WPS8602 Policy Research Working Paper 8602 Prioritizing Infrastructure Investments A Comparative Review of Applications in Chile Darwin Marcelo Schuyler House Aditi Raina Infrastructure, PPPs & Guarantees Global Practice October 2018 Policy Research Working Paper 8602 Abstract Governments worldwide face the difficult challenge of reservoir subsectors, respectively. The results show that the deciding which infrastructure projects to prioritize and Infrastructure Prioritization Framework has application select for implementation, given the limits of available beyond its original proposition and can complement a tra- funding and the need to attain their developmental goals. ditional cost-benefit analysis by directly considering social The key objective of this report is to conduct a compar- and environmental policy goals that are otherwise diffi- ative exercise between the World Bank’s Infrastructure cult to quantify in a cost-benefit analysis. The analysis also Prioritization Framework, a multicriteria analysis–based finds that in Chile there is a discrepancy between the stated methodology to project prioritization, and a more complex goals and objectives of the appraisal system and the actual cost-benefit analysis–based approach. The report focuses implementation. In the case of transport sector projects, on Chile, which has a well-institutionalized evaluation there is an evident deviation between cost-benefit analysis– process that uses cost-benefit analysis to assess projects on based selection policy and actual decisions made for project their quality and ability to generate value for money. The implementation. In the case of water catchment selection, analysis compares the results of the Infrastructure Priori- there is a bias toward projects with higher financial-eco- tization Framework alongside Chile’s current cost-benefit nomic performance as compared to social-environmental analysis–based and multicriteria analysis approaches to the performance, despite policy intentions to afford consider- same subsets of projects in the road transport and water ation to environmental and social development goals. This paper is a product of the Infrastructure, PPPs & Guarantees Global Practice. It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://www.worldbank.org/research. The authors may be contacted at dmarcelo@worldbank.org. The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. Produced by the Research Support Team Prioritizing Infrastructure Investments: A Comparative Review of Applications in Chile Darwin Marcelo, Schuyler House and Aditi Raina Keywords: Infrastructure prioritization, infrastructure planning, public investment, principal component analysis, multi-criteria analysis, transport, water JEL Classification Codes: R42, O18, O21, O22, H54, C38 Table of Contents Table of Contents ................................................................................................................................................ 2  Abbreviations ................................................................................................................................................... 4  Acknowledgements ........................................................................................................................................ 5  Chapter 1. Introduction ...................................................................................................................................... 6  Approaches to Infrastructure Appraisal and Selection ......................................................................... 6  Comparing Approaches: A Case Study of Chile ....................................................................................... 8  Chapter 2. Infrastructure Appraisal and Selection in Chile ..................................................................... 10  Evolution of Project Appraisal and Selection in Chile ........................................................................... 10  Chile’s Project Investment Cycle ................................................................................................................ 11  Project Appraisal and Selection in Chile .................................................................................................. 13  Extending Investment Decision Support in Chile .................................................................................. 14  Multi-Criteria Analysis to Support Investment Decision-Making .................................................... 15  Chapter 3. Infrastructure Prioritization Framework: An Alternative Approach .................................17  The IPF Process ...............................................................................................................................................17  Step 1. Select Criteria ................................................................................................................................... 18  Step 2. Prepare Data .................................................................................................................................... 19  Step 3. Constructing Performance Indices – SEI and FEI .................................................................... 19  Step 4. Creating the Visual Interface: The Investment Prioritization Matrix ................................ 20  Evolution of the IPF ....................................................................................................................................... 21  Pre-Analytical Steps .................................................................................................................................... 21  Technical Improvements: Variable Specification and PCA Restrictions ........................................ 21  Sensitivity Analysis and Criteria Weighting .......................................................................................... 22  Organizational and Capacity Issues ......................................................................................................... 23  Chapter 4. Applying IPF to Chile Infrastructure Project Proposals ....................................................... 24  Water Catchment Projects in Chile .......................................................................................................... 24  Water Catchment Project Sample ........................................................................................................... 24  2 Water Catchment Project Indicators ...................................................................................................... 25  IPF Results: Water Catchment .................................................................................................................29  Water Catchment IPF Matrix ....................................................................................................................33  Comparing IPF to Selection of Water Catchment Projects ............................................................... 34  Applying IPF to Road Transport ................................................................................................................36  Transport Policy Goals and Road Project Criteria ................................................................................36  Road Transport Project Sample ...............................................................................................................36  Transport Project Indicators ..................................................................................................................... 37  IPF Results: Road Transport ..................................................................................................................... 38  Road Transport IPF Matrix........................................................................................................................ 39  Comparing IPF to Funding of Transport Projects ................................................................................ 39  Chapter 5. Conclusion ..................................................................................................................................... 40  References .......................................................................................................................................................... 42  Annex 1. Chilean Law Relevant to Project Appraisal and SNI ............................................................. 44  Annex 2. Water Catchment Raw Project Data ...................................................................................... 45  Annex 3. Water Catchment Project SEI Calculations, by Region ...................................................... 47  Annex 4. Water Catchment Project FEI Calculations, by Region...................................................... 50  Annex 5. Road Transport Raw Project Data ...........................................................................................53  Annex 6. Road Transport Project SEI Calculations, by Region (with standard poverty rate) .... 54  Annex 7. Road Transport Project SEI Calculations, by Region (with multidimensional index poverty rate) ...................................................................................................................................................56  Annex 8. Road Transport Project FEI Calculations, by Region .......................................................... 58  3 Abbreviations BIP Integrated Project Bank CBA Cost-Benefit Analysis CFA Centralized Finance Agency CORFO National Development Corporation DIPRES Budget Office ESP Social Project Evaluation IDI Investment Initiative IPF Infrastructure Prioritization Framework IRR Internal Rate of Return MDS Ministry of Social Development MIDEPLAN Ministry of Planning and Cooperation MCA Multi Criteria Analysis MCEM Multi Criteria Evaluation Methodology NPV Net Present Value ODEPLAN National Planning Office RATE Result of Technical-Economic Analysis RS Recommended Favorably (according to RATE) SCBA Social Cost-Benefit Analysis SNI National Investment System 4 Acknowledgments This report was prepared by a team of experts from the World Bank's Infrastructure, PPPs and Guarantees Group with inputs and guidance from Pilar Contreras García, Head of Unit Public Investments and Non-Financial Assets (DIPRES) at Ministry of Finance, and Eduardo Koffman, Coordinator of the Planning and Development Department at Ministry of Transport and Telecommunications. We would also like to thank the Department of Irrigation Project at Ministry of Public Works for providing all the project-relevant data information for the water reservoirs analysis. The World Bank team included Darwin Marcelo (Task Team Leader), Schuyler House and Aditi Raina. Cledan Mandri-Perrott and Jordan Schwartz provided essential guidance and oversight. The team would also like to thank the World Bank Singapore Infrastructure Hub for its support.   5 Chapter 1. Introduction Infrastructure services are significant determinants of economic and social development and are typically prominent components of national development plans. While national governments and their central finance agencies (CFAs) often consider numerous project proposals from agencies, line ministries, and various sub-national government units, financial resources are often insufficient to fund the full set of proposals, particularly in the short-term. Global estimates of infrastructure investments required to support economic growth and human development lie in the range of US$65 trillion to US$70 trillion by 2030, while the estimated pool of available funds is limited to approximately US$45 trillion. Governments worldwide face a two-pronged challenge; to increase the pool of funding available for infrastructure development and to make difficult decisions about which projects to select for implementation, given the real limits of available funding. This paper deals with the latter challenge –namely the need for CFAs, ministries, and other relevant agencies to prioritize potential infrastructure projects aligning needs with fiscal constraints while attaining their respective economic and social development goals. The key objective of this report is to conduct a comparative exercise between the World Bank’s Infrastructure Prioritization Framework (IPF), a multicriteria-based methodology to project prioritization, and a more complex Cost Benefit Analysis (CBA) based approach (i.e. more data and analytically intensive). To this end, the report focuses on the infrastructure prioritization process in Chile, a country that is recognized for the strength of its institutions and capacity of its public administration. Chile has a well-institutionalized evaluation process that uses CBA to assess projects on their quality and ability to generate value for money. The Ministry of Public Works (Ministerio de Obras Públicas – MOP) has been recognized for its capacity to prepare and implement high-quality infrastructure projects (OECD, 2017). This report explores the theoretical and practical challenges of prioritization; the robustness and integrity of current approaches; and the comparative outcomes of these approaches and their alternatives. This exercise is intended to serve practical ends. The purpose is to progress discourse on project prioritization and selection in order to validate useful and productive public administration guidance, on the one hand, and create space for alternative approaches to support sound investment decision-making and responsiveness to non- monetizable policy aims, on the other.  Approaches to Infrastructure Appraisal and Selection In addition to a growing infrastructure gap, the past 20 years have also seen a shift towards decentralized infrastructure planning. Many subnational governments, regional entities, and sector agencies have been delegated responsibility for infrastructure planning to promote local responsiveness. Moreover, while spending ceilings are defined by the centralized finance agency (CFA), allocation of funds for implementation remains at the line ministry level. At the national level, decision-makers must deal with numerous project proposals, each with varying amounts of attendant project information. These projects must be ideally appraised, compared, and selectively allocated funds for implementation. The framework on Public Investment Management (PIM), proposed by Rajaram et al. (2014), is useful for guiding governments through the processes of infrastructure planning, appraisal, investment, and implementation, with an eye to increase the effectiveness of infrastructure 6 investments. PIM identifies eight key “must-have” features of an effective public investment management system (see Figure 1). Project selection should follow first-level screening, project appraisal, and independent review.1 Figure 1. Key Features of a Public Investment Management System Making decisions as to which projects should be implemented implies grappling with efficiency and effectiveness of proposed investments, monetizable project costs and benefits, non- monetizable social and environmental impacts, and the relationship of these aspects to national and sub-national development plans. Because so many factors must be considered, the use of decision support frameworks and methods can help systematize appraisal and selection. Prioritization frameworks should be rigorous and comprehensive enough to accommodate multiple facets Source: Power of Public Investment Management (Rajaram et al., 2014) of infrastructure development, but also sufficiently practical to implement. Best practice in public management and traditional policy analysis suggest that economic appraisals (preferably full social cost-benefit analysis when the main costs and benefits are measurable and there is an economic price available for them) and feasibility studies provide sound bases for project prioritization, using highest societal net present value (NPV) (or a variation thereof) as a ranking metric, along with assessing a project’s fit with infrastructure policy guidance (Rajaram, Tuan, Bileska, & Brumby, 2014, p. 20).2  In practice, however, capacity, resources, and time are often too short in supply to support extensive social cost-benefit analysis (SCBA) across full project sets and sectors. Also, in many cases, it is not possible to value the main benefits of a project, such as cultural or health investments, even if those benefits are identified and measured. Decision-makers often only have partial information on project costs and benefits, particularly since many are difficult to quantify and monetize. The PIM approach proposes that, in cases of restricted capacity or resources, basic elements of project appraisal should be applied. This includes a good justification for a project, clearly-specified objectives, comparison of alternatives, detailed analysis of the best options, fully-estimated project costs, and qualitative assessment of project benefits to justify costs (Rajaram et al., 2014, p. 8). Facing restricted information and capacity, a risk arises of falling into unsystematic project selection. In these cases, decision frameworks based on multi-criteria analysis can help government decision-makers (a) systematize prioritization based on key development goals; (b) make best use of available information; and (c) formalize clear decision criteria to promote 1 First-level screening should be done to ensure that projects align with the development strategy and meet basic requirements for budget inclusion as a project (Rajaram et al., 2014). 2 In Chile, for example, analyses utilize the NPV index (SNPV), which equates to the NPV of future costs and benefits divided by the investment level. The ranking is then determined by sorting the highest SNVP to the lowest. 7 accountability. The World Bank’s Infrastructure Prioritization Framework (IPF) is one such multicriteria analysis (MCA) approach that condenses government-selected project indicators into composite financial-economic and social-environmental indices. The analysis may incorporate the results of financial or partial social cost-benefit analysis, but does not require full SCBA. Comparing Approaches: A Case Study of Chile The IPF has been piloted in Vietnam, Panama, Argentina, and Sri Lanka. These pilots imparted methodological and practical lessons that have been used to adjust and improve the IPF. An important unanswered question remained, however, as to how effectively the IPF can substitute for the best practice of project appraisal and selection based on SCBA. Moreover, while IPF was designed as a ‘next-best’ prioritization approach based on ‘less-than-SCBA’ appraisal, IPF may nevertheless have something to offer countries where SCBA/CBA approaches are institutionalized. For these reasons, the IPF was additionally piloted in Chile, where CBA-based analysis is a standard input to project acceptance for economic infrastructure such as roads, transfer ports, dams, railroads, etc. Chile stands apart from much of the world with respect to systematic, institutionalized project appraisal and evidence-based project selection. The Government of Chile (GoC) has a centrally-managed Public Investment System (SNI) that separates project proposal (initiated by line agencies and sub-national units) from appraisal, selection, and budget allocation performed by the Ministry of Social Development and the Ministry of Finance. The SNI is used to consolidate project information, subject proposals to policy filters, and appraise projects before inclusion in sector plans and budget requests. The Chilean SNI is likely the most systematically managed and consolidated investment appraisal system in Latin America (de Rus Mendoza, 2014) and is generally seen as a good example of a “structured and coherent framework for identifying, coordinating, evaluating and implementing public investments” (OECD, 2016, p. 93). The ready availability of CBA appraisals in Chile proffered a valuable opportunity to compare IPF outcomes with CBA-based project selection. Moreover, the IPF is relevant in the Chilean context for other reasons. For one, the government recognizes the value of additional policy considerations alongside the results of CBA and has implemented a multi-criteria approach to project selection in the water sector, indicating recognition of the value of MCA even where cost-benefit analysis is widely applied. Second, the CBAs employed for sector-level project selection deviate somewhat from the academic policy approach to SCBA, due primarily to the realities of time and resource demands. Most appraisals – particularly for small- to medium-size projects – are partial (financial) CBAs that rely on highly standardized assumptions and often yield results with limited variance across projects. This exposed the possibility for MCA approaches to help fill in the missing considerations in partial CBAs. This report presents an overview of the current system of project appraisal and selection in Chile, a summary of the IPF methodology and its evolution, and the comparative results from applying IPF alongside Chile’s current CBA and MCA approaches to the same subsets of projects in the road transport and water catchment subsectors. The report follows with a discussion of the findings of the exercise. The results of this exercise show that the IPF has application beyond its original proposition of being a stop-gap measure until more sophisticated project appraisal methods can be implemented. This is because it can complement a traditional CBA by directly considering social 8 and environmental policy goals that are otherwise difficult to quantify in a CBA analysis. In addition, this exercise led to two deeper findings that went beyond the initial aim of merely comparing the IPF and CBA results. The first was that CBA analyses are not used as the basis of prioritization in Chile. There was an evident deviation between CBA-based selection policy and actual decisions made for project implementation, in the case of transport sector projects. The second was the fairly consistent alignment of IPF- and MCA-based prioritization in the case of water catchment selection, but with a surprising inclination towards projects with higher financial-economic performance as compared to social-environmental performance, despite policy intentions to afford key consideration to security, environmental, and social development goals. Therefore, in both cases, there is a discrepancy between stated goals and objectives of the appraisal system and the actual implementation. 9 Chapter 2. Infrastructure Appraisal and Selection in Chile The SNI is essentially a set of processes, data collection mechanisms, and appraisal functions that support project selection across multiple sectors. By design, and with the overarching policy goal of promoting economic growth, project selection is generally based on social net present value. Because SCBA does not consider distributional effects and regional or territorial inequalities, the SNI is complemented by a cost efficiency analysis when ‘desired but non-quantifiable’ social or environmental outcomes are deemed significant enough to justify project costs. Chile has developed a CBA-based system for investment decisions (Candia et al., 2015). Over time, the investment system has evolved, most recently by extending the appraisal approach to consider additional factors via multi-criteria analysis (MCA). The following section provides an overview of the evolution of Chile’s infrastructure investment system and its current technical and institutional aspects. Evolution of Project Appraisal and Selection in Chile Chile’s Sistema Nacional de Inversiones (National Investment System) (SNI) is a centralized public investment system jointly administered by the Ministerio de Desarrollo Social (Ministry of Social Development) (MDS) and the Ministerio de Hacienda (Ministry of Finance), via the Dirección de Presupuestos (Budget Office) (DIPRES). MDS is responsible for ex-ante project appraisal and ex- post evaluation, as well as systematic data collection and reporting, while the Ministerio de Hacienda (through DIPRES) sets the public budget. The SNI is the latest organizational arrangement in an extended history of formalized project appraisal and selection. The genesis of Chile’s investment system was the Corporación Nacional de Fomento (National Development Corporation) (CORFO) established in the 1950s, created to evaluate the financial and social impacts of national projects and units, with a strong emphasis on state enterprises. This agency’s role in investment decision-making was assumed in the 1960s by the Oficina de Planificación Nacional (National Planning Office) (ODEPLAN), which gave rise to the first formal project appraisal system. ODEPLAN also served as a platform for developing government capacity for project appraisal, leading to the creation of a specialized Social Project Evaluation unit (ESP). Through the 1970s, Chile developed extensive guidance, processes and methodologies for project appraisal and developed an integrated investment decision-making system that specified the roles and relationships among the ministries and other governmental units. Institutionally, this would be consolidated in the 1990s with the creation of the Ministry of Planning and Cooperation (MIDEPLAN), later renamed the Ministry of Planning in 2005, and replaced by the Ministry of Social Development (MDS) in 2011. The national appraisal system employed cost-benefit analysis from its earliest days to appraise proposed projects. During the 1970s, however, it was recognized that some projects – particularly in social sectors like health and education – involve social benefits that are difficult to estimate, but which may be assumed high enough to outweigh project costs. This led to the adoption of a ‘cost-efficiency’ approach to appraisal in some sectors, wherein effort is concentrated on minimizing costs to attain the desired outcome (often the provision of basic services), with qualitative assessment of expected benefits. The cost efficiency approach also allowed the government to deal with distributional effects and regional inequalities. The government employed hybrid approaches to investment appraisal, including cost-efficiency and multi-criteria approaches, to support project prioritization. The cost-efficiency approach was 10 often applied by combining expected benefits into a single composite indicator to be compared to project costs.3 While CBA remained the standard project appraisal method, cost-efficiency approaches were used to appraise 71% of projects proposed between 2000 and 2015, accounting for 47% of proposed investments (Agostini and Razmilic, 2015). Therefore, a spectrum of appraisal techniques has been in use since the 1970s. Over the past few decades, the CBA and MCA methodologies have developed with respect to valuation approaches, applied assumptions, and analytical sophistication, but the overall methodologies and institutional frameworks have remained quite the same. Chile’s Project Investment Cycle By law, except the armed forces (which have their own systems), any public-sector institution wishing to develop an investment project must do so via the SNI. Law pertinent to this process is described in Annex 1. The proponent unit initiates the process by submitting background project information to the SNI. This information is immediately available to the public via an open digital registry called the Banco Integrado de Proyectos (Integrated Project Bank) (BIP). The BIP provides a record of all project proposals in standardized format and tracks project development from initial proposal through ex-post project evaluation. The proponent agency engages in an iterative process of submissions and approvals with MDS, via the SNI, that involves increasing levels of detail with respect to project appraisal as the project progresses through the system. Upon initial submission to the SNI, a project is assigned a unique identification code within the BIP, which can be used to track the projects’ progress through these stages. Except for some project types (e.g., projects with pre-approved designs), project proponents must submit project preparation information to the SNI, via the BIP platform, at the following stages:  Profile (concept): the policy problem is described, along with the purpose and context of the project, alternative solutions under consideration, and an assessment of the feasibility and impacts of various alternatives to inform the selection of the most viable alternative;  Prefeasibility: prefeasibility studies include additional project details, including tentative schedules, budgets, and more extensive information on expected benefits;  Feasibility: full feasibility studies, including CBA or cost-effectiveness analysis are provided;  Design: technical architectural, engineering, and construction studies are done, and the timings of investments and detailed budget are specified. Project execution plans are required to be based on specific estimates of the costs of equipment, personnel and supplies, as well as a realistic schedule to estimate the duration of the various activities required; and  Execution: the project is approved to seek funding. Some projects can apply to the design and execution stages simultaneously when the main sources of risk are known. Simple projects may not include detailed information at every step, however, and may move directly to project execution from the project profiling stage. At each stage, MDS assesses the project and approves or rejects its progression to further development, depending on whether it meets the requirements of each stage. It then issues a 3 Pilar Contreras, an economist with a long career in public investment in Chile (ODEPLAN) and currently serving as Chief of Investment (DIPRES), reported in an interview that, where CBA was not feasible for health and education projects, analysts used relevant variables common to all projects (e.g., malnutrition, education, infant mortality, etc.) to construct a weighted, combined single indicator reflecting each project’s projected impact. 11 Resultado del Análisis Técnico Económico (Economic Technical Analysis Result) (RATE), which results in one of the following RATE results: a) Recommended Favorably (RS); b) Missing Information (FI), to specify that records lack required information necessary to secure favorable recommendation; c) Technical Objection (OT), reflecting a negative assessment; d) Reassessment (RE), wherein the project is recommended for additional analysis; or e) Breach of Regulations (IN), when spending is executed without the support of MDS. Projects must attain a favorable RATE (RS) at each stage to move to the next stage of development. For typical projects, the information required to pass each stage is summarized in Table 1. Table 1. SNI Informational Requirements for Investment Project Assessment Stage Transition  Submission Requirements  Pre‐investment study containing:    Definition of the problem   Analysis of supply and demand  Profile to Prefeasibility / Feasibility   Study of solution alternatives   Initial cost estimates   Preliminary strategic and economic evaluation   Further specification of best solution  Prefeasibility / Feasibility to Design    Detailed budget   Feasibility study   Detailed line‐item budget   Design to Execution   Full engineering design   Draft bidding proposal  Sources: Government of Chile (2017).  Standards, Instructions, and Procedures for the Public Investment Process  (PIN); Presentation: Public Investment Management Conference: The Chilean Experience  Once a project moves to the Execution stage, the proponent may seek funding for the project (in the SNI, the designation ‘Execution’ simply reflects authorization to seek funding, but does not necessarily mean that the project is funded or under implementation). Projects are typically funded from the proponent unit’s annual budget allocated by the Ministerio de Hacienda. Depending on the agency, funding may also come from other sources. For example, projects formulated by municipalities are funded by the regional government through the National Fund for Regional Development (FNDR). Projects funded solely by the national government do not have to necessarily have an RS from SNI, though projects that are regionally funded do require an RS RATE by law.4 Within the limits of their respective budget allocations, public entities (ministries, government agencies, etc.) then apply their own approaches to prioritize and select projects. In the case of nationally-funded projects, agencies must select from among projects that have been positively recommended by MDS for funding (assigned a RATE of RS). In the transport sector, projects are given ‘high’, ‘medium’, or ‘low’ priority by a largely qualitative consideration of the project’s 4 Organic Constitutional Law on Government and Regional Administration, Law 19175, Article 75. 12 alignment with sectoral and national strategies, whereas in the case of developing small reservoirs, appraised projects are subject to a multi-criteria process to prioritize. Project Appraisal and Selection in Chile Project selection in Chile is notionally based on social cost-benefit analysis (SCBA), with an overarching goal of maximizing societal benefits using scarce public resources. SCBA requires the quantification and monetization of societal costs and benefits, including potential positive and negative externalities as well as social and environmental benefits and costs that may be difficult to quantify and monetize. All projected costs and benefits are discounted to determine the net present values (discounted benefits minus discounted costs) of proposed projects, from the societal point of view.5 Generally speaking, SCBA can be used either to eliminate projects whose costs outweigh benefits or that do not meet minimum internal rates of return (IRR), or to rank projects by highest net present value (NPV), benefit-cost ratio (BCR), or NPV index (the ratio of the net present value of benefits minus costs to the value of the initial investment). The great strength of SCBA is the ability to compare projects across sectors and regions based on a common metric of monetized value. In Chilean practice, prior to considering CBA outcomes, projects must first pass initial screenings for legality and strategic alignment and meet the informational requirements for every stage of the SNI. SCBA is, thereafter, used to filter projects based on a minimum internal rate of return rather than to rank projects. In other words, CBA results help decide eligibility for further development but are not necessarily used to prioritize from among RS-rated projects. Chile developed its capacity and processes for CBA appraisal involving sophisticated estimation techniques, including the use of shadow pricing, the application of various estimation assumptions and methods for various kinds of projects, and standardized use of social discount rates and conversions for values of various expenses and profits in analyses. Some recent advances in project appraisal that have been mentioned are: • Consideration of the benefits associated with decreased road traffic accidents; • Consideration of the benefits associated with the reduction of greenhouse gas effects; • Evaluation of multipurpose projects (dams) or project networks (bike paths); • Consideration of increased traffic generated by transport projects (as a benefit); • Use of hedonic pricing in the evaluation of urban parks; and • Improved measurement of social prices (e.g., fuels, carbon dioxide, travel and leisure time). Proposed investments in the transport; forestry, agricultural, and fisheries; and water sectors must be subjected to CBA to generate economic indicators such as Internal Rate of Return (IRR), Net Present Value (NPV), and Net Present Value Index (IVAN). These metrics are to be included in SNI project documentation and are used to guide investment decisions. One of the most important calculated values – IRR – is used to filter projects. Following initial project appraisal and pre-feasibility, investments that do not meet a minimum IRR of 6% are eliminated from consideration (i.e., they do not receive RATEs of RS). Exceptions to this filtering rule are made for projects with low IRRs that are nevertheless deemed strategically significant (in 5In Chile, the discount rate is approved by law and encoded in guidance on application of CBA to projects included in the SNI. See SNI instruction here (link). 13 the case of water security, for example) and/or when it is recognized that CBAs are missing key information or are unlikely to capture important social benefits (e.g., closing regional income disparities, assuring future environmental quality, etc.). In these cases, a cost efficiency analysis or MCA are used to augment the appraisal (see discussion on MCA to follow). Extending Investment Decision Support in Chile While Chile has an extensive record in systematic project appraisal, there remain some technical, policy-oriented, and procedural shortfalls related to its current use as the basis of investment decision-making. These are also recognized internally and have served as an impetus for recent government efforts to extend the processes of project selection to include additional approaches to appraisal and comparison. In the transport sector, for example, the Highway Design and Maintenance Model (HDM-III/4) has been applied since the late 1980s to extensively estimate expected full life-cycle costs (including construction and maintenance) and benefits (e.g., maintenance, fuel, and travel time savings) associated with road projects. Nevertheless, due to the time and resource demands inherent to SCBA, appraisers must employ extensive assumptions. This can reduce the variation of results across sets of similar projects, tempering the comparative power of estimated metrics. Typically, it is also only feasible to account for select costs and benefits. Some key considerations may be excluded from analysis, especially costs and benefits that are strategic, environmental, or distributional in nature. Rooted in economic optimization and efficiency, CBA inherently favors projects that generate higher revenues and, therefore, cannot account for strategic or distributional issues. CBA also does not give weight to future-oriented goals such as national security or environmental preservation and privileges more profitable9 projects in metropolitan areas and low-cost regions (such as coastal metropolitan areas). As such, infrastructure funding in Chile is often concentrated in regions that are already more developed, exacerbating territorial inequalities (Ahmad & Viscarra, 2016). It is increasingly recognized that infrastructure development must consider an extended set of goals beyond economic efficiency. As a recent OECD report on ‘Gaps and Governance Standards’ in Chile’s infrastructure development system states, “The project evaluation and prioritization system will need to accommodate transversal issues and multiple policy goals,” including sustainability commitments. The report points out that “the current system offers limited scope for incorporating transversal issues and other political objectives into the decision-making process in a transparent way. Nevertheless, changes to project evaluation methodologies and selection criteria must not come at the expense of value for money and efficiency considerations” (2017). For this reason, multi-criteria analysis can be helpful to incorporate multiple considerations in addition to maximization of economic benefits, including climate change, cost efficiency, and regional inequality, in a transparent manner (OECD, 2017). Lastly, putting issues of methodological robustness aside, the outcomes of CBA analysis (e.g., IRR, NPV, IVAN) do not, in fact, strictly guide project selection or the order of fund allocation. As mentioned earlier, CBA is used to filter projects (e.g., removing those with IRRs of less than 6%) through to ‘Execution’ status, which explicitly confers SNI approval and allows the proponent to seek funding. The capital budget submitted to Congress by the Ministry of Finance may only include projects approved by SNI. Therefore, passing the CBA filter is a necessary condition of funding and implementation. But beyond this, calculated IRRs and other SCBA metrics (NPV, IVAN, BCR) are not necessarily used to prioritize within the set of projects that attain ‘Execution’ 14 status. Rather, proponent units may use any number of approaches (which are often undocumented or based on loosely-defined qualitative criteria) to select projects for implementation from the SNI-approved set. Moreover, a budget decree issued for a sector or unit does not bind the unit to developing the specific projects included in the budget proposal to the Ministerio de Hacienda. Multi-Criteria Analysis to Support Investment Decision-Making Some efforts have been made to extend the Chilean approach to project appraisal and investment decision-making to deal with the technical, policy-related, and implementation issues discussed above. In addition to the cost efficiency approach (which assumes that project benefits will be sufficient to justify estimated costs), Chile has also institutionalized the use of multi-criteria analysis (MCA) to support investment decisions for some sectors. Specifically, MCA has been applied in the rural water sector to deal with an observed mismatch between the methodological outcomes of CBA for water catchments and strategic goals of the sector. More specifically, water security warrants the development of water catchments to ensure the long-term availability of water for agricultural, residential, and industrial use, but CBA appraisals of water catchments typically yield IRRs of less than 6%, which would result in the filtering out of most catchment projects under the prevailing SNI process. As such, a 2014 Decree on the Use of a Multi-Criteria Evaluation Methodology (MCEM) for Small Reservoirs was issued by the government, based on a set of criteria and weights approved by the National Irrigation Commission. The MCEM applies the Analytic Hierarchy Process (AHP) to aggregate five criteria and 20 sub- criteria into an overall score associated with each water catchment project.6 The methodology is applied in the final stage of progression through the SNI, which starts with an initial filtering to eliminate projects that require resettlement, impose environmental threats, or exhibit various technical difficulties. Thereafter, projects pass through the pre-feasibility, feasibility, and design stages in SNI as in other infrastructure sectors. At any time, projects may be filtered out if major environmental, technical, legal, or political difficulties arise. In the last stage, catchment projects that pass filtering are subject to the small-reservoir MCEM analysis for final selection. The criteria and sub-indicators applied to select reservoir projects are detailed in Table 2, along with weights used to combine criteria into a single score. Values associated with the sub-criteria are not measured as continuous variables. Rather, sub-criteria are scored ordinally. Most are given an ordinal score across a range (often 0, 5, or 10), though some are simple binaries [0,1]. An additional criterion – technical complexity – is used in the ultimate selection of projects, though this is not weighted along with the MCE criteria in Table 2. This additional consideration covers technical issues such as soil mechanics, location of the catchment with respect to natural channels, proximity of materials earthworks and dumping sites for excavated soils. The degree to which technical complexity influences the ultimate selection of projects, relative to the MCE scoring, is not specified. 6 For technical details of AHP, see Saaty, Thomas L. (1990). How to make a decision: The analytic hierarchy process. European Journal of Operational Research, 48(1), 9-26. 15 Table 2. Reservoir MCEM Criteria and Sub-Criteria Criteria / Weight  Indicators / Sub‐criteria  Economic  Social VAN ($ million)  18.4%  Investment ($ million) / hectare  Investment ($ million) / land plots  Social  % households under poverty line  34.1%  Surface area of subsistence farms / small farms <12 ha  Number of beneficiaries (population in irrigable zone)  Indigenous communities in the territory [0,1]  % growth of rural population during last inter‐census period  Extreme zone, border region, or undeveloped area [0,1]  Strategic  Number of water shortage decrees in past five years  22.6%  Number of jobs generated (landowners and relatives)  Number of irrigation association systems that can be connected7  Electricity generation capacity (MWh/year)  Environmental /  Number of people required to relocate  Territorial  Number of archaeological site affected  9.2%  Hectares of native forest in flood zone  Management  Interest/support of beneficiaries  15.7%  Economic  contribution  of  regional  government  (regional  government  contribution / total project investment)  Organization (1‐4 indicating degree of incorporation / legal standing)  Number of land parcels required to expropriate  Source: Ministerio de Hacienda, 2014. Minuta Matriz Multicriterio Plan de Pequeños Embalses   7 This measures the number of ‘asociación pequeños regantes’ (APRs) or small irrigation associations that can be linked. 16 Chapter 3. Infrastructure Prioritization Framework: An Alternative Approach The Infrastructure Prioritization Framework is a quantitative multi-criteria approach to within- sector infrastructure project comparison. The IPF condenses project-level indicators (selection criteria) into two composite indices – a financial-economic index (FEI) and a social-environmental index (SEI) – and considers these alongside the budget constraint for a particular sector. Results are displayed graphically to map the projects’ relative expected performance along these two dimensions. While the IPF is quantitative in nature, it is also policy-responsive, since the government specifies the set of project selection criteria that reflect the sectoral and developmental policy goals. These criteria may include social, strategic, and environmental considerations alongside traditional financial and economic factors. In fact, the IPF was developed by an infrastructure team within the World Bank in response to government demand for alternative decision support approaches that could directly consider key policy goals; be feasibly applied across large projects sets within the resource means of the government; and remain systematic and evidence-based (Marcelo et al., 2016). IPF is designed to employ quantitative measures to the greatest extent possible to systematize project comparison and limit subjectivity in selection. While the IPF is a multi-criteria approach, it can utilize the results of CBA analysis as a key decision factor. While the approach was initially envisaged for low-capacity governments, its relevance to the Chilean context is demonstrated in the results of this report. In this section, we summarize the IPF prior to presenting the results of its application to the transport and water catchment sectors in Chile. The IPF Process Implementing the IPF is relatively straightforward and follows five steps: (1) selecting decision criteria; (2) gathering related project indicator data; (3) calculating social-environmental and financial-economic indices; (4) plotting projects and budget limits; and (5) comparing projects (see Figure 2). In this section, we summarize IPF application in terms of these steps.8 8 An extensive technical description of the IPF methodology is detailed in Marcelo et al., 2016, and Marcelo et al., 2015. 17 Figure 2. IPF Process Map Step 1. Select Criteria The first IPF step is to identify and select criteria used to compare projects. Selected variables may vary in different contexts based on the policy goals shaping decision-making, but will generally include indicators of value, efficiency, and social and environmental impact. This step is an opportunity to leverage professional knowledge and allow policy makers, experts, and other key stakeholders to reach consensus on the decision-making factors most important to project selection. In this way, this step helps crystallize a government’s infrastructure policy goals. Variables are organized into two general categories: social-environmental and financial-economic. Infrastructure projects are meant to improve quality of life; therefore, several direct social and environmental benefits are relevant, including factors like improved access to public services and job creation. These benefits come at a cost, however. Engineering works may require clearing forested areas, polluting and endangering natural environments, or resettling communities. The IPF directly considers these relevant social and environmental benefits and costs without requiring their monetization. In Panama, for example, the SEI initially consisted of five indicators: the number of direct beneficiaries; direct jobs created; people affected by repurposing of land use; poverty rates; and environmental impact (categorized as negative, neutral, or positive), all measured in their ‘natural’ units. The financial and economic effects of a project are also central to infrastructure decision-making. These can be assessed using outcomes of CBA or partial analyses that, at the very least, estimate project costs. In Panama, for example, four indicators were initially selected to comprise the financial-economic index (FEI): the internal rate of return, economic multiplier effects, monetizable externalities, and implementation risk. In other cases, a single economic indicator (e.g., NPV or IRR) may constitute the FEI. One key lesson from past pilots is that the set of selected indicators may require adjustment. An indicator may be found to be analytically problematic due to lack of sufficient data, calculation problems, or other issues, such as imprecision in variable specification. This is an iterative process, and indicator problems are likely to be discovered during data collection or index calculation. 18 Step 2. Prepare Data The second IPF step is to gather and transform raw data so that they are usable for calculating SEI and FEI project scores. A simple Excel environment can be constructed to populate a prioritization database with raw data. Because selection criteria variables may have different units of measurement (i.e., they are not all dollarized costs and benefits) but are combined in additive models, two types of data transformation are required. First, any qualitative data must be transformed into either ordinal or scalar data.9 Second, observations are standardized to deal with disparate units of measure, transforming all measurements to have a transformed value between [-1] and [1], with the set having a zero mean and unit variance.10 Step 3. Constructing Performance Indices – SEI and FEI Indices are used to combine information from multiple variables into composite indicators. In IPF, variables are organized into two classes to construct the social-environmental index (SEI) and the financial-economic index (FEI). The FEI may be the standardized value of a single component derived from CBA (e.g., NPV or BCR) or a combination of several factors, but the SEI is typically constructed by combining a number of key social and environmental variables. This is done via an additive model, wherein each indicator’s contribution to the overall index score is determined by weights (the coefficients associated with each variable). For example, if the SEI variables selected include (a) number of beneficiaries (BEN), (b) number of poor served (POOR), and (c) number of jobs created, the function may be expressed as , where weights, , are associated with each social-environmental indicator. The weights used to combine variables can be set subjectively or objectively, such as using some form of Principal Component Analysis (PCA). PCA is an information reduction procedure that seeks redundancies in sets of variables. These redundancies can be expressed as linear combinations or ‘principal components’ of the variables comprising the set. One key characteristic of PCA is the ability to calculate coefficients (weights) based solely on the statistical relationship between variables. While other weighting schemes may be used, PCA is particularly useful when there is a preference to objectively assign weights. Some significant advances have been made with respect to using PCA to determine weights. These changes have been made to deal with policy preferences regarding the relative importance of criteria. Over the course of the Sri Lanka and Chile pilots, for example, calculation methods were developed to add restrictions to PCA that can attain the following: require a particular coefficient sign (+/-) associated with specified variables; require that criteria are weighted to reflect a pre-set order of importance (a weighting order); and require a minimum weight for a variable. 9 This can be done with either scaling methods or via approaches like ALSOS. The transformation of categorical and ordinal qualitative and quantitative data into usable numerical data may be done using the Alternating Least Squares Optimal Scaling (ALSOS) algorithm, a widely-accepted transformation approach. Within a quantified ordinal variable, the numbers assigned by the ALSOS algorithm to each category reflect the distance between categories, revealing the implicit metric of the variable (Perreault & Young, 1976). 10 Numerical values are standardized via a standardization formula that can be coded into Excel. The standard score z of a raw score x is z_ij= (x_ij-μ_j)/ _j, where μ is the sample mean and is the standard deviation of the variable j. 19 After SEI and FEI variables are combined in the additive model, resulting values are normalized and rescaled to generate SEI and FEI scores between 0 and 100 for each project. The rescaled score can be expressed as / 100, where is the minimum value for variable and is the maximum value. These rescaled scores are used as the SEI and FEI scores for plotting in Step 4. Step 4. Creating the Visual Interface: The Investment Prioritization Matrix To create a visual comparison, projects are plotted on a two-dimensional Cartesian plane, with axes representing the SEI and FEI. The budget limit for the sector is also imposed along each axis (intercepting the axis where funds are exhausted). First, however, the budget must be hypothetically allocated separately to the SEI- and FEI-ranked project lists to determine the fundable sets in each. In other words, the budget limit is hypothetically allocated to the top-ranked projects on each list (as if selection were based only on SEI or only on FEI) until resources are exhausted. The resulting fundable sets are compared simultaneously on the investment matrix. A ‘good’ project in terms of financial and economic performance may nevertheless be undesirable from a social and environmental perspective, and vice versa. As such, decision-makers must consider projects along both dimensions. Projects can be compared by their respective SEI and FEI scores on a visual interface called the Infrastructure Prioritization Matrix (Figure 3). Once projects are plotted, the budget limit is imposed onto the plane – perpendicular to each axis – at the point where the budget would be exhausted if funding were determined solely by each index. The plane is intersected by the dually-imposed budget limit, creating four quadrants. Figure 3. Example Investment Prioritization Matrix, Panama Water and Sanitation Pilot, 2015 50 45 P15 P3 40 P23 P10 P1 P24 35 30 P27 25 P29 P4 P16 P17 SEI P12 P22 P28 20 P7 P35 P19 P13 P21 15 P18 P26 P25 P31 P5 P11 P34 10 P32 P6 P20 P30 P33 P14 5 0 0 10 20 30 40 50 60 70 80 90 100 FEI Source: Prioritizing Infrastructure Investments in Panama: Pilot Application of the World Bank Infrastructure Prioritization Framework (April 2016) 20 Projects that fall inside the budget constraint along each axis represent the ‘Investment Possibilities Set’ for each dimension. Projects in the upper right quadrant fall in the Investment Possibilities Set for both SEI and FEI and are then categorized as ‘High Priority’ projects. Evolution of the IPF Through pilot applications of the IPF in Vietnam, Panama, Argentina, Sri Lanka, and Chile, the framework has been refined. This section discusses key areas of progression, including lessons learned on important pre-analytical steps and capacity and institutional requirements, technical aspects of variable specification and weighting, and the improved use of sensitivity analysis. Pre-Analytical Steps Pre-analytical processes can help filter projects to reduce the analytical burden of prioritization and ensure sufficient comparability of data during project comparison. One of the challenges of early pilots was that some data, even from within feasibility studies, was either opaquely determined or had limited comparability across projects (Mandri-Perrott, Marcelo, and Haddon, 2015). Feasibility studies should follow clear rules, guidelines, and standards of appraisal to ensure quality and comparability of data (particularly financial estimations) across projects. Additionally, filters are helpful to ensure that projects meet basic informational, policy, or strategic requirements and/or align with key sector goals. Filters may also be useful when there are inherent biases observed in the set of projects proposed or where the government aims to break regressive patterns. In Vietnam, for example, it was observed that projects in poorer regions tended to score lower on inputs to the FEI or SEI. This observation justified use of an initial filter to target areas with higher poverty rates (Mandri-Perrott, Marcelo, & Haddon, 2015). Technical Improvements: Variable Specification and PCA Restrictions A second set of lessons is that special consideration should be given to the selection and definition of variables as well as the weights assigned via PCA. For one, metrics must be carefully specified to deal with regressive biases. As in Vietnam, Panama revealed an inherent bias towards infrastructure projects in wealthier urban regions due to better scoring on project indicators. If development plans aim to improve rural areas, however, this can yield adverse results. Alternative to using a filter, this problem can be overcome by careful indicator specification and/or the inclusion of additional indicators to capture development goals. Variable specification is also important to balancing considerations of efficiency and efficacy. For example, one could use the absolute number of beneficiaries as an input to the SEI to consider policy effectiveness where service expansion is a priority. On the other hand, ‘beneficiaries per dollar spent’ may be more appropriate if the key goal is fiscal efficiency. In the case of Panama, where development of rural services is an important policy goal, the decision was made not to control indicators by project size to avoid privileging urban projects with greater economies of scale (Marcelo, Mandri-Perrott, & House, 2015). Another lesson on variable specification relates to the appropriate use of financial and economic indicators under conditions of low information, particularly regarding project benefits. If only project costs can be estimated, additional variables must be considered to construct the FEI (i.e., FEI should not be based on cost only). Another technical issue arose regarding the use of PCA for weighting. Since PCA synthetizes information based on correlations between variables (which may yield positive or negative 21 coefficient signs), it is important to make sure that weights reflect the desired relationship between a variable and the composite indicator. A problem arises, for example, if PCA assigns a negative weight to a variable that should be positively rewarded in selection. In some cases, this can be resolved by alternative specifications of a component variable, but an important methodological development has been the imposition of a coefficient sign restriction in PCA. This development allows the user to restrict PCA results to ensure that variables that should positively contribute to the SEI are assigned positive-signed coefficients, and vice-versa for variables that should be scored negatively. Sensitivity Analysis and Criteria Weighting Another important improvement has been the addition of sensitivity analysis to test the robustness of results with different variable specifications and criteria weightings. In the Chile pilot, for example, there was an expressed goal of focusing transport investments in areas with higher poverty. Since poverty rates may be measured in several ways, however, a sensitivity analysis applied two alternative poverty rate approaches to compare IPF results. Since criteria weighting is one of the most important methodological decisions for building indices, it is another important area of sensitivity analysis. The results of IPF are determined, in part, by the weights associated with each indicator. Though a significant lesson from Panama was that composite indices were far more sensitive to indicator values than to the weights used to combine them.11 While this must be further tested, this suggests that PCA may be a useful way to weight variables if time and objectivity are important factors in selection. In practice, the use of subjective weighting can give rise to several problems, including lack of transparency, manipulation of weights to privilege ‘pet projects’, or index scores with low variation (and then, limited value for comparison). On the other hand, subjective weighting is more intuitive and directly responsive to policy preferences. Developments have focused on finding a compromise between the responsiveness of subjective weighting and the objectivity of PCA. In addition to the sign restrictions discussed above, the need arose to adjust the mechanics of PCA to better capture policy preferences. The use of PCA was further refined to allow several additional restrictions on coefficients. These include the ability to specify a minimum value for a coefficient and the ability to specify the order of weighting (order of importance to the overall score). These restrictions define a spectrum from purely objective to purely subjective weighting (Figure 4). Figure 4. Spectrum of Index Weighting Approaches 11A sensitivity analysis was performed to compare PCA indices against composite indices using subjectively established weights. Two subjective weighting schemes (equal weighting and hypothetical policy-determined) were tested to calculate alternative SEI composite indices. The categorization of projects changed only minimally when using policy- determined or equal weights (Marcelo, Mandri-Perrott, & House, 2015). 22 Positive Normative Objective Sign-constrained Minimum- Preference- Subjective Subjective PCA PCA value PCA ordered PCA weighting Neutral Responsive Organizational and Capacity Issues To improve robustness of results and foster concurrent application with other supportive analytical tools (including CBA and expert assessment), users must have sufficient technical capacity to understand the mechanics and implications of key decisions regarding the use of IPF, including decisions about criteria selection, indicator specification, and weighting. Further, to extoll the benefits of the responsiveness inherent to the tool, the proposed methodology should not be a one-off exercise. Rather, it should be utilized as a progressive approach, intended to ‘live and grow’ with the country's infrastructure needs and policy objectives. As such, the prioritization program should involve continuous refinement of the decision-support tool, based on informed deliberation regarding criteria selection and any pre-decisions of a policy nature (Mandri-Perrott, Marcelo, & Haddon, 2015). Last, planning offices and decision makers must be familiarized with the multi-criteria approach to build credibility of the decision support tool itself, establish familiarity with its use, and legitimize the results of analysis. 23 Chapter 4. Applying IPF to Chile’s Infrastructure Project Proposals This section documents the results of the IPF-based project prioritization for water catchment and road transport projects and compares these results with actual project selection outcomes. The greatest value of the Chile pilot is the opportunity to compare IPF results to project selection informed by CBA to test the analytical demands of IPF inputs, the robustness of outputs, and degree of alignment of IPF results with the outcomes of other approaches. The comparative results show that prioritization outcomes are affected not only by the methodologies in use, but also by the practices and policies of project selection – i.e., how the results of analyses are applied in decision-making. In this chapter, we first present the IPF mechanics and comparative results of IPF- and MCE-based prioritization for water catchments and follow with IPF construction and comparative results of IPF- and SCBA-based prioritization for road transport projects. Water Catchment Projects in Chile Water resource management is an important policy priority in Chile with direct impacts on rural and agricultural development and environmental sustainability. Despite an abundance of water resources (overall availability of around 50,000 m3 per capita per year), Chile faces water stress due to geographic distribution patterns. Most of Chile’s population lives in arid and semi-arid areas where water availability is low (less than 1,000 m3 per capita year) and demand exceeds surface water supply. An increasing need to offset unmet demand by groundwater extraction has led to a significant increase in annual freshwater withdrawals, which has become a key sustainability concern.12 Moreover, Chile is projected to move from a level of medium water stress to extremely high stress in 2040 due to the impacts of climate change.13 The erosion and desertification of soils also present a recognized sustainability challenge related to development of the Chilean forestry and agriculture sectors. Deforestation, overgrazing, inadequate crop management, and irrigation practices have resulted in soil degradation affecting nearly half of the territory and 75 percent of productive soils. This increased water stress and soil degradation disproportionally affects the poorest populations, who rely on small-scale agriculture as a critical income source; depend on natural resources for food, fuel, and building materials; and are typically located in arid rural regions most affected by climate variability and drought. As such, there is increased pressure to better manage water resources to ensure water security, meet growing agricultural and industrial demand on water resources, support adequate provision to the poorest communities, and deal with increased water stress associated with the effects of climate change. Water Catchment Project Sample Projects are typically prioritized within regions (as opposed to the national level) to ensure the disbursement of funding across regions. Since projects are allocated funds region by region, the IPF was also applied at the regional level in four areas: Biobio, Maule, O’Higgins, and Valparaiso. 12 Chile’s annual freshwater withdrawals as percentage of internal resources went from 2.3 percent in 1992 to 4.0 percent in 2014. 13 Maddocks, Young, & Reig (2015). 24 The regions and samples of proposed projects considered via MCE and IPF are summarized in Table 3. Table 3. Water Catchment Project Sample Region  Number of Projects in  IPF Sample  Biobio  14  Maule  13  O’Higgins  22  Valparaiso  12  To compare IPF and MCE prioritization outcomes, projects were first assigned scores and ranked in regional groups, as in practice. They are also presented in ranked order as a pooled (all regions combined) group in the IPF analysis. Water Catchment Project Indicators The selection of indicators used to construct the SEI reflects some of the key policy goals of the government with respect to developing water catchments. As discussed in Chapter 2, many of the intended benefits associated with improving water resources are strategic, environmental, and social – and these are often difficult to quantify and monetize. Therefore, economic analyses often result in calculated internal rates of return (IRRs) below the 6% threshold required for favorable recommendation. These low IRRs are likely due to undervaluation of some long-term benefits. Recognizing the importance of water resource development, however, the government implemented the Multi-Criteria Evaluation approach to assess these projects. The IPF draws from the MCE’s components to select the sets of input indicators for the SEI and FEI. Data previously gathered for the MCE were simply re-organized to fit the format of the IPF approach, with three important changes. First, the selection of input variables required paring down the extended list of MCE indicators. The resulting IPF indicator set excluded some MCE variables either because (a) they brought redundant information to the analysis or (b) exhibited no variation across projects. Second, some indicators used in IPF drew directly on project data in natural units rather than the ordinal score (e.g., 0, 5, 10) assigned to projects in the MCE approach. Third, some values maintain their ordinal scores, but are transformed so that more positive scores are recorded with higher values (in the MCE, lower scores are attributed to better performance on sub-indicators). The resulting indicators are described in Table 4, and relevant transformations are described in Table 5. 25 Table 4: Variables Included as SEI and FEI Indicators Type  Indicator  Included Variables  English  Use / Relevant Calculations  Investment ($m)  Inversión inicial  Initial Investment  Direct from MCE data  Other  Predios (u)  Número de predios   # Land Plots  Direct from MCE data  Inputs  Surface (ha)  Superficie equivalente  Surface  Direct from MCE data  Drew from MCE data, but  Porcentaje de hogares en  directly used the percentage  Poverty  Poverty  pobreza comunal  value rather than an assigned  ordinal score   # Land Plots / Surface  Beneficiaries  # Predios / superficie  Direct from MCE data  Area  Comuna extrema, fronteriza  Underdeveloped  SEI  o rezagada  Community  Jobs  See Table 5  Generation of  Generación de empleo  Agricultural  agrícola  Employment  Relocalización de vivienda  Household Relocation  Territorial  See Table 5  Afecta bosque nativo  Native forest affectation  i_NPV  VAN social / inversión inicial  NPV / Investment  Direct from project data  Expropriations  Expropiaciones  Expropriations  See Table 5  FEI  Community  Endorsement  Interés de beneficiarios  See Table 5  Endorsement  Legal Standing of  Legal   Desarrollo organizacional  See Table 5  Organizations  26 Table 5. Calculation of Select SEI and FEI Indicators Indicator  Calculation / Indexation Rule  (a) For Comuna Extrema (underdeveloped community 'UC')  (i)  If UC = Yes assign 1   (ii) If UC = No  assign 0  (b) Generación de Empleo Agrícola (Agricultural employment 'AE')  (i)  If AE = Yes  assign 1   (ii) If AE = No   assign 0   Jobs  (c) Jobs Index = a + b        Converted  Original MCE Score  Underdeveloped Community  Agricultural Employment  Yes  1  1  1    No  10  0  0  (a) For Relocalizacion de vivienda (household relocation 'HR')  (i)  If HR = Yes assign 0   (ii) If HR = No  assign 1    (b) Afecta Bosque Nativo (native forest affected 'FA')  (i)  If FA = Yes assign 0  (ii) If FA = No  assign 1   Territorial  (c) Environmental = a + b      Converted    Original MCE Score  Relocation Required  Affects Native Forests  Yes  1  0  0    No  10  1  1  (a) The more expropriations 'EX', the higher the risk to the project.   (b) We use the following formula for expropriations:  Expropriations                       / ′   Therefore, the larger the number of expropriation in original data, the smaller value taken in the  expropriations index.  Interest  (a) This stands for interest of beneficiaries 'IB' in the project  (community  (i)  If IB = YES (the community endorse the project)                Assign 1   endorsement of  (ii) If IB = No  (the community doesn't endorse the project)  assign 0   project)  (a) This variable stands for the soundness of legal standing of organizations  (i)   If original value = 1, legally constituted organization (best)             assign 3  (ii)  If original value = 5, "de facto" constituted organization                  assign 2   (iii) If original value = 10, non‐constituted/ non‐existing organization  assign 1.  Legal Standing of Organizations  Legal    (in English)  Original  Converted  Organización con personalidad jurídica  Legally constituted organization  1  3  Organización de hecho constituida  "De Facto" constituted  5  2  Organización de hecho no constituida  "De Facto" non‐constituted  10  1    Inexistencia de organización  Non‐existing organization  10  1  Water Catchment Project Indicator Weighting While projects were ranked region by region, the weights associated with criteria were calculated via PCA based on the full (combined) sample of projects. This decision was made due to the low degree of variation of many variables within each region. In other words, criteria weights were 27 calculated via PCA (or a restricted PCA) using all projects across these four regions. This common set of weights was applied to score projects in Biobio, Maule, O’Higgins, and Valparaiso, separately, based on the following formulas: , and , where , …, are the weights associated with each criterion. The weights , …, used to combine SEI and FEI indicators are described in Tables 6 and 7. Table 6 includes four weighting schemes for SEI. The first is determined by PCA with no restrictions; the second is determined by PCA with the restriction that all weights must have a positive value; the third is determined by PCA with the restriction that all weights must be positive and have a minimum value of .10; and the fourth is a simple equal weighting of all variables. Table 6. Weighting and % Variance Explained, Water Catchment SEI Calculations    Factor loadings (x vector) and % of variance explained  PCA weights,  No restriction  PCA weights >=0  minimum requirement  Simple average     (10%)  Poverty  0.011  0.011  0.316  0.500  Beneficiary  0.619  0.619  0.642  0.500  Jobs  0.416  0.416  0.400  0.500  Territorial  0.666  0.666  0.573  0.500  Retained variance  1.608  1.608  1.559  1.455  % explained  40%  40%  39%  36%  The weighting adopted to calculate the SEIs for all regions is that determined by PCA with the restriction that all criteria contribute at least 10% to the overall score. Then, the additive model is: .316 .642 .400 .573 Similarly, Table 7 includes four weighting schemes for FEI. Table 7. Weighting and % Variance Explained, Water Catchment FEI Calculations    Factor loadings (x vector) and % of variance explained  PCA weights, minimum  No restriction  PCA weights >=0  Simple average     requirement (10%)  NPV_inv  ‐0.047  0.000  0.316  0.500  Expropriations  0.395  0.369  0.316  0.500  Endorsement  0.711  0.710  0.635  0.500  Legal  0.580  0.600  0.630  0.500  Retained variance  1.808  1.807  1.732  1.457  % explained  45%  45%  43%  36%  Again, the weighting adopted to calculate the FEIs was that determined by PCA with a 10% minimum contribution restriction, with the resulting additive model: .316 .316 .635 .630 28 IPF Results: Water Catchment After calculating the SEI and FEI scores for each catchment project according to the weighted additive models above, the projects are ranked by each score, by region. The calculations for SEI and FEI scores are detailed in Annexes 3 and 4, while the graphical results of rankings by each index are summarized in Figures 5, 6, 7, and 8. Figures 5 (SEI) and 7 (FEI) display results region by region, whereas Figures 6 (SEI) and 8 (FEI) display the full set of projects pooled into a single ranked group according to each index. In each ranking chart, the projects that were actually approved for funding by the government are marked with a checkmark above the project’s score. Generally speaking, irrigation projects with higher SEI and FEI scores tended to be those approved for funding, though some projects with outlying low SEI scores were among the approved set. Water Catchment SEI Results Project rankings by SEI for each region are detailed in Figure 5. In each ranking chart, the projects actually selected for funding are marked with a checkmark. Figure 5. Water Catchment Project SEI Scores and Ranking, by Region Biobio SEI Maule SEI 100 100 90 90 ✔ ✔ 80 80 ✔ 70 70 ✔ 60 60 ✔ 50 ✔ 50 ✔ 40 40 30 ✔ ✔ 30 20 20 10 10 0 0 2 1 8 5 3 4 10 9 7 13 14 20 15 24 17 23 25 27 19 11 12 6 22 21 16 26 18 Project ID Project ID O'Higgins SEI ✔✔ Valparaiso SEI 100 100 90 90 80 ✔ ✔ 80 ✔ 70 70 60 60 50 ✔ 50 ✔ ✔ 40 40 30 30 ✔ 20 20 ✔ 10 10 0 0 57 50 55 54 60 53 56 59 51 61 52 58 47 40 44 42 46 41 43 49 45 30 37 34 28 48 38 29 32 36 39 35 31 33 Project ID Project ID ✔ - Project approved for funding by GoC Source: Authors’ calculations 29 Pooled SEI Results Figure 6. Water Catchment Project SEI Scores and Ranking, Pooled Regions Pooled Region SEI 100 90 80 70 60 50 40 30 20 10 0 56 36 26 29 6 59 20 57 47 51 24 21 4 44 50 10 31 41 1 14 22 23 27 7 17 54 30 40 34 5 58 15 48 45 2 13 35 55 3 37 25 42 43 12 8 32 33 28 18 52 53 38 46 16 49 9 19 39 11 60 61 Project ID Biobio   14 projects O’Higgins   22 projects Maule   13 projects Valparaiso   12 projects Source: Authors’ calculations 30 Water Catchment FEI Figure 7. Water Catchment Project FEI Scores and Ranking, by Region Biobio FEI ✔ Maule FEI ✔ 100 ✔ 100 90 90 ✔ 80 ✔ 80 ✔ 70 70 ✔ 60 60 50 50 40 ✔ ✔ 40 30 30 20 20 10 10 0 0 5 7 4 14 9 10 3 24 20 15 25 27 17 19 12 11 1 2 6 21 16 22 26 8 18 13 23 Project ID Project ID O'Higgins FEI ✔ ✔ ✔ ✔ 100 90 80 ✔ 70 ✔ 60 50 40 30 20 10 0 35 33 34 43 44 30 40 37 45 47 36 39 49 32 42 31 41 46 38 48 28 29 Project ID Valparaiso FEI ✔ ✔ 100 90 ✔ ✔ 80 70 60 50 40 30 20 10 0 50 53 59 57 55 54 60 51 56 52 61 58 Project ID ✔ - Project approved for funding by GoC Source: Authors’ calculations 31 Figure 8. Water Catchment Project FEI Scores and Ranking, Pooled Regions Financial and Economic Indicator (FEI) 100 90 80 70 60 50 40 30 20 10 0 56 26 50 33 34 51 53 24 7 13 4 21 20 5 35 32 52 12 2 18 15 42 43 31 30 57 27 37 54 22 25 55 8 3 23 38 48 58 45 28 59 36 9 6 16 29 39 49 46 19 44 14 11 1 10 40 41 17 61 47 60 Project ID Biobio   14 projects O’Higgins   22 projects Maule   13 projects Valparaiso   12 projects Source: Authors’ calculations 32 Water Catchment IPF Matrix Figure 9 shows the results of plotting projects (by their SEI and FEI scores as x and y coordinates) onto a water catchment prioritization matrix, region by region. Yellow project points denote projects that were actually approved for funding. Interestingly, these results suggest different patterns with respect to funding preferences. In Maule and Valparaiso, for example, approved projects are those that had the highest FEIs, whereas project selection in O’Higgins and Biobio does not directly reflect high FEI or SEI. Figure 9. Water Catchment IPF Matrices, by Region Biobio Maule 100 P14 100 P18 90 90 80 80 P19 P25 P27 70 70 P23 60 60 P26 P6 P16 SEI SEI 50 50 P17 P24 P21 40 P1 40 P15 P8 30 P12 30 P13 20 P7 P11 P2 P3 20 P10 P20 P22 P4 10 10 P5 P9 0 0 0 10 20 30 40 50 60 70 80 90 100 0 10 30 40 50 60 70 80 90 100 20 FEI FEI O'Higgins Valparaiso P49 100 P44 P45 100 P59 90 P38 90 80 P40 P39 80 P48 P47 70 P32 P42 P30 70 P33 P41 P28 60 60 P53 P34 P37 SEI SEI P43 P29 P31 50 50 P52 P61 P60 40 40 P35 P55 P54 30 30 P50 P51 20 20 P57 P56 P58 10 10 P36 P46 0 0 60 0 30 40 50 60 70 80 90 100 0 10 20 30 40 50 70 80 90 100 10 20 FEI FEI Source: Authors’ calculations 33 Comparing IPF to Selection of Water Catchment Projects Table 8 offers more detail on the comparative outcomes of IPF versus MCE ranking in each region. The MCE-ranked projects marked with an asterisk are those that were not recommended for implementation due to location in an area where environmental restrictions were in place. Highlighted projects are those that were recommended for funding. Table 8. Comparing IPF Maps and MCE Project Rankings, by region Biobio IPF Mapping MCE Ranking Biobio ID Name Score 1 Kaiser 3.0 100 P14 9 Quidico 1 3.3 90 3 Ranquil 3.5 80 8 Mirihue 3.5 70 14 Leoneras 3.5 60 2 Las Puentes 4.1 P6 SEI 50 10 Quidico 2 4.3 40 P1 6 Pichi Bureo* 4.4 P8 30 P12 P13 5 Laguna El Pillo* 4.6 20 P7 P11 P2 P3 P10 12 Vegas de Itata 4.6 P4 10 7 Rumena 5.0 P5 P9 0 13 Tauco 5.3 30 40 50 60 70 90 0 10 20 80 100 4 Tranaquepe* 5.4 FEI 11 Perales 5.9 Maule IPF Mapping MCE Ranking Maule ID Name Score 19 La Bruja 2.38 100 P18 17 Peralito* 3.12 90 18 Peralito 2 3.17 80 P19 25 El Molino 3.22 P25 P27 70 26 Derivado Porvenir 2 3.31 P23 60 P26 21 Sauzal 4.50 P16 SEI 50 P17 24 Manantiales 5.01 P24 P21 40 P15 27 Limávida 5.11 30 23 Vaquería 5.45 20 20 El Guindo 5.55 P20 P22 10 22 Botacura 6.20 0 15 Huencuecho I 6.65 30 40 50 60 70 90 0 10 20 80 100 FEI 16 Huencuecho II 6.65 * Not recommended due to environmental restriction 34 O’Higgins IPF Mapping MCE Ranking ID Name Score O'Higgins 29 San Francisco 1 (Zonada) 6.2 P49 100 P44 P45 30 San Francisco 2 (Zonada) 6.2 90 P38 31 San Francisco 3 (Zonada) 6.2 28 Codegua (CFGD) 6.4 80 P40 P39 P48 P47 37 Manquehue 1 (Zonada) 6.5 70 P32 P42 P30 P33 P41 P28 38 Manquehue 2 (Zonada) 6.5 60 P34 P37 41 Estero Seco (CFGD) 6.7 SEI P43 P29 P31 50 36 Ucúquer (CFRD) 7.6 40 32 El Maiten 1 (Zonada)* 7.8 30 P35 33 El Maiten 2 (Zonada)* 7.8 20 34 El Maiten 3 (Zonada)* 7.8 10 P36 P46 35 Huehuinco (CFRD)* 7.8 0 40 Las Palmas 2 (RCC) 8.2 30 40 50 60 70 90 0 10 20 80 100 42 Cementerio 1 (Hormigón) - FEI 43 Cementerio 2 (Hormigón) - Valparaiso IPF Mapping MCE Ranking ID Name Score Valparaiso 55 Pullally 3.04 100 P59 54 Santa Marta 3.26 90 80 60 Santa Julia 4.48 70 59 Valle Hermoso* 4.85 60 P53 56 Cuncumen 1 5.11 SEI 50 P52 P61 P60 58 El Zaino 2* 5.18 40 P55 P54 30 P50 52 Chalaco* 5.60 P51 20 P57 P56 P58 57 Lo Zárate 2 5.62 10 53 Las Carditas 2* 6.18 0 40 50 60 70 90 0 10 20 30 80 100 50 Vitahue* 6.31 FEI 51 Paihuen* 6.50 * Not recommended due to environmental restriction Note: Some projects that entered the database later do not have a corresponding MCE ranking, such as Project 39, 44-49 in O’ Higgins and Project 61 in Valparaiso 35 Applying IPF to Road Transport Regional proposals for transport projects are initiated with the submission of draft regional investment pre-project proposals (pre-ARI), which are developed at the regional level with the participation of representatives of the Ministerio de Obras Públicas (Ministry of Public Works) (MOP) subnational offices, Secretarías Regionales Ministeriales (SEREMI). These initial proposals are considered jointly by regional and national planners. SEREMI sends several regional projects on for consideration to MDS to follow the process of ongoing project appraisal and RATE. While the national government may decide to fund projects with a RATE other than RS, regional projects require an RS designation.14 The MOP determines which of the projects that received an MDS favorable recommendation (RS) will be sent further for funding to the Ministry of Finance. In the MOP's prioritization methodology, projects are classified as medium, high or low priority based on considerations of strategic relevance, completeness of information, and alignment with regional development plans. Projects selected by the MOP are sent to the Ministry of Finance for review. On approval, a decree is issued by the Ministry of Finance to commit financial resources and initiate project execution. To modify the list of projects, MOP must send its revised prioritized list to the Ministry of Finance to be approved and included in a new decree. Transport Policy Goals and Road Project Criteria The processes of transport project prioritization described above hinge on two primary inputs: project cost-benefit analysis and qualitative assessment of project alignment with medium- and long-term regional and national transport development plans. The former – CBA – is based on a sophisticated approach to assessing road project lifecycle costs, namely the Highway Development and Management (HDM-III/4) tool. The latter involves consideration of Chile’s transport policy goals. Road density in Chile lags behind other Latin America and OECD countries, partly due to low overall population density but high concentration in Santiago and Valparaiso. The road network’s size (approximately 78,000 kms in 2010) has not significantly changed since 1990, though the share of paved roads has increased from 13.8 to 23.3 percent. This pilot application of IPF considers inter-urban road projects intended to extend the road network and link urban areas. Specifically, the pilot was applied to 50 projects in the BIP that met basic requirements (mentioned below) to be included in the IPF pilot, including HDM-3 data and the existence of a CBA.15 Road repositioning projects were excluded from the pilot. Road Transport Project Sample The sample of 50 interurban road transport projects to which IPF was applied includes only a set that has reached Execution status in the SNI. Additionally, this set has passed the basic profitability test of having IRRs of at least 6% and have been judged to align with transport policy goals. 14 The Organic Law on Financial Administration of the State, Article 19, requires only that a project’s profitability be analyzed, but does not require project profitability. The Organic Constitutional Law on Government and Regional Administration (19175), Article 75 states that projects financed with regional funds require an RS RATE. 15 Special classifications exempt some projects from requiring CBA (e.g., projects in extreme zones, projects with expected volumes of less than 400 vehicles per day). 36 Transport Project Indicators SEI indicators were selected to consider beneficiaries, job creation, extension of service to areas with higher poverty levels, and savings associated with reduced gas consumption.16 Each of these input indicators is controlled for project size, yielding the following criteria inputs: beneficiaries per dollar invested (Ben_inv), jobs per dollar invested (Jobs_inv), and gas savings per dollar invested (Savings_inv). Beneficiaries is calculated by estimating traffic volumes, considering the occupancy rates for different types of vehicles. This information is taken directly from project proposal data included in the BIP. Data on the creation of jobs directly associated with construction and operation are also calculated based on BIP data, which were in turn derived from an estimation model based on the macroeconomic input-output matrix. Gas savings is an estimate of the annual savings in gas, measured in liters. The initial set of indicators also considered the extension of service to indigenous populations, but this criterion was later dropped due to limited variation. Lastly, the SEI includes the poverty rate and tests the use of two alternative specifications to measure poverty. One takes a simple poverty rate (Pov_c) reflecting the percentage of people in the service area classified as “income poor”. It is important to note that it is not necessary that the users of the road are the same as the population that lives nearby the road that is under appraisal. It is possible that the locals could have a low vehicle per capita ratio, and the usage of the road is composed by transportation trucks, tourists, drive through or long-haul trips, etc. The other applies a multi-dimensional composite poverty index (MPI-CL) developed by MDS (Pov_multi) that accounts for five poverty dimensions: education, health, employment and social security, housing, and networks and social cohesion. The FEI takes into account only one measure – the net present value index (IVAN) – derived from project CBAs. The IVAN is equivalent to a ratio of the net present value of future project costs and benefits to the project’s initial investment cost. The formulas for SEI and FEI are as follows: _ _ _ _/ where , …, are the weights associated with each SEI criterion. Input project data are detailed in Annex 5. Transport Project Indicator Weighting The criteria weighting of SEI variables was done using PCA, with the two specifications for measuring poverty. Table 9 gives the weights associated with each variable using each approach. Table 9. Criteria Weighting, Transport SEI Ben_inv  Jobs_inv  Savings_inv  Pov_c  Simple poverty rate  0.614  0.224  0.724  0.224  Multi‐dimensional  Ben_inv  Jobs_inv  Savings_inv  Pov_multi  poverty index  0.442  0.224  0.839  0.224  Note: The weights have been determined by PCA with the restriction that all weights must be  positive and have a minimum contribution of 10%  16 In economic ($) terms, this would likely be considered in the FEI; however, here, our interest is to reduce consumption for environmental reasons. Therefore, the savings is measured in liters and included in the SEI. 37 IPF Results: Road Transport As with water catchment projects, transport projects are ranked by SEI and FEI scores separately. The calculations for SEI and FEI scores are detailed in Annexes 6, 7, and 8, while the graphical results by each index are summarized in Figures 10, 11 and 12. Figures 10 and 11 show the SEI ranking based on the simple poverty indicator and the multi-criteria poverty index, respectively. Road Transport SEI and FEI Rankings Figure 10. Road Transport SEI Ranking Results from Input and Weighting with Pov_c 100 90 Social and Environmental Indicator (SEI) - Pov_c 80 70 60 50 40 30 20 10 0 P5 P8 P42 P48 P3 P2 P45 P46 P19 P16 P6 P9 P49 P10 P4 P23 P13 P14 P17 P44 P43 P20 P11 P33 P41 P47 P1 P50 P27 P40 P7 P24 P12 P28 P25 P18 P22 P35 P32 P15 P38 P26 P39 P29 P36 P31 P34 P21 P30 P37 Figure 11. Road Transport SEI Ranking Results from Input and Weighting with Pov_multi 100 Social and Environmental Indicator (SEI) - Pov_multi 90 80 70 60 50 40 30 20 10 0 P46 P6 P9 P16 P10 P14 P44 P17 P4 P1 P11 P41 P47 P40 P7 P12 P5 P18 P13 P48 P3 P42 P2 P8 P43 P15 P45 P19 P26 P39 P29 P49 P36 P31 P34 P28 P23 P21 P20 P30 P50 P27 P37 P24 P25 P35 P22 P32 P33 P38 Figure 12. Road Transport Project FEI Ranking Results 100 Financial and Economical Indicator (FEI) 90 80 70 60 50 40 30 20 10 0 P8 P5 P2 P19 P9 P46 P16 P49 P6 P14 P10 P17 P7 P44 P40 P4 P41 P47 P11 P1 P42 P3 P25 P22 P12 P18 P48 P23 P32 P45 P35 P33 P28 P15 P43 P13 P38 P39 P29 P26 P36 P21 P27 P31 P50 P34 P30 P20 P37 P24 Source: Authors’ calculations 38 Road Transport IPF Matrix Figure 13 shows projects by their SEI and FEI scores. Projects are color coded according to the duration between a project reaching Execution status in the SNI and the allocation of project execution funding. This indication of urgency with respect to funding is taken as a loose proxy of project prioritization. Projects coded in yellow were allocated funds within a year of reaching Execution status, and projects coded in orange were funded between one and two years of reaching Execution status. Green projects were either funded two or more years after reaching Execution status or are still awaiting funding, indicating lowest priority. Figure 13. Road Transport Project Prioritization Matrix 100 P24 90 P7 80 P45 70 60 P40 P27 P37 P36 SEI 50 P15 P38 40 P2 P3 P32 P42 P22 P9 P1 P16 30 P50 P35 P47 P25 P43 P8 P49 P33 P41 P11 P5 20 P48 P6 P20 P30 P21 P18 P29 P14 P44 P28 P39 10 P17 P13 P23 P19 P4 P34 P46 P26 P12 P10 0 P31 0 20 40 FEI 60 80 100 Funded <1 year ⏺ ; Funded within 1-2 years ⏺ ; Funded in 2+ years or not yet funded ⏺ Comparing IPF to Funding of Transport Projects Figure 13 demonstrates no clear pattern with respect to the link between project prioritization and the outcomes of CBA analysis. If prioritization of projects that met the basic requirements were decided according to highest profitability, this would have resulted in high-FEI projects being funded first and, therefore, a concentration of yellow- and orange-coded projects towards the upper limits of the x (FEI) axis. Nor does the figure demonstrate prioritization of projects’ funding order by joint consideration of CBA with other social and environmental goals. Prioritization based on CBA plus alignment with social and environmental goals would have resulted in a concentration of yellow and orange projects in the northeast quadrant of the matrix. Interestingly, three projects –24, 36, and 38– with the highest SEI and FEI scores were not allocated funds for implementation until 2, 9, and 9 years, respectively, after having reached Execution status in the SNI. These results suggest that, while CBA is an important filtering mechanism for road transport, it is not necessarily the basis of prioritizing investments for projects that pass the profitability test. Moreover, while the government does engage in consultation to discuss the social, environmental, and strategic implications of various projects to inform selection, much of this discussion and consultation is informal and largely unstructured. 39 Chapter 5. Conclusion The pilot of the IPF in Chile produced valuable overarching findings beyond the outcomes of the actual prioritization exercise. First, the outputs of CBA were found to be useful inputs to the FEI – inputs that can also be effectively complemented by the simultaneous and direct consideration of social and environmental policy goals that are otherwise difficult to quantify and value. The extensive economic appraisals (partial SCBAs) in place are useful to generate important economic measures associated with projects, including IRRs and NPV index (IVAN) scores, which should be used as key inputs to project prioritization and, in the case of IPF, inputs to the FEI. In this case, the IPF complements rather than substitutes traditional CBA appraisal. The Chile pilot has shown that IPF has application beyond the 'stepping stone' approach previously proposed (Marcelo, et al, 2016), which couched IPF as a stop-gap measure until more sophisticated project appraisal methods could be implemented. Such CBA outputs as benefit-cost ratios, internal rates of return, and net present value indexes can be used to construct the FEI – either as the sole input or in combination with other relevant financial and economic factors, such as risk. As such, the IPF has relevance to a wide array of government capacity levels. It can serve to systematize consideration of key factors, where only limited project information is available, or complement full economic appraisals with additional policy considerations. Moreover, CBA analyses are not necessarily used as the basis of prioritization. In Chile, for example, CBA analysis is used as a filter to eliminate projects with IRRs below a profitability threshold (6%). Once projects pass this filter and are given a positive endorsement for development (RS) within the SNI, they may be prioritized in a number of ways (often unspecified) by the proponent agency. The reservoir MCE directly responds to the inability of the CBA to fully account for critical yet undervalued strategic issues, which resulted in poor CBA results for projects that were required to ensure water security. So then, even where CBA is extensively applied and used to filter out ‘bad’ projects, there is a demand for systematic consideration of other important policy factors in the actual prioritization of investments. Also, in the last couple of years, some road projects that were traditionally appraised using the CBA methodology have been appraised instead by the CEA approach, due to political pressures and considerations, because for those low traffic flows were not allowing the project to satisfy the positive NVP criteria. These kinds of projects were not considered in the prioritization exercise, as the main economic efficiency index (IVAN) values were not possible to get. This can be seen as a signal of how economic efficiency is losing influence in the investment and prioritization decision by the authorities, as the objectives are politically driven. IPF recognizes that, in addition to economic considerations, projects may be valued by governments and other stakeholders due to non-economic considerations such as reducing income disparity or territorial inequality, promoting social cohesion, safeguarding the environment, preserving culture, or managing disaster and climate risks. CBA analyses are often unable to capture such developmental policy goals, which are inherently difficult to valuate. While SCBA should certainly be maintained as the gold standard in project appraisal and a key input for decision making, IPF can help prioritize projects under a variety of analytical conditions and according to multiple policy goals. 40 Recognizing the importance of these additional facets to decision making related to infrastructure development, there is an expressed demand to shift away from purely efficiency-based approaches to approaches that can consider strategic and social goals, that are often undervalued in traditional CBA-like assessments. For example, the Governments of the United Kingdom, Australia, and Chile, and many U.S. state governments have published notes and guidance on the application of multi-criteria decision analysis (MCDA), expanding the discourse to suggest structured ways of employing MCDA to incorporate key policy criteria. Some countries, such as Ireland, have imposed thresholds to guide when government should apply SCBA, multi-criteria analysis, or more simple assessments, depending on the size of the proposed investment. In relation to the PIM framework, the IPF tool is meant to be one possible option in a basket of analytical tools that can be used to gather the best appraisal data available and organize them to support prioritization. In Chile, where multi-criteria approaches to project selection have already been developed and institutionalized, this benefit is clearly recognized. 41 References Agostini, C. and Razmilic, S. (2015). Enfoques complementarios para la evaluacion social de proyectos. Propuestas de Politica Publica CEP. October 2015. Retrieved October 2016, from http://www.cepchile.cl/enfoques-complementarios-para-la-evaluacion-social-de- proyectos/cep/2016-05-13/155921.html. Ahmad, E. and Viscarra, H. (2016). Public investment for sustainable development in Chile. Institutions for Development Sector, Inter-American Development Bank. Discussion Paper IDP-DP-469, September 2016. Candia, J., Perrotti, D. E., & Aldunate, E. (2015). Evaluación social de proyectos Un resumen de las principales metodologías oficiales utilizadas en América Latina y el Caribe. Retrieved from https://repositorio.cepal.org/handle/11362/37954 de Rus Mendoza, G. (2014). The economic evaluation of infrastructure investment. Some inescapable tradeoffs. Documentos de trabajo (FEDEA), (16), 1-31. Gómez-Lobo, A. (2012). Institutional safeguards for cost benefit analysis: lessons from the Chilean national investment system. Journal of Benefit-Cost Analysis, 3(01), 1-30. Maddocks, A., Young, R., and Reig, P. (2015). Ranking the world’s most water-stressed countries in 2040. World Resources Institute (blog), August 26, 2015. http://www.wri.org/blog/2015/08/ranking-world%E2%80%99s-most-water-stressed- countries-2040. Mandri-Perrott, C., Marcelo, D., and Haddon, J. (2014). A methodology to prioritize and select infrastructure investments. Report to the Vietnam Ministry of Planning and Investment. The World Bank. Marcelo, D., Mandri-Perrott, C., and House, S. (2015). Prioritizing infrastructure investments in Panama: Pilot application of the World Bank Infrastructure Prioritization Framework. Report to the Panama Ministry of Economy and Finance. The World Bank. Marcelo, D., Mandri-Perrott, X. C., House, S., and Schwartz, J. (2016), Prioritizing infrastructure investment: A framework for government decision-making. Policy Research Working Paper, 7674. Washington, DC: World Bank. MDS (2014b). Socio-Economic Assessment Methodology for rail transport projects, Division of Social Investment Evaluation (Ministry of Social Development) and Ministry of Transportation Planning - SECTRA (Ministry of Transport and Telecommunications). Retrieved September, 2016, from http://goo.gl/EQHc6w. Minestario de Hacienda (2014). Minuta Matriz Multicriterio Plan de Pequenos Embalses. OECD (2016). OECD Development Pathways Multi-dimensional Review of Uruguay Volume 2: In- depth Analysis and Recommendations. OECD Development Pathways. OECD Publishing, Paris. http://dx.doi.org/10/1787/9789264251663-en. OECD (2017). Gaps and governance standards of public infrastructure in Chile. http://www.oecd.org/gov/ethics/public-infrastructure-in-chile-2017.htm Perreault, W., and Young, F. (1980). Alternating least squares optimal scaling: Analysis of nonmetric data in marketing research. Journal of Marketing Research, 1-13. 42 Rajaram, A., Le, T., Biletska, N., Brumby, J. (2010). A diagnostic framework for assessing Public Investment Management. World Bank Policy Research Working Paper, 5397. Washington, DC: World Bank. Rajaram, Anand; Minh Le, Tuan; Kaiser, Kai; Kim, Jay-Hyung; Frank, Jonas. 2014. The Power of Public Investment Management : Transforming Resources into Assets for Growth. Directions in Development--Public Sector Governance;. World Bank Group, Washington, DC. World Bank. https://openknowledge.worldbank.org/handle/10986/20393 Ruiz-Nuñez, F., & Wei, Z. (2015). Infrastructure investment demands in emerging markets and developing economies. World Bank Policy Research Working Paper, 7414. Washington, DC: World Bank. Saaty, Thomas L. (1990). How to make a decision: The Analytic Hierarchy Process. European Journal of Operational Research, 48(1), 9-26 43 Annex 1. Chilean Law Relevant to Project Appraisal and SNI National Investment System  According to current legislation, investment initiatives to be financed with public funds must have a Ministry of  Social Development record and approval, which must be based on a technical and economic assessment to analyze  profitability.  "The institutions authorized to directly present the SNI investment initiatives are those that are part of the public  sector, in particular services and institutions as defined in Article 2 of the Organic Law of Financial Administration."  Process for Submission of  Projects submitted to SNI are initially assessed in the following ways:  Investment Initiatives   Admissibility  (projects)   Analysis and issuance of the First RATE   Analysis and emission RATE  The presentation of investment initiatives can be made continuously throughout the calendar year.  Investment  initiatives  (independent  of  the  source  of  funding  proposed)  whose  area  of  influence  is  regional,  provincial or communal, and for which competition analysis is regional, must apply via the Ministerial Regional  Secretariat of Social Development.  Investment initiatives whose area of influence is national, international or interregional, and for which competition  analysis is a national must apply to the central‐level Ministry of Social Development.  Investment Programs (Item 03): "These are the expenses for investment initiatives designed to increase, maintain  or regain the ability to generate profits from human or physical resources, and which do not correspond to those  inherent in the institution formulates the program. "  Guidance on Issuance of  The technical economic analysis process begins with the receipt of the investment initiative and culminates with  Economic Analysis and  the issuance of the result of analysis by the Ministry of Social Development. This analysis reviews whether the  Technical Result (RATE)   initiative  was  properly  formulated  and  evaluated,  and  if  it  contains  the  required  technical  and  economic  background indicated in the rules of SNI and sectoral information requirements. The responsibility for this process  Ministry of Social  Development,  lies with the Ministry of Social Development at the central or regional level, as appropriate.  Undersecretary of Social  The analysis of investment initiatives must prove the technical economic benefit of carrying them out, based on an  Investment Evaluation   assessment to analyze their social and economic returns, issuing for that purpose a report in terms set out in Article  Ministry of Finance Budget  19a of Decree Law 1,263, 1975, which is expressed through the Result of Technical Analysis ‐ Economic (RATE) in  Office Rules, Instructions and  the IDI tab of BIP.  Procedures of the National  Assessments of favorably recommended projects (RS) should include:  Investment System  a. The problem to be solved and / or addressed,  b. Analyzed alternatives that would allow to solve the problem, with their corresponding indicators,  c. Alternative selected,  d. Assumptions, results and estimates incorporated in the evaluation,  e. Sensitization of variables, where appropriate,  f. Estimated operating costs and annual maintenance (separately) taken into consideration for evaluation,  g. Certifications pending and that make possible the next step, and  h. The regional sectoral policy strategy to which the initiative contributes.  Projects are given a Technical Objection (OT) RATE due to one of the following reasons:  a. The initiative is badly formulated,  b. The initiative does not use the general evaluation methodology or specific sector concerned,  c. It does not conform to the policies defined for the sector, institution and / or region,  i. The initiative is not a profitable investment (socially or economically) or is technically not feasible,  j. The information presented does not adequately support the quantification and / or valuation of benefits  and / or costs,   k. The background includes simultaneous support for more than one type and / or stage fit, or  l. The investment initiative postulated doubles with an initiative previously entered into the system.  Information Requirements  In  the  state  of  pre‐investment,  proponents  must  prepare  and  evaluate  the  project  to  determine  whether  it  is  for Projects  desirable to implement. The pre‐investment stage includes preparation and appraisal of a project. Analysis must  include market research (supply and demand), as well as technical, economic, environmental, legal and financial  appraisal. Notably, an investment project will not necessarily go through each sub‐stage of pre‐investment status;  this will depend on the complexity and the amounts involved in the project to execute. The requirements to apply  for the stages of pre‐investment status may be different depending on the sector and type of project; therefore, it  is important to review the specific requirements for investment projects contained in the respective sector  http://sni.ministeriodesarrollosocial.gob.cl/evaluacion/exante/requisitos/  44 Annex 2. Water Catchment Raw Project Data Investment  Territorial  Approved  Comarca  Endorse  SRegion  Poverty  Surface  Exprop  Region  i_NPV  Name  LPlots  Benef  Legal  Jobs  NPV  ID  Coihueco,  1  Kaiser  Ñuble  4578  ✔  82.00  ‐1069  140.00  32.86  0.586  2  0  ‐0.23  0.50  0  1  Pinto  2  Las Puentes  Arauco  Arauco  14147    87.00  ‐860  537.56  24.32  0.162  1  0  ‐0.06  0.50  0  1  3  Ranquil  Ñuble  Ranquil  6349  ✔  38.00  ‐4510  46.00  15.10  0.826  2  0  ‐0.71  0.75  1  1  4  Tranaquepe  Concepción  Hualqui  12388    20.00  ‐7464  54.23  30.81  0.369  0  0  ‐0.60  0.50  0  1  5  Laguna El Pillo  Biobío  Laja  1713    10.00  ‐4419  12.67  18.55  0.789  0  0  ‐2.58  0.50  0  1  6  Pichi Bureo  Biobío  Mulchén  13507    120.00  ‐101  1282.32  22.95  0.094  2  1  ‐0.01  0.50  0  1  Biobio  7  Rumena  Arauco  Arauco  18087    11.00  ‐19352  30.54  24.32  0.360  1  0  ‐1.07  0.50  0  1  8  Mirihue  Biobío  Antuco  3244  ✔  57.00  ‐2937  10.60  22.21  5.377  2  0  ‐0.91  0.50  1  1  9  Quidico 1  Arauco  Arauco  4793  ✔  57.00  ‐831  188.58  24.32  0.302  1  0  ‐0.17  0.50  0  1  10  Quidico 2  Arauco  Arauco  20913    293.00  ‐602  1134.95  24.32  0.258  1  0  ‐0.03  0.50  0  1  11  Perales  Ñuble  Coelemu  11294    21.00  ‐3582  103.01  24.80  0.204  1  0  ‐0.32  0.50  0  1  12  Vegas de Itata  Ñuble  Coelemu  7641    19.00  ‐2746  90.92  24.80  0.209  2  0  ‐0.36  0.50  0  1  13  Tauco  Ñuble  Coelemu  4077    16.00  ‐2667  14.67  24.82  1.090  1  0  ‐0.65  0.50  0  1  14  Leoneras  Ñuble  Coelemu  7128    189.00  ‐1326225  5.69  24.80  33.193  2  1    0.50  0  1  15  Huencuecho I  Talca  Pelarco  11949    69.00  ‐829  301.33  12.10  0.229  1  0  ‐0.07  0.00  0  2  16  Huencuecho II  Talca  Pelarco  8646    69.00  ‐687  180.84  12.10  0.382  1  0  ‐0.08  0.00  0  2  17  Peralito  Talca  San Clemente  2451  ✔  46.00  609  327.00  14.90  0.141  2  0  0.25  0.00  1  3  18  Peralito 2  Talca  San Clemente  7322    46.00  ‐666  148.40  14.90  0.310  2  1  ‐0.09  0.00  0  2  19  La Bruja  Talca  San Clemente  2863  ✔  75.00  ‐1778  172.00  14.90  0.436  2  0  ‐0.62  0.75  1  3  20  El Guindo  Talca  Río Claro  3286    26.00  ‐1314  131.35  8.40  0.198  0  0  ‐0.40  0.00  0  2  Maule  21  Sauzal  Talca  Empedrado  4539    17.00    90.40  17.30  0.188  1  0    0.00  0  2  22  Botacura  Linares  San Javier  7130    8.00    140.02  17.30  0.057  0  0    0.25  0  2  23  Vaquería  Linares  San Javier  3655  ✔  130.00  ‐2768  280.00  17.30  0.464  0  0  ‐0.76  0.00  1  3  24  Manantiales  Linares  San Javier  2050    31.00  ‐2850  177.78  17.30  0.174  1  0  ‐1.39  0.00  0  2  25  El Molino  Linares  Retiro  4585    144.00  4546  803.97  15.40  0.179  1  1  0.99  0.00  0  2  Derivado  26  Linares  Parral‐Retiro  6198  ✔  123.00  2762  362.00  15.40  0.340  1  0  0.45  0.50  1  3  Porvenir 2  27  Limávida  Talca  Curepto  10577  ✔  58.00  ‐507  163.28  22.90  0.355  1  0  ‐0.05  0.25  1  2  45 Investmen Territorial  Approved  Comarca  Endorse  SRegion  Poverty  Surface  Exprop  Region  i_NPV  Name  LPlots  Benef  Legal  Jobs  NPV  ID  t  28  Codegua (CFGD)  Cachapoal  Codegua  10477  ✔  1200.00  ‐8944  1113.00  8.00  1.078  1  2  ‐0.85  0.75  1  3  29  San Francisco 1 (Zonada)  Cardenal Caro  Litueche  8355    156.00  ‐4654  156.00  15.00  1.000  1  1  ‐0.56  0.25  1  1  30  San Francisco 2 (Zonada)  Cardenal Caro  Litueche  9324    142.00  ‐5186  142.00  15.00  1.000  1  1  ‐0.56  0.25  1  1  31  San Francisco 3 (Zonada)  Cardenal Caro  Litueche  11261    148.00  ‐7215  148.00  15.00  1.000  1  1  ‐0.64  0.25  1  1    ‐ 32  El Maiten 1 (Zonada)  Cardenal Caro  Navidad  14980  183.00  183.00  8.00  1.000  1  2  ‐0.83  0.25  0  1  12461  33  El Maiten 2 (Zonada)  Cardenal Caro  Navidad  6552    182.00  ‐5474  182.00  8.00  1.000  1  2  ‐0.84  0.25  0  1    ‐ 34  El Maiten 3 (Zonada)  Cardenal Caro  Navidad  17838  133.00  133.00  8.00  1.000  1  2  ‐0.83  0.25  0  1  14819  35  Huehuinco (CFRD)  Cardenal Caro  Navidad  3945    400.00  ‐3363  400.00  8.00  1.000  1  1  ‐0.85  0.25  0  1  36  Ucúquer (CFRD)  Cardenal Caro  Navidad  5518    73.00  ‐4592  73.00  8.00  1.000  1  0  ‐0.83  0.25  0  1  37  Manquehue 1 (Zonada)  Cardenal Caro  Litueche  14980    133.00  ‐6945  133.00  15.00  1.000  1  1  ‐0.46  0.00  1  3  O'Higgins  38  Manquehue 2 (Zonada)  Cardenal Caro  Litueche  17838    84.00  ‐6122  84.00  15.00  1.000  1  2  ‐0.34  0.00  1  3  39  Los Tricahues (RCC)  Colchagua  Lolol  7776  ✔  1317.00  ‐6753  1317.00  18.00  1.000  1  1  ‐0.87  0.50  1  1    ‐ 40  Las Palmas 2 (RCC)  Colchagua  Lolol  12252  370.00  370.00  18.00  1.000  1  1  ‐0.84  0.50  1  1  10243  41  Estero Seco (CFGD)  Cardenal Caro  La Estrella  6292    80.00  ‐5235  80.00  15.00  1.000  1  1  ‐0.83  0.25  1  2  Cementerio 1    42  Cardenal Caro  La Estrella  5518  30.00  ‐4582  30.00  15.00  1.000  1  1  ‐0.83  0.75  0  1  (Hormigón)  Cementerio 2    43  Cardenal Caro  La Estrella  5518  30.00  ‐4582  30.00  15.00  1.000  1  1  ‐0.83  0.75  0  1  (Hormigón)  44  Manquehua (Zonada)  Cardenal Caro  Litueche  1122  ✔  25.00  ‐9305  5.00  15.00  5.000  1  0  ‐8.29  0.00  1  3  45  La Virgen (Zonada)  Cardenal Caro  La Estrella  1045  ✔  30.00  ‐868  6.00  15.00  5.000  1  0  ‐0.83  0.50  1  3  46  La Palmera (Zonada)  Cachapoal  Pichidegua  1031  ✔  123.00  ‐883  123.00  8.00  1.000  1  0  ‐0.86  0.50  1  3  47  Pulin (Zonada)  Cardenal Caro  Litueche  1013  ✔  20.00  ‐841  4.00  8.00  5.000  1  0  ‐0.83  0.50  1  3  48  Rapel (Zonada)  Cachapoal  Rapel  594    10.00  ‐493  2.00  8.00  5.000  1  0  ‐0.83  0.75  1  2  49  Licancheu (Zonada)  Cardenal Caro  Navidad  987    10.00  ‐819  2.00  15.00  5.000  1  0  ‐0.83  0.75  1  2    ‐ 50  Vitahue  Petorca  Cabildo  14865  40.00  9.14  12.24  4.378  0  1  ‐0.93  0.25  0  1  13779  51  Paihuen  Petorca  Cabildo  16685    6.00  ‐8130  17.38  12.24  0.345  0  1  ‐0.49  0.25  0  1  52  Chalaco  Petorca  Petorca  33316    22.00  ‐9794  36.48  15.49  0.603  2  1  ‐0.29  0.25  0  1    Valparaiso  53  Las Carditas 2  Petorca  Petorca  14022  54.00  ‐4853  24.47  15.49  2.207  2  1  ‐0.35  0.25  0  1  54  Santa Marta  Petorca  La Ligua  3992  ✔  79.00  ‐602  23.00  26.31  3.435  1  0  ‐0.15  0.50  1  3  55  Pullally  Petorca  La Ligua  4529    170.00  ‐522  93.10  26.31  1.826  1  0  ‐0.12  0.25  0  3  56  Cuncumen 1  San Antonio  San Antonio  6245    8.00  ‐2347  28.53  20.41  0.280  0  0  ‐0.38  0.25  0  1  57  Lo Zárate 2  San Antonio  Cartagena  6891    57.00  ‐1110  66.59  24.40  0.856  0  0  ‐0.16  0.25  0  3  58  El Zaino  San Felipe  Sta María  5013  ✔  164.00  ‐1878  48.00  7.00  3.417  1  0  ‐0.37  0.25  1  3  59  Valle Hermoso  Petorca  La Ligua  4618    562.00  ‐4024  9.72  26.31  57.819  1  1  ‐0.87  0.25  0  1  46 60  Santa Julia  Petorca  Petorca  2444  ✔  73.00  ‐1370  13.00  15.50  5.615  2  0  ‐0.56  0.75  1  3  61  Pedegua  Petorca  Petorca  2024  ✔  179.00  ‐1149  17.00  15.50  10.529  2  0  ‐0.57  0.25  1  3  Annex 3. Water Catchment Project SEI Calculations, by Region VARIABLES  POVERTY  BENEFICIARY  JOBS  TERRITORIAL  APPLIED WEIGHTS  0.316  0.642  0.400  0.573  ID  Project Name  Poverty  Beneficiary  Jobs  Territorial  pov_std  ben_std  job_std  terr_std  SEI_STD  SEI    0    15.1  0.094  0  0  ‐2.12  ‐0.35  ‐1.77  ‐0.39  ‐1.83  0  1  Kaiser  32.86  0.586  2  0  2.01  ‐0.29  0.98  ‐0.39  0.62  42  2  Las Puentes  24.32  0.162  1  0  0.02  ‐0.34  ‐0.39  ‐0.39  ‐0.59  21  3  Ranquil  15.1  0.826  2  0  ‐2.12  ‐0.26  0.98  ‐0.39  ‐0.67  20  4  Tranaquepe  30.81  0.369  0  0  1.53  ‐0.32  ‐1.77  ‐0.39  ‐0.65  20  5  Laguna El Pillo  18.55  0.789  0  0  ‐1.32  ‐0.27  ‐1.77  ‐0.39  ‐1.52  5  6  Pichi Bureo  22.95  0.094  2  1  ‐0.29  ‐0.35  0.98  2.36  1.43  56  Biobio  7  Rumena  24.32  0.360  1  0  0.02  ‐0.32  ‐0.39  ‐0.39  ‐0.58  21  8  Mirihue  22.21  5.377  2  0  ‐0.47  0.26  0.98  ‐0.39  0.19  35  9  Quidico 1  24.32  0.302  1  0  0.02  ‐0.32  ‐0.39  ‐0.39  ‐0.58  21  10  Quidico 2  24.32  0.258  1  0  0.02  ‐0.33  ‐0.39  ‐0.39  ‐0.59  21  11  Perales  24.8  0.204  1  0  0.14  ‐0.33  ‐0.39  ‐0.39  ‐0.55  22  12  Vegas de Itata  24.8  0.209  2  0  0.14  ‐0.33  0.98  ‐0.39  0.00  31  13  Tauco  24.82  1.090  1  0  0.14  ‐0.23  ‐0.39  ‐0.39  ‐0.49  23  14  Leoneras  24.8  33.193  2  1  0.14  3.43  0.98  2.36  3.99  100  0    8.4  0.057  0  0  ‐2.03  ‐1.68  ‐1.41  ‐0.41  ‐2.52  0  15  Huencuecho I  12.1  0.229  1  0  ‐0.96  ‐0.30  0.00  ‐0.41  ‐0.73  39  16  Huencuecho II  12.1  0.382  1  0  ‐0.96  0.94  0.00  ‐0.41  0.06  57  17  Peralito  14.9  0.141  2  0  ‐0.15  ‐1.01  1.41  ‐0.41  ‐0.36  47  18  Peralito 2  14.9  0.310  2  1  ‐0.15  0.36  1.41  2.25  2.04  100  19  La Bruja  14.9  0.436  2  0  ‐0.15  1.38  1.41  ‐0.41  1.17  81  Maule  20  El Guindo  8.4  0.198  0  0  ‐2.03  ‐0.55  ‐1.41  ‐0.41  ‐1.79  16  21  Sauzal  17.3  0.188  1  0  0.55  ‐0.63  0.00  ‐0.41  ‐0.46  45  22  Botacura  17.3  0.057  0  0  0.55  ‐1.68  ‐1.41  ‐0.41  ‐1.71  18  23  Vaquería  17.3  0.464  0  0  0.55  1.60  ‐1.41  ‐0.41  0.40  64  24  Manantiales  17.3  0.174  1  0  0.55  ‐0.74  0.00  ‐0.41  ‐0.53  44  25  El Molino  15.4  0.179  1  1  0.00  ‐0.70  0.00  2.25  0.84  74  26  Derivado Porvenir 2  15.4  0.340  1  0  0.00  0.60  0.00  ‐0.41  0.15  59  47 27  Limávida  22.9  0.355  1  0  2.18  0.72  0.00  ‐0.41  0.92  75  48   ID  Project Name  Poverty  Beneficiary  Jobs  Territorial  pov_std  ben_std  job_std  terr_std  SEI_STD  SEI  0    8  1.000  1  0  ‐1.15  ‐0.53    ‐1.21  ‐1.40  0  28  Codegua (CFGD)  8  1.078  1  2  ‐1.15  ‐0.49    1.45  0.16  74  29  San Francisco 1 (Zonada)  15  1.000  1  1  0.67  ‐0.53    0.12  ‐0.06  64  30  San Francisco 2 (Zonada)  15  1.000  1  1  0.67  ‐0.53    0.12  ‐0.06  64  31  San Francisco 3 (Zonada)  15  1.000  1  1  0.67  ‐0.53    0.12  ‐0.06  64  32  El Maiten 1 (Zonada)  8  1.000  1  2  ‐1.15  ‐0.53    1.45  0.13  73  33  El Maiten 2 (Zonada)  8  1.000  1  2  ‐1.15  ‐0.53    1.45  0.13  73  34  El Maiten 3 (Zonada)  8  1.000  1  2  ‐1.15  ‐0.53    1.45  0.13  73  35  Huehuinco (CFRD)  8  1.000  1  1  ‐1.15  ‐0.53    0.12  ‐0.63  36  36  Ucúquer (CFRD)  8  1.000  1  0  ‐1.15  ‐0.53    ‐1.21  ‐1.40  0    O’Higgins  37  Manquehue 1 (Zonada)  15  1.000  1  1  0.67  ‐0.53  0.12  ‐0.06  64  38  Manquehue 2 (Zonada)  15  1.000  1  2  0.67  ‐0.53    1.45  0.70  100  39  Los Tricahues (RCC)  18  1.000  1  1  1.45  ‐0.53    0.12  0.19  75  40  Las Palmas 2 (RCC)  18  1.000  1  1  1.45  ‐0.53    0.12  0.19  75  41  Estero Seco (CFGD)  15  1.000  1  1  0.67  ‐0.53    0.12  ‐0.06  64  42  Cementerio 1 (Hormigón)  15  1.000  1  1  0.67  ‐0.53    0.12  ‐0.06  64  43  Cementerio 2 (Hormigón)  15  1.000  1  1  0.67  ‐0.53    0.12  ‐0.06  64  44  Manquehua (Zonada)  15  5.000  1  0  0.67  1.80    ‐1.21  0.68  99  45  La Virgen (Zonada)  15  5.000  1  0  0.67  1.80    ‐1.21  0.68  99  46  La Palmera (Zonada)  8  1.000  1  0  ‐1.15  ‐0.53    ‐1.21  ‐1.40  0  47  Pulin (Zonada)  8  5.000  1  0  ‐1.15  1.80    ‐1.21  0.10  71  48  Rapel (Zonada)  8  5.000  1  0  ‐1.15  1.80    ‐1.21  0.10  71  49  Licancheu (Zonada)  15  5.000  1  0  0.67  1.80    ‐1.21  0.68  99  0    7  0.280  0  0  ‐1.70  ‐0.46  ‐1.17  ‐0.81  ‐1.76  0  50  Vitahue  12.24  4.378  0  1  ‐0.90  ‐0.20  ‐1.17  1.13  ‐0.23  32  51  Paihuen  12.24  0.345  0  1  ‐0.90  ‐0.45  ‐1.17  1.13  ‐0.39  28  52  Chalaco  15.49  0.603  2  1  ‐0.40  ‐0.44  1.17  1.13  0.71  51  53  Las Carditas 2  15.49  2.207  2  1  ‐0.40  ‐0.34  1.17  1.13  0.77  53  Valparaiso  54  Santa Marta  26.31  3.435  1  0  1.26  ‐0.26  0.00  ‐0.81  ‐0.23  32  55  Pullally  26.31  1.826  1  0  1.26  ‐0.36  0.00  ‐0.81  ‐0.30  30  56  Cuncumen 1  20.41  0.280  0  0  0.35  ‐0.46  ‐1.17  ‐0.81  ‐1.11  14  57  Lo Zárate 2  24.4  0.856  0  0  0.97  ‐0.42  ‐1.17  ‐0.81  ‐0.90  18  58  El Zaino  7  3.417  1  0  ‐1.70  ‐0.26  0.00  ‐0.81  ‐1.17  12  59  Valle Hermoso  26.31  57.819  1  1  1.26  3.12  0.00  1.13  3.05  100  60  Santa Julia  15.5  5.615  2  0  ‐0.40  ‐0.12  1.17  ‐0.81  ‐0.20  32  49 61  Pedegua  15.5  10.529  2  0  ‐0.40  0.18  1.17  ‐0.81  0.00  37  Annex 4. Water Catchment Project FEI Calculations, by Region VARIABLES  NPV  EXPROPRIATIONS  ENDORSEMENT  LEGAL  APPLIED WEIGHTS  0.316  0.316  0.635  0.630  ID  Project Name  NPV_inv  Expropriations  Endorsement  Legal  NPV_std  exp_std  endorse_std  Legal_std  FEI_STD  FEI    0    ‐2.579  1  0  1  ‐2.89  ‐0.27  ‐0.39    ‐1.25  0.0  1  Kaiser  ‐0.233  1  0  1  0.52  ‐0.27  ‐0.39    ‐0.17  28.5  2  Las Puentes  ‐0.061  1  0  1  0.77  ‐0.27  ‐0.39    ‐0.09  30.6  3  Ranquil  ‐0.710  1  1  1  ‐0.17  3.47  2.36    2.54  100.0  4  Tranaquepe  ‐0.603  1  0  1  ‐0.01  ‐0.27  ‐0.39    ‐0.34  24.0  5  Laguna El Pillo  ‐2.579  1  0  1  ‐2.89  ‐0.27  ‐0.39    ‐1.25  0.0  6  Pichi Bureo  ‐0.007  1  0  1  0.85  ‐0.27  ‐0.39    ‐0.07  31.2  Biobio  7  Rumena  ‐1.070  1  0  1  ‐0.69  ‐0.27  ‐0.39    ‐0.55  18.3  8  Mirihue  ‐0.905  1  1  1  ‐0.46  ‐0.27  2.36    1.27  66.4  9  Quidico 1  ‐0.173  1  0  1  0.61  ‐0.27  ‐0.39    ‐0.14  29.2  10  Quidico 2  ‐0.029  1  0  1  0.82  ‐0.27  ‐0.39    ‐0.08  30.9  11  Perales  ‐0.317  1  0  1  0.40  ‐0.27  ‐0.39    ‐0.21  27.4  12  Vegas de Itata  ‐0.359  1  0  1  0.34  ‐0.27  ‐0.39    ‐0.23  26.9  13  Tauco  ‐0.654  1  0  1  ‐0.09  ‐0.27  ‐0.39    ‐0.36  23.4  14  Leoneras    1  0  1    ‐0.27  ‐0.39    ‐0.33  24.1  0    ‐1.390  0  0  2  ‐1.94  ‐0.56  ‐0.76  ‐0.64  ‐1.67  0.0  15  Huencuecho I  ‐0.069  0  0  2  0.14  ‐0.56  ‐0.76  ‐0.64  ‐1.02  15.9  16  Huencuecho II  ‐0.079  0  0  2  0.13  ‐0.56  ‐0.76  ‐0.64  ‐1.02  15.8  17  Peralito  0.248  0  1  3  0.64  ‐0.56  1.22  1.44  1.71  81.8  18  Peralito 2  ‐0.091  0  0  2  0.11  ‐0.56  ‐0.76  ‐0.64  ‐1.03  15.7  19  La Bruja  ‐0.621  1  1  3  ‐0.72  2.54  1.22  1.44  2.25  95.1  Maule  20  El Guindo  ‐0.400  0  0  2  ‐0.38  ‐0.56  ‐0.76  ‐0.64  ‐1.18  11.9  21  Sauzal    0  0  2    ‐0.56  ‐0.76  ‐0.64  ‐1.06  14.8  22  Botacura    0  0  2    0.48  ‐0.76  ‐0.64  ‐0.73  22.7  23  Vaquería  ‐0.757  0  1  3  ‐0.94  ‐0.56  1.22  1.44  1.21  69.7  24  Manantiales  ‐1.390  0  0  2  ‐1.94  ‐0.56  ‐0.76  ‐0.64  ‐1.67  0.0  25  El Molino  0.992  0  0  2  1.82  ‐0.56  ‐0.76  ‐0.64  ‐0.49  28.7  26  Derivado Porvenir 2  0.446  1  1  3  0.96  1.51  1.22  1.44  2.46  100.0  27  Limávida  ‐0.048  0  1  2  0.18  0.48  1.22  ‐0.64  0.58  54.4  50 51 ID  Project Name  NPV_inv  Expropriations  Endorsement  Legal  NPV_std  exp_std  endorse_std  Legal_std  FEI_STD  FEI    0    ‐8.291  0  0  1  ‐4.46  ‐1.53  ‐1.43  ‐0.84  ‐3.33  0.0  28  Codegua (CFGD)  ‐0.854  1  1  3  0.15  1.44  0.67  1.33  1.77  100.0  29  San Francisco 1 (Zonada)  ‐0.557  0  1  1  0.34  ‐0.54  0.67  ‐0.84  ‐0.17  62.1  30  San Francisco 2 (Zonada)  ‐0.556  0  1  1  0.34  ‐0.54  0.67  ‐0.84  ‐0.17  62.1  31  San Francisco 3 (Zonada)  ‐0.641  0  1  1  0.29  ‐0.54  0.67  ‐0.84  ‐0.18  61.7  32  El Maiten 1 (Zonada)  ‐0.832  0  0  1  0.17  ‐0.54  ‐1.43  ‐0.84  ‐1.55  34.9  33  El Maiten 2 (Zonada)  ‐0.835  0  0  1  0.17  ‐0.54  ‐1.43  ‐0.84  ‐1.55  34.8  34  El Maiten 3 (Zonada)  ‐0.831  0  0  1  0.17  ‐0.54  ‐1.43  ‐0.84  ‐1.55  34.9  35  Huehuinco (CFRD)  ‐0.852  0  0  1  0.16  ‐0.54  ‐1.43  ‐0.84  ‐1.56  34.8  36  Ucúquer (CFRD)  ‐0.832  0  0  1  0.17  ‐0.54  ‐1.43  ‐0.84  ‐1.55  34.9  O’Higgins  37  Manquehue 1 (Zonada)  ‐0.464  0  1  3  0.40  ‐1.53  0.67  1.33  0.90  83.1  38  Manquehue 2 (Zonada)  ‐0.343  0  1  3  0.47  ‐1.53  0.67  1.33  0.93  83.5  39  Los Tricahues (RCC)  ‐0.868  1  1  1  0.15  0.45  0.67  ‐0.84  0.08  67.0  40  Las Palmas 2 (RCC)  ‐0.836  1  1  1  0.17  0.45  0.67  ‐0.84  0.09  67.1  41  Estero Seco (CFGD)  ‐0.832  0  1  2  0.17  ‐0.54  0.67  0.25  0.46  74.4  42  Cementerio 1 (Hormigón)  ‐0.830  1  0  1  0.17  1.44  ‐1.43  ‐0.84  ‐0.93  47.1  43  Cementerio 2 (Hormigón)  ‐0.830  1  0  1  0.17  1.44  ‐1.43  ‐0.84  ‐0.93  47.1  44  Manquehua (Zonada)  ‐8.291  0  1  3  ‐4.46  ‐1.53  0.67  1.33  ‐0.63  52.9  45  La Virgen (Zonada)  ‐0.831  1  1  3  0.17  0.45  0.67  1.33  1.46  93.9  46  La Palmera (Zonada)  ‐0.856  1  1  3  0.15  0.45  0.67  1.33  1.45  93.8  47  Pulin (Zonada)  ‐0.830  1  1  3  0.17  0.45  0.67  1.33  1.46  93.9  48  Rapel (Zonada)  ‐0.830  1  1  2  0.17  1.44  0.67  0.25  1.09  86.7  49  Licancheu (Zonada)  ‐0.830  1  1  2  0.17  1.44  0.67  0.25  1.09  86.7  0    ‐0.927  0  0  1  ‐1.87  ‐0.40  ‐0.68  ‐0.96  ‐1.75  0.0  50  Vitahue  ‐0.927  0  0  1  ‐1.87  ‐0.40  ‐0.68  ‐0.96  ‐1.75  0.0  51  Paihuen  ‐0.487  0  0  1  ‐0.20  ‐0.40  ‐0.68  ‐0.96  ‐1.22  13.4  52  Chalaco  ‐0.294  0  0  1  0.54  ‐0.40  ‐0.68  ‐0.96  ‐0.99  19.2  53  Las Carditas 2  ‐0.346  0  0  1  0.34  ‐0.40  ‐0.68  ‐0.96  ‐1.05  17.7  Valparaiso  54  Santa Marta  ‐0.151  1  1  3  1.08  1.21  1.35  0.96  2.19  99.6  55  Pullally  ‐0.115  0  0  3  1.22  ‐0.40  ‐0.68  0.96  0.43  55.2  56  Cuncumen 1  ‐0.376  0  0  1  0.23  ‐0.40  ‐0.68  ‐0.96  ‐1.09  16.8  57  Lo Zárate 2  ‐0.161  0  0  3  1.04  ‐0.40  ‐0.68  0.96  0.38  53.8  58  El Zaino  ‐0.375  0  1  3  0.23  ‐0.40  1.35  0.96  1.41  79.9  59  Valle Hermoso  ‐0.871  0  0  1  ‐1.65  ‐0.40  ‐0.68  ‐0.96  ‐1.68  1.7  60  Santa Julia  ‐0.561  1  1  3  ‐0.47  2.82  1.35  0.96  2.20  100.0  52 61  Pedegua  ‐0.568  0  1  3  ‐0.50  ‐0.40  1.35  0.96  1.18  74.0  Annex 5. Road Transport Raw Project Data Project  ID  Stage  Allocation  Allocation2  Invest2  Length  BEN  Ben_inv  Jobs  Jobs_inv  Savings  Savings_inv  POV  POV2  Indig  IVAN  P1  30106685  1  0  1  11,698  16.0  15,605  1.334  481  0.041  321,133  27.5  28.45  22.90  1  1.72  P2  30080831  1  3  1  21,371  29  5,891  0.276  1,649  0.077  2,090,057  97.8  25.24  26.27  0  0.57  P3  30096463  1  2  1  1,776  7.8  1,427  0.803  122  0.069  94,958  53.5  23.60  29.20  1  0.06  P4  30076931  1  0  1  12,262  6.0  3,614  0.295  510  0.042  1,104,082  90.0  7.10  20.10  1  0.63  P5  30172722  1  2  1  4,898  14  1,597  0.326  313  0.064  273,980  55.9  17.60  19.20  0  0.09  P6  30107176  1  0  1  9,530  20  1,525  0.160  635  0.067  683,701  71.7  23.60  29.20  0  0.59  P7  30124028  1  9  0  625  12  1,949  3.118  444  0.710  45,796  73.3  17.60  19.20  0  0.30  P8  30090914  1  9  0  26,913  32.1  1,727  0.064  1,530  0.057  1,491,948  55.4  26.83  36.62  1  0.05  P9  30083777  1  1  1  6,747  14  6,010  0.891  444  0.066  598,069  88.6  9.70  21.00  0  0.49  P10  30071390  1  1  1  14,026  35  754  0.054  994  0.071  323,581  23.1  16.80  22.60  0  0.12  P11  30131878  1  2  1  3,735  3.8  4,907  1.314  237  0.063  145,979  39.1  16.10  23.20  1  1.15  P12  30137944  1  9  0  28,397  29.5  4,784  0.168  1,986  0.070  988,411  34.8  5.05  19.18  1  0.23  P13  30427024  1  9  0  36,791  110.0  1,495  0.041  2,534  0.069  4,482,591  121.8  5.40  17.20  0  1.35  P14  30104149  1  0  1  6,123  9.0  4,039  0.660  380  0.062  299,852  49.0  12.00  18.20  1  0.06  P15  30073279  1  0  1  893  6  1,843  2.064  56  0.063  108,178  121.2  21.90  22.49  0  0.76  P16  30091314  1  1  1  4,982  15.5  2,560  0.514  211  0.042  217,797  43.7  23.27  36.13  1  0.11  P17  30078400  1  0  1  4,513  7.4  2,298  0.509  234  0.052  261,893  58.0  14.93  22.79  1  0.18  P18  30070422  1  5  1  4,130  7.8  3,117  0.755  218  0.053  157,887  38.2  17.60  19.20  1  0.26  P19  30393223  1  9  0  5,202  8.8  1,432  0.275  366  0.070  223,331  42.9  14.54  15.79  1  0.47  P20  30135925  1  0  1  9,056  2.8  10,919  1.206  637  0.070  823,735  91.0  6.55  13.47  0  1.38  P21  30276122  1  9  0  8,221  16.0  1,687  0.205  545  0.066  299,093  36.4  23.60  29.20  1  0.01  P22  30064808  1  9  0  3,775  8.4  2,509  0.665  266  0.070  177,032  46.9  22.38  28.91  1  0.12  P23  30071804  1  2  1  3,970  12.1  1,944  0.490  237  0.060  211,746  53.3  10.77  21.18  1  0.33  P24  30089927  1  2  1  16,483  33  9,668  0.587  1,107  0.067  13,594,274  824.7  5.40  17.20  0  1.70  P25  30080601  1  1  1  14,921  23  6,640  0.445  999  0.067  923,499  61.9  16.80  22.60  0  0.10  P26  30080632  1  2  1  22,698  40  1,202  0.053  1,199  0.053  835,561  36.8  17.67  23.31  0  0.08  P27  30275022  1  0  1  280  2  1,102  3.934  20  0.071  14,321  51.1  16.10  23.20  0  0.08  P28  30083882  1  2  1  11,133  20  4,430  0.398  1,056  0.095  584,572  52.5  11.12  18.85  0  0.72  P29  30106248  1  0  1  14,297  39.0  2,786  0.195  955  0.067  2,167,634  151.6  5.40  17.20  1  1.25  P30  30106756  1  0  1  26,905  44  6,212  0.231  1,885  0.070  3,523,633  131.0  11.12  18.85  0  0.95  P31  30122219  1  2  1  11,453  9  3,084  0.269  813  0.071  213,608  18.7  4.40  9.10  0  0.13  P32  30061863  1  1  1  25,894  21  24,354  0.941  1,602  0.062  3,220,696  124.4  7.10  20.10  0  0.53  P33  30081072  1  1  1  5,903  12  6,222  1.054  492  0.083  634,376  107.5  9.70  21.00  0  0.65  P34  30106138  1  0  1  5,784  3.0  2,018  0.349  352  0.061  81,484  14.1  16.80  22.60  0  0.83  P35  30044558  1  1  1  12,246  24.3  1,808  0.148  825  0.067  1,046,948  85.5  25.20  25.58  1  0.56  P36  20079319  1  9  1  26,910  28.2  9,218  0.343  1,716  0.064  8,659,859  321.8  17.60  19.20  1  3.19  P37  30091479  1  0  1  12,611  11.2  39,110  3.101  653  0.052  2,151,890  170.6  8.86  18.05  1  1.57  P38  30381293  1  9  1  8,004  19.0  10,313  1.288  508  0.063  1,982,596  247.7  13.80  23.90  1  2.53  P39  20184422  1  1  1  7,428  16  827  0.111  479  0.064  128,461  17.3  23.60  29.20  0  0.07  P40  30134178  1  1  1  1,569  2.8  6,951  4.430  112  0.071  143,378  91.4  8.86  18.05  1  0.54  P41  30108830  1  0  1  23,258  28  20,502  0.882  1,573  0.068  3,387,199  145.6  9.93  15.53  0  0.94  P42  30080314  1  1  1  1,719  4.5  2,174  1.265  97  0.056  98,927  57.6  15.82  25.38  1  0.02  P43  30077537  1  2  1  15,406  29  7,062  0.458  890  0.058  2,560,370  166.2  4.62  27.82  0  1.29  P44  30131496  1  1  1  10,821  12.0  10,000  0.924  761  0.070  675,083  62.4  6.50  16.90  0  0.44  P45  30218272  1  1  1  10,540  4  61,374  5.823  544  0.052  837,548  79.5  12.00  18.20  0  0.53  P46  30081190  1  0  1  2,968  8  1,454  0.490  198  0.067  87,199  29.4  13.70  23.00  1  0.02  P47  30070762  1  1  1  15,225  26  9,619  0.632  1,020  0.067  2,092,546  137.4  16.10  23.20  0  1.04  53 P48  30123462  1  1  1  1,526  6.5  1,724  1.130  97  0.064  101,585  66.6  16.10  23.20  1  0.26  P49  30122528  1  1  1  4,942  21.48  2,809  0.568  348  0.070  289,395  58.6  23.60  29.20  1  0.42  P50  30217972  1  9  0  1,650  4  3,083  1.868  89  0.054  111,427  67.5  12.00  18.20  1  0.25  Annex 6. Road Transport Project SEI Calculations, by Region (with standard poverty rate) BEN_INV  JOBS_INV  SAVINGS_INV  POV_C  0.614  0.224  0.724  0.224  Project  ID  Ben_inv  Jobc_inv  Savings_inv  Pov_c  Ben_std  Jobs_std  Savings_std  Pov_std  SEI_STD  SEI  P1  30106685  1.334  0.041  27.5  28.4  0.31  ‐0.39  ‐0.57  2.03  0.15  26.37  P2  30080831  0.276  0.077  97.8  25.2  ‐0.58  0.00  0.02  1.55  0.01  23.57  P3  30096463  0.803  0.069  53.5  23.6  ‐0.13  ‐0.09  ‐0.35  1.31  ‐0.06  22.18  P4  30076931  0.295  0.042  90.0  7.1  ‐0.56  ‐0.38  ‐0.05  ‐1.15  ‐0.72  9.15  P5  30172722  0.326  0.064  55.9  17.6  ‐0.53  ‐0.14  ‐0.33  0.41  ‐0.51  13.40  P6  30107176  0.160  0.067  71.7  23.6  ‐0.67  ‐0.11  ‐0.20  1.31  ‐0.29  17.68  P7  30124028  3.118  0.710  73.3  17.6  1.81  6.89  ‐0.19  0.41  2.61  75.25  P8  30090914  0.064  0.057  55.4  26.8  ‐0.75  ‐0.22  ‐0.34  1.79  ‐0.36  16.40  P9  30083777  0.891  0.066  88.6  9.7  ‐0.06  ‐0.12  ‐0.06  ‐0.76  ‐0.28  17.97  P10  30071390  0.054  0.071  23.1  16.8  ‐0.76  ‐0.07  ‐0.61  0.30  ‐0.86  6.50  P11  30131878  1.314  0.063  39.1  16.1  0.30  ‐0.15  ‐0.47  0.19  ‐0.15  20.48  P12  30137944  0.168  0.070  34.8  5.0  ‐0.67  ‐0.08  ‐0.51  ‐1.45  ‐1.12  1.28  P13  30427024  0.041  0.069  121.8  5.4  ‐0.77  ‐0.09  0.22  ‐1.40  ‐0.65  10.54  P14  30104149  0.660  0.062  49.0  12.0  ‐0.25  ‐0.16  ‐0.39  ‐0.42  ‐0.57  12.20  P15  30073279  2.064  0.063  121.2  21.9  0.93  ‐0.16  0.21  1.05  0.92  41.75  P16  30091314  0.514  0.042  43.7  23.3  ‐0.38  ‐0.38  ‐0.43  1.26  ‐0.35  16.56  P17  30078400  0.509  0.052  58.0  14.9  ‐0.38  ‐0.27  ‐0.31  0.02  ‐0.52  13.18  P18  30070422  0.755  0.053  38.2  17.6  ‐0.17  ‐0.26  ‐0.48  0.41  ‐0.42  15.13  P19  30393223  0.275  0.070  42.9  14.5  ‐0.58  ‐0.07  ‐0.44  ‐0.04  ‐0.70  9.62  P20  30135925  1.206  0.070  91.0  6.6  0.20  ‐0.07  ‐0.04  ‐1.23  ‐0.19  19.61  P21  30276122  0.205  0.066  36.4  23.6  ‐0.64  ‐0.12  ‐0.49  1.31  ‐0.48  13.90  P22  30064808  0.665  0.070  46.9  22.4  ‐0.25  ‐0.07  ‐0.41  1.13  ‐0.21  19.25  P23  30071804  0.490  0.060  53.3  10.8  ‐0.40  ‐0.19  ‐0.35  ‐0.60  ‐0.68  10.06  P24  30089927  0.587  0.067  824.7  5.4  ‐0.32  ‐0.11  6.06  ‐1.40  3.86  100.00  P25  30080601  0.445  0.067  61.9  16.8  ‐0.43  ‐0.11  ‐0.28  0.30  ‐0.43  14.95  P26  30080632  0.053  0.053  36.8  17.7  ‐0.76  ‐0.26  ‐0.49  0.43  ‐0.79  7.83  P27  30275022  3.934  0.071  51.1  16.1  2.50  ‐0.06  ‐0.37  0.19  1.29  49.12  P28  30083882  0.398  0.095  52.5  11.1  ‐0.47  0.19  ‐0.36  ‐0.55  ‐0.63  10.95  P29  30106248  0.195  0.067  151.6  5.4  ‐0.64  ‐0.11  0.46  ‐1.40  ‐0.40  15.58  P30  30106756  0.231  0.070  131.0  11.1  ‐0.61  ‐0.08  0.29  ‐0.55  ‐0.31  17.41  P31  30122219  0.269  0.071  18.7  4.4  ‐0.58  ‐0.07  ‐0.64  ‐1.55  ‐1.18  0.00  P32  30061863  0.941  0.062  124.4  7.1  ‐0.02  ‐0.16  0.24  ‐1.15  ‐0.13  20.84  P33  30081072  1.054  0.083  107.5  9.7  0.08  0.07  0.10  ‐0.76  ‐0.04  22.73  P34  30106138  0.349  0.061  14.1  16.8  ‐0.52  ‐0.18  ‐0.68  0.30  ‐0.78  7.96  54 P35  30044558  0.148  0.067  85.5  25.2  ‐0.68  ‐0.10  ‐0.09  1.54  ‐0.16  20.28  P36  20079319  0.343  0.064  321.8  17.6  ‐0.52  ‐0.14  1.88  0.41  1.10  45.32  P37  30091479  3.101  0.052  170.6  8.9  1.80  ‐0.27  0.62  ‐0.88  1.29  49.15  P38  30381293  1.288  0.063  247.7  13.8  0.27  ‐0.15  1.26  ‐0.15  1.02  43.62  P39  20184422  0.111  0.064  17.3  23.6  ‐0.72  ‐0.14  ‐0.65  1.31  ‐0.65  10.57  P40  30134178  4.430  0.071  91.4  8.9  2.91  ‐0.06  ‐0.04  ‐0.88  1.55  54.22  P41  30108830  0.882  0.068  145.6  9.9  ‐0.07  ‐0.10  0.41  ‐0.73  0.07  24.92  P42  30080314  1.265  0.056  57.6  15.8  0.25  ‐0.22  ‐0.32  0.15  ‐0.09  21.66  P43  30077537  0.458  0.058  166.2  4.6  ‐0.42  ‐0.21  0.59  ‐1.51  ‐0.22  19.07  P44  30131496  0.924  0.070  62.4  6.5  ‐0.03  ‐0.07  ‐0.28  ‐1.24  ‐0.51  13.28  P45  30218272  5.823  0.052  79.5  12.0  4.09  ‐0.28  ‐0.14  ‐0.42  2.25  68.18  P46  30081190  0.490  0.067  29.4  13.7  ‐0.40  ‐0.11  ‐0.55  ‐0.17  ‐0.71  9.47  P47  30070762  0.632  0.067  137.4  16.1  ‐0.28  ‐0.11  0.35  0.19  0.10  25.42  P48  30123462  1.130  0.064  66.6  16.1  0.14  ‐0.15  ‐0.24  0.19  ‐0.08  21.89  P49  30122528  0.568  0.070  58.6  23.6  ‐0.33  ‐0.07  ‐0.31  1.31  ‐0.15  20.47  P50  30217972  1.868  0.054  67.5  12.0  0.76  ‐0.25  ‐0.24  ‐0.42  0.15  26.39  55 Annex 7. Road Transport Project SEI Calculations, by Region (with multidimensional index poverty rate) BEN_INV  JOBS_INV  SAVINGS_INV  POV_MULTI  0.442  0.224  0.839  0.224  Project  ID  Ben_inv  Jobs_inv  Savings_inv  Pov_multi  Ben_std  Jobs_std  Savings_std  Pov_std  SEI_STD  SEI  P1  30106685  1.334  0.041  27.5  22.9  0.31  ‐0.39  ‐0.57  0.16  ‐0.39  15.92  P2  30080831  0.276  0.077  97.8  26.3  ‐0.58  0.00  0.02  0.80  ‐0.06  21.32  P3  30096463  0.803  0.069  53.5  29.2  ‐0.13  ‐0.09  ‐0.35  1.35  ‐0.07  21.15  P4  30076931  0.295  0.042  90.0  20.1  ‐0.56  ‐0.38  ‐0.05  ‐0.37  ‐0.46  14.83  P5  30172722  0.326  0.064  55.9  19.2  ‐0.53  ‐0.14  ‐0.33  ‐0.54  ‐0.67  11.37  P6  30107176  0.160  0.067  71.7  29.2  ‐0.67  ‐0.11  ‐0.20  1.35  ‐0.19  19.24  P7  30124028  3.118  0.710  73.3  19.2  1.81  6.89  ‐0.19  ‐0.54  2.06  56.28  P8  30090914  0.064  0.057  55.4  36.6  ‐0.75  ‐0.22  ‐0.34  2.76  ‐0.05  21.54  P9  30083777  0.891  0.066  88.6  21.0  ‐0.06  ‐0.12  ‐0.06  ‐0.20  ‐0.15  19.91  P10  30071390  0.054  0.071  23.1  22.6  ‐0.76  ‐0.07  ‐0.61  0.11  ‐0.84  8.58  P11  30131878  1.314  0.063  39.1  23.2  0.30  ‐0.15  ‐0.47  0.22  ‐0.25  18.23  P12  30137944  0.168  0.070  34.8  19.2  ‐0.67  ‐0.08  ‐0.51  ‐0.54  ‐0.86  8.21  P13  30427024  0.041  0.069  121.8  17.2  ‐0.77  ‐0.09  0.22  ‐0.91  ‐0.39  16.00  P14  30104149  0.660  0.062  49.0  18.2  ‐0.25  ‐0.16  ‐0.39  ‐0.73  ‐0.64  11.84  P15  30073279  2.064  0.063  121.2  22.5  0.93  ‐0.16  0.21  0.09  0.57  31.72  P16  30091314  0.514  0.042  43.7  36.1  ‐0.38  ‐0.38  ‐0.43  2.67  ‐0.02  22.03  P17  30078400  0.509  0.052  58.0  22.8  ‐0.38  ‐0.27  ‐0.31  0.14  ‐0.46  14.74  P18  30070422  0.755  0.053  38.2  19.2  ‐0.17  ‐0.26  ‐0.48  ‐0.54  ‐0.66  11.51  P19  30393223  0.275  0.070  42.9  15.8  ‐0.58  ‐0.07  ‐0.44  ‐1.18  ‐0.91  7.45  P20  30135925  1.206  0.070  91.0  13.5  0.20  ‐0.07  ‐0.04  ‐1.62  ‐0.32  17.04  P21  30276122  0.205  0.066  36.4  29.2  ‐0.64  ‐0.12  ‐0.49  1.35  ‐0.42  15.44  P22  30064808  0.665  0.070  46.9  28.9  ‐0.25  ‐0.07  ‐0.41  1.30  ‐0.18  19.41  P23  30071804  0.490  0.060  53.3  21.2  ‐0.40  ‐0.19  ‐0.35  ‐0.16  ‐0.55  13.28  P24  30089927  0.587  0.067  824.7  17.2  ‐0.32  ‐0.11  6.06  ‐0.91  4.72  100.00  P25  30080601  0.445  0.067  61.9  22.6  ‐0.43  ‐0.11  ‐0.28  0.11  ‐0.43  15.27  P26  30080632  0.053  0.053  36.8  23.3  ‐0.76  ‐0.26  ‐0.49  0.24  ‐0.76  9.92  P27  30275022  3.934  0.071  51.1  23.2  2.50  ‐0.06  ‐0.37  0.22  0.83  35.95  P28  30083882  0.398  0.095  52.5  18.9  ‐0.47  0.19  ‐0.36  ‐0.60  ‐0.60  12.41  P29  30106248  0.195  0.067  151.6  17.2  ‐0.64  ‐0.11  0.46  ‐0.91  ‐0.12  20.28  P30  30106756  0.231  0.070  131.0  18.9  ‐0.61  ‐0.08  0.29  ‐0.60  ‐0.18  19.41  P31  30122219  0.269  0.071  18.7  9.1  ‐0.58  ‐0.07  ‐0.64  ‐2.45  ‐1.36  0.00  P32  30061863  0.941  0.062  124.4  20.1  ‐0.02  ‐0.16  0.24  ‐0.37  0.07  23.53  P33  30081072  1.054  0.083  107.5  21.0  0.08  0.07  0.10  ‐0.20  0.09  23.77  P34  30106138  0.349  0.061  14.1  22.6  ‐0.52  ‐0.18  ‐0.68  0.11  ‐0.81  8.95  P35  30044558  0.148  0.067  85.5  25.6  ‐0.68  ‐0.10  ‐0.09  0.67  ‐0.25  18.25  P36  20079319  0.343  0.064  321.8  19.2  ‐0.52  ‐0.14  1.88  ‐0.54  1.20  42.00  P37  30091479  3.101  0.052  170.6  18.1  1.80  ‐0.27  0.62  ‐0.75  1.09  40.22  P38  30381293  1.288  0.063  247.7  23.9  0.27  ‐0.15  1.26  0.35  1.23  42.53  P39  20184422  0.111  0.064  17.3  29.2  ‐0.72  ‐0.14  ‐0.65  1.35  ‐0.59  12.60  56 P40  30134178  4.430  0.071  91.4  18.1  2.91  ‐0.06  ‐0.04  ‐0.75  1.07  40.01  P41  30108830  0.882  0.068  145.6  15.5  ‐0.07  ‐0.10  0.41  ‐1.23  0.02  22.66  P42  30080314  1.265  0.056  57.6  25.4  0.25  ‐0.22  ‐0.32  0.63  ‐0.06  21.29  P43  30077537  0.458  0.058  166.2  27.8  ‐0.42  ‐0.21  0.59  1.09  0.50  30.59  P44  30131496  0.924  0.070  62.4  16.9  ‐0.03  ‐0.07  ‐0.28  ‐0.97  ‐0.48  14.42  P45  30218272  5.823  0.052  79.5  18.2  4.09  ‐0.28  ‐0.14  ‐0.73  1.47  46.47  P46  30081190  0.490  0.067  29.4  23.0  ‐0.40  ‐0.11  ‐0.55  0.18  ‐0.62  12.08  P47  30070762  0.632  0.067  137.4  23.2  ‐0.28  ‐0.11  0.35  0.22  0.19  25.51  P48  30123462  1.130  0.064  66.6  23.2  0.14  ‐0.15  ‐0.24  0.22  ‐0.13  20.27  P49  30122528  0.568  0.070  58.6  29.2  ‐0.33  ‐0.07  ‐0.31  1.35  ‐0.12  20.37  P50  30217972  1.868  0.054  67.5  18.2  0.76  ‐0.25  ‐0.24  ‐0.73  ‐0.08  21.03  57 Annex 8. Road Transport Project FEI Calculations, by Region Project  ID  IVAN  IVAN_std  FEI_STD  FEI  P1  30106685  1.718  1.62  1.62  53.66  P2  30080831  0.568  ‐0.10  ‐0.10  17.49  P3  30096463  0.060  ‐0.86  ‐0.86  1.50  P4  30076931  0.634  0.00  0.00  19.56  P5  30172722  0.086  ‐0.82  ‐0.82  2.34  P6  30107176  0.591  ‐0.07  ‐0.07  18.21  P7  30124028  0.303  ‐0.50  ‐0.50  9.16  P8  30090914  0.054  ‐0.87  ‐0.87  1.32  P9  30083777  0.489  ‐0.22  ‐0.22  15.01  P10  30071390  0.115  ‐0.78  ‐0.78  3.24  P11  30131878  1.152  0.77  0.77  35.87  P12  30137944  0.230  ‐0.61  ‐0.61  6.86  P13  30427024  1.353  1.08  1.08  42.17  P14  30104149  0.055  ‐0.87  ‐0.87  1.36  P15  30073279  0.758  0.18  0.18  23.46  P16  30091314  0.107  ‐0.79  ‐0.79  3.00  P17  30078400  0.184  ‐0.68  ‐0.68  5.42  P18  30070422  0.261  ‐0.56  ‐0.56  7.84  P19  30393223  0.474  ‐0.24  ‐0.24  14.54  P20  30135925  1.385  1.12  1.12  43.19  P21  30276122  0.012  ‐0.93  ‐0.93  0.00  P22  30064808  0.125  ‐0.77  ‐0.77  3.55  P23  30071804  0.326  ‐0.46  ‐0.46  9.86  P24  30089927  1.698  1.59  1.59  53.05  P25  30080601  0.105  ‐0.80  ‐0.80  2.91  P26  30080632  0.081  ‐0.83  ‐0.83  2.18  P27  30275022  0.084  ‐0.83  ‐0.83  2.26  P28  30083882  0.716  0.12  0.12  22.16  P29  30106248  1.254  0.93  0.93  39.07  P30  30106756  0.953  0.48  0.48  29.62  P31  30122219  0.130  ‐0.76  ‐0.76  3.70  P32  30061863  0.528  ‐0.16  ‐0.16  16.24  P33  30081072  0.654  0.03  0.03  20.20  P34  30106138  0.830  0.29  0.29  25.74  P35  30044558  0.559  ‐0.12  ‐0.12  17.19  P36  20079319  3.191  3.83  3.83  100.00  P37  30091479  1.574  1.41  1.41  49.16  P38  30381293  2.532  2.84  2.84  79.29  P39  20184422  0.071  ‐0.85  ‐0.85  1.84  P40  30134178  0.539  ‐0.14  ‐0.14  16.58  P41  30108830  0.943  0.46  0.46  29.27  P42  30080314  0.025  ‐0.92  ‐0.92  0.40  P43  30077537  1.291  0.98  0.98  40.23  P44  30131496  0.443  ‐0.29  ‐0.29  13.57  P45  30218272  0.533  ‐0.15  ‐0.15  16.40  P46  30081190  0.023  ‐0.92  ‐0.92  0.35  P47  30070762  1.044  0.61  0.61  32.47  P48  30123462  0.261  ‐0.56  ‐0.56  7.84  P49  30122528  0.415  ‐0.33  ‐0.33  12.68  P50  30217972  0.250  ‐0.58  ‐0.58  7.49  58