Report No. 46120-CZ Czech Republic Improving Employment Chances of the Roma October 17, 2008 Human Development Sector Unit Europe and Central Asia Document of the World Bank CZECH REPUBLIC Improving Employment Chances of the Roma CONTENTS Page Executive Summary.......................................................................................................... 1 1. Overview: Towards a New Approach to Improve Employment Chances of the Roma .................................................................................................................................. 4 Labor market profile of Roma in marginalized localities............................................... 6 The challenge: Barriers to Roma employment in the Czech Republic......................... 13 A dual framework for a solution: Employment activation and educational integration19 Tackling current joblessness: Employment activation for disadvantaged job-seekers ................................................................................................................................... 19 Preventing future joblessness: Improving Roma education and school-to-work transition ................................................................................................................... 31 A new approach to enhancing employment prospects of jobless Roma....................... 33 2. The Challenge: Access of Roma to Employment in the Czech Republic........... 42 Labor market outcomes for Roma in marginalized localities....................................... 42 Roma employment profile ............................................................................................ 47 Roma employed, unemployed and inactive: How do they differ?................................ 49 Skills and Roma Labor Market Outcomes................................................................ 51 Profile of employed Roma........................................................................................ 53 Profile of Roma jobseekers....................................................................................... 56 The interactions of Roma with labor and welfare offices............................................. 59 Conclusions................................................................................................................... 63 3. The Labor Market in the Czech Republic and Opportunities for low skilled workers............................................................................................................................. 65 Recent labor market trends and challenges................................................................... 65 Regional variations in labor market conditions ............................................................ 66 Demand for low skilled workers................................................................................... 68 Conclusions................................................................................................................... 73 4. Employment Incentives and the Social Welfare System ..................................... 74 The Czech taxation and social benefit system and recent policy changes.................... 76 Taxes......................................................................................................................... 76 Social Benefits .......................................................................................................... 77 Employment incentives in the tax and benefit system.................................................. 80 The impact of recent policy changes ........................................................................ 81 The impact of personal indebtedness........................................................................ 84 Conclusions and Policy Directions............................................................................... 86 5. Public Employment Services and Employment Activation ................................ 87 Employment activation ­ focusing on disadvantaged job-seekers ............................... 87 Employment service organization and methods ........................................................... 92 Labor market programs............................................................................................... 101 Conclusions and Policy Directions............................................................................. 106 6. Roma and Youth-Centered Employment Activation ........................................ 109 Education and school-to-work transition.................................................................... 110 Youth ­focused employment activation ..................................................................... 115 Conclusions and Policy Directions............................................................................. 117 Background Reports for this Study............................................................................. 120 References...................................................................................................................... 121 ANNEX .......................................................................................................................... 125 ACKNOWLEDGEMENTS This study is the result of a joint research project of the Government of the Czech Republic and the World Bank conducted at the initiative of H.E. the Minister of Human Rights and Minorities Dzamila Stehlíková. It was prepared by a World Bank team consisting of Christian Bodewig (lead author) and Jan Rutkowski and in close cooperation with Jaroslav Marousek from the Ministry of Labor and Social Affairs and Gabriela Hrabaová from the Office of the Government of the Czech Republic. The study builds on a number of background reports prepared by researchers from the Czech Republic and elsewhere. First, Stpán Jurajda and Jozef Zubrický (CERGE-EI, Charles University Prague and Academy of Sciences of the Czech Republic) prepared a report on employment incentives resulting from the taxes and benefits system. Second, a research team from the Institute of Ethnology of the Czech Academy of Sciences consisting of Zdeek Uherek, Tereza Pojarová, Renata Weinerová, Zuzana Korecká conducted focus groups with unemployed Roma as well as interviews with experts from labor offices, social welfare administrations and representatives from non-governmental organizations active in social inclusion work. Third, a research team from the Institute for Research on Social Reproduction and Integration, Faculty of Social Sciences, Masaryk University Brno, consisting of Michal Vasecka, Irena Kasparová, Michaela Bartosová, Lenka Slepicková, Helena Bartáková, Blanka Plasová, Ondej Hofírek, Michal Nekorjak, Kateina Dvoáková and Hana Dokoupilová, conducted structured interviews with employers on their experiences with and views on hiring Roma workers. Fourth, Frank Kavanagh (Consultant) prepared a background report on the public employment service's role, and Lucia Kurekova (Consultant) conducted detailed analysis of vacancy data. The report also presents data from a specially designed Labor Force Survey in marginalized Roma communities which was conducted for the purposes of this report in May 2008 by the team from the Institute of Ethnology of the Czech Academy of Sciences with support by Milada Horaková, Milos Kucera, Ondrej Nyvlt and Jii Vinopál. The study's preparation benefited from guidance of Minister Dzamila Stehlíková and Radka Soukupová (Office of the Government of the Czech Republic) as well as Orsalia Kalantzopoulos, Tamar Manuelyan Atinc, Arup Banerji and Gordon Betcherman (World Bank). The team is grateful for comments from peer reviewers David Grubb (OECD) as well as Dena Ringold and Wendy Cunningham (World Bank) and feedback from an advisory group formed for the purposes of this report, consisting of David Beák (Department of Social Welfare and Health, Prague 14 municipality) Renata Kottnerová (Roma Coordinator in Olomoucký county), Klára Laurencíková (Ministry of Education, Youth and Sports and Ministry of Labor and Social Affairs), Marek Podlaha and Jií Sandera (Department for the Social Integration in Roma Excluded Localities (Social Inclusion Agency)), Abe Stank (non-governmental organization ESTER), Ivan Veselý (civic association Dzeno), Hana Volfová (community center Ceského Svýcarska), Margita Wagener (Office of the Government) and Jana Zapletalová (Institute for the Pedagocial Psychological Consultations). EXECUTIVE SUMMARY Roma living in marginalized communities in the Czech Republic have not benefited from the buoyant labor market and are facing continued severe and multiple forms of labor market exclusion. While the magnitude of the challenge of overcoming social exclusion among Roma is well appreciated in the Czech Republic, there is less clarity about the extent of labor market exclusion of Roma, the exact nature of barriers to employment and solutions to improve their employment chances. Key findings Prepared jointly by the Government of the Czech Republic and the World Bank, this study reviews results of a special labor force survey conducted in marginalized localities where many Roma reside. It offers a fresh picture of the depth and extent of labor market barriers facing Roma ­ not all Roma in the Czech Republic, but those residing in known marginalized localities. Moreover, it reviews the extent of labor demand for low skill occupations across the Czech Republic and assesses employment incentives resulting from the social welfare system as well as the readiness of the Czech Labor Office to provide effective support to disadvantaged job-seekers such as Roma. The study arrives at the following, key findings: · The majority of working age Roma residing in marginalized localities are not unemployed but are out of the labor force. The main reason for this is discouragement: Many jobless Roma have largely given up looking for jobs. There is also a strong gender dimension, with Roma women participating substantially less than men. · The level of education attainment and skills acquired is a key predictor of labor market success of Roma. Those with post-primary, mainly vocational, education fare substantially better in the labor market than those with only primary or even special schooling. However, even those Roma that are employed often work in short-term jobs and with very low wages, with even the top paid Roma workers earning less than the national average wage. · Roma in marginalized localities suffer from severely low educational attainment and a widespread lack of functional literacy. Early school leaving looms large: 60 percent have only primary education and 14 percent attended special school for children with learning disabilities. As a result, more than 40 percent have been found functionally illiterate, with another 40 percent having only limited literacy and numeracy skills. 1 · There is no upward educational mobility between generations, if anything the level of educational attainment among Roma appears to have been worsening since the start of the transition. There is evidence that only a minority of children with parents with primary education have received post-primary education. Moreover, more than half of children with parents with post-primary education only received primary education. · The demand for low or unskilled labor is very low across the Czech Republic, especially in regions with many marginalized localities, and Roma often compete with foreign workers. The only exception is Prague, yet few Roma workers have the resources to commute to work there. This suggests that unless their skills constraints are addressed, employment chances for Roma will remain limited and likely worsen over time. · Widespread indebtedness of Roma households in marginalized localities acts as a barrier to formal employment. Once formally employed, a sizable share of their wage will be subject to debt collection, making informal employment and welfare receipt a more obvious coping strategy. Participation in the labor market of affected individuals will remain low unless they receive support in overcoming indebtedness. It should be considered part of social and employment activation services. · Many Roma rely on social welfare benefits whose levels have until recently been high relative to wages for low and unskilled occupations, acting as a barrier to labor market participation. This is a particularly binding barrier in low wage regions of the Czech Republic where demand for low-skilled workers is low and where many socially excluded Roma reside. · While many unemployed Roma are turning to the Labor Offices for support in their job search, there is a substantial fraction who does not receive support such as job search assistance or retraining. Moreover, there is evidence that few of those who participate in training programs actually complete them and that few of those who complete them find employment afterwards. Policy directions Labor market exclusion of the Roma in the Czech Republic has been driven by unsatisfactory educational outcomes due to enrollment in special schools for children with learning disabilities as well as early school leaving, condemning many Roma to joblessness and reliance on relatively generous social welfare benefits. Overcoming this legacy will require a two-pronged approach: (i) developing effective employment activation policies for jobless Roma adults, in particular with a focus on skills upgrading and second-chance education, and (ii) effective preventive measures for Roma youth to promote access to and retention in quality education and reduce early school leaving. 2 First, a successful strategy to tackle current joblessness among the Roma in the Czech Republic suggests the need for a new employment activation approach, focusing more attention of the Labor Offices on disadvantaged job-seekers. Experience from across the OECD and the EU suggests that this should involve a mutual balance in obligations between the job-seeker and the Labor Office, with raised job search incentives for the job-seeker on the one hand and enhanced capacities of the Labor Office to effectively support disadvantaged job-seekers on the other. In a first step, the Government of the Czech Republic has already introduced measures to reduce benefits and tighten conditions for inactive long-term unemployed so as to raise job search incentives. However, this study argues that, with very low demand for low or unskilled labor and widespread indebtedness of Roma which acts as a binding barrier to choosing formal employment, this tightening of beneficiaries' incentives alone will not suffice to enhance their employment chances. In addition it requires support by the Labor Office, delivered in a culturally sensitive and trustful manner either directly or through contracted qualified service providers such as NGOs, with individualized activation services addressing the multiple barriers to employment. The study finds that further reforms in the structure and policies of the Czech Labor Office are necessary to raise its capacity to provide effective support to long-term unemployed and disadvantaged job- seekers such as Roma. Second, subsidized employment with skills upgrading and second-chance education programs will be key to enhance employment chances for Roma. Given the dramatically low levels of skills among many Roma in marginalized communities and low demand for elementary occupations, subsidized public works and community employment will likely remain an important activation tool ­ but requires inbuilt systematic skills upgrading. Moreover, closing the skills gap will also require enhanced second-chance education interventions as part of activation policy to raise literacy and numeracy skills. Third, preventing future joblessness will require a focus on children and youth and prevention of future labor market and social exclusion, and the time to act is now. A sizable share of the Roma population in marginalized localities is below the age of 25. Urgent efforts need to be taken to overcome disadvantage resulting from low quality schooling and early drop out. This involves interventions from early childhood through primary education towards extending years of schooling for Roma children and youth. Implementing a new approach to improving employment chances for the Roma would best build on pilot testing, evaluating and scaling up new tools and methods to ensure good use of tax payers' money and enhance the effectiveness of measures in improving employment chances. The availability of funding from the European Union through the Operational Program Human Resources Development allows the Labor Office to run pilot projects to test new tools and solutions and evaluate them carefully before making them element of formal policy. Moreover, the recently launched Social Inclusion Agency, set up to pilot new localized social inclusion approaches and partnerships, is a unique opportunity to test, evaluate and scale up innovative solutions. 3 1. OVERVIEW: TOWARDS A NEW APPROACH TO IMPROVE EMPLOYMENT CHANCES OF THE ROMA Roma in marginalized localities in the Czech Republic have not benefited from the recent improving employment opportunities in the Czech labor market. Employment among Roma is low and labor market participation limited, often driven by lacking labor market opportunities. The labor market status among the Roma is strongly driven by educational attainment and skills, and the vast majority of Roma in marginalized communities suffer from low attainment and lacking functional literacy and numeracy skills. As a result a majority rely on traditionally generous social welfare benefits to make ends meet. In an effort to reduce this welfare trap the Czech Government has introduced measures to reduce benefits and tighten conditions for long-term unemployed who are inactive. However, with very low demand for low or unskilled labor and widespread indebtedness of Roma which acts as a binding barrier to choosing formal employment, this tightening of beneficiaries' incentives alone will not suffice in enhancing their employment chances. It requires effective interventions by the employment services; yet in its current set-up the Czech Labor Office appears not well placed to provide effective support to long-term unemployed and disadvantaged job-seekers such as Roma. A new approach to improving job chances for socially excluded youth and adults is necessary, involving a new way of engagement through the Labor Office and contracted third sector service providers and with integrated activation services addressing multiple barriers to employment such as skills deficits, lack of child care, indebtedness and others. However, given the large skills gap of Roma and the receding demand for elementary skills in the labor market, the key long-term strategy to prevent Roma joblessness has to focus on improving educational outcomes for Roma. 1.1 The Czech labor market has been performing strongly in recent years, and high labor demand has driven down unemployment to record low levels. Yet the overall positive developments conceal deeply unsatisfactory outcomes for a relatively small group of disadvantaged job-seekers who fail to find work. Unemployment has fallen to below 5 percent in early 2008 and the employment rate is approaching the Lisbon target of 70 percent by 2010. Most remaining unemployment in the Czech Republic is now of a long-term nature, suggesting that the remaining unemployed face binding barriers to employment possibly due to lacking skills and work habits, disincentives or lack of motivation to look for work or other reasons such as discrimination. Promoting further increases in employment will require addressing labor market barriers of disadvantaged job-seekers. 1.2 Roma residing in marginalized localities have been identified as a group that is disproportionately disadvantaged in the labor market. While this comes largely as 4 a result of low educational attainment, there is a sense that the traditional system of employment policies has proven unable to promote access to the labor market for Roma. This study was prepared at the request of and in close collaboration with the Government of the Czech Republic. It aims to explore the drivers of unsatisfactory labor market outcomes among Roma in marginalized localities in the Czech Republic and lay out an agenda for a systemic solution to promote employment among jobless Roma. The study builds on a specially designed Labor Force Survey, including a literacy skills assessment, which was carried out in 12 marginalized Roma localities in May 2008 (see Box 1). It also reviews regional patterns in labor demand for low-skilled workers, employment disincentives resulting from the tax and benefit system for low-wage workers and assesses the readiness of the Czech Labor Office in dealing effectively with disadvantaged job-seekers. Box 1: The 2008 Labor Force Survey of Roma in marginalized localities of the Czech Republic The data on Roma employment presented in this study are from a specially designed Labor Force Survey conducted in May 2008 in 12 marginalized localities where many Roma reside, six of them in Bohemia and six in Moravia, with ten towns and two micro- regions. The sample of surveyed localities was drawn from a list of marginalized localities assembled by the Office of the Government of the Czech Republic to guide the work of the Agency on Social Inclusion established in 2008. The survey is, therefore, not representative of the Roma community in the Czech Republic as a whole, but representative of those Roma residing in identified marginalized localities. Therefore, when referring to "Roma" this report implies those Roma who reside in known marginalized localities. The surveyed localities are mainly, yet not exclusively, inhabited by Roma, and the Roma population shares vary. The survey, therefore, also covered a limited number of non- Roma residents of excluded communities. Although their number was small, in some cases the comparison between Roma and non-Roma workers was still possible. Roma were identified using an answer to the following question: "This is a survey of the Roma community. Do you consider yourself Roma?". According to this self-identification criterion, there were 1050 Roma in a total sample of 1150 individuals. Although the overall sample size is not small, there were instances of wide error margins in responses to some questions due to small sub samples, e.g. the unemployed. Cases of wide error margins are indicated in the report. The questionnaire follows a standard Labor Force Survey structure and captures all aspects of labor market experience of surveyed individuals. In addition, it includes a simplified skills assessment to assess actual levels of functional literacy and numeracy skills. 5 LABOR MARKET PROFILE OF ROMA IN MARGINALIZED LOCALITIES The primary challenge for promoting Roma employment is overcoming a lack of labor market participation... 1.3 Roma in marginalized localities in the Czech Republic suffer from highly unsatisfactory labor market outcomes, and it is lack of labor market participation (termed inactivity1) rather than unemployment that is the biggest problem. As many as 56 percent of Roma of working age (15-64) are out of the labor force, that is neither employed, nor actively looking for a job (Figure 1, left hand panel). The unemployed, i.e. those actively looking for employment, account for only 5 percent of the working age population. Close to 40 percent are employed, however 12 percent have only casual, non-regular jobs2. Figure 1 The majority of working age Roma in excluded communities are out of the labor force and discouraged Roma population in working age by labor force status, 2008 Reasons for not working by gender, 2008 100% 90% 26.9 80% 70% 60% 50% 40% 30% 55.8 20% 10% 12.2 0% Men Women 5.2 Maternity Education/training Disability, retirement Employed Casually employed Unemployed Not in labor force No jobs available Other Note: Population of working age (15-64). Source: Excluded Roma Labor Force Survey 2008; Bank staff calculations. 1.4 The primary reason for the lack of labor market participation is the scarcity of job opportunities, and many inactive are discouraged workers. Roughly three in four men and one in two women say they are out of the labor force because they could not have found a job (Figure 1, right hand panel).3 This means that many of the Roma who are out of the labor force are actually so called discouraged workers, i.e. the unemployed who ceased searching for a job once their efforts proved futile4. 1Inactivity is the technical term used to denote the status of being outside the labor force. 2Casually employment is defined as a situation in which the survey respondent declared that he/she does not have a job, but reported income during the reference week. 3It should be borne in mind that these results reflect the subjective perception of labor market conditions. 4Discouraged workers are defined here as those who did not actively look for a job in the reference week (and thus were categorized as inactive rather than unemployed) but reported the duration of their job search. 6 ...mainly driven by lacking skills to compete in the labor market... 1.5 The educational attainment of Roma in excluded communities is very low, as are actual literacy and numeracy skills. Eight out of ten Roma of working age have no more than primary education (Figure 2, left hand panel)5. There is a substantial proportion ­ 14 percent ­ of persons who graduated from special schools for children with special educational needs ­ a specific feature of Roma schooling in the Czech Republic for decades. This large group has no formal certified vocational skills. Only two Roma in ten have some formal vocational training or secondary education. Actual literacy and numeracy skills are very low, too. As many as 44 percent of working age Roma in excluded communities can be considered functionally illiterate (Figure 2, right hand panel). Another 44 percent have only some basic literacy and numeracy skills. Only 12 percent can be considered as functionally literate, i.e. able to answer most of the relatively simple questions that require primary school-level knowledge.6 Figure 2 The vast majority of Roma have no more than primary education and little skills Distribution of Roma working age population by education, 2008 Distribution of Roma working age population by skill level, 2008 18% 12% 3% 5% 44% 60% 14% 44% Level 0 Level 1 Level 2 Less than primary Special Primary Basic vocational Secondary Note: Level 0 = little or no literacy/numeracy skills Level 1 = some literacy and numeracy skills Level 2 = basic literacy/numeracy skills (primary school level) Source: Excluded Roma Labor Force Survey 2008; Bank staff calculations. 1.6 Moreover, the inter-generational upward educational mobility is extremely limited among Roma, and there is evidence for a worsening of educational attainment. Only 21 percent of sons of fathers who have primary education or less received post-primary education (mainly basic vocational training). Even more worrying, more than half of the sons (54 percent) of fathers who have post-primary education 5This report uses the Czech terminology in which primary education refers to basic, compulsory education with duration of nine years, usually from age 6-15. This includes the internationally used primary (ISCED 1) and lower secondary (ISCED 2) levels. Basic vocational education refers to ISCED 3c, while secondary education refers to ISCED 4 (upper secondary education). 6The results were obtained using a simple skill assessment survey conducted as part of the Labor Force Survey in marginalized localities. 7 received less education.7 Thus, there is very little upward educational mobility among Roma but substantial downward mobility. This suggests that the transition might have had a negative effect on educational prospects of many Roma. The generation of fathers who acquired their education under the old regime tends to be better educated that the generation of sons who received their education after the transition to a market economy ­ a finding confirmed by employers interviewed for this study8. 1.7 Despite poor formal education many Roma workers obtained job-specific vocational skills after completing school, typically in employment. But gender differences are pronounced. While six in ten men obtained some vocational skills, only 4 in 10 women did so. Expectedly, the acquisition of vocational skills critically depends on employment. Among those who were employed in the last 3 years the proportion of workers with vocational skills (obtained mostly through on-the-job training) approaches 60 percent, while among those who were not employed it is slightly less than 40 percent. Thus employment matters for the acquisition of skills and, through this channel, is likely to increase future employment chances. ... resulting in dramatically worse employment outcomes for Roma compared to non- Roma... 1.8 Few Roma in marginalized localities are employed. Only 27 percent of working age Roma are employed, compared with the national average of 66 percent (Table 1). An additional 12 percent are only casually employed. This employment gap is dramatic: As many as 27 percent of Roma would need to find employment in order to raise the employment rate among Roma to the national average. However, the gap results largely from the low labor force participation by Roma rather than from high unemployment.9 Only 44 percent of working age Roma participate in the labor force compared with 70 percent of Czechs overall. Thus many Roma are detached from the labor market. The main policy issue to be addressed is thus discouragement and non- participation, rather than ineffective job search. The situation differs by different population groups: · Prime-age Roma women: The low employment rate among Roma is largely accounted for by labor market detachment of prime-age women. The labor force participation rate by Roma women is half that of the Czech women overall: 31 against 62 percent (Table 1). This is likely to reflect cultural factors (the role of women in Roma society) and a lack of job opportunities, but poor access to services such as nursery and kindergarten may also play a role. Indeed, utilization of kindergartens among Roma children appears low, 7A very similar pattern emerges when on compares educational attainment of daughters with that of their mothers. 8Vasecka et al. (2008) 9This is because, as demonstrated earlier few Roma are actively looking for jobs and thus are categorized as economically inactive rather than unemployed. In order to be categorized as unemployed persons who have no job need to meet two additional conditions: (a) be actively looking for a job, and (b) be able to take a job if offered. 8 with only 36 percent of Roma children in households with children in pre- school age going to kindergarten or zero classes. · Prime-age Roma men are also much less likely to be economically active than the prime-age Czech population overall. The labor force participation rate by prime age Roma men at 63 percent ­ although significantly higher than that of women ­ compares unfavorably to the national average of 95 percent. To what extent this low economic activity of prime-age Roma men reflects poor skills, lack of demand or discrimination is not quite clear. · Roma youth and older workers: In sharp contrast to prime-age workers, Roma youth and older workers are more, not less, economically active than their ethnic Czech counterparts. 42 percent of Roma youth (aged 15-24) are part of the labor force, compared with the national average of 32 percent (Table 1). This large difference is explained by much higher rates of educational enrollment of young Czechs overall than that of young Roma. For older workers (aged 55-64) the difference in labor force participation rates is smaller (2 percentage points in favor of Roma). The likely cause is that in contrast to the Czech older population overall many (if not most) older Roma lack pension insurance and thus have an incentive to seek employment. Lastly, the high labor force participation rates of young and older Roma go hand in hand with high unemployment rates. Table 1 Indicators of labor market outcomes: Romaa) against the national average. Unemployment Labor force Employment/population rate participation rate ratio Czech Czech Czech Roma Roma Roma Republic Republic Republic Population of working age b) 11.7 5.4 44.2 69.9 39.1 66.1 Gender Men 10.3 4.3 61.3 78.1 55.0 74.8 Women 14.0 6.8 30.5 61.5 26.3 57.3 Age 15 ­ 24 (youth) 24.7 10.7 42.4 31.9 31.9 28.5 25 ­ 54 (prime-age) 7.8 4.8 44.2 87.8 40.7 83.5 55 ­ 64 (older) 15.2 4.6 50.0 48.2 42.4 46.0 a) Roma in excluded communities covered by the Roma Labor Force Survey 2008. b) 15 ­ 64 Note: data for the Czech Republic refer to 2007, data for Roma refer to May 2008. Source: Roma Labor Force Survey 2008 and Eurostat; Bank staff calculations. 1.9 Roma fare worse in the labor market compared to other low-skilled and socially excluded Czechs, in particular in terms of labor force participation. The data presented in Table 1 compares labor market outcomes for Roma against national averages. However, it appears that Roma are also faring slightly worse even compared to 9 low skilled adults in the Czech Republic on average. The labor force participation rate among Roma aged 25-64 residing in marginalized localities in 2008 was 44.7 percent and the employment-to-population ratio 40.9 percent. As a comparison, in 2007, low skilled Czechs between the age of 25 and 64 faced a labor force participation rate of 56.4 percent and an employment -to-population ratio of 45.7 percent.10 This suggests that there are Roma-specific barriers to employment in addition to skills. Moreover, they also fare worse compared to their immediate non-Roma neighbors: Table 2 presents a comparison of the labor force status between Roma and non-Roma in excluded communities. The non-Roma control group is small, with only 103 persons, suggesting that the estimates are subject to a wide margin of error. With this caveat in mind, the key difference between the two groups is not in the unemployment rate, but in the labor force participation rate. Working age Roma are less likely to look for jobs than their non- Roma counterparts in excluded communities. Whether this reflects discouragement due to discrimination or other factors is unclear. Table 2: Labor force status: Roma vs. Non-Roma in marginalized localities in percent Non-Roma Roma Unemployment rate 12.5 11.7 Labor force participation rate 54.4 44.2 Employment/population ratio 47.6 39.1 Source: Roma Labor Force Survey 2008; Bank staff calculations. Population of working age (15-64) residing in marginalized localities ...but with marked heterogeneity driven by skills and labor market experience 1.10 There is a notable difference among the employed, the unemployed and those outside the labor force with respect to educational attainment and skills. Figure 3 shows the difference between the employed, the unemployed and the inactive along several socio-demographic dimensions. The employed and the unemployed are predominantly men, while the inactive are predominantly women (Figure 3, Panel A). The employed are more often prime age workers whereas the unemployed and the inactive are more often younger persons (Figure 3, Panel B). Overall, the employed are better educated than the unemployed, who in turn are better educated than the inactive (Figure 3, Panel C). Stronger attachment to the labor market goes hand in hand with better educational attainment. The employed have also better skills than the unemployed and the inactive. Roma with literacy and numeracy skills account for over 70 percent of the employed and only for around 50 percent of the unemployed and the inactive (Figure 3, Panel D).11 Accordingly, while functional illiteracy is widespread among the unemployed and the inactive, it is relatively rare among the employed. 10There may be a bias in the comparison, as the national averages for low-skilled workers may include Roma. 11Based on the results of the skills assessment, which was part of the Roma Labor Force Survey, (see Annex 1) 10 Figure 3 The employed, the unemployed and the inactive differ widely by education and skills Panel A Panel B Gender Profile Age Profile 100% 100% 6 11 6 80% 41 35 80% 69 60% 60% 61 80 72 40% 40% 59 65 20% 31 20% 29 14 21 0% 0% Employed Unemployed Not in labor force Employed Unemployed Not in labor force Male Female 15 - 24 25 - 54 55 - 64 Panel C Panel D Educational Profile Skill Profile 100% 100% 19 13 28 80% 39 80% 53 47 60% 60% 60 64 40% 57 50 40% 35 40 20% 20% 21 24 11 15 0% 12 13 0% Employed Unemployed Not in labor force Employed Unemployed Not in labor force Special- Primary Vocation+ Level 2 Level 1 Level 0 Note: The unemployed include casually employed. Source: Roma Labor Force Survey 2008; Bank staff calculations. 1.11 Vocational training greatly increases Roma employment chances, while special education condemns Roma to unemployment. Among the Roma who received vocational training 50 percent are employed. This is more than double of those who received primary education and over three times more than among those who received special education or less (Table 3). Stated otherwise, Roma who have special education are two times more likely to be inactive than Roma who received vocational training. This shows that the type and level of education received has a decisive impact on Roma employment prospects. Roma who are better educated and have vocational skills fare much better on the labor market than those who are poorly educated and lack vocational skills. Obviously, this finding has far-reaching policy implications. Better education and the provision of vocational skills are key to improving Roma labor market outcomes. 11 Table 3 Labor force status by educational attainment Basic vocational Special or less Primary Labor force status traininga) Percent Employed 15.2 22.6 50.3 Unemployedb) 18.5 17.6 15.5 Not in labor force 66.3 59.8 34.2 Note: Roma population of working age (15-64) a) Including secondary b) Including casually employed Source: Roma Labor Force Survey 2008; Bank staff calculations. 1.12 In addition to formal education, actual skills matter, too. Among those who are functionally illiterate (skill level 0) few are employed (only 17 percent) and a majority are unemployed or inactive (Table 4). Literacy and numeracy skills (levels 1 and 2) increase the probability of employment by a factor of two.12 Even if other factors ­ such as discrimination or job attitudes play a significant role ­ better education and thus better skills help to overcome Roma labor market disadvantage. Table 4 The more skilled are substantially more likely to be employed Labor force status by skill level, 2008 Labor force status Level 0 Level 1 Level 2 Percent Employed 17.3 34.8 29.9 Unemployeda) 21.6 14.1 15.4 Not in labor force 61.1 51.1 54.7 Note: Roma population of working age (15-64) Level 0 = little or no literacy/numeracy skills Level 1 = some literacy/numeracy skills Level 2 = basic literacy/numeracy skills a) Including casually employed Source: Roma Labor Force Survey 2008; Bank staff calculations. 1.13 There is a two-way relationship between skills and labor market outcomes. For example, those Roma who work are likely to have an easier access to internet or to newspapers. Employment provides earnings and thus allows one to own a car. So, causality runs in both directions: better skills improve employment chances and being employed is conducive to acquiring additional skills. Accordingly, this analysis may overestimate the effect of skills on employment. Still, there is no doubt that there is a close relationship between skills and employment among Roma and thus supporting investment in human capital and facilitating labor market entry are key policies to overcome their labor market and social exclusion. 12There is not much difference between the employment chances between Roma with skill levels 1 and 2. If anything, Roma with skill level 1 seem to fare on the labor market somewhat better than those with skill level 2. But this may be due to the small sample size and/or imprecise measurement of the actual skill level. 12 1.14 The employed differ from the unemployed and the inactive also in terms of their labor market experience. As expected, the employed are more likely to have a previous employment record. Only a minority of the inactive (27 percent) had previous job experience, compared to over half of the unemployed (56 percent) and a vast majority of the currently employed (79 percent). Thus previous labor force status is a good predictor of the current labor force status ­ a phenomenon known as path dependence: The employed reap the benefits of past employment (which was likely to be associated with skill acquisition, greater motivation, etc.), whereas the unemployed and still more so the inactive are often locked in the joblessness trap (associated with the erosion of skills and morale). 1.15 While most jobs of employed Roma are short-term and with very low pay, there is substantial variety in employment conditions between Roma. In fact, there is evidence of a dual market for Roma labor. While some Roma have better, more secure and better paid jobs others have only casual, precarious and often badly paid jobs. · The vast majority of Roma work as unskilled workers, most of them in manufacturing and construction (57 percent of all employed Roma), others in services (16 percent) and in agriculture (6 percent). Only one Roma in five works as skilled worker, mostly in construction. · The average job tenure of Roma workers is short indicating high turnover and job instability. Nearly 50 percent of all Roma workers have been holding their jobs for less than a year. At the same time, however, there is still a substantial fraction of Roma workers who hold long-term jobs. The average job tenure is 3.5 years (thus significantly higher than the median of just 12 months). · Roma earnings are very low and there is high wage inequality between Roma. The average Roma worker earns some 40 percent of the national wage and slightly more than the minimum wage .13 Those who have regular jobs earn somewhat more at about 35 percent above the minimum wage, but still only about half the national average. There is also high inequality of earnings, driven primarily by extremely low wages of Roma workers at the bottom of the wage distribution. THE CHALLENGE: BARRIERS TO ROMA EMPLOYMENT IN THE CZECH REPUBLIC 1.16 Employment chances of Roma in the Czech Republic are negatively affected by a range of barriers both on the demand and supply side. Barriers on the demand side 1.17 The biggest demand side barrier to employment of Roma residing in marginalized localities is low demand for elementary occupation workers in the 13The ratio of Roma earnings to national average (minimum) wage is most probably underestimated. 13 Czech Republic. In general, there is a strong demand for high and specialized ­ both white collar and blue collar ­ skills, coupled with a relatively weak demand for lower and more generic skills. Figure 4 presents the ratios of unemployed to vacancies by occupation. Job opportunities for less skilled workers, i.e. in elementary occupations, are markedly worse than those for more skilled workers. Figure 4: Demand for less skilled and elementary occupations is weak Employment-vacancy ratios by occupations, 2007 (second quarter) Unemployment - Vacancies Ratio 8.0 6.0 4.0 2.0 0.0 1. Legislators, 2. Professionals 3. Technicians 4. Clerks 5.Service 6. Skilled 7. Craft and 8. Plant and 9. Elementary U-V ratio 1.9 2.4 2.5 6.2 4.8 4.2 1.2 1.1 5.4 U-V ratio Source: Ministry of Labor and Social Affairs. Bank staff calculations 1.18 Job opportunities for low skilled are particularly bad in traditional high unemployment regions where many socially excluded Roma reside, although there is some variation at the sub-regional level. Figure 5 (upper panel) shows that the ratio of unemployed to vacancies ("U/V ratio") for elementary occupations is dramatically high in those regions which traditionally have had high unemployment. This is notable given the fact that the tightening labor market in the Czech Republic has also triggered a reduction in unemployment in those high unemployment regions. However, the large U/V ratios in high unemployment regions also hide substantial variation at the sub- regional level. The lower panel of Figure 5 presents a breakdown of U/V ratios for different counties in Ustecky kraj, and a similar picture can be found across other high unemployment regions (see Annex). Identified Roma localities can be found both in counties with high and low demand for low-skilled workers, suggesting that the local employment conditions vary for Roma workers across the Czech Republic, but also across the high unemployment regions where many Roma reside. Lastly, Prague stands out in terms of substantial excess demand for elementary education workers. 14 Figure 5: Job prospects for low education workers are particularly bad in high unemployment regions, although there is variation at the sub-regional level Elementary occupations: U-V ratio 50.0 40.0 30.0 20.0 10.0 0.0 ja j ý ý j ký ah kr ýkcets skze ajrk ajrk kra jark jark kra jark Pra slo j ýks ý cke ckedar jark ja vsaro ja anicos kr Vy kra kr kýsec ýksec Ú skvaro arvolra erbiL yksnílZ ýkcuo mo éhv ýkcibudr álor mohiJ hoiJ ýksnez do Pl K Ol Pa K retS M Ústecký kraj: Elementary occupations: U/V ratio Ustecky kraj: Unemployment by duration 120.0 100% 100.0 80% 80.0 60% 60.0 40% 40.0 20% 20.0 0% 0.0 Most Teplice Chomutov Ústí nad Louny Dcín Litomice Teplice Dcín** Chomutov Ústí nad Most Louny Litomice Labem Labem 12 + months 6 - 12 months 1 - 6 months Source: Staff calculations based on MoLSA data for Q2, 2007.Note: A starred county indicates that less than 50 vacancies were registered there, and therefore a cautious interpretation of the results is needed 1.19 Roma jobseekers are often rejected by employers, but substantially less so if they have higher levels of education and previous work experience. As many as 64 percent of the unemployed, casually employed and discouraged workers were rejected by the employer when applying for a job. As compared to that, only 31 percent of the currently employed Roma previously experienced a rejection by an employer (Figure 6 Panel A). The experience of the unemployed Roma is apparently quite different from that of the employed. Moreover, Roma jobseekers with special education or less are by one-third more likely to be rejected by the employer than those with vocational or secondary education (Figure 6, Panel B). 15 Figure 6 Labor market experience and better education lessen the risk of the rejection of a Roma jobseeker by the employer Panel A Panel B Roma job seekers rejected by employer by labor force status, Roma job seekers rejected by employer by edcuation, 2008 2008 60 80 70 50 60 tnecr 50 tnecr 40 30 40 pe pe 30 20 20 10 10 0 0 Special or less Primary Vocational or Employed Casually Unemployed Discouraged secondary employed Source: Roma Labor Force Survey 2008; Bank staff calculations. 1.20 Interviews with employers conducted for the purposes of this study confirm the predominance of lacking skills as an explanation for rejection of Roma job- seekers. In the interviews, employers reported a number of barriers to employing more Roma, involving the interplay of low educational attainment and missing skills among Roma, lacking work motivation, strong competition from foreign workers and a higher perceived riskiness of hiring Roma vis-ŕ-vis non-Roma workers14. 1.21 To what extent is there discrimination among employers against Roma? The available data does not allow a conclusive answer. In order to rigorously test the hypothesis of discrimination one would need to have a comparison group of non-Roma workers with similar skills and other relevant characteristics. While systematic discrimination likely plays some role in explaining the labor market outcomes among Roma, the data suggest that lacking skills and prior work experience are an even more binding barrier. After all, the likelihood of rejection of Roma by employers is not uniform, but varies with education and skill level. Employers interviewed for the purposes of this report highlight skills as a major detriment to hiring Roma, but also allude to communication barriers and mistrust which raise the perceived risk and costliness of hiring Roma compared to non-Roma. This suggests that employers may demonstrate a form of "statistical discrimination" related to information asymmetries and lacking information about individuals. Barriers on the supply side 1.22 The main barrier to employment for most Roma residing in marginalized localities is the dramatically low educational attainment and lacking skills. The 44 percent of surveyed individuals with little or no literacy and numeracy skills, and even the additional 44 percent with some basic literacy and numeracy are hardly employable in long-term employment even for basic tasks. Chances are better for the few with basic vocational or secondary education ­ and many of those have already found employment. 14Vasecka et al. (2008) 16 1.23 A large share of Roma are on social welfare benefits which are relatively well targeted at those who are without a job and thus do not have a source of earnings. 65 percent of working age Roma residing in marginalized localities are recipients of social welfare benefits. As Figure 7 shows, some 75 percent of benefit recipients are either inactive or unemployed. Additional 11 percent of the recipients are only casually employed. But still 15 percent of the recipients have regular jobs. More importantly, however, benefit receipt and labor force status are not independent. Those who do receive benefits have weaker motivation to look for a job, a phenomenon known as the "inactivity trap". Thus although benefits are targeted at the jobless, some of them are jobless because they claim welfare benefits. Figure 7 It is mainly the inactive and the unemployed Roma who receive social benefits, but some of the employed receive benefits, too Roma social welfare beneficiaries by labor force status, 2008 15% 68% 11% 6% Employed Casual job Unemployed Not in labor force Source: Roma Labor Force Survey 2008; Bank staff calculations. 1.24 Social welfare benefits received by Roma are relatively high, aggravating the risk of an inactivity trap. The average benefit reported by surveyed Roma accounts for close to 90 percent of the minimum wage, and the median benefit accounts for 75 percent (Table 5). Given that majority of Roma workers earn around the minimum wage this is likely to create significant labor supply disincentives, as long as benefits are means tested. That is, compared to when the benefit is withdrawn, the average Roma gains nothing in monetary terms when taking low-paid formal job. This is a classic case of the inactivity trap. Formal employment does not pay-off and this gives rise to welfare dependency. Table 5 Social welfare benefits received by Roma are high relative to the minimum wage, Summary of the distribution of social assistance benefits awarded to Roma, 2008 Benefit amount As a percentage of minimum wagea) Average 88.8 Top decile 168.8 Median 75.0 Bottom decile 25.0 a) The statutory minimum wage is equal to the median wage of Roma workers (including casual). Source: Roma Labor Force Survey 2008; Bank staff calculations. 17 1.25 Heavy indebtedness is another major barrier to formal employment. One Roma in two report being indebted, and the debt amount tends to be large. On average, indebted Roma owe 9.3 times the minimum wage to the lending institution. The average reflects heavy indebtedness concentrated among a relatively small fraction of Roma. The median debt is substantially lower, and accounts for 2.5 times the minimum wage. Still, for a considerable fraction of Roma workers formal employment would imply a loss of earnings due to debt repayment. While the survey of marginalized localities did not capture the sources of indebtedness among Roma, previous analysis points toward the widespread availability of small and unsecured but high interest loans for non- creditworthy individuals as well as instances of usury in marginalized Roma localities15. 1.26 Under circumstances of heavy indebtedness, informal employment is a coping mechanism to protect one's earnings. Debt collectors can only enforce debt collection provided an indebted individual earns formal income in excess of the minimum living standard (MLS) which is set by law as the guaranteed minimum income of an individual. So long as an indebted individual remains on social welfare benefits which guarantees income to the level of the MLS, debt collectors cannot easily enforce the debt. In effect, even when a formal sector job pays more than the combination of social welfare benefits and the minimum subsistence level, an individual faces little incentive to take that job, because of the threat of debt collection. Figure 8 shows a graphic presentation of the labor supply disincentive effect of debt and debt collection. Figure 8: Personal indebtedness acts as a key formal labor supply disincentive CZK debt repayment informal income MLS Benefit income Wage income Note: This is a simple graphic representation not based on actual data 1.27 Few Roma workers commute to work. Three out of four Roma workers do not commute to work, which means that they work in their immediate neighborhood. The low proportion of those who do commute (only 25 percent) suggests that a lack of affordable transportation may be an important barrier to Roma employment. In fact, remote job location is an important reason Roma jobseekers turn down job offers. Being unable to commute, many Roma have access only to a small, strictly local job market, which naturally limits their job chances. 15Uherek et al (2004) 18 A DUAL FRAMEWORK FOR A SOLUTION: EMPLOYMENT ACTIVATION AND EDUCATIONAL INTEGRATION 1.28 Roma joblessness is strongly driven by lacking skills and unsatisfactory education outcomes. Improving employment chances for the Roma will therefore require a dual approach to tackle current Roma joblessness through employment activation and labor market programs focusing on skills upgrading, while preventing a repeat in the future by overcoming educational disadvantage of Roma children and youth. This section focuses on approaches to promote Roma employment through employment activation policies as well as break the cycle of social exclusion through early childhood intervention and educational integration. Tackling current joblessness: Employment activation for disadvantaged job-seekers 1.29 In designing a strategy to enhance employment chances of the Roma, the Czech Government can benefit from experience across countries in the OECD and the EU that 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 · Provision of income support as well as access to public employment services (PES) with improved and more focused service for the disadvantaged unemployed and 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. It also involves the adjustment of social welfare benefit levels or instruction of in-work tax benefits so that "work pays". 1.30 Employment activation and mutual obligations open the opportunity for a renewed focus on hard to place Roma long-term unemployed. Traditional active labor market programs and activation schemes need modification in order to work for long- term unemployed and disadvantaged Roma. Activation policies include a range of new management approaches and services for the unemployed. Key elements of modern activation policy include: · Profiling of clients according to their distance from the labor market and priority attention by the employment office staff to those furthest from the labor market (particularly long-term unemployed). Profiling involves assessing the job-seeker's background and employability at or prior to the first interaction with the PES staff and typically leads to the categorization of the individual's distance from the labor market and then differentiated services. · An individualized approach for long-term unemployed and at-risk job-seekers, with individual reintegration action plans. Individual action plans describe an 19 individual pathway towards employment, involving training and addressing the client's multiple social needs (including health, debt, childcare etc). · New management and administrative approaches in the PES. These include greater specialization of labor office staff, including on dealing with hard to place clients and investments in training and retraining as well as rotating staff. It also crucially involves culturally sensitive service provision, centered around the individual's needs and abilities, involving dedicated advisers for minority job- seekers in those areas with large minority communities. · Service integration, typically either involving a merger of the traditional PES with social welfare offices or introduction of integrated computer systems. Service integration builds on the recognition that job-seekers typically have multiple needs, for example including indebtedness, that are best addressed in an integrated, "one stop shop" manner. · Focus on prevention and early interventions and youth. Most countries have introduced systems to detect risk groups early on and make them subject to prioritized and individualized attention, in particular for youth through career counseling and professional orientation at school. This can also involve early drop outs from school at a time prior to becoming long-term unemployed, typically after six months of joblessness, with directions to remedial and second-chance education or work placement and apprenticeship schemes. · Enhanced contracting out of services to private sector and NGOs under performance-based contracts and collaboration with private employment services. Many countries have introduced contracting out partnerships with private sector service providers and/or community based organizations and NGOs, including to facilitate the contact between the employment office and the client and to provide services. The rationale is that outside partner are positioned to deliver more effective services to highly disadvantaged job-seekers or services at lower cost than the public employment service infrastructure. · Regular evaluation of the effectiveness of the employment offices in placing the long-term unemployed, including through regular client surveys, to ensure high quality and focused service provision. · Empowering of clients to demand high quality services by introducing client satisfaction report cards where clients can provide feedback on the quality of service they have received. 1.31 Incomplete implementation of activation policy and the mutual obligations principle triggers substantial risks to the prospects for disadvantaged long-term unemployed. The challenge, therefore, is to get the policy mix right. As noted, activation policies build on mutually balanced obligations between the individual job- seeker and the employment office. They risk failure if one side is not fulfilling its obligation. The most effective employment service will not succeed in placing disadvantaged unemployed into jobs if the individual unemployed does not cooperate. On the other hand, tighter conditions on the disadvantaged job-seeker will not deliver 20 employment unless they receive effective support from the employment service. Failure to provide adequate services or barriers to actual job placement risk that the tightening of the beneficiary's obligations is not matched and that the beneficiary remains worse off ­ with lower benefits, a higher risk of poverty and without a job. Ineffective service provision and organization, insufficient attention to the disadvantaged client and a lack of quality interventions on offer which fail to result in a subsequent job ­ all can be binding barriers to making activation work. 1.32 The risk associated with imbalance is more binding the further distant the job-seeker is to the labor market and the more difficult he or she is to place, for example Roma. Not everybody is "activatable" and activation policy, if unevenly and mechanically applied, risks deepening poverty. For example, there is some evidence that employment activation policy in Slovakia, including the tightening of benefit eligibility, may have contributed to a deepening of poverty among poor and socially excluded communities16. Less disadvantaged job-seekers can more easily respond to tightened obligations and reduced benefits by utilizing their networks even if the employment services are of little help. The risk of imbalance also stems from the fact that it is easier to enhance the beneficiary's obligations than that of the public employment service: Tightening of benefit eligibility is done through simple change in legislation, while enhancing the obligation of the labor office requires time-consuming and complex institutional change. Obligations and incentives of job-seekers in the Czech Republic have recently been tightened 1.33 The Czech Government has recently introduced a reform to the benefit and tax system that have strengthened pro-work incentives, including for long-term unemployed. The reform, put in effect on January 1, 2008, transfers some of the tax burden from capital to consumption and affects labor taxation. Amongst others, it puts (at least a temporary) halt to indexation of social welfare benefits, expands child tax credits and lowers the income eligibility threshold for child benefits. Crucially, it lowers welfare support for inactive long-term unemployed by introducing a lower eligibility threshold for social welfare benefits, the Existential Minimum (EM), which applies for those that have been unemployed for more than 12 months. The Existential Minimum is set at 65 percent of the Minimum Living Standard (MLS) which is the traditional eligibility threshold, now applicable for those unemployed for less than 12 months. 1.34 How do the policy changes and other factors impact on the labor supply decisions of individuals? The effect of the changes are analyzed using a simple measure of the monetary incentives built into the set of main government labor-market programs that affect labor supply decisions: the Net Replacement Rate (NRR), explained in detail in Box 2. It concentrates on family types typical of socially excluded, with simulations of the NRR for one-earner families with 2 and 4 children and incomes ranging from close to the minimum wage (8,000 CZK) up to the average production wage (20,000 CZK). It 16World Bank (2005) 21 furthermore compares current institutional settings to those in place in 2007, quantifying the combined effects on pro-work incentives mentioned above. Box 2: Analyzing labor supply incentives: The Net Replacement Rate The Net Replacement Rate (NRR) is defined as the ratio of net income when unemployed to the net income when employed. The ratio takes values from 0 to 100. The higher the ratio, the lower the incentives to look for an employment opportunity. For example, at a NRR of 100, there are no monetary incentives to look for a job, since a given household receives the same level of income independent of the employment status. Since households enjoy not only consumption but also leisure and also face search costs and fixed costs of participating in the labor market (transportation to work, higher cost of food outside of the household, etc.), it is reasonable to expect that even net replacement rates significantly below 100 may not provide sufficient incentives for job search. Furthermore, the actual level of NRR that can be expected to effectively generate labor-supply incentive depends not only on valuation of leisure and transaction costs, but also on outside options such as shadow-economy employment opportunities and others. For all household types, the analysis contrasts the so-called short-term and long-term NRR. Short-term rates assume that one family member is unemployed for a short time period such that s/he is receiving unemployment benefits, and that the other family member is long-term unemployed, thus not receiving unemployment benefits. Unemployment benefits are calculated assuming that they are based on a wage that corresponds to the wage opportunity considered in the NRR calculations. For example, when the potential wage is CZK 10,000, it is assumed that the previous wage (before unemployment) was also CZK 10,000. Moreover, the MLS rather than the EM is used as the testing level for Social Necessity Benefits when calculating short-term rates. On the other hand, long-term NRR rates, which are relevant for long-term unemployed, correspond to the situation when the family does not receive unemployment benefits any more. In 2008 it is also assumed that the EM instead of the MLS serves as the testing level for Social Necessity Benefits of long-term unemployed. Finally, a number of assumptions are made on the family status. It is assumed that children are between 6 and 15 years old (so we abstract from parental allowances). It is also assumed that they live in a municipality with up to 100,000 inhabitants and that their actual cost of housing is the same as the socially respectable cost of housing. This means that they do not pay more than the level defined by law. It is also assumed that households satisfy all conditions to receive social necessity benefits. 1.35 Previously very high Net Replacement Rates have declined as a result of the 2008 reforms, suggesting that labor supply disincentives have been reduced, in particular for long-term unemployed. Figure 27 shows short term and long-term NRRs for different family size configurations for 2007 and 2008. Short-term replacement rates depict the situation for families whose breadwinner has been unemployed for less than 12 months, i.e. receive unemployment benefits and are subject to the MLS. The long-term replacement rates present the picture for those who have been unemployed for more than 22 12 months, are not eligible for unemployment benefits and are subject to the lower EM instead of the MLS. There are two main observations from the simulations: · The high NRRs of close to 100 percent in 2007 suggest that the previous system had powerful labor supply disincentives, in particular for those whose prospective labor earnings are low and for those who are able to earn income from informal jobs on top of social benefits. For example, the simulations suggest that in 2007 a couple with 4 children was as well off on benefits (NRR of 100 percent) as with a salary of CZK 13,000 which is 65 percent of the average production wage in the Czech Republic in 2007. This confirms the evidence from the Roma labor force survey presented above. · The 2008 reforms have decreased NRRs, in particular for the long-term unemployed. The decline is more pronounced for lower income levels and is typically in the order of 20 percentage points for long-run net replacement rates. The fall in NRRs is greater for the long-run rates, reflecting the powerful effect of the application of the lower EM for long-term unemployed. Also, as shown in Figure 9, the drop in NRR is typically more pronounced for low income families with a higher number of children. The sudden increases in short-term NRRs for families with 2 and more children, which are moving to higher wage levels as the number of children increases, are caused by the availability of social and child allowances (which grow as a function of family size)17. Figure 9: Previously high NRRs have fallen as a result of the reform, reducing work disincentives NRR, Couple 2 children, Short-term NRR, Couple 4 children, Short-term 105% 105% 100% 100% 95% 95% 90% 90% 85% 85% 80% 80% 75% 75% 8000 10000 12000 14000 16000 18000 20000 8000 10000 12000 14000 16000 18000 20000 Income (CZK) Income (CZK) 2007 2008 2007 2008 17Once family earned income exceeds the eligibility threshold, the denominator of NRR, the total family income (earned income plus allowances) decreases because allowances are no longer available. Hence the increase in the NRR. 23 NRR, Couple 2 Children, Long-term NRR, Couple 4 Children, Long-term 105% 105% 100% 100% 95% 95% 90% 90% 85% 85% 80% 80% 75% 75% 70% 70% 65% 65% 60% 60% 8000 10000 12000 14000 16000 18000 20000 8000 10000 12000 14000 16000 18000 20000 Income (CZK) Income (CZK) 2007 2008 2007 2008 Source: Jurajda and Zubricky (2008) 1.36 Moreover, the unemployed face tighter rules for cooperation with the labor Offices. Job-seekers are required to report to the Labor Office on a regular basis to demonstrate their job search requirements and can be deleted from the register for failure to cooperate with the Labor Office. However, there is a question as to whether this is fully enforced. Since 2004, long-term unemployed have been obliged to accept temporary job offers or perform public works tasks. 1.37 Summing up, the Czech Republic has introduced measures to tighten responsibilities of job-seekers and make work "pay" more. The recent changes to the Czech tax and benefit system, in particular the introduction of the Existential Minimum threshold, should reduce employment disincentives related to social welfare benefits. This, in turn, should raise long-term unemployed social welfare beneficiaries' readiness to look for employment. The next section will review the extent to which the Labor Offices are able to meet their part of the mutual obligations bargain. The Labor Office needs a new approach to provide effective activation services to disadvantaged job-seekers 1.38 The Czech Republic has introduced reforms to its employment promotion policies. However, compared to the tightening of obligations and incentives on the beneficiary's side, reforms on the Labor Office side to raise its ability to more effectively deal with disadvantaged and long-term unemployed have remained more limited. According to the new Act on Employment of 2007, individual action plans with increased attention for the job-seeker by the employment office are available for job-seekers below the age of 25 and for university graduates. Moreover, the Czech Government has proposed to merge the social assistance functions of the municipalities and the Labor Office functions from 2009 onwards by establishing a National Office for Labor and Social Affairs. 24 1.39 Yet more fundamental change in the Labor Office appears necessary to respond to the changing demands of the labor market and to enhance placement of disadvantaged job-seekers, for example Roma. This study argues that enhancing services for disadvantaged job-seekers would require the development of a number of elements of modern employment activation through the Labor Office as well as changes to the type of active labor market programs offered to Roma and the way they are administered. The following section reviews international best practice and assesses where the Czech Labor Office stands relative to that. It is worth noting that these are good practice policies relevant for the modernization of the Czech Labor Office overall and irrespective of the Roma employment challenge. At the same time, it is clear that not all presented elements can be implemented over night without the risk of overwhelming the Labor Office structures and raising, at least in the short-term, the cost of their operations. However, it is worth pilot-testing several approaches and rigorously evaluating their impact prior to a country-wide introduction over time. Employment service operations 1.40 Profiling of clients at registration allows better tailoring of service provision to job-seekers' needs, particularly the hard-to-place such as Roma job-seekers. Typically, EU public employment services categorize their client groups using a graduated scale from those easy to place to those most distant from the labor market, based on factors that predict the ease or difficulty of placement such as age, length of unemployment, education, literacy and numeracy, ethnicity, disability and other personal characteristics. Clients are then divided into separate categories, for example four groups like in the Netherlands, and job counselors' time and resources are applied in a differential way to each group. For example, a category one job-seeker (close to the labor market) would be assigned to self-service options and a category four job-seeker (most distant from the labor market) to interventions such as intensive counseling or specially required programs focusing on raising his/her employability. The profiling of job-seekers allows for service targeting and enhanced services for the disadvantaged without requiring, in theory, extra resources. The rationale is that service prioritization allows for a more efficient allocation of resources available for counseling: Staff time which is currently allocated to relatively easy-to-place job-seekers is freed up and focused on the more difficult cases. · Job-seeker profiling and service differentiation ­ status in the Czech Republic: While Czech Labor Offices record jobseekers' details for benefit payment purposes and on a caseload management computerized system, they do not utilize profiling and caseload prioritization. Research conducted for the purposes of this report indicates that labor offices are not profiling clients according to their distance from the labor market, and do not allocate proportionately more time to the long-term unemployed. Each job-seeker is registered in the Labor Offices's database, but not characterized according to his or her distance from the labor market. As a result, Labor Offices do not systematically categorize job-seekers according to their distance from the labor market and degree of counseling requirements. Essentially every job- seeker gets equal service. 25 1.41 An individualized approach to counseling job-seekers, typically through individual action plans, is at the core of modern activation policy and particularly useful for multiply disadvantaged job-seekers such as long-term unemployed Roma. Individual action plans lay out an individual pathway to employment agreed between the job-seeker and the labor office, involving training and addressing the client's multiple social needs (including health, debt, childcare etc) as well as personal aspirations: Ideally they activate and empower. The individual action plan is also often used when activation services for the most difficult-to-place clients are contracted out based on performance related payments to the contractors18. The European Employment Strategy stipulates that all those jobseekers under 25 should be engaged with before they pass the six month threshold and that the over 25s be engaged with before they pass the 12 month threshold. If fully enforced, individual action plans are a particularly useful tool for those job- seekers who have multiple barriers to employment, such as a lack of skills, a lack of prior labor market experience and social needs. · Individual action plans ­ status in the Czech Republic: The Czech Labor Office has introduced a mandatory individual action plan regime for unemployed under the age of 25, but it remains optional for those over 25. This follows the introduction of a pilot in the Moravskoslezsky kraj, where individual action plans have become a much utilized tool. However, staff in Labor Offices interviewed for the purposes of this report stated that they are generally making limited use of individual action plans for jobseekers over the age of 25, and sometimes not even for those under 25. The reasons given range from insufficient staff time to a lack of interest on the side of the job- seeker, suggesting that the tightened obligations of the job-seekers to cooperate may not be systematically enforced. 1.42 Effective implementation of the activation agenda implies a differentiated operational model for the public employment service, with a greater concentration of resources on the hard-to-place. The nature of employment activation is a departure from the traditional one-size-fits-all public employment service model and, through its inherent specialization and differentiation of services, requires staff retraining and operational restructuring. For example, it requires freeing up PES staff with back office duties to increase the number of frontline staff. While it needs to involve substantial retraining of staff, it also requires bringing in insufficient or missing skills such as psychologists, social workers and others. A greater focus on long-term unemployed and hard-to-place job-seekers requires the inclusion of social work functions into the service mix. Some OECD countries have merged employment service and social work functions to provide integrated services to job-seekers, and the Czech Government has proposed to do the same. For example, in the UK, the Jobcentre Plus (JCP) combines the previously separate job placement and benefit administration functions into a one-stop shop for employment service and income support19. 18Individual action plans are used in contracting out of activation services in the Job Network in Australia or the Employment Zones in the United Kingdom. 19Tergeist and Grubb (2006) 26 · Employment services management ­ status in the Czech Republic: The Czech Labor Office is the central institution in the implementation of national labor market policies and separate from municipal social welfare services and benefit administrations. It is tasked with a wide range of functions from the payment of State Social Support benefits to providing employment services to jobseekers. According to data from December 2007 on "frontline" staff-client ratios from each regional Labor Office in the Czech Republic, the average ratio across the country is 1:200 ­ above the EU average of around 1:15020 and the ILO-recommended 1:100. This suggests that there are not enough staff freed up to focus on frontline duties. 21 1.43 Outsourcing of activation services is particularly useful in the case of highly disadvantaged job-seekers such as long-term unemployed Roma who require more specialized and intensive interventions. The rationale is straightforward: Long-term unemployed and marginalized job-seekers require highly individualized and time- consuming services which include specialized social work services, and traditional employment services are typically not well-placed to deliver such services. It may simply be more efficient for the Labor Offices to contract out highly specialized and intensive services rather than delivering them in-house. Key requirement for effective outsourcing are good outcome measurement frameworks and pay-per-performance arrangements to help track performance quality and provide the right incentives to providers. The development of effective partnership and service outsourcing models has been an important element in the modernization of a number of OECD countries' public employment services. Over the last few years Australia has been at the forefront of the rollout of private provision of activation and placement services22, with encouraging results on the employment activation of its disadvantaged Aboriginal population23. · Contracting out of employment reintegration services and partnership with the private and NGO sectors ­ status in the Czech Republic. The Czech Labor Office has experience with outsourcing retraining and public works programs, but is not systematically outsourcing integrated comprehensive activation services such as practiced for example in Australia. As acknowledged by Labor Office job counselors interviewed for this report, the Czech Labor Offices are facing difficulties in promoting access to the labor market for Roma. The reasons are many; however, it is clear that the Labor Office alone cannot adequately address the labor market needs of the Roma. The key to success both in the Czech Republic and elsewhere is the development of close, performance-based contractual collaboration with other, more experienced and specialized agencies to deliver employment services of 20It is worth noting that this EU average figure hides a lot of variation, e.g. with Germany having a ratio of about 1:200 and the Netherlands 1:60. 21Staff-client ratios are typically difficult to define in PES generally, as an accurate ratio depends on the definition of who are the frontline staff who actually deal with the public. Here "frontline" staff are defined as mediation associates and counsellors, i.e excluding staff working on benefit administration and highly specialized staff such as psychologists. 22Tergeist and Grubb (2006) 23Department of Employment and Workplace Relations (2006) 27 relevance to Roma. There is promising experience in the Czech Republic, including funded from the European Structural Funds, which merits evaluation and further roll-out. However, systematic outsourcing of core activation services requires raising the Labor Office's capacities in setting service standards and monitoring services under performance-based contracting as well as large scale tendering and contracting. 1.44 Systematic program performance monitoring and impact evaluation of employment programs has become widespread across many OECD and EU countries. They are essential elements of public policy in general in that it allows reviewing whether programs work ­ whether they are well targeted and effective and whether money is well spent. It allows evidence-based decision making ­ scaling programs up if they work and adjusting them if they do not work as expected. Recently, the design of employment activation policy in many countries has benefited from quantitative evaluations, for example in the United States, the United Kingdom, the Netherlands, Australia and Germany. · Performance measurement and evaluation ­ status in the Czech Republic: While the is some experience with performance monitoring for active labor market policies, the Czech Republic has seen little rigorous impact evaluation of active labor market programs as yet. Raising capacities in this area will be a useful investment to ensure better knowledge about the effectiveness of employment policies, and the budgetary resources spent on them. Labor market programs 1.45 Active labor market programs most commonly used in the Czech Republic are requalification and training, public works and socially beneficial jobs. Figure 10 presents the breakdown of ALMPs across regions, suggesting that the main interventions are requalification, public works and socially beneficial jobs and confirming the findings for Roma beneficiaries reported above. However, there are variations: Moravskoslezsky kraj, the region with the highest incidence of unemployment and long-term unemployment, has been placing a strong focus on individual action plans, with more than 50 percent of ALMP participants enrolled in such plans. As opposed to that, Ustecky kraj, the second most deprived regional labor market, follows almost the same pattern as the Czech Republic averages. Interviews with job counselors conducted for the purposes of this report suggest that most disadvantaged job-seekers, including Roma, are placed into public works programs which typically last up to a year. They confirmed that there is some churning on public works programs, and training on such programs seems to be limited to basic operations required for the particular job rather than systematic up-skilling to a national qualification standard. 28 Figure 10: The content of ALMPs varies across regions ALMP participants structure across regions, 2005 100% 80% t 60% ercenp 40% 20% 0% ci aha ja ja kr kr jark ja ja ja ja kr kr kr kýc ja a kr kr kr puble Pr ý inc dea kýsva ja R hcez kýseco kýsecoh so ký or ky kýszel skez kýsr kýcets kýcer hr ja kýc ja ja kr Vy kr ouc nsílZ kos kr C ed Ji Pl ovalra Ú beiL ovélár St K dubiraP homiJ oml vsa O or K M requalification public utility works socially purposeful jobs graduate training sheltered work self-employment individual action plan Source: Research Institute for Labor and Social Affairs (RILSA), 2007. Note: Individual Action Plans are not typically an active labor market program, but are listed as such in the RILSA statistics. 1.46 While the provision of active labor market programs is concentrated in high unemployment regions across the Czech Republic, programs are not targeted to the most disadvantaged job-seekers. As Figure 11 reveals, the current programs are not targeted towards the disadvantaged groups, such as those with low educational attainment and long histories of joblessness. Over 30 percent of unemployed with complete secondary education participate in active labor market programs, compared to only 14 percent of those with only elementary education. Furthermore, 35 percent of unemployed with 6 to 12 months of unemployment participate, but only 20 percent of those with between 12 and 24 months and only 17 percent of those with more than 24 months of unemployment. Figure 11 ALMP participation varies across regions, but are typically not targeted to the most disadvantaged job-seekers Participation in ALMPs by length of unemployment, 2005 Particiaption in ALMPs by education, 2005 100% 100% 90% 90% 80% 80% 70% 70% t 60% t 60% en rcep 50% 50% 40% ercenp 40% 30% 30% 20% 20% 10% 10% 0% 0% < 6 months 6-12 months 12-24 months >24 months primary vocational secondary higher not covered covered not covered covered Source: Research Institute for Labor and Social Affairs (RILSA), 2007 1.47 Unemployed Roma rely heavily on the Labor Offices in looking for jobs, but there are also many registered who are not looking for work. Nearly two-thirds of the unemployed reported that they use services provided by the Labor Office. This is in 29 line with international experience which shows that public employment services are typically used as a main channel for job search. The main implication of these results is that, given their prominent role, better tailoring public employment services to the needs of Roma unemployed has a potential of improving their chances to find a job. At the same time, the majority of Roma registered with the Labor Office are not actively looking for jobs. This suggests that many of them may register for purposes other than finding employment. Two out of three Roma registered with Labor Office were economically inactive, i.e. not looking for jobs. The unemployed and the casually employed ­ who can be regarded as "legitimate" clients of Labor Office ­ account for only about 25 percent of all registered Roma unemployed. Given the heterogeneity of Roma registered with Labor Offices, there is a need to develop a range of services and approaches, including benefit sanctions, tailored to the different needs of different Roma groups. 1.48 Most Roma who visit the Labor Office receive job search assistance or vocational counseling. The frequency of visits varies from once per month (47 percent of registered) to once every 3 months (25 percent). Altogether, job search assistance is provided to some 60 percent of all registered Roma. Not surprisingly, it is provided mainly to those who are part of the labor force (the employed and the unemployed) and thus have close attachment to the labor market. It was received by close to 70 percent of the Roma that are currently employed, and by 65 percent of the unemployed (Figure 12). However, job search assistance is also provided to many discouraged workers (60 percent). This is a positive sign insofar as it can help to motivate discouraged Roma to actively look for jobs. Although a majority of Roma benefit from job search assistance provided by Labor Office, there is still a substantial fraction of those who do not, including those who are economically active. Thus there is room for improvement and for providing assistance to more Roma jobseekers. Figure 12 Most Roma who are registered with the Labor Office receive job search assistance, especially those who are actively looking for a job Roma registered with Labor Office who received job search assistance by labor forces status, 2008 80 70 60 tnecr 50 40 pe 30 20 10 0 Employed Unemployed Discouraged Not in labor force Source: Roma Labor Force Survey 2008; Bank staff calculations. 1.49 Many Roma are offered vocational training but few complete it. Out of all Roma registered with the Labor Office 36 percent were offered to participate in a training 30 course in the last 3 years.24 However, only 62 percent of those who were offered training actually got enrolled, while the remaining part refused to participate. And 77 percent of those who were enrolled in training completed the course. As a result only 17 percent of Roma registered with the Labor Office completed a training course in the last 3 years. The fact that almost 40 percent of Roma who are offered training refuse to participate is a negative symptom. It either implies that a substantial fraction of Roma workforce is not interested in upgrading their skills, or that training offered by the Labor Office is not tailored to Roma needs. Indeed, expert interviews conducted for the purposes of this study confirm these findings and reveal skepticism as to the usefulness of training typically offered to Roma, with some experts highlighting the fact that training seldom leads to actual employment for Roma, thereby reducing Roma jobseekers' interest and motivation to participate25. 1.50 Training is the main program used by Labor Offices to promote Roma employment, yet is not particularly effective in improving Roma employment chances. Only six percent of registered Roma were covered by active labor market programs other than training. However, only 27 percent of those who completed training managed to get a job afterwards. Most of those who found a job did it on their own (53 percent); the rest got a job through the Labor Office (43 percent) or the training provider. The low success rate is worrisome, and there seems to be much room to better tailor training to Roma needs and thus to improve the success rate. 1.51 Summing up, while the Czech Republic has introduced measures to tighten responsibilities of job-seekers and make work "pay" more, the Labor Office currently appears not fully prepared to take on its share of the mutual obligations bargain, and there is evidence that support to Roma is not very effective. The Labor Offices have not yet been restructured to provide more individualized, intensive services to disadvantaged job-seekers or to contract out such activation services to qualified third sector service providers. Moreover, as currently managed, traditional active labor market programs such as retraining do not appear to be effective to enhance employment chances of Roma. Unless the capabilities of the Labor Office are enhanced and programs are adjusted, there is a risk that the tightened obligations on the job-seeker do not result in more employment, but rather in lower household income for socially excluded families through reduced or lost social welfare benefits. Preventing future joblessness: Improving Roma education and school-to-work transition 1.52 In the long-term, employment chances of the Roma will only sustainably improve if their educational disadvantage is overcome. Ensuring their full educational integration and better education outcomes through addressing inequities in the education system requires preventive action now, in order to reap the benefits over the coming decades. International experience suggests that the educational integration of Roma, 24Training courses are organized mostly by Labor Offices themselves; only 12 percent of trainees participated in courses organized by private training providers or NGOs. 25Uherek et al (2008) 31 rather than segregation into separate schools, is the solution to unsatisfactory education outcomes. Equal opportunities in education for all is a basic human right. However, integration rather than separation is also good education policy, following mounting evidence on the benefits from instructing children with differing backgrounds and capabilities in the same classroom and avoiding early streaming into classes and schools for less and more successful students. Integration provides each child with an equal opportunity to explore his or her won potential and not be held back by poor learning conditions and instruction. 1.53 Effective and sustainable social inclusion policies need to focus on early intervention and preventive measures to improve education outcomes for Roma. Experience from around the world suggests that overcoming social and labor market exclusion of the Roma requires a life-cycle approach that starts by investing in early childhood development and translating into improved school outcomes and lifelong learning. Research shows that family background has profound impact on an individual's development. Children who grow up in poverty and disadvantaged circumstances have lower life chances: They enter school insufficiently prepared, do worse in school, leave school earlier and have lower employment chances. The stress of experiencing poverty, lacking stimulation in particular at very young age, lacking positive role models and a lacking learning support environment at home all contribute to worse school outcomes and subsequent failure in the labor market. Low educational attainment and high drop-out rates are often linked to insufficient preparation at the time of entering primary school. Research shows that preparation fosters cognitive, language and behavioral skills, which are vital to exploit one's full potential in later school education. 1.54 The large share of youth among the Roma population in the Czech Republic suggests the need for a youth focus of policies supporting Roma employment. Young people in general have been identified as a vulnerable group in the labor market in the Czech Republic, but young Roma are particularly at risk. According to the labor force survey in marginalized Roma localities, a sizable share of the population of jobless Roma is below the age of 25. With a high share of Roma having only primary education and less, young Roma are particularly at risk of being neither in employment nor education or training (NEET). The Czech education system does not appear to provide equal chances for Roma children 1.55 Roma in the Czech Republic have been suffering from unsatisfactory education outcomes for a long time, worsening their labor market prospects. The evidence from the survey in marginalized Roma localities confirms the low education status of socially excluded Roma in the Czech Republic, with the majority having completed only primary education ­ or even graduated from special schools for children with special learning needs ­ which leaves them without the skills needed in the modern Czech economy26. The legacy of segregation of Roma children into special schools has recently triggered a decision of the European Court of Human Rights against the Czech Republic over a case in Ostrava, ruling that the case of disproportionate streaming of 26Special schools were transformed into "basic practical schools" in 2004. 32 Roma children into special schools in Ostrava constituted as violation against fundamental rights.27 1.56 At the same time, it appears that access to services such as nursery and kindergarten for Roma in marginalized localities is poor. According to the survey of marginalized Roma localities, utilization of kindergartens among Roma children appears very low, with only 36 percent of Roma children in households with children in pre- school age going to kindergarten or zero classes. This contrasts with a high enrollment overall of children in pre-school in the Czech Republic overall: In 2006 more than 86 percent of Czech 4 year-olds were enrolled in pre-primary or primary education28. 1.57 While Czech pupils have consistently performed above average in successive international student assessments, the Czech education system retains high variation in education outcomes between schools. Data from the OECD's 2006 Program in International Student Assessment (PISA) indicates that the Czech Republic is one of the OECD countries with the highest variation in student performance between schools29. Moreover, a lot of this variance can be explained by economic, social and cultural status of students or schools in the Czech Republic. At the same time the variance in student performance within schools is substantially below the OECD average. This suggests that, while having highly successful schools which produce high learning outcomes, the Czech Republic also maintains highly unsuccessful ones with low learning outcomes and where the lacking success owes to the children's social status, among others. In effect, this means that the Czech education system does not appear to make up for unequal starting positions resulting from social-economic inequities. A NEW APPROACH TO ENHANCING EMPLOYMENT PROSPECTS OF JOBLESS ROMA 1.58 Improving job chances for socially excluded youth and adults, including Roma, will require a new approach, with a new way of engagement through the Labor Office and a focus on raising skills relevant in the labor market and activating and empowering marginalized individuals. Jobless Roma are a diverse group and require varying solutions 1.59 Roma in marginalized localities are a heterogeneous group, suggesting that enhancing employment chances will require different solutions for different groups. As described in this report, jobless Roma can be divided into inactive and unemployed, with inactive divided into discouraged and "truly inactive/disinterested". Moreover, skill levels vary across the Roma population living in marginalized localities. Lastly, it is important to differentiate policy interventions between adults and youth who have just left school. Table 6 lays out a simple taxonomy of jobless Roma, with differing main 27This decision follows research in 1999 by the European Roma Rights Center (ERRC) which found that Roma children in Ostrava were substantially more likely to be in special schools than noon-Roma children. While representing less than 5 percent of all primary school children in the Ostrava region, Roma children made up more than 50 percent of special school pupils. 28European Commission (2008) 29OECD (2007c) 33 challenges, to guide decisions on the type of primary intervention. The type of engagement and support for jobless Roma will depend on whether looking at those in the labor force (unemployed and casual workers) and those outside the labor force (discouraged workers and inactive). Moreover, the defining differences are the level of skills as well as the extent of labor market attachment. In addition, it is worth differentiating youth unemployed, in particular those on the verge of leaving school and those that have recently done so, and who are not in employment or training (the so called "NEETs" ­ Neither in Employment Education nor Training). Table 6: A simple taxonomy of jobless Roma in the Czech Republic and primary interventions Category Main challenge Primary intervention 1. Unemployed and casually Skills barriers and lacking work Individualized action plans with employed (in the labor force) ­ experience training and subsidies, managed 17 percent of the working age by LO or contracted out to population service providers Unskilled and low-skilled Skills · Public works, subsidized employment with skills upgrading; · Targeted retraining linked to new employment; · Second chance literacy programs; · Transport vouchers Skilled Work experience · Job counseling and search assistance · Transport vouchers 2. Individuals not in the labor Skills and attachment to the labor Employment activation with force ­ 56 percent of the market mutual obligations, integrated working age population activation services contracted out to service providers Discouraged (out of the labor Skills and attachment to the labor · Improving incentives and force) market tailored support through employment activation services with training and job search assistance, contracted out Inactive (out of the labor force) Motivation and multiple · Community social work interacting barriers preventing outreach to address barriers labor market participation (debt, lack of kindergarten access and others), contracted out 3. Youth and recent graduates ­ Skills and early drop-out from Measures to retain in or return to 20 percent (age 15-24) of the school education and training working age population · Professional and career orientation at school · Further education and vocational training programs beyond compulsory education ­ supported with conditional cash incentives; · Second-chance education 34 The Labor Office will require further change to better deal with disadvantaged jobseekers 1.60 Enhancing employment chances for Roma will benefit from the adoption of a modern employment activation approach based on mutual obligations ­ with strengthened incentives to cooperate for the jobless and a greater focus on effective support through the Labor Office. While incentives for the long-term unemployed to seek work have strengthened in recent years, the Labor Office at the beginning of 2008 remains insufficiently prepared to help disadvantaged job-seekers. In order to retain policy coherence and avoid that disadvantaged job-seekers are adversely affected by recent changes in the benefit system, the Labor Office needs to revamp its approach to help disadvantaged job-seekers into jobs. Policy directions can be differentiated into two groups, (i) a change in the modus of engagement with disadvantaged long-term unemployed and (ii) promotion of new types of active labor market programs. 1.61 First, the proposed new agenda would require a fundamental change in the management and delivery of public employment services. Based on international good practice, this could involve the following elements: · Change the mode of engagement with disadvantaged Roma job-seekers, offering a highly individualized and complex service, addressing their multiple employment, financial social and other needs that affect their employability as well as reflecting their individual aspirations. Individualized service provision starts with profiling of new clients into different categories depending on distance from the labor market and varied service packages for the different category job-seekers. · Restructure the Labor Office to free up more staff for individual counseling of vulnerable job-seekers. This entails training of staff to deal with this new challenge, including training on overcoming the cultural and communications divide, as well as the hiring of new skills, such as social work skills. It could also involve merging Labor Offices with social welfare offices to offer the job-seeker an integrated service window addressing his/her multiple needs. This would free up staff for frontline interactions with disadvantaged job-seekers and allow providing integrated services, including traditional employment services and social work services (e.g. debt advice). · Consider a substantial roll-out of outsourcing of activation services, including job counseling and referral, to qualified private and non-governmental agencies and with performance-based contracts. This is relevant in particular for the large group of discouraged inactive individuals. While it is important to ready the Labor Offices to deal with disadvantaged clients, outsourcing entire service caseloads for the most disadvantaged job-seekers is likely to be more effective and cheaper than recruiting highly specialized service providers in-house. · Make debt advice and financial literacy programs a core element of employment activation for long-term unemployed social welfare benefit recipients. International experience suggests that this may be best done through qualified 35 · non-governmental and non-profit agencies, for example through making debt work-out part of the menu of outsourced activation services. Debt advisory services are standard elements of social policy in many OECD and EU economies, and one which is crucial to develop for the Czech Republic, if a primary barrier to formal employment for Roma is to be overcome. Based on experience with such debt advisory services, it may also be worth considering introducing the institution of personal bankruptcy. Moreover, the provision of financial literacy programs in marginalized localities may help prevent further indebtedness of households and individuals30. Box 3: Dealing with personal indebtedness ­ financial literacy and debt counseling International evidence shows that personal indebtedness and joblessness, social exclusion and poverty often go hand in hand. Personal indebtedness has been identified as a social challenge, and in many advanced economies a debt counseling sector has emerged, with debt counseling services typically offered by non-governmental and non-profit organizations, charities or church-based organizations and often with public financing. Debt counseling typically involves personal advice to indebted individuals and help in paying off the debt and mediating between debtors and creditors. A typical approach builds on debt management plans which allow consolidating multiple monthly payments to different creditors into a single one as well as negotiating a reduced interest with lenders. In the EU, the Irish Money Advice and Budgeting Service (MABS) has received a lot of attention from other Member States. The MABS provides free, independent and confidential one-to-one counseling through trained money advisers. Launched as a pilot in 1992, MABS has since been rolled out across Ireland and operates through 52 independent local companies limited by guarantee and financed through the Department of Social and Family Affairs.31 MABS operates in close partnership with local authorities, community welfare offices, NGOs and credit unions, all of which are also represented on the management boards of the local companies. The majority of MABS clients are social welfare recipients. Meanwhile, financial literacy and consumer protection in financial services has recently received increased attention across the European Union32. It reflects a recognition that financial literacy ­ the ability of individuals to take adequate decisions related to purchasing financial products ­ has not kept pace with the growing availability of ever more sophisticated financial products. However, it is also motivated by the fact that low income households and less educated individuals are at particular risk of becoming unsustainably indebted in contexts of weak consumer protection and insufficient knowledge of financial products and the risks involved. Efforts have focused on developing curricula for financial education in schools as well as adult financial literacy programs. Under national financial literacy policies, trainings for young people are typically provided in conjunction with schools and, as adult programs, often by charities and non-governmental organizations. For example, in the UK, the Financial Services Agency oversees a National Strategy for Financial Capabilities and works through various non-profit organizations as well as with online tools to target school children, young people Not in Education, Employment or Training (NEETs), young parents, students in tertiary education and individuals in their workplaces. 30See World Bank (2007b) and World Bank (2008) for specific recommendations on financial literacy programs and consumer protection and financial services. 31See Korczak (2004) for the Irish MABS experience and a summary of policies in place in other EU Member States 32For a survey of financial literacy programs across EU Member States see Habschick et al (2007). 36 · Introduce routine evaluation of employment and social welfare policy, particularly of any new services focused on disadvantaged job-seekers. While the Czech Republic can benefit from international experience in guiding the development of new policy solutions, their effectiveness will depend on the circumstances in the country. It is unavoidable, therefore, to experiment with solutions, for example through piloting promising new programs and then rigorously evaluating their impact before scaling them up or rolling them out nationally (see Box 4). Moreover, it is important to assess the poverty and employment effect of the introduction of the lower Existential Minimum threshold. This would include both the effect on the share of the poor as well as the depth of poverty for those affected. It could also look at flows of registered unemployed, i.e. whether there has been an increased flow of long-term unemployed on EM into employment. Box 4: The policy laboratory: Employment program pilots and impact evaluation It is important to recognize that there is no single or certain solution to employment and skills shortage challenges discussed in this note. Countries across the OECD and EU have been dealing with the challenges of long-term unemployment and for years and continuously introduced and tested new approaches. In recent years many countries have placed a greater focus on piloting promising new programs and then rigorously evaluating their impact before scaling them up or rolling them out nationally. Like fellow new EU Member States, the Czech Republic has the opportunity to adopt a similar "reform laboratory" approach, taking advantage of the availability of funding from the Operational Program "Human Resources Development" (OP HRD). This would involve carefully reviewing international experience, adapting it to the Czech context and designing promising pilot interventions with in-built impact evaluation, all financed by OP HRD. The Social Inclusion Agency is an ideal vehicle for such an experimental approach. Impact evaluations establish the causal effect of a particular policy or program on outcomes, comparing a "treatment" group--those who receive an intervention--and a "control" group--those who do not receive the intervention for the duration of the evaluation. It can therefore, directly attribute the contribution of the new policy to change in outcomes. Conducting program pilots and evaluating their impact prior to a national roll-out can save money and help tailor programs most effectively to meet their objectives. 1.62 Second, it will also require a fresh look at the utilization of active labor market programs for disadvantaged long-term unemployed. · Explore options to shift ALMP spending and concentration away from skilled secondary school and university graduates towards low-skilled and disadvantaged job-seekers. In today's tight labor market in the Czech Republic the majority of skilled individuals are highly likely to be able to find employment without much support from the Labor Office either in form of extended 37 interaction with job counselors (i.e. beyond simple and cost-effective job referral services) or in form of participation in Labor Office-financed active labor market programs. Resources could rather be shifted to those who have trouble finding employment on their own and who therefore do require such support ­ the disadvantaged, low skilled and long-term unemployed. · Build on subsidized employment, public works programs, or "community employment", as a central intervention for unskilled Roma and in low labor demand regions, but include systematic skills upgrading. In particular, in addition to transmitting work habits and experience, they should involve strategic vocational skill upgrading through training and aim at empowering individuals. In this respect a public works assignment should have an inbuilt formal identification of training needs for each client as well as an agreed training plan linked to the national qualifications framework. Given that many municipalities are overwhelmed with managing public works programs, actually limiting the supply of such positions, the Labor Office could experiment with contracting qualified NGO service providers. · Given the widespread functional illiteracy among Roma in marginalized localities, second-chance education and literacy programs should become a core part of the retraining programs offered through the Labor Office. Creating a basis for subsequent vocational training, they should be seen as a priority engagement for Roma long-term unemployed within the framework of employment activation. They not only raise employability but also empower marginalized individuals to more fully participate in society. · Link retraining programs to actual employment and build on client choice. Retraining programs on average appear to have not delivered employment results for jobless Roma. Successful programs should ideally be closely linked to actual employment and take account of the skills needed in the labor market as well as individual aspirations. If not linked to employers' needs, they risk being wasteful spending and undermining job motivation of training beneficiaries. Focusing on youth and prevention are key to stemming the flow of new disadvantaged jobseekers 1.63 The Roma employment agenda is to a large part a youth employment agenda, and it is primarily with young workers where a difference can be made. Improving access of Roma to sustainable and high quality jobs over the long-term requires greater attention to making the education system work for Roma and developing a youth-centered employment activation approach. Policy directions include the following: · Address inequities in the education system affecting Roma by systematically implementing an educational integration policy from early childhood to tertiary education. This entails an expansion of childcare supply and measures to work with Roma parents to motivate them to send their children to child care and 38 kindergartens so as to prepare them better for school and avoid early channeling into low quality basic practical schools. Incentive measures could also include experimenting with conditional cash transfers (CCTs) linked to kindergarten and pre-school attendance, for example through adjustments in the MLS threshold for children and youth, depending on whether they go to school (higher threshold) or not (lower threshold). · Incentivize young Roma to stay in school beyond the minimum mandatory school age, i.e. into upper secondary education, instead of entering the pool of registered Labor Office clients. The Education Maintenance Allowance (EMA) program from the United Kingdom could serve as an example, and a Czech adaptation could entail applying a higher minimum subsistence level for a dependent child aged 15-26 that remains in non-compulsory education for a maximum of, say, two years and depending on retention and completion of education levels. Alternatively, scholarships for post-compulsory education for children whose parents are on social assistance could help boost retention of Roma children in education and training. · Promote early outreach through school counseling and professional orientation to identify and counsel those youth at risk of drop out. Prevention of early school leaving, and retention in formal education and training requires early outreach to and advisory services for those at risk of drop-out. The experience of the United Kingdom's "Connexions" services tasked with providing guidance to 13-19 year olds, in particular disadvantaged youth, suggests that the Social Inclusion Agency could, in a formalized manner, support schools in engaging youth at risk and counseling them about their options to remain in school or training beyond compulsory schooling. · Develop the individual action plan approach for young people further into a mandatory and intensive youth-centered activation approach focused on NEETs. Following the UK example, this would be centered around intensive counseling, with job placement services, training and remedial or second-chance education for older youth and back-to-school programs for the younger. · Pilot and test apprenticeship, internships and wage subsidy programs for young workers. Facilitating the school to work transition and preventing the NEETs phenomenon will require testing proactive measures like apprenticeships, internships, placement and job subsidies programs for young people to help them not only get into the labor force, but build relevant skills. This is an agenda that requires active contributions and partnerships between the Government and employers and trade unions. For example, experience from across the OECD shows that wage subsidy programs can have positive employment effects for young workers33. 33Kluve, J (2006) 39 1.64 Employment policy reform does carry a cost, but a lot can be achieved by reallocating resources. It is beyond this study to analyze in detail the fiscal implications of employment policy reform. However, the argument is that costs can be contained. First, the introduction of service differentiation ­ rather than equal service for all ­ allows freeing up and reallocating resources to those jobseekers who are more difficult to place. This applies both for expenditure on programs as well as management and staff cost. Second, adjusting existing programs which appear to have not worked well for Roma, such as retraining, will ensure money is more effectively spent. Moreover, experimenting with new solutions and evaluating them before a full roll-out can help save money otherwise lost on ineffective policies. Lastly, analysis in neighboring countries suggest that policies to promote Roma inclusion "pays", and successful employment activation and prevention of long-term unemployment and social welfare dependency can yield fiscal savings through reduced expenditure on welfare payments and increased revenues from taxes34. The Social Inclusion Agency provides a major opportunity for change 1.65 The Czech Government has launched a pilot Social Inclusion Agency to promote innovative partnerships between public services and NGOs in select marginalized communities which provides an entry point for a new approach to promote employment of Roma. The Agency has been created as an instrument to promote complex solutions to multiple and interwoven forms of social exclusion at the local level and in individual socially excluded Roma localities. It is to foster partnerships at the local level involving the municipal authorities, schools, the Labor Offices and other locally-provided public services as well as non-governmental organizations and the private sector. The creation of the Agency opens the opportunity for developing innovative social inclusion programs individually tailored to local conditions and taking advantage of locally available know-how and actors. While the Agency will be acting as a service provider itself in the pilot localities, it will outsource the bulk of activities. Making the Agency a success will entail adopting a deliberate experimental approach, piloting new approaches, such as performance-based contracting, partnerships with local stakeholder and others and will benefit from an inbuilt culture of evaluation. The Agency has the potential to test and evaluate innovative solutions and, based on proven success, make the case for scaling up. Dealing with discrimination requires legislation, its enforcement and work to change public opinion 1.66 Enhancing employment chances for Roma also requires modern anti- discrimination legislation to combat cases of discrimination in the labor market and in service provision. While there is legislation to address discrimination in the labor market, this legislation needs to be enforced. Moreover, the Czech Republic still has not adopted EU-compatible anti-discrimination legislation. While this report's findings suggest that discrimination in the labor market is not the only barrier to employment of Roma, discrimination is likely to play some role in explaining the high rejection rate of 34Bogdanov and Angelov (2007); Kertesi and Kézdi (2006) 40 Roma job-seekers. In addition, reducing discrimination also requires a proactive policy of changing public opinion. Enhancing employment chances of socially excluded Roma is a major challenge 1.67 Lastly, a word of caution: Not everybody is "activatable". Given the extent of the educational and labor market disadvantage of the marginalized adult Roma population, the challenge of employment activation is substantial. Experience from other OECD and EU countries shows that not every disadvantaged jobless adult can be activated, and it is to be expected that­ even with the best policies ­ not all Roma will move from welfare into employment in the Czech Republic. It is, therefore, important to ensure that benefit sanctions in case of non-participation of the jobless adult do not result in adverse effects for the family, and in particular children. Again, here is where the Social Inclusion Agency will play an essential role ­ in helping to address the multiple dimensions of inclusion of families and in breaking the inter-generational transmission of exclusion by focusing on inclusion policies for children 41 2. THE CHALLENGE: ACCESS OF ROMA TO EMPLOYMENT IN THE CZECH REPUBLIC Analysis of employment and unemployment patterns among Roma is difficult due to the lack of data disaggregated by ethnicity ­ in the Czech Republic as well as in other countries in Central and South-Eastern Europe35. To fill this gap a specially designed Roma Labor Force Survey was carried out for this study in May 2008.36 This chapter analyzes the survey results to provide a basis of policy discussions in subsequent chapters, focusing on working age Roma in socially excluded communities. It finds highly unsatisfactory labor market outcomes for Roma in the Czech Republic, defined by labor market inactivity rather than unemployment. The main reason for inactivity is the inability to find a job and discouragement, especially for women and prime age workers37. Moreover, this chapter finds that Roma disadvantage in the labor market is driven by extremely low educational attainment and widespread lack of functional literacy and numeracy skills. The labor market outcomes among Roma are highly heterogeneous, often ­ but not always ­ explained by skills differences. While most employed Roma work as unskilled workers and have short-term low paid jobs, there is substantial heterogeneity in terms of employment duration and pay. The chapter identifies a number of barriers to Roma employment: On the demand side, low skills appear to be the most important barrier to employment. On the supply side, barriers include reliance on social welfare, personal indebtedness and lacking affordable transportation. Interestingly, Roma jobseekers when looking for jobs rely heavily on assistance provided by Labor Offices, suggesting that the Labor Offices play a key role in overcoming the employment gap for Roma. LABOR MARKET OUTCOMES FOR ROMA IN MARGINALIZED LOCALITIES 2.1 Roma in the Czech Republic suffer from highly unsatisfactory labor market outcomes, although it is non-participation in the labor market rather than unemployment that is the biggest problem. As many as 56 percent of Roma of working age (15-64) are out of the labor force, that is neither employed, nor actively looking for a job (Figure 13, left hand panel)38. The unemployed account for only 5 percent of the working age population. Close to 50 percent are employed, however 12 35Ringold et al (2005) 36See Roma Labor Force Survey ­ Technical Report for the description of the survey's methodology, including sample selection. 37Active job search is a precondition for categorizing a person as unemployed. Persons of working age who are not actively looking for jobs are categorized as economically inactive (out of the labor force). 38Being outside the labor force is often also termed as "economically inactive". 42 percent have only casual and not regular jobs39. After all, in many respects the casually employed are closer to the unemployed than the employed (see below). Figure 13 Majority of working age Roma in excluded communities are out of the labor force Roma population in working age by labor force status, 2008 Reasons for not working by gender, 2008 100% 90% 26.9 80% 70% 60% 50% 40% 30% 55.8 20% 10% 12.2 0% Men Women 5.2 Maternity Education/training Disability, retirement Employed Casually employed Unemployed Not in labor force No jobs available Other Note: Population of working age (15-64). Source: Excluded Roma Labor Force Survey 2008; Bank staff calculations. 2.2 The primary reason for inactivity is the scarcity of job opportunities, and many inactive are discouraged workers. Roughly three in four men and one in two women say they are out of the labor force because they could not have found a job (Figure 13, right hand panel).40 This means that many of the inactive Roma are actually so called discouraged workers, that is the unemployed who ceased searching for job once their efforts proved futile (see also Box 5 for a discussion of the concepts of unemployment, inactivity and discouragement) 41. This in turn implies that the traditional unemployment rate may be a misleading indicator of Roma labor market outcomes, as it underestimates the actual degree of unemployment among Roma workers. This chapter looks at this issue in more detail below using the so-called augmented unemployment rate. 39Casually employment is defined as a situation in which the survey respondent declared that he/she does not have a job, but reported income during the reference week. 40It should be borne in mind that these results reflect the subjective perception of labor market conditions. 41Discouraged workers in the survey are defined here as those who did not actively look for a job in the reference week (and thus were categorized as inactive rather than unemployed) but reported the duration of their job search. 43 Box 5: The unemployed, inactive and discouraged In order to understand the nature of Roma joblessness it is important to divide the concepts of unemployment, inactivity and discouragement. Unemployed workers are individuals working age who are in the labor force and are actively looking for a job. In order to be categorized as unemployed, persons who have no job need to meet two additional conditions: (a) be actively looking for a job, and (b) be able to take a job if offered. Inactive persons or those that are not participating in the labor market are individuals of working age who are not employed, but also not actively looking for a job. Discouraged workers are a subgroup of the inactive who have given up actively looking for a job due to their actual or perceived inability to find employment. They remain interested in finding employment, if it were possible, and in principle respond to activation policies. They contrast with those inactive who are not actively looking for work, for example mothers who choose to stay at home to care for children. 2.3 The lack of job opportunities for Roma workers may reflect a number of factors. First, it may reflect weak labor demand. Second, it may be due to the fact that Roma workers lack the skills required by employers. Finally, it may be the result of discrimination. Available data do not allow rigorously testing these tree hypotheses. Instead this chapter presents some circumstantial evidence suggesting that poor skills of Roma workers are likely to be an important factor limiting their employment chances. Chapter 3 of this study will show that, although currently the labor market in the Czech Republic is tight, the demand for low skilled labor is limited. So, both factors ­ low skills of Roma workers and week demand for low skilled labor ­ reinforce each other, barring many Roma from formal employment and driving them into inactivity. Discrimination against Roma by employers likely matter, too (see the section below on unemployment), but still economic factors seem to play the predominant part. 2.4 The educational attainment of Roma in excluded communities is very low. Eight out of ten Roma of working age have no more than primary education (Figure 14, left hand panel) 42. There is a substantial proportion ­ 14 percent ­ of persons who graduated from special schools for children with special educational needs ­ a specific feature of Roma schooling in the Czech Republic for decades. This large group has no formal certified vocational skills. Only two Roma in ten have some formal vocational training or more than compulsory education. 2.5 Actual literacy and numeracy skills are very low, too. As many as 44 percent of working age Roma in excluded communities can be considered functionally illiterate (Figure 14, right hand panel). Another 44 percent have only some basic literacy and 42This report uses the Czech terminology in which primary education refers to basic, compulsory education with duration of nine years, usually from age 6-15. This includes the internationally used primary (ISCED 1) and lower secondary (ISCED 2) levels. Basic vocational education refers to ISCED 3c, while secondary education refers to ISCED 4 (upper secondary education). 44 numeracy skills. Only 12 percent can be considered as functionally literate, i.e. able to answer most of the relatively simple questions that require primary school-level knowledge.43 Figure 14 The vast majority of Roma have no more than primary education and little skills Distribution of Roma working age population by education, 2008 Distribution of Roma working age population by skill level, 2008 18% 12% 3% 44% 5% 60% 14% 44% Level 0 Level 1 Level 2 Less than primary Special Primary Basic vocational Secondary Note: Level 0 = little or no literacy/numeracy skills Level 1 = some literacy and numeracy skills Level 2 = basic literacy/numeracy skills (primary school level) Source: Excluded Roma Labor Force Survey 2008; Bank staff calculations. 2.6 The inter-generational upward educational mobility is extremely limited among Roma, and there is evidence for a worsening of educational attainment among Roma. Only 21 percent of sons of fathers who have primary education or less received post-primary education (mainly basic vocational training). Even more worrying, more than half of the sons (54 percent) of fathers who have post-primary education received less education.44 Thus, there is very little upward educational mobility among Roma but substantial downward mobility. This suggests that the transition might have had a negative effect on Roma educational prospects. The generation of fathers who acquired their education under the old regime tends to be better educated that the generation of sons who received their education after the transition to a market economy. 2.7 Despite poor formal education many Roma workers obtained job-specific vocational skills after completing school, typically in employment. But gender differences are pronounced. While six in ten men obtained some vocational skills, only 4 in 10 women did so. Expectedly, the acquisition of vocational skills critically depends on employment. Among those who were employed in the last 3 years the proportion of workers with vocational skills (obtained mostly through on-the-job training) approaches 60 percent, while among those who were not employed it is slightly less than 40 percent. Thus employment matters for the acquisition of skills and, through this channel, is likely to increase future employment chances. 43The results were obtained using a simple skill assessment survey. The survey is described in Annex 1. 44A very similar pattern emerges when on compares educational attainment of daughters with that of their mothers. 45 2.8 Most Roma have a short record of formal employment, limiting their access to old age pension in the future.45 This particularly refers to younger generations, i.e. those which entered the labor market after the fall of the communist regime in 1990. For example, the average length of formal employment is 23 years for older workers (aged 55-64) compared with less than 9 years for prime age workers (aged 25-54), (Table 7). This appears to be a disproportionate difference, suggesting that the likelihood of formal employment for Roma workers may have been higher under central planning than it is under market economy. But this is likely to be a general pattern. The question whether the transition affected the Roma formal employment chances more than those of non- Roma is still open. However the bottom-line is that given the short spells of formal employment among younger Roma most of them will not be eligible to pension insurance. This may increase the incidence of poverty among Roma in the future, and increase the burden on the social welfare system. Table 7 Duration of formal employmenta) by age Years in Age formal employment 15 ­ 24 0.9 25 ­ 54 8.4 55 ­ 64 23.1 a) Employment covered by pension insurance. Note: Roma population of working age (15-64). Source: Excluded Roma Labor Force Survey 2008; Bank staff calculations 2.9 As to the recent labor market history, most Roma worked only few months in the last three years. On average men worked for 13 months and women for 7 months (Table 8). But the average values mask substantial variation across individuals. While 25 percent of Roma men worked for at least two in the last three years, another 25 percent did not work at all. And 25 percent of Roma women worked for at least 11 months in the last three years, while 50 percent did not work at all. These data show considerable inequality in labor market outcomes among Roma even within what seem to be relatively homogenous communities. While a large number of Roma fare well on the labor market (in terms of employment), an equally large number fare poorly. Table 8 Duration of employment in the last three years by gender (summary of the distribution) Duration of employment in the last 3 years Men Women Months Average 12.7 7.0 Top decile 36 27 Top quartile 24 11 Median 9 0 Bottom quartile 0 0 Bottom decile 0 0 Note: Persons aged 18 or more. Source: Excluded Roma Labor Force Survey 2008; Bank staff calculations 45Formal employment was defined as that covered by pensions insurance. 46 ROMA EMPLOYMENT PROFILE 2.10 Few Roma in excluded communities are employed. Only 27 percent of working age Roma are employed, compared with the national average of 66 percent (Table 9). An additional 12 percent are only casually employed. This employment gap is dramatic: As many as 27 percent of Roma would need to find employment in order to raise the employment rate among Roma to the national average. However, the gap results largely from the low labor force participation by Roma rather than from high unemployment.46 The unemployment rate among Roma is 11.7 percent, roughly twice as high as the national average. It is high, but not extremely high by the standards of transition economies. After all in many European countries the unemployment rate hovers around 10 percent but at the same time the employment/population ratio is much higher than that among Czech Roma. It is the labor force participation rate which is much lower among Roma than the national average. Only 44 percent of working age Roma participate in the labor force compared with 70 percent across the Czech population. Thus many Roma are detached from the labor market. The main policy issue to be addressed is thus discouragement and inactivity, rather than ineffective job search. 2.11 The low employment rate among Roma is largely accounted for by labor market detachment of prime-age women. The labor force participation rate by Czech Roma women is half that of Czech women overall: 31 against 62 percent (Table 9). This is likely to reflect cultural factors (the role of women in Roma society) and lack of job opportunities, but poor access to services such as nursery and kindergarten may also play a role. Indeed, utilization of kindergartens among Roma children appears very low, with only 36 percent of Roma children in households with children in pre-school age going to kindergarten or zero classes47. This contrasts with a high enrollment overall of children in pre-school in the Czech Republic overall: In 2006 more than 86 percent of Czech 4 year- olds were enrolled in pre-primary or primary education48. 2.12 But prime-age Roma men are also much less likely to be economically active than their Czech counterparts. The labor force participation rate by prime age Roma men at 63 percent ­ although significantly higher than that of women ­ compares unfavorably to the national average of 95 percent. To what extent this low economic activity of prime-age Roma men reflects poor skills, lack of demand or discrimination is not quite clear. Still, this chapter will show below that skills play an important part in accounting for Roma poor labor market outcomes among Roma in marginalized localities. 2.13 In sharp contrast to prime-age workers, Roma youth and older workers are more, not less, economically active than young and older Czech workers overall. For 46This is because, as demonstrated earlier few Roma are actively looking for jobs and thus are categorized as economically inactive rather than unemployed. In order to be categorized as unemployed persons who have no job need to meet two additional conditions: (a) be actively looking for a job, and (b) be able to take a job if offered. 47Close to 90 percent of those going to kindergarten or zero classes report going to formal, rather than informal kindergartens. 48European Commission (2008) 47 illustration, 42 percent of Roma youth (aged 15-24) are part of the labor force, compared with the national average of 32 percent (Table 9). This large difference is explained by much higher rates of educational enrollment of young Czechs overall than that of young Roma. For older workers (aged 55-64) the difference in labor force participation rates is smaller (2 percentage points in favor of Roma). The likely cause is that in contrast to their ethnic Czech counterparts many (if not most) older Roma lack pension insurance and thus have an incentive to seek employment. 2.14 The high labor force participation rates of young and older Roma go hand in hand with high unemployment rates. The Roma youth unemployment rate approaches 25 percent and is more than twice as high as the national average. The unemployment rate among older Roma is 15 percent and is three times higher than the national average. Whereas it is usual that youth unemployment rates are double the average, the high unemployment rate among older workers is rare. It is consistent with the earlier conjecture that older Roma workers cannot afford to retire because of the lack of pension insurance and thus are compelled to look for jobs. Table 9 Indicators of labor market outcomes: Romaa) against the national average. Unemployment Labor force Employment/population rate participation rate ratio Czech Czech Czech Roma Roma Roma Republic Republic Republic Population of working age b) 11.7 5.4 44.2 69.9 39.1 66.1 Gender Men 10.3 4.3 61.3 78.1 55.0 74.8 Women 14.0 6.8 30.5 61.5 26.3 57.3 Age 15 ­ 24 (youth) 24.7 10.7 42.4 31.9 31.9 28.5 25 ­ 54 (prime-age) 7.8 4.8 44.2 87.8 40.7 83.5 55 ­ 64 (older) 15.2 4.6 50.0 48.2 42.4 46.0 a) Roma in excluded communities covered by the Roma Labor Force Survey 2008. b) 15 ­ 64 Note: data for the Czech Republic refer to 2007, data for Roma refer to May 2008. Source: Roma Labor Force Survey 2008 and Eurostat; Bank staff calculations. 2.15 Despite unsatisfactory labor market outcomes, high economic activity pays off for Roma workers. Although they face a high unemployment rate, Roma youth are more often employed than Czech youth overall (employment/population ratios are 32 and 29 percent, respectively). And older Roma workers are almost as likely to be employed as Czech older workers overall (42 and 46 percent, respectively). In contrast, despite the relatively low unemployment rate, prime age Roma are much less likely to be employed than the prime-age Czech population overall due to economic inactivity. This suggests that well designed activation policies may bear fruit and improve Roma employment prospects. 48 2.16 Given inactivity and discouragement, the standard unemployment rate is a poor measure of Roma labor market outcomes. Augmenting the unemployment rate by the discouraged shows that 46 percent of Roma are without a job but available. The relatively low unemployment rate masks the fact that few jobless people are actively looking for jobs. And the main reason for this is discouragement, i.e. the belief that no jobs are available and thus job search is pointless. To address this shortcoming one should use the so called "augmented" unemployment rate, which accounts for the discouraged worker effect.49 By this measure unemployment among working age Czech Roma reaches 46 percent (37 percent for men and 57 percent for women). It is almost 4 times as high as the standard unemployment rate, indicating that there is a huge pool of available Roma workers who are without a job. ROMA EMPLOYED, UNEMPLOYED AND INACTIVE: HOW DO THEY DIFFER? 2.17 Skills are the key characteristic that distinguishes the employed from the unemployed and from those out of the labor force (inactive). The better the skills the stronger the attachment to the labor market. The employed have higher skills than the unemployed, who in turn tend to have higher skills than the inactive. Figure 15 shows the difference between the employed, the unemployed and the inactive along several socio- demographic dimensions. The employed and the unemployed are predominantly men, while the inactive are predominantly women (Figure 15, Panel A). The employed are more often prime age workers whereas the unemployed and the inactive are more often younger persons. 49The augmented unemployment rate shows the percentage of the total pool of available workers who are without a job 49 Figure 15 The employed, the unemployed and the inactive differ along many dimensions Panel A Panel B Gender Profile Age Profile 100% 100% 6 11 6 80% 41 35 80% 69 60% 60% 61 80 72 40% 40% 59 65 20% 20% 31 29 14 21 0% 0% Employed Unemployed Not in labor force Employed Unemployed Not in labor force Male Female 15 - 24 25 - 54 55 - 64 Panel C Panel D Educational Profile Skill Profile 100% 100% 19 13 28 80% 39 80% 53 47 60% 60% 60 64 40% 57 50 40% 35 40 20% 20% 21 24 11 15 12 13 0% 0% Employed Unemployed Not in labor force Employed Unemployed Not in labor force Special- Primary Vocation+ Level 2 Level 1 Level 0 Panel E Panel F Driving Skills Employment Record 100% 100% 21 80% 80% 44 60 73 60% 80 60% 92 40% 40% 79 56 20% 40 20% 27 21 8 0% 0% Employed Unemployed Not in labor force Employed Unemployed Not in labor force Driving license No driving license Job in last 3 years No job in last 3 years Note: The unemployed include casually employed. Source: Roma Labor Force Survey 2008; Bank staff calculations. 2.18 There is a notable difference among the employed, the unemployed and the inactive with respect to educational attainment. The employed are better educated than the unemployed, who in turn are better educated than the inactive. For illustration, the employed have post-primary education (largely basic vocational training) twice as often as the unemployed, and three times as often as the inactive (Figure 15, Panel C). Simultaneously, the employed are roughly half as likely to have less than primary education (including special education) as the unemployed and the inactive. Hence, 50 stronger attachment to the labor market goes hand in hand with better educational attainment. 2.19 The employed have also better skills than the unemployed and the inactive. Roma with literacy and numeracy skills account for over 70 percent of the employed and only for around 50 percent of the unemployed and the inactive (Figure 15, Panel D).50 Accordingly, while functional illiteracy is widespread among the unemployed and the inactive, it is relatively rare among the employed. 2.20 The marked skill gap between the employed and the unemployed and the inactive is also notable when looking at everyday skills. For example, the employed Roma are twice more likely to have a driving license than the unemployed. The unemployed in turn are 2.5 times more likely to have a driving license than the inactive (Figure 15, Panel E). A similar pattern emerges when examining other activities requiring skills, such as using the internet or reading newspapers. Roma with those skills are found much more frequently among the employed than among the unemployed and all the more so than among the inactive. The greater the actual distance from the labor market, the lower the skills. 2.21 The employed differ from the unemployed and the inactive also in terms of their labor market experience. As expected, the employed are more likely to have a previous employment record. Only a minority of the inactive (27 percent) had previous job experience, compared to over half of the unemployed (56 percent) and a vast majority of the currently employed (79 percent). Thus previous labor force status is a good predictor of the current labor force status ­ a phenomenon known as path dependence: The employed reap the benefits of past employment (which was likely to be associated with skill acquisition, greater motivation, etc.), whereas the unemployed and still more so the inactive are often locked in the joblessness trap (associated with the erosion of skills and morale). Such a clearly marked pattern of path dependence poses considerable policy challenge of taking people out of the joblessness trap and bringing them to the labor market. Skills and Roma Labor Market Outcomes 2.22 Vocational training greatly increases Roma employment chances. At the same time, special education condemns Roma to unemployment. Among the Roma who received vocational training 50 percent are employed. This is more than double compared to those who received primary education and over three times more than among those who received special education or less (Table 10). Stated otherwise, Roma who have special education are two times more likely to be inactive than Roma who received vocational training. This shows that the type and level of education received has a decisive impact on Roma employment prospects. Roma who are better educated and have vocational skills fare much better on the labor market than those who are poorly educated and lack vocational skills. Obviously, this finding has far-reaching policy 50Based on the results of the skills assessment, which was part of the Roma Labor Force Survey, (see Annex 1) 51 implications. Better education and the provision of vocational skills is key to improving Roma labor market outcomes. Table 10 Basic vocational training significantly raises the chance of employment, labor force status by educational attainment, 2008 Basic vocational Special or less Primary Labor force status traininga) Percent Employed 15.2 22.6 50.3 Unemployedb) 18.5 17.6 15.5 Not in labor force 66.3 59.8 34.2 Note: Roma population of working age (15-64) a) Including secondary b) Including casually employed Source: Roma Labor Force Survey 2008; Bank staff calculations. 2.23 In addition to formal education, actual skills matter, too. Among those who are functionally illiterate (skill level 0) few are employed (only 17 percent) and a majority are unemployed or inactive (Table 11). Literacy and numeracy skills (levels 1 and 2) increase the probability of employment by a factor of two.51 Even if other factors ­ such as discrimination or job attitudes ­ play a significant role, better education and thus better skills help to overcome Roma labor market disadvantage. Table 11 The more skilled are substantially more likely to be employed Labor force status by skill level, 2008 Labor force status Level 0 Level 1 Level 2 Percent Employed 17.3 34.8 29.9 Unemployeda) 21.6 14.1 15.4 Not in labor force 61.1 51.1 54.7 Note: Roma population of working age (15-64) Level 0 = little or no literacy/numeracy skills Level 1 = some literacy/numeracy skills Level 2 = basic literacy/numeracy skills a) Including casually employed Source: Roma Labor Force Survey 2008; Bank staff calculations. 2.24 Everyday skills are also a good predictor of Roma labor market outcomes. Having a driving license almost triples the chance of employment (Table 18). Similarly, those Roma who use internet or are reading newspapers are significantly more likely to be employed than those who do not. 51There is not much difference between the employment chances between Roma with skill levels 1 and 2. If anything, Roma with skill level 1 seem to fare on the labor market somewhat better than those with skill level 2. But this may be due to the small sample size and/or imprecise measurement of the actual skill level. 52 Table 12 Everyday skills raise employment chances Labor force status by everyday skills, 2008 Skill indicators Employed Unemployeda) Not in labor force Percent Driving license Yes 57.5 19.0 23.6 No 19.8 17.0 63.2 Using internet Yes 34.1 13.8 52.1 No 24.0 18.7 57.4 Reading newspapers Yes 29.6 17.2 53.2 No 20.3 16.9 62.8 Note: Roma population of working age (15-64) a) Including casually employed Source: Roma Labor Force Survey 2008; Bank staff calculations. 2.25 There is a two-way relationship between skills and labor market outcomes. For example, those Roma who work are likely to have an easier access to internet or to newspapers. Employment provides earnings and thus allows one to own a car. So, causality runs in both directions: better skills improve employment chances and being employed is conducive to acquiring additional skills. Accordingly, this analysis may overestimate the effect of skills on employment. Still, there is no doubt that there is a close relationship between skills and employment among Roma and thus supporting investment in human capital and facilitating labor market entry are key policies to overcome their labor market and social exclusion. Profile of employed Roma 2.26 The vast majority of Roma work as unskilled workers, most of them in manufacturing and construction (57 percent of all employed Roma), others in services (16 percent) and in agriculture (6 percent). Only one Roma in five works as skilled worker, mostly in construction (nearly 5 percent of all employed Roma), others as operators, attendants and shop assistants. 2.27 The majority of Roma work in the formal sector. About 30 percent of Roma workers do not have written employment contract. But still ­ contrary to prior expectations ­ formal employment seems to be the norm rather than the exception (provided that having a written employment contract is a valid indicator of the extent of formal employment). 2.28 Few Roma run their own business or are self-employed; the vast majority are wage workers. Specifically, the self-employed/business owners account for less than 8 percent of total Roma employment. Wage workers are roughly equally split between those who have long-term jobs and those with only temporary jobs. 53 2.29 Few Roma workers commute to work. Three out of four Roma workers do not commute to work, which means that they work in their immediate neighborhood. The low proportion of those who do commute (only 25 percent) suggests that a lack of affordable transportation may be an important barrier to Roma employment. In fact, remote job location is an important reason Roma jobseekers turn down job offers (see the next section). Being not able to commute, many Roma have access only to a small, strictly local job market, which naturally limits their job chances. 2.30 The average job tenure of Roma workers is short indicating high turnover and job instability. Nearly 50 percent of all Roma workers have been holding their jobs for less than a year. This is a very high proportion; in OECD countries the share of new hires (i.e. workers with tenure less than one year) rarely exceeds 20 percent52. At the same time, however, there is still a substantial fraction of Roma workers who hold long- term jobs. The average job tenure is 3.5 years (thus significantly higher than the median of just 12 months). This is consistent with the earlier observed inequality in labor market outcomes among Roma workers. It seems there is a dual market for Roma labor. While some Roma have better, more secure jobs others have only casual, precarious jobs. 2.31 Underemployment does not seem be a major problem in Roma community. Few Roma have less than full-time jobs, and it may be the case that some of them prefer to work part-time.53 On average Roma work 41 hours per week. Only 17 percent work no more than 30 hours per week, which usually provides the upper bound estimate of the extent of underemployment. 2.32 Roma earnings are very low. The average Roma worker earns some 40 percent of the national wage (Table 13).54 Those who have regular jobs earn somewhat more, but still only about half the national average. Put differently, the average Roma worker earns only slightly more than the minimum wage, whereas the regular workers earn about 35 percent above the minimum wage. Even the top paid Roma workers earn less than the national average. The earnings of top decile regular Roma workers account for only 75 percent of the national average wage. On the other extreme, the earnings of the bottom decile regular worker account for about 20 percent of the national average, which is less than 60 percent of the minimum wage. Hence, virtually all Roma workers in excluded 52OECD (1997) 53The incidence of part-time employment is an indicator of underemployment only if part-time work does not reflect voluntary choices of workers. Unfortunately, it was not possible to test whether those Roma who hold part-time jobs do this voluntarily or not. Nonetheless, given the relatively low incidence of part- time employment the conclusion is valid that it is unemployment and inactivity rather than underemployment that are the main issue in the Roma community. 54The ratio of Roma earnings to national average (minimum) wage is most probably underestimated. First, the national average (minimum) wage is the gross wage, while survey respondents reported the take-home pay (net wage), which is lower than the gross wage. Second, self-reported earnings tend to be lower than the official earnings (wage) data coming from the employer based survey. Third, some surveyed workers did not work full-month full-time in the reporting month, thus their earnings cannot be compared with the average monthly wage of a full-time worker. 54 communities occupy the bottom end of the earnings distribution, and the incidence of low-pay is extremely high.55 Table 13 Earnings of Roma workers are very low, 2008 As % of national averagea) As % of minimum wageb) Regular Regular All workers All workers workers workers Percent Average (mean) 39.2 46.6 113.8 135.0 Top decile 67.2 75.4 195.0 218.8 Median 34.5 40.9 100.0 118.8 Bottom decile 8.6 19.4 25.0 56.3 a) Average gross monthly wage for the 1st quarter of 2008 (koruna 23,235) b) Minimum wage (gross) of koruna 8,000. Source: Roma Labor Force Survey 2008, national sources; Bank staff calculations. 2.33 Moreover, there is very high inequality in earnings among Roma. The top decile worker earns 7.8 times more than the bottom decile worker (Table 14). Among regular workers earnings dispersion is smaller but still substantial. The top decile regular worker earns 3.9 times more than the bottom decile regular worker. For comparison, the top decile Czech worker earns only 3 times more than the bottom decile worker. The average for the OECD countries is 3.4. Thus earnings dispersion among Roma is high also by international standards. This is surprising, given that Roma residing in marginalized localities are often considered a homogenous community. But it is consistent with the duality hypothesis advanced earlier: some Roma fare much better on the labor market than others. Table 14 Earnings inequality of Roma is wide Summary of Roma earnings distribution, 2008 Regular National Measures of earnings dispersion All workers workers averagea) Top decile/bottom decile 7.80 3.89 3.01 Top decile/median 1.95 1.84 1.77 Bottom decile median as share of median wage 0.25 0.47 0.59 a) 2005 Source: Roma Labor Force Survey 2008, OECD (2007) for the Czech Republic; Bank staff calculations. 2.34 High earnings inequality among Roma is driven primarily by the extremely low wages of workers at the bottom of the wage distribution. The bottom decile regular Roma worker earns less than 50 percent of the median wage, which is beyond the OECD Europe range (Table 14).56 For comparison, Czech workers overall earn close to 60 percent of the median wage. This represents a considerable difference in the relative 55Low-pay is conventionally defined as wages below two-thirds of the average wage. 56In the US, where the earnings dispersion is the largest among the OECD countries, the bottom decile worker earns 48 percent of the median wage. 55 earnings position of the low-paid Roma and ethnic Czech workers. Moreover, the distance between the low-paid Roma and the median Roma is very large. The distance becomes larger still when looking at all workers, including those who have casual jobs. Then the earnings of the bottom decile worker account for only one-fourth of the earnings of the median worker. This points toward an important policy issue. There is a group of Roma who are employed but earn extremely little and well below the minimum wage. Most of them have causal jobs. For this group employment is hardly a way out of poverty. Profile of Roma jobseekers 2.35 Most of the unemployed Roma became unemployed because they lost their previous jobs. The main causes were business closure (27 percent of all unemployed) and redundancy (25 percent). In addition, many became unemployed due to the termination of the temporary job. Only 11 percent quit their job voluntarily. This structure is consistent with the earlier result showing a high turnover market for Roma labor, and a high proportion of short-term jobs. 2.36 Long-term unemployment looms large in the Roma community. Close to 50 percent of Roma have been jobless for over a year.57 The proportion is still higher ­ 65 percent ­ when taking into account the discouraged workers. Thus it appears that many Roma workers manage to get a temporary job, but once the job is finished, they experience long spells of unemployment before finding another job. At the same time, nearly 40 percent of Roma workers are unemployed for less than 6 months which points to high inflows into unemployment, and is consistent with the notion of a high turnover labor market. 2.37 In looking for jobs Roma rely heavily on the Labor Offices. Nearly two-thirds of the unemployed reported that they use services provided by the Labor Office. This is noteworthy, because international experience shows that public employment services are seldom used as a main channel for job search. This result suggests that the unemployed from excluded communities, such as Roma, use public employment services more often than the unemployed at large. The second most popular job search method is using the network of friends and relatives (46 percent). International experience indicates that this is the primary job search method for the unemployed at large. The fact that Roma use this method relatively less frequently may indicate that they have less social capital they can rely on when looking for jobs. The third most important job search method is to directly contact employers, which is used by over 40 percent of Roma jobseekers. The main implication of these results is that, given their prominent role, better tailoring public employment services to the needs of Roma unemployed has a potential of improving their chances to find a job. 2.38 What kind of jobs are Roma looking for? There is a common perception that Roma are predominantly interested in short term jobs. However, it is not borne out by 57The sample of the unemployed is very small which renders the estimates reported here subject to a wide margin of error. 56 the evidence. A vast majority (nearly 75 percent) say that they prefer regular long-term jobs. Only about 7 percent say they would like to have a short-term job and less than 4 percent would like to run their own business. But a non-negligible fraction ­ around 10 percent ­ would like to work part-time. 2.39 Wage expectations of the unemployed Roma do not seem excessive. Indeed, the consistency between the reservation wages (wage expectation) and actual wages is striking.58 The reservation wage of the average unemployed Roma is virtually the same as the actual wage earned by the average Roma worker (Table 15). This indicates that the Roma unemployed have adequate information and correctly perceive labor market conditions.59 One exception is the least productive Roma unemployed (bottom decile of the reservation wage distribution), who have the lowest reservation wages, but still apparently higher than the market wage rate for low-skilled Roma workers (bottom decile of the actual wage distribution). Interestingly, the top decile reservation wages are lower than top decile actual wages. This may mean that the unemployed with the highest reservation wages recognize that they have lower skills than the most productive employed workers. Table 15 Wage expectations of Roma are close to actually paid wages in the labor market Ratio of the reservation wage of the unemployeda) to actual wages of regular Roma workers by decile Reservation wage/Actual wage Percent Average (mean) 99.5 Top decile 85.7 Median 105.3 Bottom decile 133.3 a) Unemployed include casually employed and discouraged workers. Source: Roma Labor Force Survey 2008; Bank staff calculations. 2.40 Although wage expectations of Roma workers are moderate, they reject jobs that pay too little. One Roma worker in five rejected a job offer in the last three years.60 A low wage offered was the most important reason along with poor working conditions (Figure 16). However, a remote job location is an important reason for turning down a job, too. As many as 20 percent of those who rejected a job offer did so because the job was too far away or too costly to get to. Thus, as already mentioned, the need to commute is likely to be an important barrier to Roma employment due to lack of affordable transportation. 58The reservation wage is the lowest wage offer that the unemployed is going to accept. 59This is under the assumption, that the unemployed have similar skills and other relevant characteristics as the employed. In reality, as demonstrated earlier, the unemployed have on average lower skills than the employed and thus their wage expectations may be to some extent excessive, especially at the bottom of the skill distribution. 60The term "workers" refers to members of the labor force, i.e. the employed and the unemployed. 57 Figure 16 Roma turned down job offers because of low wages and poor working conditions Resons for rejecting job offer 5% 19% 39% 37% low wage working conditions remote location skill requirements Source: Roma Labor Force Survey 2008; Bank staff calculations. 2.41 Roma jobseekers are often rejected by employers, yet less so if they have more skills and a previous employment record. As many as 64 percent of the unemployed, casually employed and discouraged workers were rejected by the employer when applying for a job. Compared to that, only 31 percent of the employed Roma experienced a rejection by the employer (Figure 17 Panel A). Thus, twice more unemployed Roma were rejected by the employer than employed Roma. The experience of the unemployed Roma is apparently quite different from that of the employed. Figure 17 Labor market experience, better education and higher skills lessen the risk of the rejection of a Roma jobseeker by the employer Panel A Panel B Roma job seekers rejected by employer by labor force status, Roma job seekers rejected by employer by edcuation, 2008 2008 60 80 70 50 60 t t 40 enc 50 30 40 percen per30 20 20 10 10 0 0 Special or less Primary Vocational or Employed Casually Unemployed Discouraged secondary employed Panel C Roma job seekers rejected by employer by skill level, 2008 60 50 t 40 encr 30 pe 20 10 0 Level 0 Level 1 Level 2 Source: Roma Labor Force Survey 2008; Bank staff calculations. 2.42 Better education and higher skills lower the risk of the rejection by the employer, too. Roma jobseekers with special education or less are by one-third more likely to be rejected by the employer than those with vocational or secondary education 58 (Figure 17, Panel B). A similar pattern emerges when looking at the role of skills. Roma workers with higher skills are much less likely to be rejected by the employers than those with lower skills (Figure 17, Panel C). Hence, whether or not a Roma worker is rejected by the employer is largely determined by the skill level as well as labor market experience. 2.43 To what extent is there discrimination among employers against Roma? The available data does not allow a conclusive answer. In order to rigorously test the hypothesis of discrimination one would need to have a comparison group of non-Roma workers with similar skills and other relevant characteristics. While systematic discrimination likely plays some role in explaining the labor market outcomes among Roma, the data suggest that lacking skills and prior work experience are an even more binding barrier. After all, the likelihood of rejection of Roma by employers is not uniform, but varies with education and skill level. THE INTERACTIONS OF ROMA WITH LABOR AND WELFARE OFFICES 2.44 Strikingly, the majority of Roma registered with Labor Office are not actively looking for jobs. This suggests that many of them may register for purposes other than finding employment. Figure 18 shows that out of three Roma registered with Labor Office two were economically inactive, i.e. not looking for jobs. It is also noteworthy that one in ten Roma registered with Labor Office already had a regular job. The unemployed and the casually employed ­ who can be regarded as "legitimate" clients of Labor Office ­ account for only about 25 percent of all registered Roma unemployed. Given the heterogeneity of Roma registered with Labor Offices, there is a need to develop a range of services tailored to the different needs of different Roma groups. Apparently, those who are actively looking for jobs need different services than those who are inactive. In the former case, it is job search assistance, vocational counseling or training that are the most effective forms of support. In the latter case activation and from "welfare to work" policies are more adequate. Figure 18 The majority of Roma registered with Labor Office are not actively looking for jobs Roma registered as unemployed by labor force status, 2008 11% 16% 66% 7% Employed Casual job Unemployed Not in labor force Source: Roma Labor Force Survey 2008; Bank staff calculations. 59 2.45 Most Roma who visit the Labor Office receive job search assistance or vocational counseling. The frequency of visits varies from once per month (47 percent of registered) to once every 3 months (25 percent). Altogether, job search assistance is provided to some 60 percent of all registered Roma. Not surprisingly, it is provided mainly to those who are part of the labor force (the employed and the unemployed) and thus have close attachment to the labor market. It was received by close to 70 percent of the Roma that are currently employed, and by 65 percent of the unemployed (Figure 19). However, job search assistance is also provided to many discouraged workers (60 percent). This is a positive sign insofar as it can help to motivate discouraged Roma to actively look for jobs. Although a majority of Roma benefit form job search assistance provided by Labor Office, there is still a substantial fraction of those who do not, including those who are economically active. Thus there is room for improvement and for providing assistance to more Roma jobseekers. 2.46 Many Roma are offered vocational training but few complete it. Out of all Roma registered with the Labor Office 36 percent were offered to participate in a training course in the last 3 years.61 However, only 62 percent of those who were offered training actually got enrolled, while the remaining part refused to participate. And 77 percent of those who were enrolled in training completed the course. As a result only 17 percent of Roma registered with Labor Office completed a training course in the last 3 years. The fact that almost 40 percent of Roma who are offered training refuse to participate is indeed a negative symptom. It either implies that a substantial fraction of Roma workforce is not interested in upgrading their skills, or that training offered by Labor Office is not tailored to Roma needs. Indeed, expert interviews conducted for the purposes of this study reveal skepticism as to the usefulness of training typically offered to Roma, with some experts highlighting the fact that training seldom leads to actual employment for Roma, thereby reducing Roma jobseekers' interest and motivation to participate62. Figure 19 Most Roma who are registered with the Labor Office receive job search assistance, especially those who are actively looking for a job Roma registered with Labor Office who received job search assistance by labor forces status, 2008 80 70 60 tnecr 50 40 pe 30 20 10 0 Employed Unemployed Discouraged Not in labor force Source: Roma Labor Force Survey 2008; Bank staff calculations. 61Training courses are organized mostly by Labor Offices themselves; only 12 percent of trainees participated in courses organized by private training providers or NGOs. 62Uherek et al (2008) 60 2.47 Training provided by the Labor Office is not particularly effective in improving Roma employment chances. Only 27 percent of those who completed training managed to get a job afterwards. Most of those who found a job did it on their own (53 percent); the rest got a job through the Labor Office (43 percent) or the training provider. The low success rate is worrisome. However, proper interpretation requires a comparison with the job placement rate of non-Roma workers. Still, there seems to be much room to better tailor training to Roma needs and thus to improve the success rate. 2.48 Training is the main program used by Labor Offices to promote Roma employment. Only six percent of registered Roma were covered by active labor market programs other than training. In this context the low success rate of training courses provided to Roma should be of particular concern. 2.49 Roma receive social benefits more frequently than non-Roma in excluded communities. Figure 20 shows that social benefits are received by 65 percent of Roma and "only" by 50 percent of non-Roma. The difference is significant and suggests that Roma face greater barriers of exclusion and greater welfare dependency than their non- Roma neighbors residing in the same marginalized locality. This may also be driven by differing socio-demographic characteristics which are not analyzed here. Figure 20 Roma receive social benefits more often than non-Roma in excluded communities Social benefit recipients in marginalized localities, Roma and non-Roma, 2008 100% 80% 35 49 60% 40% 65 51 20% 0% Roma Non-Roma Social benefit receipients Non receipients Source: Roma Labor Force Survey 2008; Bank staff calculations. Note: Shares express social benefit recipients as a percentage of all adults of working age residing in marginalized localities. 2.50 Social welfare benefits are relatively well targeted at those who are without a job and thus do not have a source of earnings. As Figure 21 shows, some 75 percent of benefit recipients are either inactive or unemployed. Additional 11 percent of the recipients are only casually employed. But still 15 percent of the recipients have regular jobs. More importantly, however, benefit receipt and labor force status are not independent. Those who do receive benefits have weaker motivation to look for a job, a 61 phenomenon known as the "inactivity trap". Thus although benefits are targeted at the jobless, some of them are jobless because they claim welfare benefits. Figure 21 It is mainly the inactive and the unemployed Roma who receive social benefits, but some of the employed receive benefits, too Roma social welfare beneficiaries by labor force status, 2008 15% 68% 11% 6% Employed Casual job Unemployed Not in labor force Source: Roma Labor Force Survey 2008; Bank staff calculations. 2.51 Welfare benefits received by Roma are relatively high, aggravating the risk of an inactivity trap. The average benefit accounts for close to 90 percent of the minimum wage, and the median benefit accounts for 75 percent (Table 16). Given that the majority of Roma workers earn around the minimum wage this is likely to create significant labor supply disincentives, as long as benefits are means tested. That is, once the benefit is withdrawn, the average Roma gains nothing in monetary terms when taking low-paid formal job. This is a classic case of the inactivity trap. Formal employment does not pay-off and this gives rise to welfare dependency. Table 16 Social welfare benefits received by Roma are high relative to the minimum wage Summary of the distribution of benefits awarded to Roma, 2008 Benefit amount As a percentage of minimum wagea) Average 88.8 Top decile 168.8 Median 75.0 Bottom decile 25.0 a) The statutory minimum wage is equal to the median wage of Roma workers (including casual). Source: Roma Labor Force Survey 2008; Bank staff calculations. 2.52 Heavy indebtedness is another barrier to formal employment. One Roma in two is indebted, and the debt amount tends to be large. On average, indebted Roma owe 9.3 times the minimum wage to the lending institution (see Table 17). The average reflects heavy indebtedness concentrated among a relatively small fraction of Roma. The median debt is substantially lower, and accounts for 2.5 times the minimum wage. Still, for a considerable fraction of Roma workers formal employment would imply a loss of 62 earnings due to debt repayment. Under such circumstances, informal employment is a coping strategy meant to protect one's earnings. One possible way of removing disincentives for formal employment resulting from indebtedness is to put in place the institution of personal bankruptcy. In addition, Social Welfare Offices should be more active in providing counseling to Roma families, including on debt management issues. Currently they are not: Only 14 percent of Roma welfare clients reported that they had received counseling in the last year. Table 17 Roma face widespread and often high debt Summary of the distribution of debt owed by Roma, 2008 Debt amount As a multiple of minimum wagea) Average 9.3 Top decile 18.8 Median 2.5 Bottom decile 0.5 a) The statutory minimum wage is equal to the median wage of Roma workers (including casual). Source: Roma Labor Force Survey 2008; Bank staff calculations. CONCLUSIONS 2.53 The majority of Roma in excluded communities fare poorly in the labor market, and the key issue is inactivity and discouragement. Many working age Roma are out of the labor force, because they believe there are not jobs available for them and therefore they have gave up job search. Relatively few of the jobless are actively looking for jobs. Therefore, according to the standard definition, unemployment is relatively low. It is inactivity that looms large. The key findings are the following: · There is great variation in employment outcomes among Roma: Although excluded communities seem relatively homogenous in terms of socio- economic characteristics (such as education and skills), some Roma fare much better in the labor market than others, with regular jobs and relatively high earnings. One important factor that greatly improves Roma labor market prospects is vocational education and better literacy and numeracy skills. Better educated and higher skilled Roma are much more likely to be employed than the less educated and skilled ones. · The majority of Roma have little formal education and lack basic skills, including literacy and numeracy, and they prevail among the economically inactive and the unemployed. Thus a lack of adequate education and skills seems to be the root cause of poor labor market outcomes of most Roma. · There are constraints to labor mobility. Most Roma do not have access to a wider labor market because of lack of affordable transportation. For this reasons they are bound to take jobs located in the close neigbourhood. This 63 obviously limits their employment chances. How important is labor mobility for Roma employment is an important issue that merits further research. · Welfare dependence and the inactivity trap discourage employment and push Roma into inactivity or into informal employment. Most Roma of working age claim social welfare benefits, which creates a severe labor supply disincentive because benefits are high relative to Roma potential earnings (which are low). Taking low-paid formal job and in consequence losing welfare benefit does not pay-off. · Heavy indebtedness is a key barrier to formal employment. Many Roma accumulated large debt and thus have a strong disincentive to reveal their actual earnings. This pushes them into informal employment and temporary, precarious jobs offering no social insurance. · In looking for work Roma rely primarily on services provided by the Labor Office. Most do receive job search assistance, although a large proportion of those registered is not looking actively for jobs. Roma often are offered participation in training courses; however the effectiveness of this policy in improving Roma employment prospects is limited. Many Roma refuse to participate, relatively few complete training, and a minority find jobs afterwards. This is likely to be a significant contributor to widespread discouragement. 64 3. THE LABOR MARKET IN THE CZECH REPUBLIC AND OPPORTUNITIES FOR LOW SKILLED WORKERS The Czech labor market has been performing strongly in recent years, and strong labor demand has driven down unemployment to record low levels ­ below 5 percent in early 2008. Most remaining unemployment in the Czech Republic is now of a long-term nature, suggesting that remaining unemployed face binding barriers to employment possibly due to lacking skills and work habits, disincentives or lack of motivation to look for work or other reasons such as discrimination. As shown in Chapter 2, the overall positive developments conceal deeply unsatisfactory outcomes for the relatively small group of Roma in marginalized localities. While most such Roma suffer from low educational attainment and skills and would only qualify for elementary occupations, demand for labor in elementary occupations is low in most regions, and in particular in those regions with traditionally high unemployment and long-term unemployment. It is precisely in these regions where many excluded Roma localities can be found. This suggests that strong economic growth, triggering a fall in unemployment overall, alone is unlikely to improve the labor market outcomes for low skill workers, particularly the young and those in underdeveloped regions in the Czech Republic, but that proactive measures are needed to overcome barriers to employment. RECENT LABOR MARKET TRENDS AND CHALLENGES 3.1 The Czech labor market has been performing strongly in recent years, with low and falling unemployment and high employment rates compared to EU 27 averages. Like in other new EU Member States, the Czech labor market has substantially tightened in recent years63. As Figure 22 (left panel) reveals, the Czech Republic continues to outperform many of its EU peers. Like in previous years, the employment rate in the Czech Republic in 2007, at 66.1 percent, was higher than the EU27 average of 65.4 percent and close to the Lisbon target of 70 percent in 2010. Equally, unemployment has been lower (5.3 compared to 7.1 percent in 2007) and has continued to fall to below 5 percent in the first quarter of 2008. Labor demand is strong and there are plenty of job openings. There are only three unemployed for one job vacancy, which points to a tight labor market, and the vacancy rate is high indicating labor shortages64. After all, many Czech employers see skills shortages as a major obstacle to the operation and growth of their firms65. 63See World Bank (2007) 64Ibid. 65Ibid. 65 Figure 22: The Czech Republic faces better employment outcomes than EU27, but much of its unemployment is long-term, 2006 Comparative Labor Market Outcomes, 2007 Share of long-term unemployment (>12 months) in total 80 unemployment 70 60 60 50 50 t 40 en t 40 erc cen 30 P er 30 P 20 20 10 10 0 0 Employment Rate Unemployment Rate Activity Rate 1994 2003 2004 2005 2006 EU 27 Czech Republic Czech Republic OECD Sources: Eurostat (left panel), OECD (right panel) 3.2 At the same time, the Czech Republic has been unable to tackle long-term unemployment which is substantially above the OECD average and keeps rising as a share of overall unemployment. Figure 22 (right panel) shows the evolution of the share of long-term unemployment in total unemployment in recent years. More than half of the unemployed in the Czech Republic have been unemployed for more than 12 months. This relatively high share of long-term unemployment points to the problems of labor market disadvantage and social exclusion. Apparently some workers have problems with finding a job despite the overall strong demand for labor. Two external factors contribute to employment problems faced by some worker groups: · Regional variation in labor market conditions. While the overall labor market conditions are good, there are some regions of the country where labor demand is much weaker and accordingly, unemployment substantially higher. · The changing skill profile of labor demand. Technological progress tends to be biased towards higher skills, which implies that the demand for less skilled labor is gradually declining. This gives rise to the skills mismatch: shortage of some, usually higher skills coupled with excess supply of other, usually lower skills. REGIONAL VARIATIONS IN LABOR MARKET CONDITIONS 3.3 The regional variation of labor market conditions in the Czech Republic is quite substantial. The unemployment rate varies from 2.5 percent in the capital region (Prague) to over 10 percent in the Ustecky kraj (Table 18). And higher unemployment goes hand in hand with the higher share of long-term unemployment (Figure 23).66 In the Ustecky region the incidence of long-term unemployment is twice as high as in the Prague region (52 and 25 percent, respectively). High unemployment reflects weak 66It should be noted that the high unemployment does not necessarily imply a high share of long-term unemployment. In fact, an increase in the share of long-term unemployment tends to be associated with a fall in the unemployment rate, as it is the short-term unemployed who are more likely to find jobs and leave the ranks of the unemployed. 66 labor demand. The unemployment/vacancy (U/V) ratio in the Ustecky region at over 8 is dramatically higher than in the in the Prague region (less than one). These are just two extreme examples. Differences in labor market conditions among other regions are less pronounced but still significant. Labor market slack is substantial also in Moravskoslezky and Karlovarsky regions. At the same time, labor markets are tight in Jihocensky, Plznensky and Pardubicky regions. Table 18: Main indicators of labor market conditions by region, Q2 2007 Unemployment rate U/V ratio Incidence of LTU (%) (%) Czech Republic 5.3 3 43 Praha 2.5 0.9 24.9 Jihoceský kraj 2.8 2.1 32.7 Plzeský kraj 3.4 1.4 36.8 Stedoceský kraj 3.5 1.8 34.3 Královéhradecký kraj 4.2 2.6 34.1 Pardubický kraj 4.2 1.8 38.8 Vysocina 4.9 3.1 41.2 Jihomoravský kraj 5.4 3.8 43.1 Zlínsky kraj 5.8 3.2 42 Liberecký kraj 6.3 3.1 36.4 Olomoucký kraj 6.4 4.4 43 Karlovarský kraj 8.1 4.1 46 Moravskoslezský kraj 8.2 7.7 51.7 Ústecký kraj 10.1 8.2 52.5 Source: Statistical office of the Czech Republic and Ministry of Labor and Social Affairs. Bank staff calculations. 3.4 Regional differences in job opportunities point to a segmentation of the Czech labor market. Regions where labor demand is strong coexist with those where it is weak, and this segmentation persists due to limited labor mobility. Accordingly, workers in the depressed labor markets face a higher risk of labor market detachment and social exclusion. Figure 23: Long-term unemployment looms large in high unemployment regions (Q2, 2007) Unemployment by duration, Q2/2007 100% 80% 60% 40% 20% 0% aj ký aj ýk aj ja a aj aj aj aj ý aj kr kr kr kr kr kr kr kr kr aha ezsl avs ckýets kos aj kýsr Pr or aj kr kr oucký kysníl incos cký ký decka Vy ecký Z esc aj kr Ú vsaro ovalra homiJ kýsez oml O dubiraP Pl beri éhrvo kýsecoh L edo álr Ji M K St K 12 + months 6 - 12 months 1 - 6 months Source: Ministry of Labor and Social Affairs. Bank staff calculations 67 DEMAND FOR LOW SKILLED WORKERS 3.5 The Czech Republic has seen a strong variation in the demand for skills. In general, there is a strong demand for high and specialized ­ both white collar and blue collar ­ skills, coupled with a relatively weak demand for lower and more generic skills (Figure 24). For illustration, the unemployment/vacancy ratio is low ­ indicating a strong demand ­ for professionals as well as for skilled manual workers (craftsmen, machine operators and assemblers). It is much higher for clerks and service workers as well as for elementary occupations. Thus, job opportunities for less skilled workers are markedly worse than those for more skilled workers. And labor demand is particularly weak for the unskilled manual workers (laborers). Figure 24: Demand for less skilled and elementary occupations is weak (Q2 2007) Unemployment - Vacancies Ratio 8.0 6.0 4.0 2.0 0.0 1. Legislators, 2. Professionals 3. Technicians 4. Clerks 5.Service 6. Skilled 7. Craft and 8. Plant and 9. Elementary U-V ratio 1.9 2.4 2.5 6.2 4.8 4.2 1.2 1.1 5.4 U-V ratio Source: Ministry of Labor and Social Affairs. Bank staff calculations 3.6 There is little regional variation in the skill profile of labor demand. The general pattern of the excess supply of administrative and service workers and the shortage of skilled manual workers prevails in virtually all regions. However, there are some exceptions, and in some regions the skill shortage/surplus is more pronounced than in others. The capital region stands out as the only region where is a strong and unmet demand for elementary occupations, i.e. unskilled manual workers (Table 19). The U/V ratio for elementary occupations in the Prague regions is as low as 0.3 and is lower than for any other category of labor. Such a low U/V ratio unambiguously points to a marked shortage of simple manual labor. So apparently there is a large market for low skilled workers in the capital region. 3.7 The regional differences in demand for low-skilled labor imply that labor mobility could alleviate the imbalances between the supply and demand. Low- skilled workers could improve their employment chances by moving from regions where demand for low-skilled manual labor is low to regions where the demand is higher, in particular to the Prague region. However, mobility is costly and low-paid workers often cannot afford to move to the booming regions, where housing costs tend to be particularly high. In addition, demand for low-skilled labor is highly concentrated in just one ­ the capital ­ region, which has a limited absorption capacity. These two factors imply that the problem of the excess supply of low-skilled labor can hardly be addressed 68 by regional mobility alone. Indeed, as presented in Chapter 2, few Roma report commuting for work. Table 19: Index of labor shortages/surplus by occupation and region, 2007 Region/Occupation 1 2 3 4 5 6 7 8 9 Praha -50 -25 -30 -64 -40 -67 86 23 155 Stedoceský kraj 11 -11 -15 -64 -50 3 209 206 -36 Jihoceský kraj 22 -31 -12 -43 -32 -41 114 139 -60 Plzeský kraj 19 -24 -16 -63 -41 -44 235 222 -43 Karlovarský kraj 200 202 155 1 24 -38 325 295 -77 Ústecký kraj 138 114 102 -47 -49 -40 162 212 -81 Liberecký kraj 68 -20 35 -40 -26 -30 114 94 -78 Královéhradecký kraj 8 29 30 -62 -8 -10 211 124 -62 Pardubický kraj 34 -4 8 -56 -32 -61 90 257 -71 Vysocina 5 -9 -4 -73 -49 -40 138 102 -43 Jihomoravský kraj 48 28 10 -55 -40 80 134 214 -33 Olomoucký kraj 148 20 39 -45 -20 -64 312 196 -53 Zlínsky kraj 116 25 11 -66 -40 1 274 247 -68 Moravskoslezský kraj 265 53 38 -55 -45 -46 167 217 -81 Occupational codes are as follows: 1 = Managers; 2 = Professionals; 3 = Technicians and associate professionals; 4 = Clerks; 5 = Service workers and sales workers; 6 = Skilled agricultural workers; 7 = Craftsmen; 8 = Plant and machine operators and assemblers; 9 = Elementary occupations. Note: The index of relative skill shortage/surplus is defined as [(vi / u ­ 1)*100] where for each region vi i is the share of vacancies in occupational group i in the total number of vacancies in the region, and u is the i share of unemployed in occupational group i in the total number of unemployed in the region. A positive value of the index indicates a shortage of workers in the given occupation, and a negative value indicates a surplus. Source: Ministry of Labor and Social Affairs; Bank staff calculations. 3.8 Job opportunities for low skilled are particularly bad in traditional high unemployment regions, although there is some variation at the sub-regional level. Figure 25 (upper panel) shows that the ratio of unemployed to vacancies ("U/V ratio") for elementary occupations is dramatically high in those regions which traditionally have had high unemployment. This is notable given the fact that the tightening labor market in the Czech Republic has also triggered a reduction in unemployment in those high unemployment regions. However, the large U/V ratios in high unemployment regions also hide substantial variation at the sub-regional level. The lower panel of Figure 25 presents a breakdown of U/V ratios for different counties in Ustecky kraj, and a similar picture can be found across other high unemployment regions67. Identified Roma localities can be found both in counties with high and low demand for low-skilled workers, suggesting that the local employment conditions vary for Roma workers across the Czech Republic, but also across the high unemployment regions where many Roma reside. Lastly, Prague stands out in terms of substantial excess demand for elementary education workers. 67See Annex for more information of U-V rates across different regions. 69 Figure 25: Job prospects for low education workers are particularly bad in high unemployment regions, although there is variation at the sub-regional level Elementary occupations: U-V ratio 50.0 40.0 30.0 20.0 10.0 0.0 ja kr ýk jark jark ja kr jark ý jark ký ajrk ja ja aha kr kr ýkcets j Pr Ú ezslsoksvaro j kra ýksravorl ýkcereb yk ýkc ckedar j kýci vsaro anicos ýk ýk nsílZ kra Vy ou kra ýksec Li oml éhvo álr omhiJ cesohiJ nsezlP Ka O ubdraP K doertS M Ústecký kraj: Elementary occupations: U/V ratio Ustecky kraj: Unemployment by duration 120.0 100% 100.0 80% 80.0 60% 60.0 40% 40.0 20% 20.0 0% 0.0 Most Teplice Chomutov Ústí nad Louny Dcín Litomice Teplice Dcín** Chomutov Ústí nad Most Louny Litomice Labem Labem 12 + months 6 - 12 months 1 - 6 months Source: Staff calculations based on MoLSA data for Q2, 2007.Note: A starred county indicates that less than 50 vacancies were registered there, and therefore a cautious interpretation of the results is needed 3.9 The problem of long-term unemployment is more pronounced in counties where the demand for low-skilled labor is weak. However, there is also some tentative evidence that the incidence of long-term unemployment is higher in those counties where many Roma reside. Figure 5 illustrates this pattern.68 The incidence of long-term unemployment increases with the increase in the U/V ratio for elementary occupations, that is with the fall in demand for low-skilled labor. Interestingly, the U/V ratio for elementary occupations is a much better predictor of long-term unemployment that the overall U/V ratio (correlation coefficients of 0.63 and 0.41, respectively).69 This means that it is the demand for low-skilled labor rather than the overall labor demand that is the main determinant of long-term unemployment. This is consistent with the earlier observation that long-term unemployment persists in the Czech Republic despite the tight 68Regression analysis indicates that the "Roma" dummy variable is not statistically significant at 10% percent level (the p-value is 0.16 in a linear specification). This result may be due to the small sample size, however. 69The U/V ratio for elementary occupations is also a significantly better predictor of the incidence of long- term unemployment than the U/V ratio for jobs requiring only primary occupation. 70 overall labor market. Put differently, the low-skilled workers such as Roma do not necessarily benefit from the overall increase in labor demand70. Figure 26: The stronger the demand for low-skilled labor the lower the incidence of long-term unemployment R 60 nte o m o oy R R o o pl R menu 50 o oo R oo o o o mret- o oR R o o ngolfo 40 oooo oR o oR R e o nc R deicnI 30 o o 20 0 50 100 150 200 U/V ratio, elementary occupations Note: Counties with marginalized Roma localities are indicated by "R". Source: Ministry of Labor and Social Affairs. Bank staff calculations. Box 6: Hiring Roma ­ what do employers think? Analyzing barriers to Roma employment in the Czech Republic requires understanding the views of employers with respect to hiring Roma. To shed light on the employer's perspectives, the research for this study involved structured qualitative interviews with 20 employers across the Moravia and Bohemia regions, small and big companies and foreign and Czech-majority owned ­ all of which had experience with hiring Roma workers 71. In the interviews, employers reported a number of barriers to employing more Roma, involving the interplay of low educational attainment and missing skills among Roma, lacking work motivation, strong competition from foreign workers and a higher perceived risk and costliness of hiring Roma. The expressed views can be summarized as follows: Skills: The primary barrier to hiring more Roma cited by the interviewed employers is lacking skills among many Roma workers. Employers reported that there are fewer and fewer manual non-qualified jobs available and that most jobs required more skills and qualifications than most 70Obviously, the positive correlation between the unemployment/vacancy ratio and the share of long-term unemployment is only a statistical regularity and there are deviations from this pattern. For example, Most is a county where the incidence of long-term unemployment at over 60 percent is the highest, while the U/V ratio is relatively low. This example shows that the incidence of long-term (registered) unemployment is also affected by factors other than labor demand (incentives and propensity to register with employment offices, demographic factors, industry mix, etc.). Still, it is quite clear that job opportunities for the low- skills are the decisive influence. 71This box summarizes a qualitative background report on employers' views on Roma employment conducted by the institute for Research on Social Reproduction and Integration of the Faculty of Social Sciences at Masaryk University Brno for the purposes of this study. 71 Roma applicants are able to provide. This applies in particular to limited knowledge of Czech language and functional literacy ­ for example the ability to adjust to new production methods and read manuals of new machines ­ confirming the findings of Chapter 2 of this study. Meanwhile, employers reported a lack of skilled workers on the one hand and their lacking readiness to hire and invest in training of not sufficiently skilled, yet available workers. The reasons employers give are lacking resources due to the competitive market environment and their worry to lose trained workers to competitors ­ two barriers to training found in many countries in Central and Eastern Europe. Work motivation: Employers reported experiences with lacking work motivation and motivation to train and seek more education among Roma. As primary reasons they identified the level of social benefits, but many also cited less well understood, cultural barriers to work motivation. Competition from foreign workers: Czech companies have benefited for several years from an inflow of often seasonal foreign workers, mostly from Ukraine but also from Poland, Slovakia and other places. Employers interviewed for this study report that foreign workers have been strong competition for Roma, and indeed ethnic Czechs, in the labor market. According to employers, foreign workers have not only often been more qualified than Czech workers, but also typically more motivated and reliable and, crucially, ready to work for a lower salary than Czech workers. Risk and costliness of hiring Roma: Interaction with Roma in the hiring process and in work was seen as more costly by most interviewed employers. This is due to perceived and actual communication barriers with Roma and often apparent mutual mistrust and suspicion. Employers also cited the risk of "substandard" behavior of Roma once employed, for example irregular job attendance and high fluctuation. Some employers also made reference to the fact that integration of Roma in the work place is often not straightforward and requires extra efforts on the side of the management. For example, one employer reported the need to ensure that Roma workers were not grouped together in the operation (suggesting this was lowering work morale), but integrated in predominantly non-Roma teams. Taking such precautions is associated with Roma workers, but not with non-Roma workers. Youth: Employers noted a "generational divide" among Roma, with older workers being more skilled and motivated than younger Roma. Such "older workers" are defined as those that received the bulk of their education and training before 1989, and have been with their firm for some time. As a result, older Roma workers are more integrated, and more attractive workers than the young. Looking ahead, employers' expectation vis-ŕ-vis Roma workers can be summarized as "integrate, educate, work" ­ with the primary responsibility for change resting with the Roma rather than with the majority population. Some, however, highlighted the responsibility of the State to make the first step in providing incentives for behavioral change, rather than just relying on keeping people passive in a social welfare trap. Only a minority accepted a responsibility among the employers for promoting change. There was consensus among interviewed employers that eventually greater chances in the labor market require better skills which require full integration, pointing strongly towards educating the next generation to ensure young Roma leave school with better qualifications than current and recent graduates and more ready for integration. 72 CONCLUSIONS 3.10 Overall, there is low demand for elementary occupation workers in the Czech Republic ­ which amounts to a major barrier for employment of marginalized Roma. The demand for low-skilled labor ­ which is a key factor determining Roma employment prospects ­ is limited, and in virtually all regions (except Prague) there is excess supply of low-skilled workers. Even when actively looking for employment, their skill level often will not permit them to readily find employment. This is also a finding confirmed by employers. In addition, they face strong competition for such positions from foreign workers who are often more qualified than Roma. 3.11 However, the demand for low-skilled labor varies within regions, and some Roma workers live in counties where the demand for low-skilled labor is less depressed, while others live in counties where it is extremely limited. Thus, different Roma communities face different labor market prospects; there is no uniform pattern of Roma settlements. In those counties where the demand for low-skilled labor is relatively high the incidence of long-term unemployment is significantly lower. Given that the incidence of long-term unemployment is particularly high among Roma workers, this suggests that Roma employment outcomes are closely linked to local labor market conditions. 3.12 Migration and commuting is one way to overcome local low demand for low- skilled workers, but it is costly. Some Roma workers can improve their employment chances by moving (either through commuting or through migration) to regions where the demand for low-skilled labor is higher. However, for many poor workers both migration and commuting are costly options. More importantly, migration only is a short-term coping strategy. The demand for low-skilled labor is low relative to the supply, and declining due technological progress, which is biased towards higher skills. Consequently, the only sustainable long-term way to improve Roma employment prospects is investment in human capital. To succeed, the next generation of Roma workers needs to possess skills and competences that enable them to successfully compete for skilled jobs. 73 4. EMPLOYMENT INCENTIVES AND THE SOCIAL WELFARE SYSTEM The motivation of the unemployed to seek formal employment depends crucially on whether work "pays" in comparison to the receipt of social benefits. In short, it needs to be obvious to beneficiaries that they become better off as a result of finding employment. While this is typically obvious for well-paid jobs and highly skilled job-seekers, it often is not for those at the bottom of the skills distribution and who can expect wages close to the minimum wage as a result. Indeed, as Chapter 2 showed, social welfare benefits received by Roma in marginalized localities appear close to the minimum wage, suggesting that many Roma may suffer from an inactivity trap. This chapter formally analyzes the incentives for unemployed resulting from the social benefit and tax system, comparing the benefit and tax systems in place in 2007 and, after a recent reform, in 2008. It pays particular attention to long-term unemployed and families of different sizes. It finds that, while in 2007the tax and benefit system retained substantial work disincentives in particular for unemployed expecting to earn just above the minimum wage, employment incentives in 2008 have been significantly strengthened, in particular following a tightening of benefits for long-term unemployed. 4.1 With a view to improving labor market outcomes OECD and EU countries have increasingly introduced revisions to the tax and benefit system to "make work pay", either through tightening access to and level of "out of work" social welfare benefits or through increasing "in work" benefits via tax breaks and other incentives or through a combination of both. For example, Germany under its "Hartz" labor market reforms, has tightened benefit rules for long-term unemployed and reduced benefit levels. Slovakia has reduced social assistance benefit levels, as has Switzerland. Portugal and Hungary have decreased the duration of receipt of unemployment benefit72. Bulgaria has introduced a time-limit for the receipt of the Guaranteed Minimum Income benefits for able-bodied working age individuals. Often complementing a tightening of benefit eligibility, France, Hungary, New Zealand, Netherlands, Switzerland and the UK have strengthened in-work benefits. 4.2 The Czech Government has recently introduced a reform to the social welfare and child protection system as well as the tax system to strengthen pro-work 72OECD (2007a) 74 incentives. The reform, put in effect on January 1 2008, transfers some of the tax burden from capital to consumption and affects labor taxation.73 Specifically, it (a) introduces a cap on social security contributions, (b) lowers the corporate income tax from 24% to 19% by 2010 (and broadens tax base), (c) increases the lower VAT rate from 5% to 9%, (d) puts (at least a temporary) halt to indexation of social benefits (but not pensions), (e) introduces an unusual flat personal income tax (PIT) rate (applied to the so-called `super gross wage', including employer contributions), (f) expands child tax credits (bonuses deducted from tax, not from tax base), (g) lowers the income eligibility threshold for child benefits, (h) abolishes joint taxation of married couples and allows parents to choose different combinations of entitlement length and benefit level for parental allowances (but it keeps the quantitatively more important spouse tax deduction), (i) introduces tax credits for retirees, and (j) lowers welfare support for inactive long-term unemployed. 4.3 This chapter aims to shed light on the nature of the social support system in the Czech Republic­the combination of unemployment and social benefits offered with income taxation, and whether it creates incentives to supply labor and participate in the official economy. To this end, it presents two simple measures of the monetary incentives built into the set of main government labor-market programs that affect labor supply decisions: the Net Replacement Rate (NRR) and the Marginal Effective Tax Rate (METR). The Net Replacement Rate is the ratio of net income of the household when unemployed to the net income of the same household under the alternative situation when one of its members has a job. The Marginal Effective Tax Rate, on the other hand, asks how much of a given (small) increase of the gross wage is taken from a household by both explicit income taxation and the implicit taxes of the social benefit system74. The chapter extends existing calculations in two dimensions: it concentrates on family types typical of socially excluded, and compares current institutional settings to those in place in 2007. Specifically, the calculations in this chapter quantify the combined effects on pro-work incentives of reform steps (e), (f), (g), (h), and (j) above.75 It is important to note that the measurements cannot fully reflect one key dimension of the Czech social support system, unaffected by the reforms, but important for socially excluded, namely the complicated nature of the system resulting in uncertainty about benefit level and availability. It also does not factor in any income that the unemployed benefit recipients may earn from informal employment. 73 A useful memorandum item is that the Czech tax structure is highly unusual in international comparison. It relies heavily on social security contributions (which are among the highest in the world) and the corporate income tax, and less on taxation of personal income, consumption and property. 74 For a detailed description see the Annex. 75 These changes were introduced on top of the expanded accommodation allowance (regionally set housing support) introduced already in 2007. 75 THE CZECH TAXATION AND SOCIAL BENEFIT SYSTEM AND RECENT POLICY CHANGES 4.4 The Czech Government has introduced substantial changes to its tax-benefit system in 2007 and 2008 aimed at strengthening work incentives and making work "pay". This section reviews the main changes introduced. From January 2008, the system is to follow the following sequence of policy rules: · Persons receive earnings, unemployment benefits, sickness benefits and pensions. · Gross earnings determine the income tax base from which social and health insurance contributions and other tax exemptions are deducted; next, personal income tax (PIT) is determined. In the next step, child tax deductions are deducted depending on the presence of children. If the resulting tax amount is negative, the household is eligible for a tax bonus. · Net household income for the purpose of income testing for means-tested state social support benefits equals income net of taxes and contributions, plus stipends, housing supplement from employers, alimony, sickness benefits, unemployment benefits, pensions and income from abroad. The child tax bonus is not included. · Based on information on the composition of the household, the household-level amounts of so-called Minimum Living Standard (MLS) and Existential Minimum (EM) are determined. · An income test for child allowance is performed using the net household income. · An income test for social allowance and housing benefit is performed using the net household income from the previous step, including child allowance. · An income test for social necessity benefits (described below) is performed using the income from the previous step including social allowance, housing benefit and child bonus. Taxes Payroll taxes 4.5 Payroll taxes ­ health and sickness, employment and pension fund contributions ­ are paid from the gross wage of households or from gross profit of entrepreneurs. The contributions are divided between employees and employers, with the employee paying 12.5 percent and the employer 35 percent of the gross wage. There was a significant change in the structure of personal income tax (PIT) in 2008, as this tax is now calculated from the so call "super-gross" wage which includes not only the worker's gross tax-base (as defined in 2007), but also the payroll contributions paid by the employer. 76 Income Tax 4.6 The Czech Republic introduced a flat income tax in 2008. Until the end of 2007, the Czech Republic had a progressive taxation system. Table 20 depicts monthly gross wage income brackets as well as marginal tax rates for every bracket. In 2008 this system was replaced by a 15% flat tax. However, payroll taxes including those paid by employers are not deductible from the tax base anymore. Table 20 Monthly gross wage income brackets and marginal tax rates in 2007 Net Tax Base (CZK) Marginal Tax Rate +(%) 0-10,100 12 10,100-18,200 20 18,200-27,600 25 27,600- 32 4.7 In 2007 married couples with at least one child (all living in the same household) could choose to fill out a joint tax return. The tax base of the couple would then equal the sum of their personal tax bases. The base would be split evenly among both taxpayers and the (equal) tax rate determined separately. Given the progressive tax schedule, joint taxation reduced the taxation of married couples as the income of the spouse with higher earnings was taxed at a lower marginal rate. In 2008 joint taxation was abolished due to the introduction of the flat tax rate. In 2007 tax deductions from actual tax were introduced to replace deduction from tax base. With the introduction of the flat tax, the level of these tax deductions was dramatically expanded. Table 21 lists monthly deductions from tax duty in 2007 and 2008. Table 21 Monthly deductions from tax duty, 2007 and 2008 In CZK 2007 2008 Personal Deduction 600 2,070 Deduction for Wife with low income 350 2,070 Child Bonus 500 890 4.8 The Child Bonus works differently from other deductions. If the tax duty after the deduction of the Child Bonus is negative, the difference is called a tax bonus and is paid to the taxpayer as a bonus, but only to families with at least one member working. This bonus is not tested as an income for state social support benefits. Social Benefits Unemployment Benefits 4.9 Unemployment benefits are available for individuals actively searching for a job who were employed for at least 12 months in the previous three years. The basis for calculating unemployment benefit includes income net of social insurance contributions and income tax, i.e. the average net monthly wage in the previous job. The amount of unemployment benefit is determined as 50% of the previous income in the first three months and 45% in the following three months of the unemployment spell, but not 77 more than 2.5 times the MLS of an adult one-member household (CZK 7,815 = 2.5 times CZK 3,126, see table below). There were no changes to unemployment benefits between 2007 and 2008. State Social Support 4.10 State Social Support defines three main types of means-tested benefits: Child Allowance, Social Allowance and Housing Allowance. The level of available social support depends on the household income relative to the household level of Minimum Living Standard (MLS), which equals the sum of MLS of all family members. Table 22 presents the thresholds for the calculation of State Social Support. They remained unchanged between 2007 and 2008. Table 22 Minimum Living Standard thresholds, 2007and 2008 MLS level in CZK Single Household 3,126 First Adult 2,800 Other Adult 2,600 Child below 6 years 1,600 Child between 6 and 15 1,960 Child between 15 and 26 2,225 Child Allowance 4.11 Child Allowance has been limited to low income families in 2008, yet it is less generous for the least well off. In 2007 a family was eligible for child allowance if it had an income including net salary, unemployment benefits, and excluding child bonus below 4 times MLS. In 2008 the ceiling was limited to 2.4 times MLS. Table 23 summarizes the levels of the benefit. Another difference vis-ŕ-vis 2007 is that families with income below 1.6 MLS receive the same level of benefit as those with higher income, as long as it stays below 2.4 MLS. The eligibility for child allowance is conditional on the child attending compulsory schooling. Table 23 Child Allowance benefit structure, 2007 and 2008 In CZK 2007 2008 Income < 1.6 MLS < 2.4 MLS < 4.0 MLS <2.4 MLS > 2.4 MLS Child < 6 576 496 256 500 0 Child 6-15 706 608 314 610 0 Child 15 ­ 26 810 698 360 700 0 Social Allowance 4.12 The level of Social Allowance has declined between 2007 and 2008. If family with at least one child has an income including net salary, unemployment benefits, child allowance and excluding child bonus below 2.0 MLS in 2007 (1.6 MLS in 2008), it is eligible for Social Allowance. The levels of benefits are summarized in the following table. 78 Table 24 Social Allowance benefit structure, 2007 and 2008 In CZK 2007 2008 Income < 1.0 MLS < 1.6 MLS < 2.0 MLS <1.0 MLS < 1.6 MLS Child < 6 873 437 146 800 320 Child 6 -15 1,070 535 179 980 392 Child 15 -26 1,228 614 205 1,125 450 Housing Benefits 4.13 The level of Housing Benefits has increased in 2008 over 2007. The benefit depends on the actual amount spent on housing compared to total income, and the so- called "Socially respectable cost of living". Total income includes net salary, unemployment benefits, child allowance, but not social allowance and child bonus. Socially respectable costs of living depend on the number of family members and the size of the municipality where the house or flat is located. Table 25 depicts socially respectable cost of living for rented apartments across municipalities of different size. The Housing Benefit is then calculated as socially respectable cost ­ (Income * 0.3). The previous table shows that between 2007 and 2008, socially respectable costs of living have been raised. The coefficient 0.3 can be interpreted as saying that housing expenditures as high as 30% of income are socially acceptable. Table 25 Housing Benefit structure, 2007 and 2008 2007 2008 Population > 100,000 > 50,000 > 10,000 > 100,000 > 50,000 > 10,000 1 member 2,893 2,659 2,518 3,383 3,155 2,895 2 members 4,233 3,913 3,721 4,998 4,686 4,331 3 members 5,858 5,440 5,188 6,971 6,563 6,099 4 and more 7,453 6,948 6,644 8,824 8,332 7,772 Social Necessity Benefits 4.14 Social Necessity Benefits serve as an income source of last resort. It is intended for families with net income (including state social support) below the household Minimum Living Standard level ­ even after receiving State Social Support benefits. Social Necessity Benefit includes two types of benefits. The Housing Supplement increases income to meet the level of actual cost of housing, where necessary. The Livelihood Benefit levels up income of household to MLS or Existential Minimum (EM, to be defined below) after costs of housing are paid. This means that after paying for housing, a given family ought to have income as high as MLS or EM. However, the right for social necessity benefit is not automatic. It is up to each municipality to decide whether to grant these benefits or not. The decision depends on the subjective state, activity, etc. of the family in question, e.g., their willingness to move to a cheaper flat. 4.15 Set significantly below the level of the MLS, the Existential Minimum serves as a testing income level for the purpose of determining social necessity benefits for family members unemployed for more than 12 months. In these cases, the Existential Minimum replaces the Minimum Living Standard. The Existential Minimum (EM) 79 threshold is currently CZK 2,020 ­ 65 percent of the MLS level for a single adult and 70 percent of the MLS level for the first adult person of a jointly assessed household. The Existential Minimum was originally introduced in 2007 to "punish" inactivity; however, it was not used in practice as the Labor Offices found it difficult to determine whether long-term unemployed beneficiaries were or were not actually searching for jobs. Since 2008, the EM, instead of MLS, automatically serves as the benchmark for determining minimal income of those with more than 12 months of unemployment. The EM is not applied to dependent children, people with a disability and to workers on pension or over 55 years of age. However, the EM is applied to LTU adults in families with children. The simulations for 2008 below take EM into account when calculating long-term NRR and long-term METR. Table 26 Summary of Changes to the Tax and Benefit system 2006-2008 2006 2007 2008 Tax rates, percent 12; 19; 25; 32 12; 19; 25; 32 15 Minimum Living Standard Incl. accommodation Excl. Excl. accommodation accommodation Child Allowance, threshold 1.1; 1.8; and 3.0 1.5; 2.4; and 4.0 2.4 times MLS times MLS times MLS Social Allowance, threshold 1.6 times MLS 2.2 times MLS 2.0 times MLS Separate accommodation Moderate Expanded Expanded allowance Cap on social security No No Yes contributions Tax credits Moderate Moderate Expanded Existential Minimum for LTU Not defined Defined, not used Used Joint taxation Yes Yes No EMPLOYMENT INCENTIVES IN THE TAX AND BENEFIT SYSTEM 4.16 How do the policy changes and other factors impact on the labor supply decisions of individuals? This section presents results from simulations of NRR and METR for one-earner families with 0 to 5 children, with incomes ranging from close to the minimum wage (8,000 CZK) up to average production wage (20,000 CZK). The key assumptions for the analysis are stated in Box 7. It also reviews the effect on work incentives of widespread indebtedness of Roma long-tern unemployed. Box 7: Net Replacement Rates and Marginal Effective Tax Rates The Net Replacement Rate (NRR) is defined as the ratio of net income when unemployed to the net income when employed. The ratio takes values from 0 to 100. The higher the ratio, the lower the incentives to look for an employment opportunity. For example, at a NRR of 100, there are no monetary incentives to look for a job, since a given household receives the same level of income independent of the employment status. Since households enjoy not only consumption but also leisure and also face search costs and fixed costs of participating in the labor market (transportation to work, higher cost of food outside of the household, etc.), it is reasonable to expect that even net 80 replacement rates significantly below 100 may not provide sufficient incentives for job search. Furthermore, the actual level of NRR that can be expected to effectively generate labor-supply incentive depends not only on valuation of leisure and transaction costs, but also on outside options such as shadow-economy employment opportunities and others. The Marginal Effective Tax Rate (METR) shows which part of a change in earnings is "taxed" away from the combined effects of income taxation and social benefits taken from a household that experiences an increases in its gross earned income. Because the maximum level of benefits is often received by families with no income and because benefits are sometimes reduced almost CZK for CZK with additional earnings, the welfare system is expected to discourage labor force participation and hours of work. It is therefore important to measure the strength of such disincentives. METR rates of over 100% indicate strong inactivity traps where individuals have negative incentives to increase their gross income. This chapter presents METRs for CZK 200 rises-- corresponding to a gross-income increase of 2 percent at (monthly) income level of CZK 10,000. For all household types, the analysis contrasts the so-called short-term and long-term NRR and METR. Short-term rates assume that one family member is unemployed for a short time period such that s/he is receiving unemployment benefits, and that the other family member is long-term unemployed, thus not receiving unemployment benefits. Unemployment benefits are calculated assuming that they are based on a wage that corresponds to the wage opportunity considered in the NRR calculations. For example, when the potential wage is CZK 10,000, it is assumed that the previous wage (before unemployment) was also CZK 10,000. Moreover, the MLS rather than the EM is used as the testing level for Social Necessity Benefits when calculating short-term rates. On the other hand, long-term NRR and METR rates, which are relevant for long-term unemployed, correspond to the situation when the family does not receive unemployment benefits any more. In 2008 it is also assumed that the EM instead of the MLS serves as the testing level for Social Necessity Benefits of long-term unemployed. Finally, a number of assumptions are made on the family status. It is assumed that children are between 6 and 15 years old (so we abstract from parental allowances). It is also assumed that they live in a municipality with up to 100,000 inhabitants and that their actual cost of housing is the same as the socially respectable cost of housing. This means that they do not pay more than the level defined by law. It is also assumed that households satisfy all conditions to receive social necessity benefits. The impact of recent policy changes 4.17 Previously very high Net Replacement Rates have declined significantly as a result of the 2008 reforms, suggesting that labor supply disincentives have been reduced. Figure 27 shows short term and long-term NRRs for different family size configurations for 2007 and 200876. Short-term replacement rates depict the situation for families whose breadwinner has been unemployed for less than 12 months, i.e. receive 76See Annex for the full range of simulations. 81 unemployment benefits and are subject to the MLS. The long-term replacement rates present the picture for those who have been unemployed for more than 12 months and are not eligible for unemployment benefits and are subject to the lower EM instead of the MLS. There are two main observations from the simulations: · The high NRRs of close to 100 percent in 2007 suggest that the system so far has had powerful labor supply disincentives, in particular for those whose prospective labor earnings are low and for those who are able to earn income from informal jobs on top of social benefits. For example, the simulations suggest that in 2007 a couple with 4 children was as well off on benefits (NRR of 100 percent) as they would have been with a salary of CZK 13,000 which is 65 percent of the average production wage in the Czech Republic in 2007. · The 2008 reforms have decreased NRRs, in particular for the long-term unemployed. The decline is more pronounced for lower income levels and is typically in the order of 20 percentage points for long-run net replacement rates. The fall in NRRs is greater for the long-run rates, reflecting the powerful effect of the application of the lower EM for long-term unemployed. Also, as shown in Figure 27, the drop in NRR is typically more pronounced for low income families with higher number of children. As opposed to that, the decline remains limited for single adult households, where NRR levels were low already prior to the reform. The sudden increases in short-term NRRs for families with 3 and more children, which are moving to higher wage levels as the number of children increases, are caused by the availability of social and child allowances (which grow as a function of family size)77. Figure 27: NRRs have fallen substantially as a result of the reform, reducing work disincentives NRR, Couple 2 children, Short-term NRR, Couple 4 children, Short-term 105% 105% 100% 100% 95% 95% 90% 90% 85% 85% 80% 80% 75% 75% 8000 10000 12000 14000 16000 18000 20000 8000 10000 12000 14000 16000 18000 20000 Income (CZK) Income (CZK) 2007 2008 2007 2008 77Once family earned income exceeds the eligibility threshold, the denominator of NRR, the total family income (earned income plus allowances) decreases because allowances are no longer available. Hence the increase in the NRR. 82 NRR, Couple 2 Children, Long-term NRR, Couple 4 Children, Long-term 105% 105% 100% 100% 95% 95% 90% 90% 85% 85% 80% 80% 75% 75% 70% 70% 65% 65% 60% 60% 8000 10000 12000 14000 16000 18000 20000 8000 10000 12000 14000 16000 18000 20000 Income (CZK) Income (CZK) 2007 2008 2007 2008 Source: Jurajda and Zubricky (2008) 4.18 The METR calculations suggest that few inactivity traps remain. Figure 28 presents METRs for couples with 2 and 4 children78. The analysis suggests that some of the inactivity trap income regions have moved to lower income levels. Some of the "spikes" in METR are extreme thanks to the unification of thresholds for social and child allowance. Similarly, the increase in METR for single adult childless households at low wage levels is due to the unified levels of social allowance. However, the overall number of inactivity trap regions has declined for some household types, which is due to greater simplicity in the tax code. To illustrate the magnitude of some of the spikes in METR presented in Figure 28, which correspond to 200 CZK wage increases, the family head would have to obtain a wage gain of about 800 CZK per child in order for the spike to be eliminated, i.e., in order to stay at the same total net income level after the wage increase. Figure 28: METRs reveal continued inactivity traps at certain wage levels METR, Couple 2 Child, Long-term METR, Couple 4 Children, Long-term 500% 1000% 450% 900% 400% 800% 350% 700% 300% 600% 250% 500% 200% 400% 150% 300% 100% 200% 50% 100% 0% 0% 8000 10000 12000 14000 16000 18000 20000 8000 10000 12000 14000 16000 18000 20000 Income (CZK) Income (CZK) 2007 2008 2007 2008 Source: Jurajda and Zubricky (2008) 78The full range of scenarios is presented in the Annex to this report. 83 4.19 The previously high replacement rates were a particularly binding barrier in high unemployment and rural regions where wage levels are lower than in the booming urban settings. The national averages presented have very different implications in different parts of the Czech Republic. Based on the calculations one can also focus on these comparisons taken at the actual regional average construction wage in 4 different regions: low unemployment rural, high unemployment urban and low unemployment urban. For example, using (i) Znojmo for high unemployment rural (12 percent unemployment rate, 42 percent of population in towns), (ii) Mlada Boleslav for low unemployment urban (2 percent unemployment, 70 percent of population in towns), (iii) Rychnov for low unemployment rural (3 percent unemployment, 58 percent of population in towns) and (iv) Usti nad Labem for high unemployment urban (11 percent unemployment, 85 percent of population in towns), one obtains the following mean (median) monthly salaries in construction from 2006: 16,460 CZK (13,540) in Znojmo, 24,352 CZK (21,366) in Mlada Boleslav, 22,026 CZK (18,225) in Rychnov and 21,039CZK (15,280) in Usti. The specific NRR and METR rates can now be read from the enclosed graphs. However, these wage levels are unrealistically high for socially excluded. Focusing on the same wage statistic for construction workers in these regions with only elementary education level, we obtain the following median wages: 9,510 CZK in Znojmo, 20,829 CZK in Mlada Boleslav, 13,287 CZK in Rychnov and 13,841 CZK in Usti. The comparatively low wages levels available in high unemployment regions as well as in rural settings suggest that the labor supply disincentives were particularly strong there. The impact of personal indebtedness 4.20 Personal indebtedness appears widespread among Roma welfare beneficiaries in the Czech Republic, as are incidents of usury. As presented in Chapter 2, many surveyed Roma reported that they are indebted, a finding confirmed by the experts interviewed for the purposes of this study. According to interviewed experts, most frequently they face debt towards commercial credit companies on loans with disproportionate interest growth. They are also in debt to municipal authorities, generally for rent and services connected with housing. The experts also indicated widespread debt to loan sharks. Widespread indebtedness among Roma welfare beneficiaries is not surprising, as debt is a typical feature of long-term welfare benefit recipients in many countries, with beneficiaries turning to various forms of credit to make ends meet. While informal debt within the Roma communities is common in many countries, many Czech Roma welfare beneficiaries have formal debt ­ to utilities companies, financial institutions and others. In particular, there has been anecdotal evidence of usury and active marketing by financial credit institutions for high interest short term loans in Roma communities. 4.21 Formal indebtedness has a powerful impact on labor supply decisions. Debt collectors can only enforce debt collection provided an indebted individual earns formal income in excess of the minimum subsistence level. So long as an indebted individual remains on social welfare benefits which guarantee income to the level of the MLS, debt collectors cannot enforce the debt. In effect, even when a formal sector job pays more 84 than the combination of social welfare benefits and the minimum subsistence level, an individual faces little incentive to take that job, because of the threat of debt collection. Because of the debt and the threat of debt collection, an individual is as well off working in the formal sector as he/she is on benefits. Under such circumstances, the individual may even be worse off in a formal job if he/she manages to earn informal income on top of the welfare benefits. Only if the wage income on top of the MLS is larger than the expected debt repayments the individual will be attracted by a job. With most individuals facing high degrees of uncertainty about possible debt repayments or fearing that they will be infinitely large at the beginning, this calculation is often impossible to make and the safest bet is to remain on benefits. Figure 29 shows a graphic presentation of the labor supply disincentive effect of debt and debt collection. Of course, this disincentive does not apply in the case of informal debt, such as debt within the community, since debt collectors are aware of any informal income the debtor is earning. 4.22 Because of debt uncertainty, the application of a lower Existential Minimum (EM) threshold for social benefits for inactive long-term unemployed may not have the desired effect on labor supply decisions. As noted above formal work would only be attractive if the expected wage income is going to be larger than the expected debt repayment. The EM does not remove the uncertainty about the level of possible debt repayments. Rather, individuals may choose to make up for the loss of income due to the lower EM by enhancing labor supply in the informal sector. Figure 29: Personal indebtedness acts as a key labor supply disincentive CZK debt repayment informal income MLS Benefit income Wage income Note: This is a graphic representation not based on actual data 4.23 With affected individuals unable to break out of the debt spiral without help, debt support needs to become a core element of employment activation policy. Debt support has been an important element of social policy in many countries in the European Union, owing to the fact that many welfare beneficiaries resort to credit. However, the labor supply disincentive effect of indebtedness has been underestimated, and it is important to recognize that employment activation policies as well as revisions to the benefit eligibility and benefit levels are likely to fail to promote greater employment among disadvantaged long-term unemployed, unless the debt issue is addressed. 85 CONCLUSIONS AND POLICY DIRECTIONS 4.24 Work disincentive effects in the Czech social support and unemployment insurance system have been gradually lowered over the last few years. The minimum living standard has been lagging behind wage growth since 2001, the child tax bonus has replaced tax base deductions and there was a reduction in personal income tax rates for the two lowest income brackets during 2005-2006. Finally, in 2008 a number of tax and benefit changes were introduced with the stated aim of increasing labor supply incentives. Average effective tax rates declined in 2008 for both low and high income groups (thanks to tax credits and housing allowances for the former and thanks to cap on social security and flat PIT for the latter group). This section documents significant declines in most NRR rates, especially for families with several children and for long- term unemployed. The 2008 changes therefore make work for low wage more attractive compared to inactivity. 4.25 With the measures to reduce work disincentive effects from the tax and benefit system under way, it is important to monitor their effects prior to introducing further change. Policy directions include the following: · Assess the extent to which the new rules, in particular the Existential Minimum eligibility threshold for long-term unemployed are enforced across the Czech Republic. A tightening of benefit eligibility is always difficult to implement, and there may be difficulties in enforcement. · Evaluate the poverty and employment effect of the introduction of the lower Existential Minimum threshold. This would include both the effect on the share of the poor as well as the depth of poverty for those affected. It could also look at flows of registered unemployed, i.e. whether there has been an increased flow of long-term unemployed on EM into employment. · Make debt workout and financial literacy programs a core element of employment activation for long-term unemployed benefit recipients. International experience suggests that this may be best done through non-governmental and non-profit agencies, for example through making debt work-out part of the menu of outsourced activation services. Debt advisory services are standard elements of social policy in many advanced OECD and EU economies, and one which is crucial to develop for the Czech Republic, if a primary barrier to employment for Roma is to be overcome. Moreover, the provision of financial literacy programs in marginalized localities may help prevent further indebtedness of households and individuals79. 79See World Bank (2007b) and World Bank (2008) for specific recommendations on financial literacy programs and consumer protection and financial services. 86 5. PUBLIC EMPLOYMENT SERVICES AND EMPLOYMENT ACTIVATION Measures to "make work pay" are only one part of necessary inputs to enhancing access to employment for disadvantaged job-seekers ­ effective employment policies targeted at the disadvantaged job-seekers is the other. Many OECD countries have adopted a mutual obligation principle at the core of their employment activation policy which requires both enhanced cooperation from the job-seeker and improved service provision of the public employment services at the same time. Indeed, as Chapter 2 showed, the majority of Roma jobseekers turn to the Labor Office for support in funding employment. This chapter finds that, while obligations of job-seekers have recently been tightened, the Czech Labor Office with its current structure and policies is not well placed to promote greater access to the labor market of disadvantaged long-term unemployed and keep its side of the mutual obligations bargain. An examination of its tools, organizational structure and programs suggests that the Czech Labor Office requires further reforms to meet the changing needs of the unemployed and in particular long-term unemployed and disadvantaged job-seekers. The reform message for the Czech Labor Office is to follow the example of other OECD countries in restructuring its approach to disadvantaged job- seekers by offering an intensive, individualized activation services as well as expanded outsourcing of services for disadvantaged job-seekers to private and non-governmental expert agencies with experience in working with Roma. 5.1 This chapter examines the Czech Labor Office's ability to effectively promote access to employment for disadvantaged job-seekers such as Roma long- term unemployed. It further reviews the elements of modern employment activation policies and its applicability for the promotion of Roma employment in the Czech Republic and provides recommendations for a new approach to activating disadvantaged job-seekers. EMPLOYMENT ACTIVATION ­ FOCUSING ON DISADVANTAGED JOB-SEEKERS Employment activation and recent OECD and EU experience 5.2 Countries across the OECD and the EU that 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 unemployed80. Activation policies typically build on a "mutual obligations" approach combining 80For an overview of activation policies in the OECD see OECD (2007), Employment Outlook 2007, Chapter 5 Activating the Unemployed: What Countries Do; OECD, Paris; and Tergeist and Grubb (2006) Activation Strategies and the Performance of Employment Services in Germany, the Netherlands and the United Kingdom, OECD Working Paper 42 87 · 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 ­ as an obligation against receiving income support such as social welfare benefits81, with · Provision of income support as well as access to public employment services (PES) with focused service for the unemployed, particularly those hardest to place, with. 5.3 Employment activation and mutual obligations open the opportunity for a renewed focus on hard to place Roma long-term unemployed. Traditional active labor market programs and activation schemes need modification in order to work for long- term unemployed and disadvantaged Roma. Activation policies include a range of new management approaches and services for the unemployed. Elements of enhanced services include individualized back-to-work agreements and individual action plans acknowledging diversity (age, experience) and relevant to the individual person's needs, wishes and priorities, regular monitoring and review of the client's job search status, in particular through regular meetings between the client and employment officer and a widened set of active labor market programs, covering training, education, subsidized employment, work placement, group activities, language-learning skills, literacy, etc. It is also usually accompanied by tightening of social benefit eligibility and sanctions in case of non-cooperation by the unemployed so as to incentivize job search. Key elements of modern activation policy include: · Profiling of clients according to their distance from the labor market and priority attention by the public employment service staff to those furthest from the labor market (particularly long-term unemployed)82. Profiling involves assessing the job-seeker's background and employability at or prior to the first interaction with the PES staff and typically leads to the categorization of the individual's distance from the labor market. Rather than spread equally across all job-seekers, PES staff time will be focused on those hardest to place, while those easy to place take advantage of job postings in the labor office or of "virtual PES" through internet, job banks, self-registration via the internet, call centers and others. · An individualized approach for long-term unemployed and at risk job-seekers, with individual reintegration action plans. Individual action plans describe an individual pathway towards employment, involving training and addressing the client's multiple social needs insofar as they deepen labor market exclusion 81 This involves that the employment legislation defines the following indicators to avoid benefit sanctions and exclusion from the roster of registered unemployed: (i) "suitable work", (ii) occupational protection (i.e. allowing unemployed people to refuse a job offer that involves a change of occupation), (iii) requirements for independent job search, (iv) frequency of contacts with the PES, and (v) compulsory participation in programs after a certain period of unemployment has elapsed. 82 Profiling is currently one of the key areas of experimentation and study in European PES (It is important to note that there is already a strong history of using profiling in the US). Some PES have piloted econometric profiling as a scientific way of identifying those clients in most need when they register with PES so that PES resources are properly and cost-effectively targeted. 88 (including health, debt, childcare etc). The individual action plan can also involve the entire contracting out of activation services for the most difficult to place clients-based on performance related payments to the contractors83. · New management and administrative approaches in the PES. These include greater specialization of labor office staff, including on dealing with hard to place clients and investments in training and retraining as well as rotating staff. It also crucially involves culturally sensitive service provision, centered around the individual's needs and abilities, involving dedicated advisers for minority job- seekers in those areas with large minority communities. Some countries have also introduced incentives to employment office staff to focus on the long-term unemployed and reward successful placements of difficult cases. · Service integration, typically either involving a merger of the traditional PES with social welfare offices or introduction of integrated computer systems. Service integration builds on the recognition that job-seekers typically have multiple needs that are best addressed in an integrated, one stop shop manner. It is also typically exploiting synergies and generating savings that can be reinvested in new and more intense service provision and programs. · Focus on prevention and early interventions and youth. Most countries have introduced systems to detect risk groups early on and make them subject to prioritized and individualized attention, in particular youth through career counseling and professional orientation at school. This also involves early drop outs from school at a time prior to becoming long-term unemployed, typically after six months of joblessness, with directions to remedial and second-chance education or work placement and apprenticeship schemes. · Enhanced contracting out of services to private sector and NGOs under performance-based contracts and collaboration with private employment services. Many countries have introduced contracting out partnerships with private sector service providers and/or community based organizations and NGOs, including to facilitate the contact between the employment office and the client and to provide services. The rationale is that outside partners are better positioned to deliver more effective services than the public employment service infrastructure ­ for example NGOs with experience in working in Roma communities. However, outsourcing is likely only effective when coupled with a performance monitoring and measurement system which allows tracking the individual's progress in finding and retaining a job (e.g. through monitoring social insurance contributions). · Regular evaluation of the effectiveness of the employment offices in placing the long-term unemployed, including through regular client surveys, to ensure high quality and focused service provision. 83This shift is reflected for example in Australia where the PES has effectively been outsourced to private and NGO type agencies and a quasi-market established for Employment Services (see Box 9). 89 · Empowering of clients to demand high quality services by introducing client satisfaction report cards where clients can provide feedback on the quality of service they have received. 5.4 Incomplete implementation of activation policy and the mutual obligations principle can trigger substantial risks to the prospects of disadvantaged long-term unemployed, in particular those suffering from severe forms of social exclusion, low skills and discrimination. The challenge, therefore, is to get the policy mix right. As noted, activation policies build on mutually balanced obligations between the individual job-seeker and the employment office. They risk failure if one side is not fulfilling its obligation and no improved employment outcomes for disadvantaged job-seekers. On the beneficiary side, failure to cooperate entails the risk of losing social benefits or getting lower benefits which can have an adverse impact in particular for dependent children in social excluded households. On the employment service's side, failure to provide adequate services or barriers to actual job placement risks that the tightening of the beneficiary's obligations is not matched and that the beneficiary remains worse off ­ with lower benefits and without a job. Ineffective service provision and organization, insufficient attention to the disadvantaged client and a lack of quality interventions on offer which fail to result in a subsequent job ­ all can be binding barriers to making activation work. It has been argued that there is a risk of an anti-Roma bias in modern activation programs precisely for reasons of discrimination, culturally insensitive approaches to Roma clients in the employment offices and low quality interventions84. Moreover, the risk of imbalance also stems from the fact that it is easier to enhance the beneficiary's obligations than the labor office's: Tightening of benefit eligibility is done through simple change in legislation, while enhancing the obligation of the labor office requires time-consuming and complex institutional change. Recent reforms in the Czech Labor Office 5.5 While the Czech Republic has taken first steps to reforms its employment promotion policies, further change is necessary to respond to the changing demands of the labor market and to enhance placement of vulnerable job-seekers. Like its neighbors, the Czech Republic's recent reforms have followed the principle of enhanced mutual obligations between job-seeker and labor offices. · Job-seekers are required to report to the Labor Office on a regular basis to demonstrate their job search requirements (every 2 weeks85, although it is not clear whether this is applied consistently across the country) and can be deleted from the register for failure to cooperate with the LO. 84For a discussion of obstacles for Roma resulting from activation schemes in France and Portugal see Bedard (2007); for a discussion about the situation in the Czech Republic see European Roma Rights Center and Numena Centro de Investicao em Ciencias Sociais e Humanas (2007) 85OECD (2007) Employment Outlook 2007, Chapter 5, Activating the unemployed: what countries do. 90 · Since 2004, long-term unemployed have been obliged to accept temporary job offers or perform public works tasks86. · The government has also introduced measures to underscore the increased conditionality for jobseekers. This effectively places an imperative on the benefit/assistance beneficiaries to engage fully with the PES action planning process or risk losing benefits. Unemployed social assistance beneficiaries will be subject to a lower eligibility threshold for their benefits if they fail to find work within 18 months of first signing on for a benefit (see Chapter 3). · Legislative amendments in early 2007 have introduced a new job search bonus scheme and strengthened in-work benefits within the system of "assistance in material need". · According to the new Act on Employment of 2007, individual action plans, with increased attention for the job-seeker by the employment office, are available for job-seekers below the age of 25 and for university graduates. They are accompanied by heightened obligations to cooperate by the job-seeker. · The Czech Government has proposed to merge the social assistance functions of the municipalities and the PES functions from 2009 onwards by establishing a National Office for Labor and Social Affairs. 5.6 While the beneficiary's obligation to cooperate with the LO has been tightened, the Labor Office remains not fully prepared to fulfill its side of the bargain, and there is recognition by the Government that further reforms are required to raise the effectiveness of the Labor Office in dealing with vulnerable job-seekers. Following the recent reforms, failure of the beneficiary to cooperate with the Labor Office triggers penalties in form of lower social assistance benefits. The Czech Operational Program Human Resources Development (HRD OP) lists a series of proposed reforms and actions in order to achieve enhanced placement of vulnerable job- seekers. Some main lines of proposed actions include more intensive mediation for those who are at risk of long-term unemployment or are already long-term unemployed, including more intensive diagnostics, retraining, support for the creation of new and targeted jobs, support for community work and short-term jobs, organization of work experience, support for self-employment and other measures to enhance the employability of the target groups. With respect to organizational implications, emphasis is placed on staff training, increased action planning capabilities, evaluation mechanisms and development of more elaborate labor market information systems. 5.7 The Czech Government has launched a pilot Social Inclusion Agency to promote innovative partnerships between public services and NGOs in select marginalized communities which provides an entry point for a new approach to promote employment of Roma. The Agency is an instrument to promote complex 86See OECD (2006) Czech Republic Economic Survey 2006, Volume 6/2006, OECD, Paris; Website of the Ministry of Labour and Social Affairs in the Czech Republic. 91 solutions to multiple and interwoven forms of social exclusion at the local level and in individual socially excluded Roma localities. It is to foster partnerships at the local level involving the municipal authorities, schools, the Labor Offices and other locally-provided public services as well as non-governmental organizations and the private sector. The creation of the Agency opens the opportunity for developing innovative social inclusion programs individually tailored to local conditions and taking advantage of locally available know-how and actors. While the Agency will be acting as a service provider itself in the pilot localities, it will outsource the bulk of activities. Its work will be funded through the State budget and the European Social Fund (ESF) and the European Regional Development Fund (ERDF) through various operational Programs87. The following sections of the report aim at providing input into the Government's reform efforts, by presenting and discussing international good practice on employment activation policies. EMPLOYMENT SERVICE ORGANIZATION AND METHODS Job-seeker profiling and service differentiation 5.8 Profiling of clients at registration allows better tailoring of service provision to job-seekers' needs, particularly the hard-to-place. There are a number of approaches to the profiling of job-seekers at the time of their registration at the public employment service, including based on an econometric model. Some EU public employment services categorize their client groups using a graduated scale from those easy to place to those most distant from the labor market. They do this through the use of checklists of factors that they have been identified as predictors of ease or difficulty of placement. These can include factors such as age, length of unemployment, education, language skills, ethnicity, disability, literacy and numeracy, family status and other personal characteristics. Clients are then divided into separate categories, for example four groups like in the Netherlands, and job counselors' time and resources are applied in a differential way to each group. A category one job-seeker (close to the labor market) would be assigned to self-service options and a category four job-seeker (most distant from the labor market) to interventions such as intensive counseling or specially required programs focusing on raising his/her employability. Similar to systems in the US and Australia, the Danish "Job Barometer" is a statistical profiling instrument that, based on information about the job-seeker's individual characteristics, calculates the probability of finding employment within six months. The Job Barometer aims to standardize the otherwise highly varying assessments by job counselors of the individual job-seekers' re- employment chances ­ and therefore be more objective88. 5.9 Profiling of job-seekers allows for service targeting and enhanced services for the disadvantaged without requiring, in theory, extra resources. The rationale is that service prioritization allows for a more efficient allocation of resources available for counseling: Staff time which is currently allocated to relatively easy-to-place job-seekers is freed up and focused on the more difficult cases. Germany divides its job-seekers into 87Office of the Government (2007), Agency for removing social exclusion and its prevention in socially excluded Roma localities 88Rosholm et al (2004) 92 four groups ­ (i) "market clients" that are closest to the labor market, (ii) "clients for counseling and activation which need job search assistance, (iii) "clients for counseling and support" that need designated programs and (iv) "clients in need of supervision" that are furthest from the labor market and need special attention. Under the new German system, training is limited to group (i) and top group (ii) job-seekers ­ those with a minimum of 70 percent chance of finding a job after the training. Job creation measures and public employment are only available for group (iv) clients. Traditional active labor market programs are targeted to group (ii) and (iii) clients. This targeting of measures has led to a reduction in caseloads as well as spending89. 5.10 Managing by objectives and performance incentives for public employment services and their staff can help focus services on the hard-to-place. Critics of profiling have argued that categorization of job-seekers according to distance from the labor market essentially focuses job counselors attention on the easy to place rather than the difficult cases (the so-called "creaming effect"). This can be countered by making available performance incentives tied to the placement of disadvantaged job-seekers under a management by objectives (MBO) system. MBO implies the setting of mutually agreed objectives between, for example, the central Labor Office and regional and municipal offices. Budgets are awarded according to the objectives and outcomes at the end of the year are measured against these objectives. Under such an arrangement, labor offices can receive additional resources aimed to meet an objective of placing more long- term unemployed into jobs. MBOs are particularly suited to a regionalized PES structure, with budgets allocated tied to certain agreed performance benchmarks. The addition of staff rewards is a feature in some PES and has proved to be a successful motivating factor in the achievement of objectives90. The UK Jobcentre Plus and the Swiss public employment service have experimented with such incentives. The UK Jobcentres Plus are subject to a job entry target based on a point system that gives disproportionate weight to disadvantaged job-seekers. Job-seekers with the highest points are lone parents as well as people with a disability, with additional points awarded for people residing in marginalized neighborhoods, defined as having a high share of ethnic minority residents, long-term unemployed and low income households91. 5.11 While Czech Labor Offices record jobseekers' details for benefit payment purposes and on a caseload management computerized system, it does not utilize profiling and caseload prioritization. Research conducted for the purposes of this study has revealed that labor offices are not profiling clients according to their distance from the labor market, and do not allocate proportionately more time to the long-term unemployed. Moreover, labor offices remain institutionally separate from social welfare benefit administrations, and the case workers do not follow clients during active labor market interventions. Each job-seeker is registered in the Labor Offices's database, but not characterized according to his or her distance from the labor market and degree of counseling requirements. Essentially every job-seeker gets equal service. Any interactions with the clients are recorded on the computer system, and a jobseeker must 89For a detailed description of the new German system see Jacobi and Kluve (2006) 90Wissenschaftszentrum Berlin für Sozialforschung (2001) 91For details see Karagiannaki (2006) 93 visit the counselor on a regular basis upon registration, typically at least every two months to avoid sanctions. This system is not fully interlinked both ways with the social assistance database operated by the municipalities where jobseekers register to obtain social assistance. Individual action plans 5.12 An individualized approach to counseling job-seekers, typically through individual action plans, is at the core of modern activation policy and particularly useful for multiply disadvantaged job-seekers such as long-term unemployed Roma. Individual action plans lay out an individual pathway to employment agreed between the job-seeker and the labor office, involving training and addressing the client's multiple social needs (including health, debt, childcare etc). The individual action plan can also involve the entire contracting out of activation services for the most difficult to place clients-based on performance related payments to the contractors (see below). The European Employment Strategy stipulates that all those jobseekers under 25 be engaged with before they pass the six month threshold and that the over 25s be engaged with before they pass the 12 month threshold. Individual action plans are a particularly useful tool for those job-seekers who have multiple barriers to employment, such as a lack of skills, a lack of prior labor market experience and social needs. 5.13 The Czech Labor Office has introduced a mandatory individual action plan regime for unemployed under the age of 25, but it remains optional for those over 25, according to the Act on Employment92. This follows the introduction of a pilot in the Moravskoslezsky kraj, where individual action plans have become a much utilized tool (see below). However, staff in labor offices interviewed for the purposes of this report stated that they are generally making limited use of individual action plans for jobseekers over the age of 25, and sometimes not even for those under 25. The reasons given range from insufficient staff time to lack of interest on the side of the job-seeker. The services to jobseekers are delivered mainly though the job counselors, and there is little specialization and prioritization. Some offices have access to a psychologist on staff who can take more difficult cases. Labor office staff interviewed for the purposes of this report reported that in some cases this service is used for crisis situations rather than enhanced career guidance. Employment services management 5.14 Effective implementation of the activation agenda implies a different operations model for the public employment service, with a greater concentration of resources on the hard-to-place. The nature of employment activation is a departure from the traditional one-size-fits-all public employment service model and, through its inherent specialization and differentiation of services, requires staff retraining and operational restructuring. For example, it requires freeing up PES staff with back office duties to increase the number of frontline staff. It requires bringing in insufficient or missing skills such as psychologists, social workers and others. A greater focus on long- 92Act on Employment, Section 33, para 2. 94 term unemployed and hard-to-place job-seekers requires the inclusion of social work functions into the service mix. Figure 30: Staff-client ratios vary across Czech Labor Offices, with high unemployment municipalities having relatively fewer staff (December 2007) ffats 500 eniltnorf 450400 350 erp 300 stnei 250 200 clfo 150 erb 100 mu 50 N 0 0 10 20 30 40 50 60 70 80 Municipalities ranked by number of unemployed (left highest) Source: Ministry of Labor and Social Affairs; Note: frontline staff defined as mediation associates and counselors; unemployment figures from Q2 2007 5.15 The Czech Labor Office is the central institution in the implementation of national labor market policies and separate from municipal social welfare services and benefit administrations. The Czech Labor Office is tasked with a wide range of functions from the payment of State Social Support benefits to providing employment services to jobseekers. The LO is managed directly by the Ministry of Labor and Social Affairs (MOLSA) and consists of 77 Labor Offices with about 5,500 staff. Staff-client ratios are typically difficult to define in public employment services generally, as an accurate ratio depends on the definition of who are the frontline staff who actually deal with the public. Such ratios do not take into account job changers and unregistered jobseeker numbers. Figure 30 presents data from December 2007 on staff-client ratios from each regional Labor Office in the Czech Republic. It defines "frontline" staff as mediation associates and counselors, i.e excluding staff working on benefit administration and highly specialized staff such as psychologists. The average ratio across the country is 1:20093; however, as Figure 30 shows, there is a lot of variation, with regions with higher incidence of registered unemployment on average facing the less advantageous staff-client ratios. Within the European Union, the average figure is around 1:15094, while the figure recommended by the ILO is 1:100. This suggests that the staff- client ratio for "frontline" staff charged with client interaction in the Czech Labor Office is above EU averages and substantially above the ILO recommendation. 93The Joint Assessment Papers (JAPs) produced in collaboration with the EU Commission just prior to accession, put the staff client ratio in the Czech PES at 1:236. Based on the December 2007 data, the ratio is 1:270 if only counting mediation associates and not counsellors. 94It is worth noting that this EU average figure hides a lot of variation, e.g. with Germany having a ratio of about 1:200 and the Netherlands 1:60. 95 5.16 Some OECD countries have merged employment service and social work functions to provide integrated services to job-seekers, and the Czech Government has proposed to do the same. In the UK, the Jobcentre Plus (JCP) combines the previously separate job placement and benefit administration functions into a one-stop shop for employment service and income support95. In focusing on both inactive and unemployed clients, the service mix of the JCP includes social work functions to address the multiple social needs of marginalized job-seekers. Initial evaluations find that service integration in the JCP has had a positive impact on job entry outcomes, a neutral effect for client service outcomes (speed, accuracy, proactivity) as well as a negative impact on benefit processing accuracy96. Germany has sought to build organizational linkages between the Federal Labor Offices which traditionally managed the employment programs and provision of unemployment benefit and unemployment assistance and the municipal social welfare departments which managed social assistance benefits. Following the experience in other EU countries, the Czech Government has proposed to merge the Labor Office and the municipal social welfare service and benefit administration; however, it is uncertain at this point as to whether this proposal will be adopted. 5.17 In order to deliver high quality individual action planning it will be necessary to invest in staff training and necessary new skills. Staff training needs to be systematic, providing both short and long-term options. Many PES have developed long-term training programs for staff and some of the ideas would seem to be transferable (see Box 8 with an example from Ireland). Consideration should be given to the development of part-time and distance based training for staff delivered over a period of one or two years to provide staff with the skills to do their jobs more effectively. Box 8: A third level course for PES staff in Ireland The Irish PES developed a training course for staff that lasts over two years, based on a combination of monthly 2-day formal sessions, coupled with distance learning methods and group work. The course is delivered by a University in collaboration with the PES, and successful participants obtain a post-graduate diploma in guidance, action planning and counseling. The course extends over two academic years and is delivered in an open learning format, incorporating e-learning materials, and workshops. Course hours are (i) home based learning, 212 hours, (ii) workshops, 216 hours. The course content includes the following modules: Psychology of Human Development: transition to adulthood, developmental stages in adulthood, development and change in the context of social exclusion. Perspectives on Work and Unemployment: work and unemployment, community development, labour market trends. 95Tergeist and Grubb (2006) 96Karagiannaki (2006), Corkett et al (2005) 96 Vocational Guidance: theory and practice of career development and behaviour, career information and information technology, theory and application of psychometric testing. Professional issues in Adult Guidance and Counseling Theory and Practice of Guidance and Counselling: applied guidance skills, applied counselling skills, theoretical perspectives on counselling. Group work-practice of group work, experiential approaches to group dynamics and process, theoretical perspectives on group work. Research and Evaluation Methodologies: quantitative approaches to research, qualitative approaches to research, applied research techniques. Psychology of Work and Working Life: organisation systems and dynamics, human resource management, occupational psychology Source: National University of Ireland Maynooth Contracting out of employment reintegration services and partnership with the private and NGO sectors 5.18 The development of effective partnership and service outsourcing models has been an important element in the modernization of a number of OECD countries' public employment services. Over the last few years Australia, and the Netherlands have been at the forefront of the rollout of private provision of activation and placement services and have progressively refined their approaches based on the results of evaluations97. Their example has been followed more recently, yet less ambitiously, by Germany and the United Kingdom. However, despite the common theme of contracting out of services, their models have varied substantially between countries. For example, in comparison with the standard type of outsourcing training or public works programs, Australia and other countries have gone further by outsourcing the entire activation service to private providers, with everything that is required to place highly disadvantaged job-seekers - from job counseling, motivational services, social work to training and referral. Box 9 summarizes the experience of several OECD countries. 5.19 Outsourcing of activation services is particularly useful in the case of highly disadvantaged job-seekers such as long-term unemployed Roma, which require more specialized and intensive interventions. The rationale is straightforward: Long- term unemployed and marginalized job-seekers require highly individualized and time- consuming services which include specialized social work services, and traditional employment services are typically not well-placed to deliver such services. It may simply 97Tergeist and Grubb (2006) 97 be more efficient for the Labor Offices to contract highly specialized and intensive services out rather than delivering them in-house. As acknowledged by Labor Office job counselors interviewed for this report, the Czech Labor Offices are facing difficulties in promoting access to the labor market for Roma. The reasons are many; however, it is clear that the Labor Office alone cannot adequately address the labor market needs of the Roma. The key to success both in the Czech Republic and elsewhere is the development of close, performance-based contractual collaboration with other agencies to deliver employment services of relevance to Roma. There is promising experience in the Czech Republic, including funded from European Structural Funds, which merits evaluation and further roll-out. Indeed it is envisaged under the Social Inclusion Agency and the Operational Program Human Resources Development. Box 9: International experience with contracting out of activation services Australia has been using outsourcing of services since the 1990s. Under its Job Network, hundreds of licensed Job placement Organizations in more than 2,700 locations across Australia offer placement services to unemployed. In 2003, the Australian Government introduced the Active Participation Model (APM) for job-seekers, adopting a more intensive and individualized approach to placing job-seekers through Australia's Job Network service outsourcing system. The APM has been found to deliver encouraging results for disadvantaged job-seekers, such as indigenous people. Under the APM service providers are offered incentives through payment for placing job- seekers in work. Payments for the more difficult to place clients are higher than those for short-term unemployed clients. Furthermore, Job Network applies a "star" rating, i.e. a comparative performance assessment model based on providers' Key Performance Indicators. Under the APM, each job-seeker has a single responsible Job Network service provider who provides close job counseling and placement services for the job-seeker. Disadvantaged job-seekers, for example those unemployed for more than three months, qualify for a more intensive support. APM specifies two main forms of support: (i) Job Search Support, focusing on direct placement of job-seekers, and (ii) Intensive Support which includes, based on an individual participation agreement, individualized training and advisory needs for long-term unemployed job-seekers with a graduated degree of customization of service for disadvantaged job-seekers. Job-seekers typically move from Job Search Support to Intensive Support if they have failed to find employment within three months after registration. During this period, job-seekers participate in "Mutual Obligations" activities, such as community work ("Work on the Dole"). If after 12 months Intensive Support job search training has not resulted in placement, job-seekers become eligible to "Customized Assistance", an intensive and highly flexible six-month counseling program. "Customized Assistance" is also available for job-seekers at day one of registration if they are deemed highly disadvantaged. Under the Job Network model placement of clients is contracted out to either private or NGO organizations and fees are paid in stages so that a client who remains in the placement for 13 weeks attracts the full placement fee and those that leave the job after fewer weeks only attract a lower fee. The fees are also based on the client's distance from 98 the labor market on entry to the program, providing incentives to placing disadvantaged job-seekers. A person unemployed for less than three months attracts a lower total fee for the agency compared to a long-term unemployed person. Under the Customized Assistance, Job network service providers can purchase training and cover incidental costs of job-seekers' placement efforts, e.g. interview preparation and travel, through a dedicated Job-seeker Account. In line with the modern employment activation approach, APM involves penalties for non-participation of job-seekers, but also Job Network service performance standards and supervision. Job Network also applies a "star" rating, i.e. a comparative performance assessment model based on providers' Key Performance Indicators. In the United Kingdom, 'Employment Zones' have been established to address long-term unemployed clients' needs, and Temporary Work Agencies (TWA) as well as other private providers are involved in delivery of services. The Jobcentre Plus has been organizing open tenders for the provision of integration services in these `Employment Zones'. In the initial roll-out contracts were given to a private employment agency in six zones. Contracts were also given to 'Working Links' in 9 areas, which is a joint venture company of Jobcentre Plus and two agencies, one of which is a significant TWA. In other cases under the New Deal initiative individual jobseekers can be referred to private employment services for training and/or reintegration or temporary work placement. In the Netherlands employment reintegration services for the unemployed who are not expected to find work within six months is outsourced to private providers. While the traditional public employment service (CWI) continues to manage job-seeker registration and profiling as well as initial job referral services as well as services for easy to place job-seekers. In contrast, all those job-seekers that are further from the labor market are redirected to the social insurance agency (UWV) for those that have unemployment insurance entitlements or to the municipalities for those who only have access to non- contributory social assistance benefits. Municipalities and the UWV are then in charge of managing benefit payments and transferring job-seekers to reintegration services. While UWV is mandated to outsource the provision of such services to private sector agencies, municipalities can manage such services themselves. Unlike in the UK and in Australia, there is no single central contractor, but a multitude of municipal contractors. Contracting periods are also typically shorter than in Australia or the UK. Outsourcing is conducted through tendering, and private and non-profit agencies have participated in several tendering rounds since the introduction of the new system in 2004. Municipalities receive lump-sum budget allocation from the central budget based on local demographic and labor market conditions, consisting both a benefit and a reintegration element. Municipalities have a strong incentive to save under the budgetary element designated for benefits, as they are free to reallocate it across budget lines. Source: Tergeist and Grubb (2006) 5.20 Service contracts with private providers typically include a performance based element with placement incentives for providers. The rationale is to cover a 99 provider's base costs but provide an incentive for placement through an outcome- dependent bonus. In the UK and Australia, a typical measure of success is whether an individual remains continuously employed in a referred job after 13 weeks. There have been some critics, pointing as a draw back to the risk that performance-based funding leads to a "job first" approach in which service providers try to place job-seekers quickly into jobs, often low productivity jobs, whereas targeted training could raise job-seekers' chances for higher productivity and higher pay employment98. 5.21 Several evaluations of outsourcing programs aimed at hard-to-place job- seekers such as ethnic minorities have found positive effects on employment outcomes. While the overall experience on outsourcing of employment reintegration services in OECD countries has been mixed so far99, there are some evaluations which show positive effects on placement of disadvantaged job-seekers. For example, Australia's Job Network evaluation and performance monitoring shows substantial positive effects of its Active Participation Model (see Box 9) for highly disadvantaged job-seekers, such as indigenous job-seekers or those from culturally and linguistically diverse backgrounds100. 5.22 Contracting out requires adjustment to the governance model of the public employment service. Under a system of outsourcing of services, the public employment service needs to assume a number of new regulatory functions: · Setting service standards: The management and quality control of such an approach would lie firmly in the hands of the Labor Office. In order to provide consistency across regions in this area of contracting out service provision for disadvantaged job-seekers, in particular Roma, it will be necessary to develop national quality standards for such services including a consistent performance based graduated and staged payment schedule for contractors such as exists in Australia. · Monitoring service standards under performance-based contracting: Under an outsourcing system, the Labor Office will need to develop the capacities to effectively monitor service standards across the country and different providers according to the defined service standards. Under mutual obligation principle, the maintenance of adequate service quality is crucial to guarantee enhanced probabilities of placing disadvantaged job-seekers into employment. The Labor Office would also need to manage the commissioning of external impact evaluation of private service provision. · Tendering and contracting: Experience in countries with a longer history of outsourcing has shown that the quality of service provision under a privately provided system relies heavily on the quality of the contracts negotiated between 98Tergeist and Grubb (2006) 99While evaluations show positive results in Australia and the Netherlands, they have shown little effect in Germany, however, this may possibly be due to the fact that the reform implementation has been short (see Tergeist and Grubb, 2006). 100Department of Employment and Workplace Relations (2006) 100 the public employment service and the private contractors. While the Czech Labor Office has experience with outsourcing of training, the outsourcing of more complex integrated activation services is likely to require further enhancement of contracting and tendering capacity Performance measurement and evaluation 5.23 Systematic performance monitoring and impact evaluation of employment programs has become widespread across many OECD and EU countries. Both are essential elements of public policy in general in that it allows reviewing whether programs work ­ whether they are well targeted and effective and whether money is well spent. It allows evidence-based decision making ­ scaling them up if they do work and adjusting them if they do not work as expected. Recent employment activation policy in many countries has benefits from quantitative evaluations, in the US, the UK, the Netherlands, Australia and Germany. Germany made rigorous impact evaluation a core element of its Hartz labor market reforms. 5.24 While there is some performance monitoring of active labor market programs, there is no rigorous impact evaluation of such programs in the Czech Republic as yet. Raising capacities in this area will be a useful investment to ensure better knowledge about the effectiveness of employment policies, and the budgetary resources spent on them. LABOR MARKET PROGRAMS 5.25 Currently, the Czech Labor Office is providing the standard set of services of a traditional PES. Pursuant to the Act on Employment measures of active employment policy by the Czech Labor Office include, in addition to job counseling: · retraining according to a retraining agreement concluded between the Labor Office and the job-seeker and paid by the responsible Labor Office. Retraining can help the job-seeker acquire new skills or deepen existing skills with the aim of enhancing his/her employability. Legislation does not specify limitation for the duration or cost of retraining courses. According to the Act on Employment the retraining agreement commits the job-seeker to repay retraining costs should he/she refuse, without serious reasons to conclude the retraining101; · investment incentives for employers to create new jobs and/or retrain new employees in a geographical are where the unemployment rate in the previous 30 months has been 50 percent above the average rate in the Czech Republic; · community service, a public works program providing part-time work in maintaining public areas and infrastructure, for a maximum of 12 months and on a renewable basis. The LO may provide a contribution up to the amount of the minimum wage plus social security contributions; 101Some LO staff interviewed for this report reported that this clause is not consistently enforced. 101 · socially beneficial jobs established by employers for job-seekers who cannot otherwise find employment whose cost are contributed to by the Labor Office for a period of up to six months (12 months for job-seekers under the age of 25 and over 50, persons with a disability and university graduates over two years after graduation and up to the age of 30). The level of contribution depends on the level of unemployment in the geographical catchment area relative to the average Czech unemployment rate; · a bridging contribution for self employed up to a period of three months in a monthly amount equal to 0.12 times the national average wage. · transport contribution for employers if they secure daily transport in areas not covered by public transport, with a contribution of up to 50 percent of actual costs (100 percent for people with a severe disability); · contribution towards training costs to employers of disadvantaged job-seekers up to a period of three months and up to half the minimum wage per employee102; and · contribution for a transition to a new business program for employers who go through a restructuring or transition to a new business and who can, for a limited time, not guarantee a full work program for their employees, up to half the minimum wage and up to a period of six months . 5.26 While the provision of active labor market programs is concentrated in high unemployment regions across the Czech Republic, programs are not targeted to the most disadvantaged job-seekers. As Figure 31 (upper left panel) shows, there is wide variation in the shares of registered unemployed who are covered in active labor market programs (ALMPs). While the Czech Republic average share of those covered is just over 21 percent, as little as 4.5 percent are covered in Prague, compared to more than 30 percent in Moravskoslezsky kraj. This is reflective of the variation in unemployment across regions. However, there is also wide variation among those regions with the highest incidence of long-term unemployment and lowest demand for low-skilled workers (see the circles). Out of the pool of Czechs that do participate in ALMPs, the vast majority are located in the difficult regional labor markets: Almost 30 percent of all those that participate in ALMPs are residing in Moravskoslezsky kraj (top right panel). However, as Figure 31 (bottom panels) reveals, the current programs are not targeted towards the disadvantaged groups, such as those with low educational attainment and long histories of joblessness. Over 30 percent of unemployed with complete secondary education participate in active labor market programs, compared to only 14 percent of those with only elementary education. Furthermore, 35 percent of the unemployed with 6 to 12 months of unemployment participate, but only 20 percent of those with between 12 and 24 months and only 17 percent 102Disadvantaged job-seekers are defined in the Act on Employment as persons with a disability, persons under the age of 25 and over 50, university graduates over two years after graduation and up to the age of 30, pregnant and nursing women for a period of nine months after giving birth, persons caring for a children under the age of 15 and persons who have been unemployed for more than 6 months. 102 of those with more than 24 months of unemployment. The age group most enrolled in ALMPs is the group of 19-24 year olds, reflecting the Czech Republic's focus on youth unemployed. Figure 31 Participation in ALMPs varies across regions, but they are typically not targeted to the most disadvantaged job-seekers Participation by region, 2005 ALMP participants across regions, 2005 100% 90% 1510, 1% Praha 80% 7639, 6% Stedoceský kraj 70% t 60% 6613, 5% Jihoceský kraj en rcep 50% Plzeský kraj 40% 4227, 3% Karlovarský kraj 39553, 28% 30% 3077, 2% Ústecký kraj 20% 10% Liberecký kraj 0% ci Královéhradecký kraj aha ja ja kr kr jark ja ja ja ýk ja a ký ja ja 19288, 14% kr kr kr kr kr kr Pardubický kraj puble Pr ý R Vysocina hcez kýseco kýsecoh incos vsaro ký ky kýszel skez kýsr kýcets ýkcer ecdar kýc 7330, 5% Jihomoravský kraj Ú éhvo Vy ouc nsílZ kos 4931, 4% C ed Ji Pl ovalra beiL álr Olomoucký kraj St K dubiraP homiJ oml O vsaro K M 4820, 3% 10683, 8% 5844, 4% Zlínsky kraj not covered covered 13199, 10% 9909, 7% Moravskoslezský kraj Participation in ALMPs by length of unemployment, 2005 Particiaption in ALMPs by education, 2005 100% 100% 90% 90% 80% 80% 70% 70% t 60% t 60% en rcep 50% 50% 40% ercenp 40% 30% 30% 20% 20% 10% 10% 0% 0% < 6 months 6-12 months 12-24 months >24 months primary vocational secondary higher not covered covered not covered covered Source: Research Institute for Labor and Social Affairs (RILSA), 2007 5.27 Active labor market programs most commonly used are requalification and training, public works and socially beneficial jobs. Figure 32 presents the breakdown of ALMPs across regions, suggesting that the main interventions are requalification, public works and socially beneficial jobs. This finding holds for Roma job-seekers as well, as presented in Chapter 2. However, there are variations: Moravskoslezsky kraj, the region with the highest incidence of unemployment and long-term unemployment, has been placing a strong focus on individual action plans, with more than 50 percent of ALMP participants enrolled in such plans. As opposed to that, Ustecky kraj, the second most deprived regional labor market, follows almost the same pattern as the Czech Republic averages. Interviews with job counselors conducted for the purposes of this report suggest that there is some churning on public works programs which last up to a year, and training on such programs seems to be limited to basic operations required for the particular job rather than systematic up-skilling to a national qualification standard. 103 Figure 32: The content of ALMPs varies across regions ALMP participants structure across regions, 2005 100% 80% ntecr 60% pe 40% 20% 0% cilbup aha ajrk ajrk ja ja kr kr ajrk ja ý ja a ký ja ja kr eck kr kr kr Pr ýk ýk inc ý Re adr kýc so vsaro ja ýk kýszel hce esco escoh ýksez ýksr eckts ýkcer ja kr ja Ú éhvo kr Vy ouc yksnílZ kos kr Cz ed Ji Pl ovalra beiL álr St K ubidraP homiJ oml O vsaro K M requalification public utility works socially purposeful jobs graduate training sheltered work self-employment individual action plan Source: Research Institute for Labor and Social Affairs (RILSA), 2007 5.28 In order to work for jobless Roma, Czech ALMPs need to be targeted more to the most disadvantaged and be placed in an individualized support program around intensive job counseling and placement efforts. The Czech Labor Office can utilize existing ALMP instruments but utilize them strategically as part of an individual action plan framework. A first step has been made in introducing the Individual Action Plan (IAP) which appears to be heavily used in Moravskoslezsky kraj but less elsewhere. The next step would be to strengthen its role by making it the mandatory primary engagement method for all long-term unemployed. It is through the individual action plan that other programs should be made available to long-term unemployed, involving an agreed contract between the job-seeker and Labor Office. 5.29 Ensuring that activation programs reflect individual needs requires building on client choice. This can involve (i) training vouchers to allow motivated and well- informed or well-advised clients to "buy" their own training or career guidance and to enhance choice and client ownership of the intervention103; (ii) development of community employment programs managed by NGOs or the local community but coupled with formal identification of training needs for each client. This can be developed into an agreed training plan linked to the national qualification framework and the delivery of such training (much of it in the workplace) by external training contractors hired by the sponsoring NGO or community. Typically clients get a premium payment above the level of their social assistance and attend such programs for 20 hours per week and are allowed to work in the open labor market for the rest of their work week without any penalties. This provides an incentive for clients to stay in the measure and provides for a gradual re-entry into mainstream employment. Durations are typically 12 months 103 Critics of vouchers argue that they do not work well in case of insufficiently informed and skilled clients who may not be able to make the right choice without external guidance and may be subject to sales and promotion activities of providers. 104 with an exit placement rate of 30-35 percent; or (iii) development of sub-contracted special dedicated community VET centres in collaboration with the education ministries, for early school leavers; (iv) focus on training and retraining and second-chance education programs in cases where lack of education is a binding barrier. 5.30 Public works programs may very well remain the only option for activation of low skilled long-term unemployed Roma in depressed regional labor markets ­ however with some necessary modifications to enhance their skills. Public works" might better be described as "Community Employment"104 in order to underline that the employment is for the individual rooted in the community. If these programs are to have a real long-term effect, it is crucial that a formal identification of training needs (ITN) is developed for each client coupled with the development of an agreed training plan linked to the national qualification framework. The sponsors/managers of such community based employment programs will need to be trained in the development of ITNs, staff management and staff development. Regular monitoring of such programs by suitably trained Labor Office staff is crucial to ensure the maintenance of high HR standards. This approach should thus avoid "programs" that focus on short-term, low-skilled employment and provide participants with neither enhanced skills nor better long-term employment prospects. The quality and training content of public works needs to be improved so that participants gain transferable skills. Such project interventions can also overcome barriers between non-Roma and Roma by building confidence through on-the-job training and employment experience. There are already examples of successful public works projects in Bulgaria which show non-Roma contractors that Roma can be reliable, effective employees. 5.31 Training and requalification are ideally linked to a subsequent job placement or risk being unattractive and de-motivating for job-seekers and lead to discouragement. Experts from non-profit NGOs interviewed for the purposes of this report expressed doubts about the meaningfulness of the requalification programmes as offered so far by the Labor Offices. If the requalification course does not directly allow entrance to the labour market, i.e. leads directly to employment, it does not have any effect on the client's standard of living. Indeed, Roma job-seekers participating in focus groups for this report reported their frustrations about prior experience of participating in retraining programs which was not rewarded through subsequent employment. Participation, in particular repeated participation, without successful subsequent employment risks having a negative psychological impact ­ when a client's aspirations and hopes are not fulfilled on the labour market after having absolved the requalification. 104In Ireland Community Employment provides training, development and work experience to the long- term unemployed (and other key groups at risk of social and economic exclusion) in community and voluntary projects and in public bodies. The measure provides work for up to one year to: long-term unemployed persons aged over 21 years; Travellers (Roma); lone parents; and persons with a disability. It also offers a three-year work option to the same categories of persons over 35 years of age. Employees, termed `participants', are obliged to provide 39 hours work over a two week period to the projects involved. In return, the sponsor must provide the participant with development and training which will improve their chances of accessing the open labor market. For projects employing more than 11 participants, the sponsor is funded to employ a supervisor and must produce a structured development and training plan for each of the individual participants. This plan must have an explicit objective of helping the participant to progress into mainstream employment. 105 It may be one key contributing factor to the widespread labor market discouragement of jobless Roma. 5.32 Specific contracted skills training for Roma should ideally be based on the potential of the trainees rather than based on education level. The general experience of Roma is that their education level excludes them from the type of economically useful courses available such as welding and any other trades. Specialist literacy and preparatory modules and/or mentoring will need to be built into the training. This provides a challenge for providers that they must be tasked to meet, through performance payments based on client selection criteria set by the PES and on placement outcomes. 5.33 "In work benefit schemes" can help facilitate the transition from welfare into work, in particular those that have been on benefits for a long time. Elements have already been tested in the Czech Republic. In a project entitled "Casual Registered work" longer term unemployed are allowed to do some work to earn extra money up to a limit of half the minimum wage. There is a case for enhancing this approach by the development of a "Back to Work Program" so that long-term unemployed clients with an unemployment spell of more than 3 years' duration who are placed in legal employment or self-employment are allowed keep their social assistance allowance for three years on a graduated scale of 75 percent for the first year, 50 percent for the second year and 25 percent in the third year. This would encourage genuinely inactive long-term unemployed back into work while receiving a premium. It would also encourage those drawing social assistance while working in the shadow economy to legalize their work by coming on to the program. Because such workers will start paying income tax, the scheme will pay for itself over time and would help to reduce the numbers of registered unemployed in the shadow economy. There would be no upper limit on the money earned. Anecdotal evidence suggests that this type of program may achieve even better results where there is an associated marketing campaign to promote the program and on the penalties for informal work. CONCLUSIONS AND POLICY DIRECTIONS 5.34 While incentives for the long-term unemployed to seek work have strengthened in recent years, the Labor Office at the beginning of 2008 remains insufficiently prepared to help disadvantaged job-seekers. In order to retain policy coherence and avoid that disadvantaged job-seekers are adversely affected by recent changes in the benefit system, the Labor Office needs to revamp its approach to help disadvantaged job-seekers into jobs. Cutting support for long-term unemployed in the absence of sufficient supply of effective and targeted ALMP programs is unlikely to achieve the desired results of greater employment among welfare recipients, whether Roma or non-Roma. Policy directions can be differentiated into two groups, (i) a change in the modus of engagement with disadvantaged long-term unemployed and (ii) promotion of new types of active labor market programs. 5.35 First, promoting greater employment chances for disadvantaged job-seekers, including Roma in marginalized localities, requires a fundamental change in the 106 management and delivery of employment services. This involves the following elements: · Introduce a mutual obligations-based employment activation approach across all elements of employment policy. While obligations of the jobseekers have been raised through reduced social welfare benefits and tighter requirements for cooperation with the Labor Offices, the Labor Offices's capacity to deliver effective services for disadvantaged job-seekers such as Roma need to be raised. · Change the mode of engagement with disadvantaged Roma job-seekers, offering a highly individualized and complex service, addressing their multiple employment, financial social and other needs that affect their employability. For example, make individual debt work-out a core element of activation policy. Individualized service provision starts with profiling of new clients into different categories depending on distance from the labor market and varied service packages for the different category job-seekers. · Restructure the Labor Office to free up more staff for individual counseling of vulnerable job-seekers. This entails training of staff to deal with this new challenge as well as the hiring of new skills, such as social work skills. · Consider merge Labor Offices with social welfare offices to offer the job-seeker an integrated service window addressing his/her multiple needs. This would free up staff for frontline interactions with disadvantaged job-seekers and allow providing integrated services, including traditional employment services and social work services (e.g. debt advice). · Consider substantial roll out of outsourcing of activation services using performance-based contracting, including job counseling and referral, to qualified private agencies and with performance-based contracts. While it is important to ready the Labor Offices to deal with disadvantaged clients, outsourcing entire service caseloads for the most disadvantaged job-seekers is likely to be more effective and cheaper than recruiting highly specialized service providers in- house. 5.36 Second, it will also require a fresh look at the utilization of active labor market programs for disadvantaged long-term unemployed. · Shift ALMP spending and concentration away from skilled secondary school and university graduates towards low-skilled and disadvantaged job-seekers. In today's tight labor market in the Czech Republic the majority of skilled individuals are highly likely to be able to find employment without much support from the Labor Office either in form of extended interaction with job counselors or in form of participation in Labor Office-financed active labor market programs. Resources could rather be shifted to those who have trouble finding employment on their own and who therefore do require such support ­ the disadvantaged, low skilled and long-term unemployed. 107 · Subsidized public works programs, or "community employment", will remain an important intervention, especially for unskilled Roma and in low labor demand regions, but they require modifications. In particular, in addition to transmitting work habits and experience, they should involve strategic vocational skill upgrading through training. In this respect a public works assignment should have an inbuilt formal identification of training needs for each client as well as an agreed training plan linked to the national qualifications framework. · Given the widespread functional illiteracy among Roma in marginalized localities, second-chance education and literacy programs should become core part of the retraining programs offered through the Labor Office. Creating a basis for subsequent vocational training, they should be seen as a priority engagement for Roma long-term unemployed within the framework of employment activation. · Link retraining programs to actual employment and build on client choice. Retraining programs on average appear to have not delivered employment results for jobless Roma. Successful programs should ideally be closely linked to actual employment and take account of the skills needed in the labor market. If not linked to employers' needs, they risk being wasteful spending and undermining job motivation of training beneficiaries. 5.37 In the pilot localities, the Social Inclusion Agency can provide an integrative role to link employment activation to other social inclusion activities. The Agency is tasked with developing and implementing local inclusion strategies in partnership with municipal authorities, the labor office, schools and education services, health services as well as the NGO sector and private businesses. Having to address many interwoven barriers, employment activation will have to be firmly anchored in the local social inclusion architecture. 108 6. ROMAAND YOUTH-CENTERED EMPLOYMENT ACTIVATION The Roma population in the Czech Republic is young, suggesting the need to focus particular attention on education and the school to work transition. Roma have been suffering from substantially worse education outcomes, being over-represented in schools for children with special needs and leaving school often with only nine years of schooling and less. The legacy of low educational attainment among young Roma persists, leaving them at greater risk of being neither in employment nor in education or training and therefore heading straight into long-term unemployment. Indeed Czech youth with low educational attainment fare significantly worse in the labor market than their peers elsewhere in Europe. While youth has been identified as a risk group among the unemployed in the Czech Republic as in many other OECD and EU countries, effective early intervention mechanisms are still lacking to prevent long-term unemployment among the young, in particular those with low educational attainment. Given the demographics among the Roma, a new approach to promote employment among Roma has to have a strong youth focus to stem the flow of young Roma into long-term unemployment. The policy directions are threefold: First, to improve access to quality education for Roma at all levels from early childhood education through tertiary education; Second, to work on preventing early school leaving of young Roma to keep them in school for as long as possible; Third, to focus on developing an intensive youth- centered activation approach in the Labor Offices, with outsourcing of the full range of activation services to qualified agencies. 6.1 Children and youth from disadvantaged background are more at risk of leaving school early and heading into unemployment. Overcoming social exclusion of the Roma in the Czech Republic requires a life cycle approach that starts by investing in early childhood development and translating into improved school outcomes and lifelong learning. Research shows that family background has profound impact on an individual's development. Whether or not children and young persons grow up in a poor family determines later chances in life. Children who grow up in poverty and disadvantaged circumstances have lower life chances: They enter school insufficiently prepared, do worse in school, leave school earlier and have lower employment chances. The stress of experiencing poverty, lacking stimulation in particular at very young age, lacking positive role models and a lacking learning support environment at home all contribute to worse school outcomes and subsequent failure in the labor market. For example, the experience of sustained joblessness of parents affects work motivation of their children. Low educational attainment and high drop-out rates are often linked to insufficient 109 preparation at the time of entering primary school. Research105 shows that preparation fosters cognitive, language and behavioral skills, which are vital to exploit one's full potential in later school education. Insufficient school readiness is also associated with the large enrollment of Roma children in schools for children with special educational needs. 6.2 The large share of youth among the Roma population in the Czech Republic suggests the need for a youth focus of policies supporting Roma employment. Young people in general have been identified as a vulnerable group in the labor market in the Czech Republic, but young Roma are particularly at risk. According to the 2008 labor force survey in marginalized Roma localities, a sizable share of the population of jobless Roma is below the age of 25, in particular among the unemployed, i.e. those in the labor force (see Figure 33). With a high share of Roma having only primary education and less, young Roma are particularly at risk of being neither in employment nor education or training (NEET). This chapter sheds light on aspects of the school to work transition of Roma youth and provides recommendations, based on international examples, for elements of a youth focused activation agenda. Figure 33: A sizable share of jobless Roma are below the age of 25 Demographic profile of Roma by labor market status, 2008 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Employed Casual Unemployed Inactive Discouraged Total 15-24 25-54 55-64 Source: Excluded Roma Labor Force Survey 2008; Bank staff calculations EDUCATION AND SCHOOL-TO-WORK TRANSITION 6.3 Roma in the Czech Republic have been suffering from low education outcomes for a long time, worsening their labor market prospects. The evidence from the survey in marginalized Roma localities presented in Chapter 2 confirms the low education status of socially excluded Roma in the Czech Republic, with the majority having completed only primary education ­ or even graduated from special schools for children with special learning needs106. The legacy of segregation of Roma into special schools has recently triggered a decision of the European Court of Human Rights against the Czech Republic over a case in Ostrava, ruling that the case of disproportionate 105Campbell FA; et al (2002 ); Sylva, K., Melhuish, E. C., Sammons, P., Siraj-Blatchford, I. and Taggart, B.(2004); Carneiro P. and J. Heckman: Human Capital Policy (with James J. Heckman), in J. Heckman, A. Krueger, Inequality in America: What Role for Human Capital Policies, MIT Press, 2003. 106Special schools were transformed into "basic practical schools" in 2004. 110 streaming of Roma children into special schools in Ostrava constituted as violation against fundamental rights. This decision follows research in 1999 by the European Roma Rights Center (ERRC) in the Ostrava region on segregation of Roma. The research found that Roma children in Ostrava were substantially more likely to be in special schools than noon-Roma children. While representing less than 5 percent of all primary school children in the Ostrava region, Roma children made up more than 50 percent of special school pupils. Further ERRC research in 2002 confirmed the situation for the towns of Teplice, Kladno and Sokolov107. 6.4 While Czech pupils have consistently performed above average in successive PISA assessments, the Czech education system retains high variation in education outcomes between schools. As presented in Figure 34, data from the 2006 PISA assessment indicates that the Czech Republic is one of the OECD countries with the highest variation in student performance between schools. Moreover, a lot of this variance can be explained by economic, social and cultural status of students or schools in the Czech Republic. At the same time the variance in student performance within schools is substantially below the OECD average. This suggests that, while having highly successful schools which produce high learning outcomes, the Czech Republic also maintains highly unsuccessful ones with low learning outcomes and where the lacking success owes to the children's social status, among others. In effect, this means that the Czech education system does not appear to make up for unequal starting positions resulting from social-economic inequities. The evidence presented in Figure 34 is also entirely consistent with the continued existence of "basic practical schools" ­ former schools for children with special learning needs with a disproportionate enrollment of Roma children. This study does not analyze the effect of the education system on Roma education outcomes. However, given the finding on dramatically low educational attainment among Roma in marginalized localities, further analytical work on identifying the drivers for bad education outcomes is of high priority. 6.5 The Czech education system appears to have been unable to break the intergenerational transmission of exclusion among the Roma: There is very little upward educational mobility among Roma but substantial downward mobility. As reviewed in Chapter 3, the current generation of Roma youth has not been able on average to achieve higher educational attainment than their parent's generation. If anything, there is evidence of a worsening of the situation. Only 21 percent of sons of fathers with primary education or less received post-primary education (mainly basic vocational training), yet more than half of the sons (54 percent) of fathers with post- primary education received less education.108 This finding confirms the picture from the PISA assessment on the strong role of the economic, social and cultural status of students or schools in explaining variation in learning outcomes between schools. 107European Roma Rights Center (2005) 108A very similar pattern emerges when on compares educational attainment of daughters with that of their mothers. 111 Figure 34: The Czech education system remains unequal Variance in Student Performance Between Schools and Within Schools, OECD 2006 PISA Assessment, Science Results between-school variance within-school variance 100 80 60 40 20 0 20 40 60 80 100 120 Bulgaria Germany Slovenia Czech Republic Austria Hungary Netherlands Belgium Argentina Chile Japan Italy Greece Chinese Taipei Israel Brazil Slovak Republic Turkey Uruguay Switzerland Romania Serbia Hong Kong-China Croatia Tunisia Korea Kyrgyzstan Luxembourg United States Portugal Thailand Mexico Lithuania Colombia Russian Federation United Kingdom Montenegro New Zealand Australia Jordan Indonesia Macao-China Azerbaijan Canada Ireland Estonia Denmark Latvia Spain Poland Sweden Norway Iceland Finland Total between-school variance Total within-school variance Between-school variance explained by Within-school variance explained by the the PISA index of economic, social and PISA index of economic, social and cultural status of students and schools cultural status of students and schools Source: OECD (2007). 6.6 Unsatisfactory education outcomes and early drop out of Roma students appear to be driven also by the lack of positive role models for Roma youth as well as a lacking supportive environment for education. Educators and social workers from the majority of localities investigated for this study have noted in their work pressure on 112 the part of parents for children not to continue with further education after completing elementary school. While schooling is often seen as costly ­ given expenses such as for school lunches, clothes and books ­ Roma youth in marginalized localities are instead typically encouraged to join the labor market and contribute to family income. This is confirmed by the finding on higher labor force participation among Roma youth than youth on average in the Czech Republic (see Chapter 2). In turn, they typically also register at the Labour Office in order to qualify for free health insurance and continued assistance in material need. Some experts interviewed for this study knew of multiple cases where Roma youth applied for apprenticeship school after completing elementary school and, while accepted, never began attending after the summer holidays or dropped out within the first year. While the financial burden of studying may account for such decisions to some extent, experts primarily blamed the lack of personal motivation as a result of a number of basic factors. On the one hand it is the negative example of the parents, who, themselves being without work and education, do not encourage their children's learning and, in some cases, even systematically discourage their children from becoming educated. On the other hand, experts also reported of discouraging cases where young Roma graduated from secondary school yet still failed to find employment. 6.7 Low educational attainments and bad labor market outcomes among Roma youth suggest parallel policy measures to improve supply of quality education for Roma and stimulate demand. This would imply the following: · On the supply side, the Czech Government's attention should focus on a sustained effort to raise the quality of education services available for children from disadvantaged backgrounds. This would entail enhanced accessibility to and provision of high quality services at pre-school and primary and secondary school levels and, crucially, involves addressing the disproportionately high share of Roma in "basic practical schools". Expanding access to pre-school as well as early childhood education programs can raise school readiness of Roma children and reduce the early streaming into basic practical schools. · On the demand side, there is a need to do outreach work in marginalized localities to work with parents to incentivize them to send their children to kindergarten, pre-school and school and to create a more conducive learning environment at home. The Social Inclusion Agency will be a crucial actor in promoting and facilitating such community outreach work. Family cash transfers conditional on behavioral change (i.e. whether or not children go to school) can raise the demand. This could either involve adjusting the eligibility threshold of the Minimum Living Standard (MLS) for children depending on whether they go to school (lower if they do not, higher if they do) or experimenting with a designated conditional cash transfer (CCT) which has been used in many countries to provide incentives for behavioral change and break the inter- generational transmission of social exclusion109. Targeted youth work in cooperation with the schools, including through providing good role models, can help address low performance and lacking motivation of Roma youth in school. 109For a detailed review of the conditional cash transfer experience see Fiszbein and Schady (2008). 113 6.8 Making cash benefits to youth conditional on continued school enrollment and completion may help raise employability of Roma youth. Demand side measures should be focused on retaining young Roma in school beyond compulsory education. International experience suggests that this requires a complex approach, linking community outreach work focused on youth and parents as well as monetary incentives. A monetary incentive could include linking the level of social benefits for the family to whether the child enrolls and remains in secondary education even beyond the compulsory level, for example through raising the monthly subsistence amount for a dependent child from age 15-26. The United Kingdom has introduced an education subsidy program for youth from low-income families to incentivize them to stay in education beyond compulsory education ­ the Education Maintenance Allowance (see Box 10). Meanwhile, the Netherlands have recently extended mandatory schooling until the completion of upper secondary education (ISCED 3) or until the age of 18. Moreover, there are plans to introduce an obligation to study or work ("leerwerkplicht"), according to which youth aged 18 to 27 who have not completed the equivalent of upper secondary education (ISCED 3) are required to continue education or work. Under this rule those that are neither in employment not in education or training (NEET) can be considered ineligible to social benefits110. Compared to the UK, the Netherlands model therefore sanctions failure to remain in school beyond lower secondary education. Box 10: The Education Maintenance Allowance (EMA) in the United Kingdom: A Conditional Cash Transfer to reduce the number of youth NEET In 1999 the United Kingdom Government introduced a pilot program in ten Local Education Authorities to address the low participation of youth in post-compulsory upper secondary education as well as the high share of youth not in education, employment or training (NEET). The Education Maintenance Allowance (EMA) is a means-tested conditional cash transfer (CCT) available to 16-18 year olds from low income families who remain in full-time education, academic or vocational, after the 11th year of schooling, typically after 16 years of age, when schooling is no longer compulsory. Available for a maximum of two years, EMA consists of (i) a weekly allowance available during term time, (ii) an annual retention bonus and (iii) a one-time achievement bonus paid at the end of the course the person has enrolled on, provided he/she meets the outcomes of a learning agreement signed at the time of entry into EMA. EMA is available to youth whose parents' incomes were below a certain threshold. The level of the allowance depends on family income and at the maximum level the allowance is the equivalent of about a third of average net earnings for the target age group. The pilot was associated with a series of rigorous multi-year evaluations of the impact of EMA on participation in education programs as well as retention and completion. All studies show unanimously that EMA has had a significant impact on the participation in post-compulsory upper secondary education. EMA was found to have raised participation rates by between 4.5 and close to 6 percentage points for the first year of participation and by even more for the second and subsequent year, suggesting that the program was 110OECD (2008), Jobs for Youth: Netherlands 114 effective in retaining those young people that had enrolled under EMA and even after the end of the allowance eligibility. EMA was found to be particularly effective for youth coming from the poorest socio-economic backgrounds as well as for young people at year 11 who had previously been low achievers. Moreover, the program drew as many young people from inactivity (NEET) as from work or training, suggesting it was a very effective tool to reduce youth inactivity, particularly among the most disadvantaged. Recent studies have also found a substantial increase in attainment of participants relative to a control group. Given the demonstrated success of the pilot program, the UK Government rolled out the program nationwide in 2004. Source: Chowdry et al (2008); Dearden et al (2005); Middleton et al (2004) YOUTH ­FOCUSED EMPLOYMENT ACTIVATION 6.9 Many countries have adopted a targeted activation framework for young people not in employment or in education or training to address educational deficits and promote sustainable employment. International experience suggests that youth employment activation policies require careful design to be effective. Evaluations of international best practice suggest a number of key design lessons111. First, job-search assistance programs are typically most cost-effective for youth, leading to higher earnings and employment, as are wage and employment subsidy programs. Second, reaching those youth that are most at risk of social exclusion requires making participation in programs compulsory for youth after a period of job search of no more than six months. Third, it is important to introduce a differential targeting approach for youth employment activation programs, in particular to early school leavers, distinguishing between those close to school age and young adults. Lastly, training programs that are tailored to local or national labor market needs have proven more effective. 6.10 In line with international good practice, the Czech Labor Office has introduced mandatory action plans for job-seekers below the age of 25 but implementation appears to have been patchy and not focused on the most disadvantaged. Initially piloted as a program called "First Chance" in 2004, a priority focus on the below 25 is now stipulated in the Employment Act. However, implementation appears uneven, according to research conducted for this report (see above). At the same time, individual action planning is not targeted to the most disadvantaged youth, but universally applied to the below 25s as well as to recent university graduates ­ who should stand a good chance of finding employment on the Czech Republic's currently buoyant labor market. The experience from several EU Member States suggests that a mandatory application of individual action plans, tied with intensive job counseling and focused training or work experience interventions are needed to have an impact (see Box 11). 111see OECD Employment Outlook 2006, Martin and Grubb, 2001; and Betcherman et al., 2004 115 Box 11: Youth Employment Activation in the United Kingdom, Ireland and Germany The United Kingdom, through its "New Deal for Young People", emphasizes early, continuous activation and contact as well as more specific help after 6 months of unemployment. Participation in the New Deal for Young People (NDYP) is mandatory for young people between the age of 18 and 24 who have been unemployed for 6 months and who receive Job-seeker Allowance (JSA). The program has multiple stages, with an initial four months period of intensified job search and job referral through job counselors ("gateway") to place the job-seeker in the primary, unsubsidized labor market. If unsuccessful, the "gateway" is followed by a period of up to eight months of targeted programs and a shorter period of "follow-through", again devoted to intense job search (see Figure 35). As for programs, job-seekers can choose from a menue of subsidized employment, voluntary sector employment, full time training and education and work on environmental task forces. In an innovative feature, the NDYP promotes partnerships with employers to encourage the training and employment of young people, facilitated by training grants (up to Ł750) and employment subsidies (up to Ł60 for six months). An individual training plan is agreed by the Jobcentre Plus adviser with the young person. In addition the employer signs an agreement that he/she will be expected to pay the going rate for the job and that states the employer will (i) keep the employee on as long as they show the aptitude and commitment needed and provide or arrange for their training as appropriate and (ii) monitor and record their progress and identify areas of action, in the same way that the employer would for any other employee to help them settle in and make progress. A range of evaluations have found that the program has been returning young unemployed to the labor market faster than without. Meanwhile, several long-term studies have found evidence of churning between NDYP participation and unemployment spells and some worsening of performance. As a result, the UK Government is in the process of reforming the program to strengthen its ability to place clients into sustainable jobs with more focus on employment retention and progression through more personalized and flexible service provision, in particular for disadvantaged job-seekers and NEETs, and through greater partnerships in service provision with private and third sector providers112. Figure 35 : The New Deal for Young People in the United Kingdom subsidized employment 6 months voluntary sector ugho Gateway, unemployment, thr 4 months aged 18-24 full-time training and education w-ol fol environmental task force 6 7 8 9 10 11 12 13 14 15 16 17 18 months 112OECD (2008), Jobs for Youth: United Kingdom, Paris 116 In Ireland the PES has been putting emphasis on preventing early school leaving and encouraging young drop-outs back into training and education, through a program called "YouthReach". The program aims at improving employability through training and education rather than placement into employment in the first instance. The `YouthReach' program is intended to facilitate young people's return to learning and preparation for employment and adult life. It is run in special community training centers dedicated to the training of Early School Leavers. Its general objectives are as follows: (i) personal and social development and increased self-esteem; (ii) second-chance education and introductory level training; (iii) the promotion of independence, personal autonomy, active citizenship and a pattern of lifelong learning; (iv) integration into further education and training opportunities and the labor market; (v) the promotion of social inclusion. In the experience of YouthReach, the key features of successful interventions with early school leavers are a focus on the development of the individual, his or her independence and integration. It is to follow a safe, structured yet challenging learning environment through a participant-centered and participant-led process. Programs are seen as most effective if organized on the basis of a curricular matrix in which each teacher or trainer is implementing a range of cross-disciplinary curricular objectives (such as communications skills development, health and safety awareness, etc). Moreover, they are oriented at closing with an appropriate assessment and certification. In YouthReach, the maintenance and in-career development of staff involved is a priority. In Germany the Federal Employment Agency has been considerably increasing the resources available for enhanced career guidance to target the excluded and marginalized. However, these measures must by law be 50 percent co-financed by a third party. In addition, particularly intensive support will be given ­ in model projects ­ to students who seem unlikely to complete school or incapable of completing vocational training. The Federal Government has set the objective of having no young person unemployed for over three months. Under current legislation on the basic provision for jobseekers, young people have a binding right ­ laid down by law ­ to have work, training or an employment opportunity provided as follows: Within one week of their application, young people must have been given their first counselling session and have their profiles drawn up. Within three weeks of their application, an integration agreement must have been completed. Within four weeks of completion of the integration agreement, the young person must be given a specific offer of a job, training, training preparation, further training or an employment opportunity. In the German system, unemployed young people are considered as a single target group for integration work. Under current regulations, an integration specialist has a case-load of a maximum of 75 young people to ensure a high quality, targeted service for each young person in their care. Current legislation makes special provision for counseling of young people on an employment-oriented case management basis if there are multiple barriers to their integration (e.g. lack of school-leaving qualifications and drug addiction and debt problems). Employment- oriented case management is provided by specially-trained case managers (partly certified by the German Association for Care and Case Management), who can access a wide network of help and counseling institutions to provide effective support to these most marginalized clients. CONCLUSIONS AND POLICY DIRECTIONS 6.11 The Roma employment agenda is to a large part a youth employment agenda, and it is primarily on young workers where a difference can be made. Improving access of Roma to sustainable and high quality jobs over the long-term requires greater attention to the making the education system work for Roma and 117 developing a youth-centered employment activation approach. Policy direction include the following: · Address inequities in the education system affecting Roma, by systematically implementing an educational integration policy from early childhood to tertiary education. This entails an expansion of childcare supply and measures to work with Roma parents to incentivize them to send their children to child care and kindergartens so as to prepare them better for school and avoid early channeling into low quality basic practical schools. Incentive measures could also include experimenting with conditional cash transfers (CCTs), for example through adjustments in the MLS threshold for children and youth, depending on whether they go to school (higher threshold) or not (lower threshold). · Incentivize young Roma to stay in school beyond the minimum mandatory school age, i.e. into upper secondary education, instead of entering the pool of registered Labor Office clients. The Education Maintenance Allowance (EMA) program from the United Kingdom could serve as an example, and a Czech adaptation could entail applying a higher minimum subsistence level for a dependent child aged 15-26 that remains in non-compulsory education for a maximum of, say, two years and depending on retention and completion of education levels. Alternatively, scholarships for post-compulsory education for children whose parents are on social assistance could help boost retention of Roma children in education and training. · Promote early outreach through school counseling and professional orientation to identify and counsel those youth at risk of drop out. Prevention of early school leaving, and retention in formal education and training requires early outreach to and advisory services for those at risk of drop-out. The experience of the United Kingdom's "Connexions" services tasked with providing guidance to 13-19 year olds, in particular disadvantaged youth, suggests that the Social Inclusion Agency could, in a formalized manner, support schools in engaging youth at risk and counseling them about their options to remain in school or training beyond compulsory schooling. · Develop the individual action plan approach for young people further into a mandatory and intensive youth-centered activation approach focused on NEETs. Following the UK example, this would be centered around intensive counseling, with job placement services, training and remedial or second-chance education for older youth and back-to-school programs for the younger. It could also entail outsourcing of the full range of activation services to qualified agencies with experience in working with disadvantaged youth. · Pilot and test apprenticeship, internships and wage subsidy programs for young workers. Facilitating the school to work transition and preventing the NEETs phenomenon will require testing proactive measures like apprenticeships, internships, placement and job subsidies programs for young people to help them not only get into the labor force, but build some relevant skills. This is an agenda 118 that requires active contributions and partnerships between the Government and employers and trade unions. For example, experience from across the OECD shows that wage subsidy programs can have positive employment effects for young workers113. 113Kluve, J (2006) 119 BACKGROUND REPORTS FOR THIS STUDY Jurajda and Zubrický (2008), Employment Incentives in the Czech Republic: Comparing Tax and Social Benefit Systems in 2007 and 2008 Kavanagh (2008), The role of labor market policies and institutions in the employment of Roma in the Czech Republic Rutkowski (2008), Roma in the Labor Market in the Czech Republic: Evidence from the Labor Force Survey in marginalized Roma localities Uherek et al. (2008), Qualitative analysis of Roma employment in the Czech Republic: Expert interviews and focus groups with jobless Roma Vasecka et al. (2008), Employers' views on Roma employment in the Czech Republic 120 REFERENCES Bedard (2007), Employment Activating Social Assistance Schemes Not Working For Roma and Travelers, in: European Roma Rights Center, Roma Right Quarterly, 1- 2/2007; Betcherman et al. (2004), Impacts of Active Labor Market Programs: New Evidence from Evaluations with Particular Attention to Developing and Transition Economies. In: Social Protection Discussion Paper Series, Nr. 0402, World Bank, Washington D.C. Bogdanov and Angelov (2007), Roma Integration in Bulgaria: Necessary Reforms and Economic Effects, Open Society Institute Bulgaria: Sofia Campbell FA; et al (2002 ) Early childhood education: young adult outcomes from the Abecedarian Project, Featherstone, H. "Pre-school: It Does Make a Difference." Principal 65(1986): 16-17. Carneiro P. and J. Heckman: Human Capital Policy (with James J. Heckman), in J. Heckman, A. Krueger, Inequality in America: What Role for Human Capital Policies, MIT Press, 2003. Chowdry et al (2008), Education Maintenance Allowance: Evaluation with Administrative Data, Learning and Skills Council Report, Coventry Corkett et al (2005) Jobcentre Plus evaluation: Summary of evidence, IAD Social Research Division Research Report 252 Dearden, Emmerson, Frayne and Meghir (2005), Education Subsidies and School Drop- Out Rates, Institute for Fiscal Studies Working Paper WP05/11 Department of Employment and Workplace Relations (2006), Customized Assistance, Job Search Training, Work for the Dole and Mutual Obligation ­ A Net Impact Study, EPPB Report 1/2006, Canberra European Commission (2008), Progress towards the Lisbon objectives in education and training, Indicators and benchmarks, Commission Staff Working Document European Roma Rights Center and Numena Centro de Investicao em Ciencias Sociais e Humanas (2007), Social Inclusion Through Social Services: the Case of Roma and 121 Travelers, ERRC/Numena, Budapest; European Roma Rights Center (2006), The Glass Box: Exclusion of Roma form Employment, Budapest European Roma Rights Center (2005) Stigmata: Segregated Schooling of Roma in Central and Eastern Europe, a survey of patterns of segregated education of Roma in Bulgaria, the Czech Republic, Hungary, Romania, and Slovakia, ERRC, Budapest Fiszbein and Schady (2008), Conditional Cash Transfers for Attacking Present and Future Poverty, World Bank Poverty Research Report, Washington DC: World Bank Habschick, Seidl and Evers (2007), Survey of Financial Literacy Schemes in the EU27, Evers and Jung Financial Services, Hamburg Jacobi and Kluve (2006), Before and After the Hartz Reforms: The Performance of Active Labor Market Policy in Germany, IZA Discussion Paper No. 2100, Bonn: IZA Karagiannaki (2006) Exploring the effects of integrated benefit systems and active labour market policies: Evidence from the Jobcentre Plus in the UK, CASE discussion paper 107, Centre for the Study of Social Exclusion, London: London School of Economics Kertesi and Kézdi (2006), Expected Long-Term Budgetary Benefits to Roma Education in Hungary, Roma Education Fund, Budapest Kluve, J (2006), The Effectiveness of European Active Labor Market Policy, IZA Discussion Paper No. 2018, Bonn: IZA Korczak (2004), The Money Advice and Budgeting Service Ireland ­ A service to help people with financial problems and to tackle over-indebtedness, Synthesis Report, peer Review in the Field of Social Inclusion Policies: Ireland 2004, mimeo Martin and Grubb (2001), What works and for whom: a review of OECD countries' experiences with active labour market policies, Working Paper Series 2001:14, IFAU - Institute for Labour Market Policy Evaluation Middleton, S., Perren, K., Maguire, S. and Rennison, J., Battistin, E., Emmerson, C. and Fitzsimons, E. (2005) Evaluation of Education Maintenance Allowance Pilots: Young People Aged 16-19 Years, Final Report of the Quantitative Evaluation, Department for Education and Skills Research Report No. 678. Nottingham: DfES Publications OECD (2008a), Jobs for Youth: Netherlands, Paris: OECD OECD (2008b), Jobs for Youth: United Kingdom, Paris: OECD OECD (2007a), Benefits and Wages 2007, Paris: OECD OECD (2007b), Employment Outlook 2007, Paris: OECD 122 OECD (2007c), PISA 2006: Science Competencies for Tomorrow's World, Volume 1: Analysis, Paris: OECD OECD (2006), Czech Republic Economic Survey 2006, Volume 6/2006, Paris:OECD Office of the Government of the Czech Republic (2007), Agency for removing social exclusion and its prevention in socially excluded Roma localities, mimeo Quintini and Martin (2006) Starting Well or Losing their Way? The Position of Youth in the Labor Market in OECD Countries, OECD Social, Employment and Migration Working Papers No.39, Paris: OECD Research Institute for Labor and Social Affairs (2006), Hodnocení efektivity program aktivní politiky zamstnanosti v CR, Brno Ringold, Orenstein and Wilkens (2005), Roma in an expanding Europe: Breaking the Poverty Cycle, World Bank: Washington DC Rosholm et al (2004), A Danish Profiling System, IZA Discussion Paper No 1418, Bonn: IZA Sylva, K., Melhuish, E. C., Sammons, P., Siraj-Blatchford, I. and Taggart, B. (2004),The Effective Provision of Education (EPPE) Project: Technical Paper 12 - The Final Report: Effective Education. London: DfES / Institute of Education, University of London. Tergeist and Grubb (2006), Activation Strategies and the Performance of Employment Services in Germany, the Netherlands and the United Kingdom, OECD Social, Employment and Migration Working Papers No 42, Paris: OECD Uherek, Weinerova and Plochova (2004), Lending Money in Roma in Roma Communities: The Problems of Credit Granted by Legal Entities to Socially Deprived Clients with Subsistence Level Income, Institute of Ethnology, mimeo Wissenschaftszentrum Berlin für Sozialforschung (2001), Management by Objectives in European Public Employment Services, Discussion Paper FS I 01 -203 World Bank (2008), Good Practices for Consumer Protection and Financial Literacy in Europe and Central Asia: A Diagnostic Tool, Consultative Draft August 2008, World Bank: Washington DC World Bank (2007), Labor Markets in EU8+2: From the Shortage of Jobs to the Shortage of Skills, World Bank: Washington DC 123 World Bank (2007b), Czech Republic: Technical Note on Consumer Protection in Financial Services, World Bank: Washington DC World Bank (2005), The Quest for Equitable Growth in the Slovak Republic: A World Bank Living Standards Assessment, World Bank: Washington DC 124 ANNEX The report builds on a range of data sources and special survey work as well as qualitative research a. Labor force survey in marginalized localities The data on Roma employment presented in this study are from a specially designed Labor Force Survey conducted in May 2008 in 12 marginalized localities where many Roma reside, six of them in Bohemia and six in Moravia, with ten towns and two micro- regions. Bohemia Moravia 1. Microregion Tolstejnsko 1. Microregion Jesenicko 2. Ústí nad Labem 2. Brno 3. Most 3. Perov 4. Cheb 4. Beclav 5. Roudnice nad Labem 5. Holesov 6. Broumov 6. Slezská Ostrava During the research a 13th town was added, the town of Jirkov in Northern Bohemia, due to the fact that in some localities enumerators did not find enough Roma families to fill up the desired number of questionnaires. The sample of surveyed localities was drawn from a list of marginalized localities assembled by the Office of the Government of the Czech Republic to guide the work of the Agency on Social Inclusion established in 2008. The survey is, therefore, not representative of the Roma community in the Czech Republic as a whole, but representative of those Roma residing in identified marginalized localities. Therefore, when referring to "Roma" this report implies those Roma who reside in known marginalized localities. The surveyed localities are mainly, yet not exclusively, inhabited by Roma, and the Roma population shares vary. The survey, therefore, also covered a limited number of non- Roma residents of excluded communities. Although their number was small, in some cases the comparison between Roma and non-Roma workers was still possible. Roma were identified using an answer to the following question: "This is a survey of the Roma community. Do you consider yourself Roma?". According to this self-identification criterion, there were 1050 Roma in a total sample of 1150 individuals. Although the overall sample size is not small, there were instances of wide error margins in responses 125 to some questions due to small sub samples, e.g. the unemployed. Cases of wide error margins are indicated in the report. The questionnaire follows a standard Labor Force Survey structure and captures all aspects of labor market experience of surveyed individuals. In addition, it includes a simplified skills assessment to assess actual levels of functional literacy and numeracy skills. b. Skills Assessment The Labor Force Survey included a simple skills assessment test for surveyed individuals. The aim of the test was the identification of general competence of understanding and using information in everyday life situations. The structure of the test had to be adjusted to the sample size, its characteristics and above all the timetable. For these reason the test could not be realized in usual scope. Nevertheless, its reduced version fulfills all the principles of functional literacy measurement ­ which is to examine three basic components (literary, documentary and numeric literacy). In order to obtain sufficient number of data concerning employment and eligibility of the target group, the test was focused on the examination of competence in using information, in particular through solving model situations concerning official communication, travelling, orientation in the legal system and orientation in time. Test outcomes of each respondent were evaluated in the following way: 0 was given to no answer or an incorrect answer, 1 to a correct answer. Questions 4 and 6 were rated as follows: 0 was given to no or an incorrect answer, 1 to partly disposed, and 2 to completely disposed. The respondent could then gain 0 ­ 8 points. According to the total score of each respondent the outcomes were divided into three categories: the total of 0-2 ­ Category 3, the total of 3-5 ­ Category 2, and the total of 6-8 ­ Category 1. The skills assessment categories represent the following characteristics: · Category 3 ­ Almost zero capability of understanding the language of institutions, competence of understanding hypothetical questions less than 40%, numeric literacy almost zero, inability of solving simplest arithmetic operation. · Category 2 ­ Basic orientation in institutional language, competence of understanding hypothetical questions 40% at least, numeric literacy and legal conscience at a substandard level. · Category 1 ­ Ability of abstraction and hypothetic cogitation at a level of at least 80%, basics of numeric literacy and legal conscience. Skills Assessment Questionnaire 1. You are about to buy a carton of cigarettes at a reduced price of 50%. The original price was 200 CZK. How much will you pay? 126 2. Situation at the post office. Imagine you need to pay a post remittance. Which of the given counters you choose? Parcel service Letter services EMS ­ express message service Financial services Insured letter receipt Bank deposits and wages 3. Your physician has prescribed you a medicine which is to be taken three times a day every eight hours. What time do you take it? 4. Imagine you have to arrange several documents necessary for social benefit. In this matter you have to visit the appropriate office, which is not situated in the location of your residence or your workplace. Which of the given bus lines would you choose in order to visit the office during morning hours and return to your workplace until the beginning of the afternoon shift (until 13:30)? Your departure place is the place of your residence (all the three locations will be specified) OPENING HOURS M. 8:00 ­ 17:00 T. 8:00 ­ 13:00 W. 8:00 ­ 17:00 T. 8:00 ­ 13:00 F. 8:00 ­ 13:00 5. Imagine you have bought goods in total of 17,470 CZK on hire-purchase. To pay off the debt you can choose one of the given alternatives: A) Maturity of credit: 15. 1. 2006 ­ 15. 12. 2006, monthly payment of 751 CZK, additional charge of 9,046 CZK until 1.1.2007 B) Maturity of credit 15. 1. 2006 ­ 15. 12. 2008, monthly payment of 751 CZK, additional charge of 0 CZK. Which of the two given alternatives is cheaper in total? 6. Please specify terms of written (formal) job contract. 127 c. Qualitative research The research for this report also included qualitative research: (i) focus groups with jobless Roma, (ii) structured interviews with experts in the twelve survey localities and (iii) structured interviews with employers. Two focus groups took place, with ten participants each, in the municipality of Velké Kuntice in the Jeseník microregion and in the Pedlice locality in Ústí nad Labem. The focus groups covered views of jobless Roma with respect to their situation and barriers in finding work, options for overcoming such barriers as well as the roles of various actors, from social services, the Labor Office, schools, employers and businesses, NGOs and others. Expert interviews were conducted by trained interviewers in all 12 localities. The interviewers' questions were primarily focussed on the issue of experience with Roma employment, personal motivation, and barriers to entering the labour market. The experts represented a range of specialists on the issues of mediating employment, long-term unemployment, child and adult education, matters of social legislation and regulation and labour recruitment. The expert interviews also included employers representing state institutions, non-governmental non-profit organisations, as well as businesses. Experts included Roma as well as non-Roma. Overall the number of people who commented on the employment of Roma included 42 Labour Office representatives ­ managers, employment agents, and counsellors; 24 representatives of non-governmental, non-profit organisations ­ mostly organisations involved in providing social services, often targeted to Roma; 6 representatives of local governments; 24 workers from municipal departments of social services ­ managers, field social workers, social assistance specialists, and Roma consultants; 2 representatives from regional authorities ­ among them one regional coordinator for Roma; 14 experts from primary schools ­ educators, directors, and teaching assistants; 2 employees of personnel agencies who deal with the recruitment of employees; 8 potential employers, and 1 municipal police officer. Roma made up about 20 percent of the experts, with the majority of them working in the non-governmental, non-profit organisations and in the positions of teaching assistants at primary schools as well as field social workers at the relevant social affairs offices. Structured interviews with 20 employers were conducted to determine views about and experience with hiring Roma as well as hiring decisions and patterns in general. The interviews were conducted in small and large companies as well as locally and internationally owned firms and across big cities, medium-sized towns and rural areas in Bohemia and Moravia. All firms surveyed are firms which have a record of hiring low qualified workers. d. Analysis of employment incentives The report also presents Net Replacement Rates and Marginal Effective Tax Rates for different family configurations and different net wages. 128 The Net Replacement Rate (NRR) is defined as the ratio of net income when unemployed to the net income when employed. The ratio takes values from 0 to 100. The higher the ratio, the lower the incentives to look for an employment opportunity. For example, at a NRR of 100, there are no monetary incentives to look for a job, since a given household receives the same level of income no matter the employment status. Since households enjoy not only consumption but also leisure and also face search costs and fixed costs of participating in the labor market (transportation to work, higher cost of food outside of the household, etc.), it is reasonable to expect that even net replacement rates significantly below 100 may not provide sufficient incentives for job search. Furthermore, the actual level of a NRR that can be expected to effectively generate labor- supply incentive depends not only on valuation of leisure and transaction costs, but also on outside options such as shadow-economy employment opportunities, etc. The report therefore does not attempt to argue whether a specific value of NRR is sufficiently high or low; instead, it shows how the level of NRR changed over time. NRR calculations are routinely available from the OECD. It is therefore important to explain how this exercise differs from that conducted by the OECD. First, the OECD calculates NRRs at two earnings levels: at the Average Production Wage (APW) and at a wage level corresponding to 67% of the APW. However, it is important to note that even 67% of APW may be much above the potential wage levels in low-skill service and laborer jobs available to low-educated socially excluded workers, i.e. workers strongly affected by the incentive effects of social support schemes. This report therefore offers NRR calculations for a wider set of income levels. Second, as social benefits depend on family structure, the NRRs must be calculated for specific family types. In its calculation of NRRs, the OECD considers the following four family types: a single adult, a couple, a couple with 2 children and a single parent with 2 children. For married couples the potential earned income relates to one spouse only while the other spouse is assumed to be inactive with no earnings. The children are assumed to be aged 4 and 6. In contrast, in the calculations for this report families with more than two children are covered. The report follows OECD (2004) and calculates two types of NRRs: the so-called short-run NRR, which is the rate applicable in the period when an individual receives unemployment and social benefits, and the so-called long-run NRR, which applies after the expiration of unemployment benefits and therefore reflects social benefits receipt by long-term jobless. The Marginal Effective Tax Rate (METR) captures the combined effects of income taxation and social benefits taken from a household that increases its gross earned income. Because the maximum level of benefits is often received by families with no income and because benefits are sometimes reduced almost crown for crown with additional earnings, static labor-supply theory predicts the welfare system to discourage labor force participation and hours of work. It is therefore important to measure the strength of such disincentives, i.e. the slope of the implicit tax schedule. METR rates of over 100% indicate strong inactivity traps where agents have low incentives to increase their gross income. This report calculates METR for increases by 200 CZK -- corresponding to a gross-income increase of 2% at (monthly) income level of 10,000 CZK. 129 Short-term net replacement rates NRR, Single, Short-term NRR, Couple no children, Short-term NRR, Couple 1 child, Short-term 80% 105% 100% 100% 75% 95% 90% 70% 90% 85% 65% 80% 80% 60% 70% 75% 70% 55% 60% 65% 60% 50% 50% 8000 10000 12000 14000 16000 18000 20000 8000 10000 12000 14000 16000 18000 20000 8000 10000 12000 14000 16000 18000 20000 Income (CZK) Income (CZK) Income (CZK) 2007 2008 2007 2008 2007 2008 NRR, Couple 2 children, Short-term NRR, Couple 3 children, Short-term NRR, Couple 4 children, Short-term 105% 105% 105% 100% 100% 100% 95% 95% 95% 90% 90% 90% 85% 85% 85% 80% 80% 80% 75% 75% 8000 10000 12000 14000 16000 18000 20000 8000 10000 12000 14000 16000 18000 20000 75% 8000 10000 12000 14000 16000 18000 20000 Income (CZK) Income (CZK) 2007 2008 Income (CZK) 2007 2008 2007 2008 NRR, Couple 5 children, Short-term 105% 100% 95% 90% 85% 80% 75% 8000 10000 12000 14000 16000 18000 20000 Income (CZK) 2007 2008 Long-term net replacement rates NRR, Single, Long-term NRR, Couple No Children, Long-term NRR, Couple 1 Child, Long-term 90% 110% 105% 100% 100% 80% 95% 90% 90% 70% 80% 85% 60% 80% 70% 75% 50% 60% 70% 65% 40% 50% 60% 40% 55% 30% 8000 10000 12000 14000 16000 18000 20000 8000 10000 12000 14000 16000 18000 20000 8000 10000 12000 14000 16000 18000 20000 Income (CZK) Income (CZK) Income (CZK) 2007 2008 2007 2008 2007 2008 130 NRR, Couple 2 Children, Long-term NRR, Couple 3 Children, Long-term NRR, Couple 4 Children, Long-term 105% 105% 105% 100% 100% 100% 95% 95% 95% 90% 90% 90% 85% 85% 85% 80% 80% 80% 75% 75% 75% 70% 70% 70% 65% 65% 65% 60% 60% 60% 8000 10000 12000 14000 16000 18000 20000 8000 10000 12000 14000 16000 18000 20000 8000 10000 12000 14000 16000 18000 20000 Income (CZK) Income (CZK) Income (CZK) 2007 2008 2007 2008 2007 2008 NRR, Couple 5 Children, Long-term 105% 100% 95% 90% 85% 80% 75% 70% 65% 8000 10000 12000 14000 16000 18000 20000 Income (CZK) 2007 2008 Short-term marginal effective tax rates METR, Single, Short-term METR, Couple No Children, Short-term METR, Couple 1 Child, Short-te rm 60% 120% 200% 50% 100% 150% 40% 80% 100% 30% 60% 50% 20% 40% 20% 10% 0% 0% 0 0% 800 000 000 8000 10000 12000 14000 16000 18000 20000 10000 12 14000 16000 18000 20 8000 10000 12000 14000 16000 18000 20000 Income (CZK) Income (CZK) Income (CZK) 2007 2008 2007 2008 2007 2008 METR, Couple 2 Children, Short-term METR, Couple 3 Childre n, Short-term METR, Couple 4 Children, Short-term 700% 1000% 500% 900% 450% 600% 800% 400% 700% 350% 500% 300% 600% 400% 250% 500% 200% 400% 300% 150% 300% 100% 200% 200% 50% 100% 100% 0% 0% 0 0 0% 8000 000 00 8000 10000 12000 14000 16000 18000 20000 10 12000 14000 1600 1800 200 8000 10000 12000 14000 16000 18000 20000 Income (CZK) Income (CZK) Income (CZK) 2007 2008 2007 2008 2007 2008 METR, Couple 5 Children, Short-term 1200% 1000% 800% 600% 400% 200% 0% 8000 10000 12000 14000 16000 18000 20000 Income (CZK) 2007 2008 131 Long-term marginal effective tax rates METR, Single, Long-term METR, Couple No Children, Long-term METR, Couple 1 Child, Long-term 60% 350% 120% 50% 300% 100% 250% 40% 80% 200% 30% 60% 150% 20% 40% 100% 10% 20% 50% 0% 0% 0% 8000 10000 12000 14000 16000 18000 20000 8000 10000 12000 14000 16000 18000 20000 8000 10000 12000 14000 16000 18000 20000 Income (CZK) Income (CZK) Income (CZK) 2007 2008 2007 2008 2007 2008 METR, Couple 2 Child, Long-term METR, Couple 3 Childre n, Long-term METR, Couple 4 Children, Long-term 500% 700% 1000% 450% 900% 400% 600% 800% 350% 500% 700% 300% 600% 400% 250% 500% 200% 300% 400% 150% 200% 300% 100% 200% 100% 50% 100% 0% 0% 0% 8000 10000 12000 14000 16000 18000 20000 8000 10000 12000 14000 16000 18000 20000 8000 10000 12000 14000 16000 18000 20000 Income (CZK) Income (CZK) Income (CZK) 2007 2008 2007 2008 2007 2008 METR, Couple 5 Children, Long-term 600% 500% 400% 300% 200% 100% 0% 8000 10000 12000 14000 16000 18000 20000 Income (CZK) 2007 2008 132 12°E 13°E 14°E 15°E 16°E 17°E This map was produced by the Map Design Unit of The World Bank. CZECH REPUBLIC The boundaries, colors, denominations and any other information shown on this map do not imply, on the part of The World Bank Group, any judgment on the legal status of any territory, or any endorsement or acceptance of such boundaries. SELECTED CITIES AND TOWNS REGION (KRAJ) CAPITALS NATIONAL CAPITAL To To Gorlitz RIVERS 51°N To To MAIN ROADS Dresden To To POLAND RAILROADS Dresden To To Jelenia Gora Decin REGION (KRAJ) BOUNDARIES Liberec S u d e t y To To Ceska Lipa Jelenia Snezka INTERNATIONAL BOUNDARIES GERMANY Teplice Gora 1,602 m To To L I B E R E C K Y Usti Nad Labem Chemnitz Most Trtnov E lbe Turnov ur To To Plauen Ergz Gebirge Chomutov Nachod 18°E 19°E U S T E C K Y Mlada To To Boleslav Jicin Klodzko Zatec K R A L O V E H - To To Karlovy Var ary Louny R A D E C K Y Opole Oder Hradec To To Klodzko K A R L O - Kladno Kralove PRAGUE To To Cheb Podebrady Opole To To 50°N VA R S K Y Rakovnik To To E lbe Pardubice Jesenik Rybnik 50°N Marktrdwitz PRAHA Usti Chrudim Marianske Kralovice Beroun Kutna Ceska O Sumperk Opava Lazane Trebova Benesov Hora PA R D U B I C K Y LO Zabreh Ostrava To To Bielsko-Biala Plzen S T R E D O C E S K YSazava Hlinsko Svitavy M Frydek Cesky Tasin Bor Rokycany O Moravia M O R AV S K O S L E Z S KOder Mistek To To P L E Z E N S K Y Pribram Y Nurnberg Havlikuv Olomouc Novy Frenstat Brod U C K Jicin Y Domazlice Nepomuk To To Klatovy Vltava Lipnik Pelhrmov Prerov Tabor Prostejov Jihlava Zilina To To Z L I N S K Y Regensburg Cesky Picek V Y S O C I N A Kromeriz Strakonice Les Brno J I H O C E S K Y Trebic Zlin To To Jindrichuv Hradiste 49°N To Deggendorf Sumav a Hradec ChribyUherske Trencin 49°N Horni Vitavice Ceske J I H O M O R AV S K Y Budejovice To To Hodonin Deggendorf Znojmo To To Cesky Nitra Velenice To To To To Cesky Zwetti Breclav Passau Krumlov Moravia SLOVAK REPUBLIC CZECH To To To To Vienna Vienna To To REPUBLIC Bratislava Danube To To AUSTRIA Linz 0 10 20 30 40 50 Kilometers DECEMBER IBRD 0 10 20 30 40 50 Miles 33395 2004 12°E 13°E 14°E 15°E 16°E 17°E 18°E