ICRR 13660 Report Number : ICRR13660 IEG ICR Review Independent Evaluation Group 1. Project Data: Date Posted : 02/09/2012 Country : Haiti Project ID : P104690 Appraisal Actual Project Name : Haiti Catastrophe US$M ): Project Costs (US$M): 9.00 8.99 Insurance L/C Number : CH278 Loan/ US$M ): Loan /Credit (US$M): 9.00 8.99 Sector Board : UD US$M): Cofinancing (US$M ): Cofinanciers : Board Approval Date : 03/08/2007 Closing Date : 12/31/2010 12/31/2010 Sector (s): Non-compulsory pensions insurance and contractual savings (100%) Theme (s): Natural disaster management (100% - P) Prepared by : Reviewed by : ICR Review Group : Coordinator : Ridley Nelson Robert Mark Lacey Soniya Carvalho IEGPS1 2. Project Objectives and Components: a. Objectives: The project development objective stated in the PAD (p. 5) was: "to reduce Haiti's financial vulnerability to natural disasters through insurance coverage against earthquakes and hurricanes ". The PAD added: "This will be achieved by providing financing to Haiti to allow it to join the Caribbean Catastrophe Risk and Insurance Facility (CCRIF)". The Financing Agreement reverses this formulation, placing the means first : "The objective of the project is to assist the recipient in participating in the CCRIF for the purpose of reducing the recipient's financial vulnerability to catastrophic events resulting from hurricanes and earthquakes ". While the ends and means feature in both formulations, this review uses the PAD version because it states the anticipated outcome up front. It should be noted here that this project and the OECS Countries Catastrophe Insurance project P 094539 were both closely related to a larger, still ongoing project (ICR due later this year), that established the CCRIF. The project team advised that this and the OECS project were added to the CCRIF project to enable the poorer Caribbean countries to participate in the insurance program and to enable them to understand and test its value for them . b.Were the project objectives/key associated outcome targets revised during implementation? No c. Components: Component 1. Payment of the participation fee to the CCRIF . (Appraisal US$2.57 million; Actual US$2.57 million). This financed the borrower's participation fees which needed to be at least equal to the annual premium . Component 2. Payment of the annual insurance premium . (Appraisal US$6.42 million; Actual US$6.42 million). This financed the borrowers' annual insurance premiums at 100% of the country specific annual insurance premium for the first two years (2007-08 and 2008-09) and half the premium for the third year (2009-10). d. Comments on Project Cost, Financing, Borrower Contribution, and Dates: The financing was provided in four installments issued to CCRIF at the request of the Ministry of Finance . The financing was withdrawn by the borrower from its credit account for direct payment into the Facility . 3. Relevance of Objectives & Design: a. Relevance of Objectives: Rated Substantial . The project is consistent with the 2009 - 20012 CAS which has as Pillar 3 Reducing Vulnerability to Disasters. The insurance program supported is referred to . The ICR (p. 7) reports that the Bank had been working with the borrower over the prior years on the implementation of an Emergency Recovery and Disaster Management Project aimed at reducing the country's vulnerability and strengthening emergency management . Thus, this project complemented an ongoing activity in disaster risk management in Haiti at the time . The ICR also notes that the Project contributed to the Section 4 Millennium Development Goal underscoring the need "to intensify our collective effort to reduce the number and effects of natural and man -made disasters". b. Relevance of Design: Rated Modest . There were a number of innovative and well -designed elements in this relatively simple project . This was the first regional parametric catastrophe risk pool in the world . (Parametric insurance is insurance that pays based on the occurrence of an event defined by a trigger measurement .) There were, however, weaknesses in the relevance of the design to the stated objectives revealed in the logic of the results framework (p. 21). It lacks specificity and focusses excessively on the insurance means rather than on the post -disaster country financial viability end which was clearly the intent of the objective . The indicator for the objective of reducing financial viability only measures the eligibility for insurance payment and the receipt in the event of a payment, a means rather than the specified end. The intermediate outcome: "CCRIF is created as a sustainable facility ." was only weakly connected to the indicator "total claims paying capacity". The indicator for the intermediate outcome : "The country benefits from partial coverage against hurricane and earthquake risks was : "The total sum insured ... up to 20% of total losses". This focused on an output as opposed to an outcome . It was not clear to what extent the 20% ceiling or the restriction to government losses and immediate impact were consistent with the objective . 4. Achievement of Objectives (Efficacy): The extent to which the project reduced Haiti's financial vulnerability to natural disasters (earthquakes and hurricanes). Rated Modest on balance due to questions about the significance of the scale of financial support in relation to overall post-disaster financial vulnerability and due to lack of evidence on impact on financial vulnerability . The ICR argues (p. 8) that the objective was fully met and that by joining the CCRIF and purchasing insurance, Haiti reduced its financial vulnerability to hurricanes and earthquakes . However, simply purchasing insurance is a narrow specification of the objectives . The project was supply-driven in that insurance was the modality on offer when the PAD PDO appear to invite a broader demand -driven approach. The ICR (p. 10) presents evidence that the payment of US$ 7.7 million for the January 12, 2010 earthquake was believed to be the first significant financial inflow after the earthquake . (The total coverage ceiling for Haiti for 2010-11 under the insurance policy was US$ 51.8 million (ICR p. 8), an amount that lay within the "up to 20%" limit.). Possibly influenced by the earthquake of 2010, Haiti later increased its earthquake coverage and reduced its hurricane coverage. The ICR also argues that the project achieved its intermediate outcome that the CCRIF became a sustainable facility as indicated by its strong claims paying capacity . However, this was not an explicit objective of this project . The ICR does not indicate what percentage of the total losses the 2010 payment represented. Thus, the extent to which the payouts met the objective of reducing financial vulnerability in terms of some percentage of the total losses remains unclear. The payout for the earthquake (ICR p. 11), according to the project team achieved within weeks, enabled Haiti to continue to pay civil servants, and, while such funds are, of course fungible, keeping civil servants working was probably important in getting quick action for recovery . However, this is only indirectly linked to the financial vulnerability objective. The ICR does not show whether the payment systems, such as salary payment, functioned adequately following the disaster . Additional benefits cited in the ICR include the TA that the CCRIF was able to provide on storm and flooding modelling and the project contribution to increasing awareness of natural disaster issues in the Ministry of Economy and Finance, although no evidence is presented to show this change in awareness . 5. Efficiency: Efficiency is rated Substantial . The ICR (p. 9) reports that the operating costs of the facility have remained within the agreed guideline of 5% of the annual premium volume. A comparator industry standard is quoted for the non -life insurance market of close to 30%. In addition, there appear to have been efficiency gains from the regional risk diversification. The ICR reports, but without giving a source, that it was estimated that for the 2008 - 2009 CCRIF insurance portfolio, the aggregate probable maximum loss from a 1 in 1,500 year event was 74% lower than the sum of the member countries individual policy limits, indicating that the insurance premium cost has been 74% lower when countries pool their risks than if they were to insure separately and individually . It was also estimated that the use of the facility for spreading risk and getting better terms on the reinsurance market reduced insurance costs by between 70% and 50% over simply the use of own reserves to address such disasters (ICR p. 9). Furthermore, the facility has been able to reduce its premium rates by 30% since its establishment. Most participant countries chose to increase their coverage or lower their deductible as a result of this improved pricing rather than lowering their premium payment. ERR )/Financial Rate of Return (FRR) a. If available, enter the Economic Rate of Return (ERR) FRR ) at appraisal and the re -estimated value at evaluation : re- Rate Available? Point Value Coverage/Scope* Appraisal % % ICR estimate % % * Refers to percent of total project cost for which ERR/FRR was calculated. 6. Outcome: The relevance of objectives was Substantial . However, relevance of design was Modest due to a disconnect between the objective of reducing financial vulnerability and the results framework for achieving it . On balance, efficacy is rated Modest mainly due to limited evidence on the impact on financial vulnerability . Efficiency is rated Substantial due to low operational costs and the significant savings from pooling risk . a. Outcome Rating : Moderately Satisfactory 7. Rationale for Risk to Development Outcome Rating: As noted by the ICR (p. 12), the main risk to the facility as a whole and the participation by the borrower is continuing fiscal constraints among the beneficiary countries including Haiti . The ICR anticipates that while the countries will probably continue seeking outside financing for their CCRIF premiums, it is believed that they will make every effort to continue their participation in CCRIF . The ICR also notes that the loss of a small number of beneficiaries would not seriously threaten the facility since it has been estimated that only about 8 countries are needed to obtain acceptable terms on reinsurance and to ensure affordable coverage . There is a proposal to add an excess rainfall product which is expected to bring in additional country clients . a. Risk to Development Outcome Rating : Moderate 8. Assessment of Bank Performance: a. Quality at entry: The ICR argues that the project was a path -breaking effort as the first regional parametric catastrophe risk pool in the world. It notes the concerns mentioned earlier about the results framework . It argues on page 13 that, in the indicator "total sum insured for each country (up to 20% of total losses)", the term "losses" referred to those suffered by the government rather than the physical damage in the country, and to those incurred "within the ... period ... of the country's (insurance) policies". The ICR notes the lack of definition in this area as a weakness in quality at entry. Related to this, the design and M&E do not sufficiently characterize the logical pathway towards the objective of reducing financial vulnerability . It remains unclear how far reduced financial vulnerability was to go and there is a disconnect between this quite ambitious stated intent and the means which was simply temporarily providing some modest insurance for a small part of the national risk . at -Entry Rating : Quality -at- Moderately Satisfactory b. Quality of supervision: As noted by the ICR (p. 13), this project required little direct supervision because there was no procurement and funds did not pass through the Haitian institutions or accounts but went directly to the CCRIF . Thus, the main supervision task was to supervise and support the facility itself . Over the three-year period of the project, there were three supervision missions and two midterm review reports . World Bank staff attended all but two of the CCRIF quarterly board meetings as observers . The ICR notes that the main intent was to strengthen the Operations Manual particularly with respect to financial management and procurement . The Bank also provided advice on pricing, risk, cash and asset management, stakeholder outreach, and technical assistance . Quality of Supervision Rating : Satisfactory Overall Bank Performance Rating : Moderately Satisfactory 9. Assessment of Borrower Performance: a. Government Performance: The main roles of the government was to understand and adopt the proposed insurance arrangements, to process the three or four withdrawal applications, and to liaise with the facility . Ministry of Economy and Finance official led the annual discussion with CCRIF regarding renewal of policies . Government Performance Rating Satisfactory b. Implementing Agency Performance: There were no implementing agencies at the borrower country level although there was the CCRIF that performed satisfactorily. However, this was not a borrower entity . Implementing Agency Performance Rating : Not Applicable Overall Borrower Performance Rating : Satisfactory 10. M&E Design, Implementation, & Utilization: a. M&E Design: There is limited evidence on M&E. The design essentially relied on the account and record keeping of the CCRIF . No evidence is presented on this aspect other than the indications in the ICR that the results framework was weak . b. M&E Implementation: According to the ICR, the Bank conducted in -depth reviews of the three years of CCRIF operation, interviewing members, donors, directors, service providers and other stakeholders . The CCRIF maintained its financial accounts . c. M&E Utilization: The accounts and reports were used by both the CCRIF and the Bank to monitor what was a fairly simple financing operation. M&E Quality Rating : Modest 11. Other Issues a. Safeguards: The project was an environmental category C project . No safeguard policies were triggered . Positive indirect impacts of the project were anticipated to include an improved ability of the government to deal with the environmental implications of disasters but there is little evidence presented in the ICR that this actually occurred . b. Fiduciary Compliance: There was a limited fiduciary role for the borrower . This was confined to submission of withdrawal applications for direct payment to CCRIF. The project did not finance works or consulting services and the only contracts were the participation and annual insurance contracts between the CCRIF and the participating countries . No procurement plan was needed. c. Unintended Impacts (positive or negative): Nothing that was clearly unintended . d. Other: 12. Ratings : 12. ICR IEG Review Reason for Disagreement /Comments Outcome : Satisfactory Moderately There were weaknesses in Relevance Satisfactory of Design. Efficacy can only be rated Modest due to questions about the extent to which the objective of reducing financial vulnerability were met. However Efficiency, in terms of cost effectiveness of the insurance, is rated Substantial. Risk to Development Moderate Moderate Outcome : Bank Performance : Satisfactory Moderately Quality at Entry was Moderately Satisfactory Satisfactory due to questions about the extent to which the project as designed could meet the objective of reducing financial vulnerability. Borrower Performance : Satisfactory Satisfactory Quality of ICR : Satisfactory NOTES: NOTES - When insufficient information is provided by the Bank for IEG to arrive at a clear rating, IEG will downgrade the relevant ratings as warranted beginning July 1, 2006. - The "Reason for Disagreement/Comments" column could cross-reference other sections of the ICR Review, as appropriate. 13. Lessons: The ICR offers a number of lessons of which the following are the most important (with some reformulation of language): 1. In designing a project supporting disaster insurance there needs to be clarity about the linkage between the objectives and how the project design will meet those objectives . If the objective is to reduce financial vulnerability then the design needs to address both means and scale . In this case, the project was supply -driven in the sense that the insurance solution was the modality on offer when the PDO seemed to invite a broader demand -driven approach. 2. Risk pooling through disaster insurance can be effective In reducing individual country financial risk, However, reducing costs and sustaining financial strength, calls for an optimum balance between reserves and reinsurance . 3. In the establishment of a catastrophe insurance facility, extensive communication with participants is important to ensure that the costs and benefits of its products are well understood . 4. The use of parametric insurance ( insurance based on immediately measurable physical parameters ) is particularly important in catastrophe insurance to ensure rapid cash injection through quick payout . However, this injection of funds can only be effective for the impacted country if the country itself has, and can sustain, an effective post-disaster budget execution system . 5. Even with a measure of insurance coverage, countries must continue to invest in risk reduction . Catastrophe insurance can only be a part of a broad risk management strategy . The cost of full financial protection would be prohibitive without proactive policy measures and physical investments to address the causes of risks and to mitigate vulnerabilities. 14. Assessment Recommended? Yes No 15. Comments on Quality of ICR: The ICR is clear and generally well argued . The lessons are particularly well formulated . However, a weakness is that the difference between the achievement of the means and the achievement of the ends (reducing financial vulnerability) could have been more clearly analyzed and differentiated . In particular, it would have helped to understand the achievement of objectives if the ICR had provided evidence on the share of the insurance payout as a percentage of total recovery resource needs, the share of the insurance payout as a percentage of immediately available recovery investment, and the application of, and social impact of, the insurance payouts to date . a.Quality of ICR Rating : Satisfactory