Document of The World Bank FOR OFFICIAL USE ONLY Report No: ICR00004612 IMPLEMENTATION COMPLETION AND RESULTS REPORT IDA 48950 ON A CREDIT IN THE AMOUNT OF SDR 19.3 MILLION (US$30 MILLION EQUIVALENT) TO THE UNITED REPUBLIC OF TANZANIA FOR A DEVELOPMENT OF A NATIONAL STATISTICAL SYSTEM FOR TANZANIA April 29, 2019 Poverty and Equity Global Practice Africa Region CURRENCY EQUIVALENTS (Exchange Rate Effective: April 17, 2019) Currency Unit = SDR SDR 1.4 = US$1 US$0.7 = SDR 1 FISCAL YEAR July 1 – June 30 Regional Vice President: Hafez Ghanem Country Director: Bella Bird Senior Global Practice Director: Carolina Sanchez-Paramo Practice Manager: Pierella Paci Task Team Leader(s): Nadia Belhaj Hassine Belghith, Elizabeth Ann Talbert ICR Task Team Leader: Habtamu Neda Fuje ABBREVIATIONS AND ACRONYMS CAS Country Assistance Strategy CPF Country Partnership Framework DHS Demographic and Health Survey DP Development Partner DFID Department for International Development of the United Kingdom FYDP II Second Five-year Development Plan GIS Geographical Information System GoT Government of Tanzania HBS Household Budget Survey HR Human Resource ICR Implementation Completion and Results Report ICT Information and Communication Technology IEG Independent Evaluation Group ISR Implementation Status and Results Report LFS Labor Force Survey LGA Local Government Authority M&E Monitoring and Evaluation MDAs Ministries, Departments, and Agencies MDG Millennium Development Goal NADA National Data Archive NBS National Bureau of Statistics NGO Nongovernmental Organization NPS National Panel Survey NSS National Statistical System OCGS Office of Chief Government Statistician PAD Project Appraisal Document PDO Project Development Objective PPI Producer Price Index SCD Systematic Country Diagnostic SCI Statistical Capacity Index SDG Sustainable Development Goal SSWG Statistics Sector Working Group STATCAP Statistical Capacity SUT Supply and Use Table TA Technical Assistance TDV Tanzania’s Development Vision ToC Theory of Change TSMP Tanzania Statistical Master Plan TABLE OF CONTENTS DATA SHEET ...............................................................................................................................1 I. PROJECT CONTEXT AND DEVELOPMENT OBJECTIVES ....................................................... 5 A. CONTEXT AT APPRAISAL .........................................................................................................5 B. SIGNIFICANT CHANGES DURING IMPLEMENTATION (IF APPLICABLE) ..................................... 11 II. OUTCOME .................................................................................................................... 13 A. RELEVANCE OF PDOs ............................................................................................................ 13 B. ACHIEVEMENT OF PDOs (EFFICACY) ...................................................................................... 14 C. EFFICIENCY ........................................................................................................................... 20 D. JUSTIFICATION OF OVERALL OUTCOME RATING .................................................................... 22 E. OTHER OUTCOMES AND IMPACTS (IF ANY) ............................................................................ 23 III. KEY FACTORS THAT AFFECTED IMPLEMENTATION AND OUTCOME ................................ 24 A. KEY FACTORS DURING PREPARATION ................................................................................... 24 B. KEY FACTORS DURING IMPLEMENTATION ............................................................................. 24 IV. BANK PERFORMANCE, COMPLIANCE ISSUES, AND RISK TO DEVELOPMENT OUTCOME .. 25 A. QUALITY OF MONITORING AND EVALUATION (M&E) ............................................................ 25 B. ENVIRONMENTAL, SOCIAL, AND FIDUCIARY COMPLIANCE ..................................................... 26 C. BANK PERFORMANCE ........................................................................................................... 27 D. RISK TO DEVELOPMENT OUTCOME ....................................................................................... 28 V. LESSONS AND RECOMMENDATIONS ............................................................................. 29 ANNEX 1. RESULTS FRAMEWORK AND KEY OUTPUTS ........................................................... 31 ANNEX 2. BANK LENDING AND IMPLEMENTATION SUPPORT/SUPERVISION ......................... 47 ANNEX 3. PROJECT COST BY COMPONENT ........................................................................... 49 ANNEX 4. EFFICIENCY ANALYSIS ........................................................................................... 50 ANNEX 5. BORROWER, CO-FINANCIER AND OTHER PARTNER/STAKEHOLDER COMMENTS ... 51 ANNEX 6. SUPPORTING DOCUMENTS (IF ANY) ..................................................................... 52 The World Bank Development of a National Statistical System for Tanzania (P107722) DATA SHEET BASIC INFORMATION Product Information Project ID Project Name P107722 Development of a National Statistical System for Tanzania Country Financing Instrument Tanzania Investment Project Financing Original EA Category Revised EA Category Partial Assessment (B) Partial Assessment (B) Organizations Borrower Implementing Agency MINISTRY OF FINANCE National Bureau of Statistics Project Development Objective (PDO) Original PDO To develop a national statistical system that effectively and efficiently delivers reliable and timely statistics in accordance with international standards and best practice. Page 1 of 52 The World Bank Development of a National Statistical System for Tanzania (P107722) FINANCING Original Amount (US$) Revised Amount (US$) Actual Disbursed (US$) World Bank Financing 30,000,000 29,871,249 27,880,147 IDA-48950 Total 30,000,000 29,871,249 27,880,147 Non-World Bank Financing 0 0 0 Borrower/Recipient 5,000,000 6,900,000 8,800,000 CANADA: Canadian International Development 4,892,607 4,892,607 4,457,724 Agency (CIDA) UK: British Department for International Development 16,170,500 10,000,000 8,067,107 (DFID) EC: European Commission 7,113,000 0 0 SWEDEN: Swedish Intl. Dev. 0 0 0 Cooperation Agency (SIDA) Total 33,176,107 21,792,607 21,324,831 Total Project Cost 63,176,107 51,663,856 49,204,978 KEY DATES FIN_TABLE_DAT Approval Effectiveness MTR Review Original Closing Actual Closing 24-Mar-2011 09-Aug-2011 26-Jun-2014 31-Oct-2016 30-Jun-2018 RESTRUCTURING AND/OR ADDITIONAL FINANCING Date(s) Amount Disbursed (US$M) Key Revisions 29-May-2015 14.22 Change in Results Framework Change in Loan Closing Date(s) 10-Dec-2015 19.78 Reallocation between Disbursement Categories 24-Feb-2017 21.31 Other Change(s) Page 2 of 52 The World Bank Development of a National Statistical System for Tanzania (P107722) KEY RATINGS Outcome Bank Performance M&E Quality Satisfactory Moderately Satisfactory Modest RATINGS OF PROJECT PERFORMANCE IN ISRs Actual No. Date ISR Archived DO Rating IP Rating Disbursements (US$M) 01 05-Dec-2011 Satisfactory Satisfactory .62 02 23-Aug-2012 Satisfactory Moderately Satisfactory 3.89 03 07-Jun-2013 Satisfactory Moderately Satisfactory 3.89 04 31-Jan-2014 Moderately Satisfactory Moderately Satisfactory 7.28 Moderately 05 29-Sep-2014 Moderately Unsatisfactory 8.59 Unsatisfactory 06 16-Jun-2015 Satisfactory Satisfactory 14.22 07 02-Feb-2016 Satisfactory Satisfactory 19.78 08 18-Nov-2016 Satisfactory Moderately Satisfactory 21.31 09 07-Jul-2017 Satisfactory Moderately Satisfactory 24.88 10 30-Jan-2018 Satisfactory Moderately Satisfactory 24.88 SECTORS AND THEMES Sectors Major Sector/Sector (%) Public Administration 80 Other Public Administration 80 Social Protection 20 Public Administration - Social Protection 20 Page 3 of 52 The World Bank Development of a National Statistical System for Tanzania (P107722) Themes Major Theme/ Theme (Level 2)/ Theme (Level 3) (%) Private Sector Development 0 Jobs 10 Job Creation 10 Public Sector Management 0 Public Administration 10 Transparency, Accountability and Good 10 Governance Data Development and Capacity Building 60 Data production, accessibility and use 60 Social Development and Protection 0 Social Inclusion 10 Other Excluded Groups 10 Social Protection 10 Social Safety Nets 10 ADM STAFF Role At Approval At ICR Regional Vice President: Obiageli Katryn Ezekwesili Hafez M. H. Ghanem Country Director: John McIntire Bella Deborah Mary Bird Senior Global Practice Director: Marcelo Giugale Carolina Sanchez Practice Manager: Kathie L. Krumm Pierella Paci Habtamu Neda Fuje, Nadia Task Team Leader(s): Thomas Danielewitz Belhaj Hassine Belghith, Elizabeth Ann Talbert ICR Contributing Author: Boniface Essama Nssah Page 4 of 52 The World Bank Development of a National Statistical System for Tanzania (P107722) I. PROJECT CONTEXT AND DEVELOPMENT OBJECTIVES A. CONTEXT AT APPRAISAL Context 1. The implementation of Tanzania’s first National Strategy for Growth and Poverty Reduction in 2005–2010 led to a strong economic growth performance. However, this growth had a limited impact on poverty. The average annual growth rate over the period hovered around 7 percent. Yet, poverty incidence remained close to the 2001 level. The country did make some progress toward achieving three targets in the Millennium Development Goals (MDGs): reducing infant and under-five mortality, combating HIV/AIDs and malaria, and addressing gender inequality. Faced with the limited success of the first National Poverty Reduction Strategy, in November 2010, the Government of Tanzania (GoT) adopted a new five-year strategy for 2010–2015 aligned with its development vision known as the Tanzania Development Vision (TDV) 2025 which aims at turning Tanzania into a middle-income country. The strategy sought to sustain high economic growth while making the pattern of growth more inclusive. 2. Effective policy making is an information-dependent process. Tanzania’s second poverty reduction strategy called for accelerating economic growth and poverty reduction. These considerations created a need for policy makers to measure the well-being of the population, monitor changes at both individual and societal levels, and evaluate the impact of policy interventions on the overall quality of life of the target population. 3. The development strategy also called for the promotion of good governance and accountability. These values are consistent with Managing for Results, a public sector management approach that entails the use of information on performance and results to improve decision making. The approach commits countries to results-oriented reporting and assessment framework and to strengthen the links between development strategies and budget processes. Effective dissemination of quality data and evidence produced in support of policy making improves transparency and enables other stakeholders such as the media, nongovernmental organizations (NGOs), and ordinary citizens to hold the government accountable. Thus, the ability to produce and disseminate reliable information about the economy and the well-being of the population on time is considered an important indicator of good policies and institutions. 4. The project design team identified several factors that led to increased demand for more and better-quality statistics in Tanzania. These include (a) commitment to Managing for Results, (b) development of democracy along with increased freedom of the press, (c) increased civil society’s awareness of the important role of reliable and timely statistics in support of transparency and accountability, (d) the need of private sector operators for business and market information, and (e) user demand for new and more advanced statistical products such as Geographical Information System (GIS) and high-frequency and subnational data. 5. The National Statistical System (NSS) was unable to fully cope with the emerging demand for high- quality official statistics. The completion report for the 2007–2011 Country Assistance Strategy (CAS) noted that weak statistical capacity made it difficult for the Government to accurately measure poverty and track progress toward development objectives. Page 5 of 52 The World Bank Development of a National Statistical System for Tanzania (P107722) 6. A situation analysis, conducted in the context of the preparation of the Tanzania Statistical Master Plan (TSMP), identified several factors constraining the ability of the NSS to produce quality data and statistics: (a) limited technical ability, (b) scarce human and financial resources, (c) dilapidated physical infrastructure, (e) lack of coordination within the NSS, and (e) deficient dissemination mechanisms. These constraints stemmed from deeper issues such as limited autonomy for the organizations in charge of the production and dissemination of statistics and a distorted incentive system facing staff. The prevailing incentive system favored data collection activities in the field rather than the use of administrative data systems and dissemination of quality statistical products. 7. To deal with the abovementioned challenges facing the NSS, the Government developed the TSMP, which was formally endorsed by the Cabinet in June 2010 to serve as a five-year national strategy for the development of statistics focusing on five dimensions: (a) Institutional Development and Legal Reform, (b) Human Resource Capacity Development, (c) Development of Statistical Infrastructure, (d) Data Development and Dissemination, and (e) Physical Infrastructure and Equipment Development. The overall objective of the TSMP was to strengthen the country’s statistical capacity and establish a well- coordinated statistical system that meets user needs and delivers quality and relevant statistics in a timely and cost-effective manner. 8. Development partners (DPs), including the World Bank Group, agreed to align their support for statistics in Tanzania and harmonize funding through a TSMP basket fund managed by the National Bureau of Statistics (NBS). The World Bank and Department for International Development of the United Kingdom (DFID) led the donor coordination efforts to support implementation of the TSMP. 1 The World Bank’s support was consistent with the alignment between the CAS and the national development strategy and with IDA priorities. Negotiations for IDA14 and IDA15 replenishments emphasized the importance of managing for development results and the central role that the NSS had to play in this process. It was, therefore, recommended that IDA supports the client countries in their efforts to improve results-based management. 9. The project was designed to contribute to the following higher-level objectives: (a) achieve the goals of the national development strategy and MDGs, (b) promote a culture of evidence-based decision making, and (c) strengthen governance and accountability. Components of the project were aligned with TSMP areas and are the following: A. Institutional Development and Legal Reform; B. Human Resource Capacity Development; C. Development of Statistical Infrastructure; D. Data Development and Dissemination; E. Physical Infrastructure and Equipment and F. Project Management. Theory of Change (Results Chain) 10. Even though the Project Appraisal Document (PAD) did not present an explicit Theory of Change (ToC) to guide project design, the Implementation Completion and Results Report (ICR) team has 1 Other major DPs supporting statistical activities at the time include International Monetary Fund EAST AFRITAC providing technical assistance (TA) for macroeconomic statistics, European Commission financing the National Panel Survey (NPS), the United Nations Children’s Fund providing funding and TA for health surveys, and the United Nations Population Fund supporting demographic statistics including preparations for the 2012 Population and Housing Census. The Canadian International Development Agency had also provided financial support for the implementation of the TSMP through a basket funding arrangement. Page 6 of 52 The World Bank Development of a National Statistical System for Tanzania (P107722) reconstructed the logical connections between the project activities and the intended outcomes, along with the underlying assumptions, based on the PAD and other project documents. 11. The NSS is the target entity of the project. While the intervention is understood to target the whole statistical system, each activity or a combination of activities may target only parts of the system. It is therefore useful to understand the basic structure of the NSS and how the activities of the intervention affect the key elements of the system. Such an understanding can facilitate the identification of the causal mechanisms at work. 12. An NSS is a network of people and organizations involved in the production, dissemination, and use of data and other statistical products and services within a given institutional environment. The institutional environment includes institutional and organizational factors that have a significant influence on the performance of the system. Clearly, the quality of statistical outputs hinges critically on the quality of the institutional environment as well as the quality of the production and dissemination processes. Quality statistics are relevant (that is, meet users’ needs), reliable, and timely. To enable the production and dissemination of quality statistical outputs, institutional arrangements must ensure professional independence, give a clear mandate to statistical authorities for the collection and dissemination of data, and ensure adequacy of available resources and commitment to quality. Similarly, the production of quality statistics requires processes based on sound methodology, appropriate statistical procedures, and cost-effectiveness. Therefore, the project is designed to enhance the capacity and processes at the NSS to enable the production of quality statistics. 13. Consistent with the TSMP, the project sought to enable the NSS to produce quality statistics for decision makers in an objective, timely, and cost-effective manner. According to the overall objective, the project is essentially a statistical capacity-building intervention meant to deal with statistical quality issues in Tanzania. There are four key determinants of the capacity of an NSS to fulfill its mandate: (a) the legal and institutional framework that governs the system, (b) organizational capacity, (c) human capital, and (d) data systems.2 The quality of governing institutions depends on the extent to which they create an enabling environment. Organizational capacity, which depends on factors such as financial resources, infrastructure, equipment, management practices, standards, and quality assurance mechanisms, determines the extent to which processes function with quality and efficiency. Human capital entails knowledge, technical skills, attributes, and attitudes. Recruitment, training, changes in attitudes, and increased motivation can improve human capacity. 14. The five substantive project components listed earlier target the key aspects of the NSS:  System Reorganization (Components A and E) o Management and communications systems of the NBS, the Office of Chief Government Statistician (OCGS); ministries, departments, and agencies (MDAs); and local government authorities (LGAs) o Legal and regulatory reforms o Physical Infrastructure and equipment upgrading  Incentive Realignment (Components A and B) o New career structure 2 PARIS21. 2018. “Assessing the Capacity of National Statistical Systems: A User’s Guide.” Paris. Page 7 of 52 The World Bank Development of a National Statistical System for Tanzania (P107722) o New job profiles and salary scale o Performance management system supportive of good performance  Individual Capacity Building (Component B) o Training on management, budgeting, and procurement o Professional statistical training for staff and users in relevant organizations (NBS, OCGS, and MDAs)  Statistical Processes Improvement (Components C and D) o Development of statistical infrastructure o Data development and dissemination 15. The causal pathways associated with a ToC entail technical and behavioral relationships that link the effective implementation of the activities described earlier and the outputs generated in the form of enhanced capacity of target individuals and organizations. To translate increased individual capacity and motivation into performance/outcomes, the project sought to improve alignment of organizational commitment with the project objectives, support committed leadership to guide the staff toward the desired outcomes and improve the institutional environment. Figure 1 presents the ToC in a nutshell based on the hypothesized relationship between capacity, incentive, institutional environment, and performance. The figure also shows that the Project Development Objective (PDO) is expected to contribute to a higher-level development vision: improved policy making. 16. The key assumptions necessary for establishing effective causal links are also presented in Figure 1. For the effective enhancement of the NSS, the system reorganization would create an enabling environment and the new NBS structure would be endorsed by the GoT in a timely manner. The proposed incentive system and capacity building activities are assumed to be compatible with the delivery of high- quality statistics, and the new incentive structure (salary scheme and staff regulations) would be endorsed by the GoT. Page 8 of 52 The World Bank Development of a National Statistical System for Tanzania (P107722) Figure 1. Theory of Change To address the following A decentralized and largely uncoordinated national statistical system with CHALLENGE... limited capacity to deliver quality data and statistics that meet user needs. …to generate the The project ...which will We will implement these following ...leading to the has five achieve the ACTIVITIES... INTERMEDIATE ultimate VISION: components PDO… RESULTS… System Reorganization Well-informed  Management and A An NSS that policy making communications systems effectively Component  Legal and regulatory and A and E reforms Enhanced national efficiently  Physical infrastructure and systems for data delivers equipment upgrading production and sharing reliable and based on timely  Improved legal and statistics in Incentive Realignment institutional accordance  New career structure environment; with Component  New job profiles and salary  Enhanced international A and B scale organizational and standards  Performance management B human capacity; and best system supportive of good and practices. performance  Upgraded statistical processes Individual Capacity Building:  Training on management, Component budgeting, and D B procurement  Professional statistical training for staff Underlying Assumptions: (A) System Reorganization: is implemented as designed to create an enabling environment for the production and dissemination of high- Statistical Processes quality data. The new NBS organizational structure is endorsed by the C Improvement GoT on time. Component (B) Incentive Realignment and Individual Capacity Building: New C and D  Development of statistical incentive system is compatible with the delivery of high-quality and infrastructure timely data; and developed capacity is relevant for the implementation  Data development and of the new statistical processes. The proposed salary scheme for NBS dissemination staff is endorsed by GoT. (C) Statistical Process Improvement: this needs to conform to international standards and best practices. Source: Author’s construction. Page 9 of 52 The World Bank Development of a National Statistical System for Tanzania (P107722) Project Development Objectives (PDOs) 17. The PDO is to develop an NSS that effectively and efficiently delivers reliable and timely statistics in accordance with international standards and best practices. Key Expected Outcomes and Outcome Indicators 18. The main outcomes that were expected to lead to the achievement of the PDO included the timely availability of high-quality statistical products and services. The original outcome indicators were (a) Percentage of users reporting satisfaction with statistical products and services of the NBS; (b) Proportion of planned publications that were actually produced and made publicly available on time on NBS website; and (c) Full and timely implementation of census and surveys specified in the Intercensal Survey Calendar 2012–2022; Components 19. The project has six components with a total investment of US$64.4 million, of which US$30.0 million was financed by IDA. The first five are substantive components while the last one is project management. Each component and the original amount of resources allocated in the basket funding to it are described in the following paragraphs. Component A: Institutional Development and Legal Reform (US$1.7 million) 20. This component entails a comprehensive organizational restructuring of the management and administration of the NBS and OCGS and the financial management, human resource (HR) management, and communications systems of the MDAs and LGAs benefitting from the project. In addition, the component supports the revision of existing legislation and promulgation of draft regulations enabling the NBS to become an autonomous public office governed by an independent Executive Board of Directors. Component B: Human Resource Capacity Development (US$5.0 million) 21. This component supports the development and implementation of an HR management strategy, based on new job profiles and salary scales for the NBS and OCGS; a new career structure and a performance management system that provides rewards and incentives for good performance; and provision of management skills training for professional staff and others with managerial responsibility, and budgeting, and procurement skills training. Furthermore, the component involves the development and implementation of a professional statistical training program for staff of the NBS, OCGS, MDAs, and other relevant producers of official statistics. Component C: Development of Statistical Infrastructure (US$8.6 million) 22. This involves a program to improve and maintain the recipient’s NSS by developing a (a) central business register of establishments and business enterprises; (b) national master sampling frame for Page 10 of 52 The World Bank Development of a National Statistical System for Tanzania (P107722) household-based censuses and sample surveys; (c) compendium of concepts, definitions, and methodologies for application in statistical data collections and analysis; (d) spatial GIS for providing maps for survey fieldwork and providing the basis for dissemination of geographically disaggregated and referenced statistical data; and (e) national classification of economic and social statistics. Component D: Data Development and Dissemination (US$11.7 million) 23. This component seeks the improvement of social and economic statistics within the NSS through a program to strengthen surveys and routine data collection, storage, processing, compilation, analysis, reporting, dissemination, utilization, archiving, and communication within the NSS. Component E: Physical Infrastructure and Equipment (US$14.7 million) 24. The component involves the construction and equipping of new office buildings for the NBS and OCGS with appropriate upgraded information technology and communications infrastructure, consisting primarily of computers and networking equipment, training rooms and library facilities, and provision of appropriate vehicles and other equipment needed to collect, process, and disseminate statistical data. Component F: Project Management (US$3.4 million) 25. This component supports project management and monitoring and evaluation (M&E) activities. It ensures the provision of consultancies for TA in strategic management and NSS coordination and specialists in procurement, financial management, M&E, and information technology. 26. In addition to the basket funding, a parallel funding of US$12.5 million was originally allocated to Component D. The GoT has also initially allocated a total US$6.9 million distributed across the five components. B. SIGNIFICANT CHANGES DURING IMPLEMENTATION (IF APPLICABLE) 27. The project underwent a total of four restructurings related to the Results Framework, the closing date, a reallocation of funds, and the relocation of office building for the NBS. None of these entailed changing the PDO or the project components. Revised PDOs and Outcome Targets 28. The PDO and project components remained unchanged throughout the implementation period. Revised PDO Indicators 29. The first restructuring (approved on July 9, 2013) entailed adding missing milestones and target data to some of the indicators in the Results Framework to clarify what the indicators were intended to measure. In addition, baseline milestones and targets were set for some of the indicators; and the units of measure changed for other indicators. 30. The second restructuring (approved on May 29, 2015) involved (a) extending the closing date of the project from October 31, 2016, to June 30, 2018, to align with the new closing date of the TSMP; (b) Page 11 of 52 The World Bank Development of a National Statistical System for Tanzania (P107722) updating the Results Framework to align it with a recent revision of the TSMP Results Framework; and (c) streamlining the number of performance indicators from the 22 original indicators to 15 new indicators. Out of the 15 indicators, the project selected 10 core indicators to monitor performance. They include three PDO-level indicators and seven activity-level indicators distributed across the five project components. (a) PDO-level Indicators 1: National Geographical Information System strengthened 2: Economic statistics strengthened 3: Number of data sets uploaded to the National Data Archive (NADA) (b) Activity-level Indicators Component A: Institutional Development and Legal Reform 4: Legal framework for the NSS with clear roles, responsibilities, reporting lines, and functions developed Component B: Human Resource Capacity Development 5: Annual staff training plan implemented Component C: Development of Statistical Infrastructure 6: Implemented Quality Assessment Framework Component D: Data Development and Dissemination 7: Cumulative number of surveys implemented under TSMP 8: Performance against Annual Release Calendar 9. Traffic on NBS website Component E: Physical Infrastructure and Equipment 10: New NBS and OCGS buildings Revised Components 31. No change. Other Changes 32. The third restructuring was approved on December 10, 2015 and involved the reallocation of funds between the project components. In particular, it sought to reduce the funds allocated to the activities related to institutional development, development of human capacity, statistical infrastructure, data dissemination, and operating expenses (Components A–D and F) and increase the budget for activities related to physical infrastructure and civil works (Component E). The proposed reallocation was designed to be ‘activity-neutral’ in the sense that it did not imply any change in the scope or nature of project activities, implementation arrangements, or other amendments. Page 12 of 52 The World Bank Development of a National Statistical System for Tanzania (P107722) 33. The fourth and final restructuring was approved in December 2016. It involved moving the construction site of the new office buildings for the NBS to Dodoma instead of Dar es Salaam. Rationale for Changes and Their Implication on the Original Theory of Change 34. The revision of the Results Framework and PDO indicators was intended to reflect the fact that the project’s scope has been narrowed to more realistic and achievable outcomes. The implementing team and the Bank management realized that, like many first-generation statistics projects, the original PDO was very ambitious, and some of the outcomes and intermediate results were difficult to achieve due to factors that are beyond the control of NBS or the Bank team. It was argued that the 10 new performance indicators represented aspects of the project that were simpler and clearer, more realistic, measurable, and better reflect the project’s achievements. They were also better aligned with the revised TSMP Results Framework. Some indicators that were considered to be unrealistic were dropped, and other pragmatic indicators were added. In reality, the changes amounted to reducing the scope of the PDO. At the time, a management decision was made not to revise the PDO. This made the evaluation process difficult by creating a disconnect between the PDO and the indicators tracked. 35. The main purpose of extending the closing date was to align with the new closing date of the TSMP. This made sense, given that the project was designed to support the Master Plan. The extension would allow the GoT to complete the planned project activities and achieve the objectives of the TSMP. 36. The reallocation of funds was meant to cover deficits arising from the initial difficulty in estimating costs accurately for certain components. For example, the cost of the OCGS’s office building was almost twice the initial allocated amount of US$2 million, and the amount was later revised to US$4 million. The resources were taken from the implementation of activities related to survey data collection and capacity building. Deficits were also exacerbated by a loss of approximately US$2.8 million because of the devaluation of the SDR relative to the US dollar. Although it was argued in the December 2015 restructuring paper that the efficient implementation of data development and capacity building activities produced savings in resources, it became clear that the capacity building trainings were implemented only partially. 37. The relocation of the construction site of the new NBS buildings was related to the Government’s decision to move to Dodoma (the capital city) and the fact that the NBS was one of the institutions required to relocate to Dodoma during the first phase of the process. Relevant safeguards documents were updated, and a new Environment and Social Impact Assessment was completed. 38. The team believed that the changes from the restructuring had no impact on the underlying ToC. Therefore, there was no need to change the PDO and project components. II. OUTCOME A. RELEVANCE OF PDOs Assessment of Relevance of PDOs and Rating Page 13 of 52 The World Bank Development of a National Statistical System for Tanzania (P107722) 39. The objective of developing an NSS that effectively and efficiently delivers reliable and timely statistics in accordance with international standards and best practices is consistent with the Government’s priorities and the World Bank Group’s Country Partnership Framework (CPF), both at appraisal and closing. 40. At the time of appraisal, the country’s Poverty Reduction Strategy called for the promotion of good governance and accountability, hence for evidence-based policy making. The NSS was unable to meet the increased demand for relevant and credible statistical information linked to the development of democracy along with the increased freedom of the press, the information needs of civil society in support of transparency and accountability, and the need of economic operators for business and market information. The inability of the NSS to meet this demand for information was mainly because of a poor institutional environment combined with limited human and organizational capacity. The TSMP was therefore designed to establish a well-coordinated statistical system to meet user needs and deliver reliable and timely statistics that meet international standards. 41. At closing, the project’s objective remained well aligned with the country’s second Five-year Development Plan (FYDP II) covering 2016–2021. FYDP II is designed to accelerate the achievement of the objectives of TDV 2025 and seeks to accelerate growth and poverty reduction through structural transformation, human development and responsive governance. The underlying M&E strategy acknowledges the important contribution of the statistical production by the NBS according to the TSMP calendar and commits to the continuation of efforts to improve existing data systems to enhance the M&E of FYDP II priorities. 42. The World Bank’s CAS (2011), at appraisal, acknowledged the constraints facing the NSS of Tanzania at the time and sought to address these by supporting the development and implementation of the TSMP. The World Bank thus decided to contribute US$30 million through the Development of the National Statistical System (STATCAP) credit with an expected implementation period of 2011–2016. The project is also consistent with the current CPF (2018–2022), which is designed to support FYDP II. It emphasizes, among others, activities that strengthen governance and public institutions, and advocates for policy making grounded in knowledge and solid analytics. The CPF recognizes the importance of accurate and timely data for the design, M&E of development strategies, and interventions for poverty reduction and shared prosperity. 43. Finally, it is worth noting that the project objectives are well aligned with the 2030 agenda. According to the CPF, about 70 percent of the 282 indicators contained in the M&E Framework for FYDP II can be mapped to the Sustainable Development Goals (SDGs). The NBS and MDAs are responsible for the production of these indicators. The CPF pledges to support capacity-building efforts for institutions involved in the formulation of M&E plans for national development strategies. 44. The relevance of the development objective is therefore rated High. B. ACHIEVEMENT OF PDOs (EFFICACY) Assessment of Achievement of Each Objective/Outcome 45. As stated earlier, the PDO is a compound objective that can reasonably be represented by the following three groups of objectives which entail improvement in (a) effectiveness and efficiency of the Page 14 of 52 The World Bank Development of a National Statistical System for Tanzania (P107722) NSS in producing and disseminating statistical information, (b) reliability and accuracy of data, and (iii) timeliness. 46. Table 1 shows two sets of PDO-level indicators: those originally listed in the PAD and the revised indicators which emerged from the 2015 restructuring. As the restructuring did not change the PDO nor the target outcomes but instead involved dropping indicators and adding new ones, a ‘spilt’ evaluation is not necessary. Therefore, the efficacy is assessed based on the revised indicators. Table 1. PDO-level Results Indicators Indicators in the PAD Percentage of users reporting satisfaction with statistical products and services of the NBS Proportion of planned publications that were produced and made publicly available on time on NBS website Full and timely implementation of census and surveys specified in the Intercensal Survey Calendar 2012–2022 Revised indicators National Geographical Information System strengthened Economic Statistics strengthened Number of data sets uploaded to the National Data Archive (NADA) 47. Like most first-generation statistics projects, this project started out with an ambitious goal to overhaul the NSS of Tanzania. The design aimed to reorganize the NBS, improve institutional environment by supporting statistical legislation, change HR structure, streamline data collection by MDAs and LAs as well as support data development by NBS and OCGS. This has been proved to be overly ambitious and many results were beyond the scope of the NBS and the Bank team. As discussed above, the 2015 restructuring re-focused the project on narrow objectives and reduced the scope of the PDO without changing the PDO itself, creating a disconnect between them. Even though the indicators do not capture the full PDO, the evaluation relies on these indicators and information on implementation status of individual components and to what extent the Bank team and NBS have control over their implementation. 48. Table 2 shows the observed change in the outcome indicators as presented in the Implementation Status and Results Reports (ISRs). As shown in the table, targets were met for the first and third outcomes, while the second outcome was partially achieved. Table 2. Observed Change in Outcome Indicators Indicator Baseline Target Achievement Geographical Village shape files Begin preparatory work for Enumeration Area shape files Information produced and the 2022 Population and edited and used in preparation System disseminated Housing Census based on GIS for the 2022 Population and Strengthened information Housing Census Economic National accounts Quarterly production of Construction of the 2012 SUT Statistics series based on 2007 Producer Price Index (PPI) Strengthened Supply and Use Table based on the 2012 basket (SUT) disseminated National accounts series based on 2012 SUT Number of 9 24 24 Page 15 of 52 The World Bank Development of a National Statistical System for Tanzania (P107722) datasets uploaded to the National Data Archive (NADA) Source: ISRs. 49. When we look at the individual components, there is evidence of significant delays in implementation. The midterm review (2014) identified the slow pace of implementation as the major challenge facing the project. The review also identified some key factors accounting for implementation delays. These factors were related mainly to three components (a) Institutional Development and Legal Reform, (b) Human Resource Capacity Development, and (c) Physical Infrastructure and Equipment. As discussed in the following paragraphs, the delays in project implementation are caused by factors that are beyond the control of the implementing team and NBS. 50. With respect to institutional development, for example, delays were because of long procedures involved in the enactment and operationalization of the 2015 Statistics Act. In addition, as reported in the last ISR, the passage of the Statistics Act in 2015 also provoked strong criticism from the media, NGOs, and independent researchers. The Government responded with a pledge to amend the controversial clauses it introduced and clarify the interpretation of the law through further regulations. While the new law was adopted in 2015, the regulations providing clarifications for the implementation of certain sections of the law were published in the Government Gazette in February 2017. In addition, as will be discussed later, highly controversial amendments to the 2015 Act was passed by the Parliament in 2018. 51. The 2015 Act also stipulates a governing board for the NBS. The chairperson has been appointed, but the full board is not yet in place. 52. Furthermore, the organizational restructuring and HR policy reforms for the NBS have stalled. The reorganization plan for the NBS was approved by the Government in November 2015 and was supposed to become effective in July 2016. It has been on hold pending the completion of the ongoing review of the overall public service sector structure seeking to identify strategies for enhancing efficiency and effectiveness in the delivery of public services. The new organization structure was approved on November 5, 2018 and will become effective in July 2019. The delay in implementation is again caused by institutional factors beyond the control of NBS. 53. Progress in HR capacity development was monitored based on the implementation of the annual staff training plan. The project failed to meet the target level of training. During 2012/13–2016/17, only 45 percent of the OCGS staff who were supposed to receive training have received the planned training. Only 35 of the 65 trainings that were included in the annual plans of the OCGS have been executed. Similarly, the NBS managed to conduct only 40 of the 64 trainings that were scheduled in the annual staff training plan. Training remained below target because priority was given to other activities in the annual work plan and the difficulties in securing international expertise for the planned training. However, in addition to formal trainings included in the staff training plan, the NBS and OCGS staff members have received on-the-job trainings during survey preparation and implementation as part of the TA for data development. The NBS also offered trainings, which are typically carried out by consultants hired through TA, to the OCGS staff in some areas where the NBS has comparative advantage. Page 16 of 52 The World Bank Development of a National Statistical System for Tanzania (P107722) 54. Finally, with respect to physical infrastructure and equipment, it is worth noting that the planned office spaces became available close to the end of the project. The OCGS Zanzibar building was inaugurated in January 2018 and the NBS building in Dodoma was started late, but it was almost completed and handed over in May 2018. The delays are because of factors beyond the NBS and/or sequencing of activities outlined in the project design. The building in Zanzibar was delayed partly because of sand scarcity and the Government’s ban on sand mining. Delays in the NBS building was primarily because of project design decisions that made progress in reform of the Statistics Act a precondition for initiating the civil works activities of the physical infrastructure component. In addition, the relocation of government offices to Dodoma was another obstacle. The implementing team managed to address the latter by restructuring the project. After dealing with these challenges, the NBS was able to complete the construction project within a year. In terms of equipment, a technical support mission in January 2018 found that the NSS computer systems need modernization, in particular those in the NBS. The available computer systems did not seem to be suitable for compilation and dissemination of quality statistics. However, new computers for the staff and a computer for the library of NBS were already procured by the time of the ICR mission. It is worth noting that this is the second upgrade of computer systems under the project. The first upgrade took place in 2012. Justification of Overall Efficacy Rating 55. The information presented in table 2 suggests that dimensions of the PDO related to effective delivery of statistics in accordance with international standards had been substantially achieved. However, the three outcome indicators, while relevant, do not fully convey the development of a NSS, which is another dimension of the stated PDO. The first two outcome indicators (a strengthened GIS and strengthened economic statistics) are manifestations and therefore indirect evidence of a stronger statistical system. Furthermore, the above review of key activities confirms that some of the components were either delayed or only partially implemented. 56. As stated earlier, the project was designed to develop an NSS that effectively and efficiently delivers reliable and timely statistics in accordance with international standards and best practices. The PDO indicators only partly capture this compound objective statement and hence additional information that could augment these indicators were provided. The terms ‘effectively’ and ‘efficiently’ refer to improved capacity of the system to deliver data services that meet users’ needs, whereas the terms ‘timely’ and ‘international standards and best practices’ connote data timeliness and quality. These considerations suggest the following potential subobjectives: (a) Improve the capacity to collect, process, and disseminate statistical information (b) Improve relevance, accuracy, and reliability of data (c) Improve timeliness of data production and dissemination (d) Improve data accessibility 57. The ToC discussed earlier considers improvement in organizational and individual capacity as a critical factor which determines the achievement of the other three subobjectives. The statistical capacity outcome can be measured by the Statistical Capacity Index (SCI) generated by the World Bank. The SCI is an aggregation of 25 indicators related to methodology, data sources, periodicity, and timeliness. Ngaruko (2008) estimated that about 77 percent of the overall indicator is based on activities and outputs, Page 17 of 52 The World Bank Development of a National Statistical System for Tanzania (P107722) while 23 percent reflects aspects of capacity building per se.3 The indicator can be used along with other information to assess the statistical capacity in a given country. 58. Even though a direct comparison of statistical capacity before and after the TSMP does not provide clear causal evidence, the level of statistical capacity before the TSMP can serve as a reasonable benchmark to assess the effectiveness of the project in developing capacity. Figure 2 shows the SCI in Tanzania before (2005–2009) and during (2010–2017) implementation of the TSMP. Before 2010, the observed level of capacity ranged between 61 and 66 (of the maximum allowable score of 100). The SCI started improving during the final stages of project preparation through efforts supported by a project preparation advance. Improvements continued after the project became effective and peaked with a score of 75.6 in 2015 and then started decreasing, though still above the pre-project score. This information suggests that the project has contributed to some capacity building. Figure 2. Statistical Capacity in Tanzania, 2005–2017 72 72 72 72 76 73 68 71 71 63 66 63 61 61 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 Source: World Bank. 59. Through formal trainings and on-the-job TAs, staff members of the NBS and OCGS have acquired some knowledge and skills. A technical support mission conducted toward the end of January 2018 (less than six months before the closing date) found that while the NBS and most MDAs have the basic capacity to produce most key social statistics needed by policy makers, these organizations are unable to produce many of the key economic statistics at the recommended frequency and in accordance with relevant international standards and best practices. One of the factors accounting for this situation could be the lack of training in modern methods for collecting and compiling such statistics. The mission identified an urgent need for quality improvement in the following domains: national accounts, industrial production, commerce and distribution services, consumer price indexes, construction cost indexes, producer price indexes, labor cost indexes, balance of payments, external trade indexes, and related short-term indicators of inputs and outputs. 60. The second subobjective listed earlier is based on three dimensions of data quality: relevance, accuracy, and reliability. Relevance indicates the extent to which the available data meet the needs of the 3Ngaruko Floribert. 2008. “The World Bank’s Framework for Statistical Capacity Measurement: Strengths, Weaknesses and Options for Improvement.” The African Statistical Journal Volume 7: 149–169. Page 18 of 52 The World Bank Development of a National Statistical System for Tanzania (P107722) users such as policy makers and other stakeholders. The initial user satisfaction survey in 2011 indicated that about 70 percent of them are happy with the statistical products and services. The outcome indicator tracking users’ satisfaction was dropped by the restructuring because of unrealistic expectations regarding the target for the indicator, which was initially set at 80 percent. Instead it was decided to elicit feedback and engage users more directly through regular user-producer focus groups. 61. Accuracy and reliability assess the extent to which the available data accurately and reliably portray reality. These quality dimensions can be monitored through data quality assessments. In 2015, the NBS prepared and published a Data Quality Assessment Framework based on the 2002 Statistics Act and the then proposed new Statistics Legislation (later ratified as the 2015 Statistics Act). The framework describes various standards and guidelines that need to be applied by any institution that produces statistics and the measures that have been put in place to ensure the quality of the production process. These standards are taken from international best practices and modified to fit the country’s situation and facilitate the regional data harmonization effort. The framework is now expected to be adopted broadly, and the NBS works with MDAs to facilitate the application of the framework. Of course, it is difficult for NBS to strictly enforce the adoption of this framework by MDAs that have limited capacity to collect high quality statistics. It should also be noted that measuring data quality is quite challenging. Even though quality of certain socioeconomic statistics such as labor and poverty statistics has shown improvements, the quality of some statistics such as data inputs used for national accounts appears to have deteriorated. A team from the Bank and the International Monetary Fund/IMF that has recently worked on national accounts identified that there are data quality issues with regards to administrative data collected by MDAs. The Bank-Fund team noted that, even if the NBS uses internationally accepted methodology for national accounts, the poor quality of administrative data generated by MDAs compromised the quality of national accounts statistics. The PDO indicator used to monitor the progress in the strengthening of economic statistics did not explicitly track quality of inputs, and it only focused on updating the SUT in 2012 (from the previous SUT constructed in 2007). 62. Timeliness can be measured by lag time in production and dissemination of data relative to a preset schedule. The project has supported the implementation of surveys and censuses according to the 2012–2022 Intercensal Survey Calendar. Almost all surveys were conducted as scheduled during the TSMP period. The socioeconomic surveys such as the Demographic and Health Survey (DHS) and Household Budget Survey (HBS) have been conducted regularly according to the calendar both before and during implementation of the TSMP. However, agriculture- and establishment-related surveys and censuses such as Industrial Census, Agriculture Sample Census Survey, Updating of Business Register, and Employment and Earnings Survey have been conducted more regularly only because of the project. In addition, as noted earlier, timeliness is one of the dimensions of the SCI. Further unpacking of the SCI shows that Tanzania has improved the timeliness of poverty statistics and has achieved the recommended frequency of three to five years for poverty surveys (figure 3). Similar achievement is recorded for health surveys such as DHS. Even though this was achieved before project implementation, the resources and improved capacity availed through the project have supported timely health and poverty data collection and dissemination as well as improved the data quality. However, Tanzania remains behind in timely collection and public release of agriculture and population censuses and adoption of internationally accepted vital (birth and death) registry. The vital registration system coverage was beyond the scope of the project except for a smaller digitization pilot in a few districts in Zanzibar. Page 19 of 52 The World Bank Development of a National Statistical System for Tanzania (P107722) Figure 3. Periodicity of Key Surveys and Censuses Legend Agricultural Census Health Survey Population Census Poverty Survey Vital Registrationa 0.0 >10 years >5 years >10 years >5 years 0.5 >10 years 3–5 years >10 years 3–5 years 1.0 0–10 years 0–3 years 0–10 years 0–3 years Source: World Bank. Note: a. The score is 1 if the country completes registries of vital (birth and death) statistics by the United Nations Department of Economic and Social Information and Policy Analysis; 0 otherwise. Green denotes the achievement of the recommended periodicity or standard. 63. The project has improved data accessibility by increasing the number of databases added to the NADA. This improved data accessibility could enhance transparency and accountability. It can also stimulate data demand and use. The January 2018 implementation support mission, however, found that the existing mechanisms for coordination, collaboration, information dissemination, and data sharing needed further improvement. 64. Ultimately, this performance story, considering a very ambitious objective of overhauling the NSS of Tanzania and a difficult operational environment, suggests a Substantial achievement of the PDO. C. EFFICIENCY Assessment of Efficiency and Rating 65. Official statistics is commonly viewed as a public good, and its provision should be financed and managed by the public sector. This characterization makes it difficult to apply standard cost-benefit analysis to quantify the net economic benefits associated with the project. Thus, no such exercise was presented in the PAD. 66. There are aspects of project implementation that contributed to efficiency and others that reduced it. For example, the use of surplus funds from some activities to cover deficits in other activities has likely improved efficiency. However, as noted earlier, the project was affected by implementation delays. While some of these delays were because of contextual factors outside the project’s control, some Page 20 of 52 The World Bank Development of a National Statistical System for Tanzania (P107722) others were because of poor systems for planning and implementation and overall unrealistic planning especially related to procurement planning and execution. The midterm review (2014) found that procurement procedures were not always followed properly and were often affected by delays in the various steps of the procurement process. A procurement consultant was hired by the project to improve procurement performance and a management tool was developed to improve the day-to-day management of activities. A recommendation was also made to hire, on a fixed-term contract, a national management assistant to support the two coordinators in managing the implementation of activities. 67. Despite delays in the implementation of the physical infrastructure component of the project, the implementation cost did not increase relative to the initial budget. The construction of office buildings in Zanzibar took much longer than anticipated. Even though the process started in December 2014, the building was inaugurated in January 2018. The completion cost, however, was slightly below the revised budget of US$4 million (but, of course, above the initial budget of US$2 million). On the other hand, despite starting quite late because of delays in the ratification of the 2015 Act, which was set as a precondition for investment on physical infrastructure, the office building in Dodoma was completed within a year. The construction cost of US$5.5 million was well below the initial budget of US$7.9 million. Because of cost savings in building construction, the fund reallocated to building during restructuring was used for data development. 68. The efficient implementation of the data development and dissemination component, which accounts for 40 percent of the entire project cost, could be evaluated in terms of the cost of conducting standard surveys such as HBSs and Labor Force Surveys per interviewed household, relative to cost in other African countries. The total cost of the 2017/18 HBS in Tanzania mainland, including cost of survey preparation, fieldwork, and data processing, was US$382 per household. In Zanzibar, the cost of the 2014/15 HBS is relatively lower (US$270 per household). The reasons might be the relative ease of implementing fieldwork in Zanzibar, compared to Tanzania mainland, and the inflation during 2014–2017. To make a reasonable comparison of survey costs across countries, Kilic, T., U. Serajuddin, H. Uematsu and N. Yoshida (2017) compare the real, instead of the nominal, cost of similar surveys in Africa (figure 4). Tanzania performs relatively well compared to other countries in the region. Figure 4. Implementation Costs of Budget Surveys Per Interviewed Household (in 2014 US$) 500 (2014 US$ per household) 400 406 Cost of survey 396 300 331 297 309 200 199 100 0 Niger Malawi Uganda Tanzania Nigeria Ethiopia Source: Kilic et al. 2017. Note: The cost is stated in 2014 US$. The surveys analyzed are part of the Living Standards Measurement Study - Integrated Surveys on Agriculture and the Tanzania NPS (2008/09). Page 21 of 52 The World Bank Development of a National Statistical System for Tanzania (P107722) 69. Considering the delays in implementation of several components of the project, completion of construction of office buildings within budget, and relatively reasonable survey costs, the efficiency of the project is rated Substantial. D. JUSTIFICATION OF OVERALL OUTCOME RATING 70. The overall outcome of this STATCAP Project is rated Satisfactory due to the highly relevant PDO, and although some of the implementation difficulties prevented the delivered outputs to be fully converted into intended outcomes within the envisioned timeframe, the project improved data quality, timeliness and accessibility in Tanzania. 71. As noted earlier, the PDO is very much aligned with both the country’s and the Bank’s priorities and is consistent with Managing for Results, evidence-based policy making, and the emerging 2030 agenda. A strategy that can plausibly lead to the achievement of this objective included activities designed to enhance human and organizational capacity and to create an enabling institutional environment. The five components of the project targeted these critical domains, thus making the project design very relevant. 72. The project has made significant contribution toward the improvements of data quality and accessibility in Tanzania. Prior to the project, quality of socioeconomic statistics was poor and accessing data was difficult. The project has brought about substantial improvements in the quality of key socioeconomic indicators including labor statistics, health indicators, poverty statistics, household surveys and so on. The NBS and OCGS now follow international guidelines for collecting and compiling socioeconomic statistics. Furthermore, the project has improved data accessibility. It supported surveys and censuses to enable the NBS and OCGS achieve the targets set in the Intercensal Survey Calendar (2012–2022). The number of datasets released on the NADA has increased from 9 to 24 during the project period. Due to the project’s effort, the NBS now makes socioeconomic statistics available for public consumption on its website. 73. The project improved human capacity for collection and compilation of socioeconomic statistics through local and international short courses, study tours, and workshops. Some 77 trainings in the areas of statistics, information and communication technology (ICT) and GIS, and leadership have been supported by the project during 2012–2018. A total of 595 training participants from the NBS and OCGS have attended these trainings. At the organizational level, the project delivered two office buildings, in Zanzibar and Dodoma, and ICT equipment. The project also improved collaboration between the NBS, OCGS, and MDAs through the compilation of joint annual workplans. 74. To support the creation of an enabling environment, the project triggered the enactment of the 2015 Statistics Act which clarifies the roles and responsibilities of national stakeholders and reporting lines and functions of constituent organizations. In particular, the associated regulations make the NBS an autonomous public office and principal coordinator of the NSS. This addressed the fragmented data collection and dissemination before the TSMP. 75. While most activities have been carried out, implementation delays have led to partial outputs in some areas or late delivery of outputs: (a) The reorganization of the NBS is incomplete because the governing board is not in place and the HR policy reform is on hold. (b) The new office buildings, which Page 22 of 52 The World Bank Development of a National Statistical System for Tanzania (P107722) were supposed to improve the working conditions for staff of the NBS and OCGS, were delivered at the end of the project. 76. Going forward, these shortcomings in human and organizational capacity development might constrain the ability of the system to sustainably produce and disseminate high-quality data and statistics without further financial and technical support. Less than six months before project completion, a technical support mission recommended that, to further improve the ability of the NSS to produce high quality economic statistics in a timely manner, additional support in the area of human development, coordination among data producers, and data dissemination is needed. This assessment is consistent with the finding by the 2017 Systematic Country Diagnostic (SCD) that Tanzania still has massive needs for quality data. E. OTHER OUTCOMES AND IMPACTS (IF ANY) Gender 77. Not applicable. Institutional Strengthening 78. A key component of the project sought to improve the legal and institutional framework of the NSS. Before the reforms supported by the project, the NSS was highly decentralized and largely uncoordinated. The main organizations in the system include the following: (a) The NBS in Dar es Salaam and the OCGS in Zanzibar are primarily responsible for compiling national accounts and price statistics and conducting large-scale household surveys and establishment-based surveys, including the centennial population census. (b) MDAs collecting, processing, storing, and disseminating mostly administrative data. (c) Other producers of official statistics include the Tanzania Revenue Authority, responsible for collecting revenue statistics, and the Bank of Tanzania responsible for balance of payment statistics and banking and monetary statistics. 79. The adoption of the 2015 Statistics Act and the associated regulations improved the quality of statistical governance in Tanzania. In particular, the legislation increased the autonomy of the NBS and instituted mechanism to ensure alignment and harmonization of statistical services within the NSS. Mobilizing Private Sector Financing 80. The project did not mobilize private sector financing. Poverty Reduction and Shared Prosperity 81. The project was not designed to have a direct impact on poverty reduction and shared prosperity. However, strengthening the capacity of the NSS to produce and disseminate quality data and statistics has the potential of improving the evidence base of development interventions intended to reduce poverty and promote shared prosperity. As noted earlier, effective policy making requires quality data Page 23 of 52 The World Bank Development of a National Statistical System for Tanzania (P107722) and statistics. To fight poverty in all its dimensions and reduce inequality, policy makers need credible information on the nature and extent of poverty and inequality, their root causes, and social consequences. Reliable measures of poverty and inequality are needed to monitor changes at both individual and societal levels, compare the relative situation of policy-relevant socioeconomic groups, and assess the impact of policy interventions on the target population. Other Unintended Outcomes and Impacts 82. None. III. KEY FACTORS THAT AFFECTED IMPLEMENTATION AND OUTCOME A. KEY FACTORS DURING PREPARATION 83. The project was overly ambitious given the operational environment, and the design did not account for limited absorption capacity. As noted in a recent Independent Evaluation Group (IEG) report, building statistical capacity takes time and therefore requires a long-term engagement (eight to ten years).4 The reorganization of the NSS involved many institutions, organizations, and stakeholders. This fact created significant coordination challenges. It also meant that five years would not be enough to fully reform the system, and have it operate as intended. As a result, reaching targets defined in the Results Framework remained elusive for years. 84. The NBS noted that the TA provided by the World Bank and DPs during project preparation were uncoordinated. This probably contributed to the time lag from project inception (2008) to effectiveness (2011). 85. There was an imbalance in the design of the project to the extent that activities emphasized the supply side of the system more than the demand side. The fact that statistical output is a result of interaction between producers and data users creates a need to actively promote data use by the Government and citizens. It is not clear whether strategies to enhance human capacity targeted enough organizations and people on the demand side. B. KEY FACTORS DURING IMPLEMENTATION 86. Implementation started slowly as there were delays in setting up the TSMP basket fund, which required close coordination between the different DPs engaged in the sector and aligning and documenting the principles and procedures governing the basket fund. Second, the client faced difficulties in developing realistic implementation and Procurement Plans, which formed the basis for drafting and submitting the first withdrawal application. This led to late disbursement of funds and resulted in low levels of expenditure and activities observed over the 2011/12 financial year and the backlog of activities in the following year. 4IEG. 2017. Data for Development: An Evaluation of World Bank Support for Data and Statistical Capacity . Washington DC: World Bank Group. Page 24 of 52 The World Bank Development of a National Statistical System for Tanzania (P107722) 87. The long process associated with the enactment of the 2015 Statistics Act is one of the key factors that affected implementation and outcome. The reorganization of the system, the development of physical infrastructure, and HR management (for example, creation of a new salary scale and career profiles) all depended on the Act. 88. Inadequate planning was another key factor affecting implementation and outcome. The lack of planning capacity had frequently pushed several outputs off track. Deadlines were frequently missed. HR development was severely constrained by poor planning and delays in approving training plans and delays in finding appropriate training providers. This situation affected the capacity development component of the project. For instance, the training activities conducted in 2012/13 reached only 39 staff members out of the 89 planned. The effects of poor planning were exacerbated by the unrealistic levels of ambition and a lack of clear prioritization. The midterm review thus recommended that the TSMP coordination teams draw up quarterly plans for project implementation to facilitate workload analysis. IV. BANK PERFORMANCE, COMPLIANCE ISSUES, AND RISK TO DEVELOPMENT OUTCOME A. QUALITY OF MONITORING AND EVALUATION (M&E) M&E Design 89. While the overall PDO was clearly stated, the three revised PDO-level indicators selected for monitoring progress toward that objective do not encompass all outcomes implied by the PDO statement. As noted earlier, the PDO statement entailed at least three key outcomes: (a) capacity development (to collect, process, and disseminate data); (b) improved data quality in terms of relevance, accuracy, reliability, and timeliness; and (c) improved data accessibility. The selected outcome indicators focus only on the strength of the national GIS and economic statistics and publication of data on the NADA. With hindsight, the chosen outcome indicators do not seem to cover the full scope of the PDO. 90. The revision of the Results Framework implemented by the second restructuring of the project was supposed to streamline the number of performance indicators and improve measurability. The design of the M&E system could have been enhanced by including meaningful indicators for (a) organizational capacity, (b) human capacity, (c) the relevance of statistical products, (d) accuracy and reliability of data produced, and (e) timeliness of production and dissemination. In fact, the five categories of indicators initially envisaged in the PAD would have reflected most of these dimensions. These categories include (a) overall user satisfaction with the statistical system; (b) indicators of the availability and timeliness of data for key statistical activities and other sector-level data; (c) indicators of the quality of key statistics, including the extent to which MDAs observe national standards for key statistical activities; (d) indicators of data accessibility such as online availability and accessibility of microdata, survey report, and metadata; and (e) indicators of the performance of the statistical system as a whole and progress on core reforms and institutional policies and procedures. 91. While some of the activity-level indicators are related to the capacity and data quality issues, by and large they are not defined and measured in a way that fully satisfy the underlying information need. For instance, the indicator for human capacity development is defined as annual staff training plan implemented. This indicator cannot inform on the type and level of skills acquired by the trainees and the extent to which the capacity gap has been closed by the implementation of the annual training plan. Page 25 of 52 The World Bank Development of a National Statistical System for Tanzania (P107722) Similarly, a quality indicator is defined as Quality Assessment Framework Implemented. Again, this indicator is not very informative with respect to methodological soundness and extent of the framework’s adoption by the NBS, OCGS, and MDAs. M&E Implementation 92. The TSMP coordination team was entrusted with the responsibility of regularly monitoring PDO- level and intermediate results indicators on the basis of the Results Framework. The NBS was also required to compile and report on several performance indicators. Throughout the life of the project, information on indicators was collected for each period according to the prevailing Results Framework (original or revised). The indicators were also measured and reported in the ISRs according to the underlying structure of the Results Framework. 93. Prima facie, the stakeholders were interested in effective M&E implementation. The fourth ISR (2013) rated the overall performance as Moderately Satisfactory because of ‘the failure to meet a few ambitious targets’ related to (a) the delivery of statistical products and services in accordance with the pre-announced annual release calendar, (b) the implementation of the annual training plan, and (c) the planned organizational restructuring of the NBS. The ISR then argued that the Results Framework should be changed to improve measurement. These considerations led to the revision of indicators used to monitor project performance with stronger emphasis on production and dissemination of statistical products and services and less on legal and organization reform. This change meant that the M&E system covered fewer dimensions of the PDO. M&E Utilization 94. As noted earlier, the indicators were reported in the ISRs and used to provide evidence of achievement of outcomes. However, given the weak conceptual connections between the selected indicators and the relevant dimensions of the PDO, it seems that progress assessments reported in the ISRs were overly optimistic. Thus, from 2015 to 2018, all ISRs rated the progress toward achievement of the PDO as Satisfactory. Justification of Overall Rating of Quality of M&E 95. There were some shortcomings in the design and implementation of the M&E system which limited its ability to provide evidence to assess all dimensions of the PDO. Subsequent revisions of the framework realigned the focus of project activities consistent with discussions at the midterm review to assign more weight to the actual production and dissemination of statistical products and services and less on legal and institutional reform. This was done without changing the PDO, which as a result was not fully captured by the revised set of indicators. In light of these considerations, the overall quality of M&E is rated Modest. B. ENVIRONMENTAL, SOCIAL, AND FIDUCIARY COMPLIANCE 96. The project Environmental Assessment is category ‘B’ and triggered only one of the World Bank’s safeguard policies, Environmental Assessment (OP/BP 4.01). All environmental impacts were linked to the construction of new office buildings of the NBS in Dodoma and the OCGS in Unguja, Zanzibar. An Page 26 of 52 The World Bank Development of a National Statistical System for Tanzania (P107722) Environmental Impact Assessment was conducted for each site in accordance with national law. It was found that the anticipated impacts would be short term, site specific, confined, and reversible and could be adequately managed through the proposed Environmental Management Plans. C. BANK PERFORMANCE Quality at Entry 97. The technical design of the project was based on a detailed situation analysis of the NSS, which accurately identified the constraints limiting the ability of the system to meet user needs for high-quality statistics. The same situation analysis informed the design of the TSMP. The TSMP was prepared through an extensive consultation process involving key stakeholders such as the main data-producing units in MDAs, data users outside the government, civil society organizations, and DPs. The identified constraints included (a) dilapidated physical infrastructure, (b) limited human and organizational capacity, (c) distorted staff incentives, and (e) deficient dissemination mechanisms. While there was no explicit reference to an intervention theory, the project components collectively represented a plausible strategy to address the key issues facing the NSS. New legislation was intended to create an enabling environment by outlining the roles and responsibilities of all the NSS stakeholders. It made the NBS the apex organization of the NSS with a primary role as adviser, coordinator, and custodian of official statistics. A sound reorganization of the NBS, combined with new buildings and human capacity development, was supposed to create effective management and incentive structures that were compatible with the production and dissemination of high-quality statistics. However, as it became evident later, the project design was ambitious and did not take into account the challenging operational environment. Quality of Supervision 98. As evidenced by the series of ISRs produced, the World Bank team conducted regular supervision missions and made efforts to resolve many of the issues leading to implementation delays. For instance, to counter the effects of inadequate planning on implementation and outcome, a decision was made to hire a consultant to help the TSMP coordination team develop a management tool to improve the preparation and implementation of quarterly plans for project implementation and support day-to-day management of activities. Furthermore, a procurement consultant was hired by the project to assist the TSMP coordination team as soon as it became evident that slow procurement was causing implementation delays. 99. While the World Bank team correctly identified the need to revise the Results Framework to improve the recording of project accomplishment and tracking of project progress, the refocusing of the Results Framework meant that some dimensions of the PDO were no longer fully covered. As a result, the ratings reported in the ISRs tended to be overly optimistic. From May 2015 to January 2018, progress toward the achievement of the PDO was reported as satisfactory, although evidence for achievement of some dimensions of the PDO was not explicitly tracked. Justification of Overall Rating of Bank Performance 100. Even if the project identification and preparation was based on a detailed situation analysis of the NSS, the design was ambitious and did not take into account the challenging operational environment. Page 27 of 52 The World Bank Development of a National Statistical System for Tanzania (P107722) This points to moderately satisfactory quality at entry. In this light, the team’s efforts to rapidly respond and address issues as they arose justify an overall rating of Moderately Satisfactory. D. RISK TO DEVELOPMENT OUTCOME 101. Basically, the achievement and sustainability of the development objective depend on three key factors: (a) commitment, (b) resource availability, and (c) capacity to sustain performance over time. The PAD identified three sources of risk to the PDO: coordination difficulties, limited capacity, and financial management. It was recognized that the complexity of the project required effective coordination of activities across multiple organizations within the NSS. The establishment of Statistics Sector Working Groups (SSWGs) was supposed to mitigate the risk of coordination failure. It is not clear whether these groups did work effectively as intended. A technical support mission conducted in January 2018 found that the existing mechanisms of coordination, collaboration, and information dissemination and data sharing among producers of official statistics need to be improved. The mission found that the SSWGs were not meeting in recent years. 102. With respect to capacity, the ability of senior management at the NBS to provide strategic guidance and successfully implement the proposed institutional reforms was considered critical for the achievement of the PDO. Indeed, the NBS was meant to play the role of an apex organization within the NSS working in collaboration with a diverse set of stakeholders. Training and study tours were envisaged to ensure that the TSMP coordination team was fully equipped to manage the implementation of the TSMP. The capacity risk remains significant in view of the fact that intended institutional and organization reforms are incomplete. 103. The overall financial management risk was considered substantial at appraisal given the fact that the implementation agency had no previous experience in implementing IDA operations and the overall declining performance in the public sector financial management. Procurement and financial management training for relevant staff, in-depth fiduciary review, and the World Bank’s support to Tanzania public financial management systems were expected to mitigate that risk. This risk was consistently rated Substantial in the ISRs. 104. Another major risk to development outcomes is the governance risk. As noted earlier, to support the creation of an enabling environment, the project supported the enactment of the 2015 Statistics Act which intends to make the NBS an autonomous public office and principal coordinator of the NSS. However, the act has been amended recently. The amendments, which were approved by the Parliament in September 2018, prohibit the dissemination of any statistical information, which are contradictory to the official statistics and have not been endorsed by the Statistician General of NBS. Under the amendments, any person or institution will be held criminally liable for the production of statistical information that contradicts the official statistics. The Amendments appear to have been initiated in response to lively public debate on socioeconomic statistics galvanized by social media, data generated by private sector through digital platforms, and utilization of such statistics in the national political discourse. The 2018 Amendments have not been publicly debated, and the Bank and donor community were not given the opportunity to comment during the preparation stage. 105. The amendments would have implications on contestability of statistics and cross-validation of evidence from both official and unofficial sources. This would likely affect data quality and pose a serious Page 28 of 52 The World Bank Development of a National Statistical System for Tanzania (P107722) challenge to project outcomes. In addition, it constrains the Bank’s and other DPs’ evidence-based approach to development. Therefore, the amendments, in their current version, could hinder the Bank’s operations and analytics and advisory services. At the completion of this ICR report, the Bank is in a continuous dialogue with the Government to clarify the implications of the amendments. 106. Going forward, the Amendments, if not revised, would make it difficult for the NBS to access resources and TA from the Bank and other DPs. This could further put an additional strain on the NSS and hinder the sustainability of project outcomes. 107. In light of the above discussion, the risk to the development outcome is High. V. LESSONS AND RECOMMENDATIONS 108. Statistical capacity building takes a long time as it must effectively deal with both institutional and human capacity development. The Tanzania STATCAP Project sought to strengthen the capacity of the NSS to produce and disseminate high-quality data and statistics that meet users’ needs. The project focused on creating an enabling environment through institutional development and legal reform, strengthening organizational capacity, and developing human capacity along with data and statistical systems. However, the design team may have been overly optimistic to believe that the PDO could be achieved in five years. Some of the project’s results were outside the control of the NBS and the team. Future statistics project should focus on specific objectives that are achievable and within the purview of the national statistical offices and/or the team. 109. The implementation experience of this project supports the finding reported in a recent IEG evaluation that statistical capacity-building operations which seek to transform an NSS tend to have high transaction costs and are slow in producing intended results.5 These types of operations require a long- term engagement and an appropriate mix of instruments. The importance of institutional development also has implications for the sequencing of activities. The limited success of this operation is largely because of a failure to secure institutional and organizational reforms before capacity building and statistical development. 110. The World Bank might want to consider ways for making statistical capacity issues a permanent feature of policy dialogue. After all, effective development policy making requires credible information on key development issues facing a country. Problems with the availability, comparability, and quality of relevant data constitute foundational development constraints, which must be constantly addressed in the SCD and CPF. 111. An effective system-wide approach to statistical capacity building requires judicious targeting of the intervention within the NSS, which is a network of people and organizations involved in the production, dissemination, and use of statistics. Statistical output is an outcome of the interaction between the supply and demand sides of the system. Like most interventions, this STATCAP Project tends to emphasize capacity building on the supply side. There is a need for promoting in-country data sharing 5IEG (Independent Evaluation Group). 2017. Data for Development: An Evaluation of World Bank Support for Data and Statistical Capacity. Washington DC: World Bank. Page 29 of 52 The World Bank Development of a National Statistical System for Tanzania (P107722) and use. Such efforts must include building capacity on the demand side of the system and engaging academic and research institutions, NGOs, and private sector. 112. The project focused on data collection through surveys and supporting administrative data collection by MDAs. However, future statistics projects should consider a component on high-frequency statistics and integration from different sources. Satellite/remote sensing and use of computer-assisted personal interviewing technology could make the survey activity more efficient. Use of administrative datasets are a major untapped source of data for better policy making. In addition, high-frequency data would be instrumental in bridging the information gap between surveys and can be used for project M&E purposes. Challenges with the Project: Preparation and Implementation 113. The STATCAP Project was incepted in May 2008. However, the project preparation took several years, and it became effective only on August 9, 2011. The TA during project preparation was somewhat uncoordinated and the NBS had to deal with technical advisers coming from different institutions, at times, with inconsistent guidance. The preparation of future statistics project should coordinate TA provided by the World Bank and other DPs. It should be noted, that the TSMP basket fund was set up as a reaction to the lack of coordination and proved to be an effective vehicle to coordinate DP-supported activities in the statistics sector from 2011 onward. 114. Some of the operational challenges associated with the inadequate planning and inaccurate cost estimates could have been resolved by engaging procurement and financial management staff of both the NBS and the OCGS at the inception of the project. Initially, the OCGS office in Zanzibar was supposed to be a six-story building, but the budget allocated was only US$2 million. The final cost ended up being 100 percent higher for a four-story office building. . Page 30 of 52 The World Bank Development of a National Statistical System for Tanzania (P107722) ANNEX 1. RESULTS FRAMEWORK AND KEY OUTPUTS A. RESULTS INDICATORS A.1 PDO Indicators Objective/Outcome: NSS Developed to deliver reliable statistics in timely manner, in accordance with international standards. Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion National Geographical Text Village shape files Begin 2022 Population Begin 2022 Population EA shape files edited Information System produced and and Housing Census and Housing Census and used strengthened disseminated preparatory preparatory Preparatory geographic works EA geographic works geographic works for shape files are 2022 Population and produced and 2012 Housing Census has Census Atlas started disseminated. 30-Jun-2011 08-Apr-2015 30-Jun-2018 22-Jan-2018 Comments (achievements against targets): Target achieved. Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Economic Statistics Text National Accounts Quarterly Production Quarterly Production Construction of 2012 Page 31 of 52 The World Bank Development of a National Statistical System for Tanzania (P107722) strengthened Series based on 2007 of Producer Price of Producer Price Supply-Use Table Supply –Use Index (PPI) based on Index (PPI) based on (SUT) initiated. Table(SUT) 2012 basket. National 2012 basket. disseminated Accounts Series based on 2012 Supply –Use Table (SUT) National Accounts Series based on 2012 Supply –Use Table (SUT) 30-Jun-2011 08-Apr-2015 30-Jun-2018 22-Jan-2018 Comments (achievements against targets): Target achieved. Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Average number of visits in Text 1,700 per day 2000 2,000 per day 1,715 visits per day NBS and OCGS websites per day 30-Jun-2011 08-Apr-2015 30-Jun-2018 22-Jan-2018 Comments (achievements against targets): Target achieved. Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Increased user engagement Text None One User One User Forum Focus Group Forum/Focus Group Forum/Focus Group discussion organized discussion discussion and report under preparation. Page 32 of 52 The World Bank Development of a National Statistical System for Tanzania (P107722) 30-Jun-2011 30-Jun-2011 30-Jun-2018 22-Jan-2018 Comments (achievements against targets): Target achieved. Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Number of MDAs producing Text Six MDAs are Eight MDAs are Eight MDAs are 12 MDAs were new or improved statistics supported and supported and supported and supported by TSMP with TSMP support producing new or producing new or producing new or arrangement and improved statistic. improved statistic. improved statistic. producing new or improved Statistics during the reference period. 30-Jun-2011 29-Apr-2015 30-Jun-2018 22-Jan-2018 Comments (achievements against targets): Target achieved. A.2 Intermediate Results Indicators Component: Data Development and Dissemination Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Cumulative number of Number 0.00 18.00 18.00 23.00 surveys implemented under TSMP 30-Jun-2011 29-Apr-2015 30-Jun-2018 22-Jan-2018 Page 33 of 52 The World Bank Development of a National Statistical System for Tanzania (P107722) Comments (achievements against targets): Target achieved. Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Number of data sets Number 9.00 24.00 24.00 24.00 uploaded to the National Data Archive (NADA) 30-Jun-2011 29-Apr-2015 30-Jun-2018 22-Jan-2018 Comments (achievements against targets): Target achieved. This should be a PDO Indicator, however in the system it is entered as Intermediate indicator. Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Performance against Annual Percentage 0.00 72.00 72.00 75.00 Release Calendar 30-Jun-2011 29-Apr-2015 30-Jun-2018 22-Jan-2018 Comments (achievements against targets): Target achieved. Component: Physical Infrastructure and Equipment Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion New NBS and OCGS buildings Text Accommodation for Buildings construction Buildings construction Buildings construction NBS and OCGS is completed completed of OCGS new office Page 34 of 52 The World Bank Development of a National Statistical System for Tanzania (P107722) inadequate completed Commencement of Substructure Construction of NBS office building. 30-Jun-2011 30-Apr-2015 30-Jun-2018 22-Jan-2018 Comments (achievements against targets): Target achieved. Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion ICT equipment and vehicules Text ICT needs assessment Procurement of new Procurement of new LAN and WAN for OCGS in place ICT equipment in line ICT equipment in line equipment for NBS with ICT needs with ICT needs procured, installed assessment, four assessment, four and completed. Motor motor vehicles and 52 motor vehicles and 52 cycles procured. motor cycles motor cycles 30-Jun-2011 30-Jun-2011 30-Jun-2018 22-Jan-2018 Comments (achievements against targets): Target achieved. Component: Institutional Development and Legal Reform Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Legal framework for the NSS Text Statistics Act 2002 Statistics Bill 2013 Statistics Bill 2013 The end target has Page 35 of 52 The World Bank Development of a National Statistical System for Tanzania (P107722) with clear roles, does not give NBS approved by the approved by the been achieved in FY responsibilities, reporting professional government and government and 2014/15 lines and functions autonomy or role of submitted to submitted to developed coordinating NSS. parliament for parliament for Zanzibar Statistics Act enactment. enactment. 2007 gives OCGS professional autonomy and role of coordinating Zanzibar Statistical System. 30-Jun-2011 29-Apr-2015 30-Jun-2018 22-Jan-2018 Comments (achievements against targets): Target achieved. Component: Human Resource Capacity Development Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Reform of Organization Text Organization Reforms are Reforms are Organization structure Structure and Improved Structure, Scheme of implemented implemented was initially approved Human Resource service and Salary contingent on contingent on by the Treasury Management structures in need of enactment of 2013 enactment of 2013 Register (TR) reform Statistics Bill Statistics Bill Salary structure, scheme of service and staff regulations have been completed, but the Government suspended the reform Page 36 of 52 The World Bank Development of a National Statistical System for Tanzania (P107722) until the whole reform of public services is completed. 30-Jun-2011 30-Jun-2011 30-Jun-2018 22-Jan-2018 Comments (achievements against targets): Not achieved. This indicator was removed during 2015 restructuring. Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Annual staff training plan Percentage 0.00 75.00 80.00 61.00 implemented 30-Jun-2011 29-Apr-2015 30-Jun-2018 22-Jan-2018 Comments (achievements against targets): Partially achieved. Component: Development of Statistical Infrastructure Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Updated statistical registers Text Tanzania Mainland: Up to date sampling Up to date sampling Final Statistical and sample frames for Statistical Business frames, Statistical frames, Statistical Business Register household and establishment Register (SBR) Business Register Business Register (SBR) report is in place based surveys updated 2007. (SBR) and National (SBR) and National and uploaded in the Master Sample (NMS). Master Sample (NMS). NBS’s website The National Master Sample (NMS) is still under preparation Page 37 of 52 The World Bank Development of a National Statistical System for Tanzania (P107722) 30-Jun-2011 30-Jun-2011 30-Jun-2018 22-Jan-2018 Comments (achievements against targets): This is one of the TSMP indicators, however it was not one of the intermediate indicators for the project. Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Implemented Quality Text No quality assessment Quality assessment of Quality assessment of Quality assessment of Assessment Framework in place. three sectoral three sectoral three sectoral compendia conducted compendia conducted compendia conducted and assessment and assessment and assessment report/certificates report/certificates report/certificates Concepts definitions issued issued issued. The target was and methodologies met during FY 2015- outdated. 16. No or little documentation of business processes. 30-Jun-2011 29-Apr-2015 30-Jun-2018 22-Jan-2018 Comments (achievements against targets): Target achieved. Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Formally Revised Completion Page 38 of 52 The World Bank Development of a National Statistical System for Tanzania (P107722) Target Legal framework for the NSS Text Statistics Act 2002 Statistics Bill 2013 The end target has with clear roles, does not give NBS approved by the been achieved in FY responsibilities, reporting professional government and 2014/15 lines and functions autonomy or role of submitted to developed coordinating NSS. parliament for Zanzibar Statistics Act enactment. 2007 gives OCGS professional autonomy and role of coordinating Zanzibar Statistical System. 30-Jun-2011 30-Jun-2018 22-Jan-2018 Comments (achievements against targets): Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Reform of Organization Text Organization Reforms are Organization structure Structure and Improved Structure, Scheme of implemented was initially approved Human Resource service and Salary contingent on by the Treasury Management structures in need of enactment of 2013 Register (TR) reform Statistics Bill Salary structure, scheme of service and staff regulations have been completed, but Page 39 of 52 The World Bank Development of a National Statistical System for Tanzania (P107722) the Government suspended the reform until the whole reform of public services is completed. 30-Jun-2011 30-Jun-2018 22-Jan-2018 Comments (achievements against targets): Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Annual staff training plan Percentage 0.00 0.00 80.00 61.00 implemented 30-Jun-2011 30-Jun-2018 22-Jan-2018 Comments (achievements against targets): Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Updated statistical registers Text Tanzania Mainland: Up to date sampling Final Statistical and sample frames for Statistical Business frames, Statistical Business Register household and establishment Register (SBR) Business Register (SBR) report is in place based surveys updated 2007. (SBR) and National and uploaded in the Master Sample (NMS). NBS’s website Page 40 of 52 The World Bank Development of a National Statistical System for Tanzania (P107722) The National Master Sample (NMS) is still under preparation 30-Jun-2011 30-Jun-2018 22-Jan-2018 Comments (achievements against targets): Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Implemented Quality Text No quality assessment Quality assessment of Quality assessment of Assessment Framework in place. three three sectoral sectoral compendia compendia conducted conducted and and assessment assessment report/certificates Concepts definitions report/certificates issued. The target was and methodologies issued met during FY 2015- outdated. 16. No or little documentation of business processes. 30-Jun-2011 30-Jun-2018 22-Jan-2018 Comments (achievements against targets): Page 41 of 52 The World Bank Development of a National Statistical System for Tanzania (P107722) Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Cumulative number of Number 0.00 0.00 18.00 23.00 surveys implemented under TSMP 30-Jun-2011 30-Jun-2018 22-Jan-2018 Comments (achievements against targets): Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Number of data sets Number 9.00 0.00 24.00 24.00 uploaded to the National Data Archive (NADA) 30-Jun-2011 30-Jun-2018 22-Jan-2018 Comments (achievements against targets): Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion Performance against Annual Percentage 0.00 0.00 72.00 75.00 Release Calendar 30-Jun-2011 30-Jun-2018 22-Jan-2018 Page 42 of 52 The World Bank Development of a National Statistical System for Tanzania (P107722) Comments (achievements against targets): Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion New NBS and OCGS buildings Text Accommodation for Buildings construction Buildings construction NBS and OCGS is completed of OCGS new office inadequate completed Commencement of Substructure Construction of NBS office building. 30-Jun-2011 30-Jun-2018 22-Jan-2018 Comments (achievements against targets): Formally Revised Actual Achieved at Indicator Name Unit of Measure Baseline Original Target Target Completion ICT equipment and vehicules Text ICT needs assessment Procurement of new LAN and WAN for OCGS in place ICT equipment in line equipment for NBS with ICT needs procured, installed assessment, four and completed. Motor motor vehicles and 52 cycles procured. Page 43 of 52 The World Bank Development of a National Statistical System for Tanzania (P107722) motor cycles 30-Jun-2011 30-Jun-2018 22-Jan-2018 Comments (achievements against targets): Page 44 of 52 The World Bank Development of a National Statistical System for Tanzania (P107722) B. KEY OUTPUTS BY COMPONENT Objective 1. NSS Developed to deliver reliable statistics in timely manner, in accordance with international standards. 1. National Geographical Information System strengthened 2. Economic Statistics strengthened Outcome Indicators 3. Number of data sets uploaded to the National Data Archive (NADA) 1. Implemented Quality Assessment Framework 2. Cumulative number of surveys implemented under TSMP 3. Performance against Annual Release Calendar 4. Average number of visits in NBS and OCGS websites per day Intermediate Results Indicators 5. Legal framework for the NSS with clear roles, responsibilities, reporting lines and functions developed 6. Annual staff training plan implemented 7. New NBS and OCGS buildings Component A: 1. Statistics Bill 2013 approved by the government and submitted to parliament for enactment. Component B: 1. Organization structure was initially approved by the Treasury Register. Salary structure, scheme of service and staff regulations have been completed, but the Government suspended the reform Key Outputs by Component until the whole reform of public services is completed. (linked to the achievement of the Objective/Outcome 1) 2. Most of the trainings were implemented in Zanzibar, benefitting to OCGS staff, while trainings for NBS staff were below the targets in the training calendar due to the prioritization of other activities in the AWP and problems of planning trainings by international experts and checking their availability well in advance. However, NBS staff members received on-the job training during survey preparation and implementation as part of TA. Page 45 of 52 The World Bank Development of a National Statistical System for Tanzania (P107722) Component C. 1. Quality assessment of three sectoral compendia conducted and assessment report/certificates issued. The target was met during FY 2015-16. 2. Construction of 2012 Supply-Use Table (SUT) initiated. Component D. 1. Preparatory geographic works for 2022 Population and Housing Census has started 2. 24 data sets are uploaded to the NADA. 3. Agriculture and establishment related surveys and censuses conducted. Component E. 1. The construction of new office space for the Office of the Chief Government Statistician of Zanzibar was completed in January, 2018. The construction of NBS building in Dodoma has was completed in May 2018. Page 46 of 52 The World Bank Development of a National Statistical System for Tanzania (P107722) ANNEX 2. BANK LENDING AND IMPLEMENTATION SUPPORT/SUPERVISION A. TASK TEAM MEMBERS Name Role Preparation Thomas Danielewitz Task Team Leader, Statistician Manager World Development Neil Fantom Indicators Lucas Ojiambo Consultant Johannes Hoogeveen Senior Economist Johan Mistiaen Senior Economist/Statistician Vildan Verbeek- Demiraydin Senior Economist Jane Kibbassa Senior Environmental Specialist Philip Beauregaard Senior Counsel Luis M. Schwarz Senior Finance Officer Donald Paul Mneney Senior Procurement Specialist Gisbert Joseph Kinyero Procurement Specialist Senior Financial Management Bella Lelouma Diallo Specialist Senior Financial Management Mercy Mataro Sabai Specialist Supervision/ICR Nadia Belhaj Hassine Belghith, Elizabeth Ann Task Team Leader(s) Talbert Raymond Joseph Mbishi, Winter M. Chinamale Procurement Specialist(s) Michael Eriu Okuny Financial Management Specialist Lucas A. Ojiambo Team Member Mary C.K. Bitekerezo Social Safeguards Specialist Senait Kassa Yifru Team Member Martin Pedro Buchara Team Member Grace Anselmo Mayala Team Member Emmanuel E. Hanai Environmental Safeguards Specialist Page 47 of 52 The World Bank Development of a National Statistical System for Tanzania (P107722) Jessica Diane Adler Team Member Neema Clarence Mkundi Team Member John Ainainyi Njau Team Member Anahit Poghosyan Team Member Essama Nssah Team Member Habtamu Neda Fuje ICR Task Team Leader B. STAFF TIME AND COST Staff Time and Cost Stage of Project Cycle No. of staff weeks US$ (including travel and consultant costs) Preparation FY09 48.147 212,677.19 FY10 40.160 172,689.26 FY11 45.120 185,578.40 FY12 0 0.00 Total 133.43 570,944.85 Supervision/ICR FY12 30.149 118,052.04 FY13 50.818 77,045.45 FY14 57.785 136,313.17 FY15 56.432 80,393.97 FY16 60.795 203,313.68 FY17 36.799 83,936.65 FY18 15.147 135,308.94 FY19 9.424 71,366.31 Total 317.35 905,730.21 Page 48 of 52 The World Bank Development of a National Statistical System for Tanzania (P107722) ANNEX 3. PROJECT COST BY COMPONENT Amount at Approval* Actual at Project* Percentage of Components (US$M) Closing (US$M) Approval (US$M) A. Institutional 1.7 0.2 6% Development and Legal Reform B. Human Resource 5 1.0 26% Capacity Development C. Development of 8.6 5.4 59% Statistical Infrastructure D. Data Development 11.7 19.0 134% and Dissemination Physical Infrastructure 14.7 11.1 39% and Equipment Project Management 3.4 1.9 76% Total 45.1 38.6 68% *Amounts in this table include basket funding and do not include Government contribution. Funding Source Actual Spending PAD Allocation DFID 8,067,107 10,000,000 IDA 27,273,171 30,000,000 GAC 4,457,724 4,900,000 Total Amount 39,798,002 44,900,000 Source: Project International Financial Reporting Standards (IFR), June 2018. Note: The project was financed through basket fund with contributions from the Canadian International Development Agency and DFID. Initial allocations were USD 10.0 million from DFID and USD 4.9 million from CIDA, however, due to currency exchange fluctuations the amounts at project closing were US$8 million from DFID and US$4.4 million from CIDA. Page 49 of 52 The World Bank Development of a National Statistical System for Tanzania (P107722) ANNEX 4. EFFICIENCY ANALYSIS Not Applicable. Page 50 of 52 The World Bank Development of a National Statistical System for Tanzania (P107722) ANNEX 5. BORROWER, CO-FINANCIER AND OTHER PARTNER/STAKEHOLDER COMMENTS Page 51 of 52 The World Bank Development of a National Statistical System for Tanzania (P107722) ANNEX 6. SUPPORTING DOCUMENTS (IF ANY) None. Page 52 of 52