92724 Policy Note Promoting equity and managing conflict in development November 2014 Sub-National Spending in Timor-Leste: Lessons from Experience © 2014 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW Washington DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org Disclaimer This work is a product of the staff of The World Bank with external contributions. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions The material in this work is subject to copyright. Because The World Bank encourages dissemination of its knowledge, this work may be reproduced, in whole or in part, for noncommercial purposes as long as full attribution to this work is given. Any queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2422; e-mail: pubrights@worldbank.org. Contact Details Justice for the Poor Governance and Inclusive Institutions Governance Global Practice World Bank 1818 H Street, NW Washington, DC 20433 USA Email: j4p@worldbank.org Policy Note November 2014 Sub-National Spending in Timor-Leste: Lessons from Experience The World Bank Acknowledgements This policy note was drafted by Sakuntala Akmeemana (Task Team Leader, World Bank) and Douglas Porter (Adviser, World Bank), with substantial inputs from Roger Shotton (Consultant and Study Team Leader), Daniel Wilde and Cristino Gusmao (Ministry of Finance, Timor-Leste) and Edio Guterres (World Bank). It is based on the findings of a study undertaken by the World Bank in collaboration with the Ministry of Finance and Ministry of State Administration. Significant contributions were also made by Joanico Pinto, Armindo Santos and Nikunj Soni (Ministry of Finance), Catherine Anderson (World Bank) and a consultant team comprising Edward Rees (contractor development and political economy) and Mark Moran (engineering process and technical quality). Christina Biebesheimer (Chief Counsel) and Luis Constantino (Country Manager, Timor-Leste) provided overall guidance and supervision. Peer review comments were provided by Steven Ndegwa, Joel Turkewitz and Vincent Ashcroft. Comments were also received from Paul Keogh, Tobias Haque, David Hook and Ingo Wiederhofer at the concept stage. Hans Beck provided guidance and support throughout the development and execution of the study. Kathleen Lonergan and Rushi Fernando provided proofreading and formatting assistance, and Jihane El-Khoury Roederer and Dolie Schein provided graphic design and typesetting. Finally, the Minister of Finance, Hon. Emilia Pires, has provided strong support and guidance for the study. She has been instrumental to its successful completion and in catalyzing follow-up work. ii Sub-national Spending in Timor-Leste List of Acronyms ADN Agência de Desenvolvimento Nacional, or National Development Agency AECCOP Associação Empresários Construção Civil e Obras Públicas, or Association of Civil Construction and Public Works Entrepreneurs AMP Aliança Maioria de Parlamentar, or Alliance of the Parliamentary Majority IADE Instituto de Apaoiu ou Dezenvolvimentu Emprezarial, or Institute of Support for Entrepreneurship Development IFMIS Integrated Financial Management Information System PDD Programa de Dezenvolvimentu Desentralizadu, or Decentralized Development Program PDID Planeamento de Dezenvolvimentu Integradu Distritál, or the Decree Law for Integrated District Development Planning PDL Programa Dezenvolvimentu Lokal, or Local Development Program PFM Public Financial Management PNDS Programa Nacional do Dezenvolvimentu do Suco, or National Suco Development Program PR Pakote Referendum, or Referendum Package MAE Ministério da Administração Estatal, or Ministry of State Administration UNCDF UN Capital Development Fund Policy Note | november 2014 iii Summary brief Background and Purpose This Policy Note distils the findings and recommendations of a Study, jointly undertaken by the World Bank and the Government of Timor-Leste, that evaluates the mechanisms used by the Government to manage sub-national spending in recent years: i) the Local Development Program (Programa Dezenvolvimentu Lokal, PDL), underway since 2004; and ii) the Referendum Package (Pakote Referendum, PR), initiated in 2009, and progressively refined as the Decentralized Development Program (Programa de Dezenvolvimentu Desentralizadu, PDD) since 2010. The Study assesses how these programs have performed in respect of a number of key government policy priorities and draws conclusions and lessons for the future. It adopts a dynamic approach, so as to adequately acknowledge and account for how shifts in the government’s policy priorities over time, in response to the changing context, were reflected in the institutional arrangements that govern sub-national spending. PDL and PR/PDD have in common the delivery of small scale infrastructure at the sub-national level. However, each had its genesis in different historical contexts, were motivated by distinct policy objectives and evolved in very different ways. During the immediate post-independence period, PDL was initiated in response to an anticipated law on decentralization, to inform the systems and procedures and build the capacities needed to manage sub-national spending and local governance. PR was born in the aftermath of conflict, in a context where Government needed to deliver an ‘independence dividend’ to its citizenry. It aimed to stimulate rural development by (i) creating a local entrepreneurial class through a modality that funded contractor capitalization and capacity, and allowed Timorese companies to win Government contracts; (ii) generating employment; and (iii) delivering high quality infrastructure. These policy priorities were often in tension, and the relative emphasis given to each has shifted over time, as PR was refined through iterations of PDD. When it was first introduced in late 2009, PR’s budget of $70 million dwarfed PDL’s budget ($1.66 million). The early emphasis was distributive, on creating a local contractor class, but started to shift by the 2012 budget to a greater emphasis on technical quality. More recently, issues of poverty impact, equity and employment generation have become increasingly prominent. The Study was prompted by the Decree Law for Integrated District Development Planning (Planeamento de Dezenvolvimentu Integradu Distritál, PDID), promulgated in January 2012. While the Study evaluated two sets of programs, in the spirit of ‘learning by doing’, its primary purpose was forward looking—to contribute to ongoing efforts by the Government to define systems and procedures for district development (including responsibilities for local planning, finance management and procurement) as it implements the Decree. At the same time, the Study aimed to inform public dialogue about how resources are being distributed and used at the local level, the quality of assets being created and services delivered, and the patterns of commercial, social and political opportunities that are emerging. The Study is also relevant to a broader discussion initiated by the Minister of Finance and the g7+ Group of Fragile States about innovative approaches to public financial management and procurement in fragile and conflict-affected settings. The innovative mechanisms that the Government created for sub-national spending in iv Sub-national Spending in Timor-Leste the aftermath of the 2006 crisis, beginning with the PR, provide instructive lessons about the trade-offs needed to balance competing technical, social and political priorities, and of how these have to be adjusted as needs change over time. The Study found that: Both PDL and PDD show good-to-excellent rates of budget execution up until fiscal year (FY) 2012 and perform considerably better than those recorded before 2009—or even now—by most line ministries for spending on infrastructure. Project completion rates are generally good for both mechanisms, but a significant minority of PDD projects remains incomplete. These programs have therefore had a positive impact on the challenge of capital budget execution in Timor-Leste. Spending equity ($ per capita) under PDD varies greatly across districts, with the most favored districts receiving between three to seven times more than the least favored. Ermera has consistently received a low allocation, and Manatuto a high allocation. Spending patterns broadly (but not entirely) reflect national sector priorities. The large majority of spending is consistent with the public services (education, water, health and agriculture) prioritized in the Government’s Strategic Development Plan. The quality of PDD investments is mixed: of an illustrative sample of 13 that this Study comprehensively examined, just under half of projects (six) are of acceptable standard and of benefit to users, three exhibit varying degrees of design, construction quality and usability problems and four are not usable for the intended purpose without additional investments. The results suggest that a follow-up study of a larger sample would be useful. Such a study of around 100 investments is planned for 2014. Under PDD and PDL, the share of total investment costs going to unskilled and skilled labor payments are around 6.5% and 3.5% respectively. The employment impact depends greatly on the type of infrastructure or investment scheme. For instance, buildings generally employ much less labor than investments such as irrigation schemes and erosion protection walls. The programs provided seasonal employment to an estimated 2.5-10% of the sub-district workforce. Significantly, the benefits have been mainly local, and shared widely, predominantly amongst the rural underemployed. The impact is likely to be substantial in the poorest quintile of households, for whom the additional cash injection is a significant proportion of household income. Extrapolations for FY 2012 suggest that these programs have created some 1.2 million unskilled and 220,000 skilled workdays respectively. At sub-district level, this could provide 165 persons with 100 days of employment each or 740 persons with 25 days each. Compared to the large special purpose public works employment programs in Asia, the overall impact of PDL and PDD on employment appears modest, but it needs to be emphasized that PDL and PR/PDD are not labor creation programs, but rather have multiple policy objectives in addition to employment creation. The number of registered local contractors has expanded greatly, in the order of a three-fold increase since 2009, and a large number of these contractors have each year been awarded a PDD contract. However, the fact that contractors cannot win more than one project per year creates few opportunities and incentives for contractors to consolidate their businesses and diversify over time. Policy Note | november 2014 v The governance dividends of sub-national spending have been mixed. On the one hand, these programs have enabled large numbers of Timorese to participate in allocation of public resources and the production of assets largely in accordance with their priorities. While sub-national spending is best seen as part of a suite of measures the Government introduced (including for instance, cash transfer and pension programs) to support its political, stability and social objectives, the way in which these programs have enabled government to target important constituencies needed to build stable political settlements should not be underestimated. The PR/ PDD programs have helped to distribute wealth to rural areas, benefiting a number of constituencies to whom the Government wanted to deliver a share of the country’s new found oil wealth, including rural labor, domestic contractors and veterans. By distributing public wealth to previously disenfranchised groups, the PR/PDD programs have contributed to peace and stability. On the other hand, nine out of 13 PDD projects examined closely in this Study gave rise to some form of disputation about land, procurement, labor or contractor performance, whereas only one of the six PDL projects gave rise to a dispute (contractor-related). More deliberate attention to local consultation during planning and more intense supervision could potentially ameliorate these problems. There is little evidence that the investment assets themselves or the employment opportunities funded under either PDL or PDD are being captured by local elites. In Conclusion: PR marked the beginning of a bold but risky initiative to address a series of economic, social and political priorities. Four years on, while the results are uneven, the initial objectives of the PR and PDD have been largely achieved. After experimenting with a range of innovative arrangements to deliver public spending, the Government’s decision to ‘regularize’ the system through the PDID Decree Law 2012 is timely. The report’s recommendations are made in the spirit of this Decree. There is no evidence to support the view that sub-national spending mechanisms, compared to centralized spending by line ministries, are more vulnerable to “weak local capacity”. The relatively high quality infrastructure consistently produced by PDL spending, and produced in a significant percentage of cases under PDD, suggest that where there is poor performance, this is not caused by inherent weaknesses in local administration or contractors as compared to capacities in the central line ministries. The reasons are more complex. They lie in the need to frequently modify systems and procedures, in response to changing priorities and lessons learned. Several aspects of poor performance can be mitigated in ways anticipated by the PDID Decree, including the clarification of assignments of responsibilities (in relation, for instance, to certification and supervision) and by more consistent attention to public communication and accountability arrangements. The Study demonstrates that now is an appropriate time to build on the successes and lessons learned through these two programs. The following priority actions are recommended on the basis that they: i) are likely to have a significant impact on a wider range of challenges faced by these programs, and ii) appear to be realistic in light of the commitments reflected in the PDID Decree, and a prudent assessment of their political and administrative feasibility. vi Sub-national Spending in Timor-Leste Priority Actions 1.  Funds Allocation, Planning and Budgeting: Instituting predictable, formula-based budget allocations to districts would pay multiple dividends. Allocating funds to districts by formulae on an annual basis is a low-cost opportunity that would promote transparency and some medium term stability in the national pool of resources allocated to sub-national spending. Formula- based allocations could improve geographic equity and ensure that the allocation of funds reflects relative district poverty and need. Announcing district budget envelopes at the start of each annual cycle would improve the quality of planning and promote a focus on delivery of high quality assets. The introduction of simple ‘district performance measures’ as part of district allocation formulae could reward good performance on planning, budgeting, procurement and investment management, and perhaps also on local employment. 2.  Procurement Strategy: Introducing much simpler contractor pre-qualification procedures and competitive tendering will improve contractor performance and create significant cost savings. With a national pool of experienced contractors, created in large part by the PR/PDD and PDL programs, now may be an appropriate time to institute: (i) streamlined contractor pre-qualification procedures, centrally managed by the National Development Agency (Agência de Desenvolvimento Nacional, ADN) and the Ministry of Public Works, perhaps in co-operation with the Institute of Support for Entrepreneurship Development (Instituto de Apaoiu ou Dezenvolvimentu Emprezarial, IADE), and (ii) simple competitive tendering, managed by district authorities, with performance links to district funding to promote their compliance with procurement rules. 3.  Investment Supervision and Assignment of Responsibilities: The quality of project design, supervision and certification for payments could be enhanced by clarifying the division of responsibility between ADN and other line ministries, and by investing in adequate technical personnel and local travel budgets. ADN’s role in certification was intended to provide an independent check on quality and ensure payments were only authorized for work properly completed, but in practice there is some evidence that it has on occasion resulted in premature certification by ADN engineers. There is merit in engaging ADN engineers assigned to districts in the preliminary and also final project design and costing phase, rather than have ADN Dili review final costs alone, post facto. This would ensure that clear accountability for certification rests with district engineers from line departments responsible for the sector in which the assets are being created. The general failure to provide adequate technical personnel and local travel budgets is a false economy. More adequate resourcing will improve design and supervision, and deliver significant cost savings. The ‘cost’ of inadequate supervision amounts to around 25% of the value of investments, or some $15 million each year. 4.  District Level Payments: Establish district treasures as soon as practicable, when enabled by the functionality of the Ministry of Finance’s Integrated Financial Management Information System (IFMIS). The comparative evidence suggests that deconcentrating the treasury function (provided it is linked with IFMIS) is likely to expedite payments and impact positively on contractor performance. Reverting to district payments for contractors will help to deal with the kind of payment bottleneck experienced in FY 2012, arising both from delays in certification reporting to Dili and in central Treasury payments. Policy Note | november 2014 vii Policy Note 1.  Background and Context The last decade has seen Timorese, for the first time in the polity’s history, driving decisions about the allocation of public resources to pursue a variety of public policy and political objectives. The initial period after independence in 2002 was marked by little in the way of discretionary resources with which Government could realistically achieve policy aims and an exclusivist political dynamic. Unrealistic expectations had been generated by the prospect of independence and remained unrealized, with several social and political fractures emerging that were manifested during the widespread violence of 2006.1 Driven by a sharp rise in the revenues deposited in the Petroleum Fund since 2007 and their distribution through public spending in the last half decade, Timor- Leste has been undergoing a significant re-ordering of its political economy. A crucial aspect of Timor-Leste’s economic performance and political stability in the aftermath of the 2006 crisis has been the way the Government has managed a five-fold increase in spending, and an even greater increase in capital spending. Significantly, this was largely financed through sharply rising petroleum revenues and utilized government systems to plan and budget, spend and account for results. Annual capital and development spending peaked around $555 million in 2011. This is in marked contrast to the average of $55 million in the years prior to 2007, which was mostly donor-funded and managed under donor procedures. The sub- Figure 1.  Major Infrastructure Spending FY 2004 – FY 2013 In $ millions 1000 900 800 700 600 Budget 500 400 Actual 300 200 100 0 Budget Year 04 11 05 06 07 08 09 10 12 FY FY FY FY FY FY FY FY FY 1 The 2006 crisis was sparked by a group of petitioning soldiers (called the “petitioners”), predominantly from the country’s western districts, who charged their army superiors, who were predominantly easterners drawn from the resistance armed forces, with discrimination in the national army. The subsequent sacking of nearly 600 petitioners by the Government led to widespread unrest and violence. See‘Timor Leste’s Recovery from the 2006 Crisis: Some Lessons’, Background paper for World Development Report 2011: Violence, Conflict and Development. November 2010. http://wdr2011.worldbank.org/input-papers. Policy Note | november 2014 1 national spending programs introduced by the Referendum Package (Pakote Referendum PR) initiative of 2009 formed part of the Government’s efforts to both drive economic activity, through the creation of private sector capacity and jobs, and address social and political priorities in the aftermath of conflict. Achieving these various aims required the Government to search for a more optimal balance between conventional public financial management and procurement (which favors multiple ex ante controls on spending, and focuses on technical and financial measures of efficiency and effectiveness as well as accountability) on the one hand and political and economic pragmatism on the other (responding to the realities of institutional fragility, the need to stimulate the domestic private sector and to prevent the recurrence of violent conflict). When the newly elected coalition known as the Alliance of the Parliamentary Majority (Aliança Maioria de Parlamentar AMP)2 came into office in 2007, it was aware that its highest order challenge was to respond to a sense of injustice and disenfranchisement amongst key political, social and geographic constituencies. After mobilizing these diverse groups, the new government needed to deliver on its election promises—among others, to the veterans of the resistance struggle,3 to the petitioners4 and to people internally displaced as a result of the 2006 crisis. Public pressure was also mounting for an oil and independence dividend. Political commentary had become livelier, feeding off greater access to information through an active Parliament. The Government publicly announced a path from ‘stabilization’ to ‘recovery’ and the pressing tasks to achieve this path were to consolidate the governing coalition and maintain public order and security. At the same time, the growth in the volume of oil sector revenues far outstripped projections, including projections made as late as FY 2006. Petroleum revenue reached 340% of non-oil GDP in 2007 and almost 500% in 2008. But while 2007 GDP growth reached 8%, the private sector was faring poorly and job creation BOX 1 The World Development Report on Conflict, Security and Development 2011 (WDR) provides a useful lens through which Timor-Leste’s experience over the last half decade may be viewed. The WDR’s survey of experience and wider scholarship suggests that successful transitions out of conflict have involved Govern- ments sending “credible signs of change” to the population. These involve indications of a distinct break from the past, with a focus on three areas: - security (of person and property, but also broader political stability and social order), - justice (in terms of relative equity in access to public wealth and the proceeds of natural assets, without regard to geography or ethnicity); and - jobs (a reasonable expectation of an existence with dignity and the opportunity to accumulate wealth and assets). These credible signs in turn buy Governments the time and space for longer-term institutional transforma- tion, that will enable more inclusive political coalitions to be formed, and for the political dispensation to become popularly regarded as effective, socially and politically legitimate and inclusive. 2 FRETILIN, which had been in government since independence, found itself short of a majority in Parliament and unable to secure a coalition. CNRT, PD, and ASDT-PSD formed the Alliance of the Parliamentary Majority (AMP), which was later joined by UNDERTIM – the National Democratic Union of Timorese Resistance. The durability of this coalition required considerable political and public spending compromises to main- tain, many of which were seen in 2008. 3 See further International Crisis Group, Timor-Leste’s Veterans: An Unfinished Struggle, November 2011. 4 See footnote 2 2 Sub-national Spending in Timor-Leste was stagnant. Anecdotal evidence suggested that poverty was increasing, which the World Bank’s poverty assessment of 2008 subsequently confirmed.5 An additional 16,000 youth were entering the workforce each year with the expectation of a reasonable economic future. Meanwhile, six months into office, the external environment was rapidly changing: the onset of the global commodity crisis in early 2008 resulted in price speculation for essential commodities and a contraction in private spending. Constraints and Responses In important respects, Timor-Leste was unusual for a country that had only recently emerged from occupation and an international transitional administration. The country had established a sound architecture to capture and govern petroleum sector revenues, inter alia through the Petroleum Fund and a public financial management (PFM) system, including regulations and procedures to plan, budget, spend, and account. At the same time, it was apparent to the Government that the existing PFM system constrained the formulation and execution of projects to such an extent that it was difficult to quickly address public expectations about development, declining incomes, stagnant employment and rising poverty. Thus, the Government was concerned that if new programs to boost development were not implemented, the country might relapse into conflict. Faced with these pressures, the Government took steps to achieve far higher rates of spending than had been achieved in the past, and introduced a suite of initiatives.6 The PR, announced in 2009, was one of these measures, funded through transferring $70 million of funds from major Ministry of Infrastructure power projects whose implementation had stalled. In terms of the framework from the World Development Report highlighted in Box 1, multiple signals were being sent to the Timorese people in relation to the distribution of oil wealth through these initiatives. The underlying ‘strategic’ objective of the PR was to provide wealth to previously disenfranchised groups (including former members of the resistance who lived in rural areas) in order to develop the country and reduce the likelihood of conflict. The economic security provided to these groups would also allow them to actively participate in the democratic process. There are clear indications that the PR modality has been fine-tuned to context through various iterations of the Decentralized Development Program (Programa de Dezenvolvimentu Desentralizadu, PDD): while the de facto rules of the game have been rapidly changing, the de jure rules have shown a clear progression in the direction of greater institutionalization and accountability. 2.  Sub-National Spending Mechanisms at a Glance Until 2009, there was only one modality for decentralized planning, budgeting and capital investment spending in Timor-Leste. The Local Development Program (Programa Dezenvolvimentu Lokal, PDL) was initiated in the context of an anticipated law on decentralization in the immediate post-independence period. It began in 2004 as a pilot program of district block grants, run by the then Ministry of State Administration (Ministério 5 Ministry of Finance, the National Directorate of Statistics and the World Bank, Timor-Leste: Poverty in a Young Nation, (November 2008). 6 The Government opened many new spending lines over the next two years; greater spending powers were delegated to line ministries and cash transfer and pension programs were crafted; spending increased goods and services, including subsidies on essential commodities. Further, the Government introduced an expanded social protection scheme, including universal pensions for the elderly and people with disabilities, conditional cash transfers for school-going children in vulnerable households, support to veterans and survivor families, and ad hoc responses to internal displacement and natural disasters. Policy Note | november 2014 3 da Administração Estatal, MAE),7 initiated in Bobonaro District and subsequently scaled up to all districts.8 The program aimed to test different models for decentralization and building capabilities for local governance and political representation, starting as a donor program and having been funded by Government since 2008.9 Initially it was envisaged that PDL’s administrative and fiscal architecture could be rolled out after enactment of a decentralization law. PDL’s approach comprises a system of annual participatory planning and investment budgeting, whereby proposals are submitted by sucos, sub-districts and districts, and prioritized at sub-district and district levels. Approved proposals are implemented and monitored entirely by sub-district development committees and district representative assemblies (foreshadowing the role of possible future municipalities). Funds are allocated as a formula-based block grant to districts on an annual basis, to fund small scale public works projects implemented under tender to private contractors.10 By way of contrast, the PR and subsequent successive rounds of the PDD had their genesis in the aftermath of a period of violent conflict, and attempted to balance competing technical, social and political priorities. Responding to pressure to deliver an independence dividend to those in rural areas and support the nascent business community, the Government introduced the PR in 2009. The package transferred $70 million of funds from the Ministry of Infrastructure budget to provide financing for small to medium sized infrastructure projects. It was presented to the National Parliament in October 2009 by Prime Minister Xanana Gusmao “to fast track rural development and empower the private sector.”11 This initiative was regarded as being controversial by some observers, with critics arguing that it had not been authorized by Parliament and lacked appropriate regulatory controls. Figure 2.  PR/PDD & PDL Budgets ($ millions) 80.00 1.66 70.00 60.00 6.30 50.00 3.50 40.00 PDL 2.78 70.00 51.75 30.00 44.30 PDD 20.00 31.86 10.00 0 2009 2010 2011 2012 7 It was supported by the UN Capital Development Fund (UNCDF), through funding from the governments of Ireland and Norway. 8 This was not Timor-Leste’s first experience in decentralized spending. A prior experience in community-driven development had sought to “strength- en local-level social capital to build institutions that reduce poverty and support inclusive patterns of growth”. As distinct from the citizen-state engagement attempted through the PDL, the Community Empowerment Program (CEP) favoured community allocation decisions, and was distinct from any government administrative artifice: World Bank Independent Evaluation Group (June 2006). 9 As one part of MAE’s Local Governance Support Programme. 10 “Government in Timor-Leste allocates $3.5 million to the Local Development Fund,” UNCDF, March 7, 2011, http://www.uncdf.org/node/325 11 RDTL Press Release, 29 October 2009 4 Sub-national Spending in Timor-Leste With a budget of $70 million, the PR dwarfed the PDL budget, and was in the order of the entire national budget of FY 2005 ($86 million). Local branches of an entrepreneurs’ body, the Association of Civil Construction and Public Works Entrepreneurs (Associação Empresários Construção Civil e Obras Públicas AECCOP) took responsibility for the allocation of funds, identification of some 700 small and medium projects, selection of contractors and general oversight. At the time, Government considered that this approach offered the means to more rapidly execute investments and stimulate the local private sector, and to operate at a scale greater than could be achieved through the PDL. While the PR and PDL funded similar types of infrastructure, the Package was distinguished in two ways and had different policy objectives. First, PDL’s multi-step planning and budgeting process was removed. Second, the Package allowed for direct procurement of capital works without a competitive bidding process. Project identification and contracting functions were assigned to AECCOP,12 with a sub-structure at the district level, headed by a district coordinator. AECCOP identified the projects, selected the local civil contractors, awarded the projects, and supervised all project implementation. Proponents of this modality considered it to have several advantages over the PDL arrangements: it would reduce time and costs in terms of planning, budgeting and contracting procedures. It could also provide the means of activating the Government’s networks of district officials and AECCOP-endorsed contractors to rapidly implement and account for spending decisions. While PDL had a total annual budget at the time of $3 million, the PR and its subsequent variations aimed at near $150 million in disbursements over the next three years. Figure 3.  PDD & PDL Budget Compared to Public Overall & Capital Spending ($ millions) 1800 1674.1 1600 1306 1400 1200 958.2 1000 All 758.7 800 717.6 603.6 Capital 600 PDD+PDL 400 209.3 253.7 200 71.7 47.8 58 34.6 0 2009 2010 2011 2012 12 Chaired by the current head of the Timor-Leste Chamber of Commerce and Industry. Policy Note | november 2014 5 The PR was succeeded by the PDD in 2010. Government created new agencies to manage the PDD at national and sub-national levels and introduced several important refinements to how projects are planned, and contractors selected and supervised.13 In 2011, the PDD was split into PDD I and PDD II, the latter covering individual infrastructure projects between $150,000 to $250,000. In 2012, the upper limits of PDD I and PDD II increased to $250,000 and $500,000 respectively. PDD had points in common with PDL, although the planning process was shorter and less participatory, district allocations were not determined by formula, and procurement was administratively determined, rather than market based. While PDL had been increasing its proportion of the sub-national spend in recent years, combined sub-national spending had not kept pace as a share of total national outlays on capital or total spending. Whilst the PDL arrangements have been refined over time, there have been far more substantial and frequent changes in how PDD operates. For instance, rules for procurement have been modified, and it moved from a mix of centralized planning and decentralized implementation in 2010 and 2011, to some degree of decentralized planning but a greater degree of central control over implementation in 2012. In the latter regard, the role assigned to the National Development Agency (Agência de Desenvolvimento Nacional, ADN), created in 2011 under the office of the Prime Minister, is of special importance. The ADN has both overall responsibility for PDD II investments, but now also design, supervision and certification responsibilities for PDD I investments, even though the budget is under MAE (which is the lead agency on deciding which projects are included). To a large extent, the frequent shift in the rules have reflected Government’s commitment to innovation and learning by doing. At the same time, there is significant evidence that these procedures are becoming institutionalised in routine practices. Over 2010 and 2011, the existence of two different sub-national spending mechanisms (PDL and PDD) became an increasing source of confusion, especially for district administrations charged with handling both. As a response, the Government enacted the Decree Law for Integrated District Development Planning (Planeamento de Dezenvolvimentu Integradu Distritál, PDID) in January 2012, so that all sub-national spending would be managed under a harmonized set of guidelines. This includes the planned Suco Development Fund under the National Suco Development Program (Programa Nacional do Dezenvolvimentu do Suco, PNDS). PDID “defines the rules and regulations applicable on competency, planning, implementation and financing of State projects at the district and sub-district levels.”14 PDID also establishes the preparation process for an annual plan15 that seeks to reconcile local planning,16 PDL and PDD with the Government’s Strategic Development Program.17 The Decree’s ambit is to create a hybrid system for sub-national capital planning and investment management, engaging both PDL and the PDD modalities, fusing the participatory elements of the PDL with the fast-track elements of PDD. 13 The PDD was introduced in the 2010 budget. In 2011, the PDD was split into PDD I and PDD II, covering individual infrastructure projects up to $250,000 and $500,000 respectively. Currently, the MAE administers PDD I. PDD II is run by the newly created Agencia Nacional do Dezenvolvim- ento or ADN (National Development Agency), under the office of the Prime Minister. Projects are now planned by the Kommisaun Dezenvolvementu Distrital (District Development Commission) composed of the District Administrator, Sub-District Administrators, Representatives of Line Min- istries at the District, and Suco Chiefs. Organs of the District Development Commission are responsible to verify, evaluate and supervise contractor performance. Efforts have been made to link contractor eligibility with past performance, but it is proving difficult to monitor the many thousands of companies now engaged in or competing for PDD contracts. 14 PDID Decree Law, Art. 1 (1) 15 The Plano Investimento Distrital (PID). 16 The Plano de Dezemvolvemento Suco (PDS) or Suco Development Plan. 17 Preamble of the PDID Decree Law. 6 Sub-national Spending in Timor-Leste The MAE has issued detailed rules to govern planning for FY 2013.18 MAE has also mandated special task teams to define corresponding rules for procurement and finance management, and to define how responsibilities will be assigned to national, district and community authorities. Figure 4.  Sub-National Spending in Timor - A Chronology PDL funded PR succeeded PDID Decree by government by PDD Promulgated 2004 2008 2009 2010 2011 2012 2013 PDL pilot begins PR introduced PDD split PNDS pilot begins (donor funding) into PDDI and PDDII PDL (Programa Dezenvolvimentu Lokal) Local Development Program PDD (Programa de Dezenvolvimentu Desentralizadu) Decentralized Development Program PR (Pakote Referendum) Referendum Package PDID (Planeamento de Dezenvolvimentu Integradu Distritál) Decree Law for Integrated District Development Planning PNDS (Programa Nacional do Dezenvolvimentu do Suco) National Suco Development Program 18 MAE had issued an initial set of such rules in 2011, to govern FY 2012 planning, but the delayed approval of the PDID Decree Law prevented full implementation. This had a knock-on effect in delaying FY 2012 implementation of projects proposed by Districts, due to the need to complete designs & Bill of Quantities which could not be completed in 2011. Policy Note | november 2014 7 3.  Purpose of the Study This Study, jointly undertaken by the World Bank and the Government of Timor-Leste, evaluates PDL, PR and the subsequent successive rounds of the PDD.19 This review assesses how these programs have performed in respect of five key government policy priorities: delivering infrastructure and services effectively and equitably; implementing the priorities of the national Strategic Development Plan; creating jobs; stimulating local business; achieving social inclusion and political legitimacy and building capabilities for local governance and the representation of citizen interests. It then draws conclusions and lessons for the future. Since policy priorities have shifted rapidly over time, as have the institutional arrangements that govern sub-national spending, this study has adopted a dynamic approach which recognizes changes in context and the shift in relevant emphasis of the government’s spending priorities. The early emphasis was distributive, on creating a local contractor class, but started to shift by the 2012 budget to a greater emphasis on technical quality. More recently, issues of poverty impact, equity and employment generation have had increasing prominence. The study thus has two purposes. First, the study aims to contribute to the Government’s efforts to improve and further define roles and responsibilities for local planning, finance management and procurement as it implements the January 2012 Decree Law. Thus the study is forward looking; it aims to inform public dialogue about how resources are being distributed and used at the local level, the quality of assets being created and services delivered, and the patterns of commercial, social and political opportunities that are emerging. The study is also relevant to ongoing discussion initiated by the Minister of Finance and the g7+ Group of Fragile States about innovative approaches to public finance and procurement in fragile and conflict-affected settings. The innovative mechanisms that the Government created for sub-national spending in the aftermath of the 2006 crisis, beginning with the PR, provide instructive lessons about the trade-offs needed to balance competing technical, social and political priorities, and of how these need to be adjusted over time. 4.  Study Methodology A multi-disciplinary study team, comprised of World Bank staff and consultants and Ministry of Finance officials, adopted a “mixed methods” approach. From August to November 2012, the team analyzed available databases, visited the districts of Baucau, Ermera, Bobonaro and Ainaro (eight sub-districts) and closely examined a representative range of 22 projects. In doing so, the team held wide-ranging consultations with district and sub-district administration and line departments, sucos, contractors, users, non-government organizations and with Dili-based officials in line ministries and the ADN. Thus, whilst the sample of 22 cases is small and results should be approached with due caution in extrapolating a national picture about useability and quality of assets, the projects were chosen to be illustrative of the aggregate, and were used to trigger discussions about the performance of a much larger set of projects. Since the study aims at improving the quality of sub-national spending programs, it does not attempt to assess sub-national spending programs against alternatives that Government decided against pursuing (such as a large labor-based public works program) nor the opportunity costs of the programs. This study provides a comparative analysis of the quality of spending through PDD and PDL mechanisms against five key dimensions of expenditure quality: 19 In other words, the PDD is seen as an evolution of the PR, and hence as a single evolving mechanism. 8 Sub-national Spending in Timor-Leste 1. Expenditure Priorities: Rates of Execution, Correspondence with National Sector Goals, Geographic Equity and Poverty Goals: How do the rates of budget execution under these programs compare with the national average? Has the pattern of investments been such as to further national development goals (e.g. as spelt out in the Strategic Development Plan)? Have sub-national budget allocations achieved an appropriate balance amongst considerations of relative poverty and needs, and country-wide fairness? To what extent do PDL and PDD offer efficient and timely execution of approved budgets? 2. Quality of Investments: Are the assets being delivered at an acceptable construction quality and ‘useability’? Are they being produced in a timely and efficient manner? Are the assets being operated and maintained in ways that sustain their benefits over time? 3. Impact on Employment: To what extent have these programs stimulated local employment? How effectively have investment expenditures converted into wage income injections into the local economy? Are the opportunities perceived as being fairly distributed? 4. Promoting Local Contractors and Private Sector Development: To what degree are these programs increasing productive capacity amongst local Timorese construction businesses and creating an entrepreneurial class? 5. Governance and Accountability, Inclusion, Stability: What has been the effect of the programs on political stability at the local level? Is there evidence of elite capture of the benefits of these programs? What are the perceptions of the contracting process and the programs’ effectiveness amongst local elites? Are the mechanisms addressing or aggravating local grievance? 5.  Findings: Outcomes Against Key Policy Priorities A) Expenditure Priorities: Execution, Spending Efficiency, Alignment with National Goals, Distributional Equity Both PDL and PDD show good-to-excellent rates of budget execution until FY 2011, as demonstrated in Table 1 below, and perform considerably better than those recorded before 2009—or even now—by most line ministries for capital and development spending. They have enabled a sizeable volume of spending all across the national territory (albeit somewhat unevenly) in ways that spending by line ministries has not managed to do. Table 1.  Budget Execution Rates of Sub-National compared to Central Government Investment Programs 2009 2010 2011 2012 Sub-national Programs PR/PDD 67% 80% 94% 70% PDL 93% 103% 98% 99% Central Government Government - all capital dev’t 85% 86% 81% 46% Education - capital dev’t n/a n/a 67% n/a Health - capital dev’t n/a n/a 47% n/a Roads - capital dev’t n/a n/a 27% n/a Source: Study team analysis of data from MAE & ADN reports & MoF budget execution reports Policy Note | november 2014 9 In interpreting Table 1, a few issues need to be highlighted. Execution for PR and PDD in 2009 and 2010 was remarkable in that budgets were only approved in a mid-year rectified budget, allowing little time for implementation. On the other hand, the FY 2012 execution rate of 70%, reflects the design, procurement and payment delays encountered. Further, in using national government delivery rates as a comparator, it must be stressed that the overall rates (81-86% in 2009–2011) are greatly skewed by the performance of a few very large projects—line ministry delivery rates for smaller, more comparable projects are generally much lower, as illustrated for education, health and roads in 2011. Figure 5.  Budget Execution Measures, 2009–2012 120.00% 100% 95% 100.00% 83% 85% 77% 80% 80.00% 70% 60.00% 46% 40.00% 20.00% 0.00% 2009 2010 2011 2012 Execution District Development Execution All Capital and Development Cost efficiency: PDD contracts are awarded on the estimated budget price so there is no room for savings on the amounts budgeted for projects, whereas the savings under PDL are significant. PDL projects are put to tender—and, in 2012, this resulted in savings in the order of 6.5% of estimated costs. B) Quality of Investments In the small sample surveyed for this study, the quality of PDD investments is mixed, as illustrated by Table 2. The results suggest that a follow-up study of a larger sample might be useful. Construction quality assessments indicate that perhaps half of the projects are of acceptable standard and of benefit to users. The remainder exhibit varying degrees of design, construction quality and usability problems and almost 25% are not useable for the intended purpose without additional investments. Useability refers to quality of assets, their design and connectedness to water, power or other necessary utilities. The six PDL investments surveyed are of a significantly higher quality. Several of the recommendations aim to improve these design, quality and useability deficiencies. 10 Sub-national Spending in Timor-Leste Table 2.  Investment Sample – Efficiency, Quality and Useability Total PDD PDL Sample Size (n) 19 13 6 Delivery Time Within Year 13 7 6 Extended to Next Year 3 3 0 Still Not Completed 3 3 0 Constructed Quality Acceptable 11 6 5 Barely Acceptable 5 4 1 Poor 3 3 0 Useability Usable 11 6 5 Partly Usable 3 3 0 Not Usable 5 4 1 Source: Study team analysis of sample investments Completion Performance: Completion rates are generally good for both mechanisms, but a significant minority of PDD schemes remains incomplete. More than half of the PDD investments are completed by the year’s end and another quarter roll into the following year. But a quarter of investments are simply left incomplete by the contractor. PDL investments seem to be generally completed satisfactorily, and on time. PDL’s participatory planning steps—which have not been a feature of PDD—do not slow the delivery cycle. Rather, they merely bring forward the start of the planning process in the preceding year and indeed may actually result in PDL’s higher completion rates. Consistency with National Priorities: Spending patterns broadly reflect national sector priorities, and focus on the sorts of public goods important for poverty reduction and development. Although spending patterns vary by year and between PDD I and II, the majority—but not all20—of spending under both mechanisms is consistent with the public services prioritized in the Government’s Strategic Development Plan, namely education, water, health and agriculture. Distributional Equity: Table 3 illustrates that spending equity ($ per capita) under PDD varies greatly across districts, those with the highest allocations getting between three and seven times more than those with the lowest. While proposed projects were evaluated on a case-by-case basis, and there is no policy of favoring certain districts, Ermera has consistently received a low allocation, and Manatuto a high allocation. Further, there is no relationship between the rates of poverty in districts and the amount received. Relatively high per capita allocations have been received by Lautem (with a relatively low poverty rate) in contrast with the low allocations received by Ainaro (with a relatively high poverty rate). Dili has also consistently received one of the lowest per capita allocations, probably explicable by reference to the focus of the programs being on rural development. Allocation by district will become more equitable in the future, as a population-weighted formula approach has recently been adopted by MAE. 20 A small but significant share has been allocated to various sorts of administrative buildings and other structures, many of which though not them- selves “frontline service facilities” of direct benefit to the public are nonetheless often essential. The problem for the future is that such investments would rarely be prioritized through a local or district managed process – hence central line ministries will need a capital funding mechanism to ensure they are not neglected. Policy Note | november 2014 11 Table 3.  PDD Allocations per Capita by District ($ per Capita) PR PDD PDD I+II PDD I+II TOTAL 2011 Districts POPN. 2009 2010 2011 2012 & 2012 Aileu 45.724 n/a 29 45 63 108 Ainaro 62.407 n/a 24 37 51 88 Baucau 113.748 20 25 18 34 52 Bobonaro 93.787 37 33 29 41 70 Covalima 62.764 51 n/a 63 66 129 Dili 212.469 31 16 24 29 53 Ermera 118.671 15 n/a 12 29 41 Lautem 65.349 33 36 86 45 132 Liquica 69.925 20 31 50 50 99 Manatuto 41.217 57 49 82 87 170 Manufahi 53.995 44 21 40 72 112 Oecusse 67.736 36 23 35 68 103 Viqueque 72.950 40 33 61 75 137 Totals/Means 1,080,742 35 29 39 48 86 ST. Deviation 12.59 8.58 22.17 18.01 36.44 Highest X Lowest 2.90 3.00 7.00 3.00 4.00 Source: study team analysis of MAE and ADN reports Map 1.  Do Poorer Districts Get More Spending? Poverty Headcount 0.0 – 19.9 20.0 – 39.9 40.0 – 59.9 60.0 – 79.9 Dili 80.0 – 100.0 $53 Lautem Baucau $132 Liquica Manatuto $52 Aileu $170 $99 $108 Ermera Viqueque Bobonaro $41 $137 $77 Ainaro Manufani $88 $112 Oecusse Covalima $103 $129 Combined Spending for PDD I + PDD II (FY 2011 and FY 2012) 12 Sub-national Spending in Timor-Leste C) Employment Generation Employment impact depends greatly on the type of infrastructure. The employment impact ranges from 1.6% of investment costs (a secondary school) to 27.9% (a riverbank erosion scheme). Buildings generally employ much less labor than investments such as irrigation schemes, erosion protection walls, and other projects that involve a greater amount of moving stones and building earth structures. Figure 6.  Average % Of Total Project Cost Spent on Unskilled Labor  20 15 10 5 0 n g e ly r io n e ot on in us pp at tio at ild Ho Pr si W Su ig ec o Bu re Er I rr n er ee Bo at Gr W Access to employment opportunities: The share of total investment costs going to unskilled and skilled labor payments is around 6.5% and 3.5% respectively. Where unskilled labor is employed, this is always from the suco where the project is being implemented, and job opportunities are rotated between teams within the community, with suco and aldeia chiefs playing a large role. The benefits have been mainly local, and shared quite widely and predominantly for the rural underemployed (the urban unemployed benefit relatively little). Laborers are usually farmers or farmworkers – investment construction work takes place in the dry season (June- December) when farm underemployment is greatest. More skilled labor however is often from outside, from Dili or other districts, and in some cases is not Timorese. Total employment days generated: The programs provided seasonal employment to an estimated 2.5-10% of the sub-district workforce. Extrapolating beyond the small sample, PDD will spend in the order of $56.5 million in FY 2012, of which some $3.5 million would be expected on unskilled labor, and $1.75 million on skilled labor. This equates to some 1.2 million unskilled and 220,000 skilled workdays, respectively. This is equivalent to 18,500 unskilled work-days per sub-district; depending on local labor-rotation arrangements, this could indicatively provide 165 persons with 100 days each, 370 persons with 50 days each or 740 persons with 25 days each. These calculations were undertaken using the labor cost of $3 per day, which prevailed in past years, but has apparently since increased. Compared to other employment opportunities in rural Timor-Leste, these outcomes are significant. The results of the recent Household Income and Expenditure Survey suggest that about 40% of the population survives on $30 per month or less, with the bottom decile earning less than $10 per month.21 For low income households, 20 days of work per month, could provide a significant cash injection (of around $80 a month at $4/day or $60 a month at $3/day). 21 Ministry of Finance Timor-Leste, Timor-Leste Household Income and Expenditure Survey 2011, (National Statistics Directorate, General Directorate for Analysis & Research). Policy Note | november 2014 13 Table 4.  Labor Generated by Selected Range of Schemes Total Unskilled Labor Skilled Labor on Any Number Imputed % Number Imputed Investment Project Person Person % Project Day on any labor Cost Project on any labor Cost days days Cost day ($3/day) Cost day ($3/day) Bore Water Supply to 5 3 253 $759 2.2 2 101 $809 2.4 MAP Compound Secondary school 8 6 1,152 $3,456 1.6 2 288 $2,304 1.0 Maliana Market 32 30 6,967 $20,901 9.4 2 435 $3,483 1.6 Village Irrigation Channel 8 4 631 $1,893 11.8 4 315 $2,523 15.8 Ramaskara Raeboudas Village water supply 6 5 473 $1,419 7.9 1 79 $631 3.5 Animal quarantine building 35 25 4,620 $13,860 11.7 10 1,320 $10,560 8.9 River bank erosion 23 20 2,031 $6,092 29.6 3 265 $2,119 10.3 protection gabion wall Youth Center 6 4 473 $1,419 5.7 2 158 $1,262 5.0 Electoral office CNE building 8 6 624 $1,872 2.3 2 156 $1,248 1.6 Village water supply 22 20 1,716 $5,148 3.4 2 156 $1,248 0.8 Green house 12 5 2,222 $6,665 3.8 7 1,296 $10,368 5.9 Water supply to green 8 7 1,255 $3,765 8.0 1 157 $1,255 2.7 house Erosion site protection to 13 12 2,496 $7,488 13.1 1 192 $1,536 2.7 agriculture building Irrigation scheme 40 40 5,760 $17,280 22.1 - - - - Dormitory for health dept. 23 16 1,728 $5,184 2.2 8 864.00 $6,912 2.9 staff Irrigation scheme 22 18 1,169 $3,507 9.2 4 213 $1,701 4.5 Totals/Means 283 232 33,569 $100,707 6.6 51 5,995 $47,959 3.1 Source: study team analysis of MAE and ADN reports Comparative employment performance: In many other Asian countries, the employment share of large public works programs and labor-focused rural public investment programs, such as that established by the National Rural Employment Guarantee Act in India, ranges between 50-75%. Comparatively, PDL/PDD’s impact on employment is modest. But it needs to be emphasized that PDL and PDD are designed to serve multiple policy objectives in addition to employment creation (as compared to large public works programs that have job creation as the primary objective). The difference in the employment impact also reflects differences in the types of infrastructure being built, the higher cost in Timor-Leste of construction materials and the fact that the construction technologies in Timor-Leste are generally less labor-intensive than elsewhere in Asia. It also reflects the “thinness” of the wage labor market in Timor-Leste and the relatively high cost of wage labor.22 While employment generation has always been a goal of the programs, initially it was secondary and was neither measured nor factored into project choice. This study and anticipated follow up study with a larger sample are initial steps in generating better data about employment impact. Going forward, there has been growing interest by the Government in measuring the employment-intensivity of public spending, and for this to be factored into project choice. 22 The numbers of Timorese rural workers who work for day or piece-rate wages are only a small fraction of an overall labor force, largely devoted entirely to self-employed agriculture. Consequently, labor remuneration rates can be volatile, and also subject to changing pressures from the Dili market or from other large discrete projects. Neither mechanism creates incentives for contractors to use labor-intensive methods for construction and transport. 14 Sub-national Spending in Timor-Leste D) Contractor Capacity and Local Private Sector The number of registered local contractors has expanded greatly (data suggests a three-fold expansion since 2009), and a large number of these have each year been awarded a PDD contract. While PDD’s procurement arrange- ments have allowed a rapid and widespread out-sourcing, and effectively catalyzed the development of private contractors, the emphasis needs to shift to promote competition between existing firms and to improve capacity. PDD’s procedures for categorizing contractors and awarding contracts do not favor contractors who perform above the average. These arrangements also do not promote the consolidation or specialization of contractors, nor their diversification over time. Furthermore, while some contractors have progressed upwards to larger sized contract categories, there is evidence that the largest category of projects (category D) is the preserve of a narrow set of contractors who also received supply contracts, other public works contracts and sole source emergency contracts. While a number of these contractors have clearly gained valuable experience, it is likely that some registered firms may be simply “shells”. PDD’s “one project-to- one contractor” procurement strategy provides a strong incentive for families, clans or groups to create multiple companies. In several cases, contractors with no obvious “capacity” have been awarded large projects. Both mechanisms have encouraged some increases in contractor capacity over time to implement higher value contracts. Yet, many contractors are openly skeptical about how contractors are categorized under PDD, and believe that entry into the highest value category and the award of high value projects requires special political connections. A fair number of the new contractors seem to have enjoyed some modest capitalization—such as a brick press or vehicle—funded initially by the PR and subsequent profits. But there is limited evidence of major investments by contractors in expanding their construction capacities in capital or skilled staff or trend increases in construction- related income. This lack of major expansion is not surprising given that private sector demand for construction in Timor is limited, and that the market is dominated by public sector spending. Contractor strategy, as a result, is to utilize “networked” arrangements, whereby assets and skilled personnel are shared, loaned or hired among firms (as a more prudent, low-risk way to enhance capacity) and where the larger firms also operate opportunistically between different lines of business—construction, transport, trade, an adaptation to the rules of the market. The sustainability of ‘Category D’ businesses does not depend on PDD because they are able to access other government contracts or simply other trading opportunities. Finally, during FY 2012 many contractors stated that they completed work without advance payments, let alone final payment (i.e. 90% payment – 10% retention is for a six month liability period). Contractors argued that this delayed their work and forced them to make cheaper shortcuts. Contractors also argued that the lack of advance payments seriously affected their cash-flow and financial viability. This led to some discontent amongst these contractors, a number of whom expressed reservations about participating in PDD projects in the future. E) Governance Capabilities, Disputes, Inclusion and Stability Sub-national spending programs have enabled large numbers of Timorese to participate in the allocation of public resources and the production of assets largely in accordance with national priorities. PDD has allowed far greater local contractor involvement in the implementation of national public programs in Timor than ever before. PDL opened scope for involvement of scores of citizens and local officials in each District, not only in submitting proposals, but also in selecting and prioritizing investments and in monitoring their implementation, through procedures which have allowed a degree of consistency and transparency. Thus, while such sub-national spending remains a small share of total public expenditure (around three percent), it offers far more opportunities for citizen engagement in governance than any other item of public expenditure. Policy Note | november 2014 15 Sub-national spending is best seen as part of a suite of measures the Government introduced (including cash transfer and pension programs) to support its political, stability, social and economic objectives. These measures have enabled government to target important constituencies needed to build stable political settlements and provided openings for ordinary citizens to engage constructively with the state. A class of contractors has been created by the sub-national programs, which includes key veterans at district level, a crucial element of the political settlement that has formed since the 2006 crisis. Aside from some relatively minor incidents in the wake of the 2012 elections, the country has not experienced significant organized violence for over five years. At the same time that it played a significant role in the broader political stabilization of the country, localized disputation was associated with over half the PDD projects that were examined. Nine out of 13 PDD projects examined closely in this Study gave rise to some form of disputation about land, procurement, labor or contractor performance, whereas only one of the six PDL projects gave rise to a dispute (contractor-related). Table 5.  Conflicts in the Investment Sample Total PDD PDL Sample Size (n) 19 13 6 Disputes None 9 4 5 Land-Related 5 5 0 Labor-Related 3 3 0 Contractor-Related 9 8 1 Source: Study team analysis of sample investments Disputes generally reflect poor quality consultation during planning, weak supervision and payment autho- rization problems. PDD’s planning procedures omit several local level steps of consultation and prioritization featured under PDL.23 The involvement of chefes de sucos and chefes de aldeias in the planning process under PDL, for instance, appears to reduce the incidence of land disputes. The comparative evidence suggests that more deliberate attention to local consultation during planning, a competitive procurement process, coupled with more intense supervision, mitigate against disputes. It also appears to secure more local ownership, reduce costs and does not prolong the process of planning or completing projects. Outcomes of the PDD procurement process are also frequently disputed, a concern about which Government is responding. District officials voiced concerns about disputation around procurement. MAE reports that it receives some 60 to 80 complaints annually from disgruntled contractors; a number of these complaints also go to the Anti-Corruption Commission. PDL has adopted a more orthodox competitive tendering process,24 with district officials near unanimous that this process is quite manageable and that the outcomes tend to be better accepted by contractors. As a response, ADN has recently declared that all PDID projects will undergo a competitive tender process. Contractors themselves have mixed views as to what they prefer, although several claimed that the PDL tender process is more transparent and helps those without political connections to gain 23 PDL procedures are not problem free and the local planning manual prepared in 2012 attempts to grapple with some of these issues. For instance, a recent Asia Foundation study highlighted weaknesses of the Suco stage of the planning processes and inadequate District-to-Suco feedback in some areas under PDL. 24 That said, there is a need to tighten up PDL’s bid evaluation procedures to ensure more transparent—and efficient—selection between close bids. 16 Sub-national Spending in Timor-Leste contracts. While concerns expressed during 2008/2009 that tendering favors a few capable contractors with capacity at the expense of the majority may have been valid, the evidence from PDL in FY 2012 suggests this no longer occurs.25 The evidence from this study does not support the view that accelerated spending has accentuated existing tensions along social, economic or political lines. Indeed, there is significant evidence that those contractors who are linked with the political opposition are not being excluded from PDD projects. There is little evidence that the assets themselves or employment opportunities funded under either PDL or PDD are being captured by local elites. A basic precept of good governance is clarity and consistency in application of the “rules of the game” under which public authorities make decisions, which in turn weakens the scope for political manipulation or disputation. Yet, ‘clarity and consistency’ are in tension with the contending need to periodically adjust the rules in accordance with lessons learned and to meet changing policy priorities. The period from 2008 has seen a frequent adjustment in procedures and the assignment of functions to different players and levels in the system, thus demonstrating a commitment to innovation and learning by doing. There was a relative lack of clarity and transparent procedures in the initial PR program. While in subsequent PDD investment cycles procurement has been regulated, PDD procedures for identifying schemes, preparing estimates and supervising contracts have been frequently changing. The frequency of changes to procedures has led to a degree of uncertainty amongst officials at the local level as to how PDD is to be managed, fed confusion and perceptions that the schemes are not transparent or inclusive, and resulted in cynicism—amongst ordinary people, contractors and civil society observers as to why and how resources and contracts were allocated. These problems were likely compounded by the nature of public information on PDD budget allocations and expenditures, or on investment-specific budgets and expenditures. While notice boards provided basic information at some sites, they are far from universal. Budget Book 3 was an innovation that provides cost estimates, but unlikely to be disseminated beyond district elites, and final project budgets and actual expenditures are not so readily available. That said, the cynicism is more than is warranted, as there is little sign of “capture” of the benefits accruing to PDD. 6.  Conclusions and Recommendations PR marked a bold but risky initiative. Aimed at quickly stimulating rural development, the PR bypassed the serious constraints on sub-national public investment spending and delivery inherent in the existing public financial management system and responded to concerns that Timorese companies could not win Government contracts. It was one of several measures taken by the Government to achieve more equitable national development, consolidate the peace, and share the independence dividend and natural resource revenues more widely. In a situation where the financial markets were not working properly, these mechanisms funded contractor capitalization and capacity, while achieving respectable and at times impressive rates of execution. The innovative mechanisms that the Government created for sub-national spending in the aftermath of the 2006 crisis, beginning with the PR, provide instructive lessons about the trade-offs needed to balance competing technical, social and political priorities, and of how these have to be adjusted as needs change over time. 25 A substantial portion of those interested (71% of those who attended the pre-bid meetings) were in fact able to submit bids that were assessed as “eligible”. In ten districts, the median contractor received only one project; in the other three districts (Aileu, Manufahi, and Viqueque) the median contractor received only two contracts. Policy Note | november 2014 17 Several years on, while the results are uneven, the PR and PDD have largely achieved their initial policy objectives, and the Government’s decision to normalize or stabilize the system through the PDID Decree Law of January 2012 is timely. Mainly due to these programs, there are now many domestic private sector companies who have the capacity to compete through competitive tendering. There is little evidence that PDL-style procurement would favor a few qualified and capable contractors if implemented in 2014, or result in time-consuming delays. Indeed, the evidence suggests that both PDL-style competitive tendering and also preliminary participatory planning may often save time later on and do not at all compromise rates of execution/spending, project completion, spread of contracts, or consistency of spending with national priorities. There is no evidence to support the view that sub-national spending mechanisms, compared to centralized spending by line ministries, are more vulnerable to “weak local capacity”. The relatively high quality infrastructure consistently produced by PDL spending, and produced in a significant percentage of cases under PDD suggest that where there is poor performance, this is not caused by inherent weaknesses in local administration or contractors as com- pared to capacities in the central line ministries. The reasons are more complex. Rather, the causes of the sometimes poor performance lie largely in unclear, inappropriate, poorly communicated or frequently changing investment delivery procedures, in lack of incentives for good performance, or in the political economies around these. The recommendations below are made in the spirit of the commitment to institutional innovation and to learning by doing in the PDID Decree. As part of the implementation of the PDID Decree, a common set of well-publicized and transparent procedures must be developed for funds allocation, planning and budgeting, procurement and payments as well as the assignment of responsibilities. Key recommendations: 1.  Funds allocation, planning & budgeting 1.1  Improve district allocations by introducing formula-based district budget envelopes. This is a low- cost opportunity to promote transparency and geographic equity, and which also has multiple potential pay-offs for quality of planning and the performance of district management. ƒƒ To ensure the degree of predictability needed for a meaningful participatory District planning process it is important that there be some medium term stability in the national pool of resources allocated for PDD. The nominal $50/capita used to design the Suco Development Fund under the PNDS sets a good precedent—a similar pool (denominated in average $ per capita) could be announced for the medium term, consistent with national budgetary availability. ƒƒ Formula-based allocations could improve geographic equity and ensure the allocation of funds would reflect relative district poverty and need. ƒƒ Announcing district budget envelopes at the start of each annual cycle would improve the quality of planning and promote a focus on delivery of high quality assets. ƒƒ The introduction of simple ‘district performance measures’ as part of district allocation formulae could reward good performance on planning, budgeting, procurement and investment management, and perhaps also on local employment. ƒƒ Introducing elements of performance into district allocations will require more concerted attention to monitoring and evaluation systems, fully integrated with program management. 1.2  Clarify the priority given to employment generation through sub-national spending modalities. If employment creation objectives are overwhelmingly significant, consider the merits of arrangements that would either earmark a share of the funds for ‘employment creation’ or establish a special purpose grant dedicated to this purpose. 18 Sub-national Spending in Timor-Leste 1.3  Create a separate funding mechanism for the national investment priorities of line ministries, which district planning will inevitably miss. A more permanent mechanism is needed, once the 15% of the PDID allocations reserved for line ministry projects expires after 2013, to fund legitimate small/ medium capital investment funding requirements determined by central line ministries (such as the various “administrative buildings” funded under PDD). 1.4 Consolidate and refine the PDID planning process, to ensure an appropriate degree of local consultation and dispute pre-emption, and of investment efficiency. ƒƒ Address some of the consultative shortcomings as identified in the Asia Foundation review. Draw lessons from the ongoing PDL three year rolling strategic planning process (conducted in Viqueque). ƒƒ Tighten investment screening and appraisal for economic projects. Introduce weightings for employment created by PDD projects. ƒƒ Improve district-sector coordination to ensure investments reflect sector policy and avoid duplication. 1.5  Safeguard the principle of district approval of the District Investment Plan, backed by central level responsibility to ensure the legality of the Plan). As the new PDID Planning Manual makes clear, clarity of responsibilities to approve plans (the District) and to check the plan’s compliance with regulations (MAE) are crucial aspects of achieving accountability in subnational spending. MAE’s role is indispensable, but its ‘sign off’ should only be for “legality”—rather than second-guessing district priorities and introducing an additional layer of controls. It is mistaken to believe that this can be better done in a national forum than at district level. 2.  Procurement strategy 2.1 Shift procurement strategy toward (i) much simpler contractor pre-qualification procedures, centrally managed by ADN and the Ministry of Public Works (perhaps with IADE), (ii) introduce simple competitive tendering (but tightening the PDL procedures), with performance links in District funding to promote compliance, and (iii) review the contractual 10% retention arrangements in order to improve contractor incentives for quality work. This measure will be widely supported by districts and contractors, and will allow consolidation of the contractor sector and encourage the better performers, while also removing what is a currently a major irritant for District Administrations. 2.2  Maximize opportunities to build contractor capacity and link PDID implementation with other ongoing programs, notably International Labor Organization (ILO) support to IADE for enterprise development, and ILO and European Union training programs for contractors and for adoption of labor-based technology. These other programs offer opportunities to complement the market opportunities that PDD/PDID offers to contractors, by helping to put these contractors in a position to compete, to manage their contracts and to adopt the sort of labor-based technologies that will promote local employment. 3.  Investment Supervision and Assignment of Responsibilities 3.1  Rethink the current roles of the ADN and Line Ministries in design, supervision, and certification, as the current role of ADN district engineers appears to be ineffective in some cases. The project cost revisions by ADN Dili has in some instances caused problems in the implementation of projects. ƒƒ End the split between supervision and certification—although ADN’s role in certification was intended to provide an independent check on quality and ensure payments only authorized for work properly completed, in reality this does not always work and has, on occasion, resulted in inappropriate, premature certification by ADN engineers. Policy Note | november 2014 19 ƒƒ Co-opt ADN district engineers in the preliminary and also final project design and costing phase, rather than have ADN Dili review final costs alone, post facto (and have ADN Dili issue district ADN and other district engineers with clear design templates and up-to-date and acceptable sub-district materials and transport cost schedules) 3.2  Invest in adequate technical personnel and local travel budgets to improve input into design and supervision. Lack of such input means that 25% of PDD investments are not useful and some $15 million is wasted each year. ƒƒ Place more technical staff into the districts – funding could be through a small percentage levied from the district grants to be used for technical support.[NB Beware false economies! Compare the cost of un- useable assets with the deployment of five dedicated engineers in each District which in total might cost around $400,000 annually.] ƒƒ Define roles more clearly. As a priority, reduce blurred accountabilities for technical staff of the District Development Committees, district line ministries and the ADN. 4. Payments Revert to district level payments for contractors as soon as possible. To achieve this, establish district treasuries as soon as the Integrated Financial Management Information System (IFMIS) linked to the Ministry of Finance has been operationalized. Reverting to district payment for contractors will help to deal with the kind of payment bottleneck experienced in FY 2012, arising from both delays in certification reporting to Dili and in central Treasury payments. The comparative evidence suggests that deconcentrating the treasury function (provided it is linked with IFMIS) is likely to expedite payments and impact positively on contractor performance. 20 Sub-national Spending in Timor-Leste 1818 H St. NW Washington, DC 20433, USA