68711 World Bank Bulgaria: Mitigating the impact of the economic crisis on the poor Social Safety Net Reform Policy Note September 20091 Main messages Bulgaria has a seasoned social safety net, but the system needs a fresh look to improve its ability to respond fast to tackle the impact of the crisis on the poorest Bulgarian households and to promote active social inclusion of the poor and vulnerable. Bulgaria’s social policy has many positive elements, most notably a well-targeted social assistance program, education-conditioned child allowances, investments in employment activation an active labor market programs. However, there are also a number of areas for improvement, which constrain the effectiveness of the overall system.  Bulgaria operates two very well-targeted but small cash transfer programs – the Guaranteed Minimum Income (GMI) program and heating allowances. Both are means-tested and targeting accuracy is very impressive: 85 and 65 percent of GMI benefits and heating allowances (respectively) go to those in the lowest quintile. However, the coverage and generosity of these benefits are quite low. Coverage of the GMI has fallen dramatically since 2005 (from 140,000 to under 40,000 beneficiaries by 2009), resulting from tightened eligibility criteria and the imposition of strict time limits (now 12 months) – a policy which has recently been found by the European Committee of Social Rights (ECSR) to be in violation of Bulgaria’s obligations under the European Social Chapter. With concerns about “welfare dependency� running high, policy makers have focused more on forcing "graduation from the program" than "graduation from poverty" which could be pushing structurally-poor vulnerable populations out, without actually strengthening their assets. The program also requires certification of 9-months of unemployment prior to application (and requires monthly registration of unemployment during the program). These requirements are highly inefficient, distort incentives, and reduce the ability of the GMI to respond in an agile manner to crises.  Benefits administration is often weak with mostly paper-based processes and record-keeping, rudimentary computerization of records that generates at most a payroll listing of locally-approved beneficiaries but lacks even the most basic data on applicants or reports for policy feedback of any kind, and often dilapidated offices. Weak benefits administration reduces efficiency, results in duplications of processes and benefits as well as errors.  Bulgaria has invested significantly in introducing and modernizing employment activation programs and the Public Employment Services (PES) offices. The PES offices are quite modern and are striving for “model bureau� status in EU standards. They operate elements of the “one-stop 1 This note was prepared by a World Bank team consisting of Christian Bodewig, Kathy Lindert, Lire Ersado, Boryana Gotcheva and Yordan Dimitrov. It builds on a forthcoming World Bank report “Bulgaria: Social Assistance Programs – Cost, Coverage, Targeting and Poverty Impact�. 1 shop� model with front offices that handle routine initial queries (and requests to certify unemployment – needed for SA applications), and “back offices� for higher-end services for job seekers and employers. Client orientation typically is strong, and “clients� are profiled for skills, education, qualification, needs via an interview with a trained interviewer who immediately enters the data into the database of job seekers. They also jointly fill out a tailored “mutual obligations� personal plan (considered good practice internationally) to signal co-responsibility for finding a job. They then access a range of activation services, including: (a) training (differentiated by skill level, ranging from basic literacy/remedial education to low-skill vocational training to higher-skill technical vocational education and training); (b) counseling and job-readiness advice (for the hard- to-serve); (c) intermediation and job brokerage services, etc. The centers interface not only with job seekers but also with employers and offer employer-services to hold recruitment events. Automated registries and information systems greatly facilitate case management and tracking.  Despite some strong elements – a well-targeted GMI cash assistance program and strong activation/PES services – the system is largely disconnected. From top-down, the social assistance and employment services are institutionally completely separate. They are managed by separate central offices, and the disconnect goes all the way down to local offices which are physically separate. There is also little policy connection between the two that would support a “social promotion� agenda to encourage poor GMI beneficiaries to take-up the activation service supports (no automatic referrals or outreach to bring GMI beneficiaries to PES services, no joint MIS systems or case management, etc). This is a missed opportunity to link social policy more effectively, with a focus on forcing “graduation from the program� via the time limits on GMI without leveraging the activation services to help build assets that could actually help promote “graduation from poverty� for longer-run employability and job opportunities. Policy directions These challenges point to a structural reform agenda that would:  promote stronger linkages between (or even integrate) the social assistance and activation agendas with an emphasis on “social promotion� that would help promote graduation from poverty not just from program benefits;  overhaul the benefits administration system and refurbish offices to modernize and automate the registry, client interface, and case management processes and systems as functional tools of the trade (to reduce duplications, errors, fraud and admin costs);  strengthen performance-based management tools for both cash assistance and PES services;  revise some of the policies guiding the GMI to strengthen its capacity to quickly respond to the crisis and mitigate its effect on the poorest Bulgarian households, such as: (i) eliminating the 9 month unemployment requirement prior to application; (ii) removing requirements for monthly certification of unemployment for beneficiaries (distorted incentives); (iii) rethinking across-the- board strict 12 month time limits and developing policy measures to better link GMI beneficiaries to tailored activation services; and (iv) expanding coverage of the poorest and vulnerable groups. Some of those measures may result in a slightly lower degree of targeting accuracy, but that trade-off would be well worth it for the gains in effectiveness of Bulgaria’s social policy (and given how high the targeting accuracy is already). 2 Bulgaria: Social Safety Net Reform Policy Note Introduction Like many other EU Member States Bulgaria is strengthening social inclusion elements in social policy and has recently been refocusing its attention on the social safety net in the wake of the global economic crisis. This note aims to inform this discussion by summarizing recent World Bank analysis on the performance of the social safety net on the basis of the 2007 Multi-Topic Household Survey. It reviews the effectiveness of the main social safety net transfer programs (see Box 1) in reducing poverty and assesses how well they are placed both to address the possible social impact of the crisis and to promote active social inclusion. It also reviews the effectiveness of recently introduced limits to eligibility to Guaranteed Minimum Income (GMI), the main social assistance benefit, which aimed at raising work incentives for beneficiaries. It finally lays out a set of policy recommendations on how to strengthen the social safety net to deal with the economic crisis in the short term as well as to promote employment activation and support broader a human development agenda over the medium term. Box 1: Bulgaria’s Social Safety Net Bulgaria has a long record of social protection programs and institutional arrangements for delivering income support and other benefits to the population. Both contributory social insurance schemes and noncontributory social assistance benefits and social services are in common use: (i) social insurance programs, including pensions, sickness, disability and unemployment insurance; (ii) passive and active labor market measures; and (iii) the noncontributory cash transfer aimed at poverty reduction (social assistance) and child protection, and (iv) social care services. This note focuses on non-contributory social safety net programs. In Bulgaria, the main noncontributory safety net programs consist of two schemes for low-income and vulnerable households: (i) Guaranteed Minimum Income (GMI) and (ii) Heating Allowance (HA). There are other important noncontributory assistance programs which are categorical in nature, including (iii) child protection benefits through Monthly Benefits for Families with Children (MBFC), birth grants, paid maternity leave for uninsured mothers and other support schemes under the Law on Family Support for Children; and (iv) benefits for people with disability through Monthly Supplement for Social Integration (MSSI), which is a supplement to disability pension. The social welfare system also provides (v) social care services in institutions, communities and for families. The GMI and HA are antipoverty means-tested on income (and on assets) programs and are targeted to the most vulnerable families. They are aimed at providing protection to the poor and vulnerable and play an important role in helping individuals and families cope with income shocks and poverty. The MBFC program under the Law on Family Support for Children uses a combination of categorical targeting approach and a means test with a higher threshold for selecting beneficiary families with children. The MSSI provides a monthly income supplement for people with disabilities and supports their social integration. Spending on the social safety net Social safety net programs take a small share of the overall spending on social protection in Bulgaria. The total budgetary allocations for the four main social assistance and child allowance programs in 2007 amounted to BGN 504.6 million, slightly more than 81 percent of all social assistance outlays by the Social Assistance Agency. These expenditures represented only 7 percent of total spending on social protection (Table 1). While overall social protection spending amounted to more than 35 percent of total public spending, the four targeted social assistance and child allowance programs accounted for only 2.5 percent of overall public spending. The bulk of social protection spending is driven by contributory old age and disability pensions. 3 Table 1: Public Spending on Social Protection Programs, 2007 Spending as percent of Spending as total social insurance, Social spending Social insurance and assistance Spending percent of Total social assistance, and as percent of categories (BGN million) public spending social care spending GDP Pensions 4,691 22.9 64.8 8.3 The four targeted social assistance and child protection programs 504.6 2.5 7.0 0.9 GMI 78.4 Heating Allowance 87.1 Child Allowance 231.7 Disability Suppt. 107.3 Other social insurance, social assistance and social care 2,039 10.0 28.2 3.6 Social insurance, social assistance, and social care (total) 7,235 35.3 100.0 12.8 Source: Staff estimations based on data from the Ministry of Finance, Social Assistance Agency of the Ministry of Labor and Social Policy, and the National Statistics Institute (NSI). Bulgaria’s social safety net spending is similar, though at the lower end, in a comparison across new EU Member States and substantially below the OECD average. Overall, Bulgaria spent about 1.3 percent of the GDP on social welfare programs in 2008, compared to about 2.5 percent average for OECD countries (see Figure 1) and to 1.6 percent for countries in Eastern Europe and Central Asia. Bulgaria’s spending is of a similar magnitude as most new EU Member States, with only Poland and Romania spending less than Bulgaria. Figure 1 Bulgaria’s spending on social assistance programs compares well with neighboring countries Public spending on social assistance programs as share of GDP, (latest available year, 2007 for Bulgaria) Croatia Bosnia-Herzegovina Hungary Ukraine OECD Uzbekistan Macedonia Russia Moldova Armenia Latvia Estonia Belarus Lithuania Bulgaria Kyrgyzstan Kosovo Serbia Romania Kazakhstan Albania Poland Georgia Turkey Azerbaijan Tajikistan - 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50 5.00 Source: Social Assistance Agency data for Bulgaria; Various World Bank Public Expenditure Reviews; OECD Social Spending Database. Bosnia and Herzegovina and Croatia include categorical benefits for war veterans 4 Assessing the performance of the social safety net The performance of the social safety net can be assessed by looking at (i) coverage – what is the share of poor households who receive the benefit; (ii) targeting – what share of overall benefit spending goes to the poor; and (iv) adequacy – what share of poor household consumption does the benefit represent. Figure 2 presents data from the 2007 MTHS on these three characteristics. It suggests that Bulgaria’s social safety net has a mixed performance. Figure 2 Determinants of Bulgaria’s Main Noncontributory Programs a. Targeting performance (% of the benefit spending going to…) b. Benefit adequacy (% of HH consumption) 48.7 7.8 Heating Allowance 51.3 Heating Allowance 6.0 66.8 6.6 31.4 12.2 GMI 68.6 GMI 16.2 84.2 27.4 70.4 23.6 Disability Supplement 29.6 Disability Supplement 21.7 39.8 25.4 70.1 14.0 Child Allowance 14.2 Child Allowance 29.9 19.4 41.1 21.3 63.8 All SA/Child Allowance 22.8 All SA/Child Allowance 36.2 36.6 44.1 0.0 20.0 40.0 0.0 20.0 40.0 60.0 80.0 100.0 Nonpoor Poor Bottom 20% Poor Bottom 20% Very Poor c. Coverage (% of HH receiving the benefit) d. Coverage of child allowance 19.6 Child Allowance 36.2 very poor 41.4 61.1 2.0 Heating Allowance 12.1 15.5 32.7 poor 0.9 GMI 11.4 15.8 35.5 nonpoor 14.0 Disability Benefit 33.9 30.4 37.4 32.6 All All SA/Child Allowance 66.0 69.9 83.3 0 20 40 60 80 100 0.0 20.0 40.0 60.0 80.0 100.0 percent of households with children under 18 and in Nonpoor Poor Bottom 20% Very Poor school Source: Staff estimates based on MTHS 2007. Note: coverage data for child allowance in panel c relates to all households for comparability reasons, while coverage rates in panel d refer to households with children below the age of 18 only First, the targeting accuracy of the GMI and HA is very high. As Figure 2 shows, GMI and HA transfer around 85 percent and 67 percent of benefit spending to the bottom 20 percent households of the income distribution. This is a very solid performance even in an international comparison. The Child 5 Allowance which, given that it is not an explicitly targeted anti-poverty benefit, transfers most of its spending to the non-poor – not surprising and in line with its objective. This suggests that GMI and HA are very effective in transferring resources to those who are most in need. However, this good targeting performance may not necessarily be the result of sound targeting methods per se. It may also derive from the fact that benefit levels are low, suggesting there may be self-targeting of those who really need the benefit. This is also consistent with low coverage (see below). Second, however, benefit adequacy is low: Social assistance benefits only account for a relatively small, though important, part of household income. Heating allowance only accounted to 6 percent of monthly household consumption in 2007 for households in the bottom 20 Percent of the distribution, which is not surprising, because it is a seasonally limited benefit. However, GMI only accounted for about 16 percent of household consumption. At the same time, the non-targeted child allowance is about as important a source of income to poor households as the GMI. This suggests that social assistance benefits only play a limited role in tackling poverty. Indeed, as Table 2 below shows, a large share of households who receive GMI are poor (55.5 percent, compared to 9.4 percent among all households who do not receive GMI). Without the benefit, the poverty rate would rise to 60 percent. However, more importantly, the poverty gap, which measures the monetary amount needed to get household consumption to the poverty line, is large for GMI beneficiary households at BGN 30. That said, without the benefit, it would rise to close to BGN 50, which suggests that GMI does play an important role in reducing the poverty gap. Table 2: Poverty among GMI beneficiaries is widespread and deep, and the benefit does not bridge the poverty gap Comparative Poverty Rates and Poverty Gap, 2007 Poverty rate in Poverty rate w/o Poverty gap Poverty gap percent benefit in percent in BGN w/o benefit in BGN GMI recipients 55.5 60.1 30.4 49.4 non GMI recipients 9.4 3.9 Source: staff calculations based on MTHS 2007 Third, GMI and HA are very small programs and their actual coverage of the poor is severely limited, in particular the very poor. As Figure 2 (panel c) shows, only about 16 percent of the bottom 20 percent of the income distribution receives GMI. The situation is similar for HA, with 15.5 percent. Both GMI and HA do better on the coverage of the very poor households, with more than 30 percent of very poor household in receipt of the benefit. However, inversely this still means that around 65 percent of very poor households are not captured by one of the two programs. Indeed, as Figure 3 below shows, the number of beneficiaries of GMI has precipitously declined over the last years, from around 140,000 in early 2005 to below 40,000 in early 2009. Much of this may reflect improved employment opportunities which have pulled people off social assistance. Indeed, poverty incidence has substantially declined between 2003 and 2007 on the back of fast economic growth2. However, it also reflects an increased focus on the ‘exclusion’ of recipients using different kind of ‘filters’ before performing the means test. 2 World Bank (2008), Bulgaria: Living Conditions before and after EU Accession 6 Moreover, the decline in beneficiaries reflects a tightening in eligibility for GMI effective since January 2008 (see below). In addition, the eligibility thresholds for GMI have not been adjusted in line with real incomes, likely resulting in a reduction of eligible individuals. It is expected that demand for GMI will expand in the wake of the economic crisis, though the initial picture in 2009 is one of constant beneficiary numbers (see Figure 3). Figure 3 GMI beneficiary numbers have been steadily falling over the years 160000 GMI beneficiaries 140000 120000 100000 80000 60000 40000 20000 0 Sep Sep Sep Sep Mar May Mar May Mar May Mar May Mar May Jan Jul Jan Jul Jan Jul Jan Jul Jan Nov Nov Nov Nov 2005 2006 2007 2008 2009 Source: Social Assistance Agency Coverage of Child Allowance is substantially higher than GMI and HA, in particular of the very poor, however still rather low. Figure 2 (panel d) shows the child allowance coverage of households with children. A coverage of around 55 percent of poor households, i.e. households living under a poverty threshold of BGN 185 per month per adult equivalent in 2007, nevertheless suggests that around 45 percent of poor households with children aged below 18 did not get the child allowance benefits. The eligibility threshold for child allowance in 2008 was BGN 300 per month, suggesting that there are a lot of poor households who should be getting the child allowance but who are not. This low coverage is worrying, suggesting significant exclusion due to barriers to access. It goes beyond the scope of quantitative research to determine the reasons for low coverage, but two explanations are possible (i) access barriers, for example lack of knowledge of potential beneficiaries or too complex application rules; or (ii) benefit levels are too low to make it worthwhile for beneficiaries to apply, for example in rural areas where the next Social Assistance Office is far and mobility is limited. Insufficient budget allocation is unlikely to be an explanation for low coverage, since actual GMI benefit spending was substantially below budgeted amounts in recent years. The Monthly Supplement for Social Integration (MSSI), a supplement to disability pension, is awarded with categorical targeting of people with 71 percent or greater disability is and based on self-reported disability. In 2007 about 70 percent of people with self-reported severe disabilities (in excess of 70 percent) were covered by the disability supplement. As it is not an explicitly targeted anti-poverty 7 benefit, it transfers most of its spending to the non-poor. However, the coverage by the program of the poor is considerable, with about 35 percent of the poorest 10 percent covered by the program, while adequacy is also significant, at around 24 percent of household expenditure. However, it may be worth considering introducing a benefit eligibility conditionality for attending rehabilitation procedures or training to promote the social integration effect of the benefit. The social safety net and employment activation In recent years countries across the OECD and the EU have increasingly been introducing “activation� elements into their social protection and employment policy frameworks as well as approaches to prioritize and individualize service provision for the unemployed. Activation policies typically build on a “mutual obligations� approach combining  enhanced responsibilities of the employment offices in terms of improved and more focused service for the unemployed, particularly those hardest to place, with  enhanced responsibilities of the unemployed, including able-bodied individuals receiving social assistance and other welfare payments, to regularly visit the employment office and be available for work. Aiming to promote employment activation of those furthest from the labor market, the Government has tightened eligibility of social assistance for working age individuals who are capable to work. As of 1 July, 2006, the maximum duration of eligibility for GMI for able-bodied working age beneficiaries was limited to a maximum period 18 months. While they are eligible to enroll in remunerated employment and training programs upon losing GMI benefit eligibility, affected beneficiaries cannot re-apply for the benefit until after 12 months of break in benefit receipt. The policy measure has been enforced since January 1, 2008, resulting in a substantial drop in beneficiaries from around 60,000 to below 50,000 within just two months (see Figure 3), and the duration of GMI receipt was further reduced to 12 months as of July 1, 2008. In August 2009 the policy measure was found by the European Committee of Social Rights (ECSR) to be in violation of Bulgaria’s obligations under the European Social Chapter. The original rationale for the measure was that the buoyant labor market created an opportunity to remove employment disincentives in the social assistance system and to move people off benefits and into work. Prior to the measure’s effectiveness the Social Assistance Agency (SAA) and Employment Agency (EA) reviewed the structure and profiles of beneficiaries and developed individual activation action plans for affected beneficiaries to identify particular barriers and disadvantages of beneficiaries in particular with respect to their skill levels. This review allowed the Employment Agency, through its branch offices, to draw up individual activation action plans for affected beneficiaries. This involved setting up individual interviews with the beneficiaries to inform them of the measure and the options for enrollment on training and employment programs, including subsidized employment which by now takes the largest share of affected beneficiaries. A survey of the initial cohort of affected beneficiaries conducted by the National Statistical Institute in April 2008 confirmed their highly vulnerable profile: Almost 60 percent of the affected beneficiaries were with primary education (4 grades), less than 4 grades of education or illiterate. Moreover, close to two thirds of the affected were young, between 21 and 40 years of age, while one third is aged between 8 41 and 60 years of age. Close to two thirds of the surveyed population reside in rural areas, where the demand for labor is limited. The majority of affected beneficiaries are Roma (63 percent), regionally concentrated in Plovdiv/Pazardjik and Vidin regions. The survey revealed barriers in moving people off welfare and into work and limitations to the effectiveness of the measure. Around 70 percent of the surveyed beneficiaries reported to have remained unemployed 3 months after the stoppage of the GMI support – in a situation of overall high demand for labor. More than half of the remaining 30 percent worked as formally employed, while the rest worked in the informal sector, as self-employed, or in temporary employment. About a quarter (27 percent) of the affected stated that they had not looked for jobs after the stoppage of the GMI support. They pointed at several reasons - lack of qualification, low pay offered/higher reservation wage, family reasons (child care), long distance/travel of the offered job to the place of residence. Not looking for a job does not seem to be driven by the receipt other forms of social benefits. Barriers also include apparent insufficient activation support by the Employment Offices (EO): Despite early identification of beneficiaries and the preparation of individual action plans, many former beneficiaries reported that were left to search for jobs on their own. Another reason for low activation rate is the low share of those who were offered participation in program for subsidized employment or training: only around 12 percent of the respondents declared that they have been given such options and only two thirds of them have accepted to participate. While many affected former GMI beneficiaries enrolled onto remunerated training and subsidized employment programs, administrative data from the employment agency shows that the measure has also led to attrition from any form of employment office programs, although there is no disaggregated data available. Anecdotal evidence from local employment offices in Sofia suggests that his attrition may be concentrated among mothers of children above the age of 3 who stay at home to care for children. Removing the benefit from the mother may have disproportionate effects on children, as, according to experience the world over, the mother’s contribution to household income is more likely to be spent on children. Given data limitations, these issues can be analyzed in detail at this point, but they should. This analysis suggests the following messages:  Bulgaria’s employment activation approach appears to be unbalanced and not in line with mainstream activation policy across the EU. Stand-alone time limits are not a useful activation tool: Cuts in GMI entitlements appear not to have been complemented with commensurate institutional change in the employment offices for more effective support through a pathway to sustainable employment for disadvantaged and GMI benefit-reliant job seekers. A core component of employment activation reforms across the EU, improving the effectiveness of the employment offices in helping highly disadvantaged jobseekers find employment is challenging, as it requires institutional change and a culture of greater partnership with non-governmental and community-based organizations – for example to prevent unnecessary attrition. Moreover, employment activation policy in many EU Member States, including Germany and the United Kingdom, is not about rigid time limits in eligibility but a discretionary approach to benefit 9 sanctions to enforce cooperation of the job seeker and an agreed gateway towards sustainable employment.  The low adequacy of GMI implies that benefit receipt may not act as an important disincentive to look for employment. Rather, barriers to employment lie elsewhere, for example in low skills of jobseekers and limited employment opportunities.  At the same time, most GMI beneficiary households are poor and their poverty is deep – even with the benefit. Taking away the benefit without a realistic chance for commensurately remunerated employment risks deepening poverty of GMI recipient households, many of which have children.  There may simply be no jobs in the primary labor market for affected former GMI beneficiaries with a very low level of education and multiple employment barriers. Recent experience from the Czech Republic suggests that effective labor demand for low skilled workers remained low even in times of high labor demand overall – suggesting that very low-skilled workers fail to find employment even if they want to. Not everybody is “activatable�. Evidence from the US, for example, suggests that 5-7 percent of households there have multiple barriers to employment and are therefore not easily “activatable�. In comparison, Bulgaria’s GMI beneficiaries account for 1.5 percent of the population only.  In view of worsening employment prospects overall in the wake of the economic crisis, it is worth reconsidering the GMI eligibility cuts to ensure that there remains a basic last resort social safety net for the poorest, while raising the ability of the employment offices to effectively support highly disadvantaged job seekers. The social safety net, children and social inclusion Child poverty remains a significant challenge for Bulgaria even after years of steady poverty reduction on the back of strong economic growth. A recent World Bank Living Standards Assessment of Bulgaria found that many children live in poverty, driven mainly by a substantially higher poverty risk of households with three of more children3. Much, though not all, of child poverty is concentrated among Roma households. In an effort to promote social inclusion, many countries around the world are introducing conditional cash transfers which combine income support to alleviate poverty today with incentives for children to attend school to enhance their life chances and reduce the risk of their poverty in the future. Consistently with this approach Bulgaria introduced a school attendance requirement for child allowance eligibility in 2003 and tightened its enforcement in 2006. While there is not rigorous impact evaluation evidence on the introduction of the conditionality, there is some evidence that it may have resulted in positive effects on school enrolment. In 2007, school enrolment among girls aged 6-14 was 7.6 percentage points higher among beneficiaries than among non-beneficiaries; for boys the corresponding figure is 1.7. 3 World Bank (2008), Bulgaria: Living Conditions before and after EU Accession 10 Table 3: School Enrolment is higher among child allowance beneficiaries School enrolment In percent of relevant age group Girls age Boys age All children Girls age Boys age All 6-14 6-14 age 6-14 15-18 15-18 children age 15-18 Non-beneficiary 90.6 92.1 91.3 73.6 74.7 74.2 Beneficiary of child allowance 98.2 93.8 95.8 73.3 94.4 84.8 Difference 7.6 1.7 4.5 -0.3 19.7 10.6 Source: Staff estimates based on MTHS 2007. Leveraging the safety net during the economic crisis There are visible signs that the global economic crisis is affecting economic growth and poverty reduction in Bulgaria. After a period of strong economic growth through 2008, growth in 2009 is projected to slow down to -3.5 percent due to the crisis, with important implications for poverty. Two channels have particular relevance for poverty in Bulgaria: (i) the labor market; and (ii) foreign remittances. Due primarily to its effect on trade and investment, the global economic crisis is likely to affect household welfare through contraction in the labor market and slowdown in wage growth, both of which are under way. Remittances from abroad, which are significant contributor to consumption of nearly 7 percent of the Bulgarian households, are likely to slow down in 2009. At the same time, opportunities for informal and casual employment – the informal safety net – are likely to become more limited in line with overall worsening of employment opportunities – there will be fewer informal employment opportunities and more individuals searching for such opportunities. Bank estimates suggest that the recent poverty reduction rend has been halted by the economic crisis and poverty incidence and depth may in fact increase in the coming two years4. Bulgaria’s social safety net building blocks to deal with the social impact of the economic crisis, but the system needs some adjustment in the short-term to be more effective in the face of greater needs and over the medium-term to achieve both effective poverty alleviation and active inclusion. While the bulk of the initial crisis response for those who lose their jobs in the wake of the crisis will come through active and passive employment programs, many of the previously jobless – often the most vulnerable and the poor – will rely on the social safety net. In the short term, the government may want to consider strengthening the GMI as a primary crisis response program for the poor by offering a menu consisting of (i) expanding coverage by reducing any barriers to access, requiring an in-depth review of current benefit application procedures and possibly qualitative analysis among potential beneficiaries. This includes simplifying eligibility conditions such as the requirement to having been registered unemployed for nine months prior to applying for the benefit as well as the bar to traveling abroad; (ii) introducing a separate and temporary GMI top-up benefit, while not permanently raising the eligibility threshold. This measure would allow addressing transient poverty risk for already 4 World Bank (2009), Bulgaria: Poverty Implications of the Global Financial Crisis 11 poor and vulnerable households related to the effects of the economic crisis, for example related to a loss of informal and casual employment opportunities; (iii) canceling the eligibility limitation to 12 months to restore the role of the GMI as a benefit of last resort program. The effectiveness of the limitation in terms of employment activation without a commensurate institutional change in the employment office appears to have been limited, and the economic crisis will further worsen employment prospects of low- skilled unemployed; (iv) continuing to offer public works programs through the employment office; and (v) close crisis monitoring of which tool is most responsive and attractive for beneficiaries. While overall fiscal pressures are mounting with falling revenue, there is fiscal space for expanded coverage of GMI. In general it is important to retain sufficient fiscal space for safety net interventions during the economic crisis. Lower actual than budgeted spending on GMI in recent years (see Table 4) suggests that there is sufficient fiscal space for greater coverage and a benefit top-up if needed, and good targeting efficiency suggests that additional resources allocated to the GMI program is money well spent. It is essential to protect this relatively small budget allocation. Table 4: GMI expenditure has been falling in nominal terms in recent years, GMI budget and expenditure BGN '000 2006 2007 2008 2009 budgeted 85,850 85,850 85,850 85,850 executed 83,480 65,759 38,217 Source: Social Assistance Agency Re-energizing the social safety net for the medium term In the medium-term, and after the immediate effects of the economic crisis abate, the government may want to consider a new strategy for the social policy system by taking a fresh look at objectives, purpose of instruments and tools. Issues to consider are (i) defining objectives in line with the EU concept of active inclusion (poverty reduction, activation and social inclusion including of the next generation), (ii) better linking benefits with services, e.g. employment counseling and purposeful re- qualification, to create a gateway towards sustainable employment and better life chances for the next generation, and (iii) improving interaction of benefits. This could lead to measures to restructure the social safety net to ensure greater effectiveness by (i) revisiting the employment activation framework to prevent unnecessary welfare dependency by focusing more attention on better linkage between cash transfers and activation services to create an individual pathway towards sustainable employment and on the ability and effectiveness of the employment offices to deal with highly disadvantaged job seekers, including through greater partnerships with NGOs and community-based organizations; (ii) reconsidering the role of the last resort GMI program. Such programs have a place in all advanced economies as providing adequate protection for people who are not able to find employment. Bulgaria’s well targeted GMI program has been shrinking for a number of 12 reasons outlined in this note and is at risk of disappearing. It is important to reconsider the role of the program and its design to ensure that an adequate basic safety net for the poorest Bulgarian individuals remains available, not least in periods of economic distress, and the considerable know-how and experience associated with managing the program is not lost. This includes a reconsideration of indexation of the GMI eligibility threshold that would balance fiscal concerns with the need to provide protection to the poorest Bulgarians, for example through defining the threshold in the context of relative rather than absolute poverty ; and (iii) further strengthening registries and monitoring systems, including tracking the number of applications for programs, both accepted and rejected and by reason for rejection. wb220293 C:\Users\wb220293\Documents\Bulgaria\HD Policy Notes\Bulgaria Social Safety Net Policy Note FINAL.docx 9/3/2009 11:18:00 PM 13