ACTIVATION AND INTEGRATION: WORKING WITH INDIVIDUAL ACTION PLANS

The European Commission Mutual Learning Programme for Public Employment Services DG Employment, Social Affairs and Inclusion ACTIVATION AND INTEGRATION: WORKING WITH INDIVIDUAL ACTION PLANS Monitoring and follow-up of IAPs and their outcomes in selected EU countries March 2012

This publication is commissioned by the European Community Programme for Employment and Social Solidarity (2007-2013). This programme is implemented by the European Commission. It was established to financially support the implementation of the objectives of the European Union in the employment, social affairs and equal opportunities area, and thereby contribute to the achievement of the EU2020 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. For more information see: http://ec.europa.eu/progress Editor: Susanne Kraatz, DG Employment, Social Affairs and Inclusion, Unit C.3 Skills, Mobility and Employment Services, European Commission Author: Regina Konle-Seidl, IAB In collaboration with GHK and the Budapest Institute Please cite this publication as: European Commission (2012), Monitoring and follow-up of IAPs and their outcomes in selected EU countries, Brussels, Author: Regina Konle-Seidl. The information contained in this publication does not necessarily reflect the position or opinion of the European Commission

CONTENTS EXECUTIVE SUMMARY... I 1 INTRODUCTION... 1 2 WHAT PRACTICE CURRENTLY EXISTS AROUND IAPS?... 2 2.1 Profiling practices differ in terms of timing and segmentation... 2 2.2 Sequencing and intensity of follow up also varies and affect caseloads... 4 2.3 Sanctioning practice is linked to definitions such as reasonable job offer and varies in strictness and method of enforcement... 7 3 CORE IAP ELEMENTS INFLUENCING THE RE- INTEGRATION SUCCESS... 9 3.1 Benefits might outweigh costs of early intervention... 9 3.2 Influential role of case workers... 11 3.3 Monitoring and sanctions: positive effects on job finding vs. negative effects on job stability... 12 4 IMPLICATIONS FOR PES DELIVERY... 13 ANNEX 1: CONTENT ANALYSIS OF INDIVIDUAL ACTION PLANS... 15 ANNEX 2: REFERENCES... 16

EXECUTIVE SUMMARY Individual action plans (IAPs) are an instrument used by employment services to activate jobseekers. They are used by case workers to define the individual goal of the counselling and integration process and to control the agreed activities. Almost all EU countries have established some sort of IAP. We draw on current practices and processes around IAPs like timing and frequency of intensive interviews as well as follow-up in Austria, Denmark, Germany, Switzerland and the UK. The country review shows that practices around IAPs are getting more and more aligned. The early set up of IAPs is a common feature in the surveyed countries as well as the use of regular follow up interviews to monitor job search efforts indicating a work-first approach. Drawing on recent evaluation results we identified several IAP elements exerting a significant influence on the reintegration success. There is empirical evidence that lower caseloads, an increase in the frequency of contacts or a consequent monitoring and sanction regime have positive effects on job finding rates. However, results depend very much on client profiles and their distance to the labour market. They also depend on the objectives of activation policies. The experience of high number of repeaters of benefit claims not able to find sustainable employment in many countries indicates that there might be a trade-off between rapid placements as opposed to placements in more sustainable employment. There is no silver bullet nor a-one-size-fits all approach with respect to core IAP elements like monitoring and sanctions. Research evidence suggests that monitoring and sanctions or an intensive contact regime are important to activate jobseekers. They have a sizeable impact on fast re-integration. But to achieve higher job stability the counselling element of meetings is probably more important than the monitoring element. Although it is not possible to identify a clear best country example research findings and country experiences induce some lessons worth being considered for the wider PES delivery. i

1 INTRODUCTION Individual action plans (IAPs) are an instrument used by employment services to activate jobseekers and implement the principle of mutual obligations. Benefits and support ( carrots ) from the employment offices are linked to the obligation of the unemployed to take active steps to find a job, enforced with benefit sanctions ( sticks ). Almost all EU countries have established some sort of IAP 1 following a practice of intensive interviews between jobseeker and employment counsellor. In most countries - but not in all - the jobseeker must sign an action plan. In most - but not in all countries - an IAP has to be signed by both parties, the benefit providing employment service and the job-seeking benefit recipient. The idea behind IAPs is the assumption that by involving jobseekers more intensely in their individual re-integration paths, the level of personal commitment and activation could be raised. On the other hand caseworkers would focus and deliver more client oriented services. They also provide a better justification for sanctions than broadly and impersonally formulated commitments do. From a practical viewpoint IAPs are a smart management method: a clear and well-defined IAP can be a useful tool for case management 2 and sanctioning. IAPs are used by caseworkers to define the individual goal of the counselling and integration process and to control the agreed activities. They usually describe the job seeker's situation, lay down certain goals the job seeker has to achieve and list commitments by the jobseeker as well as by the employment service. IAPs could be seen as a set of compulsory (monitoring and sanctions) and supporting elements (intensified counselling) making the rights and responsibilities approach more binding. Counselling implies that unemployed clients are actively supported in their job search efforts, e.g. by helping them with their applications, practicing job interviews, or helping them to use the internet for job search. In contrast, monitoring implies that caseworkers observe the client's job search efforts, e.g. by requesting proof of submitted applications or other job search activities. Qualitative research on the design of IAP, however, indicates that IAPs are rarely based on preferences of benefit recipients. Instead, IAPs concentrate more on the duties of jobseekers to raise their search efforts. In many countries the IAP is 1 Their names vary across countries: integration agreement, guidance plan, jobseekers agreement, individual employment access plan etc. 2 Case management means defining a personal path to employment, individualised monitoring and provision of services such as coaching or training. The terms case worker, personal adviser or placement counsellor are used alike. The term case manager is used for special trained staff to work intensively with jobseekers facing multiple re-integration barriers. 1

recorded on a standard form in the IT system. Thus, goals and content are mainly standardized and rarely adjusted to the individual case (Schütz et al. 2011a). In this sense, the agreement is predominately an instrument to monitor the willingness of a person on benefits to cooperate in finding a job. In the next section we look at current practices around IAPs like timing and frequency of intensive interviews in Austria, Denmark, Germany, Switzerland and the UK. Section three focuses on research findings of core IAP elements aimed at increasing (fast and sustainable) re-integration, such as counselling and monitoring, job search assistance, mandatory participation in ALMP and corresponding sanctions in case of non-compliance. Finally, based on the available evidence, cautious lessons for implementation will be drawn, e.g. for whom dynamic action planning and ongoing support can add value on the path to employment. 2 WHAT PRACTICE CURRENTLY EXISTS AROUND IAPS? All countries under review have established some sort of IAP but national regulations, processes and practices still vary from country to country and they are continuously being developed given changing economic, social, political and organisational environments. In most countries an IAP should be concluded with all registered unemployed including more and more groups with a greater distance to the labour market. Since work-related conditionality of inactive people relying on social benefits like disability and sickness benefits or social assistance has been enforced, the obligation to set up IAPs has been extended to more and more heterogeneous target groups. In a few countries the individual action plan is a legally binding contract that each benefit claimant and the providing employment office have to conclude. A refusal to sign an IAP may result in a sanction on benefit payments in the UK (for all benefit claimants) and in Germany (for means-tested UB II claimants). The signature is not compulsory for example in Austria or France although it is considered desirable. In the decentralized Swiss system there is no legal rule to conclude IAPs. However, most of the cantons use their own IT based IAPs. Typical requirements of individual action plans involve participating in intensive interviews, applying frequently for job vacancies, undertaking verifiable independent search efforts, accepting offers of suitable work and attending work or qualification measures (OECD 2007). 2.1 Profiling practices differ in terms of timing and segmentation In order to work out the best integration strategy, the plan should be concluded after an assessment of the individual chances and problems of the job-seeker or their 2

potential risk of long-term unemployment ( profiling ) 3. A soft profiling system is in use in Germany (and other countries like France) but not in the UK or Switzerland 4. The Austrian and Danish segmentation approach uses rather rough categories. There is a general acknowledgment that profiling provides a more systematic framework for caseworkers to personalize assistance at an early stage. Ideally, individual action plans break down abstract goals like (fast) integration into the labor market into specific goals. These may be based on a customized integration strategy, specifying which goals have to be achieved, when the jobseeker has to achieve the goals, and what happens if goals were not achieved. In Germany there is a uniform way to profile PES target groups, insured unemployment benefit recipients (UB I) and recipients relying on means-tested unemployment benefits (UB II). Based on a software-guided assessment of clients distance to the labour market customers are segmented into six different profiles linking each profile to a specific service strategy to be followed by the case worker. In principle, nature, timing and level of intervention are strongly dependent on the profile of the individual and his particular needs. Ideally, the results of the profiling exercise should be used for setting up the individual action plan at the initial face-to-face interview. However, qualitative research has demonstrated that direct links to differentiated profiling results are rather weak when phrasing goals to be achieved. (See Annex 1 for a content analysis of IAPs in Germany). In Denmark insured and uninsured people are assessed according to the same principles. A match model distinguishes broadly between unemployed individuals ready to take on a job, people able to take part in active employment measures and those temporarily on passive measures. Rights and responsibilities differ according to the match group but there is no direct relation between the match groups and activation measures. The Austrian PES applies a three zone concept: info-zone just for information, service zone to register and get basic services and the counseling zone for more intensive support where clients who are still unemployed after 3 months are referred to receive more intensive case-management services. Specific client groups deemed to need more support, like young persons or persons with a lack of qualification, are directly transferred to the counseling zone. In the UK, traditionally no profiling and early identification of individual client needs is applied to avoid deadweight costs as statistics show that 50% of JSA jobseekers find employment within the first three months of unemployment. Interventions and resources are targeted on the basis of group eligibility. For more intensive 3 For a more detailed country review on the use of profiling for client segmentation and customized action planning see: PES to PES Dialogue: Thematic Synthesis Paper "Profiling Systems for Effective Labour Market Integration, (2011) Brussels 4 Swiss cantons are free to set up such systems, but they have not developed such an approach to any significant extent. 3

interventions (like the former New Deal Programmes) a queuing strategy was applied. Just basic services were offered until specific periods of benefit claim/unemployment have elapsed depending on age and benefit type. The allocation rules to more intensive measures have changed recently. Customers identified as having severe barriers to work by their Jobcentre Plus advisors are now referred earlier to private providers for intensive treatment usually after 3 months. Intensive treatment by the Work Programme - implemented in June 2011 and delivered by contracted service providers - is a flexible and personalized system. However, the referral point to the programme is still dependent on the type of benefit claim and age. JSA claimants aged 25+ are generally referred to the Work Programme after 12 months (http://www.dwp.gov.uk/docs/the-work-programme.pdf). 2.2 Sequencing and intensity of follow up also varies and affect caseloads As part of job-search assistance and monitoring, almost all countries follow a practice of intensive obligatory interviews between the jobseeker and a case worker. In some countries (e.g. Austria) the IAP is also available via the web for the clients to make the whole process more transparent. Table 1 provides information on the targeted and actual timing of the initial interview, the frequency of follow-up interviews and their average duration for the years 2009/2010. In most of the surveyed countries IAPs are concluded at the first interview with the personal adviser. Timing and frequency of face-to-face interviews are stipulated by law (e.g. in Denmark) or by PES internal guidelines (e.g. in Austria). In Germany an IAP could be set up even before the start of an unemployment spell as the law obliges employees to notify the employment services about the end of their contract three months in advance. In Denmark, an IAP must be concluded at the latest with the start of the so-called age-dependent activation phase : for jobseekers aged 20 to 30 after 13 weeks and for those aged 30 to 60 after 9 months. Data provided by the national PESs show that the actual timing of the first intensive interview usually differ from the objectives or recommendations. The actual timing for the first interview ranges from 5 1/2 days after registration in Germany (for UB II clients) to 11 weeks in Denmark. Once an action plan has been developed, IAPs are updated regularly at each followup interview or after a scheduled time frame (6 months in Germany, 3 months in Austria and Denmark). The frequency of follow-up meetings with the case worker varies between monthly and every three months. Some countries have set up a fixed schedule of interviews, four of them in the UK (after 6, 13, 26 and 52 weeks) as well as in Denmark (every 3 months). In Germany the frequency varies by jobseeker category and operational focus. This approach seems reasonable as job-search assistance may not be very productive for job ready market clients who can search effectively without assistance as well as for complex profiles with severe barriers to employment. The interview frequency is lower for these groups than for activation clients with motivational problems. Regular intensive interviews later in the 4

unemployment spell are an important element for maintaining off-flows and client s work focus. They are often coupled with monitoring of job-search efforts and modification of individual action plans. Additional, mandatory - but IAP-independent job-search reporting procedures - exist in the UK and Denmark where jobseekers in match group 1 (job ready) have to report weekly online about active job seeking. In the UK, JSA clients have to provide evidence of job search activity in Fortnightly Job Reviews (FJRs) personally at the jobcentre. FJRs last on average just seven minutes. This kind of strict job search monitoring ( signing on ) has proven to be costeffective at least for JSA clients (Freud 2007). As information on the quality of IAPs is rare the average duration of intensive interviews may be used as a proxy. Initial interviews last on average 30 minutes in Austria, 40 minutes in the UK and 60 minutes in Germany. In contrast to earlier surveys (e.g. the OECD surveys in 1999 and 2007) caseloads across countries are now lower and more similar across countries varying mainly between 100 to 140 clients per caseworker. For special target groups like youth (aged less than 25 in Germany) they are even more intensive. The participation in labour market programmes (ALMP) is compulsory after a certain length of joblessness in Denmark and the UK. In Denmark the referral point to mandatory activation measures is stipulated by law and depends on the age of jobseekers. However, length and type of activation measures can be arranged by the adviser. In the UK, JCP advisers have less discretion as the referral to the Work Programme follows stipulated guidelines. The possibility for advisers to arrange ALMP measures at any time are given in Austria, Germany and Switzerland. Nonetheless, once referred by the case worker, ALMP participation is obligatory as well. Although there is a huge body of evidence on the treatment effects of programme participation little is known about how case workers allocate jobseekers to specific programmes, and how the allocation criteria influence outcomes. Evaluation research has demonstrated that different programmes have different impacts for different target groups. The cost-effectiveness of ALMP measures could be increased if case workers could identify the optimal strategy for each client. Sophisticated targeting systems have been developed in several countries (e.g. SAPs in Switzerland or TrEffeR in Germany). Although they have proven to be effective they are not currently in use mainly due to case worker resistance (for more details see Thematic Synthesis Paper "Profiling Systems for Effective Labour Market Integration, 2011) 5

Table 1: Timing and follow up of IAP, caseloads and ALMP participation in selected PES, 2009/2010 Austria Denmark Germany SGB II (welfare) SGB III (Insurance) UK Switzerland Timing IAP = 1 st intensive interview Target: after 3 weeks Actual: 4 weeks after registration in 90% of all cases IAP: At latest with start of activation phase First interview actually after Ø 11 weeks Target: For aged < 25:10 days For aged < 25:15 days Ø 15,4 days after registration Ø 5,5 days after registration Target: JSA*: 6 weeks after start of a claim ESA*: 8 weeks after start of a claim Target : 15 days Ø 12,6 days Frequency of interviews Ø every 37 days Target: every 3 months Actual frequency n/a Due to client profile Ø job ready and complex profiles: every 3-4 months Ø activation clients every 1-2 months Target: after 13, 26, 52 weeks Achievement rate: 86% Target: monthly Actual frequency not available (n/a) Duration of interviews Initial Follow-up Counselling zone Ø 30 min Ø 24 min n/a Recommended 60 min Due to client profile Recommended 60 min 30-45 min FJR* ø 7min Ø 40 min n/a n/a Caseload** Service zone: 257 Counselling zone: 142 n/a Aged < 25 1:79 Aged >25 1:160 142 142 100 Participation in ALMP Adviser decision Compulsory once agreed to an IAP Mandatory agedepend activation Adviser decision Compulsory once agreed to an IAP Immediate offer of work or education for U25 welfare clients Referral to Work Programme Conditional on type/age of benefit claim Adviser decision Compulsory once referred by case worker Source: Information provided by national Public Employment Services; **caseload: actual staff-client ratio; *FJR= Fortnightly Job search Review; ESA= Employment Support Allowance; JSA= Jobseeker Allowance 6

2.3 Sanctioning practice is linked to definitions such as reasonable job offer and varies in strictness and method of enforcement Sanctions are deemed to drive desired behaviour. Sanctions play a crucial role in enforcing PES interventions, such as the obligation on jobseekers to regularly attend the PES office or to participate in ALMP. Sanctioning reasons are quite similar across the selected countries. The main reasons are: Refusal of reasonable job offers Insufficient personal effort/not actively searching for a new job Refusal to participate in /quitting of ALMP Voluntary job quit Missing early registration requirements (only in Germany) Refusal to sign an IAP (in the UK and in Germany for UB II clients) Although there is much discussion about reasonableness 5 the refusal of a suitable job offer is not the most important sanctioning reason as national sanction statistics demonstrate. Other reasons are more important. Two thirds of all sanctions imposed in Germany are due to missing early registration requirements, voluntary quits, or the failure to attend job interviews. Insufficient personal effort like not attending an agreed date is also the main sanctioning reason in Austria. In Switzerland job-search requirements are particularly strong and sanction rates high. Even at initial registration for placement, Swiss jobseekers are liable to be sanctioned if they cannot present evidence of recent job-search actions. The main motives for sanctions in 2008 were insufficient personal effort (usually lack of sufficient job search), followed by voluntary quits and non-compliance with instructions, mainly job or programme refusal (OECD 2010). The vast majority of sanctions in Denmark are of short duration and are imposed because the unemployed person misses a meeting at the PES. Table 2 shows that sanctions rates (excluding voluntary quits and early registration requirements) range from about 2% in Austria and Germany to about 6% in Switzerland and up to 12% in Denmark. Sanction rates for welfare clients are generally higher due to stricter eligibility criteria (e.g. to accept any job). They are about 4% in Germany, about 10% in the UK. Sanctions are mostly staged. Length and level of benefit suspension are progressive. Removals from the benefit register can last from 2-3 days to 26 weeks. In most countries, the case worker has the authority to impose such sanctions. In the UK such cases have to be referred to a higher authority, the Decision Maker. When 5 Reasonableness is linked to a number of factors such as education/qualifications; health status; personal situation; previous salary or status; working hours and/or length of contract offered; and distance from work. There is a wide range of definitions of reasonable job offer across countries. The definition used tends also to vary by client group. 7

Danish case workers observe that an unemployed person is not fulfilling the eligibility criteria, they submit a notification to the relevant UI fund. The UI fund then evaluates the notification and decides whether a sanction should be imposed and what kind of sanction is relevant. In Switzerland, offices may differ as to whether counsellors decide themselves on benefit sanctions or delegate such matters to other specialised staff in or outside the local office. This lightens the counsellors administrative burden, but also relieves them of the requirement considered uncomfortable by many to wear two hats, that of counsellor and policeman. On the other side, separating sanctioning decisions from monitoring job search reduces the autonomy and credibility of case workers and may imply more coordination efforts. Using the variation across local employment offices Frölich et al. (2007, p. 169) found a positive impact on the re-integration success when the decision on sanctions was delegated. One reason for the positive effect of delegation might be the fact that sanctions rates are generally higher when delegated. Table 2: Sanction practices in selected countries Country Benefit claimant Sanction rate in %* Length of benefit suspension 2007 2008 2009 min max Austria UI 2.2 2.3 2.1 6 weeks 8 weeks Germany UI 1.8 2.4 2.2 1 week 12 weeks Welfare UB II 3.2 3.7 3.6 10% cut for 3 months 60% - 100% (U25) cut for 3 months Denmark UI 12 (2005) 2-3 days until person worked 300 hours within a 10- week period Switzerland UI 5.4 6.2 5.7 1 day 60 days UK UI/UA 10 (4/09-3/10) 1 week 26 weeks Source: Own calculations based on figures provided by national PES. Figures for Denmark are taken from Svarer (2011). *Sanctions imposed in relation to the average stock of unemployed claimants; Sanctions due to non-compliance with mutual-obligation requirements; Sanctions for voluntary quits or missing early registration requirements like job-to-job rules are excluded In sum, the country review shows that IAPs are a common activation instrument and practices around IAPs are getting more and more aligned. The early set up of IAPs is a common feature in the surveyed countries as well as the use of regular follow up interviews to monitor job search efforts indicating a work-first approach. This 8

approach generally has the advantage of more rapid re-employment. However, it also involves the risk that jobseekers may have to accept jobs too quickly, resulting in lower wages or more rapid return to unemployment. In this context, the objective of job stability and job match quality deserves greater attention. 3 CORE IAP ELEMENTS INFLUENCING THE RE- INTEGRATION SUCCESS Does the mere existence of an IAP improve re-employment chances? There is not enough knowledge to answer this question as research on the effectiveness of IAPs as a single treatment is just starting 6. But since signing an IAP is related to multiple treatments, i.e. different PES processes and practices as described in the chapter above, we rely on the existing empirical evidence on advantages and costs of early intervention, frequency of interventions, and sanctions to different types of client. We draw mainly on results from social experiments which have been conducted in recent years in several European countries. Many earlier studies have shown that the broad category of services and sanctions or counselling and monitoring have a sizeable impact on rates of job-finding or exit from benefits. In a meta-analysis of the existing cross-country evidence in Europe, Kluve (2006) found that programmes on services and sanctions are 40-50% more likely to report a positive impact than traditional training programs. In another metaanalysis Card et al. (2009) find a positive effect of counselling efforts on reemployment chances. But it appears that the estimated impact on re-employment as well as on employment duration differs by country, by the type of programme evaluated, and by the participants' characteristics. For example, results on monitoring effects from the Netherlands indicate that monitoring is only effective in situations where individual employment prospects are low. In contrast, Kluve (2006) reports that monitoring seems to be more effective when employment prospects are good. Others (e.g. Lalive et al. 2005) have suggested that monitoring will be more effective if it is combined with a credible threat of sanctions to incentivize job-seekers. 3.1 Benefits might outweigh costs of early intervention The question of a personalized approach at an early stage is an important one, as the issue of deadweight needs to be addressed. Do the advantages of an early 6 There is just one study by Schneider (2009) analyzing the success of job search due to IAPs. She found that in the short run individual action plans increase search effort of welfare clients (UB II recipients in Germany) but the increased efforts do not automatically translate into better employment chances. The missing effect might be due to the particular high enrolment in a workfare programme as well as to the lower labour market attachment of welfare recipients compared to unemployment insurance recipients. 9

assessment followed by targeted assistance in order to reduce the detrimental effects of long-term unemployment outweigh potential deadweight effects of early intervention? In all countries under review IAPs are set-up at an early point in the unemployment spell but the later onset of intensive case management services in Denmark and the UK is mainly a resource issue. Both countries have emphasised that very early intervention has significant deadweight costs 7. Although there is a general trend towards earlier activation only a few impact studies have tried to quantify the possible efficiency gains of early intervention based on client differentiation ( profiles ) compared to a strategy of wait and see. More intensive face-to-face interviews are time-consuming and costly. The staff/client ratios needed for the contact density and intervention frequency are part and parcel of activation strategies. Comprehensive impact studies on successful PES processes and practices in Switzerland (for insurance clients) and Germany (for welfare clients) show that a favourable caseload has indeed an impact on re-integration success 8. Experimental studies from France and Germany confirm these findings. Cost-benefit computations of a large-scale pilot project of Germany's PES suggest that the costs from hiring additional caseworkers (caseload 1:40 in the participating offices compared to 1:100 in non-participating offices) were offset by the savings from decreased benefit expenditures after a period of about ten months (Hainmüller et al. 2011). The authors found that lower caseloads resulted in a decrease in the rate and duration of local unemployment and a higher re-employment rate in the short to medium run. The improved caseload reduced the average duration of the unemployment spell by 10 days. Evidence of the cost-effectiveness of a higher contact density at least for newly unemployed workers - is also reported from Denmark. The findings from a series of social experiments based on controlled randomized trials are very instructive in several ways. The final aim of this ongoing evaluation programme commissioned by the Danish National Labour Market Board is to find out which interventions (meeting vs. programmes vs. synergies) could obtain the same impact in the cheapest possible way. The first experiment exploited a setting with increased client contacts (individual meeting every week, group meetings every second week) as well as an earlier start of mandatory full-time activation (after 13 weeks) for newly unemployed workers. In the treatment group the unemployment duration could be reduced by 3 weeks. Moreover, the new intervention regime has turned out to be cost-effective. 7 The evaluation of a segmentation exercise in the UK showed that although the predictive power of a statistical profiling model was correct in 70% of cases, the relatively short increase in average JSA claimant durations stemming from false predictions would negate any savings as 90% of claimants leave the register within a year (Driskell 2005). 8 The main results of the second evaluation wave of ALMP in Switzerland are summarized in http://www.seco.admin.ch/dokumentation/publikation/00004/00005/index.html?lang=de; The main result of the evaluation report on the German experiment clause is available for download (in English): http://www.bmas.de/shareddocs/downloads/en/pdf-publikationen/f390-final-reportexperiment-clause-pursuant-6c-sgb-iii.pdf? blob=publicationfile 10

Net benefits of about 2,000 per unemployment spell have been the result of a costbenefit analysis conducted by the Danish Economic Council. A further analysis to find out what worked for whom demonstrated that women are positively affected by meetings while men are positively affected by (the threat of) early activation. Interestingly, absolutely no effects have been found in further trials using the same intervention regime for sick-listed and the weaker, long term unemployed (Vikström et al. 2011). 3.2 Influential role of case workers All countries under review use a unique or single case worker approach where job seekers deal with the same member of PES staff at all times. Hence, personal advisers are the sole point of contact throughout the job search process. This gives the personal adviser a critical role. Counsellor qualifications and experience affects significantly the quality of the case management process. Using statistics that link jobseeker and case-worker records, Frölich et al (2007) found many case-worker factors that influence re-employment rates: case-worker specific training 9 and experience act positively, age negatively. Behncke et al. (2010b) found a positive employment effect of about 3 percentage points when the caseworker and the unemployed person are of the same gender, age, nationality and educational background. Surveys among case workers confirm that they find themselves with a discretionary power of how to work with their clients. Case worker attitudes and behaviour are influential as well the way they provide case management services and assign clients to programmes. An analysis based on a detailed linked jobseeker caseworker data set for Switzerland revealed that case workers attitudes and behaviour have a significant impact on job-finding probabilities. Behncke et al (2010a) found that tough case workers assigning unemployed persons to jobs and labour market programmes without their consent are more successful in the short and medium term than case workers who place more emphasis on a co-operative and harmonic relationship with their clients. In line with these findings are results from the above mentioned experiment in Germany. Disentangling the causal pathways for the positive re-employment effects of lower caseloads, Hainmüller et al. (2011) found that successful offices became more proactive. They increased the monitoring and imposed more sanctions on clients with low search efforts. On the other hand, successful offices also registered more new vacancies implying that caseworkers with better contacts to local firms are more successful in providing placement services. 9 In the Swiss case this was a specific certificate: Eidgenössische Fachausweis mit Spezialisierung Personalberatung. In contrast, case workers with a university degree yielded less favourable results. 11

Consistent with the idea that case workers exert a major influence, Lagerström (2011) finds that caseworkers in Swedish employment offices vary dramatically in their effectiveness at bringing their clients back into regular employment. However, the author did not observe case workers attitudes but rather looked on different working strategies, especially on practices of allocating jobseekers to different types of programmes. Case workers who assign their clients to classroom training tend to be less successful than those who concentrate on supporting their clients in the actual process of searching for work. But as these findings are in line with previous evaluation results on programme effects, a casual interpretation is rather difficult. 3.3 Monitoring and sanctions: positive effects on job finding vs. negative effects on job stability The dominance of work first activation strategies and the experience of high number of repeaters of benefit claims not able to find sustainable employment is a major policy concern in many countries (for country experiences see Eichhorst et al. 2008). It seems that there is a trade-off between rapid placements as opposed to placements in more sustainable employment. Within a work first framework underpinned by legal regulations (i.e. to accept any job) monitoring and sanctions exert threat which might support the observed revolving door effect. The international body of evidence on sanctions, their impacts and context show that sanctions increase the probability to find regular employment. Most studies both for insurance as well as welfare clients find that benefit sanctions speed up job finding and have a positive re-employment effect regardless of the size of the penalty or the duration of the punishment. Analyzing Swiss data Lalive et al. (2005) conclude that by imposing a benefit sanction the job finding rate goes up by 45% causing the unemployment duration to decrease by roughly three weeks from 33 to 30 weeks. Svarer (2010) using Danish data finds that the unemployment exit rate increases by more than 50% following enforcement of a sanction. However, more recent evidence from Switzerland and Sweden reveals that sanctions negatively affect the quality of jobs the unemployed obtain under the threat of a benefit cut. The study of Arni et al. (2009) confirms previous findings: the threat and the use of sanctions increase job-finding rates, as well as exit rates from the labour force. On the other hand, however, the enforcement of benefit cuts lowers the quality of post-unemployment jobs, both in terms of job-duration and in terms of earnings. Instructive are also the effects on job stability of the above mentioned social experiments in Denmark. Investigating the long-term effects of the first experiment on the newly unemployed, Blasco and Rosholm (2011) found that the new contact regime significantly reduces unemployment recurrence for men, but not for women. In contrast to earlier (suggestive) findings on threat effects driving the speed up of job finding, the majority (80%) of the impact on job stability is attributed to a more efficient matching process suggesting that the counselling element of meetings is more important than the monitoring element. 12

4 IMPLICATIONS FOR PES DELIVERY How can these findings inform wider PES delivery? Although it is not possible to identify a clear best country example, research findings and country experiences induce some lessons worth being considered. Profiling and appropriate assessment tools can help to target support better and thus reduce deadweight costs. It would not be efficient to spend money on activation schemes for all jobseekers. But due to increasingly heterogeneous client groups it probably has a high cost if individuals with significant barriers to work are left without support for up to 12 months before being given effective employment support. While the average interval between intensive interviews may be important, fixed interview schedules are not necessarily preferable. Thus, the interview intensity and frequency need not be identical for all types of jobseekers, in particular where profiling has identified those groups that are likely to engage in successful job search on their own. Staff/client ratios needed for an intensive intervention regime are found to have a causal impact on shortening unemployment spells. Lower caseloads are cost-effective when higher expenditures are offset by benefit savings. There is no silver bullet nor a-one-size-fits all approach with respect to core IAP elements like monitoring and sanctions. Research evidence suggests that monitoring and sanctions are important to activate jobseekers. They have a sizeable impact on fast re-integration. But to achieve higher job stability the counselling element is probably more important. This implies that IAP practices based on mutual obligation have to be complemented by high quality counselling and professional guidance. In setting up IAPs case workers have to cope with the double role of supporting and empowering as well as demanding and sanctioning ( policing ). A way to relieve this tension is to delegate the final decision on sanctions to other specialised staff in or outside the local office. What works depends very much on the profile of the job seeker. Individuals profiled as hard to-place usually have more difficulties regarding their abilities as well as their motivation in returning to work. While more intensive meetings with the case worker and/or an earlier start of mandatory activation measures have proven to be effective for activation or counselling clients they are less effective for individuals with complex profiles (long-term unemployed, labour market reentrants like disabled or sick listed). Eventually, they may benefit more from intensified case management services and less from compulsory instruments. Another implication is that there might be some potential benefits to lifting strict monitoring and sanction regimes for those who have only slight chances of finding regular jobs on their own. But it is important to have a clear picture on who are the vulnerable with no or low capacity of changing their behaviour. 13

Too rigid requirements, such as too many pro forma job applications or pressure on jobseekers to accept quick job matches that do not fit their profiles should be avoided. This is also advisable with regard to employers. Sending inappropriate candidates to employers for interviews is rather harmful. In contrast, a careful selection of candidates helps to improve the direct interaction with employers being a pivotal factor for successful job placements. Although there are just a few consolidated findings from job retention programmes to reduce unemployment recurrence, the follow-up after a placement could be a promising approach. In most countries there is still a lack of such services although ongoing in-work support may be crucial in increasing the chances of individuals staying in employment. Job stability is going to be a key for performance reviews and remuneration for contracted service providers. Linking organisational performance targets as well as personal rewards closer to job sustainability outcomes could help to push after- placement services in the public sector, too. What works depends in the end also on the local/regional market situation. However, it is unclear to what extent the same effects, e.g. lower caseloads, could be expected if unemployment is high and there are few jobs available. But even in economic downturns there is a lot of movement in regional labour markets. Consequently, well established counselling and monitoring processes are important even in times of crisis. 14

ANNEX 1: CONTENT ANALYSIS OF INDIVIDUAL ACTION PLANS For insurance clients (SGB III) and welfare clients (SGB II) in Germany SGB III (aged > 25) SGB II (aged > 25) inflow total inflow Sample: N of IAPs 163 119 62 Formulated goal No goal formulated 4 6 3 Integration into (regular) employment/ ending or shortening the unemployment spell 143 83 47 Other goals 16 30 12 Obligations of the agency Job search support/support to find an apprenticeship place 122 10 6 To take over job search costs 92 83 47 To publish the applicant's profile on the internet 118 51 26 Referrals to job offers 138 87 47 Referral to at least one activation measure 81 83 38 Obligations of the client Preparation / Revision of written application documents 23 28 19 Applications 155 89 48 thereof: with proof of applications 105 71 42 thereof: concrete N of monthly applications 57 67 35 Ø monthly applications 7.1 5.2 to 5.4 Participation in at least one activation measure 45 54 22 Source: Schütz et al (2011b), Table 17, p 137 15

ANNEX 2: REFERENCES Arni P., Lalive, R. and J. van Ours (2009): How Effective Are Unemployment Benefit Sanctions? Looking Beyond Unemployment Exit. IZA Discussion Paper 4509. Behncke, S., Frölich, M. and M. Lechner (2010a): Unemployed and their caseworkers: Should they be friends or foes? Journal of the Royal Statistical Society: 173: 67-92 (26). Behncke, S., Frölich, M. and M. Lechner (2010b): A caseworker like me Does the similarity between the unemployed and their caseworkers increase job placements? The Economic Journal, 120 (December), 1430 1459 Blasco, S. and M. Rosholm (2011): The Impact of Active Labour Market Policy on Post-Unemployment Outcomes: Evidence from a Social Experiment in Denmark; IZA Discussion Paper No. 5631, April 2011 Card, D., Kluve, J. and A. Weber (2009): Active labor market policy evaluations: A Meta-Analysis. IZA Discussion Paper No 4002 Driskell, W. (2005): Customer Segmentation: Statistical profiling using the Work and Pensions Longitudinal Study. DWP Policy Working Paper Freud, D. (2007): Reducing dependency, increasing opportunity: Options for the future of welfare to work. An independent report to the Department for Work and Pensions Frölich, M., Lechner, M. und S. Behncke (2007): Einfluss der RAV auf die Wiedereingliederung der Stellensuchenden. SECO Publikationen Arbeitsmarktpolitik N 20 (1.2007) Hainmueller, J., Hofman, B., Krug, G. and K. Wolf (2011): Do Lower Caseloads Improve the Effectiveness of Active Labor Market Policies? New Evidence from German Employment Offices. Laser Discussion Papers Paper No. 52, September 2011 Kluve, J. (2006): The effectiveness of European active labor market policy. Labour Economics 17: 904.918. Eichhorst, W. Kaufmann, O. and R. Konle-Seidl (2008): Experiences with Activation Schemes in Europe and the US. Springer Verlag Berlin, Heidelberg Lalive, R., van Ours, J. and J. Zweimüller (2005): The effect of benefit sanctions on the duration of unemployment. Journal of the European Economic Association, Vol 3, Issue 6, 1386-1417, December 2005 Lagerström, J. (2011). How important are caseworkers - and why? New evidence from Swedish employment offices, IFAU Working Paper Series 2011:10 OECD Employment Outlook (2007): Chapter 5, Activating the unemployed: What countries do, Paris OECD (2010): Activation policies in Switzerland: OECD Social Employment and Migration Working Papers No. 112, 21 October 2010 16

Schneider, J. (2010): Activation of welfare recipients. Impacts of selected policies on reservation wages, search effort, re-employment and health. Dissertation Free University of Berlin 2010 Schütz, H., Kupka, P., Koch, S. und B. Kaltenborn (2011a): Eingliederungsvereinbarungen in der Praxis. Reformziel noch nicht erreicht. IAB-Kurzbericht 18/2011 Schütz, H., Steinwede, J. Schröder, H. Kaltenborn, B:, Wielage, N., Christe, G. und P. Kupka (2011b): Vermittlung und Beratung in der Praxis. Eine Analyse von Dienstleistungsprozessen am Arbeitsmarkt. IAB-Bibilothek 330 Svarer, M. (2011): The Effect of Sanctions on Exit from Unemployment: Evidence from Denmark. Economica, Vo 78, issue 312, 751-778, October 2011 Vikström, J., Rosholm M. and M. Svarer (2011): The Relative Efficiency of Active Labour Market Policies: Evidence From a Social Experiment and Non-Parametric Methods, WP 11-04, Aarhus University, March 2011 17