Document of The World Bank Report No. 26444-KE KENYA TRANSPORT SECTOR MEMORANDUM VOLUME II KENYA TRANSPORT SECTOR MEMORANDUM Volume 2 SECTOR PAPERS 2 KENYA STRATEGIC REVIEW Pages Section I: Kenya Roads Executive Summary i - vii 1. Overview of the Sector 1.1 Role of the Road Sector I 1.2 Size of the Network 2 1.3 Condition of the Network 2 1.4 Traffic on the Network 4 1.5 Responsibilities for the Network 6 1.6 Financing of the Network 7 1.7 User Costs and User Charges in the Road Sector 7 1.8 The Road Program 9 1.9 The Use of the Road Network 15 2. Strengths and Weaknesses in the Road Sector 2.1 Sector Strengths 19 2.2 Sector Weaknesses 20 3. Critical Issues: Development of the Kenya Roads Board 3.1 Sector Strategy 24 3.2 Financial Control 25 3.3 Procurement 26 3.4 Financial and Technical Auditing 27 3.5 Compliance 27 3.6 Manual of Procedures 28 4. Critical Issues: Sector Ownership and Responsibilities 4.1 Districts 29 4.2 Major Urban Areas 30 5. Critical Issues: Effective Road Implementing Agencies 5.1 The Roads Department 32 5.2 The Road Agency Model 33 5.3 The Time-Frame for Change 37 6. Critical Issues: Effective Road Maintenance Delivery 6.1 The Main Road Network 38 6.2 Urban Roads 39 6.3 Rural Roads - Re-launching Roads 2000 40 7. Critical Issues: The Funding Gap in the Road Sector 7.1 Maintenance Funding 43 7.2 Capital Funding Needs 45 7.3 Sources of Capital Funding 46 3 8. Critical Issues: The Roles of the Public and Private Sectors 8.1 The Traditional Model 50 8.2 An Evolving Model: Public-Private Partnership 50 8.3 Evolving Donor Assistance 52 8.4 Acceptability in the Kenyan Context: Ways Forward 52 9. Critical Issues: Road Safety in Kenya 9.1 The Problem 54 9.2 Development of an Institutional Framework 54 9.3 Possible Safety Interventions 55 9.4 Enforcement of Traffic Regulations 56 Section II: Kenya RaDlways 57 1. Background 58 1.1 Railways 58 1.2 Magadi Railways 58 2. The Decline of an Institution 60 3. Present Status of Kenya Railways 63 3.1 The Basics 63 3.2 Rail Traffic 63 3.3. KR's Operations 64 3.4 Locomotives and Rolling Stock 65 3.5 Infrastructure 66 3.6 Signaling and Communications 66 3.7 Finances 66 3.8 Insolvency 68 4. The Concessioning of Kenya Railways 70 4.1 Vertically Integrated Concession 70 4.2 Monopoly Railway Operator 70 4.3 Public Service Obligations 71 4.4 The Potential for a Concessioned Railway 71 4.5 Outlook for Concession Fees 72 4.6 Expediting the concession 72 5. Short Term Actions 75 5.1 Introduction 75 5.2 Available Wagon Fleet 76 5.3 Improved Wagon Turnaround 76 5.4 Reduce the Wagon Fleet 77 5.5 Better Scheduling of Wagon Repairs 77 5.6 Inproved Container Wagon Utilization 77 5.7 Improved Train Operations 78 5.8 Realistic Mainline Locomotive Fleet 78 5.9 Shunting 78 5.10 Infrastructure Maintenance 79 5.11 Staff Rationalization 79 5.12 Avoiding Insolvency 80 4 Section Im: The Port of Mombasa 89 1. Background 90 1.1 Competitive Position and Perfornance 90 1.2 Infrastructure 90 1.3 Traffic 91 1.4 Shipping Services 95 1.5 Inland Transport 97 1.6 Roles of the Public and Private Sector 98 2. Issues 100 2.1 Level and Quality of Port Services 100 2.2 Safety 105 2.3 Security 106 2.4 Port Capacity 107 2.5 Port Finances and Staffing 109 3. Actions 112 3.1 KPA Actions 112 3.2 KPA Plans 112 113 4. Development of the Container Terminal 117 4.1 Objective for Private Sector Participation 117 4.2 Future Shipping Patterns 118 4.3 Economic Needs and Development Constraints 119 4.4 The Capacity of the Existing Terminal 120 4.5 Investment in Equipment 120 4.6 Private Sector Management 121 4.7 A Scenario for the Container Terminal 122 5. Future Role for the KPA 125 Section IV: Civil Aviation 127 1. Introduction: Global Aviation Situation 128 1.1 Cyclical Downturn and September 11 128 1.2 The US Industry Post September 11 128 1.3 The Global Industry Post September 11 128 1.4 Trends in Air Transport Regulation 129 2. The General Structure of Civil Aviation in Kenya 130 2.1 Background 130 2.2 Existing Institutional Arrangements 130 2.3 Possible Change to the Institutional Arrangements 132 2.4 Air Transport Service 134 3. The Airline 137 3.1 Kenya Airways 137 3.2 Flamingo Airways 140 3.3 AirKenya 141 5 3.4 Regional Airlines 142 3.5 Foreign Airlines 144 4. The Airports 147 4.1 General 147 4.2 Kenya Airports Authority (KAA) 147 4.3 Nairobi Airports 148 4.4 Mombasa 148 4.5 Eldoret 149 4.6 Assessment of KAA 149 5. Aviation and the Economy 151 5.1 Air Cargo Industry 151 5.2 Cut Flowers and Fresh Produce 151 5.3 Flower Exports: an example 152 5.4 Horticultural Exports: an example 153 5.5 Tourism 154 6. Conclusions' and Recommendations 155 Exchange Rate: Ksh.78 = US$ This Transport Sector Memorandum was prepared on the basis of missions in November, 2001 and mid- 2002, by Mr. Simon Thomas (Senior Transport Economist) in collaboration with Mr. Josphat Sasia (Operations Officer, AFTIR), Mr. David Rudge (Senior Road Engineer, AFITR), Mr. Yash Pal Kedia (Principal Railways Engineer, AFTITR), Mr. John King (Aviation Consultant) and Mr. Paul Thompson (Port Consultant). The Road Sector Review was undertaken with the active participation and support of the DFID, EU, KfW and SIDA. The views and recommendations contained in the Transport Sector Memorandum are those of the review team and are not necessarily endorsed by the Management of the World Bank i SECTION 1 THE KENYA ROAD SECTOR ii EXECUTIVE SUMMARY Purpose of the Review The govemance structure of the road sector was substantially modified in 2001 by the creation of the Kenya Roads Board (KRB) and the District Roads Committees (DRC), and legislated allocation of Fuel Levy funds to different road categories. As the development partners in the road sector had not undertaken any comprehensive review for a very considerable period, this seemed a particularly opportune time to assess present conditions, the constraints faced by the sector, and the effectiveness of the existing policies and structures. The review does not propose a strategy, though it is clear that the 1997 Strategic Plan for the Road Sector needs substantial revision to take account of the KRB Act and to become an effective instrument for guiding development and expenditure. The development of a Sector Strategy is the role of the KRB working with the Ministry of Roads and Public Works (MORPW) and the road implementing agencies. While not presenting a strategy, the review does identify options for addressing the critical issues and thus hopefully contributes to the development of the appropriate way forward to sector sustainability. The Kenya Road Sector The basic characteristics of the sector can be summarized as follows: * Roads and road transport are crucial to the economic and social development of Kenya; the railway could have a role along the main corridors but road transport will remain the predominant mode for both people and freight * Data on the size, condition and use of the network are either incomplete or out-of-date but, in broad terms, the sector can be categorized as: Sector Daily Traffic Kms Network Need Main urban network Very high 3,000 High capacity and service Main road network >400 6,000 High service quality Secondary road network >75 9,000 All weather Rural road network <75 100,000+ Passability * The service condition of the main road network has improved recently, about 63% is in good/fair condition, but much of the network needs rehabilitation or reconstruction. * Road conditions in the main urban areas have also improved with rehabilitation and improvement funded by the Fuel Levy and the Kenya Urban Transport Improvement Project (KUTIP). * Constituency funding through the KRB Act is beginning to have an impact on rural roads, many of which have seen no maintenance for years. * Domestic funding for road maintenance has substantially increased in the last 10 years, and now mobilizes about US$110 million annually. But, this increase has been largely offset by reduced development funding and the net total increase has been small: 1989-95 US$137 million (KSh.10.7 billion): 1996-01 US$145 million (Ksh. 11.3 billion). * Much of the Fuel Levy funding is used for road rehabilitation/reconstruction. There is no regular routine maintenance but, recently, priority has been given to patching paved road potholes. * Substantial over-commitment of Fuel Levy funds resulted in major waste and a huge pending bill and claims liability. iii * The KRB Act redefined responsibilities in the sector but needs modification, especially for the urban road networks. * About 2,800 people are killed on Kenya's roads each year, and another 9,500 seriously injured. There is little enforcement of traffic regulations, despite the numbers of enforcement agents, and no coherent and coordinated traffic safety program. Strengths and Weaknesses of the Road Sector Strengths: > The Kenya Roads Board: provides the opportunity for major improvements in policy direction, expenditures priorities and accountability. > Road maintenance funding the resource mobilization for road maintenance is a major achievement; road users must now obtain value-for-money. > Decentralization of funds, decision-making and implementation: the DRCs are working reasonably well and provide an important level of local decision mraking. > Rural roads strategy and labor-based technology: The Roads 2000 Strategy has the potential to have a major positive impact upon rural roads, employment and poverty. > Well qualified engineers in both public and private sectors: Kenya has as good engineers as any country. The engineers know what should be done, and how it should be done. > Established construction and consulting sectors: Kenya has well established large construction companies and consulting engineers, small contractors throughout most of the country and a dynamic private sector. > Axle-load control: while not perfect, axle-load control has reduced the gross vehicle overloading prevalent in the early 1990s. Weaknesses: > No sector strategy: without a strategy, it is impossible to develop expenditure priorities and monitor impacts, and makes it very difficult for stakeholders to "buy-in". > Inadequate capitalfunding while recurrent funding is more or less sufficient, it is used for quasi- rehabilitation works to buy time until full rehabilitation is possible. Major capital funding is needed to bring the network back to a maintainable condition > Institutional malaise: the erosion of the institutional capacity of the road implementing agencies is the most pernicious weakness of the sector. The present institutions could not effectively and efficiently implement a major roads program. > Absence of management and planning systems: planning and reporting systems have ceased to function, replaced by firefighting "management by surprise". > Lack offinancial control: the over-commitment of the Fuel Levy has been a disaster, resulting in large additional costs for low priority road projects. > Delayed and inappropriate auditing: the Auditor and Comptroller-General provides little effective control in the sector. > Inadequate funding and responsibilities for urban roads: urban road users generate 50% of the Fuel Levy but urban roads receive no specific maintenance funding and implementation responsibilities are ill-defined. iv > No implementation strategy for Roads 2000: there is little evidence of pro-active leadership, ownership or a coordinated strategy to implement the approach. > Dubious construction and doubtful supervision: while much the roadwork may be well executed according to specification, there is a widespread perception that road users have not received value-for-money. There is too little accountability, and poor performance is no bar to future work. > Non-enforcement of traffic regulations: without effective enforcement of traffic and vehicle regulations, the carnage on Kenyan roads will continue. It is not a lack of skills or equipment, simply a lack of political will. Critical Issues in the Road Sector Development of the Kenya Road Board * Sector Ownership and Responsibilities * Effective Road Implementing Agencies * Effective Road Maintenance Delivery * The Funding Gap in the Road Sector * The Roles of the Public and Private Sectors * Road Safety in Kenya Development of the Kenya Roads Board Strategy and Selectivity should guide the KRB: Without a coherent sector strategy, the KRB has little chance of influencing the direction of expenditure and the achievement of consistent objectives, and little opportunity to monitor impacts and hold implementing agencies accountable. Selectivity because the resources of the KRB are limited. KRB needs a clear separation of powers and responsibilities from the implementing agencies. The core functions of the KRB should be: * Development of a strategy for the sector and then overseeing its implementation * Financial control to ensure no repetition of previous over-comnmitment * Financial and technical auditing with all reports placed in the public domain. * Ensuring compliance with the work programs and contract specifications Involvement in procurement would have very profound impacts on the KRB. There may be arguments for a no objection role, but it could blur responsibilities and seriously compromise accountability and should probably be avoided. Sector Ownership and Responsibilities Prior to the KRB Act, responsibilities in the sector were clear: * central government was responsible for the classified road network, and * local authorities were responsible for the rest. Now, the central government is responsible for a much smaller network of national roads, while DRCs have responsibility for the rest of the network. DRCs, however, do not consider that they have responsibility for paved or urban roads. Moreover, DRC priorities are established by constituency and inter-constituency roads may receive inadequate priority. Meanwhile, local authorities still perceive ownership of unclassified roads and allocate funds (admittedly limited) for road maintenance and v improvement. Amendments to the KRB Act are required to align responsibility with perceived ownership: Districts and small urban areas: realigning responsibilities for rural roads and small urban areas should not be an insuperable problem: * Determine the size and nature of the network within the districts: reliable data are very limited and an inventory is needed to identify important rural and town roads * Re-assign the network to those institutions that perceive ownership. * Re-align funding to cover those important roads, which presently receive no funding. District funding could be earmarked for designated town roads as well as rural roads which require agreement within the DRC on a work program. * Roads not included under the DRC and Fuel Levy revert to communities and/or local authorities Large urban areas: the district solution is not appropriate for Nairobi and the other large urban areas - Mombasa, Eldoret, Kisumu and Nakuru - which have extensive networks which are not 'owned' by the DRCs. These networks are costly (possibly US$10-1 1 million annually) and need to be managed as part of the overall urban area. These major urban areas need specific implementing arrangements which recognized that poor performance of local authorities. Effective Road Implementing Agencies The KRB and the Fuel Levy offer the potential for effective governance and funding but they must be supported by effective implementation. The Roads Department supervises an extensive emergency repair program but has suffered from such institutional erosion that it is extremely difficult to believe that it could effectively implement a major capital program. Roads are big business, with a turnover of at least US$150 million (Ksh.12 billion), but they are not managed as a business and road managers have neither incentives nor sanctions. Kenya needs more effective and commercial implementing agencies on a more effective and commercial basis. Many countries have faced the same problem of institutional decay and most have moved to the agency model'. Kenya needs to determine the number of agencies required and how they should be established: * Unitary non-urban road agency: the Roads Department could be replaced by a single agency. It would be understandable and might be desirable, but there could be greater advantage in having more focused agencies; o Highways agency: responsible for the main road network (A, B and C roads, or the core main paved network), with activities directed on a regional basis; o A rural roads agency: to provide implementation for the DRCs and provide the platform for re-launching Roads 2000; o RASTRA: the recent KRB workshop proposed a Rural and Small Towns Road Agency which would combine the rural roads agency and implementation support for designated urban roads outside the big five urban areas. * Urban road agencies: a specific agency for Nairobi but for the other large urban areas, there are a number of alternative approaches including a single agency, strengthening the engineers department, or network contracts with large contractors or consultants. The move to the road agency or authority model was reconmmended by the major workshop in 1992, and was included in the Interim PRSP and in many drafts of the PRSP. vi Two basic approaches can be adopted in establishing new agencies; the first is to transfer staff from the existing agency, and the second is to recruit staff on an open basis. The first approach may cause less pain and disruption but would be ineffective without new management able to replace staff unable to adjust to the new enviromnent. Renaming the Roads Department would achieve little or nothing. Effective Road Maintenance Delivery The Main Road Network2: The comprehensive force account system has collapsed but no systematic replacement has been introduced. Periodic maintenance is contracted out while recurrent maintenance is undertaken by a combination of force account, using casual labor, and single activity contracts. Comprehensive maintenance contracts have been discussed, for several years, but there has been no progress in introducing them. This situation cannot persist and a strategic choice must be made: * Return to force account: little advantage in trying to revive the system * Expand present delivery methods: possible, but for routine maintenance, much greater management effort and systems would be required * Input-based maintenance contracting: the contractor paid according to the volume of work undertaken * Performance-based maintenance contracting: the contractor paid a flat fee to meet specified standards. Failure to perform results in deductions from the monthly fee. Input-based contracting would be a major step. Most of the world has already taken this step and many have moved to performance-based contracting. Perhaps Kenya can move directly to performance-based contracting and avoid the intermediate step. Rural Roads: In view of the enthusiasm for Roads 2000, it is depressing that only 14 districts were covered by 2000. If Roads 2000 is the strategy, it must be Kenyan-led rather than donor-driven. Leadership could provide coordination, standardization of documentation and reporting, and possibly also the development of the "basket of funds" approach to funding. Leadership should also promote Roads 2000 as the appropriate technology for the Kenyan environment and rural economy. Increased donor funding is important, but re-launching of Roads 2000 needs an inclusive approach, under which even those districts without donor funding are covered. The Funding Gap in the Road Sector Maintenance Funding estimates of funding needs for the classified road network average around US$120 mnillion (Ksh 9.4 billion), lower if Roads 2000 is adopted. Maintenance needs of the urban road networks is estimated at about US$14 million (Ksh 1.1 billion). Overall, between US$120 - 130 rnillion is needed once the network is maintainable. This is not greatly different from present funding, and sufficient funding would certainly be available if vehicle licenses were incorporated into the Fund. Capital Funding: About US$1 billion (Ksh.80 billion) is required to rehabilitate the network, and additional funding would be desirable to upgrade heavily trafficked unpaved roads. It is very unlikely that either GOK or donors can provide funding of this magnitude within the time required. To bridge the gap, other sources of funding will be needed Bridging the Funding Gap: road users are the logical source for the funding as, otherwise, they will pay through deteriorated roads and higher vehicle operating costs. If governnent and donors cannot provide the funding, Kenya could try to attract private sector capital finance, serviced either directly through tolls or through an increased fuel levy. Long-term perfornance based rehabilitation and maintenance contracts 2 Thesituation is very sinilar for the urban road networks vii could be funded through an agreed schedule of annual payments, or maintenance concessions could be funded by shadow tolls. While private sector capital funding is expensive, it may be preferable to a deteriorated main road network3. The Roles of the Public and Private Sectors Until recently, the almost universal model for the road sector was public ownership and management. The government would finance, plan, and supervise the sector, often also undertaking much of the work. The private sector would construct roads financed, designed and often supervised by government. This model is now changing to a public-private partnership with the private sector undertaking more of the roles traditionally performed by government. Restrictions on public sector borrowing has led many governments to seek private finance. More generally, the private sector is playing a much larger role in road management, taking long-term responsibility for networks of roads for annual payments. The government remains crucial but its role has shifted from management and implementation to monitoring and supervision. In Kenya, the role of the private sector has expanded but entirely within the traditional model. Recent studies have indicated, however, that it would be possible to concession the entire main paved network to the private sector through either full (toll) or maintenance (shadow toll) concessions. This could transform management of the sector. However, the private sector would be unlikely to finance all the rehabilitation/reconstruction costs for maintenance concessions and might well take a very conservative view toward country risk, especially regarding the payment of shadow tolls. But, donor assistance is also evolving and donors (or at least the IDA) is prepare to provide funding for private sector concessions. New instruments have been designed to lower the risks, and thus costs of private sector finance, through the provision of partial risk guarantees to protect the private investor from government default. Under the new model, the private sector is committed for the long-term. This would be a major (and possibly unwelcome) change for the Kenyan private sector which has enjoyed profitability without accountability. Both the public and private sectors could enjoy significant benefits under the new model, unfortunately there is, as yet, too little knowledge within Kenya to have an informed debate. Road Safety in Kenya After every major bus/matatu tragedy, there is a brief outpouring of outrage, but the tragedy is soon forgotten, by all but the victims and their families, and nothing is done. Government seems resigned to impotence with respect to road accidents, deaths and injuries. Road accidents result from the interaction of human behavior, vehicle condition and road conditions, all of which can be influenced by public policy if effectively implemented. Reducing the overloading of buses and matatus, limiting speeds, and checking tire condition could have a dramatic impact on the number of road fatalities (vehicle passengers account for 40% of road deaths). Unfortunately, the underlying reasons for the lack of effective enforcement of the relevant laws are no mystery for any Kenyan. Unless there is real political will, the situation will not change and efforts to improve road safety will have to be based on the assumption that laws will not be enforced. 3.The typical cost of private sector finance is about 15% (a blend of equity and loan finance) which must be compared with 30% which is the typical economic return on investment for the rehabilitation of roads with > 500 vehicles/day. viii While enforcement of traffic laws would be the most cost-effective way to reduce accidents, other interventions can make a contribution to the reduction of road deaths and injuries: * Preventive engineering and black spot improvement programs: these interventions could have a significant impact but there is little understanding or interest within the Roads Department. Road safety audits should be undertaken for all new road projects, and police traffic data should be processed to identify black spots. * Road marking: while marking is included in all contracts, roads are rarely marked, and traditional materials have a very limited useful life. Thermoplastic paint has a substantially longer life and should be mandatory for all new contracts. * Separation of pedestrians and traffic: well maintained walkways in urban areas and well maintained shoulders on inter-urban roads could have a substantial impact * Driver licenses: introduce smart cards to reduce fraud and control bus and matatu timetabling. * Vehicle inspection: privatize to designated garages and extend to all vehicles, older than perhaps three of four years * Education: revise the curricula and then re-introduce traffic education into all schools The first steps toward a coherent road safety program should be: (a) political will to tackle the problem; and (b) establishment of a Road Safety Agency or similar institution, bringing together the key public and private sector stakeholders to develop and implement annual action plans. STRATEGIC REVIEW: THE KENYA ROAD SECTOR 1. OVERVIEW OF THIE SECTOR 1.1 The Role of the Road Sector It is hard to over-estimate the importance of the road sector to the economy and people of Kenya; also, the Northern Corridor to Uganda has critical importance for much of East Africa. Outside the main corridors, roads and road transport have provided the only means of commercial motorized transport. Along the main corridors, rail transport used to provide the main mode for longer distance transport but its role has faded; first on the branch lines and, since the 1970s, increasingly on the main lines. * Total rail transport in 1973 was 4.6 million tones, and rail carried 95% of the freight to Uganda. Byl997, total rail freight had fallen to 1.6 million tonnes, though it has now risen to about 2.4 million tonnes; * The branch lines carry no freight traffic with the exception of those to Magadi, Kisumu, Thika and very occasionally Kitale; * Rail currently moves about 1.5 million tonnes on the line between Mombasa and Nairobi, road transport moves about 6.5 million tonnes, 80% of the market; * In the early 1990s, there was a daily passenger train to Kisumu, two trains a day to Mombasa, and daily branch line services to Butere and Taveta: in 2002, there is a passenger service four times a week to Mombasa and a twice weekly service to Taveta. In comparison, there are about 35 buses and 180 matatus in each direction between Nairobi and Mombasa, as well as a frequent air service (>10 daily services). Unless the Govermment imposes public service obligations (PSO) on the rail concessionaire to provide long distance passenger services, it is unlikely that even the present limited passenger service will be provided.4 There has been' a substantial increase in the network of domestic air services to Mombasa, Kisumu and Eldoret, bus services provide a frequent, low cost service throughout Kenya, and matatus provide a slightly higher cost but even faster road service Passenger services are generally unprofitable for rail operators in developing countries, imposing substantial track costs above those needed for a freight only operation. The privatization of Kenya Railways' operations should result in a substantial diversion of long distance freight traffic from road to rail. The railway might regain its 50% share of the Mombasa: Nairobi traffic but, even with a revitalized rail freight service, roads and road transport will remain Kenya's primary mode for commercial freight transport. Kenya is acknowledged to have a very competitive road transport industry, but its road infrastructure is regularly rated as one of the country's major constraints to economic development. A constraint both to large scale industrial production and smallholder agricultural production. Improved roads may not generate faster economic growth, but they are necessary for ftat growth to take place. 4 It is difficult to see a PSO justification for inter-urban passenger rail services. Rail is always paralleled by road, and the differences in cost between road and rail are small: rail to Mombasa Ksh. 350, bus to Mombasa Ksh. 400. The recent rehabilitation of the Taveta road has removed the PSO need on this route. It is probable that the concessionaire will be required to continue the Nairobi commuter service. 5In terms of total tonnes transported, however, headloading may be more important, but the distances are relatively short. 2 1.2 Size of the Network Data on the road sector are limited, even the size of the network is not really known. Table 1 is derived from the Strategic Plan for the Roads Sector (Strategic Plan) and data from the Ministry of Local Govermment. Table 1: Kenya Road Network (kms) Paved Gravel Earth Total % Paved Roads Department 6806 4768 2648 14222 47.90/o Other Classified Roads 2131 22413 24524 49068 4.3% Total Classified Roads 8937 27181 27172 63290 14.1% Nairobi 1643 211 1854 88.6% MSA 258 88 346 74.6% Eldoret and Kisumu 166 204 370 44.9% Other KUTIP 290 1837 2127 13.6% Other Municipal 99 3367 3466 2.90/o Other Towns 35 6330 6365 0.5% Total 'Urban' Roads 2491 12937 14525 17.1% County Councils 7 116254 116261 0.0% 11435 182644 194079 5.90/o The network of urban and unclassified rural roads is reported to be twice the length of the classified road network and Kenya's total network is close to 200,000 kms. If such a network exists in Kenya, the definition of 'road' is stretched to include any path or track down which a vehicle ever passes or has ever passed. It is inconceivable that such a network could be maintained in the foreseeable fiuture; maintaining a network of half this length would be a major achievement. In reality, the unclassified network often exists only on paper (and sometimes very old paper), as it is based on the 1:50,000 Survey of Kenya maps (some prepared pre-1970), rather than a physical inventory. Verification, through field spot checks, indicated that only 30 - 50% of the links shown on the maps were actual roads6, the rest were tracks or footpaths. Only a few local authorities could provide accurate estimates of their networks, backed by maps or drawings. Basically, the total road network length is not known, but may fall within the range of 100,000 - 150,000 kms. This uncertainty is hardly conducive to rational planning or resource allocation. 1.3 Condition of the Network Comprehensive information on the condition of the network does not exist and the data, which are normally included in reports, are derived from partial surveys undertaken in the early 1990's. The last full condition survey of the classified network was made in the late 1980's and a new inventory and road condition survey is only now underway and the results are not expected until 2003. The following sections bring together the information which is available and commonly quoted. 6 Final Report: Local Authorities Road Inventory Project, Vol. V, Uniconsult (Kenya), 1975 3 (a) Classified Road Network: Table 2 Condition of the Classified Network (%) Paved Gravel Earth Year Reported: 1993 1994 1995 1994 1994 Good 8 21 29 2 0 Fair/satisfactory 56 56 37 70 42 Poor/Critical 26 17 17 25 41 Bad/Failed 10 6 17 4 16 100 100 100 101 99 With the completion of some large projects on the main roads, and a substantial number of resealing and regravelling contracts, present conditions may be broadly similar to those reported for paved and gravel roads. Earth roads deteriorated with El Nino, and the limited donor activity since the ending of the Minor Roads Programme. However, the specific flmding for minor roads under the El Nino Emergency Project (ENEP) has remedied some of the most serious damage and KRB funding to the district roads is leading to improvements. In preparation for this Review, the World Bank undertook a very rapid service and condition survey on the most heavily trafficked paved road network (Annex 1). The results, based on a sample of about 3,350 kms, are shown in Table 3. Table 3 Condition of the Main Paved Network (%) Ught Vehicle Service Visual Speed (kph) Level Condition Good > 85 37.7 33.1 Fair 75 - 85 25.4 31.0 Poor 55 - 75 28.3 16.8 Bad/very bad < 55 8.6 19.1 Total 100.0 100.0 If the limited maintenance resources have been allocated rationally, then priority would have been accorded to this network, and conditions on the less trafficked paved network may be poorer. The present conditions on the main network are broadly comparable with those reported in the mid 1990s. The substantial difference in the proportion of roads in bad/very bad condition between the service level and visual condition reflects the major efforts of the Ministry of Roads and Public Works (MORPW) to maintain the major corridors through extensive emergency patching. (b) Unclassified Network The inventory produced by the 1994 Local Authorities Road Inventory Project, also contains road condition data. The urban condition estimates may have been based on actual inspection and it is possible that, with the funding from KUTIP and the Fuel Levy, there has been some improvement to the paved networks in the main urban areas. The numbers in Table 4, for the other unclassified roads, should be largely disregarded as neither the network nor its conditions were identified by comprehensive fieldwork. Very little of the unclassified rural road network receives, or has ever received meaningful maintenance, so it must be a reasonable assumption that, for those roads that exist, their condition is likely to be, at best, poor. 4 Table 4 Condition of the Unclassified Road Network (0) Paved- Unpaved Good Fair Poor Good Fair Poor Nairobi 15.3 34.6 50.0 70.6 2.4 27.0 Mombasa 8.1 51.2 40.7 0.0 9.1 90.9 Eldoret and Kisumu 29.5 20.5 50.0 1.0 2.5 96.6 Other KUTIP 20.3 33.8 45.9 0.3 13.0 86.8 Other Municipal 12.1 67.7 21.2 5.7 48.0 46.3 Towns 0.0 94.3 2.9 4.7 38.2 57.2 All Urban Roads 15.8 37.5 46.8 5.4 35.6 59.0 County Councils No paved network 4.7 41.8 53.5 -1.4 Traffic on the Network There used to be a comprehensive traffic census for the classified road network but its coverage and processing has declined in recent years. However, the broad structure of traffic flows has probably not changed substantially, except perhaps around the major urban areas where there has been a noticeable increase in congestion. Overall fuel consumption has not increased significantly, and traffic on the major corridor has not increased since the mid-1990s. Table 5 may, therefore, provide a fairly realistic representation of flows on the classified road network. Table 5: Traffic Flows on the Classified Road Network (kms) >1000 500-1000 300-500 200-300 100-200 50-100 <50 Daily traffic: 3000 750 400 250 150 75 25 Paved 2790 1310 950 660 1085 660 1345 Gravel 5 205 490 715 2485 3445 20067 Earth 75 190 45 830 1410 24895 Total 2795 1590 1630 1420 4400 5515 46307 Veh-kms(mn) 3061 435 238 130 241 151 423 % Veh-kms 65.4% 9.3% 5.1% 2.8% 5.2% 3.2% 9.1% Cum.% 65.4% 74.7% 80.0% 82.6% 87.7% 90.9% 100.0% The World Bank survey in 2001 found similar flows on the paved network and the recent Japanese study on Road Maintenance obtained broadly consistent results using 1997 traffic census data, Table 6. Table 6: Recent Estimates of Traffic Flows World Bank Paved Road Study JIKA Study Vehicles/day % of Roads* Road Class krns ADT >2000 11.5 A 3755 1721 1500-2000 7.1 B 2799 1205 1000-1500 14.8 C 7668 290 500- 1000 49.9 D 11216 68 0-500 16.7 E 37850 33 *Excluding Northern Conidor When these results are combined with the urban and unclassified rural roads in Kenya, the structure of the network might be broadly categorized as: 5 Sector Daily Traffic Kms Network Need Main urban network Very high 3,000 High capacity and service Main road network >400 6,000 High service quality Secondary road network >75 9,000 All weather Rural road network <75 100,000+ Passability There are a few anomalies in this categorization; for example, the roads to Moyale and Taveta carry relatively little traffic but would be recognized as major strategic links. The vehicle fleet in Kenya has rapidly expanded since Independence but, according to the official statistics, its size has stagnated since about 1997, reflecting the performance of the economy7. Table 7: Kenya Vehicle Fleet ('000 vehicles) 1970 1980 1990 1995 1997 1998 1999 2000 Cars 58.5 113.6 157.7 172.8 211.9 225.1 229.0 229.6 Pick-ups 37.4 55.5 88.3 100.9 121.7 148.8 147.9 146.8 Trucks 13.7 23.6 13.2 32.6 39.5 54.2 53.8 52.9 Buses/imini-buses 2.5 5.1 13.2 29.8 34.9 46.4 45.6 44.8 Total 112.1 197.8 272.4 336.1 408.0 474.5 476.3 474.1 The composition of traffic on the network varies very considerably between the Northern Corridor, which links Mombasa through Nairobi to Uganda, and other roads, Table 8. Table 8 Traffic Distribution: Classified Road Network Car LGV Matatu Bus MGV HGV Total Northern Corridor Mornbasa-Nairobi 18% 11% 15% 3% 11% 41% 100% NakuTu-MauSummit 25% 19% 15% 11% 14% 13% 100%V MauSummit-Timboroa 26% 18% 15% 8% 18% 15% 100% Mau Sunmmit - Kericho 29% 19% 200/o 100% 11% 11% 100% Kericho - Kisumu 26% 20% 22% 12% 10% 10% 100% A Roads 29% 23% 24% 4% 9% 11% 100% B Roads 300/o 28% 290% 2% 90/0 2% 100% C, D & E Roads, 24% 39% 25% 1% 10% 1% 100% The % of heavy vehicles is somewhat overestimated as A roads include the Northern Corridor The Northern Corridor, particularly between Mombasa and Nairobi carries an exceptionally heavy percentage of heavy vehicles. This difference in the composition of the traffic flow has important implications for both the costs of road rehabilitation and the potential for private sector involvement in road financing. Heavy vehicles are the major contributor to road damage and reconstruction costs, but are usually prepared to pay substantial road tolls. 7The statistics are based on vehicle licenses and there may be anomalies 6 1.5 Responsibilities for the Network In principle, the governance of the sector was fundamentally changed by the Kenya Roads Board Act, 2000, but this has not yet been fully reflected in implementation arrangements: Kenya Roads Board (KRB): the KRB has overall supervision of the sector. It is supposed to: (a) distribute funds from the Fuel Levy to road agencies for the implementation of approved work programs; and (b) audit both the financial and technical compliance with these work programs Road Implementing Agencies: at present, only three implementing agencies are recognized - the Roads Department, the Kenya Wildlife Services (KWS) and the District Road Committees (DRC). Subsequent to the Act, it was established that the DRCs could not function as implementing agencies and their role has been restricted to the development and monitoring of district work plans which are then implemented by the District Road Engineer (DRE) of the Roads Department. Roads Department: responsible for the A, B and C roads. Clear responsibilities and clear funding arrangements. Kenya Wildlife Services: responsible for unclassified roads within the national parks, but with no funding from the Fuel Levy. Clear responsibilities and no funding. District Road Committees: responsible for other roads (D, E and Special Purpose, as well as unclassified roads). But in practice, DRCs do not see that they have any responsibility for either paved roads (even when classified as D or E) or roads within the urban areas. Local Authorities (LAs): the KRB Act gives no role to the LAs in the road sector but they do have such responsibilities under other legislation. There is no funding for LAs under the KRB Act (although some flnding is being provided for committed contracts) but LAs continue to undertake some road activities with funds derived from other sources, particularly LATF. Cess Committees: in some districts (especially those in tea growing areas) the crop cess (an agricultural output tax) generates substantial revenue and much of the funding is supposed to be used for road maintenance. Cess committees organize the work program, concentrating on roads important to farmers, using both the public and private sectors to undertake the work. The work programs are not coordinated with the programs of the other agencies. Planning, financing and institutional arrangements in the sector need to recognize the very different functional characteristics of the networks within the overall sector. Responsibilities for the sub-sectors should be devolved to where ownership is felt. The rural road network has immense local importance and is felt to be our roads, while main roads are not owned locally and should be the responsibility of them. Only unpaved roads are our roads, paved roads are always owned by them. The KRB Act has moved a long way toward a more realistic alignment of ownership and responsibilities in the non-urban sector, though there is the need for a road reclassification to ensure that roads are really designated to the right categories. There is certainly a need to establish the actual road network which should be maintained with funding from the fuel levy. The major shortcoming of the present legislation is with respect to the urban areas, in terms of both the level of maintenance funding for the main urban roads (arterial roads and central business district streets) and ownership. Just as rural DRCs do not 'own' the main roads that go through their districts, so urban DRCs do not own the arterial roads and central streets. DRC ownership relates only to those streets of most immediate and direct connection to their constituents. The other roads are the responsibility of them. Unfortunately, the them is neither defined nor funded under the KRB Act. 7 1.6 Financing of ihe Network Kenya has made quite remarkable progress, during the 1990s, in generating domestic resources for the maintenance of the road network. The Fuel Levy has transformned the financing of the sector, eliminating dependence on the size of, and uncertain releases from the Recurrent Budget. Even though the Government has been experiencing major financial problems, the Fuel Levy has been subject to neither diversion nor borrowing. Based on experience elsewhere, this is a quite massive achievement. While the overall level of maintenance funding may still be insufficient (Section 7), there has undoubtedly been a tremendous increase in recurrent funding. But, this has resulted in little overall increase in funding for the classified road sector. The increase in recurrent funding has been largely offset by a major reduction in development funding. Table 9 provides an indication of overall funding, adjusted to US$ at 2001 prices. Table 9: Funding of the Classified Road Sector (US$ million) (2001 prices) Recurrent Development Total Funding Budget Toll Total FY89 37.3 9.9 47.2 152.9 200.2 FY90 28.8 13.5 42.2 131.8 174.0 FY91 26.5 14.8 41.3 115.5 156.8 FY92 24.6 14.3 38.9 104.1 143.0 FY93 19.3 9.8 29.1 68.3 97.4 FY94 34.2 0.0 34.2 29.5 63.6 FY95 60.2 0.0 60.2 65.5 125.7 FY96 81.0 0.0 81.0 100.3 181.3 FY97 96.5 0.0 96.5 58.1 154.6 FY98 85.8 0.0 85.8 51.3 137.1 FY99 89.2 0.0 89.2 46.7 135.9 FY00 87.7 0.0 87.7 37.3 125.0 FYOI 89.9 0.0 89.9 47.7 137.6 FY89-95 33 9 42 95 137 Ksh. Bn 2.6 0.7 3.3 7.4 10.7 FY96-01 88 0 88 57 145 Ksh. Bn 6.9 0 6.9 4.4 11.3 Recurrent funding has more than doubled in real terms (FY2000 and FY2001 may be rather underestimated), but development funding has fallen substantially and the net increase in overall funding has been marginal. The flow of development funds has been seriously affected by the reductions in overall flows of aid funds to Kenya during the 1990s but, in addition, Kenya has not always taken advantage of the donor funding potentially on offer - windows of opportunity have been lost through slow project preparation and processing. 1.7 User Costs and User Charges in the Road Sector Dedicated funds and taxes are usually anathema to Finance Ministers as well as to macro-economists in the IMF and the World Bank - they limit fiscal flexibility, restrict government action and are often notorious for corruption. However, Road Funds, when established and managed under KRB type arrangements, have become increasingly accepted as an important component of the process to introduce 8 a more commercial approach to road sector management. Under this reform agenda, fuel levies are not considered as a dedicated tax but a specific price charged for the provision of road services. Dedicated taxes are bad but user charges are good! Fundamental to this concept is the role of the road user in deciding how these funds are utilized. The Kenya Roads Board has a majority of private sector members who represent either directly or indirectly the important road user groups. For the efficient allocation of resources and sustainability of the road pricing system, the charging regime should meet two criteria: (i) User charges should be related to the costs that road users impose on the road network; and, (ii) Road users should perceive that they are benefiting from the price paid. The framework of road user charges in Kenya is similar to very many countries, a vehicle license and a levy on fuel. An estimate of expected total user charges was prepared on the basis of the reported vehicle fleet, average vehicle utilization and nornal rates of fuel consumption, Table 10. Table 10: Kenya Expected Road User Charges: FY2001 Licence Fees Fuel Levys Total User Charges Vehicle Licence Revenue Fuel Use Levy/veh Revenue Fleet 000 US$/veh US$m LAlveh/year US$ . US$m US$m % Car 230 30 6.90 1300 100 23.0 29.9 22 Pick-up 148 35 5.18 3250 250 37.0 42.2 31 Matatu 40 35 1.40 4150 325 13.0 14.4 11 Bus 5 145 0.73 9000 700 3.5 4.2 3 Truck 54 180 9.72 7750 600 32.4 42.1 31 Trailer 15 205 3.08 3.1 2 Total 27.00 108.9 135.9 100 Ksh. Bn 2.1 8.5 10.6 The distribution of user charges and the subsequent use of the revenue raises important issues: > Licence Revenue. Gap: The estimate of expected revenue for the fuel levy is very close to the actual reported revenue collection. But, actual revenue from license fees are only about US$10 million, less than 40% of the expected revenue. While license revenue does not currently form part of the Road Fund, it would be the logical addition.. License fee revenue would be a small benefit at present collection rates, but would make a significant contribution, if full collection was achieved. > Inefficient User Charge Distribution: About 64 percent of the road user charge revenue is raised from light vehicles - cars, pick-ups and matatus. Only 36% of the total revenue comes from heavy commercial vehicles - large buses and trucks. However, road damage and subsequent maintenance and rehabilitation, costs are primarily caused by these heavy vehicles. Heavy commercial vehicles account for at least two-thirds and often three-quarters of maintenance costs. Commercial vehicles in Kenya are seriously undercharged for their road use. The Transit Charge for a heavy commercial vehicle is US$0. 10/kmn, the user charge for heavy commercial vehicles in Kenya is just over US$0.05/km, about 50% of the transit charge. Increasing user charges on heavy commercial vehicles would be economically justified, as well as generating additional revenue. 8 The present fuel levy is Ksh.5.8/liter for both gasoline and diesel, approximately US 5.5 cents/liter. 9 > Inequitable Revenue Distribution: Cars, pick-ups and matatus are used predomninantly for short distance urban trips. One recent study estimated that about 53 percent of the total fuel levy was generated in the urban areas by light vehicles, this report estimates a slightly lower level at 47 percent9. In broad terms, therefore, about 50 percent of the fuel levy is generated in the urban areas but no funding is, at present, specifically allocated for the maintenance of the urban roads on which the revenue is generated. There is no reason why revenue and expenditure in the road sector should be exactly equated but it is both inefficient (because of the volume of traffic and high vehicle operating costs) and inequitable (given the proportion of funding generated) if the urban roads do not receive the maintenance funding needed to keep them in good condition. 1.8 The Road Program Road Maintenance in the 1990s (a) The Main Road Network: The progressive increases in the level of Fuel Levy funding have resulted in a major increase in the number of road contracts. Unfortunately, several reports, as well as numerous complaints in the Press, suggest that road-users have not received full value for money from this funding: indifferent construction and ineffective prioritization. Less important links were rehabilitated, reconstructed or improved, while much of the main road network was neglected (with the exception of a few major contracts). Some contracts seem to involve works beyond the scope of the Levy Act which restricts funding to the repair and maintenance of roads. The large number of gravelling contracts surprised many, as maintenance strategy for unpaved roads had supposedly shifted to partial rehabilitation and spot gravelling (though this may not always be feasible on heavily trafficked gravel roads). Moreover, the funds were hugely over-committed, resulting in a very large pending bill liability as well as large outstanding claims. The Ministry of Finance has attempted to resolve the issue by converting pending bills to Treasury Bonds but about Ksh 5 billion of pending bills remain. Unless taken over by the Government, this liability will seriously restrict the level of activity financed by -the KRB and could reduce its credibility. The Fuel Levy provided the opportunity but no priority was given to increasing regular routine maintenance on the network. By the early 1990s, it was generally agreed that the established force account maintenance system had collapsed but, while new initiatives were discussed, no new systematic approach to the delivery of regular routine maintenance was introduced on the network. The 1990s was a lost decade for routine maintenance in Kenya. (b) The Minor Road Network: The Roads 2000 strategy evolved in the early 1990s from the Minor Roads Programme, to address the needs of the lower level roads for which access rather than condition was the primary objective. The strategy was endorsed by all stakeholders but, with the notable exceptions of DANIDA and SIDA, there has been little implementation. Although donor funding was limited, substantial local funding was available; but when local funding was directed at the unpaved network, it was through full rehabilitation contracts. If the strategy is now to be introduced, it will have to be re-launched. 9 Fuel consumption on the classified network was estimated from Table 7 and subtracted from total fuel consumption, leaving urban fuel consumption as the residual. 10 (c) The Urban Road Network After a long period of inadequate funding, substantial support for the urban road sector was provided during the 1990s through the Kenya Urban Transport Improvement Project (KUTIP) and the Fuel Levy. The project and associated GOK funding was expected to rehabilitate 400 krms, provide periodic maintenance on 400 kms, improve 75 kms and maintain about 2000 kms in Nairobi, Mombasa, Eldoret, Kisumu and a further 22 municipalities. The planned capital program was subsequently revised down to 344 km in order to release funds for emergency rehabilitation works, following El Nino. However, some fourteen major road contracts (totaling 208 kIn) have been completed. These major works were implemented by contractors employed by the Ministry of Local Government, rather than the Local Authorities. By 2000, regular funding for the routine maintenance of the KUTIP networks was being transferred to the LAs and systems put in place for its management and monitoring. Counterpart funding for the capital works and the routine maintenance funding was derived from the Fuel Levy. Road Maintenance in the 2000s: Road maintenance in Kenya is now very largely financed by the Fuel Levy, with the funds being allocated by the KRB. The overall Fuel Levy budget is now over Ksh 8 billion. Its distribution between the sectors is detailed in Table 11. 11 Table 11 Kenya Fuel Levy Budget Percentage (%/6) of Total Budget FY2000 FYZOOI FY2002 FY2003 A: TRUNK AND PRIMARY ROADS (A, B and C) Administrative Support and Management Services 6.6% 4.2% 3.2% 3.2% Contracted Works Outstanding Payments 28.3% 0.0% 0.0% 0.0% Paved Roads Rehabilitation 9.0% 22.2% 12.9% 12.9% Resealing/Recarpeting 15.00/% 13.9% 5.8% 5.8% Unpaved Roads Regravelling 10.2% 12.8% 8.8% 8.8% Other 3.4% 0.8% 2.9% 2.9% 65.9% 49.7% 30.4% 30.4% Direct Labour Maintenance.Paved 7.0% 9L3% 8.4% 8.4% Grading/Routine Maintenance (Unpaved Roads) 3.7% 8.3% 7.90% 7.9% Other Activities 16.8% 8.6% 8.6% 8.6% Sub total 27.5% 26.2% 24.9% 24.9% B: OTHER ROADS (D, E and Other) Administrative Support Services 0.0% 1.0% 1.7% 1.7% Contracted Works Paved 0.0% 3.2% 3.0% 3.0% Unpaved 0.0% 6.6% 6.4% 6.4% Other 0.0% 0.3% 0.8% 0.8% 0.0% 10.1% 10.2% 10.2% Direct Labor Emergency Intervention Programme 0.0%/0 3.4% 2.6% 1.3% Equipment Maintenance and Rehabilitation 0.0% 0.2% 0.3% 0.3% Constituency Funding 0.00/0 5.3% 14.7% 16.0% 0.0% 8.9% 17.6% 17.6% C: URBAN PROJECTS UNDER MOLG 0.0% 0.0% 11.8% 11.8% D: KENYA ROADS BOARD 0.0% 0.3% 1.3% 1.3% TOTAL 100.0% 100.0% 100.0% 100.0% The substantial change in the distribution of funding, between FY2000 and FY2001, reflects the introduction of the KRB and the specified allocation of funding. It is unlikely that the outstanding payments (Pending Bills) terminated in FY2000, rather they are now included within the other elements of the contracted works. A significant element of the revenue is financing on-going urban contracts. FY2002 and 2003 are expected to show very little change, other than a slight increase in constituency funding to the 16% mandated in the KRB Act. It is not yet clear whether the funds are used as allocated or whether the implernenting agencies are re- allocating funds to meet changing priorities. This will only be established by auditing. KRB will have to establish procedures for accommodating changing circumstances - activities should follow the approved work plan, but a means of providing flexibility, such as a mid-year review, will be needed. A realistic balance, maintaining financial discipline while allowing operational flexibility, is required. 12 (a) The Main Road Network (A, B and C) Fuel Levy Contracts: a large number of resealing and re-gravelling contracts are on-going, widely spread over the country (contracts in more than 38 districts). Given the high costs of the contracts (z US$100,000/km), the works involved are not simply periodic maintenance but include a substantial element of rehabilitation - maintenance resealing should cost 11 tonnes. But, the level of overloading on these axles is not always trivial: axle-loads of 16, 17 and 18 tonnes are still to be found on the road (though the numbers are small in percentage terms, some 1% of axles >13 tonnes). On the basis of the allowable axle-loads, the damaging effect (ESA) of a fully, but legally loaded vehicle can be calculated, and damaging effect of actual traffic can be compared to this level, Table 14. Table 14 Vehicle Road Damage Northern Corridor: Vehicles from Mombasa (% of vehicles) 3 axle 4 axle 5 axle 6 axle 7 axle Survey I >legal level 66 54 73 82 80 >2x legal level 18 27 27 38 42 >3x legal level 9 15 9 13 12 Survey 2 >legal level 65 54 61 70 87 >2x legal level 17 20 23 32 46 >3x legal level 9 11 8 13 16 The enforcement of axle-load regulations has had an impact. Heavy vehicle speeds have perceptibly increased, semi-trailers with a single 20 foot container are seen, and there is evidence of some reduction in average equivalence factors for commercial vehicles compared with previous surveys in the early 1990s. But the impact is relatively modest. The KRB is in the process of awarding a contract to a consultant to continue the axle-load monitoring program which was financed, for a time, by the EU. This is imnportant but needs to be accompanied by a review of the legislation and the operating instructions provided to the weighbridge officials. There has been discussion, for a long time, regarding contracting out the operation of the weighbridge stations and axle-load control to the private sector. The legal ramifications of such a move need to be clarified - the legal powers designated to the operator, the legal consequences to the operator, if overloaded vehicles are found in spot-checks by the Ministry, etc. Combining such powers with contracting out the maintenance and management of particular roads or corridors to the private sector might be a more efficient solution. The private sector would have a direct interest in ensuring that overloaded vehicles are controlled. Certainly, if Kenya moved to some form of concessioning, some measure of devolved axle-load control would be desirable. In South Africa, road concessionaires have taken the government to court for not enforcing axle-load control. Road Safety in Kenya Like many other developing countries with low rates of vehicle ownership, road safety, vehicle accident rates and the level of deaths and injuries are horrifically high in relation to the size of the vehicle fleet. On average, between 2,800 - 3,000 are killed each year and a further 8,000 - 10,000 seriously injured. The situation is bad and gradually deteriorating. The causes of accidents are quite simple: human behavior, vehicle condition and road condition, as well as the interaction between these factors. In Kenya, the high accident rate reflects the very poor driving standards of many, the financial incentives to speed by matatu and bus drivers, the poor condition of many vehicles, and poor road conditions (particularly potholes on paved roads). 17 Many of the basic elements exist, in principle, to reduce the accident rate and the unnecessary loss of life: * Unlike in many developing countries, the police maintain an accident data recording system which, though simple, is to a large extent reliable. * Kenya has a developed legal framework and updated traffic legislation. * Kenya has a vehicle licensing system and an established vehicle insurance industry. * Kenya has Traffic Police and potentially an extensive enforcement capability. * There are good engineers, a road safety unit in MORPW and manuals for road safety engineering. * Groups and individuals who are seriously concerned about the situation and who are already trying to take initiatives to tackle the problem. The problem in assembling these elements into a well articulated and targeted road safety program is the lack of political will and the low priority given to road safety by the government. Outrage is expressed after a major bus accident, but this soon fades and no real action is taken. A National Road Safety Council was established in the 1980s but has been moribund for years. A Kenya Road Safety Agency Bill has not been tabled in Parliament, although it has been prepared for the last four years. The number of road deaths pale into insignificance compared with the AIDS pandemic, but there was no priority given to road safety even before the scale of the pandemic became known. The number of deaths may be small in comparison to AIDS, but accident victims remain a major cost to the health service and take up many of the available hospital beds. With this low political priority, other constraints emerge: * No overall coherent planning process - policy in one ministry, implementation of policy in another. * Little or no funding for road safety measures. * No development of technical expertise which results in improper implementation of specific accident countermeasures. * MORPW is hampered in integrating road safety in design by funding constraints, but perhaps more importantly by the apparent lack of appreciation of the importance of such basic measures as road marking. * No attempt to ensure that enforcement enforces the regulations in an effective manner. * No processing of accident data to identify where either enforcement should be concentrated or safety engineering considered. * Lack of coordination hampers the efforts of the stakeholders, opportunities are lost and scarce resources are wasted. The police report much action, over 300,000 court cases for a vehicle fleet of 400,000, vehicle inspection centers report inspecting 120,000 vehicles, but with no apparent effect. Kenya has a very different road fatality distribution to developed countries, where public transport is very safe and vehicle drivers account for a substantial proportion of road deaths. In Kenya, 40% of the fatalities are vehicle passengers and 40% are pedestrians. Given the frequency of matatu and bus accidents reported in the Press, often caused by a combination of speed, overloading and burst tires, the high fatality rate is not surprising. The observer might wonder what is being achieved by all the police roadside checks on matatus, if they are not able to control overloading and tire condition. A task force has been established by a group of key stakeholders,. outside the official structures, and is launching campaigns and raising funds from private sector companies. In each of the last three years, 18 campaigns have been undertaken during December and significant reductions in accidents have occurred. But the campaigns are not sustained. '19 2. STRENGTHS AND WEAKNESSES IN THE ROAD SECTOR Before identifying the critical issues facing the sector and the possible solutions, it is useful to identify the main strengths and weaknesses in the road sector. Any strategy for the sector must address the weaknesses and build on the existing strengths as these form the basis for progress. Extemal assistance can be provided but, for sustainability, the solution to the sector's problems must come from within Kenya. 2.1 Sector Strengths (1) The Kenya Roads Board: the creation of the KRB is the major innovation in the road sector and is potentially the Sector's greatest strength. If the KRB succeeds in (a) developing a coherent strategy for the sector which addresses the primary road transport needs; (b) ensuring a clear division of responsibilities which cover the network; (c) requiring realistic and prioritized work programs, based on the strategy for the sector; (d) ensuring the work programs are implemented; (e) closely monitoring financial and technical compliance through external auditing; (f) ensuring transparency through the publication of all work programs and audit reports; and (g) obtaining value-for-money then there is a way forward for the Sector which can be endorsed by all stakeholders and very actively supported. If the KRB fails to achieve these objectives, then the future for the sector and stakeholder support becomes very unclear and rather bleak. (2) Assured Road Maintenance Funding: the level and regular flow of road maintenance funding in Kenya is a major achievement. While the present level of funding may not be sufficient to meet all maintenance needs (Sector 7), the equivalent of US$110 million should provide the basis for meeting most of the maintenance requirements of the network, if targeted appropriately, especially if there is additional substantial funding for the rehabilitation of the network. One major advantage of incorporating MPs into the system, through the DRCs, is that it has created a powerful constituency to ensure that the level and flow of Fuel Levy Funds is not disturbed by financial problems elsewhere in the public sector. (3) Decentralization of Funds, Decision Making and Implementation: the role and composition of DRCs may not be ideal, but they have introduced an important level of local decision making, monitoring and accountability for local roads which was clearly lacking. Like the KRB, the DRCs have the potential to play a key role in the revitalization of the sector at the local level. However, unless DRCs' responsibilities and perceived ownership can be better aligned, critical constraints and discontinuities will become increasing evident and prevent the introduction and implementation of a coherent and comprehensive strategy for the sector. (4) Rural Roads Strategy and Labor Based Technology: the Roads 2000 approach for the rehabilitation and maintenance of the rural road network is an eminently rational response to the needs of the rural sector within Kenya's financial constraints and resource endowment. It adopts a network approach, at relatively low cost, to providing the roads required for agricultural development and access for the rural population to social and economic infrastructure as well as income generating opportunities. Moreover, the approach has the potential directly to generate substantial levels of paid employment in areas where income earning possibilities are extremely limited. The Poverty Report in the mid-1990s, identified road construction and maintenance works as a major instrument for relieving poverty. Roads 2000 should logically form a core element of any road strategy as well as any Poverty Reduction Strategy Program which attempts to increase unskilled and semi-skilled employment. The lack of emphasis on the 20 implementation of the strategy should be considered as one of the main missed opportunities of the 1 990s. Labor-based road works need not, however, be confined to the rural environment. The hand pitching of stones as the base course on urban roads has proved extremely successful in Nairobi, and was successfully employed in recent KUTIP works, providing extensive low wage employment. (5) Well Qualified Engineers in both Publc and Private Sectors: Kenya has a stock of engineers who are as well qualified and experienced as those to be found anywhere. There is little doubt in anyone's mind that Kenyan engineers know what should be done and what they should be doing. The major problem is not lack of training nor knowledge, but the incentives to use that training and knowledge and to do what should be done. Further training may help introduce new ideas regarding road management, contract formulation and control but the fundamental need is not training but re-oriented systems, incentives and sanctions. (6) Substantial Construction and Consulting Sectors: Unlike its neighbors, Kenya has significant local construction and consulting sectors able to undertake substantial works on the main road network. Performance by these sectors can be extremely good and equivalent to any international contractor or consultant. Similarly, at the local level, there is no shortage, in most districts, of small contractors willing to undertake road works, nor does there appear to be any difficulty in sourcing equipment from the private sector when required. Kenya has a relatively large and dynamic private sector which has demonstrated its ability to respond positively to new opportunities. These elements could form the basis for much greater involvement of the private sector in the management and maintenance of the sector, at both the national and local levels. (7) Axle-Load Control: Kenya has demonstrated that axle-load control can be effective, and an immediate reduction in overloading was achieved. Empowering the user to participate to ensure non- discriminatory control was a major step and for a time proved very effective. Enforcement is by no means perfect, and monitoring by the private sector has faded but, even now, most of the gross overloading has been eliminated. 2.2 Sector Weaknesses (1) No Strategy for the Sector: the 1997 Strategic Plan was never owned by the sector, it was produced by MORPW mainly to satisfy external demands. In some important respects it has been overtaken by events, and is now of little relevance. But, without a clear strategy for the sector, it is difficult to determine the objectives for the sector and whether are being achieved. It also makes it very hard for stakeholders to "buy in" and provide support, whether that be road users or development partners. Without a defined strategy, there are no clear priorities for expenditure. The Strategic Plan detailed expenditure priorities for the classified road network, determined on the basis of benchmark analysis, i.e. traffic, road condition and likely costs. The priorities looked sensible but were not applied. It seems almost impossible for the KRB to evaluate work programs without the underpinning of a strategy and expenditure priorities. Even at the level of the DRCs, there may be substantial advantages, if some broad expenditure priorities are established to guide the development and evaluation of the work programs. (2) Inadequate Capital Funding: The Fuel Levy provides a reasonable flow of funds but, quite visibly to any road user, major capital funding is also required to rehabilitate or reconstruct large parts of a very deteriorated road network. The volume of major capital works, as shown in Table 9, declined during the 1990s; partly because available funding declined and partly because Kenya was not able to utilize the available funding. Without capital funding, the Fuel Levy has become practically the sole source for all road activities, both capital and recurrent. The consequences of this lack of capital funding are: 21 (a) Funds have to be diverted from recurrent to capital works, thus providing inadequate funding for maintenance with consequent additional capital requirements in the future. There is no indication that any comprehensive routine road maintenance is either being undertaken or planned. (b) Funds diverted for capital works are inadequate for full remedial works and the interventions are insufficient to remedy the underlying structural defects. Often the objective is to hold the road together until more substantial funding is available. While often rational, in the circumstances, the benefits of the interventions are relatively short-lived and achieved at high cost. The alternative to diverting maintenance funds to rehabilitation/reconstruction would more extremely bad sections of road, but improved conditions on the rest of the network, and a reduction in future reconstruction costs. There are clear trade-offs between rehabilitation and maintenance; it is arguable that too much emphasis is now placed on major works at the expense of regular maintenance. (3) Institutional Malaise: Government has a substantial cadre of very capable engineers who have demonstrated that they can perform. It is, therefore, depressing to observe the deep malaise that has enveloped the key agencies in the road sector. Depressing but not surprising given the low salaries, the very frequent changes of top management, the lack of any clear strategy or mission statements, the frequent shifts of responsibility among the engineers, the low expectations regarding perfornance, the lack of any credible incentives or sanctions, the political influence on decisions and priorities, the suspicions of major and minor corruption, the years of inadequate funding, the tolerance of poor workmanship, the erosion of professional pride and professionalism, etc., etc. Not that this situation in the road agencies is unique in Kenya, nor that Kenya is unique in the region, nor that the region is unique in the world. This malaise is pervasive in much of the public sector in much of the world. To the outside observer, there is a very strong impression that, for many of the engineers within the public service, work within the road agency is a secondary consideration to the other activities required to maintain a reasonable life style for themselves and their families. This problem is perhaps not only the most serious weakness in the sector, but also the most difficult to remedy without major disruption. Civil service reform was started but has not moved beyond the retrenchment of some support staff. Additional funding can be provided, but unless the weaknesses within the institutions are removed or the institutions replaced, it is very difficult to believe that the funding will be effectively used. (4) Absence of Management and Planning Systems: it is evident that all managern,ent information and planning systems have been eroded, in some of the agencies, to the point of extinction. In the past, systems have been established but they are not functioning. Possibly they were misconceived and/or poorly designed, perhaps they were too complex but perhaps the real underlying reason is that no-one appreciated the need for such systems. If senior management required reports, then there would be systems, though they might be simple. But such systems are required for systematic management and there really is little systematic management in the sector. Years of inadequate funding and erratic releases of funds essentially rendered forward maintenance management pointless. Political intervention in the selection of capital projects rendered investment planning meaningless. Delays or reversals in donor funding, usually the consequence of issues outside the sector, have compounded the problem. In such circumstances, it may be rational simply to concentrate on the most immediate concerns: deciding what to do with funds when they become available. A reactive firefighting approach to management of the sector has developed, which might be described as "management by surprise". The KRB is required, under the Act, to monitor the use of the funding. This will necessarily require a reliable and efficient management and information system. But such a system must encompass not only the KRB but also the implementing agencies. Certainly within the Roads Department there is no reliable management information system, nor has one functioned there for many years. 22 (5) Lack of Financial Control: The introduction of the fuel levy provided an opportunity for the award of locally financed road contracts. And, contracts were awarded and awarded and awarded without any apparent regard for the level of funding available. Those responsible for this huge over-commitment may just have been financially incompetent or not supported by adequate financial systems to indicate the level of commitments or they were simply responding to pressures that they could not resist. Whatever the motives, the effects were the same - delays in payments to contractors, downtime of equipment, suspension of works, interest charges at commercial bank rates, contractor claims and, in total, a grossly inefficient road program, very high costs and a huge pending bill liability. Basically, the sector is suffering a monumental hangover after a party which got out of control'2. The Auditor and Comptroller General's annual reports are not effective in imposing the financial control needed to prevent over-commitment of resources. There should be other controls which limit the tendering or award of contracts to the level of funding available which will allow contracts to be completed within the contract period. Without such controls, there will be no safeguard against a repetition of the over-comriiitment. Such control could be specific to the road sector and KRB, or could be a general control covering the entire government admninistration as the problem is not confined to one sector. (6) Delayed and Inappropriate Auditing: The accounts of the Fuel Levy and of the major implementing agencies are audited by the Auditor and Comptroller. While the audits provide an independent financial opinion on the accounts, they are limited in their usefulness: * The long delay between the end of the review period and the publication of the reports reduces the effectiveness of control, and follow-up action; * The scope of the audit is essentially limited to financial control - whether funds were spent according on the intended purpose, rather than whether the funds were well spent * The limited emphasis on identifying remedial measures. Moreover, no-one seems to pay any real attention to the audit reports, thus they provide little effective control, and no action ever seems to be taken to remedy the identified deficiencies. The KRB's mandate includes both financial and technical auditing and this provides the opportunity to introduce a real measure of performance auditing into the system. (7) Inadequate Funding and Responsibilities for Urban Roads: urban road networks have fallen into a deep institutional pothole. Responsibilities and ownership of networks have been clearly established for much of the network. Responsibility for urban road networks has been set by default - they are not the responsibility of the Roads Department, and they are not seen by the DRCs as their responsibility. They thus remain the responsibility of the urban authorities, but without funding. It is inconceivable that this can continue. (8) No Implementation Strategy for Roads 2000: while the Roads 2000 strategy addresses the critical issues in the rural road sector, there is little evidence of pro-active Kenyan leadership, ownership or coordinated strategy to implement the approach. There is the strong impression of donor-driving: the approach is implemented when donors funding becomes available. Where donor funds are not available, there is no implementation. Roads 2000 has not become the change in direction that many had anticipated. Rather, it has become a fragmented donor-supported project in which the details of implementation (what, how and by whom) varies widely according to donor rather than Kenyan needs. The individual projects result in extremely beneficial impacts in the local communities, but it is doubtful whether they can be sustained unless the approach really becomes the Kenyan strategy for rural roads. 12 The over-commitment of the Fuel Levy occurred well before the creation of the KRB. 23 (9) Dubious Construction and Doubtful Supervision: there is a widespread perception within Kenya that the road problem is not a shortage of funds but a problem of contractors and engineers colluding to cheat on the contracts: * Design standards are reduced once works commence; * Sub-standard materials are approved; * Defects are not remedied; and * Contracts are not finished. Roads fail well before their expected design life, road users are faced with rapidly deteriorating road conditions and the resources are wasted. The poorly performing contractors continue to be awarded contracts and no-one is held to account for inadequate supervision. It is extremely difficult to gauge precisely the extent of poor performance, malpractice and corrupt practice in the sector. Stories abound, cynicism prevails, and reports by the Auditor-General and others, including fact-finding teams from the MORPW, have commented on construction anomalies and particularly the inadequate supervision of works. Certainly, engineers on World Bank missions have been baffled by some of the road rehabilitation and improvement works. Not all road failures can attributed to the contractors and supervising engineers, poor designs have also played a part. Inadequate and ineffective supervision cannot only be attributed to Ministry engineers, there are plenty of reports regarding poor consultants. The real issues are accountability and actually holding people and companies to account. It just does not happen in Kenya. (10) Non-Enforcement of Traffic Regulations: the enforcement of axle-load regulations may be having a modest impact, but the effective enforcement of other traffic regulations is minimal. Kenya may not have the most traffic enforcement agents and the least traffic enforcement in the world, but it will be toward the top of the list. With no effective enforcement of vehicle condition standards, vehicle speeds, or the overloading of matatus and buses, no-one can be surprised at the level of road fatalities and injuries. 24 3. CRITICAL ISSUES: DEVELOPMENT OF THE KENYA ROADS BOARD The KRB is a new institution with major responsibilities and few staff. It must oversee the maintenance and development of a sector which has major problems and a poor reputation, within a difficult environment in which there are many pressures and considerable resistance to change. The Board Members and staff of the KRB face very considerable challenges but, if they succeed, they can make a major impact upon not only the road sector but also the development of Kenya. The poor state of Kenya's roads invariably comes toward the very top of the list of constraints to development - whether the list is developed from the views of overseas investors or ordinary Kenyans, as shown in the PRSP consultations. The critical issue for the KRB, at this stage of its development, is to identify and then address the core critical issues within the sector. Given the multitude of problems facing the sector, there is a danger of attention and resources being diverted down interesting but perhaps peripheral avenues. The KRB has been given overall responsibility for the road network but it is neither a detailed planning nor implementing agency. Strategy and Selectivity should be the motto for the KRB. Without a coherent strategy for the sector, the KRB has little possibility of influencing the direction of expenditure and the achievement of consistent objectives, and little opportunity to monitor realistically the impact on the overall network and to hold the agencies to account. Selectivity because the resources of the Board are limited and it needs to concentrate them where the biggest impact can be made. The KRB has to define its role in relation to the implementing agencies. The Board needs to establish the agenda, ensure, through the approval of the work programs, that the implementing agencies are working to that agenda, and then monitor that the work programs are implemented. As far as possible, the KRB should not become involved in the details of implementation, this should be the responsibility of the implementing agencies. Once the KRB becomes involved in implementation, then accountability becomes blurred. In essence, the KRB and the implementing agencies must agree a separation of powers, functions and responsibilities. 3.1 Sector Strategy It is difficult to imagine that the KRB can work effectively and meet its obligations unless there is a coherent strategy for the sector which guides the programs of the various implementing agencies and provides the basis for flnding. Development of a strategy would necessarily need the close involvement of the various implementing agencies and other sector stakeholders, and should be endorsed at the highest levels. Once prepared and agreed the Strategy should form the basis of action by KRB and the agencies. Strategies may not come in all shapes but they certainly come in all sizes and KRB has to decide how detailed and prescriptive a strategy would. best meet the Kenyan circumstances. A strategy is not a plan, but should be the framework within which plans are developed by the implementing agencies. There may be some core elements which form part of any strategy: * Overall objectives of the strategy: what does it intend to achieve * The means to achieve the objectives: policies toward institutions and implementation * The priorities in terms of actions and activities: broad expenditure priorities would seem an essential component of any action rather than shelf bound strategy * The funding needs and availability: this is an uncertain world, and a strategy should consider the implications for actions and activities of different levels of funding. It would seem prudent for any strategy to have a base case founded on assured local funding. Other funds should be considered as additional to, rather than central to this base strategy. Such an approach provides a 25 monitoring mechanism even in the worst funding scenario, and clearly demonstrates the impact of additional funding. * The indicators for success or failure: how the implementation of the strategy will be monitored The Strategic Plan took several years to prepare and moments to put on the shelf. Neither the KRB nor the road sector can afford to repeat such a process; the strategy must be owned and be the basis for action. The KRB Workshop, May 21 - 23, 2002 identified the need for an overall Transport Sector Strategy, with substantial emphasis directed toward the road sub-sector, as the highest priority. It was agreed that KRB and the implementing agencies would work to a deadline of the end of 2002 for the formulation of an agreed Sector Strategy which would then be submitted to Cabinet for approval. The preparation of the Strategy has to be the responsibility of the KRB, as the primary policy adviser to the Government on the road sector. It was agreed that it would be prepared in a consultative and participatory manner, and that process and content assistance would be provided, if needed, by development partners. This would be achieved through a phased approach requiring (i) the production and dissemination of a "strategy discussion document" by the end of September 2002, to be followed by (ii) formulation of the final agreed strategy by the end of December, 2002. These milestones have yet not been achieved. The KRB undertook to put in train a number of arrangements to facilitate the process of strategy development. These would consist of a high level "Consultative Committee" consisting of the Permanent Secretaries for Transport, Finance, Local Government, and Roads and Public Works, a representative of development partners (EU) and an independent legal adviser, (Dr Albert Mumma). It also proposed that a dedicated "process management team" (PMT) be established to facilitate and monitor all activities, which would be responsible for advising on a new stakeholder analysis, essential if the envisaged participatory activities are to be efficiently undertaken. A number of consultancy inputs are planned as follows: * Support to the PMT; * Technical assistance to the production of the strategy discussion document; * The undertaking of a transport sector and roads sub-sector strategy study. The EU has agreed in principle to provide funding. 3.2 Financial Control Given the past experience of over-commitments, one of the major issues has to be financial control within the implementing agencies, irrespective of the sector strategy. Credibility in the sector requires that such over-commitments do not recur. If financial commitments are kept to the level of the funds available, contractors will be paid, contracts will be completed to time, equipment will not stand idle, and costs in the sector will inevitably fall as interest payments are eliminated and claims are reduced. There are, however, various stages of the process at which financial control could be exercised. Work Programs: Approval of the annual work program would clearly be the first step: implementing agencies should provide not only the proposed works but also the financing plan which includes existing financial commitments, on-going works and proposed new activities. Ideally, there should be rolling work programs with a longer time horizon than a single year. Such programs should be required from the main implementing agencies, but they are probably not necessary from the DRCs. Tendering It is rather difficult to see why further controls should be required, as long as the KRB is notified of, and agrees to major changes in work programs, and the agencies can demonstrate 26 that the changes are consistent with the overall financial plan. If further control is felt to be necessary, the implementing agencies might be required to obtain KRB agreement to go out to tender. The construction companies would then come to recognize that KRB has endorsed that funding will be available for the works, and start to reduce the risk loading in their bids. Contract Award: Rather than imposing control at the tendering stage, the KRB might contemplate introducing control at the stage of contract award, only approving those contracts for which funding is available. This might well lead down a slippery slope to closer involvement in the procurement process. Unless restricted strictly to endorsement of the funding availability, it could fundamentally change the direction of the KRB's activities"3. There is also the danger that this stage is too late, with major pressures to award the contract from the contractor, the implementing agency and those benefiting from the proposed works. 3.3 Procurement There are undoubtedly serious issues throughout the procurement process. They relate to the inadequacy of design, flaws in tendering, problems with evaluation, delays in award, as well as the previous issue of over-commitment of resources. It will be very tempting for the KRB to become 'involved in procurement, as the financing agency. Its involvement in procurement, if any, is a basic decision for the KRB; what it decides will have a very major impact upon its work program and staffing requirements. There is no consensus with regard to this issue. * One group argues that KRB should adopt a role similar to that of the extemal funding agencies when dealing with their projects: establish the procurement procedures to be followed and then monitor the process through the provision of "no objections", ensuring itself that the procurement process is followed and proposed contract works will provide value. To undertake this function effectively, without becoming a major bottleneck in the process, KRB staffing would have to increase substantially, in terms of both numbers and skills. The next step along this involvement would be the provision of no objection to contract payments. * The other group believes that KRB should establish the procurement processes and then hold the implementing agencies accountable for following them, through random auditing of the procurement process. Once KRB starts providing "no objections" it becomes part of the process, *the separation of powers and responsibilities become blurred, and full accountability is lost. For its public credibility, KRB should stand above the procurement process and remain 'squeaky clean'. Given the number of contracts and number of implementing agencies, it is almost inevitable that some contracts will go sour, for one reason or another, and complicity in the procurement of these contract could then taint KRB. Many donors are themselves trying to move away from direct involvement in procurement, by shifting from project to budgetary and sector support. Some donors may desire KRB participation in the procurement process for their financed activities, but it will be up to the KRB to decide whether to take a more general mandate. The KRB should not consider taking a role in procurement lightly; it is a fundamental policy decision entailing substantial risks to the effective implementation of its overall mandate. 3.4 Financial and Technical Auditing Monitoring the use of the funds, through auditing, is a legislative responsibility of the KRB and central to its mission of obtaining value-for-money. The Road Works Inspectorate in MORPW is said to produce Control at the award stage might be on an exceptional basis, if the contract award is substantially in excess of the Engineer's estimate on which the tendering was based. 27 good work and provides a valuable performance auditing function, but the results are internal to MORPW and it is hard to quantify the extent to which the auditing has improved performance. It is vital for the transparency of KRB that the audits, undertaken either by or on behalf of the KRB, are placed in the public domain irrespective of how uncomfortable the audit reports. For the credibility of KRB, the audits should be conducted by independent auditing and engineering consultants who are well respected within Kenya. However, all the Kenyan based engineering consultants are likely to have involvement in the road sector, and it is conceivable that this may impact their objectivity. KRB should, therefore, consider introducing a measure of quality control, using external consultants with no commercial interests in Kenya. KRB will have to determine the scope and frequency of audits, and the means by which they are undertaken. Performance auditing need not necessarily cover every contract and every implementing agency, every year. Random auditing may be as effective, but provision shotild also be made for targeted auditing of problem contracts. Auditing, however, should not be restricted to the implementation of civil works, equally important is the random auditing of the procurement process, to ensure that procurement procedures are being followed. 3.5 Compliance KRB has responsibilities but its sanctions to ensure compliance need to developed and specified. Performance auditing is an important step to achieving improvements in the use of available funding, but what actions can the KRB take in cases of inadequate performance by the implementing agencies, the contractors and/or the supervising engineers? Naming and shaming is an important power in some environments but it may not be sufficient to remedy poor performance. Persistent poor performance by contractors and consultants might be handled by either suspension from eligibility for KRB financed contracts, or a restriction to certain categories of work, until performance improves. Before the introduction of such sanctions, KRB would need to review carefully the legal framework. Sanctions on poorly performing implementing agencies seem more difficult to envisage. To improve the level of activity reporting, KRB might release funds only when previous activities have been reported and the use of funds justified. In cases of persistent poor performance or misuse of funding, suspension of funding to the implementing agencies might be contemplated. Suspension of even the small agencies would be difficult and suspension of the agencies dealing with the main road network or urban areas might well be impossible. In such cases, KRB's only redress may be through Government to change the management of the agencies. It is not an issue which has to be resolved immediately, but it must be eventually faced: as with procurement processes that turn sour, there are almost certain to be contracts or other work activities which are implemented poorly and implementing agencies that fail to perform. 28 3.6 Manual of Procedures The KRB has only recently recruited staff to the Secretariat, and full staffing is only expected by the middle of 2002. Establishing a manual of procedures should take high priority in order that both the staff of the KRB and the road implementing agencies understand what is expected, what documentation should be produced to support proposed work programs, what reports on on-going or completed work activities should be submnitted, etc. Separate manuals may well be required for the financial and performance auditing roles of the KRB. Fortunately, KRB is not the first institution to be faced by the need to develop such manuals for the road sector, and KRB should be able to draw on the manuals and experience of similar institutions, possibly through the auspices of the Sub Sahara Africa Transport Program (SSATP). Rwanda, for example, has just completed the preparation of procedures manuals. 29 4. CRITICAL ISSUES: SECTOR OWNERSHIP AND RESPONSIBILITIES According to Table 1, the road network is very large, approaching 200,000 kIn, although there is a very much smaller network of engineered roads. Much of the network has minimal traffic and serves very small communities and might be better categorized as tracks rather than roads. For the effective maintenance of such an extensive network it is essential that responsibilities are devolved to where ownership is perceived. The KRB Act has changed the responsibilities but these are not always aligned with the perceptions of ownership. Unless modifications are made to bring ownership and responsibilities into alignment, there is a real possibility that crucial segments of the network will fall through the cracks and receive neither funding nor maintenance until there is major deterioration in the network. Previously, responsibilities were clear: * central government was responsible for the classified road network, and * local authorities were responsible for the rest. In reality, inadequate funding resulted in maintenance being undertaken on a limited part of the classified road network, and on an extremely small part of the unclassified network. Under the new arrangements, central government is responsible for the maintenance of a much smaller network of national roads, while DRCs have responsibility for the rest of the network. DRCs, however, do not consider that they have responsibility for either paved or urban roads. These, they believe, remain the responsibility of either central government or the local authorities. Moreover, within the DRCs, priorities are established on a constituency rather than district basis, and, in this process, important intra and inter-district roads may receive little priority. Meanwhile, local authorities still perceive ownership of parts of the network and continue to allocate some of their funds (admittedly limited) for road maintenance and improvement. 4.1 Districts The situation is complex but it is not intractable. The solution may require the underpinning of substantial preparation, a willingness to compromise, and some modification in the allocation of funds. The cost and effort needed to achieve a solution will be modest in comparison with the ultimate benefits: o Determining the size and nature of the road network: the first step toward the re-alignment of ownership and responsibilities has to be a much better understanding of the road network, particularly the network outside the classified road system. It is difficult to imagine that this can take place without a realistic inventory of, at least, the most important parts (as perceived by the local conmnunities) of the unclassified network. The inventory needs to determine not only the number of roads, but also their location, their condition (impassible, dry season, all-season, or all- weather) and function (urban or rural). A full inventory and condition survey of all 100,000+ rural unclassified roads, based upon physical inspection, would be a massive undertaking, and local consultation combined with inspection of the key unclassified roads may be the better approach. o Reclassification of the network: having established the network, it will be necessary to reclassify district roads according to perceived ownership: those road networks which should fall under central governnent, those networks which DRCs perceive to be their responsibility, and those residual urban and rural roads for which neither the central government nor the DRCs take responsibility. A basic classification already exists, and the reclassification would generally take place at the mnargin, perhaps changing the classification of some existing classified roads, adding important unclassified roads, and designating the core urban networks in the districts. The 30 objective of the classification would be to determine the road network which should receive fuel levy funding and the responsibilities within that network. The remaining network of very minor roads and tracks should probably be left to the local communities. There seems no rationale for a monolithic structure which supervises a network ranging from the Nairobi-Mombasa Road to an un-engineered, dry season track which a pick-up may very occasionally use. O Re-align funding mechanism: The 16% allocation of funds going to the constituencies might be considered fixed, similarly the 57% for the main road network may be fixed, but the 24% of funding for the districts provides a significant element of flexibility. This district funding (to be distributed equitably, according to the KRB Act) provides the opportunity to meet the needs of those elements of the network which have either been ignored or are likely to be neglected. The funding might be distributed in two components: o One component for the designated urban networks: funding would be distributed to the DRC but earmarked for the designated urban roads within the district and would require a separate work program: The funds could be allocated on a km basis. In most districts, the network of urban roads (rather than roads within an urban administrative boundary) will be very limited and works could be organized by the same agency which implements the other DRC activities. O One component for general district roads: these funds would be allocated on a district rather than constituency basis, and would mean that DRCs have to agree on a work program, rather than simply aggregating separate constituency plans. This would provide the opportunity to fund intra-district links rather han simply intra-constituency roads. The aim is not technical perfection, but a classification of responsibilities and funding which broadly reflects the priorities and which can command ownership. The magnitude of the task to inventory and then reclassify the Kenyan road network is difficult to estimate as it will require both technical survey and extensive consultation and participation. A pilot exercise in one or two districts might be an immediate 14 way forward to test possible approaches and gauge the resources required for a national exercise 4.2 Major Urban Areas The structure, described above, would appear workable for most districts and the smaller urban authorities, but it is not reasonable for Nairobi nor very probably for the other large urban areas - Mombasa, Eldoret, Kisumu and Nakuru - which have extensive paved urban road networks. These networks are not 'owned' by the DRCs and reasonable maintenance needs (routine and periodic) are substantial, possibly in the order of US$ 10 - 11 million annually. Not only are the networks large and costly to maintain, they also need to be managed both as a network and as part of the overall urban area. Unfortunately, the performance of the major urban authorities has often been more than disappointing and clearly there may be considerable concern regarding the ability of the large urban authorities to plan and implement efficiently and cost-effectively the scale of resources which are technically required. But, shifting the entire responsibility of road works in the major urban areas to the DRCs, with implementation by either the PRE or DRE, is not a plausible alternative. There are extremely strong arguments in favor of the main urban areas being designated as implementing agencies in their own right under the KRB Act. In the longer tern, this must be the solution. To achieve this objective would very probably require a classification of the network to designate the main roads/streets which would be the responsibility of the urban implementing authority and those more local roads that would remain under the DRC responsibility. As with other implementing agencies, the urban 14 SIDA at the KRB Workshop, May 21 - 23, 2002 offered to discuss with KRB the possibility of assistance to undertake such a pilot exercise. 3'1 implementing agency would be required to prepare work plans for approval by the KRB and would be subject to external technical and financial auditing. Given the reputation of urban authorities, the level of monitoring, both technical and financial, would probably need to be intense, at least, in the early stages. However, the KRB secretariat has already approached the Parliamentary Committee on Transport to designate the big five urban areas as implementing agencies and received a very negative response. Resolving this issue should be one of KRB's highest priorities as the politicians will only appreciate the need for maintenance once it is too late and much more extensive and expensive interventions are required. Without in-depth understanding of the political processes and imperatives, it is not possible to propose a solution to the present impasse, but there are some elements which may further an acceptable agreement: * Reclassification and designated funding for the main urban network could be a short-term measure which would provide the level of funding required for the maintenance of the network. If decisions on their use remain with the DRC, roads would not be integrated into overall transport and urban planning, and could lead to sub-optimal decisions; * Well defined, comprehensive monitoring and oversight arrangements will necessarily be part of any solution for these urban areas, with very clear demarcation of funds between the road implementing agency and the rest of the urban authority to ensure that the road funds can be separately monitored and there is no leakage into the general accounts of the local authority. This might be accompanied by limiting the use of fuel levy funds to contracted works, thus removing the potential for employment patronage; * The de-linking, from the Nairobi City Council, of the road implementing agency into a separate semi-autonomous government agency (SAGA) might assist the process and introduce better management and controls; * DRC participation in an oversight transport commission for the urban implementing agencies might provide an acceptable role and involvement; * Successful local government reform would provide an improved environment for responsibility and accountability. It may be the preferred long term objective but is neither under the control of KRB nor is it likely to be achieved quickly. Now that the KRB has the technical underpinning of the Secretariat, it should take determined steps to involve fully the Parliamentary Committee, as well as the urban MPs, in dialogue to resolve the urban funding issue. The KRB should take the initiative with the development of detailed proposals, supported by the necessary cost, inventory and condition data. 32 5. CRMCAL ISSUES: EFFECTIVE ROAD IMPLEMENTING AGENCIES Through the KRB Act and the Fuel Levy Act, Kenya has developed the basis of a very effective funding arrangement for the road sector. The KRB has been given overall responsibility for the road network but, while it should develop the sector strategy, it is neither a detailed planning nor implementing agency. Its role is to review and approve work plans developed by the implementing agencies, in conformity with the strategy priorities and financial position, and then monitor their efficient and cost-effective implementation. The responsibility for detailed planning and subsequent work implementation rests with the implementing agencies. The KRB is only one element of a system, and results will depend upon all the elements of the system. The effective funding arrangement must be complemented by effective implementing agencies. 5.1 The Roads Department The Roads Department (RD) has managed to undertake substantial repair activities in the last few years, and could not doubt continue to keep the main road network functioning on a fire-fighting basis, responding in a reactive manner to one emergency situation after another. It has to be recognized that the main road network has been held together in a much better condition than the networks of many of its neighbors without large scale external financial and technical assistance. The road network is deteriorated but it has been a very gradual decline rather than the collapse which has occurred elsewhere. There are, however, very severe doubts as to whether the RD, under the present institutional arrangements, could plan, effectively manage and implement a major program of rehabilitation and planned routine maintenance on the main road network. Certainly, recent experience has been uniformly disappointing. The basic engineering competence remains but the professionalismr, commitment and enthusiasm seems generally to have been lost. The result is long delays and substandard outputs. The RD has ceased to function in terms of systematic management and planning, and is now working by ad-hoc responses to outside stimuli. This is no sudden change, but rather the steady erosion of the institution over an extended period of time. It is extremely difficult to envisage that the RD can be brought back to provide the level and standard of professional management to plan and implement both a comprehensive program for the main roads and provide implementation support for the DRCs. Rather like the force account maintenance system, the time of the RD may be coming to an end and a new approach is necessary. The road sector in Kenya is big business - the turnover from the KRB alone is the equivalent of over US$100 million a year (Ksh. 7.8 billion) and a sustained program of network rehabilitation might increase this level of activity to US$200 million (Ksh. 15.6 million). The road sector is among the very largest industries in Kenya, and has a crucial role throughout the economy. It is not necessary to elaborate the theoretical arguments in favor of managing roads on a commercial basis as some very simple questions and answers are perhaps sufficient: Q. Are roads providing the level and quality of services required: No Q. Are roads a major constraint to economic and social development: Yes Q. Are roads big business in Kenya: Yes Q. Are roads managed in a business oriented manner: No Q. Do road managers work within a business-type environment: No Q. Do road managers perform as business managers: No Q. Do road managers have the incentives/sanctions to perform: No Q. Can business incentives/sanctions be introduced in the RD: No 33 Q. Could the RD implement a major program efficiently: Very Unlikely Kenya may be unusual but it is not unique and the same problems have frequently been faced elsewhere. Many countries have moved from the civil service to the SAGA model for handling the road infrastructure sector. This is not to suggest that civil service departments cannot satisfactorily manage roads (just as there are some parastatal railways which are efficient, and some even efficient and profitable), but it is very doubtful whether the successful civil departments have faced similar institutional erosion over such a sustained period. Having moved forward so effectively with the institutional arrangements for the funding and overall supervision of the sector, the Government now needs to move forward and establish the core implementing agencies on a more effective and more commercial basis. Would it be possible to restructure and re-invigorate the RD? No doubt, the introduction of substantial numbers of highly paid consultants and advisers would increase capacity in the short-term but it would be neither acceptable, desirable nor sustainable. The experience of trying to restructure and re-invigorate other state transport entities, such as railways and ports, has been consistently disappointing (in Kenya and elsewhere), but they have generally succeeded when new management and incentive structures have been introduced. 5.2 The Road Agency Model Kenya should consider very seriously a shift away from the civil service model for the delivery of road infrastructure services to one more aligned with business management, incentives and sanctions. Management of the road business should be a full-time occupation, rather than a part-time activity. The move to a road agency has been discussed, from time to time, for the last ten years, ever since the RMI workshop in 1992. The shift to a Road Agency was included in the Interim Poverty Reduction Strategy Paper and in several drafts of the PRSP. The shift away from the civil service model would be a very major step for the sector and a decision is unlikely to be taken without more intensive study. But, given that the road sector needs effective implementing agencies, such a move seems inevitable, if the government is serious about the road sector. Once it is agreed that the SAGA model offers the better prospect for the professional and business-like management of the sector, further important decisions are needed: (a) The number of agencies (b) The approach to the establishment of an agency In making these decisions, the government will need to balance the future effectiveness of the institutional arrangement against the immediate cost implications and the likely disruption. Whatever the approach, the transfornation of the sector will not be completed overnight, nor without resistance from those who will be adversely affected. (a) The Number ofAgencies: The RD is responsible for the A,B and C roads and is also implementing activities for the DRCs on D, E and other roads. The activities of the RD cover an extremely wide spectrum of activities - from large contracts on major highways to spot improvements, implemented by force account, on barely engineered minor roads. This spectrum of activities needs a wide range of skills, monitoring systems and management. It would be possible for a new roads agency to encompass such a range of activities, just as the RD currently undertakes them. Improved management systems and management culture and environment should result in substantial improvements in efficiency and effectiveness. Replacing the RD and placing the road network on a more business oriented basis would be a major step in the sector reform process, it 34 would follow the previous structure for the sector, it would be very understandable and involve the minimum of change. The approach might be desirable if it were considered desirable to maintain a highways organization in each and every district. The single agency approach would help to reduce the cost of the overhead structure; there might, for be example, be a single district engineer/manager, and two inspectors with one responsible for the main roads and another for the DRC roads. The single agency approach could have clear management functions with respect to the A, B & C roads (maybe all paved roads, irrespective of classification) and a coordinating and regulatory role with respect to the DRC roads (monitoring the technical standards and methods for all works). A duplication of supporting units would be avoided, but a certain split of responsibilities would be required at the top management level, as well as very clearly defined responsibilities at the level of district road managers. However, the needs and types of engineering activities on the main and other road networks are so different that separate agencies may well be more appropriate. (i) A Highways Agency: The mandate of a roads agency, like the present RD would remain very broad. It is possible that a more focused approach with more targeted mandates would offer significant advantages. The advantages might be very substantial, if there were moves toward substantially greater involvement of the private sector in the management and maintenance of the main road network through either road concessions or long-term performance based contracting. The role of a road implementing agency in such a management environment shifts from organizing the work to performance monitoring, a radical departure from the role played by the RD. The main road network (A, B & C) is only about 14,000 kms and, in reality, the network of 'main' roads is substantially less. 70 separate district offices, plus regional or provincial offices, to handle the management and maintenance of 14,000 kms, is excessive. Regional offices, each managing 1500 - 2000 kms, might be a more cost-effective approach, especially if they were confined to the monitoring of comprehensive, long-term maintenance contracts. (ii) A Rural Roads Agency: Separating the functions of the present RD into a Highways Agency and a Rural Roads Agency could provide benefits for rural roads and the re-launch of a comprehensive rural roads strategy. Rehabilitating and then maintaining the rural road network to provide basic access for production and the rural population, should be a very high priority. But, by the nature of the roads, individual investments are small, interventions are low-technology, and they do not command the attention (both positive and negative) given to the main roads. The closure of a minor rural road in Bomet normally generates no press attention nor general public outrage: closure of the Nairobi-Mombasa Road is major news. One of the factors, perhaps, underlying the relative failure to implement the Roads 2000 approach more widely in the 1990s was not simply the shortage of resources, but also the focus of senior management time and attention on the main road contracts. A Rural Roads Agency, with clear objectives, would be much more focused on rural roads, and would almost necessarily take ownership of the program and the re-launch of Roads 2000. With the further development of the DRCs, autonomous district agencies might be considered to undertake district works, replacing the existing DRE and organization. However, a role for a Rural Roads Agency, even under this scenario, would remain. The role would be in the level of coordination, provision of technical advice and services, professional oversight and management of donor funding and reporting requirements. The approach might give rise to some confusion regarding reporting responsibilities- would the engineer/technician report to the DRC or the Rural Roads Agency? The answer is clearly to both, but for different aspects of the activities. More complex, but this type of matrix management is increasingly common in all sorts of businesses and institutions and certainly does not present insuperable problems. Indeed the DREs in the districts are already operating under a form of matrix management and they do not seem unduly stressed by the experience. 35 Separate district implementing agencies would be the most decentralized approach, and it could have attractions. However, it could also result in extensive overheads, as each agency might attempt to obtain equipment and/or facilities which would be more cost-effectively provided at the central level. Retaining the basic structure of the DRE may be a more cost-effective alternative as well as offering a more defined career structure for the professional staff. The DRE approach, modified to meet new institutional arrangements and responsibilities, may probably be the way forward for the immediate future - only so much change can be effectively accommodated at any one time. (iii) RASTR4: the appropriate arrangements for implementing agencies were discussed extensively at the KRB Workshop. The general, but not unanimous conclusion was that the two agency approach offered the most promising line of approach. It was suggested, however, that the role of the rural road agency be expanded to include the primary networks of urban areas within the districts. A Rural and Small Towns Road Agency (RASTRA) was proposed. Such an agency would overcome the shortage of professional and technical expertise in the smaller urban areas. (iv) Urban Road Agencies: There is general agreement that specific organizational arrangements are required for the major urban areas in Kenya - the big five. This report does not review in any detail the arrangements for the implementation of urban road activities. There are strong arguments, however, in favor of providing full ownership of the designated urban networks to a single institution, rather than the present division of responsibilities between the urban local authorities (routine maintenance) and the Ministry of Local Government (capital works). The local authorities should prepare the work plans, the real issue is the organization for the implementation of the work, possibilities include: * The present city or municipal engineers departments; * An urban roads agency, providing implementation services; * A Nairobi roads agency and an urban roads agency covering the other urban areas; or * Network contracts A start has been made, through KUTIP, to strengthen the capacity and systems of the existing engineering departrnents, but it is an issue whether the process can continue to the level required for effective delivery of road services. It is also an issue whether it is really efficient to have separate road maintenance organizations for the urban centers outside Nairobi (Mombasa, Eldoret, Kisumu and Nakuru), which have networks of only 150 - 350 kms. If the road activities are not separated from the other activities of the engineers department, then accountability for the funds becomes a problem, and some form of accounting separation would certainly be required. Nairobi itself has a road network of close to 2000 kms, and may require a total annual budget, recurrent and capital, of about US$9 million. This may well be sufficient for a separate agency arrangement. The other four large urban areas have, together, a total of only 1,000 kms, and a budgetary need of only US$3 million which would be marginal for a separate agency (though the role of the agency might be enhanced by the inclusion of other urban areas). One single urban road agency would have be responsible for a network of about 3,000 kmn and a potential budget of US$12 million. In view of the size of the network, the importance of Nairobi to the national economy, the level of traffic congestion, and the need for specific traffic engineering skills a specific agency may well be the best approach. For the other urban areas, the best initial approach may be to work through the existing engineering departments, with separate accounting for roads, with the consideration of moving to a road agency at a later stage, if strengthening of the engineering departments fail. 36 The four urban road networks, outside of Nairobi, are of a size which would make them amenable to network maintenance contracts, under which all the road activities could be planned, managed and executed by a large contractor. Alternatively, the activities could be planned and managed by a major consultant, who would be responsible for procuring the implementation of the works. The role of the urban road department would then be confined to the development of the work plan, the letting of a single contract (which would include performance criteria and penalties) and monitoring the performance of the works. (b) Agency Establishment: The 1992 RMI Workshop recommended that the RD should be converted into an agency and everyone should be paid more in line with their private sector counterparts. As most of the participants came from the RD, this decision was not very surprising. Certainly, it is an approach that has been used extensively elsewhere - the UK Highways Agency, for example, or TANROADS in Tanzania. Even in New Zealand, the present structure evolved from the civil service. This transformation is perhaps the approach which can be most easily achieved (in name, at least) and causes the least disruption to existing mid and lower level staff. But changing a name and increasing salaries will not necessarily change outcomes. A substantial shift in management approach, as well as major changes to the incentives/sanctions structure, is fundamental to improving institutional effectiveness. Transferring existing senior management to the new agency would be unlikely to facilitate the shift in management approach"5. The most senior management for the new agency should be recruited from the open market. Recruitment should not necessarily be restricted to engineering professionals, the engineering skills already exist, the need is for management shills and a business approach to running what is essentially a large business. The alternative approach is to establish a new agency and recruit all management and staff from the open market. Employees of the previous institution would be eligible to apply for employment but so would individuals working in the private sector or elsewhere. The process would be more disruptive but much of the institutional baggage and practices of the past would be left behind as entirely new systems would be introduced. Staff which are not recruited by the new agency would need to retrenched, with training for entry into the private sector, or transferred to other activities within the public service, as happens at the present time. However, if both approaches are predicated on the basis of new management and direction, it is probable that, over time, the two approaches will converge. As long as the incentives and sanctions structure is in place; and the new management has the authority to manage, then gradually those of the transferred staff, who cannot adjust to new systems and practices, will be replaced by new engineers who can make the transition to the new environment. The transition time may be longer but substantial convergence would be probable though possibly not complete. What is unlikely to succeed would be a conversion of the RD with little or no new management. Under such circumstances, why would substantial change be expected? Few public sector institutions have been radically transformed successfully with existing senior management. 37 5.3 The Time-Framefor Change Structural change is urgently required within the implementing agencies in the sector. A major program of rehabilitation and reconstruction is needed to generate the asset base for a sustainable road sector. It is almost inconceivable that the present implementing agencies could plan, procure, and implement efficiently and effectively a major capital program in the road sector. Without confidence in the effective capacity of the implementing agencies, it is inevitable that assistance will be. limited and provided on a piecemeal basis rather than through a broad sector approach. If Kenya were to decide to move to the agency approach for the delivery of road infrastructure services, there is no reason why the shift should take an extended period. There is already a substantial body of experience, in Africa as well as elsewhere, from which Kenya can draw. Nor is there any reason why the approach should necessarily be implemented in a 'big bang' manner with all the desired SAGA agencies being established simultaneously. A more incremental implementation could easily be adopted with, perhaps, first priority being given to the establishment of a highways agency to be responsible for the highly trafficked main road network. If the agency approach is recommended in the Sector Strategy and accepted by Government, it should be possible to establish the legislative framework by mid-2003, and have a highways SAGA operational by the beginning of 2004. 38 6. CRITICAL ISSUES: EFFECTIVE ROAD MAINTENANCE DELIVERY 6.1 The Main Road Network There is little or no regular routine maintenance being undertaken and there is no longer a system for its delivery. It is a major improvement that attention is now, after a gap of several years, being given to pothole patching but this is on an emergency rather than systematic basis. Regular ditch and culvert cleaning, shoulder repairs, grass cutting are noticeable by their general absence. A recent MORPW survey on the Maji Ya Chumvi - Miritini section of the Nairobi - Mombasa road reported that all the culverts were either buried or blocked. Re-establishing regular routine maintenance should be a high priority, if sustainability is to be achieved. A higher priority for routine maintenance in work programs and in the allocation of funding will be required but, while necessary, this is not sufficient. In addition to proper funding, a delivery mechanism is required which can effectively convert the funding into actual works, monitor the condition of the roads and report both works and progress. At present, there is neither systematic deliver, nor regular monitoring nor reporting. This is not a new issue, the lack of routine maintenance has been evident for many years, and proposals have been made, but there has been little progress. A strategic choice must be made regarding the means to re-establish effective routine maintenance. * Return to force account: the skeleton of the previous system remains with work camps spread throughout the country, deserted except for some watchmen. There seems little advantage in trying to revive a system which became both expensive and ineffective. Very few road professionals would contemplate a return to the past, but not all senior policy makers in MORPW have been road professionals. * Expand present delivery methods: increase the flying gang approach to pot-hole repairs, using the provincial resealing units, and increase the funding for single activity contracts such as drainage cleaning and grass cutting. More funding would increase activity and could expand maintenance to include the road shoulders. Major changes would be required to turn the present approach into what might be properly termed a system. Activities are divided, on a rather ad-hoc basis, between the PRE and DRE, monitoring and reporting is difficult, and ultimnate responsibilities are blurred. A move from Provincial boundaries to more operationally convenient regions would improve the coherence of the system. * Move to input-based maintenance contracting: instead of letting single activity contracts, a single contract could be awarded for the total maintenance of either a section or network of roads. The contractor would be paid on the basis of work performed, monitored through regular inspections by the implementing agency. The role of the PRE would shift from being the implementer of works to the monitor of works. The approach was going to be tested in the mid-1990s and contract documents were developed (for both small and large networks), but the pilot projects were never implemented. Such contracts may require substantially higher payments in the first year, as the maintenance contractor remedies previous maintenance neglect. * Move to output-based maintenance contracting: input-based maintenance contracting would be a major departure for the RD, but much of the world has already used that approach and has moved on to output-based performance contracting. Paying for work performed (input-based contracting) gives no incentive to the contractor to innovate, improve maintenance methods and reduce the volume of work, indeed there is a perverse incentive to increase work in order to increase payments. The output-based approach sets the standards to be maintained, in terms of road condition, and the contractor is paid a flat monthly fee to achieve and maintain these standards. 39 The fee is proportionally reduced, if the standards are not achieved. The contracts often include a lump sum in the first year to allow the contractor to bring the road(s) to a maintainable condition. Maintenance contracting has been adopted increasingly throughout the world and it is the most promising alternative for the main road network in Kenya. The present arrangements are not delivering full routine maintenance and, though improvements are possible, are very unlikely to develop into a coherent system. The real issue is not whether maintenance contracting is desirable but whether Kenya must start with input-based contracting, learn what other countries have already learnt and then move to an output-based system. Perhaps Kenya can take advantage of the experience elsewhere, leap-frog input-based contracting, and move more directly to the output-based performance system. While engineers in the RD are receptive to the concept of contract maintenance, the need for maintaining a core in-house capacity for emergency maintenance works is always stressed. Every DRE indicated that they should retain a basic fleet of maintenance equipment (graders and bulldozers) for emergencies. The argument is plausible, when experience is based on the present system of maintenance contracting, but the need to make provision for emergency maintenance is not confined to Kenya. Maintenance contracts can include provision for emergency maintenance, through provisional sums at agreed unit rates, and included in the bid evaluation process. Performance based contracts might thus include three elements: (i) a lump sum for the rehabilitation of the network, (ii) a monthly fee for maintenance of the network; and (iii) a provisional sum for emergency maintenance and unforeseen circumstances. Long-term maintenance contracting may be perceived as applicable to paved roads on which the standards change relatively slowly and conditions can be measured accurately. But, it is also possible to apply the concept to gravel roads, though obviously the criteria against which the contractor is evaluated are rather different. Performance based contracts are, for example, now in operation on the unpaved road network in Chad. Sample bidding documents have been prepared for such contracts which are applicable to both paved and unpaved roads. Decisions on the financing of the main road sector (Section 7) will also have an important impact upon the maintenance system adopted for the network. If Kenya moves toward private sector financing of some of its main road network; it will also be moving toward a form of output-based contracting, as all concession agreements require that specified standards are maintained. Maintenance concessions can be viewed as long-term performance based maintenance contracts. The RD and the MORPW have discussed the concept of maintenance contracting for the last decade, without making any definitive decision or taking any effective action. It is surely time that decisions are taken and action initiated to re-establish a coherent program of regular routine maintenance. Involvement of the private sector in long-term maintenance arrangements should form the basis for moving forward, but different approaches are possible. Some roads have already been improved to the standard under which long-term maintenance would be feasible, financing arrangements for other roads are now being discussed. It is the opportune time to move forward with extensive piloting of the different approaches, to indicate which approach is best suited to the Kenyan environment. 6.2 Urban Roads The choices for the maintenance of urban roads are very similar to those for the main road network: continuing with a mixed force account/contracting system, moving to full input-based contracting, or moving to performance based contracting. Given the compact nature of the urban road network; there would seem to be much advantage in adopting a network contract with single contracts, rather than smaller contracts covering individual links in the network. 40 6.3 Rural Roads - Re-launching Roads 2000 Kenya has a very large unpaved road network which it would be too costly to maintain using conventional engineering maintenance practices. Kenya has a large and growing pool of un and underemployed people who desperately need income earning opportunities. In many of the agriculturally most productive areas of Kenya, suitable materials for unpaved roads are becoming increasingly scarce, either the sources are exhausted or people are reluctant to allow their excavation. As a result, haul distances are rising and maintenance costs are becoming increasingly expensive. The Roads 2000 approach to the rehabilitation and maintenance of the unpaved network'6 offers an opportunity to increase the level of accessibility of the network, within these constraints: * Low cost: partial rehabilitation and spot improvements * Employment generating: labor-based works for both rehabilitation and maintenance can generate substantial unskilled and semi-skilled employment in local areas * Material tolerant: spot rather than full re-graveling, use of quarry waste and other over-sized materials The potential benefits have been recognized but they have not materialized as the concept has, as yet, been hardly applied. The restructuring of responsibilities and funding in the road sector, through the KRB Act, offers the opportunity to re-launch Roads 2000 in districts throughout Kenya. (a) Roads 2000 - The Past Donors have assisted with the introduction of the strategy in a few districts, with very encouraging results: * SIDA: Nyeri and Kirinyaga * Danida: Coast Province The EU has also provided funding for the strategy in Eastern Province but, up to now, the results have been disappointing, partly caused by payment problems, and partly because the strategy was not really applied. Further assistance for the implementation of the strategy is in various stages of preparation: * SIDA: preparing to undertake studies in Nyanza Province * EU: preparing a second phase of assistance in Eastern Province * KfW: ready to commence in Nakuru, Nandi, Kericho, Bomet, Bureti and Nyamira * ADB: selecting consultants for Kajiado, Trans Mara, Uasin Gishu, Trans Nzoia, West Pokot, Marakwet and Narok * AFD: selecting consultants for Nyandarua, Muranga and Maragwa. While a significant number of districts are covered by donor assistance, there is no donor funding for the majority of the 70 districts. In view of the great enthusiasm for the Roads 2000 approach in the early 1990s, the coverage of only 14 districts by 2000 is deeply disappointing. The strategy for re-launching Roads 2000 should take into account both the shortcomings identified in the previous individual district programs, as well as more systemic problems. The problems identified in the SIDA and Danida projects were very similar: * Delays in disbursement of GOK flmds/ lack of efficient modalities for funds' transfer17 16 Roads 2000 approach can also be applied to the off-road maintenance activities for paved roads (shoulders, drainage, and vegetation control). Incentives for labor-intensives techniques could be included in bidding documents for comprehensive maintenance contracts. 17 This has improved with the creation of the DRCs and the establishment of specific MORPW accounts in the District. Previously, funds went through the central district account. 41 • Lack of efficient accounting and management information systems * Inadequate district capacity for supervision and monitoring * Lack of appropriate contract documentation for labor-based works * Absence of pro-active leadership by MORPW The major systemic problem in the implementation of Roads 2000 was the lack of any real system. SIDA, Danida and EU implemented their projects without real leadership or coordination at any level from MORPW. There were thus significant differences in: * Inplementation strategy * Project selection and prioritization procedures * Planning and reporting procedures and formats • Contract administration modalities * The level and scope of road works As a result, an integrated approach to road maintenance was not developed, rather a set of different donor initiatives within the overall framework of Roads 2000. When starting out with a radically different concept of road rehabilitation and maintenance, the use of a variety of different implementation approaches may be neither surprising nor a disadvantage. Different approaches provide the opportunity to learn what works and what does not, what delivers results and what is sustainable. Indeed the diversity of approaches could be considered a major advantage, but only as long as the experiences of these pilot projects are brought together, the results analyzed and lessons incorporated into a more unified approach for the future. This will require leadership by the agency responsible for rural roads and a willingness on the part of the development partners to adopt unified procedures whenever it is possible within their funding rules. (b) Roads 2000 - A Way Forward In view of the size of the network and the funding constraints, Roads 2000 or a very similar approach will be needed. If Roads 2000 is to be successfully 'mainlined' as the way forward in the rural roads sector, basic changes in the past approach will be needed: Leadership: this has to be a Kenyan-led approach. The establishment of a rural roads agency or RASTRA could help provide the focus that has been lacking. Until such an agency is established, the role of the present unit within MORPW should be enhanced to act as both the coordinator of the Roads 2000 and as the national resource center for labor-based road activities. Whatever overall institutional arrangement is adopted, it is critical that Kenya rather than development partners take the leadership role. Coordination: the approach of the different development partners needs to be coordinated through the leadership of the rural road implementing agency. It may well be desirable for a coordination forum to be established at the working level, before new initiatives are launched, in order to learn from the experience of the existing donors and prepare a more uniform approach for future implementation of the program. The MORPW is seeking assistance to strengthen its coordination unit and this is very welcome. Extension: the wider adoption of Roads 2000 and labor-based works appears constrained by the lack of awareness of policy makers, engineers both within and outside the RD, local authorities and the general 42 public about the concept and potential benefits of the approach. An education and media program would help to: * Ensure that the labor based approach is understood as appropriate * Demonstrate that the approach is technically sound and offers substantial benefits * Remove the confusion between welfare and appropriate technology Standardization: as far as possible, the activities and procedures adopted in the implementation of Roads 2000 should be standardized, irrespective of the funding source. Reporting and information systems and lines of communication should be uniform, as well as common approaches to the prioritization of works and levels of intervention. The standardization of the approach would be substantially facilitated if the development partners could agree to the pooling of donor funds through the "basket offunds" approach which is increasingly applied to the health and education sectors. In view of their very small scale nature, contractors from donor countries would not be interested in participating in the works and thus tied aid would seem an irrelevance. Inclusion: even with the additional donor assistance expected, only a relatively small number of districts will be covered. Rolling out the program to all 70 districts will take time, but an overall national plan for implementation and funding should be prepared. Possibly, the allocation of the 24% 'equitable' district funding should be employed to implement the Roads 2000 approach in those districts without specified donor assistance. The 'equitable' funds, or a proportion of the funds'8, might thus be included in the basket of funds to provide implementation of the Roads 2000. Supervision: much greater emphasis will have to be placed on orienting district supervisory staff to both labor-based works and small contract administration. Many of the DREs, outside the present Roads 2000 districts, have little exposure or training in the Roads 2000 approach. Appropriate technical and administrative guidelines should be developed for supervisory staff, operating at various levels of responsibility. Contractor Development: both SIDA and Danida have trained substantial numbers of contractors in labor-based works. There is no evidence of a shortage of potential contractor capacity in any of the districts visited. The need may be more to re-orient existing contractors to labor-based works, rather than developing contractors, per se. Appropriate specification of works and contract documents could assist substantially in extending labor-based methods. However, with the previous downsizing of the RD, there are many skilled road workers now in the community and these could form the nucleus of increased contractor capacity, if it proved necessary. '8 The 24% allocation will also be needed to fund the maintenance of the main streets in the urban areas which are, at the present time, unfunded. 43 7. CRITICAL ISSUES: THE FUNDING GAP IN THE ROAD SECTOR Most, if not all, Kenyan officials in the road sector will immediately state that funding is the most critical issue in the road sector. If more funds were available from domestic and particularly external sources, then the roads could be rehabilitated and maintained. More funding is certainly required, but unless the institutional issues are resolved, there are no guarantees that the additional funding will result in a sustainable sector. The move to a sustainable road sector must include both funding and institutional change. The more effective prioritization and use of existing resources might well have a significant impact on road conditions. The road sector is already receiving over US$100 million (>Ksh. 8 billion) a year, from the Fuel Levy, and many Kenyans, outside the road sector, consider that this is such a large amount of money that it must be enough for good roads, unless it is being grossly misused. Unfortunately, roads are expensive and Kenya has an extensive network of classified roads, without even considering the unclassified network. The funding problem combines the need for both a flow of funds for the maintenance of the network, and extensive capital funding to bring back the asset base to the standard needed for routine rather than emergency maintenance. Increasing the flow of funds, through raising the fuel levy, would certainly help but would not necessarily solve the problem unless the increased flow can be converted into capital. 7.1 Maintenance Funding Over the last 10 years, a number of studies have estimated the maintenance funding needs of the classified road network. The studies have used different methods and different assumptions but they all come to the same broad conclusion, Table 15. Table 15 Total Maintenance Costs: Classified Road Network (US$ million) Year Study $ 2001 $ MOPWH 1992 90 114 E. Rausch 1992 99 125 Roads 2000 1993 81 99 Kenya PER 1993 107 130 Expenditure Needs 1994 116 138 Strategic Plan 1997 103 113 BKS 2001 110 110 Average (mean) 118 Ksh. Bn 9.2 The studies suggest that the classified road network requires about US$120 million (Ksh 9.5 billion) annually for full routine and periodic maintenance. Several of the studies assumed 'conventional' maintenance (multiple grading and full re-gravelling) on the unpaved network and thus the adoption of the Roads 2000 Strategy might reduce the total requirement. A rather different approach to the maintenance funding issue would be to determine: (a) the level of funding which is likely to be available; (b) what roads must be maintained fully; and (c) use the remaining funds to keep the rest of the network in whatever state the funding allows. Using cost data from the most recent feasibility and design studies, the following estimates of likely maintenance requirements, assuming that a Roads 2000 type approach is adopted on roads with very low traffic flows, were made, Table 16. 44 Table 16 Annual Road Maintenance Needs (US$ million) >2000 1000 - 2000 500- 1000 300 - 500 200 - 300 100 - 200 50 - 100 < 50 Total Kms: 730 2065 1590 1630 1420 4400 5515 46307 63657 Routine 3 5 2 2 1 3 4 16 36 Periodic 12 23 7 5 4 9 7 0 67 Total 15 28 9 7 5 12 10 16 103 Ksh. Bn 1.2 2.2 0.7 0.5 0.4 0.9 0.8 1.2 8.0 The total annual costs are similar to the other estimates, taking into account the assumption made for the very low trafficked roads. If the heavily trafficked main road network (>500 vehicles/day) must be fully maintained then slightly more than US$50 million (Ksh. 4.1 billion) is required. Funding needs at the lower end of the spectrum, on the very lightly trafficked network, could be about as long as a piece of string as the network is so large and the present conditions so poor. The classified road network is not, however, the only network in need of adequate road maintenance. The main paved roads in the urban areas must also be maintained. The estimates of the possible maintenance needs for the urban paved networks are shown in Table 17. Table 17 Urban Paved Road Maintenance Needs (US$ million) Routine Periodic Total Nairobi 3.3 5.8 9.0 Mombasa 0.5 0.9 1.4 Eldoret, Kisumu & Nakuru 0.6 0.9 1.5 Other KUTIP (22) 0.3 0.7 1.0 Other Municipal (22) 0.2 0.3 0.5 Towns (60) 0.1 0.1 0.2 Total (108) 5.0 8.7 13.7 Ksh. Bn 0.4 0.7 1.1 The length of paved urban road networks outside the five largest urban areas is extremely small and the bulk of maintenance need is within Nairobi. Overall, some US$5 million (Ksh.400 million) for routine maintenance is required, and US$ 8 - 9 million (Ksh.0.6 - 0.7 billion) for the regular periodic maintenance of the network. Overall annual maintenance funding needs for the classified and urban road networks might fall somewhere in the range of US$120 - 130 million (Ksh.9.5 - 10 billion), which is not enormously different to the funding being generated from the Fuel Levy and, if the revenue from vehicle license fees were to be included in the Road Fund, adequate fimding would almost certainly be available. If an additional 30,000 kms of unclassified road were to be included in the maintainable network, as part of re- classification exercise, annual maintenance costs would be increased by a further US$10 million (Ksh.800 million), assuming annual maintenance needs of US$300 - 400/lkm. From the evidence, maintenance funding would not be a constraint, if the networks required only normal maintenance. To generate this level of maintenance funding from domestic resources must be acknowledged as a really major achievement, matched by very few developing countries. The capital funding to bring the network back to a maintainable condition is the real problem. 45 7.2 Capital Funding Needs Fewer estimates have been made of the capital requirements of the network. Everyone accepts that the amount is substantial and probably increasing as necessary works are delayed: periodic maintenance -+ renovation -e rehabilitation -e reconstruction US$40,000 US$100,000 US$200,000 US$300,000 Ksh 3.1mn Ksh 7.8 mn Ksh 15.6 mn Ksh 23.4 mn MORPW uses much of its fuel levy funding to hold the network together but often without being able to undertake the full works necessary to remedy the basic structural defects and provide a reasonable future design life. The Strategic Plan estimated that approximately Ksh 35 billion was needed to bring the network back into a maintainable condition (and substantially greater funding for network upgrading). This might approximate to a present requirement of about US$700 million (Ksh.55 billion). An earlier study (Expenditure and Funding Needs) estimated that, in 2001 prices, about US$775 million was required, though this included an element for the upgrading of earth roads. A recent study (BKS) has also made estimates of the rehabilitation needs on the classified network, Table 18. Table 18 Estimated Backlog of Interventions: Classified Network (US$ million) Paved Gravel Earth Total Northem Corrdor 336 336 Rehabilitaton 191 191 Reconstruction 393 225 24 642 Total 920 225 24 1169 The BKS costs are substantially higher than the previous estimates, but they do include the upgrading of the very heavily trafficked unpaved roads to an appropriate paved condition. It may be realistic to assume that the rehabilitation needs of the network are in the range of US$ 750 - 900 rnillion (Ksh.60 - 70 billion). A significant proportion of the capital funding will be required for the Northern Corridor, but rehabilitation and reconstruction needs are widespread over the network. This estimate of capital requirements relates only to the highly trafficked, main road network. Additional funding would be required to support Roads 2000 Strategy: the cost of partial rehabilitation and spot improvements on a network of 25,000 kms might be US$125 million. Further investment would also be desirable to continue the work started in the urban areas under KUTIP. Overall, requirements might well be in the order of US$1 billion. The rate of expenditure on the Development Account in the last few years has been below US$50 million. Continuing at this rate will mean, effectively, that Kenya never clears the backlog of rehabilitation and reconstruction - the existing backlog might take 15 years, by which time a substantial part of the remaining network would require major capital expenditure. There has to be a substantial increase in capital fumding. 46 7.3 Sources of Capital Funding (a) GOK: The government has funded road projects in the past, but there are many demands in all sectors and only limited resources. GOK has also the objective of reducing its share of GDP. In all probability, the direct contribution from GOK to the rehabilitation of the road network will be modest. Moreover, if past experience is any guide, GOK funding will be more directed to road upgrading then rehabilitation or reconstruction. GOK remains central, however, to resolving the funding issue, through its direction of other available resources. (b) Donors: There was an implicit agreement, in the early 1990s, that if GOK increased maintenance funding to the levels required for a sustainable road network, the donors would help to clear the back-log of reconstruction, rehabilitation and delayed periodic maintenance. Maintenance funding has increased substantially, but donor assistance has been more limited that anticipated. Some World Bank funding has been effectively lost as a result of (a) the slow preparation of the Strategic Plan, and (b) the general reduction in donor assistance to Kenya. Other funding has been available but its use has been beset with major problems and delays. However, substantial funding for the road sector may be expected, in the future, assurning that there is: > A resumption of normal donor assistance levels to Kenya > A road sector strategy which addresses the needs and commands the support of the stakeholders > Implementing agencies which have the systems and capacity to utilize the assistance efficiently and effectively While donor assistance will probably be substantial, in comparison with recent levels, it is unlikely that it will approach the levels required to fund the entire back-log of capital works. The aid environment has changed on the part of both the donors and GOK: > Total aid levels have declined significantly over the last 10 years. NEPAD may reverse this trend, but this is not assured; > The priorities of several donors, previously active in the road sector, have shifted toward the social sectors and more direct poverty linked activities > There is a growing trend among donors to provide budget or sector support rather than specific project support. It may be too soon in the institutional reform process to expect sector support for roads in Kenya > The Ministry of Finance is giving much higher priority to general budgetary support than specific project assistance. Assuming that there is a resumption of more normal levels of support to Kenya, donor support might fall possibly somewhere within the range of US$40 - 80 million a year. A concerted and coordinated effort, similar to that undertaken in Ethiopia or Uganda, based upon the demonstrated success of the KRB, an agreed Strategy and reform process, and a well defined and prioritized investment program would obviously increase the likelihood of assistance. The overall levels of donor funding available to Kenya are not rigidly fixed, but they are circumscribed. The amount of funding which goes to the road sector, rather than health, education, agriculture, etc. will be very materially affected by the priorities established by GOK. In all likelihood, given the magnitude of the funding required, a significant funding gap will remain after GOK funding and donor assistance. 47 (c) Road Users and Concessions: If the Government does not have the funds and donor funding is limited, it will be necessary to increase the level of financing from the road user. In one way or another, the road user will always pay; either (a) the road user can fund the necessary road rehabilitation; or, (b) the road user will pay through higher vehicle operating costs on deteriorated roads. Previously, it was indicated that an increase in the fuel levy alone would not be sufficient. An injection of capital funding is necessary, and the sooner the better. The flow of funds from road users must be converted into capital. Road users could finance the additional capital funding in a number of ways: i. The fuel levy could be increased and the KRB could raise capital on either the domestic or intemational capital markets, with the loan serviced from additional fuel levy revenue. Some form of government guarantee would most probably be necessary, and interest rates would reflect the country commercial and political risks. Presumably, the capital would be transferred by the KRB to the implementing agency to invest in the network; ii. The fuel levy could be increased and roads could be concessioned to private operators to rehabilitate, manage and maintain. The additional fuel levy would be used to finance the concession through either a shadow toll arrangement or some agreed time schedule of payments. iii. The fuel levy could be maintained at its present level, and high trafficked roads could be concessioned to the private sector on a toll basis. The fuel levy generated by traffic on the toll road could either form part of the concession agreement, thus lowering the toll level, or could be used on the rest of the network. In principle, the public sector could make the toll road investment but this would not reduce the constraints. It would not provide any additional capital, but rather another flow of revenue, which is of secondary importance. The various approaches differ with respect to the funding method, the responsibilities, and the distribution of risks. (i) Capital Bond: The implicit conversion of the fuel levy revenue into a capital loan from the capital markets would keep the road network entirely within the public sector and its implementing agencies. Some may see this as a major advantage, while others might consider it a wasted opportunity, as all the risks would also remain with the public sector. The capital loan approach is probably an unlikely scenario. A financial guarantee from Government would be required and it would result in additional public sector financial liabilities, at a time when government is trying to reduce such liabilities. It would also require changes in the legislation to allow the KRB to raise capital in this way. (ii) Shadow Road Concession: The financing of a road concession through the fuel levy would transfer some of the risks to the private sector. In the case of the shadow toll, the private sector would be bearing the construction and maintenance cost risk, as well as a risk regarding the level of traffic. With the agreed payment schedule, the concessionaire is relieved of most of the traffic risk but still bears the construction and maintenance cost risks. The use of the fuel levy would separate the construction and maintenance costs of the concession from the direct user beneficiaries; all road users would be contributing to the rehabilitation of particular parts of the network. With an general overall improvement in the network, and reduction in vehicle operating costs, this may not be a major problem. The private sector would have to evaluate the likely risks of KRB defaulting on the payment of either the shadow toll use or the fixed payment schedule. On the other hand, the KRB would need to evaluate the probability of the concessionaire failing to meet performance standards, and include some form of performance bond. (iii) Toll Road Concession: This would be a much more controversial change to the system of financing and managing roads in Kenya. Road tolls were levied on certain roads and bridges during the 1980s and early 1990s but were removed with the introduction of the Fuel Levy. The concessionaire would be 48 bearing the construction, maintenance and traffic risks and there would be a direct link between the road investrnent, the road use and the road charge. Elsewhere in the world, in both developed and developing countries, road users have accepted this method of funding roads. On most networks, elsewhere, there is an alternative free road, but this is not always the case and existing roads have been improved and converted into toll facilities, such as the Maputo Corridor, linking Maputo with Gauteng. The construction and operating costs of a tolled facility would be slightly higher than for the other alternatives; toll booths have to be constructed and manned, and there would be some suppression of potential traffic (though in absolute terms, the traffic flow might well increase as a consequence of the improved operating conditions). There are relatively few routes, in Kenya, on which toll concessions would be applicable - perhaps only the Northern Corridor and the Nairobi to Thika road. The toll approach might be deemed inequitable as road users are already paying for roads through the Fuel Levy. For light vehicles, this might be the case, but such inequity already exists as all road users pay the Fuel Levy and some sections of road are in good condition, while others remain in a very poor condition. For heavy commercial vehicles, the equity argument is not applicable, as such vehicles are currently undercharged for the costs that they impose on the network. The flow of the largest commercial vehicles is concentrated along the Northern Corridor, and would be charged directly for their use of the road and responsibility for road cost. The toll would have the advantage of correcting the imbalance which exists between the level of user charges and road costs, outlined in Section 1.6. (d) Private Sector Finance The private sector is not a charity nor a donor partner and will not invest in Kenyan roads without the prospect of substantial profit. The private investor may require a 25% return on equity (not an unlikely return in the Kenyan environment) and it has to be evaluated whether there would there be any benefit to either the Kenyan road user or the economy from attracting private finance into the road sector. Obviously, if there were no limit to the availability of donor grants and/or IDA-type Credits, then private sector financing would be a very expensive and unattractive alternative. But, if the real altemative is between private sector finance and bad roads, private sector financing may be desirable. The economic rates of return from the rehabilitation/reconstruction of poor/failed paved roads with 1500 vehicles/day would typically be about 30%: the higher the traffic, the higher the rate of retum. These rates of return are substantially higher than the overall cost of private sector financing which would be about 15% (on the basis of a blend of equity and privately raised loan finance). Moreover, privately funded and managed construction projects normally have significantly lower construction and maintenance costs than similar public sector projects. The private sector is unlikely to invest just anywhere in the Kenyan road network but will be very selective, looking for the roads with high traffic and thus substantial revenue earning potential. If Kenya sees the need for private sector finance to rehabilitate the network, then the KRB and the government will be need to prepare an overall financing strategy which takes both donor and private sector finance into account. Donors would undoubtedly provide the funding required for rehabilitation/reconstructing the Northem Corridor. The IDA has already helped finance the Mtito Andei - Bachuma Gate section and is considering a supplementary credit for the Maji Ya Chumvi - Miritini section; similarly, the EU has agreed to finance both the Sultan Hamud - Mtito Andei and Mai Mahiu - Lanet sections of the corridor. But, using donor fimding for rehabilitating the remaining sections of the Northem Corridor would largely exhaust the likely availability of donor funding for a number of years. There would be little funding available for the rest of the network. A preferable solution may be to use private sector financing to invest in the Northem Corridor. This would release available donor funding to be deployed on irmportant roads elsewhere on the network which would not be candidates for private sector finance. 49 In view of the economic returns from a road network in good condition, the use of private sector finance is potentially very attractive in comparison with the alternative of bad roads. Perhaps the much larger issue is whether the private sector would be willing to invest in Kenyan roads, given the perceived political and comnmercial risks. A private sector toll concession for the Northern Corridor, for example, would be the largest single private sector investment in Kenya for many years. Generating revenue from tolls might be seen as less risky that relying on shadow tolls, paid from the fuel levy, but substantial other risks exist: * a future government might effectively nationalize the concession; * government may not allow the concessionaire to raise the toll levels in line with either inflation or the change in the exchange rate; * government might re-introduce exchange controls and prevent the repatriation of capital and/or profits; etc. Kenya is not a risk-free environment, but nor is any developing country. The World Bank has recognized the risk problem and has recently extended its coverage of partial risk guarantees to include not only IBRD but also IDA countries. This form of financial assistance to government would eliminate many of the potential non-commercial risks bome by the private sector and both encourage private sector investment as well as possibly reducing its cost. The guarantees are, however, only partial and would not cover commercial risks such as unexpectedly high construction costs or lower traffic flows than anticipated. (e) Financing Imperatives The bottom line: If Kenya wants to mobilize the level of funding needed to protect the existing investments and ensure an improved network within the next 5 - 7 years, it will have to take a very pro- active approach to mobilize both donor and private sector finance. The funding gap is such that there is not an either/or decision regarding funding: both donor and private sector finance needs to be mobilized. The success of neighboring countries in mobilizing funding for the road sector has been based upon: (i) the strongest political support with roads as one of the highest government priorities; (ii) credible implementing institutions in which investors have confidence; (iii) willingness to make structural changes when institutions are not performing. To mobilize support, Kenya will undoubtedly have to demonstrate similar resolve and take difficult and painful decisions to demonstrate that it is determined that the road sector will be managed in the same manner and with the same professionalism as a major business. 50 8. CRITICAL ISSUES: THE ROLES OF THE PUBLIC AND PRIVATE SECTORS 8.1 The Traditional Model Until relatively recently, the almost universal model was that the road sector was owned, managed and financed by government, with the public sector often implementing much of the works. The role of the private sector was restricted to implementing contracts let by the government (central and local), to the designs and standards established by the government. This model required large public works departments, large numbers of govemment engineers and usually extensive works organizations and equipment/plant holdings. This was the model inherited by Kenya, and it provided good service for many years. Overtime there have been some modifications: * an increasing proportion of the major works, both new construction and periodic maintenance, being let to the private sector until it is now about 100% * increasing use of the private sector for recurrent maintenance activities as the capacity of the force account maintenance system has declined * increasing use of privately owned equipment/plant as MTD can no longer deliver While the private sector has played an increasing role in the delivery of services, the basic management and, until recently, financing arrangements have remained fully within the public sector. The Roads Department remains the Engineer for all road contracts and provides the supervision for all or almost all Fuel Levy financed contracts, despite attempts by the Ministry of Finance to require independent consultants. In concept, at least, the introduction of the Fuel Levy and the creation of the KRB has shifted the financing arrangements from taxation, controlled by government, to user charges controlled by representatives of the road user; but the degree to which this is recognized or appreciated by Government is unclear. The traditional model is what the public and private sectors in Kenya know and both sectors find it has many advantages, providing ample opportunity for profit while demanding limited responsibility and accountability. The model is changing elsewhere, but there is, as yet, little appreciation of these changes in Kenya19 8.2 An Evolving Model: Public-Private Partnership The problems being faced in the Kenyan road sector, by the declining effectiveness of the public sector model, have been faced elsewhere. Many countries are moving toward a new model for the sector which widens and deepens the involvement of the private sector through a partnership framework with the public sector. Responsibilities for the detailed management of networks are being devolved to consortia of consultants and engineers who both manage and maintain the networks to the standards specified by the public sector. Restrictions on public sector borrowing for capital works have resulted in the increasing use of private sector finance for both tolled and untolled roads. The private sector invests and can no longer walk away after the nominal 12 months defects liability period, it is contracted for the long-term. Many of the possible elements in such a partnership have been discussed in the previous sections, e.g. on maintenance delivery systems and private sector financing. Kenya will need to decide whether they are '9 Roundtable meetings with representatives of both the construction and consulting sectors indicated that the concepts of private concessions or performance based contracting had yet to reach Kenya. 51 applicable to the Kenyan environment. Clearly, the needs and condition of the very large rural road network is very different to the networks found in developed countries and the Roads 2000 approach, based on labor-based technology and small contractors, seems far removed from private management and maintenance. But, there is no reason to suppose that contracting out the management of district roads would raise any particular problems - in a sense, this is what is already happening under the arrangement between the DRC and DRE. Perhaps, more directly comparable to partnerships elsewhere would be the network of main roads in Kenya which carry most of the inter-urban traffic. In terms of construction and standards, if not traffic, they are similar to roads anywhere. Contracting out the maintenance of the network, using the input-based approach, would raise few technical problems. Moving to an performance-based approach, which is in many respects superior, would face the funding shortage for the initial rehabilitation works. The recent BKS report on Road Concessions in Kenya put forward the concept of a toll concession for the Northern Corridor and maintenance concessions on all the other main roads with a traffic flow >500 vehicles/day (based on payment of shadow tolls from the fuel levy)20. If Government were to adopt the proposal, the entire 'main' road network would be effectively placed under private sector management and maintenance. The public sector's role would be to monitor that the service standards specified in the concession agreements were being provided. This would be a very radical departure from the traditional model, but may well be an attractive way forward, if efficient road services are to be guaranteed. The study indicated that the potential financial returns were such that the private sector could be prepared to finance the upfront capital costs on the Northern Corridor (zUS$300 million) and that the proposed tolls to finance the investment would fall within the normal range of tolls found in other countries, Table 18. Table 18 Potential Tolls: Northern Corridor MSA - Malaba MSA - NRB US c/km Ksh/km US c/km Ksh/km Carspick-ups 3.0 2.3 2.0 1.6 MatatusAight trucks/buses 7.5 5.9 5.0 3.9 Medium trucks 9.0 7.0 6.0 4.7 Heavy trucks 12.0 9.4 8.0 6.2 US$/trip Km:- 1016 532 Cars/pick-ups 30.5 2380 10.6 825 Matatus/light trucks/buses 76.2 5945 26.6 2075 Medium trucks 91.4 7130 31.9 2490 Heavy trucks 121.9 9510 42.6 3325 The question is whether road users would be prepared to pay US$0.03/km to have the quality of road now represented by the Mtito Andei - Bachuma Gate section of the Nairobi - Mombasa Road, rather than the quality represented by the Sultan Hamud - Mtito Andei or Mau Summit - Timboroa sections 20 Im many respects, long-term performance based contracting would be a simnilar approach, differing with respect to the length of the contract period and the allocation of risks. 52 The study concluded that maintenance concessions would also be feasible, but that it was unlikely that the private sector would fund all (or even a large part) of the initial capital costs and some supplementary source of capital funding would be required, a public-private partnership. Rehabilitation and long-term maintenance contracts could be let separately but this loses the synergy of the approach - the rehabilitation contractor is only tied to the defect liability period, while the maintenance contractor would include a risk premium or contract restrictions to cover the possibility of future defects arising from inadequate construction. Dividing the responsibilities would not achieve the objective of ensuring that the construction is to the highest standard, by making the contractor responsible for the next 10 - 15 years. 8.3 Evolving Donor Assistance Just as the traditional model for road sector management is changing, so are the traditional models for donor assistance. Assistance can and is still provided through the conventional approach to financing roads - such as the present contracts on the Northern Corridor. Under this approach, a consultant designs the road, and a contractor builds the road, and after a short period of time both the consultant and contractor walk away and take little/no responsibility for any design, supervision or construction defects which appear later. There is now more flexibility in donor assistance, for example: * IDA can fund part of the capital costs involved in concessioning: as long as the concession is awarded on an agreed competitive basis, IDA standard rules do not apply to subsequent procurement contracts; * IDA can fund output-based aid: the initial lump sum rehabilitation costs, for example, for a performnance based maintenance contract or, indeed, fund the entire performance based contract; * IDA can provide partial risks guarantees to the private sector, as discussed previously. By providing this type of assistance, the cost of private sector finance can be reduced and additional capital funding from the private sector can be leveraged. With this flexibility in approach, the donor assistance can be spread over a much greater network and for a more sustained period than under the conventional model. 8.4 Acceptability in the Kenyan Context: Ways Forward Nothing is for free. Vested interests in the present model are strong and entrenched, both within the public and private sectors, and these interests would be seriously affected by a changed approach. The evolving public - private partnership approach offers a very different vision for the future of the sector, and would provide an entirely different set of responsibilities for a Highways Agency than presently performed by the Roads Department. It would also require a very different approach from the consulting and construction sectors. It is very difficult to judge acceptability because the level of knowledge/understanding of the new approaches is minimal among all the principal stakeholders in the sector, whether they be in the public or private sectors. Quite simply, at the present time, it is not possible to have an informed discussion on the roles of the public : private sectors because there is so little understanding, by either the private or public sectors, of what other countries have achieved by mobilizing the advantages of each sector. Discussions with private sector contractors and consultants suggest that they are rather content with their present roles in the sector (though they would all appreciate more funding and a larger work program). At present, they take limited responsibility, even in the short-term, for the work performed. All the new initiatives of public-private partnership give a much more extensive role to the private sector, but also demand that the private sector assumes a much greater responsibility for the work undertaken. It is certainly not clear whether either the private or public sectors appreciate the major changes that modem 53 road management would require, nor is it clear that the private sector (with some notable exceptions) would welcome the shifts in responsibility. However, it is time that pressure was placed on the private (as well as public) sectors, with substantial rewards to those consultants and contractors that perform, and substantial penalties on those companies that want to maintain the present Kenyan system of rewards without responsibility. Exposure, of both the public and private sectors, to the new concepts in road sector management and financing has to be the necessary first step. Sending selected officials and representatives from the KRB to overseas courses, such as those run by the University of Birmingham, may provide one approach but it is an extremely expensive and very restrictive. The approach is targeted basically at public sector officials and provides no exposure for the private sector consultants and contractors who also need to understand how their roles could be transformed in the future, and how they will need to adapt to play these roles. Moreover, it is questionable whether the real decision makers can afford to take the time to attend such lengthy courses. If 'decision-makers' can be allowed to attend a three/four week course, are they really the decision-makers? A more promising approach would be to run a number of much shorter and targeted courses and/or seminars within Kenya. Bringing the expertise to Kenya and opening the attendance to much larger numbers of stakeholders from both the public and private sectors would be much more effective than sending a few selected individuals to Birmingham. Short courses and seminars in Kenya would allow the real decision makers to participate and understand how others have tackled similar problems to those faced by Kenya. The cost of such seminars would be trivial in comparison with road rehabilitation (and no more expensive than sending a few people to overseas courses) and financing would be easily available. All that is required is: (a) An appreciation that Kenya might be able to learn from the experience of others (in Africa as well as the developed world) (b) An appreciation that both the public and private sectors need to understand the new opportunities (c) Some organization (Institution of Civil Engineers, KRB, MORPW) to request funding and be prepare to assist with the organization Unfortunately, the inertia in the sector, combined with the vested interests is considerable. A major study of the potential for road concessioning in Kenya was requested by the Government. PPLAF agreed to finance the study and the draft Phase I report was delivered in early December, and the Final Report in April, 2002, but there was little attempt to circulate the results and recommendations of the study within Government, let alone the private sector. 54 9. CRITICAL ISSUES: ROAD SAFETY IN KENYA 9.1 The Problem The problem is grim as demonstrated by the accident statistics, Table 19. There appears to be a reasonable correlation with the level of traffic, rising in the early - mid 1990s and then stabilizing in the same way that traffic flows have leveled off on most roads. Table 19 Kenyan Road Accidents Statistics _ _ _______Road Accident Statistics Kenya 1992 to 2000 Year 1992 1993 1994 1995 1996 1997 1998 1999 2000 Accidents 12.735 12.355 11.785 12.960 13.890 14.849 14.342 14.291 13.938 Fatalities 2.673 2.516 2.424 2.617 3.000 3.022 2.972 2.823 2.819 Senousil Injured 8.495 7.734 7.652 8.661 9.313 9.618 9.632 10.160 9.659 Slightly Injured 14.959 14.150 12.884 14.332 15.431 16.133 15.896 17.038 16.539 Deaths and InLured Total 26.127 24.400 22.960 25.610 27.744 28.773 28.500 30.021 29.017 -~ - - - - - - - . FatalMtes_Grouped _ _ _ _ _ _ _ _ _ _ _ __ _ _ _ _ _ _ Fatalities - Passengers 1.100 955 907 895 1.127 1.096 1.110 1.012 1.117 FatalIties - Pedestrians 1.044 1.079 1.039 1.141 1.240 1.296 1.265 1.194 1.125 Fatalities - Others 529 482 478 581 633 6301 597 617 577 Fatalities Total 2.673 2.516 2.424 2.617 3.000 3.022 2.972 2.823 2.819 The problem is not just the level of accidents, deaths and injuries, but the fact that so little is being done by the authorities to address the issue2 . It is almost as if the authorities are resigned to the problem, and have accepted that a high rate of accidents and road deaths are inmmutable facts of life. Certainly there is little evidence in the public sector, responsible for roads and transport, that anyone believes that anything can be done. 9.2 Development of an Institutional Framework There needs to be a central coordinating body which can develop and then monitor the impact of Annual Action Plans. Such plans are a valuable tool in targeting interventions in a coordinated and comprehensive manner. The action plans lead into awareness campaigns directed at (a) road users; (b) pedestrians; and (c) decision makers. Past camnpaigns have had a positive impact but have not been sustained. Establishing a Kenya Road Safety Agency, with a Board representing the main stakeholders, with dedicated funding from either the fuel levy, license fees or insurance premiums, may be the ideal arrangement. In the interim, the KRB should consider a specific sub-committee to try and bring coordination and coherence to the individual efforts which are being made to tackle the road safety issue. It was noticeable that in the KRB Workshop, a small element of funding for road safety was proposed by one of the workshop break-out groups. But, establishing an institutional framework alone will do little to resolve the road safety issue, unless there are identified actions which have the potential to make a positive impact upon the three causes of road accidents - human behavior, vehicle condition and road conditions. 21 A large Finnida funded Road Safety Project was implemented during the late 1980s and early 1990s. Other than the accident report system still used by the Traffic Police and the updated Traffic legislation, there seem to be few other lasting impacts. No other donor has targeted road safety since the departure of Finnida. 55 9.3 Possible Safety Interventions (i) Road Conditions: bad roads cause more accidents, good roads may reduce accidents but increase their severity as a result of faster speeds. However, road safety engineering can make a major impact: * Preventive engineering on roads under design, based upon safety audits for major highway construction and rehabilitation works. The cost of preventive engineering as compared with subsequent remedial works is minimal. * Focused black spot improvement programs can be based on interviews together with accident location data. There seems little awareness of the concepts of preventive safety engineering or safety audits within either the planning or design departments of the RD. Accident data are not processed, beyond counting, and there has been no demand for such data from MORPW. Lack of technical competencies are coupled with a lack of awareness of the need to integrate safety into design. Given the small costs involved, it is difficult to imagine that safety audits and preventive engineering could not be incorporated into all new road projects, if the design department actually had the competence and commitment. Safety audits should be mandatory for all new major contracts. Preventive engineering and black spot improvement have reduced accident rates in other countries by up to 40% (ii) Road markings: until recently, if roads were marked at all, the paint had an effective life of about 6 months. Engineers saw little purpose in road markings and consequently, while included in the bills of quantities, roads are generally unmarked. The use of thermoplastic paint offers a much more durable treatment. Though significantly more expensive, the paint not only has a higher visibility but also has a life of between 2 and 10 years, depending on the conditions and traffic. The extra investment provides a quick return and should be mandatory for all new contracts. The RD should also initiate a contract for the remedial marking on the main paved road network. The paint can also be used for rumble strips, advising drivers to be cautious and reduce speeds, and would be extremely useful in complementing the unmarked sleeping policemen which have become an epidernic on Kenyan roads. (iii) Separation of pedestrians and vehicles: 40% of road accident fatalities are pedestrians. Well maintained pedestrian sidewalks should be a high priority for all urban areas, and urban road design should give priority attention to the separation of pedestrians and vehicles at bus and matatu stops. Separate sidewalks would be too expensive outside urban areas but if road shoulders were paved, or at least maintained to a reasonable standard, then a greater separation of pedestrian and vehicle would be achieved. In many cases, pedestrians are forced to walk on the pavement, as the road shoulders barely exist. (iv) Driver Licensing: Kenya is known for fraudulent drivers' licenses and there has been talk of the Kenya license not being recognized in neighboring states. The introduction of "Smart Cards" would be a step to diminish the present high level of reported fraud. (v) Drivers Hours: Smart cards are smart and their use could be combined with the re-introduction of bus time-tabling. The smart card could be used to record the passing time at stops along the major highways, reducing speeding and perhaps also enabling the traffic police to check not only speeds but also driving hours. Again, it depends upon whether decision makers perceive a need to tackle the accident rate. There appears to be a total lack of appreciation of the dangers of extended driver hours - Kenatco taxi drivers, for example, operate 24 hour shifts (24 hours on, 24 hours off). 56 (vi) Vehicle Inspection: all secondhand vehicles, imported into Kenya, are inspected for roadworthiness, as well as all vehicles which are used as commercial or public service vehicles. However, in view of the poor condition of many of these vehicles, vehicle inspection has to be made more effective, possibly in terms of technical competencies but more probably in terms of effective enforcement. From observing the Kenyan vehicle fleet, it is by no means clear why regular vehicle inspections are not applied to all vehicles, as is usually the case in other countries. Anyone taking a taxi from outside the Nairobi railway station would appreciate the need. Several stakeholders suggested that vehicle inspection should be privatized to licensed garages and workshops, as is the case in many countries. Certainly, the results are unlikely to be less effective than the present arrangements. (vii) School education: old curricula need to be revised and then taught. The interventions outlined above have been shown to improve the accident environment and their implementation in Kenya would have an impact. But, their effectiveness would be enhanced if they were coordinated and brought within an overall framework. A Kenyan road safety coordinating body, whether it be a committee of the KRB or a separate Road Safety Agency, should act to implement what has proved effective elsewhere. A new study to undertake an in-depth review of road safety might well generate additional benefits, but action on what has already been shown to be effective should not be delayed by the desirability of further study. 9.4 Enforcement of Traffic Regulations The underlying reasons for the lack of effective enforcement of the existing traffic laws in Kenya are no mystery for any Kenyan. Providing additional equipment and vehicles to the traffic police will not solve the problem nor result in any improved enforcement. It has been tried in the past (in Kenya and elsewhere), and it has failed, and there is certainly no reason to suspect that any greater success would be achieved in the future. Unless there is the political will to ensure that traffic regulations are enforced, and to take the steps necessary to achievement that enforcement, little will change. Without that political will, actions to inprove road safety will have to be based upon the assumption of no effective enforcement. The results will be substantially less than could be achieved, but there may still be an improvement. Perhaps, the one feasible way forward is public exposure, collecting the information that demonstrates that there is no enforcement of traffic regulations despite the numbers of enforcers and then making the information public. Many traffic studies are undertaken in connection with the design of road improvements. These studies collect information on the number of passengers traveling in buses and matatus, and often indicate that the vehicles are grossly overloaded. Similarly, axle-load surveys show that the weighbridges have only a limited impact. At present, the results of these surveys are only reported in feasibility and design reports. There should be a determined effort to give the results of these surveys much higher public profile, and the information should be associated with demands for explanations of why, with so many police officers stopping vehicles, regulations are not enforced. Ultimately, there must be a radical restructuring of traffic enforcement. The situation is similar to that often found within Customs Departments. Successful restructuring of Customs Departments have often involved partial privatization and/or wholesale replacement of existing staff with new employees recruited from outside the public sector, with greater incentives and sanctions. The concessioning of the much of the main road network would provide the opportunity to introduce a new element of enforcement for axle-load and vehicle weight regulations but could be extended to include enforcement of both passenger vehicle loading regulations and even vehicle speeds. 57 SECTION 2 THE RAILWAY SECTOR 58 STRA TEGIC RE VIEW: .KENYA RAIL WA Y 1. BACKGROTND While most people associate railways in Kenya with Kenya Railways, there are in fact two rail operators: Kenya Railways (KR) and Magadi Railways (MR). 1.1 Kenya Railways: The main railway operator, owned by Govemment and operated as a parastatal. KR has large assets, in its books, including: * 156 locomotives and over 6,000 wagons * 1,939 kms of track * Major workshop complex, etc In reality, the operating assets are significantly smaller, with major railway operations being restricted to: The main lines: Mombasa - Malaba 1,083 km Nakuru - Kisumu 217 km Nairobi - Thika 60 km and social obligation services (twice a week) on the Voi - Taveta Branch line. With the exception of 4 trains/week between Nairobi and Mombasa, a rudimentary commuter service into Nairobi, and the mixed operations on the Taveta Line, KR is now a freight only railway. 1.2 Magadi Railways: MR operates a dedicated, 'own account' service moving soda ash from the Magadi Soda plant to Mombasa, and bringing back inputs to Magadi. Magadi Railways: * owns and maintains the 146 km branch line between Konza and Magadi * pays trackage to KR for use of the mainline22 * operates with leased GE locomotives and wagons23 MR was started in the mid-1990s, when Magadi Soda found that its overall profitability was at risk because of KR's high tariffs, poor service and inadequate capacity. MR is a small, efficient, low cost operation. Its owners recognize that it is only required because of KR's failure to perform. If KR became a low cost, efficient operator, Magadi Soda might reconsider the need for its own rail operating company, but the present low trackage and lease charges would make the choice difficult. Low transport costs are the key to Magadi Soda's profitability, road transport is not an alternative. 22 The trackage fee is very low: in the negotiations on trackage payments, Magadi Soda was supported by specialist advisers, KR refused all expert assistance. Magadi Soda obtained a very good deal. 23 The lease charges are reported to be low for both locomotives and wagons and the reliability conditions are more stringent than agreed between KR and GE, resulting in KR paying penalties. 59 The main issues in the railway sector concern the future role and management structure of Kenya Railways, which is still considered, by many, a major national institution, when it is actually an insolvent and inefficient transport enterprise. 60 2. THE DECLINE OF AN INSTITUTION In the early 1970's, KR (then part of East African Railways) was the largest (in terms of employment) public sector enterprise in Kenya, and reputedly one of the best managed. KR carried the dominant share of freight traffic between Mombasa and Nairobi24, and had almost a monopoly of long distance traffic into Uganda. 1973 was the peak year for rail freight in Kenya, with a total of 4.6 million tonnes; the trend thereafter has been decline, with short-lived increases whenever new locomotives were introduced into service. Kenya Railways Freight Traffic 5.00 O' 3.0 - c )0a '- Ca) 101a) - ax ) ) 10 1'- 0x a-) 1) a) 0 ):n o 2.00 - - A -' ' Year By 1997, freight traffic had fallen to only 1.6 million tonnes, and in tonne-kms KR was carrying little more than the Tanzania Railways Corporation. Traffic has subsequently increased to about 2.4 million tonnes in 2001, still well below the levels that the market requires and also well below the levels needed for profitability (generally calculated to be _3.5 million tonnes, but could be lower with major staff and asset rationalization). Passenger traffic has shown a different pattern, with major changes in the passenger market, Table 1: Table 1 Kenya Railways: Passenger Market 1971 1979- 1989 1992 199S 1998 1999 2000 2001 No.'000 1773 1763 3700 2599 1624 2,849 3,870 4,201 4,120 Pass-kmn 273 499 823 586 363 347 299 302 289 Av.kmn 154 283 222 225 224 122 77 72 70 Passenger traffic was stable in the 1 970s but journey lengths increased. En the 1 980s, both passenger numbers and kms rose substantially. Thereafter, passenger-kms have fallen steadily but, from 1995, passenger numbers have more than doubled . Long distance passenger services have either been suspended or their frequency reduced, while short distance Nairobi commnuter traffic has expanded. While having social benefits, the commuter service is the largest loss-making activity. The service was 24 TeNairobi - Mombasa Road was not designed as a heavy road freight route, because of the dominant position of KR 61 terminated in the early 1990s, when GOK refused to pay compensation, but KR was instructed by the President to resume the service a couple of months later, on May Day. The factors underlying the falling traffic have been partly political, partly increased competition from an aggressive road transport sector (especially since the early 1980s)25 but mainly KR's inability to provide the level of service and capacity demanded. KR's intemal problems were, however, closely linked to political intervention in its management, operations and pricing policies. (i) Mid-1970s: political problems with the EAC resulted in the balkanization of EAR, finally leading to the creation of national railways in 1978, following the collapse of the Community. Kenya (according to the other countries) was left with the prime assets of EAR, such as the main workshops in Nairobi. The civil strife in Uganda and virtual collapse of Uganda Railways (URC) during the 1970s lost KR much of its longest distance and most profitable freight market; average freight haul distance fell by 25%. (ii) 1980s: the weaknesses of KR's management within a Government controlled environment became increasingly manifest. Tariffs could not be increased in line with inflation, and asset renewal fell accordingly. Political involvement in the appointment and tenure of senior management increased. Salaries and benefits began to fall in real terms. Donor assistance was substantial, but it became very clear that, unless management and commercial autonomy was considerably increased, KR would not be able to respond effectively to the increasing road competition (iii) Late 1980s and early 1990s: donor assistance attempted to assist with the commercialization of the management and operations of KR, within the framework of a Memorandum of Understanding between KR and the Government (a Corporate Plan). GOK would provide compensation for non- commercial services and activities, while allowing KR the management and commercial autonomy to compete on conmmercial services. Soon after the MOU was signed, it became evident that there was no commitment by either KR management or GOK to implement its terms. KR still looked to GOK for tariff approval, compensation was only paid in three years, and then only for some non-commercial activities, GOK directed traffic priorities26, a Presidential Commission kept management in virtual limbo for a year, political management was appointed, and KR's resources were directed to activities quite unconnected with railway operations (e.g., the Numerical Machining Complex and the new Foundry). Management and staff morale declined rapidly, and very little benefit was obtained from technical assistance. Eventually, a measure of commercial autonomy was achieved but, the effectiveness of KR management to deliver efficient rail services had declined substantially, with locomotive and wagon availability falling from already low levels, despite the donors funding over US$23 million of spare parts (equivalent to US$ 30 million in 2001 prices). Management was not prepared to reduce KR's cost base, and staffing levels were only reduced from 21,000 in 1990 to 19,000 in 1994. (iv) Mid 1990s: to reverse the declining motive power capacity, DFID and IDA agreed to fund the overhaul component of integrated overhaul and long-term maintenance contracts for the mainline and shunting locomotives as part of a new assistance program. KR management could not agree to competitive tendering, preferring to negotiate on a sole source basis, and the donors withdrew. A number of contracts were negotiated but only the contract with GE for 35 mainline locomotives was signed. 25 The opening of the Mombasa-Nairobi refined products pipeline in the late 1970s took away a segment of the market, but KR failed to utilize the release capacity to domninate the heavy oils market in the 1980s and 1990s. 26 Instructions were given at different times to give total priority to fertilizer and food shipments. KR could not provide capacity for commercial traffic which invested in road capacity. 62 (v) Late 1990s: the GE contract has provided 35 mainline locomotives in very good condition, but KR was unable to fund adequately the maintenance of its other locomotives or its wagon fleee 7. KR has the mainline motive power but cannot provide an effective service through a lack of available shunting and wagon capacity. Except for the GE-maintained locomotives, all other locomotives in the fleet are runnmg well beyond scheduled overhauls and all now require replacement or major rehabilitation. The financial problems of the mid - late 1990s finally resulted in a dramatic reduction in staffing levels, from 19,000 in 1994, 14,000 in 1998, and 10,000 in 2001. It is now planned to reduce staff levels to 6,500. Sometime during the mid-1990s, KR lost the confidence of its customer base, and the situation shifted from a capacity constraint to demand constraint and it will require a major effort to bring back traffic to rail. Tariffs were increased to maintain revenue, to such levels that it became cheaper to move containers by road rather than rail - a quite extraordinary situation. 27 KR was also unable to find the full costs of the GE contract, and funding has been provided by GOK to meet the contract terms (e.g., Ksh. 400 million during FY02) 63 3. PRESENT STATUS OF KENYA RAILWAYS 3.1 The Basics: KR has always been a freight railway, but also provided a very limited number of passenger services. Though the importance of passenger revenue increased after the establishment of KR, its maximum contribution to total revenue was only about 10% (in the early 1990s) and now accounts for only 3% of revenue. Trivial in terms of total revenue, passenger traffic has had a much more substantial influence on the level of costs: • Track standards, and thus costs, have had to be maintained to a much higher standards than would be the case for a freight only railway; * Passenger trains take priority in terms of both track paths (right of way) and the allocation of the most reliable locomotives. During much of the 1980s and 1990s, the real cost of passenger traffic was the revenue lost from not carrying the available freight traffic; * Passenger traffic takes up a totally disproportionate amount of senior management time, and this has been perhaps the highest cost to KR KR is a single track, general cargo railway. Though a number of commodities are important, there is no single dominant traffic. On the mainline, the track is 80/951b and allows the operation of standard mainline locomotives (17/18 tonne axle-loads), unlike the Tanzania Railways Corporation (TRC) which has an axle load of 13.5 tonnes and has to use light mainline locomotives. It is not, however, a major heavy duty freight railway capable of moving a 10,000+ tonnes trains, as in Gabon or South Africa. Trains are restricted by the length of the passing loops and power of the locomotives to a maximum of about 1,200 tonnes and average train loads are much lower, around 400 tonnes. In the 1970s it was estimated that the mainline, without additional investment, could acconmmodate about 18 - 20 trains/day in each direction. This level of capacity has probably been reduced quite significantly west of Nairobi with the deterioration in the level of communications. However, the available capacity would still allow a substantial increase in traffic, above the present 2.5 million tonnes. Moreover, the costs of investing in modem train communications would be modest and it would be perhaps one of the first investments made by a new operator. 3.2 Rail Traffic: KR has lost much of its market share. Even in the early 1980s, KR was moving at least 50% of the freight traffic on the Mombasa - Nairobi Corridor. Estimates made from a number of recent road surveys would suggest that KR now has a 20% share of an 8 million tonne market (excluding pipeline traffic). Very similar market share is reflected 'in the movements from the container terminal in Mombasa Port, about 22% of total containers, and a slightly higher share of transit containers. As indicated in the previous section, total freight traffic has fallen from over 4 million tonnes (early 1980s) to about 2.5 million tonnes, while the total freight market has increased. Several factors underlie this decline: • Lack of effective capacity to meet available demand28 * Perverse incentives within KR to keep wagons in short supply • Decline of the parastatal sector which was a major customer of KR • Political interference in the allocation of capacity and loss of commercial customers29 28 Mainly because the utilization of the available assets is so poor 29 This was prevalent in the early-mid 1990s, but has since declined 64 * Major imbalance of traffic and lack of flexibility in competing for backhauls * Decline of some major customers KR had the opportunity in the aftermath of the El Nino rains (roads in very poor shape and axle-load control pushing up road freight rates by 50 - 100%) to get back into the market, but there was not the capacity to respond effectively. 3.3 KR's Operations: KR's freight operating pattern is relatively simple as almost all the traffic moves between an extremely limited number of stations; 18 stations account for over 85% of the total freight traffic. Block train operations have been introduced several times in the last 10 years (each time with great publicity) between Mombasa - Nairobi and Mombasa - Kampala, but after some initial success the pattern tends to revert to a less structured pattern with substantial en-route marshalling. Terminal shunting at Mombasa takes place mainly in the port, particularly in the lines behind the container terminal, rather than at the purpose built Changamwe yards. Given the need to expand the container stacking yards, it can be expected that the rail yards in the port will be reduced in the near future and a new operating pattern introduced. Long-haul operations are hampered by URC's preference to use the wagon ferry route, Kisumu - Port Bell, rather than the direct rail route through Tororo and Jinja. The wagon-ferry route has several advantages for URC: (i) the ferries are cheap to operate (if the capital costs are ignored); (ii) the Jinja - Kampala route has poor alignment (both vertical and horizontal) and is in poor condition; and (iii) wagon clearance is faster through Kisumu, with the change in mode, than through the Malaba - Tororo border crossing30. The Kisumu branch line, however, has 601b track and numerous viaducts which restrict train lengths, axle-loads and the type of locomotive. This results in wagon axle-loads being restricted to 15 tonnes, throughout the system, and trains have to be divided at Nakuru, for the Kisumu section. KR reports that they are constrained by a shortage of available wagons and have plans to purchase additional wagons, if funds are available. The turnaround of the existing fleet is, however, very poor, and has deteriorated substantially over the years, Table 2 Table 2 Kenya Railways: Wagon Turnaround (days) 1976 1979 1983 1988 1990 1994 1997 2000 2001 9 12 16 19 20 16 22 21 20 During the Railways II project in the early 1990s, wagon tumaround was reduced from 20 days to 15 days, against a target of 11 days, but the improvement was short-lived. KR manages a wagon turnaround of 20 days with an average haul of under 700 krns, TRC has an average turnaround of 12 days with an average haul of over 1,000 kms. One of the causes of the very poor wagon performance is Uganda which, from the observations by both KR and TRC managements, can be best described as a railway 'Black Hole'. Wagons enter Uganda and then take for ever to re-appear on the system - terminal facilities at Kampala are reported to be inadequate and customers use wagons for storage. However, this only explains part of KR's wagon problem. Poor management, the deliberate withholding of wagons and inadequate shunting capacity are other major contributory factors. 30 The reasons for this customs clearance through Kisumu are not clear, but the phenomnenon has persisted for at least 12 years 65 3.4 Locomotives and Rolling Stock. Some railways are efficient3', others are both efficient and commercial. The former carry large volumes of traffic at relatively low cost, but with pricing policies that do not generate adequate returns. The latter move traffic efficiently and make profits. For much of the period KR was neither very efficient nor commercial, and then became extremely inefficient while, at the same time, attempting to become commercial. Traffic fell, tariffs increased and financial results deteriorated. It is rather obvious that a railway cannot make money unless it can move traffic, and to move traffic it must have working locomotives and rolling stock. KR has large stocks of both locomotives and wagons but too few of them work Chartl. Kenya Railways: Locomotive and Wagon Availability 100 N _80 _ 7 0 X I. = 60 -D?l +Mainline Locos _ _IDUi - Shunting Locos 40 T1 _g _ < Wagons r20 -1_ 10 _ Locomotive availability has always been poor (KR has never achieved its targets of 70% availability for mainline locomotives and 75% for shunting locomotives, and even this target availability would be considered low in many railways) but has collapsed in the 1990s, reaching 40% for mainline and 28% for shunting locomotives. The recent upturn in availability of mainline locomotives is the result of the GE contract which is maintaining 35 locomotives at about 90% availability (the remaining fleet must have an availability well below 30%). Even wagon availability has fallen below 60%; international standards are >85% availability. Shortages of working locomotives has resulted in KR hiring locomotives from both South Africa and Uganda, adding significantly to operating costs (up to 10% of working expenses), while not providing the desired movement capacity because the wagons and/or shunting capacity were not available. Locomotive availability in many developing country railways is a serious problem, but improvements can be achieved - in TRC the availability of the mainline fleet is now over 75%, and the availability of the shunting fleet is moving in the same direction, following the introduction of contracted workshop management. The GE locomotives haul about 90% of the freight traffic and could haul significantly more, if their utilization was improved from the present levels of 170 kms/day. Many of the other locomotives could then be stabled. 31 For example, the railways in Morocco 66 3.5 Infrastructure: Currently, the track is in reasonable condition. There are no major problems, only a few speed restrictions, and almost no serious derailments. However, replacement of track is becoming overdue and problerms will soon develop and further speed restrictions may have to be imposed. Rail: About 10 kms of rails between Nairobi and Mombasa are worn beyond the tolerance limits and immediate replacement is needed. KR has recently received about 400 tons, i.e., about 7.5 kms of track which should relieve the most immediate problems. KR is also interchanging rails between sections of high and low density rails to even out rail wear. About 80 kms of rails are also corrugated but can be restored through grinding. The National Railways of Zimbabwe has been making use of contractors for grinding rails and KR could also try to get the rails ground by contract. Sleepers: an estimated 160,000 steel sleepers and 16,000 wood sleepers need replacement. The railways have a stock of about 40,000 dented sleepers awaiting re-pressing. Once re-pressed, these sleepers would be used to replace the dented sleepers, which will then be available for re- pressing (l I0% of the dented sleepers have to be discarded in each cycle). Some of the re-pressed sleepers could be used to replace the rotten wood sleepers but in some locations wood sleepers have to be used and KR would need to buy some wood sleepers. The rate of re-pressing sleepers needs to be substantially improved. Fastenings and Ballast. Track fastenings for about 175 kms of track are either broken or missing and need replacement. This investment is necessary to maintain the quality of track. An estimated 160,000 cubic meters of ballast is also required. The ballast cost appears high and KR should seek lower cost supplies. Track Maintenance: The track is maintained by gangs of 10-12 persons, one gang for every 7 kilometers. There may be scope for reducing gang size and the number of gangs without new investment. The Engineering Department has proposed mechanized maintenance with new tamping machines (or rehabilitation of the existing machines) as well as ballasting, screening, and track relaying machines at a total estimated cost of about Ksh. 1.0 billion. This would enable about 2,320 staff to be retrenched, reducing wage costs by about Ksh. 400 million per year. But, with severance payments, the payback period would be close to 5 years and the decision should be left to the concessionaire. 3.6 Signaling and Communications: Even though the communication system has failed, there is no danger to operations. The train control system, though technologically outdated, is still serving its purpose well and there is really no need to replace it for quite some time, certainly not until there is a substantial increase in traffic. The decision on investment in the communication system should be left to the concessionaire. 3.7 Financial: The financial position of the railways in Kenya has not been strong for the last 25 years. One of the basic measures of railway financial performance is the working ratio: the ratio between operating costs (less depreciation) and operating revenues. A working ratio of 70% is normally considered as the absolute minimum for a railway to generate the funding required for replacement of assets. KR's working ratios, are shown in Chart 2. 67 Chart 2 Kenya Railways: Working Ratio 140 . - - 120 A c60 I I. o40- 20 -' Almost throughout the period, the working ratio has been above 80%, essentially indicating that insufficient funding was generated to replace assets, let alone make new investments from retained earnings. There was some improvement in the early 1990s and, in 1994, the working ratio dipped below 70%. However, much of this improvement was generated by the reduction in materials (staff reductions were minimal) and the postponement of other expenses (such as wage increases), both contributing to the very bad results in the following years. Staff costs have always formed a major proportion of KR's operating costs, reflecting the very high staff levels rather than high wage rates for most of the artisan and unskilled groups, Chart 3 Chart 3 Kenya Railways: Staffing Levels 25 - 20 p15 0§10- 5 ' O _, . I 0 .r II ,IIII ,IIo I, I III Id 68 Employment was increased in the early 1980s, then stayed relatively static until about 1993, and finally began to decline, with total staff numbers halved in the space of 9 years. In low wage economies, labor costs on railways should be less than 30% of total operating costs. On KR, the proportion of wage costs has been generally much higher, but with some erratic movements: (i) Rising from 36% of operating costs in 1980, to 48% in 1988 (ii) Falling gradually to 29% inl 994 (iii) Escalating to >50% in 1995/96 (iv) Stabilizing thereafter at about 35% of total operating costs It is now planned to reduce staffing levels to about 6,500 but this cannot be achieved under present track maintenance practices which are both a fixed cost (largely unrelated to the level of traffic) and labor intensive. 3.8 Insolvency: The financial position of KR deteriorated markedly in the 1990s and it has been technically insolvent for several years, servicing little of its long-term debt (at the end of FY2001, KR owed GOK over Ksh.10 billion). Technical insolvency is now moving to actual insolvency, with such large current liabilities that suppliers are increasing reluctant to deliver fuel and other essential materials. More senior management time may be devoted to juggling bills and creditors than in running a railway. Not only are the current liabilities large but they are growing and further GOK financial support will be required just to maintain KR as an operating entity. Recent estimates, by IFC, indicate that KR may require Ksh.2 billion in working capital to maintain the current level of operations; but, even with this inflow, the average age of creditors would still be about 4 months. Successive managements at KR have attempted to improve the financial position by cutting back on expenditure; for much of the period by reducing purchases of materials and spare parts rather than reducing staffing levels. While improving the reported financial results, the reduction of expenditure has simply eroded the asset base of the corporation - the quality of the track as well as the stock of locomotives and wagons. Expenditure on track maintenance has fallen in real terns in the last 10 years by 40%, expenditure on the maintenance of locomotives and wagons (excluding the GE contract) has fallen similarly, and the real value of the depreciation element in the accounts has fallen from about US$18.5 million in the early 1990s to about US$3.5 million in 2001. On the other hand, the reduction of staffing levels, through early retirement, has doubled the annual pension cost, and it is now running at twice the level of depreciation. In the early 1990s, the writing was already on the wall and a World Bank aide memoire commented that "'unless present policies change, Kenya Railways will inexorably grind to a halt". It would seem that this stage has essentially been reached and that continued operations are dependent entirely upon the goodwill of the Government and suppliers. It is difficult to envisage any actions that KR management can take to extricate KR from the present situation without a massive restructuring of its debts, a major infusion of new capital and a radical restructuring of the framework of incentives and sanctions for both management and staff. Reducing costs will not necessarily solve KR's financial problems although increasing the effectiveness of expenditures and reducing the financial leakages, which are reported to be prevalent throughout the corporation, would undoubtedly help. In all railways there is a very substantial element of fixed costs (such as track maintenance and station crossing staff) which cannot be reduced without either abandoning the services or changing technology which generally requires, at least, some investment. KR's total operating costs have basically remained constant, in real terms, over the last 20+ years, despite the major changes in traffic and staffing levels, Table 3. 69 Table 3. Kenya Railways: Operating Costs 1978 1991 1995 1997 1999 2001 Freight ton-kms (billion) 1.99 1.92 1.37 1.07 1.49 1.60 Passenger kms (billion) 0.40 0.72 0.36 0.39 0.30 0.29 Total Traffic units (billion) 2.39 2.64 1.73 1.46 1.79 1.89 Operating costs US$million 74.5 76.0 74.5 72.2 81.2 70.9 US$/traffic unit 0.031 0.029 0.043 0.049 0.045 0.038 The financial position can only be improved substantially, if Kenya Railways carries much higher levels of traffic and can further reduce its staffing levels. But to be able to increase additional traffic, Kenya Railways requires additional working locomotives and wagons and the incentives to manage the assets much more effectively. In terms of the potential, very major improvements in financial performance should be possible: an additional million tonnes of freight traffic combined with radical downsizing of the staff to about one- third of its present level would reduce the working ratio to 42%. A 50% increase in traffic and 20% reduction in costs would reduce the working ratio to 50%. However, such operating results have never been achieved by KR and it is almost inconceivable that this change could be achieved by the present corporation or within the existing parastatal institutional framework. Radical restructuring and the introduction of both private capital and management appears the only option for developing a viable commercial and competitive railway in Kenya. 70 4. THE CONCESSIONING OF KENYA RAILWAYS GOK announced its intention to bring private sector management into Kenya Railways a number of years ago, but privatization has proceeded slowly32. However, a concessioning options study has been undertaken and the form of concessioning selected and approved by Cabinet. In the mid-1990s, KR was moving forward to a form of open access with individual operators. Financial constraints in the early 1 990s resulted in a number of cooperation deals with the private sector which both rehabilitated wagons and re-powered shunting locomotives, in return for discounts on freight rates and either exclusive or preferential access to the wagons. When GOK announced its intention to privatize operations, large C/F agents, shipping lines and the oil industry expressed interest in operating their own trains. The scale of rail operations is comparatively small, less than 5 million tonnes at its peak, and the options study indicated that open access would reduce the expected returns to GOK from privatization, and reduce the interest of the private sector in taking a concession for operating and maintaining the basic infrastructure. In addition, potential train operators expressed concern about exploitation by a private track concessionaire but leaving the management of the track in the public sector would substantially diminish the potential benefits from concessioning. The options study concluded that open access would only be preferable, if GOK placed an extremely high value on developing local rail entrepreneurs and a relatively low value on financial retuns to the public sector. However, separation of track and open access for operators would offer the potential for competition and lower rail tariffs. 4.1 Vertically integrated concession: It was no surprise when GOK selected the simplest form of concessioning, which would also maximize private sector interest and generate the greatest expected financial return to GOK: a single, vertically integrated concession (i.e., including both the track and operations), with no access to other operators, other than Magadi Railways which already has trackage rights over the mainline to Mombasa. It is proposed, therefore, to replace the public sector rail monopoly by a private sector monopoly. 4.2 A Monopoly Rail Operator: While conferring a rail monopoly, the vertically integrated concession does not confer a transport monopoly to the concessionaire. The concession will have to compete for traffic with a very competitive road transport industry which is well organized, operating large truck-trailers along paved roads parallel to the railway. The railway operator should enjoy a very significant competitive advantage for the long distance traffic to Uganda, though recently road container rates were cheaper than rail. But much of the potential market is in the 500 - 600 Iam haul distance, at which distance road can be very competitive, taling into account the double handling and short distance road delivery in the rail alternative. In such a market segment, rail will need to offer much improved delivery times as .well as a substantial cost discount to road. The likelihood of competition from road transport will severely limit the monopoly power of the rail concessionaire. Kenya has a very extensive road network and it is very difficult to identify any areas which are dependent on rail for transport. It is also very difficult to identify any sectors or industries which mtight be identified as captive to rail transport, with the exception of Magadi Soda which has already its own rail operations. Under these circumstances, the monopoly power of the 32 The recruitment of the transaction adviser was at a very advanced stage, when the World Bank suspended the preparation of the Privatization and Private Sector Assistance Project which was to finance the transaction assistance as well as assisting with staff retrenchment costs. 71 railways is likely to be limnited, and financial viability may be of more concern (given the investments needed in communications, wagons, and enhanced track maintenance), than financial exploitation. Some authority for economic regulation may be necessary, despite the competition from road transport, just in case circumstances change and major new rail dependent industries emerge. A regulatory authority would probably still be required, even if the period of exclusivity was restricted to five years - to regulate the terms of access agreements. 4.3 Public Service Obligations (PSO): In many countries, railways perform both commercial and public services - supplying transport to either areas or communities without alternative access to transport. The arguments for extensive PSO in Kenya are very weak, given the extensive road network and the availability of alternative transport at rates only slightly higher than those charged by KR. The one service which might be considered as a PSO would be the commuter service into Nairobi which now accounts for most of KR's passengers. The commuter service provides a low cost altemative and assists, to a possibly limited extent, in reducing peak hour traffic on the roads through a reduction in bus and matatu traffic. A concession contract could require the continuation of a specified level of commuter services and capacity, with specified fare levels (indexed to the rate of inflation). Ideally, a PSO payment should be paid by government for these services but, on the basis of past experiences in Kenya and elsewhere, the potential concessionaires will probably discount heavily the likelihood of the PSO payment being made, and will adjust their financial offers for the cost of running these loss-making services. It is difficult to see substantial justification for a PSO on inter-city services or on branch lines - such services have declined or closed in recent years without major economic or social dislocation. 4.4 The Potentialfor a Concessioned Railway: The response of the private sector for a concession of KR will depend upon its evaluation of the potential profitability. Potential profitability is crucially dependent upon the level of possible demand. It is highly unlikely that a concessionaire would see much potential for conventional passenger services, the competition from road and air is too strong on most sectors. There may be some potential for 'tourist experience' services down to Mombasa, in which transport only makes up one element of the total experience package but the railway's profitability is dependent upon the level of freight traffic. It may be possible for a private operator to make a profit with the present level of freight traffic, through running a very efficient service with very low staffing levels and overheads, but substantial profits will only be derived from increasing traffic through attracted freight back from road. There is evidence of substantial traffic potential, especially in traffics for which rail should be able to offer a very competitive service. The movement of containers to/from the port should be a core traffic for KR, but its present share of the market is small, Table 4. Table 4. Movement of Containers from the Mombasa Container Terminal 1999 2000 2001 Rail Road % rail Rail Road % rail Rail Road % rail All Containers 20150 67430 23.0% 20190 67300 23.1% 20270 83850 19.5% TransitContainers 5340 25190 17.5% 9090 24400 27.1% 10560 26720 28.3% Rail presently only moves about 20% of containers from the port, and only about 25%of the long distance transit containers. This level of market share appears to be reflected in the total freight market. A number of recent road traffic surveys indicate that there are approximately 1,000/day heavy commercial vehicles ( Ž 3axle trucks) operating between Mombasa and Nairobi. Based on the loaded/empty ratio of vehicles and 72 average load factors, this translates to about 6.5 million tonnes (4.5 million tonnes from Mombasa and about 2 million tonnes to Mombasa). KR is currently moving about 1.5 million tonnes on the sector, so it has a total narket share of just less than 20%. An efficient operator should be able to increase rail's market share to at least 50%, with total traffic of about 4 million tonnes. While traffic on the Mombasa - Nairobi section would be clearly a major target for a private concessionaire, the distance is such that the traffic is unlikely to be hugely profitable for the rail operator, unless there are substantial customers with their own sidings. For other customers, the double-handling and short distance delivery transport is likely to require the rail operator to offer a significant discount to road transport. The longer distance transit traffic is likely to generate substantially higher profit margins for the rail operator. Traffic studies in 1999, indicated that there is presently about 2 million tonnes of long distance freight being carried by road on the Northern Corridor, between Mombasa and Western Kenya/ Uganda (split 2:1 in favor of westbound traffic). A rail operator should be able to capture a very large share of this market, if efficient services are offered. The efficiency of the rail service to Uganda (the prime market) is not under the complete control of the Kenyan concessionaire, efficient services in Kenya must be complemented by efficient services in Uganda. Uganda Railways (URC) are also being concessioned and thus the present constraints imposed on the system by URC may be removed. Overall, there is substantial potential traffic for an efficient rail operator, but it is unlikely that the trucking sector will simply roll over and allow rail to take the traffic without intense competition which could substantially reduce the freight rates and profitability. 4.5 Outlookfor Concession Fees: At one time, there were expectations that the concessioning of KR would generate a very substantial upfront capital payment from the concessionaire. Though KR has a some quite valuable capital assets, primarily the rehabilitated GR mainline locomotives, the concessionaire will have necessarily to make significant investments to rehabilitate the track, the rolling stock and the communication systems though, perhaps, it may be possible to stagger the investment. A large up front payment is unlikely for an enterprise which is insolvent (but the potential is high). GOK will benefit, however, from being relieved of the need to provide continuing support to keep KR operational as well as significant royalty payments, based on the level of traffic/revenue. In view of the expectation that rail can capture substantially greater freight traffic, it may well be appropriate to include a tiered royalty into the concession agreement. 4.6 Expediting the Concession: The present management of KR asked GOK to give it time to turn around the performance of KR. In a recent meeting of the Northern Corridor Consultative Forum, KR's Managing Director, Mr. Andrew Wanyandeh, was reported as saying that privatization was not a panacea to the problems be-devilling railway transport in the region and urging the three states to be cautious about the privatization process, noting that if not carefully done, it could bring railway transport to its knees33. KR management has approached GOK with an investment plan of approximately Ksh.15 billion (US$200 million) to revitalize the railways. Several previous managements have also confidently stated that they would reverse the decline of KR, and substantial investment has been made, but with no sustained impact. Rather than agree to anything more that the minimum funding required to keep KR operational, GOK should attempt to expedite the concession of KR. 33 Increasing recognition of the magnitude of the problems has resulted in a shift in attitudes toward concessioning. 73 The first steps have already been taken by GOK through its formal decision to concession KR as one unitary concession with local participation of about 40% in the company to whom the concession is finally awarded. Transaction advisers commenced work in October 2002; the concessioning process culminating in a concession contract is hoped to take about 18 months. This will only be achieved if GOK establishes a mechanism for quick decision making during the process, viz. * finalization of the pre-qualification documents, * evaluation of pre-qualification proposals, * finalization of the bidding documents, * evaluation of concessioning proposals, * establishing a negotiating team, * negotiations with the preferred bidder, and * signing of the concession contract. Establishing such a mechanism is the more essential as there is no formal privatization agency or commission to move the process forward and there will be important decisions to be taken and GOK should already be considering how to handle the issues. Some of the issues are common to the concessioning/privatization of most insolvent enterprises, for example: (a) the retrenchment of surplus staff including arrangements financing of severance payments; (b) making good the arrears to the pension scheme; (c) dealing with the outstanding loans and accounts payable to avoid any legal challenge. There are, however, some issues which are very specific to KR and will require policy decisions by GOK, for example, (a) KR's operational land in Nairobi city center; (b) Management/disposal of KR's workshops. Nairobi: city center land: KR has, at present, very substantial operational land holdings in the center of Nairobi: the HQ buildings, the railway station, extensive marshalling yards (many of which are occupied by derelict coaches) and very extensive workshop facilities which were constructed when Nairobi maintained all the steam locomotives of East African Railways. Land in the center of the city is potentially very valuable, and much of this operational land need not be required, in the long term, by a concessioned railway • The HQ building is much too large for a streamlined concession operator, and could very possibly be sold for redevelopment as offices or a hotel. GOK will have to decide whether to include the HQ building in the concession, or require the concessionaire to develop a new HQ, whether to sell the building or allow a residual KR to occupy the building until other assets and liabilities have been satisfactorily clear up * The Railway station may necessarily have to be part of the concession, if the concessionaire is expected to maintain a level of passenger services. However, if only commuter services are to be provided, there may be other more convenient locations and the railway station could be redeveloped. * It would seem unnecessary to have marshalling yards in the city center, especially marshalling yards which have not been effectively used for 10+ years. A very large area of land could be immediately released for redevelopment, without even affecting the operational marshalling yard. At a later stage, all the freight marshalling yards could be relocated outside the city center, leaving the mainline and some passing loops within the center. 74 The workshop complex raises particular land issues, see below, but potentially much of the land could be available for redevelopment KR's workshops: The workshop facility is huge, far larger than a concessionaire would either need or want. The workshops also contains facilities such as the numerical machining complex, the sand foundry, sophisticated machining equipment, heat treatment shop and material handling equipment which have no value for the concessionaire but in which Kenya has made major investment. If the entire facility is included in the concession, the concessionaire is likely to utilize only a small part for the maintenance of the operating assets and allow the rest of the expensive facilities to lie idle and gradually decay. If the facilities are denied completely to the concessionaire, then the bidders' interest and offers could decline sharply. It is obvious that a more innovative solution is required: i. Part of the workshop area could be designated for the use of concessionaire, and attempts could be made to sellAease the rest of the workshops to other interested parties. This would require the relocation of equipment and machinery. ii. The concessionaire could be allocated the maintenance sheds, e.g. Makadara, and they could be equipped with selected machines and equipment from the workshop. The workshop complex could then be redeveloped as a separate entity iii. The concessionaire could be provided with the use of the workshops for a limited period (for example, five years), during which time new facilities would have to be developed outside the city center. The costs of constructing the new facilities could be financed on some cost sharing basis between the concessionaire and GOK, with ownership reverting to GOK at the end of the concession. iv. The machining complex and foundry could be packaged separately and sold as the nucleus of a major engineering facility. The market demand for such extensive facilities is likely to be fairly small, given the small size of the market and the scale of the facilities; v. The numerical machining equipment and the foundry could be liquidated as individual items with the buyer to dismantle and move from the site. While there might be very few bids for the entire machining complex, there might be very much more conmmercial interest in acquiring individual machines. Whatever the method of disposal, GOK has to recognize that it will realize very little of the original investment as it made a profoundly bad commercial decision. Whatever decision is taken, it has to be made before the bidding documents are issued to potential bidders who need to know, with absolute certainty, the scale and scope of the facilities which form part of the concession. The transaction adviser can elaborate on the implications of each of the above alternatives (and there may well be other alternatives) but fundamentally it is a GOK decision. The worst of all possible worlds would be to make no decision now and then start to renegotiate the facilities after the concession has started. It is unfortunate that there has been no comprehensive land-use assessment for Nairobi, leading to an agreed land-use plan to guide the transaction adviser and the GOK decision-makers. It would, for example, lead to very different valuations of the land, depending upon whether KR's operational land was considered as part of the central business district, or as part of the industrial area. In the past, private developers have expressed great interest in the KR land and, if this interest still exists, it should be possible to generate the funding necessary for the relocation of the facilities required for rail operations, as well as realizing real value to offset the liabilities inherited from KR. 75 5. SHORT TERM ACTIONS 5.1 Introduction: The almost universal experience of rail concessioning is that it takes much longer than anticipated. Even when the concession is agreed, the actual takeover of the facilities can be delayed for any number of reasons. It is quite possible, therefore, that a private operator will not assume operational control for the next two years, and it is even possible that this could extend to three years. GOK and KR management should have plans to ensure the survival, and possibly improvement, of KR during this interim period. Heavy investment in infrastructure or operating assets on the eve of concessioning, even if finance is available, is certainly not desirable: * there is often not enough time to complete the investment, prior to the commencement of the concession, and it can interfere with the operation of the concession. * investments are unlikely to provide any additional capacity in the short term and are unlikely to provide any relief to KR in resolving its current problems. * the burden of planning and implementing investments diverts management attention from improving current operations; and most importantly * investments may very probably not be valued fully by the concessionaire: o the concessionaire could have different plans to achieve the same objectives; or o the concession could obtain the same equipment/material at less cost34. It is better to leave the concessionaire to decide on all future investments; of course, the investment requirements will get factored in the bidders' financial proposals. The investment program of about Ksh. 15 billion is clearly unsupportable by the current operations. More modest estimates of Ksh. 3.2 billion (z US$42 million) have been proposed for emergency investment to remedy KR's most serious immediate problems. Most of the emergency investment can be postponed: 'including investment in wagons, communications, locomotives, etc. All major investment decisions should be left to the concessionaire, who can choose from various alternatives - leasing equipment, outsourcing key maintenance operations, renting equipment, further improvements in utilization of equipment, redesigning operations etc. Any investment by KR, at this stage, should be restricted to maintaining basic operations and safety standards While major investment is not desirable, KR will need some cash injection in order to carry the traffic on offer (and thus demonstrate the potential to possible concessionaires). The required amount would be much less than KR's estimates, possibly around US$10 million, for spare parts, small track machines, and limited equipment for improving the information systems. However, this amount does not include the funds required for staff retrenchment, current creditors, and repayment of loans. GOK may be reluctant to provide even this level of investment, given previous experience, on the eve of railway privatization. However, any delay in the injection of funds could worsen the operational perfornance and consequently the financial position of the railways. 34 Under normal public sector procurement rules, it is impossible to procure secondhand equipment or track, but such procurement may be the most appropriate for the private sector operator, achieving the same impact at a fraction of the new cost. 76 In reality, there should be considerable scope for operational improvement, even with marginal additional inputs. The strategy for the KR management should be to recognize the most critical constraints and address them in a focused manner. The following could substantially improve performance: (a) improving wagon turnaround time fromn the 19 days to 10 days and, with some effort and inputs, to 6 days; (b) improving availability and reliability of shunting locomotives which would improve wagon turnaround; (c) improving availability and reliability of main line locomotives; (d) preventing speed restrictions; (e) double loading of containers; (f) prioritizing allocation of scarce resources; and (g) retrenching surplus staff as early as possible but only after modifying the package and making it more rational. These and other actions are discussed in more detail in the following sections. 5.2 Available Wagon Fleet: Shortage of available wagons is identified by KR as the primary constraint to increasing freight traffic and revenue. The total fleet is about 6,000, serviceable wagons are only about 3,000 and, of these, about 30% are in workshops, sick-lines, or yards, mostly waiting for spare parts, most frequently wheel tires. KR has, therefore, decided to buy 4,000 tires, which will put 500 wagons back into service. The funding constraint means that tire deliveries will spread over 2+ years which, when added to the installation time, will mean a three year program; by which time, an equal number of wheels could have become due for tire replacement. The procurement will help to maintain the size of the active wagon fleet but will probably not add significant numbers of net additional wagons. The binding constraint is not the fleet but the high wagon turnaround time. 5.3 Improved Wagon Turnaround: The current turnaround is about 19 days. This level of turnaround is quite unacceptable in any efficient and competitive railway system. Management explains the high turnaround by the standard responses of any poorly performing railways: * inadequate cash allocation for spare parts; * poor locomotive availability and reliability; * track condition and speed restrictions; * poor communications; and * poor wagon reliability While all the above may contribute, the basic fact is that a high turnaround time is largely a management and control problem. It is also central to the rail problem: by forcing a large fleet to be kept in circuit, it results in almost all resources and facilities (maintenance facilities, locomotives, line capacity) being overstressed. This negative cycle can be broken only by reducing the turnaround time. If KR were to work to a turnaround time of 10 days, it would need half the wagons and thus, half the maintenance facilities and spare parts. Management attention would be more focused on monitoring fewer wagons. If turnaround time were reduced to 6 days, which is normal for the type of operations and average hauls on KR, the management of the wagon fleet would become even better. Management actions 77 should be directed at reducing turnaround, by improving the physical management of the wagon fleet and tackling the pervasive problem of perverse incentives among traffic staff to reduce wagon availability and thus increase their scarcity value. 5.4 Reduce the Wagon Fleet: Even the briefest casual survey reveals that the yards and parking places around the sick lines and workshops are packed with wagons of all types - scrap, labeled and awaiting repairs, repaired and awaiting removal, awaiting overhaul, etc. The sick line in Nairobi, for example, has about 150 wagons awaiting repairs, at any point of time, and the workshop about 800 wagons. The large number of stationary wagons has reduced mobility, delayed shunting, and eroded management control. KR needs to undertake a comprehensive survey of these wagons and take action as follows: (a) scrap wagons awaiting disposal - strip any usable spare parts, and remove to a remote site, to await final disposal (b) low capacity and old wagons in bad condition and requiring heavy repair - strip any usable spare parts, and remove to a remote site, to awaitfinal disposal; (c) low capacity and old wagons in good condition - strip any usable spare parts, and remove to a remote site, to await final disposal; (d) high capacity wagons in bad condition - repair using the spare parts strippedfrom (a) - (c) and bring back into operation, if wagons .in (e) and 69 are insufficient for the traffic; (e) high capacity wagons in good condition but in excess of the requirement on a 10-day turnaround basis; and 3 million tones of traffic; (f) high capacity wagons in good condition and forming the core of the wagon fleet required for the projected traffic. The procurement of spare parts should be guided by the requirement of wagons in category (d). The segregation of wagons and their storage in proper places will result in better wagon management, reduced maintenance effort, reduced requirement of spare parts, higher quality of wagons and improved reliability, reduced load on shunting, improved shunting operations, and better availability. The above classification should also enable better spare parts planning, which should focus entirely on maintaining the wagons in categories (e), (f), and (d), if required. Instead of investing in 4,000 tires, the available cash can be used to improve the reliability and availability of the wagons in the above categories. 5.5 Better Scheduling of Wagon Repairs: There are no scheduling rules for taking up wagons for repair in the sick line (such as first-in-first-out, flat wagons first, light repair wagons first), resulting in sub-optimal prioritization. The sick line has a quota of 50 wagon units per day and rules need to be introduced to help ensure that the maintenance activities result in the maximum impact on traffic, e.g., wagons with light repair first, followed by those which are already placed etc. 5.6 Improved Container Wagon Utilization: Most flat wagons currently carry only one container. The tare weight of the wagons varies from 12 to 14 tons. The heavier wagons are converted from low-sided wagons but without the removal of the end walls and some times even the side walls. For many of these wagons, the permissible axle load is only close to 12 tons, which leaves only 34-36 tons for the payload. Since many containers weigh 20 tons and some now up to 27 tons, the railways have been loading only one container. This means more wagons and more 78 trains and a very high cost, inefficient operation . The railways must plan to carry two containers per wagon and should examine the possibility of: (a) Removing the end and side walls: This will increase the payload by a few tons, sufficient to allow two 20-ton containers. (b) Replacing the bogies of the flat wagons with those designed for a higher axle load-of 16 plus tons which would enable a payload of about 54 tons or two containers. 5.7 Improved Train Operations: KRC operated 16,752 trains in 2000/01 for a reported freight traffic of 2.33 million tones, 139 tons/train. Even if there are no return loads, the average net load on a loaded train is only 280 tons. One can only surmise that: (a) trains are designated between intermediate yards with substantial en-route marshalling, so that a wagon from Mombasa to Malaba might form part of 4 or five trains; and/or (b) locomotive hauling capacity is grossly underutilized. Rough calculations indicate that in an efficiently operated system, with locomotives of the power of the Class 93 and 94, less than 5,000 trains should be sufficient for this level of freight traffic. Even if the' reduced train loads on the Kisumu branch line are taken into account, the present number of train appears excessive and points to a very inefficient pattem of operations. There are periodic announcements regarding the introduction of block trains as if this were a radical innovation, but block operations have been attempted for, at least, the last 10 years. It is perhaps a measure of management failure that this pattem of operations has not become established. In few of the very limited number of traffic origins and destinations, block trains have to be the efficient operating pattern. 5.8 Realistic Mainline Locomotive Fleet: On the basis of a 20% backhaul, an average haul distance of 680 kms, 80% locomotive availability (presently achieved by the GE locomotives), and a daily utilization of 450kms/locomotive, the GE fleet of 35 locomotives should be more than sufficient for a traffic of 3 million tons. The axle load limitation between Nakuru and Kisumu requires a few additional light axle-load locomotives, and a very small number of locomotives are needed'for passenger services. KR's present requirement of about 60 locomotives per day only demonstrates that locomotive utilization is dismally low, less than 200 kms/day. This level utilization is inexplicable, given the high availability and reliability of the GE locomotives. KR should consider worldng only the GE locomotives (except between Nakuru and Kisumu) and stable all other mainline locomotives. This would reduce expenditure, since the GE contract is already committed, and improve operations by reducing failures. Management emphasis should shift to less locomotives, less trains but more block trains and increased utilization of locomotives. 5.9 Shunting: The level of shunting locomotive availability is nothing short of dismal and the rehabilitation of the Class 62 locomotives will not have a major impact. Clearly more efficient train working, to reduce shunting needs, as well as more efficient use of available shunting locomotives will necessarily be required. The shunting locomotives are not very reliable, but even with limited resources a lot more could be achieved as available locomotives are used inefficiently. Management attention should be more concentrated on shunting performance: (a) all shunting delays to be recorded and discussed by the top management; (b) as a special drive, staff should be assigned to watch these delays in different locations; (c) even main line locomotives should be 79 used for shunting, if available; (d) a more intense drive for improving the reliability of the shunting locomotives. Previously, KR has hired shunting locomotives from Uganda and this it may again be desirable in the short-term. 5.10 Infrastructure Maintenance: To maintain the track and prevent the introduction of new and more extensive speed restrictions, some track materials will need to be purchased, though all major investments, such as increasing the axle-loads on the Nakuru - Kisumu line, should be postponed. 5.11 Staff Rationalization: Since 1994, KR has been using a voluntary staff retrenchment scheme to reduce its staff strength down to 10,200 (end 2001). According to KR, a further 700 staff could be immediately retrenched and another 3,500 following mechanization/computerization in different departments. The remaining staff, about 6,000, is probably much higher than a concessionaire will ultimately require (perhaps 2,000 - 3,000). The retrenchment packages vary with the nature of employment and age of the staff: (i) The permanent staff, who are less than 50-year old, are entitled to: * severance of one month of the last pay drawn, for every year of service; * a golden handshake of Ksh, 40,000; * a transport allowance of Ksh. 10,000; * one month's salary in lieu of notice; and * a maximum of one month's salary in lieu of leave outstanding. In addition, the staff are entitled to their normal pension which is the last pay drawn if the service rendered is 40 years or more and proportionally lower for lesser service. The staff are also entitled to commutation of pension to a maximum of 25%. (ii) Permanent staff more than 50-years old can be retired with three-months notice and are entitled to pension and commutation as above. They are not entitled to any severance payments but may receive the golden handshake. (iii) Casual staff are entitled to gratuity at the rate of 12% of the last pay drawn multiplied by 12 and the number of years served subject to a maximum of 20 years. (iv) Temporary staff are entitled to a gratuity at the rate of 25% of the last pay drawn multiplied by 12 and the number of years served subject to a maximum of 20 years. The retrenchment packages have been negotiated with the unions and the normal retirement and gratuity packages are defined in the service agreements. However, the severance package is anomalous and needs modification: * a 49-year old staff with 20 years of service would receive 50 months severance + other benefits + pension and pension commutation; * a 50-year old staff would receive 3 months salary + pension and pension commutation. The severance rule should have read: "Severance payment will comprise 1 month of pay per year of service or the months required to complete 50 years of service, whichever is less." Thus staff with 12 months service before retirement would only receive 12 months severance payment. If the staff refuses, then he should be retired normally at age 50 years. 80 GOK has agreed to finance severance payments while KR has the responsibility for financing normal retirement benefits. This has distorted the retrenchment program: KR has concentrated on retrenching 2000 under-50 staff (retrenchment costs of Ksh.1.7 billion) while leaving about 2000 over-50 staff (pension commutation cost of Ksh.0.7 billion) employed. If GOK had agreed to finance pension commutations, KR could have retired the 2000 staff at much lower cost to the economy. As it is very likely that GOK will have to finance pension commutation payments at the time of concessioning, substantial savings could be achieved by financing them now. 5.12 Avoiding Insolvency: The first priority for KR is avoiding a cessation of operations as a result of suppliers refusing to supply fuel, or staff refusing to work as a result of delayed wages/salaries. Quite clearly, KR is not in a position to service its long-term official debts, and these will continue to be paid by GOK. The much more immediate problem is the short-term creditors, who are owed some Ksh. 4.7 billion (US$60 million). The short term actions proposed should help to improve the financial position of KR, but additional expenditure is required for spares, track materials, etc. KR's present financial statements, bad as they are, do not reflect the true horror of the position: maintenance of assets is determined by the cash available and not the need, and the present depreciation provision is perhaps 1/8 of the level required for asset renewal. Quite simply, it is highly likely that GOK will need to continue providing cash to keep KR running, even at its present level of operations. This should be a major incentive to ensure that the concessioning of KR takes place without the delays which have been a feature of concessioning elsewhere. It should also be an incentive to ensure that the extensive operational land assets presently used (though often misused by KR) are reduced to the minimum required by an efficient, modem operator, and the remaining assets sold, or put to altemative uses to offset KR's liabilities which will necessarily remain with GOK. A concessioned railway in Kenya should be able to make very substantial profits, and GOK should benefit from substantial tax and concession payments, but it is implausible to believe that a private concessionaire would be prepared to takeover KR's existing debts or to pay a large up-front amount for the concession. 81 SECTION 3 THE PORT SECTOR 82 THE PORT OF MOMBASA 1. BACKGROUND 1.1 Competitive Position and Performance Mombasa is Kenya's only international seaport and has an effective monopoly over Kenya's sea-trade. Kenya's other sea ports are trivial in size and facilities, and Tanga, its nearest potential competitor in Tanzania, is poorly connected by road and rail to Kenya and is a lighterage facility. Mombasa is also the port for the 'Northern Corridor', handling cargo for Uganda, southern Sudan, Rwanda, Burundi and the Democratic Republic of the Congo (DRC). For some transit traffic, Mombasa faces competition from Dar es Salaam and the 'Central Corridor' and, in recent years, the level of competition has increased significantly with the Tanzania Railway Corporation (TRC) establishing a dry port at Isaka for Rwandan traffic and actively pursuing Ugandan transit traffic. Mombasa is critical to the economy of Kenya, and the decline in its efficiency and reputation during the latter part of 1980s and through the 1990s may well have acted as a substantial constraint on the growth and diversification of the economy. In many respects, the port of Mombasa is the single most important physical impediment to development. The capacity and ievels of service provided by Kenya Railways probably deteriorated to a far greater degree than Mombasa's port services, but users could always switch to road transport. For Kenya's overseas trade, there is little alternative to Mombasa, other than air transpo*t. It is noticeable that almost all new economic enterprise in Kenya in the last 10+ years has been centred around air freight. It is difficult to envisage any serious overseas investor wishing to locate a footloose industry in Kenya, opting to be dependent on the service provided by Mombasa. Raising port services to international standards and restoring the reputation of the port should be extremely high priorities for the Government. But, they should also have been extremely high priorities throughout the past 15 years, but little positive was done until the last couple of years. Worldwide shipping trends, with the shift from conventional break bulk cargo to either bulk handling or containers, have had major impacts upon operations at Mombasa. While the port has gross berth capacity well above present and likely future demands, some facilities are under substantial pressure (under prevailing levels of productivity) while other parts of the port are almost redundant. Substantial investment will be required in new capacity, while many berths are almost total idle. If raising performance standards and service levels to intemational levels are the highest priority issues, the financing and management of future investments at the port are also crucial. Unfortunately, incremental investment in additional or modified berth capacity is rarely possible, investments tend to be large and lumpy. Large investments may be necessary, at Mombasa, within the medium term and inappropriate decisions now could result in the need for extremely large investments in the longer term. 1.2 Infrastructure: The Port of Mombasa is a collection of berth facilities, almost all located on Mombasa Island. There are five main areas for cargo handling: The Old Port: located away from the main port areas, adjacent to the Mombasa Old Town, handling smaller coasters and traditional dhows and jahazis Mbaraki: located near the mouth of the entrance channel, adjacent to the Likoni ferry crossing, handling bulk coal and clinker imports and bagged cement exports. The berth is also used to berth large non-cargo vessels, such as the QE2 and foreign naval vessels 83 English Point: located on the mainland, diagonally opposite the Old Port, dedicated to handling bulk cement exports. Kilindini: berths 1 - 10 of the main port, traditionally handling conventional cargo, soda ash exports, ro-ro vessels and cruise vessels. Grain Bulk Handlers Ltd have recently established a bulk freight terminal just outside the port area and have a way leave to unload vessels at berth 3. Kipevu: berths 11 - 1835 of the main port, includes the container terminal (berths 16 - 18) but increasingly container traffic is dominating the entire facility. Kilindini and Kipevu are separated by the Makupa Creek, and the main offices of the Kenya Ports Authority (KPA) are located by the Creek. In addition to the dry cargo handling areas, oil terminals are located in both Kilindini (products) and Kipevu (crude). The old lighter wharves, in Kilindini, are no longer operated and could be redeveloped. The channel to the port is navigable 24 hours a day, but vessels entering the inner channel is not an easy task because of the strong current towards the north. Pilotage into the port is compulsory for all but the smallest vessels. The maximum berth draft in the port is 10.36 meters (container terminal, berths 16 - 8, berth 9, and Mbaraki) which limits container shipping to second generation vessels, 1000 - 2000 TEU. 1.3 Traffic (a) Total Traffic: Mombasa handled record traffic in 2001, 10.3 million tonnes, Table 1 Table 1: Mombasa Port: 2001 Traffic (million tonnes) Imports Exports Total Dry cargo: General 2.84 1.59 4.43 Bulk 1.17 0.21 1.38 Total 4.01 1.80 5.81 Uquid Bulk 4.29 0.20 4.49 Total Port Traffic 8.30 2.00 10.30 There are major import imbalances in the flows of total traffic (4: 1) and dry cargo traffic (2: 1). The substantial increases in traffic in 2000 and 2001 were largely the result of big increases in the import of fuel, partly the result of the emergency thermal power generation required in Kenya during the drought. There was, however, a 20 percent increase in dry cargo in 2001, mainly recorded as transit traffic to Uganda. Taking a much longer time perspective, traffic growth has been erratic and relatively modest, Figures 1 and 2: 35Berth 15 was never been constructed because of the very poor geological conditions 84 Figure 1: Mombasa Port Traffic 12.0 - 10.0 - 0 . 5 80 Sj W tv W iw-" s-- * J% t -4-Dry Cargo 6.0 xprsi- Bulk Liquid _ . ''* 4. __ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _- -Total Traffic - .00 .:fi _ 0.0 Figure 2: Mombasa Port: Traffic Composition 4.0 I I A- Import:Dry Cargo c 04-. a U Import:Bulk Liquid 3.00 - g Export:Dry Cargo = 2.00 tim- Export:Bulk Liquid 0.00 Average annual growth rates can be markedly influenced by the choice of the time period. To overcome this problem, time trend analysis was used to determine the growth as well as its variability, Table 2. 85 Table 2: Mombasa Port: Growth Trends 1980 - 2001 1980 - 2001 1990- 2001 Annual Growth Annual Growth Innports: Dry 4.74% 0.79 5.66% 0.56 Liquid -0.15% 0.00 0.55% 0.01 Total 2.27% 0.69 3.15% 0.75 Export: Dry 0.51% 0.07 0.000/% 0.00 Total Dry Cargo 3.07% 0.80 3.52% 0.52 Total Bulk Liquid -0.09%/0 0.00 0.02% 0.00 Total Traffic 1.41% 0.55 2.190/o 0.63 Over a 20 year period, the annual growth in total traffic at Mombasa has been less than 1.5%, and there was almost no growth in either dry cargo exports or bulk liquids. The only consistent growth sector was dry cargo imports which have increased at an average annual rate of 4.7%. The growth in dry cargo 'imports during the 1990s was rather faster but more erratic. (b) Container Traffic: Despite the slow growth in total port traffic, the changes in cargo handling patterns have resulted in a major shift in activity within the port. Traffic has moved from the conventional berths in Kilindini to the container facilities concentrated in Kipevu. The move from break-bulk to bulk handling and containers has made much of the port's cargo storage facilities redundant while placing a premium on the container stacking areas. Many of the transit and back-of-port sheds in Kilindini are empty, and there are plans to demolish some of the transit sheds at Kipevu to facilitate ship-to-shore handling along berths 12 - 14 and expand the container stacking areas. The container terminal was constructed by the Kenya Ports Authority and opened in the early 1980s36. The growth in container traffic has been the main feature at Mombasa during the last 20 years, rising from zero in the 1970s to almost 300,000 TEU in 2001, Figure 3. Figure 3 Mombasa Port: Container Traffic 350__ -- 300 250 _ 200 - _ 1 +~~~~~~~~~~~ Loaded TEU 200 200 Empty TEU o 150 100 - _ _ - Total TEU 100 50 - O~ -O . C -9 Total container traffic has increased by an annual rate of just under 8 percent, reflecting some increase in total cargo traffic but mainly the shift from break-bulk traffic, Table 3. 36 According to KPA, the terminal was built against the advice of World Bank experts who, in the 1970s, argued that Mombasa would never need a dedicated termiinal. 86 Table 3: Mombasa Port: Container Trends 1981 - 2001 1981-2001 1990-2001 Annual Growth R' Annual Growth R2 Imports 9.99 0.93 8.87 0.87 Exports 5.23 0.68 2.12 0.19 Total full containers 7.96 0.88 6.95 0.87 Total containers 7.83 0.91 7.90 0.90 Containerization in East Africa started with export cargo and, during the 1980's, there was a slight imbalance of outbound loaded containers. During the 1990's, increased container penetration in import traffic reversed the direction of flow, and there is now a significant imbalance of loaded inbound containers, Table 4. The percentage of empty containers has generally remained within the range of 20 - 25% of total containers. Table 4: Mombasa Port: Container Traffic 1981- 2001 ('000 TEU) 1981 1986 1991 1996 2001 Loaded Import 13 39 49 85 118 Export 16 49 52 69 72 Transshipment 1 5 6 11 25 Total 30 93 107 173 215 Empty containers 14 27 28 51 76 TOTAL 44 120 136 217 291 A very substantial proportion of the conventional break-bulk dry cargo has now been containerized, leaving mainly iron & steel, some bagged commodities, and equipment and vehicles to be handled at the conventional berths. (c) Transit Traffic: Mombasa serves a much larger hinterland than Kenya, acting as a primary access route to the sea for Uganda, Rwanda, Southern Sudan, Eastern DRC, and to a much lesser extent Burundi. Transit traffic forms a significant part of Mombasa's total traffic, but never to the same extent as at Dar es Salaam37. According to the KPA, transit traffic has increased significantly in recent years, rising from about 7.5% of total traffic in the 1980s, to >15% in the 1990s, and 20% in 2001. Transit flows in Eastern Africa have fluctuated markedly during the last twenty years, reflecting the impact of civil strife on both total volumes and traffic routings. Some of the apparent increase in transit traffic at Mombasa may be due more to changes in definition than total traffic38. 37 In the 1980s and early 1990s, transit traffic was >50% of total port traffic at Dar es Salaam. Fuel refined at Mombasa and then transported to Uganda is defined as Kenyan exports. If Uganda imports refined fuels directly from overseas, the fuel is in transit If Uganda switches to importing refined products there would be a big increase in transit traffic without any change in actual flows. 87 Uganda generates about 80 percent of the transit traffic through Mombasa, Table 5 Table 5: Mombasa Port: Transit Traffic 2001 Imports Exports Total Tons % Tons % Tons % Burundi 3 0 4 1 7 0 DRC 57 3 11 4 69 3 Rwanda 89 5 21 8 109 5 Sudan 67 4 0 0 67 3 Tanzania 126 7 20 7 146 7 Uganda 1452 79 218 80 1670 79 Other 49 3 0 0 49 2 Total 1843 100 273 100 2117 100 Mombasa has lost almost all the Burundi market to the Central Corridor and a substantial part of the Rwanda market. Dar es Salaam competes for Ugandan traffic but is handicapped by the longer distance and the wagon ferry services across Lake Victoria. However, the potential competition from the Dar es Salaam route has been an important factor in encouraging both KPA and Kenya Railways to reduce costs and attempt to improve services along the Northem Corridor. The stimulus has increased with the substantially improved service offered by the privatized container terminal in Dar es Salaam. When the privatization of the remaining port activities at Dar es Salaam and the Tanzania Railways Corporation are completed, the Central Corridor to Uganda may well become a much more effective competitor, although this may also depend on how the concession for the Uganda Railways Corporation (URC) is structured. 1.4 Shipping Sernices (a) Container Services: Mombasa is relatively well served by liner shipping. Direct scheduled services are provided to/from Europe and the Far East as well as services which connect with South Africa and the transshipment ports in the Middle East, Table. Table 5: Mombasa Port: Monthly Container Liner Services Route Services TEU Lines Europe 7 11,500 2' South Africa 18 13,750 6 Middle East 9 8,250 32 Far East 7 5,000 23 India 7 4,500 2 Aden 2 640 1 Israel 1 450 1 1 includes the East Africa Conference service and a vessel sharing joint service 2 includes services to regional hubs 3 includes a joint service The ending of apartheid has resulted in a switch in the import source from Europe to South Africa. The relatively short shipping time from South Africa has exposed the inadequacy of the traditional paper- based documentation systems; ships arrive before the documentation for clearing cargo has been received and processed. The increasing importance of South Africa as a trading partner also opens up the possibility of land-based alternatives to sea-trade. Trans-Africa already offers a direct rail service from 88 Gauteng to Kampala, and as the railways in the region are increasing brought under private management, the overall efficiency and attractiveness of this operation is likely to increase. Shipping patterns to Mombasa have also been modified by the worldwide changes in the shipping industry. The power of the traditional shipping conferences (cartels) has been substantially diminished by over-capacity in the sector and an increase in regular services from large independent operators. The patterns have also been changed by the consolidation within the shipping industry and the increasing adoption of hub and spoke operations by the large shipping lines. These lines utilize very large container vessels calling at a very restricted number of regional hub ports and feedering to/from other ports using much smaller vessels. Aden39, Salalah4o, Dubai, and to a lesser extent Durban, are already operating as hubs and it seems possible that this trend toward major regional ports and feedering. servces.is likely to continue. In view of the limited overall regional traffic, it is very doubtful that any port in East Africa will develop regional port status (with the possible exception of Mauritius, for certain trades). Discussions with the shipping lines serving Mombasa suggest that feeder services into the main East- West services, through a hub in the Middle East/Gulf, are expected to increase in importance. Maersk, with a dedicated feeder service to Salalah, already accounts for 29% of Mombasa's container traffic. The direct services to Europe are likely to continue, though they have lost market share to Maersk, and to a lesser extent American President Lines. Shipping lines to the Far East have already developed contingency plans to turn their direct services to feeder operations should circumstances require. As the excess capacity on the main East-West services increases with new, larger ships entering the trade, the feeder option will probably become more attractive in the near future The development of hub and spoke operations as a replacement for direct services to Europe does not necessarily mean an increase in the isolation of East Africa from the global economy. More frequent feeder services to the hub ports and reduced shipping costs on the main trunk services should result in relatively little change in the levels of either shipping times or costs. While shippers may be little affected, the future direction of shipping services in East Africa could have very major implications on the provision of port infrastructure. Direct services or feeder services could have substantial implications for future infrastructure investments. The present berth drafts limit the container vessels to about 1800 TEU and the costs of increasing drafts to allow larger vessels would be substantial. A substantial increase in the service frequency of relatively small feeder vessels would require a very different infrastructure configuration. How to plan the development of Mombasa Port in a situation of considerable service uncertainty is one of the major issues for the KPA and GOK. (b) Ro-Ro Services: while most ro-ro cargo is containerized, the operations take place well away from the container terminal, between Berths 3 and 5. The main ro-ro operators are Messina of Italy and Global Shipping Lines of New York. The normal monthly services provided are detailed below: Shipping Line Services Vessel Size Messina 3 600 -700 Global Shipping 4 400 -500 In 2001, just over 12,000 TEU were handled at the ro-ro berth but, in addition to the containers, vehicles and equipment are also carried on trailers in the ro-ro vessels and are off-loaded at the berth. The ro-ro operators have their own dedicated 500 TEU container stacks. In general, the ro-ro operations are self- contained with the operators providing both the on-board and on-shore handling equipment, relying very little on KPA. 39 American President Line feeds traffic from Mombasa to its East/West service at Aden 40 Maersk with Safinarine operates a shuttle service from Mombasa to connect with East-West services at Salalah 89 Car carriers from Japan generally use Berth 14, on the Kipevu side of the port, for the discharge of vehicles off their ramps. The vehicles are then stored at Kilindini. (c) Other Cargo Services: Most of the conventional berths (1-14) are multi-purpose and were used by 412 vessels in 2001. The shipping services are generally demand charters rather than regular scheduled services. The shipping patterns vary with the type and origin/destination of the cargo. The maximum size of the vessels are just over 30,000 tonnes, and the berth depth would have to be increased to 11 meters to allow an increase to 50,000 tonnes. The Shimanzi oil terminal can accommodate tankers up to 30,000 tonnes and is used by product tankers only. Crude oil is off-loaded at the Kipevu terminal which can handle tankers up to 82,000 tonnes, which is small for modem crude tankers. Almost 300 vessels discharged bulk liquids/gases during 2001 (269 oil tankers, and 30 other vessels). (d) Passenger Services: Mombasa has no dedicated passenger berth nor purpose built passenger terminal. In 2001, 50 passenger ships called at Mombasa, approximately one per week (very simnilar levels were recorded in 2000). The vessels were handled at either berths 1 or 3, and the passengers were processed in a converted port shed alongside berth 1, which is poorly equipped, ventilated, lighted and decorated. On average, about 300 passengers/vessel are handled at the port, i.e. approximately 15,000 passengers/year. Very large passenger vessels, such as the QE2, have called at Mombasa and are handled at Mbaraki, along with the coal and clinker. 1.5 Inland Transport Mombasa was designed to be an interface between shipping and the railways (then the dominant means of surface freight transport). The berths had rails alongside for direct loading as well as rails serving the cargo sheds and marshalling lines at the back of the port. Up to the mid-1970s, almost all up-country cargo was moved to/from the port by rail and this was expected to continue. The container terminal was designed to be a primarily rail served facility with rail-mounted gantry cranes (RMG) and extensive train marshalling lines. The rail facilities at the port container terminal were complemented by a rail-served inland container depot (ICD) at Embakasi (outskirts of Nairobi), equipped with rubber tyred gantry cranes (RTG) and a design capacity of 180,000 TEU. It was expected that most containers would be transshipped to rail and cleared at the ICD. When the KPA made large local currency surpluses in the early 1990s, KPA constructed additional ICDs at Kisumu and Eldoret. The market share of rail collapsed in the 1980s and 1990s and Mombasa became primarily a road served port, but retaining the inconveniences and disruption of an extensive rail infrastructure. The port area is still used by KR for marshalling of trains from the port, despite the extensive marshalling yards at Changamwe on the outskirts of Mombasa. Despite efforts to run dedicated container trains, KR and KPA have never been able to provide the level of service and capacity necessary to take full advantage of the Embakasi ICD. The facility was opened in 1985 and utilization rose to 35,OOOTEU in 1991, it then stagnated at this level until 1995 and has since fluctuated between 25,000 - 30,000 TEU, less than 20% of the ICD's capacity. The investments in the Kisumu and Eldoret ICDs have generated even less financial and economic benefits - traffic through Kisumu has never exceeded ;4,000 TEU and the Eldoret ICD has never opened to traffic and the buildings are leased to the university. The KPA investments in these additional ICDs do not reflect so 90 much poor market and financial investment analysis but no market and financial analysis - the decisions to construct the facilities were made on political instruction. On the basis of the limited statistics available, KR's present total market share of dry cargo port traffic is probably around 20%, with a slightly higher share (perhaps 25%) of transit traffic to Uganda. The remaining traffic is carried by road. 1.6 Roles of the Public and Private Sector in the Kenyan Port Sector (a) Port Facilities and Services: The port of Mombasa became a publicly-owned "almost full service port" in the mid 1980s when the KPA and the Kenya Cargo Handling Services4' were merged. "Almost full service port" because though the KPA provided all harbor management, most stevedoring and cargo handling for containers and conventional cargo, and almost all equipment, the private sector retained its participation in dedicated facilities at Mbaraki and English Point (built, maintained and operated by Bamburi Cement), as well as in the export of bulk soda ash. The participation of the private sector was substantially increased in the late 1990s with the construction of a bulk terminal, for grain and fertilizer, outside the port area but connected to a quayside discharging facility by overhead conveyors. During the mid-1990's, attempts were made to expand the role of the private sector within Mombasa in order to stem the decline in port performance. A contract was awarded by competitive tender to an international port operator to manage the container terminal. The contract failed and the operator withdrew after less than two years, having offered to convert the arrangement to a long-term concession. The management contract failed because: * The contract was imposed in a very hostile environment * The contract placed the operator in a weak management position * The operator had little control over the terminal staff * The operator had no control over the supply and maintenance of handling equipment * KPA could not provide the equipment availability and reliability Contracts were also awarded to the private sector to rehabilitate and maintain cargo handling equipment and marine craft. These contracts appeared to have been awarded on a sole-source basis and have not provided Mombasa with high levels of availability and reliability. In the latter part of the 1990s, when container productivity was very poor, the East African Conference came to agreement with the KPA to bring in and operate container handling equipment to supplement KPA's capacity, in return for priority berthing at the container terminal. This arrangement improved the service to the conference ships, but to the disadvantage of other lines. When new management was introduced into the KPA in 1999, the arrangement with the conference lines was terminated. (b) Associated Shipping and Cargo Services: while the KPA is similar in many respects to the Tanzania Harbours Authority, the Government of Kenya has not attempted to impose the types of restrictions on the shipping sector that were enforced in Tanzania. Shipping Agency services remained in the private sector and the Government did not try to introduce a Central Freight Bureau. Mombasa has thus a well established shipping and cargo related services sector with international shipping agencies and freight forwarders, as well as domestic companies (especially within the forwarding and clearing sector). 41 A parastatal which had earlier taken over cargo handling from private stevedoring companies. 91 2. ISSUES 2.1 Level and Quality of Port Services The central issue facing the GoK and KPA is how to raise the level of service provided at the Port of Mombasa to international standards. Mombasa is Kenya's only realist interface between land and sea transport which, despite the growing importance of air freight, remains the primary transport mode for intemational trade. The inadequate performance of Mombasa seems reflected in almost all aspects of port activities. (a) Marine Services: Shipping lines report problems with pilotage and marine services at the port. The port has 24 hour navigation but: * Not all the navigation beacons and channel buoys are working * On average, only 3 or 4 tugs are available, out of a fleet of six * The average actual power of the tugs is well below their design levels * Usually only one of the three pilot boats is working * There are 11 pilots on the roster, but shipping lines frequently report piloting delays KPA recognizes some of these problems, and it is reported that a further 11 pilots have been sent for training. Unfortunately, given the length of pilot training, there may be no quick resolution to this constraint under existing institutional arrangements. (b) Bulk Cargo: the level of port services for bulk dry handling are considered adequate. They are largely operated by private commercial interests (Bamburi Cement, Grain Bulk Handlers, and Magadi Soda). However, both Bamburi Cement and Grain Bulk Handlers would welcome draft deepening to allow vessels of 50,000 tons. (c) Conventional Cargo: Shipping lines using the conventional berths are critical of their neglect by the KPA which is concentrating its resources on the container terminal. KPA management has indicated that there is no intention to make further investrnents on the conventional berths. According to the users, this neglect has become manifest in: * the general state of disrepair, evident on the quaysides and sheds * the decline in the number of worling cranes and forklifts * the shortages of labor and the consequent inability of some vessels to work all their hatches KPA's policy of reducing pennanent labor and having a pool of casual labor is very probably appropriate. Unfortunately, according to the shipping agents, the system is not working very well. KPA's priorities are understandable, given its limited resources; conventional cargo traffic has been in decline for many years, and the levels of berth occupancy are low (<50%) and falling. However, some solution is necessary to ensure that remaining conventional traffic is provided with acceptable service standards. (d) Containers: Except for low value bulk traffics and some non-containerizable commodities (equipment, steel, vehicles, etc), containers now dominate sea-bome trade. Raw materials and components are shipped in containers, semi-manufactures are shipped in containers, final products are shipped in containers. Inter-modal transport, declining freight rates and just-in-time manufacturing techniques have transformed global manufacturing and companies now have much greater locational flexibility. No longer are manufacturers tied to raw material or markets, but they can locate on the basis of a range of factors - labor costs or skills, investment incentives, tax regimes, etc. A necessary condition for 92 countries to participate in this new global economy is low cost, efficient, reliable and predictable transport for components and finished products. Without such transport, countries are restricted to the old economy, exporting primary products or very basic goods processed from local raw materials. The present level of service provided to container traffic at Mombasa quite simply excludes Kenya and East Africa from the new global economy, beyond what can be handled by air transport. Mombasa's level of container productivity is currently between a third and a half of accepted international norms (at least 600 moves/ship/day42), Table 6. Table 6 Mombasa Port: Container Handling Rates (Moves/ship/day) 2000 2001 EACL Other Lines EACL Other Lines Low 193 133 248 175 High 311 173 392 251 Average 235 157 309 207 EACL = European Conference Lines While there was a 30% improvement in 2001, the overall levels of productivity remain extremely low, and it is understood that handling rates have fallen during 2002. The shipping lines have very modest expectations of the performance at Mombasa; the Conference only applies freight surcharges when handling rates fall below 200 moves/ship/day. But, container surcharges were imposed during the 1990s. The level of service provided to container traffic is constrained by three main problems: (i) inadequate equipment; (ii) Inadequate management systems; and (iii) cumbersome and time consuming procedures. (i) Inadequate Equipment: The terminal is equipped with adequate numbers of cranes and associated equipment but equipment availability and productivity is a major constraint to handling operations. The gantry cranes were purchased from Caillard in the mid-1980s under a French protocol and, according to KPA, were selected without reference to port management. It is said that the cranes were prototypes and, with the exception of the rail mounted gantries (for loading rail wagons) have never operated satisfactorily. Attempts were made in the mid-1990s to rebuild the rubber tired gantries (RTG) but the results did not meet expectations. More recently, in 2001, KPA commissioned ABB to refurbish one of the ship-to-shore gantry cranes (SSG) which had never really worked since its delivery. The refurbishment was expected to be completed by October 2001, but the work was not finished until mid- 2002. The levels of equipment availability are well below those expected for an efficient container operation (>85%), Table 7. 42 For the first and second generation vessels handled at Mombasa. For larger vessels, the handling rates can be substantially higher, e.g. 3,5000 TEU/day at Tanjung Pelapas, for 6,000TEU vessels. 93 Table 7 Mombasa Port: Equipment Stock and Availability 2001 2002 Owned Available % Owned Available % SSG 4 2 50 4 3 75 RTG 11 6 55 11 8 73 RMG 2 2 100 2 .2 100 Front loaders 5 3 60 10 7 70 Tractors/trailers 57 50 88 57 50 88 Reachstackers 3 3 100 5 5 100 Equipment availability has improved in the 2002, but is still poor for the core equipment of SSGs and RTGs. Not only has the level of availability of the SSGs and RTGs been a continuing problem, but their productivity, when working, has been well below intemational levels. SSGs are expected to make 20+ moves/hour; in Mombasa the levels of productivity have been generally below 10 moves/hour. The SSGs are undoubtedly one of the reasons for this very low level of productivity, but management and the organization of the terminal also contributes. Under THA management, the average productivity of the SSGs at Dar es Salaam was about 12 moves/hour, under the present leasee, the average productivity has risen to >20 moves/hour43. The results of the low levels of ship-to-shore handling productivity are: * Ship turnaround is slow, thus increasing ship's cost and freight rates * Berth occupancy is increased, resulting in more ships having to use berths 12 - 14 * Berths 12 - 14 are not equipped with SSGs, so containers are handled by ships gear which is significantly slower, =5 moves/hour * Shipping lines cannot deploy gearless vessels which are cheaper to build and maintain. Overall, the present level of SSG performance substantially increases shipping costs, which inevitably results in higher freight rates on the route. In addition, the high berth occupancy and berthing delays are leading to arguments that the terminal requires expansion. (ii) Inadequate Management Systems: an efficient container terminal requires not simply the rapid interchange between the vessel and the terminal but also the efficient handling and management of the containers within the terminal. While international container terminals moved to computer based systems for the organization/management of container stacking and storage in the 1980s, Mombasa has persisted with a manual card based system which has long proved ineffective. The introduction of a computer based system was as part of the proposed IDA project in the mid-1990s, but only recently has the KPA management announced that it intends to install the COSMOS container tracking system. The consequence of the lack of an adequate tracking system is not only inefficient management of the operation but also less control by the port management over the operations of the terminal. Ships agents and C/F agents could not rely on the KPA systems and kept track of their containers within the yards. The presence of these agents in the yards allowed a system of informal payments to develop between agents and equipment operators for priority treatment. This has now extended to agents paying the operators for simply handling their containers. KPA thus, to an extent, lost control of operations and operational staff in 43 Gulf transshipment ports, with modem high capacity SSGs average well over 30 moves/hour 94 the stacking yards44. The introduction of an effective computer-based control system will provide the opportunity to exclude non-terminal staff, regain full control of the terminal, and impose discipline over the operating staff. (iii) Time-Consuming Container Clearance Procedures: low berth productivity delays ships and increases freight rates, but the delays to the containers themselves are relatively modest. Freight rates to Africa are very high, by intemational standards, and poor port performance is a significant contributory factor, although the limited volumes of traffic may play a rather larger role (it is uneconomic for shipping lines to use high volume, low unit cost vessels). However, there is evidence that direct costs, such as freight rates and port charges, are only one of a number of factors in location and shipping decisions, the overall level of service is a much more important factor. It hardly matters to the importer whether it takes 12 or 24 hours to offload the containers, when it then takes another 10+ days to clear the container from the port. Many studies, including studies at Mombasa (the Tariff Study in the early 1 990s) have demonstrated the importance of service over cost - most shippers are prepared to pay more for better service. Like many ports in developing countries, the procedures required to clear containers through Mombasa are complex, cumbersome, time-consuming and expensive (Annex I provides a detailed account of the steps necessary to process a container through the port). The procedures contribute substantially to the very long average dwell times for containers at the port. Detailed estimnates were not available, but it is thought that dwell times of 20 - 30 days are common. The long dwell times reduce the effective capacity of the stacking yards and result in an increase in the overall level of congestion, thereby putting pressure on the port to expand the terminal. The KPA management has recognized the problem for years but, unfortunately, the resolution of the problem is not entirely or indeed mainly within its control. The clearance of containers requires the interaction of an almost bewildering array of institutions and officials. For a container to be released from the port, stamps or signatures have to be obtained from the following: Customs Long Room Customs' office attending verification Port Central Documents Office (CDO) KPA Revenue Department Port police attending the verification45 Port police, Senior Duty Office CID officer attending verfication CID, Senior Duty Officer Special Branch attending verification Special Branch, Senior Duty Ofjficer KPA Security Officer, attending verification KPA Security, Senior Duty Officer Kenya Bureau of Standards Not only does the port and Customs have to sign off on the documentation, but also four different police or security organizations, with each security organization having two officers sign off. Each organization has its own office, normally in a separate location from the others, and it is all undertaken on the basis of paper documentation. Cumbersome as the above may seem, it is probably an improvement on the situation prevailing in the early 1990s when the Bank commissioned a study on documentation at Mombasa. The study identified 24 44 A very similar situation persisted at Dar es Salaami, and the first action of the new operator was to exclude all outside agents from inside the container terminal. Agents were restricted to delivering or collecting containers to the terminal management at the terminal gate. 45 Verification to ensure that the contents of the container agree with the customs' declaration 95 different procedures/stamps. Recent estimates suggest that fast track clearance of containers (paying informal incentives) takes an elapsed time of about 15 days, of which about 5 days takes place after the container has been discharged from the vessel. If incentives are not paid, clearance time can take 25 days or more. Port management can streamline its procedures and can structure its storage charges to encourage importers to clear the containers as rapidly as possible, but unless the other organizations are prepared to streamline their operations, clearance will remain a lengthy process46. Three major obstacles exist with respect to a major improvement in container (and other cargo) clearance time: * Reliance on a paper rather computer based system * Objectives of the Customs * Distrust of importers and between organizations A computer based system would assist in allowing faster entry of the details required for cargo clearance, from the port of loading, for example. This would help solve the problem, on the short sea routes, of vessels arriving without the necessary documentation. A computer based system often forces a simplification of the procedures (one of the major advantages of a Customs based system like ASYCUDA). The computer based system also reduces the personal interfaces and requirement for informal payments. KPA is presently seeking to introduce a port community computer system, which will. allow computer entry for clearance by agents, as well improved interaction between the different groups involved in the port. Its full benefit will only be achieved with compatible computerization by Customs. By end of 2002, the KRA Long Room in Mombasa was not operating with an internal computer network, and no computer link existed with KRA Nairobi. The Customs Department in any country performs three roles with respect to overseas trade: i. Collection of duties and other tax charges on foreign trade ii. Control of banned or restricted substances and goods iii. Facilitation of foreign trade The degree of control exercised on the physical movement of cargo depends very much on the relative importance given to these three roles. In most developed countries, import duties have generally been reduced to almost nominal levels and import taxes are relatively trivial in total government revenue. Consequently, the customs departments concentrate on the control of banned substances and facilitation of trade. Procedures are directed to inspect physically a very small percentage of containers, normally guided by well developed risk targeting, based on the commodities, the exporters and the importers. In Kenya, and other developing countries, import taxes remain relatively high (giving an incentive to evade duties) and a very important component of total government revenue. Many governments distrust importers and thus there is a major effort to verify the nature, value and customs duty of imports. Often, pre-shipment inspection (PSI) of imports is obligatory as well as customs inspection, and sometimes the whole process of PSI and customs valuation is finally audited by an independent third party. 100% physical verification tends to be the norm rather than a more selective approach using risk targeting and random inspection. Government revenues are clearly important, and the IMF have often encouraged more comprehensive procedures to increase revenues, but the procedures and controls have a very negative effect upon both port productivity and the business environment. 46 The new management at Dar container terminal has managed to reduce average import container dwell time from almost 30 days to 10 days. While a major improvement, 10 days is still clearly far too long; it should be reduced to about 3 days. 96 The Customs Department has the duty to collect import taxes and control the import of banned or restricted items. The Port management has the need to collect cargo-related port charges (which may require knowledge of the cargo value). It is difficult to understand why four separate police or security organizations are also required to verify the contents of containers, beyond total mistrust within the system. It is extremely difficult to comprehend why Customs + CID + Special Branch are all needed to check for prohibited goods. Unless this can be resolved, it seems almost inevitable that cargo clearance at Mombasa will remain complex, expensive and time-consuming. And, unless cargo clearance can be streamlined, Kenya and East Africa will remain unattractive for time-sensitive, import dependent manufacturing and processing. It is reported that the Kenya Revenue Authority (KRA), which includes the Customs Department, intends to purchase container scanners, and visits have been made to various manufacturers. It is not, however, clear whether the introduction of scanners would substantially improve clearance times at the port. Scanners are slow, they are complex and they tend to be used to identify banned items. Scanners may be able to identify electrical equipment, but not the value of that equipment which is required for duty purposes. There is thus the danger that the scanners may be additional to, rather than a substitute for present procedures, with physical verification still being required. 2.2 Safet Port users raise a number of serious concerns regarding the level of safety in many aspects of port operations and the KPA's likely ability to respond effectively to major emergency incidents. (a) Oil Terminals: the oil industry is concerned about the structural condition of the two oil jetties. The companies contend that KPA maintenance and investment has been inadequate and the oil industry is having to repair and rehabilitate the jetties in order to maintain safety and meet international standards. At Shimanzi, the industry is paying for the rehabilitation of the pipes and jetty supports as well as providing firefighting equipment and installing a fire pump. Despite this work, the industry is seriously concerned about the basic structural soundness of the jetty and recommends a full structural survey as soon as possible. The oil industry considers that the basic framework of the Kipevu jetty is still structurally sound, but that extensive maintenance and rehabilitation are required immediately. The walkways on the jetty are said to be in such a poor state of repair that they are dangerous. Bollards are corroded and some are missing. Firefighting equipment for the terminal has been provided by the Kenya Pipeline Company, rather than by KPA. (b) Marine: the limited number of tugs available and their down-rated power have both service and important safety dimensions, as does the standard of pilotage at the port. In cases of grounding or shipwreck, the port tugs are not thought capable of performing any credible salvage operation due to their poor state of repair. (c) Firefighting Capability: only two of the port's four fire tenders are available, the water cannon and firefighting equipment on the tugs were removed some time ago (though there is now discussion of re- fitting the equipment), there is a general shortage of firefighting equipment and materials available on the quayside, the equipment for the oil jetties are kept in a locked store away from the terminals and not easily accessible in an emergency, no regular fire drills are held, other than those organized with the oil industry for the oil jetties, etc. It is very debatable whether KPA has the capacity to mount an effective 97 response to a serious fire on land (unless Mombasa town was able to provide substantial additional firefighting capacity, let alone a major fire breaking out on a tanker in the harbor. (d) Oil Spill Containment: the oil industry works closely with the KPA on the issue of oil pollution through the Oil Spill Mutual Aid Group (OSMAG). OSMAG is registered with the Oil Spill Response Centre in Durban, South Africa. The group has the objective of pooling available resources and coordinating responses in case of a pollution event. The KPA contributes a pollution control boat, a pollution control officer and a store in which all the anti-pollution equipment is stored. Regular oil spill drills are held with both the industry and KPA. The pollution control equipment available should be sufficient to contain a moderate spill. Port Regulations do not require that all vessels entering the port have double hulls, and a single hulled vessels grounding on a coral reef could result in a level of pollution beyond the present containment capacity, especially if weather conditions were unfavorable. (e) Working Practices and Worker Protection: it is clear that KPA's operational management has a rather more relaxed attitude to working practices and worker safety than should be tolerated: • Lack of proper protective gear when working on vessels - failure to ensure that stevedores are wearing, at all times, proper footwear and hard hats • Lack of proper protective gear when working on the quayside - failure to ensure that dockers are issued with adequate heavy duty gloves and the failure to either issue or insist that dockers wear adequate face masks and goggles when discharging bulk grains, fertilizers, etc. • Working unsafe equipment - some cranes have ladders and walkways without safety rails, and corroded floor sections on which it is dangerous to walk. * Lack of enforcement of basic working safety standards - failure to prevent staff from resting or riding on cranes (especially the SSGs) when they are either stationary or moving along the quayside. * Lack of adequate traffic control of private vehicles within the container terminal It is very possible that the basic requirements for safe working practices are established in the KPA's operational manuals but that they are simply ignored. No incentives exist to ensure that the practices are followed, and no sanctions on management, if they are not. 2.3 Security Access into the port is theoretically controlled by the port police, at the port gates, through port passes (with photograph) for individuals and movement passes for vehicles. These controls appear to apply only to those that enter the port by vehicle. In practice, most casual labor and other individuals on foot simply walk through the gates without challenge. KPA attempts to deny access to hawkers, taxi touts, and souvenir sellers when passenger ships or foreign naval vessels are berthed met with such resistance that they were abandoned. KPA management is well aware of the theft problems and has taken action: * tighter enforcement of controls has been introduced recently * better lighting in the port, particularly at the container port * some staff termiinated or transferred * conventional vessels, carrying sensitive cargo, no longer worked during the night shift 98 These actions have contributed to the reduction in the incidence of theft and pilferage which have been reported by both the KPA and the shipping industry47. The KPA has also been assessing the use of the private sector to both upgrade security measures and takeover port security. While theft has been reduced significantly, fraud appears to be on the increase with growing numbers of counterfeit bills of lading, delivery orders, and even port release orders. Containers have thus been collected by individuals with no legal right to them. This is an industry wide problem and its control will require agreement and cooperation between the shipping agents, forwarding agents, the KPA and Customs. Agreement and cooperation has yet to be achieved. Like many ports, Mombasa is potentially vulnerable to terrorism. There are no harbor patrol vessels nor water police to control movements from the Likoni side of the creek. Some navigation lights have been subject to theft and vandalism and incidents of vessels being boarded while at anchor or alongside the berths have been reported. Such incidents are thought to be the work of gangs using small boats from Likoni. 2.4 Port Capacity In terms of total capacity, Mombasa is operating well within its potential which is often quoted in excess of 15 million tonnes. But, total capacity would only be relevant if ships and cargo were homogenous. In reality, congestion at some berths can quite easily co-exist with under-utilization at others, as vessels and cargo cannot be efficiently transferred between the berths. The differences in utilization of the conventional and container parts of the Mombasa are visibly very striking. (a) Conventional Cargo: conventional, break-bulk cargo has been in decline for many years, largely replaced by containers and bulk handling. Much of Mombasa's berth capacity was constructed to handle conventional cargo and there is clearly no shortage of berth capacity nor storage. (b) Dry Bulk Handling Berth Capacity: the entry of Grain Bulk Handlers Ltd has led to a major change in cargo handling. The mechanical discharge system, connected to the GBHL terminal by overhead conveyor, achieves a much faster discharge rate than the previous system of discharge into hoppers, thus turning vessels around faster, reducing vessel costs and releasing berth space. It is an issue whether GBHL has the capacity for all cereal and fertilizer handling needs at the port or whether there is the market for a competitor. The GBHL discharge system has a large capacity but the storage facilities have a capacity of only 67,000 tons, which may be the limiting constraint on operations. The system is designed to handle both cereals and fertilizer but, up to now, it has only handled cereals, which occupy all the silos. GBHL have plans to increase their storage capacity but their terminal area is limited. Onward movement of cereals is reported as being entirely by road and the existing facility already has limited space for freight vehicles, which often queue on the public road outside, increasing general congestion. GBHL report that some importers are using the silos as a convenient storage facility rather than as an interchange facility and throughput capacity could be substantially increased by faster offtake. GBHL may have a large annual potential handling capacity, but it is a single berth facility and is thus vulnerable if the total shipments of bulk grains and fertilizers are erratic in volume48 and/or concentrated during certain times of the year. There may thus still be a niche market for another operator, particularly 47 There has also been a considerable reduction in the number of stowaways found on vessels in recent months. 4 Total fertilizer and grain imports, 1990 - 2000, averaged 1.0 million tonnes, but + 0.6 million tonnes, depending on the harvest and food aid. 99 an operator with a lower cost technology and facility. There is certainly space in Kilindini for another facility, if some of the redundant transit sheds were utilized or demolished to provide additional space. (c) Dry Bulk Vessel Capacity: the alongside draft at Mombasa restricts vessels to about 30,000 DWT. Both GBHL and Bamburi Cement would welcome an increase in berth drafts to 11+ meters to allow 50,000 tonne vessels. Given the level of grain/fertilizer traffic there might be a justification (but such large vessels would raise major problems for a silo capacity of only 67,000 tonnes), but Mbaraki handles <200,000 tonnes/year. (d) Container Terminal Handling Capacity: the major operating constraint at Mombasa is the effective capacity of the container terminal and plans for very substantial investments (=US$200 million) in additional container berths and stacking areas have been proposed. When asked the design capacity of the present container termiinal, KPA management will quote 250,000TEU. Total container traffic handled by Mombasa in 2001 was 290,000 TEU, of which 231,000 TEU were handled by the container terminal. The container terminal gives the impression of reaching capacity, with high berth occupancy and some container vessels diverted to conventional berths. It is not possible to state precisely the capacity of the present terminal. One consultant quoted "the conventional yardstick" for berth capacity as 700 - 1000 TEU per meter of berth49. Applying this rule to Mombasa would give capacity of 420,000 - 600,000 TEU/year, well above that quoted by KPA. Most efficient termiinals achieve rates at the higher end of this range, and productivity at the best ports (with large vessels) is way off this scale. Generally speaking, no self-respecting terminal operator should consider less than 600,000 TEU/year. Mombasa's capacity is constrained by a number of factors: * Insufficient and unreliable equipment: a second rule of thumb is 1 SSG per 100 - 120 meter of berth length. Mombasa should, therefore, have 5/6 SSGs rather than its 4 SSGs. Their availability should average >85%; until recently, at Mombasa, the rate of availability has been significantly below 75%5°. In effect, Mombasa has been operating with less than 3 SSGs. * Low equipment productivity: modern container terminals operate work with SSGs which have productivity levels of 30+ moves/hour. For much of the 1990s, the SSGs worked at 7 or 8 moves/hour; it is now reported to be =10 moves/hour. The equipment accounts for some of the problem, a recent study suggested that, even with refurbishment, the cranes would only average 12 moves/hour. Inefficient operations also contributes to low productivity. With 4 SSGs and 10 moves/hour, the berth capacity of the terminal would be 250,000 TEU, at 80% berth occupancy and 75% availability, assuming some use of ships gear in addition to the SSGs. * Inadequate container stacking yards: the container stacking yards at Mombasa are limited by the cliffs behind the port. Modem stacking yards normally have a width of 250 mn; at Mombasa, the yard averages about 120m, less than half the normal width. The container terminal, including emergency storage, has rather less than 8,000 TEU slots (at a normal average stacking of 2.5 boxes/ground slot). If the average container dwell time in the terminal is 10 days, the annual capacity of the terminal would be limited to about 280,000 TEU. Overall, under present operating conditions, the Mombasa container terminal is approaching capacity. Further growth in container traffic will require action to increase effective terminal capacity. 49 The world's best performance is 1.0 - 1.2 million TEU at the 300 meter terminal operated by Sealand/CSX at Hong Kong, i.e. 3,330 - 4,000 TEU/meter. 50 One SSG was never available, and the other SSGs have a reputation for unreliability. 100 (e) Container Terminal Vessel Capacity: the terminal is presently limited to the second generation of container vessels with capacities of up to 2,000 TEU. Much larger container vessels are now operating: TEU Length (m) Berth (i) SSGs Draft First generation 500 -1000 130 - 160 185 1 - 2 7.3 Second 1000 - 1999 160 - 215 245 2 - 3 10.7 Third 2000-3000 215-225 290 2-3 11.5 Fourth (Panamax) 3000 - 4500 255 - 290 320 3 - 4 13.5 There also Post-Panamax vessels operating, with capacities reaching 6,500 TEU, and even higher capacity vessels (8,500 TEU) are being built. Larger vessels are more economic to operate and lower the costs of sea freight. Container rates to Durban are about 60% of those to Mombasa, because the scale of the traffic allows 4,500 TEU vessels to be utilized, compared with 1,500 TEU vessels to Mombasa. To allow third generation container vessels, berth draft would have to be increased to about 11.5 or 12 meters. Panamax vessels (3,000 - 4,500 TEU) would require drafts of 13.5 meters, widening and straightening of the inner channel, and major modifications to crane foundations to accommodate the larger and heavier cranes. The issue is whether it is necessary for Mombasa to provide the facilities necessary to accommodate the larger container vessels. There is a world trend toward using ultra (Port-Panamax) vessels on the main trans-global routes, handling very large numbers of containers at a very few hub ports. It is perhaps too early to understand clearly the implications for ports, such as Mombasa, which are outside these trans- global routes: * Possibly, with the increasing use of ultra vessels on the main routes, more 3d and 4h generation vessels will be cascaded to the other routes. Shipping lines may wish, therefore, to operate larger vessels to East Africa, though the flow of traffic is limited. However, other ports along the East Africa conference route would also have to make major investments to allow 4h generation vessels (most of the other ports have a berth depth of 12 meters) * Altematively, with the increasing use of ultra vessels to regional hub ports, Mombasa's traffic will be increasingly handled by feeder services, operating to hub ports in the Gulf or Middle East, rather than direct services. Such a trend would reduce the average vessel size operating to Mombasa; feeder vessels with a capacity of 750 TEU could be accomrnodated at any berth in the port (though investment would be required to provide the necessary infrastructure for cranes, etc.) Certainly, no shipping line presently operating to Mombasa has plans to bring larger vessels into the trade, indeed some are considering down-sizing their existing fleets because of the low return load factors from East Africa. 2.5 Port Finances and Staffing KPA's finances have recovered in 2000 and 2001, following three years of extremely poor performance with the port recording an actual loss of Ksh. one billion in 1999. The improvement has been brought about by some increase in revenue but also by a significant reduction in operating expenditures. 101 In the 1990s, total revenue fluctuated between the equivalent of US$ 100 - 120 million. The very dramatic increase in operating surplus in 1993 and 1994 was the result of a reduction in the US$ value of KPA expenditures following a major depreciation in the Ksh. Port revenues rose substantially in Ksh terms as the tariff is denominated in US$. Except for the period 1997 - 1999, KPA has not faced the financial problems of almost all other parastatals in Kenyas5. KPA has made money - hardly surprising as it has an effective monopoly over 85% of its traffic. Container handling rates are high by international standards, but not very different to its perceived comparator, Dar es Salaam. The relatively healthy financial position of KPA, especially in the early 1990s, has had a number of adverse impacts: * KPA has been expected to finance uneconomic investments, such as the ICDs at Eldoret and Kisumu; * KPA has been expected to support parastatals in financial difficulties, such as the Kenya Ferry Services and the Kenya National Shipping Line * KPA has had to deposit funds in financial institutions which subsequently collapsed Perhaps the most serious problem, however, was that GOK gave little attention to KPA's performance, focusing much more on the loss-making parastatals. As long as KPA did not require financial assistance, GOK was largely unconcerned irrespective of KPA's service levels. KPA made a substantial operating surplus, in 2001, over Ksh 2 billion (plus a depreciation provision of Ksh 1.2 billion) but it has large contingent liabilities, for back taxes and penalties, a very large pension liability, and the need to invest substantially to improve its level of services. It is unlikely that KPA will be able to fund all its investment needs from internally generated funds. Staff costs account for about 50% of KPA's total operating expenditure. This is very much the same proportion as in 1986, when KPA had almost twice as many employees. Staff levels have fallen from about 10,900 in 1990 to 6,160 in 2001, entirely as a result of natural attrition. Salaries and wages per employee have, however, risen dramatically in both Ksh and US$ terms. In 1986, the average wage cost was the equivalent of US$2,840/employee, this had risen to US$6,850 in 2001. Increasing real wages may have contributed, but there was also a more than proportional reduction in low wage dock labor. A significant reduction in manual labor was necessary, with the decline in break-bulk traffic, but port users are now complaining about labor shortages on the conventional berths and their inability to work all the hatches of a vessel. KPA may be turning into a top-heavy organization with an oversized management and an inadequate labor force. 51 KPA's financial position was also artificially inflated as it was not paying taxes on its income, nor servicing the loans for the French container handling equipment. Subsequently, GOK demanded massive payments from KPA for back-taxes and debt servicing. 102 3. ACTIONS 3.1 KPA Actions Port users report a significant improvement in the performance of KPA during 2000 and 2001, especially with regard to container operations. These positive reports are somewhat offset by KPA's neglect of the conventional berths, in terms of both labor and equipment. KPA management is introducing major changes in the way it manages the port, especially for containers. (a) COSMOS: A computerized container tracking system is finally being introduced in Mombasa which should increase efficiency both directly and indirectly. KPA has decided to install the COSMOS system which provides a comprehensive system for container management. It is hoped that the system will be introduced in early 2003, and work is being undertaken to relocate the terminal offices to a new site and connect the terminal and port gates with the central computer system. COSMOS comes in five modules: * Traffic and operations (the Yard module) * Central database (Container Tracking Control System module) * Finance (COREBUS module) * Ship planning * EDI More effective tracking and monitoring of containers within the terminal would itself bring substantial benefits but the EDI module, if successfully implemented, offers the opportunity for a fundamental shift in the port system. The module governs the transfer of information between the terminal, shipping agents, forwarding agents, Customs and other authorities. It allows the opportunity for creating essentially a paperless system, which could remove many of the procedures and personal interfaces which create the potential for delay and rent-seeking. It could provide the basis, for example, for pre-clearance of containers. To allow all govermment, port and private enterprise systems to communicate through EDI, it will be necessary to create a Community Based System with a common platform and shared protocols. Discussions are taking place between both the public and private sector interests, but progress toward an agreement is slow, perhaps reflecting the local lack of specialized technical expertise. To be really effective, it may also require Customs to introduce computerized documentation for clearance and customs officers to operate the system. If Mombasa is really to move to international standards, such a system is essential. (b) Enterprise Resource Planning: KPA management is also introducing a comprehensive computerized management system, developed by SAP. The system has only recently been introduced and it is too soon to judge its likely success. However, the same type of system is operating in ports elsewhere and runs the World Bank. The system should provide management with much better information and the basis for modem management. It may also allow substantial cost saving, but again its success will be dependant on staff being both well trained and motivated to work with the system. The introduction of COSMOS and ERP demonstrates that senior KPA management recognize that fundamental changes must take place, if Mombasa is to provide the level of service now accepted as international standards. Unfortunately, KPA management is not operating in a vacuum, and the efficiency of a port depends not only on the port facilities and management but critically also on the other members of the port-related community. Major improvements should be within the domain of KPA management, 103 but international standards may not be achievable until such standards are adopted in customs and security. It is, for example, difficult to envisage international service standards being achieved when most import containers are physically verified, and verified by four separate security organizations. 3.2 KPA Plans (a) Commercialization and Private Sector Participation: the restructuring of the KPA has been discussed for several years and a major restructuring study was undertaken52 in the mid-1990s. The demise of the "strategic parastatal" which must be both owned and operated by the public sector has widened considerably the boundaries for structural change. KPA management has indicated that it sees the need for a basic change in the role of the KPA, from being a full service provider to being a 'landlord port authority". KPA would manage and plan the port infrastructure, but ship and cargo services would be provided by others, primarily the private sector. The broad approach is endorsed in the Poverty Reduction Strategy Paper (PRSP) which indicates the privatization of some KPA activities and specifically mentions the concessioning of the container terminal. Introducing private sector management and investment into the port sector must be the most likely way forward, public sector management has clearly failed to provide the consistently high quality of service needed for the modem economy. "Privatization" of Mombasa Port has, however, raised opposition from the unions and both local and national politicians, though it is not always clear whether it is opposition to privatization or the way that the privatization process might be undertaken. It is also not clear whether those opposing privatization recognize that the port is not being sold; the basic infrastructure remains within the public sector, only its use is leased and a concession granted to provide services. There are some who suggest that the private sector is not required and point to Singapore to show that management by the public sector can be a world leader, but the PSA is very much the exception and is itself now facing problems. No plans have been finalized for the introduction of private sector participation but KPA has made interesting proposals: (i) Container terminal: to be concessioned. It is proposed that the shareholding in the concession operating company would be split between an international port operator, the KPA, and local private and institutional investors. Such an approach has certainly been adopted in some other port concessions, such as the port of Salalah. But, the experience of the management contract with Felixstowe may make international port operators reluctant to accept minority participation, particularly in the early years of the concession. It is also hoped that a second container terminal could be developed, perhaps under a BOT arrangement to provide competition. (ii) Conventional berths: The conventional berths have lost much of their traffic to containers but some traffic remains. KPA proposes that private stevedoring companies be re-introduced, with cranes to be provided either by the KPA or a joint company owned by the KPA and the stevedoring companies. The shipping industry is very supportive of the idea of allowing private stevedoring companies, but still have concens regarding KPA's neglect of the berths and equipment. A separate concession for some of the conventional berths might be an alternative approach, but (a) it is unlikely that such a concession would be very profitable, and thus would not attract much interest; (b) a concession would introduce another monopoly into the sector, private stevedoring provides an element of competition; and (c) the private stevedoring approach would be more open to local enterprise participation. 52 Not entirely with the full cooperation of KPA's then senior management, particularly with respect to the container terminal. 104 (iii) Bulk handling these are already effectively operated under a landlord arrangement and no major changes would be required. KPA could license additional bulk handling operators, and provide operating space, as the conventional berths would not be concessioned to an operator but would be a conmmon user facility for stevedoring companies. (iv) Oil terminals: these would be operated by a joint venture of the oil companies and the KPA. Oil industry participation in the management and financing of the oil terminals seems critically necessary, given its perception of the deficiencies in the present arrangements. (v) Marine craft: tugs and pilot boats and the repair facilities would be operated either by a joint venture between the KPA and the private sector or under a management contract. Pilots would, however, remain under the KPA. KPA management have thought out a comprehensive approach to the introduction of private sector participation. One common feature is that KPA remains as either an active participant (provision of equipment on the conventional berths) or financial participant in all the activities. KPA would thus be involved as the landlord and the tenant; this could obviously raise the potential for conflicts of interest. Certainly, the arrangement has been adopted in some ports, Salalah was mentioned previously and, in Maputo port, the CFM has a financial shareholding in all the terminal concessions, as well as in the master concession53. Other countries have adopted a strict separation between the operators and the landlord authority to avoid conflicts of interest and to allow the landlord authority to act more as an independent regulator or arbiter between port users and service providers. This is the approach most commonly recommended by port experts. The potential shortcomings of the proposed KPA approach are most obvious with respect to the conventional berths. The private stevedoring companies will be dependent on the KPA for the provision of the shore cranes, and perhaps other cargo handling equipment. The efficiency of the conventional cargo handling operation, and the success of the stevedoring companies, will thus remain dependent on the KPA and its equipment. Unfortunately, the KPA has a dismal history with equipment maintenance, and there is little to suggest that this will improve. An altemative might be the sale/lease of the basic equipment to a company jointly owned by the licensed stevedoring companies. Such an arrangement would much more closely align the interests and incentives of the equipment provider with the cargo handling companies. The KPA plan sees no role for an independent regulator in the sector and assumes that the KPA would undertake this role. If the KPA is financially involved with service providers, it clearly cannot also perform the role of an independent economic and safety regulator, the conflicts of interest would be too great. Even if the KPA has no active or financial participation in the service providers, an independent regulator may well be desirable as usually some part of the landlord authority's revenue is derived as royalty on the operators revenue - both the landlord and operator thus have a common interest to increase revenue at the expense of the user. Some form of independent regulator is thus desirable, especially for those service providers who will enjoy an effective monopoly. (i) Investment: the KPA has plans for the major expansion of container capacity, recognizing the limitations of the present facilities. Expansion of the existing termninal is being discussed, together with 53 The master concession is almost the equivalent of the complete privatization of the port during the concession period. The master concessionaire acts as the landlord authority, collecting the terminal concession payments, maintaining the physical infrastructure of the port and paying a single concession fee to the goverrunent. In Maputo, the master concessionaire is also responsible for operating the general cargo berths. It is a new arrangement, frowned upon by conventional port experts, but with some attractions, in principle at least. 105 rather longer range plans for the construction of entirely new facilities on the Likoni side of the creek at Dondo Kundu: (i) Kipevu Terminal: a feasibility study was undertaken in 2000il to consider the most appropriate expansion of the existing terminal. The consultants considered three alternatives: * Low cost/capacity: the conversion of berths 13 - 14 into fully equipped container berths, providing Mombasa with a total of five container berths. Total estimated initial capital costs, US$107 million. * Medium cost/capacity: the conversion of berths 13 and 14, and the construction of the missing berth 1554. This would provide Mombasa with a six container berths. Total estimated initial capital costs, US$140 million * High cost/capacity: the construction of a straight six berth terminal between Berths 13 to 18. This solution would require the reclamation of land, providing additional space in the terminal, and would increase the berth depth, thus allowing Mombasa to service larger container vessels than can presently be handed. Total estimated initial financial costs, US$ 171 million. According to the consultants, these investments would increase the capacity of the terminal to between 375,OOOTEU (low cost + 10 day dwell time) to 725,OOOTEU (high cost + 7 day dwell time), depending on the alternative and the assumption made regarding the average container dwell time. The investments would substantially increase the present level of effective capacity and, in the case of the high capacity investment, provide for Mombasa's very long-term needs. One of the problems in determining the appropriate solution is that implementing one alternative now may pre-empt the adoption of another alternative later. It might be possible to extend the low cost option by the subsequent construction of berth 15, but it would not be feasible to move from either the low/medium capacity alternatives to the high capacity solution. Technically, it would be possible (although the investments in the low/medium alternative would be wasted) but operationally it is not feasible: 50% of the berths would be closed during the construction and the traffic could not be handled. KPA management has a preference for the high cost/high capacity solution. It provides for very substantial additional capacity, and would allow the port to handle larger vessels. The straight berth line would give the most efficient use of the gantry cranes which could be moved throughout the terminal(s). However, the capacity would be greatly in excess of present demand and the terminal(s) would initially be working at well below 50% of capacity. (ii) Expanded container stacking yards: container storage is almost as much of a constraint on the present capacity of the terminal as the ship-to-shore berth capacity. It is understood that KPA management has plans during the present financial year to increase substantially storage capacity. It is proposed to demolish the transit shed behind berth 14 and resurface the area for 833 additional ground slots. It is also proposed to demolish the existing container termiinal building and move terminal management to offices in a disused workshop facility which is being rehabilitated. The area around the existing terminal building will be paved to provide an additional 1,167 ground slots. Together, an additional 2,000 ground slots would be provided which would have the effect of almost doubling normal storage capacity (assuming an average of 2.5 high stacking) and increasing total storage capacity by some 65%, substantially easing the present constraints.. 54 Berth 15 could be costly to construct because of the poor geological conditions, the reason why it was not constructed initially. 106 (iii) Dondo Kundu: the development of port facilities on a green-field site, located on the Likoni side of the creek, has been discussed for many years. Unfortunately the costs of developing the facilities would be extremely high as major basic infrastructure would have to be constructed, in additional to the port facilities. The site would require investment in a road infrastructure and the construction of a long bridge and/causeway to connect with Kenya's main road network55. Unless there is the establishment of other major employment, income and traffic generating activities on the Likoni side of the creek, it is very difficult to envisage that Dondo Kundu will ever become more than a long term plan. The port facilities would need to form part of a much larger development to include an export processing zone and other industrial and processing facilities. Certainly, there is no imperative to develop the site simply for container capacity as there is sufficient capacity potential in the present port to handle likely traffic flows for the medium and probably long term. (c) Financing: increasing the effective capacity of Mombasa to handle container traffic will probably be expensive and unless KPA can develop a very low cost solution, some form of external financing will be necessary. There were prospects of a large donor loan with low interest and long repayment terms for the construction of additional terminal facilities. The loan would have been sufficient to finance the high capacity alternative for the container terminal, though whether this alternative for expansion is the most desirable remains an open issue. The loan would have effectively been almost a turnkey operation with both the civil works and handling equipment being provided by the donor. While potentially attractive, there are disadvantages to such an approach, especially with regard to the provision of the container handling equipment. An alternative approach might be to follow the BOT approach and look for private sector finance for the construction and then management of the new terminal. The concepts of: (a) private sector investment sounds very attractive; and (b) two competing terminals sounds very attractive. But, it is very difficult to envisage that the two concepts are compatible in the present Mombasa context. Concessioning the existing terminal would be very attractive to private terminal operators, and the concessionaire might well be willing to invest in a new terminal, or expand the existing termiinal. However, the prospects for a second operator of investing US$100+ million, to compete with an existing concessionaire, would be extremely unattractive. Attracting a competing operator to construct a terminal seems implausible, given the level of total container traffic. 55 Road connections to improve access to the South Coast have been studied for years, but no economic solution has yet been determined. A causeway/bridge to Dondo Kundu would not solve the present traffic congestion problems at the Likoni ferry as most the traffic has an origin/destination within Mombasa Island and using the causeway would involve too long a diversion. 107 4. DEVELOPMENT OF THIE CONTAINER TERMINAL The main issues facing the GOK and KPA revolve around the container terminal and its management, expansion and financing. KPA's proposals for the development of the other port 'business units' are reasonable in concept, though they needed to be refined in detail, and major new investments are not expected to be required although enhanced maintenance for the conventional berths and rehabilitation of the oil terminals is necessary. A major restructuring study of KPA was undertaken in the mid-1990s and government has adopted the principle of concessioning container handling. A major investment study has been undertaken, and alternatives developed for the expansion of the container terminal. However, in view of the crucial importance of container traffic to the development of the economy, and the likely cost of major new infrastructure, it is apparent that a further independent review would be useful to clarify both the short and longer term issues facing the container sector. 4.1 Objectives for Private Sector Participation Raising the level of port services to international standards should clearly be the overriding objective, in order that the port facilitates rather than hinders the growth and development of the economy. It may well not be the only possible objective for promoting private sector participation in Mombasa. Improved service to users is sometimes not the dominant motivation for privatization. * The port is a major public investment and government may want to maximize its the financial benefits from privatization, especially in a situation of severe fiscal constraints. * Government may also want to minimize the need for further public investment in the sector by providing attractive financial conditions for private sector investment. Such objectives might have important implications with regard to the financial benefits derived by port users from private sector participation. The government needs to take a view with regard to how they would like the potential benefits to be distributed and then develop the modalities to achieve such distribution. Benefit distribution could be influenced by, for example, the bidding terms for the container terminal concession, or by whether a second terminal is to be encouraged to promote cornpetition. (a) Bidding terms: Concessions are usually bid on the basis of maximum handling charges, annual lease payments, and royalty payments on either traffic or revenue. In order to have a transparent and determinant evaluation, the bidding documents normally fix two of the potential variables and award on the basis of the bids on the third variable. * For the Dar es Salaam container terminal the maximum handling rate was fixed at the existing THA rates, the royalty paymentTEU was fixed and bidders bid on the annual lease payment. This approach maximizes the financial value to the government. Other concessions have adopted an entirely different approach * The required lease and royalty payments have been fixed in the bidding documents, at the levels required for port costs, and the concession is awarded to the bidder who bids the lowest maximum handling rate. This approach minimizes the cost to the port user. GOK must decide how it wishes to distribute the financial benefits between the public sector (KPA and Government) and the port user. The Government would also need to decide whether it wants the benefits of increased efficiency to be shared with the port user, by having the maximum handling tariff 108 progressively reduce over time. There would appear to be great opportunities to increase productivity and reduce costs as the container terminal should need many less staff than the 1500 presently employed56. (b) Competition: the idea of two competing container terminals has attraction for the port user, though for cargo handling (rather than ship handling) the competition might be expressed more through the level of service than through prices. Port users would gain from competition, at least until the terminals approached capacity57. Profits would, however, be more uncertain for the concessionaires, than with a single terminal, and consequently the concessionaires might discount the level of fixed payments that they might offer for the concession. The concessionaires would be more reluctant to invest in the terminals, especially investment in fixed assets. (c) Private Sector Investment: Mombasa is a small container terminal, on the international scale. A single operator might well be prepared to invest in additional berth capacity, when the present terminal is congested. But, it would make no financial sense for a second operator to invest in fixed infrastructure when the present terminal is congested. Such infrastructure would have to be provided by the public sector, which the second concessionaire would then operate. Bids for the initial concession would also be lower, if there was a possibility of competition in the relatively near future. The government is thus faced with the choice: A. A single container operator, and expect that the operator will invest in any necessary expansion in facilities; or B. Competing container operators, and accept rather lower concession fees and be prepared to fund the necessary expansion in berth infrastructure. Inevitably there would be initial excess capacity as the minimum scale of a second terminal would be two berths. The public sector benefits financially under A (and the private sector might well provide the necessary facilities at lower cost than the public sector), the port user is more likely to benefit under B. The large capacity/cost alternative proposed by the recent study would be ideally structured for two competing, privately operated terminals but at high cost and substantial excess capacity for many years. 4.2 Future Shipping Patterns The future operating patterns of the shipping lines and the future structure of the shipping industry is critical to an efficient decision on future investment in container facilities at Mombasa. East Africa is off the major trade routes and does not have the potential to be a major hub. Even if either Mombasa or Dar es Salaam become a mini-hub for East Africa, the total level of traffic would be too small to justify a direct service with the latest generation vessels. The real issues are whether (a) East Africa will continue to be served by direct services from Europe and the Far East, or by feeder services from the Middle East and perhaps South Africa; (b) there will be further consolidation in the shipping industry, which would make the hub and spoke operating pattern more likely; and (c) feeder services will continue to be operated by relatively small container vessels, offering a frequent service, or whether shipping lines will cascade larger vessels offering a more infrequent service. Depending on how the shipping trades develop, Mombasa might need to provide berths for small feeder vessels, which could be located almost anywhere in the port, or provide berths with deeper draft than presently exists. Mombasa needs to respond to the shipping sector, the level of traffic is too small for the 56 The conventional standard for terminals used to be one employee/I OOOTEU, but efficient modem terminals have much lower staffing ratios, Hong Kong employs about 0.4 staff/000 TEU. 57 Two 3 berth terminals would have rather less effective capacity than one 6 berth terminal. 109 shipping sector to respond to Mombasa. In view of the high cost of making the wrong decision, GOK and KPA might well wish to delay any decision until the future direction of shipping patterns is more clearly established. The proposed high capacity/high cost alternative for the container terminal would have both the capacity and capability of handling traffic carried in feeder vessels, but it would not be the appropriate investment to make in those circumstances. 4.3. Economic Needs and Development Constraints Kenya requires an international standard container terminal to link its economy with the global market and unlock the country's undoubted human resources for exported oriented production. Without this connection, the economy will remain dependent upon tea and coffee, flowers and horticulture, basic manufacturing for the regional market and air freight dependent production. This seems an unlikely combination to generate the employment and income necessary to accommodate the very large numbers of young people coming into the job market. Kenya needs this international standard port as soon as possible. The country may well be losing opportunities to countries which are already well connected to the global economy. The construction of a major new container terminal may be attractive, for both legitimate and perhaps less legitimate reasons, but it would take five years for the detailed design, financing and construction to be completed. It is also quite possible that a decision taken today may appear inappropriate in a few years, if shipping patterns change. Kenya has also relatively poor access to capital and has enormous capital needs for improving the country's basic infrastructure, especially water, power and roads. The re-establishment of normal relations with the IFIs will undoubtedly relax the present constraints, but it appears that the Govenment would like most of the expected increase in official aid through general budgetary support rather than large public sector investment projects. Moreover, large public sector investment projects may well not be the way to achieve cost-effective improvements in the port sector. Private sector investment for the development of additional berth capacity may be an option, but the flows of Direct Fixed Investment (DFI) to developing countries have been declining and Sub-Sahara Africa has never managed to attract substantial DFI. Kenya's economic and political risk rating is much less than AAA, and the financial potential of the Mombasa container terminal has not been demonstrated. Mombasa has undoubtedly major attractions for port operators, given the present level of container traffic and its large potential market in East Africa. However, it is by no means clear that the private sector would come and immediately make major new physical investments in the port without either substantial co-financing by the public sector (as in several Gulf/Middle East Ports) or substantial government guarantees of financial returns (as in Aden, for example) It might make sense, therefore, for the Government to investigate whether decisions on substantial investment in physical infrastructure can be delayed by the implementation of appropriate policies and/or investments in equipment. Eventually, additional physical infrastructure will be required to accommodate the growth in traffic, but it might be possible to postpone major investment in physical infrastructure into the medium tenm 110 4.4 The Capacity of the Existing Terminal In most respects, according to the conventional benchmarks, Mombasa should not be facing capacity problems. A 600 meter terminal should provide berths capacity of 600,000TEU, with the right equipment and the right management58. The present container stacking yards may be a more serious constraint but it appears, from the information available, that these constraints could be relaxed: * The port already plans to expand substantially the area of the container stacking yards; with an average 10 day dwell time for containers, the planned expansion would provide a capacity of 450,000 TEU/year. Such capacity would accommodate a 5% growth rate for the next ten years * The capacity of the container stacking areas are directly dependent upon the average dwell time of containers in the yards. Most estimates for the capacity of container terminal at Mombasa are based on a container dwell time of 10 days (and that may be optimistic). Irrespective of the efficiency of the port in ship-to-shore movement, an average dwell time of 10 days cannot be considered as acceptable in the global economy. The Kenyan port community (KPA, Customs, security and users) should be targeting an average dwell time of <5 days. Even reducing the average dwell time to 7 days would increase the capacity of the stacking yards to >600,000 TEU, compatible with the potential berth capacity of the present 600 meter terminal. Such capacity would accommodate a 5% growth for 15 years, or alternatively a 7% growth for over 10 years. Potentially, therefore, the present terminal could handle likely traffic for a number of years, before major new investment in physical infrastructure is required. This time window would provide two very important benefits: (a) it would allow a much clearer view of Mombasa's role in the future structure of the international shipping - whether the port needs to accommodate larger vessels or more feeder services; and (b) it would demonstrate the financial potential of the port and thus facilitate future private sector investment in additional physical infrastructure59. 4.5 Investment in Equipment While the present infrastructure has the potential to handle traffic for several years, with little investment, the equipment presently operated by the KPA cannot provide the levels of service required to generate such capacity. The present equipment has problems providing the service for half the potential capacity of the termiinal. Introducing private management and equipment maintenance might well improve productivity and equipment availability to some extent. But, there is general agreement in the port sector that the cranes were never very good, even when new, and they are now old and need replacement. The major rehabilitation of the SSGs might be possible (retaining the basic steel structure but replacing the motors, controls, etc), and the recent overhaul of one of the cranes by ABB should provide an indication of the potential for rehabilitation, but the previous efforts on the RTGs were less than fully successful. Replacement rather than rehabilitation appears to be the most likely approach. In addition,' the number of SSGs, to provide the 600,000TEU capacity, would have to be increased to six. Replacement of the primary container handling equipment, by the public sector, would be very costly in any port. New SSGs , RTGs and similar equipment are very expensive, the consultants for the recent 58 P&O are handling 950,000 TEU through the new 600 meter terminal at Nhava Sheva, Mumbai they assumed 600,000 TEU in their 1998 Business Plan. 59 Dar es Salaam was leased for 10 years rather than concessioned for 20 years, because no major investmnent was foreseen during the period. It is hoped that the performance during the lease period will substantially increase intemational interest in a future long-term concession when major investment will be necessary. ll terminal study estimated an initial equipment investment requirement in excess of US$50 million. The purchase of six SSGs alone would cost well in excess of US$ 30 million. A much cheaper alternative for the port would be the procurement of secondhand equipment. At the present time, the container market is extremely competitive especially in the Gulf and the Middle East. Container handling rates are falling and vessels are getting larger. The ports competing to be regional hubs are being forced to upgrade constantly their handling facilities and deploy the latest and most productive equipment. High capacity regional hub ports may need to be in the forefront of technology, but the secondhand cranes from such a hub port would make an enormous impact upon the productivity of Mombasa. Hub ports may be moving from SSGs with rates 25/moves/hour to 35/moves/hour; but the 'obsolete' 25/moves/hour cranes have more than twice the productivity of those presently operating at Mombasa. The procurement of such secondhand equipment, while financially and economically attractive, is virtually impossible for the public sector. Public sector and donor funded procurement is always predicated on the basis of the purchase of new equipment. It is almost inconceivable that either government or donor-based procurement rules would allow the purchase of secondhand equipment - the choices would potentially be too open to judgment, manipulation and abuse. The market for used equipment is, on the other hand, used frequently and successfully by the private sector. The road freight sector purchases secondhand trucks, the rail sector purchases secondhand locomotives and cascaded track, and the port sector purchases secondhand cranes'. If the private sector were to re-equip the container terminal at Mombasa, it would be achieved at a fraction of the public sector cost. Such cost savings would clearly generate major benefits which would be distributed in some manner among the operator (higher total profits), the government (higher taxes of profits and/or higher concession fees) and users (lower handling charges). 4.6 Private Sector Management Equipment alone is unlikely to achieve the levels of productivity required, it has to be accompanied by both improved management systems and a very different framework of incentives and sanctions for both management and workers. The present KPA management is introducing improved management systems, but it is implausible that, within the parastatal framework, they can introduce the changes in the framework of rewards and penalties which are needed to ensure that the new management systems are complemented by a more motivated management and work force. Some form of private sector management system, with the freedom to introduce private sector motivation is urgently required. This has to be accompanied by the removal of opportunities for obtaining informal rewards for 'priority' services. Within the yards, this must mean that only terminal staff are allowed within the terminal area, rather than the present system which appears to allow access to anyone with any connection with the containers. Containers should be delivered to and collected from the terminal gates. All activity within the terminal area should be strictly the responsibility of the terminal staff and management. Motivation must come from the structures developed by management rather than through ad hoc informal payments by c/f or shipping agents. With the introduction of COSMOS, KPA management will have the potential to make this important change in terminal control, but KPA management is not private sector management. Some MPs, who went on a study tour to the Far East, stressed that private sector management is not necessary, simply look 60 West European ports have purchased shore cranes from the former Soviet block which became redundant with containerization. TICS in Dar es Salaarn was reported as considering the purchase of secondhand RTGs from the Gulf, but was able to buy new RTGs cheaply when an order fell through and the supplier was left with the cranes. 112 at the example of Singapore. However, the entire public sector in Singapore operates on a very different environment to that prevailing in Kenya and many other countries. In Singapore, the Government has, since Independence, paid salaries at least comparable to those in the private sector, and has thus been able to recruit its share of the top managers. This has been accompanied by zero tolerance for corruption and ruthless penalties for those convicted or suspected of corruption. This is not the public sector environment which presently prevails in Kenya. Even, however, within the Port of Singapore, much of the actual operations is undertaken by the private sector, e.g. the movement of containers to/from the stacking yards. There is no reason that Kenyan management and workers cannot turn Mombasa into an international standard port and perform as well as any other port. Dar es Salaam has, in a very short time, demonstrated what can be achieved within a different management structure. The performance of the port of Dar es Salaam used to be notorious but, with the change in terminal management, it is being held up as an example to the Port of Durban: "Sivi Gounden, director general of the public enterprises department, said ........ Durban was lagging behind the likes of Dar es Salaam and Hong Kong, with throughput of 9,300 containers per hectare, versus 21,000 at the Tanzanian port and 67,000 in Hong Kong" (Reuters 28/10/02) Under the right management and with the right equipment, there is no reason why Mombasa should not be able to achieve average handling rates of 600/moves/ship/day. 4.7 A Scenario for the Container Terminal Concessioning now, under the right conditions, should have many more attractions for the Government and the Kenyan economy than investing in new physical infrastructure and then concessioning. * Kenya would obtain a substantially improved connection with the global economy, much sooner (possibly four or five years earlier) and at much lower cost. * The need for major physical investnent would be substantially postponed and the time gained will allow the shipping patterns to become much clearer, and enable the government and the private sector to decide what configuration of terminal facilities will best meet the needs of Kenya and the demands of the shipping trade. * The financial benefits of concessioning during this early period should also reduce the resistance of the private sector to finance long term physical investment in the port, though very probably on an incremental approach (and possibly also with an extension of the concession period). However, this approach will mean essentially that there is likely to be only a single terminal operator at Mombasa, unless the trend is toward feeder vessels which can be handled along the conventional berths, without substantial investment. The approach should be very attractive to the Government but, because of the implied monopoly, it may be rather less attractive to users. Though, in comparison with the present situation, it would still result in very major improvements in service levels61. In view of the longer-term consequences of a monopoly, it would be essential that the right conditions are included in the concession agreement. Defining the right conditions requires skilled transaction advisers, well experienced in drawing up concession agreements as well as extremely knowledgeable about the likely consequences of different concession arrangements on the three principal stakeholders in the concession - the concession operator, the government and the termninal users (both shipping lines and importers/exporters). 61 Users are not complaining about the monopoly operator at Dar es Salaam, very much the reverse. Only the THA have reservations on the arrangement, because of the reduction in control over resources. 113 The concession agreement would clearly need to specify the performance standards required from the terminal operator as well as the penalties for not achieving them. Such standards would need to cover both the average and minimum levels of handling productivity to be achieved. Moreover, given the time period of the concession, and the possible need for future expansion of the terminal, the agreement would also have to specify the average and maximum average vessel waiting time permitted at the terminal as well perhaps as the destination of the penalty (the KPA or the vessel operator). The investment plans proposed in the business plans of concession bidders might be specified in the concession agreement, but experience suggests that this is not very useful. Circumstances change and investment plans have to be adapted. Moreover, the concession operator may find other ways to meet required performance standards than major investment or alternatively through a substitution of one type of investment (equipment, for example) for another (new berths). Crucially, the Government would need to determine, through the bidding terms and concession fees, how the benefits of the concession will be distributed between the KPA, the Government and the port user (whatever bidding arrangement is used, the concession bidders will always ensure that they receive an acceptable expected return). If the Government and potential concessionaire have the same assessment of the risks involved, the bidding terms and configuration of concession payments should affect the distribution of financial benefits between these parties, rather than the distribution between the concessionaire and the Government/users. However, it is quite plausible that the Government and the concessionaire have rather different views regarding the risks, with the concessionaire typically including a heavier risk weighting than the government. Risk sharing, by giving weight to royalty payments rather than fixed lease payments, may thus generate greater benefits to the country. The right conditions might also include: • the distribution of future terminal efficiency gains, using a formula which progressively reduces the stevedoring and handling charges, while retaining a strong financial incentive for the concessionaire to increase termninal efficiency and reduce operating costs; * the shipping and/or wharfage charges62 applicable to investment in new berth capacity, over and above the berths provided by KPA as part of the concession. The concessionaire must be able to generate financial returns on new fixed physical investment; * the arrangements in case of a major change in the value of the tariff denominator in relation to other currencies. Tariffs are presently denominated in US$, however, it is conceivable that the US$ might depreciate/appreciate against other world currencies - the concession agreement may wish to make provision for such windfall gains/losses; * the rules for equal treatment of all potential users, especially if shipping lines and/or freight forwarders, operating at Mombasa, are allowed to bid. Certainly the right conditions must include the respective roles/rights/obligations of the concessionaire, the KPA, and any regulator or arbitrator established for ports or the transport sector more generally. The concession agreement should also establish the specific circumstances under which: * port tariffs and/or concession fees can be renegotiated, * the duration of the concession agreement can be extended * the concession can be termiinated, and the rights and obligations of each party, in such an event. 62 Wharfage is a paymnent by TEU, or percentage of cargo value, charged for the fixed port infrastructure. For existing berths, such payments would normally go to the port authority, but for additional berths a different arrangement would be necessary. 114 * additional operators can be licensed, and the conditions under which additional licensees would operate. A contract cannot provide for all the possible eventualities over the 20/25 year period of a concession, but it should try to be as precise and as comprehensive as possible to reduce the role of the regulator, arbitrator or commercial courts in resolving disputes. But, undoubtedly, foreseen problems will emerge and a dispute resolution procedure will be required and must be outlined in the concession agreement, together with the overall regulatory framework. 115 5. FUTURE ROLE FOR THE KPA The KPA has operated what is essentially a full service port since the mid-1980s, providing all stevedoring and cargo handling (with the exception of some specialized bulk cargos), maintaining the physical infrastructure, providing and maintaining the port handling equipment, tugs and other marine craft, and employing a large management and labor force. The port of Mombasa has, from time to time, generated large surpluses which have provided the KPA with the ability to invest in inland container terminals, as well as the national shipping line and other activities. In addition, the KPA has developed extensive housing estates for its workers in Mombasa and provided health and other services for its staff. In many respects, the KPA has become an example of the classic parastatal enterprise, with all the attendant problems. The KPA has become the core enterprise in Mombasa and thus attracts very considerable local economic and political attention while providing a national service. Restructuring the KPA into a landlord authority will not only alter its role within the port sector, but also its role within the economic life of Mombasa. As a landlord port authority, the roles and responsibilities of the KPA would change fundamentally from a blue-collar provider of services, to what would be essentially white collar contract management. The KPA would: * monitor the implementation of the contracts; * ensure that the provisions regarding service quality and maintenance are followed by the contractors and/or concessionaires * ensure that safety and environmental standards are met; * plan the most efficient and economic use of the port assets available * promote, along with the service providers, the use of the port. Few of the KPA's present direct operational activities with respect to vessels and cargo would remain but its responsibilities toward common services at the port would continue: * the harbormaster's function would necessarily remain with the KPA; * firefighting; * environmental facilities and waste disposal; * security of the ports perimeter and common-user areas; and * the pilotage service (although other arrangements are possible). One major activity that will remain with the KPA would be the maintenance of the basic infrastructure of, at least, some of the present facilities. The sub-structure of the port (the berths) remains the property of the government and KPA would be responsible for its maintenance. Unfortunately, KPA's past record of maintenance has been less than exemplary. Users of the conventional berths complain about their neglect, and the repairs to the container berths were not implemented for several years after these serious problems were identified. The maintenance of the container berths could be included within a concession agreement, but it would be much more difficult to re-assign responsibilities for the maintenance of the berths used by private stevedoring companies and the bulk handling operators. A master concession arrangement would include infrastructure maintenance (as well as the other common services), but such an arrangement is improbable at Mombasa. Thus, either KPA has to maintain its own infrastructure maintenance capacity or contract out the function. 116 The KPA might also have some regulatory functions vis-a-vis the concessionaires/contractors and their customers. Whether the KPA should also have economic regulatory powers regarding user charges would depend upon the overall regulatory framework that the Government adopts. Certainly, many of the concessionaires/contractors would have strong monopolist positions and some safeguard regulatory framework is required to ensure that the monopoly powers are not exploited. However, it is by no means clear that the KPA would be right body to provide such economic regulation. In many similar situations, it has been decided that while contract regulation should reside with the landlord authority, economic regulation should reside with an independent regulatory, outside the port sector. The landlord authority may itself have a common interest with the concessionaires/contractors to maximize revenue. This would certainly be the situation if the KPA was a shareholder/partner in either concessions or contracting companies. Though the overall role of the KPA is reduced and becomes more focused, it will remain crucial to the efficient functioning of the port sector. However, its role with respect to the local economy will be transformed as most of the employment and control over resources will pass from the KPA to several separate service providers. KPA will still collect concession fees and other revenues, but will no longer be a major employer (especially if it were to decide to contract out port maintenance and other common services to the private sector) nor will it be responsible for the large scale procurement of equipment and spare parts. KPA's overall command over resources will shrink substantially, though its overall profits will be much less affected as long as it reduces its costs in line with its reduction in responsibilities. However, experience elsewhere suggests that there may be redistribution of the surplus away from the port authority to the government, through taxes on the profits made by the concessionaires and contractors operating in the port. 117 SECTION 4 CIVIL AVIATION 118 CIVIL AVIATION 1. THE GLOBAL SITUATION 1.1 Cyclical Downturn and September 11 During the past 30 years, worldwide passenger traffic has increased at an annual rate of 6.2%, nearly double the growth in global GDP. Between 1995 and 2000, the world airline industry (IATA members), earned profits of $39 billion and took delivery of more than 4,700 jetliners. But, in the first eight months of 2001, passenger traffic for US carriers rose by only 0.7%, a sharp fall from the average 4% growth over the previous decade. The reduction in growth came despite aggressive pricing as airlines tried to fill seats; profits vanished. The financial position of US airlines fell from net profits of almost 4% of revenue, 1998 - 2000, to losses of >3% of revenue during the first half of 2001. September 11 then caused an unprecedented drop in air travel and airline finances. Worldwide, passenger traffic dropped by more than 4%. The effect of the terrorist attacks was most acute in the United States, where passenger traffic dropped by 6.8% for the full year and earnings fell by approximately 8.5%, prompting the US government to provide $5 billion in compensation and to make available $10 billion in loan guarantees. The slump has already taken its toll. The four major US carriers, American Airlines, Delta Air Lines, United Airlines and US Airways all reported record losses of over US$1 billion in 2001. Swissair, Sabena, Midway and Ansett went bankrupt and other major airlines (including British Airways, Japan Airlines, Philippine Airlines, Scandinavian Airlines System and Varig) recorded substantial losses. Other airlines, like Air New Zealand, were rescued by their governments. 1.2 The US Industry post September11 Passenger traffic started to rise again, in the US, at the end of 2001. But, in March 2002, traffic was still 9% lower than the previous year, despite a 13% reduction in average fares. From 1995 - 2000, operating costs for the US majors stayed roughly constant, but costs rose in 2001, prior to September 11, and then rose even higher to pay for increased security and insurance (probably 2% of revenues). Cutting capacity is critical to cost reduction, but its full impact on profits takes tirne to emerge. After September 11, airlines reduced the number of seats available by 20% in the United States. More than three quarters of these reductions came from flying less, with a reduction in the utilization of aircraft; the remaining cuts coming from putting some aircraft, particularly older models, into storage in the US desert. Both measures allow airlines to lower variable costs, notably fuel and maintenance, but the fixed costs remain. In early 2002, airlines were already bringing back capacity rather than lose market share on key routes. A key element in financial recovery is revenue yield - the price per pass-km flown. Business travel has been the primary source of high yield passengers and critical to the economics of the hub-and-spoke operations of the major airlines. However, over the past several years, business travel has accounted for a reducing share of airline revenues: 35% of US revenues in 1999, but only 23% in 2001, prior to September 11. The business cycle may account for some of the decline but there may also be structural change. In order to reduce the cost of travel, companies have made several fundamental changes that are likely to have long lasting effects on business travel revenues. Large corporations have used their purchasing power to negotiate volume discounts on fares, with price reductions on some routes often as high as 20% to 30%. Furthermore, most corporations are now enforcing the travel restrictions tied to these volume agreements. 119 1.3 The Global Industry post September 11 The global picture is little different to the situation in the US, as a report of the IATA Secretary General speech at the LATA General Meeting on 2nd June, 2002 indicates Speaking during the IATA annual meeting in Shanghai, the association's director general, Pierre Jeanniot, said the world's airlines recorded pre-tax losses of USD12 billion last year on international routes - more than double the USD4.8 billion lost in 1992 after the Gulf War. He said many were still struggling to return to profitability after the September 11 attacks on New York and Washington. "The best we can hope for would be to cut last year's losses in half, and that will require us to adhere to some fairly challenging estimates, " Jeanniot said. JATA has estimated that the airline industry, including domestic routes, lost USD17 billion in 2001, with traffic falling by 5. 7% overall, the worst decline on record. International routes showed a negative US$12 billion of a gross revenue of US$144 billion. The margin was minus 8.3% Jeanniot said the decrease in traffic had now stopped, but yields remained moribund - "meaning the airline is making less money per passenger thanks to low ticket prices and high insurancepremiums, " he said. He claimed the airline industry would have struggled for profit even without the attacks, after adding too much capacity. "This industry was ill-prepared to successfully weather even a fairly mild, regular economic cycle, " Jeanniott said. " However, not all airlines are suffering financial problems. In both Europe and the US, some low-cost, no- frills airlines, operating point-to-point services, often to smaller secondary airports are still making large profits. 1.4 Trends in Air Transport Regulation The privatization of airlines and airports has led to increased commercialization and efficiency. Privatization has also resulted in the establishment of new regulation; it is perceived that airports, having significant natural monopoly power, require both price and access regulation. However, with growing experience of privatization, the tendency is to adopt a rather softer approach to price regulation, which may consist of little more than price monitoring, while strictly enforcing access regimes in order to develop competition. In the domestic sector, where capacity, route and price controls are removed, the essential air transport regulatory environment is based on safety. International airlines are partially regulated by their participation in bi-lateral or multi-lateral air service agreement regimes which typically require schedule approvals and the filing of tariffs. Most governments reserve the right to disapprove tariffs but, in reality, governments cannot enforce tariffs. In most environments, passenger fares and cargo rates vary with seasonality, direction, routing and perceived carrier quality. Where airport regulation exists and where there is regulation of potentially anti-competitive practices, such as alliances and code-share, there is continuing debate about whether there should be an industry specific regulator or a more generic competition regulator. The debate is unresolved, with some states favoring one model, some the other. 120 2. TEIE STRUCTURE OF CIVIL AVIATION IN KENYA 2.1 Background The Kenyan situation is reported in the 2002 Kenya Economic Review as follows: "The September 11, 2001 terrorist attacks on the US severely affected air transport business worldwide. This resulted in cancellation of some travel routes with some major airline companies like Sabena and Swiss Air completely ceasing operations. In the local scene, Kenya Airways terminated the Muscat route and reduced flights to Saudi Arabia but increased those to Europe. The latter was made possible following the airline 's acquisition of 3 new and modern aircrafts. Total passenger traffic through the two main airports of Nairobi and Mombasa declined marginally by 0.7% to 3,819 thousand persons from 3,846 thousand persons in 2000. The total volume of imports handled however, rose by 5.9% from 39,353 thousand tonnes in 2000 to 41,668 thousand tonnes in 2001. Exports volume dropped by 2.8% to 99,154 thousand tonnes from 101,997 thousand tonnes in 2000. Transiting passenger traffic dropped by 9.8% during the same period. Eldoret International Airport recorded an improved cargo traffic performance in 2001, as the traffic more than doubled to 10,110 thousand tonnes in 2001 from 4,968 thousand tonnes in 2000. However, passenger traffic declined by 24.8% to 32 thousandfrom 42 thousandpassengers handled in 2000.63-" It is interesting that the aviation sector is reviewed in 3 paragraphs while the tourism sector, which is dependent on air transport, has a chapter of 12 pages (including tables). The aviation review is perfunctory; it deals only with traffic levels and makes no reference to infrastructure or institutional issues. There is no mention of the financial performance of the Kenya Airports Authority nor of the Governments shareholding in Kenya Airways. 2.2 Existing Insitutional Arrangements The current institutional arrangements for aviation in Kenya, at the time when the present study was undertaken, were not ideal: * there were two Ministries involved in the sector, Transport and Communications and the Office of the President; * the Directorate of Civil Aviation (DCA) was both a service provider and a regulator; and * there were some cross-functional relationships between the DCA (which reports to the Ministry of Transport) and the Kenya Airports Authority (which reports to the Office of the President). 63 Eldoret was not visited but it is understood to have few cargo facilities: this is a surprisingly high figure and equates to a fully laden 747 every 3 days. The figures may be inbound and outbound. The passenger figure appears to verify the airports low usage: 88 pax per day, again presumably a total in and out figure 121 (i) Directorate of Civil Aviation" The Directorate was recognized by the sector stakeholders as having impressive knowledge and understanding of the aviation sector. It was not a parastatal and was dependant upon budgetary funding. Since 1998-99, the Directorate has had liquidity problems and the flow of cash from the government was often a problem. If able to retain its own revenue (when it becomes part of the new CAA), the potential revenue base should be sufficient (with charges for use of over-flight facilities and Air Operator Certificates - the safety license of the airlines). The financial challenges are in the areas of training, adequate remuneration of staff, and investment. The structure of the Directorate was: Director Deputy Director ATS Engineering Operations Flight Administration School of Operations Aviation In addition, there is an ICAO security obligation accountability, which is exercised in association with the KAA. DCA claimed that it had the best infrastructure in Africa, comparable only to South Africa. Full Instrunmental Flight Rules facilities (IFR) and radar are provided at Jomo Kenyatta International Airport (JKIA). Wilson has aerodrome control, with handover on take-off. Mombasa has its own control and radar. (ii) Kenya 's Airports The major airports are managed by an autonomous parastatal (Kenya Airports Authority - KAA), reporting to the Office of the President. Under the new government, KAA is now reporting to the Ministry of Transport and Communications. KAA has often been headed by a senior official from one of the security branches and reporting to the Office of the President was then justified by the security implications of the airports. KAA is, in principal, responsible for all airports in Kenya but in practice it appears to maintain only those with scheduled services, the remainder being left to tourism and other stakeholders to maintain, sometimes without a clear framework of responsibilities. 64 The Directorate has very recently been replaced by the Kenya Civil Aviation Authority, created through the Civil Aviation (Amendment) Act, 2002. A Director has been appointed. There will be an I Ilmember Board which includes, as ex-officio members, the permanent secretaries of Transport and Communications, Finance, and Internal Security and the CEO of the Kenya Tourist Board. 122 While KAA is responsible for the management of the airports, the DCA is the safety regulator and thus necessarily has to play a role. By Kenyan law, the aeronautical authority with responsibility for safety resides with the Minister of Transport and Communications (MTC). There is a national airport security committee and a local one for each major airport. The KAA chairs these security commnittees which include all related institutions (iii) Air Accident Investigation Accidents are investigated by the appointment of "Inspectors" from within the Directorate of Civil Aviation, as there are no full time resources. During the period of the investigation and report phase, the 'inspector' is responsible only to the Minister and not the Directorate. There is recognition that the country would benefit from a Transport Accident Investigation Bureau but there are not the funds available. The development of a full-time regional accident investigation unit, under the auspices of the EAC or COMESA might be the potential way forward. The recent reforms establishing the Kenya Civil Aviation Authority (KCAA) clearly separates accident investigations (attached to MTC) from KCAA. (iv) Licensing The DCA issues 5 categories of personal license in its role as a safety regulator. The number of licenses is an indictor of the size and resources of the aviation sector in any country. Kenya, by this measure, appears to have a vigorous aviation sector. Another indicator is the number of aircraft on the civil register: there are currently 769 aircraft registered in Kenya, which again demonstrates a substantial industry. However, there are no available data on hours flown, which would be a better measure as, in a recession, aircraft may remain registered but not be operated. f . Category .- Descriprion, Kenyan Holders - Non Kenyan Total -Holders Ia Class Air Transport Pilots Scheduled Airline 177 116 293 License Conmnercial Pilots License Charter Aircraft 234 159 393 Private Pilots License No passengers for reward 192 182 374 Flight Engineers In-flight engineering, 2 6 8 systems operation Aircraft Maintenance Licensed to certify aircraft as 198 24 222 Engineers fit to operate 2.3 Possible Changes to the Institutional Arrangements On the basis of successful reforms elsewhere, a number of changes might be considered for the institutional arrangements in Kenya's civil aviation sector: * Separate the Air Traffic Control (also known as Air Traffic Management) from the regulatory roles, and again, split the two regulatory roles (safety and economic regulation). This would eliminate the potential for some serious conflicts of interest * Retain only, but importantly, the policy functions within the Ministry of Transport and Communications. These policy functions would include domestic regulation (if appropriate) and negotiation of air traffic rights and oversight of security provision. 123 * Arrangements to remove the award of capacity on international routes between Kenya Airways and Regional Air (trading as a franchisee of British Airways) from political considerations * Kenya Airports Authority should be brought under the Ministry of Transport & Communications. 124 One possible institutional structure for the sector might be as follows: Proposed Structure Aviation Component Ministry of Transport Commission Mister |Transpc rt Safety 0 0 | < 1 ~~~~~~~In vesti gation j i l ~~~~~~~~~~Authority/Bureau .MinistTy of Transport & Conununica tions Internati4 nal Air Rights Conmmission Aviation Policy l| Aviatil l I l g~~~~~~ Operations | Domestic International Security Policy Air Traffic Air Safety Kenya & Co-ordination Management Regulatory Airport Authority Authority Authority The institutional arrangement may appear complex but all the activities are currently being undertaken except that there is a lack of separation, lack of transparency, lack of co-ordination, and ultimately, lack of accountability. The legislation to restructure the government sector of aviation to create a "Civil Aviation Authority" which will incorporate the DCA has been enacted. The new Authority will have three directorates, as shown below. Director-General of Civil Aviation Directorateof Directorate of Safety Directorate of Air Air Navigation Services Regulation Transport (economic aspects including ASAs, schedules, tariffs and general policy) Most of the civil aviation staff from the Ministry of Transport and Conmnunications will be transferred to the CAA, although a small unit may remain in the Minister's office to deal with traffic rights and accident investigations. 125 The CAA goes someway to meeting the preferred arrangement and hopefully it will soon be fully established; financial autonomy will be achieved by the end of FY2003 and staffing is ongoing under the supervision of ICAO. The decision to bring the Kenya Airport Authority under the Ministry of Transport and Communuications has now been taken. Certainly, while a parastatal, KAA should be responsible to the Minister responsible for transport. If, in the future, KAA was corporatized or individual airports concessioned, then both the Ministry of Transport and the Ministry of Finance would be involved. The Ministry of Transport would provide aviation policy and technical/safety regulation advice and the Ministry of Finance would manage the privatization process and hold any government equity. 2.4 Air Transport Services (i) Domestic Scheduled Sector Domestic air services are provided by competing carriers. The principal ones are Flamingo (Kenya Airways subsidiary) and Air Kenya. Nairobi - Mombasa is, by far, the most dense scheduled domestic operation. It has 7 departures daily in the new schedule by Kenya Airways plus Regional Air (BA) and Air Kenya (turbo-props). The increased demand on this route is partly the result of a reduction in direct charter flights to Mombasa, and tourists now transfer from long haul flights, at Nairobi, for the short haul flight to Mombasa. (ii) Domestic Non-Scheduled Sector There is an extensive non-scheduled general aviation sector, largely serving the tourism but also providing humanitarian services. These aircraft are generally small, but range to over 100 seats (in military configuration). (iii) International Scheduled Air Service Intemational air services are govemed by a series of bi-lateral agreements and the COMESA multi-lateral agreement (although, it was mentioned by several stakeholders that this agreement is not fully effective)65. Kenya has a policy of dual designation and has two carriers: the formerly state owned, Kenya Airways and Regional Air, which operates as a British Airways franchisee. Regional Air is not a separate company but, effectively, a management division of Air Kenya There are a number of foreign scheduled carriers offering direct services to Europe, the Middle East (and by connecting flights to Asia and the US), India and regionally within Africa. There are no direct or code- share services to the United States, as Kenyan airports do not meet the USA Federal Aviation Administration's Category 1 security status. Most scheduled international flights operate from the Jomo Kenyatta Intemational Airport (JKIA) in Nairobi. Since Kenya Airways was privatized, the level of connectivity has increased significantly with increased services to Europe and to West Africa. The present broad structure of the scheduled air service network from JKIA is shown in Fig. 1: 65 For examples, see discussion of Regional Air in Section 3 126 Figure 1: Major International Routes to/from Nairobi LHR AMS CPH BRU F TLV DBX CAI A AIM!BH ABJ ADD ACC S NBO - -- MBA DLA \< ~DZA LAD BLZ / HRE TNR JNB 127 There are international scheduled flights to/from Mombasa, but the network is confined to routes to/from Tanzania, with the exception of a weekly scheduled flight by Condor to Germany, Figure 2.. However, it is reported that Virgin Atlantic may be interested in serving the tourist market to Mombasa. Figure 2: Mombasa: International Scheduled Services FRA 1l*ght per week: 767 - Carrier: Condor MBA Z Carriers: Air Kenya K1M / - \ / Prestige Air / D A/ Air Tanzania (iv) International Non-Scheduled There are occasional non-scheduled flights" to Nairobi but the majority are to Mombasa. Charter operators predominate to Mombasa especially Condor (Germany), the former Balair (Switzerland) and others, including Corsair (France). The number of charter flights is highly dependant on the season and the popularity of Mombasa as a holiday destination. Mombasa's popularity is dependant upon both internal events within Kenya (the ethnic clashes in the tourist areas, for example) and the development of competing facilities in Tanzania and Zanzibar. The recent terrorist attack on the Paradise Hotel and the missile attack on a tourist charter aircraft may obviously have a serious impact on the tourist sector and thus also the level of air operations at Mombasa. (v) Air Freight Services. Air freight services are operated by charter aircraft and by scheduled services. The main scheduled operators are Lufthansa, Martinair, Air France and British Airways (using, in BA colors, an Atlas owned B.747F). Other carriers operate on demand. Air cargo is dealt with further in Section 5. 66 Non-scheduled is the terminology used by government aviation officials. The tourism sector uses 'charter' irrespective of how the services are handled. In the EU, there is no longer any no regulatory distinction between scheduled and charter operations. 128 3. THE AIRLINE INDUSTRY IN KENYA There are two Kenyan airlines operating jet aircraft on scheduled services. Both airlines have associated businesses operating turbo-prop aircraft. 3.1 Kenya Airways (i) Ownership Structure The privatised national carrier has been one of the few business successes in Africa's aviation sector, following years of dismal operational and financial performance in the public sector. The airline has the following ownership structure: * Government of Kenya 23% * KLM (Netherlands) 26% * Other foreign shareholders 4% *Staff 3% * Kenyan institutions and private shareholders 44% The company is controlled by a Board of 12 (2 nominated by Govemment, 2 by KLM, 7 others and an independent chairperson). In spite of his regulatory role in the aviation sector, the Pernanent Secretary of Transport and Communications is one of the government nominees as is the Permanent Secretary of Finance. The company is listed on the Kenyan, Ugandan and Tanzanian stock exchanges. (ii) Recent Financial Performnance Kenya Airways increased its gross revenue by 6.5% (from Ksh 6.1 billion to Ksh 6.5 billion) in the year ending March 2002 and announced a profit of Ksh 868 million. This was achieved while the world air transport industry was in financial freefall, massively impacted by the US recession and September II. The first six months of the financial year saw pretax profits rise by 14%, but this very positive result was eroded (by Ksh 600 million) during the second half of the year. Kenya Airways now faces additional war risks'insurance cover amounting to about US$4/passenger. The airline has replaced its Airbus fleet with B767-300s and a B736-700, the first fleet transition since privatization and this has involved the company in training and other introductory costs of Ksh 446 million. The airline is expanding its fleet, ordering three 777-200 ER aircraft for delivery starting in 2004. A first and final dividend of Ksh 0.60/share is to be paid, a yield of 8.5% on the current share prices. For detailed operating and financial data, see the Annex volume67.. (ii) Policy Approach Kenya Airways states that it favors an 'open skies' approach in Africa, arguing that protectionism will lead to economic failure. It believes that its Nairobi hub will benefit from new carriers and additional frequencies. While it cannot operate every route itself, it will receive transfer traffic. National carriers, like Kenya Airways, may have a lower total market share but may increase their passenger yield when they no longer have to operate low yielding routes. However, it is not clear whether this liberal attitude toward open skies and increased competition also applies to its core profitable routes. 67 The financial position of the company is complicated by a legal case brought by former employees - they were awarded massive compensation by the court. The airline is appealing the decision 129 Kenya Airways also supports the separation of the Directorate of Civil Aviation from the Ministry of Transport & Communications and the creation of the Civil Aviation Authority. It argues that an independent authority may not have incurred the recent air traffic controllers strike and would be better equipped to raise the funds necessary to improve air safety in Kenya. The airline believes that JKIA is in need of basic "redesign". (iii) Schedule & Operations Before privatization, Kenya Airways operated services on a point-to-point schedule. Since privatization, the airline has moved toward a hubbing operation. The effectiveness of hubbing is limited by the poor quality of the departure areas at JKIA, though these are being slowly improved. KQ is fine tuning its hub operation by continuing analysis of origin and destination data in order to produce optimal connections. Some routes have been discontinued, notably Muscat and Jeddah, whilst new flights are being introduced to Heathrow (additional 3 flights per week with a daylight operation in both directions). Additional flights to West Africa (Accra/Abidjan) as well as within East and Southem Africa are also being operated. Kenya Airways believes that it has advantages in its hubbing operations compared to Ethiopian Airlines(ET), previously the primary hub operator. Many of ET's services are multi-stop, whilst Kenya Airways operates non-stop services (except for the Abidjan service which stops at Accra). The airline is finding that while hubbing has increased load factors, revenue yields/passenger are declining. Kenya Airways considered bidding for the Air Tanzania Corporation (ATC) but eventually did not submit a bid. South African Airways is now in the process of acquiring ATC and developing Dar es Salaam (DIA) as a mini-hub for regional flights. This may have some impact on Kenya Airways, but JKIA is a substantially larger, better connected and more established hub airport than DIA. However, DIA is approaching compliance with FAA category 1 security status. (iv) Fleet and Services The present fleet operated by Kenya Airways is shown below. There are also three B777-200 on order for delivery from mid-2004. B.767-300 ER 3 KQ owned B.767-300 ER 2 Operating lease to be returned 2004 B.737-300 3 B.737-200 2 B.737-700 2 3rd by December, 4th June 2003 Kenya Airways flies an extensive network of services within Africa, frequent services to Europe (Amsterdam and London) and almost daily services to Dubai (Middle East) and Mumbai (India). A direct service to Bangkok has recently been announced, to start in June 2003. A full list of Kenya Airways routes and frequencies is shown below. The new aircraft on order will allow increased operations to Asia. Kenya Airways code-shares some flights with KLM, its strategic airline investor, and KLM code-shares with Northwest Airlines. Kenya Airways would also like to code-share with NorthWestern, a KLM alliance partner, and perhaps provide its own direct services to the USA (as does Ethiopian Airlines). However, the lack of the US FAA's category 1 security status at the Kenyan airports prevents code- sharing with a US carrier, let alone direct flights to the US. This is a significant handicap for Kenya Airways, and a handicap which may increase in importance as other countries intensify airline and airport security in the light of terrorist attacks. 130 Kenya Airways Routes & Frequencies To - - Weekly Aircraft En-Route Stop . Frequency Abidjan 4 737 Accra Accra 4 737 Addis AdabaW 4 737 Amsterdam 7 767 7 Code-share on KLM - Blantyre 2 737 Bujumburra 5 737 Cairo 4 737 Khartoum Dar-Es-Salaam 14 737 Douala 2 737 2 767 Dubai6 7 767 Entebbe 28 737 Harare 4 737 Johannesburg 3 737 4 767 Khartoum 4 737 Kigali 7 737 Kinshasa 5 737 Lagos 4 767 Lilongwe 1 737 2 737 Harare 2 737 Lusaka London (Heathrow) 10 767 Lusaka 3 737 2 737 Harare 1 737 Lilongwe Mombasa 48 737 Mumbai 6 767 Seychelles 2 737 Yaounde 2 737 Zanzibar 7 737 - (v) Load Factors & Break Even Kenya Airways calculates both load factors and break even points on a route by route basis as well as on a system wide basis. System wide break even load factor is 57-58%, but may be considerably lower on some routes (depending on competition and the fare levels). The Kinshasa route has a break even close to 30% because of high yields and low costs, whereas on a highly competitive route such as London, the breakeven point is near 70%'°. 68Flights will increase to 6/week in January, 2003 69Fhghts will increase to 10/week in January, 2003 70 Competition comes from British Airways, Emnirates, KLLM, and to a lesser extent Gulf air and Saudi Arabian Airlines. Swissair and SN Brussels airlines are also experienced sixth freedom operators and, as they re-establish themselves in the market after the collapse of their predecessors, they will be more aggressive competitors on Kenya Airways' European routes, putting further pressure on yields and break even points. 131 (vi) Kenya Airways Performance Management The airline is making considerable efforts to both establish performance targets in several areas and to communicate the results widely. The areas currently targeted and monitored are: * Punctuality * Customer Complaints per 2,500 Departures * 15 and Last Bag Deliveries * Pilferage * Maintenance Snags * Cabin Factor * Schedule Integrity * Overall Productivity ASKs Per Employee * Revenue for ASK * Overall Service All of these performance target results are communicated in an easy to follow monthly bulletin using advanced graphics. (vii) Kenya Airways Subsidiaries In addition to Flamingo Airways (see next section), Kenya Airways also has an airline handling subsidiary and an airfreight subsidiary. Kenya Airlines Handling Ltd (KAHL): This air cargo and ramp handling company was created by the merger of KA's cargo and ramp handling activities. Cargo handling, having high fixed costs, was suffering from low throughput as a result of the downturn in the Kenyan economy. The company has equipment to handle both passenger and wide body pure freighter aircraft (upper deck loaders etc). KAHL is the dominant handling company but does not have a monopoly. KAA provides some services including a rather down-market First Class lounge at JKIA. KAHL also operates the new (and over-sized) Nairobi International Air Cargo Centre at JKIA. Kencargo Airlines International: This company was established in 2001. It is 60% owned by Kenya Airways, with KLM and Martinair' each owning 20%. The function of the company is to market the cargo capacity of the three airlines. It has a specific mandate to market the belly capacity on Kenya Airways' short haul services within Africa, feeding cargo for the long-haul wide-body services of the three airlines. 3.2 Flamingo Airways Like most airlines flying long haul, high capacity jet services, Kenya Airways has struggled with the finances of its domestic operations. It is extremely difficult for short-haul turbo-prop services to be financially sustainable, even when they have low overheads, if they have to bear part of the overheads and staff service terms of an international airline. Typically, airlines create wholly or partially owned subsidiaries for domestic operations with their own capital, management and, especially, labor contracts. Kenya Airways has created a new brand: "Flamingo" modeling the operation on the conmmercial practices of European low cost operations with net fares being sold on the intemet, i.e. the fares are non- 71 Martinair is a charter cargo carrier but operates some scheduled cargo flights. It is controlled by KLM 132 commissionable to agents. This has created difficulties with the travel trade: consumer resistance to service charges, and a very slow internet service in Kenya, compared with Gallileo as the booking medium. The airline owns and operates 2 x SAAB 340 aircraft (34 seats) and two Beechcraft. It also leases (wet) a 24 seat GI aircraft which is really an executive or corporate shuttle aircraft: it is fast but has a very high fuel burn as it uses Rolls-Royce Spey turbo-props72. The Embraer Brasilia would seem a more attractive aircraft - lower fuel consumption, six additional seats and available second hand at a low cost (US$2 million). Air Flamingo Schedule Weekly Frequency Aircraft Nairobi - Lamu 7 Gl Nairobi Malindi 7 S.34 Nairobi - Malindi 2 G1 Kisumu 28 S.34 Eldoret 10 G1 2 S.34 Lokichoggio 14 B.19 All Flamingo flights operate ex JKIA at Nairobi and feed into Kenya Airways' international flights, whenever possible. 3.3 AirKenya This privately owned company was formed in 1985, by the merge of two other aviation companies and now provides a network of scheduled services throughout Kenya. The main administration, flight operations and engineering base is at Wilson Airport approximately 4 kms from the center of Nairobi. In May 2000, AirKenya commenced international scheduled services, under the trading name of Regional Air. Regional Air operates from Jomo Kenyatta International Airport, using Boeing 737 aircraft (see Section 3.4). AirKenya has a sister company, Regional Air Services, based in Arusha, Tanzania, which provides scheduled services to Kilimanjaro, Manyara, Seronera, Grumeti as well as charter flights throughout the country. (i) Operations In 2001, the company carried over 160,000 passengers on its scheduled services and in all operations flew over 100 million passenger kilometres. Destinations served by the scheduled services from Nairobi are: * Kenya Coast: Mombasa, Ukunda, Malindi, Lamu and Kiwayu * Kilimanjaro in Tanzania, and * Kenya game parks: Amboseli, Nanyuki, Lewa Downs, Meru, Samburu and the Masai Mara The Coast destinations of Mombasa, Malindi, Lamu and Kiwayu are also linked by the "Coastline Connector" scheduled service, operating from Mombasa. 72 There are very few GIs in scheduled service anywhere in the world. 133 These levels of activity are comparable with airlines in the US that are among the 50 largest regional airlines. The scheduled services are operated under the carrier code QP and are listed in global distributions systems such as Galileo etc. There are no details available of the financial performance of the company. (ii) Fleet and Aircraft Maintenance AirKenya presently operates 12 small/medium passenger aircraft, the details are shown below: De Havilland Dash 7 3 48 2 Shorts 360-300 1 33 2 De Havilland Twin Otter 300 3 18 2 De Havilland Twin Otter 200 1 16 2 Cessna Grand Caravan 3 13 1 Beechcraft Kingair 1 12 1 Except for overhaul of the turbo-prop engines on the larger aircraft, all maintenance is carried out in- house. The work station is approved by the British Civil Aviation Authority (AI/8692/8 1), as well as the local Directorate of Civil Aviation (K/lA/29/82), and the Tanzanian Directorate of Civil Aviation (AI/CAI.93). 3.4 RegionalAir Regional Air is an international airline operating as a British Airways Franchisee. It operates as a division of Air Kenya which holds. the licenses and is the legal entity. Regional Air only maintains management accounts and does not file formal accounts with financial regulators or taxation officials. Regional Air aircraft carry the British Airways name and colors, the services have the BA 2 letter code and the airline participates in the BA frequent flyer program. Regional Air's only domestic operation, Nairobi - Mombasa, is operated as a BA route. Regional Air sells in US$, except in Asmara, where local currency sales are required73. (i) Services Regional Air's present weekly schedule is as follows: Destination Weekly Frequencies Notes Mombasa74 25 Harare/Lusaka 4 (no local rights) Asmara 3 2 via Khartoum; 1 via Djibouti (no local rights Khartoum - Asmara) Djibouti I Khartoum 2 In November 2002, the airline commnenced a Johannesburg service, 5 times weekly, and, subject to adherence to the COMESA Agreement, will add services to Blantyre (Malawi) and Addis Ababa 73 There are also ftmds remittance problems ex Asmara. 74 According to a report in the East African, December 9, 2002 134 (Ethiopia). Regional Air has become a major competitor to Kenya Airways on the Mombasa route, though KA has 50 flights/week. (ii) Aircraft Regional Air operates 2 x 737-200 series aircraft (non-hushkitted) with stage 2 noise compliance. The aircraft are configured 12 Business and 88 Y Class. A third aircraft is being acquired to expand the network and increase frequencies. The 737-200 Chapter 2 aircraft is still accepted in Africa but cannot be used in USA, Europe and Australia (because of the noise level). Its low capital price (approx. US$3 million) gives it a low operating cost, despite higher fuel consumption than newer, noise compliant aircraft. (iii) Load and Break Even Factors Regional Air claims a break even load factor of 47%. Employees are only 60, with 24 pilots. The overheads are extremely low, as is the marketing budget. But, as the airline expands, its overheads will rise. It is unlikely that it will be able to sustain a larger fleet and more extensive network with such a slim management and low planning and other indirect cost activities. (iv) Issues Regional Air management identified five main problems in the present civil aviation sector:- * High aviation charges * Poor management of Kenyan airports * Inadequate security at Kenyan airports. * Poor safety, especially at Mombasa (non-operating equipment) * The non-adherence to the COMESA Agreement by some countries. The Yamoussoukro Decision is also not effectively implemented High aviation charges: Regional Air believes that the air navigation charges (for air traffic management and meteorological services) are high and that little service is provided. Management of airports: AirKenya and Regional Air consider that the major airports are badly managed and that the smaller airports are not managed at all. In many cases, AirKenya negotiates with tourism stakeholders to have basic maintenance undertaken (grading of gravel runways, etc) at the mninor airstrips. Security at airports: There is a need to modemize security equipment, especially for baggage screening, and to ensure baggage reconciliation. Regional Air (as well as several other carriers) conducts tarmac identification of baggage and conducts additional passenger and hand luggage screening, after the official airport screening. Regional Air also considers it is necessary to separate departures and arrivals - they presently intermingle at JKIA. Non-conformance with COMESA Agreement: There appears to be a lack of uniformity and consistency in the application of the agreement. Some provisions are either not applied or applied differently as between airlines. There are apparent differences in the treatment of the 'national airlines' (even when privatized) and private sector airlines. In some cases, multiple designation has not been approved (Ethiopia, for example) or there have been difficiulties is obtaining 5* Freedom Rights (Sudan) or even sufficient 135 frequencies75. In most countries, national concern to protect the market of the national carrier is overriding the COMESA multi-lateral framework (Uganda being a notable exception). 3.5 Foreign Airlines The Galileo booking system list some 17 foreign airlines operating into Nairobi's JKIA. This may rather underestimate the total number of airlines as there some regional operators who operate outside the international booking system. A list of all foreign airlines, recorded in the Galileo system, together with their approved frequencies and aircraft is provided in the Annex. Representatives from three of the main operators were interviewed. (i) Ethiopian Airlines (El) This carrier provides a strong hub operation from Addis Ababa. This operation now links: * North America to Africa * Europe to Africa * Asia to Africa, (BJS, HKG, BKK and BOM) and serves multiple destinations in East, South and West Africa. ET flies 737-200 aircraft daily to Nairobi. The volume of Addis - Nairobi traffic is limited but ET is exploiting its 6e freedom rights for its Addis hub. ET markets itself as a network carrier: the only African carrier to serve Beijing and Bangkok (KA will soon compete), and one of only two airlines serving Hong Kong. ET also offers services to New York and Washington and has a strong network to West Africa. ET also flies 757-F (freighters) to Nairobi, on demand, participating in the cut flower trade to Europe. Issues: ET has enough traffic rights on the route but is concemed about * Security at the airport * Security in Kenya generally, which is making Nairobi a more difficult destination to market. * Airport facilities: the JKIA passenger terminal is being renovated but ET does not regard it as good. * Aggressively competitive pricing by other operators (ii) South African Airways (SAA) Inbound Market: about 30% are high yield business passengers. SAA has a significant portion of the leisure traffic originating in US as it can offer an immediate connection to its US flights and tourists tend to stop in South Africa on their return journey. Outbound Market: Kenyan businessmen are discovering that it is cheaper to do business in South Africa because of low value of Rand: SAA has only about 30% of outbound market to South Africa, as it was represented by a General Sales Agent who tried to sell all the business itself and did not market aggressively through sub-agents. SAA has only 30 - 40% of the market, but it is hopeful of expanding its share by establishing its own marketing operation 75 Rwanda refused Kenya Airways request to increase the frequency of flights from one to two/day. 136 Air Cargo: The 737-800 aircraft has a long narrow body without containerization or mechanical roller floor. Cargo is difficult and requires extra handling. SAA has a joint-venture on a Lufthansa MD-l 1 pure cargo aircraft which operates daily JNB - NBO - FRA and vice-versa. The joint venture operates on the following sectors. JNB SA only SA-LH J-V NBO LH only * FRA On the Frankfurt - Johannesburg leg, Lufthansa and SAA share the costs and revenues. Between Johannesburg - Nairobi, SAA accepts the commercial risk, while on the Nairobi - Frankfurt section, Lufthansa takes the commercial risk. This is an efficient way to utilize widebody capacity in markets which could otherwise not sustain the capacity. SAA believes that there is potential to improve cargo loads. One of the main consignments is wine into Nairobi. SAA cannot access the flower trade via JNB as shippers will not accept the increased trip time, but does carry high security cargo ex NBO. Traffic Rights: The carrier expressed no real difficulties. There is a current entitlement of 14 frequencies for each side, but only 7 are being exercised for passenger service76 Competitive Pressures: Kenya Airways operates the B.737-300 5 days per week but on peak days uses a B.767-300, with about 80 additional seats. This gives an important advantage, especially with higher yielding business traffic. There are supposed to be agreed fares but Kenya Airways is very aggressive and often has "guerilla" fares. (iii) British Airways (BA) Passengers: BA operate passenger services with B747 aircraft, 7 times weekly to London, continuing twice a week to Seychelles and once a week to Lilongwe. However, in the medium term, Regional Air may be asked to fly the legs to the Seychelles and Lilongwe. Despite the downturn in tourism, there has been no reduction in loads, the US tourist market was slowly returning to Kenya. Air Cargo: BA also operates a freighter (wet leased from Atlas but in BA colors) once a week. It also carries bellyhold cargo on its passenger services, but it does not have much capacity because of the high passenger load factors. JKIA: The airport was perceived as having sound design but there is disruption during the renovation". There is a need for upgrading of navigation aids but, overall, BA did not think that it was "too bad". Security had improved since 100% hold baggage screening was introduced. 76 Presumably the daily LH MD-l l operating JNB - NBO - FRA w uses 7 frequencies and so, effectively, capacity is fully utilised by SAA). 137 Regional Air Franchise: It has been slow in its expansion because of the unexpectedly high costs of developing the fleet and network. BA has not yet had the traffic feed that it had expected. " It would appear that the contractors are only working one shift - most airports under such circumstances would work two or three shifts. 138 4. THE KENYA AIRPORTS 4.1 General The management of the Kenya Airports Authority felt unable to cooperate with the study. The rationale, for this, is not clear: perhaps the management perceive the airports as more a part of national security than the transport sector. This section has, therefore, had to be based on secondary sources and information provided by stakeholders. A summary of the main physical characteristics of the major airports is provided in the Annex as well as a list of minor airports. There are also many small airstrips owned and maintained by private tourism operators as well as by the Kenya Wildlife Service. 4.2 Kenya Airports Authority (KAA) The Authority was created by act of parliament in 1991 and is 100% state owned. The KAA is a parastatal falling under the responsibility of the Office of the President. The Authority is regarded as under-funded but, on the basis of the published information it is not possible to confinn or deny this. KAA has not given a breakdown of its revenue sources since 1991. It reports everything under the heading, "Air Passenger Service Charge" and no income against rental of buildings/land/parking area/landing and aircraft parking fees. This total lack of transparency on the income side and absence of cost data makes it impossible to say anything about the Airport Authority as a business. It is thought that the KAA earns about 80% of its revenue from airports (the remaining 20% comes from land rentals). About 50% of the airport revenue comes from aircraft landing and parking fees. Air navigation fees are paid to the DCA and not to KAA, and passenger fees are channeled directly to the Ministry of Finance. The KAA, at JKIA and Moi International Airport, charges every vehicle that accesses the airport zone. The charge applies not only to visitors to the airport but all vehicles entering the airport, whether a private car or vehicles essential to the operation of airport based businesses, including delivery vehicles to export facilities, persons visiting the DCA and KAA and the staff vehicles of airport businesses. This is, if not unique, highly unusual and adds considerable cost to airport based businesses and thus to the consumers of those services. KAA is using its monopoly power to raise revenue from a relatively inelastic source but it may inhibit potential retail or hotel or food and beverage operations from establishing on airport land. It is more normal to charge: (i). car parking; (ii) an access fee for taxis. This would generate revenue without the negative consequences to business development. The huge area of land under the JKIA is slowly being developed through BOT projects. The Nairobi Air Cargo Center was developed on this basis78. The center is operated by the Kenya Airways subsidiary, KARL, but apparently the commercial risk is taken by the BOT developer KAA operates 3 international gateway airports, Jomo Kenyatta International Airport (JKIA) at Nairobi, Moi International Airport at Mombasa and the little used Eldoret. Lokichoggio near the Sudanese border provides a limited gateway to Southern Sudan. The principal domestic airports are Malindi, Kisumu, Lamu and, importantly at Nairobi, its second (domestic only) airport: Wilson. 71 The center was apparently built ahead of demand and the major shippers and forwarders have declined invitations to utilise it. The air freight users complain that there are too many restrictions on the leases at the facility and they prefer to operate outside. 139 4.3 Nairobi Airports Jomo Kenyatta International Airport (JKIA): JKIA is served by a total of 39 scheduled carriers. The 50 daily departure terminal was designed for 2.5 million passengers per annum; it is currently handling 3.0 million. The facility will come under intense pressure from increased passenger throughput with Kenya Airways using JKIA as a hub point - there would be less impact from a more linear network of services. A feasibility study is being undertaken for a fourth terminal, to be built in 3 years time. At the time of a new termninal, it is proposed that an on-airport hotel, to be operated by Marriott, would be constructed.79 Additional aerobridges, some modemization and the installation of CUTE (common user terminal equipment) is planned80. There are many reports from the airlines and other stakeholder regarding the poor state of the infrastructure, lack of security, duty free shops offering little choice, and insufficient passenger accommodation. The basic shortcoming of JKIA is its lack of FAA security class 1 status. Departures and Arrivals are not segregated but co-mingle in the terminal. While it is generally accepted that segregation of departures and arrivals is necessary, it is not an absolute requirement for meeting US requirements. Auckland Airport in New Zealand has only one concourse and arriving and departing traffic can mingle, but there are extra security checks at the departure lounge gates. As JKIA has only one runway, but the use of Wilson airport for most domestic flights and all general aviation allows JKIA to operate on an uncongested basis. Although, because of Kenya Airways wave scheduling for its hubbing strategy, there are peaking issues. Wilson Airport: Kenya's extensive general aviation sector, AirKenya and the police aviation wing are all based at Wilson. A brief inspection of Wilson shows a large number of aircraft and a multiplicity of types, ranging from single engine piston aircraft through to GIs and DC-3s! There were numerous helicopters, as well. Wilson has 2 runways and more aircraft movements than any other airport in Kenya. This is typical of airports which have a large general aviation base. The landside of Wilson is congested with many buildings and poor taxi arrangements. Concems were raised about construction of buildings, including a supermarket close to the runway end. This has adverse safety and security implications. 4.4 MoiInternationalAirport, Mombasa The new Mombasa terminal was built with a Japanese aid loan in the mid-1990s and has a high level of Japanese sourced equipment. The facilities are under-utilized due to a downtum in the tourism arrivals. Whilst aerobridges are available for international operations, domestic operations use tarmac emplanement and deplanement. There are no marked walkways and passengers from BA 737s stream across the tarmac, unsupervised to the baggage hall. This is not a good airport safety or security procedure. 79Until a much larger hub operation is developed, a four star airport hotel seems a doubtful investment. A hotel was built previously at the airport was constructed but never opened as a hotel. 80 C[TE allows any passenger for any flight to be checked in and baggage processed at any terminal counter 140 The international passenger facilities are in the new terminal. The interior of the domestic terminal is not well maintained, has a shabby appearance and the toilets are not well serviced, etc. However, many domestic flights are handled at the international terminal. Internal security at the terminals can be lax (the x-ray and metal detectors do not always operate and the visual inspection of hand luggage is often cursory). Access to the tarmac for photography is (or perhaps was) freely available. KAA reports some medium term maintenance problems on the runway, and complains about the very poor state of the town access to the airport, offering a poor image of Mombasa to the arriving passengers, and triggering security problems (only one access). 4.5 Eldoret The international airport at Eldoret was constructed in the mid-1990s, using domestic financial resources, at a cost of about US$100 million. The airport has no scheduled international services, and very few domestic flights. A few charters have been operated, as have "relief' flights. It is in the west of Kenya and nearer to Sudan than Nairobi, and close to the Uganda and Rwanda borders. There is talk of making the airport and adjacent land into a "free port". It has been mandated that Eldoret will have 25% lower landing fees, free aircraft parking and low termiinal rents than the other major airports. In terms of air freighting, the runway is said to be 500 meters too short, which restricts the gross take-off weight, and there are no cool rooms for fresh produce. However, a relatively high cargo volume is reported for the airport, and press reports of the Kenya Revenue Authority stopping vehicles from Eldoret Airport with undeclared cargo are not infrequent. Whatever economic or financial benefits were expected from the construction and opening of the airport, it is very difficult to envisage that they have materialized 4.6 Assessment of KAA KAA management is not easy to assess and the Authority apparently fails to communicate with its primary clients, let alone its secondary clients. This is the unanimous view of the primary users of the main airports, but such users may not be the most objective observers. However, simple observation of the main passenger terminals and their condition leads to the inevitable conclusion that the facilities have not been well maintained and have been allowed to decay; that both passenger and baggage security is lax (JKIA has a notorious reputation for baggage pilferage and theft); and that the commercial opportunities of a relatively busy terminal have not been exploited to their maximum potential. Tourism is one of Kenya's main revenue earners and employment generators, JKIA is the gateway for most of the tourists, and the first impression of Kenya is a rundown, inefficient, rather disorganized facility. JKIA was constructed in the 1970s and the facilities are showing their age. The runway and taxiways were rehabilitated recently but there has been little effort to update the design of the passenger facilities. The renovation, which is underway, seems largely confined to the main passenger concourse, though this may only be the first stage. One question might be "where has all the money gone" because the passenger and aircraft charges have been at least comparable to those at other international airports in the region8". The channeling of the passenger fares, which probably generate a revenue of around 30 M US$ annually, to the Treasury may also be a reason for the apparent under-funding of the sector, as it usually represents 81 Landing fees for international flights are rather lower than most international airports in East and Southern Africa, landing fees for domestic services are rather higher than the average. The Passenger Services Charge for international passengers is about the highest in the region. 141 more than half of the revenues of most Sub Saharan African airports. Indeed, almost all the freight operators using JKIA complain of much higher fees and charges than at other airports, particularly compared with Johannesburg82. But perhaps this question has little relevance for the future, the past mistakes cannot be rectified, nor the waste reclaimed. JKIA needs substantial investment to remodel the facilities to meet international standards for passenger handling accommodation and modem security (especially to meet FAA security status requirements). There is nothing to suggest, on the basis of past performance, that continuing under the present management and financial model will lead to sustained and financially viable improvements. KAA should provide the same standards of financial and operating reporting as other parastatals; at least, this would improve the transparency of KAA's operations. However, the financial and operating performance of almost all parastatals in Kenya has been disappointing, and the visual evidence suggests that the KAA is certainly no different. Corporatization of the KAA might be one alternative, but this provides no guarantee in the Kenyan environment that there would be substantial improvement. The introduction of private sector finance and management of, at least, the termninal facilities would offer much greater potential. Concessioning of the existing facilities, tied with private sector financing of the necessary remodeling and/or expansion of the terminals would provide a much more certain way of ensuring the quality of airport facility necessary to support Kenya's civil aviation sector. The Government must however take into account the issue of attribution of the passenger fare to the airport sector if the sector is to be profitable for a private or public operator. JKIA could very obviously be concessioned to the private sector, given its present level of traffic. Moi International Airport would be a more marginal proposition, and there seems no way that any concessionaire would be willing to take Eldoret without a large operating subsidy. An alternative, and probably very plausible approach, would be to concession the three airports together, with the profits from JKIA cross-subsidizing Eldoret. Under such an arrangement, the government would retain the ownership of the airports, and all the facilities (both existing and constructed) would revert to the government at the end of the concession period. KAA, or similar, would act as the landlord and would receive concession fees and monitor the performance of the concession contract. The state could still undertake the security functions (as now in the US) or ensure that security standards are enforced (as in the UK). In objective terms, there seems little to lose and much to gain from concessioning the main airports. However, if the capacity expansion of JKIA can be postponed for a few years, there may be more to gain from initially leasing rather than concessioning the terminals. A concession is a long-term arrangement, requiring substantial investment. In the circumstances of Kenya, with a total lack of transparency under the KAA, there is no convincing financial track record on which bidders can base concessioning proposals. It would be expected that bidders would take a very risk adverse approach and submit low bids, and probably require generous concessioning legal terms. A more attractive solution could be, therefore, a shorter term leasing arrangement, with financial results reported under International Accounting Standards, which would establish the financial track record and provide the basis for future bidding on a longer-term concession. 82 Representative of several companies and industry groups have suggested that the costs/charges at JKIA should be the subject of a specific study and benchnarked against competing airport facilities. JKIA has a virtual monopoly but there is no independent regulator of its monopoly power. 142 5. AVIATION AND THE ECONOMY 5.1 Air Cargo Industry The revised IATA global forecast for air freight over the next 4 years shows an average 2% p.a. growth. However, air freight between Europe and Africa is forecast to grow rather faster with an average annual rate of 3.8%, peaking in 2003 with growth at 5%. The air cargo industry is composed of four main operations: * Integrators, such as DHL, UPS or Fedex who provide a complete door-to-door service, within very defined time lines. Often the consignments are very small (letters, computer tapes, mail order goods etc). * General air freight forwarders who use both scheduled passenger "belly hold" capacity and freighter aircraft to move consignments. The consignments may be small, they may be regular or irregular and they may be containerized (called "tin boxes") or palletized. * Specialized carriage of commodities that is in sufficient volume to obtain annual contracts for charter aircraft as well as guaranteed space and rates from scheduled carriers. This is the situation of the fresh flower and produce industry in Kenya. * The fourth broad category is the one-off very large shipment. This business is highly specialized and is dominated by USSR built freighters although 747 F's are sometimes used. The items are often industrial plant but Formula I cars are also moved in this way. Given the poor level of service provided at the Port of Mombasa, air freight has becoming increasingly important to the Kenyan economy, providing the opportunity to diversify the economy and take advantage of the opportunities available for non-traditional exports (including garments under the AGOA program). There has been a major improvement in the availability of air freight capacity for Kenyan exporters in the 1990s. The change is well explained by the Executive Director of the Kenya Flower Council: "Airfreight capacity used to be a big problem in the 70s and 80s when rates were controlled by the state. Then liberalization came, and we became part of the problem... .there was a big boost in capacity by chartering... there used to be aforum to determine the next week's capacity, now it is a commercial decision" While capacity may not be an issue, cost remains a problem especially for the smaller producers unable to contract a regular charter. For such producers, the air freight cost for beans from Nairobi may be US$1.50/kg, compared with US$0.50/kg from Johannesburg or US$0.10/kg from Morocco which is close enough to Europe to use sea or land transport. 5.2 Cut Flowers and Fresh Produce This export industry is totally dependant upon air freight. In a recent World Bank study, it was reported that, in 1998, about 80% of the cut flowers exports were carried on charter flights, with Lufthansa dominating the remainder. Lufthansa has since discontinued its passenger operations but maintains a daily freighter service. Air France (the third highest scheduled carrier) has also withdrawn its passenger service but operates a daily 747F. Ethiopian Airlines, a transshipment carrier (Addis Ababa) was the fourth largest scheduled carrier. 1.43 The logistics of the cut flower trade are more complex than many other industries. Cut flowers are extremely perishable as well as easily damaged. From the moment they are cut, the flowers are dying. The exporters, in order to facilitate the appropriate handling have established joint-ventures with established freight forwarders such as Kuhne and Nagel, or have established their own subsidiaries. About 90% of the export trade is handled by four non-airline freight forwarders. The export of flowers and fresh horticultural products ranks third in Kenya's total export trade and by far the largest export by air. In 1999, the value of fresh flower exports alone was US$110 million. Except for avocados (shipped by sea) all the products are air freighted. Flowers (about 40,000 tonnes) and green beans (about 30,000 tonnes) dominate the trade. Europe is the destination for almost all the flowers and fresh produce. Flowers are mainly destined for the auctions in Arnsterdam, and horticultural products to the UK 3. This report deals only with the aviation aspects of the fresh cut flower export trade and not the whole supply chain". 5.3 Flower Exports: An Example Airflo acts as the export management for 55 flower farms owned or managed by Oserian. They use charters from Martinair, D.A.S. (Uganda), Air France's scheduled freighter and KLM's Combi (passengers and upper deck cargo) but not Lufthansa which is too rigid. The company is reluctant to use passenger carriers because cargo becomes a second priority; when there are high passenger loads and/or bad weather, cargo is off loaded in favor of passenger baggage. Flower exports require first priority The logistic chain for flower exports has been tightened. Post-harvest treatment used to commence 24 hours from cutting in the field. Now flowers are harvested, packed and graded the next day. Airflo provides documentation and cold storage but does not pack, thus reducing VAT costs: the consigning farm pays only air freight costs, which are VAT free. The farms arrange delivery to the airport where Airflo provides cold storage and aircraft palletization and delivery to aircraft. (Retuming trucks are used to move farm supplies from Nairobi to the farms). After arrival in Amsterdam, flowers can be at auction in 3 hours. The flower market is beginning to change and this is affecting the logistic chain. There is now growth in direct supply to retailers, generated by the increase in flower retailing by supermarkets who purchase directly and source custom bouquets. They do not purchase at auction. Flowers for UK supermarkets are delivered to London and for Germany to Frankfurt. There are also wholesalers in the UK and Europe who are bypassing the Dutch auction system. The integrated international logistics system allows an Oserian company to undertake the customs and quarantine clearances and then to truck the flowers from the airport to auction. The company was concerned about the charges for airport access, about commercial terms on leases and about the reliability of the airport power supply. It can never be without power, so it maintains generating capacity sufficient to maintain its cool rooms at correct temperatures and a separate capability, with UPS, for its computer network85. Overall, the airport related difficulties at JKLA that exporters encounter are 83 About 80% of flowers and fresh produce are exported to France, the Netherlands and the UK 84 See "Equatorial Rose: The Kenyan and European Cut Flower Supply Chain", Thoen, Joffee, Dolan and Ba 85 While the costs of maintaining standby power generation is not critical for the success of the export operation, it is surprising as airports are normally provided with very high quality power supplies and standby facilities. Such power reliability is essential for airport and airline safety as well as customer service. 144 relatively minor and have been overcome by the exporters at their own initiative and with their own resources. The remedial actions add marginally to the export cost base, but are not major constraints to Kenya's export capability. Similarly, the vehicle access fee at the airport is not critical, only an irritant. 5.4 Horticultural Exports - An Example East African Growers is a privately owned business and is an example of the vertically integrated operation in the horticultural industry. The company handles every step of the production and logistics chain from growing, transporting by road, packing for wholesale and retail, through to exporting. It operates its receipt, packing, chilling and forwarding activities on land leased from the Airport Authority on a 99 year lease. The company both owns and manages farms. The only product which it purchases are avocados which are trucked to Mombasa and then shipped to Marseilles for European distribution. Avocados, because of their weight and density, cannot support the costs of air freight. The company currently exports about 200 tonnes/week of fresh produce through JKIA86. Some customers want produce picked in the moming, and airlifted out of Nairobi in the evening. The company believes that the products last longer (longer shelf life) if chilled for rather longer and would prefer to ship the next day. However, the company uses both procedures, depending upon customer requirement. The company grows and exports an extensive range of fresh produce: Fine Beans Tomatoes Leeks Extra Fine Beans White Onions Patty Pans Bobby Beans Red Onions Passion fruit Runner Beans Bullet Chilli Water Melon Snow Peaks Sprint Onions Sweet Melon Sugar Snaps Cherry Tomatoes Mango Baby Corn Green Chilli Pineapple Baby Carrots Okra Okra Coloured Capsicum Baby Courgettes Baby Aubergine Various Herbs Courgettes Aubergine On its farms, the company has air conditioned packing houses, cold rooms, cold storage facilities and grading sheds (the range of facilities is farm specific). The company also maintains facilities at JKIA: The EAGA Pack House: Built in 1994 at the present location to provide an efficient product flow within the cold chain under strict hygiene standards. It can be staffed by up to 200 workers/shift and has a minimum of two shifts/day, with a throughput of 15 tonnes of pre-packed produce per shift. Produce is processed under a chain of command that ensures quality checks and traceability. Each processing table has a table head as well as a Quality Controller, with on-line checks carried out by the Supervisors and records confirmed by the Quality Control Manager. Every procedure involving quality control and product flow within the pack house is documented and kept on record. Cold Storage Environmental friendly cool rooms, ammonium based coolants, CFC free, ozone friendly. Total floor area round 400 square metres. 86 The company hopes to increase its export levels to 250 tonnes over the next 12 months 145 Temperature controlled palletizing area: the products are kept in a temperature controlled environment from harvesting to loading of the pallets on the aircraft. The company maintains its own standby generators to ensure security of power supply. The company faces no problems with air freight capacity. It uses a variety of carriers and is well served by the air freight industry. While the company generates enough volume to justify charters it is risk averse and, because of the multiplicity of markets served, it is prefers to negotiate annual prices with the scheduled carriers. It uses national carriers to specific markets as it does not want goods to be trucked long distances at the receiving end, i.e. KLM or Kenya Airways to Amsterdam, Lufthansa to Frankfurt and Air France to Paris. All airfreight is handled by the wholly owned IATA-accredited airfreight company: Maya Freight. 5.5 Tourism Tourism is of major importance to the Kenyan economy, generating foreign exchange as well as extensive direct and indirect employment. Kenya is operating in the increasingly competitive market of the budget tourist as well as the higher income sector. The tourist is market is sensitive to: * Global economic trends: particularly the rate of growth in the economies of the developed countries * Global events: September 11, and terrorism directed at tourist facilities e.g. the Bali disco bombing * Domestic events: the ethnic clashes at the Coast in 1997, and now the bombing of the Paradise Hotel X Security perceptions: Kenya is increasingly perceived as having a low level of security, whether in Nairobi, in the game parks, or at the Coast * Competition: Tanzania is rapidly expanding its tourist facilities for both safari and beach tourism; Mozambique may be the next competitor. Air travel is crucial for the Kenyan tourist trade, but it does not seem a major constraint on its development. Europe is Kenya's primary tourist market, and Kenya is well served with almost 6 direct daily scheduled flights to destinations in Europe, and other airlines providing indirect flights through the Middle East. Charter operators will also provide supply, if there is the demand. There has been, however, some loss of connectivity in the scheduled passenger sector in recent years with the withdrawal of Air France, Alitalia and Lufthansa from the Kenyan route. International air transport frequency and capacity do not appear to be major constraints to the sector. Kenya is developing a much improved domestic scheduled air passenger sector, and its general aviation sector is very strong. In terms of basic transport, Kenya is well served and there is little doubt that the sector could and would accommodate increased tourist traffic. The problems of the tourist sector are to be found in areas other than transport. However, as the aviation industry becomes more global with transnational alliances between the major airlines, it is crucial that the Kenyan industry has the opportunity to participate. This requires that the airport infrastructure (both airside and landside) and security meet accepted international standards. The US market for tourism is potentially very large, and Kenya Airways should have the opportunity to exploit the market through either direct flights or, at least, code sharing with US airlines. Until the public sector raises the security standards at Kenya's airports, this participation is not possible. 146 6. CONCLUSIONS AND RECOMMENDATIONS In most respects, Kenya's civil aviation sector is well established and has provided the basis, over the last 10 years, for the creation of a major new economic sector (flowers and fresh horticulture produce) as well as supporting the tourist sector and the more traditional sectors of the economy. Kenya has three international airports (though this is probably one too many), possibly the most successful airline in Africa, a second private airline which is expanding both domestically and intenationally, and a thriving general aviation sector. The experience of the flower and horticulture sector demonstrates that Kenya can take advantage of the opportunities for export, if the private sector is allowed to develop without government intervention (it is noticeable that the sector has grown dramatically since Govenmment withdraw from trying to control air freight tariffs). If Kenya can successfully operate an intemational logistics chain which must ensure delivery of flowers and fresh produce to intenational markets within 24 hours of harvesting, there is no reason to suppose that other logistics chains can not also be managed87. The sector has thrived in those areas from which the public sector has withdrawn. The problems in the sector, particularly regarding the development and management of the airports, are found in those areas where the public sector remains dominant. There is the general conclusion that the civil aviation sector and the industries dependent on air transport have succeeded despite the public sector rather than having been facilitated by the public sector. It is unfortunate that the civil aviation demonstrates the public sector at its least transparent, making public no information on either its operational or financial performance, beyond the basic traffic flows at some, but not all of the airports. While the public sector cannot withdraw entirely from the sector, as there will always remain residual safety and security oversight functions which must necessarily be taken by the State, the role of the public sector should be reduced substantially. The creation of an autonomous Civil Aviation Authority is a welcome move, but its potential effectiveness will be substantially diminished if it has to face the problems endemic in other Kenyan parastatals - government interference, patronage, inadequate management and fmnancial autonomy, etc. Concessioning the Airports: The KAA is now the major constraint to the functioning of the civil aviation sector. There is unanimity among the stakeholders in the sector that the management of the KAA is both poor and non-commercial. The airports function, but fail to meet their potential. Substantial investment is needed to expand and modernize JKIA and bring security to the standards that are increasingly being demanded by the airlines and partner countries. The introduction of private sector management into the airport sector should be a very high priority for the Government. Corporatizing the KAA into a government owned public limited company, and then gradually selling the shares to the private sector may be one approach, but the conversion of the KAA into a landlord authority and concessioning the airports may be preferable in the Kenyan environment. The concept of a landlord authority and private sector operation is now well accepted in the transport sector (the concept has been accepted for both Kenya Railways and the Kenya Ports Authority) and would enable a much faster access to both private sector expertise and the private sector capital needed to bring JKIA to international standards. 87 The flower and horticulture sectors are not, however, dependent on the just-in-time delivery of inputs. Developing export orientated production using large inputs of foreign sourced materials and components may raise more problems, given the cumbersome nature of customs clearance procedures in Kenya. 147 A study of the management of the civil aviation infrastructure in Kenya is urgently required. It is not simply a question of the management of the main airports but also of the ownership, management and financing of all the airports - international, domestic and secondary airstrips. The present structure is clearly not delivering the quality of infrastructure and management required. There are alternative models, these should be explored. PPIAF financing was successfully obtained for a similar study in Tanzania, there seems little reason why such financing could not be obtained for Kenya. Open Skies: There has been much talk of an open skies policy for Africa, and various declarations have been made in political fora endorsing the approach. However, it is not clear that national policy makers have really accepted the concept as applicable to their own country rather than just to other countries and are prepared to abandon the protection of national carriers. Kenya Airways is now a financially successful airline, but aspects of government protection, in terms of grandfathered traffic rights, still seem to linger. Kenya , with a successful aviation sector, a large tourist market, and a vibrant air-based export sector, has everything to gain from the introduction of an open skies policy (even though Kenya Airways might find the consequences uncomfortable on some routes). The open skies policy should be extended to both passenger and freight sectors, allowing sabotage by foreign airlines, and to allowing charter passenger aircraft to utilize empty belly capacity for the uplift of cargo, if available88. Policymakers in the sector should ignore the presence of Kenya Airways as the 'National Carrier', it is now a privately owned and managed company, able to operate commercially and adjust its level and pattern of services to the market's demands. Rather, the policymakers should be assessing the policies and agreements that will maximize the contribution of the civil aviation sector to the economic and social development of Kenya. This seems inevitably to lead to providing a low cost sector, with access for all operators in order to maximize competition and reduce the costs for both passengers and freight. saafike01l M:\Rosemary\ANIL\T-S-M-VoI 2-Sector Papers Apiil 2003.doc May 30, 2003 1:46 PM 88 This would help to provide the air freight capacity, for fish and tropical fruits, that Monibasa presently lacks. 147 A study of the management of the civil aviation infrastructure in Kenya is urgently required. It is not simply a question of the management of the main airports but also of the ownership, management and financing of all the airports - international, domestic and secondary airstrips. The present structure is clearly not delivering the quality of infrastructure and management required. There are alternative models, these should be explored. PPIAF financing was successfully obtained for a similar study in Tanzania, there seems little reason why such financing could not be obtained for Kenya. Open Skies: There has been much talk of an open skies policy for Africa, and various declarations have been made in political fora endorsing the approach. However, it is not clear that national policy makers have really accepted the concept as applicable to their own country rather than just to other countries and are prepared to abandon the protection of national carriers. Kenya Airways is now a financially successful airline, but aspects of government protection, in terms of grandfathered traffic rights, still seem to linger. Kenya , with a successful aviation sector, a large tourist market, and a vibrant air-based export sector, has everything to gain from the introduction of an open skies policy (even though Kenya Airways might find the consequences uncomfortable on some routes). The open skies policy should be extended to both passenger and freight sectors, allowing sabotage by foreign airlines, and to allowing charter passenger aircraft to utilize empty belly capacity for the uplift of cargo, if available"8. Policymakers in the sector should ignore the presence of Kenya Airways as the 'National Carrier', it is now a privately owned and managed company, able to operate commercially and adjust its level and pattern of services to the market's demands. Rather, the policymakers should be assessing the policies and agreements that will maximize the contribution of the civil aviation sector to the economic and social development of Kenya. This seems inevitably to lead to providing a low cost sector, with access for all operators in order to maximize competition and reduce the costs for both passengers and freight. 88 This would help to provide the air freight capacity, for fish and tropical fruits, that Mombasa presently lacks.