Reform of placement services

MUTUAL LEARNING PROGRAMME: HOST COUNTRY DISCUSSION PAPER - GERMANY Reform of placement services Peer Review on Systematic Preventive Integration Approach (Support) for Jobseekers and Unemployed Germany, 28-29 October A paper submitted by Hugh Mosley in consortium with GHK Consulting Ltd and CERGE-EI Date: 10/09/2010

This publication is supported for under the European Community Programme for Employment and Social Solidarity (2007-2013). This programme is managed by the Directorate-General for Employment, Social affairs and Equal Opportunities of the European Commission. It was established to financially support the implementation of the objectives of the European Union in the employment and social affairs area, as set out in the Social Agenda, and thereby contribute to the achievement of the Lisbon Strategy goals in these fields. The seven-year Programme targets all stakeholders who can help shape the development of appropriate and effective employment and social legislation and policies, across the EU-27, EFTA- EEA and EU candidate and pre-candidate countries. PROGRESS mission is to strengthen the EU contribution in support of Member States' commitments and efforts to create more and better jobs and to build a more cohesive society. To that effect, PROGRESS will be instrumental in: providing analysis and policy advice on PROGRESS policy areas; monitoring and reporting on the implementation of EU legislation and policies in PROGRESS policy areas; promoting policy transfer, learning and support among Member States on EU objectives and priorities; and relaying the views of the stakeholders and society at large For more information see: http://ec.europa.eu/social/main.jsp?catid=327&langid=en The information contained in this publication does not necessarily reflect the position or opinion of the European Commission.

CONTENTS 1. QUICK SUMMARY... 1 2. LABOUR MARKET SITUATION IN THE HOST COUNTRY... 2 3. POLICY MEASURES... 5 4. RESULTS... 9 5. DIFFICULTIES AND CONSTRAINTS... 13 6. SUCCESS FACTORS AND TRANSFERABILITY... 14 ANNEX 1: REFERENCES... 17 ANNEX 2: SUMMARY TABLE... 19

1. QUICK SUMMARY This paper has been prepared for a Peer Review within the framework of the Mutual Learning Programme. It provides information on the policy example of the Host Country for the Peer Review in this case, Germany For information on the views of the countries participating in the Peer Review, please refer to the relevant Peer Review Comments Papers. The present model for German placement services was established by the so-called Hartz Reforms in 2003 to 2005. The major elements of this belated PES modernisation strategy in employment services are similar to those in other OECD countries: A new customer-oriented service delivery model; Profiling and segmentation in service delivery; New emphasis on activation; Increased reliance on external provision of services and competitive tendering. The changes that are the focus of this Peer Review, i.e. (1) the Four-Phases Model, (2) the Placement Budget and (3) Placement-related contract services, are essentially elements of a post-reform optimisation strategy after several years of experience and comprehensive evaluation research. The Four-Phases-Model provides a more systematic framework, especially for profiling, for both PES agencies and the Jobcentres for unemployed on social assistance. The new Placement Budget and the new regulation on Placement-related Services are not new programmes but consolidations of much more detailed older regulations. Instead of the former statutory catalogue that defined the type, level and prerequisites for services to be offered in some detail, the new general statutory provisions define broad purposes and give broad discretion to individual PES agencies regarding the type and level of cash assistance or the content and duration of placement related services. Because all three reforms are relatively new a definitive assessment of their impact is not yet possible. Interviews conducted in Berlin with staff in PES agencies and Jobcentres indicate that the 4-Phases-Model and the IT support provided is generally regarded as a further development and significant improvement over the previous model, in particular because it provides a basis for much more differentiated profiling (six instead of four client profiles) with IT-based links to strategies and products. The impact on placement processes in the Jobcentres is even greater because it has served to focus placement activity on a clear integration goal, in contrast to the rather diffuse survey of client needs that was previously practiced. The new statutory flexibility provided by the Placement Budget for ad hoc financial assistance to assist jobseekers in meeting the incidental costs related to job search and starting new employment and the new regulation on placement-related contract services appear to have had only limited impact in practice. Staff interviews and early impressions from on-going implementation research indicate that, despite statutory deregulation, PES agencies and Jobcentres have not responded with innovation but, with some exceptions, largely oriented-themselves on past practice. 1

2. LABOUR MARKET SITUATION IN THE HOST COUNTRY 1 2.1 The labour market situation The impact of the strong economic recession on the German labour market has been ameliorated by labour market programmes and, in particular, heavy use of short-time work. Thus while annual average Euro zone unemployment rose from 7.9% to 9.4% between 2008 and 2009, the German rate increased only slightly from 7.3 to 7.5%. Men and young people were most strongly affected by the crisis. By June 2010 seasonally adjusted unemployment had fallen below the pre-crisis level to 7%. Long-term unemployment has also declined from 5.5% to 3.4% of the labour force but still remains high as a percentage of all unemployed. Although unemployment is still too high, especially in Eastern Germany, there is a considerable dynamic in the employment service register: While there were statistically 9.25 million entrants to registered unemployment in 2009, 9 million persons left unemployment. Persons leaving the register had been unemployed for 33 weeks on average. Much of this turnover in the register is, however, for statistical or administrative reasons: only about 1/3 of exits from the register actually entered employment. 2 Expenditure for active and passive labour market policies have declined sharply between 2005 and 2008 with the decline in registered unemployment, rising again in the current crisis. There has been a marked improvement in the case load in placement services, which declined from 320 to 135 clients per job counsellors in the PES agencies over the same period. Case loads are even lower (165 for adults and 85 for youth) in the Jobcentres. Table 1: German Labour Market Indicators 2005-2009 2005 2006 2007 2008 2009 GDP (% change) Employment (1000s) Active policies (%GDP) Passive policies (%GDP) 0.8 3.2 2.5 1.3-5 26,236 26,365 26,942 27,510 27,503 0.89 0.86 0.72 0.81-2.02 1.71 1.30 1.10-1 About the same number entered education or training (including labour market measures) or were stricken from the register for other reasons (e.g. health reasons, non-renewal of registration, not being available for work; BA 2009) 1 Thanks are due to Dr. Yvonne Kaiser and Frank Festner in the German Labour Ministry for their support as well as the Berlin PES Regional Office and PES placement staff in Berlin for interviews and documentation. The author alone is responsible for the contents. 2 About the same number entered education or training (including labour market measures!) or were stricken from the register for other reasons (e.g. health reasons, non-renewal of registration, not being available for work; BA 2009) 2

Unemployment 2005 2006 2007 2008 2009 Unemployment (1000s, registered.) Unemployment (%) Long-term (>12m.) (%) 4861 4487 3776 3268 3423 10.7 9.8 8.4 7.3 7.5 5.7 5.5 4.7 3.8 3.4 Female (%) 10.1 9.5 8.3 7.2 6.9 Male (%) 11.2 10.2 8.5 7.4 8 Youth (<25) % 14.2 12.8 11.1 9.9 10.4 Low education (%) 19.8 19.6 17.7 16.2 16.4 Sources: Unemployment rates: Eurostat Labour Force Survey. GDP, employment and registered unemployment: Federal Employment Agency. Low education = pre-primary primary and lower secondary education only- levels 0-2 (ISCED 1997); OECD Labour Market Statistics 2.2 New model employment services: The Hartz Reforms A major reform of the German employment services took place between 2003 and 2005, the so-called Hartz Reforms named after the chairman of the reform commission who proposed them. The reform took place in two distinct waves: (1) reform of the public employment service (Bundesagentur für Arbeit; hereafter PES) beginning in 2003 (German Social Code III hereafter SGB III) and (2) the reform of employment services for social assistance recipients, Basic Income Support for Jobseekers (SGB II), beginning in 2005. These reforms established a dual system of service delivery: (1) the PES, which provides employment services primarily to the insured unemployed, and (2) the Jobcentres of the Basic Income Support programme, which serves the unemployed on social assistance. The most recent changes that are the focus of this Peer Review (1.Four-Phases Model, 2 Placement Budget and 3.Placement-related contract services) are essentially elements of the post-reform optimisation strategy after several years of experience and comprehensive evaluation research. The following section provides an overview of the original Hartz reforms before discussing the most recent changes, which can only be understood in this context. 2.2.1 The Public Employment Service (PES) Even prior to the Hartz reforms, the PES was a highly centralised organisation. The first wave of Hartz reforms (Hartz I-III) in 2002-2003 and internal organisational changes, in which external management consulting firms such as McKinsey played a central role, further strengthened centralisation along new public management principles. The major elements of this belated PES modernisation strategy in employment services are similar to those in other OECD countries (see WZB/infas 2006; Schütz and Oschmiansky 2006): A new customer-oriented service delivery model 3

Profiling and segmentation in service delivery; New emphasis on activation; Increased reliance on external provision of services and competitive tendering New service delivery model. Employment services for both employers and for the unemployed were reorganised; This included, for example, a single entry point in which simple requests can be resolved immediately, separate departments for processing benefit claims and for job counselling and placement, counselling by appointment, a special employer service unit for employer contacts and referring applicants to vacancies. Profiling and client segmentation: An IT-based service management system (the so-called Handlungsprogramme or service guidelines) was introduced that standardises interaction between placement counsellors and clients. In an initial profiling stage, assessment, jobseekers were sorted into four customer groups (Kundengruppen) based on their distance from the labour market. The customer group to which jobseekers are assigned determines what specific reintegration services (employment and training schemes, job search assistance etc.) are offered and the frequency of client contact. The service guidelines focused employment counselling and reintegration services mainly on the middle customer segment of customers needing counselling. So-called market clients, i.e. those with good employment prospects, were to receive only limited assistance, as they will in all likelihood find work on their own without help from the PES. The same cost-benefit logic was also applied to the fourth category of support clients deemed to have little chance of finding regular employment before their PES benefits expire after one year, resulting in those facing the greatest problems in the labour market benefiting least from early intervention, being excluded from most reintegration measures until they became the responsibility of the Jobcentres. Activating the unemployed: In the course of the Hartz-Reforms, there were various changes in the rules for benefit recipients designed to activate the unemployed and thus to speed up the placement process. In addition to the shortening of the duration of unemployment insurance benefits and reductions in benefit levels for the long-term unemployed (see 1.2.2), there have been other changes in placement services, for example, a new flexible sanctions regime. 3 Furthermore, there has been a strong emphasis on reducing the case load of individual job counsellors. Outsourcing of reintegration services: One of the most far-reaching changes was the restructuring of outsourcing with a new emphasis on vouchers and a restructuring of the tendering process for PES providers. The German PES has always outsourced a large share of activation, training, job creation and other reintegration services. Whereas in the past the tendering process for labour market services was managed by the local PES agency, after 2003 a highly formalised process of competitive tendering through five PES regional purchasing agencies was introduced. There has also been an increased reliance on contracting out of placement services. In 2002 a placement voucher ( 421g SGB III) was also introduced, to which the jobseeker is entitled after two months of unemployment. 4 2.2.2 Basic Income Support for Jobseekers The labour market reform legislation on Basic Income Support for Jobseekers (SGB II) that came into effect in January 2005 ( Hartz IV ) provides a framework for integrated provision of benefits and labour market services to the long-term unemployed and other employable social assistance recipients. All needy employable persons not eligible for unemployment 3 The new flexible sanctions regime was intended to reinforce work and job search requirements by introducing graduated, initially less drastic, sanctions and shifting the burden of proof to the jobseeker. See 2.1 below. 4 The PES pays a commission to the private employment service ( 2000), if it successfully places the job-seeker in a regular employment. 4

insurance benefit receive the new Unemployment Benefit II, 5 a consolidated benefit near the social assistance level, and reintegration services from the new local agencies of the Basic Income Support programme (SGB II). These services and benefits are largely funded by the federal government but administered jointly by the PES and the local authorities in special Jobcentres for this client group. Thus alongside the PES a new and separate means tested delivery system for job seekers on social assistance was created The principal rationale was to provide labour market services for all social assistance recipients from a single agency. In addition to PES labour market services, the Jobcentres can also provide typical municipal social services (e.g. child care, debt & drug counselling, family and psychological counselling) not available from the PES. 6 3. POLICY MEASURES Two of the new policy measures, the Placement Budget ( 45 SGB III) and Placementrelated Services ( 46 SGB III) examined here are part of the Reform of Labour Market Instruments ( Instrumentenreform ), which came into effect in January 2009. Both new regulations are not really new programmes but a consolidation of several existing programmes on a new statutory basis. The changes were intended to simplify and rationalise the bloated portfolio of statutory instruments, making them easier to administer, and to provide greater flexibility in implementation for both PES agencies and job counsellors. The so-called 4-Phases Model was developed, parallel to the broader reform, within the public employment service in response to perceived shortcomings in the placement process, especially in the Jobcentres. Moreover, the Four-Phases-Model provides, for the first time, a common framework for reintegration services for jobseekers in both the PES agencies and the Jobcentres. Because the Four-Phases-Model represents the PES core placement strategy, it is discussed first and thereafter the two new instruments in support of the placement process, the Placement Budget and Placementrelated Services ( 45 & 46 SGB III). Because the new policy measures are very recent and no systematic evaluations are yet available, the findings reported here should be regarded as preliminary and not necessarily representative. They are based on documentary analysis, available PES data and on a small number of semi-structured interviews with placement staff in a PES agency and two Jobcenters, in a Regional PES office, and in the German Labour Ministry. 3.1 Four-Phases Model for Reintegration Services Active intervention under the 4-Phases Model commences with the initial registration as a jobseeker. There is, moreover, an early registration requirement for jobholders at risk. 7 The objective for the latter is to use the time span before the beginning of unemployment (the so-called action time ), to achieve a job-to-job placement without any intervening spell of unemployment. Phase I: Profiling Profiling in some form is a widespread practice in European employment services. In general it aims, based on an analysis of known characteristics of a jobseeker, to analyse employment prospects, to predict reintegration chances and to prepare the reintegration plan. As such, profiling has long been a feature of labour market policy in Germany, and not only since the Hartz-Reforms. A new development includes increased emphasis on the jobseekers strengths - versus his or her shortcomings both in staff training and in the 5 Unemployment insurance benefit recipients with low benefit levels may also qualify. 6 On a trial basis 69 local authorities (municipalities and counties) were given full responsibility for implementing employment services along with the 353 local Jobcentres initially established. In 2010 this local option was made permanent and expanded to include up to 110 localities. 7 Jobholders are obligated to register as jobseekers three months before the end of an employment or, for those with shorter notice, within 3 days after receiving notice of dismissal, under threat of a sanction (denial of benefit for 1 week; 38 & 144 SGB III. 5

placement software that structures interaction with clients ( Verbis ). This shift, which is mandated by statute, 8 is intended to broaden the matching process to include types of work beyond the jobseeker s previous occupation. The profiling with potential analysis should identify clients individual support needs and provide a prognosis on the client s distance to the labour market and probable duration of job search according to six client profiles: Market profile: Clients without placement-related deficits (1-6 months); Activation profile: Clients with deficits in motivation (1-6 months); Promotion profile (Förderprofil): Clients with simpler deficits in capabilities, or in their life circumstances (1-12 months) qualifications, Development profile: Clients with deficits in qualifications, capabilities, or in their circumstances as well as in a further dimension, or a severe deficit in one of the former (> 12 months); Stabilisation profile: Clients with deficits in capabilities as well as at least in two other dimensions, or a severe deficit in one of the former (1-12 months); Support profile: Clients with deficits in circumstances as well as at least in two other dimensions or a severe deficit in circumstances (>12 months). For job-ready persons profiling would normally take place during the first interview, for which a total of 60 minutes are recommended, including preparation and post-interview recordkeeping. There are a number of software-supported tools to assist job counsellors in profiling and assigning clients to profile groups (e.g. Calculating Labour Market Chances (BAC) and Profiling Criteria Catalogue ) but the responsibility lies ultimately with the individual job counsellor or case manager, who may ultimately override the software recommendation. 9 Phase II: Goal definition On the basis of the profiling results, the job counsellor or case manager develops jointly with the jobseeker a realistic labour market goal, which may be supplementary employment, training, stabilisation of existing employment (especially for those on social assistance) or an alternative to regular employment for those with little chance of finding employment in the next 12 months (e.g. workfare). The agreed goal should be thematised in all subsequent client interviews and any changes documented. Phase III: Selection of strategy On the basis of the strengths and deficits identified in profiling and the reintegration goal a concrete action plan is agreed with the client. Based on his or her deficits (Handlungsbedarf) the placement software (Verbis) suggests a number of very general and more or less self-evident service strategies (Handlungsstrategien) and for each a number of concrete strategy options and the steps and time frame that they entail in support of the counselling process. A product catalogue provides guidelines for job counsellors on what labour market programmes are recommended for each service strategy. As in the past, assignment of clients to active measures should also be guided by considerations of effectiveness and efficiency. Referral of clients to active measures is, with few exceptions (e.g. the placement voucher after two months of unemployment), a discretionary decision of the PES for which the standardised service guidelines provide norms. In practice budget restrictions and available capacities also limit the options available. Phase IV: Integration agreement and implementation The integration plan agreed between the job counsellor and the jobseeker is given written form in a formal agreement signed by both the jobseeker and the placement counsellor. If 8 A potential analysis for newly registered jobseekers must be conducted without delay ( 37 SGB III). 9 This occurs not infrequently in case of personal or family problems that may not entered into administrative records without the client s permission (e.g. alcohol, drug or psychological problems). 6

no agreement is reached obligationsmay be imposed on the client by an administrative act. Previously obligatory only for social assistance recipients (SGB II), the new regulation ( 37) now makes conclusion of an integration agreement obligatory for PES clients as well. The intent is to make the plan more transparent and to enhance commitment on the part of the jobseeker. The agreement also serves as an orientation for follow-on interviews with the same client. For adult clients (>25 years) the integration agreement is normally concluded for a period of 6 months and becomes a reference point for subsequent interviews. After the conclusion of the integration agreement the software prompts a search of job openings based on the results of the profiling. The interview ends with the documentation of the results in the client s electronic file, which is generated semi-automatically based on the previous entries during the course of the interview but can be amended or supplemented before storing. Follow-on interviews with the same client are likewise structured by the Verbis software with several core elements: 1) controlling adherence to the integration agreement; 2) reexamination of the profiling, goal, and choice of strategy, which are amended as necessary; 3) discussion of job search activity; 4) agreement on next steps and on the date for the next interview. Whereas there was previously a central norm for the frequency of contact with clients, this is now at the discretion of the local agency or Jobcentre and the job counsellor. Refusal of a job offer or an intervention measure, dropping out of a measure or insufficient own efforts, may lead to a suspension of benefit payments being imposed by the job counsellor: Three weeks for the first, six weeks for the second, and twelve for the third offence. Failure to actively search for work can result in a suspension for 2 weeks. 3.2 Placement budget ( 45) Objectives: Flexible, individualized and unbureaucratic financial assistance for job search and start up costs. Target groups: Unemployed or those at risk of unemployment seeking insured employment or an apprenticeship Timeframe: January 1, 2009 - Geographical and sectoral scope: National, all persons seeking insured employment Procedures and staff resources: Discretionary benefit available in all PES agencies and Jobcentres Financial framework: No separate budget; local agency discretion within global reintegration budget (2009 = 311 million). Participants 2009: 2.35 million Legal framework: German Social Code, 45 Book III and PES regulations Institutional framework: Implementation by PES and Basic Income Support (SGB II) agencies. Legal supervision by Ministry of Labour and Social Affairs The goals of the new Placement Budget are to make new types of assistance for jobseekers possible, to achieve efficiencies by shifting the focus away from what benefits are available, towards what assistance is necessary and to reduce the administrative burden and increase transparency for PES staff. 10 It consolidates in one regulation a number of existing separate regulations under which job counsellors can provide flexible financial assistance to job seekers in meeting the incidental costs incurred in searching for or starting a new job. Prior to the reform, financial assistance could be provided for specific purposes such as the costs of job applications, travel costs for job interviews, and moving expenses, work clothing and tools etc. when starting a new job. Instead of a detailed statutory catalogue that defined permissible types of assistance, the new regulation gives the PES agency or Jobcentre broad discretion in determining whether and what type and level of assistance is necessary and appropriate. The Placement Budget can, however, 10 See Begründung in Bundestag 2008 Entwurf eines Gesetzes zur Neuausrichtung der arbeitsmarktpolitischen Instrumente, Dt. Bundestag 16. Wahlperiode, Drucksache 16/10810, pp. 31-32. 08.11.2008. 7

only be used for incidental financial assistance related to finding and starting a new job and not for intervention measures, which are regulated separately. It should be noted that the Placement Budget is not an individual budget available to jobseekers nor is there separate national funding for it, as the name might imply. 11 Individual PES agencies and Jobcentres allocate resources for it from their own global integration budgets for active measures. The size of their budgets depends on the number of jobseekers they serve and on the size of the overall budget approved by the legislature. The Placement Budget thus competes with other expenditure for active measures and is subject to local discretion. 3.3 Placement-related Services ( 46 SGB III) Objectives: More flexible contracting of placement-related services Target groups: Unemployed or those at risk and youth seeking apprenticeship Timeframe: January 1, 2009- Geographical and sectoral scope: National Procedures and staff resources: Available in all PES agencies and Jobcentres according to national and local guidelines. Entitlement after 6 months unemployment. Financial framework: No separate budget; local agency discretion within global reintegration budget (2009= 760 million) Participants: 1.4 million in 2009 (435,000 with an employer) Legal framework: German Social Code, 46 Book III and PES regulations Institutional framework: Implementation by PES and Basic Income Support (SGB II) agencies. Legal supervision by Ministry of Labour and Social Affairs The new regulation for placement-related services authorises five general types of labour market measures without further detailed statutory regulation: (1) activation and labour market orientation; (2) assessment and reduction of placement handicaps; (3) private placement services; (4) support in transition to self-employment; (5) stabilisation in a new employment. These may be either individual or group measures and may take place either with a provider or as on the-job training with an employer. Both content and duration of the measure are flexible. Like the Placement Budget, the regulation is not a new programme but consolidates a number of existing separate programmes in a more flexible general regulation. 12 In the past so-called training and activation measures were authorised for a variety of specific purposes, for example, job search assistance (e.g. for preparation of job applications, job interviews), aptitude and skills assessment, in addition to short-term training (e.g. computer course, or language training), on-the-job training with an employer ( 48 52 SGB III) as well as referral to private placement services. In the past the content and maximum duration of the measures was largely defined by statute (Kopf and Wolf 2009). Which jobseekers are referred to placement-related contract services and the particular type of measure to which they are referred is, according to PES internal guidelines, at the discretion of the PES agency or Jobcentre and should be governed by the integration agreement with the client. How long jobseekers participate is likewise at the discretion of the job counsellor or case manager, except that training with an employer is limited by statute to four weeks and vocational training in a labour market measure to eight weeks Importantly, contracting out of placement-related services under the new regulation has to be carried out under statutory provisions on competitive tendering. 13 The PES own internal 11 The notion of individual budgets was considered but discarded because it might have led to an entitlement mentality. 12 The most important were Commissioned Placement Services ( 37 SGV III), Integration measures ( 421i); Assessment & Training measures ( 48); Youth Activation ( 241 Abs. 3a SGB III) and the so-called Personnel - Service-Agencies ( 37c), subsidized temporary work agencies for the unemployed. 13 This was also the case for the previous measures that were merged into the new regulation. 8

guidelines require that PES agencies contract all measures with providers through the highly formalised procedures of the Regional Purchasing Offices (REZ), which is an important limitation on local flexibility 14. As a rule, training and integration measures are selected from a menu of approximately 12-15 REZ standard products for the PES agencies and Jobcentres respectively. Because of its central role in the tendering process, the REZ is thus de facto an important regulator of the type of measures offered. 15 4. RESULTS 4.1 The German Placement Service Model We focus here on the new German model of placement services as it has developed since the Hartz reforms in 2003-2005 (see 1.2 above) and not exclusively on the most recent changes associated with the 4-Phases Model. Because the latter was first implemented in July 2009 in the Jobcenters and in December in the PES agencies, there is only limited experience with it and no systematic evaluation. 16 According to earlier evaluation results for the Hartz organisational model (2002-2006) in placement services (Kundenzentrum, Kundengruppen and Handlungsprogamme), it contributed to a more effective work organisation and better structured counselling, improving the quality of services for both jobseekers and employers (iso/ochs 2006; WZB/infas 2006). Noteworthy in particular are the following: Jobseekers requests and questions are now pre-filtered in an Entry Zone and in the Service Centre (call centre); a large percentage of customer concerns can be handled there and dealt with immediately, relieving counsellors of tasks which have little to do with their actual mission. Scheduled appointments with jobseekers made possible uninterrupted counselling and avoid longer waiting periods. Job counselling was given a new structure, focusing on carefully analysing the needs of customers and job counsellors were relieved of many administrative tasks, allowing them to spend more time in actually counselling jobseekers. Local PES agencies were required to allocate at least 20% of all placement counsellors working time to services for employers and quality standards for employer placement services were introduced, for example, agreeing with the employer the maximum number of jobseekers to be referred and only referring applicants willing to accept the job. Staff in PES agencies regards the 4-Phases Model as a further development of the socalled service guidelines (Handlungsprogramme) and client groups (Kundengruppen) that have been the core of the placement process since 2006. For them the principal innovation is the much more differentiated set of categories (now six client profiles instead of four client groups ) in which the profiling process culminates. The new version of the placement process also offers a correspondingly more differentiated categorisation of client deficits and potential action strategies to address them as well as links to products that may be used to address these needs. This more complex profiling approach is supported by an improved version of the Verbis software that guides the placement counsellor better through the entire process. The new profiling approach begins with an inventory of the 14 The Jobcentres are not required to tender contracts through the REZ but do so in most cases. 15 PES agencies may in principle design up to 25% and Jobcentres an unlimited share of their tenders individually, i.e. outside the standard products of the REZ. However, only a small fraction of all programme slots tendered through the REZ in recent years have been in such individualised measures. 16 Several pilots were conducted before the new model was introduced but no detailed results are available. 9

jobseekers strengths or potential and so takes, at least psychologically, a more positive approach to the jobseekers labour market situation. Aside from this greater differentiation, the basic process of profiling, counselling and conclusion of an integration agreement with clients remains largely unchanged in the PES agencies. The impact of the new model is greatest in the Jobcentres. The placement staff who were interviewed, had a very positive view of the 4-Phases Model, which was said to be a significant improvement over the previous support level approach (Betreuungsstufen), because it focuses the process for the job counsellor on the employment goal ( Zielberuf ) instead of on the diverse needs of clients, which had previously led to a very unstructured approach. The relevant PES products and services are now linked to the agreed goal and strategies that are subsequently documented in the formal integration agreement. Dissatisfaction with the integration concept previously used in the Jobcentres was an important reason for the adoption of the new approach. The changes in the placement process are said to be largely invisible for clients. Working with the sophisticated software is a problem for some placement staff and using it efficiently requires a relatively long learning period. This is especially a problem in the Jobcentres in which staff turnover is high. The new model and software also require more copious documentation of the placement process. From a management perspective, this is important because the exercise of discretion by placemen staff is subject to challenge in courts. The use of a common profiling and integration process in the PES agencies and Jobcenters also reduces the workload for Jobcenter staff for Jobseekers transitioning from the PES to the Jobcenters (about 10% of new entrants and 25% of all benefit recipients; BA 2009c). A transition management module in the 4-Phases Model requires the PES agency to inform the client and update the client profile in preparation for the transition. The client s electronic records can be accessed and, after an initial re-examination and eventually revisions, continued by the responsible placement counsellor in the Jobcentre. Our overall impression is that the 4-Phases Model and its supporting software revision is an improved version of the existing placement process in the PES agencies and a major change in the Jobcentres. Only a systematic evaluation, which is not currently planned, can confirm these impressions and show whether the reform has addressed the systemic problem of exclusion tendencies in the previous PES service guidelines (Handlungsprogramme) that had led to about 30% of clients, with a considerable distance to the labour market, being in practice excluded from most PES reintegration services. The more differentiated profiling and the fact that product recommendations are now keyed to client needs, rather than to client groups based on distance to the labour market, is an important step toward a more inclusive practice. 17 4.2 Placement Budget ( 45) The results of the reform cannot yet be definitively assessed because no evaluation results, or even detailed data on types of expenditure, are available. 18 Our own interviews with PES staff in Berlin and a PES internal report provide some initial evidence. 19 While there are examples of innovative practice, the overall pattern is not very different from the past. The PES agencies are expected to provide guidance to placement staff in order to establish consistent criteria for granting assistance. In the absence of detailed statutory regulation, local PES agencies have adopted internal guidelines oriented on past practice. They usually authorize the same types of assistance as in the past (job application costs, 17 The distribution of client profiles in PES agencies in Berlin and Brandenburg in August 2010 was as follows: market profile 21%, activation profile 12%, promotion profile 40%, development profile 25%, stabilisation profile 1.3%, support profile 1%. No information is available on the access of the latter three client profiles to services. Data from PES Regional Office. 18 The results of an evaluation of the implementation of the Placement Budget ( 45) and Placement-related Services ( 46) conducted by the IAB, the PES research institute, will not be available until the Fall (Pilger and Steinke 2009). 19 The unpublished PES report summarizes the results of an internal PES workshop on experience during the first year of the reform. 10

travel costs for job interviews, moving costs, special work clothing and tools not provided by the employer). In most agencies job counsellors can decide independently up to a locally set maximum amount per client and year (e.g. 260 for job applications). Greater sums have to be approved by team leaders. This is also similar to past practice. Some innovative types of financial assistance are also foreseen: the costs of job-related certificates (e.g. health certificates, language tests, or tests for taxi drivers), personal assistance (e.g. haircuts, dry cleaning) as well as other assistance. The latter, which requires the approval of the team leader, has been used, for example, to subsidise the cost of attaining a driver s license, if necessary for employment. More conventional uses, however, still predominate. In Berlin in 2009, assistance was provided largely for purposes that were explicitly authorized under the previous statute: job search costs (82%), mobility assistance (6%), work clothing and tools (1%), whereas only 2% of all grants was for jobrelated certificates, >0.1% for personal assistance and 3.2% for other assistance. 20 Even these latter new types of assistance were also possible in the past under an earlier regulation for innovative individual measures (Freie Förderung 10 SGB), which was eliminated by the reform. Internal PES regulations encourage individualised assistance at the discretion of the local PES agency and job counsellor. Exercise of their discretion should be orientated on the client s integration agreement as well as his or her need for financial assistance (BA 2009). There is no clear evidence that granting assistance under the new regulation may be easier to administer, except that the number of forms and procedures involved have been reduced under the consolidated regulation. 21 Like any major change in program regulations, there are large upfront costs in staff training, placement software and data systems. New uses for the Placement Budget may increase with time and experience. It is possible for staff to propose innovative new tasks for the Placement Budget but it is difficult to get them adopted in local guidelines. The PES Agency or Jobcentre is understandably concerned to have as a rule the same support available in all placement teams (and there is even Berlin-wide coordination), so innovations have to be adopted at the Agency or Jobcentre level. This requires broad approval of a new concept. Moreover, innovative uses compete with more conventional uses for budget resources. In this setting, the chances of innovative suggestions from placement staff winning acceptance are low. A total of ca. 2.35 million jobseekers received some type of financial assistance through the Placement Budget in 2009 making it one of the largest labour market programs. Total expenditure under the Placement Budget for the PES (i.e. with the Jobcentres for social assistance recipients), amounted, however, to only 311 million, only about 3.5 of all expenditure for active measures since most assistance was small (on average 130.00 per case in 2009). 22 4.3 Placement-related Services ( 46) The reform gives the PES and Jobcentres in theory new flexibility in developing placementrelated support services supplied by external providers. Like for the Placement Budget, there are as of yet no evaluation results available. First impressions from an ongoing IAB evaluation suggest that there has been, as of yet, relatively little change in the type of measures contracted despite the greater statutory flexibility under the new regulation. Placement staff and other interviewed for this paper also report little change. Data on current contracting for placement-related services through the REZ indicate that around 650,000 programme places were contracted in 2009 under the new regulation. The most important contracted measures were the so-called standard Module measures, emphasizing either assessment, job search training, or preparation for self-employment (38%), Intensive Placement (Ganzil/UVgA) with activation, qualification, and individual counselling (15% ); Intensive Job to Job counselling and training for displaced workers 20 Data provided by the PES Regional Office Berlin-Brandenburg. 21 PES expenditure for this type of assistance was less in 2009 than under the previous regulations ( 217 vs. 265 million) but this can have many causes ( BA 2010). 22 Data scource: BA Statistics and Labour Ministry. 11

(12%); Placement with Intensive Counselling and Compulsory attendance (7%) and Job Search Centres & Activation Centres (5%). An additional relatively large share (16%) was for individualised measures, i.e. not REZ standard products. 23 There are unfortunately no data available for previous years that would permit us to examine whether this pattern has shifted significantly since the adoption of the new regulation. There are a number of possible explanations in the implementation process for the apparently as yet limited impact. First, as in the case of the Placement Budget legal uncertainty is a contributing factor - for many months until the autumn of 2009, PES agencies and Jobcentres lacked implementation guidelines for the new regulation. Second, most measures are contracted not directly by local agencies but as REZ standard products. Change thus depends on the introduction of new REZ standard products, which is itself a lengthy process. Moreover, when new REZ products are introduced, changes work their way only slowly through the system in which existing contracts with providers are for 1-2 years, and in many cases longer. Finally existing measures presumably meet important needs. In 2009 there were approximately 1.88 million entrants (about 435,000 with an employer) in placement-related services, including those entering older programs being phased out and merged into the new instrument. With total costs of 761.2 million (about 7.5% of all active expenditure) and costs per entrant was at about 404 relatively low. The average duration of measures under the new instrument was less than 2 months, with considerable variation according to the type of measure. 24 23 Contract data for both SGB III and SGB II kindly provided by the BA-Service House, Nürnberg. Some additional measures contracted directly by Jobcentres as well as all measures in the 69 municipally run SGB II agencies are not included. 24 Source: BA Statistics and Labour Ministry. 12

5. DIFFICULTIES AND CONSTRAINTS This section reflects on the recent reforms in the context of the development of German labour market policy and some the basic challenges and choices any employment service faces. Centralisation vs. local flexibility The German PES has undergone in international comparison a belated modernisation process, strongly embracing management by objectives and related new public management approaches. It has become a highly centralised organisation with regard to its management structures and work organisation and to a large extent, even structures client interaction through IT-based standard operating procedures. While these changes have had in many ways a positive impact on service delivery in the PES, they also entail a farreaching standardisation of work processes in local PES agencies (Schütz 2009). 25 The new 4-Phases Model continues this tendency insofar as it established a common reference model for placement services in both the Jobcentres and PES offices. The Placement Budget ( 45) and the new regulation of Placement-related Services ( 46) represent promising efforts to reduce the density of regulation and grant local agencies and job counsellors greater autonomy. Changes in law are, however, only a necessary but not a sufficient condition for innovative practice. Policy vs. implementation. The impact of legislative changes in programmes on practice in public organisations, depends ultimately on a complex process of implementation. The impact of the new Placement Budget and Placement-related Services ( 45 and 46 SGB III) on the placement process appears initially to be markedly less than one might expect from the very flexible formulation of the relevant statute. It may be that it is too soon to see a real impact. Although the reform legislation came into force in January 2009, its practical implementation in PES operative guidelines and IT system did not take place until into the autumn of 2009. Innovation takes time to work its way through organisations. On the other hand, we should probably not expect too much innovation in response to a more flexible statute. In a hierarchical and rule-oriented organisation like the PES the elimination of detailed statutory regulations may lead to uncertainty and risk adverse behaviour. 26 Dissemination of information on best practice should complement deregulation by legitimizing innovation. Moreover conventional forms of placement assistance and contract services have their justification in clients needs and shrinking budgets are not a fertile ground for innovation. Hard vs. soft profiling Two basic types of profiling can be distinguished in OECD countries: Statistical (hard) profiling based on client characteristics and the experience of similar jobseekers in the past in the local labour market (e.g. in the USA) and more qualitative (soft) profiling that relies on check lists or other guidelines on client skills and deficits and the relevant labour market as assessed by the placement counsellor. This is the approach that has been practiced in Germany in various forms. In more complicated cases profiling is supported by external services now contracted under the new general provision for Placement-related Services ( 46). 27 25 The same is true for the internal work organization and processes in the Jobcentres, which largely follow the PES model, whereas the 69 agencies managed by the local authorities are entirely independent of the PES (Mosley 2010). 26 As one staff member said: Whatever new idea you have, there s a reason why you can t do it. 27 Other PES agencies do not attempt to use profiling as a tool of early intervention at all but resort to profiling and offer active measures only after the unemployment spell has lasted (typically) six months (e.g. UK). This practice is justified with the argument that early predictions of unemployment durations are not sufficiently reliable to serve as a basis for selecting jobseekers for referral to active measures. 13

Profiling provides a systematic framework for the placement counsellors, that is clearly an improvement over more informal approaches. It is, however, also a small wonder when one considers that the placement advisor has only 45 to 60 minutes to go through a complex profiling check list as well as to the other phases of the placement model, and typically has no specialized knowledge of the client s occupational field because clients are assigned alphabetically. It is, therefore, appropriate that soft profiling is regarded by its practitioners as an ongoing process rather than a prediction, subject to revision and deferential to the jobseekers own articulated goals. Early intervention, profiling and client segmentation in service delivery The initial implementation in the new IT-based service management system (the so-called Handlungsprogramme or service guidelines) in the PES agencies was not unproblematic. The service guidelines focused reintegration services on the middle customer segment of customers needing counselling. So-called market clients, i.e. those with good employment prospects, received only limited assistance, on the assumption that they will in all likelihood find work on their own. The same also applies to the fourth category of support clients who, based on profiling, were deemed to have little chance being integrated into the regular labour market. This means that persons who have the greatest problems in the labour market had less frequent contact with placement counsellors and tended to be excluded from most reintegration measures until they became the responsibility of the Jobcentres (Mosley 2008). In response to criticism, a number of special target- group oriented programmes were introduced (e.g. Training for Low Skilled Employees in Firms, 2005; Integration Steps for Support Clients, 2007; Intensive Integration Support, 2008) outside the logic of the service guidelines for hard to place jobseekers. The 4-Phases-Model is itself, in part, a response to this criticism. Its client-profile typology represents a synthesis of the old PES client support groups (Kundengruppen) and the service support steps (Betreuungsstufen) in the Jobcentres. It is conceived as a much more differentiated and a common framework for integration services in both the PES and the Jobcentres, which could, in principle, be the basis for earlier and better planned intervention for job seekers with considerable distance to the labour market. The more differentiated profiling concept and the linkage of recommendations for intervention measures to client needs rather than reintegration prospects, is an important step toward addressing this problem. There is at present too little experience and data to judge whether access to programs and services has become more needs-oriented. The changes in profiling process described above are a necessary but not a sufficient condition for earlier intervention in the unemployment spells of clients with considerable distance to the labour market (or for preventive labour market policies). Also required is a change in incentive structures within the PES, which has in the past been strongly oriented-toward job-ready clients, as well as adequate financial resources, which would reduce the need for rationing. In practice, both PES agencies and Jobcentres see themselves confronted with many clients for whom they see no realistic prospect of reintegration under current labour market conditions and limited resources to provide more intensive services or alternatives to regular employment. This dilemma is in some respects even more acute in the Jobcentres, where the minimalist definition of employability (i.e. capability of working at least three hours per day) blurs the line between labour market and social policy. Many unemployed in Germany would no longer be expected work in other countries. Centralised purchasing vs. deregulation There is, in European employment services, not only an increased reliance on external service provision, but also on centralised competitive tendering conducted at a regional or national level (Mosley and Sol 2005). Since 2004 the PES Regional Purchasing Offices (REZ) have carried out this task in Germany. Centralised purchasing is, whatever its merits, also an important regulator and brake on decentralisation. It is justified in particular by the need to ensure compliance with complex regulations on public tendering. Before 2004, local PES agencies in Germany had developed active measures in cooperation with local service providers and contacted directly with them for services without competitive tendering. The new centralized approach did lead to cost savings and increased competition by opening up existing local provider networks, but also to longer time-lags in planning and less local 14