Document of The World Bank

DRAFT

Report No: ICR0000831

IMPLEMENTATION COMPLETION AND RESULTS REPORT (TF-57623 TF-90001)

ON A

GRANT

IN THE AMOUNT OF US$ 6.64 MILLION

TO

COOPERATIVE HOUSING FOUNDATION INTERNATIONAL

INTERNATIONAL ORGANIZATION FOR MIGRATION

FOR A

YOGYAKARTA AND CENTRAL JAVA EARTHQUAKE ROOF STRUCTURE PROJECT

June 30, 2008

Urban Development Sector Unit Indonesia East Asia and Pacific Region

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

CURRENCY EQUIVALENTS

(Exchange Rate Effective October 19, 2006)

Currency Unit = Indonesian Rupiah 1.00 = US$ 0.109

US$ 1.00 = IDR 9,109

FISCAL YEAR

January 1 – December 31

ABBREVIATIONS AND ACRONYMS

ADB Asian Development Bank BKM Local Board of Trustees CAS Country Assistance Strategy CHF Cooperative Housing Foundation (CHF) International CSRRP Community-based Settlement Reconstruction and Rehabilitation Project DIY Daerah Istimewa Yogyakarta (Yogyakarta Special Administrative Region) ERC Early Recovery Cluster FY Fiscal Year GoI Government of Indonesia IBRD International Bank for Reconstruction and Development ICR Implementation Status and Completion Report IOM International Organization for Migration JRF Java Reconstruction Fund M&E Monitoring and Evaluation MTR Mid-Term Review NGO Non Governmental Organization PAD Project Appraisal Document PDLAR Preliminary Damage and Loss Assessment Report PDO Project Development Objective RSK Roof Structure Kits RSP Roof Structure Project RSU Roof Structure Unit RT Rukun Tetangga (Community Leader) SA Special Account SCG Shelter Cluster Group TF Technical Facilitators TS Transitional Shelter (T-Shelter) UGM Universitas Gadjah Mada (Gadjah Mada University) UPP Urban Poverty Project

Vice President: James W. Adams

Country Director: Joachim von Amsberg Sector Manager: Sonia Hammam

Project Team Leader: Indira Dharmapatni ICR Team Leader: Indira Dharmapatni

INDONESIA Yogyakarta and Central Java Earthquake Roof Structure Project

CONTENTS

Data Sheet A. Basic Information B. Key Dates C. Ratings Summary D. Sector and Theme Codes E. Bank Staff F. Results Framework Analysis G. Ratings of Project Performance in ISRs H. Restructuring I. Disbursement Graph

1. Project Context, Development Objectives and Design............................................... 12. Key Factors Affecting Implementation and Outcomes .............................................. 63. Assessment of Outcomes ............................................................................................ 84. Assessment of Risk to Development Outcome......................................................... 115. Assessment of Bank and Borrower Performance ..................................................... 126. Lessons Learned ....................................................................................................... 147. Comments on Issues Raised by Grantee/Implementing Agencies/Donors .............. 14Annex 1. Project Costs and Financing.......................................................................... 20Annex 2. Outputs by Component ................................................................................. 21Annex 3. Grant Preparation and Implementation Support/Supervision Processes....... 22Annex 4. Beneficiary Survey Results ........................................................................... 23Annex 5. List of Supporting Documents ...................................................................... 26

MAP

i

A. Basic Information Country: Indonesia Project Name:

Yogyakarta and Central Java Earthquake Roof Structure Project

Project ID: P104357 L/C/TF Number(s): TF-57623,TF-90001 ICR Date: 07/11/2008 ICR Type: Core ICR Lending Instrument: ERL Grantee: CHF AND IOM Original Total Commitment:

USD 6.6M Disbursed Amount: USD 2.3M

Environmental Category: U Implementing Agencies: International Organization for Migration (IOM) Cooperative Housing Foundation (CHF) International Cofinanciers and Other External Partners: B. Key Dates

Process Date Process Original Date Revised / Actual Date(s)

Concept Review: Effectiveness: Appraisal: Restructuring(s): Approval: 12/22/2006 Mid-term Review: Closing: 08/31/2007 08/31/2007 C. Ratings Summary C.1 Performance Rating by ICR Outcomes: Satisfactory Risk to Development Outcome: Low or Negligible Bank Performance: Satisfactory Grantee Performance: Satisfactory

C.2 Detailed Ratings of Bank and Borrower Performance (by ICR) Bank Ratings Borrower Ratings

Quality at Entry: Satisfactory Government: Satisfactory

Quality of Supervision: Satisfactory Implementing Agency/Agencies: Satisfactory

Overall Bank Performance: Satisfactory Overall Borrower

Performance: Satisfactory

ii

C.3 Quality at Entry and Implementation Performance Indicators Implementation

Performance Indicators QAG Assessments (if any) Rating

Potential Problem Project at any time (Yes/No):

No Quality at Entry (QEA):

None

Problem Project at any time (Yes/No):

No Quality of Supervision (QSA):

None

DO rating before Closing/Inactive status:

D. Sector and Theme Codes

Original Actual Sector Code (as % of total Bank financing) Housing construction 100 100

Theme Code (Primary/Secondary) Access to urban services and housing Primary Primary E. Bank Staff

Positions At ICR At Approval Vice President: James M. Adams James M. Adams Country Director: Joachim von Amsberg Andrew D. Steer Sector Manager: Sonia Hammam Keshav Varma Project Team Leader: Indira Dharmapatni Indira Dharmapatni ICR Team Leader: Indira Dharmapatni ICR Primary Author: Indira Dharmapatni F. Results Framework Analysis Project Development Objectives (from Project Appraisal Document) The project development objective was to respond to the most basic needs of the households affected by the earthquake through the provision of roof structures that transition to permanent housing. Achievement of the PDO: Although the numerical target was not achieved as planned, the Project was able to cover all of the remaining needs of roof-structure. The Project was carefully designed to anticipate the need and rapid changes on the ground to ensure that all of the affected people are provided by shelters prior to the construction of permanent houses. As planned in the project design, the team conducted a mid-term review (MTR) three months after the Project commencement and adjusted the target needs according to the situation,

iii

whereby government permanent house reconstruction program was implemented successfully. The project scale was then significantly reduced from a total initial target of 24,000 units to 6,141 units. Towards the closing date, the needs of roof structures depleted further as a result of rapid government permanent house reconstruction. The provision of the roof structures was racing with the provision of the permanent houses. It should be noted that the implementation of the permanent houses reconstruction in Yogya was unprecedented in terms of speed and coverage. Over 150,000 houses were completed within one year. Thus, eliminate the needs of the roof structures. Revised Project Development Objectives (as approved by original approving authority) Not available (a) PDO Indicator(s)

Indicator Baseline Value

Original Target Values (from

approval documents)

Formally Revised Target Values

Actual Value Achieved at

Completion or Target Years

Indicator 1 : Percentage of beneficiaries households' (man/women) satisfaction on ability to resume the normal households activities

Value quantitative or Qualitative)

0 80% (MTR and by project closing)

IOM (DIY, male: 96.77%, female 94.74%; Central Java, male: 92.86%, female: 100%). CHF (DIY, male: 93.94%, female: 100%; Central Java, male: 97.14%, female: 100%)

Date achieved 12/31/2006 02/28/2007 08/31/2007

Comments (incl. % achievement)

By MTR (March 2007) the target values achieved were as follows: IOM (DIY + CJ: male 88.24%, female 87%); CHF (DIY+CJ: male 93.65%, female 94.17%); Both at the MTR and project closing, the target achievements are higher than the target indicators.

Indicator 2 : Percentage of beneficiaries of roof structure units (RSUs) that transition to permanent houses

Value quantitative or Qualitative)

0 60% (MTR), 80% (project closing)

IOM (DIY: 52%, Central Java:24.3%); CHF (DIY: 80.6%, Central Java: 46.6%)

Date achieved 12/31/2006 02/28/2007 08/31/2007

Comments (incl. % achievement)

By MTR (March 2007),target achieved were: IOM (DIY: 37.5%, Central Java: 50.9%); CHF (DIY: 39.4%, Central Java: 42.9%). Values at MTR & closing are lower than the intended target due to the rapid realization of gov't permanent housing program

iv

Indicator 3 : Percentage of beneficiaries of roof structures that reuse their materials in permanent house

Value quantitative or Qualitative)

0 60% (MTR), 80% (project closing)

IOM (DIY: 26%, Central Java:40.54%); CHF (DIY, 10.19%, Central Java: 3.45%)

Date achieved 12/31/2006 02/28/2007 08/31/2007

Comments (incl. % achievement)

By MTR (March 2007),target achieved were: IOM (DIY: 21.8%, Central Java: 35.3%); CHF (DIY: 24.3%, Central Java: 20.7%). Target at MTR & closing are < than the intended target due to most beneficiaries have permanent houses & only rooftiles can be reused

(b) Intermediate Outcome Indicator(s)

Indicator Baseline Value

Original Target Values (from

approval documents)

Formally Revised

Target Values

Actual Value Achieved at

Completion or Target Years

Indicator 1 : Number of prefabricated roof structures kits (RSKs) produced that meet satisfactory basic technical quality (cumulative)

Value (quantitative or Qualitative)

0 24000 6141 4886 (79.56% of revised target values)

Date achieved 12/31/2006 08/31/2007 08/31/2007 08/31/2007

Comments (incl. % achievement)

At the MTR, a survey estimated that remaining needs of Roof Structure Units (RSUs) were 3,500 units. The project scale was reduced to 6141 units. Rapid implementation of Gov't permanent housing progr. has further reduced the RSKs needs.

Indicator 2 : Number of RSKs delivered to eligible beneficiaries that meet satisfactory basic technical quality

Value (quantitative or Qualitative)

0 24000 6141 4790 (78% of revised target values)

Date achieved 12/31/2006 08/31/2007 08/31/2007 08/31/2007 Comments (incl. % achievement)

Rapid implementation of Gov't permanent housing progr. has reduced the project scale to 6141 RSKs. In the field, the geographical distribution of the beneficiaries are more spread & in smaller groups.

Indicator 3 : Number of Roof Structure Units (RSUs) constructed that conform to pre-agreed specification (cumulative)

Value (quantitative or Qualitative)

0 24000

6141 (IOM: 974 and CHF: 1667, plus 3500)

4434 ( 72.22% of revised target values)

Date achieved 12/31/2006 08/31/2007 08/31/2007 08/31/2007 Comments (incl. %

Since May 2007, CHF&IOM could no longer found beneficiaries due to rapid decreasing needs. Materials equivalent to 150 & 141 RSUs of CHF and IOM,

v

achievement) respectively, were distributed to eligible beneficiaries thru UPP-Housing.

G. Ratings of Project Performance in ISRs

No. Date ISR Archived DO IP

Actual Disbursements (USD millions)

H. Restructuring (if any) Not Applicable

I. Disbursement Profile

1

1. Project Context, Development Objectives and Design 1.1. Context at Appraisal 1. Background: On May 27, 2006, an earthquake registering 6.3 on the Richter scale damaged surrounding areas of Yogyakarta and Central Java (the “Affected Areas”). While a significant number of houses were destroyed or damaged by the earthquake; societal coping mechanisms were also severely disrupted, with widespread trauma and means of earning a living buried in the rubble. The scale of the physical destruction caused by the disaster was belied by the relatively low death toll, owing to a traditional building culture that was complacent about earthquake risk. Masonry walls made of burnt bricks connected with weak mortar – a reality exacerbated by unwitting negligence and poverty – were no match for the strong earthquake. In addition, a lack of adequate anchoring, poor workmanship, and minimal compliance with building codes all contributed to the extent of the damage. 2. In the immediate aftermath of the earthquake, in response to a request from the Government of Indonesia (“GoI”) for housing assistance and in cooperation with government, local and international stakeholders, various organizations, individuals and community groups distributed emergency shelter materials (e.g. tarpaulins, mattresses, blankets and other emergency-shelter non-food items). While close to 100% of the needs for immediate emergency shelter were covered at the time of project preparation, emergency tents and tarpaulins would not provide adequate shelter to cover transitional needs through the anticipated two to three years required for preliminary reconstruction. Many families were thought to be likely to wait much longer for the completion of permanent housing built to earthquake-resistant standards. In the meantime, although some families may have had the resources to recover on their own and others may have received assistance from government, civil society or the humanitarian community, poverty and insufficient resources highlighted the need for transitional shelter support. The lack of adequate housing also impacted the capacity of people to resume their livelihoods. A large percentage of the affected population was and remains engaged in small home-based industries. Hence, in the absence of durable shelter, these communities would not be able to resume their previous economic activities. Finally, the lack of adequate housing causes severe health problems, most of which worsen during the rainy season, imminent at the time of project preparation.

3. GoI is presently focusing all available resources on providing permanent houses to affected households, while the NGOs working under the coordination of the United Nations Early Recovery Cluster (“ERC”) are focusing on meeting immediate needs for housing in the form of “Roof Structures”. A typical Roof Structure is a wind and weather-proof bamboo structure with a tile roof, assembled on the existing concrete foundation of the original collapsed house where possible, and is considered durable in that it is built to last for a minimum of two years while permanent houses are being constructed. All Roof Structures should be built to SPHERE standards1. Some key features of Roof Structures are that: (i) they do not endanger a family’s entitlement for permanent housing assistance under the broader GoI shelter response to the disaster; (ii) almost all components used in the

1 For further detail on Roof Structures, refer to Annex 5: Technical Annex of the two Project Appraisal Documents attached as listed in Annex 6.

2

construction can be reused in the construction of the permanent house; and (iii) no permits or authorizations are required for construction. 4. Preliminary estimate of damage and response: In June 2006, weeks after the earthquake, BAPPENAS, the provincial and local governments of the Affected Areas and a team of international partners led by the World Bank and the Asian Development Bank (ADB) produced the Preliminary Damage and Loss Assessment Report (“PDLAR”). The PDLAR showed that over 300,000 homes were either totally destroyed or heavily damaged; with an average family size of 4.3, approximately 1.3 million people were rendered homeless by the earthquake.

5. Assessments conducted by the key members of the ERC at the time of project preparation showed that 93,000 families were living with other families, were in shelters with a roof height below two meters, or were in shelters that would otherwise not provide sufficient shelter from the rainy season. At the time of preparation, overall commitments from various agencies providing Roof Structures amounted to approximately 50,000 – showing a gap of approximately 47,300 shelters (including a 10% contingency).

6. Project Rationale: The Activity Initiation Summary for the Yogyakarta and Central Java Earthquake Roof Structure Project (“RSP”) was signed off in August 2007 and, as it was an emergency operation, was able to reach effectiveness by December 2007. GoI had released its Action Plan for rehabilitation and reconstruction for the Affected Areas and had committed US$270 million over the period of 2006-2007 for the reconstruction of permanent houses. One of the objectives of the multidonor Java Reconstruction Fund is to respond to the basic needs of the households affected by the earthquake, notably in terms of housing.

7. While GoI had just begun distributing the first phases of cash grants to start the process of permanent reconstruction, this resource could not have been mobilized in time to mitigate the potentially devastating effects of the rainy season upon settlement conditions in the Affected Areas. The Provincial Governments of the Affected areas together with the ERC developed a joint and complementary strategy aimed at accelerating the process of reconstruction to address the need for housing, whilst adhering to the plan for permanent housing reconstruction established by GoI. 1.2. Original Project Development Objectives (PDO) and Key Indicators 8. The project development objective was to respond to the most basic needs of the households affected by the earthquake through the provision of roof structures that transition to permanent housing. 9. The key outcome indicators for the project were as follows:

• Beneficiaries households’ (men/women) satisfaction on ability to resume their normal household activities, e.g. children can study, family member/s can conduct informal economic activities, protected from rain and adverse weather conditions.

• Percentage of beneficiaries of roof structures that transition to permanent houses. • Percentage of beneficiaries of roof structures that reuse their materials in

permanent house.

3

1.3. Revised PDO (as approved by original approving authority) and Key Indicators,

and reasons/justification 10. While Intermediate and Output results were revised downwards to reflect changing circumstances on the ground (detailed in section 7), the Key Outcome Indicators and their targets remained unchanged. 1.4. Main Beneficiaries 11. The RSP was designed to provide approximately 24,000 temporarily homeless families in targeted villages in the Affected Areas with safe and durable housing through the provision of Roof Structures while their permanent house was being rebuilt. 1.5. Original Components (as approved) The project components varied slightly between the Implementing Agencies. IOM had been operating a Roof Structure mass-production facility in the Affected Areas for some time when the RSP began implementation, while CHF established their facility specifically for the RSP. All components listed below applied to CHF, while components (ii) through (v) applied to IOM.

(i) Set-up of the prefabrication facility. CHF was to set up a fabrication facility and subcontract five prefabrication teams of 15 people to operate for one shift per day producing 110 Shelter Structure Kits per day; or 660 kits per week, (based on a 6-day production week). Workers were specialized according to the different aspects of the production to optimize speed and quality of construction.

(ii) Preparation of Roof Structure Kits (RSK). This component involved prefabrication of the most labor intensive and structurally crucial elements of the building including: roof trusses, columns, tarpaulins for the roof, bracing and pre-cutting of wall components. Further, roofing elements (the frame that held the roof-tiles) were precut to facilitate transportation of materials and improve on-site construction efficiency. Pre-fabrication was also utilised in enhance quality control procedures, which can be more easily implemented at a pre-fabrication stage. (iii) Targeting and distribution of kits to beneficiaries. The Implementing Agencies undertook selection of end-users/beneficiaries according with the following procedures:

a) Site identification. Using the SCG database, outreach staff identified locations with heavy damage that were not receiving any assistance from other organizations. This information was to be reviewed and crosschecked with various other sources including local government. Both Implementing Agencies aimed for full coverage in each targeted village. Once an area is selected, CHF’s outreach staff met with community leaders (RTs) to introduce the program. IOM socialized the program with community leaders and the community at large, in addition to conducting a rapid assessment. In addition, IOM’s Community Assessment Teams carried out a rapid

4

assessment plan. b) Finalization of beneficiaries list and raising awareness. In consultation with the community leaders, the Implementing Agencies established a list of eligible households, based upon a set of standard eligibility criteria agreed upon with the World Bank. Potential beneficiaries were then to be surveyed by a team of student volunteers under the supervision and guidance of Implementing Agency field staff to verify eligibility for assistance. Following the volunteer assessment, the beneficiary lists were to be vetted with the entire community via community-wide meetings. c) Ex-post distribution/monitoring. After delivery of RSKs, CHF provided technical support to ensure the quality of RSKs delivered, and conducted technical audit of approximately 30% of beneficiaries to verify structural quality. Meanwhile, IOM conducted beneficiary follow-up with all beneficiaries of RSK’s. The follow-up included two components: firstly, a technical quality control inspection, in which any defects, maintenance, etc. was conducted on the spot. Secondly, an assessment of beneficiary satisfaction and survey of shelter function and material usage for permanent shelters.

(iv) Assistance to beneficiaries during construction. In the field, the Implementing Agencies oversaw and provided technical guidance for the construction of Roof Structures, in cooperation with volunteer students from local universities, mainly the Department of Architecture, and Civil Engineering of the Universitas Gadjah Mada (UGM) and skilled construction workers who were recruited and trained for the task. In the construction process, CHF used volunteers and community-based labor groups supervised by the outreach department and field construction manager to ensure adherence to work plan, material control requirements and construction standards. Both Implementing Agencies provided technically skilled volunteers and hired professionals to support in the construction of the Roof Structure; both RSKs included a construction manual and information on safe building materials. IOM also used teams of volunteers which reported to specialized units within IOM, including Community Liaison and Assessment, Prefabrication, Distribution, Construction and Beneficiary Follow-up.

(v) Quality Control and Reporting. The Implementing Agencies had a quality control system for the input and the output. The tracking procedures cover incoming material, production, field distribution, end-user tracking and community handover of the structure kits, quality control during the construction and certification upon completion of the construction.

1.6. Revised Components The original components approved by the Board remained unchanged for the duration of the RSP. 1.7. Other significant changes Project End-Date Extensions

5

12. The JRF RSP grant agreement was signed on December 12, 2006 and became effective on December 28, 2006, however, the first advance payments were not disbursed until early January 2007. While the Implementing Agencies proceeded with beneficiary identification prior to the drawdown of funds, full project implementation could not begin until early January – a month behind the intended schedule. As a result of the shift in the Project start date, the Implementing Agencies requested a one month extension of project implementation to June 2007 for CHF and July 2007 for IOM. The task team raised no objection as the extension still fell within the project closing date as specified in the grant agreement. Project Design Adjustment 13. One issue facing the project was the overlap with the JRF Community-based Settlement Reconstruction and Rehabilitation Project (CSRRP) which was intended to fund infrastructure and shelter reconstruction in areas administered by the World Bank’s Urban Poverty Project (UPP). The PAD clearly stated that the RSP will have to work outside the UPP villages. In the UPP areas, the JRF fund channelled through the CSRRP (Re-Kompak) that had been allocated for infrastructure in the villages should be diverted and prioritized for the TS using the UPP mechanisms. Other parties, including IOM and CHF can cover the remaining TS needs in the UPP villages whenever the TS needs in a particular village was no longer covered by this infrastructure funds, that is, beyond 100 TS’ per village. Local boards of trustees (“BKMs”) were loathe to spend on shelter for individual families at the expense of common infrastructure for the whole community. The UPP task team moved to resolve this through the imposition of penalties for BKMs who were found to have been negligent in beneficiary identification. Furthermore, the BKM’s definition of community membership sometimes excluded families within their geographical scope for a variety of reasons, most commonly due to their recent arrivals from other villages since the earthquake. Both of these issues resulted in leaving families without shelter in some UPP villages, even in the UPP villages where BKM had covered all of the needed TS’. 14. As a solution to this problem, it was agreed that both IOM and CHF could operate in the UPP areas also in the villages where BKM had provided TS less than 100 units, if there’re still TS need there. As both implementing agencies expanded into these new areas, they coordinated with representatives from JRF-CSRRP (Re-Kompak) and the community-based BKMs in selecting and verifying beneficiaries, sharing information, and ensuring that all those in need received shelter assistance while avoiding duplication of assistance between the potentially overlapping projects. Project Scale Reduction 15. Prior to the mid-term review, an independent survey showed that the remaining need of roof structure units was drastically decreasing due to the rapid implementation of the GoI-funded permanent housing program which covered most of the earthquake affected people. Within a year of the earthquake, the GoI Permanent Housing Program was able to provide 150,000 permanent houses for the earthquake-affected people. 16. Given the unexpected speed of the GoI response, it became clear that the scale of the project needed to be drastically reduced. The RSP task team conducted a mid-term review (MTR) mission in March 2007 to assess program progress and evaluate the outstanding need for transitional shelter. The mission and independent review team concluded that as of the end of March, 2007, there remained only approximately 3,500 families in need of shelter. As

6

a result of this, the overall targeted number of roof structure for the project was revised downwards from 24,000 to 6,1412 as presented in Annex 2. Reduced Grant Allocation 17. As a result of the reduced scale of project operations, it was clear that total grant allocation for both CHF and IOM would not be fully utilized. At the time of the mid-term review mission, when the decision was made to reduce the scale of the project, US$0.27 million of the CHF grant and US$1 million of the IOM grant had been disbursed. An additional US$0.64 million was scheduled to be disbursed for CHF. It was recommended at that point that for CHF the total US$0.91 million and for IOM the disbursed funds would be sufficient to cover remaining need and, therefore, further disbursements should be suspended. This total was later revised upwards to cover extra need as well as to cover increased per-unit costs. 2. Key Factors Affecting Implementation and Outcomes 2.1. Project Preparation, Design and Quality at Entry 18. The RSP was identified as a Rapid Response Emergency Operation given the numbers of people homeless and the imminent rainy season at the time of preparation. In hindsight, it appears that the need for transitional shelter was vastly over-estimated, however this is understandable given the lessons learnt in the response to the Aceh tsunami in 2004 where the permanent reconstruction of houses took over two years and local institutional capacity was severely disrupted. 19. The RSP was conceived as a response to a transitional shelter provision gap identified through assessment conducted by the ERC and prepared by the World Bank in cooperation with the provincial governments of the Affected Areas and the two Implementing Agencies. The biggest unknown at the time of project preparation was the ability of GoI to deliver on its promise to construct 150,000 houses in the year following the earthquake. As implementation of the GoI program had not begun in November when the RSP was approved, the scale of the project appeared reasonable. 20. The capacity of the local communities to band together to form community work teams was well utilized by the RSP permitting minimal paid work in the construction phase of the RSP and ensuring value-for-money. 2.2. Implementation (including any project changes/restructuring, mid-term review, Project at Risk status, and actions taken, as applicable) 21. Project implementation proved creative and adaptable to changing circumstances as is vital for work in a disaster recovery situation. The project was able to reduce scale, move

2 At the MTR, the total number of TS constructed by both CHF and IOM was about 2,641 units. This figure consists of the number of the constructed TS and the estimated remaining demand, based on the independent survey.

7

into new areas and transfer excess building materials to suit the demands of the new situation quickly and relatively efficiently. 22. Over the course of the RSP, it became apparent that the GoI reconstruction program was moving ahead at a rate far beyond what the RSP had anticipated, eventually succeeding in constructing the approximately 150,000 houses it had promised within one year of the earthquake. This highly satisfactory outcome had certain implications for the RSP, IOM had begun procurement materials for a larger number of RSKs than were required for the final number of RSKs delivered. CHFs more conservative approach to targeting and RSK construction resulted in a more modest, but still significant over-procurement of materials. This issue was raised by the Implementing Agencies to the task team at the MTR and discussed the different options for the excess materials. After some consultations, the Implementing Agencies were advised that the preferred option was to pass on remaining materials on to the UPP housing program. The Implementing Agencies held discussions with the local UPP and ReKompak representatives and with task team the amounts and locations in need and organized the distribution of the materials in question to the dedicated UPP locations in Bantul and Klaten. 2.3. Monitoring and Evaluation (M&E) Design, Implementation and Utilization 23. The World Bank’s field office in the Affected Areas permitted close supervision of the Implementing Agencies and played a major role in the timely and creative responses to issues facing the projects. Overall the M&E design of the RSP and its implementation and utilization from the Bank side was excellent. However, concerns were raised over the Bank’s combined role as Trustee, Monitor, and Implementer of multi-donor fund activities, as this could be perceived as creating a conflict of interest and creating bias in monitoring. 2.4. Safeguard and Fiduciary Compliance Safeguard 24. As all TS’ were constructed on the same landplots of the damaged houses, there was no land acquisition or resettlement. Similarly, there was no on-site environmental impacts. The use of bamboos in the RSP was also considered as not having significant impacts to the environment, as they are widely planted in Java and in many, cases are grown without wildly. Bambo is also considered as an environmentally friendly materials that can be easily degraded by the environment. 25. To ensure all safeguards concerns of the populace were taken into account, CHF established a telephone hotline, monitored during working hours by volunteer, to receive and respond to telephone or mobile-text-message concerns or complaints. CHF also placed complaint boxes in each community, so that community members could submit anonymous written complaints. These boxes and complaints were checked by CHF staff and community leaders each week. All complaint policies, complaints and its follow up action were tracked in database and posted in public places within the community to ensure transparency. IOM established a partnership with local NGOs and provided training and instruction to maximize the outreach of the beneficiary identification process. Further IOM set up a monitoring and spot check system to assess the identification process of local partner NGOs. During the

8

course of scheduled monitoring visit, it became clear that many of the families recommended in the assessment reports to receive shelter assistance did not, meet IOM’s eligibility criteria. Partner implementation activities were immediately placed on hold as IOM and IOM’s partners took action to address the issue of recipient eligibility Fiduciary. 26. Procurement. CHF and IOM generally followed procurement procedures as specified in the grant agreement and had appropriate procurement controls in place. The RSP task team noted that CHF and IOM used a merit point system in evaluating bidder’s quotations, which was not acceptable as per the PAD Annex 4, Section B: Procurement and the Bank’s guidelines. Following on from this CHF provided the Bank with detailed information about its approach and submitted a revised procurement plan and a redesign bid evaluation system. However, IOM procurement process for roof tiles was significantly delayed in the initial project implementation, however, it delivered all shelters within project implementation timeframe. 27. Financial Management. CHF had a satisfactory financial management system based on the Bank’s review of their internal controls, payment verification and documentation under financial transactions in the field office. During the mid-term review mission, the Bank suggested that CHF to establish an internal audit function which includes a review of internal control over project implementation. Over the course of the project the task team noted that CHF had improved their project financial management system to have better documentation, systems and controls. 28. IOM was evaluated as having adequate accounting systems and internal controls. However, there was delayed in submission of financial management report due to dispute in the foreign exchange rate in the special account report. During the mid-term review, it was noted that IOM had an unsettled issue regarding initial deposit expenditures which again delayed submission of the financial management report and withdrawal application. Towards the project closing, IOM improved its reporting system. 2.5. Post-completion Operation/Next Phase (including transition arrangement to post-completion operation of investments financed by present operation, Operation & Maintenance arrangements, sustaining reforms and institutional capacity, and next phase/follow-up operation, if applicable) 29. There is no post-completion operation 3. Assessment of Outcomes 3.1. Relevance of Objectives, Design and Implementation 30. The RSP promoted the transfer of earthquake-resistant construction knowledge to beneficiaries as well as strengthening local institutional capacity through implementation partnership, strongly supporting the final pillar of the FY2004-2008 Indonesia Country Assistance Strategy (“CAS”) – that is, to enhance disaster risk preparedness and response (by strengthening institutional capacity and framework). 31. The JRF’s main objective is to respond to the most basic needs of the households

9

affected by the earthquake, and housing is one of the most important basic needs. The JRF RSP was designed to meet JRF objectives, responding to affected persons’ basic needs, through the versatility of the roof shelters and the reusability of their component materials. 3.2. Achievement of Project Development Objectives 32. The Project Development Objective of the RSP was to respond to the most basic needs of the households affected by the earthquake through the provision of roof structures that transition to permanent housing. Three key outcome indicators, two intermediate results indicators and two output indicators were proposed to measure success against that objective. However, only the key outcome indicator concerned with recipient satisfaction was met (and significantly exceeded). The number of RSKs produced, delivered, and constructed did not meet the original plans, as the TS needs decreased significantly since the MTR. However, all of the TS needs was successfully covered by CHF and IOM during project implementation. The lack of success in meeting outcome indicators does not point to a failure of the RSP, rather a combination of the sorts of unexpected circumstances that arise in fast-changing disaster relief situations. 33. The most significant effects to the indicators were due to the outstanding success of the GoI permanent housing reconstruction program that substantially decreased the total number of Roof Structures required, increased the number of recipients transitioning to their permanent houses and, consequently, decreased the occupancy rate of the Roof Structures by the time of the final survey undertaken by the World Bank in July 2007. 34. Another significant factor affecting the occupancy rate (ranging from 24-81% with a target of 60%) was the extent to which the Roof Structure could be used for purposes other than dwellings. Even at the time of the mid-term review, the task team noted substantial numbers of Roof Structures that had been converted into small shops or other business places and community meeting places. The usefulness of the Roof Structures as places of business, warehouses, animal shelters and other community functions decreased the percentage of beneficiaries that re-used the materials used in the Roof Structure to construct their permanent house ranges from 3.45% to 40.54% (Annex 2). 35. The most significant negative factor affecting occupancy rates was the differing interpretations of the eligibility criteria by the Implementing Agencies and the World Bank in the beginning stages of the RSP. Shelter eligibility criteria was agreed upon in the initial stages of the project; however the implementing agencies applied that criteria based on humanitarian need, which differed from the Bank’s interpretation of those criteria. Targeting of the beneficiaries by both Implementing Agencies following the mid-term review greatly improved. 36. For further detail on the project development outcomes and indicators, please refer to Annex 2. 3.3. Efficiency (Net Present Value/Economic Rate of Return, cost effectiveness, e.g., unit rate norms, least cost, and comparisons; and Financial Rate of Return)

10

37. The RSP experienced significant cost overruns in the per-unit costs as compared with the projections during project preparation. The realized per-unit cost for CHF was 37% greater than the anticipated per-unit cost (US$446 vs.US$297) and for IOM was 150% greater than the anticipated per-unit cost (US$658 vs.US$266). A large part of the increased costs were largely as a result of the loss of economies of scale when the project was downsized. Furthermore, IOM procured construction materials sufficient for the construction of 3,000 Roof Structures in the early stages of the project before the reduced demand became apparent. By project close, IOM had constructed 1,586 Roof Structures and dealt with the oversupply by transferring the remaining materials to UPP villages throughout the Affected Areas as approved by the task team. 38. The role of the RSP in permitting earthquake affected persons to resume their lives and enhancing their ability to cope while their permanent houses were being constructed is quite high. The internal rate of return for thousands of children returning to school, beneficiaries returning to work and social wellbeing is impossible to calculate, but it is surely substantial. 3.4. Justification of Overall Outcome Rating (combining relevance, achievement of PDOs, and efficiency) Rating: Satisfactory 39. While most of the project indicators were not met numerically, the project development objective of responding to the most basic needs of households affected by the earthquake through the provision of Roof Structures that transition to permanent housing was certainly met in substance. The RSP was able to complement the outstandingly successful GoI permanent housing reconstruction program by bridging the shelter gap between emergency shelter and permanent housing for those thousands of families that would have otherwise had to remain in tents or other sub-standard dwellings throughout the rainy season. 40. The Implementing Agencies in coordination with the local and provincial governments of the Affected Areas, GoI, World Bank, the NGO community in the Affected Areas, the academic community and the people of the Affected Areas showed flexibility and initiative in dealing with a rapidly changing situation resulting in an effective project. 3.5. Overarching Themes, Other Outcomes and Impacts (if any, where not previously covered or to amplify discussion above) (a) Poverty Impacts, Gender Aspects, and Social Development 41. In accordance with the eligibility criteria that the Implementing Agencies used to select beneficiaries, those who received Roof Structures were mostly poor households without the means to rebuild an adequate dwelling before the mobilization of permanent structure assistance from GoI. Mobilization of this assistance permitted them to resume their lives and protected them from falling further into poverty 42. Based on the beneficiaries assessment conducted by IOM, 396 of the households that received assistance were headed by females, representing 25% of the total 1586 recipient households. Data collected by CHF shows that 51.25 % of their recipient beneficiaries were

11

female (5975 out of 11,658 beneficiaries). 43. In addition to quantitative analysis, the survey of beneficiary satisfaction conducted by the World Bank shows that satisfaction to be able to resume daily activities after receiving Roof Structure assistance is higher for females than for males. It is difficult to analyze these positive responses without taking into account the traditionally polite Javanese culture, however, it may be concluded that the Roof Structure program benefited females in general. (b) Institutional Change/Strengthening 44. Local Institution Contribution. Positive development impacts of this project include contributions to the strength of local institutions and support (through knowledge exchange) of community self-help traditions. 45. Javanese culture has an established tradition of voluntary community self-help, known as gotong royong, where individuals work together to help other members of their community with large or difficult tasks, such as rice-harvesting or construction projects. CHF International utilized this tradition by pairing field construction teams, UGM volunteers and one paid professional carpenter with gotong royong labor groups to erect the shelter frames. In addition to the knowledge transfer from the paid professionals, gotong royong groups have proven themselves more than capable of playing a major role in future community development projects. 46. Further, the CHF International facilitator team worked in close coordination with volunteers from UGM (a major university in the earthquake-affected region). The durable shelter used for the RSP was designed immediately following the earthquake by a team of CHF engineers working in conjunction with the faculty of Architecture and Urban Planning and the faculty of Engineering at UGM. Furthermore, UGM volunteers participated in or were present at many stages of the roof structure delivery process, from beneficiary identification and verification through construction completion and progress monitoring. The resultant capacity building will have positive knock-on effects, as this local institution is now positioned to take on greater responsibility in future disaster reconstruction efforts. IOM also worked closely with the Engineering and Architecture Departments of the UGM in the development of the design. (c) Other Unintended Outcomes and Impacts (positive or negative) 47. Given the speed with which the GoI permanent housing program occurred, the vacated Roof Structures constructed by the project were able to be used for other economic activities that played a significant role in the resumption of affected persons’ livelihoods in the Affected Areas. 3.6. Summary of Findings of Beneficiary Survey and/or Stakeholder Workshops (optional for Core ICR, required for ILI, details in annexes) 48. Refer to Annex 5. 4. Assessment of Risk to Development Outcome Rating: The overall risk to the development outcome is rated Negligible to Low (L).

12

49. The shelter crisis caused by the Central Java earthquake has passed due to the shelter programs conducted by the government, private individuals and the efforts of the donor community, of which the RSP was a major component. As such, there is no further risk to the development outcome. Furthermore, the skills and knowledge transfer from the academic, government and donor community on earthquake resistant construction techniques should mitigate the scale of future destruction from a similar size earthquake. 5. Assessment of Bank and Borrower Performance (relating to design, implementation and outcome issues) 5.1. Bank Performance (a) Bank Performance in Ensuring Quality at Entry Rating: Satisfactory 50. While the demand for Roof Structures was vastly over-estimated, the project design remained flexible enough to adapt to the changing circumstances to ensure the development objective was met. (b) Quality of Supervision Rating: Highly Satisfactory 51. The close coordination between the World Bank and the Implementing Agencies both in their respective field offices and head offices in Jakarta permitted early detection and appropriate solutions to issues as they arose. This close relationship resulted in the improvement of targeting, the expansion of the program into UPP villages (solving a political issue fast approaching crisis point as discussed in section 1.7), and an appropriate down-scaling of the project before too much wastage occurred. (c) Justification of Rating for Overall Bank Performance Rating: Satisfactory. 52. The Bank’s performance is rated satisfactory because of its rapid response to emergency, intensive supervision and monitoring, and strong coordination with other stakeholders. 5.2. Borrower Performance (a) Government Performance Rating: Satisfactory 53. GoI’s main role in the RSP was as coordinator and supervisor in the overall reconstruction program in Yogyakarta and Central Java, although its involvement was minimal due to its focus on its own permanent housing reconstruction program. The Yogyakarta Provincial Government was responsive in the reconstruction assistance and actively undertook coordination with other actors in the overall reconstruction program. The Central Java Provincial Government was slightly slow in processing the grant agreement and involved only at the district level.

13

(b) Implementing Agency or Agencies Performance Rating: Satisfactory The following table shows rating of the implementing agencies performance:

Task CHF IOM (i) set-up of the prefabrication facility S N/A (ii) preparation of Roof Structure Kits HS HS (iii) targeting and distribution of kits to beneficiaries MS MS (iv) assistance to beneficiaries during construction S S (v) quality control and reporting HS S

Note: HS=Higly Satisfactory; S=Satisfactory; MS=Moderately Satisfactory 54. Over the course of the project CHF International was able to successfully modulate the level of production from the prefabrication facility to match the demand. After the MTR, the pre-fabrication capacity has been constantly adjusted to the needs and demands of the JRF project 55. Further, RSKs delivered were standardized and of high quality. Monitoring teams rarely encountered instances of RSK inadequacy over the course of this JRF RSP. 56. CHF International and IOM used their local NGO networks, student volunteers and outreach staff to identify beneficiaries in the field. At the time of the first supervision mission, in January 2007, targeting was recognized as a major problem, with frequent, cases of ineligibility. Following this supervision mission, both CHF International and IOM dramatically improved their arrangements as well as capacities of NGOs and volunteers in order to improve the process of selecting beneficiaries. 57. The rating for assistance to beneficiaries during construction is largely based upon field monitoring and beneficiary feedback. The very high level of beneficiary satisfaction (see outcome indicators) justifies the satisfactory rating. 58. CHF coordinated with a number of different stakeholders and maintained internal monitoring mechanisms to ensure the quality measures were being achieved. They submitted high quality reports to the task team in a generally timely manner. IOM quality control activities were judged satisfactory, but they were judged moderately satisfactory for this component due to difficulties adhering to the reporting schedule. (c) Justification of Rating for Overall Borrower Performance Rating: Satisfactory 59. Both Implementing Agencies provided good quality Roof Structures under difficult circumstances with high beneficiary satisfaction. Even for those beneficiaries who have received permanent housing assistance, Roof Structures continue to play a role in the livelihoods of 90% of the recipients pointing to continued satisfaction with the structures. The prefabrication process, training, collaboration and coordination with other stakeholders were all conducted in a satisfactory manner. This process was accelerated by the local institutional capacity and community participation.

14

6. Lessons Learned 60. Although the JRF RSP successfully achieved its development objective, it would have been structured and implemented much differently if the actual level of need had been more accurately estimated prior to inception. Specifically, the estimated gap in coverage of housing need required a response that was beyond the capacity of any one agency, so both IOM and CHF were contracted to build roof structures. In actuality, however, the level of need was low enough to easily be covered by only one group. 7. Comments on Issues Raised by Grantee/Implementing Agencies/Donors (a) Grantee/Implementing agencies (IOM and CHF) 61. According to UN cluster data, the international aid community provided more than 80.000 temporary shelter (t-shelter) to the victims of the May 2006 earthquake. As the biggest single contributor, IOM build and delivered more than 15.000 t-shelter all together, including those funded under this project. JRF’s contribution to the t-shelter programme was relevant and needed to fill the remaining gap and to provide those still in need with adequate living space. 62. As the JRF supported t-shelter programme started relatively late (January 08, or 8 months after the earthquake), the recovery efforts were quite progressed by that time and the need for further t-shelter assistance was smaller than initially anticipated. However, the results show that IOM identified more than 1.500 families that were still in need and which could have not been assisted without the JRF contribution. Hence, the JRF project enabled IOM to fill a critical gap, which might have been smaller than anticipated but still existent. 63. It is also a common phenomenon in large scale relief operation, that the “last 10%” of beneficiaries to be reached in need of support are harder to reach, more scattered and more problematic to identify and this was exactly the population targeted by the project. This increases the time and costs involved compared to early stages of an intervention and contributes to higher unit costs. Nevertheless, those supported by the JRF project were in need of assistance and thanks to the project, could receive shelter support. 64. Since the start of the project, IOM worked closely together with the task team to adopt to the project to the fast changing environment and both sides were flexible enough to adjust the project accordingly. When it became clear that the number of t-shelter needed was smaller than initially assessed, the task team and IOM quickly reduced the project scope so that more than ¾ of the initially budgeted funds could be reallocated to other needs. IOM also worked closely together with the JRF in order to find a solution for the unused materials. Thanks to the task team’s assistance and quick reaction, the materials could be delivered to the Re-Kompak programme within the project timeframe. 65. The Yogakarta and Central Java emergency relief operation is widely regarded as one of the most successful interventions of the international aid community when it comes to natural disaster responses. Provision of temporary shelter was the main focus of international aid and UN data suggests that close to 99% of the t-shelter need was matched. The JRF funded project turned out to be smaller than anticipated but greatly contributed to the vital need and a successful intervention. 66. Overall CHF achieved the project goals and objectives as stated and approved in the project design document. CHF experienced a number of challenges during project implementation, namely: Differences in perceptions of eligibility criteria; restrictions on areas of

15

operation and overlap with UPP programming; reduction in the scale of the project; and changes in procurement procedures. However, CHF was able to work efficiently with the Bank in overcoming all of these challenges quickly and in a manner agreeable to all parties. All changes recommended in the MTR report, particularly as related to beneficiary targeting and procurement procedures, were immediately addressed and resolved. 67. The issues surrounding beneficiary identification had a significant impact on project implementation. Both implementing agencies, in conjunction with the World Bank, established clear eligibility criteria for shelter recipients in the initial program set-up. However, these criteria differed from the criteria used for beneficiary identification under the World Bank-implemented CSRRP program. This led to confusion among parties and disagreement over the appropriateness of eligibility criteria. All parties worked to resolve this issue through extensive discussions on criteria and the basis for the criteria (i.e. international humanitarian standards, such as SPHERE) and ultimately reached agreement on beneficiary eligibility and targeting benchmarks. 68. As noted previously, changes in demand for transitional shelter, based on the rapid onset of the GoI housing program, also created implementation challenges. CHF addressed this decrease in demand by reducing production numbers and deploying additional field teams to identify beneficiaries as rapidly as possible over a wide geographic area. In spite of these efforts, CHF retained some unused materials and shelter kits at the end of the program, which were in turn provided to the CSRRP program for continued use in communities identified as vulnerable or particularly in need. (b) Cofinanciers/Donors 69. Based on the reporting materials submitted to the donors, the following questions were raised: (i) 4,886 kits were produced, 4,790 delivered, and 4,434 constructed. Why were some kits prefabricated not used for construction? In the beginning, especially IOM, procured a large amount of materials at once with the intention to be able to deliver fast. It was not anticipated that the government permanent housing program was delivered quite fast, and therefore the TS projects had to “compete” in terms of delivery pace with it. Large amount of materials had been bought and prefabricated, but the demand has been declined rapidly. Consequently, IOM and CHF had difficulties in finding eligible beneficiaries, resulting in the discrepancy of the kits produced and the units delivered. CHF slowed production of kits to match declining demand, but even with reduced production, excess kits remained in stock that could not be delivered to eligible beneficiaries. As the beneficiaries were prioritizing to finish their permanent houses instead of the TS, not all units were completed at the closing of the projects. This resulted in the discrepancy between the units delivered and the units constructed. Many units were nearly complete (i.e. frame, roof, and walls constructed) but not finished out as the beneficiaries focused on constructing their permanent homes. Our survey in October 2007 indicated that some of the materials or units delivered were used as parts of the permanent houses or used as supporting materials during the permanent house constructions. (ii) Low occupancy rate: please provide explanation on reasons behind. The construction of the TS was carried out in parallel with that of the government permanent housing. At the MTR, about 77% of the government permanent housing funds was disbursed to the community. Within 6 months of the disbursement,, it was expected that the community would have completed their permanent houses. The construction of the TS was completed at about the

16

same time with that of the permanent houses. Therefore, at the closing of the RSP, people had moved to their permanent houses and no longer used the TS as their main place to live. This is the reason why the occupancy rate of the TS at the closing was low. However, as the survey had indicated, some of the TS were used and/or reused for their permanent houses. Furthermore, even when families moved into their permanent houses, many continued to use the TS as small shops or production rooms for livelihoods facilities, or as additional rooms (such as kitchens) for their permanent homes. Many families also used the permanent home during the day but continued to sleep in the TS at night, due to lingering fears of earthquakes and that fact that many felt safer in a lightweight bamboo TS than a brick-and mortar permanent home. (iii) Mid-term review assessed "moderately satisfactory: on IOM and CHF's procurement processing and IOM's financial management and follow-up actions were recommended. At completion, how did you assess IOM/CHF's procurement and financial management? Were there improvements and remedical actions successful? The fact was that for IOM, at the MTR stage procurement was rated moderately satisfactory because it used a merit point system. As a result, the winning supplier was not the lowest bidder as has been suggested by the Bank (in the MTR’s aide-memoire). IOM procured only once, so that since the MTR, there has been no procurement. Therefore, at the closing its rating for procurement remained moderately satisfactory. In the case of CHF, the procurement was rated moderately satisfactory in the MTR for the same reason as that of IOM. Following this rating, CHF submitted a revised procurement plan which was subsequently approved by the Bank. Based on this revised plan, for any procurement undertaken after the MTR, CHF no longer used a merit system and consequently, they selected bidders with lowest bids. CHF had smaller packages so that they easily adjusted the procurement sizes in accordance with the declining demands. For financial management, IOM was rated moderately satisfactory during the MTR and satisfactory during the project closing. This was because it had submitted the financial report not in timely manner at the MTR, but in timely manner at the project closing. While the CHF’s financial management ratings were satisfactory both during MTR and at project closing. (iv) As of September 2007, project implemeted by IOM spent USD 1.04 million to build 1,586 units while CHF spent USD 1.2 million to build 2,848 units. Please elaborate on this. Both IOM as well as CHF proposed a project cost consisting of staff costs, goods and materials, labor costs and wages, operating costs and unallocated (contigencies). The amount of goods and materials as well as the labor costs and wages reflected the number of units produced, however, the staff and operating costs relatively the same regardless of the units produced. Therefore, the lesser the number of units produced, the higher the costs per unit would be which mainly contributed by the staff and operating costs. Below are the breakdown of the costs in the original plans:

17

IOM LINE ITEMS Costs

(in US$) Costs per unit

(in US$)

Staff costs and consultants services 396,000 Goods and materials 2,800,000

Labor costs and wages 464,000 Operating costs 426,000

Unallocated 172,000

Total 4,258,000 266.00 CHF

The above tables showed the costs per unit for the size of total production of 16,000 units and 8,000 units for IOM and CHF, respectively. At the project closing, the costs per unit for IOM and CHF was $ 658 and $ 446 , respectively. These are 2.5 and 1.5 higher than the original costs, for IOM and CHF, respectively. As noted previously, with the decrease in the scale of the project, and over-procurement of materials based on anticipated demand, the individual unit cost increased. (v) Was TS project good idea? Do you recommend for similar situations in the future? Learning from the project, the appropriateness of the TS provision approach will depend on the situation and external factors. First, the population and the local government function. In the case of Yogyakarta, population remained intact and they have the capacity and social networks that can rebuild their own houses. In this case, TS delivery will be better if carried out by the people with their existing social networks/capacity. The local government functions was also in place and for the case of Yogyakarta, by working together with the existing community social networks, it is possible to deliver the TS immediately after the disaster. Second, TS provision should be delivered very fast (as part of the immediate relief actions), it has to start immediately after the disaster, while the community is waiting for permanent housing program. The challenge is the identification of the beneficiaries. Again, this would be addressed by using the community social networks, particularly using the existing community-based projects which have been there prior to the disaster. To some extent, IOM and CHF used the community social networks, one of the reasons that they could deliver relatively fast, particularly in identifying beneficiaries. Third, government (Central and Local) programs for emergency and reconstruction. It is very important to not underestimate the government capacity and commitment to deliver its emergency and reconstruction housing programs. Yogyakarta has been always considered as one of the very important regions in the country. It has a unique historical , social and economic position in the country, so that Central Government will always do as best as it can to maintain Yogyakarta’s development. In the case of having strong commitment and capacity from the central government,

Object Class Categories Cost (inUS$)

Costs per unit (inUS$)

Staff Costs and consultants’ services 244,000

Goods & Materials 1,668,000

Labor costs and wages 100,000

Operating costs 336,000

Unallocated 32,000

Total amount 2,380,000 297.5

18

TS approach will only be an appropriate approach if it does not have to compete with the government permanent housing program. Or in other words, it should be delivered as part of the relief efforts immediately after the disaster. Both the Bank, as the trustee of the grants, and the implementing agencies have to be equipped with a project preparation procedures that are tailored to rapid relief-kind of requirements and mechanisms. Both CHF and IOM are equipped with these procedures, which allowed for rapid mobilization, beneficiary identification, production, and delivery of shelter kits immediately upon receipt of Project funds. (vi) Was implementing through NGO beneficial? Why? why not? Working with NGOs and intergovernmental organization has advantages and shortcomings. The Bank has had limited experiences in working with NGOs for project investments. The Bank has mainly counterparted with the Central and local governments. Learning from the RSP, the main advantage is once we agreed upfront on the procedures and requirements, funds disbursement could be done relatively fast. Furthermore, an NGO like CHF and an intergovernmental organization like IOM have well-established networks with government, and local and international NGOs and universities to work with, and therefore, could draw upon these resources to assist them in accelerating the TS delivery. As CHF and IOM are two international organizations with good reputations, they have a well-established operational system, namely, financial management, procurement, reporting system, etc. and their system and standards requirement are easily can be adjusted with those of the Bank’s. Adjustment of IOM and CHFs procedures and requirements was made quite easily and this also contributed to the relatively fast project preparation. If we worked with the government for the TS project, perhaps it would take longer, as the project needs to be included in the national/local budget cycle, and in the situation of rapid relief, the government may not be able to do the procurement rapidly. However, there are some shortcomings for the project implemented by NGOs and intergovernmental organization. First, again, in the situation where there was strong government capacity to deliver the permanent housing program, the Bank had to monitor and supervise the project to ensure that beneficiary criteria for TS reflected the realities in the field and were applied appropriately. Second, there was a risk to have less flexibility in reducing procurement of materials as a respond to the declining TS needs, because the NGO and the intergovernmental organization already design a large-scale procurement upfront. This will affect the efficiency of the TS delivery, leading to higher operational and overhead costs compared to the material and construction costs of the TS. (vii) Was deliverying Prefab good approach? Delivering prefab units would be appropriate for a situation when there is no houses and people in the disaster areas. In the case of Yogyakarta, there were still many houses (although with various damage levels) and people living on-site in the disaster affected areas. People wanted to stay on-site and built/rebuilt their houses on-site and it was a fact that land availability was in some cases limited and its shape may not accommodate the pre-designed TS. (i) The fixed design of roof structures required a certain size of land be available, which was not always possible. This restriction on size did, if rarely, limit the selection process of beneficiaries; (ii) The usage of bamboo as standard material affected the roof structures delivery and utilization as follows:

• During the reconstruction stage, the demand of bamboo was quite high in DIY and Central Java which affected the availability of bamboo for the JRF Roof Structures. To be able to obtain bamboo, both IOM and CHF had to search to other districts (beyond DIY and Central Java).

19

• Bamboo is a material that has a high level of ‘starch’. Therefore large scale and rapid procurement contributed to a risk of getting a lower quality bamboo due to limited availability in the market. A lower quality bamboo affected the utilization of roof structure by the beneficiaries, especially near the project completion time due to bamboo scarcity.

(iii) Many beneficiaries did not construct the TS as soon as it received the units from the implementing agencies, because they prioritizing the construction of the permanent houses. During this period, the bamboo’s quality declined. However, communities in Java are accustomed to working with bamboo, and therefore could more easily construct the shelters out of locally-available materials that are familiar to them. Furthermore, communities could easily replace or reuse the bamboo in a number of ways, contributing to the overall objectives of the project. In above mentioned situation, the TS as a prefab unit is less flexible to serve the needs of the people. It would be more useful if in during the early reconstruction period, people were given flexibility to rebuilt/built temporary shelters while waiting for permanent houses to come. Also, people should have been given flexibility of type and quality of materials to build temporary shelters. This approach can be done in a large scale without waiting for the prefab TS units are produced. This would only be possible if the community-based approach is used for the delivery. However, delivery of prefabricated units ensured rapid production and standardized high quality of shelters. By delivering a standard model that communities worked to construct in the field, the implementing agencies ensured the quality of the shelters while still allowing communities some flexibility to modify the structures as needed, and re-use materials as appropriate. IOM regards the pre-fabrication approach as appropriate and based its strategy in this respect inter alia on lessons learnt from the Tsunami reconstruction operation. It turns out, that with a centralized pre-fabrication approach a higher economy of scale and therefore cost efficiency can be achieved. Furthermore, the pre-fabrication approach makes construction and project implementation less dependent on local construction companies and other local partners. The experience in Aceh showed that this can become a major bottleneck for timely progress and construction standards as the project depends on the performance of such local partners, which are difficult to control and only to a low extent, controllable through legal means, given the complexity and length of the legal system in the country. (viii) Results achieved, especially for cross-cutting issues (gender, environment, DRR) (a) Results achieved in terms of gender issue is presented in the Annex 2. The satisfactory level for male and female at the Mid-term Review (MTR) was > 87%, while at the closing or the project was > 92%. In fact, satisfaction level for women in Central Java (Klaten) both for IOM and CHF’s TS was 100%. (b) There is relatively no environmental issues or impacts for the TS projects. Bamboo is widely cultivated in Java and it is renewable plants. For a short period during early project implementation stage, IOM and CHF had to obtain bamboo from outside Yogyakarta because of the high demand. But this was not a problem. The project was preventing the people from illegally harvesting wood from the forest, as it provided them with bamboo-based TS, while waiting for the permanent houses from the government. (c) Bamboo-based TS is resistant to earthquake, as it is light and flexible, i.e. responsive to a wind proof, and vertical load test indicated load of 860kgs and resistance up to 2.2 tons. IOM and CHF designed a structurally strong unit and tested in the Gajah Mada University’s laboratory. For the Javanese, a bamboo-made houses psychologically give them a feeling to be safer than that of concrete houses when there was earthquake.

20

Annex 1. Project Costs and Financing

(a) Project Cost by Component (in USD Million equivalent)

Components Appraisal Estimate (USD millions)

Actual/Latest Estimate (USD

millions)

Percentage of Appraisal

Total Baseline Cost 0.00 0.00

Physical Contingencies 4.468

1.092

24.44

Price Contingencies 2.57

1.223

47.59

Total Project Costs 6.638 2.315 Project Preparation Costs 0.0030 0.08 27 0.0035 0.35 .100

Total Financing Required 6.733 02.745

(b) Financing

Source of Funds Type of Cofinancing

Appraisal Estimate

(USD millions)

Actual/Latest Estimate

(USD millions)

Percentage of Appraisal

Trust Funds 0.00 0.00 Indonesia - Java Reconstruction Program 6.64 2.315 34.86

21

Annex 2. Outputs by Component

Result Indicator Grant Agreement

IOM

CHF

Project Development

Objective Key Outcome

Indicators MTR Closing

Post-MTR Target MTR*) Closing**) MTR*) Closing**)

Male: 88.24%; Female:

87%

Male: 93.65%; Female: 94.17%

DIY: Male: 93,94%-

female: 100%; CJ: Male: 97,14%;

female: 100%

Beneficiaries households’ (men/women) satisfaction on ability to resume their normal household activities, e.g. children can study, family member/s can conduct informal economic activities, protected from rain and adverse weather conditions

80%

80%

NA

98.1% (DIY); 74.7% (CJ)

DIY: Male: 96,77%- female:

94,74%; CJ: male:

92.86 %- female: 100 %

90.8%

(DIY); 93% (CJ)

95.37% (DIY); 98.28% (CJ)

Percentage of beneficiaries of roof structures that transition to permanent houses

60%

80%

NA

37.5% (DIY); 50.9% (CJ)

52 % (DIY); 24.32% (CJ)

39.4% (DIY);

42.9% (CJ)

80.56% (DIY);

46.55% (CJ)

To respond to the most basic needs of the households affected by the earthquake through the provision of roof structures that transition to permanent housing.

Percentage of beneficiaries of roof structures that reuse their materials in permanent house

60%

80%

NA

21.8% (DIY); 35.3% (CJ)

26 % (DIY);

40.54 % (CJ)

24.3% (DIY);

20.7% (CJ)

10.19% (DIY);

3.45% (CJ)

Number of prefabricated Roof Structures Kits produced that meet satisfactory basic technical quality

Total: 22,400(IOM:14,400 CHF: 8,000)

Total: 24,000 (IOM:16,000CHF: 8,000)

6141

1,416

1,586

2,015

3,300

Households receive Roof Structures Kit Number of Roof

Structures Kit delivered to eligible beneficiaries that meet satisfactory basic technical quality

Total: 20,480(IOM:12,480 CHF: 8,000)

Total: 24,000 (IOM:16,000CHF: 8,000)

6141

1,228

1,586

1,845

3,204

Number of roof structures constructed that conform to pre-agreed specifications

Total: 15,940(IOM: 8,640 CHF: 7,300)

Total: 24,000 (IOM:16,000CHF: 8,000)

Total: 6,141 (IOM: 974 CHF: 1,667 +

3,500)

974

1,586

(CJ: 838 and DIY: 748)

1,667

2,848

(CJ: 1000 and DIY: 1,848)

To provide earthquake affected families with safe and durable housing through the provision of roof structures while permanent housing is being rebuilt

Percentage of roof structures occupied by beneficiaries

90%

90%

NA

65.4% (DIY); 55.4% (CJ)

16 % (DIY);

24.32 % (CJ)

58% (DIY); 70.1% (CJ)

54.63% (DIY); 43.10% (CJ)

*) MTR: Mid-Term Review: Feb, 20 to March 30, 2007 **) Figure in this column is based on the survey conducted by the World Bank in July-August 2007

22

Annex 3. Grant Preparation and Implementation Support/Supervision Processes

(a) Task Team members Name Title Unit Identification/Preparation Ida Ayu Indira Dharmapatni Task Team Leader EASUR Hongjoo Hahm Infrastructure Sector Coordinator/Lead

Infrastructure Specialist EASUR

Raj Soopramanien Senior Counsel LEGEA Yogana Prasta Senior Disbursement Officer EACIF Joe Leitmann Lead Environmental Specialist EASEN Rajiv Sondhi Senior Financial Management Specialist EAPCO Unggul Suprayitno Financial Management Specialist EAPCO Rizal Rivai Senior Procurement Specialist EAPCO Sarosh Khan Senior Infrastructure Specialist ConsultantGottfried Roelcke Urban Specialist ConsultantJohn Holdaway Research Assistant Consultant Supervision Ida Ayu Indira Dharmapatni Task Team Leader EASUR Hongjoo Hahm Infrastructure Sector Coordinator/Lead

Infrastructure Specialist EASUR

Unggul Suprayitno Financial Management Specialist EAPCO Yogana Prasta Senior Disbursement Officer EACIF C. Irma Dona Financial Management Specialist ConsultantPaulus Bagus Tjahjanto Procurement Specialist EAPCO Sutantyo Purnomo ConsultantTri Dwi Budirianto Consultant

(b) Staff Time and Cost Staff Time and Cost (Bank Budget Only)

Stage of Project Cycle No. of staff weeks USD thousands (including travel and consultant costs)

Lending PreparationFY07 Not available Not available

Total: Supervision/ICR FY07&08 10.5 51,200

Total: 51,200

23

Annex 4. Beneficiary Survey Results The survey was conducted by the World Bank during July-August 2007 to evaluate the performance of the JRF RSP and identify the impact of the Roof Structures on the beneficiaries based on the indicators stated in the Grant Agreement and PAD. The survey measured six indicators as set out below:

CHF IOM No Indicator DIY Central Java DIY Central Java

1. Eligibility 50.93 % 74.14% 76% 78.38 % 2. Process

Source of Information Head of community group 51.85% 65.52% 68% 89.19%

a.

Implementing agencies staff 54.63% 41.38% 32% 10.81% b. Beneficiary knows the eligibility criteria 31.48% 46.55% 66% 56.76%

Registration mechanism Through implementing agencies staff 82.41% 93.10% 92% 97.30% In community meeting 4.63% 6.90% 2% 2.70%

c.

Others 12.04% 0% 8% 0% Announcement of recipient In community meeting 76.85% 81.03% 52% 75.68% Through public (religious) places 0% 8.62% 0% 0%

d.

Others 1% 25.86% 48% 24.32% Complaints submitted 2.78% 1.72% 4% 2.7% e. Complaints responded 100% 100% 100% 100%

3. Delivery Quality a. Material received is complete 94.44% 96.55% 68% 94.59% b. Timeliness delivery 85.19% 93.10% 2% 43.24% c. Accompanied by hand-over document 79.63% 98.28% 96% 97.30% d. Material received in good quality 79.63% 70.96% 78% 62.16% 4. RSU Construction a. Roof structures completed with manuals

(book, training and posters) 75.93% 94.83% 100% 100%

b. Provided assistance from the implementing agencies staff

92.59% 87.93% 96% 91.89%

c. Provided assistance from the facilitator 38.89% 82.76% 96% 75.68% d. RSU constructed by community group 74.07% 60.34% 12% 18.92% 5. Quality of RSU a. Good quality of main structure 84.26% 75.86% 90% 89.19% b. Good quality of overall structure 68.52% 62.07% 46% 64.86% 6 Outcome a. RSU occupancy duration (in months) 3.65 3.95 3.24 5.03 b. Current occupancy rate of RSU 86.11% 91.38% 82% 100% c. Change to permanent house 80.56% 46.55% 52% 24.32% d. Reuse RSU materials for permanent house 10.19% 3.45% 26% 40.54%

Level of satisfaction to the RSU Male 93.94% 97.14% 96.77% 92.86% Female 100% 100% 94.74% 100%

e.

Total 95.37% 98.28% 96% 94.59% f Made modification to the RSU 24.07% 12.07% 6% 5.41%

24

Source of eligibility information. The table shows that community received information about assistance mostly from the head of community group (or RT). It is in line with the method of beneficiary identification and selection that the implementing agencies undertaken consultation with the head of villages or head of community group for preliminary identification and selection of beneficiaries. Further, strong local institution in DIY and Central Java contributed to greater role of local leaders in the reconstruction assistance, including in the JRF RSP of CHF and IOM. Knowing eligibility criteria. The eligibility criteria was disseminate among communities through head of community groups and community meeting as well as use of public information boards and leaflets. The implementing agencies field staff was provided by eligibility criteria and conducted household assessments and community general meetings for socialization of the program. Further, the implementing agencies undertake consultation to the community leader and follow up with community meeting for final selection of beneficiaries. CHF announced the list of beneficiary to community as implementation of transparency principle in the JRF RSP program. Meanwhile IOM undertook final verification as follow up of community meeting through deployment of field staff for final decision and site inspection. The IOM Information System to ensure the community in general was aware of the program and eligibility criteria. In some cases, IOM visited the villages in several occasions to verify additional beneficiaries even during the construction stage, when the village beneficiary identification process was technically already closed. Delivery quality. Delivery quality is seen from four criteria: completeness of received material, timeliness of delivery, provided by hand-over documents, and quality of received material. Beneficiaries of JRF IOM in DIY surveyed said that the delivery time was not accurate. The problem appeared due to delay in procurement of roof tiles in the earlier stage of project implementation, therefore the beneficiaries received roof structure kits incomplete and the delivery could not be done in timely manner. However, IOM’s own assessment results and beneficiary satisfaction process carried out towards the project closing, conducted for all JRF beneficiaries, suggested a high satisfaction with the delivery of assistance. Further, IOM procured materials of roof tiles slightly behind schedule although the other materials for roof structure kits were procured in advance. The roof structure kits delivered to the beneficiaries without roof tiles although it could be used because complemented with the tarpaulin to cover the roof. But all of the materials was delivered within the project implementation. Concerning the quality of materials received, small percentage of beneficiaries felt that the materials received were not in good quality. As emergency program, it was required to implementing agencies to deliver the program rapidly to allow the community to recover prior assistance of permanent housing. Near the completion time, the implementing agencies had difficulties to find a good quality material, especially for bamboo. Bamboo needs to be cut in exact time and method to be used as material for structures meanwhile the need should be fulfilled in such short time. Therefore, small portion of bamboos were received not in good quality as it is easily to break. IOM purchased, stored and controlled the quality of materials continuously at a warehouse/production facility. A quality control team was deployed during distribution and construction as well to ensure only good quality materials would be utilized for the construction process. Broken bamboo or straps were returned to the production site were they would be replaced; at a later stage, IOM would provide extra material for each

25

truck with prefab elements so the QC team on site could repair the light damages if any. This created greater efficiency in delivery. RSU Construction. The survey identified the performance of roof structure unit construction. Implementing agencies applied different approach for construction, for example CHF involved community groups (in gotong-royong system) to construct the roof structure supervised by the field technical staff. Meanwhile, IOM constructed the roof structure by professional worker accompanied by facilitator and witnessed by recipient community. The result of the survey shows that although it was not obligatory for community to involve in construction of RSU IOM, but in most cases, the beneficiaries voluntarily involved in the construction process. RSU Quality. As previously mentioned that near to the project completion, the implementing agencies had difficulties to find a good quality material especially bamboo. Therefore, they had to search ouside the DIY province since the bamboo availability in DIY and Central Java was drastically reduced. Outcome. The survey measures achievement of outcome as the following: beneficiary’s satisfaction, number of beneficiary’s who changes their RSU to permanent house and number of beneficiaries who uses RSU materials for their permanent house. It is shown in the above table that the transformation of RSU to permanent house is happened to small percentage of beneficiaries. The reason is during the survey was undertaken many respondents have received permanent house assistance from the Government. The RSU, therefore, is utilized for other purposes (compliment to permanent house), such as: mini market, animal stall or even it becomes part of the permanent house (kitchen). In addition, few beneficiaries who have not yet received permanent housing assistance utilize the RSU as living place. Concerning the reuse of RSU materials for permanent house, it is only small percentage of beneficiaries who reused the materials. It is because of among all materials, only roof tiles which can be re-used. Further, as previously mentioned, most of beneficiaries used the RSU as complimentary to their permanent house.

26

Annex 5. List of Supporting Documents 1. Project Appraisal Document for A Java Reconstruction Fund (JRF) in the Amount of US

$ 4.258 Million to Republic of Indonesia for the International Organization for Migration (IOM), Yogyakarta and Central Java Earthquake Roof Structure Project, December 22, 2006

2. Grant Agreement (Yogyakarta and Central Java Earthquake Roof Structure Project) among International Organization for Migration (IOM), Republik of Indonesia, respresented by the Governor of the Yogyakarta Special Region and Central Java Province, and International Bank for Reconstruction and Development/International Development Association, acting as Administrators of Grant Funds under the Java Reconstruction Fund, December 12, 2006

3. Project Appraisal Document for A Java Reconstruction Fund (JRF) in the Amount of US $ 2.38 Million to Republic of Indonesia for the Cooperative Housing Foundation (CHF) International, Yogyakarta and Central Java Earthquake Roof Structure Project, December 22, 2006

4. Grant Agreement (Yogyakarta and Central Java Earthquake Roof Structure Project) among Cooperative Housing Foundation (CHF) International, Republik of Indonesia, respresented by the Governor of the Yogyakarta Special Region and Central Java Province, and International Bank for Reconstruction and Development/International Development Association, acting as Administrators of Grant Funds under the Java Reconstruction Fund, December 12, 2006

5. Mid-term Review Mission February 27-March 30, 2007, Aide Memoire 6. Independent Survey Final Report, LPPM University of Gajah Mada, April 2007 7. Monthly Reports of IOM and CHF International 8. Final Report of International Organization for Migration, period of December 28, 2006-

July 27, 2007 9. Project Completion Report of CHF International, October 31, 2007 10. Independent’s Auditor Report for International Organization for Migration, October 31,

2007 11. Independent’s Auditor Report for CHF International, September 19, 2007