Public Investment Management in Colombia 1 World Bank Group Governance Global Practice March 2018 1 This document was prepared by Tomás Martín, Pablo Andres Guzmán, Eguiar Lizundia and Pedro Arizti. It is based on the final reports from a Reimbursable Advisory Services (RAS) agreement between Colombia’s National Planning Department and the World Bank Group (WBG), which was implemented between 2015 and 2016. The WBG team was co-led by Pedro Arizti and Jorge Luis Silva from the WBG Governance Global Practice and comprised Tomás Martín and Eguiar Lizundia. The RAS final outputs and reports included comments from several units and colleagues from the National Planning Department. Contents I. Introduction .............................................................................................................................. 3 II. Framework for Public Investment Management .................................................................... 3 III. Peculiarities of Public Investment in Colombia ..................................................................... 5 IV. Institutional Arrangements for Public Investment Management ............................................ 7 V. The Diagnostic ...................................................................................................................... 10 A. Fractioning by source of funding ......................................................................................... 11 B. Confusion between management, fiduciary, and auditing controls ..................................... 11 C. Limited differentiation between project types and implementing agencies......................... 12 D. Formalistic controls ............................................................................................................. 13 E. Weak ex post review and evaluation .................................................................................... 14 F. Fragmentation of technical assistance .................................................................................. 15 VI. Recommendations .................................................................................................................. 15 A. Integral management and control throughout the public investment cycle ......................... 15 B. Clearly separate the management control from the fiduciary and auditing functions ......... 16 C. Differentiate project types and implementing agencies ....................................................... 16 D. Emphasizing results ............................................................................................................. 18 E. Strengthen ex post review and evaluation functions ............................................................ 18 F. Taking an integrated approach for providing technical assistance to implementing agencies .................................................................................................................................................. 19 References ..................................................................................................................................... 19 Acronyms and Abbreviations CADO Collegiate Administrative and Decision Organ IMF International Monetary Fund IPFD Investment and Public Direction (NPD) NPD National Planning Department MCES Monitoring, Follow-up, Control and Evaluation System MCS Monitoring, Follow-up, Control Strategy PIM Public investment management 2 I. Introduction 1. The Colombian Government has made significant progress in the implementation of effective processes, practices, and methodologies for public investment management (PIM). A 2011 IMF-published report ranked Colombia among the top 5 countries with best PIM practices out of a sample of 71 low- and middle-income countries (Dabla-Norris and others, 2011). The report assessed the following 4 key features: (a) strategic guidance and project review, (b) project selection and budgeting, (c) project implementation, and (d) project evaluation and auditing. Colombia was deemed to have achieved a good performance on 3 out of the 4 key features. 2. Nevertheless, the country’s PIM system still faces the need for areas of improvement identified by the World Bank’s PIM framework. 2 This framework allows for identification of not only the presence of must-have features in PIM systems that are key to promoting effective and efficient public investment but also if those features are implemented in the field and if they have achieved their intended purpose. In the case of Colombia, the main challenges identified were related to the broad implementation of tools designed — particularly for management of public investment projects as well as with the fractioning of standards and practices — and the overemphasis placed on formal controls during the formulation and project implementation stages. In addition, the country PIM processes had weaknesses in terms of the review and ex post project evaluation. 3. This document summarizes the main findings of a PIM study carried out in Colombia in 2015. The document includes an analytical framework for assessing public investment management and presents the assessment results as well as improvement recommendations to strengthen the institutional set-up to increase the effectiveness and efficiency of public investment in Colombia. II. Framework for Public Investment Management 4. The analytical PIM framework is based on the premise that a system is only as strong as its weakest link. In this regard, it seeks to strengthen each step of the public investment cycle and minimize risks that can generate inefficient investment or an inadequate use of public resources. During the 1980s, several countries invested heavily on generating capacity to perform cost-benefit analysis and established special units focused on the prior review of projects. However, these efforts were not matched by the development of mechanisms to ensure that the analysis effectively rejected unfeasible projects. Similarly, a feasible project loses may lose its rationale if delays on its implementation or operation change the cost-benefit calculations, either by an increase of costs or a reduction of expected benefits. Hence, every step of the project life cycle is crucial for effective and efficient public investment management. 2 For a detailed presentation of the PIM framework, see Rajaram and others (2014). 3 5. The PIM framework identified 8 must-have features that are essential to achieve efficiency in public investment management. Figure 1 illustrates the 8 features of the PIM cycle: Figure 1. PIM Cycle 1 2 3 4 5 6 7 8 Independent Guidance Appraisal Selection Implementation Adjustment Operation Evaluation review Ajustes Source: Rajaram and others (2014). 6. The first requisite for an efficient PIM system is the existence of specific and coherent national strategy documents that are useful to prioritize and guide public investment projects. The use of general guidelines allows for anchoring government decisions and guides sector decision-makers to align themselves with national priorities. These guidelines can comprise the National Development Plan, sector plans, or development plans from subnational governments. A critical feature of these documents is that they need to inform the budgeting exercise to serve as a credible guide for public investment. 7. Once any given public investment project is deemed to be satisfactorily aligned with government priorities, it must undergo a feasibility assessment. This will allow for evaluating its consistency and establish if the proposing ministry or government agency can proceed. This process requires complying with a series of pre-established steps, including preparation of pre- feasibility and feasibility studies as well as preliminary assessments of the project design, and its potential social and environmental impact. These steps must be completed before securing funding for the project, and must have clear and transparent guidelines. 3 8. After the formal project review is completed, it is advisable to submit it for an independent evaluation to mitigate potential over-optimistic biases in cost and benefits estimations. The introduction of independent reviews mitigates the risks of biases in the feasibility assessments and provides an objective mechanism for project selection. 9. Another critical characteristic of an efficient PIM system is establishing formal linkage between the project review and selection process with that of the budget preparation. While the project lifecycle may not necessarily coincide with the budget preparation schedule, the medium-term fiscal framework and yearly budget must establish budgetary ceilings for public investment (at an aggregate sector level). This will facilitate the establishment of a sustainable public investment program. 10. Once a project has been revised and given a budgetary allocation, monitoring during project implementation is key to ensure its success. Monitoring should cover both the physical and financial progress. In this regard, it is important to prepare a project operations manual that 3 In the case of Colombia, there is a differentiation between project “viability” and “feasibility”. Overall, the term “viability” is often used during the ex ante project evaluation. However, the use of this term across countries is referred to the financial assessment, whereas the term “feasibility” comprises a broader series of studies. For comparison purposes, the PIM framework refers only to the term “feasibility” during the ex ante evaluation stage to facilitate international comparison. 4 outlines clear implementation and monitoring arrangements as well as a project schedule. The capacity of project implementing agencies should also be assessed since they will oversee gathering of the necessary information as well as construct procurement plans and carry out procurement processes. With respect to financial monitoring, it would be ideal to have a comprehensive accounting system in place to manage costs at the project level and track the execution of contracts. 11. During implementation, there must be sufficient flexibility to modify the project’s annual operating budget, adjusting disbursements in line with possible changes on the field. However, a new project review should be carried out if significant changes occur both in terms of costs and schedule as well as on the expected impact on beneficiaries. If unexpected changes in the field, for example, reduce the project cost-benefit ratio, there must be a mechanism to reconsider the project rationale and whether the case demands suspending implementation. This flexibility implies that budget appropriations need to be tied to specific project phases, and each new disbursement request should be accompanied by an updated cost-benefit analysis. 12. The PIM cycle however does not end with project completion. Effective arrangements must be made for the transfer and asset maintenance before they become operational. In the case of infrastructure projects after civil works have been completed, there needs to be a process in place to ensure that the projects are ready for operation and that the necessary conditions for service delivery are met. In this regard, it is important to verify that the completed work does not require additional investments to begin operations. 13. Finally, a desirable but largely absent feature of PIM systems is the ex post evaluation of projects. Overall, a basic review should be systematically carried out for all PIM projects shortly after completion. This basic review should assess if the project was completed within the planned timeframe and budget (considering both original and amended if needed), and if it delivered the expected products stated at the approval stage. This review should be viewed as complementary to the financial audits usually carried out by the Comptroller’s Office. III. Peculiarities of Public Investment in Colombia 14. A critical peculiarity in Colombia is the agreement on an actual definition of what is “public investment”. The definition is broad and includes not only capital expenditure but also transfers and operating expenses and transfers. Under this definition, the wages of teachers and medical personnel are considered as public investment. The rationale for this classification is that investment made on human and social capital is as important as the one on fixed assets. 4 Box 1 elaborates further on the definition. 4 The official definition of public investment in Colombia is provided in Article. 41 of Decree 2710 issued in 2014. 5 Box 1. Definition of Capital Expenditure What constitutes capital expenditure is differently defined across the world. Generally, it refers to expenditure on physical assets with a lifetime that extends beyond a year, including improvements or the rehabilitation of physical assets that improve or extend its lifespan. This is different from the usual repairs and maintenance expenditures required to ensures that an asset remains operational throughout its expected lifespan. Capital expenditure is often associated with investment or development expenditure, which typically produce benefits that extend throughout several years in the future. This definition includes physical assets for government use (such as office buildings) as well as public assets that improve and support the development of the private sector (roads, hydric systems, etc.) and intangible assets like research and education. This represents a complexity and blurs the lines for classification of capital expenditure, particularly in cases where it receives preferential treatment on the yearly budgeting processes. This can lead to classifying both recurrent and non-recurrent expenditures as investments. Usually, each government establishes an arbitrary line to distinguish between capital and current (or operating expenses) expenditures. Current expenditure includes the procurement of assets that will be consumed during a year regardless of the case. Small expenditure items (e.g., those below US$10,000) are considered as current expenditure regardless if the asset could be consumed in a period beyond a year. Source: Jacobs (2009). 15. The Colombian definition of public investment establishes conditions to limit what can be considered as a public investment expenditure. Specifically, it requires investment projects to have a temporal deadline for implementation and must be associated with the state’s capacity to provide goods and services. Nevertheless, there are still investment projects that finance operating expenses and whose duration is rather undefined or whose duration is defined by norms, even though they could indefinitely continue operations given their nature. 5 16. Beyond the challenges entailed by the budgetary classification, the manner in which public investment and the concept of public investment is defined in the legal framework has serious implications for management and control. Specifically, tools and methodologies developed for the management and control of projects are implemented on activities that do not follow a project logic. Some significant implications of this type of inclusion errors are the following: a) Treating operating expenses as projects does not generate value added in terms of their management. b) Adjusting existing tools and methodologies for project management to non-project operations might disrupt said methodologies and tools and render them ineffective for actual projects. For example, in the nation’s general budget, good data regarding delays and cost overruns is lacking in project implementation. This happens because many investment projects are recurrent operating expenditures, and any information on delays in execution would lose relevance in comparison with cases of actual investment projects only. c) There is unnecessary administrative workload on part of the National Planning Department (NPD) and the project implementing agencies since investment projects 5 The definition of what constitutes an investment project can be found in Art. 5 of Decree 2844 issued in 2010, and in Art. 6 of Decree 1949 issued in 2012, for the purposes of the Colombia´s General Budget and the National Royalties System. The definition is the same for both sources of funding. 6 are only a fraction of the operations that comprise the public investment budget regardless of its funding source. d) The units and staff assigned to management and control of public investment lose their focus due to resources now being scattered across recurrent spending operations that should be administered with other tools. IV. Institutional Arrangements for Public Investment Management 17. In Colombia there are several funding sources for expenditure on public investment. This has also been reflected in diverse institutional arrangements. Currently, the main sources of funding are the following: a) Colombia’s General Budget, which is prepared by the central government. b) General Participation System, which is a fiscal transfers scheme to subnational governments. c) General Royalties System, which is an additional transfer scheme to subnational governments funded by revenues from non-renewable resources. d) Own revenue of subnational governments. Colombia’s General Budget, the General Participation System, and own local revenues sources can finance both operating and investment expenditure, but the General Royalties System can only fund investment expenditure. In addition, authority over the budget is split between the Ministry of Finance and the National Planning Department. Although final budgetary decisions are taken by the Ministry of Finance, the National Planning Department prepares and monitors the budget for public investment. 18. There are different institutional arrangements depending on the source of funding for public investment, with many actors involved. While the National Planning Department is a common stakeholder as an institution, the actual task corresponds to different units within this Department. This is particularly important given that there exist different practices and standards among NPD units. For example, the NPD Investment and Public Finance Direction (IPFD) plays a fundamental role in the management of investment projects funded by resources from the Colombian General Budget and instructs sector ministries to include maintenance costs in project formulation. In contrast, the Royalties Oversight Direction does not require that same information from subnational governments. Table 1 shows the different revenue sources with different responsible agencies and different implications for resource management across the PIM cycle. 7 Table 1. Responsible institution by source of funding across the PIM cycle Colombian General General Royalties General Participation Own revenue Budget transfers transfers sources • Fiscal Support Direction (MoF) • Subnational • Central government • Ministry of Education governments Guiding of public • National Planning • Ministry of Health • Subnational • Technical investment Department • Ministry of Housing governments secretariats and • Line ministries • National Planning CADOs 6 Department • Subnational governments Formal project • Line ministries • Royalties Oversight • Subnational government´s • Subnational review • Head of sector Direction (NPD) Planning Secretariats government´s ministries Planning Secretariats Independent • NPD IPFD • Royalties Oversight review • Technical NPD Direction (for selected Directions cases only) Selection • Line ministries • CADOs • Subnational government´s • Subnational • Head of sector Planning Secretariats government´s ministries Planning • NPD IPFD Secretariats Implementation, • Line ministries • Subnational • Subnational governments • Subnational adjustment and governments (in most governments project operation of cases) Monitoring and • Line ministries • Royalties Oversight • Subnational governments • Subnational control during • NPD IPFD Direction (NPD) • Fiscal Support Direction government´s implementation (Mof) Planning • Ministry of Education Secretariats • Ministry of Health • Ministry of Housing • Direction for Territorial Sustainable Development (NPD) Ex post review • Royalties Oversight and evaluation Direction (NPD) 19. In addition to the roles outlined, the National Planning Department also establishes basic project preparation methodologies and provides management tools and support information systems regardless of the funding source. The IPFD designed a methodology to guide the formulation of public investment projects irrespective of the funding and developed an on-line support tool to formulate projects with a standardized and consistent approach. Moreover, IPDF administers an information system to register public investment projects of both central and subnational governments as well as on-line monitoring tools for projects funded by General Budget resources and royalties. 20. The actors and responsibilities also vary when a project is co-financed by national and subnational governments. Generally, if a project receives any funding from the General Budget, it is immediately monitored by the National Planning Department. This is less likely the 6 The Collegiate Administrative and Decision Organs (CADO) are a committee usually comprised by representatives of the subnational government and the central government Municipalities and Departments have two representatives, whereas the central government has a single representative. Although decisions related to the strategic alignment of public investment are usually taken by consensus, the interests of subnationals have more weight in the decision. 8 case for projects financed through royalties or the general transfer systems, and even less so for those funded by subnational governments’ own local revenues. 21. Table 1 also includes the monitoring and control during implementation that, while not formally included in the PIM framework, is highly significant in Colombia. The National Planning Department has developed tools to monitor implementation of projects financed by the national budget and royalties. However, the focus of this control is on the fulfillment of formal requisites; and while there exists information regarding the achievement of project goals, the use of such information for management purposes is highly limited, especially in the case of general budget-funded projects. 22. In this regard, both the General Royalties System and the General Participations System have explicit control functions. In the case of the General Royalties System, this function is part of the Monitoring, Follow-Up, Control and Evaluation System (MCES), whereas the General Participation System has the mandate of the Monitoring, Follow-Up and Control Strategy (MCS). According to the legal framework (Art. 102 of Law 1530), the objective of MCES is to “prevent or correct the inadequate, inefficient, or ineffective use of resources, and the unfulfillment of legal requisites by beneficiaries and executing agencies”. In this regard, MCES includes preventive and corrective measures as well as sanctions triggered by system-generated alerts during project implementation. 7 The alerts are activated where there is no information reporting or when the provided information is inconsistent or incomplete. On the other hand, the control function within the MCS is to “ensure the proper execution of resources from the General Participation System as well as the fulfillment of continuity goals, coverage, quality, and proper delivery of the services entrusted” (Art. 3 of Decree 028, dated 2008). In this regard, the control function includes preventive and corrective measures associated with “risk events” identified during project monitoring and follow-up. Each risk event is in turn related either with noncompliance with information requirements or budgetary and administrative regulations for the use of public resources. Some of these risks include unauthorized financial transactions, errors in accounting registries or procurement processes, and failure to achieve coverage or quality and service delivery goals. 23. In contrast, while the control function in the nation´s General Budget is not completely comparable to those of the General Royalties System and the General Participation System, controls are specifically defined for the formulation stage of investment projects. These controls are placed to cover the phases of formulation and the feasibility assessment. These controls are performed directly by the line ministries, the head of sector ministries, and the National Planning Department. Overall, they aim to verify the following elements: • Alignment of investment projects with the national development plan’s objectives, sector policies, and the functions of each ministry; • Consistency with the medium-term expenditure framework; • Project financial closing; and 7For both the General Royalties System and the General Participation System, monitoring is referred to the systematic recollection of data to measure progress of policy, program, or project. Follow-up implies the selective supervision of a policy, program, or project in the field based on the monitoring information. 9 • Methodological consistency. 24. Finally, the independent review of subnational governments and ex post evaluation of projects are significantly weaker stages, except for projects funded by resources through the General Royalties System. While it is possible that some line ministries and subnational governments have processes in place to carry out these tasks, there are no common guidelines in this respect. The establishment of mandatory guidelines from the central government has been focused on the project formulation and on the implementation stages. The low emphasis given to the ex post review and evaluation of projects might be, among others, related again to the Colombian definition of what constitutes public investment, as many activities classified as such are operating and recurrent expenses, without a defined future end-date. V. The Diagnostic 25. The PIM system diagnostic through the PIM framework emphasizes the administration and management processes of public investment. The key underlying assumption is that the quality and the effectiveness of public investment depend on existing institutional arrangements and on processes in place to manage the whole PIM cycle. The diagnostic aims to assess the institutional arrangements in place taking into consideration the must- have features set by the framework for the system to be deemed effective and efficient. To this end, a questionnaire was developed to assess the system components and identify its functioning features. 26. Ideally, information obtained based on the questionnaire should be complemented with a collection of quantitative data that would empirically verify if the designed features are effectively fulfilling their functions on the ground. In the Colombian case, however, given the peculiar definition of public investment and what constitutes a project as well as the formalist character of ex post evaluation, the availability of this kind of information is quite limited. Commonly used data for contrasting is the alignment between a country’s infrastructure needs and the resources allocated to address them. Among others, the goal is to establish if the resources are sufficient and if the execution is spread across multiple sectors, thus diluting the expected impact. 27. In Colombia, it is difficult to obtain accurate budget figures of projects that can be classified as pure public investment. On the other hand, in the case of ex ante evaluation, the gatekeeping function is related to the existence of processes and procedures to ensure that financing is only granted to projects that are consistent with the government’s priorities and have proven positive benefits, for example, through a cost-benefit analysis. This function can be assessed by reviewing data of rejected projects during the formal review and selection stages. However, in Colombia, it is yet again difficult to obtain such information since the available data generally does not correspond to rejected projects but rather to projects that were returned to ministries for adjustment or corrections to comply with the established requirements. 28. Lastly, information on cost overruns and delays is key to assessing the functioning of a PIM system. In this regard, data from Colombia includes data from non-pure public investment projects, which significantly distorts the assessment. In addition, while there is some information 10 available for projects financed through General Royalties System transfers, this is difficult to obtain; and there are no available data for projects financed through General Participation System transfers or by local revenues in the case of subnational governments. 29. The diagnostic of Colombia’s PIM system has revealed several challenges that limit its capacity for an effective and efficient management tool. Some of the key identified challenges include (a) fractioning by source of funding; (b) confusion among the management, fiduciary, and auditing control functions; (c) limited differentiation between project types and implementing agencies; (d) formalistic in controls; (e) weak ex post review and evaluation; and (f) fragmented technical assistance to subnational governments. Some of these challenges were explicitly identified by the National Planning Department and other government entities while others are based on the analysis of the Colombian framework vis-à-vis international experience. A. Fractioning by source of funding 30. The PIM system is fractioned by the source of funding used for a given project. Each source of funding has its own management and control mechanisms for public investment. This entails a high risk for coordination as the different mechanism and tools available are not sufficiently articulated. Fractioning also generates significant operational challenges for the management and control of public investment resources since a given project can be funded by a combination of multiple sources. In addition to different procedures and schedules for each source of funding, the other challenges such as the following have emanated from fractioning: • Different standards for public investment management and control. Different criteria for managing public investment projects and the lack of common standards do present a challenge for projects financed with a combination of sources, particularly at the subnational level. This might not be the case for projects financed solely with funds from the General Budget, the independent nature of the Royalties’ budget, and the fragmented management and control responsibilities stated in the General Participation System. • Duplicate controls and management inconsistencies. The emphasis given on the source of funding rather than its use leads to duplicate controls and inconsistencies. This is reflected in the existence of multiple information systems for the monitoring and control of public investment projects in Colombia. • Confusion and work overload on subnational governments. Subnational governments face an unclear and confusing framework for public investment projects, which exacerbates the effect of their already limited institutional capacity. In addition, they are forced to deal with different systems that often entail duplicate reporting requirements and whose operation has human resources and workload implications. B. Confusion between management, fiduciary, and auditing controls 31. The current framework for public investment in Colombia lacks a clear separation between management control and fiduciary and auditing control functions. Given the available resources, management control should be based on effectiveness and efficiency and focused on production of results. On the other hand, fiduciary control and auditing functions should focus on ensuring the proper use of public money. In this regard, the management control 11 function should be under the responsibility of the Executive, whereas the fiduciary and auditing functions should be under the responsibility of independent entities. 32. The control exercised by the national government over public investment in Colombia has features that are generally a competence of Supreme Auditing Institutions and other fiduciary control institutions. The Colombian Constitution and the current legal framework 8 provide a mandate to the Executive to control and ensure the proper use of Royalties’ resources. However, this mandate has led to configuring a system excessively concerned with avoiding misappropriation of funds, which has important implications in the field. The following 3 short summaries identify problems with this approach: • Limited efficiency of control efforts. The lack of separation between management control and fiduciary and auditing control functions can duplicate and overlap control. Having several entities simultaneously focus on control in the proper use of public funds generates unnecessary workload and raises the issue whether this is the most efficient use of public resources. • Less control effectiveness. The proliferation of fiduciary control actions reduces the effectiveness of management control as it becomes a marginal concern. Moreover, auditing control is also weakened as resources for this purpose are spread across several low impact actions rather than on strategic targeted actions. • Attrition of control entities. The confusion on control functions produces an attrition effect on the responsible institutions. In the case of the National Planning Department, using the same system both to ensure the proper use of public resources and to carry out results- based budgeting management distorts the instruments and leads to administrative overload. C. Limited differentiation between project types and implementing agencies 33. The Colombian PIM system lacks a differentiated management approach depending on the project type. The Colombian definition of what constitutes a public investment project differs from international practice. This implies that recurrent expenditures are treated as investment projects and must undergo all the formulation, monitoring, and evaluation requirements applicable to capital expenditure. This lack of distinction is also replicated when looking at the costing of projects, which would allow to focus and target control efforts on those projects with the largest costs. 34. According to IPFD estimations, only 21 percent of projects executed with investment resources from the General Budget are focused in construction, improvement, and maintenance of infrastructure. These projects account for 20.8 percent of the overall investment budget. The remaining projects in the portfolio are not related to investment expenditures in a strict sense, but some of them do include the formation of gross fixed capital. To carry out its estimation, IPFD considered 22 types of expenditures, including administrative expenses and support to state services as well as support to policies and programs, technical assistance, and training. 8 Among others, the Legislative Act 05/2011, Law 1530 (2012), and Decree 1118 (2014). 12 35. The NPD’s Royalties Oversight Direction carried out a similar exercise for projects funded with resources from the General Royalties System. However, its estimation was not comparable since it considered different expenditure categories. The estimation considered only recurrent expenditures, expenses on studies and designs, goods, and investment, which in turn includes civil works, the prevention of diseases, higher education, and agroforestry programs. The comparison across funding sources requires defining a common methodology as well as stronger effort to identify operations that can be classified as pure capital expenditure but are in a sense “hidden” within larger projects corresponding to a different type of expenditure. 36. Likewise, there is no differentiation based on the type of executing agency despite the heterogeneous institutional capacity across subnational governments. In this regard, there is no differentiated approach for management and control that considers the technical and institutional strengths of the executing agencies. While the National Planning Department has been providing technical assistance to both central and subnational institutions, there are no streamlined guidelines to assist executing agencies with limited institutional capacity. 37. Applying the same methodologies and processes across the board on pure investment projects and recurrent projects alike is problematic. Likewise, the disregard given to the different institutional capacities of executing agencies can lead to a result such as the following: • Heterogeneity in project’s quality. The non-differentiated treatment implies that project formulation is carried out with tools and methodologies inappropriate for its nature and the capacity of the executing agency. Currently, a project for the construction of a primary road in a high-capacity urban department is subject to the same management and control procedures as the ones for paying transport subsidies on a low-capacity rural municipality. This implies that the latter must be formulated, monitored, and evaluated using a methodology that does not fit the type of project and its operation, which only entails paying a subsidy. • Lack of focalized management and control. Additionally, the lack of differentiation dilutes management and control and promotes excessive scrutiny on low-risk areas and actions, which can also lead to implementation delays. D. Formalistic controls 38. An additional challenge relates to the overemphasis placed on formal scrutiny and administrative control procedures on projects during the formulation and implementation stages. The PIM framework establishes that management controls should be focused on results as a good international practice. In contrast, a considerable share of the control activities in Colombia are centered on monitoring compliance with formal legal requisites during the preparation and throughout implementation. As a result, little attention is paid to actual efficiency or project efficacy. While it is important to monitor and control compliance with the legal framework, this type of control should be tasked to an independent entity rather than to an institution of the Executive branch like the National Planning Department. Overemphasizing formal controls within the Executive deviates resources and efforts that could otherwise have been focused on assessing project performance, vis-à-vis its declared objectives, through performance indicators or ex-post evaluations. 13 39. Moreover, the current formalistic nature of management and controls affects the quality of investment projects by generating incentives to focus on compliance with legal requirements rather than on ensuring the effective use of project formulation methodologies. Specifically, it can generate the following challenges: • Reduction of a project’s potential impact. During its initial stages, the focus on formal issues opens the door for the approval of projects with limited impact on policy priorities, or with uncertain benefits, but in full compliance with the legal framework. • Administrative fatigue. The formalistic approach on government entities generates attrition on the staff, which need to devote a considerable amount of time to the completion of forms and to input data on different information systems. E. Weak ex post review and evaluation 40. Weakness on ex post evaluation is another challenge that affects public investment management in Colombia. The effective use of ex post evaluations is a major bottleneck to substantially improve the quality and results of the PIM system. While the implementation of such evaluations differs by source of funding, there are no streamlined guidelines at the national level in this regard. Similarly, there are no guidelines for using ex post evaluation information for feedback on the programming and resource allocation. On the other hand, the emphasis is placed on implementing complex and profound impact evaluation methodologies, which are quite costly, overlooking more economical options that could contribute in a more effective way to improve project implementation. 41. The lack of an evaluation program not only generates management programs but also affects accountability and prevents the use of feedback to improve public investment projects. In addition, it causes: • Difficulty to establish fulfillment of the project schedule at closure. The absence of a basic review upon project completion implies that there is no information on delays and deviations with regard to the original budget, which are critical inputs to assess a project’s performance. • Uncertainty regarding the real impact of public investment. The evaluations currently carried out are selective and do not allow assessment if the initial cost-benefit estimations were accurate in measuring a project’s impact. • Low accountability of implementing institutions regarding results and project impacts. The lack of systematized information has consequences on transparency and the accountability of institutions responsible of project implementation. Without proper information regarding the project results, it is difficult for the national government to hold subnational governments accountable. 42. Overall, the weaknesses in ex post evaluation are missed opportunities for learning and the development of institutional capacity. Since information on project results is not 14 systematized and does not feed the decision-making process, it is not used for the preparation of new projects nor for providing tailored technical assistance to strengthen institutional capacity. F. Fragmentation of technical assistance 43. The provision of technical assistance to improve institutional capacity is fragmented. Currently, the efforts to construct capacity at the national, departmental, and municipal levels are not articulated and are not streamlined into a coherent strategy. On the one hand, capacity-building efforts are fundamentally restricted to project formulation and the preparation of reports. The National Planning Department has provided technical assistance to subnational governments on project structuring. However, this assistance has not covered subnational needs in terms of monitoring and management of projects. 44. The lack of a streamlined approach for technical assistance is reflected on the persistent low institutional capacity of many project implementation agencies, particularly for the implementation and ex post evaluation stages. This is one of the key challenges that need to be addressed to improve the efficiency and effectiveness of public investment in Colombia as the quality of projects greatly depends on the capacity of the implementing agencies. VI. Recommendations A. Integral management and control throughout the public investment cycle 45. There is some urgency in establishing guidelines, criteria, and common tools through the PIM cycle irrespective of the source of funding. The current differentiation, based on the source of funding, makes management dependent on the legal requirements established for the nation’s General Budget, General Royalties System and General Participation System funds as well on the regulatory framework of implementing agencies. A first step toward the implementation of a comprehensive approach is streamlining guidelines, manuals, and processes. 46. One way to avoid fractioning is centralizing management and control over public investment. The current organizational structure of the National Planning Department will require a significant effort to coordinate and streamline procedures, tools, and methodologies among the different units and actors involved in management and control of public investment. Although it is not immediately possible to discard the alternative of keeping the current organizational structure and engaging in a comprehensive coordination effort, centralizing several functions into a single unit could be a more feasible and, above all, efficient solution. This would enable obtaining results in a cost-effective way without the need of changing the whole organizational structure. This approach in public investment management has been implemented by several countries’ leaders who have a centralized unit for the PIM system with responsibilities for establishment of guidelines; control and monitoring and evaluation functions; and training and technical assistance. In case of deciding to centralize all or some functions related to the management and control of public investment in a single unit, the National Planning Department can begin with the functions related to project formulation and ex ante evaluation. These functions can be assigned to an already existing unit within the current organizational structure, or to a new unit or decentralized agency under the National Planning Department. 15 47. The purpose of a centralized unit would be to promote the effectiveness and efficiency of public investment through integral management and control methodologies, procedures, and systems. The potential impact of creating a centralized unit is mainly related to operational efficiency gains at both national and subnational levels. These gains are due to the creation of economies of scale, the simplification of the interaction between implementing agencies, and the standardization of processes as well as due to the consolidation of resources for monitoring and technical assistance. The decision regarding the centralization of functions in an existing or new unit is up to the Colombian Government. B. Clearly separate the management control from the fiduciary and auditing functions 48. The separation of management control and fiduciary control are key features of a modern PIM system. This would enable focusing efforts and resources on ensuring the efficient use of public resources and monitoring the actual impact of public investment. In this regard, the central government should delegate fiduciary and auditing control, as well as the faculty to issue penalties, to the Office of the Comptroller General. The role of the National Planning Department should be focused on improving the execution of public investment (efficiency), and its results (effectiveness). These functions should be concentrated on a single unit as discussed above. However, given the functions and activities defined for General Royalties System projects in the legal framework, the separation of control functions would entail the changes to the legal norms in the medium term. Nevertheless, in the short run, the National Planning Department could study ways to steer the control and penalties mechanisms established in the General Royalties System Law toward a results-based management. Resolving the confusion between functions would allow strengthening the management, control, and evaluation functions and would reorient incentives within the National Planning Department from a formalistic paradigm to one focused on promoting the effectiveness and efficiency of public investment. C. Differentiate project types and implementing agencies 49. Differentiating the types of projects and implementing agencies is key to improve public investment management. The establishment of management and control procedures differentiated by project type is a feature present in the systems of the United Kingdom, South Korea, and Ireland (Box 2). In the case of Colombia, the implementation of such a model would raise three key actions that would address the current limitations: (a) establish a definition of public investment and investment projects more aligned with the economic function and international best practices, (b) define thresholds to differentiate between project types and management processes, and (c) develop a methodology for technical assistance that explicitly considers the different capacities of implementing agencies. 50. Differentiated management by project type. As was previously discussed, several projects that comprise the current public investment portfolio in Colombia are not capital expenditure in a strict sense. While a substantial legal reform would be required to adjust the definition to international best practices, an interim solution would be to classify projects on two broad categories. The first of these projects are those aimed at gross fixed capital formation or sustainable human development, which have a clear start and end date and where there is a clear value chain linked to expected results. These projects would undergo all PIM framework stages. Within this category, an additional distinction would have to be made to classify projects based on their cost, which would serve as a risk proxy to focus control. To this end, it is recommended to 16 establish monetary thresholds. The differentiation of projects by costs would allow the National Planning Department to identify those projects whose failure would entail a significant loss for the government and thus require closer supervision. The second broad type of project would include those that seek to support the production capabilities or the provision of goods and services within a program value chain. This category would include operations such as subsidies, transfers, and acquisition of assets as well as other operating expenditures. This type of projects should not undergo all PIM framework stages. Box 2. Thresholds for differentiating public investment projects in selected countries Several countries across the globe have implemented differentiated treatment of projects based on their cost. Below is a summary of some examples and their implications. United Kingdom. Road construction projects require approval from the Ministry of Finance when they surpass 500 million pounds. The involvement degree of the Ministry of Finance in other transport projects depends on scale and complexity. South Korea. All projects with a cost equal or above 50 KRW billions (US$50 million) require pre-feasibility studies. This was originally required for infrastructure projects only. The National Finance Act expanded this requirement to other types of projects, including research and development. The projects prepared by subnational governments or by public-private partnerships also require pre-feasibility studies if the share of central government participation exceeds 30 KRW billion. Ireland. The Ministry of Finance has developed guidelines for the review of public investment projects based on their scale, according to the following schedule: • All projects under 500,000 Euros require a simple review only. • Projects between 500,000 and 5 million Euros are required to undergo a single technical review process that includes elements of a preliminary detailed appraisal. • Projects between 5 million and 30 million Euros require a detailed review with using a multi-criteria approach. • For projects with a cost beyond 30 million Euros, it is necessary to carry out a complete cost-benefit analysis. This requirement is also applicable to innovation projects over 5 million Euros, which involve specialized sectors and the use of new technologies. This requirement is also applied if the project is a pilot for a larger program and if the use of new technologies might generate substantial additional operating and maintenance costs. Source: Authors, based on Rajaram and others (2014) 51. Differentiated control by implementing agency. To improve the efficiency and efficacy of the current system, it is key to acknowledge the differences in the institutional capacity across project implementing agencies. This is important not only within central government institutions but also at the subnational level. It is also important to clearly distinguish between the institutional capacity risk of a given agency and the inherent project risk. In this regard, the feasibility assessment plays a key role in the ex ante project evaluation and includes a careful study of risks that an individual project might entail. However, institutional capacity risks related to the implementation agencies refer to a different kind of event or condition that may affect not only an individual project but potentially the project portfolio executed by a given agency. In the first case, the unit of analysis when determining the risks is the project; whereas in the latter, it is the implementing agency. A low institutional capacity in each entity can affect the development of projects under its responsibility and thus the achievement of the expected results and outputs in an indirect way. 17 52. A differentiated approach can bring management and control improvements across the whole PIM cycle by adjusting the formulation, programming, implementation, and evaluation according to the capacity of the implementing agency. This approach would allow achieving better results in terms of quality in the formulation stage and increase the efficiency in the allocation of control and management resources to maximize their impact. Likewise, by considering the different institutional capacities, the planning and management of small projects is simplified as well as for other expenditures incorrectly classified as public expenditure, which entails a significant reduction of the current administrative workload. D. Emphasizing results 53. It is crucial to reorient the focus of control on public investment management toward the achievement of results. PIM systems are increasingly focused on ensuring that projects achieve their expected outcomes and impact. In Colombia, this requires linking projects to budget programs and ensuring alignment with the government’s policy priorities. Additionally, it is important to monitor delays and costs overruns as part of the usual reporting requirements, concentrating on high-risk projects implemented by low-capacity agencies. Likewise, it is necessary to provide technical assistance to subnational governments on project implementation. 54. It is crucial to increase the impact of public investment by producing and embedding performance information on decision-making throughout a project’s implementation. This means that the National Planning Department would need tools to ensure that the formulation and approval of projects is coherent with the policy priorities. Performance information should include data on inputs and activities but must also emphasize the data regarding products and results. The monitoring system should at least contain information on expected products and also on the variables associated with the achievement of expected results or impacts, which provided the rationale for implementing the project in the first place. For example, when building the infrastructure for a school, there must be certainty with regards to the dates on which the school will begin operating as well as the number of students that attend. In addition, it is necessary to have information regarding the quality of the services provided and the students’ performance. Underlying this data requirement, linkages must be established among inputs, activities, products, and outcomes to establish whether an intervention has been successful and to assess whether the products have effectively had an impact on variables associated with the outcomes or results. This information should be available to the National Planning Department and the public to enhance accountability and provide incentives for project managers to improve their performance. A final step in this regard would be associate budget allocation decisions to entities performance. E. Strengthen ex post review and evaluation functions 55. Strengthening ex post review and evaluation is crucial for an efficient and results- oriented PIM system. The implementation of basic ex post review processes will provide a quick and cost-effective way to identify gaps in the project planning and formulation process. To this end, standardized selection criteria and ex post review and evaluation methodologies need to be developed to focus on assessing results, costs, and delays of projects and to provide inputs to strengthen the institutional capacity of implementing agencies during implementation. Establishing a specific team focused on these tasks within a central unit, would greatly aid to improve these processes. 18 56. An improved ex post evaluation in Colombia would have a positive impact on transparency and accountability. Having better and timely information upon project completion would allow assessment of the degree of responsibility of the implementing agency in achieving results and enhancement of accountability both within the government and toward citizens. Moreover, greater feedback based on the gathered information will provide valuable inputs to improve forthcoming projects as well as identify and address institutional capacity limitations. F. Taking an integrated approach for providing technical assistance to implementing agencies 57. A technical assistance program is an important feature of any PIM system. The National Planning Department should promote the creation of an integral, institutional technical assistance program covering the whole project cycle and targeting underperforming implementing agencies. It is important that the provision of technical assistance be the responsibility of a single specialized unit to avoid the current fragmentation. 58. An integrated approach would be reflected in greater institutional capacity of implementing agencies, not only in terms of planning and project formulation but also regarding management practices and monitoring and evaluation practices. While the impact of the technical assistance will yield benefits in the medium term, it is key to ensure quality in the PIM system, especially at the subnational level. VII. References Dabla-Norris, Era, Jim Brumby, Annette Kyobe, Zan Mills, and Chris Papageorgiou. 2011. Investing in Public Investment: An Index of Public Investment Efficiency. IMF Working Paper. Washington, DC. Jacobs, Davina F. 2009. Capital Expenditure and the Budget. IMF Technical Guidance Note. Washington, DC. Rajaram, Anand, Tuan Mihn Le, Kai Kaiser, Jay-Hyung Kim, and Jonas Frank (editors). 2014. The Power of Public Investment Management: Transforming Resources into Assets for Growth. Washington, DC: World Bank Group. 19