70032 Albania Social Assistance Policy Note: Key Challenges and Opportunities June 2010 Human Development Sector Unit South East Europe Country Unit Europe and Central Asia Region 1 Acknowledgements This policy note was prepared by a World Bank team led by Melis U. Guven (task manager and author, ECSHD) as part of the Albania Social Protection Program Technical Assistance. Ramya Sundaram and Natsuko Kiso (ECSHD) provided data analysis and technical inputs. The team benefited from discussions and comments from Caterina Luggeri Laderchi (Senior Economist, ECSP3), and Lorena Kostallari (Senior Operations Officer, ECSHD). The report benefited from the background report prepared by Florian Tomini and Melis U. Guven. The note was prepared under the overall supervision of Kathy Lindert, Sector Manager, Social Protection (ECSHD). The note was processed by Katerina Timina (ECSHD). . 2 Contents I. Context ........................................................................................................................................................ 8 II. Overview of the Social Protection System in Albania .............................................................................. 10 III. Assessing the Performance of Ndihma Ekonomike ................................................................................ 15 Targeting Accuracy ................................................................................................................................... 15 Coverage .................................................................................................................................................... 16 Adequacy of Benefits ................................................................................................................................ 17 IV. Critical Design and Implementation Features for the NE Program......................................................... 18 Targeting Mechanisms: Room for Improvement ...................................................................................... 18 Administration and Monitoring Capacity and Fraud and Error Controls ................................................. 21 V. Conclusion with Policy Recommendations .............................................................................................. 23 Improving Coverage, Targeting and Adequacy of Ndihma Ekonomike ................................................... 23 Improving Administration of Benefits, Monitoring Capacity, and Fraud Controls .................................. 25 Identifying Clear Roles for Social Assistance and Social Insurance ......................................................... 27 ANNEX: ......................................................................................................................................................... 28 Registration Process ...................................................................................................................................... 28 Management of Registers/data bases ............................................................................................................ 31 REFERENCES ............................................................................................................................................... 32 LIST OF FIGURES Figure 1: Poverty Headcount by Region ......................................................................................................... 8 Figure 2: Extreme Poverty Headcount by Region ........................................................................................... 9 Figure 3: Poverty Headcount Increase due to Lower Growth ......................................................................... 9 Figure 4: Social Protection System in Albania .............................................................................................. 11 Figure 5: Total Spending on Social Protection .............................................................................................. 12 Figure 6: Spending on Social Assistance in the ECA Region ....................................................................... 14 Figure 7: Social Assistance Spending (in % of GDP) ................................................................................... 15 Figure 8: Targeting Accuracy of Ndihma Ekonomike .................................................................................. 16 Figure 9: Targeting Accuracy of Specific Poverty Focused Benefits ........................................................... 16 Figure 10: Total NE Spending and Average Monthly NE Benefit ................................................................ 17 Figure 11: NE Targeting at Two Levels ........................................................................................................ 20 Figure 12: The Effect of Filters in Excluding the Urban Poor (bottom 10 percent) in Ndihma Ekonomike 21 LIST OF TABLES Table 1: Coverage of Social Protection Benefits ........................................................................................... 13 3 Executive Summary Reducing poverty continues to be one of the main priorities of the Government of Albania. Albania experienced economic growth averaging about 5.5% during 2000-2008, resulting in a drastic reduction in poverty rates nationwide. Due to increasing output growth, the poverty rate fell from 25.4% in 2002 to 12.4% in 2008, and all areas of the country experienced a decline in poverty, though at varying rates. While there is consensus that Albania has been spared most of the hardship that the global crisis inflicted on many of the economies in the region, the slowdown in growth, the losses in terms of employment and a decline in remittances from abroad are likely to have affected the pace of poverty reduction. Reducing poverty further is a priority for the Albanian Government, as is the related objective of improving the effectiveness of social assistance systems. Objectives. This policy note summarizes the findings of on-going analysis and technical assistance discussions with the Government of Albania over the past two years. 1 It seeks to identify ways in which the Government could strengthen the equity, efficiency and transparency of social assistance programs, with a focus on the main poverty-targeted program, Ndihma Ekonomike, NE, so as to improve effectiveness for poverty reduction. A companion paper details diagnostics and recommendations for disability benefits. Broader Context of Social Protection System. Currently, Albania has a reasonable platform to provide social protection to its citizens through social insurance, social assistance and employment programs. However, the lack of clarity in roles and incentive compatibility between “social insurance� (contributory) schemes and “social assistance� benefits – and a lack of effective linkages across these pillars (e.g., between social assistance and employment programs) reduces the effectiveness of social policy. At the same time, low participation in the social insurance system along with demographic shifts are likely to alter the relative importance of social insurance vs social assistance schemes in the future. Specifically, while coverage of the elderly population under social insurance programs is currently high, it can be expected to decrease in the future since a low share of the working age population is currently contributing to the pension system, meaning that a major share of the elderly will not have access to a pension income in the future. 1 The other activities under this Technical Assistance Program include (i) an overview report on the social protection system in Albania, (ii) two workshops which were held in Tirana on the priority areas identified in identification discussions and during a workshop with government stakeholders; and (iii) a functional review of the administration processes of Ndihma Ekonomike. The overview report was intended for background purposes in preparation for the Technical Assistance Program. The workshops-attended by government official from headquarters and regional offices- covered the following three key themes: (i) better targeting of cash benefits, (ii) implementation matters (good housekeeping and more efficient administration); and (iii) making social protection pro-employment The functional review provides a clear description of the architecture of NE Program through assessing the operational aspects of NE and includes a detailed diagnostics on information flows, financial flows, functional responsibilities and institutional. 4 As such, the role of non-contributory social assistance schemes is likely to increase in importance in the future. This approaching shift further emphasizes the importance of improving the effectiveness of social assistance programs, along with improving the pension system to encourage participation and improve fiscal sustainability. Social Assistance Overview. At 1.6% of GDP, total spending on social assistance in Albania is comparable to the regional average for Europe and Central Asia (ECA). Two main social assistance schemes account for the majority of this spending: the NE program, which is the main poverty targeted social assistance benefit, and disability assistance benefits. The relative balance in spending on these two programs has shifted significantly over time, however. Outlays on disability benefits increased from 0.4% of GDP in 2000 to 0.9% of GDP in 2008, arising both from an increase in certified beneficiaries and from an increase in the size of the benefits. An initial diagnostic suggests that disability legislation is scattered, medical eligibility criteria are outdated, and the system for assessing and certifying disability cases is deficient and open to conflicts of interest.2 Moreover, this ballooning of expenditures on the (more generous) disability assistance benefits has been accompanied by a contraction in spending on the poverty-targeted NE program, which has fallen from 0.8% of GDP in 2000 to 0.3% in 2008. Mixed Performance Outcomes for the NE Program. In terms of outcomes, performance of the NE program is mixed:  Targeting accuracy is respectable in terms of relatively low leakages to the non-poor. Over half (56%) of benefits paid out are received by those in the poorest quintile, and 82% of benefits go to those in the poorest two quintiles. In terms of leakages, the share of benefits captured by the wealthiest quintile is 6%. The performance of NE is about average for targeting accuracy in the ECA region: it performs better than programs in Latvia and Russia (for which about 30% of benefits go to those in the poorest quintile) but not as well as guaranteed minimum income programs in Kosovo, Serbia, Bulgaria, Ukraine and Romania (transferring 70-80% of benefits to the poorest quintile).  However, coverage of the poor is low, due fiscal constraints (exacerbated by ballooning expenditures on disability benefits) and due to a high degree of errors of exclusion. Specifically, NE only reached 22% of the poor in 2008, which presents a decline from a coverage of 36% in 2005. This reduction in coverage of the poor occurred due to the rigorous application of a system of “layered filters� for eligibility. Our analysis shows that these “layered filters� introduce distortions in the eligibility process and results in severe errors of exclusion, as discussed further below.  Furthermore, benefit adequacy (generosity) is quite low. Benefits of the NE program account for just 14% of total post-transfer consumption of the poorest beneficiaries. The level of benefits has eroded over the period from 2000-2007, with inadequate indexation of benefits to price inflation and the contraction in overall spending on the program (squeezed out by an 2 A more detailed assessment of disability benefits is currently underway as a companion to this paper, which focuses primarily on the NE program. 5 increase in spending on the relatively more generous disability benefits).Average disability benefit per person was three times the NE household benefit in 2008.  As a result of low coverage and benefit adequacy, impact of the NE program on poverty is limited. Overall, the NE program only brings 0.4% of the pre-transfer poor out of poverty. Design and Implementation Features Contributing to Mixed Outcomes. In addition to increasing budget constraints (exacerbated by the expansion of outlays on disability benefits), there is a number of design and implementation features in Albania’s NE program that contribute to these mixed outcomes:  Two-Tiered Targeting: Geographic and Individual Assessment. Albania targets spending under the NE program at two levels: (a) geographic allocation of block grants to local governments; and (b) screening of applicant families by local governments on the basis of a set of eligibility filters. While a combination of geographic and individual assessment methods is often recommended for effective targeting, both sets of criteria could be improved in Albania for the purposes of improved equity, efficiency and transparency: o Geographic targeting allocations (block grants) are currently based on a combination of estimates of poverty (from combined Census 2001 and 2008 LSMS data) and historical numbers of beneficiaries for each municipality. The exact criteria are not clear however and appear somewhat ad hoc. Equity and transparency could be improved based on the use of objective poverty-based criteria from an updated poverty map. o Individual eligibility screening is carried out by municipalities using “multi-layered� criteria that differ for urban and rural localities. In principle, these criteria form a rudimentary “means test.� In practice, however, there is no objective “weighting� across numerous eligibility filters that are applied in an ad hoc way and result in excluding otherwise poor households if they fail to meet any one of these filters. Examples of these binary filters include: possession of certain assets, participation in economic activities, etc. These “multi-layered� filters introduce distortions into the eligibility process both within and across urban and rural areas, result in large errors of exclusion of the poor, and could potentially generate adverse incentives. Equity, efficiency and transparency could be improved by shifting to a consistent and unified scoring formula for eligibility on the basis of explicit weights derived from objective correlations with poverty status.  Inadequate indexation of benefit levels and eligibility thresholds, as well as inconsistency with other social assistance benefits (such as the relatively more generous disability benefits). These features have contributed to eroding the value of NE benefits and a decline in coverage (and in relevance of the program).  Benefits administration is also hampered by weak capacity, information systems, and oversight and controls. Albania currently lacks an automated national registry of applicants and beneficiaries and benefit management is largely paper-based at the local level. This lack 6 of an effective management information system leads to (a) inefficiencies in applying for and awarding benefits (with high transactions costs); (b) weaknesses in oversight and controls of fraud and errors; and (c) ineffective monitoring and evaluation of social policy. Options for Reform: Improving Equity, Efficiency and Transparency Building on the existing foundation of social protection – and particularly the NE social assistance benefit, there are a number of measures that the Government of Albania could undertake to improve the equity, efficiency and transparency of benefits:  Basing geographic allocations on objective poverty criteria. Transparency and equity could be improved through a standardization of the geographic allocation of block grants across municipalities on the basis of an updated poverty map.  Strengthening and standardizing household screening criteria using a continuous scoring formula. We recommend that the Government of Albania consider standardizing and improving eligibility criteria by: (a) eliminating the use of one-off binary filters; and (b) strengthening individual assessment criteria using a single continuous scoring formula that incorporates objective weights and variables that are empirically associated with poverty status (hybrid means testing), which could be adapted for urban (non-farming households) and rural (farming households) areas. Such a move could reduce errors of inclusion, reduce errors of exclusion (poor families denied benefits due to application of ad hoc filters), and improve transparency in the system.  Introducing indexation of benefits and eligibility thresholds to price inflation.  Strengthening benefits administration by: (a) automating information systems and creating a national registry of all applicants and beneficiaries; (b) simplifying and standardizing intake and application procedures; (c) improving monitoring, oversight and controls to reduce fraud and errors; (d) possibly shifting eligibility decisions to the central government. These improvements would not only improve effectiveness (equity, efficiency) but would also boost transparency and thereby serve a political role, helping the government be more accountable to the public. Moreover, improving the links between the NE and other programs, both on the administration and policy side, would further contribute to the intended objectives of efficiency and effectiveness. Integrating social assistance and employment services, or strengthening the bridges between such services (for example, linking safety nets beneficiaries to activation services), could improve effectiveness. In addition, clarifying the roles of the social insurance and social assistance programs, and ensuring that there is no overlap or blurring of responsibilities and functions would strengthen the consistency of those programs. This elucidation would also allow for a reduction of pension system deficits by making benefits provided to individuals consistent with the contributions received. Resulting savings from the reduced deficit would create fiscal space in the budget that could be used to improve and expand social assistance programs. 7 I. Context Since the year 2000, Albania has experienced a drastic reduction in poverty rates nationwide as a result of economic growth.3 In fact, since its transition from a centrally planned to a market economy, major structural and economic reforms have helped Albania achieve, on average, a growth rate of 5.5% (during 2000-2008) and have reduced poverty from 25.4% in 2002 to 12.4% in 2008, effectively pulling over 410,000 people out of poverty. At the same time, the rate of extreme poverty fell sharply over the period from 2002-2008, from 4.7% to 1.2%, and the number of people living in extreme poverty dropped from 146,000 to 36,000. Economic growth contributed to poverty reductions in all regions of Albania, although different areas of the country saw more improvement than others. The most significant improvements were realized in the capital, Tirana, where the poverty rate dropped by 55%, followed by the Mountain areas (43%), coastal areas (21%), and central areas (17%). Both urban and rural poverty rates halved between 2002 and 2008, falling from 19.5% to 10.1% (urban) and 29.6% to 14.6% (rural); nevertheless, 60% of all Albania’s poor still reside in rural areas countrywide. After falling steadily from 2002 to 2005, poverty rates continued to decline in coastal and central areas after 2005, but went up marginally in the Mountain areas and Tirana. 4 Figure 1: Poverty Headcount by Region 50 45 40 35 30 25 20 15 10 5 0 Tirana Coastal Central Mountain Total Urban Total Rural Total 2002 2005 2008 Source: LSMS data Despite the marked improvement in the extreme poverty rate throughout Albania over the period of 2002-2008, extreme poverty remained relatively high in the mountain areas (Figure 2). Extreme poverty – defined as the state in which one’s overall consumption is lower than what 3 Based on Albania LSMS data 4 According to Albania Poverty Assessment (World Bank, 2007) migration explains, in part, the observed changes in poverty rates in Albania. First, there is clear evidence that Tirana and the Mountain rural regions are the areas where the largest increases in the share of households receiving remittances are observed. Second, not only did the fraction of households receiving remittances increase, but the amount of remittances also increased substantially. Specifically, the flows to Tirana more than doubled while those to the Mountain region increased in the order of 50%. Third, the Mountain area has witnessed the strongest pace of poverty reduction in no small part because it is the only region whose outflow of new permanent international migrants continues to grow. 8 is required to meet nutritional needs – fell from 5% to 1.2% during 2002-08 in all of Albania, but only decreased to 3.8% in the Mountain areas.5 The majority of extremely poor people live in the central and coastal regions, and the Mountain areas – which are relatively sparsely populated – account for only 18% of people living in extreme poverty. Figure 2: Extreme Poverty Headcount by Region 14 12 10 8 6 4 2 0 Tirana Coast Coast rural Central Central Mountain Mountain Total urban urban rural urban rural 2002 2005 2008 Source: LSMS data The pace of the poverty reduction is likely to have been affected in the aftermath of the crisis. Albania was one of the few countries in Europe that maintained positive growth with a GDP growth rate of 2.2% in 2009. While there is consensus that Albania has been spared most of the hardship that the global crisis inflicted on many of the economies in the region, the slowdown in growth, the losses in terms of employment and a decline in remittances from abroad are likely to have affected the pace of poverty reduction. The results of simulations on the effects of lower growth on the rate of poverty reduction shows that by 2011 when the economy is expected to go back to pre-crisis growth levels poverty could be 3.4 percentage points higher than it would have been in the absence of the crisis.67 (Figure 3) Figure 3: Poverty Headcount Increase due to Lower Growth8 Source: World Bank Staff Estimates 6 Other simulations suggest that the loss in remittances alone could account for between 1 and 2.6 percentage points of higher poverty incidence in 2011, while the losses in terms of employment 8 The Impact of the Crisis on the Poor and Vulnerable Goups in Albania: Some Preliminary Findings, WB Programmatic Poverty Assessment for Albania(2010) 9 Reducing poverty continues to be one of the main priorities of the Government of Albania. Currently, Albania has an ample platform to provide social protection to its citizens through social insurance, social assistance and employment programs. However, these programs are not efficiently linked to each other, which can lead to unclear and occasionally overlapping roles among the programs. Among these social protection programs, Ndihma Ekonomike (NE) is the only poverty-targeted social benefits program available to the Albanian Government, but despite the continued efforts of the Government to improve the program’s impact,, the poverty impact of NE remains low. Against this backdrop, the objective of this note is to assist the policy makers in Albania with identifying reforms that would improve the overall effectiveness and targeting of Ndihma Ekonomike. Specifically, the note will present key challenges faced by the Government, as well as outline the reforms that would contribute to a more efficient and effective social assistance system in Albania, resulting in a process with improved transparency and accountability. The note is organized as follows: Section 2 presents a brief overview of the social protection system in Albania while at the same time discussing inefficiencies that arise due to lack of linkages between these programs; Section 3 assesses the current effectiveness of Ndihma Ekonomike; and Section 4 concludes with recommendations. II. Overview of the Social Protection System in Albania Albania has an ample social protection platform, but with lower overall spending than ECA and EU countries. Currently, social protection is provided through three main programs, which consist of social insurance, social assistance, and employment assistance (Figure 4). Overall spending on social protection in Albania is somewhat low. In 2008, Albania spent only 7% of GDP on social protection, much lower than the EU countries.9(Figure 4), or for the average for Eastern Europe and Central Asia (ECA) region (about 10%). One of the main reasons for this relatively low spending is that the level of pension benefits (which provide the highest coverage) is low, despite the fact that the pension system has to finance several generations of the elderly, most of whom have full pension rights. The Albanian pension system avoided a pension system fiscal crisis (despite the low share of working-age population participating in the current system) by compressing benefits; consequently, the average pension in the urban sector amounts to 1.1 times the minimum pension.10 Effectiveness of social policy and benefits is compromised by the disconnect among the institutions that administer different parts of the programs. Effective connections between the 9 Of the total spending on social protection, 70.8% of was on pensions; 23.3% was on social assistance; 2.3% was on employment programs and the remaining 3.5% was on short term benefits. 10 The main reason for such a compressed benefit structure is the fairly stringent limit on the maximum pension. Pensions cannot exceed twice the minimum pension or 75 percent of the net average wage of any 3 successive years in the last 10 years of employment, whichever is less. This suggests that participants who earn above the average wage will reach the limit very quickly. The pension system includes significant aspects of redistribution towards lower income workers and therefore provides a relatively low rate of return for high income individuals, creating weak incentives to participate and/or declare full earnings. 10 existing social protection systems are weak at both the policy level and the operational level, and this leads to missed opportunities. For example, Social Insurance is administered by the Social Insurance Institute (with urban contribution collections taken by the Tax Administration, along with other social contributions, including unemployment insurance). The Tax Administration transfers the pension and unemployment contributions to the SII, which uses the funds to finance pension benefits. Meanwhile, the employment agency receives separate transfers, disconnected from unemployment insurance contributions, to finance unemployment benefits and programs. Finally, local governments administer social assistance, with funding provided by the Ministry of Labor. All three institutions have local offices throughout the country. The current process of verifying the status of the applicants for NE with the Social Insurance Institute is paper based. Moreover, the links between the NE program and employment offices is limited to verifying the status of the applicants for NE beneficiaries, which is also a paper-based process. Clearly, the lack of automated databases that would improve the communication of these very inter-related services results in unnecessary slowdowns. At the same time lack of linkages between employment services and NE program potentially leads to less effective administration of benefits, as well as missed opportunities for coherency in social policy. Figure 4: Social Protection System in Albania Social Protection System Social Assistance Employment Programs Social Insurance Cash Benefits Social Care Services Labor Market Programs Pensions (Urban/Rural) Ndihma Ekonomike Maternity Benefits Employment Promotion Monthly allowance for Program Sickness Benefits disabled Professional Training Price compensation Work accidents and illness Unemployment Unemployment Benefits Insurance 11 Figure 5: Total Spending on Social Protection 20 18 16 14 12 10 8 6 4 2 0 a a a ia a nd ia 27 25 15 R. ia ia y ni tvi ni i ak an ar en an ar la to h ba EU EU EU La ng lg ov hu ec m Po ov Es Al Bu Ro Hu Sl Cz Lit Sl Social Protection Spending % of GDP in 2006 Source: EUROSTAT, SII and MoLSA Low participation in the social insurance system along with demographic shifts are likely to alter the relative importance of social insurance vs social assistance schemes in the future. The social protection system provides coverage to more than half of the population. The three main programs cover 53% of the total population (Table 1). About 50% of the urban population and 56% of the rural population live in a household that receives some sort of social protection benefit. Social insurance (urban and rural pensions) covers 46% of the total population while social assistance programs cover almost 10% percent11. Ndihma Ekonomike (NE), the main poverty-targeted social assistance program, covers slightly more than 7% of the population. Currently, a significant percentage of the elderly are covered under the social insurance system. According to the Albania Social Insurance Review, about 40% of the elderly are projected to have no access to retirement income in the future, given the low number of working age individuals making active contributions to the pensions system at this time. Therefore, social assistance programs can be expected to play a larger role in the future because the current low coverage of the pension system (contributors) will leave a large proportion of the elderly without access to pension benefits. This approaching shift further emphasizes the importance of improving the effectiveness of social assistance programs, along with improving the pension system to encourage participation. 11 Social insurance coverage does not include the current contributors of the pension scheme. 12 Table 1: Coverage of Social Protection Benefits Share of Share of Share of Program Total Urban Rural Population Population Population Social Protection 52.8 49.5 56.1 Social Insurance 45.9 43.4 48.4 (Pensions) Social Assistance 9.7 7.2 12.1 Ndimhe Ekonomike 7.3 5.8 8.8 Source: LSMS 2008 One of the main reasons weaknesses in the pension system is that the roles of social insurance and social assistance programs are not clear. Over time, the social insurance and social assistance branches of the system have evolved in such a way that rural pensions have essentially become a social assistance benefit (since contributions are largely paid by the Government). The rate of return on contributions by rural pensioners is actually much higher than that of urban pensioners, as their benefits are relatively higher than what they pay to the pension system. Benefits received do not necessarily reflect the income level of individuals, and this weak link between contributions and benefits makes the average rural pension closer to a social assistance benefit. Clarifying these roles would help both systems function more efficiently, enabling the pension system to focus mainly on the social insurance role, while the social assistance programs focus on those citizens who cannot afford to participate in the social insurance system. This would also reduce the deficits of the pension system by improving the link between contributions paid and benefits received. This, in turn, would free up resources from financing the pension system so they could be put to use in poverty-targeted social assistance programs that do a better job in identifying the poor. For example, analysis shows that it costs almost twice as much to reduce the poverty gap by one Albanian LEK through a rural pension scheme as it does through the NE. From 2008 LSMS data, we compute that five LEKs have to be spent on rural pensions to reduce the poverty gap by one LEK. On the other hand, to reduce the poverty gap by one LEK, three LEKs have to be spent through the NE. This amount could be expected to go down as NE becomes more efficient. In fact, total spending on social assistance alone compares well with the average spending in the ECA region. Albania spent 1.6% of GDP on social assistance in 2008, which is similar to the ECA average of 1.7% (Figure 6). 13 Figure 6: Spending on Social Assistance in the ECA Region Hungary 06 Croatia 08 Bosnia 08 OECD 05 Belarus 08 Uzbekistan 07 Ukraine 08 Serbia 08 Russia 06 Albania 08 Romania 07 Estonia 06 Lithuania 08 Kosovo 08 Poland 07 Moldova 08 Kazakhstan 07 Bulgaria 08 Kyrgyzstan 08 Georgia 07 Montenegro 08 Armenia 08 Macedonia 08 Latvia 08 Azerbaijan 08 Turkey 07 Tajikistan 08 0.0% 0.5% 1.0% 1.5% 2.0% 2.5% 3.0% 3.5% 4.0% Source: Europe Central Asia Region Social Protection Database Two main social assistance schemes account for the majority of social assistance spending. The NE program, which is the main poverty targeted social assistance benefit, and disability assistance benefits. The relative balance in spending on these two programs has shifted significantly over time, however (Figure 7). Outlays on disability benefits more than doubled as a share of GDP over the period from 2000-2008, arising both from an increase in certified beneficiaries and from an increase in the size of the benefits. Moreover, this ballooning of expenditures on the (more generous) disability assistance benefits has been accompanied by a contraction in spending on the poverty-targeted NE program, which has fallen from 0.8% of GDP in 2000 to 0.3% in 2008. Spending on disability programs, which accounts for more than twice the spending on NE, should be re-evaluated. During 2000-2008, both the number of disability beneficiaries and the level of average disability benefits increased significantly, leading to a total disability spending of 0.9% of GDP12 (Figure 7). Spending on disability benefits accounted for 56% of total spending on social assistance programs in 2008. An initial diagnostic of the disability program shows that disability legislation is scattered, medical eligibility criteria are outdated, and the system in place 12 Number of beneficiaries increased from 44.4 thousand in 2000 to 73 thousand in 2008, an increase of 64.4%. Over the same period, the average disability benefit increased from 4060 LEK to 8700 LEK, an increase of 114 %. 14 for assessing disability cases is deficient and open to conflicts of interests (as oftentimes doctors assessing disability cases are also on the reviewing commission).13 Figure 7: Social Assistance Spending (in % of GDP) 2.50% 2.00% 1.50% 1.00% 0.50% 0.00% 2000 2001 2002 2003 2004 2005 2006 2007 2008 Ndimhe Ekonomike (Economic Asssitance) Disability Allowance Care Allowances Other Source: MoLSA III. Assessing the Performance of Ndihma Ekonomike The NE program is Albania’s main poverty-targeted social assistance scheme. It is financed by the central government, which channels block grants through municipal governments who then carry out implementation on all aspects: applicant intake, eligibility determination, verification, and payments (see Annex). As discussed above, spending on the program has contracted since 2000, and currently represents 0.3 % of GDP. In terms of outcomes, performance of the NE program is mixed. While the program performs fairly well in terms of targeting accuracy (low leakages to the non-poor), coverage is low due to a system of filters that causes serious errors of exclusion, and the adequacy (generosity) of the value of benefits has eroded over time. As a result, the impact of the NE program in reducing poverty is not significant. Overall, the NE program lifts out of poverty only 0.4% of the pre-transfer poor. Targeting Accuracy The NE program produces respectable outcomes in terms of targeting accuracy (leakages to wealthier families). The poorest 20% of the beneficiaries receive 56% of all NE transfers. (Figure 8) A large share of the total benefits (82%) is captured by the poorest 40% of the population. Overall, the targeting accuracy of NE improved from 2005 to 2008, with a higher share of the benefits going to poorer households. The share of the benefits captured by the wealthiest households (fifth quintile) is at 6%. Comparison of NE with similar poverty-focused programs in other countries shows that NE is about average for targeting accuracy in the ECA region (Figure 13 A more detailed assessment of disability benefits is currently underway as a companion to this paper, which focuses primarily on the NE program. 15 9). It performs better than programs in Latvia and Russia (for which about 30% of benefits go to those in the poorest quintile) but not as well as guaranteed minimum income programs in Kosovo, Serbia, Bulgaria, Ukraine and Romania (transferring 70-80% of benefits to the poorest quintile). Figure 8: Targeting Accuracy of Ndihma Ekonomike 60.0 50.0 40.0 30.0 20.0 10.0 0.0 Q1 Q2 Q3 Q4 Q1 2008 2005 Source: LSMS data Figure 9: Targeting Accuracy of Specific Poverty Focused Benefits 90 Share of Benefits Received by Poorest Quintile 80 70 60 50 40 30 20 10 0 Lithuania S. Benefit Armenia FB Prog Romania GMI Poland SW benefits Estonia MT Benefits Croatia S. Allowance Uzbekistan SA for low income Bosnia-Herzegovina CSW Bulgaria GMI Macedonia SFA Hungary Regular SA Kyrgyzstan UMB Serbia MOP Latvia GMI + dwelling Ukraine XP program Russia CA Serbia CA Kazakhstan TSA Georgia TSA Azerbaijan TSA Kosovo SA Montenegro FMS/MOP Albania NE Source: Europe Central Asia Region Social Protection Database Coverage Coverage of the NE program is low due to high errors of exclusion. The number of households benefiting from NE has declined over the past years, mainly because of MOLSA’s efforts to tighten the process for determining eligibility. The share of the population covered by Ndihma Ekonomike was 15% in 2005 and went down to 7% of the population in 2008due to MOLSA’s removal of beneficiary households that failed to meet all criteria under the official eligibility rules (one-off filters, as discussed below). For example, the NE beneficiary households that also receive urban old-age and survivor pensions accounted for more about 18% of NE beneficiaries in 2005, 16 but had declined to 7.4% of the total NE beneficiary households by 2008. However, along with the declining number of beneficiaries, the percentage of poor covered by the NE program also fell. As a result, NE reached only 22% of the poor in 2008 as opposed to 36% of the poor in 2005. At the same time, coverage of the non-poor by NE’s program was at 5.2 %. This reduction in coverage of the poor occurred due to the rigorous application of a system of “layered filters� for eligibility. Our analysis suggests that these “layered filters� introduce distortions in the eligibility process and results in severe errors of exclusion, as discussed further below. Adequacy of Benefits Furthermore, benefit levels of the Ndihma Ekonomike Program are quite low. Benefits from NE account for a modest adequacy, and are responsible for just 14% of the total post-transfer consumption of the poorest beneficiaries. Moreover, there are significant differences in the adequacy of the NE benefit across the regions. In Tirana, the NE benefit accounts for 37% of the post-transfer consumption of households, while it represents only 9% of the post-transfer consumption of the poor in the mountain areas. This is despite the fact that mean consumption in the mountain areas is significantly lower than in other areas. At least two factors cause this low (and declining) level of benefits. First, benefit levels have not been indexed properly to price inflation, and have thus seen erosion in real values. Second, the decline in benefit levels has been accompanied by the overall contraction in NE spending (squeezed in part by the expansion in disability benefits). Average disability benefit per person was three times the NE household benefit in 2008. Spending on NE followed a declining trend until 2008. (Figure 11) During 2002- 2007, total NE spending and average NE benefit went down by 37% and 22% respectively. In the same period, the number of NE beneficiary households declined by 19%, which was less than the decline in total NE spending, leading to an erosion of the NE benefit. In 2008, more funds were allocated for NE and the declining trend reversed. Despite the improvement, NE’s spending in 2008 is only comparable to its spending in 2002 when the decline in the number of beneficiaries is taken into consideration. Figure 10: Total NE Spending and Average Monthly NE Benefit NE Spending (million ALL ) 5,000 3,500 4,500 3,000 4,000 3,500 2,500 3,000 2,000 2,500 2,000 1,500 1,500 1,000 1,000 500 500 - - 2000 2001 2002 2003 2004 2005 2006 2007 2008 Total NE Spending in 2000 prices Total NE Spending NE Average Benefit in 2000 prices NE Average Benefit 17 Source: MOLSA IV. Critical Design and Implementation Features for the NE Program Design and implementation features contribute to mixed outcomes. In addition to increasing budget constraints (exacerbated by the expansion of outlays on disability benefits), there are a number of design and implementation features in Albania’s NE program that contribute to these mixed outcomes, including flaws in targeting mechanisms (design features) and weaknesses in benefits administration (implementation aspects). Reforming and strengthening these design and implementation features would contribute to improving the equity, efficiency and transparency of the NE program. Targeting Mechanisms: Room for Improvement The targeting mechanism currently adopted by the NE has two levels. The first level is the (geographical targeting mechanism used in the allocation of central budget funds to local governments. The second level is the targeting mechanism adopted by the local governments to determine eligible households on the basis of a series of filters (Figure 12). The methodology for geographically allocating block grants to the local governments could be formalized for improved transparency and targeting accuracy. Until 2010, MOLSA allocated block grants, taking into account some demographic and socio-economic characteristics, but did not consider any direct poverty indicators and, consequently, did not account for regional differences in poverty rates. In 2010 MOLSA adopted a new methodology,14 which is currently based on a combination of estimates of poverty (from combined Census 2001 and 2008 LSMS data) and historical numbers of beneficiaries for each municipality. The exact criteria are not clear however and appear somewhat ad hoc. Based on the data from MOLSA, it is not possible to identify the criterion followed in combining those two numbers, confirming the impression that the process relies on the experience and judgment of MOLSA officials. By breaking the link between targeting and the poverty status of a given commune, and by doing so in what appears to be a rather ad hoc way, however, this second stage reduces the intended impact of the geographical targeting of NE. Equity and transparency could be improved based on the use of objective poverty-based criteria from an updated poverty map. Eligibility at the household level is determined at the local level, based on centrally-designed criteria and a final decision by local councils. Eligibility at the household level is administered by local governments using a decentralized institutional structure.15 NE relies on an “on-demand� registration process where households are required to come to social assistance offices to apply for benefits. The social administrators under the local governments are the first point of contact for the applicants. They 14 Reports show that the regional poverty (as measured from LSMS and Census of 2001) is not always consistent with the allocation of block grants to local governments. Consequently, local government units that have a high number of poor households do not receive an adequate level of block grants (Social Protection Sector Strategy, 2007). 15 See Annex for a detailed description of the registration process and eligibility criteria. 18 are responsible for collecting applications, verifying information, and assessing eligibility according to centrally-determined criteria (as discussed below). Social administrators then prepare the draft decision for approval by the Council of the commune/municipality. Based on Council discussions, the draft decision can be changed and the decision by the council on approval or rejection of benefits is final. While involvement of local councils in eligibility decision can be viewed as optimizing “local� community information for special circumstances,16 this discretionary aspect is not transparent in relation to other eligibility criteria (no use of transparent “weighting� across criteria, for example) – and in practice, anecdotal evidence suggests that political considerations can enter into these Council decisions. Official eligibility criteria are determined centrally, and include a series of “one-off� filters. According to the official rules, NE is dispensed on the basis of beneficiary identification rules and administrative arrangements identified at the central level. Officially, eligibility for NE is determined on the basis of a rudimentary means test that relies on a series of qualifying “filters.� These criteria differ between urban and rural (farming) areas: 17  For urban households: Eligibility for NE is determined on the basis of a rudimentary means test that relies on a series of filters. The means test operates based on a system where the social administrators apply a series of one-off “binary� filters to screen eligibility. Specifically, a household will be denied benefits if any one of the following filters applies: Households that have a member who (a) owns any form of capital; (b) is economically active (disabled are exempted); (c)resides outside of the country; (d) is not registered as unemployed and looking for work (farmers and disabled are exempted); (e) refuses to work when a job is offered by the employment offices; (f) refuses to take the land given; (g) does not collect the benefit within six working days (except in special cases as defined in the legislation), or (h) undertakes fraudulent actions to benefit from NE are not eligible for NE benefits. Those families with a member receiving a pension (except disability pension) are also excluded from NE. 18  For rural (farming) families: NE uses a hybrid means test where the household income is calculated as the difference between the maximum NE benefit and the sum of the income from economic activity, social protection transfers, and the estimated unverifiable income (imputed income) from land and livestock.19 The adequacy of the benefits provided to households 16 See early evidence from Harold Alderman (1998). “Social Assistance in Albania: Decentralization and Targeted Transfers.� World Bank Living Standards Measurement Paper. 17 According to Law 9355, dated 10.03.2005 on Social Assistance and Services 18 According to LSMS, 20% of the NE beneficiary households have a member that receives a disability pension. LSMS 2008 does not include specific information on non-contributory disability programs. According to the administrative data, about 10% of the NE families also receive a disability benefit. 19 Farming families that have land in their possession are eligible for partial NE benefits.19 The level of partial NE is calculated as the difference between the maximum level of NE and the total income generated by the household from economic activity, income from social protection programs (excluding non-contributory disability benefits and disability pensions), and revenues from any type of capital and imputed income from land and livestock. Proposals by the social administrators need to be approved by the Local Council at the Municipality. The decision by the council is considered final. (Annex) 19 involved in farming activities is lower than the average NE benefit, raising the question that the methodology used in the calculation of the imputed income for rural households could be overestimating the income from farming activities. Figure 11: NE Targeting at Two Levels This non-continuous, multi-layered filters approach distorts eligibility decisions and results in errors of excluding the poor. Conceptually, there is no transparent weighting across variables (filters): classification on any single filter can completely override other indicators of need. A family could be quite poor, but still have a job (albeit low paying or seasonal), have some form of capital (but with low value or liquidity), or have a family member abroad (but with low paying job), and nonetheless still be excluded from program eligibility due to any one of these ad hoc filters. Our initial analysis suggests that these filter variables do indeed result in significant errors of exclusion of the poor. Examining the characteristics of those in the first decile of the population living in urban area,20 we find that strictly applying these filters would lead excluding 92 % of the of the poorest decile of the population. (Figure 13) However, in practice it may be difficult to verify informal employment, and as result the filter may be more precise in excluding those families with a member with a job in the formal sector. Simulating these differences between formal and informal employment status, we find that if only the formal employment filter is used, then 41% of households in the poorest 10% would remain eligible. Based on actual implementation results, we find that 28% of the poorest decile receive NE benefits, suggesting that some informal employment is being captured in implementation and used to exclude potential beneficiaries (in addition to formal employment). Analyzing and simulating the impact of individual filters, we find, for example, that about 82% households that have at least one member who is employed (formally or informally) are excluded from NE. These households are poor due to low earned incomes (as measured by the household survey), but are excluded from eligibility for the NE program due to their participation in economic activities (formal or informal employment). Similarly, about 15% of families in the poorest decile receive some remittances from family members working abroad, but again, these families – who are poor overall (classified 20 Poverty headcount is about 12.5% in Albania so this is a relevant group for analysis. 20 in the poorest decile) – would be excluded from the program due to the filter for receipt of remittances. Figure 12: The Effect of Filters in Excluding the Urban Poor (bottom 10 percent) in Ndihma Ekonomike Total Percentage of Individuals, Bottom Decile, % Filter Urban 100% (=122,172 individuals) % Percent % Have rental income Not eligible for NE Have rental income? % 0.8 Yes No 99.2% 0.8% % Receive Survivor Pension or Special Receive Survivor Pension or Special Merit Pension Merit Pension or Supplementary or Supplementary Pension? Pension % 0.4 Yes No 98.7% 1.3% Own a car? % Own a car Yes No 95% % 3.7 5.0% Receive Remittances from Abroad? % Receive remittances from abroad Yes No 81.6% % 14.9 18.4% Receive Old-Age Pension? % Receive Old-Age Pension Yes % 25.9 No 60.4% 39.6% % Have family member employed in the Have family member employed in the LM? LM (formal or informal) Yes No % 81.8 92.1% Percentage remaining as eligible for NE after applying all filters 7.9 % Source: World Bank Staff Estimates based on LSMS 2008 Note: The figure on the left indicates the cumulative effect of the “one-off� filters in excluding individuals in the bottom decile in urban areas. The table on the right presents the exclusionary impact of each filter individually. Moreover, such filters could introduce distortions in behaviors and potential disincentives to work. By design, some of these one-off binary filters could distort applicants’ behaviors by reducing incentives to work. Excluding families with any member that is economically active, for example, could potentially have this distortionary effect. Poor families could alter their behaviors to work informally (harder to monitor and verify) or not work at all, in order to qualify for benefits. Since the adequacy of the value of NE benefits is low, it is unlikely that these adverse incentives effects are particularly high, but the potential for distorting behavior is inherent in the use of such filters. Administration and Monitoring Capacity and Fraud and Error Controls 21 Implementation of the Ndihma Ekonomike program spans central and local governments, with decentralization of many functions. As discussed above and in the annex, the central government is responsible for program financing, design (including eligibility criteria / filters), geographic allocations across localities, monitoring and oversight. The central government 21 The work on the functional review of the administration processes of Ndihma Ekonomike is ongoing. The functional review intends to provide a clear description of the architecture of NE Program through assessing the operational aspects of NE and includes a detailed diagnostics on information flows, financial flows, functional responsibilities and institutional. 21 transfers “block grants� to local governments, which are then responsible for implementing the program, including determining household eligibility, maintenance of records (largely paper- based), and payments to approved beneficiaries. Eligibility decisions are made at the local level on the basis of application of eligibility filters by local social administrators and then a final decision by the local “Council� (which also introduces discretionary considerations, as discussed above). Benefits administration is hampered by weak capacity, information systems, and oversight and controls. Albania currently lacks an automated national registry of applicants and beneficiaries and benefit management is largely paper-based at the local level. This lack of an effective management information system leads to (a) inefficiencies in applying for and awarding benefits (with high transactions costs); (b) weaknesses in oversight and controls of fraud and errors; and (c) ineffective monitoring and evaluation of social policy. The lack of automated databases and a national registry is a serious constraint. The main difficulties reported relate to (i) the limited skills of the statistical staff at the regional and local level (many mistakes are reported in aggregating the tables); (ii) the paper-based data22 from the regional offices; (iii) the inability of current software and templates from the central SSS office to link the information across the indicators, and (iv) delays in transferring the information to the central SSS statistical office. As a result, monitoring is limited to evidencing the payments and number of beneficiaries. A large amount of information is collected at the local level, but is not used properly for developing indicators and observing trends over time at the household level. Computers are available at the regional and local offices, but are mainly used for Excel tables and word processing. The lack of automated databases and registries leads to inefficiencies in applying for and awarding benefits. Applying the current means test for NE without electronic databases and sufficient means to document incomes has resulted in a complex system, from the perspective of both implementers and beneficiaries. The transaction costs for applicants are inflated by numerous documentation requirements, and administrators are slowed down by verification and cross-check requirements with multiple agencies at different levels of public administration. Conducting simultaneous cross checks with other programs is burdensome on all institutions involved. Oversight and controls for reducing fraud and errors suffer from weak inspection capabilities and the lack of cross-communicable databases. The primary duty of the inspectorate is verification of documentary evidence (beneficiary files); cross-checks of 22 Every two months municipalities/communes collect and send to the regional directorate of the SSS, statistical data on the number of benefiting families and the amount of funds used. It is the responsibility of the regional directorates to establish the link with other corresponding institutions, and to exchange information regarding beneficiaries on a regional level. The data are processed by the regional offices of SSS and aggregated at the central statistics office. These data, together with another set of indicators on the demographic and economic situation of households, are then used by MoLSA to prepare the fund allocation plan for local government units 22 information (exchange of letters with other public offices like tax, cadastre, car registry, pensions, unemployment); registry checks to make sure that beneficiary names match with those approved for eligibility; and field visits to check on the living conditions of beneficiaries. Cross-checks are done on paper without the benefit of automated databases. Moreover, the budget for inspection is inadequate. There are often no means of transport, especially to remote areas, which reduces the inspectorate’s ability to perform home visits in remote communes. Furthermore, the information from inspections is used only for summary reports, rather than being utilized to develop profiles of fraudulent clients, poorly performing offices, etc. At the same time, sanction policy is ineffectively formulated and implemented. SSS inspectors have the power to impose fines on the front-line staff of local SA offices, but they rarely do. Additionally, evidence from field interviews reveals that inspections are performed mainly at random (tip-offs are rare, as the infrastructure for such is lacking), and that inspectors have not received any particular training on the job. SSS inspectors can only recommend that local governments terminate benefits. No attempt to recover benefits is carried out, even when fraud is detected. The consequence of these inefficiencies is continuous leakage of NE benefits to non-poor households; thus, opportunities for considerable improvement in the pro-poor impact of the program are clear. V. Conclusion with Policy Recommendations As discussed above, the objective of this note is to provide recommendations to policy makers in Albania to improve the poverty impact of Ndihma Ekonomike by improving equity, efficiency and transparency. Policy makers in Albania can build upon existing programs as they attempt to increase the capability of Ndihma Ekonomike as part of the poverty reduction strategy. This could be achieved by maximizing coverage of the poor (or minimizing errors of exclusion) and minimizing leakages to the non-poor, thereby ensuring that a greater share of the resources spent on the NE program reaches the poor; additionally, improving administration, monitoring, and transparency would enhance credibility and reduce fraud. This section presents reforms to achieve these objectives. Improving Coverage, Targeting and Adequacy of Ndihma Ekonomike Coverage provided to the poor by NE could be improved even within the existing budget envelope through improved targeting mechanism, administration, and fraud and error controls. With improved targeting mechanisms supported by better administration, the ability to identify fraud and errors of inclusion (providing benefits to non-poor households) could be reduced. Reducing errors of inclusion and fraudulent claims on benefits could make space for expanded coverage of the poor. Errors of exclusion could be reduced by revamping eligibility criteria to correct design flaws associated with the use of one-off filters that unfairly exclude otherwise poor households and introduce potential distortions in behaviors and disincentives for work. Therefore, improvements in targeting and administration of NE could result in a reduction in fraud and errors, as well as errors of inclusion – and thereby make way for extended coverage of the poor (reduced errors of exclusion). 23 Transparency, equity, and efficiency could be improved through targeting mechanisms that base eligibility on a single scoring formula to determine “need.� We recommend that the Government of Albania consider standardizing and improving eligibility criteria by eliminating the use of one-off binary filters and non-transparent discretionary criteria. Instead, we recommend that they revise program criteria such that eligibility is based on a single continuous scoring formula that incorporates objective weights and variables that are empirically associated with poverty status. Such criteria could be adapted for urban (non-farming households) and rural (farming households) areas, taking into account the different structural situations of these contexts. This revised formula could incorporate: (a) easily observed and verified incomes (formal incomes); (b) other proxy variables (such as the imputed value of assets and possession of consumer durable goods) that represent observable characteristics with empirical correlation with poverty status and/or help predict “hard-to-measure� incomes. Special circumstances (“discretionary considerations�) can also be considered, as long as these are incorporated into the single scoring formula with transparent weights relative to other eligibility criteria.23 The overriding principles in developing revised eligibility formula are that the scoring criteria: (a) target the poor as effectively as possible (to reduce errors of exclusion and inclusion); (b) are transparent in terms of explicit standardization of variables and relative weights; and (c) are administratively feasible for implementation purposes (keep it simple!). Such reforms could reduce errors of inclusion (leakages to the non-poor since the program currently benefits 136,000 non-poor families), reduce errors of exclusion (poor families denied benefits due to the current practice of applying of ad hoc filters), and improve transparency in the system. They could also potentially improve economic efficiency by reducing potential disincentives to work. Geographic targeting could be improved through standardizing the systematic use of a poverty map to determine allocations across localities. Linking the allocation of block local government NE grants to objective poverty criteria based on a poverty map would improve geographic allocation of NE funds. MoLSA has already started to planfor 2010 block grant allocations using some poverty indicators. This approach could be further improved by establishing and institutionalizing a clear methodology and allocation formula, using the poverty map as a guide. This would improve the geographic targeting of resources by ensuring that local governments receive a share of the available resources based on regional poverty, subject to the annual budget envelope. This methodology would need to be updated on the basis of data from the new census, planned in 2011, and the most recent LSMS data, to account for changes in population and regional poverty rates. Preliminary estimates show that the targeting accuracy of 23 These “objective� variables and their associated weights are usually defined through regression analysis using household survey data to design the eligibility criteria. In some instances, policy makers may want to add additional variables to the scoring formula that do not derive from household survey analysis. For example, a category (variable) for “special circumstances� to allow for additional local knowledge (discretionary variable) to capture idiosyncratic shocks to households such as catastrophic health shocks or other circumstantial considerations. Inclusion of additional variables is possible but these should be added to the scoring formula wit explicit (subjective) weights so as to keep them in relative perspective with the other variables under consideration (rather than applying them as one-off ad hoc filters). 24 NE would increase from 56% to 61% if the block grant allocation was to be done based on the poverty map within the available budget envelope. Furthermore, the design of NE could be modified to account for disability benefits in the household income. Coincidentally, this single change could save enough in misdirected funds to eliminate extreme poverty, if the monies saved were redirected toward the poorest households. Some current NE beneficiary households include individuals already receiving a social protection transfer through disability programs. One option would be to change the rules so that the disabled household members receiving a disability benefit are not counted in the household size when calculating the NE benefit for a household. A rough estimate suggests that about 7% of the total NE spending could have been saved in 2008 if disability beneficiaries were not included in the calculation of benefits for NE beneficiary households.24 The resulting 7% savings is roughly equal to the amount of funding needed to eliminate extreme poverty. 25 Another option would be to include income from disability transfers as part of the means testing for NE. Erosion in the generosity of the NE benefit could be avoided by introducing a transparent indexation rule that would link the benefit indexation to inflation. While the MoLSA has been working on identifying and eliminating fraudulent cases in NE to better target the benefits to the poor, these savings have been spent on the increasing number of disability beneficiaries under the non-contributory disability program. As a result, the real value of the benefit has declined, along with the coverage of the NE program, leading to a low adequacy and therefore, a low poverty impact. Indexing the benefit level to inflation would maintain the real value of the NE benefit. Improving Administration of Benefits, Monitoring Capacity, and Fraud Controls Irrespective of the targeting mechanism chosen, registries and databases need to be improved to support implementation. Intake processes need to be carefully designed to ensure transparency, outreach to the potentially poor, cost efficiency, and administrative feasibility. Albania could consider establishing a unified household information registry to help avoid duplications and track beneficiaries. As proper identification of beneficiaries is crucial, a unique social identification number could be used to link registry information and beneficiaries with other systems and programs. Furthermore, data management should be designed to flexibly respond to the changing policies and updates. Developing a national registry/database would significantly improve the monitoring/oversight capacity of the institutions and would improve transparency in all aspects to enhance credibility and reduce fraud. In Albania, oversight instruments are even more important, since program implementation and management is decentralized to local governments. The goal should be to develop feasible and cost-efficient systems to minimize the occurrence of fraud. A solid monitoring system would also serve a political role, helping the government be accountable to the public. 24 Disabled members are assumed to be at working age. 25 The extreme poverty gap was estimated at 256,000 ALL in 2008. 25 Improving the Proactivity of Social Assistance Programs Albania could also strengthen the “proactivity� of social assistance programs by linking beneficiaries to human capital and/or activation measures. Currently, employment and social assistance programs are completely disconnected, except for the requirement of social assistance applicants to register as unemployed before application (which can create disincentives for work). Many middle-income and EU countries have adopted various mechanisms to enhance the “proactivity� of social benefits to get beyond passive cash handouts. Some options include:  Linking child beneficiaries to human capital. Many countries (low-income, middle-income and even high-income OECD countries) link receipt of cash transfer assistance to school enrollment and attendance to improve the long-run impacts for strengthening human capital and reducing poverty. These mechanisms are generally referred to as “conditional cash transfers.� Such conditioning of benefits could make sense in Albania’s context, where, contrary to most neighboring countries, Albania has not achieved universal enrolment in basic education. Net enrollment is about 90% for basic education and only 60% overall for secondary education, while the figure is significantly lower among vulnerable groups. According to administrative data, net enrollment in basic education is only 52% among children in families benefitting from NE cash assistance. The NE program could be enhanced to incorporate incentives for beneficiary families to enroll school-aged children in school and ensure daily attendance, thus increasing the “proactivity� of the benefits beyond simple cash handouts.  Activation links. Many EU and other OECD countries have been enhancing the “proactivity� of social benefits programs by linking beneficiaries to incentives, social contracts, and services to promote employability and employment opportunities. These efforts take many forms, depending on the profile of the poor (barriers to work), the programs (incentives aspects), and institutional capacities, including: o Incentives to promote work efforts by able-bodied adults. Such incentives are usually attached to receipt of cash assistance benefits, and can include: (a) design features to reduce potential disincentives to work by (i) avoiding eligibility criteria that requires registration as unemployed and/or bars economic activities – as is the case with the one-off filters in Albania’s NE program); (ii) revamping eligibility criteria to reduce disincentives to work (e.g., by introducing higher exit thresholds for income testing, earnings disregards for earned incomes, time limits, etc.); and/or (iii) introducing “work requirements� that require able-bodied adults to participate in “work-related activities� for a certain number of hours per week (e.g., training, job search activities, or actual employment / work activities). o Social contracts established between the social worker and the beneficiary that identify the family’s assets, constraints, and goals and map out a set of actions that the beneficiary family must comply with to work towards reducing their barriers to work or improving other social conditions facing the family, usually in conjunction with social worker “accompaniment� and monitoring. o Service links such that beneficiaries of cash assistance are actively linked to employment- related services (or other social services for hard-to-serve families), such as training and education, job readiness training, job search services (information, matching and brokerage), and actual job placement. Depending on the institutional context, these 26 service links can be promoted via “one-stop shop� models for integrated service provision (linking Centers for Social Work with Public Employment Services into single one-stop- shop offices) or via monitored referral systems. Identifying Clear Roles for Social Assistance and Social Insurance To increase efficiency, impact, performance and sustainability, Albania should rationalize the incentive-compatibility of the safety net within the context of reforms to social insurance for a more consistent overall social protection system. This policy trend calls for greater consistency and incentive compatibility between non-contributory benefits and contributory social insurance benefits, particularly with respect to benefit levels and eligibility requirements (e.g., for disability insurance and assistance benefits). Separating the contributory and non-contributory benefits could improve incentives to participate and also help achieve fiscal sustainability. This could allow for fiscal space to be created in the budget that could be used to improve and expand social assistance. 27 ANNEX: Registration Process NE relies on “on-demand� registration to determine eligible households. To be considered for eligibility for benefits, households must come to apply at social assistance offices at the corresponding commune or municipality between the 1st to the 15th day of each month. The applicants fill in the “Statement of Family Social-Economic Situation (FSES),� in which they provide personal data and information on household income (from all sources)26, housing, demographic data of all the household members (education, profession, work, health), and other social family information. The application is submitted by the head of the household and signed by all other members. This application is submitted together with a number of documents that certify socio-economic status and/or asset ownership. (Table 4) Table: Documents Presented at Application for NE Benefits List of documents presented for NE benefits  Family status certificate  Certificate of land/real-estate  Socio-economic statement  Monthly socio-econ. update  Ownership act or certificate deed to land  A certificate of “Unemployed seeking work� issued by the Labor office every three months.  Other specified documents for migrants, orphans and invalids. Additional documents for electricity subsidy  No. of contract with cash  Proof of past electricity payments.  Certificate for not having any hh member benefiting NE/invalidity.  Proof of having only one KESH contract for hh. Verification of information provided by the applicant households is done by Social Administrators through home visits. For each (new) NE application, a social administrator is required to visit the family to make sure the FSES is filled in correctly. Self-reported socio- economic statements and “home visit� inspections are meant to be complementary to each other. One of the crucial aspects of both administrative procedures is the cross-checking of the declared 26 Households are required to self-declare their income from: 1) employment-generating activities (employment, self- employment, registered work, work with a self-employed family member, small private business); 2) social benefits (old-age pension, invalidity entitlement, family pension, other pensions, unemployment benefit); 3) emigration; 4) shares and bonds; 5) bank deposits; 6) inheritance in monetary values; 7) use of properties by the owner or other persons; 8) agriculture; 9) livestock; and 10) other. 28 property and income. For urban applicants the process is strictly bureaucratic, as they are assumed not to have income from agriculture or livestock. In this case, the checking is done only through the other institutions (see above), and by checking that none of the adult members of the household has migrated abroad (individual presence of each is required). In the rural areas an assessment of the income from the agriculture and livestock is needed, as most of the households possess land based on Law 7501 “On the land reform.� After the social economic statement is compared with the actual situation, the social administrator forms his or her opinion on the type of social assistance for which the household is eligible. The social administrator can deny the applicant’s eligibility if one or more income sources from the situations in Table 5 is observed to be in practice. Income from disability benefits are not taken into consideration when defining eligibility for NE. Table: Check List for Denying Eligibility of NE benefits List of eligibility checks  Owns any form of capital (with the exception of abode or farming land)  Is economically active (disabled are exempted).  Resides outside of the country.  Is not registered as unemployed looking for work (farmers and disabled are exempted).  Refuses to work when a job is offered by the employment offices.  Refuses to take the land given  Does not collect the benefit within 6 working days (special cases are foreseen).  Undertakes fraudulent actions to benefit from NE. The application and verification processes are paper-based. The social Administrator verifies the documentation presented by the applicant households with the relevant institutions between the 15th to the 25th of every month. Upon completion of collection of required documentation, the social administrator compiles the list of eligible candidates. He then sends the list to the relevant institutions27 and receives confirmation of receipt. 28These institutions are required to check the lists and provide relevant information to the social assistance office at the commune/municipality within ten days. Both the applications filed by the local SA offices, and the other NE candidate lists sent out for verification to other institutions, are based on paper work. 27 The institutions include: 1) labor offices, 2) tax offices, 3) social insurance office, 4) offices for the registration of immovable properties, 5) offices for vehicle registration, 6) relevant sections in the municipality and, 7) employment inspectorates. 28 The lists that are checked with the other institutions include: 1) name, father’s name and surname of the individual; 2) file number; 3) date of birth; and 4) current address. Evidence from field interviews shows that in their current form, the lists lack the necessary information needed for cross-checking. 29 Figure: Application Process and Administration flows for NE Real Estate Registration Office Vehicle Registration Office Real Estate Regional SSS Registration Office Office Soc. Ins. office Tax office Final decision Employment office Draft decision sent to commune/municipality Applicant hands in the council (see Table FSES* (see Table 1) 3)*** Social Local Govn. Applicant Administrator Council  Visit the hh the first time Final Decision and then two times a year**.  Verifies socio-economic data and living conditions and classifies the family.  Rejects application if one of hh members if elegibility  The draft-decision criteria are not met (see  The decision Table 2)  The payroll list  The registry  The legislation file Archive * - This is the first step in applying for NE benefits ** - The visit and the verifications are the second step of the application *** - The decision by the local govn council is the final step in applying for NE The local government council makes the final decision regarding households eligible for NE. After verifying all the required documents, the social administrator files the documents and prepares the draft decision for approval by the council of the commune/municipality. The draft decision includes the beneficiaries list and should also be accompanied by a relevant report signed by the head of the economic assistance department in the local government, and the social administrator. The SA also participates in the meetings of the council to provide information on the proposal. Based on discussions, the draft decision can be changed. Following the approval of the council, the decision taken is archived, together with all the necessary documents, and the list is sent to the financial department, which prepares the payroll. Payments to beneficiaries are made through banks in urban areas, and through post offices in rural areas. Table: Information on draft decisions sent to municipality council Draft decision for NE benefits  The ordinal number, full name of the hh head, file number, number of heads, type of benefit (full or partial), the period of entitlement and the amount of assistance.  Full names of applicants who do not receive economic assistance for given reasons.  The names of families introduced to the scheme for the first time, explaining the reasons why.  The names of families reintroduced in the scheme, explaining the reasons why. Farming families that have land in their possession are eligible for partial NE benefits. 30 Management of Registers/data bases Management of databases is the responsibility of local administrators. Names of NE’s beneficiary households (all family members over 18 years old) are kept in the local SA office registry. After the decision has been passed from the local council, social administrators archive the following documents: 1) draft-decision, 2) decision, 3) payroll list, 4) registry, and 5) other legislative documents. The registry usually refers to the information on the decision and payroll list and includes: 1) name, father’s name, last name; 2) number of file; 3) number of household members; 4) type of benefit; 5) benefit amount; and 6) identification number (passport, identification certificate, etc.). The information and the statistical tables are transmitted every two months to the regional office of SSS, from where the information is transferred to the Central Office and then to MoLSAEO. Local governments collect information on some other characteristics of the households. The social administrator collects and reports information on: 1) number of families benefiting from NE within the central or local budget (identifying also the number of members, number of families with disabled members, and funds per family); 2) number of families entering or leaving the scheme (and the specific reasons); 3) number of families with triplets (or more) receiving NE; number of orphans (with the orphan status), and family heads who receive subsidies for the increase in energy prices; and 4) age information for children in families receiving NE. 31 REFERENCES Alderman, H. 1998. “Social Assistance in Albania: Decentralization and Targeted Transfers.� World Bank Living Standards Measurement Paper�. Washington, DC. World Bank. World Bank. 2006 “Albania Social Insurance Review�. Washington, DC. World Bank. World Bank.2007.Albania:Urban Growth, Migration and Poverty Reduction, A Poverty Assessment. Washington, DC. World Bank. World Bank. 2008. Background Report on the Overview of the Social Protection System in Albania. Washington, DC. World Bank. World Bank. 2010. The impact of the crisis on the poor and vulnerable groups in Albania: Some Preliminary Findings,WB Programmatic Poverty Assessment for Albania. Mimeo. Washington, DC. World Bank. 32