Uganda
Technical Assistance Report-Initiating the Stock-Taking of the Public Investment Program

This Technical Assistance Report discusses the initiation of the stock-taking of the public investment program in Uganda. This stock-taking will provide a basis for better budgeting by providing information on the existing multi-year project commitments, and the incremental recurrent costs for operation and maintenance of the assets delivered. It will also identify a basic information structure for each project and subsequently collect a data baseline, providing a foundation for more robust project monitoring. It will aid the management of the overall project portfolio. By identifying the scale of existing multi-annual commitments, it will avoid adding projects to the investment pipeline, which cannot be financed under the Medium Term Expenditure Framework.

Abstract

This Technical Assistance Report discusses the initiation of the stock-taking of the public investment program in Uganda. This stock-taking will provide a basis for better budgeting by providing information on the existing multi-year project commitments, and the incremental recurrent costs for operation and maintenance of the assets delivered. It will also identify a basic information structure for each project and subsequently collect a data baseline, providing a foundation for more robust project monitoring. It will aid the management of the overall project portfolio. By identifying the scale of existing multi-annual commitments, it will avoid adding projects to the investment pipeline, which cannot be financed under the Medium Term Expenditure Framework.

I. Introduction

1. Public investment is one of the cornerstones in the development strategy of Uganda. The five-year National Development Plan (NDP II)—up to FY 2019/2020—includes large-scale public investments in capital intensive sectors such as roads and energy, as well as in social sectors such as education and health. However, projects are not being delivered on time, on budget and with the planned impact.

2. Uganda’s Public Investment Program (PIP) is likely overcommitted relative to the available medium-term fiscal space.2 The PIP includes the projects that are authorized to receive funding, either from the National Budget or from Development Partners in the current fiscal year. Over-commitment contributes to slow budget releases, underfunding of projects with respect to the original investment plans and falling behind the execution schedule. This situation drives up costs and reduces the expected impact.

3. Information available in the Ministry of Finance, Planning and Economic Development (MoFPED) on implementation status and required expenditures for PIP projects is also quite weak or unreliable, especially for the outer years. This is despite some recent efforts to improve the information. For example, the Standard Gauge Railway project that is estimated to cost circa USD2bn, appeared in government databases with a much lower value; some investment initiatives do not have start and end dates (4 projects); others that were supposed to end in FY 2016/2017 are still in the PIP for FY 2017/2018 (12 projects); and the units used to input the total value of projects varies from billions to thousands of Shillings between projects.

4. Thus, the government is undertaking a stock-taking of its public investment projects. The benefits of the stock-taking are: first, it will provide a basis for better budgeting by providing information on the existing multi-year project commitments, and the incremental recurrent costs for operation and maintenance of the assets delivered. Second, it will identify a basic information structure for each project and subsequently collect a data baseline, providing a foundation for more robust project monitoring. Third, it will aid the management of the overall project portfolio. By identifying the scale of existing multi-annual commitments, it will avoid adding projects to the investment pipeline, which cannot be financed under the MTEF. Furthermore, the stock-taking will prepare the grounds for a comprehensive review of the PIP and identify under-performing projects for review and corrective action.

5. The Project Analysis and Public Investment Department3 (PAP) of the MoFPED plays a central technical role in managing the overall project portfolio, and is the body responsible for preparing and implementing the stock-taking.4 PAP is leading the process for strengthening the selection process of investment projects, and it acts as the secretariat for the Development Committee. PAP is also leading the process of comprehensively reviewing the PIP, which includes identifying projects that may have to exit the PIP. PAP has already identified projects that should not be in the PIP because of a high content of recurrent expenditures. The stock-taking will provide additional baseline information.

6. As part of the strengthening of its Public Investment Management (PIM) capabilities, the authorities also plan to build a project database—an Integrated Bank of Projects (IBP) over the next 12–18 months. This IBP should become a tool for the MoFPED to monitor all public investment projects—whether they are under development or in implementation—across all line ministries and to facilitate budgeting.5 Ideally, the IBP will become a tool which has value beyond the MoFPED for all government officials involved in implementing and monitoring projects and the wider project portfolio. The government is currently developing the terms of reference for the hiring of the consultants that will assist in the design and roll-out of the IBP. The stock-taking exercise will provide valuable inputs for the design phase of the IBP,6 by identifying a first set of information desirable for project monitoring, providing an understanding of what information is readily available, and identifying the information shortfalls that have to be addressed by the MoFPED and line ministries in the medium term.

7. An important challenge for stock-taking and subsequently the IBP is to ensure that the information collected is reliable. The comprehensive review of the PIP will only be credible if stakeholders and decision makers trust that the baseline information obtained in the stocktaking is correct. And similarly, the IBP will only be adopted, used as a one-stop source of information on investment projects by stakeholders if the information is considered reliable. Quality assurances in the stock-taking will be essential.

8. The mission focused on developing templates for collecting and aggregating key information on public investment projects with PAP and pilot ministries. Two excel-based files were developed to achieve this goal. The first one is a template that will allow the MoFPED and line ministries to capture standardized information on individual investment projects (see Annex II). The second is a spreadsheet that will aggregate the key data on projects to allow the authorities to analyze its investment portfolio from different perspectives (see Annex III). It was decided to work with an Excel based format to reduce development time and cost to provide a short-term operational solution prior to developing a robust IBP.

9. The templates were developed in consultation and collaboration with internal stakeholders in MoFPED, pilot MDAs, members of the DC sub-committee and external stakeholders. Officials from the Office of the Prime Minister and the Ministries of Health, Energy and Minerals Development and the Uganda National Roads Authority provided useful guidance and input for the template design. A dry run of the template was conducted with one infrastructure and one social ministry.7

10. This report summarizes key observations and recommendations of the mission relevant for the finalization of the template and roll-out of the stock-taking. It also briefly provides advice on related issues on which the authorities requested advice.

  • Chapter II describes key observations and lessons from developing the template;

  • Chapter III describes the template for individual projects and the aggregation file that were developed in collaboration with the authorities;

  • Chapter IV provides some insights on the scale of the challenge ahead some ideas on how the stock-taking exercise can be undertaken and the trade-offs that the government will have to make decisions on;

  • Chapter V provides briefly an outlook on the comprehensive review of the PIP that is planned to follow the stock-taking; and

  • Chapter VI summarizes comments on the draft TOR for developing an IBP, which were requested by the authorities.

II. Observations from Developing the Template

A. Engaging with Stakeholders

11. In order to ensure the feasibility of the template, it was developed in consultation with internal MoFPED stakeholders and pilot ministries. PAP was invited to select suitable MDAs in the economic infrastructure and social sectors in order to assess the potentially different challenges in gathering stock taking data across sectors and to see whether a single common template would be adequate to assist in the task. The choice of MDAs proved to be helpful in guiding and advising the mission team in the design of the template.

12. Awareness of the intended stock-taking exercise and intended review of the PIP was limited and in some cases non-existent. This was the case both for MoFPED internal stakeholders, and external stakeholders. PAP, as the coordinating unit for the investment project portfolio, will need to develop stronger practices for identifying, engaging with, informing and coordinating with all stakeholders.

13. While MDA’s were not aware of PAP’s intentions ahead of the mission’s meetings, once informed about the stock-taking, MDAs reacted generally positively and were willing to cooperate. PAP should develop a narrative and speaking points to communicate about the stock-take and the wider context. The narrative should emphasize that, if undertaken appropriately, the stock-take will be a win-win situation for MoFPED and line ministries.8

14. Where information already exists within MoFPED there is no value added by attempting to collect it again through project teams in line ministries. Discussions with internal stakeholders revealed that some information already exists on projects within MoFPED. Useful information is already available on the Aid Management Platform (AMP) and the Debt Management Financial Analysis System (DMFAS). Random checks by the mission team revealed a reasonable level of reliability in the data held. The challenge for the stock taking teams in these cases is to verify the data rather than gathering it again (see below).

B. Review of IT Systems

15. MoFPED maintains some PIP information already in IT systems. However, these systems are setup to address specific issues and information needs within specific departments of the MoFPED9 and not to work as a “one-stop shop” for investment projects. An individual who wants to gather data on a specific project can use these systems as a source of partial information, but will need to tap into more than one to get a complete picture of the data. Given their focus on a specific subject (e.g., aid or debt management) they are not set up to hold all the information desired for the stock-taking.

Suggested Narrative for Communicating with Stakeholders

  • → The PIP is likely over-subscribed, resulting in delays in budget releases, but MoFPED information is unreliable and MoFPED cannot make a proper assessment—too many projects chasing the same limited resource—this is a recognizable situation for all line ministries in Uganda.

  • → The quality and reliability of information contained in the PIP is questionable. There are some inconsistencies and potentially incorrect data—this leads to uncertainty about the validity of information.

  • → It would not be easy to aggregate the data contained in the existing PIP, meaning important questions regarding the scale of the GoU’s total commitments for projects and the structure of the project portfolio cannot be accurately assessed.

  • → New projects will be subject to the DC guidelines so they will be better prepared in the future but this leaves the legacy projects—those that are already on the PIP – it is these projects that will be subject to the stock-take exercise.

  • → All public bodies with a project in the PIP will be required to complete a template for each project, with the support of MoFPED.

  • → The data gathered will allow for more accurate information for decision-making on the PIP, including better planning and necessary budget releases to ensure the project can move forward as planned.

  • → The data collected will be shared with sectors and MDAs, so they can make use of them for their own requirements.

  • → Information gathered will include information on the expected O&M costs of projects especially those that are nearing completion.

  • → A better understanding of the future budgetary requirements of line ministries will lead to more accurate planning at both the MoFPED and the line ministries, resulting in better and more efficient allocation of resources, and ultimately timely completion of projects.

  • → Better budget planning based on more accurate input data will provide line ministries with a much improved predictability of funding for projects.

  • → Improved budget allocations for projects should lead to faster project completions and subsequent benefit realization.

16. The Program-Based Budgeting System (PBS) is currently MoFPED’s main reference system for presenting project information. Besides the information on budget allocations,10 project managers in MDAs must present information on outcomes, achievements and also descriptive information on a project’s background and objectives. However, the main issue to be addressed with this information is the annual budget commitments for the upcoming fiscal year, limiting the attention placed on the quality of other information available, which explains why existing project information is considered unreliable. Furthermore, the system is still under the latter stages of development with some issues of functionality such as the use of differing units of measurement and allowing data to be inputted with either numbers or text. The information from this system can be used to identify the projects to do the stock-taking on or to determine how to phase the exercise if it is decided to do so.

17. The current set-up of the PBS has several shortcomings for PIP management. Users in line ministries upload PIP information into the system annually. However, the system is not designed to facilitate aggregation, easy analysis or data checks, making meaningful analysis almost impossible and aggregation extremely time consuming. Many of the inputs are not restricted (e.g., text or numbers) or structured (for example through drop-down menus) so there is no consistency in the way data is entered. Units for numbers are not defined, and different projects use different units (e.g., Subprogram 1444 – Agriculture Value Chain Development appears to be in Shillings, while most of the others are in Billions of Shillings). Data records do not retain certain reference information (e.g., the original project completion date), nor are previous data submissions available for comparison. Thus, the current data management practices make it difficult to determine such important questions as when or for how long a project is delayed, or changes to budget requests.

18. BPED as the owner of the PBS should take some of the ideas developed through the stock-take template and feed them back in to improvements in the PBS. Upgrades to the PBS are currently underway in conjunction with a change from output-oriented budgeting to program-based budgeting, providing an easy opportunity to implement systems improvements. In particular, the use of open dialogue boxes in the PBS should be restricted and replaced instead with drop down menus that would facilitate deeper analysis of the inputted data each budget year. It could also lead to better management of planned and revised dates in the PBS. Moreover, the dialogue with BPED initiated during the mission needs to be continued by PAP. This would ensure that improvements in the functionality of the stock-take template can be shared through the final stages of PBS development. Moreover, this ongoing dialogue over data-sharing should extend to cooperation over the design and development of the IBP. Annex I presents some projects for which the information uploaded into the Program Based Budgeting System (PBS) should be reviewed.

19. The Debt Management Financial Analysis System (DMFAS) and the Aid Management Platform (AMP) already capture key data on projects supported by development partners. The systems have sufficient data on loans and grants assigned to specific projects that could be used by PAP to collect information for the stock-taking. When a contract with a development partner is signed, information on the amount of resources committed to the project, the disbursement profile and the financial conditions are uploaded into the system. This information is updated by a project’s sponsor and double checked with the Development Partner. However, the focus of these systems is not to monitor the projects but to track the use of the resources from development partners. Therefore, a change in the domestic resources committed to the project would not be captured by the system.

20. The DMFAS and AMP have a limited reach on projects that are financed by non-traditional donors. This is more common in what are considered social sectors, such as health and education, which are not as capital intensive. The Ministry of Health, for example, suggested to the mission team that as much as 30 percent of projects in the health sector receive direct financing from donors and are not known by the ministry or uploaded into the existing systems. This situation makes it more difficult to coordinate initiatives for the ministries, adding an additional line of complexity to their jobs.

21. Managing a project with a unique project identifier is critical for understanding the evolution of projects. It appears that the existing PIP includes some projects twice with double entry of projects under different lines. As an example, a quick analysis by the mission team revealed that of the 727 input lines, approximately 423 might be considered as public investment projects11 using the current classifications under the Chart of Accounts. This shows that it is difficult to analyze the information provided by the PBS due to the different filters that need to be applied.

C. Lessons from the Template Dry-Runs

22. Dry-runs were conducted with MDAs as a test of the practicality of the questions included in the template and the functionality of the template itself. They offered the opportunity to measure the amount of time taken to complete the template and thereby give an early indication of the scale of the task involved in gathering information on all PIP projects during the stock-take. Since the template was new to the pilot MDAs, an element of the time taken was consumed by familiarization. In order to accelerate the completion of the work, and to improve the quality of data submission, formal training on the template should be provided to those who are tasked with completing the template.

23. It might take one stock-take team the majority of one day to complete an entire template. This is extrapolated from the time it took to complete just the Summary Sheet of the template, which was between 50 minutes and 1½ hours. With more than 400 projects to cover, it can be envisaged that more than 400 team days12 might be consumed by the exercise.

24. Even with the right documentation available, it is easy to make simple mistakes— the value of one project was entered originally as UGX179bn with an updated estimate of UGX181bn. Double checking of the documentation by members of the team revealed that the number reported was in fact a sub-total and that the cost that should have been recorded was UGX215bn. A verification process should be an integral part of the stock-taking exercise.13

25. The template requests information that is structured by the Chart of Accounts, but this is not the reporting structure MDAs and their supervising engineers use to monitor their projects. The Interim Payment Certificates (IPC), seen by the mission team and which are milestones in project implementation, did not refer to classifications in the Chart of Accounts but instead used classifications that might be better aligned to the reality of the projects industry. Ensuring that these expenditure items align with the classifications in the Chart of Accounts is likely to add significant time to the completion of each template and should be supported by MoFPED staff to ensure accuracy for accounting and fiscal reporting.

26. Project information should be completed following consultation of project documents, not solely from memory. A key purpose of the exercise is to enhance information reliability, which requires—the somewhat time-consuming—re-validation of the information provided from official or project documents. MoFPED staff working with MDA teams during the roll-out should be vigilant that appropriate verification is being done as part of the on-the-day process.

27. The stock-taking dry runs prompted project teams to self-assess projects’ weaknesses and identify missing information; this needs to continue during the roll-out. During the dry-runs the project teams became aware of various inconsistencies and outstanding issues simply because the template forced them to answer questions that may not have previously been asked. Since one of the purposes of the stock-take is to identify projects that may need to be adjusted, teams engaged in the stock-taking should take note of projects that demonstrate particular difficulties in finding information or in completing the template.

28. Some updated cost information is available and it will be important to verify and capture this information during the stock take. Where cost information is clearly out of date, stock taking teams should persist in obtaining the most recent estimates available from recent project documentation if it is available or if not from a consensus estimate of the project teams in MDAs. Where this is the case, a note should be added to the template to the extent that a number is estimated. This becomes a red flag signaling that the project should undergo a new feasibility analysis during the clean-up stage of this process.

29. It was the considered view of the mission team that the in-house resources available to PAP would not be sufficient to achieve a fully reviewed and cleaned PIP in time for FY 2018/2019. Furthermore, the mission team was not fully convinced that even the stock-take part of the exercise could be achieved in time using only existing in-house resources. This observation points towards a need for external support. Therefore, MoFPED needs to consider its options for undertaking the tasks ahead. These options might include:

  • Hiring external consultants to support the entire stock-take – this will require a substantial budget and would also eat into the already constrained timescale, in order to conduct the necessary procurement exercise. Further external support might also be necessary to conclude further stages such as criteria development and possibly supporting the reviews of under-performing projects;

  • Phasing the stock-take and project reviews either by tackling the largest projects first and focusing on them for the 2018/19 budget14 or;

  • Conducting the stock-take on a sector by sector (or line ministry) basis calculated to be commensurate with available resources.

D. Summary of Recommendations

Recommendation 2.1: Prior to the stock-take PAP needs to develop a strong and persuasive narrative for conducting the stock-take and subsequent clean-up of the PIP.

Recommendation 2.2: Stakeholder engagement with internal and external stakeholders at all levels should begin as soon as possible.

Recommendation 2.3: Information that already exists should not be collected twice—simply verified.

Recommendation 2.4: As a quick-win, PAP should verify the mission team’s estimate of the number of public investment projects using the Chart of Account classification in order to fully quantify the stock taking task.

Recommendation 2.5: The functionality of the PBS should be adjusted in line with the functionality of the template in order to improve the quality and consistency of input data.

Recommendation 2.6: Training and familiarization with the template is an essential part of conducting the stock take in the most efficient way possible.

Recommendation 2.7: Additional time should be factored in for double checking and verification of data during the stock take. This should be done using available documentation.

Recommendation 2.8: Additional time should be factored in for ensuring that financial values shown in Interim Payment Certificates are consistent with classifications in the Chart of Accounts.

Recommendation 2.9: PAP should urgently consider its options for external support in the conduct of the stock take exercise and further additional support for the subsequent PIP cleanup.

III. Templates for the Stock-Taking Exercise

A. The Templates for the Stock-Taking

30. Two Excel-based templates were designed to help guide the stock-taking exercise and the subsequent PIP review and clean-up. A project template will assist in capturing individual project data and is the one that will be used throughout the stock-taking, during the meetings with the MDAs, and as a source of information detailed data when reviewing specific projects. An aggregation file will create a simple standardized database of the information captured through the individual templates. Its objective is to facilitate analysis of the PIP information across different projects, sectors and MDAs. It will prove more useful as individual project templates are completed on a larger share of projects in the PIP and it can provide answers to understanding, for example, the total value of all projects for a certain MDA.

31. The templates were developed in Excel because it provides quick-wins for the stock-taking exercise. Developing the files in Excel reduces the design and formulation times, in comparison to a web-based system, which is very important, given the time constraints. Since it is a widely-used system, it reduces the time and cost of training other users in the use of the templates. Likewise, it can be easily modified by PAP to address issues that were not included during the design phase. However, Excel does have some limitations in terms of how information can be stored and of the access to users. As this is a one-off exercise and the fact that the development of an IBP is planned for the near future, these limitations do not present a hindrance to achieving the goals of this exercise.

32. The project template is designed to force data standardization. This is achieved through drop-down menus and limitations on the type of data that can be inputted in each field. Only with such standardization will it be possible to easily group projects within specific categories such as the phase of the project or the link with the national development strategy. During training it will be important to convey to stock-taking teams that they are not permitted to manually override the data input restrictions or standard categories. If the categories provided are inappropriate for their projects, they need to contact PAP for guidance on how to address the issue.

33. In order to support use of the templates as intended and reliable data provision, an instruction manual on the template should be prepared. The manual need not repeat the template, which will be shared in full, but should highlight key issues in using the template and for providing the data. The dry-runs with MDAs (see Chapter II), training sessions and the early rounds of using the template provide insights into issues that may be challenging for MDAs.

B. Template for Individual Project Information15

34. The individual project template for the stock-taking exercise focuses primarily on the resources necessary to complete projects and to operate and maintain them afterwards. This was one of the main deficiencies of project information identified during the March mission, and acknowledged by the MoFPED. Information demanded by the template include total investment costs of the project and a forecast of annual budget allocations that are needed for its completion; the source of the funds (Government of Uganda, Development Partner, or from a third source); as well as operating and maintenance costs that will arise once the project is completed (e.g., employee costs, purchase of goods and services, subsidies or social benefits granted).16 The inclusion of operating costs into the template was specifically requested by the Budget Director, given that correctly estimating the recurrent costs of operating projects is one of the weaknesses in the budget forecasts.

35. A key output of the stock-taking exercise desired by MoFPED is to understand how a project’s costs have changed since its initial estimate. Thus, the template requires, first, the original budget with which the project was approved, and second, the most recent forecast of project costs. The second set of information will allow MoFPED to determine if the information in the PBS reflects the most recent cost estimate for a project, as is required when the PIP is being prepared. It will also become the cost baseline for comparing any future changes requested by the entity sponsoring the project. Having these two data points will also allow authorities to determine which projects have deviated considerably form their original execution plan and could be subject to additional review.

36. The stock-taking template also collects improved qualitative information about each project, so as to allow MoFPED to better understand the structure of its investment portfolio. Project sponsors are required to provide information on the link of each initiative with the national development strategy NDP II. This data will allow a check for whether resources are being deployed accordingly to the stated development priorities and how each initiative will contribute towards achieving specific development goals. In addition, information on relationships and dependencies with other projects within the same or other sectors is requested, as one project falling behind can cause delays in other projects. And not least, basic information such as start and end dates or the region impacted by the project must be validated and re-submitted to ensure that the information available to MoFPED is accurate.

37. The design of the template focuses on strengthening information that is considered unreliable and avoiding duplication of already reliable data. Therefore, for projects supported by development partners, detailed information on loans and grants provided by development partners is not being requested from MDAs as this information is already available in the DMFAS and AMP systems and can be captured internally.

38. Although unit costs are an important yardstick with which to assess the soundness of project costs, these are not requested in the final template adopted for the stock-taking exercise (as of end of the mission). These were removed from the final version of the template because in both “dry runs,” the project owners acknowledged that they did not have this data and were concerned that producing this information could deviate attention from the main objective, which is to determine overall project costs. They also questioned the usefulness of these parameters, especially for social projects which are very diverse and difficult to compare through unitary costs. Finally, given the already expanded timeframe for completing the stock taking exercise, requiring that project sponsors develop these parameters could protract it even more.

39. The stock-taking template will be an important input into building the IBP, but it will not provide all the information needed for project analysis and management. Experiences in other countries have shown that building a project database is an ongoing exercise that requires many iterations. It is important to start with a core set of data that can be provided fairly easily and reliably, demonstrate usefulness and success in managing this information, and then expand. In Uganda, avenues for expanding structured project information maintained in databases include: (i) the information demanded in the current guidelines approved and published by the Development Committee for approval of new projects; (ii) information demanded in the annual budget process; and (iii) the implementation of an IBP, when the MoFPED will be able to consolidate in a single source all the information on investment projects, from the moment they start the approval process, through execution and operation.

C. Template for Aggregating Project Information

40. A key component of the stock-taking is to aggregate the data collected so that analysis can take place. At a macro level, answers to questions that have to date eluded MoFPED can be achieved. Significant questions might include:

  • How many projects are currently under implementation through the budget17 in Uganda?

  • What is the total commitment cost of all these projects?

  • How much has been spent to date on these projects?

  • How much will it cost and how long will it take to complete all of these projects or a subset of projects (e.g. for a sector)?

  • How much will be financed from GoU, from Development Partner sources?

  • What will be the additional O&M costs resulting from projects reaching completion in the next three budget years?

  • How does the project portfolio overall align with the growth and development strategy set out in the NDP?

41. The aggregate data management tool developed by the mission allows aggregation of the data collected in the many individual project templates; but it is only an interim short-term solution. It can be used to achieve a better understanding of the investment portfolio and provide answers to questions like those above.18 It should also be used to review data in relation to each other in the context of each sector, including conducting validation exercises on the data collected, and to determine which information is readily available for all projects and on which additional focus and follow-up work might be needed. Further, the tool will assist in undertaking a review of the project portfolio. PAP has already issued a first set of guidelines on which projects should be moved out from the investment budget and into the recurrent budget. By providing an overview of all projects the tool will also facilitate determining additional criteria by which funding decisions might be made and which projects require a more detailed review. Finally, given that the IBP will not be in place for the construction of next year’s budget and PIP, this file should provide the baseline information on ongoing projects for building next year’s project budget.

42. Not all the information that is being captured in the individual project template is reflected in the aggregate file. Some information is only relevant at the individual project level, especially certain qualitative information. Another important consideration is to avoid a ballooning file size once the information on all projects starts coming in. This will ensure that the file is easy to handle and share with stakeholders. The IBP eventually will be able to handle much larger databases and also can display multiple relationships within different variables of a project. Annex III presents the list of variables from the template that are aggregated.

D. Summary of Recommendations

Recommendation 3.1: Request the original and updated information on project costs so that it is possible to determine the actual change in investment commitments.

Recommendation 3.2: Use the information on loans and grants from Development partners available in DMFAS and AMP to complement data captured through the stock-taking.

Recommendation 3.3: Continue with the development of the IBP system to replace the aggregation template to develop a platform for future project analysis.

IV. Roll-Out of the Stock-Taking

A. Key Elements of the Road Map

43. Completing the stock take will require strong awareness of the scale of the task, good planning and strong leadership. The key elements involved are project management, training and awareness in the structure of the template; stakeholder communication as well as the stock-taking data gathering and verification activities at MDA level. Further ex post validation of data is likely to be needed as a next step or confirmation that data are reliable enough as the basis for making reliable decisions. Additionally, a better understanding of the information already in the PIP, will guide the stock taking teams to better challenges and verification during the stock-take meetings. An initial work-plan initiated by PAP has already recognized some of these elements.

B. Project Team and Management

44. MoFPED should appoint a senior official to take overall responsibility for the stock-take. This official should be senior enough to command respect not just with internal stakeholders but across a wide range of external ones. The stock-take would be supported by two teams and the people who are members should be clearly identified. Firstly, a senior level coordinating team, which can reach out to and inform MDA counterparts at the PS and senior management level, and who would be available for resolving difficulties. Participation by the PM’s office and the Accountant General’s Office may also be useful. This senior level team which could be chaired by PS/ST and with membership at director level can reach out to more senior levels in MDAs. Secondly, a technical coordinating team would be led by and heavily represented by PAP, but there should also be representation from BPED, the Budget Office as well as sector contact points. A representative from the NPA should participate, to reference linkages with national development policies. This team would be responsible for the planning and execution of the stock-take. The same technical coordination team will also be responsible for analyzing and assessing the information from the stock-take and for drawing observations and conclusions.

45. Each MDA should nominate its own coordinator for gathering project data ahead of the stock take and for guiding the stock take team through the data and completing the template. MDA coordinators should identify those officials with the greatest knowledge and insight of each project and should ensure that these officials are available for between half and one and a half days (depending on the scale and / or complexity of the project) in order to support the completion of the template. The MDA coordinator should also ensure that all relevant documentation related to each project is made available to the stock-taking team on the appointed day(s) pointing out that absence of this documentation might result in the project receiving a “red-flag” in a subsequent review.

46. Training and awareness of the template will be an important means of reducing the overall time-scale for the stock-take. PAP is already taking steps to ensure that a substantial body of officials is trained in the detail of the stock-take template. The core PAP team is envisaged to be 10 strong, supported by the training of 140 officials in MDAs. Since a PAP core team member will need to be present at each template exercise this will be the main constraint. If all 10 core team members are fully deployed in the stock-take supported by the right project officials and documentation in MDAs, it would theoretically be possible to complete 50 projects per week based on an average of 1 project per team per day. This would indicate a period of eight-nine weeks of intense activity but, given the rate of progress may slow due to unforeseen circumstances in some projects such a rate of progress might be seen to be ambitious and would require an exceptional quality of planning to achieve.

C. Stakeholder Communication

47. MoFPED needs to develop a narrative and communication strategy with internal and external stakeholders which presents a convincing case for carrying out the entire exercise. They will need to convince a wide range of stakeholders of the benefits that can accrue from a successful outcome. Awareness of the stock-taking and the intention of reviewing comprehensively the PIP should also be raised at the most senior levels among Uganda’s key decision makers.

48. Further important steps in building a stakeholder consensus are also necessary. PAP should present the road map for the stock taking and the template to top management in MoFPED to ensure support for its intentions internally. In order to achieve buy-in from line ministries, Accounting Officers in line ministries need to be informed, followed by comprehensive engagement with Planning Departments. Given the scale of the engagement required, this should begin immediately. A letter could be sent from PS MoFPED to counterparts in MDAs seeking support for the exercise and detailing the necessary form of collaboration and requesting a nominated coordinator. The key message to be transmitted is that the objective is to acquire the right detail on projects so that the “right” budget can be designed for the PIP. Additionally, the PM’s office could provide support, for example, the exercise could be discussed and promoted at a Cabinet meeting through a Cabinet paper on the entire exercise— subsequently a Cabinet paper could discuss progress and results.

D. Procedures for MDA-Level Stock-Taking

49. Each stock-take team will comprise as a minimum a trained representative from MoFPED /PAP, the MDA coordinator and at least one senior project representative. The scale of representation should be proportionate to the scale (or complexity) of the project involved. If the MDA coordinator is not from the Planning Department then a representative should also attend each stock-take assessment.

50. It is recommended that the stock-taking planning builds in at least one-week notice to MDAs for each project. Requests for individual appointments should be coordinated and agreed between PAP and the MDA coordinator with the understanding that the entire team involved should be prepared to dedicate an appropriate amount of exclusive time during the appointed day in order to complete the template satisfactorily. In order to save time on the day, the template should be shared by the MDA coordinator with the project teams well in advance. This is to ensure that there is sufficient awareness of the information required and to allow time in advance to gather it. Project teams should also be encouraged to complete some of the basic information in advance which again saves time on the day and allows for the allocated time to be dedicated to verifying more critical areas of the template.

51. At the conclusion of each stock-take exercise, the MDA coordinator and the project team should agree that they are content for the information to be finalized and locked. Each completed template should be then printed and signed by the required signatories with one copy kept by the MDA coordinator and another by PAP. At the end of each day all data collected should be stored securely and backed up at PAP.

52. Quality control and assurances about the reliability of the information will be essential to ensuring that there is reasonable confidence in the data collated. Given the substantial efforts that will be necessary to undertake the stock-take, it is right that every effort is made to check and double check the information provided during the completion of each template. Each stock-taking team should, during the course of the template completion, scrutinize project documentation to ensure that the latest information is provided as well as original estimates. Discrepancies or anomalies should be searched for by PAP afterwards as part of the verification process and those found should be subject to re-checking with the relevant MDA.

E. Speed vs. Comprehensiveness of MDA-Level Stock-Taking

53. A decision on the timeline and scale of the stock-taking for the 2018/19 budget will be needed soon. Dependent on the number of teams and personnel available, it may be preferential to stratify the PIP and focus on the most important/largest projects this year, and conduct a second stock-taking next year. As data reliability is a key objective of the stock-taking, no compromises should be made in monitoring the presentation of information for specific projects and in ex post validation. A strong preference should be chosen in favor of the quality of the data gathered rather than accepting lower quality or unverified data in the interests of expediency.

54. A MoFPED participant in the stock-taking team will need to be present at all times while the templates are being filled in. This is important for assuring that the information is prepared rigorously, to understand the challenges encountered and to provide guidance.

F. Validation of Data Collected

55. Once the MDA-level stock-taking is completed the coordinating team will need to review, analyze and assess the data collected. The coordinating team will prepare the aggregation file and run data summaries by sector. Plausibility checks on individual project information should also be conducted. Inter alia, previous information in the PBS, and the DMFAS and the AMP can be used. As PAP has already identified in its concept note, the training of desk/sector officers, with their specialized knowledge will be an important additional resource in this task. Also, the support of staff so far not directly involved with the stock-taking but knowledgeable about the project universe (e.g., debt management staff from AG office, macro-fiscal team) could be enlisted.

56. Observations on the ease of each MDA in completing project templates also need to be gathered and analyzed. MoFPED staff participating in the teams with MDAs should prepare a short report on key observations, results achieved and share any concerns about the data collected. The coordinating team may also develop a simple template for this report to make it easier to compare and summarize.

57. The coordination team should prepare a summary report on the stock-taking exercise identifying objectives achieved and identified weaknesses. The summary report would be part of an “after action review” and will provide an important input into subsequent decisions on next steps.

58. Once the results of the stock-taking have been assessed, the coordinating team in consultation with the senior-level team will need to reach decisions on next steps. The “after action review” should in particular reach a conclusion on whether the data collected to date are sufficiently reliable to understand the scale of multi-annual project commitments, and whether they can be used as a basis for the review of the PIP and essential decisions on the structure and future evolution of the PIP. Decisions should also be reached on additional data collection, whether certain data on some projects may have to be revalidated a second time, and what the process will be. PAP should share the data collected analyzed with MDA’s including the summary information.

G. Summary of Recommendations

Recommendation 4.1: The Road Map should be used to make a detailed estimate of the time and cost viability of cleaning the PIP, to identify where additional resources may be required and how these resources might be funded.

Recommendation 4.2: PAP should develop a chronology for the activities shown on the road map based on a decision on whether or not the overall objective is to arrive with a cleaned PIP for the 2018/19 budget.

Recommendation 4.3: MoFPED should appoint a senior official to take responsibility for the entire stock-take with support from senior level and technical committees.

Recommendation 4.4: All MDAs should appoint a stock-take coordinator through whom the technical committee led by PAP can plan, organize, and execute the stock take.

Recommendation 4.5: At least one-week notice should be provided before any stock-take exercise to allow the right team and the right documentation to be organized.

Recommendation 4.6: Completing templates should not take place without the presence of a trained MoFPED official.

Recommendation 4.7: All data gathered should be secured and backed-up at PAP at the end of each day and should be checked and validated afterwards for plausibility.

V. Outlook: Review of the Public Investment Plan

A. Key Steps in the Comprehensive Review of the Public Investment Plan

59. The review of the PIP can only be confidently undertaken with up to date cost and time data. Without it the whole exercise is undermined. The review of the PIP should not begin until MoFPED are sufficiently confident in the quality of the information resulting from the stock-take. MDAs should have been previously informed that a lack of information about updated costs will result in an inevitable “red flag” for the project. This stance would be fully justified if the project representatives are unable to provide the necessary cost information.

60. Once the quality of information from the stock-taking has been ascertained, the comprehensive review of the PIP and clean-up can be initiated. (For a road map see Figures 1 and 2). The review will start with development of criteria for identifying underperforming projects, which will draw on the data collected and the observations during the stock-taking. Once criteria are developed agreed and approved, the project data need to be reviewed to identify the projects to which they apply. Then these projects can be reviewed—an undertaking that might include the following possible resolution actions:

Figure 1.
Figure 1.

Steps to be Taken from the Stock-Take Template to Cleaned PIP – Stage 1

Citation: IMF Staff Country Reports 2018, 180; 10.5089/9781484362686.002.A001

Figure 2.
Figure 2.

Steps to be Taken from the Stock-Take Template to Cleaned PIP – Stage 2

Citation: IMF Staff Country Reports 2018, 180; 10.5089/9781484362686.002.A001

a. Cancelling the project

b. Postponing

c. Re-phasing

d. Early termination at a specified point in the future

e. Re-appraisal of the project (or possibly appraisal for the first time)

f. Rectification of issues prior to re-entry in to the PIP at a later date

61. An important element of the comprehensive review will be an assessment of the available fiscal space for projects, and ensuring that the updated project portfolio does not exceed it. Therefore in parallel MoFPED will need to prepare fiscal space estimates for the PIP against the scale of the multi-annual commitments. The overall result will be a PIP that “fits” the available fiscal space.

62. Cabinet will have to be involved in reaching decisions on the PIP. Cabinet could be asked to approve the clean-up criteria after they have been designed at a technical level. By approving clean up criteria, the subsequent work on individual projects could be conducted at less senior or technical level.

B. Summary of Recommendations

Recommendation 5.1: The review of the PIP should not begin until the MoFPED has sufficient confidence in the quality of the data from the stock-take. Projects that were unable to provide updated costs during the stock-take should not be included until they have been subjected to further appraisal and review.

Recommendation 5.2: Data gathered from the stock-take should be analyzed and used to model a range of criteria for classifying, and thereby identifying under-performing projects. The criteria will need to be coordinated with the expected available resource (internal and/or external) to conduct a review of the projects identified.

VI. Outlook: Review of the Terms of Reference for Integrated Bank of Projects Development

63. The authorities have resolved to move forward with developing the IBP. Having received internal approval for the development of the IBP, PAP has developed a draft Terms of Reference (ToR) that they intend to form part of the procurement documentation for sourcing a software developer. The mission was asked to review a draft ToR.

64. The specifications in the draft TOR for the IBP are extensive and seem unnecessarily complex and therefore unduly expensive to realize. The mission recommends the following:

  • A conceptual design of what is needed should first be developed. At that point a decision can be made on what serves the most immediate needs and what can wait for further development phases. This will allow MoFPED to manage both time and cost and align them with “need” rather than “want.” Once short-term needs have been mastered additional desirable features can be added.

  • There is a need to design in the relationship between the IBP and existing sources of data within MoFPED. These include: the PBS, DMFAS, AMP, and IFMIS. This requirement needs to be emphasized. It should also be mentioned that one of the first data entry exercises will be the data collected through the stock-take exercise—this will be a specific task.

  • In order to protect the medium and long-term interests of the MoFPED, the ToR should state that the consultancy should leave the system code as a deliverable so that future amendments can be made without undue cost.

  • It will be necessary to state what is acceptable and not acceptable in terms of physical presence of the consultants in the MOFPED premises or even in the country. If international consultants are used, a desire to have them present in the country at all times will incur additional costs. It is possible that much of the design work could be achieved remotely— having full-time experts on site involves a different cost base to having experts present only at key stages.

  • The ToR should be more specific and descriptive, avoiding general terms that may cause bidders to price in unnecessary eventualities.

  • It might be a good value proposition to hire someone to do the conceptual design of the IBP and then separately hire someone to build it.

  • The template’s data categories should form a core component of the IBP with the data inputs from the stock-taking (and subsequent updates) providing its baseline data.

Annex I. Projects in the Public Investment Plan to be Reviewed

This annex presents some cases of issues that arose when analyzing the information in the PIP that raise doubts on the reliability of project data. Addressing these issues will help improve quality of data and identify quick wins for improving data management systems.

article image
article image
article image

Annex II. Structure of the Template

The following table presents a short description of the sheets were users will have to input project information. A more detailed explanation is available in the “Instructions” sheet of the individual project template.19

article image

Annex III. Information Included in Aggregation File

The following table presents a summary of the different fields that will be captured by the aggregation file from the project templates. It was decided not to aggregate 100 percent of the information captured by the template considering that not all of this will be comparable across projects and to avoid ending up with a file could become difficult to manage. The integrated bank of projects will have the capacity to capture all the information in the template and any other considered relevant by the authorities.

article image

Fiscal Affairs Department

International Monetary Fund

700 19th Street NW

Washington, DC 20431

USA

http://www.imf.org/capacitydevelopment

2

As of March 2017, the PIP multi-year commitments were deemed to be approximately $ 17bn, an amount similar to the 5-year medium-term envelope for development and investment projects. Yet, this number does not include large new projects, for example, the Standard Gauge Railway ($ 2bn), which questions the reliability of this data. See C. Roehler, J. Charaoui, M. Darcy, K. Khasiani, A. Navarro: “Uganda: Enhancing the Performance of Public Investment Management, Section II: Establishing Baseline Information on the Public Investment Plan,” IMF TA report, May 2017.

3

Which was known as Project Analysis and Public Private Partnerships at the time of the March 2017 mission.

4

Both PAP and the DB had initially expected that the entire exercise could be completed in time to prepare the FY 2018/2019 budget. However, given the challenges involved in this process, which are described in more detail in Chapter V, the authorities are reviewing what should be the best approach.

5

It is also intended to include local government projects in the IBP.

6

Refer to Chapter VII, Section C of the document C. Roehler, J. Charaoui, M. Darcy, K. Khasiani, A. Navarro: “Uganda: Enhancing the Performance of Public Investment Management,” May 2017.

7

Uganda National Roads Authority and Office of the Prime Minister.

8

This is further discussed in Chapter IV.

9

The PBS is focused on the annual allocation of resources for both recurrent and investment expenses. The AMP or DMFAS focus only on the resources coming from development partners but does not keep track on an initiatives milestones or future operating expenses.

10

All projects sponsors must provide information on a project’s total cost and distinguish between previous years’ commitments, those required for the next fiscal year, and forecast for outer years.

11

Illustrative only – this number should not be relied upon.

12

‘Team days’ means the number of days taken if only one team is employed. Therefore, the number is halved if two teams are employed etc.

13

It also demonstrates the value of the review function in conventional ex ante appraisal.

14

A quick review by the mission team indicated that there are in the region of 87 projects with an estimated value of more than UGX100bn.

15

Annex 3 provides more detail on the different components of the template developed for the stock taking exercise.

16

These categories were identified with PAP as the key items for project operation.

17

The stock-take will be limited to those projects in the PIP. It is well-known that there are other ‘development projects’ that are outside the budget, mainly in the health and education sectors, that are funded and managed by NGOs and charities which operate directly through local governments without recourse to the line ministries.

18

Personal instructions were given to selected PAP staff on how to use the aggregation tool.

19

In addition to these sheets, the template also has a Cover Page, an Introduction page explaining the stocktaking exercise, an Instructions page with more detail on the components of the template, a classifications page where PAP can modify the drop-down menus to be used during the exercise, and three “calculation” sheets (Strategy, Votecalc and Location) which should not be modified to ensure the appropriate operation of the template.

Uganda: Technical Assistance Report-Initiating the Stock-Taking of the Public Investment Program
Author: International Monetary Fund. Fiscal Affairs Dept.