Document of. The World Bank

Transcription

1 Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized Municipal and Social Services Division Country Department IV Europe and Central Asia Region Document of The World Bank PROJECT APPRAISAL DOCUMENT ON A PROPOSED INVESTMENT LOAN IN THE AMOUNT OF DM 30.4 MILLION TO THE REPUBLIC OF LATVIA FOR A WELFARE REFORM PRO.IECT April 16, 1997 Report No.: LV

2 CURRENCY EQUIVALENTS Currency Unit = Lat Lat I = US$1.82 US$1 = Lats 0 S5 WEIGHTS AND MEASURES Metric System LATVIA - FISCAL YEAR January 1 - December 31 CAS - Country Assistance Strategy CPI - Consumer Price Index EU - European UJnion GDP - Gross Domestic Product ICB - International Competitive Bidding IMF - International Monetary Fund MOF - Ministry of Finance MOW - Ministry of Welfare NCB - National Competitive Bidding PCU - Project Coordination Unit SPRP - Social Policy Research Program PHRD - Policy and Human Resources Development Fund (formerly Japanese Grant Facility) SAD - Social Assistance Department (Ministry of Welfare) SID - Social Insurance Department (Ministry of Welfare) SSIF - State Social Insurance Fund SOE - Statement of Expenditure TOR - Terms of Reference UNDP - United Nations Development Program USD - United States Dollar Vice President, Europe and Central Asia Region: Johannes F. Linn Director, Europe and Central Asia Country Departmnent IV: Basil G. Kavalsky Chief, Municipal and Social Services Division, Country Department IV: Thomas Blinkhorn Staff Member/Project Task Manager: Louise Fox (Senior Economist, EC4MS)

3 Republic of Latvia Welfare Reform Project Project Appraisal Document Table of Contents PROJECT FINANCING DATA BLOCK 1: PROJECT DEp CtIPTION Project Development Objectives Project ' Thmponents Benefits and Target Population... 2 Institutional and Implementation Arrangements... 4 BLOCK 2: PROJECT RATIONALE CA O S bjective(s) Supported by the Project.5 6. Main Sector Issues and Government Strategy.5 7. Sector Issues to be Addressed by the Project and Strategic Choices.7 8. Project Alternatives Considered and Reasons Rejected.8 9. Major Related Projects Financed by the Bank and/or Other Development Agencies Lessons Learned and Reflected in the Project Design... 9 ]1. Indicators of Borrower Commitment and Ownership Value Added of Bank Support... 0 BLOCK 3: SUMMARY PROJECT ASSESSMENTS Economic Assessment Financial Assessment Technical Assessment Institutional Assessment Social Assessment... 1 I1 18. Environmental Assessment... ; Participatory Approach.11...I 20. Sustainability Critical Risks Possible Controversial Aspects BLOCK 4: MAIN LOAN CONDITIONS Effectiveness Conditions Other BLOCK 5: COMPLIANCE WITH BANK POLICIES... 15

4 Anne ii ANNEX 1: PROJECT DESIGN SUMMARY AND MONITORING INDICATORS ANNEX 2: DETAILED PROJECT DESCRIPTION ANNEX 3: ESTIMATED PROJECT COSTS ANNEX 4: COST EFFECTIVENESS ANALYSIS SUMMARY ANNEX 5: FINANCIAL SUMMARY ANNEX 6: PROCUREMENT AND DISBURSEMENT ARRANGEMENTS ANNEX 7: PROJECT PROCESSING BUDGET AND SCHEDULE ANNEX 8: DOCUMENTS IN THE PROJECT FILE ANNEX 9: STATUS OF BANK GROUP OPERATOINS ANNEX 10: LATVIA AT A GLANCE TEXT B BOX 1: LATVIAN PENSION REFORM MAP IBRD No

5 INTERNATIONAL BANK FOR RECONSTRUCTION AND DEVELOPMENT INTERNATIONAL DEVELOPMENT AsSOcIATION Europe and Central Asia Regional Office Country Department IV Project Appraisal Document Republic of Latvia Welfare Reform Project Date: April 16, 1997 [] Draft [x ] Final Task Manager: Louise Fox, Senior Economist, EC4MS Country Manager: James Q. Harrison, Chief, EC4CI Project ID: LV-PA Sector: Social Lending Instrument: Specific Investment Loan PTI: [x] Yes [I No Project Financing Data [x] Loan [] Credit [ Guarantee [] Other [Specify] For Loans/Credits/Others: Amount: DM 30.4 million (US$18.1 million equivalent) roposed Terms: [j Multicurrency [xj Single currency Grace period (years): 3 [ I Standard [x] Fixed [] LIBOR-based Variable Years to maturity: 15 Commitment fee: 0.75% Service charge: none , Financing plan: (US$ millions) Source Local Foreign Total Government IBRD Donors TOTAL Borrower: Republic of Latvia Guarantor: N/A Responsible agencies: Ministry of Welfare; Ministry of Finance; State Social Insurance Fund; Kandava Municipality Estimated disbursements (BankFY/US$M: Annual Cumulative Expected effectiveness date: June 30, 1997 Closing date: December 31, 2002

6 2 Block 1: Project Description 1. Project development objectives (see Annex 1 for key performance indicators): The project objective is to support the development of a more efficient and effective social welfare system. As a result of the policy reforms already approved but not yet implemented, as well as those to be developed and implemented during the project period, Latvia should have a modem, fully affordable social insurance and old-age security system. Progress in achieving this development objective will be evaluated on the basis of the following: (a) improved financial health of the social insurance system resulting from successful implementation of the Government's reform program; (b) a client-oriented social insurance administration will be implemented, measured by improvements in cost efficiency, processing speed and accuracy, improved external accountability and transparency, effective internal monitoring and evaluation, uniform handling of legislation and good client service, and flexibility to introduce new regulations; (c) private pension plans will become professionally supervised, financially stable, and viewed by the public as a viable savings alternative; (d) more efficient and effective delivery of social assistance will occur, including an increase in the number of clients of programs that provide alternatives to institutionalization of vulnerable groups; (e) high quality and professional policy monitoring and evaluation will become the norm, with Government and Parliament using applied social policy research outputs in their work; and (e) efficient and timely implementation of the project will occur. 2. Project components (see Annex 2 for a detailed description and Annex 3 for a detailed cost breakdown): Contingemcies (S$M %f Component Caego tal 1. Social Insurance Development: to improve Policy; institution building; physical the development and implementation of social insurance policy by the Social Insurance Department, and the management and administration of the social insurance budget by the State Social Insurance Fund 2. Regulation of Privately Managed Institution building Pensions: supports the provision of safe and supervised options for the Latvian public to make personal provisions for retirement through privately-managed savings schemes. 3. Social Assistance Administration: to Policy; institution building; physical improve the performance of the Social Assistance Department in developing social assistance policy, and the municipalities in delivering local level services. 4. Monitoring and Evaluation of Social Policy; institution building Policy: supports the development, implementation, monitoring and evaluation of the Ministry of Welfare's social policy agenda. 5. Project Coordination: supports the Project management effective provision of support and efficient management of the project in order to achieve the welfare reform project goals. Total I Benefits and target population: Benefits: * Social insurance: The component will allow the State Social Insurance Fund to implement the social insurance reform program. This reform program improves the affordability and equity of the social safety net and also

7 3 improves resource allocation while increasing public savings. As a result, the income level of the population should increase. Implementation of most of this reform would be impossible without the proposed investments. The expected benefits are: a) After ten years, and given successful implementation, the expenditures on pensions are projected to be one-third lower than they would have been under the old system (see "Latvian Pension Reform" in project file). This represents a savings of about 2 percent of expected GDP. This is primarily the effect of delayed retirement, and decreased benefits for early retirement, but it also results from switching to price (from wage) indexation of benefits. These savings will be gradually allocated to privately managed second pillar accounts. By 2005, it is expected that about ten percent of total contributions will be privately-managed and invested -- a potential increase in national savings of about I percent of GDP per year. Because of the strong incentives to delay retirement, the system dependency ratio will fall steadily from eighty-five percent to below fifty percent by the end of the period. At the same time, the payroll tax rate will be progressively reduced. Implementation of these reforms could raise economic growth by percent over the project period, and even more thereafter. b) The new organization and administration of social insurance will be significantly more effective, flexible enough to adapt to future change in a cost-effective way, and focused entirely on serving the clients. Under the new organization (1) new regulations and procedures will be easily introduced; (2) the time of clients (employers and beneficiaries) will be saved, with some benefit decision times reduced from days or weeks to the first client visit; (3) standard treatment of clients will be provided under the law; (4) effective payment systems will be in place; (5) new accounting and financial management systems will improve external accountability and transparency; and (6) internal monitoring and evaluation will be more effective. Fast and reliable public information on rights and responsibilities would be made available. The new system will also increase the collection of tax revenue. * Regulation ofprivately managedpensions: This component will increase old age security by providing a safe and secure place for private savings for old age. The system should be of benefit to all labor force participants and their families. The privately managed pension system in Latvia may have as much as I percent of GDP under management beginning in This is expected to grow and will reach an estimated 20 percent of GDP by the year 2025 through mandatory and voluntary contributions. This development will introduce institutional investors to the capital markets of Latvia. Their appetite for long-term investments will provide the vehicle for financing options that are not available to local entrepreneurs at present, as well as a supply of local financing for large projects such as infrastructure requirements. The financing burden of the state government will thus be sharply reduced. The development of pension funds, with their long-term savings nature, will lead to a deepening of the capital markets in Latvia.. Social assistance: The project activities will strengthen'the capacity of the national government to assist localities in targeting and assisting the poor, including providing poor and vulnerable groups with services that are more responsive to their needs during the transition to market. Successful replication of the Kandava pilot will improve the quality of life and productivity of individuals in danger of isolation, homelessness, or institutionalization. The project will also support the development of an appeals mechanism that will provide benefit adjudication rights consistent with Latvian law. * Monitoring and Evaluation of Social Policy: In 1995, total government expenditures on social insurance and assistance amounted to 14.9 percent of the Latvian Gross Domestic Product (GDP), but the effectiveness of this spending is not known. Weak evaluation capacity results in imprecise assessments of social policy options, resulting in policies that fail to achieve their goals. Improved social policy monitoring and evaluation will allow policy-makers to identify groups that are the most vulnerable and to design and implement effective policy responses. Better targeted social policy should reduce poverty and improve the efficiency of public spending. Beneficiaries: This project will benefit social insurance contributors, employers, poor and other vulnerable groups, pensioners, Kandava municipality, the Ministry of Welfare, the State Social Insurance Fund, and the Inspectorate of Private Pensions.

8 4 Poverty focus: The social assistance and monitoring and evaluation components specifically target the poorest and most vulnerable populations in Latvia. Gender focus: Project design has incorporated a gender focus. In the Monitoring and Evaluation component, training programs will include an analysis of gender issues in transition, and research projects in this topic are eligible for funding under the project supported Policy Research Program, which should assist policy-makers to assess gender related issues and appropriate policy responses. In the Social Assistance component, the Kandava pilot center will undertake a needs assessment, mid-term and final evaluation that will pay special attention to gender issues, including the needs and characteristics of women in the region, and the development and impact of regional programs on women. 4. Institutional and implementation arrangements: Implementation period: June 30, June 30, 2002 Executing agencies: Ministry of Welfare; Ministry of Finance; State Social Insurance Fund; Kandava Municipality Project coordination: A Project Coordination Unit was established in July, Currently, there are 5 full-time positions, which are all filled. With financing from the Swedish Government and UNDP, the PCU has received extensive training, and has begun handling complex procurement, disbursement, and accounting tasks associated with the project PPF and project trust funds. The PCU will be responsible for the timely implementation, monitoring, and evaluation of project activities and will ensure effective administration of critical project activities. For this purpose an Operational Manual describing the organization of the PCU, its functions, and responsibilities, and the linkages between the different entities, as well as report formats and administrative steps to be followed in implementing and monitoring the project is being prepared by the PCU. The PCU will also be the focal point for public information about the project and donor activity coordination. Project oversight (policy guidance, etc.): The project Steering Committee will provide overall policy guidance. The Steering Committee is chaired by the Minister of Welfare, and includes representatives from the senior management of the MoW and State Social Insurance Fund (SSIF). The PCU will report directly to the Steering Committee and serves as its secretariat. Accounting, financial reporting and auditing arrangements: The PCU will maintain separate accounts and records for each project component, and consolidate them into the project accounts and SOEs. Supporting documentation will be made available to Bank missions and independent auditors as required. It will also prepare and provide the Bank semi-annual reports on implementation progress along with a financial statement of project expenditures. The PCU will be responsible for monitoring the project's financial information, ensuring that all project expenditures and commitments are consistent with the project's agreed allocations for expenditure categories. All project-related accounts will be prepared in accordance with internationally accepted accounting procedures. Project accounts, including the Special Account and SOEs, will be audited annually in accordance with appropriate auditing principles applied by external private independent auditors acceptable to the Bank, with terms of reference for auditors and reports approved by the Bank. The costs of the consultant services to be employed for such annual audits will be fully incurred by the Govermment. Audit reports will be furnished to the Bank within six months of the end of the Government's fiscal year. The Operational Manual will provide guidance on accounting and auditing to the Government in the implementation of Bank assisted projects, drawing from information provided in the Financial Accounting Reporting and Auditing Handbook (FARAH). The Operational Manual will also provide template formats for financial statements, standard forms for audit documentation, and sample generic terms of reference for audits. Monitoring and evaluation arrangements. The Steering Committee and the PCU will be responsible for monitoring project implementation activities. The Monitoring Plan (included in the Operational Manual) includes component perfornance indicators developed by the working groups which will be used to track progress made towards objectives (see Annex I). The PCU will ensure that all reporting requirements under the project are carried out in accordance with the appropriate formats and with the agreed submission schedules. Semiannual project implementation reports will be submitted to the Bank by June 30 and December 31 of each calendar year, with the latter also including a draft Annual Action Plan for the upcoming calendar year for Bank approval. The Government and the Bank will conduct joint reviews in an annual supervision mission on the progress made in the implementation of the project of each year. During these reviews, the Bank and the Government would agree upon proposed Action Plans incorporating the results of the process evaluations. Only those activities in the approved plans would be

9 eligible for project financing. A formal mid-term review will be conducted by the Bank and the Government, where progress toward all targets will be assessed. Project design will be adjusted if needed. An Implementation Completion Report would be submitted to the Bank promptly after the completion of the project, but in any event no later than six months after the Loan Closing Date. Block 2: Project Rationale 5. CAS objective(s) supported by the project: Objective: Improved delivery and sustainability of the social safety net and basic social services. Document number and date of latest CAS discussion: Category 4 country, assistance strategy update discussed December 19, 1996, in the context of the Structural Adjustment Loan (R96-246). 6. Main sector issues and Government strategy: By 1993, Latvia realized that the social protection system inherited from the Soviet period was: (a) too expensive given current fiscal resources; (b) inequitable; (c) non-transparent, encouraging rent-seeking behavior; and (d) lacking the flexibility needed during the transition. In 1994, the Government prepared a comprehensive new set of legislation governing cash benefits and non-cash social assistance to families, covering: (i) social assistance, (ii) unemployment benefits; (iii) pension benefits; (iv) work injury benefits; (v) sickness and maternity leave, and (vi) social insurance financing. This legislation all passed Parliament in In social assistance, the legislation was designed to implement Latvia's plan to decentralize social services. In social insurance, the goal was to tie benefits more closely to contributions. This required removing non-insurance benefits from the insurance program financing, placing them in the state budget. Included in this legislative package was the first step in a fundamental reform of the pension system: creating a new, three tier system with increased reliance on individual savings efforts (privately managed) for income replacement in old age (see Box 1: Latvian Pension Reform). In 1996, three additional pieces of legislation regulating the welfare system were submitted to Parliament. The first was legislation creating the regulatory framework for licensing and supervising privately-managed pensions. Initially, most of these programs will be employer sponsored plans. However, it is envisaged that in 1998, a portion of the mandatory contributions to the Pay-as-you-go (PAYG) scheme will be channeled into privately-managed funds. The second was a piece of consolidating legislation, rationalizing coverage and contribution rates across insurance programs and adopting standard definitions (consistent with other legislation, including other tax codes). Reduced payroll tax rates (corresponding to reduced insurance coverage) will be offered beginning in 1998 for farmers, the self-employed, and those over 60 who continue to work. This legislation also sets the stage for the transfer of the administration of social insurance contribution collection (the social tax) to the State Revenue Service (SRS). In 1997, legislation completing this transfer will be prepared, as will legislation regulating the assignment of contributions from the public PAYG system to privately-managed funds (the second tier). Finally, a new disability pension formula was put in place, harmonizing this system with the rest of the reforms. This policy agenda is also supported by the Latvia SAL. Implementing this ambitious policy agenda will be a five year task. Policy-making and administrative agencies need to be transformed to handle their new responsibilities. For example, in order for the incentives in contribution-based system to function, individual accounts must be established and maintained by the social insurance agency. This infornation must also be available to contributors on a timely basis. Regulatory agencies must be strengthened to insure that financial intermediaries investing pension funds are protecting the safety and security of assets. Municipalities need support to assume their new responsibilities. The Ministry of Welfare needs to develop effective monitoring and evaluation mechanisms in order to track progress towards policy objectives, and to improve the ability to develop effective policies in the future. Social insurance. The Ministry of Welfare Social Insurance Department is formally responsible for developing policy, drafting legislation and submitting it to Government and Parliament, monitoring the impact of policy, and overseeing the Social Insurance Fund. The State Social Insurance Fund is the agency responsible for administering benefits, system budgeting, providing financial and management information to the Social Insurance Department, and assessing the administrative implications of new legislation. The SSIF prepares its own plans and budgets and submits them to the SID.

10 6 The current organization, facilities and equipment of the SSIF are inadequate to administer the new social insurance laws. * Inadequate financial management systems have resulted in no differentiation between Fund assets and liabilities for benefits and Fund administration to date, and no ability to properly analyze administrative costs. Accountability is thus limited under the existing system. * The antiquated computer system does not permit the maintenance of individual accounts as envisaged by the legislation. Nor does it allow for sharing of information on clients among offices and across services, slowing benefit processing and limiting the cross-checking needed to prevent fraud and abuse of the system. * Local office organization is functionally disjointed, so that the functions have specialized administrative systems. This results in poorly developed information and work flows between the systems and with the central office. Each administrative function is undertaken by an overly bureaucratic apparatus that engages in activities that frequently do not add value or that are duplicative. The organization has too many levels to operate efficiently, and system controls are inefficient. Social assistance. The Ministry of Welfare's Department of Social Assistance oversees policy design, most state programs, licensing of institutions, and adjudication of disputes over local benefits. Since the passage of the 1994 decentralization and social assistance legislation, the financing and administration of social assistance and care is divided between the national government, 26 regions, the 7 largest cities in Latvia, and 562 municipalities. The seven cities perform all of the functions of both the region and the municipality. Each municipality/city is required by law to open a social assistance office, but the size and organizational structure of that office is left to its discretion. Institutions are mostly managed by the state or the region, as defined in the legislation. The social assistance office administers the legally mandated means-tested benefit and any other benefits at its discretion. Municipalities have autonomy in determining eligibility and assessing benefit levels for the meanstested benefit within the criteria set by state law. The social assistance reform is ambitious in its objectives to move away from the highly centralized Soviet-era social assistance system, towards a system which responds to the needs of the community and individual. However, the design of the system has a number of shortcomings.. A national social safety net to ensure universal coverage and facilitate the transition to market is not realized. Under the 1994 decentralization program, municipalities were granted full autonomy in spending and full responsibility for financing and administering the safety net. The Government has attempted to address the gaps in provision to the truly needy through the issuance of detailed regulations. However, these are generally not followed at the local level. As a result, Latvia has an uncoordinated network of 595 local, city and regional offices, and a mixed provision of social assistance with no guarantee that the poor or categorical groups will receive equivalent treatment. * The new financing system does not encourage the development of community-based alternatives to institutional care. Multiple financing sources for institutions (all with input-based budgeting) do not encourage the substitution of community based alternatives for institutional care. Many municipalities are too small to develop and effective social assistance system. Most municipalities range in size from 500 to 5,000 residents, and spend some 15 percent of their budgets on administration alone. Small municipalities do not have the resources to provide a full range of services -- in some cases, there are only one or two staff who may lack professional training. Because many municipalities are so small, branch offices are reportedly overburdened by the services they are expected to provide, and they are also largely unable to develop new forms of care or service due to the administrative demands that are placed upon them (municipal staff engaged in means tested and other administration of cash benefits have little or no possibility to develop a community based senior citizens program). A baseline assessment in the Municipality of Kandava in late 1996 showed that four poorly trained social workers were serving 124 children in families with severe problems such as battering and alcohol abuse, 240 residents over the age of 80 and 243 other elderly living alone, 252 disabled adults and 21 children, and 991 children in families with less severe hardships but on the margin.

11 7 The Government has developed a strategy to address these key institutional development priorities. * In the Ministry of Welfare, the two relevant policy departments, (Social Insurance Department and Social Assistance Department) have both reviewed their activities, processes, and organization in light of their new responsibilities. Business plans for their development have been prepared. * Two new agencies are being created under the supervision of the Ministry of Welfare. The existing State Social Insurance Fund (SSIF) is being transformed into the independent Social Insurance Services Agency (see Section 7), with a new organizational structure focusing on delivering services in an efficient, client-oriented manner. An independent Social Assistance Fund has been created to oversee the increasingly decentralized network of social care institutions and to support the development of community-based alternatives to institutionalization for vulnerable groups. Both of these agencies have also prepared business plans describing their institutional goals and means for achievement. * The Ministry of Welfare has created a new Strategic Department, responsible for overall policy monitoring and evaluation as well as for assisting specialized departments to perform their own monitoring and evaluation functions. A project coordination unit has been formed within the Strategic Department to oversee project implementation.. The Ministry of Finance is strengthening the Department of Insurance Supervision to handle the enhanced responsibility of supervising pensions. * A local government reform program is being prepared to improve the efficiency of subnational governments. The regional level of government will be abolished by 1999, and smaller municipalities will be consolidated. Implementation of this reform should go a long way to improving the efficiency of social service delivery at the sub-national level. Once this reform is prepared, the Ministry of Welfare will propose a new financing system for social assistance which will move from an input-based system, favoring institutionalization, to one where "money follows the client," favoring least-cost, community-based solutions. [For further information, see documents in project file and the President's Report on the Structural Adjustment Loan (P-7019, LV)] 7. Sector issues to be addressed by the.project.andistrategic oices T e ch first str aegic choice faced by the Government was on the principles, and timing and sequencing of the welfare reform prograrn. The government chose the very ambitious approach of reforming all entitlements at once. Timing and sequencing was motivated by: (a) the strong desire to decentralize social service delivery as far as possible (social assistance); and (b) by concerns about the affordability, efficiency effects, and equity of the system of social insurance inherited from the Soviet period. In reforming entitlements, the Government sought to create a system which would serve Latvia in the transition to a more open society, integrated into Western Europe, and a market economy through: (a) separating insurance benefits from assistance ones; (b) moving the responsibility for administering the assistance benefits to as close to the target population as possible through decentralization; (c) basing entitlements to insurance on contributions (reducing the inefficiency associated with social insurance); and (d) increasing the role of private and public savings in assuring old-age income support. Once the Government developed the main policy proposals for the reform and drafted the legislation, the institutional development effort began. Ideally, this effort would have commenced simultaneously with the development of policy. Inexperience typical of a transition economy, combined with the sheer ambitiousness of the Latvian effort, probably caused the early neglect of implementation capacity. However, the Government has now moved rapidly on this issue. The key strategic choice has been what kind of agency or agencies should be responsible for administering contributions and benefits, and the financial management of the separate social insurance budget. * With respect to contributions, the Government has decided to stop investing in two separate agencies (SSIF and State Revenue Service (SRS)), and instead to consolidate all revenue administration functions in one agency: the SRS. This transfer of responsibilities will occur over the next 18 months, and is also supported by the SAL. * With respect to benefit administration and financial management, the Government has decided to create one independent public sector agency for this purpose, rejecting the alternative of separate agencies for each program.

12 8 By 1998, the agency will be independent of the civil service, with its own personnel structure, and a new organizational and management structure will be in place, focusing on client service and flexible, efficient and professional service delivery. A new financial management system will also be put in place, following International Accounting Standards (IAS) principles. To regulate privately-managed pensions, the Government has decided to expand the Department of Insurance Supervision to include pensions, rather than to create a new agency or to assign the responsibility elsewhere in the Government. Major changes in the institutional framework of the social assistance system will depend on the outcome of the reform of local governments, including the decision on what, if anything, will replace the regional level. However, the business plans of the national agencies both recognize the decentralization goals and propose increased emphasis on support to municipalities through training and dissemination of best practice, moving away from the traditional "control" approach of the Soviet system Project alternatives considered and reasons rejected: Type ofproject: The government requested a broad-based investment project, supporting the whole program of welfare system reform. A more narrowly-focused project was therefore rejected. Also rejected was a hybrid loan to support both policy and institutional development, as the Government requested a SAL which covered this area (and more). Guarantee: The option of including Bank support for a guarantee scheme for the privately-managed pension system was discussed in the Bank and with the Government. This option was rejected. If Bank financing for such a scheme is eventually needed, it will be considered in a subsequent operation. Social assistance component: Various options for Bank support to the evolving social assistance system were discussed. However, development of this system is dependent on the on-going local government reform, since the responsibility for social assistance has been devolved to the sub-national level of government. The MoW and the Bank agreed that this operation would support a small pilot effort to develop new program models, and limited training and technical assistance in the Ministry. Once more viable political units are created at the local level, these models could be replicated with support from other donors or local funds. A number of sites were considered for the pilot location. The locality of Kandava was chosen because this area was already working with the Ministry of Local Government Affairs on consolidation of social service delivery. This pilot then formed a part of this on-going effort. 9. Major related projects financed by the Bank and/or other development agencies (complete, ongoing and planned): Sector Issue Project Latest Form 590 Ratings IP DO BANK-FINANCED PROJECTS Transition to the market oriented Structural Adjustment Loan. S S economy supported by strengthening US$60.0 million. LV-PE fiscal management. Signed 12/20/96, closes 12/31/97. Technical, financial, and economic Municipal Services Development. S S management of municipal services US$27.3 million. LV-PE under decentralization supported. Signed 1/26/96, closes 6/30/01. OTHER DEVELOPMENT AGENCIES Support for Ministry of Welfare Social Insurance Reform Project to prepare and implement a Swedish International Development welfare reform program Cooperation Agency (Sida). TF and TF SEK 15,422,400 ( ).

13 9 Sector Issue Project Latest Form 590 Ratings IP DO Support for social work training Social Work Training Program. Sida SEK 871,800. ( ). Improving the efficiency and Strategic Support to Mo W Program effectiveness of the MoW organization EU-PHARE. LE ECU 600,000 in carrying out its core functions ( ). Developing management capacity Social Insurance Management ProgramL of the social insurance organization EU-PHARE. LE ECU 836,000 ( ). Developing Social Assistance Department - Social Assistance Department/NGO Program NGO cooperation EU-PHARE. LE ECU 540,944 ( ). Training a supervisory body for Private Pension Supervision Project private pensions and developing proper UK Know-How Fund. US$288,028 organization, administration, and staffing to be finalized ( ) Developing capacity to monitor and evaluate Social Policy Monitoring and Evaluation-- -- the impact of social policy during the Project. Netherlands/UNDP/World Bank. transition to market LAT/95/003. US$410,000. ( ) Lessons learned and reflected in the project design: To date, the Bank has implemented a Rehabilitation Loan in Latvia (FY96). The lessons learned from that project (Report 15173) include:. Strong and sustained Government commitment to policy reform is needed along with aflexible and active policy dialogue. Government commitment to social welfare system reform has been amply demonstrated by the legislation already adopted (including the first major reform of a public pension system in Eastern Europe). The project will provide the Government with the resources necessary to sustain its commitment to a new, innovative social welfare system.. The appointment of well qualif ed procurement and disbursement advisers to the Project Coordination Unit is important given the inexperience of this new borrower: The Project Coordination Unit (PCU) was established and fully staffed immediately after pre-appraisal. Extensive technical assistance and training has been provided to set up the necessary procurement, disbursement, and accounting systems. The project will provide continued support to the PCU during implementation.. Higher than average resources are needed when working with new borrowers, particularly in loans that include policy reform components. Project preparation activities included substantial technical assistance and training programs supported by donor funds (Sida and PHRD). These activities will continue during the project period. Supervision requirements during the first year are estimated at 20 staff weeks. Successful projects in the social sectors are usually characterized by (i) a supportive policy framework, assessed by the Bank prior to project preparation; (ii) a high degree of local ownership, including of project stakeholders; and (iii) well-prepared implementation plans, with clear and monitorable goals. In this project, preparation activities focused on developing ownership by the counterparts of project objectives, activities, and outputs. Stakeholders are involved at the local and central levels in the preparation of all components. All components have developed clear implementation plans with measurable monitoring indicators (see project files).

14 11. Indicators of borrower commitment and ownership: * During preparation, two project workshops were held with all MoW management and staff involved in the project, as well as representatives from other concerned ministries and stakeholders, to discuss and agree on project objectives, activities, and monitoring plans. One of these workshops was facilitated to encourage counterpart participation. * All main components have prepared their own strategy documents as part of the preparation process. 10 * A project PCU has been established and is managing the project well, including the execution of a special trust fund.. H V * A PPF (US$1.99 million, dated October 18, 1996) has been agreed and signed. 12. Value added of Bank support: This project continues World Bank support for the design and implementation of the Region's first significant pension reform, which is serving as a model for other countries in the region. The Latvia pension reform is based on the innovative "notional defined-contribution" approach to public pension entitlement. The Bank is uniquely placed to support the ongoing reform given our international experience in this area. Especially important is the assistance to the Government across sectors and ministries to support development objectives, such as the implementation of regulated privately managed pension options. Implementation of this reform is a complex task requiring a coordinated program of support including technical assistance, training and equipment...i i...; a... 1' There is an essential public role to be played in the provision of a social safety net and the insurance of risks that are either uninsurable privately or that require a national program for efficient risk pooling. No private agency can provide this type of financial assistance, nor is any other donor agency able to respond in this sector. This project has attracted significant donor support, with approximately 20 percent of project costs covered by donor financing, allowing the Bank to serve as the financier of last resort. Most donors active in this area are now coordinating their programs with the Bank through parallel financing of project activities. Block 3: Summary Project Assessments (Detailed assessments are in the project file. See Annex 8) 13. Economic Assessment [ Cost-Benefit Analysis : NPV=US$ [xl Cost Effectiveness [] Other (see Annex 4): million; ERR= % Analysis: [Specify] Fiscal impact (for allprojects): Detailed economic analysis was only carried out for component 1 (83.4 percent of project costs). The fiscal impact by the end of the project period from full implementation of the welfare reform program is expected to be about 50 million 1997 Lats per year. After 10 years, this impact should rise to about 100 million Lats per year, or about 2.5 percent of projected GDP. Investments in component 1 will result in administrative cost savings equal to the investment cost after 6.7 years. In addition, benefit processing time will fall from up to 2 weeks, to less than I day, with a lower error rate, saving clients' time. The automated system will also save taxpayers time, as forms will be simplified and filed electronically. Benefits to the economy as a whole are summarized in section 3 above. Cost recovery measures will be applied where appropriate. For example, once the pension regulatory agency is set up and in full operation, fees will be imposed on fund management companies to cover the costs of running the agency. 14. Financial Assessment: Not applicable Technical Assessment: This project is technically sound. * Organizational changes to be implemented in this project reflect best practice in public sector service delivery agencies in OECD countries. The client-service principles followed in this design will lead to more flexible and efficient organizations, saving time of clients and staff.

15 11 * The computer solution chosen for the social insurance agency represents a good balance between functionality and cost. It uses a combination of object-oriented and traditional approaches to data-base management. All software will be fully tested in pilot operations before implementation nation-wide. * The automation and civil works investment programs were designed only after the core business processes had been reviewed and assessed to ensure that efficient, effective, client service principles were applied. A full information model was constructed to consider how to effectively introduce automation into the system, and what were the technical needs. This project preparation sequence insures that the project will not involve automating manual processes, but will instead use automation as a tool to improve business. Investment and recurrent cost estimates for the project are based on estimates of prevailing market unit cost (including recent experience in procurement under Bank-financed project in other transition economies), with appropriate allowance for inflation. An adequate allowance for physical contingencies has also been included in project costs. 4 ~.l... W... I Institutional Assessment: a. Executing agencies: All executing agencies are young agencies, created after Latvia regained independence. Despite this relative inexperience they have designed and implemented one of the most far-reaching reforms in any transition economy. They have benefited from substantial technical assistance and training during this period, which has built strong technical and managerial capacity in the executing agencies. This has been demonstrated by the high quality of the project preparation documents prepared by the agencies. It is expected that the need for technical assistance and training will decline over the project period. However, staff turnover is always a risk, given low public sector salaries relative to opportunities in the private sector. Ample technical assistance and training are included in project plans to mitigate this risk (especially important in the IT area). b. Project management: The PCU has done an excellent job in managing the project to date, despite the MoW's limited experience in implementing externally-financed projects. The proposed technical assistance and training for the PCU under the project will help to deepen its capacity to manage more complex tasks in the areas of financial management, procurement, accounting and reporting. 17. Social Assessment: Social assessments were u dertaken during project preparation to ensure that the project design and strategy are the most appropriate. A nationwide survey on societal knowledge and attitudes towards social insurance was commissioned and undertaken in the summer of 1996 (about 6 months after the initial wave of reforms). This survey found a very weak knowledge base in both employers and employees on the provisions of the new legislation. The project communication strategy is designed to address this knowledge gap. The effects of new organizational changes piloted in the Ziemelu social insurance office were evaluated rather informally in Clients expressed satisfaction with the changes, including the reduced waiting times which resulted from improved organization. A full client survey will be undertaken in 1997 by SSIF. A thorough sociological mapping of the population of the Kandava region was undertaken in 1996 in order to design the pilot. This survey identified a number of vulnerable groups who were not yet being served by social assistance services. An assessment of public knowledge and attitudes towards the newly developing private pension industry will be undertaken in early 1997 by the Government. The project will support follow-up social assessments of both social insurance/private pension clients (ongoing) and the Kandava municipality (middle and end of project). To build up local capacity, training in beneficiary assessment will occur as part of the Monitoring and Evaluation component. nvironmental Assessment: Environmental A B [x] C Category The proposed project is not expected to have a negative environmental impact. i..c-~ i... l~ Participatory Approach*: IdenU&fatioazPrearation Implementation Operation Beneficiaries/community groups COL COL COL Intermediary NGOs/Academic institutions COL COL COL Local government COL COL COL Other donors COL COL COL *COL(collaboration); IS (information sharing); CON (consultation)

16 Sustainablity: This project will develop the institutional capacity of the Ministry of Welfare and the State Social Insurance Fund. It is expected to be both cost-saving and quality enhancing. Borrower ownership is high, owing to the extent of participation in preparation activities. * Social insurance: The improvements in systems, processes and facilities will permit improved performance at a lower cost per standardized benefit. These efficiency improvements, combined with an improved financial position of the social insurance budget from implementation of the policy reforms, will ensure project sustainability (see Annex 4).. Privately-managedpensions: Initially, the costs of the supervisory system will be paid for by the state budget. However, by the end of the project period, licensing fees will be imposed on pension funds to recover the annual recurrent costs of the system. * Social assistance: The component supports the development of more cost-effective approaches to assisting vulnerable groups. Ongoing monitoring and evaluation activities will help to ensure sustainable models are developed. * Monitoring and evaluation: At the end of the project, a research network will be established that is sustainable beyond the life-time of the proposed activities. The acquired skills and attitudes in the training will be passed on to new civil servants and will be useful in the future, although it may be necessary to update the knowledge continuously. The Social Policy Research Program (SPRP) is expected to be continued beyond the lifetime of this project with increasing support from the Latvian government for local researchers and declining need for donor funds to finance foreign technical assistance. The need for policy analysis will remain, and changes in policies will have to be monitored and evaluated. Researchers will be trained in performing social research and the increase in private research institutes will raise the demand for research funds. The increased supply of high-quality research should also help ensure the sustainability of the SPRP. 21. Critical Risks:* Project outputs to development objectives Risk Risk Aa#ig Risk Mininuzation Measure Policy change: A major shift in policy NR A number of activities are in place to strengthen direction in Latvia could undermine the institutionalization of the policy reforms: achievement of the development objectives.. The SSIF has prepared and is implementing a communications program on the new pension system and the SSIF organizational development. A similar strategy is also under preparation for the privately-managed pension system (to be launched once the regulatory system is ready) and the social assistance reforms.. Monitoring and evaluation activities are being strengthened to track the fiscal and social impact of welfare policy. This information will be used to identify needed policy adjustments as well as to keep policy makers and the population informed of social developments. Technical assistance activities have strengthened the policy making and implementation capacity in the Ministry. Finally, should policy adjustments occur, SSIF business is now being organized in a way that will allow the maximum possible flexibility in benefit delivery. Organizational response time to policy change should be rapid. Staffchanges: The large difference between MR All components have been designed by Ministry staff, who are public sector and private sector salaries always committed to these activities. The opportunity to participate in leads to a risk of staff turnover slowing international training and otherwise learn on the job are also

17 13 implementation or even disrupting project activities. factors which help to retain staff. In the new SISA, a new, more flexible salary structure and consultant contracting procedure will be put in place to help retain core staff, including in particular IT staff. Monitoring and evaluation: Better information MR Technical assistance activities will include effective and analysis may not be used by policy makers. communication strategies. Bank participation in the policy dialogue may also lessen this risk. Social assistance: Proposed changes in the MR The project will provide substantial technical support for the financing of institutional care, (to be developed implementation of the White Paper on Social Assistance Policy following the regional reformn) may not take (January, 1997), which proposes a phased program to transfer place. If the state continues to finance much responsibility for financing of care to the municipalities once institutional care, the incentives to provide the regional reform is complete. alternative social care provision consistent with the lessons of the Kandava pilot will be weakened. Project components to outputs lu A Risk Ratin Risk Minimization Measur Social insurance: The Wide Area Network NR The SSIF will be actively involved in the development of the under development in Latvia and envisaged as network. Alternative technical strategies for internal system part of the automation plan would not be communications, particularly with some rural areas, are under developed sufficiently in the next five years, preparation. slowing planned communications between offices. A new banking crisis would disrupt the NR Emergency benefit payment procedures will be developed. envisioned automated payment links. The chosen technological level in the IT NR The new IT-systems will be thoroughly defined in light of systems could be too high, which would lead requirements and fully tested. Substantial training in this to complicated systems that are not working project component will be provided. The new SISA might not be able to attract MR The independent SISA will not be subject to civil service skilled people thus could not manage the rules, and will have flexible compensation and staff organization and the project. development policies designed to attract and retain high quality staff. The replacement of the existing computer NR center and staff could go wrong and pensions A special organization for employees who have to leave the and benefits would not be paid properly. organization will be set up in order to get all to feel safe even if they are dismissed. Private pensions: Due to the lack of MR The Inspectorate staff will be trained so they can serve in a experienced financial managers in Latvia and supervisory capacity in respect of both private pension low salaries in the Government, it may be insurance companies and private pension funds. The difficult to recruit and retain competent staff for Inspectorate will not be subject to the normal civil service the Inspectorate for Private Pensions. salary scale, and will have more flexibility in staff compensation (allowing it to compete with the private sector). Social assistance: Replication of lessons MR The pilot's final evaluation will include a financing study learned at the pilot could be hampered by aimed at facilitating suitable replication, including the municipalities incorporating only some pilot examination of appropriate cost recovery mechanisms features in order to save costs

18 14 Monitoring and evaluation: Inadequate internal NR All Ministry departments and agencies concerned have been communication and coordination may limit the involved in the preparation of this component. The Ministry effectiveness of training and technical has also requested technical assistance from EU PHARE to assistance activities. improve communication in the Ministry Project management and disbursement: The NR Intensive technical assistance has been provided to the PCU disbursement projections are optimistic targets. and the component I team to help them reach these targets. Lack of experience with Bank projects could Adequate contingencies have been included to insure sufficient cause a disbursement delay initially. project funds. Disbursement perfornance will be reviewed annually and targets revised if necessary. Overall project risk rating Risk ariskng Risk Minimiation Measur Overall project risk rating NR Overall project risks are mitigated by the well-developed policy dialogue in the social sectors, and the Government's clear commitmento a medium term reform that is consistent both with Latvia's transition to a market economy (as it is fiscally sound) and democratic society (as it is socially responsible, and consistent with the intent of the municipal reform). The risks are further mitigated by strong coordination with other external donors, and by a well-designed supervision strategy. *HR(high risk, >75%); SR (substantial risk, 50-74%o); MR (modest risk, 2549o); NR (low/negligible risk, <25%). 22. Possible Controversial Aspects: None Block 4: Main Loan Conditions 23. Effectiveness Conditions As conditions of loan effectiveness, the Government would cause: (a) the SSIF Subsidiary Loan Agreement to have been executed on behalf of the Borrower and SSIF; (b) an implementation protocol to have been executed on behalf of MOF and MoW on terms and conditions satisfactory to the Bank; and... (c) the Borrower to have adopted an Operational Manual satisfactory to the Bank. 24. Other [classify according to covenant types used in the Legal Agreements database.] 1. Project Management The Borrower shall maintain the Project Coordination Unit within MoW. 2. Flow and Utilization of Project Funds The Borrower shall cause SSIF to have its records, accounts and financial statements for each fiscal year audited in accordance with appropriate auditing principles consistently applied by independent auditors acceptable to the Bank. 3. Project Monitoring, Review and Reporting The Borrower shall and shall cause SSIF and the Municipality of Kandava to maintain policies and procedures adequate to enable it to monitor and evaluate on an ongoing basis, in accordance with indicators acceptable to the Bank, the carrying out of the Project and the achievement of the objectives thereof.

19 15 4. Project Implenentation (a) the Borrower shall apply an Operational Manual satisfactory to the Bank. (b) a Kandava Subsidiary Agreement satisfactory to the Bank shall be executed on behalf of the Borrower and the Municipality of Kandava prior to any loan withdrawals for this component. (c) The Borrower shall ensure that the automation plans for computer equipment, agreed with the Bank, are followed during Project implementation. (d) The Bank shall receive civil works plans satisfactory to the Bank for civil works to be financed under the Project. Block 5: Compliance with Bank Policies [x] This project complies with all applicable Bank policies. [The following exceptions to Bank policies are recommended for approval: applicable Bank policies.] The project complies with all other T'ask Manager: L use Fox, enior Economist (EC4MS) Country Manag^: BasilG sy ietr EC4

20 16 Box 1: Latvian pension reform Latvia has become the first country in Eastern Europe to begin implementation of a major reform of the public pension system. In November 1995, Parliament approved legislation creating a new public system for those who retire after 1995 (pensions for current pensioners were not affected). The Latvian pension concept provides an entirely new formula for calculating future pensions, one in which benefits are directly linked to individual contributions and to the retirement age. It also provides for the introduction of funded pension programs, initially as a supplement to the public system. In 1998, contributors will be allowed to allocate a portion of their payroll tax to individual, privately managed accounts (the second pillar). The new pension system is a "notional defined contribution" one, It is designed to mimic the contribution-based pension that would be offered in the private sector by an efficient insurance company. The system starts by giving everyone paying the social tax an individual account. As contributions earmarked for the pension system are paid, the account is credited, as if it were a savings account. The "capital" in the account earns a rate of return just like a savings account. This rate of return is equal to the growth of the sum of wages on which contributions are collected (the contribution wage base). At retirement, the pension paid is equal to the total capital in the person's account divided by the expected post-retirement life span for all those of that person's age. For example, if a person has 10,000 Lats in their account at retirement, and is expected to live 10 more years, the pension would be 1000 Lats per year or 83 Lats per month. The pension will be indexed, adjusting for price changes. In the new system, there is no mandatory retirement age and no "full pension." The minimum retirement age will be 60 years for most participants. But the system provides strong incentives to work longer. Longer working years reduces the number of payments, allowing each to be higher, and increases the initial capital. The new system includes a minimum guaranteed pension for all those who reach the age of 60. How Do the Actual Benefits Compare Under the Old and New Systemns? Assuming that under the old system, the required quarterly indexation would have taken place in 1996, most new pensioners with full years of service retiring after 60 do not suffer a benefit decline compared with the old system. For those who work longer, benefits increase -- by over 200 percent for those who work another 10 years. Benefits are also larger for those with higher incomes (who in the future will also contribute more). Early retirees will be penalized. For example, a woman who retires at 55 will experience on average a 25 percent decrease in her pension under the new system if she stops contributing. Those who retire even earlier will suffer an even greater loss of benefits. It is expected that these strong incentives to work and contribute will drastically reduce the number of early retirees. Macroeconomic Impact. After 10 years, the expenditures on first tier pensions are projected to be 1/3 lower than they would have been under the old system (see also Block 1.3 and Annex 4). The system dependency ratio will fall steadily to below 50 percent by the end of the period. Provided Latvia realizes at least some economic growth, the affordability of the system should be achieved. Other Economic Effects. As benefits are completely and fully dependent on contributions in the new system, a large part of the disincentive effect of a traditional social insurance tax disappears. When benefits are unrelated to contributions, the social insurance contribution becomes a tax, and like any other tax embodies a loss of income and utility to the payee. The closer benefits are related to contributions, the less loss of income and utility the system implies, and the easier administration becomes. There is always some loss of income and utility associated with a mandatory contribution system, however. The new system also permits a flexible adjustment of benefits to changes in life expectancy. This avoids the acrimonious debate over retirement age. The system also has the advantage of containing one simple fornula which covers all circumstances. In doing so, it provides a wider range of options for older workers. Anyone over the age of 60 can take a pension if they desire. Or they can continue working and get a higher pension later. Or they can do both, and get a slightly higher pension later. Or they can stop working for a while, and then start up again. These options can be especially important for women (who tend to have greater responsibility for taking care of dependent family members), and those whose health may be weakening, but still wish to work part time and require limited income support.

21 17 Annex I Latvia Welfare Reform Project Project Design Summary Narrative Summary Key Performance Indicators Means of Verification Critical Assumptions CAS Objective To improve the delivery and I Sustained improvements in I Annual reviews of financial sustainability of the social the Social Insurance Fund performance safety net and basic social balance sheet services 2 Improved client satisfaction 2 Annual client surveys Project Development Social Insurance System (PDO to CAS Objective): Objectives To develop a more efficient I Affordability of the social I Annual reviews of financial Economic growth is and effective social welfare insurance system is improved, performance sustainable system as indicated by reduced pension expenditures and Lessons learned in the increased revenues into the Kandava pilot fully replicated Social Insurance budget. by municipalities 2 Standard service cost per 2 Agency Annual Report benefit or contribution reduced by 20 percent from Rate of client satisfaction 3 Annual client surveys increases by 10 percent per year from Number of residents well 4 Annual sample surveys by informed about rights and the new SSIF Evaluation responsibilities increases by Division 10% by 1999 and a further 15% by 2000 Privately Managed Retirement System 5 Private pension plans viewed 5 Annual review of volume of by public as viable savings alternative private pension savings; and mid-term and final project evaluation Social Assistance 6 Introduction of a streamlined 6 Site inspections benefit system 7 Introduction of central 7 Mid-term and final dissemination of best practice evaluation in social assistance delivery that is supportive of community-based programs to I help the poor and vulnerable.

22 18 Narrtiv Summary 01Key Performance Indicators Means of Verification Critical Assumptions I 8 Introduction of new 8 Annual review of SAD financing system for institutional monitoring system institutional care leads to more and municipalities' programs non-institutional care being provided by By December, 1999, 9 Mid-term and final recurrent costs per person evaluation served at the Kandava pilot are lower than the recurrent costs per institutionalized person. 10 By December 1999, 10 Client survey at mid-termn as improved service delivery in compared with baseline survey pilot area leads to 50% rise in satisfaction levels. Monitoring and Evaluation Project Outputs: 11 High quality policy 11 Independent review of monitoring and evaluation the Annual Social Reports, norm Research reports (Project Outputs to PDO) 1 Improved development and 1.1 Policies furthering the 1.1 Annual policy review No major shift in welfare implementation of social Government's social insurance policy insurance policy by Social policy agenda are developed Insurance Department, and and successfully implemented Ability to recruit and retain improved management and competent staff for the PCU, administration of the Social 1.2 SID and SSIF business 1.2 Annual review of Business Inspectorate for private Insurance Budget by the State plans implemented, including Plan targets by new SSIF Pensions and the new SISA. Social Insurance Fund introduction of the new Evaluation Division organization 1.3 Object oriented databases 1.3 Site inspection are functional and connected to central databases by 12/31/ Social tax collection 1.4 Site inspection transferred to State Revenue Service (SRS) by 1/1/ New and transparent 1.5 Site inspection accounting system in place for SSIF by 1/98 2 Improved regulation of 2.1 Number of licenses 2.1 Reports published by the privately managed pensions granted, voluntary participants, Inspectorate (supervision and publication of an annual mission confirmation of report by the Inspectorate. quality) 2.2 Staff of lnspectorate and 2.2 Staff skills assessment at of funds use newly acquired mid-term and final project skills in their work stages. Policy makers interested in social policy research. Proposed changes in the financing of social care take place. Consolidation of local government units proceeds according to schedule.

23 19 Narrative Summary Key Performance Indicators [ Means of Verification J Critical Assumptions 2.3 Number and nature of 2.3 Inspectorate records complaints received and reviewed. 2.4 Public relations materials 2.4 Site visits/press reviews disseminated to public beginning in Inspectorate employs 2.5 Site visits administrative tools effectively, such as supervision manuals, reporting formats, and newly automated fund monitoring systems 3 The performance of the 3.1 New organization and 3.1 Site inspections Social Assistance Department administration of SAD in place in developing social assistance by 1/98 policy improves 3.2 Over 70 percent of SAD 3.2 Staff skills assessment staff apply new skills and administered by SAD in 1/98 knowledge by 12/01 and annually thereafter 3.3 Submission to Parliament 3.3 Departmental reports of legislative package on new principles for financing institutional care by end- 1997, and streamlined package of means-tested and categorical benefits by mid SAD develops and 3.4 SAD report on impact of implements a new adjudication new system by end-2000 system by mid-1999 that leads to more uniform benefit delivery 3.5 MoW and municipalities 3.5 Seminar and workshop cooperate to develop and evaluation. site visits implement best practice in service delivery 3.6 Kandava pilot office 3.6 Mid-term and final begins operations by end-1997, evaluation site visits a mid-term evaluation is carried out by end- 1999, and a final evaluation including a plan for replication by mid MoW's development, 4.1 Municipalities report 4.1 Mid-term and final review implementation, monitoring monthly to national level using and evaluation of its social new administrative monitoring policy agenda improves system

24 20 Narrative Summary Key Performance Indicators Means of Verification Critical Assumptions 4.2 Ministry provides quarterly 4.2 Site visits feedback report to municipalities on data system 4.3 Municipalities employ 4.3 Site visits information to improve social assistance administration 4.4 Annual Social Report 4.4 Site visit published regularly beginning in MoW staff trained in 4.5 Site visit evaluation methods 4.6 Evidence of active 4.6 Parliamentary record, press references to the Report by review policymakers and the public 4.7 Staff use skills in social 4.7 Staff surveys policy design, monitoring and evaluation 4.8 At least 50 percent of 4.8 Press record major conceptual proposals of social policy issues are discussed in public by Capacity of research 4.9 Independent review of institutes developed; high social policy research quality social policy research supported by ministry meeting international standards is performed 5 Effective provision of 5.1 Resources procured in 5 Supervision mission support and efficient conformity with all interviews with PCU staff and management of the project specifications, and in stakeholders, reported accordance with World Bank periodically in Form 590 guidelines 5.2 Project accounts provide a reliable basis for financial management and planning 5.3 Satisfaction by all stakeholders with PCU performance is high

25 21 Narrative Summary Key Performance Indicators Means of Verification I Critical Assumptions Components: Inputs: (Components to Outputs): (See Annex 2 - Detailed (See Annex 3 - Estimated Project Description) Project Costs) Quarterly disbursement reports Ability to recruit and retain I Social insurance: 1 US$ 32.2 million for each component competent staff for the PCU, * improving policy design Inspectorate for private * developing system Pensions and the new SISA. administration * support for organizational The Wide Area Network change (WAN) is developed on * support for information schedule. technology Accounting system in place to 2 Private pensions: 2 US$0.4 million track standard service cost per * legislative development benefit. * system supervision * training in new methods Internal communication and * communication strategy coordination will be adequate to achieve training and 3 Social assistance: 3 US$2.9 million technical assistance objectives. * improving social assistance policy development and implementation * community based social services * Kandava pilot evaluation and dissemination 4 Monitoring and evaluation: 4 US$1.9 million * support the Kandava pilot data collection system and evaluate its suitability for other municipalities 5 Project coordination: 5 US$1.2 million * provide timely and accurate flows of essential information * provide support to working groups in: financial management: management of technical assistance and training; and procurement

26 22 Annex 2 Latvia Welfare Reform Project Detailed Project Descriptioni Project Component 1 - Social Insurance Development US$ million (total cost of component). This component of the project would help reform and modernize the Ministry of Welfare's Social Insurance Fund by supporting: (i) a flexible and efficient system for managing and administering social insurance; (ii) the capacity of the social insurance administration to manage the social insurance budget under the forthcoming reforms, including the development of core skills for system pers;onnel; (iii) a client orientation, by developing during project preparation a pilot in a branch office at Ziemelu to demonstrate new ways of providing service, evaluating the pilot, and expanding its positive elements to other offices during project implementation; and (iv) improved accountability and transparency through better reporting and handling of information, clearer lines of managerial and financial subordination, and improved control mechanisms SubcomponentA: Improving Policy Design This subcomponent supports the design and implementation of strategically oriented policy by the Social Insurance Department (SID), and to develop and implement the second tier legislation. It will assist the SID to present clearly and transparently the strategic goals of policy, to revise the policy strategy on an annual basis, to draft appropriate system regulations in a timely fashion, and to implement a Business Plan aimed at (a) clarifying the respective roles of the SID and the SSIF; and (b) improving the efficiency and effectiveness of the SID's work methods. tinder this subcomponent, the project will finance local and foreign technical assistance and training, civil works t) upgrade the offices, furniture and equipment. i. SID Business Plan: The subcomponent will assist the SID to implement a Business Plan to improve the organization and management of the Department. It will also assist the SID to implement a long term policy plan for developing and amending the current legislation. Technical assistance and training will be provided to improve the quality of policy analysis (such as through the development of tools to analyze policy scenarios and trends), and policy formulation and presentation to clients (primarily Government, Parliament, and public). Subcomponent B: Developing System Administration This subcomponent supports the implementation of the social insurance reforms by improving the recording of revenues and benefit administration. It will develop and introduce administrative procedures that are consistent with the new legislation, and revise the layout of the local social insurance offices in order to improve client service (based on the results of a pilot office in Ziemelu that was evaluated during project preparation). A business plan has been drafted by the SSIF. Under this subcomponent, the project will finance technical assistance, training and study tours, some equipment, and minor works. i. Contributions recording: This subcomponent will develop the capacity of the administration to record contribution related information in a timely, reliable, and accurate manner. It will support the eventual transfer to the Ministry of Finance of contribution collection functions.

27 23 ii. Benefit administration: The subcomponent will help ensure that benefits are administered properly and on time, within a more efficient administration that can manage multiple benefits and variable rates of contribution. Subcomponent C: Supporting Organizational Change The objective of this subcomponent is to support the ability of the SSIF to undertake its core functions. A business plan has been drafted by the SSIF, and a preliminary audit and accounting prestudy is to be revised. A communications strategy has been adopted and is soon to be under implementation. Under this subcomponent, the project will finance technical assistance, training and study tours. i. Project planning and management: This will assist the SSIF to employ new planning and management tools to improve the efficiency and effectiveness of its daily work. ii. iii. iv. Financial management: These activities will support the introduction of a new financial management system, including a new accounting system and an improved planning/audit cycle. Communication/PR: These activities will support more effective SSIF communication with clients, and the development of a coherent PR strategy. HR Development: These activities will support a more flexible and effective HR system, including recruitment, staff development/training, etc. v. Introduction of Object Oriented Information Base: These activities will support the introduction, testing and upkeep of an object oriented information base that will provide for the information needs of SSIF management through a system that requires minimal maintenance, while allowing the efficient transfer of information and data across regions concerning a wide range of benefit types. Through improved technology, it is expected that service to the client will improve through much faster and more effective processing. The new system will be designed in accordance with management requirements. The information model will consist of data 'objects' that correspond to a particular individual, piece of legislation, etc. The system will select the appropriate objects and algorithms for a particular calculation or processing event, and will also process system-wide data in order to provide information to managers. Subcomponent D: Supporting Information Technology The objective of this subcomponent is to support the introduction of hardware and software in order to meet the automation needs of the new system. A detailed draft automation plan has been completed. Under this subcomponent, the project will finance technical assistance, training and study tours, equipment, and minor works. i. Technology platform: This will assist the SSIF to introduce a new hardware platform, with networking within the local offices. ii. Telecommunications/networking: These activities will support the introduction of a new wide area network (WAN) through linkages with the Post Office system. iii. Information/programming: These activities will support more effective programming, including programming necessary for the introduction of the new information model Project Component 2 - Regulation of Privately Managed Pensions USS 0.44 million (total cost of component) This component supports the provision of safe and supervised options for the Latvian public to make personal provisions for retirement through privately-managed personal savings schemes.

28 24 Subcomponent A. Legislative Development This subcomponent will provide technical assistance and training to support the monitoring and evaluation of the private pension law and related legislation and regulation, and assist with the necessary policy development over the project period. Subcomponent B. System Supervision This subcomponent will assist the Inspectorate of Private Pensions to become an effective supervisory body. It will support the implementation of an automation plan prepared during project preparation that will allow the Inspectorate to function effectively. Technical assistance will be provided to assist in the organization and training of staff, the development of work instruments such as procedural manuals, and the introduction of appropriate accounting standards. Subcomponent C. Training in New Methods This subcomponent will support the implementation of a detailed training plan for those who will be involved in the management and supervision of the system (accountants, auditors, asset managers, fund managers, and others). Subcomponent D. Communication Strategy This subcomponent will provide technical assistance and training to support the implementation of a program to communicate to target groups the merits and advantages of privately-managed pension plans. This strategy should assist in overcoming the lack of confidence that is perceived to exist in the minds of potential participants in private pensions. Project Component 3 - Social Assistance Administration US$2.90 million (total cost of component) The objective of this component is to ensure that all needy receive effective and efficient social services. The project will improve the performance of the Ministry of Welfare Social Assistance Department in developing social assistance policy, and the municipalities in delivering local level services. A pilot social assistance office in Kandava region will seek to demonstrate best practice in service delivery, and disseminate lessons learned throughout Latvia. Under this subcomponent, the project will finance civil works, equipment, training, technical assistance, and furniture. SubcomponentA: Improving Social Assistance Policy Development and Implementation This subcomponent will support the development of an effective and efficient central level social assistance policy development and administration through improved legislation, a clearly defined organization, and stronger institutional capacity and performance of the Social Assistance Department. i. Supporting Legislative Development: This set of activities will provide technical assistance and training to the Social Assistance Department to implement the plan for legislative and regulatory development over the medium term outlined in the White Paper. This will include support to develop and introduce the new financing system for institutional care and simplified benefit system, training for the social assistance department on policy development, and technical assistance to develop safe and appropriate regulation of care. ii. Implementation of a Business Plan for Social Assistance: This subcomponent will support the implementation of the Business Plan, including the new SAD organization. Technical assistance and training will assist the MoW to develop a monitoring system for the Business Plan and a mechanism for annual review and adjustments, implement a program to privatize MoW social enterprises (such as those producing prostheses) and to monitor the quality and costs of institutional care, develop and introduce an effective adjudication system, and develop and introduce a training program that assesses gaps in social assistance skills and supports the dissemination of information nationally on best practice in delivery.

29 25 Subcomponent B: Community Based Social Services: Kandava Pilot The objective of this subcomponent is to encourage the development of a uniform cash benefit system to improve targeting of those in need, the efficient use of staff, and to develop a less expensive and more effective community social service alternative to institutions. Technical assistance, training, civil works, furniture and vehicles will be provided to develop a pilot office in Kandava (in western Latvia) which has been selected as a model region for the proposed municipal reform, and which covers a population of some 6,000. The municipality will demonstrate the advantages of consolidation of local administration across all sectors, including social assistance, and will show how social care can be responsive to local needs and result in fewer individuals being placed in institutional care. The activities under this subcomponent will support pilot design (including agreement on service provision and financing plan for investment and recurrent costs), development, facility renovation, staff training and implementation of a community social assistance center. The services to be delivered are: (i) a benefit information and payment department will provide a full range of information services to clients on available cash and non-cash assistance, including the selection and purchase of institutional care, and that will administer cash assistance. The client base of the department is expected to cover some 900 residents of the municipality, many of whom receive little or no service today. Four full-time social workers will provide the assistance. Of those residents who are already institutionalized, it is expected that this department may be able to facilitate the return of a small number to their homes. (ii) a home care program that will provide those in need with limited mobility (such as disabled and elderly) with access to trained social workers at home (five will be employed by the program). The client base of the program is expected to cover some 160 residents of Kandava, mostly the very old or very disabled living alone and with little or no support from their families. The social workers will primarily work to ensure that the clients receive adequate support to meet their nutritional, mnedical, and other needs. (iii) a program to support families at risk, which aims to provide a range of services such as support to battered women, and to children who may be at risk of institutionalization because of parental unemployment/family poverty, or alcohol abuse. The client base of this program covers some 46 families (124 children). One social worker and one family therapist will work with those fanilies (perhaps 30 percent) that may only require assistance with finding employment or other sources of income to overcome crisis. The other families will be provided with information on and support for options aimed at protecting, for example, the wife or children from further abuse. (iv) a day center for the elderly to provide elderly in the community with assistance on problems they face on issues such as health care and personal finances, and to improve their quality of life by providing social and recreational opportunities. A social worker and a nurse will be on site to provide assistance, which will include support to those who suffer from dementia such as Alzheimer's disease. Some elderly are expected to visit the center each week. The day center activities will be coordinated with the home help program. (v) a day center to provide rehabilitation and support to the handicapped and support and training to their home care-givers. This center will provide follow-up support for those who have received rehabilitative and occupational therapy in a residential center, in order to support independent living, a return to work (where possible) and avoid further institutionalization. The client base is expected to be a large fraction of the 270 handicapped individuals currently living in Kandava. Subcomponent C: Kandava Pilot Evaluation and Dissemination The subcomponent will support the monitoring and evaluation of the implementation of the Kandava pilot. Monitoring and evaluation will draw off of a baseline study of the demand for social services in the region and currently available service provision, and will include the following: (i) establishment of a case monitoring system and bi-annual site reports; (ii) annual conferences to review the pilot projects; (iii) a mid-term evaluation of service

30 26 provision; and (iv) a final evaluation including the plan for replication. These activities will be used to determine if the pilot projects are providing cost effective and replicable services to the region. The subcomponent will finance local and foreign technical assistance and staff training. Project Component 4 - Monitoring and Evaluation of Social Policy USS1.89 million (total cost of component) The objective of this component is to ensure that Government, Parliament and society are regularly informed about the impact of social policies by developing adequate organizational potential, and human and material resources in the MW to design, implement, monitor and evaluate social policy. The component will support the development, implementation, monitoring and evaluation of the MW social policy agenda. It will improve the allocation of human and financial resources within the Ministry through improved internal and external organization and communication. The component will strengthen the capacity to develop a workable social policy strategy. It will also strengthen local capacity to prepare high quality social policy analysis through training for the staff of the MW and for researchers in the field. Subcomponent A: Improving Local Monitoring This subcomponent will support the envisaged pilot social assistance office in Kandava to establish a more effective administrative data collection system and to replicate the lessons learned throughout Latvia. It will assist the municipalities with the monitoring of social assistance benefit administration by providing technical assistance, training and standardized procedures to deal with revenues, benefits and clients and by avoiding the wasting of resources. After implementation, the local authorities and the MW will be provided with the information needed to monitor social assistance policy and to evaluate its impact. i. Determination of data needs for monitoring and evaluation. the Strategic Department and the Social Report Group will determine their data needs with regard to clients, target groups and the Kandava population. ii. Standardized procedures: within the Kandava pilot, standardized procedures to manage information on budgets, benefits and clients will be developed, allowing the municipality to register its records in an easy and accessible way that can be used for policy monitoring and evaluation. This activity includes a baseline study of the current situation in Kandava, the development of new~ information procedures, a local data management system, and data communication management. iii. Evaluation and dissemination: after the first stage in the Kandava pilot, the implementation of the tools and the devices will gradually be extended towards others mtnicipalities in Latvia as warranted. Subcomponent B: Annual Social Report This subcomponent will assist the Ministry of Welfare to produce an annual report on social policy for Government, Parliament and the public (first one to be published in 1997). This subcomponent will take the form of an ongoing activity during the entire training and assistance period. It will create a working group that should start working on an Annual Social Report. At the end of the period this Report will review (a) the actual social situation in the country, using empirical information from in house analyses and academic research. (b) the policy actions undertaken during the reportperiod; (c) an evaluation of the policy measures taken; and (d) an action plan for future policy initiatives. Subcomponent C: DevelopingAnalytical Capacity This subcomponent will support the implementation of a training and technical assistance progran for users of the new household budget survey and other data related to social policy. Users will include the Departments within the Ministry, research institutes, universities, NGO's and local consultants.

31 27 i. Applied empirical policy analysis: This subcomponent encompasses training and updating skills and knowledge of MW Officials as well as outside specialists in the area of applied policy analysis. It will support exercises with large survey data sets, intermediate and advanced econometric analyses and panel data exercises. ii. Rapid appraisal: Training in rapid appraisal studies on new policy initiatives, in merging data and in advanced software applications will be supported. iii. Seminars: Seminars and workshops on designing evaluation programs, on modem developments in OECD Welfare States and on evaluating Social Security and Social Assistance Reform in Latvia will be supported. Subcomponent D: Social Policy Research Program This subcomponent will support competitively selected research proposals in applied social policy analysis. It will provide the MW with the resources to commission out social policy research to researchers and research institutes, using a process of competitive tendering, in order to assist the Ministry to improve the quality of policy monitoring and evaluation activities. The Social Policy Research Program will consist of a dedicated fund, financed out of donor financing and other resources. The procedures and bylaws will guarantee the policy focus of the projects, the independence of researchers as well as their academic integrity. Competitive tendering, peer review and academic discussion will improve the quality of the research. Project Component 5 - Project Coordination US$1.18 million (total cost of component) This component will ensure the effective provision of support and efficient management of the project in order to achieve the welfare reform project goals.

32 28 Annex 3 Latvia Welfare Reform Project Estimated Project Costs (US$ Million) % of Local Foreign Total Base Cost A. Social Insurance Development 1. Policy Design (Legislation) Administration Development Organizational Changes Informational Technology Development Subtotal B. Privately Managed Pensions C. Social Assistance 1. State Administration & Organization Pilot Project Kandava Kandava Pilot Evaluation & Dissemination Subtotal D. Monitoring & Evaluation of Social Policy 1. Improving Local Monitoring Annual Social Report Training Program (Improving Analytical Capacity) Social Policy Research Program Subtotal E. Project Coordination Unit Total Baseline Costs Physical Contingencies Price Contingencies Total Project Costs Annex3.xls

33 29 Annex 4 Latvia Welfare Reform Project Cost Effectiveness Analysis Summary A detailed economic analysis was only carried out for component 1 (83.4 percent of project costs). For other components, the direction of the fiscal impact was estimated. The assumptions and results are summarized in the table. The text below explains the analysis, including the key assumptions. Project Fiscal Impact (millions of 1997 Lats) Fiscal Impact: Component Cost Revenue Fiscal Impact: Expenditure' 1: Social insurance development : Regulation of privately-managed 0.2 +/0 0 pensions 3: Social assistance administration : Monitoring and evaluation of social 1.0 +/0 unclear policy _- 5: Project coordination Social Insurance Development Fiscal impact. This component will assist the State Social Insurance Fund (SSIF) to develop an administration that can implement the Government's Welfare reform. At present, the main elements of the reform (contribution related benefits based on individual accounts; increased savings in privately managed pension programs) cannot be implemented, as administrative systems cannot perform their responsibilities. The main fiscal gain, therefore, is from implementing the reform and the behavioral response. The non-project scenario, therefore, implies that key provisions of the new legislation are not implemented. On the revenue side, the projected 5-year increase of 72.4 million Lats into the social insurance budget comes from two behavioral assumptions. * Longer working life provides more system revenue: given information about the new benefit structure and the incentives to work longer, average retirement age will rise gradually. * Better information systems combined with new benefit formulas gradually increases contributions: Direct linkage of benefits to wages, including accurate information on contributions provided regularly to employees, will encourage voluntary compliance. Improved tracking of individual contributions under the automated system will allow faster detection of tax evasion. On the expenditure side, the projected 5-year savings of 38.4 million Lats comes from decreased benefits for early retirees under the new formula and postponed retirement. Other positive expenditure effects from the new legislation (e.g. price indexation instead of wage indexation) were not considered here, as new information and administration systems are not needed to realize these savings. The savings to other programs (e.g. social assistance programs, unemployment benefits, etc.) from better monitoring of individuals permitted by the nationwide computer network is also not included, although this will be an externality of the project. Also not considered is the revenue effect on other tax bases of lowered incentives to join the informal sector. If all of these effects are considered, the total fiscal impact after 10 years is expected Over 5 year peri')d, exclusive of -roject costs and depreciation of investme it

34 to be about 2.5 percent of GDP per year. (For further information, see Fox, Palmer, and Mclsaac, 1996, in the project file). IWhat is the most cost-effective approach to the design of the new system? The administrative reform will be primarily based on the principle of client orientation, in order to develop an administration that is significantly more effective, and flexible enough to adapt to future change in a cost-effective way. Client orientation is based on the following three principles, * All business is focused on serving the client * One inspector handles one benefit at a time * All necessary information must be available during the handling time 30 Introducing these principle will require that SSIF's business be reorganized, new systems installed, the premises reconstructed and the personnel trained. A successful reorientation should lead to an organization that can handle benefits in minutes instead of weeks, higher quality and accuracy, and a dramatically more effective organization. Currently, the SSIF employs about 1500 people, with a total administrative budget in 1996 of 6.2 million Lats, or 1.9 percent of revenues. During the design phase, three approaches to system improvement were considered for SISA as it implements the current reform. One alternative considered was marginal modifications to the current organization and administration. Two involved the introduction of a client-oriented system. These scenarios are summarized in the component working paper. The costs and benefits of each scenario, as well as the non-project scenario are summarized in the table below. The alternatives were judged for effectiveness in according to the following criteria: - It will permit the implementation all provisions of the new legislation It will improve cost-efficiency (lower administrative costs per standardized benefit) - Uniform handling of the laws and good client service will be provided - Effective payment systems will be in place - It will provide transparent accounting - Effective monitoring and evaluation will occur

35 31 Summary of Costs and Effectiveness of Each Scenario Invest -ment Admin. cost Cost Other Meets Scenario ($mil) Savings Impacts criteria Comments 1. Non-project scenario 1,0 None Can not administer No The administration will be modem welfare system; overwhelmed within a year by difficult to control fraud massive information system and abuse; clients' time failure. is wasted Option 1. Keep current 10,0 None Can not implement new No A system similar to other FSU organizational structure; pension system; difficult countries. No transparency; modest upgrade for IT to control fraud and information to clients is very system; no new abuse; clients' time is limited and thus behavioral accounting system wasted incentives do not work. Qption.2. Limited 20,8 Pay-off in Modem welfare system Partly Weak client service and investments in a client 6,5 years is implemented, leading information in non-networked oriented organization; (not to lower expenditures offices; weaker financial new IT system but including and higher growth; management; fewer efficiency without full national fiscal fraud and abuse control gains and weaker network impact) tightened in 70 percent sustainability. of the country Option 3. Full 24,7 Pay off in Modern welfare system Fully A modem system meeting all investments in a client 6,7 years is implemented, leading requirements of the reform. oriented organization; to lower expenditures Nationwide network allows new IT system with full and higher growth; better financial management national network and new fraud and abuse control and improve fraud control for financial management tightened in 100 percent all welfare system programs. system of the country (for full discussion, including financial analysis, see component working paper). Key assumptions in the analysis * timeframe: 10 years * interest costs: not included in pay-off cost. Net present value over 10 years is 0 with 6 percent real rate of interest for Options 2 and 3. Net present value is negative for Option 1. * depreciation (options 2 & 3): replacement /maintenance investments are projected at $1 million in year 5 and $2 million in year 7. * staff cost savings: none in non-project scenario and option 1 (more staff will be needed to manually process information); 550 jobs (net) are eliminated over the first three years of the project in option 2; 800 jobs (net) are eliminated over the first three years of the project in option 3. Layoffs are assumed to be costless. * externalities: economic effects such as the time savings of clients, or reductions in administrative costs for other programs are excluded from the financial analysis, as these are very difficult to quantify. Based on this analysis, Option 3 was chosen for this component. Option 1 does not yield any cost savings. It actually wastes money. Option 2 is viable and cost-effective. However, it would leave roughly 1/3 of the country out of the network. Processing time in these offices would be slow, wasting clients' time. Nationwide cross-checking of individuals to detect fraud would be much slower, and less rigorous. Local social assistance offices would not be able to tap into the

36 32 network to check income information during means-testing. More staff would be needed than for option 3, and as a result, the total savings is not much different from Option 3. Regulation of privately managed pensions Fiscal impact. Plans for the development of a private pensions system in Latvia call for the creation of funds that will be receiving up to 5 percent of covered payroll for all workers in Latvia. From the first year of operation of such funds, they may have as much as I percent of GDP under management. This is expected to grow and will reach an estimated 20 percent of GDP by the year This development will introduce very substantial institutional investors to the capital markets of Latvia. They will have an appetite for long-term investments, providing the vehicle for financing options that are not available to local entrepreneurs at present and providing a supply of local financing for large projects such as infrastructure requirements. The financing burden of the public sector will thus be sharply reduced. Latvia has previously established a State Insurance Supervision Inspectorate (SISI). Due to the similarity in the functions required to carry out the two types of supervision, it has been decided to assign the pension fund supervisory responsibilities to SISI. It is estimated that the incremental cost to the Ministry of Finance to establish the pension supervision will involve the addition of two staff-members to SISI for an annual cost not exceeding 30,000 Lats. The pension supervisory system would share the overhead and infrastructure costs of SISI, thus producing economies of scale for both parties. Cost recovery. Once the system is set up, it is envisaged that a licensing fee will be collected from all pension funds each year. Once the system reaches maturity, as early as five years from the start, the fees should be sufficient to completely defray the costs of supervision of pension plans incurred by the Ministry. Is the component the best and most cost-effective alternative? The alternative to establishing a regulatory and supervisory regime for privately-managed pension funds is not to do so. Individuals would have to depend on their own resources to monitor the accounting for their deposits and for the safe handling of the funds. The risk of financial loss to savers in such a system is high. It would not be advisable under these circumstances to channel any mandatory contributions into privately-managed accounts, so the pension reform would not be possible. In addition, the retirement income of participants could turn out to be less than promised. Another option would be to create a separate agency to carry out the supervision of private pensions. This would be a more costly altemative than the one proposed. The new agency would require a complete staff and infrastructure, certainly more than two persons. Such a choice would also magnify the difficulty of finding competent officials to staff the regulatory agency. Social assistance administration Fiscal Impact of Providing Alternatives to Institutional Care. A baseline survey in Kandava region (the pilot locality) in late 1996 is indicative of the gap in social service provision that will increase institutionalization in the absence of reform. In this time of rapid economic and social change, the survey found that a total of four poorly trained social workers are available to serve 124 children in families with very severe problems (battering, alcohol abuse, etc.), 240 residents over the age of 80 and 243 other elderly living alone, 252 disabled adults and 21 children, and 991 children in families facing less severe hardships but on the margin. Little is offered these groups other than direct cash assistance or referral to an institution -- only 8 residents receive home care. Successful implementation of a reform requiring municipal payment for institutional care for residents in state care will mean that municipalities will face the choice of (a) institutional bed-month costs that range from roughly US$180 per month to US$300 per individual (1996 figures) depending on type of care; or (b) diverting institutional referrals into altemative care that would roughly cost US$40 per individual served per month, based on preliminary figures from the Kandava pilot. While it is impossible to predict how high demand for institutional care will rise, a one percent increase in institutionalization in Latvia would lead to some 120 new placements.