivity and competitiveness in SMME firms to facilitate their integration into larger firm value chains. Among the measures to be taken would be to: 150 · Improve and refine a range of programs such as the CF, BBSDP, MACs, LBSCs in business planning, operations, technology, investment, financial control, export and domestic marketing, packaging, branding, quality certification, etc.; · Develop regional/local support programs, to support business services to upgrade firms, parallel to above; · Introduce program refinements such as targeted selection (for additionality), more effective promotion, improved management subject to control of costs, improved monitoring and controls, better design of services to meet specific local needs; · Introduce or refine alternative productivity improvement vehicles through e.g. enhancement of large-small firm business linkages and spillovers including incubators, outsourcing, subcontracting; · Scale up and improve access to existing export promotion programs such as EMIA/TISA, broaden scope and step up promotion for SMMEs; an · Intensify `country marketing' programs (e.g. like the `Proudly South African' program) in key export markets, through target market identification, trade fairs and promotional campaigns. E. THE PATTERN OF GOVERNMENT ASSISTANCE 5.50 The range of support activities to industry and services, justified by current economic conditions, includes continuing supply-side measures to encourage small firm and black, women and youth business entry, and also labor-use, skills development, productivity enhancement, infrastructure investment, R & D, technological development, BDS market growth, and targeted and transitional incentives towards specific industries. 5.51 A difficulty in designing the overall program is the fact of the conflicting triple objectives of economic efficiency (competitiveness), welfare (poverty alleviation) and political power (black empowerment), and of the appropriate allocation of funding and effort between these objectives. The objectives of simultaneously pursuing economic efficiency and, say, unskilled labor-intensive investment and micro-enterprise development, along with the dilution of European ownership and management, are subject to considerable tension, especially for economic activities under severe competitive pressure (for example, the textile industry as discussed in chapter five). While SMMEs were seen as agents of employment promotion, redistribution and even improvement in global competitiveness, clear guidelines for resource allocation and a hierarchy of importance were absent, and the SMME strategy has struggled to meet this range of different and sometimes conflicting objectives. 5.52 The main recommendation here is for the Government to revisit the rationale for the split of expenditure and management effort between its assistance programs in terms of both developmental prioritization and cost effectiveness. The purpose of this would be to review disproportions in cost and funding, and the potential for transferring of funds between programs. It is recommended that part of the SMEDP funding, which is currently going to fixed investment in medium scale firms, be diverted to skills-building, while the SMEDP itself be reintroduced with a focus on small firms. 151 5.53 Skills training may be conducted through a variety of mechanisms, including matching grants schemes similar to the Competitiveness Fund, which can stimulate linkages between firms needing skills and private skill providers. These links should be both to formal skill training institutions and to informal providers, particularly large enterprises able to provide market sensitive and state of art practical training and skill transfer within specific value chains. The next generation of matching grants funds in South Africa, following those assessed in this report, could form part of the urgently needed skills development effort, in which case they should be designed with the above international cost benchmarks and these principal lessons of experience in mind to ensure cost-effectiveness. 5.54 There are several international examples of good practice that can be drawn from, including, for instance, the Skills Development Fund in Singapore and the Human Resource Development Fund in Malaysia. The Singapore Fund provides financial incentives for workers, through a Skills Development Levy (of 1% of monthly remuneration) imposed on employers with workers earning $1500 or less a month. Incentives are offered on a cost-sharing basis and the amount of training is not specifically tied to the levy contribution. Malaysia's HRDF is similarly a training levy reimbursement scheme operating on a matching grant basis. Though not specifically targeted at SMEs, the scheme appears to have been successful in promoting productivity growth (especially through repeated rounds of training).142 F. IMPROVING THE SMME SUPPORT INFRASTRUCTURE The Key Issues 5.55 From the evidence of this study, we conclude that the organization of the institutional structure is extremely important to improve the alignment of capacity with objectives, sharpen focus, ensure stable funding and prevent wasteful duplication. The Government has taken rationalization steps, with the creation of SEDA in 2004 incorporating NAMAC, Ntsika, CPPP and more recently GODISA, but, at the same time, it is also developing new programs for business process outsourcing, film development, and the micro apex fund. 5.56 We support the rationalization of Ntsika and NAMAC. The central issue is how to organize the overall support framework and how to design and operate public support programs to achieve their objectives cost-effectively. 5.57 The key questions are: · Is a solution to the improvement of program management to spin it off from Government to independent fund managers? · What role should Government take if such a spinning off were to occur? 142See: Hong Tan: "Malaysia's Human Resource Development Fund: An evaluation of its effects on training and productivity". 152 · Should Government focus on leveraging the capacity of large private firms to assist small firms, or, private initiatives to help micro-businesses ­ in effect using the first economy to boost the second economy? · If so, should government incentives aim at larger firms to compensate them for the risks (e.g. poor quality or late delivery) of accepting a small firm into their supply chain, or at smaller firms to make the necessary adjustments to strengthen their skills and competitive efficiency? RECOMMENDATIONS: Consolidation and Rationalization 5.58 It is not easy to identify the appropriate program structure without a clear understanding of the relative cost-effectiveness of each type of program. It is equally not obvious that programs are too numerous or overlapping without detailed knowledge of each. 5.59 There are complex issues in deciding whether transaction cost savings from merging programs are more important than the benefits of specialized program targeting, but the evidence of this report strongly suggests that duplication is a problem, and that compartmentalization of assistance efforts into numerous program areas creates serious bureaucratic challenges both for the providers and for the recipients, which reduce the value of the assistance. For similar reasons, we would support re-integrating Bank programs such as the SPF and CF into single funds. The more specialized cluster- building aims of programs like the SPF could be retained within the larger program through a `firewall' allowing amended selection procedure, sub-project objectives and M & E. The Government's initiative in setting up SEDA, which consolidates NAMAC, Ntsika and GODISA, is consistent with progressive rationalization. Spin-Off and Independent Fund Management 5.60 On the experience of programs like the CF and the SPII, we recommend spinning off program management from Government departments and, possibly, also from the public sector through independent fund managers. The GODISA incubators and the Ntsika LBSCs were both implemented by private/NGO groups, in the first case chosen through a competitive tender process and, in the second, through an accreditation process. We recommend revisiting the management selection model of the GODISA and Ntsika to see whether it is possible to use these procedures more successfully than they have been up to now. 5.61 As explained in chapter three, the TEO was established within DTI in 1999 with a view to potential spin-off, but this did not take place. It is recommended that DTI seriously considers reviving this idea. The reason for spinning off management and implementation from Government is one of alignment of interests and objectives and also 153 one of effectiveness. This is because the oversight role and mentality of Government is generally not appropriate to the more entrepreneurial approach required to operate a support program successfully. Centralization and Decentralization 5.62 While spinning off may, in many cases, mean decentralization of program management, decentralization is, in itself a separate assistance strategy issue which is worth considering more closely. 5.63 Decentralization has potential benefits. Key among these are the better targeting of clients (especially BEE clients) and client ownership, better definition and delivery of services required by client firms, and more effective follow-up. There are also potential difficulties which have emerged, as in the case of Ntsika. These have included in particular variable resources, capacity and service quality. The judgement about decentralization has to take into account both these potential benefits and difficulties. 5.64 Decentralization of assistance program management, if introduced, could take place in the following ways: - within State Government (e.g. DTI), over a transition period, to provincial offices as the capacity of those offices is developed; - to the provincial government, city or local development agency level (e.g. under control of Provincial Economic Development Agencies); - within semi-governmental agencies, to local management centers under a central coordinating organization (the Ntsika LBSC model); - within semi-governmental agencies, to local management centers under a more hands-off central apex institution (the NAMAC model); - to independent non-government and private sector fund managers operating within a partnering or contractual relationship to central or local Government agencies; - to semi-private sector agencies (e.g. section 21 entities), on the possible model of Business Partners SA. 5.65 Some of the semi-Government agency schemes involve a public-private partnership. Of these, the NAMAC approach, which allowed considerable autonomy to well established local centers, partly privately or donor funded, may be the one worth special attention. The Role of DTI 5.66 The principal recommendation is for the DTI to divest itself of all or most operational program management. This recommendation recognizes that such divestment will place a greater oversight, monitoring and evaluation burden on the Government, but this may in fact be a more appropriate function than direct management. Ideally, DTI and Government in general should attend to policy, regulation, and oversight functions, 154 allowing independent agencies to be responsible for implementation and management of programs, which would be the responsibility of semi-independent Government or private agencies. DTI would, however, provide: a) policy analysis and development, b) formulation and development of regulations, and c) internal capacity for monitoring and impact evaluation of programs managed outside. Public-Private Initiatives 5.67 The recommendation is to strengthen Government-industry partnerships to promote industrial activities with long-term productivity potential. Examples include Khula Finance's initiation of private partnerships for micro-finance institutions, and schemes linking large private firms and public agencies to provide finance to micro and small (black-owned) firms such as the Khula-Anglo-American mining fund, a Shoprite venture for setting up retail shops, Kickstart' with SA Breweries, the Eskom fund, Semele, and the ACSA fund. Such partnerships should be designed in such a way that a net financial cost is not placed on the sponsoring private firm such that its competitiveness is reduced. Using Large Scale Business to Support Small Business 5.68 It is also recommended that redoubled efforts be made to strengthen linkages between micro-small and large enterprises, such that the capacity of the first economy is used to support the second economy. There are two main approaches, as follows. - Formal-to-informal enterprise links: Intensified steps may be taken to promote business development in the small/informal sector using the knowledge of the first economy to develop incentive schemes. It is worth looking at schemes to provide a financial incentive to large enterprises to take the initiative and design intensified assistance programs for their small/micro clients funded by, for example, matching grants. These schemes could be organized through large supplier firms who would provide management/technical support to micro/small firm clients who are purchasers of their equipment, materials and services, and training support in business management to those same firms. - Informal to formal enterprise links: Based on the positive evidence of this report, it is worth considering incentives to strengthen forward linkages from MEs to larger formal firms through subcontracting and outsourcing. This approach has been endorsed by the Government as, for example, in its State of Skills Report 2005 (p8 and others). The idea would be to provide financial incentives to the larger firms to actively develop subcontracting and outsourcing to small/micro firms coupled with management and technical assistance to those firms. 5.69 Examples of the latter include outsourcing to garment CMT units, or the spinning off of small startup firms from motor vehicle OEMs to sell parts back to the OEM. The 155 earlier Bank-supported SPF program had focused on setting up outsourcing clusters, and this model could be developed, with improved design. Subject to avoiding placing undue burden on larger enterprises already facing severe competition from imports (which would be avoided through the financial incentives), there is an apparently viable development approach in South Africa of large-to-small firm subcontracting through business transaction links, technology and know-how transfer links. Final thoughts on program design 5.70 The following are main principles that we would recommend for setting up assistance programs: At the institutional level: · Identify a clear economic rationale for programs of support; · Simplify the institutional structure supporting that rationale by rationalizing, cutting or merging agencies; · Separate service provider agencies from (central) policy, regulatory and oversight agencies, or Government departments; and · Focus on and test good program management models using, where possible, specialized local management resources outside the public sector. At the program level: · Develop program models and ensure effective performance before scaling up ­ i.e. learn by doing; · Develop cost-effective publicity and promotional measures maximizing coverage per rand invested; · Maintain relatively simple program objectives and reasonable workload; · Select projects according to additionality and public benefit criteria; · Negotiate, at time of project preparation, to anticipate operational problems arising from Government regulations (e.g. tax clearance conditions); and · Install unified databases and simple, but informative, monitoring and follow-up procedures with, if possible, comparative data on assisted and non-assisted firms. Some Final Comments on the Objective of BEE 5.71 Central to the whole effort must be support to black business, to start to bring it closer to the productivity levels of European-owned firms. The large majority of the measures discussed above have BEE as an implicit objective, and some as an explicit objective. These include investment grants and loans, small loan guarantee products, microfinance institution-building, productivity and networking funds, small business incubators, technical and management advisory services, Government procurement advice services, training, skills and innovation support funds, and workplace reorganization. Sub-sector support to textiles, motor vehicles and other industries have 156 an indirect BEE effect through sustaining employment and helping the spin-off or outsourcing of ancillary production into small and micro-production units. 5.72 The suggestions for SMME support infrastructure also explicitly take into account the needs of BEE through the mechanisms of: a) decentralization of support programs to local areas, b) the establishment of special private-public funds that could support small and micro-enterprises, and c) the concept of using the support of larger enterprises to upgrade small enterprises, both on the input and the output side, including training, after- sales service, subcontracting and outsourcing. 5.73 However, the social necessity of BEE is not necessarily going to be consistent in the short run with the economic necessity of increasing international competitiveness, especially for enterprises under acute competitive pressure such as textiles and garments. In the short run, BEE aims will have to be especially carefully integrated into the imperative of competitiveness, innovation and economic growth. In this respect, the new broad-based BEE effort uses a flexible and fairly constructive, if somewhat complex, methodology, whereby BEE compliance can be achieved through a combination of ownership change, management and labor force change, and training investment, allowing firms the opportunity to qualify for access to Government contracts through a number of different routes. 157 ANNEX 1 KEY LABOR LEGISLATION Table A1.1: Key Labor Laws and Regulations for the Manufacturing Sector Key Laws and Provisions Regulations Education and Training · Promotes a skills development system that is accessible, equitable Authority, 2000 and and promotes the sustainable formation of the industry's skills base. The Skills Development · Provides for the levying of one percent of payroll through the Levies Act, 1999 General employment requirements to finance a skills development system Employment Equity · To achieve equity in the workplace through equal opportunity, fair Act, 1998 treatment and the elimination of unfair discrimination, and through affirmative action to redress the employment disadvantages of designated groups. Labor Relations Act, · Establishes a single industrial relations system for all employees, No. 66 of 1995 promotes collective bargaining, establishes new procedures and institutions for resolution of disputes and provides for workplace forums. Preferential · Provides for the creation of categories of preference in the award of Procurement Policy contracts to promote development objectives including the Framework Act, 2000 advancement of enterprises owned, managed and controlled by historically disadvantaged South Africans Broad-based Black · Establishes a legislative framework for the promotion of black Economic economic empowerment and empowers the Minister to publish Empowerment Act, transformation charters and issue codes of good practice relating to 2004 procurement criteria, indicators, weightings and guidelines. 158 ANNEX 2 CHARACTERISTICS OF ICA FIRM SURVEY SAMPLE The sample contained 803 firms, with about 75 % (603) in manufacturing. About two-thirds of firms in the sample were based in Gauteng, with the remaining firms in Western Cape, KwaZulu-Natal, and Eastern Cape. The sample was mainly composed of small (10-49 employees), medium (50-99 employees) and large (100-499 employees) enterprises, although about 14 % were very large (over 500 employees). There were few microenterprises (fewer than 10 employees) in the sample, especially in the manufacturing sector. Table A2.3: Characteristics of Sample for Investment Climate Survey Total Number of Firms 803 Geographic Distribution: Enterprise Size: Gauteng 63% Micro (1-9 workers) 5% Western Cape 23% Small (10-49 workers) 27% KwaZulu-Natal 9% Medium (50-99 workers) 23% Eastern Cape 5% Large (100-499 workers) 31% Very Large (500 or more 14% workers) Ownership: Corporate 40% Sectors: Individual ­ 49% Manufacturing 75% European/Caucasian Individual ­ Asian 6% Construction 14% Individual ­ African/Other 5% Trade (Wholesale and Retail) 11% Source: Investment Climate Survey Most firms were owned either by corporations (i.e., other firms) or Caucasian/European individuals and families. Only 5 % of firms were owned by African or colored individuals or families. The small number of African-owned firms, however, appears to reflect the distribution of formal firms. Previous studies have also found relatively few African owned firms in these size classes. For example, in a survey of enterprises in Johannesburg in 1999, whereas 97 % of informal micro-enterprises were black-owned, only 7 % of formal micro, small and medium-scale enterprises were.143 This pattern appears to have persisted through 2004. 143Chandra and others (2001a, 2002). 159 ANNEX 3 SOUTH AFRICAN EXCHANGE RATE TRENDS Figure A3.1: South Africa's real exchange rate has varied, especially against the US dollar. 120 100 80 60 1998 1999 2000 2001 2002 2003 2004 Euro Area Namibia United States Trade Weighted Source: IMF (2005). Note: Bilateral REERs are calculated relative to 2000 (i.e., it is equal to 100 in 2000 for all currencies). The bilateral exchange rates are calculated using the formula from Hinkle and Montiel (1999, p. 45) using CPI as measure of inflation. Real exchange rate differences with Namibia are due to differences in inflation in South Africa and Namibia. 160 ANNEX 4 MICA SAMPLE CHARACTERISTICS About half the sample is located in Tshwane in Gauteng. Microenterprises in this region were interviewed in the central business district of Pretoria and two other areas Soshanguve and Mamelodi. Another third of the sample was from Ekurhuleni, also in Gauteng province. Firms were interviewed in Alberton/Germiston, Tembisa and Katlehong/ Tokoza. The final sixth was from Stellenbosch in Western Cape. Over three-quarters of the firms in the sample were owned by individuals or families of African descent. The remaining enterprises were owned by individuals of European, Asian or other descent (10 %, 7 % and 6 % respectively). Most firms owned by individuals of `other' descent described their owners as `colored'. Enterprises were included in the sample if they had at least one full-time employee (in addition to the owner) and had no more than 10 employees (i.e., to restrict the sample to microenterprises). Most enterprises were on the lower end of this scale, with about three quarters of the sample had between 1 and 4 employees. Firms owned by individuals of European or other descent tended to be slightly larger than firms owned by individuals of African or Asian descent (median size is 4, 4.5, 3 and 2 employees respectively). Firms in Stellenbosch were slightly larger (4 employees) than in the other two regions (3 employees). The median enterprise has 3 employees. Most firms were relatively young--only about one-quarter of the sample were operating before 1994. The median firm was 5 years old. There weren't any noticeable differences in age with respect to ownership, although firms in Stellenbosch were slightly younger than firms in the other regions (4 years for the median firm compared to 5 in the other regions). Table A4.1 Sample Characteristics Total Sample: 240 firms Location Ownership Tshwane 50% African 77% Soshanguve 17% European/Caucasian 10% Mamelodi 17% Asian 6% Pretoria--Central Business District 16% Other 7% Ekurhuleni 34% Sector Tembisa 13% Services 26% Katlehong/ Tokoza 13% Light manufacturing 23% Alberton / Germiston CBD 8% Retail trade 34% Stellenbosch 17% Construction 17% CBD & Other 17% Age Size Less than 5 years old 41% 1-4 employees 76% Between 5 and 10 years 32% 5-10 employees 24% Older than 10 year 28% 161 Most enterprises were in the retail trade and service sectors (about 60 % of the sample in total). An additional 23 % of enterprises were involved in light manufacturing, while 17 % of enterprises were in the construction sector. To ensure that we had enough firms in the manufacturing and construction sectors to compare with firms in the large firm survey (the Investment Climate Survey), firms in the light manufacturing and construction were oversampled. Without this restriction, most of the sample would have been made up of enterprises in the retail trade and service sectors. 162 ANNEX 5 PRODUCTIVITY MEASURES BY FIRM TYPE, MEs IN SOUTH AFRICA Table A5.1 Value Added Wages per Unit Labor Capital per Capital Employ- Sales per Worker Worker Costs Worker Productivity ment Growth Growth All R 9,455 R 2,820 28% R 1,343 4.0 0.0 0.0 Services R 10,999 R 3,453 39% R 2,015 4.1 0.0 0.0 Retail R 7,010 R 2,417 22% R 806 2.7 0.0 0.0 Manufacturing R 8,595 R 1,934 25% R 1,612 2.1 0.0 0.0 Construction R 17,123 R 2,753 33% R 1,220 6.7 0.0 0.0 Registered R 22,160 R 5,802 29% R 3,022 5.3 0.0 0.0 Not Registered R 3,143 R 1,450 26% R 672 2.5 0.0 0.0 Black R 6,446 R 1,934 25% R 806 2.8 0.0 0.0 European R 58,958 R 18,466 36% R 8,058 9.6 0.0 10.0 Asian R 18,936 R 7,252 30% R 6,715 2.8 0.0 0.0 Other R 34,534 R 11,886 35% R 3,022 10.4 0.0 0.0 Tshwane R 14,827 R 3,868 19% R 1,007 2.5 0.0 0.0 27% Ekurhuleni R 2,820 R 967 R 1,007 2.4 0.0 0.0 Stellenbosch R 46,082 R 16,188 36% R 4,029 11.3 0.0 0.0 0-5 employees R 8,259 R 2,417 25% R 1,249 3.0 0.0 0.0 6-10 employees R 24,174 R 7,655 31% R 2,015 7.6 8.3 0.0 Source: ME Investment Climate Assessment Note: All numbers are medians 163 ANNEX 6 BENEFICIARY ASSESSMENT SURVEY METHOD The study took place over December 5th to 15th, 2005. Representative sampling of firms was not a key factor, since the idea was to get more understanding of the specific processes at work in accessing funds. However, to ensure adequate representativeness a randomized approach was used. The survey was restricted to the Gauteng province (where 70% of industrial and commercial firms are located) because of the time limitation and logistical requirements. This may have caused some bias since upcountry firms might tend to be more isolated from Government programs. However the results provide valid indicators. The pool of firms from which the selection was made consisted of the CF impact review control group of 30 firms in Gauteng province (firms that had not accessed the CF), and a pool of about 50 randomly selected firms that had accessed either the CF or the BBSDP programs.144 The sample frame consisted of the following: Table A6.1: Total Non- Participants Number of programs in which firms Participants participated. 145(Total sums to more than 16 * because four firms participated in more than one program. Average participation rate per participating firm is 1.45) CF BBSDP SMEDP SPII Khula SPII EMIA Size > 50 9 3 6 3 wkrs Size <= 50 14 4 0 3 wkrs Age < 10 14 5 9 4 yrs Age >= 10 9 2 7 2 yrs Total 23 7 6 4 4 3 1 1 4 sample * Non-participants were firms not applying for, rejected by, or withdrawing from a program before a grant was disbursed. Some firms were rejected for one or more programs but accepted for others. E.g. two firms were rejected for the NEF but they participated in other programs. 144CF= Competitiveness Fund; BBSDP = Black Business Suppliers Development Program. 145SMEDP = Small Medium Enterprise Development Project; SPII = Support Program for Industrial Innovation; Khula = Khula guarantee; SSP = Skills Support Program; EMIA = Export Marketing and Investment Scheme. 164 Sector Participant Non-participant Food prods 1 Textile goods 1 Printing 1 Plastic prods 1 Chemical prods 3 Mech/electric equipment 3 3 IT products 1 1 IT services 1 1 Other services 4 2 Total 16 7 This table shows the sub-sector breakdown of firms interviewed. Of the two largest categories, metal products and equipment included mining equipment, two heating equipment firms, electronic security devices and auto parts. `Other services' were consulting, therapy, graphics and communications. These sectors are also those with the bulk of non-participant firms. 165 ANNEX 7 QUALITATIVE PERFORMANCE SCORES Table A7.1 ­ Khula RF1 Criterion Comment Score 1. Economic Satisfactory. The strengthening of RFI capacity to achieve 3.0 Rationale sustainability remains a priority in view of its potential to mobilize micro firms. However, a narrower focus on a smaller group of larger RFIs may be more likely to achieve that objective. 2.Target RFI eligibility is based on track record, including appropriate 2.0 beneficiaries/ activity, adequate management, existence of an adequate business eligibility and plan, and disadvantaged group ownership. 25% of applicants selection rejected on the basis of weak management, inadequate business plans or inadequate rollout. However, rejection rate seems too low and criteria should be applied more aggressively. 3. Operating Management appears competent and on top of issues, but 3.0 efficiency: questionable whether RFI unit staffing is adequate for intensive support to future program of up to 12 RFIs. Some problem of organization - e.g. duplication between the institutional support and Khula operations department in assisting RFIs from loan approval to post-investment stage. Some delay in approval of loans to RFIs. Loan applications taking between two and nine months for decision. Need to review the approval and disbursement process. 4.Economic Weak additionality and impact reflected in RFI failures. Khula 2.0 impact, cost- overall depends on DTI capital injections, and subventions to effectiveness fund subsidized loans, write-offs, and technical assistance to clients. Unsatisfactory experience with RFI loan write-offs. RFI scheme not cost-effective up to now, pending consolidation. 5.Verification, Satisfactory performance overall. Management information 2.5 inspection, system in place. Need to strengthen monitoring and evaluation of Monitoring and use of RFI funds, including measurable financial targets for the Evaluation RFIs, and for the ISS program. Probably need to strength credit management system through improved loan tracking. 166 Table A7.2 ­ Ntsika LBSCs Criterion Comment Score 1. Economic Small business support remains a key area for generation of 3.0 Rationale externalities in entrepreneurship and skill development, contingent on well defined approach (e.g. wholesale vs. retail support). 2. Target LBSC accreditations were based on sound principles (e.g. 2.0 beneficiaries/ adequate capacity and sustainability) but were slow and eligibility/ subject to policy shifts (e.g. on link between accreditation selection and financing). Target of LBSC numbers missed. LBSCs selected tended to be distributed unevenly and were too diverse in terms of capacity, race, and gender. 3. Operating With 12 HQ staff trying to manage a program of up to 100 2.0 efficiency LBSCs, the program did not achieve aim of demand driven local service provision. Funding constraints and uncertainty also weakened LBSC enthusiasm. Existing LBSCs did not receive adequate support from Ntsika as it focused on new LBSCs, and Ntsika generally lacked knowledge of issues on the ground. Key partnerships with local stakeholders including local business were variable. Training programs were rated quite highly but more appropriate management and marketing tools and materials were needed, in local languages. 4. Economic Limited additional economic activity or externalities. 2.0 Impact, cost- Outcomes of new LBSCs, new businesses, entrepreneurs effectiveness and employment created were well below targets set out. Combining this with the broad operational problems suggests that cost-effectiveness was weak. 5. Inspection, M For Ntsika as a whole reporting requirements multiplied by 2.3 and E the number of programs required significant dedicated staff. However half of the LBSCs were not submitting reports to Ntsika on time. Quarterly reports and customer surveys were prepared but insufficient attempt to evaluate results. The monitoring and evaluation system was in place but faced problems of underreporting. Too top-down (inflexible) ­ needed revision to encourage `bottom-up' reporting. 167 Table A7.3 ­ The Competitiveness Fund (CF) Criterion Comment Score 1. Economic The objective of entrepreneurship capacity, technology and 3.5 Rationale small business information development remain central to the country's strategy 2.Target Comprehensive set of selection criteria, but without a 2.5 beneficiaries systematic indicator of economic impact. Actual selection eligibility and highly skewed to urban areas. Cancellation rate of 21% selection relatively high compared to rejection rate of 8%, suggesting need to tighten selection 3.Operating Overall management cost ratios reasonable, and responses to 3.0 efficiency project management overall positive. Delays in approval and payment not serious. However, disbursement lag built up from year four resulting in three year extension. 4.Economic Overall satisfactory in terms of direct outcome effect on 3.0 impact, Cost- beneficiaries and spillover benefits; modest BDS market effectiveness development effect. But qualitative results ambiguous on additionality with assisted firms doing worse on job creation and sales growth but better than control group on productivity increases (leading to longer term growth). 5. Inspection, These systems were not fully in place. Verification through ex 2.0 Verification, M ante checks and reimbursements process but not on site. M and and E E system never in place despite plan to introduce it. 168 Table A7.4 ­ The Sector Partnership Fund (SPF) Criterion Comment Score 1. Economic The objective of cluster development remains valid as a tool 2.5 Rationale for collective investment in productivity and information development. 2. Target Selection criteria were reasonably comprehensive but did not 22.5 beneficiaries/ include judgment of likely economic impact. Selection failed eligibility and to identify adequately self-managed partnerships with high selection potential collective investment ideas requiring technological or market links. 3. Operating Management capacity not sufficient at start and systems not in 2.0 efficiency: place. Problems of multiple databases and records, inadequate application, marketing of concept, and delays in approval and documents, disbursement. payment, management quality 4.Economic Sub-marginal result due to relatively high cost of grants (65% 2.2 impact, Cost- of total cost) and management cost compared with modest effectiveness firm productivity results and relatively low linkage and spillover development. 5. Inspection, Verification only carried out ex-ante and through 2.2 Verification, M reimbursement process. M & E system not in place despite and E plans to introduce it. Table A7.5 ­ Black Business Suppliers Development Program (BBSDP) Criterion Comment Score 1.Economic The objective of supporting micro-scale, black-owned 3.5 Rationale businesses to avail themselves of local business services is important both for the improved performance of the businesses themselves and the wider development of the local businesses services market 2.Target Adequate selection criteria were established but there was no a 2.8 beneficiaries/ priori judgment about economic impact, The high (over 50%) eligibility and applicant rejection rate and the low proportion of funding selection requests actually approved (60%) suggest that selection criteria were applied quite aggressively. 3. Operating Reasonably speedy application and processing from 3.0 efficiency: application through approval to payment. There was above average satisfaction with the program, and 90% would re- apply for such a program. 4.Economic Additional activity occurred over and above the Na impact, Cost- counterfactual. Data on derivative potential public benefits not effectiveness sufficient to judge economic impact but a platform existed for potential spillover and market creation benefits. 5. Inspection, M and E system not in place during project. 2.0 Verification, M & E 169 ANNEX 8 ERODING LABOR SKILLS, DECLINE OF APPRENTICESHIPS AND LACK OF INVESTMENT IN RESEARCH AND DEVELOPMENT Skills Development In efforts to increase employment the Government has focused on stimulating the growth of the labor intensive sectors. In industries such as spinning and weaving of textiles, and a range of engineering products including pumps, however, relatively advanced technology (such as automatic looms and computer controlled machinery) is required in order to remain competitive in the global market. The same is true for a broad range of construction equipment and industrial manufacturing. This translates into less labor intensive processes and works against direct increases in employment, particularly of unskilled and minimally skilled workers. Only in limited areas of manufacturing, where low productivity is offset by low wage costs, would low-tech processes remain internationally competitive. Historically in South Africa the construction and manufacturing industry has largely relied on a core of highly skilled staff (generally white and often expatriate) to supervise a largely semi-skilled and unskilled workforce (generally black). However, over the past few decades there has been a reduction in available skills training capacity and closure of industry training institutions in the 1990s. In addition, more reduction in training capacity is foreseen as only about 70 percent of the available training capacity is currently being utilized and there is a continuing decrease in the numbers of people entering technical and skilled professions. The types of skills in demand, as illustrated in the analyses of textiles and engineering products for this study, are as follows: Technical: · Mathematics; · Drafting and design; · Ability to interpret technical documents and apply on the production floor; · Mechanical and structural engineering; · Machine repair, maintenance and operation; · Machining; · Welding and casting; · Metalworking and related crafts skills; · CAD/CAM operations; and · Tooling and dyeing. 170 Administrative and Management: · Mid-level management; · Shop floor supervision; · Cost accounting and finance; · Marketing; and · Quality assurance. Internationally competitive leather seat cover production would require a combination of craft skills and semi-mechanization. The VCA findings are also consistent with findings by Kraak (et al)146 where an assessment was made of areas where employers have the most difficulty recruiting workers. As the figure below indicates, technicians, craftsmen, and managers were identified by firms as the most difficult to recruit. In the absence of an influx of a young, well-trained, technically skilled labor force, the construction and manufacturing sectors are largely reliant on an ageing skills base. In every manufacturing facility visited during in-field interviews there were no journeyman level craftsmen under the age of 50. Much of the industry relies on a semi- skilled and unskilled workforce, under increasingly less capable supervision. This results in slow delivery, significant rework to rectify defects, and associated materials waste which must be built into the tendering and project costs. The absence of skilled labor is contributing to increasing the cost of production to an uncompetitive level on a global basis. 146Kraak, A, Paterson, A. Visser, M G Tustin, D (2000) Baseline Survey of Industrial Training in South Africa, Report commission by the European Union's `Labor Market Skills Development Program', Pretoria: Department of Labour. 171 Table A8.1: Technical Apprenticeships in South Africa 1986 1998 2003 Number of Apprentices 29,800 16,500 19,500 Ave Age of Journeymen 36 years 46 years 51 years Source: Industrial Manufacturing Organization of South Africa and the State of Skills Report (South African Department of Labor) The current lack of skilled workers can be attributed to the decline in apprenticeships and the "poor quality" of technical training at colleges. Research by the Industrial Manufacturing Organization of South Africa has shown a decline in apprenticeships from 29,800 in 1986 to 16,500 in 1998, rising somewhat to around 19,500 in 2003 across the whole of South Africa. While it may be that the upturn sine 1998 signifies an underlying improvement, the overall decline of some 30% in the number of apprenticeships in 15 years, contrasts with a desirable growth in numbers. At a reasonable 3% annual growth rate since 1986 the current total number of apprenticeships would exceed 45,000. On this basis there is a deficit of around 60%. However, taking account of the need for replacing loss of skilled workers through natural attrition, it is estimated that in fact over 40,000 new apprentices would be required now to respond to the skilled labor needs of the sector. In South Africa today, over 40% of the construction and manufacturing sector relies on informal labor, where the sector is increasingly reliant on subcontracting and casual employment. The lack of in-house technical capability and skills base is undermining the ability of SMEs to expand. As more workers are shifted off payroll, the quality of human resource management has also dropped, resulting in poorer wages, less skills development and consequently poorer work quality. The shift from formal to informal and/or contract employment in the construction and manufacturing sector has distanced the workforce from the principal contractors, resulting in an erosion of the employer's direct responsibility for skills enhancement. Many companies indicated that they do not have the certainty of workflow to be able to commit to the contractual commitments involved in apprenticeships. The Government has attempted to address this problem indirectly by formalizing tax allowances of up to 70% of the apprentice's pay through the training levy system.. Employers contributing to the training levy and registered with the government have however cited complicated procedures required to access the fund and the time lag in trying to recoup from the fund. Thus many employers are discouraged from even attempting to access it. The result is that the skills levy has created a significant fund that is underutilized. GDS interviews with companies indicated that 18 to 24 months following application, the employer is still waiting for government reimbursement of costs of hiring apprentices that have now moved on. After one or two experiences like this, manufacturing employers do not hire apprentices with this fund in mind anymore. If they do hire them they do so with funding out of their own company with no plans for reimbursement because of the immediate need for workers. 172 The first National Skills Development Strategy (NSDS) report issued by the South African Department of Labor in September 2002 indicated that approximately 7% of the small businesses paying levies and employing less than 50 employees are participating in the Skills Development Strategy. The most recent data gathered by GDS interviews indicates that not even 1% of companies now participate. This percentage figure is reflective of the realities and difficulties faced by small companies trying to access the fund. In this regard, between 2001 and 2004, approximately 90% of the fund went un-accessed or unused each year, and was reallocated to other government uses the following year. In almost all sectors the transition to the new skills training environment has been accompanied by system bottlenecks such as the development of the new qualification standards and framework. With a total of 8,365 levy paying construction and manufacturing companies in the construction sector, only 56 apprenticeships had been registered in that specific sector by mid 2003. This represents only about 12 percent of the 478 apprenticeships registered across all sectors of the economy, which is grossly inadequate to support an industrial base the size of South Africa. Recognizing relatively short training opportunities that will result from the project-based nature of employment, the Department of Labor has undertaken to keep a record of training delivered, thus creating a platform for further training and a pool of skills potentially available to the industry. = Table A8.2: Proportion of vacancies by major occupational group April 2003 April 2003 (% of total vacancies) Legislators, senior officers and managers 10.0 Professionals 38.3 Technicians and associate professionals 22.4 Clerks 20.4 Service workers and shop and market sales workers 4.5 Skilled agricultural and fishery workers 0.0 Craft and related trades workers 3.0 Plant and machinery operators and assemblers 1.5 Elementary occupations 0.0 Source: State of Skills, South African Department of Labor = 173 = Table A8.3: Proportion of vacancies in professional unit groups April 2003 April 2003 (% of total) Computing professionals 9.1 Engineers 14.3 Accountants and related accounting occupations 39.0 Personnel and careers professionals 10.4 Economists 10.4 All Others 16.8 Source: State of Skills, South African Department of Labor Skills formation in light engineering faces a range of challenges requiring commitment by industry and stakeholders. The low-tech image of the industry, together with deteriorating profitability, is discouraging bright young people from entering the construction and manufacturing professions. Particularly hard-hit is the engineering profession and the potential intake to tertiary education is also restrained by the low percentage of matriculates with higher grade mathematics and science. It is encouraging to note that the percentage of students with higher grade math and science has increased. What is alarming is that even with the improving overall pass rates in math, science and other technical skills trades, the employers in the manufacturing industry have seen a decrease in the actual capability of applicants to the point that they give their own private entrance tests to gauge math and science skills and are subsequently shocked when an applicant with a high school graduation certificate cannot add fractions together. Based on interviews with manufacturing plant operators and managers, it appears that approximately 90% of the people who apply for entry level positions requiring at least a secondary education and/or a matriculation are not capable of basic mathematics. When an individual is initially hired as a basic laborer, there is generally an expectation that over the medium and long term, the individual will make a career progression as a craftsmen and perhaps even higher. Thus all applicants in a manufacturing company are required to possess basic skills that will allow them to progress up the professional ladder to take on newer responsibilities such as machine set up and operator, and quality assurance and control officer. Lack of Investments in Research and Development South Africa is not adapting rapidly enough to the changing production and manufacturing environment. There are global shifts in the development of new processes, services and learning programs designed to improve customer economics and efficiencies. Industry currently invests little into Research and Development (R&D) in either the construction or the manufacturing industries, as no immediate competitive benefit is perceived. Thus, industrial manufacturing processes tend to be particularly labor intensive. 174 The average R&D investment to sales turnover ratio in the United States was 4.3% in 2003, according to the National Science Board's Science and Engineering Indicators. This benchmark has been selected for comparative purposes in assessing South African R&D investment to sales turnover ratio. The current South African total (public and private sector) expenditure on R&D amounts to less than 0.7% with 0.29% of that amount coming from public sources. Similarly, the average OECD country expenditure is 2.15%. Finland, for example, with an economy the same size as South Africa, spends 3.5% from private sources and 1% public sources. The ratio of R&D investment to sales turnover varies considerably in the different sectors, due to the nature of products in the sectors, the sophistication of the products, the pressure level on margins, the manufacturing process and the different types of technologies in the process. Table A8.4: Average R&D Investment to Turnover (TO) Ratio in South Africa Sector Average ratio of R&D Invest/TO Agriculture 1.54% Mining 1.29% Base Metals 0.47% Pulp & Paper 0.60% Power Generation 0.17% Petro-chemicals & Chemicals 0.62% Rubber & Plastic 0.26% Civil & Construction 0.20% Textiles & Footwear 2.06% Glass & Non-metallic 0.63% Food & Beverage 0.24% Electrical & Electronics 0.73% Medical & Pharmaceutical 10.21% Water 0.11% Automotive 2.11% Metal products & Machinery 1.94% S.A. All Sectors Average 0.63% By Comparison - U.S. All Sectors Average 4.30% Source: National Science Board's Science and Engineering Indicators, 2004 175 In considering the possible R&D investment within each sector, based on the most closely related measurement parameter - the ratio of R&D investment to sales turnover and adjusting for any known anomalies in the sector such as high R&D spending for justified reasons, the following broad indications of R&D expenditure can be estimated. As can be seen from the table above, all sectors in South Africa fall considerably below the `US All Sectors Average' of 4.3%, except for the medical and pharmaceuticals industry. The responses of the companies interviewed confirmed the low level of R&D investment level, not only for the manufacturing industry, but also in the country as a whole. Reduced public sector funding over many years has caused an erosion in research capability that has reached critical levels, and which could impair the national capacity to respond to the developing market. In those occasions where industrial manufacturing does engage in research and development of either new products or new production/manufacturing processes the funding is private and usually comes out of ongoing operating funds of the company. This provides little ability to later capitalize the costs of Research and Development against a true product and results in limited operating funds. 176 per three ation over stoc grant only million w funding enterprise/ of ne R3 million to Max per organiz R3 annum years 30% Service 75% machinery up 2 01-05 By 03-04. R39 1999. p.a. over pa phase in total bn bn mn mn since R Financing Government R1.5 R1.4 R11 Cumulative: mn inception for es)fe building projects nda SA East firms in fortn funding ainingrt ulting a,g Bay gra onsc operating rations projects nt grant labor pment workplace and infrastructure opee customs Coe­ Richards international 9 quie NPI workshops, for fre uresd to improve PROGRAMS Description Investme and DTI/ approved (e.g. material to /management practices. Grant improvement Duty simplified proce London, Grant establishing ANNEX nt/ 177 Benefit INCENTIVE smallot veloping relations essn foreign neur no nda tegrationni to investme de exports labor busi sfern Public strategic costs ethnic by to and tra barriers ss vei infrastructure barriers inward entrepre ote busine Intended Reduce firm/ startup/expansi Strengthen growth/employment/export sector Reducing improving in construct Reduce constraints expansion Reduce investment Prom technology INDUSTRY duty AND ins foreign Industry zones TRADE Sector General Motor General Critical Infrastructure Export-oriented industrie free General investment OF Scheme ce Support DEPARTMENT SMEDP MIDP Workpla Challenge CIP IDZ FIG of per to R10 write- of 30% to 006R tax value services x services hip uptn ainingrt to egaw al 25% up to of ma value mn1Rx gra R100,000 of up nu to to ma Up million off 50% advisory up R100,000 65% advisory or partners 80% max 50% costs anfo per bill. 15% rebates mn application xsi d nm overd disbursenm years six 03-4nm 04-5nm pa budget mn mn 140 42 R disburse years R over R360 R490 DOL R45 R35 04 r/ s ning or trai to films export taxd cluste sect6 to program skills eligible of business for fortn Action on base ders gra available produced programming nt- services andsr incubators Fund grant; IA? provi building beneficiaries ntoring ME ssociations/a, and Network sharing me of incentives Investme allowance advisory produce service Matching network Cost and DOL/DTI development SMEDP Part groups councils Tax domestically Strategy Innovation/technology support. specialized 178 in and private entry and portsxe raise improve tion to increase ot black to stment nda rategicts flow, and for training. to strategic returns in port/industryxe inve growth upgrading of s technical vity coordina barriers echnicalt vity barriers of businessesd (priority) oting inducing production and Support growth industrie Improve information producti competitiveness Increase flow information producti Reduce entry/success owne Increasing investment Prom development Reducing and film Target subsectors Entry and and firms scheme sector strategic s uring uring (recipients owned PDE TV Specific ons s MS and Selected industrie General manufact services General manufact services Black General of funding) Sector Assistance Film producti Customized program e SIP CF SPF BBSDP Enterpris Development SSP SSAS Film CSP GODISA to mn RFIs 85% to to 10 mn R R5 fornm facility to required up up no indemnity max 80% 80% 60% to less equity million million R1 for experienced RFIs; R100 experienced Max 90%; R100,000 Max of Max indemnity R1 Max indemnity R5 50-80% indemnity 05/06 06/07nm mn 05-06nm 06-07nm R20 R24.4 R100 R106 (2004) up p.a. nsion bank bank R20 subject for for expa owner- specified llateraloc from R25mn or RFIs for where where for outsourcing, centers loans self-funding for acquisitions/ ailableva ailableva lacking fors call 10,000R eligible NGOs/CBOs micro-enterprise services finance for E to to finance nso not not guaranteed guaranteed guaranteed MS provide stage of max time-bound Incentive including Will to Loans to targets Loans providing advisory Direct startup/acquisition 2nd managers Expansi Loans collateral borrowing Loans collateral borrowing Loans type EIDD) nce nce nce k nk nk nk 179 by fina fina startups fina ris ba ba ba to to to to to to to nitoredo rriers rriers rriers rriers rriers rriers rriers m- ba se ba ba ba ba ba ba s micro- rural/peri-urban SME Reducing for Enterpris Reducing for Reducing for Reducing finance Reducing finance Reducing finance Reducing finance (Independent of rriers small ng ba for micro finance micro- se types enterprises. FINANCE entry Facilitating outsourci projects Reducing to micro General business Smallest business (rural/informal sector) Emerging entrepreneurs SMEs SMEs SMEs Specific SME fund Process ing ENTERPRISE subvention Scheme Scheme neur Apex DTI undF loans vidual Business Outsourc Micro KHULA Total RFI KhulaStart Emerging Entrepre Scheme Equity Indi Guarantee Empowerment Scheme Portfolio Guarantee risk Cover on ficiary R600,000 limit. bene service service limit No depends Max per Free Free Free Free Free Free Free 05/06 06/07nm p.a. mn mn 137 R145 R50 for nks guarantee ba services to missions/ market sory accessing loans nters dvia financing institutions 04 2004 trade ce Ntsika/NAMAC in services building, services and Development with Nov ofr opportunities financial Wholesale to Concessional onlending Started Merge functions Advisory Exhibitions/ fairs Network development Market information Advisory Training Materials Assistance SETAs of EIDD) for ICT 180 by SME for in to growth of entry contract costs inter-firm business business nce to operations to back- firm fits: to tion ent to to facilitate facilitate nitoredo fina SMMEs support nda growth growth bene and and m- funding rriers (SEDA) small/ small bank barriers tryne ba for tion barriers transac barriers barriers enable firmsd na developm market market cost cost To lenders Enable projects Reduce black-owned Consolidating SMME AGENCY Reducing markets owne Facilitating Coordi reducing transactions Reducing market Reducing services Reducing services Reduce Training Reduce training (Independent taln esp. services services FINANCE financial PROMOTION SME sector Rural/environme ICT SMMEs SMMEs- disadvantaged SMMEs SMMEs SMMEs Business Business SMMEs SMMEs s ENTERPRISE fund ENTERPRISE support form Linkage promotion provider provider building ss Re loan ntee DTI ne KHULA Institutional Guarantee Scheme Land Facility ICT guara SEDA NTSIKA Total Tender Procurement Support SMME Business Market opportunities Service networks Service capacity Materials development Aware training Fee firms. to service Free to providers investors and mn communities R80 rural in ation, es service rvice se rnment between (SEDA) services servic services heme heme sc sc nda ryo informe business ce top) advisory visoryda govela links in advis information nchis dvia loc (one-s agencies to award award CENTERS startup and fra entrepreneurs and economicla advisory business nters holding loc ce young National National Business operational 17 Referral network Business training Centralized comprehensive Hand for Assistance municipalities/ development Runs development providing Forging ADVISORY for ent 181 NG growth rural firms entry entry; protecting promotion firm to to ploymme models in business of iph of groupsd to of ACTURI small s)r small role communities reducing costs to rriers PROGRAM costs ba barriers youth ownership neurs ss rural-urban and entry benefits ANUF worke firms services M disadvantage community busine Social disabled Entrepre OF Facilitating (<200 Reducing information and Reducing small Reducing promoting as Public Empowerment disadvantaged in Building networks business communities PARTNERSHIP and black- p. VATE banks es providers ASSOCIATION d croim human proProject Firms, academia SMMEs MSEs owne SMMEs Service Youth business Municipal authorities SME resources Rural Investors PUBLIC-PRI NATIONAL­ support excellence DTI Pilot Disabled entrepreneurs awards SMME NAMAC Total MAC BRAIN FRAIN PROJECT Youth Project GAIN SEHD COMMUNITY CPPP g (2004) cultivatin roles including mn outwarddna black-owned 29 ingr for R expansion, projects historically up enginee and for portersxe and especially entry setting biotech, skills rural for rketsam SA ncee in finance, in to vation and for sci iness entry products skri lly suppliers inno housing) of export entry inton bus to mining,­ and of innovation technical sts D communities co access access ulture,c centers especia for innovation innovative service wome & costs . to/ ccessinga nvestorsi of training to of R and rural incubators flori tech of of groups of (excludingse rriers of Six ICT Two Funding production, upgrading communities product sales access entry ba stsoc foreign costs costs cost/facilitate se costs business tate to Reducing investors Reduce Facili Reduce Facilitating Facilitating Reducing Facilitating Facilitating Reducing business Facilitating/lowering entrepreneurship, disadvantaged Reducing entry firmsll vation inno sma to and D 182 SSAS & investment support R of research nt research assistance trade/ startup costs exhibitions pavilions market investme Specific Intensive Reducing vidual rd/outward National Indi Export Foreign Sector Inwa missions SMMEs ssing, roads, proce network mining, uring, T food IC support services Multisector Rural forestry, manufact water, upport:s access for support ogy access access community Trust neurial ogy and bationu ogy Inc Generator Accelerator Transformer EMIA Technol Market Training Integrated Techno-girl Entrepre a) b) c) Rural GODISA Technol development and Centers Technol Development mn mn1 R80 R industries competitive in entrepreneurs for Financing & & R R of of 183 costs costs product and Reducing D innovation Reducing D LABOR) OF CORPORATION uringt (DEPT Manufac FUND fund DEVELOPMENT for oni in ibilitys Propellere fea YOUTH funds PROVINCE mining Innovat program research development goods Enterpris venture Businessplace for fund fund INDUSTRIAL Support Industrial (SPII) Fund industry Capital study USUBOMVU GAUTENG Gauteng Private-public Anglo-Khula Shoprite ACSA Investec ANNEX 10 TARIFF ON TEXTILES AND TEXTILE GOODS Table A10.1: Duty Rates for SADC Type of Product Normal Rate 2005 2006 2007 Cotton Fiber 1.60 R/Kg 0 0 0 Polyester Fiber 7.5% 0 0 0 Other Fiber 0 0 0 0 Yarns 15% 0 0 0 Woven Fabrics 22% 0 0 0 Knitted Fabrics 22% 0 0 0 Household Textiles 30% 6% 0 0 Clothing 40% 10% 5% 0 Table A10.2: Duty Rates for EU Type of Product Normal Rate 2005 2006 2007 Cotton Fiber R1.60/kg R1.41/kg R1.2/kg R1.01/kg Polyester Fiber 7.5% 6.6% 5.6% 4.75 Other Fiber 0 0 0 0 Yarns 15% 8% 7% 5% Woven Fabrics 22% 13% 12% 10% Knitted Fabrics 22% 13% 12% 10% Household Textiles 30% 21% 18% 15% Clothing 40% 26% 23% 20% Source: Global Development Solutions, LLC 184 ANNEX 11 COTTON PRODUCTION AND MARKET FLOW IN SOUTH AFRICA Producer of Seed Cotton Grinner Seed Lint Imported cotton 35% 65% Oil presser Spinner Cake Oil Yarn Weaver Knitter Textiles Textiles Manufacturer Manufacturer Fodder Products Products Trade Source: Global Development Solutions, LLC 185 ANNEX 12(A) COTTON GINNING VALUE CHAIN The ginning process can generally be divided into the following value adding activities: Drying/cleaning > Ginning > Cleaning/packing > Transportation Diagram 4.12 : Ginning Value Chain for South Africa Municipal Yard Salaries Insurance Travel Rent Costs Vehicles Other 47.5% 28.7% 3.0% 5.9% 3.0% 4.0% 7.9% Seed Drying/ Cleaning/ Cotton Cleaning Ginning Packing Transport Admin 78.3% 4.3% 3.8% 2.8% 1.5% 9.2% Labor Electricity Maintenance 46.9% 34.7% 18.4% Apart from seed cotton, (78.3% of the total value), administrative costs comprise 9.2% and ginning 4.3%. The GOT is approximately 37%. The remainder (60%) consists of cotton seeds, and (2%) waste. Hence, producing one kg of lint cotton requires approximately 2.7kg of seed cotton. Table A12.1: Cotton Ginning Costs (ZAR/kg Lint Cotton) Cotton Ginning Costs 10.97 ZAR/kg $1.68/kg Seed CottonDrying/ CleaningGinningCleaning/ PackingTransportAdminTOTAL Unit Cost 8.59 0.47 0.42 0.31 0.17 1.01 10.97 % of Total 78.3% 4.3% 3.8% 2.8% 1.5% 9.2% 100% Source: Global Development Solutions, LLCTM 186 ANNEX 12(B) COTTON TEXTILE SPINNING AND WEAVING VALUE CHAIN Diagram A12.1 : South Africa's Textile Value Chain Material Labor Electricity Maintenance Overhead Wastage Profit 21.8 % 28.1 % 10.9 % 4.5% 26.3 % 1.4% 6.9% Lint Cotton Combing Twisting Weaving Dyeing Finishing 28.5 % 23.7 % 9.9% 1.4% 17.2% 7.4% Labor Electricity Maintenance Overhead Profit 27.6 % 7.2% 13.8 % 36.9 % 14.4% Depreciation Admin Financing Other 16.8 % 53.5 % 12.7 % 17.0 % Table A12.2: Key Performance Indicators for Clothing and Textiles Clothing Textile International Norm (Clothing) Customer return rate 0.49% 1.9% 0.09% Customer delivery 86.7% 88% 89.4% reliability Absenteeism 7.9% 4.0% 4.7% B&M Analysts 187 ANNEX 13 COTTON TEXTILE SPINNING AND WEAVING VALUE CHAIN : BREAKDOWN OF COSTS The breakdown of costs, by activity, within spinning and weaving is as follows: Table A13.1: Textile Production : South Africa ($/Kg) % of Lint Cotton Combing Twisting Weaving Dyeing Finishing Total Total Unit Cost $ 1.83 $ 1.52 $ 0.64 $ 0.86 $ 1.10 $ 0.47 $ 6.42 100.0% % of Total 28.5% 23.7% 9.9% 13.4% 17.2% 7.4% 100.0% Table A13.2: Benchmarking Textile Production Costs Distribution ($/kg) Labor Utilities Inputs Other Total South Africa $ 1.26 $ 0.45 $ 0.31 $ 2.57 $ 4.59 Cambodia $ 0.44 $ 0.47 $ 4.24 $ 0.42 $ 5.56 Kenya $ 1.27 $ 1.65 $ 1.27 $ 1.24 $ 5.43 Mali $ 0.79 $ 1.68 $ 0.60 $ 1.45 $ 4.52 Global Development Solutions, LLC Table A13.3: Breakdown of "Other" Costs Overhead $/kg % of Total Depreciation 0.21 8.2% Administration 1.01 39.3% Finance Costs 0.28 10.9% Other 0.08 3.1% Maintenance 0.47 18.3% Profit 0.52 20.2% Total 2.57 100.0% Global Development Solutions, LLC 188 ANNEX 14 TRADE TRANSPORT LOGISTICS South Africa's transport infrastructure is largely controlled and operated by the Government-owned Transnet. Several private-sector companies also provide trucking services, and a number of private airlines provide services in competition with South African Airways. Approximately 80% of all freight carried in the country is via a network of national roads (7,200 km). Companies interviewed all report excellent transport access domestically and to neighboring countries. Transport and logistics are not perceived as a barrier to trade. South Africa also has the best-equipped, and most efficient, ports on the African continent, and there are major upgrades currently under way to increase handling capacity.147 The 21,300-km rail route network is the largest narrow gauge national railway network in the world, connecting all major inland agricultural, mining and industrial areas with major ports. The railway infrastructure is easily the best in Africa and approximately 45% of the network is electrified. Spoornet operates 6,000 km of main routes and over 70 branch lines totaling over 12,000 km, with some concessioning. The Airports Company of South Africa (ACSA), privatized in 1998 to a consortium led by Aeroporti di Roma, currently handles 90% of the country's aviation.148 South African Airways (SAA) is the largest commercial airline operating on the African continent. Inland shipment costs average $0.20/km/ton (metric) to major ports. All companies interviewed relied on international customers to provide ship transport from the port, the cost of which is included in contract terms and prices. 147Durban handles about 20% of South Africa's total port traffic and is a hub port for cargo to and from the Far East. Cape Town serves as a hub between Europe and the Americas, and Africa, Asia and Oceania. Port Elizabeth caters to fruit exports, and serves as an entry/exit point for the vehicle industry. 148At the three international airports (Johannesburg, Cape Town and Durban) and six national airports. Johannesburg is equipped to handle all types of cargo through a mechanized terminal. Cape Town and Durban airports have smaller scale facilities. 189 Table A14.1: Cost of Inland Transportation Country US$/Km/Ton India 0.019 Indonesia 0.023 Pakistan 0.024 Ethiopia 0.038 Tanzania 0.057 Kenya 0.059 Mozambique 0.146 South Africa 0.200 Ethiopian Roads Authority, CIA World Factbook, Cement Manufactures Assoc of India Although South Africa's transport infrastructure is relatively well developed and maintained, and is superior to others in Africa, high surface transport operating costs, including labor, and fairly long distances to the major location of industrial activity (Gauteng), translate into relatively high transport costs, as shown in the above table. 190 ANNEX 15 A NOTE ON THE STEEL SECTOR IN SOUTH AFRICA For products like towing balls, steel is a dominant feature in the value chain, and a brief assessment of the steel sector in South Africa is called for. There are several producers of flat products, but Mittal Steel S.A. dominates the market. Table A15.1: Manufacturers of Long Products (including forgings) in South Africa 191 The production of steel in South Africa is expected to be about 9.8 million tons in 2006. Of this, approximately 6.2 million tons is produced supplied by Mittal Steel, SA. Table A15.2: Steel Production (1,000 mt) Year Production Capacity 2001 8,557 10,821 2002 8,758 10,846 2003 9,106 10,846 2004 8,303 10,866 2005 9,840 11,087 2006 9,887 11,118 2007 9,374 11,188 South Africa Iron & Steel Institute For long products, Mittal is still a significant player in the market, but shares the position with a number of other producers. In a global context, production of steel exceeds 1 billion metric tons, thus placing South Africa as the 20th largest producer of steel. However, labor productivity in the steel sector in South Africa continues to remain well below the major global players. It is estimated that it takes approximately 3.7 hours to produce a ton of steel in South Africa while countries like China, India and the U.S. are able to produce steel at a fraction of the time required for a South African facility (refer to the table below). Table A15.3: Benchmarking Labor Productivity of Steel Production Country Hours/ton China 2.8149 USA 0.6 ­ 2.8150 India 0.96 ­ 5.0 South Africa 3.7151 149Investor's Business Daily, February 3, 2004. 150Monthly Review Foundation, January 6, 2006. 151Mittal Steel, 2005. 192 Low productivity in South Africa's steel sector is generally thought to be the result of outdated equipment and poor labor skills. In this context, Mittal Steel has been actively engaged in a rehabilitation program to upgrade some its major facilities in South Africa. While change has been slow, labor productivity has improved from 4.1 hours/ton in 2003 to 3.7 hours/ton in 2004 (refer to the data below). Mittal (Ispat/Iscor) CorporateProfile Financial Profile(millionUS$) Geographic Sales 2003 2004 %Change Revenue 18,487 23,053 25% South Africa HeadlineEarnings 1,605 4541 183% OperatingProfit Rest of Africa Vanderbijilpark 2,179 4,129 89% Saldanha 379 1,147 203% Total Africa LongProducts 777 1,769 128% Coke&Chemicals 172 462 169% Far East Other 10 42 320% OpereatingProfit 2,439 6,668 173% European Union LiquidSteel Production(tons) 2003 2004 %Change North America Vanderbijilpark 3,681,000 3,628,000 -1% 2003 Saldanha 1,251,000 1,227,000 -2% Middle East 2004 LongProducts 2,153,000 2,178,000 1% TOTAL 7,085,000 7,033,000 -1% % 0% 10% 20% 30% 40% 50% 60% 70% 80% SalesVolume 2003 2004 Exports Domestic Total %Change Exports Domestic Total%Change Vanderbijilpark 1,310,000 1,863,000 3,173,000 70% 943,000 2,224,000 3,167,000 42% Saldanha 833,000 354,000 1,187,000 235% 637,000 503,000 1,140,000 127% LongProducts 943,000 956,000 1,899,000 99% 743,000 1,151,000 1,894,000 65% TOTAL 3,086,000 3,173,000 6,259,000 97% 2,323,000 3,878,000 6,201,000 60% Key Performance Indicators 2,003 2,004 % Change Number of full-time employees 12,539 11,416 -9% Man hours/ton (Steel) 4.1 3.7 -10% Revenue/head (Rand) 1,474,000 2,019,000 37% Percentage value-added exports Flat 41 57 39% Long 61 74 21% Key Competitors Market Share (% Total Production) Arcelor 4.8% LNM 3.8% In v o ic e d E x p o rt P ric e s Nippon 3.3% Posco 3.1% Is p a t Is c o r in v o ic e d p ric e s (c & f) U S $ /t Shanghai 2.2% Iscor <1% 7 0 0 H o t ro lled c o il L o w c arb o n w ire ro d Cost Reduction Strategy 6 0 0 Objectives by 2007 $43 - $58/ton Operating efficiency $23 - $31/ton 5 0 0 Raw material & procurment initiative $13 - $20/ton Labor productivity $4/ton 4 0 0 Newcastle PCI Project $3/ton 3 0 0 Throughput Strategy Overall objectives by end 2007 1 Mtpa 2 0 0 2 new DRI kilns (Vanderbijlpark) 325 ktpa Efficiency improvements 660 ktpa 1 0 0 Capital expenditures 1994 1995 1996 1997 1998 1999 200 0 2001 2002 2003 2004 Expand sinter capacity (Vanderbijlpark) R460 million Blast furnace D reline (Vanderbijlpark) 445 kpta S o u rc e : Is p a t Is c o r Blast furnace C reline (Vanderbijlpark) 355 kpta 2 new DRI kilns (Vanderbijlpark) R600 million Steel Pricing in South Africa Given the dominant position held by Mittal Steel in South Africa's steel sector, a lively debate has been taking place over whether the company has been exercising unfair pricing practices. While a number of well-known cases have been heard by the Competition Commission, none have been referred to the higher court ­ the Competition Tribunal ­ as no actual infringement of law was found to have taken place. 193 Table A15.4: Recent petitions for unfair pricing brought before the Competition Commission of South Africa, 2004-2005 Company Petition Ruling Cadac ­ barbecue Unfair pricing practices Import Parity Price is international manufacturer (dominant position) of steel best practice and does not exhibit manufacturers (Mittal) and monopoly pricing producer-merchant collusion (Macsteel International ­ a JV between Mittal and Macsteel) Harmony Gold Ltd ­ Abuse of dominant position "Iscor's prices are comparable to Leading Gold Producer by Iscor (now Mittal) by prices of competing prices available Durban Roodeport Deep locally, as well as to prices charged (DRD) charging excessive prices in other countries' domestic Conveyor Manufacturers markets.... Accordingly the Association Volkswagen South Africa investigation found no evidence that Iscor was abusing its dominance by charging excessive prices." A closer look at figures brought forward by petitioners supports the estimation in the Table A15.4 above that the mark-up of local steel has been in the range of 10% - 20%. At the heart of the submissions from Cadac against Macsteel International152 were Macsteel invoices to Cadac for February 2005, as well as a `sales contract' and pro forma invoice, dated July 19, 2005, generated by Macsteel International for the same steel grade (LPG 275) and issued to a cylinder manufacturer in Ghana. In February, Cadac was being charged R7,388/ton (MacSteel invoiced price on 22/02/2005) while in July, Macsteel International's pro forma invoice to the Reltub Company, in Accra, quoted a price of $498/ton, free on board (fob) Durban (approximately R3,200 /ton) and $563/ton cost, insurance and freight Accra (approximately R3,500/ton). The fob Durban quote was over 50% lower than prices charged to Cadac a few months earlier. While the plaintiffs accepted that some of this price decrease to a foreign buyer compared to a local buyer was related to world steel price decreases, the gap was wide enough to suggest that for anything other than monopolistic and discriminatory pricing had taken place. Cadac also documented and submitted a detailed pricing comparison (on steel type LPG 275 dimension 930 x 2.2) for the months February 2004 and February 2005 respectively. Using an R/$ exchange rate of 6.78 for February 2004 and 5.99 for February 2005, Cadac recorded a South Africa `base price' from Mittal of R4,435/ton for 2004 and R4,899/ton for February 2005. The `base IPP price' was 37% higher than the domestic price and, after discounts provided by Mittal, that gap was reduced to 20% higher than the IPP. 152The formation of MacSteel International BV as a joint venture by Iscor and MacSteel, South Africa's largest steel merchant, the acquisition of Saldanha by Iscor, and subsequently of Iscor by Mittal, as well as the acquisition of the operations of Baldwins Steel by Trident Steel, as well as the merging of the third and fourth largest steel merchants, Baldwins/Kulungile and Abkins Steel. 194 Given the volatility of the Rand and fluctuations in the steel sector, price adjustments have not been immediate, and there is some level of `stickiness' observed in the price movement, particularly downwards. Table A15.5 below illustrates the estimated difference in raw material cost for imported hot-rolled steel, at prevailing average spot export prices of US$570/ton between January to July 2005, and the price provided by local steel producers. It can be seen that the price difference very dependent on the Rand/$ exchange rate. In periods when the Rand was weak relative to the dollar, as in the first six months of 2005, the local steel price was virtually the same as the import price when import parity of transportation and duty is taken into account ­ the difference being 2.4% in favor of imported steel. In periods when the Rand was strong, however, the local price of steel was almost 27% higher than the import price, suggesting that import parity pricing exhibits price `stickiness' and prices have not come down as readily when the exchange rate recovered (local currency strengthened). Since steel-making raw material base (ore and coke) is local, and therefore Rand based, Mittal has often been accused of exercising its dominant position to exploit this `stickiness'. Table A15.5: Hot-Rolled Steel Imported vs. Local Steel at Import Parity Price -IPP, South Africa, Jan-July, 2005 Strong Rand* Weak Rand* Average* Rand/ton FOB Port of Export 1 3,192.0 3,955.8 3,568.2 Shipping and Inland Transport (15%) 478.8 593.4 535.2 Duty (5%) 2 159.6 197.8 178.4 Total Import Price 3,830.4 4,747.0 4,281.8 Average Sales Price, Locally Produced Steel priced at IPP 3 4,860 4,860 4,860 Cost Difference 1,029.6 113.0 578.2 % Cost Difference (over and above import price compared to local price based on IPP) 26.9% 2.4% 13.5% *As per historical forex rates Strong Rand/US$=5.6, Weak Rand/US$=6.94, Average Jan-Jul 2005=6.26 1 Average Spot Export Price for Hot-rolled Coil, Jan-July 2005, US$ 570/ton. Source: World Steel Dynamics. 2 South African Customs Tariffs, Chapter 72, Article 72.0.810 3 Average Sales Price Jan-Jul 2005, priced based on Import Parity Pricing of local steel producers as per field interviews, 2005 195 More importantly, the steel-maker-cum-trader can use informal, non-price, import restraints such as delivery delays and inadequate remedies for problems with quality of imported products, which artificially increase effective import prices and permit inflation of domestic supply prices.153 For exporters of automobiles and components another factor that may lead to overpriced steel is the IRCC rebates under the MIDP. These rebates have not been passed on in lower prices to the consumer but have been retained within the cartelized industry as an extra source of income (and profit) which is in effect a form of subsidy arising from exports.154 This extra income may be partly passed back up the chain since the steel supplier exercises some control over chain pricing and there is an incentive to increase local content value to maximize rebates. Table A15.6: Hot-Rolled Steel Imported vs. Local Steel at Import Parity Price -IPP, South Africa, Jan-July, 2005 (Rand per ton) Strong Rand* Weak Rand* Average* FOB Port of Export 1 3,192.0 3,955.8 3,568.2 Shipping/Transport (15%) 478.8 593.4 535.2 Duty (5%) 2 159.6 197.8 178.4 Total Import Price 3,830.4 4,747.0 4,281.8 Local Steel priced at IPP 3 4,860 4,860 4,860 Local steel supplier discount (675) (675) (675) IRCC rebate 5 (195) (195) (195) Price to component makers 3,990 3,990 3,990 Cost Difference 159.6 (756.9) (291.84) % difference of import and local price incl. all rebates 4.2% (15.9%) (6.8%) Likely Improbable Improbable * Average Jan-Jul 2005=R6.26 per US$. Min US$=R5.6, Max US$=R6.94, 1 Average Spot Export Price for Hot-rolled Coil, Jan-July 2005, US$ 570/ton. 2 South African Customs Tariffs, Chapter 72, Article 72.0.810 3 Average Import Parity Price Jan-Jul 2005, as per field interviews, 2005 4 The discount is based on processed steel quantities exported. A conversion rate of 75% assumed from hot-rolled steel to forged towing balls (approx 326 pieces per ton). This price discount is discretionary up to R900/ton of processed steel. The R675/ton is the specific discount to towing ball manufacturers (at conversion rate of 75%). 5 IRCC rebate estimated at 17.5% of sales price. This rebate system does not apply to other local users of steel products 153According to a recent competition policy analysis, the mark-up over and above the international price is typically around 30 percent. See `The Role of Competition Policy in Economic Development: The South African Experience", Development South Africa Vol. 21, No.1, March 2004. 154See D. Kaplan op cit p23. Since it is permissible to sell these certificates, the value of the IRCCs can then be deduced from their price on the secondary market, which is usually at a discount from face value. 196 Key Automotive Component Market Drivers Despite record global production, many automotive component suppliers are struggling, at breakeven levels. Four major industry trends and market drivers presenting suppliers with both challenges and opportunities include: · Supplier profitability squeeze--unrelenting price reduction pressures from OEMs is a constant feature of the industry and in periods when such pressures coincide with increasing material costs, the effects for suppliers can be devastating. · Systems and module supply--automotive manufacturers are demanding more integrated products and services from suppliers who increasingly emerge as R&D spin-offs of OEMs; these Tier 1 suppliers are then expected to optimize the design and conception of particular components. · Supplier rationalization--as integrated supply chains become standard (in terms of flow of information and business processes rather than vertical integration), the market trend is towards fewer suppliers providing more of the products. · The three largest companies (Delphi, Bosch and Johnson Controls) are responsible for generating 13% of the market's revenues. The top seven companies control almost one quarter of the global market, leading to consolidation of the supply base, collaboration between supply chain partners, and increased specialization. · Longer warranties and increased product reliability pressures the aftermarket-- Decreasing demand and margins in the independent aftermarket segment are forcing many suppliers to reevaluate their product portfolios. · The combination of these four factors as well as general trend of globalization has lead to suppliers being given an increasing share of tasks and responsibilities in manufacturing of vehicles, including their parts. OEMs increasingly focus on car- related services and on reducing the number of suppliers. As a result, they are becoming less involved in manufacturing and assembly, passing the responsibility of developing, manufacturing, and assembling important sections of the car on to their suppliers. Options for Growth The growth of the sector and the presence of multinational heavyweights in the country appear to be, at first glance, a strong enough indicator that despite its distance from some of the major markets, the South African automotive sector provides cost and quality competitiveness in global terms. Yet, as it demonstrated in the value chain analysis, the growth has essentially been driven by the MIDP which masks cost and productivity inefficiencies. 197 The major reason why OEMs have been purchasing from local component suppliers is related to the OEMs' ability to obtain duty rebates on automotive sales in the local market which could then be used to import vehicles for the local market. An example is the export sales of BMW 3 Series, right-hand drive vehicles produced in South Africa. Any component manufacturer supplying BMW 3 Series components to the OEM has been highly dependent on the growth of the local market for other BMW series (5 and 7) that are imported by the OEM and for which duty reductions can be obtained from component manufacturers. In general, the component manufacturer's growth has not had much to do with its intrinsic price competitiveness, but has more to do with the ability of its buyers (OEMs) to sell imported vehicles and spare parts in the local market. Perversely, component manufacturers' exports are thus less related to the foreign demand than to local demand for imported vehicles and parts (except for some special components for which materials are produced from unique local mineral resources - e.g. catalytic converters containing rare metals such as rhodium). The tight correlation between level of imports and exports is pervasive. Any meaningful growth of the auto component sector, outside the boundaries placed on it by the size of the local market, and the MIDP rebates that it generates, depends on price and quality competitiveness, without reliance on OEM duty rebates. Steel-based component manufacturers' may also gain some benefit from the discounts available from Mittal Steel Company under a largely voluntary scheme with DTI to provide a so-called Developmental Pricing Model for steel-based processors in the country. The rebates of about R900/ton vary depending on market conditions and are, by and large, at the discretion of Mittal Steel. The rebates appear, however, to bring prices to about IPP levels, so they do not constitute a subsidy, rather a passing down the chain of the rents available to Mittal through its monopolistic position in steel supply. According to Mittal's Annual Report for 2004, the company contributed R45 million to South Africa Iron and Steel Institute (SAISI) export promotion fund, and a total of R450 million for promoting secondary exports (automotive and non-automotive) of steel-based products and import substitutions. The breakdown of the industry by product categories shows that large-scale OEMs create the bulk of employment and turnover in the industry (27% and 68% respectively). Apart from these the sector has a significant number of small companies (31% classified as other) that produce a range of products but contribute barely 1% of industry turnover. As such, the growth of the South African automotive industry faces risks that come with dependence on a few large multinationals, that is the risk of relocation to lower cost centers and dependence on international developments of the industry. 198 This industry setup also suggests that along with the high dependence of the industry on OEMs, the emergence of a strong SME supply chain, at least in so far as contribution to the overall turnover of the industry is concerned, is yet to be achieved. Few notable exceptions are the vehicle interior product category, with contribution of 10% in employment and 5% in industry turnover, as well as tires and wheels and fuel supply and exhaust systems clusters. Strengthening and supporting the emergence of these clusters is an important growth option of the sector, an option which could significantly reduce the high dependence on OEMs and all the risks that come with it. Figure A15.2: Correlation Trends, Automotive Sector Imports and Exports, South Africa $ 5 . 0 0 $ 4 . 5 0 )no $ 4 . 0 0 illib$SU( $ 3 . 5 0 $ 3 . 0 0 $ 2 . 5 0 lueaVeda $ 2 . 0 0 $ 1 . 5 0 $ 1 . 0 0 Tr $ 0 . 5 0 $ 0 . 0 0 2 0 0 0 2 0 0 1 E x p o r t e d C a r s , P a r t s a n d A c c2 e s s o r i e s 2 0 0 2 0 0 3 2 0 0 4 I m p o r t e d C a r s , P a r t s a n d A c c e s s o r i e s $ 2 . 5 0 ) $ 2 . 0 0 illionb$SU(e $ 1 . 5 0 luaVedarT $ 1 . 0 0 $ 0 . 5 0 $ 0 . 0 0 2 0 0 0 2 0 0 1E x p o r t e d P a r t s a n d A c c e s s o r i e s 2 0 0 2 2 0 0 3 2 0 0 4 I m p o r t e d P a r t s a n d A c c e s s o r i e s Source: Data from UN Comtrade. Analysis by Global Development Solutions LLCTM 199 Table A15.7: Employment and Turnover Contribution by Category, Auto Industry, South Africa, 2004 Product Company Employment Turnover Category Distribution Distribution Distribution Engines 6.57% 3.94% 0.98% Fuel supply and exhausts systems 8.39% 5.38% 6.39% Transmission systems 3.28% 2.97% 1.58% Steering, suspension and breaking 11.31% 4.22% 1.01% systems Tires and wheels 4.38% 12.00% 8.77% Electrical and electronic equipment 9.12% 9.03% 2.77% Body equipment 8.39% 7.38% 2.33% Interior 9.12% 10.19% 5.10% Tooling 2.19% 1.32% 0.22% Other 31.02% 10.56% 1.04% General and specific engineering 4.38% 5.28% 1.67% OEMs 1.82% 27.72% 68.13% Source: DTI 200 ANNEX 16 KEY LABOR LEGISLATION Table A16.1: Key Labor Laws and Regulations for the Manufacturing Sector Key Laws and Regulations Provisions Education and Training · Promotes a skills development system that is accessible, Authority, 2000 and The Skills equitable and promotes the sustainable formation of the Development Levies Act, industry's skills base. 1999 · Provides for the levying of one percent of payroll through the General employment requirements to finance a skills development system Employment Equity Act, 1998 · To achieve equity in the workplace through equal opportunity, fair treatment and the elimination of unfair discrimination, and through affirmative action to redress the employment disadvantages of designated groups. Labor Relations Act, No. 66 of · Establishes a single industrial relations system for all 1995 employees, promotes collective bargaining, establishes new procedures and institutions for resolution of disputes and provides for workplace forums. Preferential Procurement · Provides for the creation of categories of preference in the Policy Framework Act, 2000 award of contracts to promote development objectives including the advancement of enterprises owned, managed and controlled by historically disadvantaged South Africans Broad-based Black Economic · Establishes a legislative framework for the promotion of black Empowerment Act, 2004 economic empowerment and empowers the Minister to publish transformation charters and issue codes of good practice relating to procurement criteria, indicators, weightings and guidelines. 201 REFERENCES AND SOURCES USED Economic Reports IMF Staff Report Article IV Consultation 2004 SA Budget Review 2004, 2005 `Medium Term Industrial Strategy' DTI 2002 `Accelerating Growth and Development: The contribution of an integrated manufacturing strategy'- DTI 2002 `Support Measures for the enhancement of the international competitiveness of South Africa's Industrial sector'. Submission to NEDLAC, Nov 1995 `SME Access to Finance ­ Supply Side Regulatory Review' - Treasury Task Group Report. November 2001 `South Africa's Export Performance: Determinants of Export Supply' ­ L. Edwards and P.Alves. World Bank Informal Discussion Paper 2005/03. University of Cape Town. March 2005 `Investment Climate Reform in South Africa' ­ Brendan Vickers, Presidents Office. Nov 2003. (Contribution to WDR 2005) `Manufacturing Performance and Policy in South Africa ­ A Review.' D. Kaplan. University of Cape Town. TIPS Forum 2003. `Competitiveness and Private Sector Development in Africa' B. Eifert and V. Ramachandran World Bank RPED Oct 2004 World Bank. Industrial Competitiveness and Job Creation Project. Staff Appraisal Report, May 8, 1997 `Counting the Cost of Red Tape ­ for business in South Africa'; SBP, June 2005 Statistical Sources Census of Manufacturing 1996 ­ Statistics SA 1996 Large Sample Survey of the Manufacturing Industry ­ Statistics SA 2001 SA Manufacturing Trends 1993-2003 - Statistics SA Productivity Statistics NPI 2004, 2005 Annual Review of Small Business in SA, 2003, 2004 Papers Barrientos, S and Kritzinger A The global value chain and changing employment relation in South African Fruit. IDS/ Stellenbosch Flatters, Frank `The economics of MIDP and the South African Motor Industry', Queens Univ., Canada, November, 2005 Flatters, Frank `From Import Substitution to Export Promotion: Driving the South African Motor Industry,' Queens Univ., Canada, 2002 202 Kaplan, D Manufacturing Performance and policy in South Africa ­ a review TIPS, DPRU, Univ. Cape Town 2003 McCormick, D and Rogerson, C. "Clothing and Footwear in African (Ed.) Industrialization", Africa Institute of South Africa, 2004 Moodley, S. and Velia, M. `An Analysis of the Recent Trajectory of the South African Clothing Value Chain: Upgrading or Downgrading?' Research Report No. 54; School of Development Studies, Univ. of Natal, November 2002 Roberts, S. and Thoburn, J Globalization and the South African Textiles Industry SDS, UEA, Globalization and Poverty. Disc paper no 9 Rogerson, Christian `Successful SMEs in South Africa: the case of clothing producers in the Witwatersrand', Development South Africa, Vol.17, No.5, December 2000 Rogerson, Christian The impact of the South African Government's SMME programs: a ten year review (1994 -2003) Development South Africa, Vol. 21, No 5, December 2004 Rogerson, Christian `Tourism SMMEs in South Africa: a case for separate policy development', TIPS, Univ Witswatersrand working paper 2 ­ 2005 Rogerson, Christian `Pro-poor local economic development in post- apartheid South Africa: The Johannesburg fashion district', 2004 Westhuizen, C Trade and Poverty: a case study of the South African clothing industry Program evaluation reports `Support Measures for the Enhancement of the International Competitiveness of South Africa's Industrial Sector' Government of SA Report to NEDLAC, Nov 1995 `Impact Review of the SMEDP' BSSA, LHA, and BSSA. For the DTI. July 2004 `Supporting South Africa's SMME Strategy' Khula EF and Ntsika EPA; Discussion points and the Way Forward'. Report to the World Bank. ICC. July 1999 `Impact Study of Khula Programs', Source 2000? `Synthesis Report on LBSC Evaluation', Ntsika October 2004 203 `Review of SA Local Business Service Center Program' report for the IDRC Canada, R. Bloch and S. Daze, June 2000 .Annual Reports, Brochures SMEDP Program Guidelines NEF Annual Report 2004 Khula Annual Reports ­ 2002, 2003, 2004 Ntsika Promotion file SPII Annual Report 2003, 2004 SPII Promotion File Workplace Challenge Annual Report 2003-4 DTI/EMIA National Pavilions Guidelines. April 2005 GAUMAC annual report 2004 GEP Promotion File GEDA promotion file GODISA promotion file Softstart Promotion File 03/01/2007 4:00 PM 204