Report No. 55672-UG Uganda Public Expenditure Review Strengthening the Impact of the Roads Budget Poverty Reduction and Economic Management 2 Africa Region Document of the World Bank Monetary Equivalents Exchange Rate Effective January 2010 Monetary Unit = Uganda Shillings (Ush) 1.00 dollar US = Ush 1950 ABBREVIATIONS AND ACRONYMS APG Advance Payment Guarantee CEM Country Economic Memorandum DLP DUCARIP District, Urban, and Community Access Roads Investment Plan FIDIC International Federation of Consulting Engineers GAAP Governance and Accountability Action Plan GoU Government of Uganda IMTC Inter-ministerial Technical Committee on Roads KM Kilometer KPI Key Performance Indicators LIC Low Income Country M&E Monitoring and Evaluation MIS Monitoring Information System MoFED Ministry of Finance and Economic Development MoFPED Ministry of Finance, Planning, and Economic Development MoWT Ministry of Works and Transport MTEF Medium Term Expenditure Framework MTRA Multi-Sector Transport Authority NRSA National Road Safety Authority NTMP National Transport Master Plan PA Price Adjustment PDU Procurement and Disposal Unit PEAP Poverty Eradication Action Plan PER Public Expenditure Review PPDA Public Procurement and Disposal of Public Assets Authority R2 Regression Coefficient RAI Rural Access Index R&D Research and Development RSDP Road Sector Development Program SG Solicitor General SSA Sub-Saharan Africa SMART Specific, Measurable, Attainable, Realistic, and Timely TSDP Transport Sector Development Project UNHS Uganda National Household Survey UNRA Uganda National Roads Authority URF Uganda Road Fund USD United States Dollar VO Variations Orders Vice President: Obiageli Katryn Ezekwesili Country Director: John McIntire Country Manager Kundhavi Kadiresan Sector Director Sudhir Shetty Sector Manager: Kathie Krumm Task Team Leader: Dino Merotto/Jos Verbeek Table of Contents ACKNOWLEDGEMENTS .......................................................................................................... i Executive Summary .......................................................................................................................1 A. Introduction ..........................................................................................................................1 B. Key messages .......................................................................................................................1 1. Uganda's Rural Roads Investment Strategy ....................................................................8 A. Introduction ..........................................................................................................................8 B. The impact of investment in roads on agricultural production and consumption in rural areas............................................................................................................................................12 C. What is the transport requirement from a farmer's and trader's perspective? ...................15 D. How are public resources allocated across districts? .........................................................19 E. Conclusions and key recommendations .............................................................................23 2. Value for Money and Absorptive Constraints in Procuring and Implementing Roads Contracts.......................................................................................................................................25 A. Introduction ........................................................................................................................25 B. Value for money, absorption and the project cycle............................................................26 C. Conclusions and key recommendations .............................................................................39 References .....................................................................................................................................63 List of Tables Table 1.1: Secondary Road Network Density (in km/1,000 km2) ................................................11 Table 1.2: Financing Plan of the Medium Term Expenditure (in billion Shillings) .....................11 Table 1.3: Transport determinants of income derived from agricultural sales .............................14 Table 1.4: Difference between Sales (at the local price) and Transport Costs per Mode of Transport, Commodity Value, Distance and Tonnage (in USD) ...................................................15 Table1.5: Transport Price per Mode of Transport and Distance ..................................................16 Table 1.6: Selling Price Discount Needed to Compensate Operating Costs for a Truck for Various Quantities and Commodity Values ..................................................................................17 Table 1.7: Catchment Area (in numbers of farms and villages) for the Equivalent of 5 and 10 Trucks Traffic ................................................................................................................................17 Table 1.8: Main Determinants of Spending for Rural Roads .......................................................20 Table 1.9: Share of Maintenance and Rehabilitation Needs (for districts roads) Covered by the Current Maintenance Allocation (in percentage) ...........................................................................22 List of Figures Figure 1.1:Consumption Compared to the Time to General Agricultural Producers' Market ......12 Figure 1.2:Consumption Compared to the Time to a Paved Road ................................................12 Figure 1.3:Food and Beverage Consumption Compared to the Distance to the Nearest City.......13 Figure 1.4:Road condition compared to road maintenance funds of selected districts in Uganda 21 Figure 1.5:Coffee Potential at International and Local Prices Compared to Road Maintenance Grants (USD) .................................................................................................................................22 Figure 1.6:Maize Potential at International and Local Prices Compared to Road Maintenance Grants (USD) .................................................................................................................................22 Figure 2.1:Cost Savings of Effective Project Management ...........................................................28 List of Boxes Box 1.1: What factors affect transport cost in rural areas? .............................................................9 Box 1.2: Uganda's Road network .................................................................................................10 Box 1.3: How to foster loads consolidation ..................................................................................18 Box 2.1: The road project cycle in Uganda ..................................................................................27 Box 2.2: Current procurement practices for roads projects ..........................................................31 Box 2.3: Additional technical assistance and training needs for UNRA ......................................37 ACKNOWLEDGEMENTS This report was prepared by a team led by Dino Merotto during the initialization and preparation stage and by Jos Verbeek during the preparation and finalization of this document and its dissemination. In preparing the report, Gael Rabbaland, Patricia Macchi, Dino Merotto, and Carly Petracco were responsible for the analysis of the rural roads strategy in Uganda. Paul Wade, Nick Gardner, Godfrey Kalikabyo, and Amina Kyabangi analyzed opportunities to improve value for money and absorptive capacity. Dieter Schelling and Victor Ocaya provided valuable insights from their experience in the road sector. Invaluable administrative support was received from Rosemary Mugasha. Rebecca Simon was very helpful with the organization of the dissemination efforts. The team received guidance from Kathie Krumm, Sector Manager in the World Bank and Kundhavi Kandiresan, Country Manager for the World Bank in Uganda. Their guidance and insights are gratefully acknowledged. The internal peer reviewers for the report at concept stage were Ed Mountfield (Economic Advisor, OPCCE ) and Vivian Hon (Senior Economist, FEUUR ), while Vivian Hon was joined by Dieter Schelling (Lead Transport Specialist, AFTTR) and Vivien Foster (Lead Economist, AFTSN) to review the final document. The team is also grateful to the support it has received from the various agencies of the Government of Uganda. The team would like to thank in particularly officials from the Ministry of Finance and Economic Development, UNRA, and URF for their feedback and overall assistance. Chapter 2 was prepared in collaboration with DFID who financed part of the background paper on value addition and absorptive capacity. Various Government agencies have provided valuable time to meet with mission members and provided comments on earlier drafts. The PER benefited enormously from the comments and support provided by many stakeholders in Uganda. The initial results were discussed in Uganda in March 2010. i Uganda Public Expenditure Review EXECUTIVE SUMMARY A. INTRODUCTION i. Uganda needs to focus on improving the effectiveness of its roads investment strategy for rural Uganda and improving the manner in it procures and implements roads contracts at the national level. In recent years the Government of Uganda has shifted the priorities in its national development strategy as there was accumulating evidence that infrastructure deficiencies had become a binding constraint to economic growth and poverty reduction. ii. Consequently the Government of Uganda increased in particular the budget allocation for the road sector substantially as a means to tackle this constraint to growth and poverty reduction: · By investing in rural roads it aims to facilitate market access for farmers, which will allow them to increase their earnings capacity; and · By improving the national roads network, transport cost will be reduced, competitiveness enhanced and additional income generated. iii. However, to ensure the highest economic return for its investment, it is advised to rebalance the way allocations are set for rural roads and to increase absorptive capacity to efficiently utilise the augmented budgetary resources for the national roads sector. The remainder of the executive summary will discuss how this can be done. B. KEY MESSAGES Improving Uganda's Rural Roads Investment Strategy iv. Rural household consumption is positively related to connectivity to a market. Therefore, Government policies that decrease rural related trade cost, including transport (service) cost will promote domestic trade and export and thus reduce rural poverty. Consequently, the Government of Uganda in its ten year District, Urban and Community Access Roads Investment Plan committed itself to increase the allocation for rural roads significantly. v. A higher rural access index, which measures the proportion of rural people who live within two kilometres of an all-season road and which is used as a benchmark to assess the need for rural transport investment in most low income countries, is not a significant determinant of higher consumption by rural households in Uganda. Therefore, a more refined approach on how much to allocate to a district needs to be designed. Such an approach should include the need for transport, including options for consolidation of loads, length and condition of existing roads network, population, area and above all agriculture potential. 1 vi. The need for transport by a rural farmer is rather limited. Taking into account the fact that plot size is limited to less than one hectare, the average farmer's transport requirement in Uganda is quite small. More detailed analysis of production and yields in three districts, Bushenyi, Masindi, and Tororo, show that without consolidation of loads transport by bicycle and/or motorcycle is adequate and economical. Consequently, existing infrastructure is probably sufficient to link farmers with the nearest market place and emphasis should thus be on maintenance of rural and less on new rural roads. vii. Consolidation and deconsolidation of freight at hubs becomes optimal when the capacity of the most efficient mode of transport, i.e., truck, is large compared to the average freight being transported from origin (farm) to its destination (market). However, consolidation of loads has proven difficult due to existing coordination failures. The Government could assist farmers and transporters testing various models such as the producer groups' model developed in Poland, the IT based e-choupal model pioneered in India or supporting contract farming where appropriate to overcome the existing coordination failures. viii. Currently budget allocations to districts for roads are based almost in its entirety on network length. Road condition and area of a district do not explain why some districts benefit more from higher funding than others. At the same time agricultural output and potential do not appear to be considered either when allocating the road maintenance budget in Uganda. Henceforth, the incentive for a district is squarely to increase network length as that determines its budget allocation. ix. Estimating agricultural potential using export parity prices as well as local prices showed for the 2006 budget that districts with high agricultural potential receive too little in road maintenance grants while other with low agricultural potential received significantly more. As a household that is able to produce for local markets or even better for export, is likely to have a higher income and therefore is less likely to be poor, connecting local farmers, through investments in rural roads, to local and export markets contributes to poverty reduction. x. Resources are scarce and therefore, it is important to prioritize the allocation of resources to those districts that have the ability to take advantage of this improved connectivity of farmers to agricultural markets. Henceforth, the need to take into account agricultural potential of a district when allocating public resources for rural roads. This could require additional reallocation of resources away from certain districts in the South West to districts in the North. 2 Value for Money and Absorptive Constraints in Procuring and Implementing Roads Contracts xi. Historically the road sector suffered from underfunding and poor management. The road sector has now become a very clear priority for Government and consequently there has been a quantum step up in funding in the last 2 years. No major changes in policy and approach are required at this stage, but rather an emphasis on policies that can facilitate the absorption of the increased funding without unreasonable increases in unit cost and practices that will assist with timely and efficient implementation of the planned roads investments. xii. Given that the institutional structure in the public sector for roads is rather young, with Uganda National Road Agency (UNRA) established in 2008 and Uganda Road Fund (URF) having been established this year, targeted technical assistance, primarily for UNRA, is needed to have an impact on the successful and efficient implementation of roads projects. It is clear that, whilst UNRA has achieved a lot in its first year, it has also been consolidating the new procedures, new staff and setting up new systems. To help this process it is proposed that UNRA appoint Associate Directors in each Directorate and establishes a dedicated program management unit that will manage the design process and the tendering of the works contracts. xiii. There are a number of areas where action can be taken to increase absorption and value for money in the sector by focusing on reducing risks and delays in the implementation of projects, and on increasing the efficiency of procedures. Projects are still taking too long to implement, and cost and time overruns are significantly higher than should be the case. xiv. One of the main limiting factors is procurement and the time it takes, which greatly impacts on value for money and absorption capacity on all contracts. In Uganda under current procurement Public Procurement and Disposal of Public Assets Authority (PPDA) rules, it takes at least three years to initiate a new project from taking on a new consultant, undertaking design and starting works on site. To procure a consultant for a GoU project takes 10.5 months with another 1.5 months required for them to mobilise, a total of 12 months (typical timeframe). Of this 12 months, 5 months (42 percent) input is from bidders input and 7 months (58 percent) is required for preparation and approval. A similar time of 10.5 months is required to procure a contractor plus another 3 to 6 months for mobilisation. In addition a contractor can not immediately start "high volume work" (high volume of spend), it takes time to set, undertake initial works and build up momentum. xv. One area that appears to be least appreciated is the cost implications of a project that is awarded late i.e. the time between tender and award is protracted. Slow award of contracts will also significantly impact on the rate of increase in absorption i.e. delay to works starting on site, that can be achieved as well as increased cost in part due to price adjustment clause which uses the base price submitted by the contractor at tender stage, not start of contract works on site. In simple terms any slippage in time during the 62 months procurement process can result in overall increased cost through the price adjustment clause. Conversely if the time to award a contract can be reduced real savings may be achieved. 3 xvi. To reduce the risk of delays and claims, it is far more cost effective to spend additional funds in improving the quality of designs and improving the quality of the works contracts in order to prepare improved tenders. Therefore, UNRA should undertake technical audits on all works contracts before they are tendered and use experienced experts when preparing the works contracts. This could best be done by a dedicated unit within UNRA. xvii. The current approach of not commencing with the landtake procedure until the final construction contract has been given the go-ahead delays absorption. It takes approximately six months for a 40km road and the acquisition of registered land can take up to 15 months for 100km of road. UNRA should determine how it can complete landtake prior to signing a works contract to avoid additional claims from contractors because of delays. Addressing the capacity of the office of the Government Chief Valuer and the Land Titling Office are also needed to ensure a more timely process of landtake. xviii. Currently all variations orders are sent to the SG for approval. This has been a particular cause of delay. The variation orders need to be processed timely. The current process falls well short of best practice. In many countries variation orders are processed by the Roads authority up to a certain value where upon they obtain approval from other authorities. Such a process is transparent and much faster and Uganda could usefully implement it. xix. To contain the rise in unit prices the Government could usefully, in addition to measures it is already contemplating, review costs of materials and investigate potential of alternative road construction materials (based on local availability) and review how GoU may help reduce costs by facilitating expanded domestic supply of inputs. For example, based on projected volumes of cement requirements, information could be provided to potential investors and/or cement companies about the prospective multi-year demand for major input for public construction works. xx. The data reporting system for measuring performance in the road sector is weak. Some Key Performance Indicators (KPIs) are reported in the Ministerial Budget Statement, but these are fairly general in nature. Furthermore, many sector performance indicators desired by Government in order to more accurately assess performance cannot yet be reported due to deficient or lacking data reporting systems. The M&E department in UNRA needs further investment and support. It is recommended that monitoring needs to be extended from simply recording progress to making predictions for future modelling and providing indicators of potential problems. 4 Pretext 1. In recent years the Government of Uganda (GoU) has shifted the priorities in its national development strategy. There was accumulating evidence that infrastructure deficiencies were binding constraints to economic growth, in particular deficiencies in the transport and energy sectors. If bottlenecks in infrastructure were not removed the rate of economic growth could slow down in the next few years which would also impact on the progress in poverty reduction. 2. This marked a reorientation from the Poverty Eradication Action Plan (PEAP-3), i.e., the national development strategy at the time. Since the bulk of transport in Uganda is done by road (trucking), the most immediate transport infrastructure deficiencies to be addressed were in the road sector. The most pressing needs were identified as improving through investment in rural roads market access for farmers and improving road quality (through rehabilitation, upgrading and better maintenance) and better linking some regions of the country to the national road network and to neighboring countries for regional trade through new investments at the national level. 3. In light of this the Government started a shift of budget resources towards infrastructure in last fiscal year's budget (see table 1). Up until FY06/07, GoU spend on average 2.3 percent of GDP or approximately 12 percent of total expenditure o managing its roads sector. In FY08/09, this number increased significantly to 3.6 percent of GDP for that year, equivalent to 18.8 percent of total expenditures. For the ongoing and upcoming fiscal years, these numbers are anticipated to remain at these alleviated levels. The Medium Term Expenditure Framework (MTEF) allocation is projected to drop back to pre-FY08/09 levels in FY11/12. Table 1: Recent and Projected Developments of the Roads Budget in Uganda Expenditures in million of Budgeted Planned Uganda Shillings 2005/06 2006/07 2007/08 2008/09 2009/10 2010/11 2011/12 Expenditures on Roads 400.1 497.0 657.9 1,083.8 1,133.7 1,409.3 953.8 o.w financed from GoU resources 182.5 210.7 322.3 736.7 774.5 871.0 554.4 Current 30.4 80.5 152.9 196.1 179.0 180.54 215.0 Development GoU 130.2 107.8 113.6 473.4 480.3 575.29 200.2 Local 21.8 22.4 55.8 67.2 115.2 115.2 139.1 Donor 217.7 286.4 335.7 347.1 359.2 538.29 399.5 In percentage (%) % of Total Budgetted Expenditures excl interest. 11.5 12.4 14.8 18.8 16.9 19.1 11.1 % of GDP 2.3 2.3 2.7 3.6 3.1 3.4 2.1 Source: MoFED, IMF and World Bank staff estimates 4. However, the government has difficulties executing the national roads development projects. Since FY05/06 actual spending on national road development projects has hovered around 50 percent indicating that serious planning issues and limited absorptive capacity in the sector (see table 2). In general, the limited ability to absorb effectively the budgeted resources was limited to donor funding as government's own resources often had to be increased due to delays in disbursements of donor funding. However, this has radically changed in FY08/09 when 5 Government increased the availability of its own resources for national roads projects manifold. Even though absorption increased in absolute figures in FY08/09, the government and its agencies were unable to handle the programmed increase pointing towards additional bottlenecks with project implementation. To date in FY09/10, these implementation difficulties have continued. The under execution of the national roads budget is also having an adverse impact on timely execution of the Government's fiscal stimulus program in response to the impact of the global financial crisis which was included in the FY09/10 budget. Table 2: Budget and donor funding and actual spending on national road development projects 2004/05 2005/06 2006/07 2007/08 2008/09 I. Government of Uganda funded roads works (in Ush billions) Budget 34.3 19.6 28.9 31.4 411.2 Actual Spending 37.2 16.6 33.8 45.5 156.2 Absorption (% Actual/Budget) 108.6 84.7 116.7 144.7 38.0 II. Donor funded parts of road network (US$ millions) Budget 97.7 90.0 121.8 147.8 175.3 Actual Spending 92.6 25.4 52.0 68.8 104.7 Absorption (% Actual/Budget) 94.8 28.2 42.7 46.6 59.7 III. Total GoU and donor funded road works (US$ millions) Budget 116.7 100.9 137.4 164.4 380.9 Actual Spending 113.3 34.6 70.2 93.3 183.8 Absorption (% Actual/Budget) 97.1 34.3 51.1 56.8 48.3 Source: MoFPED, UNRA, Authors calculations. See Appendix A of background paper for details. 5. As revenues are anticipated in the short term to remain basically at their current levels, i.e. 15.4 percent, it is clear that such an increase has to be accommodated by an adjustment in the composition of expenditures (see table 3 and 4). The initial increase of expenditures on roads in FY08/09 was accommodated within a tight budget as overall expenditures declined by one percent of GDP. The initial surge in expenditures on roads was accommodated for a large part by a reduction of expenditures on security, justice, and governance, whose allocation declined by 0.8 percent of GDP between FY07/08 and FY08/09. To be able to keep the alleviated levels of roads expenditure for the ongoing and next FYs, allocations to security, justice, and governance in conjunction with a reduction in expenditures on human development, all measured as a percentage of GDP, are planned. 6. Note that this does not mean that their allocations in nominal terms are reduced as can be seem from the bottom part of table 4, which show how the nominal increases in expenditures are allocated. For example, of the total increase in expenditures, for FY09/10, 18.9 percent of additional nominal expenditures were allocated to Agriculture, 37.3 percent to Energy, transport and economic competitiveness, 37.5 percent to human development and 6.3 percent to security, justice, and governance. In FY11/12, roads expenditures are anticipated to fall back to their levels before the surge of around 2 percent of GDP, allowing for a direction of resources to human development, which will absorb 75 percent of the total increase in expenditures for that fiscal year. 6 Table 3: Fiscal Developments and Outlook Table 4: Composition of Expenditures Actual Estimated Planned Budgeted Planned % of GDP 2006/07 2007/08 2008/09 2009/10 2010/11 2011/12 % of GDP 2006/07 2007/08 2008/09 2009/10 2010/11 2011/12 Revenues 20.2 15.5 15.2 15.4 15.4 15.6 Agriculture and Value Addition 0.8 0.9 0.8 0.8 0.9 0.9 Domestic 13.6 12.8 12.5 12.8 13.2 13.7 Energy, Transport and economic competitiveness 4.0 4.6 5.3 4.7 4.6 3.4 Grants 6.6 2.7 2.7 2.6 2.2 1.9 o.w. Roads 2.3 2.7 3.6 3.1 3.4 2.1 Total Expenditures 22.3 17.9 16.9 17.8 19.2 19.2 Human Development 5.9 5.6 5.7 5.0 4.7 4.9 Overall Balance Security, Justice, and Governance 6.7 7.2 6.5 5.4 5.3 5.5 Increase in Expenditure (as % of Including grants -2.1 -2.4 -1.7 -2.4 -3.8 -3.6 total and y-o-y) allocated to: Exluding Grants -8.7 -5.1 -4.4 -5.0 -6.0 -5.5 Agriculture and Value Addition 0.1 5.2 2.2 18.9 11.5 16.9 Financing 2.1 2.4 1.7 2.4 3.8 3.6 Energy, Transport and economic External 3.1 2.5 1.8 2.4 3.0 2.6 competitiveness 25.0 34.7 45.3 37.3 30.7 -78.8 Domestic -1.0 -0.1 -0.1 0.0 0.8 1.0 o.w. Roads 10.0 20.6 42.0 16.1 43.5 -114.1 Human Development -11.2 17.5 32.2 37.5 21.0 75.2 Nominal GDP Growth 22.8 15.7 22.3 21.2 14.2 11.8 Security, Justice, and Governance 86.1 42.6 20.3 6.3 36.8 86.8 Real GDP Growth 8.7 7.1 6.3 6.6 7.0 Total Increase 100.0 100.0 100.0 100.0 100.0 100.0 Source: IMF, GoU MTEF publication, and Bank staff estimates 7. This report is a logical extension of the previous PER which analyzed if whether there existed room in Uganda's budget for more spending in particular on health to promote faster economic growth without weakening the macroeconomic environment. The previous PERs main messages were that Uganda needs to increase investment in infrastructure and budget appropriate levels for spending on maintenance if its current impressive growth is to continue. This PER, therefore, analyzes two basic questions: (i) one of allocative efficiency of Uganda's rural road budget i.e. how it can improve upon the method used to allocate resources for managing rural roads; and (ii) second of absorptive capacity and value for money i.e. how it can improve the cost effectiveness of the national roads sector while facing a major increase in budget and thus pressure to spend resources without significant cost increases. 7 1. UGANDA'S RURAL ROADS INVESTMENT STRATEGY A. INTRODUCTION 1 Uganda has experienced positive economic growth over the past two decades, since 1988 annual GDP growth has ranged from 3 to 12 percent with an average of six percent growth over the time period (World Development Indicators). Despite this overall growth, poverty remains prevalent in rural areas of Uganda (Deininger and Okidi 2003; Fan et al. 2004, World Bank 2007). Several articles address this issue in the contexts of roads and access to markets. The prevailing notion is that as household distance from roads increases (on roads which eventually lead to markets), the income/consumption expenditure of household decreases.1 2 By facilitating market access, through roads, the Ugandan government can encourage participation in export leading to decreased rural poverty. Utilizing district level data from 1992, 1995, and 1999, it is established that government expenditures on roads have a significant impact on poverty reduction in rural Uganda (Fan et al. 2004). Further research has shown that market availability increases household participation in export crops which leads to higher income among these households (Balat et al. 2008). By utilizing instrumental variable regressions the authors concluded that farmers with fewer markets for agricultural export crops are poorer than those with access to markets. 3 Government policies which decrease trade costs, including transport service costs, will promote export and domestic trade and lead to poverty reduction. Stifel and Minten (2008) site high transportation costs as a reason for the positive relationship between poverty and isolation. The high costs may be connected either to a lack of passable roads or a lack of a mode of transportation, including public transportation. As noted previously, Balat et al. (2008) shows that market availability in Uganda increases household participation in export cropping leading these households to be less likely poor compared to non-export crop households. 4 The rural access index (RAI) measures the proportion of rural people who live within two kilometers (km), typically equivalent to a 20-minute walk, of an all-season road2 and is used as a benchmark indicator to assess the need for rural transport investments in most Low Income Countries (LIC). This indicator, also used in Uganda, is said to be a compromise between those who find any distance even less than one km too great a struggle (i.e., the elderly and disabled) and those who are accustom to walking great distances because of their remoteness. 1 This chapter draws from the background paper prepared by Raballand et al, 2009. 2 An all-season road is a (gravel or bitumen paved) road that is motorable all year by the prevailing means of rural transport (often a pick-up or a truck which does not have four-wheel-drive). Predictable interruptions of short duration during inclement weather (e.g. heavy rainfall) are acceptable, particularly on low volume roads. 8 Box 1.1: What factors affect transport cost in rural areas? Across LIC, governments and their development partners now face a dilemma: should they open more rural roads to achieve a complete RAI in a country or should they drop the aim of the full RAI because of sustainability problems and concentrate more resources on maintenance of the existing road network? To answer such a question it is important to understand the various elements in play. Road impact does vary. Over the last thirty years various road impact studies have been carried out and in view of different country circumstances, a wide range of impacts have been found. Impacts range from a negative one on agricultural production, where the area under crops fell by 52 percent on the project road and 44 percent on a control road, to situations of very substantial positive impact in the case of a new 85 km mountain road in Madagascar where rice production increased by 160 percent and coffee by 70 percent (Mitchell and Rakotonirina, 1977). The impact of roads investment on the economy seems to depend upon a range of factors: (i) the magnitude of the change in transport costs, (ii) the competitive nature/current behavior of the transport and distribution markets, and (iii) the response of different parts of the economy to changes in transport costs and quality of transport. Change in transport costs. Large changes in transport costs could occur with new construction if this involves a large proportionate change in trip length for diverted traffic or a change in transport mode, say from human or animal transport to truck. Smaller changes in transport costs (per km) will occur with rehabilitation or maintenance. A change in impassability can have a major effect on transport cost. For example, a bridge over a river can reduce the need for a detour. Weak soils, or an area prone to landslides, can make vehicle transport impassable during the wet season. In accessing a major market, a 20 km reduction in trip length could give transport cost savings that are 20 times more than 5 km road maintenance improvement. Similarly, because human transport is so expensive, it has been estimated for the forest zone in west Africa that converting a footpath to a navigable track could change transport costs by over 100 times compared with the effect of upgrading the same length of track to a gravel standard road. (Hine, Riverson and Kwakye, 1983). The competitive nature of transport and distribution markets. For transport cost reductions to have the maximum effect on other sectors of the economy, it is important for transport cost savings resulting from road improvements to be passed on to producers and consumers. In theory this depends upon the nature of competition among transporters and others involved in marketing and distribution. Concerns about the competitive nature of transport operators have long been recognized and most recently in a study on international corridors (Teravaninthorn and Raballand, 2008). A number of earlier studies have pointed high cost monopolistic transport operations in Africa for many years. Similarly there is also plenty of evidence of high marketing margins, restrictions of supply and other monopolistic food marketing practices in Africa (Romanik, 2007, Shepherd 2005, Balat et al. 2008). The response of different parts of the economy to changes in transport costs and quality of transport. Increases in personal mobility are often the most noticeable change resulting from transport improvements. Reduced transport costs will often lead to an increased frequency and availability of transport services. Bulky low value commodities have, almost by definition, high transport intensity. Hence mineral production can be very sensitive to transport costs. Similarly bulky low value agricultural products, e.g. sugar, coconuts and melons also have a comparatively high transport cost component of their final market price. However grain has a relatively high value to weight ratio and as a result there may be little impact on farm gate prices from reduced transport costs. For example, assuming all the transport cost reductions are passed on the farmer, it has been calculated, in an example from Ghana, that improving a 5 km earth road to gravel standard would only increase farm gate prices of maize by 0.1 per cent. (Hine, Riverson and Kwakye, 1983). 9 5 In Uganda, contrary to some other countries in Sub Saharan Africa (SSA), the government has invested heavily in the road sector, and especially in rural roads. Currently, the GOU guides the development of the entire road network in Uganda (see box 1.2 for an overview) in line with the Road Sector Development Program (RSDP), which has two components: (i) the RSDP is a 15-year National Transport Master Plan (NMTP, issued in Nov. 2008) to be implemented by Uganda National Roads Authority (UNRA), and (ii) the Ten-Year District, Urban, and Community Access Roads Investment Plan (DUCARIP), to be implemented by local governments, both rural (district) and urban authorities. However, due to the fact that Uganda is still a predominant rural country, and despite investments, the RAI has not even reached 30 percent in Uganda. Therefore, thousands of additional km of rural roads would need to be built in Uganda to achieve a rural access index of 100 percent. Box 1.2: Uganda's Road network The road network in Uganda is about 78,000 km comprising of: National roads (also known as trunk roads): 10,800 km of which 2,870 are bitumen standards and 7,930 km are gravel surfaced. National roads connect major towns and districts with one another and link Uganda to the neighboring countries; the national roads have expanded in size over the years, not by construction of new roads, but through re-classification of district roads into national roads network (e.g. the national road network was 9,300 km in 1996). The paved national road network has also expanded. In 1996 only 2,200 km or 24 percent of the national network was paved. Since then the paved network has expanded to 2,650 km by 2003 and 3,050 km by 2008. The length of paved national roads is expected to increase to 4,100 km by 2013, and to 7,100 by 2023. The 2023 figure would represent 37 percent of projected national network of 19,000 km. District roads (also known as rural/feeder roads): The district roads are about 27,500 km. This will reduce to 20,000 km due to re-classification and transfer of some 8,000 km to national road network. District roads are predominantly gravel and earth surfaced. About 12,322 km of district road network is in good condition, 6,161 km is in fair condition, and 8,939 km is in poor condition. Community access roads (also called economic roads) are small tracks and footpaths which link communities to social and trading centers, and connect to district and national roads. There are about 30,000 km of community access roads. Access roads are predominantly earth surface with carriage width ranging from 1 to 3 m. Access roads are the responsibilities of Local Council III Governments/sub-county governments, which are sub-division of district governments. No inventory has been taken on community access road condition. The estimated road network of 30,000 km was based on the assumption that links in the range of 2 to 5 km, and a sub-county has 8 to12 links. 6 Uganda has, however, one of the highest overall road density in SSA and the highest secondary road density (see table 1.1). In the last 15 years, the GOU has made substantial investments in rehabilitation and maintenance of District, Urban and Community Access Roads (DUCAR), is estimated at 740 billion Uganda Shillings (US$400 million). The impact of the investment in the last 15 years has been substantial since the proportion of district roads from fair to good condition has increased from 15 percent in 1990 to 65 percent in 2007. 10 Table 1.1: Secondary Road Network Density (in km/1,000 km2) Countries Density of all roads Density of secondary roads Uganda 385 136 Rwanda 568 72 Malawi 165 71 Lesotho 196 50 Ghana 187 33 South Africa 300 31 Kenya 111 30 Tanzania 62 25 Cote d'Ivoire 82 24 Nigeria 174 23 Benin 142 21 Namibia 77 15 Madagascar 51 11 Cameroon 72 11 Senegal 94 10 Mozambique 61 6 Burkina Faso 39 6 Zambia 50 5 Ethiopia 46 5 Chad 27 5 Niger 13 2 Average 138.19 28.18 Median 82.00 21.23 Source: Carruthers et al. (2008). 7 The GOU has, nevertheless, adopted an ambitious investment plan for rural roads for the period FY08/09 to 2017/18. In March 2008, a Ten-Year District, Urban and Community Access Roads Investment Plan (DUCARIP) with a corresponding financing plan were announced. In the upcoming ten years, GOU has committed itself to invest a total of 1,594 billion Shillings or USD862 million, of which 953 billion Shillings for district roads for ten years from fiscal year 2008/09 to fiscal year 2017/2018. Over the medium term, 2008/09 to 2012/13, the estimated shortfall in financing the medium-term investment plan for both DUCARIP and MTEF is around 365.5 billion Shillings (USD197.7 million equivalents; see Table 1.2 for details). Table 1.2: Financing Plan of the Medium Term Expenditure (in billion Shillings) Expenditure estimates 2008/09 2009/10 2010/11 2011/12 2012/13 Total DUCARIP projection 125.5 141.2 156.9 174.2 171.7 769.5 MTEF projections 55.8 75.8 90.8 90.8 90.8 404.0 Shortfall 69.7 65.4 66.1 83.4 80.9 365.5 Source: MOFPED MTEF Ceiling FY 2007/08 ­ FY2012/13. 11 B. THE IMPACT OF INVESTMENT IN ROADS ON AGRICULTURAL PRODUCTION AND CONSUMPTION IN RURAL AREAS 8 This section uses empirical analysis to test out the causal relationships between investments in roads and agricultural production and consumption.3 It uses the Uganda National Household Survey (UNHS) data from 2005/06 to look at the relationships between: a) whether the household's share of their crop that is marketed increases per capita consumption - to answer the question of whether there are gains from trade; b) whether household road access and proximity (distance, time use) to markets affects the share of households' crop that is marketed; c) whether the mode of transport used to access markets (and the time taken) depends upon access to roads [triangulated with traffic counts]; d) whether there are remoteness effects on percentage marketed and mode of vehicle use to access markets. Figure 1.1:Consumption Compared to Figure 1.2:Consumption Compared to the the Time to General Agricultural Time to a Paved Road Producers' Market 1 40 140 1 00 1 20 Total Con sump tio n (1 ,0 00 Sh illi ng s) Total Co nsu mption (1,0 00 S hil ling s) 100 120 80 80 60 60 40 40 20 20 0 0 0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14 16 Tim e to Ma rket (Days) Time to Ta rma c Ro ad s (D ays) Note: Red represents walk, green a taxi, blue a bus, purple a motorcycle, yellow a bicycle, and brown a boda-boda. The upper five percent of time to market and of total consumption has been dropped. 9 Less consumption is correlated with distance to agricultural markets and distance to a tarmac road (see figure 1.1 and figure 1.2). The story emerging from figure 1.1 and 1.2 is that different conclusions about time to markets can be drawn. An overall trend across all the graphs is a downward relationship between consumption and access to an agricultural market and a tarmac road. In addition, the mode of transportation does not seem to vary much with distance or consumption. Bikes, motorcycles, taxis, buses, and boda-bodas (bike taxis) also appear to be time and income invariant. Therefore, in Uganda there does not appear to be a strong relationship between the mode of transportation and the consumption level of the household. 3 Integration of agricultural market prices, which were analyzed in the recent CEM (World Bank 2007), is not analyzed. Nor is the impact considered on farmer incomes of multiple markets or options for sale. Barat et al. (2008) show that having more options increases sales and prices received. 12 10 Using distance of a household to a city confirms that consumption is negatively related to isolation. A city will have all the markets, i.e. for inputs and outputs, and therefore provides distance between the household dwelling and the city therefore provides for more accurate and precise measure of isolation than individual markets. The results are depicted in Figure 1.3. To give an overall picture the households were broken down by quintiles of distance to the city. For the overall graph the downward trend is again apparent (for more details see Annex X). An important point to note is the large drop in consumption between the first quintile and the second quintile. After 4.5 km from a large city/market the consumption of the household drops greatly, and more than it changes between any other two quintiles after that. This finding is consistent with with Stifel and Minten (2008). Figure 1.3:Food and Beverage Consumption Compared to the Distance to the Nearest City T o ta l C o n su m p ti o n (1 ,0 0 0 S h i ll in g s ) T o ta l C o ns u m p ti o n (1 ,0 0 0 S h il l in g s ) T o ta l C o n s u m p ti on (1 ,0 0 0 S h il li n g s ) O v e ra ll Q u intil e1 Q u intile 2 150 150 150 100 100 100 50 50 50 0 0 0 0 20 40 60 80 0 1 2 3 4 5 4 6 8 10 12 14 D is ta n ce to C i ty ( km ) D i s ta n ce to C ity (k m ) D i st a n ce to C i ty (k m ) T o ta l C o n su m p ti o n (1 ,0 0 0 S h i ll in g s ) T o ta l C o ns u m p ti o n (1 ,0 0 0 S h il l in g s ) T o ta l C o n s u m p ti on (1 ,0 0 0 S h il li n g s ) Q u in til e3 Q u intil e4 Q u intile 5 150 150 150 100 100 100 50 50 50 0 0 0 1 3 1 4 1 5 16 1 7 18 1 9 2 0 20 2 2 2 4 26 2 8 3 0 32 30 40 50 60 70 80 D is ta n ce to C i ty ( km ) D i s ta n ce to C ity (k m ) D i st a n ce to C i ty (k m ) 11 The main conclusions are that there is an overall downward trend of consumption as people move further away from markets, with regards to both time and distance, and, on average, consumption is highest closer to the large cities/markets, but sharply declines for those households more than 4.5 km away. This indicates that road network expansion is very relevant to reduce poverty. However, the picture is more complex. It appears to be a distance/transport time ceiling; income generation is marginally more constrained beyond one to two days walking distance from the markets. Moreover, the mode of transportation does not really impact income. 13 Table 1.3: Transport determinants of income derived from agricultural sales Dependent variable: income (`0,000 Shillings) Basic Controls Density Tororo Greater than (1) (2) (3) (4) 2km (5) Sell direct 150.638*** 144.447*** 148.927*** 124.201*** 126.209*** (39.679) (40.170) (39.805) (39.235) (39.326) Crop type 122.013** 61.243 95.355 257.234*** 249.087*** (59.587) (62.682) (62.990) (77.457) (78.050) Yield 0.176*** 0.187*** 0.207*** 0.218*** 0.219*** (0.058) (0.058) (0.059) (0.057) (0.057) HH size 1.636 2.951 3.239 2.869 (3.625) (3.658) (3.543) (3.570) Secondary 16.303** 8.304 6.09 5.584 (6.256) (7.143) (6.949) (6.977) Gender of head 61.288 39.307 18.331 11.391 (42.884) (43.044) (42.145) (42.890) # of bikes owned 27.646 22.141 21.175 (22.737) (22.081) (22.123) Passability -0.604 0.001 -0.046 (0.532) (0.545) (0.548) Road density 440.951* 680.394*** 693.383*** (247.416) (249.809) (250.403) Tororo 127.105*** 123.665*** (37.474) (37.699) Greater than 2km 22.574 (25.407) Constant 3.078 -74.553 -153.226** -291.41*** -288.192*** (24.811) (51.275) (65.691) (75.549) (75.686) # of obs. 173 170 169 169 169 R2 0.2209 0.2631 0.3021 0.3494 0.3527 Significance: 10%*, 5%**, 1%***. Note: Standard deviation in parenthesis. 12 A more in-depth analysis of three Uganda districts4 indicate that crop type, yield, ability to directly sell on a primary market, and road density is what matters for rural consumption (see table 1.3). The three districts chosen are Bushenyi, Masindi, and Tororo. Even though rural density of roads is important, bikes owned, passability of the road, and distance to a road of less than two kilometers do not pass the relevant statistical test for significance i.e. t-value larger than 2. Hence, rural consumption does not seem to be influenced by these indicators. 4 A resource consuming exercise was carried out because the standard Uganda household survey does not provide enough information to allow a detailed analysis of the impact of transport services and infrastructure on household production and income level. A special survey derived from Sieber (1996) was designed to deal with these shortcomings and used in three districts. The three districts were chosen given that they provide a diverse but representative sample of the districts in Uganda, i.e. one with a high density road network, one with high agricultural potential but low road density and one with low agricultural potential and low road density. 14 C. WHAT IS THE TRANSPORT REQUIREMENT FROM A FARMER'S AND TRADER'S PERSPECTIVE? 13 This section will explore what the transport requirements are from a farmers' and traders' perspective and consequently demonstrate why the RAI has a minimal positive impact on income. It will explore the consequences on transport requirements given average lot size and produce produced. The farmer's perspective 14 Taking into account the fact that plot size is limited to less than one hectare, the average farmer's transport requirement in Uganda is quite small. The average farmer does not necessarily require massive investments in rural infrastructure to connect to primary markets from the village, homestead or farm gate because they can afford neither to hire a truck nor load it sufficiently to break even. Even if their agricultural productivity was significantly higher, most smallholder farmers would not approach the production threshold they would need to justify hiring a truck. 15 Current production volumes and yields in the three districts under investigation show that transport by bicycle and/or motorcycle is the most economical in nearly all cases. Volume is the critical factor in determining need for transport and in the case of Uganda; the current volumes marketed are low. This finding is especially critical to design the infrastructure requirement for farmers to be linked to markets and, in most cases, it can explain, despite a rural road, transport by truck is not much needed. Table 1.4: Difference between Sales (at the local price) and Transport Costs per Mode of Transport, Commodity Value, Distance and Tonnage (in USD) Low Medium High Transport 60 kilos, 10 kms value value value Bicycle 8.5 19.2 45.2 Motorcycle 8.1 18.8 44.7 Truck -2.1 8.6 34.6 Transport 110 kilos, 10 kms Low Medium High Bicycle n/a Motorcycle 15.7 35.3 82.9 Truck 5.6 25.2 72.7 Transport 1 ton, 50 kms Low Medium High Bicycle n/a Motorcycle n/a Truck 96.6 274.7 707.3 Notes: Low commodity selling price (cassava) is declared at 300 Shillings per kilogram, medium commodity selling price (maize in Tororo district) is declared at 650 Shillings per kilogram, high commodity selling price (beans in Tororo district) is declared at 1,500 Shillings per kilogram. Transport costs include the value of time. 16 Mode of transport needed depends critically on volume and distance. Based on various selling prices (low, medium and high), the difference between sales and transport costs per mode of transport for different distances and volumes is computed. Unsurprisingly, for one ton and 50 km transported, the margin is the highest for a truck (actually the other modes of 15 transport are not suitable); even more interesting is the fact that for 110 kilos, transport by motorcycle is more profitable than transport by truck and for 60 kilos, transport by bicycle is always the most profitable (see Table 1.4). 17 The main implication for road planning and design is that, in most cases, infrastructure for bicycles and motorcycles in rural areas is sufficient to link economically farmers and the first market. For a farmer producing low quantities and without cash to purchase a means of transport, transport per bicycle is the cheapest mode of transportation; for a vast majority of farmers, they cannot load a 5 ton truck and do not have the cash to pay for USD30 (which is over 15 times more expensive than bicycle and 10 times than motorcycles, see Table 1.5). Table1.5: Transport Price per Mode of Transport and Distance Distance to Commodities Bicycle Motorcycle Pick-up Lorry Tororo 60 kg 110 kg per 1 ton per 5 to 7 tons Market per trip trip trip per trip (km) 8 Ground-nuts, fruits 3,000 5,000 15,000 50,000 5 Rice, maize 2,000 5,000 15,000 40,000 14 Onion, millet, tobacco 4,000 7,000 30,000 50,000 14 Onion 4,000 7,000 30,000 50,000 20 Pineapple, oranges, mangoes 5,000 7,000 40,000 80,000 23 Rice, pineapples, groundnuts 5,000 8,000 55,000 100,000 18 In case of significant increase of agricultural productivity, with an average of 1 hectare per household, annual production would hardly reach 8-10 tons, which is still not equivalent to a truckload per year. Therefore, even though a season would last only a couple of months, the transport equivalent would be limited to 300-400 kilograms per week, which means that infrastructure-wise, a paved, all-weather road would not be necessarily needed and IMTs5, with appropriate infrastructure, could bridge the last mile gap. The trader's perspective 19 Without 250/500 kilos, running a truck over 50 km in rural areas is not profitable at all. Using trucking services starts to be really profitable for the trader from 500 kilos of load (see Table 1.6). That is also why, consolidation of loads is critical for a trader as without consolidation, the needed discounted selling price is so high and in that case most farmers are interested in selling their small quantities to traders (see Box 1.3 on options for consolidation). 5 Intermediate means of transport (IMT) can increase the carrying capacity and speed, reducing transport costs. If markets are too far to walk (one way 10 ­ 15 km) is often regarded as the threshold for access to markets. A pack animal can extend the distance to 20 km in hilly areas, a bicycle to 30 km in flat terrain and a single-axle tractor with trailer covers up to 50 km (Hine and Ellis, 2001). Thus, IMT make new markets accessible where producer prices might be higher; new products might be demanded, or inputs might be cheaper. For long distances the use of motor vehicles is essential. 16 At the farmer average production level, transport or marketing margins are high to compensate a lack of economies of scale. Table 1.6: Selling Price Discount Needed to Compensate Operating Costs for a Truck for Various Quantities and Commodity Values 10 km, old truck 60 kilos 110 kilos 250 kilos 500 kilos 1,000 kilos Low value 100% 67% 29% 15% 7% Medium value 57% 31% 14% 7% 3% High value 24% 13% 6% 3% 1% 10 km, new truck Low value 100% 100% 46% 23% 11% Medium value 88% 48% 21% 11% 5% High value 38% 21% 9% 5% 2% 50 km, old truck Low value 100% 100% 100% 73% 37% Medium value 100% 100% 68% 34% 17% High value 100% 67% 29% 15% 7% 50 km, new truck Low value 100% 100% 100% 100% 57% Medium value 100% 100% 100% 53% 26% High value 100% 100% 46% 23% 11% 20 When the capacity of the most efficient vehicle is large compared to the average origin­ destination freight flow, then consolidation and deconsolidation of freight at hubs becomes optimal. Like Smart (2008) demonstrates and what is relatively well known, "when origin­destination freight flows are large compared to the capacity of a standard vehicle, then the optimal routing is point-to-point because all standard vehicles are likely to achieve high load factors, and the point to-point routing minimizes travel distance. In such an optimal network, smaller, less efficient vehicles would be used to feed freight into hubs and distribute it from hubs to final destinations while large efficient vehicles would perform the interhub haulage". In most cases this point is overlooked when discussing mode of transport in rural areas. Table 1.7: Catchment Area (in numbers of farms and villages) for the Equivalent of 5 and 10 Trucks Traffic Need for 5 trucks- equivalent Need for 10 trucks-equivalent Traffic (3 times a week) traffic (3 times a week) Case 1: 1 tonne per hectare Number of farmers 600 1200 Number of villages 3.0 6.0 Case 2: 5 tonnes per hectare Number of farmers 120 240 Number of villages 0.6 1.2 Note: computations are made for a 5 ton-truck transporting goods over 30 km, with return load, USD 4,000 of fixed costs and charging at USD 1.2 per kilometer. 17 Box 1.3: How to foster loads consolidation This box discusses three models of consolidation, which occur at different levels among the farmers themselves, i.e. the producer groups in Poland, or at a higher level in the chain where the farmers output is consolidated at a single point by an outsider, i.e. the e-Choupal model or contract farmers/outgrower schemes. Producer groups: the experience of Poland. After the end of the communism rule in Poland in 1990, many farmers were lost without the direction and reliable purchasing by the government. In the free market economy many farmers suffered, especially because of their small land holding and their inability to compile with quality standards. In response, the Polish farmers organized producer groups. In producer groups all farmers retain control over their land and the group only exists to act as a market intermediary who coordinates sellers and buyers in the hopes of obtaining higher prices for their output (Banaszak 2007). The benefits from the group stemmed from diminished transaction costs to the sellers; instead the group manager searches, negotiates, communicates, contracts and monitors the transaction. By consolidating their output, the producer groups could now organize, pick up, and transport of their crops to buyers and utilize their size to negotiate for higher prices (Adamowicz and Lemanowicz 2006). The producer group acts as a point of consolidation of agricultural output, where the large size of the output is used as a marketing strength. In fact, on average group members received a premium of 6.2 percent on their products, with some groups reported premiums as high as 39 percent. The lessons learned for the experience of producer groups in Poland is the need for groups to be developed by those directly involved in the production, farmers who already have business ties. The producer groups should also establish a selection process for members and seek to create legal recognition of the group. There is also the need to recruit more members in order to increase market share and bargaining power with purchasers. Consolidation through ITC: the e-choupal model: The e-Choupal is the brain child of the Indian Tobacco Company (ITC)'s International Business Division. The idea came in response to the challenges of acquiring agriculture in Indian, problems that included small size/fragmented farms, multiple intermediaries, and poor infrastructure (Indian Planning Commission). To overcome these problems ITC developed the e-Choupal, which means village meeting place in Hindu, as a way to connect directly with the farmers using internet kiosk. On average 600 farmers from 10 villages within 5 km are served by one e-Choupal. Once the village is identified, a sanchalak is selected; he is also a farmer (Annamalai and Rao 2003). The sanchalak is the operator of the e-Choupal. Once installed, the sanchalak accesses information from the e-Choupal regarding weather, new and best farming practices, and market price information, which is gathered from mandis (local market). With this information the farmers are now capable of making an informed decision; they can either sell their produce to ITC or at the mandis. The result of the e-Choupal system has been a win-win for farmers and ITC. With greater information and understanding of prices, farmers have become more aware of what they should/can receive for their crop. When farmers sell to ITC through the e-Choupal, prices are 2.5 percent higher on average then if sold at the mandis (Annamalai and Rao 2003). And even though ITC is paying more for the produce and compensating farmers for transport, ITC is paying less than before (Prahalad and Hammond 2002). Because ITC cut out the intermediaries the mark up paid by ITC has decreased from 5 to 2.5 percent. Contract farming/Outgrower schemes. Contract farming or outgrower schemes are methods that firms employs to utilize the existing assets of small rural farmers. "Contract farming is a vertical coordination between a central processing or exporting unit on the one hand, and growers of agricultural products" (Al- Hassan et al. 2006). The coordination is based on a contract that outlines the purchase of the crop being grown, beforehand. In general, inputs (seeds, fertilizer, pesticides) and extension services are provided by the firm to the farm free or at a lower cost to the farmer, who in turn grows the crop and sells it to the firm at the previously agreed upon price (Kindness and Gordon 2001). Specific elements of the contract can vary, such as the extent of control over the farmer by the firm or if a certain amount of output was agreed upon, etc. 18 21 At the current production level of approximately 1 ton of cash crop per year per hectare, trucks would need to consolidate the production of at least 600 farmers. Assuming that competition in the trucking industry requires at least five trucks on the same route, it is possible to compute what is the catchment area needed to make the operation of these trucks a viable one. This would mean that a truck could probably serve only one out of three villages in the production area. The non served villages would have to transport their production by IMTs to the served village. It is obvious that for 10 trucks equivalent, the number of non-served would increase tremendously (see Table 1.7). This phenomenon is worth being noted because there is trade-off between individual traffic (for roads and trucks) and catchment area, usually neglected on the assumption that traffic will grow coupled with a smaller catchment area. In reality and in the short and medium term, increase in individual traffic (for a road) can only come at the expense of a larger catchment area, which explains why investments in large infrastructure and services in rural areas should be prioritized carefully. D. HOW ARE PUBLIC RESOURCES ALLOCATED ACROSS DISTRICTS? 22 The GOU adopted an ambitious investment plan for rural roads (see table 1.2). However, regarding investment in roads to ensure rural growth, constructing or maintaining roads in areas with high agriculture potential (See Box 1.4 for a description of agricultural potential) is a recommended policy. As this plan entails further massive investments in rural roads, it is crucial to know if the current investments are achieved according to road condition and agricultural potential. Indeed, if allocation to roads maintenance is assigned independently of road condition, future roads investments may face the same problem of spending efficiency. Box 1.4: How is Agricultural potential computed Computation of agricultural potential is possible thanks to the agroecoloogical zone (AEZ) model developed by the United Nations Food and Agricultural organization. It consists of two main steps. First, AEZ provides a standardized framework for the characterization of climate, soil, and terrain conditions relevant to agricultural production. Second, AEZ matching procedures are used to identify cropspecific limitations of prevailing climate, soil, and terrain resources under assumed levels of inputs and management conditions. This part of the AEZ methodology provides maximum potential and agronomically attainable crop yields of basic land resource units (grid-cells). Finally, agricultural potential values are computed by multiplying the output of the best-suited crop in ideal conditions (in terms of inputs) by the crop value. Source: United Nations Food and Agriculture Organization and Raballand, Macchi, and Petracco (2010) 19 Table 1.8: Main Determinants of Spending for Rural Roads Dependent variable: Released funds for feeder roads maintenance in 2006 (per capita) (1) (2) (3) (4) (5) (6) (7) - Road 1.44 1.20 5.08 2.93 condition 3.11 3.14 5.18 3.44 4.7E+ 4.8E+0 Network 4.7E+05 ** 05 ** 5 ** 4.8E+05 ** length 5.1E+ 4.7E+0 per capita 5.1E+04 04 4 5.4E+04 Number of 1.9E+ con- 3.8E+06 07 ** 5.4E+06 stituents 6.4E+ per capita 4.2E+06 06 4.4E+06 4.2E+ NRM 06 2.0E+07 ** constituents 5.2E+ per capita 06 7.9E+06 0.01 0.01 0.01 Area 0.01 0.01 0.01 Poverty 2.07 rate 2.23 738. -3.25 5.46 64 ** 97.7 431.67 ** 488.78 ** -127.08 Constant 97.0 101.35 99.79 4 62.46 92.12 83.47 174.94 # of obs. 55 55 55 55 55 55 52 R^2 0.68 0.68 0.02 0.66 0.14 0.11 0.70 Notes: (**) implies significance at the 5 percent level and (*) at the 10 percent level. Standard error is reported in italics. 23 Taking into account the high correlation between network length and budget allocation for roads maintenance, network length defines the allocation per district. Based on data per district, Table 1.8 and Figure 1.4 demonstrate that road condition and district area do not explain why some districts benefits from higher funding than others. Due to the current investment strategy in rural roads, it seems better for a local authority to expand its network than maintain it due to the fact that increased allocation probably mainly depends on the network length and can explain why local authorities now strive to upgrade many community roads to district roads. 24 On the link between spending in roads and agricultural potential, results show that agricultural output or agricultural potential do not appear to be a major consideration when allocating the road maintenance budget in Uganda. Using the 2006 figures of the amount of money released to the districts under the heading of Road Maintenance Conditional Grants, a simple correlation test was run with the agricultural potential data. The results show that there is no correlation between the agricultural output of a district and the amount of road grants received: 0.05 for the correlation coefficient between coffee potential and road grants 0.05; -0.02 between cotton potential and road grants; 0.02 between maize potential and road grants; and -0.04 between soy bean potential and road grants. 20 Figure 1.4:Road condition compared to road maintenance funds of selected districts in Uganda 45 800 40 6 99 Percentage of Non- Major Constraints (% ) 40 700 36 35 34 35 Uganda Shi lli ngs (1,000,000) 32 600 30 48 1 480 500 25 400 20 300 15 239 2 11 216 1 80 183 1 90 200 10 1 45 5 3 100 3 2 2 1 0 0 ge le i wa ra o an er gi i ra am b bu ra pi en Li eb d or um Pa ba M ba ng ch N w dj am m tu M ap A se N K K S Dis tr ic t N o ne m a yo r co n stra i nts (% ) R o ad M a in ten a n ce Source: Ministry of Finance, Planning and Economic Development. Data sorted by percentage of roads none considered as a major constraint by household interviewed. 25 Estimating agricultural potential using export parity prices and local prices shows that districts with high agricultural potential receive too little in road maintenance grants while others with low agricultural potential receive significantly more. The potential value of coffee and maize is compared to the road maintenance allocation of districts in Uganda in Figures 1.5 and 1. 6. 6 Potential output in international and local prices is presented on the left vertical axis and the amount of road maintenance grant on the right vertical access. Note that the agriculture potential is in millions of dollars, while road maintenance is in thousands of dollars. Only, a subsample is provided, including the five largest and smallest potential producing districts. The Nakapiripirit district has the potential to generate almost USD1 billion from coffee at international prices, but receives less than half the road allocation that Jina, a district that has little potential to produce one of Uganda's largest exports, receives. Figure 1.5 shows that even though Kotido has the potential to produce three times the maize of Arua, they are allocated almost the same amount in road maintenance grants. 6 See the background paper, revising the Roads Investment Strategy in Rural Areas: An Application for Uganda, for information on methodology and actual calculations. 21 Figure 1.5:Coffee Potential at International Figure 1.6:Maize Potential at International and Local Prices Compared to Road and Local Prices Compared to Road Maintenance Grants (USD) Maintenance Grants (USD) 90 0 500 1 ,8 00 50 0 80 0 450 1 ,6 00 45 0 A gr i c u ltu r a l P o te n t ia l ($ 1 ,0 0 0 400 Agricultrual Potential ($1,000,000) 40 0 70 0 R o a d M a in te na nc e ( $ 1 , 1 ,4 00 Road Maintenance ($1,000) 350 35 0 60 0 1 ,2 00 300 30 0 50 0 1 ,0 00 250 25 0 40 0 8 00 200 20 0 30 0 6 00 150 15 0 4 00 20 0 10 0 100 2 00 10 0 50 50 NA K A P I R I P I RI T 0 0 0 0 M AYU G E M O RO T O B UG I R I MB AL E KIT G U M KISO R O J I N JA K O T I DO A RU A JINJ A N AKAPIR IP IR IT M UKON O GUL U PA DER KITGU M MA YUGE KOTIDO WAKISO TOR ORO Intern ati on al Pri ce In te rn atio na l Pr ice Ho us ho ld Pri ce H ou shol d Pri ce Ro ad Mai nten an ce R oa d M ai nte na nce District D istrict 26 Based on the current size of the road network (in selected districts), the present allocation only covers routine maintenance needs (for districts roads) (see table 1.9).7 In the two districts, Tororo and Masindi, the grant allocation fully covers the need for routine maintenance. However, the allocations are unable to cover periodic maintenance and rehabilitation in the three districts, which means that even without further expansion of the district road network;8 the sustainability of the district networks is questionable, without increases in grant allocations. Table 1.9: Share of Maintenance and Rehabilitation Needs (for districts roads) Covered by the Current Maintenance Allocation (in percentage) Bushenyi Masindi Tororo Routine maintenance 88% 108% 138% Routine maintenance + periodic 29% 36% 46% maintenance (every six years) Rehabilitation 3% 4% 5% Source: MOFPED for maintenance allocation per district; needs computed from road unit costs9 and the size of the network. 7 Our selected districts are not among the lowest in terms of road maintenance allocation. 8 We exclude community roads in our discussion, assuming that it is a second priority order. 9 Following data were used for our computations: Road unit costs (in USD per km) Routine maintenance 319 Routine maintenance + periodic maintenance (every three years) 1,278 Periodic maintenance 3,836 Rehabilitation 9,204 Low cost sealing 17,297 Source: Ministry of Public Works. 22 27 A revised allocation mechanism for each districts taking into account agriculture potential, population, area, length and condition of the district road network should be considered. As stated earlier, a household that is able to produce for local markets or even better for export, is likely to have a higher income and therefore is less likely to be poor. Therefore, connecting local farmers, through investments in rural roads, to local and export markets contributes to poverty reduction. However, as resources are scarce it is important to prioritize the allocation of resources to those districts that have the ability to take advantage of this improved connectivity. Henceforth, the need to take into account agricultural potential of a district when allocating public resources for rural roads. In addition, the Government could consider using community and labor-intensive maintenance methods for rural roads such that off farm employment is created at the same time. 28 It is important though to realize that such an allocation mechanism will require increased coordination among line ministries and a strong capacity within the District, Urban and Community Access Road Department (DUCAR) with the Ministry of Works and Transport to undertake the analysis needed to prioritize budgetary allocations to districts using the proposed formula of allocating resources. As DUCAR will be the main agency within the MOWT to take on the task to change the allocation mechanism, it will be important to increase its capacity to undertake the needed analysis but aloes to play a coordinating role with other ministries involved such as the Ministry of Finance and Economic Development and Ministry of Agriculture. E. CONCLUSIONS AND KEY RECOMMENDATIONS 29 The WDR 2009 recommends that in lagging areas countries should invest in people, while in leading areas should invest in place. This combination provides people in lagging areas with education in enhancing their opportunities, while the improved infrastructure will allow mobility of people, agricultural goods and information to and from the leading area. This chapter concludes that the average farmer does not necessarily require massive investments in rural infrastructure from primary markets to the village, homestead or farm gate because they can neither afford to hire a truck nor load it sufficiently to break even if they could. Even if their agricultural productivity was significantly higher, most smallholder farmers could not approach the production threshold they would need to reach to justify hiring a truck. Consequently the conclusions are the following: · Maintenance and upgrading of existing rural roads rather than construction of new roads should be given priority in most cases; new roads might still be needed in the North as a consequence of the civil war. However, priority should be given to those proposed rural roads that facilitate the opening up of areas with high agricultural potential. · Attention should be given to innovative marketing models from other countries such as India where smallholder loads are consolidated through consolidators. · An alternative objective and strategy is proposed for rural transport policy and investments, which would take into account much more strongly agricultural potential. This would be a two-pronged approach: 23 (i) Define the road allocation per district as a direct function of agricultural potential; contemplating the economic benefits of areas with strong agricultural potential, and (ii) Define a minimal road connectivity measure per region such as connectivity at less than 8 or 10 km for Ugandan rural population based upon resource availability. · Implementation of this new approach and based on 2006 data indicate that rehabilitation, upgrading, and/or new construction of rural roads should be done in some districts in the North and resource allocation for rural roads should be reduced for some districts in the South West. 24 2. VALUE FOR MONEY AND ABSORPTIVE CONSTRAINTS IN PROCURING AND IMPLEMENTING ROADS CONTRACTS 30 This chapter analyzes what measures the Government of Uganda could implement to increase value for money and absorption of its budget for (national) roads against the backdrop that the allocation for roads has increased significantly in the last few years. Various analytical pieces, including the 2007 Country Economic Memorandum, have shown that infrastructure, including roads, is a major constraint to sustainable development in Uganda. Consequently, the GoU decided to address this issue and increase investment in the sector. The ramping up of public expenditure posed challenges for the agencies involved in the execution of the roads budget (see tables 1 and 2 above). Various efforts to improve outcomes in the sector are ongoing. This chapter focuses on how improvements in procurement and implementation of roads projects can assist the various agencies involved to increase absorption and value for money without rapid escalation of cost and loss in quality. A. INTRODUCTION 31 In mid-2008, as part of the Government's increased emphasis on upgrading the quality and coverage of the road network, a decision was made to transfer some 8- 10,000km of district roads into the national network, over a number of years. This will bring into the network roads connecting district capitals, roads to borders, heavily trafficked district roads, strategic security roads, roads to central government institutions, and key tourist routes. 32 At the same time a new institutional structure was put in place to re-organize the manner in which the Government managed the sector. It is the government's vision that Ministry of Works and Transport (MoWT) focuses on policy formulation, strategic planning, sector oversight and monitoring, and delegates executive functions, including regulatory functions, to specialized entities which have been, or are being created. The new entities created are Uganda National Roads Authority (UNRA), which is responsible for the management of national roads and commenced operations on July 1, 2008; and the Uganda Road Fund (URF), planned to be fully operational by July 1, 2010, which is responsible for funding of maintenance of all roads. The source of revenues for URF will be road user's charges including fuel levy, license fees and other road related charges. Still to be created as part of the reorganization of the sector is the National Road Safety Authority (NRSA) and Multi-Sector Transport Regulatory Authority (MTRA), which will take on the surface transport regulatory functions, currently still with MoWT.10 33 Against the current ramping up of public expenditure in roads and the new implementation structures created, this chapter addresses the following concerns that · The value for public money on road investment and maintenance spending is low; and 10 See IDA's project appraisal document for its Transport Sector Development project, Report No 50977-UG for more information. 25 · The capacity to absorb the augmented budget resources for the road sector and turn them into effective investment in a timely fashion is constrained. 34 Consequently, a review of all projects in the development project portfolio during 2005/06 ­ 2008/09 was undertaken.11 As such this chapter focuses on the main short- and near- term issues that impede absorption and reduce value for money. The chapter, unlike other reports, goes into detail in an effort to drill down and identify the underlying problems in order to ensure that appropriate, specific actions are taken in the short- and near-term that will result in increased absorption and value for money. B. VALUE FOR MONEY, ABSORPTION AND THE PROJECT CYCLE 35 Value for Money relates to the efficiency in completing quality designs and quality works contracts to specification on time, ensuring works contract bids are competitive (item prices as opposed to unit prices per km are reasonable) and there are minimal claims from contractors e.g. for extensions. In addition, one has to be wary of collusion between contractors and of political interference, which can make the cost of road works artificially high. 36 Currently, it takes 61 months as a minimum for a road to be designed, constructed, completed and handed over (see box 2.1: The road project cycle in Uganda). As seen in the project cycle above, procurement is undertaken at various stages to procure under PPDA laws and regulations all consultants and all contractors. The effort currently needed to procure a consultant or contractor should not be underestimated in terms of time and GoU manpower (all the stakeholders) to administer. A key finding of this report is that the current process using PPDA regulations is overly time consuming. 37 Final road project cost in Uganda are greatly affected by contractors' claims for extension of time as a result of poor design, obstructions on site because the landtake process was not completed before the works contract started, significant increases in the price of commodities, and late issue of variation orders. Much of this is due to slow decision making by those responsible for the project itself, quality of the work delivered by the design consultants, and procurement. In order to minimize claims from contractors quality designs are required, all necessary project preparation (e.g. landtake) completed before the project starts with prompt payments to the contractors as the project is executed. 11 A review of project files was undertaken of ongoing, recently completed and stalled road development projects. The review was done in close cooperation with the Ministry of Works and Transport, MOFPED and the Uganda National Roads Authority (UNRA). It analyzed a statistically significant sample of projects from the inventory of road development projects in the period 2005/06 ­ 2008/09 and reviewed the processes for project planning, preparation and implementation. 26 Box 2.1: The road project cycle in Uganda All projects have to pass through a project cycle from inception to completion for each and every road scheme. The main stages of a new development project and what this entails in terms of time to procure consultants and contractors are as follow: Stage 1 ­ Selection and planning. Planning the development of the network which also includes Project Identification required to identify an individual project that should be undertaken as it will provide a suitable return on investment and improve the efficiency of the network. Planning is done in UNRA by the Planning Directorate in order to develop and improve the network making use of consultants as required. The Directorate also manages the collection of suitable data on the network (road condition, traffic etc) and is currently building up an asset database. Time currently needed on each step: Procure consultant ­ 9 months Pre-feasibility Study ­ 6 months Approval to move to next stage ­ Varies greatly Stage 2 ­ Design & Development of Works Contract includes procurement of Consultant who subsequently produces a Feasibility Study, Detailed Design and the Works Contract Tender Documents while the process is managed by the Planning Directorate. Procure consultants ­ 9 months Consultancy for feasibility study and detailed design - 12 months Approval to move to next stage ­ varies greatly but can be short Stage 3 ­ Implementation & Supervision. This phase includes construction of the road and supervision. All construction work is outsourced to contractors and supervision to external consultants. Currently the contractor is procured first and later the supervising consultant. Implementation is undertaken by UNRA's Operations Directorate that receives the engineering design and tender documents from its Planning Directorate, procures the works contractors and supervising consultants, and coordinates and manages the implementation. Procure contractor ­ 10 months (Records show more than 1 year) Construction ­ 2.5 years (this will include a 3 to 6 month mobilisation period) Selection and Planning 38 In the Ugandan context it pays for itself to spend additional funds on improving the quality of designs i.e. using better designers and undertaking technical audits on all works contracts before they are tendered as well as expediting landtake in order to prepare improved tenders and reduce potential claims by contractors. On the quality of design and other project preparations (see Figure 2.1), it is worth considering that design fees are only about five percent of the construction costs and that revisions to design made during the design phase cost a small fraction of the cost of any changes made during construction. 39 Cost implications of a late awarded contract, i.e. the time between tender received and contract awarded, can be significant. As contracts include a price adjustment clause, the longer the projects takes to award, the more the overall cost increases as a result of this price adjustment clause. The price adjustment clause is a method to share risk of price escalation between the contractor and client and under the right circumstances should result in cost 27 effective project costs. Without this clause contractors would have to bid even higher to absorb the potential risk and uncertainty with the likely consequence the client pays more. Figure 2.1:Cost Savings of Effective Project Management The Need for Effective Project Management Effectiveness Cost the effectiveness and relative costs of making changes during a civil works project Operations: The costs of Design: corrective actions soar after An effective design should consider construction because of the loss the 'constructability' of the proposed of income when operations are Feasibility: works as well as how the new interrupted This stage provides an infrastructure will be operated and opportunity to confirm maintained the appropriateness of Construction: a project; only a Once construction starts it becomes relatively small budget ever more expensive to correct is required conceptual and design errors. Feasibility Design & Procurement Construction & Operations Study 40 The funding implication of the decision to upgrade 10,000 kilometers (kms) of District Roads to National Road status and hand them over to UNRA needs to be appropriately dealt with such that UNRA can indeed maintain them. As resources will remain scarce, it is recommended that UNRA undertake an exercise to balance/prioritize the scope of the road network against available funding. UNRA should review the total budget required to develop and maintain their expanded road network of 20,000 kms on a technical needs approach and discuss resource implications with MoFED as well as donors. While go through this exercise UNRA is well advised to improve it works contracts estimates by implementing a budget costing system that considers not only base costs, but also incorporates variations of price and projects risk in combination with a appropriate contingency for cost over runs. 28 Design, Development of Works Contract, and Land Take 41 While the approach largely follows best practice, to date it still has not ensured that the standard of works documentation such as tender for contractor developed by design consultants is adequate. As stated earlier, many cost overruns can be contained through a comprehensive design process. It is recommended that UNRA initiates the start of the design process for future roads earlier to provide for adequate design period and withstands pressure to get designs tendered expediently. More thorough investigation of the project site could also improve the quality of the design as inadequate site investigation has led to wrong assumptions e.g about available suitable cut for fill and consequently to additional claims from contractors leading to cost overruns. 42 The poor performance of design consultants has resulted in cost overruns in the construction phase in some cases. Primary examples include errors in quantities, change to pavement, hurried design reviews, unforeseen ground conditions, and addition to scope. It is recommended that UNRA implement Technical Audits during the design process to ensure timely feedback to consultants as well as improved quality of designs and works contracts. In addition UNRA should pro-actively guide consultants in the design process. In both cases this requires that UNRA raises its internal capacity, possibly with external support. GoU should implement a performance monitoring system for design consultants in order to provide structured feedback to consultants on their performance and over time weed out poor performing consultants in selection for new design work. 43 The majority of road projects currently planned and under construction consists of upgrading of existing routes. Hence, they should not require significant new land take (acquisition of land including, surveys, identification of owner, payment of compensation and registration of the land in the name of UNRA). However, not all existing gravel roads have a designated road reserve for future development. Hence, when those roads are upgraded, the road reserves will need to be surveyed, titled, and owners compensated. It should be noted that any realignment of the road requires additional land take. 44 The current land take practice delays road works. The current approach is to not commence land take until the final construction contract has been given the go-head in order to make maximum use of inadequate funds for land compensation. This delays work progress on site since acquisition of unregistered land takes around 6 months for 40km road and acquisition of registered land can take up to 15mths for 100km of road. It is recommended that UNRA determine whether and/or how it can complete land take prior to signing a Works Contract to avoid the high risk of claims from contractors delayed. UNRA could also request assistance from donors. 45 Lack of capacity in government agencies is a bottleneck to processing of land take cases. This is one of the reasons for the long ­ and costly - duration of the land acquisition process. The land acquisition department of UNRA consists of just one land acquisition expert and one assistant. Between them, they need to manage and supervise the surveys, valuations, compensation payouts, dealing with complaints and approving Resettlement Action Plans. This department would benefit from 2 additional staff and additional space. The other clear bottleneck 29 is in getting approval for Valuation reports from the Government Chief Valuer which has very limited resources to meet its processing requirements. 46 Planning ahead, it is recommended that UNRA assesses the likely extent of land take works required, staffing implications and the cost of compensation over, e.g. the next five years and that it develops a plan to meet these challenges. The plan should propose how to improve the processing capacity of all stakeholders i.e. UNRA, Government Land Valuer's Office, and the Land Titling Office ­ both the number and capability of staff and of consultants. The plan should also consider whether it is possible and practical for the design consultants to undertake more of the Land-Take process at design stage than currently envisaged, and determine whether land purchase can be completed for certain projects, possibly donor funded projects, before the works contract is let. Procurement 47 Delays in procurement, whether on new or existing projects result in cost increase and reduced value for money, not to speak of slower absorption. In essence service delivery is delayed and made more costly by any delays caused by stakeholders12 in the procurement process. However, some stakeholders take long periods to process documents and do not always make it clear what information they require or notify others if data is missing. This indicates a lack of appreciation that any delay in the procurement process is costly and appears endemic with departments blaming each other. The underlying reasons are weak accountability - since there is no systematic tracking of the time spent at each stage and hence delays are not attributed to individual stakeholders - and lack of incentives and sanctions to induce stakeholders to process procurements faster while of course maintaining the integrity of the process. 48 UNRA needs develop an Actionable Governance Indicators database to improve monitoring of procurement process including time each stakeholder takes to process, for all projects and changes to contracts in order to identify and quantify bottlenecks. The proposed monitoring and information system to be implemented by the Monitoring and Evaluation department of UNRA should be designed to capture enough data to include (i) dates of key events in the project cycle e.g. identification, feasibility study, detailed design phase, tender, award of contract and completion; and (ii) detailed dates of each stage of the procurement cycle itself i.e. date sent to Contracts Committee, date approved by Contracts Committee, date sent to Solicitor General, date approved by the Solicitor General etc. 49 UNRA, using the information gathered, should discuss and agree with other Ministries and other stakeholders how the procurement process can be simplified, streamlined and made more users friendly. UNRA should also develop closer ties with PPDA and Solicitor General and meet on a regular basis, i.e. each month to discuss problems and issues. It should produce agreed guidelines on how best to manage the high volume of bids and initiate "joint" training so different Ministries understand the problems others face and the 12 Stakeholders involved in processing procurements include: the User Department, the Contracts Committee, Evaluation Team, Solicitor General's office, donors if foreign aid financed, and in the case of EU projects, also the National Accounting Officer. 30 consequence of delays. It is advisable that UNRA makes one person responsible in UNRA to improve this process. In addition, the Interministerial Technical Committee on Roads (IMTC) should monitor the time taken for each ministry to provide the necessary approvals and, wherever a delay is noted, be it in UNRA, PPDA, or SG, provide the necessary authority to get things moving and streamline the process at all stages Box 2.2: Current procurement practices for roads projects The procurement process is based on the PPDA Law and associated Implementation Regulations requiring UNRA to use an internal Procurement and Disposal Unit (PDU) and Contracts Committee who are legally responsible for ensuring projects are procured in line with PPDA rules. The Procurement and Disposal Unit (PDU) within UNRA has managed the procurement of some US$250 million of government funded and donor funded projects in its first year of operation, FY2008/09. This is no small accomplishment. A study has analyzed the process that is currently followed for procuring a consultant or contractor, breaking it down into detailed steps to estimate where there are undue delays (see chart below for details). Currently the process involves 25 steps. Even with the suggested rationalization where it is assumed some tasks can be undertaken in parallel the process involves at least 20 steps for government own-financed projects. · To procure a consultant for a GoU project takes 10.5 months with another 1.5 months required for them to mobilise, a total of 12 months (typical timeframe). Of this 12 months, 5 months (42%) input is from bidder's input and 7 months (58%) is required for preparation and approval. · A similar time of 10.5 months is required to procure a contractor plus another 3 to 6 months for mobilisation. In addition a contractor can not immediately start "high volume work" (high volume of spend), it takes time to set, undertake initial works and build up momentum. The approval periods take up approximately 60% of the time to procure a consultant or a contractor or about 7 months. It is estimated that this time can be cut in half. The approval process is governed by the current Ugandan PPDA Law and associated implementation regulations. It should be remembered that donor funded projects will take longer to procure as approvals have to be sought from the Development Partner as well. The project cycle requires procurement more than once, first for design consultants etc., later for contractors and supervision consultants, and every time it takes at least 10 months, extending the total project time. 31 Desig n M th s 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 S ho rt list co nsu ltan ts 3 O bt ain ap pro val t o b id R FP Te nd er RFP ( Bid ders c om plet e te nd er) 2 Des ign E valu ate an d aw ard t end er 2 Com p lete S ign Co nt ract 4 Co nsu ltan t M ob ilises - up to 56 d ays 2 Un dert ake d esign 12 S ho rt L ist o f C on tract ors FO LLO W E D B Y C ON ST RU CTI ON 25 Co nst ruct ion M th s 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 60 61 62 O bt ain ap pro val t o b id w o rks co nt rac t 1 Te nd er W orks co nt ract 2 E valu ate an d Aw ard T end er 2 Cons tr uct ion Com plete S ign Co nt ract 4 I y ear D LP rem ains Co nt racto r M ob ilises 3 to 6 m on th s 4 Co nst ruct ion - G en erally tw o ye ars 24 37 50 The administrative procedures within UNRA, which have been set up to comply with PPDA implementation regulations, appear overly complex and onerous. Consequently, they require a high amount of effort that could be better used elsewhere in managing design projects and works contracts. In order to increase the efficiency of road sector procurement it is recommended that a taskforce, set up by UNRA, not only looks at the process of dealing with approvals from stakeholders, but among others also reviews UNRA's own internal capacity and processes in collaboration with PPDA. This should lead to more streamlined and flexible procurement guidelines while maintaining the integrity of the overall process. 51 Processing timelines need to be developed and made general knowledge so that all stakeholders know the expected processing times for each stage in the process. In addition, the following can be done to improve the process within the existing procedures: (i) early identification of the Evaluation Committee and approval by the Contracts Committee before tenders are received; (ii) timely preparation of bidding documents; (iii) preparation by UNRA of a 3-year procurement plan rather than the current 1-year plan, now that medium term funding is clear, to assist in the overall planning of the procurement process; and (iv) in the short term focus efforts on increasing the capacity of the PDU within UNRA to deal with the current procedures and to cope with the increasing workload. PPDA, which has the capacity, should be requested to expand its technical support and training to the PDU in UNRA, to assist them in efficient compliance with the procedures. 52 UNRA as one of the single largest procurement entities should have a closer and more regular relationship with PPDA. The role of the PPDA is to ensure that procurement rules are followed and to safeguard bidders and agencies involved. The agency also has a role in capacity building and has carried out some capacity building in the regional offices of UNRA. It is clear that given that PPDA deals with some 120 central government and 93 local government institutions that is has its own capacity constraints though. Notwithstanding, UNRA has an 32 urgent short term aim to get 1,940 kms of roads designed and implemented during 2010. Therefore, a temporary solution to provide for a fast-tracking of the process for these projects is required. Both PPDA and UNRA should consider allocating specific teams to handle the procurement of only these projects. 53 It is important that UNRA, PPDA and Office of the Solicitor General (SG) office work closely together to provide mutual support to guarantee successful and timely completion of procurement processes. The SG is involved in all legal aspects of procurement and contract preparation, including tender award, negotiations, contract signing, variations, extensions of time and any other changes to the contract throughout the implementation period. Under RAFU, the SG's office was represented on the Contracts Committee. This ensured that the SG was involved early in the procurement process. Since UNRA came into being, the SG is no longer represented on the Contracts Committee, and only becomes involved in the early stages of the procedure at the invitation of UNRA. In order to avoid problems and hence delays at the late stages, the SG office should be encouraged to become involved early in the process regularly, and to assist in the evaluation process, so that major issues can be identified early on, before formal approval is requested later. Consultancy contracts 54 It is generally better to have the designer carry out the supervision. Traditionally, separate consultants have been contracted to carry out feasibility studies, detailed design and construction supervision. There is also an argument that, if a separate consultant carries out the supervision, the original designer is not available to respond to any problems in the design, and the supervision consultant has a tendency to blame any problems on the designer. If the same consultant carries out the design and supervision, then he cannot escape responsibility for any design problems, as he/she will also be responsible for the implementation of the design (incentive compatibility). The counter argument is that the consultant may during construction try to cover up any problems in the design. 55 With existing procedures and implementation regulations, the current approach for separating design and supervision leads to 3 periods of procurement of 9 months each. This extends the planning and design period and delays service delivery. The procurement processing time could be reduced while also improving incentive compatibility by using the same consultant for design and supervision. It is recommended that UNRA discuss with Donors the advantages and disadvantages of this approach and how the contract would need to be revised. Contract management and implementation 56 If UNRA outsources the project management function it should ensure that the expertise in project management is also used to train and transfer experience to local staff to strengthen project management capability within UNRA. The Project Management function covers the management of the whole process of bidding, contract award, supervision of land acquisition, project management of the construction supervision consultants, instituting suitable internal quality audits, overseeing design changes, reviewing UNRA management 33 systems and developing systems, manuals, guidelines, template and forms to be used on the project as necessary. For some large and complex projects the Project Management function is contracted out to a consultant team. 57 Currently all variation orders (VO, change to a contract), once approved by the contracts committee, are sent to the SG for approval, while once the VOs amount cumulative to more than 25 percent of the contract all future variations, however small, need to be processed through the PPDA. This has been a particular cause of delays resulting in reduced value for money, i.e. the total time taken for approval of VOs. It is particularly important with VOs that they are processed efficiently and in a timely manner, as delays in issuing valid VOs can lead to further delays and claims from the contractor and subsequent further costs. The practice in Uganda falls short of best practice. In many other countries VOs are processed by the road authority up to a certain value where upon they obtain approval from other authorities. This process is transparent and much faster. 58 UNRA should be allowed to handle any VOs up to a limit of, say, 25% internally without having to seek approval of the SG. The delay in processing VOs can result in claims from the Contractor for late notification of change to works where the claim would be for extension of time (EOT) with associated costs. EOT costs can vary between 1.5 to 2.5 percent of contract value per month. It is recommended that UNRA discuss the issue with PPDA and Solicitor General and propose a process that allows entities such as UNRA to approve VOs up to an agreed aggregate value of the contract value in a transparent manner that may include copying all VOs to Solicitor General for information. 59 Any delay through the price adjustment clause, for example due to delayed land acquisition or increased scope of earth works can have cost implications disproportionate to the actual delay. Note that the expensive work i.e. the bitumen pavement layers, are undertaken at the end of the contract period. The price adjustment clause, which becomes generally operational after 18 months, can thus trigger a price evaluation of the more expensive part of the contract and lead to serious cost increases, and which is part of the works contract. Henceforth, it becomes even more pertinent to ensure timely implementation of the works contract. 60 UNRA should undertake a full scale review of the price adjustment clause. This review should address the following issues: (i) the limit of contract period before this clause should be applied, e.g. 18 months at present; (ii) a common set of price and cost indices that determine the overall price adjustments, acceptable to the contractors and UNRA alike, needs to be agreed upon; (iii) consider whether assistance to Bureau of Statistics is needed to improve the data collection and analysis of these construction indices to a standard that can be acceptable to international contractors; (iv) analyze and define exactly what proportion of the price adjustment should be considered as "cost overrun" and how much as part of the contractual price; and (v) Review the accuracy of Engineer's Estimates. This should include a check on how the budget is built up, what risk assessment and money for risk items is included etc. In addition, UNRA should also examine the benefits of smaller contracts to allow contracts to be completed in a reasonable time and reduce price adjustment issues. 34 61 Discrepancies between engineer's cost estimate prepared the design consultants and actual tender price creates problem downstream as funding is insufficient and can cause delays in project execution. While the lowest cost bid is typically selected as the winning bid under current procedures, still the winning bid is often higher than the engineer's cost estimate and the associated budget allocation for the project. The key sources for the difference in total costs are the imputed unit prices for the major items, such as bitumen, rocks etc., the assumed volumes needed of those items, the assumed profit margin and assumed project risks. It is recommended that UNRA develop a process for improving engineer's cost estimates by focusing on these factors, among others. 62 UNRA should monitor the imputed unit prices for the major items for all projects and provide all data to consultants. This is done by Road Authorities in many countries. In addition, design consultants should be required to prepare engineer's cost estimates that cover basic unit rates, sufficient allowance in the budget for price adjustment i.e. potential item inflation, identified project risks as all roads have various levels of financial risks which should be assessed, and allocation of risks, and consideration of exchange rate risk and its cost impact. 63 The assumed profit margin, even for the lowest bid can be high due to tacit collusion. UNRA can ensure increased competition between contractors by widening number of companies who bid. For example, UNRA can establish good links with contractor associations in a number of countries and give advance notice on future projects for which members of these contractor associations can prepare bids. 64 Unit prices for roads in Uganda have been rising. While often not fully anticipated in the engineer's cost estimates, depending on the specific case the increased costs could reflect genuine cost pressures as rising unit rates on major items internationally and not necessarily collusion.13 In addition to other measures proposed Government can: (i) review costs of materials and investigate potential of alternative road construction materials (based on local availability); (ii) review how GoU may help reduce costs by facilitating expanded domestic supply of inputs. For example, based on projected volumes of cement requirements, information could be provided to potential investors and/or cement companies about the prospective multi- year demand for major input for public construction works. Hence, improved knowledge by investors could facilitate investment decisions to open factories in Uganda. Over time as a benefit of government investment in improved transport links between Uganda and neighboring countries, the cost of importing materials into Uganda could decline as well. Monitoring and evaluation 65 Government has a weak data reporting system for measuring performance in the road sector. Some Key Performance Indicators (KPIs) are reported in the Ministerial Budget Statement, but these are fairly general in nature. Furthermore, many sector performance indicators desired by Government in order to more accurately assess performance cannot yet be 13 The increased costs are due to many reasons which are well documented in the "Workshop to Review Unit Costs of Road Construction in Uganda", February 10, 2009, Kampala. It is recommended that Government implements the action plan on reducing unit costs of road construction agreed at the workshop. 35 reported due to deficient or lacking data reporting systems. The Government should develop and implement as soon as possible actions plans to establish and run these data reporting systems, if needed with donor assistance. 66 The efficiency of UNRA's M&E department is impeded by lack of automation and efficient organization. The M&E department in UNRA needs further investment and support. In general the department is lean but is hindered by too limited space and too little provision for physical filing of documents. It is recommended that monitoring needs to be extended from simply recording progress to making predictions for future modeling and providing indicators of potential problems. The monitoring systems to be set up need to assist in identifying problems early on, predicting possible overruns at an early stage, so that pro-active action can be taken. The systems should use a red-flag system that enables attention to focused on possible problem areas. 67 The external audit approach is in line with good practice but its coverage needs to be extended. The Auditor General is primarily responsible for auditing of all government projects. These audits, usually carried out by a subcontracted firm of private auditors, focus on the financial administration of the projects. In the road sector financial and engineering audits are carried out for National Roads Maintenance on a regular basis i.e. typically every 3-4 years. These audits are performance orientated and provide an independent opinion from outside on processes of delivery of services, focusing on areas of weakness and suggestions on possible areas of improvement, including Value for Money, Resource Utilization, Planning and Budget Performance. 68 Financial and technical and/or value for money audits have not yet been carried out on development projects. It is therefore recommended that UNRA undertake external financial and technical audits on new development projects on a regular basis. Institutional capacity and technical assistance 69 With the ongoing, massive increase in project spending and need to mobilize new designs and works contracts in a short timescale, UNRA is clearly under-resourced, in particular on human resources. For example, UNRA urgently needs seven key experts to provide additional support during the design/procurement phases. These experts would cover the fields of transport economics, highways and/or hydrology, pavement and/or materials, structures, road safety, environmental and social, and contracts and are essential for the following reasons. Their work would reduce the risk of cost over-runs on projects due to expensive delays and claims. 70 To provide effective project management at the right time requires sufficient staff. It is far more cost effective to spend additional funds in improving the quality of designs through technical audits undertaken on all works contracts before they are tendered and improving the quality of the works contracts through the use of experienced contracts experts, internal review by a new UNRA contracts department in order to prepare improved tenders. In addition, the landtake process and other processes must be expedited before a works contract is signed in order to reduce potential claims. The additional support from the mentioned experts would be 36 essential in the work to ensure value for money and would reduce the real risk of not increasing absorption. 71 In order to cope with the increased workload and the need to undertake landtake on a number of roads in parallel, it is imperative that additional staff resources are provided to the Chief Land Surveyor. If not, appropriation will not be complete. In that case either projects are delayed or they go ahead and at a later stage GoU ends up paying substantial amounts to settle claims. 72 As is clear from previous sections, there is substantial need for technical assistance and capacity building to help UNRA meet both the challenging near-term and medium- term goals set out by Government. While it is an option to request development partners to finance these experts through technical assistance programs ­ which can be time consuming to set up ­ it is also important to consider the opportunity cost of not getting the experts timely. It is less time consuming if Government pays for the technical assistance from its own budget. The cost would be minimal when compared to the overall road budget and could result in substantial future savings on road projects. Some of the technical assistance needs have already been identified by UNRA and assistance requested from development partners or otherwise listed in other documents. Box 2.3. Additional need for TA provides an overview of TA not covered in other documents or through other projects. There are both short-term and medium-term needs for increased capacity. This can be provided through technical assistance and capacity building by development partners. Box 2.3: Additional technical assistance and training needs for UNRA Technical Assistance: A number of existing TAs are planned for UNRA. As a result of this review the need for Technical Assistance in the following areas is highlighted. The list follows the procurement cycle and is not prioritised. UNRA should prioritise the TAs in discussion with their Development Partners. Planning 1 ­ Consultant to develop prioritisation process and Donors to agree how they wish to be involved with development of the prioritisation process and ensure it meets GoU and Donor needs to avoid duplication (9 months) 2 - Provide support to determine total budget requirements on NEEDS approach, impact of availability of funds and options to manage the network (18 months) 3 ­ Development of budget costing system (6 months) 4 ­ UNRA to propose what additional Institutional support they need Design 1 - Analysis of project cost increases (develop a system to record meaningful data), Review of Price Adjustment Clause, development of improved Engineer's Estimates and training to UNRA and local consultants (6 months) 2 - Develop a system and implement Performance Monitoring of Consultants (18 months) 3 - Provide technical support in UNRA, at least 6 experts for possibly up to 2 years are required (Development Partners to request UNRA what they need in view of this report) 4 ­ Provide support in developing low cost road construction options setting up a soils laboratory to undertake and manage R&D (18 months) Design - Landtake 1 - Develop additional capacity in UNRA and local consultants (12 months) 2 - Develop additional capacity of local land valuer consultants (6 months) 3 - Develop additional capacity in Government Land Valuers and Land Titling Offices (12 37 months) 4 ­ Development of Land Acquisition Regulations to provide UNRA legal backing to the execution of land acquisition activities Procurement 1 - Support procurement process mapping and streamlining of the approval processes and provide training (6 months) 2 - Support development of MIS (12 months) 3 - Review award process for contractors and use of value driver selection process (3 months) Develop and implement performance monitoring for supervising consultants (18 months) 4 ­ UNRA is implementing a `Governance and Accountability Action Plan' (GAAP) under the World Bank financed Transport Sector Development Project (TSDP) intended to improve transparency and good governance in the road sector. Procurement Alternative Options 1 - Examine the relevance of alternative options, develop suitable system and provide training in their implementation (6 months) 2 ­ UNRA to hire a consultant under the TSDP to assist in the preparation of standard bidding documents for alternative project delivery Contract Management / Implementation ­ Contractual 1 - Provide contracts expert and set up contracts department (18 months) 2 - Develop and implement performance monitoring for supervising consultants (18 months) Contract Management / Implementation ­ Price Adjustment Clause 1 - Review PA clause for FIDIC and all donor contracts, examine if Uganda Statistics Bureau can provide data, revise clause Contract Management / Implementation ­ Engineer's Estimate and Unit rates 1 - Develop a method to record average unit rates for different road types to improve estimating at feasibility stage and to monitor cost of roads to GoU, - (12 months) 2 ­ Develop process and provide training in improving estimating of works contracts including risk assessment and management (6 months) Training: UNRA is in the process of developing a training programme for its staff based on training needs identified in its business plan. It is recommended that this training programme includes such elements as: · Day to day contract management · FIDIC and other Contract forms, contractual issues · General Project Management including systematic filing procedures, data management and reporting · Training should be focused more in UNRA, using specialist trainers to carry out training in-house, rather than to send staff overseas. This would reduce time staff are away from the job and make training better targeted to the specific needs in Uganda and hence more cost effective. 38 Capacity of the local road construction sector 73 Annual turnover of the Ugandan registered international companies is of the order of US$15 million to US$60 million only. The annual turnover of the Ugandan-owned companies is considerably less. Major road construction works are currently mostly carried out by major foreign-owned companies. A number of the largest construction companies in the country, mostly foreign owned, have been in the country generally for longer than the locally owned ones. Some foreign companies have been in Uganda since the 1940s whereas one of the oldest Ugandan companies, Babcon, was set up in 1984. 74 Local contractors need to strengthen their capability and capacity but will only invest if there is continuity of work. The challenge of unlocking private sector investment in capacity is two-fold: on one hand it is an information problem and on the other it is a competitiveness problem. Government can address the former by providing information to the Contractors' Association about its multi-year spending plans in the road sector. On the latter it is recommended that GoU works with the Contractors' Association and determines what actions can be taken that will help develop and promote a sustainable local contracting capacity. Options include that GoU investigates and agrees with donors on letting a larger number of smaller contracts that local contractors can bid for while maintaining strict supervision procedures. In addition it would be useful if donors could provide technical support to these contractors in areas such as quality control, planning and plant maintenance to ensure projects are completed to quality on time using best practice. 75 One key issue affecting local contractors is problems with working capital financing from the domestic financial sector. This limits the cash flow available to local contractors. In general, the local contractors do not have large turnover, and rely on a positive cash flow from the project to implement the works. All contracts include an advance payment, usually for 10% of the contract price. However, the contractor has to provide an Advance Payment Guarantee (APG) for the full amount of this advance. This is not difficult for international contractors to do from international banks. However, for local contractors this presents a significant problem, as the local banks are not familiar with providing such guarantees without further guarantees from the contractor. Hence, the contractor is often required to provide cash backing to obtain the required guarantee, thus reducing the benefit of obtaining an advance payment. The contractor must also provide a performance guarantee, usually for a further 10 percent of the contract price. Hence, he must tie up further cash backing in order to secure this guarantee. If there is any delay in submitting, approving and receiving payment for his interim certificates, the contractor will suffer severe cash flow constraints and the rate of work will be duly limited. C. CONCLUSIONS AND KEY RECOMMENDATIONS 76 Historically the road sector suffered from underfunding and poor management. However, the road sector has now become a number one priority for Government and there has been a quantum step up in funding in the last 2 years. Coupled with this, major policy changes have been implemented in recent years, with the establishment of UNRA on July 1, 2008 and the recent establishment of the Road Fund. Road sector projects always require a significant project 39 life cycle, counted in years, rather than months. It will therefore take some time before the real benefits of these major changes will start to be felt. 77 No major changes in policy and approach are required at this stage, but rather consolidation of the policies already being implemented and increased support to those agencies striving to move the sector forward and bring dramatic improvements in the standard of roads in Uganda. Primarily this is UNRA, but support is also needed to other major players in the sector, who have an impact on the successful and efficient implementation of projects. It is clear that, whilst UNRA has achieved a lot in its first year, it has also been consolidating the new procedures, new staff and setting up new systems. To help this process it is proposed that UNRA appoint Associate Directors in each Directorate as described below. 78 There are a number of areas where action can be taken to increase absorption and value for money in the sector by focusing on reducing risks and delays in the implementation of projects, and on increasing the efficiency of procedures. Projects are still taking too long to implement, and cost and time overruns are significantly higher than should be the case. The set of detailed recommendations discussed in this chapter, covering all stages of the project cycle, are summarized in Appendix I along with an analysis of the associated performance impediments and underlying causes. In this section a core set of strategic measures is put forward that focus on how to ensure that the set of detailed recommendations are effectively implemented. Short Term Strategic Recommendations 79 Establish a Dedicated Project Management Unit ­ Required by April 2010 to manage current designs and the tender of Works Contracts. Taking into account the need to focus on: · completing the design of current projects (1813 kms) and ensuring the designs are high quality (e.g. obtain feedback on issues from current projects and feed into design process) · completing the landtake process before Works Contract is signed, as such Government Land Valuer's Office needs additional support · finalizing and improving Works Contract Documents, where design is already complete, requires Contract Specialist · the need to procure new Works Contracts (planned start date on site end- 2010) at competitive rates · the need to rapidly "start up" works contracts as soon as possible · increasing the level of absorption · maintaining / increasing value for money 80 It is essential that UNRA create a dedicated team who has the responsibility and time to focus on current SMART14 tasks to be undertaken now to raise the level of absorption and maintain value for money, i.e., the team members have no other responsibilities and are not diverted. The team leader and his team have a lot to do if they are to 14 SMART stands for Specific, Measurable, Attainable, Realistic, and Timely. 40 overcome the many problems associated with the current procurement process in order to meet the short timescales while ensuring value for money is achieved. Prime tasks include:- · Develop an agreed fast track process for the current projects that are scheduled to start on site this year where the designs are not due to be complete until June 2010 while taking into account it currently takes a year (from when a design is complete) to bid a works contract and for work to start on site. The fast track process must be developed and agreed through dialogue with all partners (let others know quantity of work coming through the system) and a pro-active approach to identify areas where delays will occur and if they are not under UNRA control so inform GoU in order they help. · Request help from Development Partners (early provision of the 7 staff noted above to further assist UNRA in the development of quality designs) · Request other Stakeholders to provide shorter approval periods as required and that approvals may be requested in parallel rather than in series (currently all parties are required to approve one after the other) · Review the detailed recommendations of this report and other reports in developing a short term plan that only includes tasks necessary to achieve the agreed immediate goal, increase absorption and maintain value for money · Undertake a technical audit on all projects; ensure more contractors bid for the new projects. 81 The above approach requires support from GoU as the problems are not all within UNRA, e.g., Government Land Valuer's Office is seriously under staffed which means there is a high risk landtake is not complete before a works contract starts. Many stakeholders have an impact on the procurement process and can help reduce the timescale but only if GoU takes leadership and also provides them with additional resources. The role of the Inter-Ministerial Technical Committee on the Road sector (see below) is instrumental in this regard to ensure cooperation across government agencies. In addition to the above, it is essential to implement other recommendations which as a minimum include: · Establish a Contracts Department to provide support to Planning and Projects Directorates (revise the price adjustment clause, consider extending DLP to 24 months). · The need to undertake external Financial and Technical Audits on construction works on a regular basis before the works contract is completed (for instance at 50% complete and 100% complete) to ensure quality is maintained and resolved any problems before the contract ends. · UNRA does not wait for the new MIS to be developed. By April 2010 UNRA prepares a detailed procurement program (if not already done) for increasing absorption and sets up a simple "tracking system" that monitors the progress of all requests for approval during all stages of the procurement process. In addition UNRA meets with stakeholders on a monthly basis to discuss issues met, e.g. delay in approvals of the fast track projects. · UNRA to also review the capacity of all departments involved in this fast track process to cope with the increased workload and consider need for increased numbers over that fixed by the wage bill, including urgent need for increased workspace. To consider taking on temporary staff to help overcome expanding the service level up to that 41 required to cope with the increased budget levels. The effort required should not be underestimated as the aim is to accelerate the process. · Variation Orders are issued by the Road Authority and copied to all stakeholders for information (within given magnitudes it will not be necessary to seek SG's approval) 82 Inter-Ministerial Technical Committee on the Road sector (IMTC). A number of issues affecting road sector performance are cross-cutting or originate outside the purview of UNRA and even MoWT. It will be critical that the Inter-Ministerial Technical Committee gets the necessary authority across agencies to move actions where necessary. It remains important that the final recommendations be approved by the IMTC and that it assists in realizing them. In addition to the detailed set of actions in Appendix G these also include at a higher level: (i) Autonomy of the Road Fund so that it operates as envisaged, (ii) Autonomy and continued development of UNRA so that it operates as envisaged, (iii) Ensuring that the spirit of the National Construction Industry Policy is enforced, including strengthening the local construction industry, and (iv) Current Transport Sector Review. Medium Term Strategic Recommendations 83 National Land Acquisition Program for the Road Reserve. Consideration should be given to establishing a national program of land acquisition to secure the road reserves for all National Roads. Currently this is only done as part of rehabilitation projects, and is not initiated until the project is given the go ahead for construction. If the road reserve for all National Roads was secured, then further acquisition for individual projects would be minimized. 84 UNRA is faced with issues related to the project preparation and approval processes in Government outside its control. It is recommended that UNRA appoints a senior person to be made responsible for resolving the external problems. · In the first instance examine issues that limit UNRA's ability to cope with the increased budget, propose improvements and ensure they are instigated. This could for instance include encouraging other government units/agencies to speed up their approval timeframes and to agree that approvals are run in parallel, variation orders are approved by UNRA etc. · In the second instance examine longer term issues as noted in the preceding summary and in detail in the report e.g. how to improve value for money and reduce unit costs. 85 With regard to improving its internal systems, UNRA has received a lot of attention and advice in recent months with recommendations made under RSDP, by World Bank and other Development Partners' investigations as well as from ongoing audits. It is recommended that UNRA considers the issues raised and prioritises them into one list (backed up by more detail) focusing in the first instance on achieving the required level of absorption and delivering value for money through setting SMART objectives. 86 To this end it is recommended UNRA focus on examining internal problems in each Directorate using an Associate Director (new position) with the authority and time to improve internal efficiency. The Associate Directors should propose how improvements may 42 be achieved, agree actions with the Director in each Directorate, spearhead obtaining support from Development Partners where applicable and be responsible, and ensure improvements are implemented in an agreed timescale. They should also form a sub-committee to discuss cross cutting issues and keep each other informed. In addition they should make use of external consultants to help them structure the interventions so they may focus on implementing them. Development Partners could assist by accelerating the provision of help that in the immediate future provides the support UNRA requires e.g. provide the staff requested. 87 It is of critical importance to realize that technical assistance will be required to take up the recommendations put forward in this report. Again Box 2.3 provides a useful overview of what is required. 43 Appendix I: Summary of Analysis of Chapter 2: Value for Money and Absorption Constraints in Procuring and Implementing Road Contracts Description and Assessment Current Practice Adequacy Against Analysis Main Recommendations Current UNRA Plans And Best Practice Proposed Donor Assistance Planning In general, Analysis shows HDM has been used for many years and UNRA PLANS approach currently resulted in plans that include National roads · HDM used to prioritise UNRA intend to undertake planning considered appropriate roads for development as such selected on basis of · RSDP 1 and 2 not exercise on network and develop a satisfactory as there is clearly a planning process in place. overall network completed because of Road Network Plan and further HDM is used to There is though always room for connectivity lack of GoU counterpart develop in house capacity, this prioritise improvement. All projects funding includes updating RSDP assessed using Problems may come when decisions have to UNRA are developing a roads · UNRA now responsible multi criteria for 20,000 km of roads be made on order of development of the databank and asset management analysis including (100 % increase), it is secondary roads and which District roads system, tools for preparing and HDM not clear if UNRA's to upgrade requiring a road investment updating annual plans budget has been prioritisation framework as proposed by UNRA follows UNRA intends to extend the road increased accordingly MoFPED. It is essential donors be involved MoFPED inventory/ condition and spatial in its development as it may be used as the guidelines in · Project budget estimates referencing surveys to the new need improving "Gatekeeper". preparing project 10,000 kms. profiles for A high risk is that GoU do not allocate · Requirements of Road UNRA intends to establish a Cost consideration of Fund must be sufficient funds to develop and maintain Estimating Unit which will also Public Investment considered roads on a NEEDS basis. UNRA must implement or use the World Bank Plan examine the realistic cost of maintaining Road Costs Knowledge System the national road network and the impact GOU has A ­ All projects reviewed (ROCKS) on the network if funds are not available. formulated included HDM analysis, a UNRA plans to undertake an In addition to improve the process of Transport standard process to determine institutional diagnostic study that selecting and prioritizing roads in order to Masterplan RSDP1 the Internal Rate of Return will determine capacity gaps determine realistic budgets in 1996, RSDP 2 with regard to the investment. affecting institutional efficiency and under mid- term A - MoFPED are preparing a ToR for a Road Further prioritisation criteria mandate implementation review in 2001/02 Investment Prioritisation Framework Study. will be required especially UNRA and Donors to be consulted about ToR RSDP3 developed when looking at secondary and involved in development of process, in for 2009/10 to roads particular how it may be used to manage Proposed Donor 2018/19 includes B ­ Most projects under the development funding. This will have to Assistance(Duration) National Roads, RSDP 1 and 2 were left encompass the development of a framework District etc undone for 10-15 years in part 1 ­ Consultant to develop or criteria for prioritization and selection prioritisation process and Donors to 44 Description and Assessment Current Practice Adequacy Against Analysis Main Recommendations Current UNRA Plans And Best Practice Proposed Donor Assistance Current investment due to lack of counterpart community access and District roads for agree how they wish to be involved strategy focuses on funds. upgrading. with development of the Northern Corridor B - 10,000 kms of District B - UNRA to review the total budget required prioritisation process and ensure it Roads have recently been to develop and maintain their expanded road meets GoU and Donor needs to Donors have their upgraded to National Road network of 20,000 kms on a NEEDS approach avoid duplication (9 months) own priorities and pick from proposed status and handed over to not only a budget approach; discuss 2 - Provide support to determine priority list UNRA who now have implications with MoFPED and Donors. Then total budget requirements on responsibility to maintain produce a multi-year programme to involve NEEDS approach, impact of them in addition to the 10,000 appropriate stakeholders such as horticulture, availability of funds and options to kms they currently manage. tourism and agriculture, especially important manage the network (18 months) UNRA developing B - Considering the increase for District roads. In addition when the yearly 3 ­ Support development of budget roads in network size and increase in budget is confirmed and insufficient, UNRA costing system (6 months) database and asset length of asphalt concrete are to advise GoU of the reduction in asset 4 ­ Ask UNRA what Institutional system roads (may cost more to value due to insufficient maintenance. support they need maintain), the sufficiency of B - UNRA to undertake an exercise to future budgets for new works balance/prioritise Uganda Road network and maintenance are not clear against available funding. It is imperative the and may be totally insufficient issue of insufficient maintenance leading to to preserve asset value. the "new build followed by road failure cycle" C - Works contracts estimates be avoided, if maintenance finding is not (which are used to develop available do not rebuild the road. Exercise will UNRA road budget) are involve determine optimal network size, core sometimes well below bid network and likely resulting impacts on prices and final contract costs network performance of the lack of funding in (which include claims and the future based on past experiences. payments for price C ­ UNRA to implement an improved budget adjustnment). This impacts on costing system that considers Base Cost, the adequacy of funding Variation of Price and Project Risks plus D ­ In July 2009 the road fund appropriate contingency level. To include was established. RF no doubt preparation of a manual and training for will wish to develop a process UNRA and Consultants for prioritising work , it is D ­ MoFPD, MoWT, Road Fund, Donors and important this is undertaken in UNRA should agree a road investment and conjunction with MoWT, maintenance prioritisation process that meets 45 Description and Assessment Current Practice Adequacy Against Analysis Main Recommendations Current UNRA Plans And Best Practice Proposed Donor Assistance MoFPD and UNRA so there is all their needs and avoid duplication and not a third approach. wasted effort E- GoU ramped up the road E ­ To cope with increased size of network budget by a factor of 13 last and budget whilst ensuring value for money year and doubled the network plus implement above, temporary staff will be UNRA is responsible for. required as well as Institutional support. This does not appear to be recognised by GoU. In order to avoid low absorption UNRA to examine staffing levels and expand on a temporary basis as required. UNRA to request what it needs from GoU and Donors who must both respond QUICKLY if the desired level of absorption is to be achieved. It is unrealistic to expect UNRA with its present capacity to deliver this dramatic increase in workload without more resource (staff and office space). Design The approach Analysis shows UNRA to develop realistic timescales for UNRA Plans follows best design projects or risk they will be inferior. Feasibility Studies · Insufficient time allowed UNRA have initiated design of practise, the Further improve the design ToR's, include are undertaken for design 1,800 kms of network in order to problem is the all Site Investigation as a Provisional Sum. including economic · Inadequate site "have construction projects standard of works Increase range of experts and internal analysis using investigation available to meet increased budget" documentation staffing to allow closer management of HDM UNRA have tendered for a (Tender for · Costs of roadworks are design process and ensure issue identified Preliminary designs Contractor) increasing, (value for Technical Audit consultant to audit are not repeated. It is imperative to are undertaken to developed by money falling) in part designs in an effort to ensure the produce high quality designs "Correct First determine optimum design consultants due to Price Adjustment standard of Design and hence Time". terrain alignment is not up to and poor design Works Contracts are improved and cost Monitor cost increases, problems in standard and must There will be a need for low UNRA intend to extend an existing applying Price Adjustment Clause and Detailed designs be improved to cost road construction consultancy contract to collect data increase the level of competition for Works are undertaken to reduce the options for District roads on the District roads recently Contracts produce contract problems especially handed to UNRA so planning may documents and experienced on site The aim must be to employ "better" proceed A - Insufficient time allowed Engineers Estimate with subsequent consultants, as such GoU to develop for design phase results in claims and Performance Monitoring System. poor design and subsequent Proposed Donor 46 Description and Assessment Current Practice Adequacy Against Analysis Main Recommendations Current UNRA Plans And Best Practice Proposed Donor Assistance corresponding delays in the completion of With the upgrading of 10,000 kms of Assistance(duration) reduction in value construction of the road District roads to national roads status, 1 - Analysis of project cost for money resulting in an overall increase UNRA to examine ways to build low cost increases (develop a system to in time to deliver the finished roads appropriate to level of service record meaningful data), Review of road. Get the design right, envisaged. Price Adjustment Clause, then tender. development of improved A ­ UNRA to initiate the start of design B ­ Clearly inadequate site process for future roads earlier to provide for Engineer's Estimates and training to investigation has resulted in adequate design period UNRA and local consultants (6 poor design and subsequent months) B ­ UNRA to include in the ToR's for design, claims from contractors 2 - Develop a system and implement adequate Provisional Sum items for all site C ­ Roadwork costs are investigation work to ensure sufficient site Performance Monitoring of increasing in Uganda and in investigation is undertaken Consultants (18 months) the region. 3 - Provide technical support in C ­ UNRA to undertake further investigation D - Price adjustment clause on an ongoing basis into cost increases (refer UNRA, at least 6 experts for provides for contractors to to recommendations made in the Action Plan possibly up to 2 years are required claim back increased costs of Matrix derived at the Unit Cost Workshop of (Development Partners to request materials as per clause in February 2009. UNRA what they need in view of conditions of contract in order this report) and compare against regional costs in order to to share this risk with the 4 ­ Provide support in developing ensure Value for Money is being achieved and employer- some contractors low cost road construction options need to increase level of competition for aim to manipulate this clause setting up a soils laboratory to Works Contracts to their benefit. To block this undertake and manage R&D (18 practise will require careful D ­ UNRA to investigate the various price months) examination and changes to adjustment clauses and how they can be contract documents improved E ­ GoU to implement a Performance Monitoring system for design consultants in E - Poor performance (design) an effort to provide structured feedback to by consultants results in cost consultants on their performance. Use the overruns. Primary examples data in the procurement process for new include errors in quantities, design consultancies so that better consultants change to pavement, hurried are rewarded with new work and poor design reviews, unforeseen performing consultants do not receive work ground conditions, addition to scope. F ­ UNRA to implement a Technical Audit 47 Description and Assessment Current Practice Adequacy Against Analysis Main Recommendations Current UNRA Plans And Best Practice Proposed Donor Assistance F - Errors in design in part due Process to be undertaken during the design to poor consultants and lack of process to ensure timely feedback to guidance by UNRA consultants as well as improved quality of designs and Works Contracts G - There is a lack of feedback of site issues to the designers F ­ UNRA to improve internal capacity to so they may improve the pro-actively guide the design process e.g. on contract documents traffic growth rates and to also undertake H ­ UNRA are currently technical audits. Development Partners to examining options to build provide additional support to UNRA to lower costs roads, not clear if improve staffing to requested levels as soon the options are suitable for the as possible. 10,000 kms of District roads G ­ UNRA to develop a process that provides (recently transferred to feedback on site issues to Planning Directorate UNRA) which are to be who must then feed issues to Design upgraded, but to what level Consultants to ensure continued improvement service? H ­ UNRA to undertake further investigation into the use of marginal road construction materials and construction of low cost sealed roads including application to the National Network in particular the "new" District roads to be upgraded. UNRA to examine need for in house soils laboratory to undertake R&D. Review advances made by AFCAN Design Best practice Analysis shows that Land UNRA to quantify for next five years the UNRA PLANS Land Acquisition requires all land for Acquisition procedures are extent of landtake works and cost UNRA intend to expand their land project to be not completed before Works (consultant input and compensation acquisition department and The majority of procured before a Contracts start. This is due payments for landtake) and examine distribute elements of the landtake road works include project starts, GoU to lack of capacity of the capacity of all stakeholders, thereafter manual to design consultants so the upgrading of should follow this various Government propose to GoU how capacity to be they improve their process roads to a higher Policy. Departments and lack of improved and agree how funds will be standard that UNRA now require their design In cases where land funding which results in released early enough to ensure normally includes consultants to undertake a more has not been delays in paying compensation payments are made prior to an increase in thorough process in identifying land acquired and it is compensation resulting the start of a Works Contract. width to carry and buildings affected as well as essential a project obstructions in the ROW A ­ UNRA to determine how they can higher levels of determination of the compensation. 48 Description and Assessment Current Practice Adequacy Against Analysis Main Recommendations Current UNRA Plans And Best Practice Proposed Donor Assistance traffic. In brief it must proceed, the when the Works Contract complete land take prior to signing a Works requires some contract documents starts. Contract Proposed Donor Assistance- additional land. must reflect this so B - UNRA to prepare a paper indicating the (duration) Utility companies the contractor is likely extent of landtake works required and A - In the past landtake 1 - Develop additional capacity in rely on UNRA to fully aware he has the cost of compensation (possibly use data (acquisition of land including, UNRA and local consultants (12 have acquired the to work around the from completed to date) in order to quantify surveys, identification of months) Right of Way problem and as the problem area and develop mitigation plans owner, payment of (ROW) so they such may not claim e.g. 1 how best to improve the capacity of all 2 - Develop additional capacity of compensation and registration may install their PPDA law is very stakeholders to deliver the process faster and local land valuer consultants (6 of the land in the name of services detailed, 2, To determine if for certain projects, land months) UNRA) has started when a prescriptive and works contract starts, far too purchase can be completed before the Works 3 - Develop additional capacity in provides a process late resulting in greatly Contract is let e.g. Donor funded projects Government Land Valuers and Land to tender projects. increased risk of claims from where funding is assured and timetable more Titling Offices (12 months) contractor for non access certain 4 ­ Development of Land B ­ UNRA to consider if they need to further Acquisition Regulations to provide A - It is understood UNRA increase their internal capacity UNRA legal backing to the intend to continue to follow execution of land acquisition the same approach of paying B ­ UNRA to consider how capacity within activities compensation for land when the Government Land Valuers and Land the Works Contract is Titling Office can be increased tendered, in order to use the B ­ UNRA to consider how capacity in the limited finds wisely (projects number and capability of the land Valuer may be cancelled) consultants can be increased B - Acquisition of land can B - UNRA to determine if it is possible and take from 6 to 15 months practical for the design consultants to (longer if owner objects) undertake more of the Land Take process at depending if the land is design stage than currently envisaged registered or not registered, owners are in the country etc B - The process can take longer if the Government Chief Valuer is unable due to lack of internal resources review the proposed valuations in a reasonable 49 Description and Assessment Current Practice Adequacy Against Analysis Main Recommendations Current UNRA Plans And Best Practice Proposed Donor Assistance timeframe B - Another limitation is the limited capacity of the private companies who are registered to undertake land valuations Procurement This detailed The current procurement The National Transport Plan clause 6.2.3, General process is not process takes too long notes extended procurement cycles as a considered best because all stakeholders problem citing 10 years as against the Open competitive practice when (including Donors) do not potential 5 years. bidding is used in compared to process approvals in a The TSDP November 20009 notes problems line with PPDA simpler processes timely manner. The cost to with Procurement and includes an agreed law using re- in use in other GoU is high (EOT on a Action Plan to mitigate Procurement Risks. measurable countries. works contract can typically Neither go far enough as they do not contracts UNRA have an cost between 1 to 2% of include all the stakeholders nor promote Contracts are operational PDU in Contract Value per month) simplifying the onerous process. awarded on basis of line with PPDA and greatly impacts on "Lowest Evaluated UNRA to examine the impact and cost of law Value for Money. Bid Wins" the delays in procurement (the processing In addition improved of new contracts and changes to existing Governance is required to contracts) and make all stakeholders aware reduce the impact of of how this greatly impacts on Value for Political interference and Money and absorption capacity i.e. the collusion. longer it takes to procure contracts the lower value for money and lower absorption capacity. UNRA to determine how improvements may be achieved within UNRA, simplify the process, implement targets (e.g. process A to be completed all request within five days first time) for PDU and Contracts Committee A - Various stakeholders Develop agreed guidelines (timescale for UNRA PLANS appear to have different views stakeholders to process requests) with other UNRA have had some discussion on the extent and complexity stakeholders. GoU to support this initiative with regard to PPDA law but of the required administration and not block it. without any improvement 50 Description and Assessment Current Practice Adequacy Against Analysis Main Recommendations Current UNRA Plans And Best Practice Proposed Donor Assistance process to record and Although MIS database is planned it will UNRA intend to develop a MIS, implement the PPDA laws. take time to develop. In the meantime this should have the capacity to B ­ Some stakeholders take UNRA should set up a process to monitor monitor a project through the full long periods to process all approvals, any beyond the allocated time cycle, record relevant data and documents and do not always to be followed up, this will require a produce output data in a meaningful make it clear what information dedicated person. format they require or notify others if A - UNRA to develop a bespoke database to data is missing improve monitoring of procurement process (time each stakeholder takes to process) for all projects and changes to contracts in order to identify and quantify bottlenecks. B - UNRA to discuss and agree with other Ministries and other stakeholders how the procurement process can be simplified, streamlined and made more user friendly. Produce agreed guidelines on how best to manage the high volume of bids and initiate "joint" training so different Ministries understand the problems others face and the consequence of delays e.g. expensive claims from contractors, lower value for money. A ­ The lack of appreciation A ­ UNRA to develop closer ties with PPDA Proposed Donor Assistance- that any delay in the and Solicitor General and meet on a regular (Duration) "procurement process", costs basis to discuss problems and issues 1 - Support procurement process money is endemic with B - UNRA to simplify contract management mapping and streamlining of the departments blaming each of Works Contracts e.g. issue of VO's. approval processes and provide other. All delays whether on Consider using engineers only in the process training (6 months) new projects (more money of assessing engineering projects 2 - Support development of MIS (12 will be paid out under the C ­ UNRA to agree how to achieve months) price adjustment clause) or accreditation under PPDA law and how it will 3 - Review award process for existing projects (EOT with help UNRA contractors and use of value driver costs can typically cost between 1 to 2% of Contract D ­ UNRA to review their internal capacity selection process (3 months) Value per month) result in not only in PDU but in other departments and 4 ­ UNRA is implementing a consider increasing level of resource to allow 51 Description and Assessment Current Practice Adequacy Against Analysis Main Recommendations Current UNRA Plans And Best Practice Proposed Donor Assistance cost increase and reduced processing of the documents in a timely `Governance and Accountability value for money. In essence manner Action Plan' (GAAP) under the service delivery is delayed by E - Examine other options of selecting World Bank financed Transport any delays occasioned by contractor and not just "lowest evaluated bid" Sector Development Project (TSDP) procurement. e.g. investigate value driver selection process intended to improve transparency plus taking into account the past performance and good governance in the road of the contractor to ensure the "better" sector. B - UNRA have to process a large volume of tenders, contractors are employed contracts and changes to F ­ UNRA to ensure there is no collusion. contracts e.g. VO's. which One approach is to advertise contracts early requires a lot of manpower to and widely to ensure maximum competition process C - It is understood that the proposed revisions to PPDA law are targeted to other areas and do not help improve the procurement challenges of large turnover organisations like UNRA. D - UNRA has insufficient office space, staffing, a poor archive system and referencing of correspondence, each letter must have a unique reference E- "Lowest evaluated bid wins" approach does not always result in appropriate contractor for a project. F ­ There is concern that Good Governance is not always employed, that collusion and Political interference leads to 52 Description and Assessment Current Practice Adequacy Against Analysis Main Recommendations Current UNRA Plans And Best Practice Proposed Donor Assistance appointment of inappropriate contractors and consultants Procurement Less complex A - The administrative process A - UNRA to not only look at the process of UNRA PLANS approaches are within UNRA has been set up dealing with approvals from Stakeholders (see PDU in UNRA UNRA to conduct study tours to used in other to comply with PPDA rules. above), but must :- Under PPDA rules, other countries with more efficient countries yet It appears overly complex and UNRA have a PDU · review their internal capacity and need to procurement systems with a view of provide a onerous absorbing a high unit that operates increase office space and staffing borrowing best practices transparent volume of effort that could be within UNRA and (possibly only temporary in order to ramp UNRA to implement a Procurement procurement better used elsewhere in up capacity) and make a case to GoU manages the Improvement Action Plan proposed process for high managing design projects and procurement of all · review internal process and agree with under the TSDP intended to volume works contracts contracts, this PPDA how the process may be simplified improve procurement effectiveness organisations includes services · determine from PPDA how to obtain (design accreditation with PPDA and implement consultancies), works contracts and · make contracts committees work more efficiently and process request first time supplies by setting targets · set up a process to prioritise / fast track projects, run approval processes in parallel agree with PPDA additional internal training for all staff Procurement Currently these PPDA do not have standard UNRA should be aware of the complications Proposed Donor Assistance- Long approaches are not documents for these of entering such new territory and should not Term Alternative used but are being procurement options underestimate them. If it is decided to follow Options 1 - Examine the relevance of considered a new route obtain advice on how to set up Early Contractor alternative options, develop suitable and build a system. Also consider option of system and provide training in their Involvement in paying bonuses to contractors for early implementation (6 months) Design (consultant completion. 2 ­ UNRA to hire a consultant and contractor identify simpler With regard to PPP, it is also recommended under the TSDP to assist in the ways to achieve that for each project it be determined if it is preparation of standard bidding goal and hence more cost effective for the Government to documents for alternative project improve value for borrow the funds and use the normal form of delivery methods money), Design, procurement 53 Description and Assessment Current Practice Adequacy Against Analysis Main Recommendations Current UNRA Plans And Best Practice Proposed Donor Assistance Build and Operate (DBO), PPP Procurement of Approach is As a result of PPDA laws and UNRA to discuss with Donors the advantages Consultants acceptable and duration it takes to procure and disadvantages of using the same varies in different services, the current approach consultant for design and supervisions and Procurement countries leads to 3 periods of how the contract would need to be revised (Selection) of procurement of 9 months each (Overriding aim should be procure better Consultants is which extends the planning consultants) undertaken for each and design period thus It is noted UNRA have moved to 90:10 stage delaying service delivery by as criteria in evaluating consultants in order to A detailed much procure improved services evaluation process is used by UNRA to select a consultant as stated in the RFP Contract The split of Analysis shows that there is UNRA to UNRA PLANS Management / responsibility is reduced Value for Money · Ensure decisions are made in a timely UNRA are preparing an internal Implementation / considered (overspend) due to manner project management manual General acceptable. · Delays in making · Provide structured feedback on UNRA has two One downside is decisions resulting in problems to Planning Directorate directorates the problem of claims feedback on site · Implement further training in FIDIC involved with · Lack of feedback of site problems and in Uganda in UNRA on an ongoing procurement of and issues to Planning basis using real examples implementation of mitigation in future Directorate development designs · Implement Performance Monitoring of · Lack of understanding Supervising consultants and feedback projects of FIDIC and issues during the project so the UNRA's Planning implications on project consultant may improve, note GoU to Directorate · Weak Supervising develop system procures and Consultants · Investigate extending DLP and agree manages the design · Poor contractors with Donors process, production delivering poor · Agree with PPDA and Solicitor and tender of workmanship General to take Contract Management Works Contracts. UNRA's Projects · Delays in issuing VO's out of the realms of procurement. As 54 Description and Assessment Current Practice Adequacy Against Analysis Main Recommendations Current UNRA Plans And Best Practice Proposed Donor Assistance Directorate are · Complicated Price such UNRA may approve VO's and involved in tender Adjustment Clauses copy stakeholders for information only evaluation and · Simplify the Price adjustment Clause · Poor determination of thereafter project budgets · Improve development of budgets manage the implementation. · M&E unit has no · Increase capacity of internal M&E unit systems and has low · External audits to be undertaken on capacity development projects · No external audits on development projects A - UNRA on an ongoing basis to review their contract management process in order to speed up their internal approval processes and examine how increased value for money may A ­ UNRA has delayed be achieved. making some decisions due to B - UNRA to set up a system to feed back site lack of understanding of the problems in a structured manner to Planning issues raised by a contractor Directorate so Works Contracts may be with subsequent impact on the improved. progress of the Works C - UNRA to instruct supervising consultants B - Structured information is on an ongoing basis to feed back problems not fed back to Planning experienced on site and propose solutions in Directorate on problems order that Planning Directorate may improve experienced on site the Works Contracts Contract FIDIC is a well A - As with all contracts, A - UNRA to establish a FIDIC Contracts UNRA PLANS Management / known conditions FIDIC must be carefully Department employing contract specialists Training in FIDIC conditions of Implementation of Contract and implemented with appropriate with extensive experience in highway contract to be included in UNRA's Contractual provides a sound letters and instructions issued contracts (require the technical overview to Annual Training Plan. base at various stages. UNRA staff analyse claims) whose brief includes International have a good understanding of providing support to UNRA project managers The FIDIC Consultants the requirements but when it on site on issues arising from Works Contracts conditions of supervising the comes to claims, depending on and claims. In addition to review any changes Proposed Donor Contract are used Works should the complexity are not aware to the Conditions of Contract proposed by Assistance(Duration) for Works provide the level of of the full implications Design Consultants, to monitor all VO's and 1 - Provide contracts expert and set Contracts ­ for expertise required Claims and feedback proposals to Planning B - Some International up contracts department (18 55 Description and Assessment Current Practice Adequacy Against Analysis Main Recommendations Current UNRA Plans And Best Practice Proposed Donor Assistance World Bank this Consultants are not pro-active (who manage the design consultants) on how months) applied to in identifying problems and they may be mitigated. 2 - Develop and implement Contracts > 10 resolving them, nor managing A - UNRA to increase the level of internal performance monitoring for MUSD and advising on contractual training on FIDIC with courses run in UNRA supervising consultants (18 months) International issues. by outside experts, making use of examples of Supervising C ­ Some projects have failed problems UNRA face rather than send staff on Consultants very early on in their design overseas courses which by their nature are Supervise the major life but outside of the DFL general projects period, hence the contractor is B ­ GoU to implement Performance Currently the not responsible monitoring of Supervision Consultants to defects liability include their capacity and capability on period is 12 months dealing with contractual issues, feedback during the project in order that their past performance impacts on whether they are awarded new supervision contracts or not C ­ UNRA to examine options of extending the DLP to 24 months and/or letting contracts that include a five year maintenance period. There are many issue with both proposals which must be examined carefully and agreed with Donor Contract In many other A - Analysis shows that this A - UNRA to discuss the issue with PPDA UNRA PLANS Management / countries VO's are onerous process can take a and Solicitor General and propose a process Looking in to how to speed up the Implementation ­ processed by the long time with records that allows entities such as UNRA to approve process Road Authority up showing some VO's took 7 VO's up to an agreed aggregate value of the Variation Orders to a certain value months to process. The delay contract value in a transparent manner that Variation Orders where upon they may include copying all VO's to Solicitor in processing VO's can result (change to works obtain approval General for information in claims from the Contractor contract) on current from other for late notification of change contracts are authorities, the to works where the claim prepared by process is would be for extension of time Supervising transparent and (EOT) with costs Consultant, much faster B - Taking into consideration submitted to there are various audit teams UNRA Contracts 56 Description and Assessment Current Practice Adequacy Against Analysis Main Recommendations Current UNRA Plans And Best Practice Proposed Donor Assistance Committee, then (UNRA and MoFPED) who submitted to also look at projects, the Solicitor General as current PPDA and Solicitor it is a change in General approval process for contract. In VO's appears unnecessary addition for EU contract the VO is finally submitted to National Authorising Officer. If value greater than 15% per item or 25% of the overall contract value, the VO however small to be submitted to PPDA Contract This is in line with A - It is not clear if all UNRA A - UNRA to undertake a full scale review of UNRA PLANS Management / good practise in project budgets include a PA PA costs to date on various projects and Not known Implementation ­ order to share the element to allow for the investigate if - 1 the current Price Adjustment risk of increases in expected increase in project Clause is fair and reasonable and is being used Price Adjustment materials and cost above tendered amount as intended i.e. to share some of the risk on The majority of changes in oil (which does not include the increase of cost of materials between the client Contracts with prices etc PA) The impact of PA should and contractor and 2 if the clause is too Proposed Donor Assistance- (3 durations in excess not be underestimated complicated to administer with goods being months) of 18 months A - Price increases from this imported now from many countries 1 - Review PA clause for FIDIC and include a Price Adjustment (PA) adjustment that occur during A ­ UNRA to present findings to Donors, all donor contracts, examine if Clause based on periods of EOT add to the consultants and contractors, obtain feedback Uganda Statistics Bureau can FIDIC Conditions overall cost increase and and alter clause. provide data, revise clause of Contract become a secondary cost if A - UNRA to make all stakeholders more EU Contracts EOT is awarded aware that any delays to the start of a project include a similar A - Some contractors appear will result in increased costs as the contract to manipulate the clause in 57 Description and Assessment Current Practice Adequacy Against Analysis Main Recommendations Current UNRA Plans And Best Practice Proposed Donor Assistance clause based on EU order to maximise profit, as provides for paying increased cost of materials contract such the clause requires based the price of goods at tender stage updating to take account of A ­ UNRA to examine benefits of smaller this problem and procurement contracts to allow contracts to be completed in of goods from many countries a reasonable time and reduce PA issues Contract Unit Rates for Although unit rate costs per Increased costs in road works is not unique to UNRA PLANS Management / Roads are rising in km were investigated the Uganda, many countries are effected UNRA in their new design ToR Implementation ­ many countries as results are not conclusive. To mitigate the impact UNRA must analyse in have included a requirement that the commodity and oil Engineer's The data available was not more detail project out-turn costs and break Engineer's Estimate consider more prices rise plus as Estimate and Unit broken down sufficiently as down the costs as listed in order to understand factors as noted the volume of road Rates listed below to allow why prices increase and how to combat the works increases. Unit Prices for determination of what problem. roads in Uganda proportion of the outturn cost A - In order to keep costs down are rising, increases are a result of · Increase competition between contractors Engineer's A ­ increased costs Proposed Donor Assistance- by widening number of companies who estimates Need to (Duration) · Lack of competition bid be more accurate 1 - Develop a method to record · Increase in cost of · Review costs of materials, investigate average unit rates for different road materials potential of alternative road construction materials (based on local availability) and types to improve estimating at B ­ Poor Budget Estimates feasibility stage and to monitor cost how the GoU may help reduce costs e.g. · Poor budget estimates based on projected volumes of cement of roads to GoU, (tender price higher than requirements, encourage cement (12 months) Engineer's Estimate) companies to open factories in Uganda 2 ­ Develop process and provide · Insufficient allowance in · Reduce costs of importing materials into training in improving estimating of budget for Price Uganda e.g. open up more transport works contracts including risk Adjustment routes assessment and management (6 · Lack of allowance for B - In order to improve budget estimates months) project risks (all roads have various levels of · UNRA to monitor unit rates for all financial risks which projects and provide all data to should be assessed) consultants Impact of exchange rate Design Consultants to prepare Engineer's changes Estimates that cover basic unit rates, potential 58 Description and Assessment Current Practice Adequacy Against Analysis Main Recommendations Current UNRA Plans And Best Practice Proposed Donor Assistance inflation (PA), identified risks and allocation of risks, consideration of exchange rate risk Contract M&E is in place A - The M&E unit in UNRA A - UNRA to increase the resources of the UNRA PLANS Management / which is an does not have systems and has M&E unit and further develop the capability UNRA are reviewing KPI's Implementation ­ important start too low a capacity of the unit UNRA intend to let consultancies to Monitoring & The limited An M&E unit should be able B ­ Develop a MIS that can record all the develop overall M&E Policy, M&E Evaluation capacity of the unit to make predictions and red KPI's and produce the reports for each framework, M&E Plan plus an does not provide flag problem areas and then Stakeholder to now include Road Fund Monitoring and integrated MIS the manpower to follow up to see they are Evaluation within also make addressed UNRA is predictions for B ­ The different requirements undertaken by a future modelling, and range of KPI's GoU, small team in the dig down into Donors and UNRA use results Planning problem areas in the need to write different Directorate mainly concerned with reports for different racking progress, stakeholders reviewing, analysing and pointing out issues The KPI's at sector level in the Ministerial Budget Statement are limited Donor's also have their own KPI's they wish to see met Donors also monitor projects at different frequencies to each other, monthly, quarterly and 59 Description and Assessment Current Practice Adequacy Against Analysis Main Recommendations Current UNRA Plans And Best Practice Proposed Donor Assistance yearly UNRA uses KPI's at project level to monitor contracts Contract The external audit A - An audit of road A - UNRA to determine if there are plans to UNRA PLANS Management / approach is in line maintenance spending extend the audit process to development UNRA intends to develop Implementation ­ with good practice undertaken by independent projects guidelines for technical audits under but needs to be consultants was reviewed, it the TSDP Audits GoU to review the process used by extended to focused on Value for Money, The Auditor independent auditors to check the process Development Resource Allocation, Planning General is investigates in sufficient detail the quality Projects, not only and Budget Performance Primarily achieved on site in order to ensure they maintenance It did not cover development Responsible for adequately report back on Value for Money auditing all projects Government Projects Local Consultant Use of A - Local consultants are A ­ UNRA to provide information on future UNRA PLANS Capacity International concerned about investing in workload to the Consultant's Association to UNRA shall support the For major projects consultants in this building up their capacity as help local consultants plan implementation of the Local the GoU make use way is an there is not a steady demand B ­ GoU to undertake further investigate the Construction Industry policy that of local consultants acceptable for their services problems experienced by local consultants and has recently been approved by associated with approach B - The quality of work why some consultants do not deliver the Cabinet. International Overall aim should produced by local consultants quality required. Determine how the problems Consultants in be to increase is variable may be addressed and capability improved, capacity and Proposed Donor Assistance- order to provide the C - A number of local experts agree a way forward required capacity capability of local work for themselves. They 1 ­ Support investigation into B ­ GoU to establish and implement a and missing skills consultants to work for many consultants capability of local consultants, agree Performance Monitoring process for local e.g. materials undertake a higher and appear in many proposals. a way forward and assist in the consultants and reward good consultants with engineers volume of the If the percentage is too high it development of the local consultants more work, so better consultants are able to work, this though impacts on performance of the (6 months) expand. This requires a performance must be based on consultant evaluation process that is implemented by performance experienced staff C ­ UNRA to examine if the issue of local contract experts working for many consultants impacts on delivery of quality designs 60 Description and Assessment Current Practice Adequacy Against Analysis Main Recommendations Current UNRA Plans And Best Practice Proposed Donor Assistance Local Contractor Use of A - Local contractors have Capacity International problems with cash flow, For major projects contractors is an which are not resolved with GoU makes use of acceptable advance payments as local International approach Banks will not provide their Contractors who Problem remains Guarantees without substantial may associate with how to improve the cash deposits. a local contractor local capacity and A - It is understood there is a Local contractors capability ­ the commitment to initiate leasing on their own overall aim to mechanisms for local struggle to meet strengthen the local contractors to obtain plant donor requirements capacity based on B - Local contractors need to performance strengthen their capability but will only invest if there is continuity of work Overall Budget With the changes in In view of the increase in roads allocated to MoWT has a National Road Network size, UNRA and changes in the Districts, MoWT responsibility to it is assumed the to update the investment plan based on the provide sector development and higher service level of these roads and oversight. This maintenance budget for NEEDS analysis should include an UNRA will have to increase Current plans include undertaking backlog overview of the maintenance of National Roads and upgrading budget required to Table 18.1 and Table 18.2 of 10,000 kms of District Roads to National develop and the integrated Transport level. maintain Uganda's Investment Plan indicates the It is also understood that 5,000 kms of road network budgets up to 2017/18 community roads are to be upgraded to With the upgrade of District road level and it is intended to 10,000kms of District roads to undertake backlog maintenance of some National level, it is assumed 10,000 kms of road over the next five years. the development and MoWT to update the investment plan based on maintenance budget for the higher service level of these roads and UNRA will have to increase NEEDS analysis. In the event future maintenance budgets are not sufficient to maintain the roads (routine 61 Description and Assessment Current Practice Adequacy Against Analysis Main Recommendations Current UNRA Plans And Best Practice Proposed Donor Assistance and periodic), the overall approach should be reconsidered Financing Road Annual budgeting The annual budgeting UNRA to propose a plan for using UNRA Plans Development is not appropriate process is not appropriate dedicated project fund account and discuss Projects for multi ­ year given that road development with MoFPED. UNRA to request MoFPED to set Government development projects are multi ­ year up a Development Project Account greatly increased projects as investments. GoU should consider that the funding for into which development funds will funding for road implemented by Development Projects is undertaken through be deposited to support road development. This UNRA. Many An analysis of UNRA's dedicated development project fund account/s. development projects. funding is provided Development financial performance in FY Similarly, an appropriate indicator other than through annual Partners have in the 08/09 shows that out of the annual development funding outturns should budgeting process past created and total amount released by the be developed to evaluate UNRA's which requires used Special MoFPED, whilst UNRA was performance, consideration of a rolling annual budgets to Accounts in which able to commit 65% it was indicator that takes account of the likelihood be expended in the funding for only able to spend 35%. It of achieving the plan/level of absorption FY they are projects is reflects the nature of the should be investigated provided. deposited to support the project development projects and their as it proceeds. financing. Given that road development projects are multi ­ year investments, the annual budgeting process is not appropriate. It places undue pressure to UNRA to spend which is not possible because expenditure is supposed to be based on contractors' productivity 62 REFERENCES Adamowicz, M. and Lemanowicz, M. (2006), "Producer Groups as a Form of Entrepreneurship and a Tool for Strengthening the Competitive Position of Private Farms in Poland," presented at 7th International Conference on Management in AgriFood Chans and Networks, Ede, The Netherlands, 31 May ­ 6 June 2006. Al-Hassan, R., Sarpong, D., and Mensah-Bonsu, A. (2006), "Linking Smallholders to Markets," Ghana Strategy Support Program Background Paper No. GSSP 0001. International Food Policy Research Institute: Washington, DC. Annamalai, K. and Rao, S. (2003), "ITC's e-Choupal and Profitable Rural Transformation," World Resources Institute. Balat, J., Brambilla, I, and Porto, G. (2008), "Realizing the Gains from Trade: Export Crops, Marketing Costs and Poverty," Policy Research Working Paper no. 4488, World Bank. Banaszak, I. (2007). "Testing Theories of Cooperative Arrangements in Agricultural Markets: Results from Producer Groups in Poland," presented at IAAE-EAAE Seminar: Agricultural Economics and Transition. Deininger, K. and Okidi, J. (2003), "Growth and Poverty Reduction in Uganda, 1999­2000: Panel Data Evidence," Development Policy Review 21 (4):481-509. Fan, S., Zhang, X., and Rao, N. (2004), "Public Expenditure, Growth, and Poverty Reduction in Rural Uganda," DSGD Discussion Paper 4, International Food Policy Research Institute. Hine J., Riverson, J. and Kwakye, E. (1983), "Accessibility, Transport Costs and Food Marketing in the Ashanti Region of Ghana", TRRL Report SR 809. Crowthorne: Transport and Road Research Laboratory. Kindness, H. and Gordon, A. (2001), "Agricultural Marketing in Developing Countries: The Role of NGOs and CBOs," Social and Economic Development Department, Natural Resources Institute, Policy Series. 13. University of Greenwich: London, UK. Mitchell, B. and Rakotonirina, X. (1977), "The Impact of the Andapa-Sambava Road: A Socio- Economic Study of the Andapa Basin Madagascar", Washington, World Bank. Prahalad C. and Hammond, A. (2002), "Serving the World's Poor, Profitably," Harvard Business Review 80(9): 48-57. Raballand, G., Macchi, P., Merotto, D., and Petracco, C. (2009), "revising the Roads Investment Strategy in Rural Areas: An Application for Uganda," Policy Research Working Paper no. 5036, World Bank. 63 Raballand, G., Macchi, P., and Petracco, C. (2010), "Rural Road Investment Efficiency," Directions in Development Series, the World Bank, Washington, D.C. Romanik, C. (2007), "An Urban-Rural Focus on Food Markets in Africa," CIAO Working paper. Shepherd, A. (2005), "Associations of Market Traders", AGSF Occasional Paper 7, Rome: FAO Sieber, N. (1996), "The Contribution of Transport Investments to Regional Development in the Rural Areas of Sub-Saharan Africa," PhD thesis, Universität Karlsruhe, Karlsruhe, Germany. Smart, M. (2008), "Transport Demand and Spatial Equilibria," Journal of Transport Economics and Policy 42(2): 323-348 Stifel, D. and Minten, B. (2008), "Isolation and Agricultural Productivity," Agricultural Economics 39(1):1-15. World Bank (2007), "Uganda: Moving Beyond Recovery: Investment and Behavior Change for Growth," in Country Economic Memorandum, Washington: World Bank. World Bank, (2008), World Development Report 2009 ­ Reshaping Economic Geography, Oxford University Press. 64