Document of The Word Bank

Report No. ICR00003723

Public Disclosure Authorized IMPLEMENTATION COMPLETION AND RESULTS REPORT (IDA-4147 IDA-4148 IDA-4149 IDA-H2020 IDA4977)

CREDIT IDA-4147 IN THE AMOUNT OF SDR 18.2 MILLION (US$25.6 MILLION EQUIVALENT)

CREDIT IDA-4148 IN THE AMOUNT OF SDR 83.03 MILLION (US$117.3 MILLION EQUIVALENT)

CREDIT IDA-4149 IN THE AMOUNT OF SDR 25.5 MILLION Public Disclosure Authorized (US$35.9 MILLION EQUIVALENT)

GRANT IDA-H2020 IN THE AMOUNT OF SDR 10.4 MILLION (US$14.6 MILLION EQUIVALENT)

PARTIAL RISK GUARANTEE IDA-B0090 IN THE AMOUNT OF UP TO US$45 MILLION FOR THE KENYAN SUBSIDIARY OF RIFT VALLEY RAILWAY HOLDINGS LIMITED.

PARTIAL RISK GUARANTEE IDA-B0100 IN THE AMOUNT OF UP TO US$15 MILLION FOR THE UGANDAN SUBSIDIARY OF Public Disclosure Authorized RIFT VALLEY RAILWAY HOLDINGS LIMITED.

AND

CREDIT IDA-4977 IN THE AMOUNT OF SDR 18.6 MILLION (US$26.2 MILLION EQUIVALENT)

FOR AN

EAST AFRICA TRADE AND TRANSPORT FACILITATION PROJECT

September 27, 2016 Public Disclosure Authorized

Transport and ICT Global Practice Country Departments AFCE1 and AFCE2 Africa Region CURRENCY EQUIVALENTS (Exchange Rate Effective September 30, 2015)

Currency Units = Shilling (KES), Rwanda Franc (RWF), Shilling (TZS), Shilling (UGX) XDR1.00 = US$ 1.40 US$ 1.00 = KES 104.9 US$ 1.00 = RWF 712 US$ 1.00 = TZS 2160 US$ 1.00 = UGX 3695

FISCAL YEAR July 1 – June 30

ABBREVIATIONS AND ACRONYMS CAS Country Assistance Strategy CCTA Central Corridor Transport Authority CCTTFA Central Corridor Transit Transport Facilitation Agency CU Customs Union DFID Department for International Development, United Kingdom EAC GoK Government of Kenya GoU Government of Uganda IBM International Business Machines ICR Implementation Completion and Results ICT Information and Communications Technology IDA International Development Association IEG Independent Evaluation Group IFC International Finance Corporation IFMS Integrated Financial Management System IRAP Implementation Recovery Action Plan ISPS International Ship and Port Facility ISR Implementation Status and Results Report KE Kenya KeNHA Kenya National Highways Administration KenTrade Kenya Trade Network Agency KRA Kenya Revenue Authority KSH Kenyan Shillings MOI Ministry of Infrastructure (Rwanda) MOTI Ministry of Transport and Infrastructure (Kenya) Mtce Maintenance MoWTC Ministry of Works, Transport and Communications (Tanzania) MWT Ministry of Works and Transport (Uganda) NC Northern Corridor NCTTCA Northern Corridor Trade and Transport Coordination Authority NSW National Single Window NTB Non-tariff Barrier OSBP One Stop Border Posts PAP Project Affected Person PCS Port Community System PDO Project Development Objectives PRG Partial Risk Guarantee QAG Quality Assurance Group RADDEx Revenue Authorities Digital Data Exchange RAP Resettlement Action Plan REMU Resettlement Management Unit RVR Rift Valley Railways RRA Rwanda Revenue Authority RTDA Rwanda Transport Development Agency SCT Single Customs Territory SUMATRA Surface and Marine Transport Regulatory Authority (Tanzania) TANROADS Tanzania National Roads Agency TEU Twenty Foot Equivalent Unit Tkm Ton kilometers TMEA TradeMark East Africa TPA Tanzania Ports Authority TRA Tanzania Revenue Authority TRL Tanzania Railways Limited TZ Tanzania UG Uganda URA Uganda Revenue Authority VAT Value Added Tax WCO World Customs Organization WiM Weigh in Motion XDR Special Drawing Rights

Vice President: Makhtar Diop Country Director: Ahmadou Moustapha Ndiaye Senior Global Practice Director: Pierre Guislain Practice Manager: Aurelio Menendez Task Team Leader: Solomon Muhuthu Waithaka ICR Team Leader: Richard Martin Humphreys REPUBLIC OF KENYA, REPUBLIC OF RWANDA, UNITED REPUBLIC OF TANZANIA, AND REPUBLIC OF UGANDA East Africa Trade and Transport Facilitation Project

CONTENTS DATA SHEET ...... i A. Basic Information ...... i B. Key Dates ...... ii C. Ratings Summary ...... ii D. Sector and Theme Codes ...... ii E. Bank Staff ...... iii F. Results Framework Analysis ...... iii G. Ratings of Project Performance in ISRs ...... xiii H. Restructuring (if any) ...... xiii I. Disbursement Profile ...... xiv 1. Project Context, Development Objectives and Design ...... 1 2. Key Factors Affecting Implementation and Outcomes ...... 6 3. Assessment of Outcomes ...... 15 4. Assessment of Risk to Development Outcome ...... 20 5. Assessment of the Guarantee ...... 21 6. Assessment of Bank and Borrower Performance ...... 22 7. Lessons Learned ...... 27 Annex 1: Project Costs and Financing ...... 30 Annex 2: Outputs by Component ...... 32 Annex 3: The Kenya-Uganda Railways Concession ...... 38 Annex 4: Bank Lending and Implementation Support/Supervision Processes ...... 46 Annex 5: Summary of Borrower's ICR and/or Comments on Draft ICR ...... 49 Annex 6: Comments of Cofinanciers and Other Partners/Stakeholders ...... 54 Annex 7: List of Supporting Documents ...... 55 MAP ...... 56

DATA SHEET

A. Basic Information

AFCC2/RI-East Africa Country: Africa Project Name: Trade and Transport Facilitation Project IDA-41470, IDA- P079734 41480, IDA-41490, Project ID: L/C/TF/GU Number(s): IDA-49770, IDA- H2020, B0090, B0100 ICR Date: 10/13/2016 ICR Type: Core ICR KENYA, UGANDA, Lending Instrument: SIL Borrower: TANZANIA AND RWANDA Rift Valley Railway Lending Instrument: PRG Borrower: Company Original Total XDR 137.40M Disbursed Amount: XDR 150.18M Commitment: Revised Amount: XDR 156.18M Environmental Category: A Implementing Agencies: Ministry of Transport and Infrastructure (MOTI), Kenya Kenya Revenue Authority (KRA) Kenya Ports Authority (KPA) Kenya Trade Network Agency (KenTrade) Kenya National Highways Authority (KeNHA) Ministry of Infrastructure (MININFRA), Rwanda Rift Valley Railways (RVR) Rwanda Revenue Authority (RRA) Rwanda Transport Development Agency (RTDA) Tanzania Ports Authority (TPA) Surface and Marine Transport Regulatory Authority (SUMATRA), Tanzania Tanzania Revenue Authority (TRA) Tanzania National Roads Agency (TANROADS) Ministry of Works and Transport (MOWT), Uganda Uganda Revenue Authority (URA) Cofinanciers and Other External Partners: (AfDB) Department for International Development, United Kingdom (DFID) TradeMark East Africa (TMEA)

i B. Key Dates Revised / Actual Process Date Process Original Date Date(s) Concept Review: 09/14/2004 Effectiveness: 06/05/2006 06/05/2006 05/27/2011 Appraisal: 06/15/2005 Restructuring(s): 09/30/2014 Approval: 01/24/2006 Mid-term Review: 12/08/2008 12/08/2008 Closing: 09/30/2011 09/30/2015 C. Ratings Summary C.1 Performance Rating by ICR Outcomes: Moderately Satisfactory Risk to Development Outcome: Moderate Bank Performance: Moderately Unsatisfactory Borrower Performance: Moderately Unsatisfactory

C.2 Detailed Ratings of Bank and Borrower Performance (by ICR) Bank Ratings Borrower Ratings Moderately Moderately Quality at Entry: Government: Unsatisfactory Unsatisfactory Moderately Implementing Moderately Quality of Supervision: Unsatisfactory Agency/Agencies: Unsatisfactory Overall Bank Moderately Overall Borrower Moderately Performance: Unsatisfactory Performance: Unsatisfactory C.3 Quality at Entry and Implementation Performance Indicators Implementation QAG Assessments Indicators Rating Performance (if any) Potential Problem Quality at Entry Project at any time Yes None. (QEA): (Yes/No): Problem Project at any Quality of Yes None time (Yes/No): Supervision (QSA): DO rating before Moderately

Closing/Inactive status: Satisfactory D. Sector and Theme Codes Original Actual Sector Code (as % of total Bank financing) General public administration sector 13 13 General transportation sector 38 38 Ports, waterways and shipping 12 12 Railways 37 37

ii

Theme Code (as % of total Bank financing) Improving labor markets 14 14 Income Support for Old Age, Disability & Survivorship 14 14 Regional integration 29 29 Regulation and competition policy 14 14 Trade facilitation and market access 29 29 E. Bank Staff Positions At ICR At Approval Vice President: Makhtar Diop Gobind T. Nankani Country Director: Ahmadou Moustapha Ndiaye Mark D. Tomlinson Practice Aurelio Menendez C. Sanjivi Rajasingham Manager/Manager: Project Team Leader: Solomon Muhuthu Waithaka Jean-Francois Marteau ICR Team Leader: Richard Martin Humphreys ICR Primary Author: Christopher Willoughby F. Results Framework Analysis Project Development Objectives (from Project Appraisal Document) The original project development objective was defined as: (a) improve trade environment through the effective elimination of tariff barriers in the EAC Customs Union area; (b) enhance logistics services efficiency along key corridors by reducing non-tariff barriers and uncertainty of transit time; and (c) improve railway services in Kenya and Uganda. Revised Project Development Objectives (as approved by original approving authority) The revised project development objective was defined as: (i) enhance efficiency of customs agencies’ clearance processes, for the EAC Partner States participating in the East Africa Customs Union, to facilitate trade; (ii) improve efficiency and reliability of transport and logistics services along the key corridors; and (iii) enhance safety in identified areas and reduce the Recipient’s fiscal transfers to railway institutions by rationalizing the work force on the Kenya-.

(a) PDO Indicator(s)

Actual Value Original Target Formally Achieved at Values (from Indicator Baseline Value Revised Target Completion approval Values or Target documents) Years Indicator 1: The EAC Customs Union is established and is applied consistently. Value Applied with almost (quantitative CU not established Achieved. no exceptions. or Qualitative) Date achieved 06/15/2005 03/31/2011 05/27/2011 Comments Target achieved (100%). This indicator was achieved prior to restructuring in (incl. % 2011.

iii achievement) Indicator 2: EAC Customs Management Law is implemented within EAC region. Value Law enforced and (quantitative Law not implemented effectively Achieved. or Qualitative) implemented Date achieved 06/15/2005 03/31/2011 05/27/2011 Comments Target achieved (100%). The Law was enforced and implemented by the 2011 (incl. % restructuring. achievement) Indicator 3: The Regional interconnection of Customs IT systems operates. Value Interconnection (quantitative Non existent Achieved. functioning or Qualitative) Date achieved 06/15/2005 03/31/2011 09/30/2015 Comments Target achieved (100%). Interconnection achieved between four countries in (incl. % 2014. achievement) Indicator 4: A common database of Customs Union is established. Value Not quantitative or Not established Established established Qualitative) Date achieved 06/15/2005 03/31/2011 09/30/2015 Comments Target not achieved (0%). This indicator was unattainable due to the difficulty (incl. % in establishing a common customs Union database. achievement) Rwanda has developed and implemented a comprehensive action plan towards Indicator 5: accession to the EAC Customs Union. Value quantitative or Not Prepared Prepared Achieved Qualitative) Date achieved 06/15/2005 03/31/2011 05/27/2011 Comments Target achieved (100% achievement). This indicator was achieved prior to (incl. % restructuring in 2011. achievement) Transit time through main Northern Corridor decreases by 30 percent (from Indicator 6: to Kigali). Value quantitative or 19 days 13 days 5 days 5 days Qualitative) Date achieved 06/15/2005 03/31/2011 05/27/2011 09/30/2015 Comments Original target achieved (146%). Revised target achieved (100%). Target (incl. % was revised as part of the 2011 Restructuring. achievement) Indicator 7: Variance of transit time through the Northern Corridor significantly reduced. Value quantitative or 14 days 7 days 2 2 days Qualitative) Date achieved 06/15/2005 03/31/2011 05/27/2014 09/30/2015

iv Comments Original target achieved (100%). Revised target achieved (100% (incl. % achievement). Indicator target achieved and revised from seven days to five in achievement) 2011, then further revised to two days and achieved in 2014. Policies and regulations regarding transport and transit are harmonized along the Indicator 8: Northern and Central corridors. Value No single customs Single customs quantitative or document used from document used from DROPPED Not achieved Qualitative) Mombasa to Kigali Mombasa to Kigali Date achieved 06/15/2005 03/31/2011 05/27/2011 09/30/2015 Comments Target not achieved (0%). Indicator dropped in order to simplify results (incl. % framework. achievement) Through implementation of cargo tracking system, control and monitoring of Indicator 9: transit goods improves so that the Kenyan escort system is removed from the Northern corridor. Effective Value No tracking system Implemented; escorts cargo quantitative or REVISED available removed tracking in Qualitative) place Date achieved 06/15/2005 03/31/2011 05/27/2011 09/30/2015 Original target achieved (100%). Revised target achieved (100% Comments achievement). Indicator revised in 2011 restructuring to reflect the removal of (incl. % the escort system, but the regional cargo tracking system was not introduced, but achievement) a national scheme was introduced. Indicator 10: Increased security in Port of Dar es Salaam in accordance with ISPS Code. Value Installation quantitative or Initial Assessment Final security audit Achieved Complete Qualitative) Date achieved 06/15/2005 03/31/2011 05/27/2011 09/30/15 Comments Original target achieved (100%). Revised target achieved (100% (incl. % achievement). Indicator revised in 2011 restructuring and replaced with more achievement) precise indicator. Indicator 11: Increased security in Port of Mombasa in accordance with ISPS Code. Value Installation quantitative or Initial Assessment Final security audit Achieved Complete Qualitative) Date achieved 06/15/2005 03/31/2011 05/27/2011 09/30/15 Comments Original target achieved (100%). Revised target achieved (100% (incl. % achievement). Indicator revised in 2011 restructuring and replaced with more achievement) precise indicator. Indicator 12: Reduced dwell time at the port of Dar es Salaam, from 17.5 days to 10 days. Value quantitative or Average 18 days 10 days REVISED 8 Qualitative) Date achieved 06/15/2005 03/31/2011 05/27/2014 09/18/2014 Comments Original target achieved (100%). Revised target achieved (100% (incl. % achievement). Target revised to eight days in 2011 the dropped in the 2014 achievement) restructuring. Indicator 13: Reduced dwell time at the port of Mombasa, from 13 days to 8 days.

v Value quantitative or Average 13 days 7 3 3 Qualitative) Date achieved 06/15/2005 03/31/2011 05/27/2011 09/30/2015 Comments Original target achieved (100%). Revised target achieved (100% (incl. % achievement). Target revised in 2011, to 3 days, and fully achieved. achievement) Indicator 14: Kenya/Uganda border- Malaba crossing time decreases by 35 percent. Value quantitative or 45 hours 30 hours Dropped Qualitative) Date achieved 06/15/2005 03/31/2011 05/27/2011 Comments Original achieved (100%) then dropped as a PDO indicator. Target revised (incl. % to 5 hours in 2011, and kept as an Intermediate Output Indicator. achievement) Indicator 15: Kenya/Uganda border - Busia crossing time decreases by 25 percent. Value quantitative or 7 hours 5 hours DROPPED Qualitative) Date achieved 06/15/2005 03/31/2011 05/27/2011 Comments Target achieved (100%) then dropped as a PDO indicator. Indicator dropped (incl. % in 2011 Restructuring. achievement) Kenya/Tanzania Border – Isebania – Sirari border crossing time decreases by 20 Indicator 16: percent. Value quantitative or No baseline specified No target specified DROPPED Qualitative) Date achieved 06/15/2005 03/31/2011 05/27/2011 Comments Target not achieved (0%). Indicator had no baseline or target specified and was (incl. % dropped as a PDO indicator in 2011 Restructuring. achievement) Kenya/Tanzania Border – Lunga-Lunga - Horohoro border crossing time Indicator 17: decreases by 20 percent. Value quantitative or No baseline specified No target specified DROPPED Qualitative) Date achieved 06/15/2005 05/27/2011 Comments Target not achieved (0%). Indicator had no baseline or target specified and was (incl. % dropped as a PDO indicator in 2011 Restructuring. achievement) Uganda/Rwanda Border – Gatuna/Katuna border crossing time decreases by 30 Indicator 18: percent. Value quantitative or 13 hours 9 hours 3 hours 2 hours Qualitative) Date achieved 06/15/2005 03/31/2011 05/27/2011 09/30/2015 Comments Original target achieved (100%). Revised target surpassed (160% (incl. % achieved). Target revised to 3 hours in 2011 Restructuring. achievement) vi Rwanda/Tanzania Border – Rusumo border crossing time decreases by 20 Indicator 19: percent. Value quantitative or No baseline specified No target specified DROPPED Qualitative) Date achieved 06/15/2005 03/31/2011 05/27/2011 Comments Target not achieved (0%). Indicator had no baseline or target specified and was (incl. % dropped as a PDO indicator in 2011 Restructuring. achievement) Indicator 20: The RVR Concession is effective and financially sustainable. Value quantitative or Concession not started Concession Started DROPPED Qualitative) Date achieved 06/15/2005 03/31/2011 05/27/2011 Target partially achieved (50%). Target partially met. The concession did Comments commence operations but was far from financially sustainable. Indicator dropped (incl. % in 2011 as project activities did not directly influence performance of the achievement) concession. Indicator 21: Railway costs for the Government diminish. Value Net transfers from quantitative or Government to railways Net transfer negative DROPPED Qualitative) 1.4 billion schs Date achieved 06/15/2005 05/27/2011 Target achieved (100%). Target was met in the sense that the concession did Comments commence paying fees to the Governments. The indicator was dropped in 2011 (incl. % because project activities were considered disconnected from the performance of achievement) the concession. Rail market share on the Northern Corridor has increased from the current 16 Indicator 22: percent to 30 percent. Value quantitative or 16 percent 30 percent DROPPED Qualitative) Date achieved 06/15/2005 03/31/2011 05/27/2011 Comments Target not achieved (0%). Dropped in 2011 Restructuring; project activities (incl. % were felt to have no influence on the performance of the concession. achievement) Indicator 23: Rail efficiency improves (Transit time Mombasa to ). Value quantitative or 14 days 7 days DROPPED Qualitative) Date achieved 06/15/2005 03/31/2011 05/27/2011 Comments Target not achieved (0%). Dropped in 2011 Restructuring; project activities (incl. % were felt to have no influence on the performance of the concession. achievement)

New PDO Indicators added in 2011:

Average time from ship readiness to unload to final destination for an imported Indicator 24: container, on the Northern Corridor. Value 9 days 5 days 5 days vii quantitative or Qualitative) Date achieved 05/27/2011 09/30/2014 09/30/2015 Comments (incl. % Target achieved (100%). This indicator was added in 2011 Restructuring. achievement) Indicator 25: Rwanda joins the EAC CU protocol. Value Substantially Rwanda is a full Rwanda is a quantitative or Application made complete member full member Qualitative) Date achieved 05/27/2011 09/30/2014 03/31/2015 09/30/2015 Comments Target achieved (100%). This indicator was added in 2011 Restructuring. (incl. % Rwanda was a full member as of March 2015. achievement) Indicator 26: Kenya Railways Pension Scheme operational and financially sustainable. Operational Value Operational and Established; has financial and quantitative or financially problems financially Qualitative) sustainable sustainable Date achieved 05/27/2011 09/30/2014 09/30/2015 Comments (incl. % Target achieved (100%). achievement) Indicator 27: Safety operational zone established in Kiberea and Mukuru. Value Implementati RAP prepared and quantitative or Established on of RAP disclosed Qualitative) ongoing Date achieved 05/27/2011 09/30/2014 09/30/2015 Comments Partially achieved (30%). This indicator was added in 2011 Restructuring to (incl. % monitor resettlement in the expanded safety zone. At project close, achievement) implementation of resettlement programs were partial and ongoing. Indicator 28: Wagon Ferry service resumed in Uganda. Value Rehabilitation of MV Partially quantitative or Ferry services resume Kaawa started Achieved Qualitative) Date achieved 05/27/2011 09/30/2014 09/30/2015 Partially achieved (50%).This indicator was added at 2011 Restructuring. The Comments wagon ferry has been rehabilitated and is available for service, however the (incl. % condition of the railway connections means there is little traffic, and hence no achievement) service.

(b) Intermediate Outcome Indicator(s)

Original Target Formally Actual Value Values (from Revised Achieved at Indicator Baseline Value approval Target Completion or documents) Values Target Years Indicator 1: EAC Customs Management Law is implemented within EAC region.

viii Value Law enforced and (quantitative Law not implemented effectively Achieved or Qualitative) implemented Date achieved 06/15/2005 03/31/2011 05/27/2011 Comments Target achieved (100%). The Law was enforced and implemented by the 2011 (incl. % Restructuring. achievement) Indicator 2: The EAC Customs Union is established and functions. Value (quantitative CU not established CU established Achieved or Qualitative) Date achieved 06/15/2005 03/31/2011 05/27/2011 Comments Target achieved (100%). This indicator was achieved prior to the 2011 (incl. % Restructuring. achievement) Customs procedures and documents are simplified and WCO indicators show Indicator 3: that efficiency of customs agencies increases. Value (quantitative - Yes Yes Achieved or Qualitative) Date achieved 06/15/2005 03/31/2011 05/27/2011 09/30/2015 Comments (incl. % Target achieved (100%). Indicator refined in 2011 Restructuring. achievement) The Central Corridor Transit Transport Facilitation Authority (CCTTFA) and its Indicator 4: permanent Secretariat is established and functions. Value CCTTFA (quantitative Non-existent Yes established in 2010. or Qualitative) Date achieved 06/15/2005 03/31/2011 09/30/2015 Comments Original target achieved (100%). (incl. % achievement) The Port Community Systems (PCS) are established and operational in both Indicator 5: Ports of Dar es Salaam and Mombasa. Value Established and REVISED to PCS established in (quantitative Non-existent starts to function drop Dar Mombasa or Qualitative) Date achieved 06/15/2005 09/30/2006 05/27/2011 09/30/2015 Comments Target partially achieved (50%) A PCS was established in Mombasa port, but (incl. % due to procurement difficulties, and a misunderstanding over scope, was never achievement) established in Dar port. Indicator 6: Regional cargo tracking system is operational along the Northern Corridor. Value No regional CTS in (quantitative Non-existent Yes place or Qualitative) Date achieved 06/15/2005 09/30/2009 09/30/2015 Comments Target not achieved (0%). The regional CTS was never introduced, but Kenya (incl. % introduced its own scheme achievement) ix Indicator 7: National cargo tracking systems are established in four countries. Kenya – no Yes Value Uganda – no Yes (quantitative DROPPED Rwanda – no Yes or Qualitative) Tanzania - no Yes Date achieved 06/15/2005 09/30/2009 05/27/2011 Comments Target achieved (100 %) Indicator dropped in 2011 restructuring as part of (incl. % simplification of results monitoring. achievement) Indicator 8: Harmonized load control policy is implemented in four countries. Value (quantitative - Yes DROPPED or Qualitative) Date achieved 06/15/2005 09/30/2009 05/27/2011 Comments Target not Achieved (0 %). Indicator dropped in 2011 restructuring; (incl. % implementation of harmonized load control was found to be outside the scope of achievement) the project. Indicator 9: Weighbridge crossing-time in Mariakani Value (quantitative 6 hours 4 hours 0.5 hours or Qualitative) Date achieved 06/15/2005 09/30/2010 09/30/2015 Comments Target achieved (100%). New weighbridge was installed and crossing time has (incl. % reduced to a fraction of the project target. achievement) Indicator 10: Kenya/Uganda border-crossing time decreases. Value Malaba: 45 hours (15 KE; Malaba: 4 (2 Malaba: 30 Malaba: 2.1 (1.5 quantitative or 30 UG) KE; 2 UG) Busia: 5 KE; 0.6 UG) Qualitative) Busia:7 hours Busia - Date achieved 06/15/2005 09/30/2010 05/27/2011 09/30/2015 Comments Target achieved (100%). Improvement exceeds targets significantly, (incl. % particularly on Uganda side. Busia border crossing dropped from monitoring in achievement) 2011 Restructuring. Indicator 11: Kenya/Tanzania border-crossing time decreases. Isebania/Sirari: 5 hours (2.5 per Value Isebania/Sirari: - 7 hrs - side) quantitative or Lunga-Lunga/Horohoro: - - Lunga-Lunga/ Qualitative) Horohoro dropped. Date achieved 06/15/2005 - 05/27/2011 09/30/2015 Comments Target not achieved (0%). No baseline or target in PAD for Lunga, and (incl. % indicator was dropped in 2011 Restructuring – Baseline refined in 2011 for achievement) Isebania/Sirari – but no information on attainment at project close. Indicator 12: Uganda/Rwanda (Katuna-Gatuna) border-crossing time decreases by 30 percent. Value quantitative or 13 hours 9 hours 2 hours 2 hours Qualitative) Date achieved 06/15/2005 09/30/2010 05/27/2011 09/30/2015 Comments Original target achieved (100%). Revised target achieved (100%). x (incl. % achievement) Indicator 13: Tanzania/Rwanda (Rusumo Falls) border-crossing time decreases by 20 percent. Value quantitative or - - - Qualitative) Date achieved 06/15/2005 - Comments Target not achieved (0%). No baseline or target provided in the PAD. (incl. % Indicator dropped without explanation in 2011 restructuring. Subsequently JICA achievement) funded the development of this border crossing.

Intermediate Outcome Indicators added in 2011:

Regional Interconnection of ICT system is functioning efficiently in all four Indicator 14: participating countries. Value Optimal functionality not (quantitative Yes Achieved achieved yet or Qualitative) Date achieved 05/27/2011 09/30/2014 09/30/2015 Comments Target achieved (100%). Added in 2011 following establishment of the (incl. % Customs Union. achievement) Customs procedures and documents are simplified in accordance with WCO Indicator 15: standards. Value (quantitative Work in progress Yes Yes or Qualitative) Date achieved 05/27/2011 09/30/2014 09/30/2015 Comments Target achieved (100%). (incl. % achievement) Both corridor authorities (CCTTFA, NCTTCA) set a transit traffic monitoring Indicator 16: database. Value Database not (quantitative Non-existent Yes created. or Qualitative) Date achieved 05/27/2011 09/30/2014 09/30/2015 Comments Target partially achieved (50%). At project close, CCTTFA did not have a (incl. % functioning transit traffic monitoring database. achievement) Indicator 17: National Single Window ICT backbone is established in Kenya. Value Specifications prepared; (quantitative Yes Achieved Achieved capacity building starts or Qualitative) Date achieved 05/27/2011 09/30/2014 09/30/2014 09/30/2015 Comments Target achieved (100%). (incl. % achievement) Indicator 18: CBS ICT backbone is established in Dar es Salaam.

xi Value Specifications prepared; (quantitative Yes Not established capacity building starts or Qualitative) Date achieved 05/27/2011 09/30/2014 09/30/2015 Comments Target not achieved (0%). There was no ICT backbone for a CBS in Dar port (incl. % at project close. achievement) Indicator 19: Decrease in crossing time at the Mutukula border. Value 5 hours (2.5 h each quantitative or Not available DROPPED TZ and UG) Qualitative) Date achieved 05/27/2011 09/30/2014 09/30/2014 Comments Indicator not reported as component dropped. Improvement to this border (incl. % crossing financed by TMEA. OBSP operational by August 2015. achievement) Indicator 20: Railway ICD in Uganda built and functional. Value Land identified and ICD fully (quantitative ICD constructed designs completed. established or Qualitative) Date achieved 05/27/2011 09/30/2013 09/30/2015 Comments Target achieved (100%). Replaces Intermediate Indicator 15. (incl. % achievement) Indicator 21: Updated railway RAP is implemented in Kiberia and Mukuru Value Fully 30 percent (quantitative Updating of RAP Fully implemented Implemented implemented or Qualitative) Date achieved 05/27/2011 09/30/2014 09/30/2015 09/30/2015 Comments Target partially achieved (30%) at project close. (incl. % achievement)

New Intermediate Output Indicators added in 2014:

OSBP established and functional at Malaba (KE/UG), Isebania/Sirare (KE/TZ), Indicator 22: and Lunga-lunga/Horohoro (TZ/KE) borders. Modern and Value integrated Not achieved by (quantitative Not available infrastructure project close or Qualitative) supported by IBM system Date achieved 09/30/2014 09/30/2015 09/30/2015 Comments Target partially achieved (50%). Construction of new facilities complete by (incl. % project close, but necessary public sector reform was not complete. achievement) Modern weighbridge stations installed in: Vigwaza (TZ), Athi River, Mariakani Indicator 23: (KE), and Malaba (UG) Value Enhanced work Vigwaza: 2016 (quantitative Not available environment Athi River: 2016 or Qualitative) supported by Mariakani: 2016 xii modern axle load Malaba: 2015 control equipment; wait time <30 minutes. Date achieved 09/30/2014 09/30/2015 09/30/2015 Comments Target partially achieved (25%). Baseline as of 2005 was poor work (incl. % environment, outdated equipment, and waiting times >60 minutes. Athi River, achievement) Vigwaza and Mariakani weigh stations not complete by project closing.

G. Ratings of Project Performance in ISRs

Date ISR Actual Disbursements No. DO IP Archived (USD millions) 1 05/21/2006 Satisfactory Satisfactory 0.00 2 10/05/2006 Satisfactory Satisfactory 8.20 3 04/05/2007 Satisfactory Satisfactory 33.45 Moderately 4 06/26/2007 Satisfactory 42.04 Satisfactory Moderately 5 12/07/2007 Satisfactory 54.01 Satisfactory Moderately 6 05/26/2008 Satisfactory 58.77 Satisfactory Moderately 7 09/12/2008 Satisfactory 60.63 Satisfactory Moderately Moderately 8 03/04/2009 67.83 Satisfactory Unsatisfactory Moderately Moderately 9 10/16/2009 75.45 Satisfactory Unsatisfactory Moderately Moderately 10 03/25/2010 77.43 Satisfactory Satisfactory Moderately Moderately 11 11/03/2010 82.90 Satisfactory Satisfactory Moderately Moderately 12 04/23/2011 86.78 Satisfactory Satisfactory Moderately Moderately 13 09/14/2011 92.34 Satisfactory Satisfactory Moderately Moderately 14 05/13/2012 100.39 Satisfactory Satisfactory Moderately Moderately 15 12/11/2012 111.01 Satisfactory Satisfactory Moderately Moderately 16 07/09/2013 120.07 Unsatisfactory Unsatisfactory Moderately Moderately 17 03/06/2014 143.50 Satisfactory Unsatisfactory Moderately Moderately 18 10/04/2014 174.14 Satisfactory Unsatisfactory Moderately Moderately 19 05/27/2015 192.79 Satisfactory Unsatisfactory

H. Restructuring (if any)

Restructuring Board ISR Ratings at Amount Reason for Restructuring & Key Date(s) Approved Restructuring Disbursed at Changes Made xiii PDO Restructuring in DO IP Change USD millions Corrective restructuring: Additional Financing to cover cost overruns and scaling up of some activities 05/27/2011 Y MS MS 87.19 (particularly resettlement); revise PDO to better reflect project scope; Safeguard Category changed to A. Corrective restructuring: address cost increases from inflation; reallocate among categories; revise 09/30/2014 MS MU 169.15 implementation schedule and financing plan; cancel some activities; and revise the results framework. If PDO and/or Key Outcome Targets were formally revised (approved by the original approving body) enter ratings below: Outcome Ratings Against Original PDO/Targets Moderately Unsatisfactory Against Formally Revised PDO/Targets Moderately Satisfactory Overall (weighted) rating Moderately Satisfactory

I. Disbursement Profile

xiv 1. Project Context, Development Objectives and Design 1.1 Context at Appraisal

1. The three main eastern African countries, Kenya, Tanzania and Uganda – long associated with one another in a formal community which had broken up under political pressures in 1977 – came together again towards the end of the 1990s to sign a major new treaty for mutual cooperation. The three countries committed themselves to deepening their economic cooperation, through inter alia, the establishment of a permanent East African Community (EAC) Secretariat at Arusha, Tanzania, and the formation of a customs union by 2004. This was to be followed further steps towards the creation of a single market and monetary union, and potentially, by 2020, a political federation.

2. In December 2002, Kenya, the largest country of the group, elected a new government determined to break the economic mismanagement and income stagnation that the country had suffered. At the same time the Development Partners and leading advisory bodies such as the United Nations Economic Commission for Africa (UNECA) had been advocating trade and economic integration as key dimensions of income generation and growth. This synergy resulted in quick action in Eastern Africa: The ratification of the Protocol for the new Customs Union and the finalization of an EAC Customs Management Law (replacing the old East African Customs Act of 1960) were completed in December 2004, and new import tariffs enforced from March 2005.

3. Despite intra-regional trade growing significantly, especially between Uganda and Kenya, more than 80 percent of the main trade flows were still in to and out of the region. The key international trade routes in East Africa are the Northern and Central Corridors (see map at end of document). The Northern Corridor runs from Mombasa port via to Kampala, with extensions to Democratic Republic of Congo (DRC), Rwanda and Burundi. The Central Corridor runs from Dar-es-Salaam port via Dodoma to North- Western Tanzania, with extensions to DRC, Burundi, Rwanda and Uganda. At the time of project preparation, transport and inventory costs in East Africa were amongst the highest in the world, and recognized as a major cause of limited integration and low competitiveness of the tradable sectors of the economy.

4. The high cost of transport along the key trading corridors, reflecting the condition of transport infrastructure, underdeveloped transport and logistics services, slow and costly bureaucratic procedures, and other non-tariff barriers (NTBs), was recognized as a major impediment to economic development. This contributed to the emergence in the Bank of an ambitious new concept: A multi-country regional project addressed broadly to lowering the costs of trade – especially transport and trade costs – through interventions to improve both the physical infrastructure, and the soft infrastructure. An initiative along these lines was mentioned in the Bank’s Country Assistance Strategy (CAS) for Kenya, issued in 2004. The proposed project was also consistent with the CAS’s for Rwanda, Tanzania and Uganda, which prioritised infrastructure investment, private sector growth, and economic competitiveness.

5. In terms of the physical infrastructure, while the quality of the road infrastructure was improving, and further road improvements were expected, the performance of the

1

railway operators on both the Northern and Central transport corridors was deteriorating. As result, rail was losing significant levels of traffic to the road mode, including those bulk and unitized traffics where it could have had a competitive advantage. The result from the reduced competition was increased transport costs overall, but also a significant increase in the social and economic disbenefits of road traffic, increased congestion and pollution, higher road maintenance funding needs, and fatalities and injuries from road traffic crashes.

6. The World Bank was already engaged with the agencies involved in the implementation and the establishment of an effective customs union, and accordingly was well positioned to provide the necessary support and technical assistance. In addition, the World Bank had considerable experience in implementing similar regional trade and transport facilitation projects in other parts of the world, which it could bring to bear for the benefit of the project design and countries in the region.

7. The rationale for the World Bank to support the Kenya Uganda railway concession was that a long-term concession was the only effective way to ensure efficient railway operations and thereby potentially contribute to a reduction in transport costs. At that time of appraisal, both railway operations were in precarious financial situation, requiring government subsidies to remain operational, and the degree of effective competition with road was limited. Obsolete equipment and lack of adequate maintenance resulted in poor operational performance that in the face of stiff competition from the road sector caused declining traffic levels. Lack of coordination between systems led to delays in cargo movement. The benefits from an efficient railway were perceived as substantial, and the World Bank was uniquely placed given its experience in supporting railway reform and concessioning in many other countries. The CAS for Kenya and for Uganda indicated that improving management of the joint Kenya-Uganda railways, through concessioning, was expected to be a major catalyst to regional integration and growth.

8. The preparation of the project commenced the same year (2004) with discussions with all three EAC countries, following their decision in March 2004 to establish the customs union and introduce a common external tariff. The discussion was extended to include Rwanda which had shown interest in joining the emerging Customs Union

1.2 Original Project Development Objectives (PDO) and Key Indicators

9. The original project development objective was: (i) to improve the countries’ trade environment through effective implementation of the EAC Customs Union Protocol; (ii) to enhance transport and logistics services’ efficiency along key corridors by reducing non- tariff barriers and the uncertainty of transit time; and (iii) to improve railway services in Kenya and Uganda.

1.3 Revised PDO and Key Indicators, and reasons/justification

10. The PDO was revised as part of a project restructuring in 2011. The opportunity was taken to refine the three elements of the PDO and strengthen the linkage between them, the actual interventions in the project, and the expected outputs and outcomes. The revised PDO was: (i) to enhance the efficiency of customs agencies’ clearance processes for the countries participating in the EAC Customs Union; (ii) to improve efficiency and reliability

2

of transport and logistics services along the key corridors; and (iii) to enhance safety in identified areas and reduce the Recipient’s fiscal transfers to railway institutions by rationalizing the work force on the Kenya-Uganda railway.

11. The first element of the PDO was amended to reflect the fact the Customs Union had been established and was functioning, and hence the indicator was sharpened to focus on improving the efficiency of the respective Customs Authorities. The second element of the PDO was also revised to reflect the limited ability of the project to address NTBs that were considered to be outside the scope of the project. Finally, the third element was refined to reflect the project was now only financing the establishment of safe zones and the resettlement of people who had established informal residence within the right of way (RoW) on some sections of the railway line in the vicinity of Nairobi, and the resolution of financial problems associated with labor-force retrenchment and future pension obligations.

12. The restructuring of the third element of the PDO limited the scope of the railway component in the PDO, and thereby effectively excluded the Partial Risk Guarantees (PRGs) subcomponent. Hence all indicators related to the efficiency, operation or management of the railway concessions were dropped without adding any new indicator to monitor and assess the progress of the PRG subcomponent. As a result, the results framework for the project did not have any meaningful indicator to monitor the performance of the railway under the two PRGs, when those guarantees became effective in June 2012.

1.4 Main Beneficiaries

13. The primary beneficiaries were expected to be those already engaged in international trade – either production of goods for export, logistics of their movement, or receipt/distribution of goods imports – the increasing numbers of people who should find employment in these fields, and eventually the consumers of imported goods. A small sub- group of beneficiaries drew special benefit from the project because of their particular situations were those released by the railway reform, and those residing in the right of way a few yards from passing trains.

1.5 Original Components (as approved)

14. The components of the project were divided into four groups: Component 1, Support to Create Regional Customs Union, included (a) long term support to the EAC Secretariat for the implementation of the CU;1 (b) equipment to implement a modem customs integrated system and common database linking the customs departments in the member states and Rwanda to the EAC Customs directory in Arusha; and (c) support the strengthening and modernization of national customs departments.

15. Component 2, Institutional Support to Transport Facilitation included: (a) strengthening of the Northern Corridor Transit and Transport Coordinating Agency (NCTTFA); (b) supporting the respective Governments to establish an appropriate

1 This sub-component was subsequently funded by AfDB in parallel. 3

management mechanism, the Central Corridor Transit Transport Facilitation Agency (CCTTFA) for the Central Transport Corridor connecting Dar es Salaam with the Great Lakes countries; and (c) helping improve the regional transport policy and harmonize transport regulations.

16. Component 3, Investment Support for Trade and Transport Facilitation, included: (a) enhancing security and facilitation (e.g. Community Based IT Systems) in the ports of Mombasa and Dar es Salaam; (b) improving goods security by financing a regional/national electronic cargo tracking system from the ports of Mombasa and Dar es Salaam throughout the EAC and Rwanda; (c) financing the establishment of key joint border posts at main cross-border posts within the region; (d) investing in Inland Container Depots (ICDs), intermodal infrastructure; and (e) supporting the implementation of a common Weighbridge policy.

17. Component 4, Support to the Concession of the Railways, included the following: (a) technical support to the Kenya Railways Company (KRC) and the Uganda Railways Company (URC); (b) support for the retrenchment and social mitigation of KRC staff; (c) support for establishment of a Pension Fund for the staff of KRC; (d) support for the Kenya Relocation Action Plan (RAP) implementation; (e) investment support for Uganda Railways Corporation (URC); and (f) support to the Joint Railways Concession through the provision of PRGs to Rift Valley Railway (RVR) Corporation subsidiaries (concession companies in Kenya and Uganda set up to enter into and implement the Concession Agreements) to backstop Government of Uganda (GoU)/URC and Government of Kenya (GoK)/KRC contractual obligations to the concession companies under the two Concession Agreements.

18. The coverage of the PRGs was and remains limited to contractually stipulated termination payments owed to either of the concession companies by GoK or GoU in the event of an early termination by the concession company (companies) of the relevant Concession Agreement(s) as a result of one of the following Guaranteed Events: (a) termination due to a breach (including expropriation and the imposition of materially adverse changes in law) of the Concession Agreement by GoK or GoU; (b) government breach of the exclusivity provisions of the Concession Agreement; and (c) termination due to force majeure events. The first component was intended to support the attainment of PDO sub-objective 1, the second and third components intended to support the attainment of sub-objective 2, and the fourth component the attainment of PDO sub-objective 3.

1.6 Revised Components

19. Over the nearly ten-year period of project implementation, whilst the scope and detail of some sub-components, and the associated costs, changed markedly in some cases, as design work progressed and experience identified useful improvements, a significant amount of progress was made. No new components were introduced, and some were dropped and funds reallocated. In face of cost overruns, the financing of several of the proposed One-Stop Border Posts (OSBP) originally envisaged to be financed by IDA, was

4 taken over by TradeMark East Africa (TMEA)2 and funded in parallel with a UK£20 million (US$30 million equivalent) grant. TMEA also provided parallel technical support to a number of the project’s implementing agencies. TMEA also subsequently took over the financing of the Port Community Systems (PCS) in Dar es Salaam port in Tanzania. A summary of the main changes to components is provided in Table 3.

Table 3: Changes to Project Components: By Country Kenya 2011 Scope of resettlement was increased from about 10 meters alongside railway tracks (US$11 million) to about 60 meters (US$39 million) after request from the Government of Kenya; Increase in retrenchment costs for KRC employees; Increase in security investment in Mombasa Port (US$10 million to US$21.4 million); Construction of new weighbridges at Mariakani and Athi River; Malaba bridge construction dropped from project; and Port Community System (PCS) changed to National Single Window (NSW) system. 2014 NSW dropped due to government decision to implement proprietary system provided by Singapore; and Funds reallocated to cover increases in costs of railway resettlement. Tanzania 2011 Increase in funds required for port security investment; Port Community System (PCS) changed to National Single Window (NSW) system; Increase in construction costs for Inland Container Depot (ICD) at Isaka; Change in implementing agency for OSBP component from TANROADS to TRA; and Change in weighbridge location from Kibaha to Vigwaza. 2014 NSW and remaining port security activities dropped, or taken over by TMEA, due to delays in implementation; and Funds reallocated to border post activities. Uganda 2011 Only one (of two) wagon ferries to be rehabilitated due to increased costs – GoU choose to use own resources to rehabilitate MV Pamba; Relocation of planned ICD from Kampala to Mukono and increase in costs; and Reallocation of funds to cover increase in border crossing improvements and costs of electronic cargo tracking. Rwanda 2011 Change in Implementing Agency, with establishment of RTDA; Electronic Single Window dropped from the project (Government choose to fund from own resources); Pipeline feasibility study dropped; and Increased support to monitoring and evaluation and border crossing facility investment.

2 TMEA was an agency created in 2009 specifically to support trade reform and expansion in East Africa, and supported by a number of Development Partners including the United Kingdom Department of International Development (DFID). 5

2014 Electronic Cargo Tracking dropped due to decision to implement under COMESA grant; funds reallocated to border post activities.

1.7 Other significant changes

20. The 2011 restructuring (Level 1) also included the provision of Additional Financing (AF) and extension of the project closing date from September 30, 2011 to September 30, 2014. The scope of resettlement also increased significantly, following an explicit request to accommodate a wider safe operation zone of 60 meters. This increased the scope of resettlement from approximately 3,400 structures or households to around 10,000 structures or households (an estimated 30,000 people). The AF was intended to provide funding to cover cost overruns for delayed activities, fund the expanded scope of resettlement, and give additional time to complete project activities. The safeguard category was also changed from B to A as part of the same restructuring, reflecting the increased safeguard risks from the increased resettlement. A further restructuring (Level 2) was approved on September 30, 2014, extending the closing date of the project to September 30, 2015, together with a further reallocation across categories.

2. Key Factors Affecting Implementation and Outcomes 2.1 Project Preparation, Design and Quality at Entry

21. The Project Appraisal Document (PAD) notes that the East Africa region at the time of preparation had extremely high transport costs compared to other regions of the world and that the benefits of the prospective EAC Customs Union hinged upon successfully eliminating non-tariff trade barriers and establishing effective joint customs administration. Meaningful gains in regional freight movement would require simultaneous improvements to customs, transit, and transport (port, rail, and highway) across multiple countries. Given this, the scope of interventions appeared well considered and comprehensive enough to tackle the nature of the challenges at hand. The project activities were to be supported by regional components, implemented by the EAC Secretariat, and the establishment and development of a corridor committee on the central corridor, which was financed in parallel by the African Development Bank (AfDB).

22. The project was innovative and path-breaking in many ways, and contributed in a substantive way to overcoming some of the trade facilitation challenges in the region, the nature of the challenge led to a complex and excessively broad project design. The PAD was explicit in highlighting the associated risks, and recognizing the complexity as a Substantial risk to achievement of the PDO. It also noted past experience with regional corridor projects which suggested key measures of success would be: (i) strong political will; (ii) stakeholder engagement; and (iii) efficient monitoring tools. The PAD also noted that extensive coordination would be required. A regional steering committee was established in September 2005, during project preparation. It was intended to ensure the establishment of a lead ministry and focal person within that Ministry, establishment of country Project Steering Committees, and establishment of Project Implementation Teams (PITs) within each Implementing Agency. Joint implementation teams would be established to carry out tasks that required joint planning and execution by multiple countries. The project design relied heavily on coordination mechanisms and the Bank

6 team placed substantial faith on the ability of the EAC Secretariat to take a proactive role in facilitating collective action.

23. The support to the joint railway concessions built on the IFC’s Advisory Group work as transaction adviser to the Government of Kenya on structuring of the Joint Kenya Uganda railway concession.3 The 25 year concessions were structured on the basis that the railway system could be commercially viable if run efficiently. Consequently, the commercial risks, principally the operational, investment, and traffic risks, were transferred to the Concessionaire. Nevertheless, the concessions included minimum investment levels for the first five years, minimum growth targets for freight traffic, and safety, health and environmental requirements among other covenants. The railway restructuring carried the risks of negative social impact, particularly on retrenchment and resettlement, and political risk (e.g. Governments’ failure to honour their commitments under the Concession Agreements), particularly regarding the payment risk in the event of early termination or at expiry of the concession.

24. IDA support to the railway concession (Component 4) was structured to mitigate those two risks. The PRG sub-component mitigated political risk related to termination or expiry payments while the Credit sub-component helped manage the negative impact on retrenchment and resettlement and build capacity to monitor concession compliance. A Quality Assurance Group’s (QAG) ‘Assessment of the Quality at Entry of IBRD/IDA Guarantees Fiscal Year 2005-2006’ found Component 4 well-prepared, with the PRG being suitable and innovative. QAG review also indicated that the PRGs and credit sub- components under Component 4 complemented each other. Certainly without this support, railway services would very likely have continued to deteriorate, with further traffic diverting to the road mode, further increasing the concomitant social and economic costs.

25. During preparation, the Bank team contemplated the division of the project into railway and non-railway components. The concession for the railway, the necessary resettlement along the right of way, the retrenchment of the railway labor, the establishment of the pension scheme, and the provision of the two IDA financed PRGs, essentially represented a high risk project on their own. However, Bank management essentially decided against this proposed division for four reasons: Firstly, separation would have required an increase in preparation and supervision resources, and would have increased the share of transport operations in the IDA portfolio. Secondly, the transit and customs aspects of the project were felt to support rail operations. Thirdly, a successful railway concession increasing corridor competition for freight with road was seen as central to lowering transport costs along the Northern Corridor. And finally, dividing the project was seen as a potential risk to the concession process.

26. This decision resulted in serious shortcomings: Firstly, a number of the key activities (the concession, and supervision of the railway), sub-elements of the PDO, and related PDO indicators were outside the scope of activities in the project. Secondly, whilst important to the corridor, the railway components were incorporated into a complex project

3 Canarail, a leading railway engineering consulting firm from Canada, was the adviser to the Government of Uganda 7 not well suited to tackle those specific challenges. Whilst the QAG assessment, mentioned above, found the proposed PRG design suitable and innovative, it strongly disagreed with their incorporation into this project due to the added complexity. The PRG sub-component overstretched the team’s ability to support implementation of an already complex project.

27. The already wide scope of the project was broadened beyond reasonable limits by the resettlement activities along the rail corridor, and the higher than envisaged associated risks. The proposed resettlement, even on the smaller scale initially envisaged, was a risky and challenging undertaking for a number of reasons: (i) the sheer scale of the expropriation which was originally estimated at 3400 structures or households, involving possibly 10,000 people; (ii) the socio-economic context of the Mukuru (with an estimated 250,000 residents) and Kibera (with an estimated 1 million residents) settlements, the political economy of which little was known a priori; (iii) the spatial nature of the context, as the area to be resettled was part of an integral urban fabric, which represented the upgrading of an informal settlement, rather than the simpler removal of piecemeal structures along a linear intervention; and (iv) the initial disagreement with the Government of Kenya (GoK) as to the scope of the resettlement, with the latter seeking the clearance of a 60 meter wide zone. The World Bank Group, in accordance with policy (OP 4.12), wanted to minimize resettlement to what was considered absolutely necessary to ensure personal safety, suggesting a right of way (ROW) of 10.4 meters or 5.2 meters on each side of the tracks.

28. Overall, whilst the PAD rated the risk to the achievement of the PDO as Substantial, this rating substantially underestimated project risks, and did not explicitly recognise others. At approval the risk associated with resettlement was rated only as Moderate. Project design and preparation, while innovative, timely, and strong in many respects, was far too ambitious and ultimately overlooked or underemphasized key risks.

2.2 Implementation

29. The project was innovative and path-breaking in many ways, and contributed in a substantive way to overcoming some of the trade facilitation challenges in the region. Some of the many positives of the project were the increased emphasis on the soft aspects of trade and transport facilitation, the physical infrastructure improvements and resulting time savings at the One Stop Border Posts (OSBP), the improved compliance with the ISPS code in the two ports, the port community system in Mombasa, and the increased focus on monitoring performance and reducing transit time along the corridors.

30. However, the multi-country nature of the project also introduced challenges and added to the complexity: The cross-effectiveness clauses in the legal agreements meant that the program initially progressed at the pace of the slowest country. To the extent that each government brought its own concerns and priorities, the reverberations sometimes had broader impacts. Limited capacities in the myriad implementing entities required more time in undertaking the necessary preparations, which included bringing on-board and building capacity in project management, project coordinators, project accounts, procurement specialists, and project implementation teams. And some of the cross-border activities initially proved difficult to co-ordinate, due to the sovereign nature of credit

8 provision which required contracts to be procured and implemented by country, and necessitating parallel contracts for singular activities; the establishment of the OSBP and the introduction of a regional cargo tracking system as two examples. Even after parallel implementation was agreed, complementary readiness proved difficult, as evidenced by the OSBP at Gatuna, where ground conditions on the Rwanda side led to delay, when the Government of Uganda (GoU) was ready to proceed on the Katuna side.

31. By the middle of 2007 the Bank was also concerned about slow progress in Kibera and Mukuru informal settlements in Nairobi. Delay in the procurement of consultants to prepare and implement the Resettlement Action Plan (RAP) led to a downgrading of the project’s overall safeguard rating to Moderately Satisfactory. It was eventually agreed with the GoK that action to mobilize the consultancy would be taken by early January 2008. The unfortunate post-election violence in 2008 engendered further delay.

32. The Mid-Term Review (MTR) was scheduled for 24 months after the effectiveness of all three credits. But as the Uganda credit was not declared effective until April 2007, the MTR was not carried out until late 2008. The MTR found the PDO and components still relevant and achievable and did not recommend restructuring. Delays and cost over- runs were identified as concerns. All four countries were flagged for ‘slow progress’ or ‘slow disbursement’. For example, disbursements were yet to exceed 48 percent, 19 percent, 12 percent and 5 percent for Kenya, Rwanda, Uganda, and Tanzania, respectively. Kenya, which had disbursed nearly half its credit, was actually substantially behind schedule; by the MTR, 88 percent of Kenya’s disbursements related only to retrenchment while other activities lagged. Implementation Progress ratings were downgraded from Moderately Satisfactory to Moderately Unsatisfactory for Uganda, Kenya, and Tanzania. The MTR also anticipated that delays to implementation would lead to cost overruns and called upon the three governments to consider the magnitude and how they could be accommodated.

33. The MTR called for preparation of implementation recovery action plans (IRAPs) and for substantially stepped-up internal coordination by all countries except Rwanda. The IRAPs would aid project monitoring and speed implementation through the first half of 2009. Monthly meetings between Project Implementation Teams and the Bank were to begin late January 2009 and the Project Steering committee was to meet three times in the six months following the MTR. Country-specific issues raised in the MTR primarily focused on delays to implementation and the very slow progress of engineering and design work for the more substantial civil works elements of the project, such as the border posts and highway weigh stations. For instance, project management Technical Assistance (TA) activities in Uganda would focus on improving procurement capacity of MOWT to implement the Uganda IRAP, and MOWT would accelerate efforts to have these consultants in place by March 2009. These efforts appear to have helped make up for the time lost and to increase progress on many of the issues covered by the project.

34. The 2009 train derailment caused several fatalities in Kibera and allowed the GoK to return to its earlier position of requiring a wider right of way, despite the opposition of the Bank technical staff. A Lead Social Development Specialist was placed in Nairobi to oversee the resettlement process. By January 2011 a new RAP had been prepared

9 confirming a right of way of 40 meters, 20 meters on each side of the tracks and the resettlement of those households and other buildings within the next 10 meters on each side for a total work area of 60 meters in two defined three-kilometer-long safety zones, one in Kibera and the other in Mukuru. Substantial changes to the security infrastructure were also incorporated including a nine meter high wall between the settlements and the tracks. The number of households requiring resettlement grew from 3,400 to about 10,000, involving some 30,000 people. The social safeguard rating was raised back to Moderately Satisfactory but the implementation of the RAP was considered high risk. The increased cost of this new plan, together with significant increases in cost-estimates for most of the civil-works elements of the project in Kenya, required the provision of Additional Financing (AF) of US$30 million, a change in the safeguard categorization from B to A, and an extension of the project closing date to September 2014. These changes were approved in June 2011. A Resettlement Management Unit (ReMU) was established and staffed by the GoK, and the RAP census verified by early 2012.

35. The weak performance of the railway concession was also a serious shortfall of the project’s early years and the project PDO was revised as part of the 2011 restructuring to try and simplify the results framework and focus on results that directly reflected project activities. The restructuring of the third element of the PDO from improving railway services in Kenya and Uganda to enhancing safety in identified areas and reducing the Recipient’s fiscal transfer to railway institutions by rationalizing the work force in practice excluded the PRG subcomponent from the PDO. Accordingly all indicators relating to the efficiency, operation or management of the railway concessions were dropped without adding any new indicator to monitor and assess the progress of the PRG subcomponent. The justification contained in the 2011 Project Paper for eliminating the indicators was that the two Concessions were performing poorly for reasons, which by design, were not controlled by the project. As a result from this point on, the Results Framework did not have any meaningful indicator to monitor the performance of the railway under the two IDA PRGs, when those guarantees became effective in June 2012.

36. In addition, the immobilizing RVR Board divisions which had such negative effects on the functioning of the Kenya-Uganda Rift Valley Railway (RVR) concession were at last resolved when an Egyptian company Qalaa, bought out in 2010 the full shareholding of the South African Sheltam Rail Company (Pty) Ltd., the leading original investor.4 Thereafter, an IDA Guarantee Agreement with Rift Valley Railway Company was signed on May 14, 2012 and became effective on June 22, 2012. In July 2012, the rehabilitation of the dedicated train ferry Motor Vessel (MV) Kaawa was completed. This was intended to complement the mostly privately owned ferries which, since 2005, had been the sole source of service (mostly for petroleum products and grains) on . However, the operation of the wagon ferry, which was still based on old technology, was the responsibility of the concessionaire RVR. Without a functioning railway connection to , or , full utilization of the rehabilitated wagon ferry was unlikely, an issue that was still pending at project close.

4 Qalaa became the controlling shareholder with Transcentury due to the nature of their shares (A controlling shares) while DFIs (DEG, AMC, IFC, Proparco and FMO) who hold the majority of the shares combined only had B – non controlling shares. 10

37. Simultaneous occurrence of new delays on several dimensions of the project early in 2013 gave rise to worries about effective completion of the project by the revised September 2014 closing date. The award of one contract to develop an OSBP in Uganda was challenged by a losing bidder. There were delays in appointing consultants to prepare the next stage of work on the Gatuna OSBP in Rwanda. Kenya had been able to award only two of the four contracts required for the RAP works in Kibera and Mukuru. The project’s performance rating on both Development Objectives and Implementation Progress was downgraded in July 2013 to Moderately Unsatisfactory. A further restructuring was approved on September 30, 2014, extending the closing date by one year to September 30, 2015.

38. Attention now turned to the completion of the RAP implementation, and the two outstanding contracts were signed in August 2013. A delicate issue nonetheless arose between the GoK and the Bank as to how to treat absentee owners and multiple structure owners in the affected area: The GoK position was to provide support to squatters occupying public land but not to compensate people who had profited from illegal building and leasing. Bank policy requires compensation to be provided to any owners/occupants of lands taken over for public purposes – so that nobody should be worse off due to the project unless they have been speculating through acquisition of land or other assets to benefit from such assistance.

39. The matter was eventually resolved by agreement, as in one or two other countries where the issue had arisen, to provide some degree of compensation to owners of multiple plots but on a less generous basis than for single units – in the Nairobi case, a certain cash level for six months of revenues for each structure leased. When it became clear that the RAP could not be fully implemented before project closure at the end of 2015, a detailed Addendum (to the 2011 RAP) was agreed, strengthening the ReMU, clarifying grievance mechanisms, spelling out the provisions for multiple-plot owners, and providing for continued World Bank monitoring of RAP implementation. It was also agreed that Government would establish an escrow account in the amount of some US$860,000, drawn from the remaining balance of the IDA Credit and being the estimated total amount required for the multiple-plot owners. As at the end of January 2016, 51 percent of the necessary residential units, and 29 percent of the necessary business units had been built, and a total of 1,256 households relocated.

40. The World Bank has been closely supervising RAP implementation through regular meetings with senior management of the Kenya Railways and Ministry of Transport and Infrastructure and site visits with the most recent meeting being on September 15, 2016. At that time it was established that the most of the Absentee Structure Owners and (ASO) and Multiple Structure Owners (MSOs) have been compensated. However a total of KES 2,052,000.00 is yet to be paid due to claimants who are yet to provide their account numbers as at August 31, 2016.

41. A social audit was undertaken followed by a through vetting process through it has been determined that a total of 7,429 residential, institutional and business units will be required, down from the original projection of 9,005 units. A total of 3,214 of PAPs have been relocated. As of September 7, 2016, a total of 70 grievance cases were pending before

11 the Grievance Appeals Board (GAB). A total of 324 learners relocated from the informal schools to 10 public schools. The schools requested the project for assistance with the following requirements, whose purchase will be initiated soon: Desks and chairs; Improvement to water services; Ablution blocks; and Improvement to kitchens. Construction is ongoing and it is anticipated that construction will be completed in January 2017.

2.3 Monitoring and Evaluation (M&E) Design, Implementation and Utilization

42. The chosen indicators covered institutional changes relating to development of the Customs Union and important samples of the stages in fulfilment of an international trade transaction that were expected to be accomplished more quickly thanks to use of electronic equipment and software appropriate for trade and customs. The selected indicators were mostly ones that were already widely gathered by the sectoral institutions involved, especially the ports, the national revenue authorities responsible for collection of customs duties, and the respective line ministries for transport/works. The selection of these indicators helped to unify definitions among all participating countries and to improve capture of hard-to-measure issues of importance like the proportion of cases taking more than normal average time (e.g., container dwell-time at port before release, or number of days required to cross a transit-country) and the scale of the performance shortfall – so critical to reliability of shipment delivery date. However, there were some missing baselines and targets for some PDO indicators, and a confounding of PDO and Intermediate Output Indicators in the Results Framework (RF) at approval.

43. With few exceptions, the data carefully selected and identified in the Results Framework in the PAD were collected from the start of the project by the institutions directly concerned, especially ports and customs in the early years. After analysing them it was observed that there was less consistency between the RF in the PAD at approval, and the reduced set of indicators reported in the ISRs until the time of the Additional Financing in 2011. After that date, the RF was revised to improve the clarity of indicators, improve alignment with the revised PDO, and increase the focus on key outcomes. Following those revisions, the results reporting in the ISRs improved markedly, although the number of beneficiaries continued to go unreported.

44. Over the project period, there has been increased expansion of Monitoring and Evaluation efforts, and of the range of entities involved in them, for these large-scale trade and transport services. Important contributions were made from the beginning by specially recruited public-private teams with EAC Secretariat participation, in the form especially of so-called Seamless Committees – amongst other things to check the degree of seamlessness in fact achieved on major EAC routes, or, as on many occasions over the years, to provide a sound mechanism for serious follow-up on all complaints submitted about Non-Tariff Barriers on the trade corridors. The latest of the latter reviews (2015) showed that a cumulative 70 cases had by then been satisfactorily resolved, 20 were still being pursued, and four new cases had been submitted. They are mostly matters of adjusting old laws or practices to the new overall legal framework of a Customs Union.

12

45. The inter-governmental Northern Corridor Transit and Transport Coordination Authority (NCTTCA), founded in the 1980s, has long fulfilled a Monitoring and Evaluation role for the Corridor whose management it coordinates for the different countries involved, and in 2012 it added a Transport Observatory Facility providing a continuous flow of up-to-date information. The countries involved in the Central Corridor, originating at Dar es Salaam, used the project to advance the long-discussed idea of creating a similar body on behalf of its member countries and with somewhat more participation of the private sector. The Central Corridor Transit Transport Facilitation Agency (CCTTFA) was agreed among the countries concerned in 2006 and began operations in 2010, initially supported by AfDB, but subsequently self-financing after a small levy was introduced on consignments to member countries.

46. But perhaps potentially one important outcome of the Monitoring and Evaluation effort in connection with the project has been the extent to which it has encouraged broader community initiatives in the same direction, generating at least the possibility of wider and stronger commitment by all the parties involved in a service to fulfil their part in the total. The Mombasa Port Community Charter was developed in 2013-14 mainly by the initiative of Kenya Port Authority, Revenue Authority and Bureau of Standards, with strong support from TMEA and the Shippers Council, and extensive participation of the wide variety of private users of the port. It thus committed its many participants to their particular part in the overall improvement of the performance of both the port and the wider regional freight transport system.

2.4 Safeguard and Fiduciary Compliance

47. The most serious risks anticipated in execution of the project were connected with the railways, and especially the small stretches of track, mainly in Nairobi, where dense informal settlements had developed dangerously close to the railway line. The staff retrenchment and pension plans were consistent with past practice and existing agreements, and had been extensively discussed with the labor unions, the quality of the preparation reflecting the lack of discord engendered by these major social changes. The process also benefitted from earlier concession arrangements in Africa, involving an independent auditor to review actual payment and validate that no fraud was taking place. The auditor issued a Retrenchment Completion Report in 2014.

48. The concession companies were required to have competent health and safety officers. They had to maintain and follow up-to-date environment protection plans. And they had to recognize the same responsibilities as the existing government companies and the IFC for protection and gradual easement of the situation of the people living along the track. Preparation and implementation of the needed resettlement action plans were on the other hand fully the responsibility of the governments, mainly Kenya, with the help of the Bank. The Resettlement Management Unit created in 2011 to handle the enlarged program set effective grievance and appeal procedures and will prepare an assessment of the program’s social quality and effectiveness when resettlement is complete.

49. The financial management and procurement capacities of the main implementing agencies involved in all four countries were assessed during project preparation and found

13 to be satisfactory. Most of these bodies had already been involved in Bank-financed projects. A few specific measures were identified for action before the start of project implementation, mainly in Rwanda and Tanzania, for institutions which had not previously worked with the Bank. Tanzania, Kenya and Uganda were found to be at advanced stages of implementing procurement reform aimed at enhancing transparency and value for money from their procurement systems. Rwanda was at an early stage of the process, with a shortage of adequately trained staff and a lack of established procedures. Procurement risk was overall acceptable, provided special attention was given to this aspect in Rwanda and in one of the Tanzanian sectoral entities which was newly established.

50. Despite these positive assessments at the time of project appraisal, a number of disagreements arose in the course of project execution that were sufficiently serious to cause several sub-components of the original project – in three of the four countries – to be dropped from IDA financing. The reasons appear to have been most often concerns that the borrowers had not always followed fully the procurement guidelines, or had failed to secure approval from the Bank for specific steps to move the procurement process forward for a particular item. All the problems involved the commissioning of consultants for execution of studies, the development of software, or design and delivery of computer systems. They were generally resolved by finding alternative financing sources, internal or external, for the specific procurement in question.

2.5 Post-completion Operation/Next Phase

51. The main outstanding priority at the time of project close related to the completion of the resettlement of people living within the operational safety zone in Mukura and Kibera. As at the end of January 2016, 51 percent of the necessary residential units, and 29 percent of the necessary business units had been built, and a total of 1,256 households relocated. Given continued progress at this rate, the resettlement process is expected to be completed around December 31, 2016.

52. A second high priority at project close relates to the lack of clarity about future prospects for the problems in the railway sector – the performance of the concession on the one hand and, on the other hand, the continued exposure of IDA via the PRGs. In terms of the former, improving the performance of the railway sector remains crucial to reduce the costs of transport, but equally importantly, reduce some of the social and economic costs associated with the current imbalance in favour of road transport. In terms of the latter, IDA supervision of the PRGs will need to continue for as long as they remain in place.

53. A third high priority is full introduction of Integrated Border Management, to enable efficient use of the new facilities and ensure border post officials from both sides of the border work together to get best results from the One-Stop Border Posts. Stronger and wider efforts are also required, for example, to better prepare truck drivers on documents they should carry, on electronic cargo tracking arrangements, and, in particular, on adherence to vehicle weight limits and the importance of good bracing and blocking of goods inside containers to avoid slips causing overloading of particular axles. In all of these and similar areas, it is urgent now to update and broaden skills and flexibility of

14 people to enable them, and of course the country more broadly, to get full benefits from the large investment made in improving technology and buying equipment.

3. Assessment of Outcomes

3.1 Relevance of Objectives, Design and Implementation

54. The relevance of objectives is substantial. Fuller economic integration with other East African countries, and strong increase of international trade with the neighbors and equally with more distant countries, remain widely supported priorities in all the EAC countries. The frequency of highly publicized Presidential initiatives related to the Customs Union, neighboring countries, and the role of their own nation in providing them vital access infrastructure and services, is testimony to widespread popular support for the EAC link. The Country Assistance Strategies for all the EAC countries have maintained an emphasis on trade expansion as a key basis for wider growth of employment, incomes and social well-being.

55. The relevance of design and implementation is moderate. The complex project design was reflective of the scope of the objectives, and ambitious even without the railway sub-component. It is clear the addition of the latter, and the extension of the required resettlement, was too much, and by closing date, the design required simplification. In addition, the implementation arrangements were not adequate to support the breadth of the proposed interventions, and the complementary policy initiatives, were not implemented in parallel, or even ex ante, as would have been optimal.

3.2 Achievement of Project Development Objectives

3.2.1 Achievement of the original PDOs by the first restructuring in 2011, when PDO indicators were modified Rating: Moderately Unsatisfactory

56. The original PDO had three main sub-elements: (i) to improve the countries’ trade environment through effective implementation of the EAC Customs Union Protocol; (ii) to enhance transport and logistics services’ efficiency along key corridors by reducing non- tariff barriers and the uncertainty of transit time; and (iii) to improve railway services in Kenya and Uganda.

57. Sub-Objective 1: to improve the countries’ trade environment through effective implementation of the EAC Customs Union Protocol. Rating: Moderately Satisfactory

58. It is clear that substantial progress was made, initially (2006-09) through the rationalization and simplification of documentary requirements, along with increased use of scanners, and, more importantly, through the electronic systems innovations and thorough review of all customs management processes, led by Kenya and Uganda. Much increased joint work between the countries, including joint offices and joint inspections at country borders and ports, was also noted. Four out of the five PDO indicators were met by the time of the 2001 restructuring.

15

59. Sub- Objective 2: to enhance transport and logistics services’ efficiency along key corridors by reducing non-tariff barriers and the uncertainty of transit time. Rating: Moderately Satisfactory.

60. The wealth of data that has been generated on these issues, as referred to early, particularly since 2009, and especially by the Northern and Central Corridor Authorities, permits a conclusively positive verdict on this aspect up to 2011. The removal of the reference to NTBs in the 2011 restructuring is acknowledgement of the failure of the project to realize any improvements in removing these barriers. Nine out of the thirteen PDO indicators were met by the time of the 2001 restructuring.

61. Sub-Objective 3: to improve railway services in Kenya and Uganda. Rating: Unsatisfactory.

62. No substantive improvements were realized in the performance of the railway under the project prior to the first restructuring. The third element of the PDO was rewritten at the time of the 2011 restructuring in order to be more realistic as to how much the Bank might be able to contribute to what the countries would achieve in the area of railway reform and renewal. Only one fully and one partially, of the five PDO indicators had been met by the time of the 2011 restructuring.

3.2.3 Achievement of the revised PDOs achievement by the project closing date in 2015 Rating: Moderately Satisfactory

63. The revised PDO was: (i) to enhance the efficiency of customs agencies’ clearance processes for the countries participating in the EAC Customs Union; (ii) to improve efficiency and reliability of transport and logistics services along the key corridors; and (iii) to enhance safety in identified areas and reduce the Recipient’s fiscal transfers to railway institutions by rationalizing the work force on the Kenya-Uganda railway.

64. Sub-Objective 1: to enhance the efficiency of customs agencies’ clearance processes for the countries participating in the EAC Customs Union. Rating: Satisfactory

65. It is clear that substantial progress has been made, initially (2006-09) through some rationalization and simplification of documentary requirements, along with increased use of scanners, and, more importantly in the last five years, through the electronic systems innovations and thorough review of all customs management processes, led by Kenya and Uganda. Much increased joint work between the countries, including joint offices and joint inspections at country borders and ports, was also noted.

66. The results of all the innovations are very likely to further improve in the next two- three years from fuller use of what has already been done. What few specific results indicators were included in the 2011 Board Paper for the Additional Financing (mainly container transit-time Mombasa-Kigali, defined in a specific way and projected to drop over the project execution period from nine days in 2010 to five days in 2014) have in fact been fulfilled, as shown in Annex 3. But World Customs Organization indicators that were also mentioned in the appraisal turn out to be available only for two of the countries, and

16 even for them only very partially. The only indicator not achieved under this element was the creation of the regional Customs valuation database.

67. Sub- Objective 2: to improve efficiency and reliability of transport and logistics services along the key corridors. Rating: Satisfactory.

68. The second element of the PDO was to improve efficiency and reliability of transport and logistics services along the key corridors. The wealth of data that has been generated on these issues, as referred to above, particularly since 2009, and especially by the Northern and Central Corridor Authorities, permits a conclusively positive verdict on this aspect. The latest data with regard to the specific indicators selected to illustrate impact show 2014 dwell time for a standard (20 foot equivalent unit, or TEU) container at Mombasa of three days (Seamless Committee) and at Dar es Salaam of 6.75 days (Central Corridor Transport Observatory), close to the targets adopted in provision of the Additional Financing: reduction from the 2010 bases of eight days at Mombasa to three days by project completion and at Dar es Salaam from 14 days to eight.

69. The Bank’s 2011 Project Paper for the AF also wisely gave special attention to the important subject of reliability and predictability of deliveries, especially for the last leg of import journeys, from port to final destination. The target adopted in the 2011 appraisal was to reduce variance of transit time on the Northern Corridor from 2010 baseline of five days (additional, for Kigali, to the nine days normal average transit time) to a 2014 performance of two days (additional, for Kigali, to the five days average transit time). Records of the Seamless Committee show that these targets were met. Overall only 4 out of the 15 indicators for this element of the PDO were not fully achieved, and of those three reflect limited or incomplete data, or relate to components dropped from the project, funded in parallel by other development partners.

70. Sub-Objective 3: to enhance safety in identified areas and reduce the Recipient’s fiscal transfers to railway institutions by rationalizing the work force on the Kenya-Uganda railway. Rating: Moderately Unsatisfactory.

71. The railway Pension Scheme set up under the first credit has continued to operate satisfactorily, and the project investments related to railways (mainly the large Uganda Inland Container Depot (ICD) plus the rehabilitation of the wagon ferry on Lake Victoria) were completed. A further benefit is that the respective Governments have not been providing subsidies to railway operations, but rather receiving concession fees and significant tax revenues. However, only 1 out of 5 indicators has been fully achieved, although two others have been partially achieved.

3.3 Efficiency

72. Efficiency is rated substantial. The economic analysis in the 2005 PAD delineated project activities into three main areas: (i) the impact of the support for the CU and customs authorities on regional growth. The latter was rightly seen as a challenging task for project level analysis, and hence the economic analysis excluded the benefits from the quantitative analysis, but included the costs. (ii) The second element was a detailed financial analysis of the RVR concessions in Kenya and Uganda, which clearly indicated

17 the viability of the proposed concession. For the purposes of the ICR, this financial analysis has not been recalculated. The third and final element related to the economic costs and benefits of the trade facilitation measures, which would reduce cost, time, uncertainty, or the need for additional inventory on the part of traders in the region. This section presents a short summary of the original model, and presents an ex post analysis as a validation of the original work, and a justification of the activities in the project.

73. The expected impact of the project was divided in two stages: in the short term, the project would result in: (i) reducing transit time; (ii) reducing transport unpredictability; and (iii) reducing non-logistic costs, such as inventory and storage costs. In the medium term, the project was expected to contribute to higher transport quality and lower transportation tariffs. Indeed, reduced transport time was expected to lead to reduced transportation costs for transportation companies and increased transport supply. The decrease of transportation costs for transportation companies would increase the competitiveness of those transportation companies which bear high fixed costs i.e. those companies offering a better quality of service. To reach this objective, countries would have to improve their roads condition, but also most critically reduce delays and uncertainty caused by poor transport logistics, including increased dwell time in the ports, the time consumed at the border crossings, and at the inland clearance facilities, and the additional time and cost required to traverse poor quality infrastructure along the main corridors. The results of the economic analysis reported in the PAD, using a 12 percent discount rate, was a Net Present Value (NPV) of US$21.44 million, and an Economic Internal Rate of Return (EIRR) of 26 percent.

74. In 2014, when the project was nearing completion, an overall review of accomplishments was conducted for the project, specifically including development of global estimates of benefits generated, and drawing on the unusually comprehensive data that happened to have been generated for port and corridor operations in 2012. The results were not put in final form because they were after all not needed for further decision, but using the same logistical framework, showed that the economic benefits generated by the actual investments resulted in a NPV and EIRR that greatly exceeded those reported above, with the latter estimated to exceed 50 percent. While the final project costs were nearly twice as much as originally expected for OSBP construction or reconstruction, equipping new joint offices with extensive electronic systems, and refurnishing, this cannot be considered a poor investment. It seems rather to reflect limitations in the original cost estimates, little provision for necessary electronic equipment, and little allowance for the significant adjustments that would probably be needed when the two countries concerned would come into the picture to develop a locale well adapted to joint work and a variety of technical bureaus. The data available about construction costs of non-IDA-financed Border Posts give no indication of greater economies achieved.

75. In October 2014, acting at the request of its member states – both the EAC countries and their landlocked neighbors – the Northern Corridor Authority commissioned an independent evaluation of the real effects of the customs, trade and transport policy changes of the last four or five years on the performance of the corridor and its contribution to lowering costs of business and improving integration. The study built on a 2010 assessment, which had also emphasized wide canvassing of users’ experience and

18 extensive analysis of costs to users. The findings of the study were published in May 2015. For the period of interest 2010-14, it found that overall costs had fallen substantially (23- 56 percent for all six destinations) and some 40-50 percent for , Democratic Republic of the Congo and Kenya. But the composition was very informative: The direct trucking-costs component had fallen nearly as much (some 15-50 percent) as overall costs, though not in the same places. Needs for extra inventories had fallen very substantially (more than 50 percent in all but one case) due to reductions in delays and improvements in reliability. But marginal truck capacity had become more costly, due to pressures of strong demand against limited increase in fleets (except in Nairobi itself).

3.4 Justification of Overall Outcome Rating

76. The overall Outcome Rating is Moderately Satisfactory, based on the ratings of achievement of original and revised PDOs and rating of efficiency.

Table 5: Assessment of the Overall Outcome Against Against Revised Against Revised Overall Original PDOs PDOs (2011) PDOs (2014) 1. Relevance of: Substantial Substantial Substantial a) PDOs Substantial Substantial Substantial Substantial b) Design Substantial Substantial Substantial c) implementation 2. Moderately Moderately Moderately Achievement of the PDOs - Unsatisfactory Satisfactory Satisfactory 3. Efficiency Substantial Substantial Substantial Substantial

4. Moderately Moderately Moderately Overall rating - Unsatisfactory Satisfactory Satisfactory 5. Rating value 3 4 4 - 6. Weight (% disbursed 44% 66% 83% 100% before/after PDO change

7. Weighted value (3x4) 1.32 2.64 3.32 4

8. Moderately Final Rating - - - Satisfactory Satisfactory=5; Moderately Satisfactory=4; Moderately Unsatisfactory=3; Unsatisfactory=2

3.5 Overarching Themes, Other Outcomes and Impacts

(a) Poverty Impacts, Gender Aspects, and Social Development

77. Poverty Impact. The project did not measure its impact on poverty reduction. However, the expansion of trade with other countries is one of the most promising routes to spreading more productive employment in East Africa and thereby reducing poverty. In addition, there are clear indications that the project’s direct contribution to reducing poverty – through improving the housing conditions of substantial numbers of truly poor

19 people who had to be removed from locations only meters away from the railway lines in specific sections of Nairobi – has ultimately proved a considerable success.

78. Gender Impact. The project did not measure its impact on gender.

79. Social Development Impact: The project did not measure the Social Development Impact. But there are clear indications that the project’s direct contribution to social development – through improving the housing conditions of substantial numbers of truly poor people who had to be removed from locations only meters away from the railway lines in specific sections of Nairobi – has ultimately proved a considerable success.

(b) Institutional Change/Strengthening

80. Delivery of much increased volumes of trade and transport service in more timely and reliable manner and at lower unit cost obviously depended on significant strengthening of the many institutions responsible for the different parts of the service – and especially the Port and Highway authorities, Customs and the coastal countries’ Maritime Authorities. This was largely carried out by the leaders and managers of these national institutions, with some consultant assistance financed out of the IDA Credits, as for instance for port security at Mombasa, financial management in some bodies in Tanzania and Uganda, and the Customs service in Uganda and Rwanda.

(c) Other Unintended Outcomes and Impacts (positive or negative)

81. One unintended outcome of the project is the clear failure of the railway to increase either the volume of freight or the modal share of total freight traffic carried by rail transport on the corridor. The consequence of this outcome was the failure to contribute to a reduction in the significant externalities associated with the growth in road transport in transit along the corridor. The extreme congestion in many of the major urban areas, including Nairobi and Mombasa, and on the trunk highways is visible to all. Now the more comprehensive worldwide report5 on the health effects of motorized road transport that has just been produced shows that, even at its current relatively low levels of vehicle ownership but taking account of environmental pollution as well as traffic accidents, roads and their vehicles are one of the most serious sources of disease in Kenya, paralleled in few other parts of the world. These are outcomes that the railways concession had been very explicitly intended to help the region avoid.

4. Assessment of Risk to Development Outcome

Rating: Moderate

82. The customs, trade and trunk transport improvements of the last ten years are well established by now and likely to be sustained. The most serious uncertainty relates to Rift Valley Railways, given its financial performance, which raises questions regarding the financial health of the concessionaire. While the concessionaire invested US$139 million

5 Global Road Safety Facility and Institute for Health Metrics and Evaluation 2014. The Global Burden of Disease from Motorized Road Transport. 20 in network rehabilitation in 2011-2015, freight traffic, the main source of revenue, just started to increase in 2015 and passenger traffic continued to decline. RVR has reported losses of more than 10% of its revenue every single year since 2011. Equally important, but on a more limited geographical scale, is the risk that the solutions so carefully developed over the last few years to the Kibera/Mukuru resettlement problems might run into new difficulties in the course of execution. Both problems warrant the highest-priority attention, to help things stay on track and reinforce confidence in the region’s future.

5. Assessment of the Guarantee

5.1 Impact of the Guarantee in mobilizing Private Sector Finance

83. The PRG support was offered as an option to the bidders to enhance bid responsiveness at the recommendation of the governments and their transactions advisers. This strategy was adopted by IDA to mitigate the risk of a bid failure experience around that time by countries such as Republic of Congo and Tanzania. The government-related risks were a major source of concern due to the perception of high political risk in Africa, particularly for such long term arrangement. The selected bidder, Rift Valley Railway Consortium required the PRGs to undertake the railway concessions as it considered it crucial to its ability to mobilize long-term debt and equity. While valid for the entire concession term, the guarantee agreements allowed for elapsing the guarantee support anytime that the Concessionaire would become confident that it could manage the political and government related risks.

84. The capital expenditure required by the two concessions over their term was US$400 million. Financing was expected to come through equity, debt, and internal cash flows. IFC and KFW provided loans for a total of US$64 million while the sponsors agreed to provide US$24 million in equity. The Concessionaire took over both concessions in November 2006, but soon after ran into financial difficulties and was immerse in internal disputes. Concessionaire was unable to make the Guarantee Agreements effective, undertake the necessary investments, or turn around the railway network line. The Concessionaire was facing termination notices from both governments by January 2010. To prevent termination, lenders brought Citadel Capital of Egypt as a new lead investor. Citadel signed restructuring agreements with both governments in October 2010. The restructuring agreements allowed for some adjusting to the concession agreements (such as timing for achieving freight volume targets), but kept the key features in place.

5.2 Role and Value of the Guarantee in addressing critical risks and improving the overall sustainability of the transaction

85. The restructuring included a new financing plan which contemplated additional equity of US$82 million and debt of US$110 million. IFC was one of the main providers of the additional debt and equity financing. With the restructuring, the concessionaire finalized the PRG agreements which were signed in May 2012 and became effective soon after. The concessionaire indicated that they had invested US$139 million in 2011-2015.

21

6. Assessment of Bank and Borrower Performance

6.1 Bank Performance

(a) Bank Performance in Ensuring Quality at Entry

Rating: Moderately Unsatisfactory

86. The Bank responded with enterprise and imagination to assist the East African countries in embarking on a wholly new turn in their development, and the results already achieved show that the initiative was timely, innovative and worthwhile. Two distinct features of the initiative are particularly noteworthy: the broad logistics-based approach to the selection of investments and prioritization among them, and the simultaneous initiation of actions in interdependent fields. The logistics-based approach was vital to the sizing and packaging of interventions in different transport services, customs and communications. Simultaneity of action was of particular importance for reinforcing the political feasibility of sustaining action in concessioning the railways and paying off their excess staff, negotiating acceptable solutions to the Kibera/Mukuru safety and social problems, implementing the customs reform to the level of the Single Customs Territory, and modernizing port management with emphasis on port community mobilization.

87. However, this strength was itself a fundamental weakness as it led to a project design that was too broad, overly ambitious, and unnecessarily complex: Essentially, the project was six projects rolled into one. The aggregating of so many distinct components was required by the institutional incentive of minimising the number of projects, and hence preparation and supervision funds, whilst delivery was dictated less by technical readiness, and more by commitment demand. These issues were then further compounded by the decision of Bank management, against the advice of the team, to include the railway concession, the increased resettlement, and associated activities, within the project, and subsequently, the increased resettlement. Successful implementation was almost unattainable from the start.

88. A further weakness in the Bank’s approach which became visible in retrospect was the provision made for so many new initiatives, each in four countries, and not only in the specific areas highlighted above, and without any firm requirements for advance joint action by the countries to identify the best ways to bring the different initiatives to life in the region. The problem was to some extent increased by the admirable emphasis of the project designers on intensive and open discussions with each of the countries on their home ground, with encouragement to suggest their own projects for inclusion in the proposed program. There is also the structure of the project – 4 credits with cross- effectiveness guarantees, so the project went at the pace of the slowest

89. Once the IDA Credits became effective, some important customs initiatives, training efforts and small-scale studies envisaged in the country programs went ahead rapidly, but most of the work required of the countries was initiating the procurement of the necessary preparatory and design work for the wide variety of civil works. The volume of work seems in most cases to have been at or above the upper limit of their small teams’ capacities. This meant that the first 18 months of project implementation had to focus on

22 civil works procurement. So by 2008 the Bank was already concerned by lags in sub- project preparation and the slow progress of disbursements. Time that should have been devoted to serious joint efforts by the few qualified people from the EAC countries to discuss how the region could shape joint, mutually supportive solutions in areas where they were most needed was lost to the heavy work of preparing too many projects, not always of highest priority, in traditional single-country ways.

(b) Quality of Supervision

Rating: Moderately Unsatisfactory

90. The heavy load of civil works preparation, contracting and supervision in the countries – indeed, not infrequently, overload relative to the limited experienced staff available in the Ministries – took up most of the Bank’s implementation support effort, largely handled from Nairobi after 2009, with the aid of staff visiting from headquarters. Support for complex and sensitive RAP planning work for Kibera and Mukuru built up significantly over time, from inadequate initial levels, and came mainly from Washington, as also did other Environmental and Safeguards-related supervision.

91. The most disappointing aspects of the Bank’s implementation support was the failure to respond to broader problems that seemed to be emerging due to earlier errors for which the Bank was not directly responsible but in which it had nonetheless shared. Almost from the earliest days of the railways concession it become clear that the original projections were grossly over-optimistic and that the financial framework provided for the railways’ functioning was inconsistent with economic and financial efficiency. A proactive response would have involved the project being radically restructured/simplified so that the stakeholders involved could have focused on a much narrower range of achievable objectives/components/activities.

92. The question has been raised whether neglect of this vital topic might not have been avoided with more effective use by the Bank of the PRGs granted along with the Credits. More important than the shortfalls alleged by the 2006 QAG (para. 23 above) may have been the failure to require from the governments, in return for the PRG, a written commitment to pricing and costing approaches consistent with fair competition between the modes. And the question has already been raised whether, as long as the PRGs were maintained, it was permissible for the Bank to downgrade (as it did in connection with the 2011 Additional Financing) immediate objectives of the support for the railways from full economic and financial success to effective completion of the particular programs (retrenchment, pensions and RAP) supported.

93. While the PRGs have certainly played a useful role in attracting and retaining potential concessionaires (including Qalaa), the consistency of financial arrangements for transport infrastructures with sound intermodal competition will continue to require priority attention in a wider range of cases. With or without PRGs, the main objective of any Bank-supported railway operation must clearly be full success in serving the traffic for which it has comparative advantage, and this continues to apply to RVR. The failure of the Bank’s supervision to give more attention to these matters, and the lack of supervision

23 of the railway component despite the IDA support for the PRGs, were significant shortcomings.

(c) Justification of Rating for Overall Bank Performance

Rating: Moderately Unsatisfactory

6.2 Borrower Performance

KENYA

(a) Government Performance

Rating: Moderately Unsatisfactory

94. Within the East African Community, the GoK provided strong leadership to the broad objectives of the project. Project implementation nonetheless moved very slowly because of the weakness of government agencies and the novelty of the issues raised. But the last five years of project implementation saw many significant developments, directly promoted by the government of Kenya including inter alia: combined port community efforts to increase transport efficiency and reliability, management of import duty payments on a SCT basis, development of acceptable solutions to the difficult housing and safety problems suffered by many thousands of people in Kibera and Mukuru, to take only three examples.

(b) Implementing Agency or Agencies Performance

Rating: Moderately Unsatisfactory

95. The Ministry of Transport and Infrastructure and its various agencies had the broadest role in project execution, with responsibility for contracting design and construction of most of the civil works, while the Kenya Port Authority played an important role in regulating and managing some of the transport services most important for international trade. Both had their own problems of shortage of qualified staff, often old-fashioned office procedures, and difficulty in adhering to tight deadlines. The Ministry of Transport’s weaknesses were a major cause of slow progress in design and contracting of works, especially in the early years, and it was not an effective interlocutor with the Bank on project costs or results monitoring, but it played a valuable role in later stages, especially with the Kenya National Highways Authority (KeNHA) in promoting wider use of Weigh-in-Motion technologies to reduce truck transit delays. The port was naturally blamed when traffic delays did arise, but Mombasa managed to maintain generally positive performance trends throughout the ten years. Illustrative of its broader contributions were its lead in developing, the Kenyan Electronic Single Window, and the Port Community Charter.

24

UGANDA

(a) Government Performance

Rating: Moderately Unsatisfactory

96. After a delayed and rather weak start – due to lack of sufficient project-preparation and procurement capacities – the Uganda government maintained strong commitment to the project’s objectives and achieved steady progress. The main building work which remained outstanding at the end of 2015 was only the one-stop border post for the Katuna side of the Gatuna crossing to Rwanda, delayed by difficult environmental conditions (swamp) but now firmly under development (financed out of the country’s regular budget) for completion within 2016. The railway still falls far short of the intended (15-30 percent) share of total freight traffic from/to Mombasa but it has regularly paid the concession fee due government as well as other taxes, and imposed no financial burdens on the government budget.

(b) Implementing Agency Performance

Rating: Moderately Unsatisfactory

97. The principal Implementing Agency has been the Ministry of Works and Transport and associated transport-sector bodies, and it has worked closely, as necessary, with the Uganda Revenue Authority. The Ministry made systematic use of the Results Framework and monitoring system proposed by the Bank, and it effectively managed the combination of funding sources – IDA, other aid (mainly UK) and government budget – available. With Kenya it set, at Malaba, the first example of converting an old-fashioned border post to an OSBP. Uganda carried out limited but effective expansion of mobile weighbridges and weigh-in-motion installations and had probably the best experience among all four countries in the development of electronic cargo tracking (ECT), improving security and eliminating need for expensive convoys and escorts. The Ministry developed jointly with RVR (who paid for the detailed design) the major ICD at Mukono. By careful use of available resources, the Ministry managed to fund a much higher proportion of the US$13.0 million works proposed for Mukuno than had seemed possible when the first detailed proposals and economic justification were put forward in 2009.

TANZANIA

(b) Government Performance

Rating: Moderately Unsatisfactory

98. The Government of Tanzania moved ahead well in the project’s early years on a variety of important study and training initiatives, principally at the port of Dar es Salaam but also inland, as for example on vehicle weight control. It was slow, however, to progress preparation and contracting of the more major works envisaged, principally modernization of border posts, addition of weigh-stations and improvement of port security. A problem that further delayed progress was constant threat of corruption, involving also public

25 servants, and lengthy consequent delays on occasion in the completion of bidding processes, specifically on the Integrated Security System and procurement of six patrol boats at the Dar es Salaam Port under the Tanzania Port Authority (TPA). By the beginning of 2013, i.e. in the eighth year since approval of the IDA credit, still less than one-third had been disbursed. The last three years, however, saw a large amount of progress, effective completion of many works, substantial fulfilment of project objectives and increased disbursement to about 95 percent of the amount originally committed. The first large accomplishment was completion of four OSBPs, with two of them co-financed by TMEA as well as IDA, and the launch of their effective operation in partnership with neighboring countries in 2013-14.

Implementing Agency or Agencies Performance

Rating: Moderately Unsatisfactory

99. The institutions under the Ministry of Transport, Ministry of Works, and Ministry of Finance, were responsible for most of the project components, with coordination and follow-up handled by the general transport regulatory agency Surface and Marine Transport Regulatory Authority (SUMATRA), but with Tanzania’s Revenue Authority also taking over some of the works responsibility in 2012 to accelerate progress. Tanzania Port Authority was responsible for the effort to develop, as a Community Based Service, a national Single Window. And with insufficient IDA credit resources available by the time this became possible in 2013, TMEA kindly stepped in – as it had already done for some of the border posts. The major works implemented for upgrading and modernizing vehicle weight controls were carried out in 2013-14 under contracts let by Tanzania National Roads Agency (TANROADS).

RWANDA

(a) Government Performance

Rating: Moderately Satisfactory

100. Despite severe weaknesses in the staffing of some of its concerned agencies, such as RTDA, the government of Rwanda normally ensured timely actions to enable the country to get good results from its membership in the EAC Customs Union. The country thus benefited from acceleration of customs and other processing of trade even in the very small and outdated buildings of the old border posts. They were assisted by eventually effective use of the scanners bought under the project and, in more recent years, by well managed and effective development and use of electronic cargo tracking and Single Window facilities.

(b) Implementing Agency Performance

Rating: Moderately Satisfactory

101. The main implementing agency was RTDA and its predecessors, all of them subsidiaries of the Ministry of Infrastructure. Seemingly due to shortage of experienced

26 staff and repeated periods of high staff turnover, this agency remained slow in decision- making and progressing the upgrade of the Gatuna border post to become an OSBP, the only Rwandan civil works project foreseen in the IDA-assisted EAC Trade and Transport program. These delays were certainly partly due to the complexity of the project, located in a very narrow valley alongside an important small river and a swamp, protection of both of which was an important issue for the country’s environmental authorities – and for the country’s Cabinet when the issue was referred to it. The slow process of reaching a decision, combined with the lack of advance planning for any alternative use of the IDA grant which had been provided, meant that the Gatuna project still had a long way to go when the EATTF project closed in December 2015 and that nearly a third of the IDA grant which had been offered was lost to the country.

Justification of Rating for Overall Borrower Performance

Rating: Moderately Satisfactory

102. This was the rating scored by all the borrower governments and executing agencies other than the Ugandan Ministry of Works and Transport, which very effectively executed a diversified program of works, and the Rwandan RTDA, responsible for some elements in the smallest country program, including one that may well take another two years to finish – at full cost to the government – and therefore cannot be considered at this point as a Development Objective Delivered.

7. Lessons Learned

103. Whilst a regional project almost by definition will be complex, project designers need be realistic in terms of scope, timeline and outcomes. The sheer scale and complexity of this project contributed in a deleterious way to both implementation and outcomes. This is not to suggest that regional projects should not be ambitious, and include an appropriate mix of hard and soft elements, merely that teams have to be realistic both in defining the scope and the expected implementation timeline. In this case it would have been more appropriate to focus on either the trade and transport facilitation activities, or the railway activities, or develop distinct, but linked, projects as part of the same program. The QAG raised substantive concerns about the complexity of the Project, and interviews with staff involved during the preparation phase, support these concerns. Regional projects are inherently challenging and require close implementation support; they often progress at the pace of the slowest, as all countries face different implementation contexts, and coordination with different cultures and languages is a challenge. A clear lesson is that the designer of any regional project needs to be very realistic in terms of what can be achieved within a defined project period, and what can be expected from external partners, and the steps to the attainment of both are clearly mapped out, ideally by the appraisal stage.

104. The financing of preparatory work for regional activities and cross-border infrastructure itself remains problematic for a sovereign lender. For any project, robust preparation work is essential. This is particular true for cross-border infrastructure, whether a road, bridge, a border crossing itself, or works that straddle a national border. There were a number of instances in the implementation of the project, where either a component was

27 delayed, (a number of border crossings), or dropped from the project entirely (the bridge at Malaba), as funding could not be found to undertake either the necessary preparatory work, or the necessary civil works, as financial responsibilities straddled the border. As a further example, the optimal approach to establishing an OSBP would involve one procurement for the preparatory work, comprising physical and institutional improvements, and one for the subsequent construction and parallel institutional reform. With a sovereign lender, this is difficult as most clients are unable from using their IDA resources on the territory of another state, with the result that necessary preparatory works and then implementation, has to take place independently, and often to a different timeline, on each side of the side of the border. This is hardly ideal when the objective is a seamless facility. This problem remains very relevant in regional operations today, and highlights a need for a regional fund focusing on the delivery of improvement of cross border infrastructure, or potentially regional public goods.

105. The Bank should have tried to promote more serious joint effort by the countries themselves to figure out the best way to develop multi-national solutions in some of the key areas – and to limit its own initial financing to projects where the path ahead was clearer or a pilot project (e.g., in development of a one-stop border post) was defined. Paragraph 21 noted that the Bank placed considerable faith in the ability of the EAC Secretariat to facilitate agreement and collective action ex post. With hindsight, a more productive approach might have been to encourage the EAC Secretariat and member countries to undertake the necessary groundwork a priori to ensure agreement in advance. As two examples, (a) a well thought-out plan as to how best to build towards an eventual EAC-wide Single Window for all trade-related documentation, clearances and duty payments, with specific attention to the arrangements to be made for consultancy work, system development and hardware, and the financing of each, and, separate from that, for any national initiatives; (b) a similar plan for the development of an EAC-wide Electronic Cargo Tracking system, again with specific attention to financing for consultancy, system development, system management and financing arrangements. While such an approach would have increased the transaction costs and resource requirements initially, it would have facilitated implementation and led to a better outcome.

106. Despite the aspiration, the collaboration between the World Bank and the IFC needs improvement. The World Bank-IFC collaboration in the EATTFP was good during the preparation of the Project. The original RAP began with an IFC-funded environmental and social due diligence study. The components indicated for public financing and World Bank support were appropriate and a good example of what is required in such cases. The initial collaboration seems to have been a manifestation of the rapport and mutual understanding among the two task teams at that time. However, as the composition of the task teams evolved and the EATTFP was implemented, there was less and less cross- communication. More importantly, there seemed to be a lack of access to each other’s internal documents. If it was in the Operational Portal, IFC staff had access. Bank staff, on other hand, has very limited access to IFC documents due to the private sector confidentiality concerns, despite the need on the part of IDA to supervise the implementation of the PRGs. A clear lesson is that there shouldn’t be any access constraints to key documents when both parties are equally involved so that supervision of guarantee operations (or any other project operation) is properly conducted.

28

107. There is a need for the Bank to take a different approach to resettlement risks in projects. The project approached this as a standard linear resettlement typically associated with a road or rail investment. Originally, the objective was to clear as minimally required right of way as necessary to ensure safety as well as operational efficiency. The implementation showed in complex informal settlements such as Kibera and Mukuru, it is difficult to apply such an approach. Moreover, the Government clearly wanted to clear the whole reserve for expanding rail services and other development. Treating the required resettlement as a slum upgrading initiative would have respected the complex social, economic and political dynamics of these settlements. Working in such a context is analogous to working in a fragile state situation and requires the continuous monitoring and adaptation to respond to a situation in which the governance framework is volatile, institutions weak or relatively informal, security is tenuous and the overall vulnerability of the population is substantial. Such a situation goes well beyond compliance with OP 4.12; rather than dealing with the specific PAPs, the resettlement should be conceived as an interaction with an entire community, and handled as such in future.

29

ANNEX 1: PROJECT COSTS AND FINANCING Table 1-1: E.A. Trade & Transport Facilitation Project Costs & Financing (US$ millions) Components Appraisal Revised Actual Final Cost December Document Estimate 2015 & Financing 2005 2011 Total IDA Gov’t TMEA 1. Support to Create Customs Union 15.84 16.09 16.64 15.71 0.93 0.00 Rwanda 6.20 6.45 5.96 5.55 0.41 0.00 Uganda 1.89 1.89 2.33 1.89 0.44 0.00 Kenya 7.75 7.75 8.35 8.27 0.08 0.00 2. Institutional Support for Transport Facilitation 10.36 9.85 7.60 7.51 0.09 0.00 Tanzania 2.70 2.70 2.48 2.48 0.00 0.00 Kenya 3.50 3.50 1.79 1.79 0.00 0.00 Rwanda 3.50 1.64 1.32 1.23 0.09 0.00 Uganda 1.66 2.01 2.01 2.01 0.00 0.00 3. Investment Support for Transport Facilitation 97 139.19 152.26 86.31 346.81 31.15 a. Port Safety/Security Compliance 23.00 37.05 30.62 15.78 14.84 0.00 Kenya 11.50 22.90 29.38 14.54 14.84 0.00 Tanzania 11.50 14.15 1.24 1.24 0.00 0.00 b. Cargo tracking systems 3.65 4.52 5.94 2.29 1.65 2.00 Uganda 1.05 1.62 5.25 2.00 1.25 2.00 Kenya 0.40 0.40 0.40 0.00 0.40 0.00 Rwanda 1.20 1.50 0.00 0.00 0.00 0.00 Tanzania (for Railway) 1.00 1.00 0.29 0.29 0.00 0.00 c. Weighbridges 19.36 30.43 25.26 22.96 2.30 0.00 Kenya 10.60 19.03 6.33 4.03 2.30 0.00 Tanzania 7.50 8.00 16.90 16.90 0.00 0.00 Uganda 1.26 3.40 2.03 2.03 0.00 0.00 d. Joint Border Posts 41.21 51.66 77.04 34.40 13.49 29.15 Malaba (Kenya/Uganda) 8.67 10.20 12.50 10.23 2.27 0.00 Busia (Kenya/Uganda) 5.51 8.24 12.86 0.00 2.39 10.47 Lunga Lunga/Horohoro 5.38 7.18 10.96 9.55 1.41 0.00 Taveta/Holili 4.68 3.77 9.83 0.50 1.53 7.80 Isebania/Sirari 4.38 6.68 10.55 9.40 1.15 0.00 Mutukula (Tanzania/Uganda) 5.83 6.18 12.51 0.50 1.13 10.88 Katuna, Uganda 4.06 5.20 4.84 1.23 3.61 0.00 Gatuna, Rwanda 2.70 5.21 2.99 2.99 0.00 0.00 e. Inland Container Depots 3.41 6.45 10.20 8.87 1.33 0.00 Mukono, Uganda 2.31 5.20 8.95 7.62 1.33 0.00 Dar es Salaam, regional plan 1.10 1.25 1.25 1.25 0.00 0.00 f. Community-based Systems & other 6.60 9.08 3.21 2.01 1.20 0.00 Electronic Single Window, Kenya 3.20 3.20 1.00 1.00 0.00 0.00 ESW (TANCIS-based), - Tanzania 2.00 4.68 1.86 0.66 1.20 0.00 ESW - Rwanda 1.40 1.20 0.35 0.35 0.00 0.00 4. Support to Railways Concession 74.19 103.54 163.94 106.42 57.52 0.00 a. Retrenchment cost in Kenya 46.00 46.00 52.20 49.00 3.20 0.00 b. Pension start-up & arrears 12.00 12.00 10.20 10.20 0.00 0.00 c. Rehab. assets given to Pension Fund 1.00 2.40 2.40 2.40 0.00 0.00 d. Tech. & Investment Support 5.19 6.44 6.34 5.72 0.62 0.00

30

Components Appraisal Revised Actual Final Cost December Document Estimate 2015 & Financing 2005 2011 Total IDA Gov’t TMEA KRC 1.20 1.20 1.42 1.12 0.30 0.00 URC 3.99 5.24 4.92 4.60 0.32 0.00 d. Implement Relocation Action Plan 11.00 39.10 92.80 39.10 53.70 0.00 TOTAL 197.62 268.67 340.45 213.55 93.35 31.15

31

ANNEX 2: OUTPUTS BY COMPONENT Component Outputs Delivered Current Status/Comments Component 1: Support to Create Regional Customs Union Kenya Customs Assist rollout new procedures & IT Action sustained 2008-13 across all customs stations; staff training; purchase scanner, solar panels, computer equipment Uganda Customs reform/modernization Completed August 2010. Revenue (equipment and interconnectivity Authority across country); purchased Disaster Recovery System. Rwanda Training on EAC & Customs Union Completed by 2010. Revenue for private sector, incl. translation key Authority documents to French and Kinyarwanda, and purchases office equipment. Component 2: Institutional Support for Transport Facilitation Kenya Ministry Creation of national Transport Data Contract to University of Nairobi of Transport and Center Enterprise & Service Unit (2013), Infrastructure but further results unclear Uganda Staff Training/Reinforcement, Project team trained in areas of Ministry of especially on transport services World Bank procurement, M&E, Works and economics and enforcement of financial management and Transport regulations disbursement. Rwanda Two major consultant studies for Both studies largely completed by Ministry of long-run transport infrastructure 2011, identifying possibilities Infrastructure development: Dry port at Isaka, relevant for future transport Tanzania to improve link with Dar es development for Rwanda and Salaam port and Akagera River Great Lakes area. navigability for Lake Victoria traffic. Tanzania SUMATRA Integrated Financial Follow-up legislative action still Ministry of Management System (IFMS), Freight required to complete Safety, Transport Forwarders Training and Pollution, Search and Rescue, and Certification, Safety at Sea, Marine Accreditation of Freight Pollution, Search and Rescue, Staff Forwarders Profession. capacity building. Central Corridor Established CCTTFA Secretariat with Database for monitoring transit Transit appropriate staffing and reliable traffic was yet to be established Transport source of funding. and is still required. Facilitating Agency Component 3: Investment Support for Trade and Transport Facilitation a. Port Safety & Security Kenya Ports Extensive anti-terrorism and other Training on Integrated Security Authority security training for Port’s security System (ISS) and Maritime staff. Security was achieved as per International Maritime

32

Organization (IMO) standards, the security system is in place, and 375 KPA security staff and port police officers were trained. There were KPA personnel who were assigned/seconded to the Project continuously and are part of the maintenance team. This is a hybrid system which was developed specifically for KPA and attaching the staff as explained above was a major effort to ensure they understood the system; technical staff were trained on site and in Israel. Continuous improvement and knowledge transfer is part of the maintenance. Expert review of needs, followed by The needs assessment was done major security upgrade in Mombasa, which enabled KPA to gain a completed by August 2013 at cost of thorough understanding of the US$21.4 million. business plan, operations and facilities. This helped KPA to isolate and identify security concerns, issues, threats and vulnerabilities. Hence the design, supply, installation and commissioning of the Integrated Security System, which is successful. Tanzania Ports Some boats procured, but with Serious faults in boats accepted Authority malfunctions. Security studies and departures from acceptable eventually developed plans that were procedures in revising specs in implemented in 2015, independent of course of ISS bidding process the IDA project, for major security prevented IDA use. After taking upgrade at Dar es Salaam using outside advice as IDA techniques similar to Mombasa. recommended, TPA is using own funds to pay for US$6.4 million seven-month ISS contract, and also to replace boats. b. Cargo Tracking Systems Kenya Revenue Developed GPS-based Electronic KRA initiative without use of IDA Authority Cargo Tracking System and requires funds in the end. Initial progress all import, export & transit freight delayed by running-in problems movements in Kenya to be covered, and lengthy legal challenges but using ECT instruments hired from system was firmly established by private supplier. late 2013. Uganda Has also established a reliable ECT Completed Oct. 2014. Generates Revenue system, for which it recovers its costs improved freight movement data. Authority from users.

33

Rwanda Development of ECT system, now Finally implemented independent Revenue fully operating on Northern Corridor of project, benefiting from 2014 Authority roads in Rwanda. . COMESA initiative for all Northern Corridor, not requiring IDA funds but some TA from TMEA. Tanzania Good progress on developing and Following successful Dar es Railways launching railway cargo tracking Salaam-Morogoro pilot under the Limited (TRL) experiments in context of possible IDA project, TRL plans to roll out broader rehab and spread of rail the system using own funds. network. c. Weighstations Kenya National Athi River Axle Load Control Stn. Completion of construction Highways (25 km SE of Nairobi): expected for mid-2016. Authority Modernization and addition of bridge + WiM pads on other side of road. Mariakani Axle Load Control Station Completion of construction (50 km from Mombasa): expected for May 2016. Rehabilitation and modernization, plus addition of weigh bridge on other side of road. Uganda Four mobile weighbridges procured, Completed in December 2009, Ministry of for better supervision: -; handed over to UNRA Jan. 2010. Works & -Lira; Kyotera-Mutukula, and Transport Mbarara- roads; and main international itineraries. Procurement/installation four WiM Construction contracted early 2013 bridges: Malaba, Busia, Katuna, and and completed works handed over Mutukula border posts. to UNRA August 2015. TANROADS New main weighstation (including Construction contracts finally (Tanzania WiM) built at Vigwaza (which signed early 2013 and all stations National Road governs all traffic in/out of Dar es now functioning. Agency) Salaam), and existing stations Horohoro and Sirari much upgraded. d. Border Crossing Improvements Kenya National Five One-Stop Border Posts on main Taveta/Holili is the main border Highways routes, of which two were undertaken post already fully built and Authority by TMEA, including Taveta where it furnished, and run by fully trained also took full responsibility for the staff (benefiting from TMEA/ station on the other side of the JICA-developed Integrated Border Tanzanian border, Holili, and has Management training), operating been able to set impressive example fully on One-Stop basis since June of construction and joint preparation 2015. Final construction inspection for border operations by the September 2015. concerned staff from the two governments. The other Kenyan OSBPs largely Malaba was first EAC post to completed now are Busia (also operate as an OSBP – in mid-2009 financed by TMEA); Malaba, under thanks initial building work and IDA financing (in many ways the effective Kenya/Uganda electronic

34

most important because of the high data systems link. EU committed volume of traffic to Uganda and road-program grant 2008 for landlocked countries beyond); Lunga- essential new bridge that is now Lunga (Nov. 2014); and Isebania nearing completion. (Jan. 2015). Uganda Five OSBPs, including the Ugandan All operating and close to National Roads sides of the above-mentioned Malaba construction completion, but with Authority and Busia posts, and at Katuna, substantial delays still in view at Mutukula and Mirama Hills, the last Katuna due to environmental two financed by TMEA. difficulties. Rwanda One OSBP, at Gatuna, across from Difficult environmental issues of Ministry of Ugandan Katuna: new bridge already the site where the main road runs Transport completed, but new building, & existing border post stands, roadways and parking still needed between marshland and steep hill, despite environmentally constrained and differences of opinion among location. Rwandan bodies responsible delayed contracting and execution of studies financed by IDA to start only in 2010. Difficulties reaching agreement continued. Withdrawal of initial firm and recruitment in 2013 of another company. Govt. managed to reach firm decision on project March 2014, signed May 2015 US$7.5 million works contract financed internally. Government hopes to complete works within 2016, keeping border open throughout. Tanzania Two OSBPs – the Tanzanian sides of Responsibility for Tanzanian side Revenue the Sirari and Horohoro crossings of Mutukula border post which, Authority from Kenya, which were contracted like Holili, had originally been by TRA in 2011 soon after included in IDA project, was taken government had decided to transfer to over by TMEA at time of 2011 it responsibility for such works – project revision to release funds to were duly completed with final cover feared overruns. refinements in November 2015. e. Master Planning for Inland Container Depots Uganda Design and construction (finally This center, substituting (with Ministry of contracted November 2012) of large strong support of RVR) for revival Works and ICD at Mukono (some 25 km from of cramped facilities in Kampala, Transport Kampala) close to main road and rail finally completed December 2015, lines and next door to a large area will open for business early 2016, reserved for planned export managed by RVR in line with processing zone in Namanwe. railway concession. Tanzania Ports Port Master Planning exercise for Dar Study – by consultan