Document of The World Bank FOR OFFICIAL USE ONLY Report No: ICR00004764 IMPLEMENTATION COMPLETION AND RESULTS REPORT ON A SMALL GRANT IN THE AMOUNT OF USD 4.8 MILLION TO THE REPUBLIC OF LIBERIA FOR LIBERIA - STRENGTHENING THE NATIONAL STATISTICS SYSTEM - HOUSEHOLD INCOME EXPENDITURE SURVEY (P145709) MARCH 26, 2019 Poverty And Equity Global Practice Africa Region ABBREVIATIONS AND ACRONYMS AfDB African Development Bank AfT Agenda for Transformation CPI Consumer Price Index EU European Union GDP Gross Domestic Product GIS Geo Information System GoL Government of Liberia HIES Household Income Expenditure Survey IDA International Development Agency IMF International Monetary Fund LISGIS Liberia Institute of Statistics and Geo-Information Services MDTF Multi-Donor Trust Fund MFDP Ministry of Finance and Development Planning MoFA Ministry of Foreign Affairs NAAS National Accounts Annual Survey NEC National Establishment Census NSSP National Statistics Strengthening Project PAPD Pro-poor Agenda for Prosperity and Development PDO Project Development Objectives PIU Project Implementation Unit SDG Sustainable Development Goals SIDA Swedish International Development SUT Supply and Use Table TFSCB Trust Fund for Statistical Capacity Building USAID United States Agency for International Development WB World Bank Regional Vice President: Hafez M. H. Ghanem Country Director: Henry Kerali Senior Global Practice Director: Carolina Sanchez-Paramo Practice Manager: Andrew L. Dabalen Task Team Leader(s): Daniel Kwabena Boakye, Ayago Esmubancha Wambile ICR Main Contributor: Daniel Kwabena Boakye, Ayago Esmubancha Wambile TABLE OF CONTENTS DATA SHEET ....................................................................... ERROR! BOOKMARK NOT DEFINED. I. PROJECT CONTEXT AND DEVELOPMENT OBJECTIVES ....................................................... 4 II. OUTCOME .................................................................................................................... 12 III. KEY FACTORS THAT AFFECTED IMPLEMENTATION AND OUTCOME ................................ 12 IV. BANK PERFORMANCE, COMPLIANCE ISSUES, AND RISK TO DEVELOPMENT OUTCOME .. 13 V. LESSONS LEARNED AND RECOMMENDATIONS .............................................................. 13 ANNEX 1. RESULTS FRAMEWORK AND KEY OUTPUTS ........................................................... 15 ANNEX 2: ASSESSMENT OF OUTCOMES ............................................................................... 19 The World Bank Liberia - Strengthening the National Statistics System - Household Income Expenditure Survey (P145709) DATA SHEET BASIC INFORMATION Product Information Project ID Project Name Liberia - Strengthening the National Statistics System - P145709 Household Income Expenditure Survey Country Financing Instrument Liberia Investment Project Financing Original EA Category Revised EA Category Organizations Borrower Implementing Agency Liberia institute of Statistics and Geo-Information Ministry of Finance and Development Planning Systems (LISGIS) Project Development Objective (PDO) Original PDO The objective of the Project is to improve the capacity of the national statistics system in the provision of national accounts, price and poverty data and baseline indicators for the medium-term development strategy. Page 1 of 20 The World Bank Liberia - Strengthening the National Statistics System - Household Income Expenditure Survey (P145709) FINANCING FINANCE_TBL Original Amount (US$) Revised Amount (US$) Actual Disbursed (US$) Donor Financing TF-16277 3,559,483 3,279,652 3,279,652 TF-A0983 500,000 500,000 500,000 TF-A7007 458,350 458,350 458,350 TF-A7061 324,870 324,870 324,870 Total 4,842,703 4,562,872 4,562,872 Total Project Cost 4,842,703 4,562,872 4,562,872 KEY DATES Approval Effectiveness Original Closing Actual Closing 10-Jul-2013 31-Jan-2014 30-Sep-2018 30-Sep-2018 RESTRUCTURING AND/OR ADDITIONAL FINANCING Date(s) Amount Disbursed (US$M) Key Revisions 19-Oct-2017 3.58 Change in Results Framework Change in Loan Closing Date(s) Change in Implementation Schedule 14-May-2018 3.78 Additional Financing Change in Loan Closing Date(s) Change in Implementation Schedule KEY RATINGS Outcome Bank Performance M&E Quality Moderately Satisfactory Moderately Satisfactory Substantial Page 2 of 20 The World Bank Liberia - Strengthening the National Statistics System - Household Income Expenditure Survey (P145709) RATINGS OF PROJECT PERFORMANCE IN ISRs Actual No. Date ISR Archived DO Rating IP Rating Disbursements (US$M) 01 24-Aug-2017 Satisfactory Satisfactory 3.43 ADM STAFF Role At Approval At ICR Regional Vice President: Makhtar Diop Hafez M. H. Ghanem Country Director: Yusupha B. Crookes Henry G. R. Kerali Senior Global Practice Director: Marcelo Giugale Carolina Sanchez Practice Manager: Pablo Fajnzylber Andrew L. Dabalen Daniel Kwabena Boakye, Ayago Task Team Leader(s): Daniel Kwabena Boakye Esmubancha Wambile ICR Contributing Author: Senait Kassa Yifru Page 3 of 20 The World Bank Liberia - Strengthening the National Statistics System - Household Income Expenditure Survey (P145709) I. PROJECT CONTEXT AND DEVELOPMENT OBJECTIVES Context 1. The National Statistics Strengthening Project (NSSP) was a Multi-Donor funded project which was approved on July 10, 2013 and implemented over 5 year period upto September, 2018. The project was financed by two sources: (i) the Multi-Donor Trust Fund for National Statistics System Strengthening (MDTF NSSS); and (ii) the Trust Fund for Statistical Capacity Building (TFSCB) with the objective of improving the capacity of Liberia’s statistics system in provision of national accounts, price and poverty data, and baseline indicators for the Agenda for Transformation (AfT). The following surveys were conducted under the project; (i) the Household Income & Expenditure Survey (HIES 2014), (ii) the Second Household Income & Expenditure Survey (HIES 2016), (iii) Crop Cutting Survey-2014 &2015, (iv) Agriculture Recall Survey 2016, (v) the National Establishment Census (2017) and (vi) the National Accounts Annual Survey (NAAS 2018). Following the successful completion of these key surveys/census, the Liberia Institute of Statistics & Geo-Information Services (LISGIS) was able to achieve the key development objectives of the project outcome , including a detailed poverty analysis for both HIES 2014 and 2016, the construction of the Consumer Price Index, an update of the National Accounts and the estimation of socio-economic indicators to track the progress in the implementation of the country’s development plan : “Agenda for Transformation (AfT)- 2013-2017�. 2. Additionally, the HIES 2016, provided detailed poverty, employment, health, gender and agricultural analyses which served as baseline information for the “Pro-Poor Agenda for Prosperity and Development (PAPD)�, which was launched by the Government of Liberia in October 2018, and provided benchmarks for tracking the implementation of the Sustainable Development Goals (SDGs). Other components of this project included capacity building and cross-country knowledge sharing, alongside efforts to improve survey methodologies in Liberia. 3. The project was implemented by Liberia Institute of Statistics and Geo-Information Services (LISGIS), with support from Government of Liberia (GoL), and four main donors; United States Agency for International Development (USAID), European Union (EU), World Bank (WB), African Development Bank (AfDB) and Swedish International Development Agency (SIDA). Project Development Objectives (PDOs) 4. The objective of the Project was to improve the capacity of the national statistics system in the provision of national accounts, price and poverty data, and baseline indicators for the medium-term development strategy. Page 4 of 20 The World Bank Liberia - Strengthening the National Statistics System - Household Income Expenditure Survey (P145709) Key Expected Outcomes and Outcome Indicators 5. Achievement of the overall development objective was to be assessed through the following PDO level indicators: (i) Poverty Status of Liberia based on a 12-month HIES and the development of baseline Indicators for monitoring the implementation of the Country's Development Agenda- AfT (2013-2017). (ii) Development of a new CPI basket based on the consumption data from the Household Income and Expenditure Survey data. (iii) Development of benchmark data for the compilation of national accounts statistics (iv) Baseline indicators for the Agenda for Transformation (i) Components (as approved) 6. The project was structured under three Components. Component I: Support to the 2013 HIES • Sub-component 1(a): Preparation of the HIES • Sub-component 1(b): Training and Field Survey • Sub-component 1(c): Data Management • Sub-component 1(d): Support to the HIES Data Analysis and Dissemination Component II: Strengthening capacity of statistical users and producers to foster long-term Sustainability • Sub-Component 2(a): Capacity building for statistical users in carrying out evidence-based policy analysis; • Sub-Component 2(b): Capacity building for producers of statistics. Component III: Project Management • This component supported operating costs for the project management. The project management team comprised the following: (a)International Technical Advisor, (b) financial management services carried out by the Project Financial Management Unit at the Ministry of Finance, and (c) an international procurement specialist. 7. Descriptions of project achievements by components and sub-components: Page 5 of 20 The World Bank Liberia - Strengthening the National Statistics System - Household Income Expenditure Survey (P145709) Key Expected Outcomes and Outcome Indicators Component I: Support to the 2013 HIES 8. Household Income & Expenditure Survey (HIES) 2014: In 2014, LISGIS implemented the first Household Income and Expenditure Survey since 1964. The survey was designed such that fieldwork would be conducted over a period of 12 months, to capture the effect of seasonality on household incomes and consumption. The sample design for the HIES utilized a two-stage clustered sampling technique, comprising of a nationally representative sample of households every quarter and was obtained using the 2008 National Housing and Population Census sampling frame. The sample size, set at 8,350 households from 835 enumeration areas was deemed sufficient for also providing disaggregates at the county level and by urban/rural strata. 9. The questionnaire and survey tools were all prepared in-house at LISGIS through extensive consultations with various stakeholders such as line ministries and agencies, development partners and NGOs. Fieldwork officially started on January 26, 2014 with twelve teams deployed across Liberia. Frequent monitoring and field based first data entry ensured that the quality of data is high. 10. Unfortunately, due to the Ebola Virus outbreak, data collection activities had to be halted in August 2014, which meant that the full data collection for 12 months could not be completed. Fortunately, 409 out of 418 enumeration areas were completed in the first two quarters. This was deemed to be a sufficient sample size to obtain national level estimates and breakdown by urban/rural strata and by region for key indicators. The sample size, however, could not provide county level estimates as originally planned. 11. Given the lack of nationally-representative agricultural data in Liberia, a comprehensive agricultural module was incorporated. After extensive discussions between LISGIS, the Ministry of Agriculture and the World Bank, a Crop Cutting Survey was launched in July 2014. The goal was to compare estimates from a Crop Cutting Survey to a Recall-Based Agricultural Survey, which would be implemented in January 2015. Unfortunately, the work was halted due to the Ebola Virus outbreak and data is missing for the north-western counties. 12. First Data Entry took place on the field. Each team was equipped to enter, verify and troubleshoot information collected by the enumerators. Second Data Entry took place at LISGIS headquarters. Data cleaning was conducted in collaboration with the World Bank poverty team. A data validation and training workshop took place towards the end of the cleaning process where the World Bank and LISGIS verified all input. 13. Analysis activities began after the data was cleaned. LISGIS spearheaded analytical activities through coordination of thematic experts and technical partners, and benefited through various World Bank missions, intended to raise capacity of LISGIS staff in such areas as sampling and weighting procedures, National Accounts and poverty analysis. The focus in each component of the analysis stage of the HIES was on capacity building and transfer of knowledge so that LISGIS can continue with such work independently in the future. The survey weights were constructed by a sampling expert using background Page 6 of 20 The World Bank Liberia - Strengthening the National Statistics System - Household Income Expenditure Survey (P145709) information provided by LISGIS on survey non-response and replacements. Interim CPI weights and basket were constructed using the six month HIES data despite excluding seasonal impacts on consumption. 14. National Accounts was also updated using the HIES 2014 data; the benefit of adding household expenditure information and data on informal sector was substantial, given that both were not in the System of National Accounts. The previous series from 2008 to 2013 has also been revised based on 2010 prices and to incorporate a back-cast estimates of the household contribution to expenditure based on HIES 2014 data. However, given the limitations of the 2014 HIES to incorporate all seasonal effects for the entire 2014 year, estimates from the HIES 2014 could not be fully use for updating GDP estimates. 15. Several indicators within the monitoring framework for the Agenda for Transformation have been constructed successfully from the survey. Finally, the data has been used to conduct Poverty Analysis. Preparation of final reports and data for publication and dissemination was carried out from October 2015 through March 2016. The results were accepted by the President of Liberia and all key stakeholders. 16. LISGIS has successfully completed the HIES 2014 and made the results publicly available, including the microdata, reports, and other dissemination materials such as power point presentations, flyers and 2-page briefing notes. The national launch of HIES 2014 key results was carried out in April 2016 during the National Development Summit in Gbarnga, Bong County. Other dissemination activities carried out included radio talk shows, workshops, press briefings, dialogues with local communities, stakeholders’ meeting, digital publications, and hard copies. 17. Household Income & Expenditure Survey (HIES) 2016 Re-run: After deliberations between LISGIS, the GoL, and the development partners, it was agreed that a full 12-month rerun of the HIES would be required to meet the objectives originally set out for the HIES 2014. The HIES 2016 aimed to: capture impacts of seasonality on consumption data; to construct the CPI basket and weights and calculation of poverty numbers; to allow for county level estimates to be obtained for key indicators including poverty; and for the collection of comprehensive nationally representative agricultural statistics. 18. As such, resources were pooled to provide supplementary funding for the HIES re-run to begin in 2016. The re-run uses the same sample methodology and design as the 2014 survey, which allows for nationally representative data to be collected each quarter, and inclusive of all twelve months, and enough observations for key indicators to be estimated at the county level. The agriculture module of the HIES encompassed the use of two methods, the crop-cutting method and the farmer recall method, allowing for comparison between the two methods and for the results to inform future best practices in agricultural data collection in Liberia. The Agriculture Recall questionnaire was administered alongside the Household questionnaire to reduce the number of visitations to a household, reduce attrition, and to take advantage of cost sharing through implementation of two questionnaires in the same set of logistics. 19. A comprehensive recruitment and rigorous training process was completed and all field teams were deployed on 14th January 2016. Field teams successfully collected data for the full year, ending on 15th January 2017. Six field monitoring trips were completed, during which LISGIS staff trained in the questionnaire, GIS and data entry activities. The field team visited all teams to assure the quality of their work. Results of the monitoring trips were used to inform decision-making regarding logistics for the following quarter as well as to provide further feedback training during the monitoring. Page 7 of 20 The World Bank Liberia - Strengthening the National Statistics System - Household Income Expenditure Survey (P145709) 20. Additional data entry clerks who had been pre-trained in November were hired to enter data a second time beginning in March 2016 and finalized by March 2017 in LISGIS Headquarters. Post-data collection activities began early; while data collection was still on-going, draft cleaning codes were written for all modules of all questionnaires allowing for the six-month analysis work to begin ahead of time, and the twelve-month cleaning to be completed within a reduced timeframe. 21. The World Bank poverty team finished working with LISGIS on producing poverty estimates based on six months of 2014 survey data and twelve months of 2016 survey data. Production of the Statistical Abstract was completed in July 2017. The approval process from governmental stakeholders was completed by February 2018, unfortunately, delayed by the national elections and government transition. National dissemination activities started in April, with the combined Executive and National launch of all publications and results on the 9th of April at the City Hall ball room. This major event was attended by key government officials, donor representatives, members of civil society and the general public. Over 150 persons congregating in the city hall ball room with a cross section of dignitaries from government and different ambassadorial representatives. On the platform was Deputy Minister for fiscal Affairs MFDP, Deputy Minister for institutional and man power development in the MOL and the delivery of the HIES launch address as proxy on behalf of the President by the Deputy Minister at the Ministry of Foreign Affairs (MoFA). 22. The LISGIS Director General, presented an overview on the HIES. A host of representatives from the donor community including student leaders from universities and high schools. Moreover, local leaders from community and faith- based organizations also attended the event. Media representatives from lead media institutions like Truth FM, Liberia Broadcasting Corporation (LBC) and Fabric Radio, as well as 10 local print and electronic institutions covered the event. The deputy director general for LBCwas also present at the program. Initially there were two talk shows on Truth FM and LBC signaling the launch of the HIES national dissemination exercise. 23. Furthermore, through the project, LISGIS staff were able to attend international trainings and conferences especially in the sub-regions to share and gather knowledge. Examples of trainings attended include, a short sampling methods course in Tanzania and the Africa Open Data Conference in Accra, Ghana. 24. Agriculture Crop Cutting and Recall 2014, 2015, 2016 (Funded under the TFSCB): Given the lack of nationally-representative agricultural data in Liberia, a comprehensive agricultural module was incorporated. A Crop Cutting Survey was launched in July 2014 which was interrupted by Ebola Virus Outbreak but was continued in 2015. 25. In 2016, the agriculture module of the HIES utilised the recall methodology to capture household agriculture information. The two methodologies, namely crop cutting and recall, were later compared in a comparative analysis report. The Agriculture Recall questionnaire was administered alongside the Household questionnaire to reduce the number of visitations to a household, reduce attrition, and to take advantage of cost sharing through implementation of two questionnaires in the same set of logistics. Page 8 of 20 The World Bank Liberia - Strengthening the National Statistics System - Household Income Expenditure Survey (P145709) 26. National Accounts: The National Accounts Section, under the Economics Statistics Division, was responsible for compiling macroeconomic statistics, such as GDP estimates, to provide a clearer picture of the economy for evidence-based fiscal planning. 27. The last fully rebased GDP figures were produced in 1984. Due to civil war in the intervening years, there has been no formal update for the last three decades. In 2010, there was an attempt to rebase the GDP through implementing the National Economic Census (NEC) and National Accounts Annual Survey (NAAS) and collecting administrative data for compilation. LISGIS was, however, advised to include the results from the 2014 and 2016 HIES before publication of the GDP numbers. As such, estimates from 2008-2013 have been produced but are awaiting more in-depth data and further confirmation. The production of the 2014 to 2016 estimates, based on administrative data compilation, are currently underway with technical support from the International Monetary Fund (IMF). 28. With the completion of the HIES 2014 and 2016, the parts of the GDP related to household final consumption expenditure, the informal sector, agriculture, owner occupied dwellings, and other estimates were fully updated. In addition, the CPI basket will also be re-weighted by the HIES 2016 data and the latest estimates will be included. The CPI basket is currently weighted using the 2014 data. National Establishment Census (NEC) 2017 : The National Establishment Census (NEC), funded completely by USAID independently of the MDTF or TFSCB, was successfully completed between July and September 2017. In total, data was collected from about 18,000 establishments throughout the country. Data cleaning for the NEC was completed. In particular, the classification of economic activities by industries was completed using the International Standard Industrial Classification of All Economic Activities, Rev.4 (ISIC Rev. 4), in line with the practices at Liberia Business Registry and Liberia Revenue Authority. The NEC 2017 was used as a sampling frame for the National Accounts Annual Survey (NAAS) to be conducted in 2018. 29. Ultimately, LISGIS seeks to produce annual and quarterly GDP numbers by all three methods of production, expenditure, and income. For this process to be efficient, compilation of administrative data for external trade, government consumption, domestic revenue, balance of payments, production of agriculture, fisheries, forestry, CPI, and other topics must be systematized and streamlined. In addition, the GDP should be rebased every five years to ensure accuracy of the estimates produced. While it is recommended that the National Accounts Annual Survey (NAAS) is completed annually. If that is not possible, then with each rebase, the NAAS must be completed. 30. LISGIS aims to rebase Liberia’s GDP to 2016, using data from the Household Income and Expenditure Survey 2014-2016, the Labor Force Survey 2016, National Population and Housing Census 2008, the National Accounts Annual Survey 2018, and administrative data from line Ministries and Agencies. 31. National Accounts Annual Survey (NAAS) 2018: The NAAS is foundational to rebasing the GDP and is one of the main surveys Liberia uses to capture data from key economic players stratified by industry and size. The survey results will be used to construct a Supply and Use Table (SUT), calculate GDP estimates, and improve on past GDP estimates through back-casting. Page 9 of 20 The World Bank Liberia - Strengthening the National Statistics System - Household Income Expenditure Survey (P145709) 32. NAAS preparation started in January and field deployment commenced in May 2018. Field work was half-completed by the end of June. It is predicted that the NAAS field world will finish mid of August. Follow up work on establishments that refused interviews is in progress. The National Accounts team at LISGIS worked closely with consultants from the World Bank and IMF to ensure the survey design and implementation is according to international standards. A national sample of 3000 businesses was enumerated, with particular focus on large- and medium-size companies. Component II: Strengthening capacity of statistical users and producers to foster long-term Sustainability 33. Significant capacity has been built at LISGIS at various stages of the survey cycle. Through technical assistance missions, workshops, video-conferences and constant communication between local staff and development partners, tangible skills were transferred including sampling methodologies, questionnaire design, CSpro programming, GIS mapping, STATA coding and analysis, and report writing skills. International standards were followed throughout the survey and LISGIS has successfully conducted its first twelve-months survey; a significant milestone in the development of statistics in Liberia. 34. A big part of the project is the dissemination of the survey findings in order to foster adoption and integration of evidence-based policy making. In 2014 this was done extensively, not only with Ministries but also with the University of Liberia and with the general public in order to raise awareness on the socio- economic indicators. Attention to the dissemination of the information is important to increase future demand and end-use of surveys and their findings. The HIES 2014 and 2016 provided a wealth of data to report of the previous development plan, the Agenda for Transformation. Additionally, the HIES 2016 is acting as an important baseline in the development of the new Pro-Poor Agenda for Prosperity and Development. Component III: Project Management 35. Project Financial Management Unit at the Ministry of Finance carried out financial management and fiduciary risk management of the project for the whole implementation period. A procurement specialist and finance officer were hired to assist in the implementation of the project. LISGIS staff promoted to Project Coordinator and support staff have benefitted in improving their management skills in delivering a sizable project such as the HIES. Changes During Implemenation 36. The Liberia Strengthening the National Statistics System project (P145709) was approved on July 10, 2013. The project had Satisfactory project development and implementation ratings. The project was Page 10 of 20 The World Bank Liberia - Strengthening the National Statistics System - Household Income Expenditure Survey (P145709) financed by two sources: (i) the Multi-Donor Trust Fund for National Statistics System Strengthening (MDTF NSSS); and (ii) the Trust Fund for Statistical Capacity Building (TFSCB). The original end- disbursement date of the NSSS MDTF was June 30, 2016. An 18-month extension was processed to extend the end-disbursement date to December 31, 2017. The extension was required to enable the Liberia Institute of Statistics & Geo-Information Services (LISGIS) to rerun the 2014 Household Income and Expenditure Survey (HIES), which was abruptly halted because of the Ebola Virus Disease (EVD) Outbreak in March 2014. It took Liberia almost a year and half to recover from the Ebola Outbreak. Consequently, the HIES rerun was started in 2016. 37. A second extension of the end-disbursement date from December 31, 2017 to December 31, 2018, was required in order to utilize the remaining trustee balance of $455,000 to complete the following activities; (i) To conduct the National Accounts Annual Survey (NAAS) exercise by March 2018, (ii) Dissemination of the HIES findings, and (iii) To continue the capacity building for staff who were committed under the project funding. These activities had to be postponed to the first half 2017, due to the intensity of political activities leading to the general and presidential elections on the October 10th 2017 and the post-election litigation at the supreme court. Table 1. Project Financing Breakdown Allocation by Original Amount Revised Amount Actual Disbursed Source of Funding (US$) (US$) (US$) TF-16277 (MDTF) 3,559,483 3,279,652 3,279,652 TF-A0983 (TFSCB) 500,000 500,000 500,000 TF-A7007 (MDTF) 458,350 458,350 458,350 TF-A7061 (MDTF) 324,870 324,870 324,870 Total 4,842,703 4,562,872 4,562,872 Changes to Original Procurement Plan 38. The only change introduced in implementation was the re-run of the HIES in 2015/2016 that was as a result of the halting of data collection activities for the 2014 HIES due to the ebola virus disease. The full 12-months rerun of the HIES was agreed by the Government of Liberia and Development partners. Details of the agreement included the recruitment of field staff, the recruitment of an international CSPro Expert to provide trainings to field staff, and trainings as well as stationery and an assortment of other supplies for the rerun. A total of 14 teams were deployed, each consisting of a driver, a supervisor, a data entry clerk, 4 enumerators, a GIS Specialist all of which were reflected in the Procurement Plan as agreed. Page 11 of 20 The World Bank Liberia - Strengthening the National Statistics System - Household Income Expenditure Survey (P145709) II. OUTCOME Assessment of Achievement of Each Objective/Outcome 39. The project ensured that new national accounts is estimated annually based on the Establishment Survey, the NA benchmark and the HIES; CPI calculations based on the revised weight for computing the basket & weights from HIES; Key poverty statistics estimated from the HIES. Overall the PDO was achieved effectively (see Anex 2 for detail descriptions of PDO and Intermediate indicatorsachievements. Overall Outcome Rating 40. The outcome rating is moderately satisfactory. Other Outcomes and Impacts 41. The funding of the project provided leverage for the World Bank to engage in many current and future in-depth technical discussions on data collection and analysis. The project also built the foundation for obtaining comparable poverty estimates in Liberia having collected year-long data in 2016. The project enabled the mobilization of IDA resources for statistical capacity building, data production and dissemination activities. The lesson learnt are also useful for the preparation of the future projects. III. KEY FACTORS THAT AFFECTED IMPLEMENTATION AND OUTCOME 42. There were some factors that affected the implementation of the project. 43. Throughout the project, access to liquid funds and delays in payment caused significant constraints in the smooth and timely running of the project. This encompasses payments for field-staff, data-entry staff, and available funds for procuring required project items 44. High turnover of staff required interim handover periods, which would create time lags and additional administrative work for the project team. The handing over of documentation, adjustment of the new staff and hiring process added to the workload of the project management team. 45. Difficult field environments including poor road conditions, extreme rains and hard to reach locations. 46. Significant delays leading to significant incomplete activities at the original closing date of the project. Page 12 of 20 The World Bank Liberia - Strengthening the National Statistics System - Household Income Expenditure Survey (P145709) IV. BANK PERFORMANCE, COMPLIANCE ISSUES, AND RISK TO DEVELOPMENT OUTCOME 47. No safeguard policies were triggered. The grant’s activities were rated Category “C� for Social and Environmental Screening as it was not expected to have any direct social or environmental impacts. 48. One of the TTL was country- based, so issues were dealt with rapidly and support to fiduciary issues was organized rapidly. The poverty and equity team provided technical support during implementation particularly on the Agricultural surveys, poverty analysis and dissemination. 49. The World Bank provided hands-on training and follow-up support on World Bank procurement regulations. Frequent hands-on support was provided by local WB procurement specialists to Project Implementation Unit (PIU) procurement staff. Procurement plans had to be continuously updated to reflect delays in implementation. V. LESSONS LEARNED AND RECOMMENDATIONS 50. LISGIS acknowledges donors as key partners in the success of the project, namely the United States Agency for International Development (USAID), European Union (EU), World Bank (WB), African Development Bank (AfDB) and Swedish International Development Agency (SIDA). Donor support was critical in ensuring the successful completion of project activities as Government funding was both inadequate and unpredictable. To provide visibility and appreciation, the project endeavored to honor its donors in in all documents, radio shows, presentations and final reports. 51. LISGIS recognized the need to improve administrative data aggregation to produce GDP estimates annually and quarterly. Currently, the collection of administrative data is ad-hoc and through personal relationships with various coordinators in different Ministries. There are efforts at setting up working groups, such as with external trade, in which data is aggregated and cleaned using the same method across governmental stakeholders. 52. There is the need to institutionalizing the process of routine administrative data collection. The process has already been initiated through draft Memorandums of Understanding with related Ministries and Agencies. In addition, LISGIS staff have received ongoing training in using database software to store and manage the data collected, so that more of the compilation process can be automated. 53. Building in-house statistics development capacity is critical for ensuring sustainable production of national statistics. LISGIS as an institution is better placed in conducting household budget surveys than before; the implementation of year-long 2016 HIES for example, was smoother than that of 2014 HIES, not only because in 2016 there was no ebola but partly because of improved human capacity and better management of physical assets needed to carry out the survey. 54. Looking forward, LISGIS has three main goals for building the administrative data collection system. First is to ensure that statistics produced in partner Ministries and Agencies are compliant with Page 13 of 20 The World Bank Liberia - Strengthening the National Statistics System - Household Income Expenditure Survey (P145709) international standards. Second is to improve partnerships within the National Statistics Producers so that data can be shared and used collaboratively. Third is to ensure the data given by line Ministries and Agencies are in a consistent with the format that can be automatically integrated into the National Accounts database. . Page 14 of 20 The World Bank Liberia - Strengthening the National Statistics System - Household Income Expenditure Survey (P145709) ANNEX 1. RESULTS FRAMEWORK AND KEY OUTPUTS A. RESULTS INDICATORS A.1 PDO Indicators Objective/Outcome: Development of Benchmarks for National Accounts Compilation Objective/Outcome: Development of Baseline Indicators for monitoring the implementation of the Country's Development Agenda- AfT (2013-2017) Objective/Outcome: Assess the Poverty Status of Liberia Unit of Formally Revised Actual Achieved at Indicator Name Baseline Original Target Measure Target Completion Poverty Status of Liberia based Yes/No N N Y Y on a 12-month HIES 28-Jun-2013 02-Jan-2017 31-Oct-2017 30-Sep-2018 Comments (achievements against targets): Objective/Outcome: Development of a new CPI basket based on the consumption data from the Household Income and Expenditure Survey data. Page 15 of 20 The World Bank Liberia - Strengthening the National Statistics System - Household Income Expenditure Survey (P145709) A.2 Intermediate Results Indicators Component: Compilation of the CPI basket Unlinked Indicators Unit of Formally Revised Actual Achieved at Indicator Name Baseline Original Target Measure Target Completion Completion of the 12 month Yes/No N N Y Y Household Income and Expenditure Survey 28-Jun-2013 31-Dec-2017 31-Oct-2017 28-Jun-2017 Comments (achievements against targets): Page 16 of 20 The World Bank Liberia - Strengthening the National Statistics System - Household Income Expenditure Survey (P145709) A. ORGANIZATION OF THE ASSESSMENT OF THE PDO Objective/Outcome 1 1. Development of Benchmarks for National Accounts Compilation: New national accounts estimated annually based on the Establishment Survey, the NA benchmark and the HIES 2. 3. Access to the Poverty Status of Liberia - Key poverty statistics estimated from the HIES 2014 and 2016. Statistical Abstract with Outcome Indicators Poverty status and detailed information on other socio-economic indicators were produced for both suurveys (HIES 2014 and 2016) The results were publicly dissiminated. 4. 4. Development of a new CPI basket based on the consumption data from the Household Income and Expenditure Survey data - Revised CPI basket with selected commodities produced. 1. Completion of the 12 month Household Income and Expenditure Survey. The main survey report produced in 2017 following the Intermediate Results Indicators complition of the survey. Preparation of the HIES - Status Reports documenting progress, Basic Information Document, Cleaned dataset Training and Field Survey - Training completion reports and workshop Key Outputs by Component materials (linked to the achievement of the Objective/Outcome 1) Capacity building for statistical users in carrying out evidence-based policy analysis - Training workshops on key thematic areas targeting statistical producers and users conducted Page 17 of 20 The World Bank Liberia - Strengthening the National Statistics System - Household Income Expenditure Survey (P145709) Project Management – Project fiduciary aspects exercised successfully. Page 18 of 20 The World Bank Liberia - Strengthening the National Statistics System - Household Income Expenditure Survey (P145709) . Annex 2: Assessment of Outcomes PDO Level Results Target Date of Indicators Description Reached Deliverables Completion New national accounts estimated annually based on 1 National Accounts the Establishment Survey, the NA benchmark and the HIES 1.1 National A complete census of all Yes NEC dataset, December 2017 Establishment Census establishments operating in NEC report 2017 Liberia- this forms the sample frame for the National Accounts Annual Survey 1.2 National Accounts An in-depth survey on the Yes NAAS dataset, September 2018 Annual Survey 2018 accounts of 3,000 businesses to NAAS report form the basis of an updated National Accounts for GDP estimation 1.3 Updating Supply and An updated Supply and Use Yes SUT report September 2018 Use Table using the HIES table using administrative data, 2016 available household information from the HIES CPI calculations based on the revised weight for computing 2 CPI the basket & weights from HIES CPI calculated based on the Yes Weights being September 2017 2.1 Consumer Price Index revised weight for a used for CPI Estimation consumptions basket from 2014 HIES Consumption aggregates Yes Aggregates for July 2018 2.2 Consumption extraction to be used for weights Aggregates extraction weights computation construction CPI basket for the country is Yes Revised CPI September, 2.3 CPI basket being rebased with average basket with 2018 configuration using 2016 consumables by average wage selected HIES earners commodities CPI weights for basket Yes Revised CPI September, 2.4 CPI estimation using commodities (group & sub- weights 2018 2016 HIES pending Page 19 of 20 The World Bank Liberia - Strengthening the National Statistics System - Household Income Expenditure Survey (P145709) groups), county weights and national weights Key poverty statistics estimated Yes 3 Poverty from the HIES 3.1 In-depth Poverty Yes Poverty Report March 2016, Analysis 2016 and 2018 August 2017 3.2 Comparative Study Yes Comparative August 2017 between HIES 2014 and Report 2017 HIES 2016 3.3 At least 10 key Revised M&E and results Yes AfT report February 2016, baselines indicators framework for the AfT with 2016 and 2017 August 2017 required for monitoring of baseline indicators drawn from the AfT became available the HIES data 3.4 In-depth Agriculture Yes Agriculture August 2017 Analysis Report 2017, Crop Cutting Report 2017 3.5 In-depth Gender and Yes Gender April 2018, May Employment analysis Analysis 2018, 2018 Employment Analysis 2018 Intermediate Indicators/ Target Date of Milestones Description Reached Deliverables Completion HIES Data collection •Pre-testing and listing of Yes Status Reports March 2016, completed households completed documenting August 2017 •Field manuals for supervisor, progress, Basic enumerator, data entry Information operator prepared Document, •Field survey, data collection, Cleaned first and second data entry dataset completed in 15 counties •Data cleaning completed CPI weights created and CPI calculations based on the Yes September 2017 consumption basket revised weight for computing revised the basket & weights from HIES Page 20 of 20 The World Bank Liberia - Strengthening the National Statistics System - Household Income Expenditure Survey (P145709) Consumption aggregates •Household consumption Yes Consumption March 2016, at the county and national aggregates for food and non- aggregates data August 2017 levels calculated food consumption file for 2014 •Consumption aggregates and 2016 quintile stratas (1-5), 1 being the poorest and 5 being the richest consumption quintile Training workshops on Yes Training Throughout key thematic areas completion project timeline targeting statistical reports and producers and users workshop conducted materials Page 21 of 20