EUROPEAN COMMISSION

Strasbourg, 13.12.2016

SWD(2016) 460 final

COMMISSION STAFF WORKING DOCUMENT

IMPACT ASSESSMENT

Initiative to partially revise Regulation (EC) No 883/2004 of the European Parliament and of the Council on the coordination of social security systems and its implementing Regulation (EC) No 987/2009

Accompanying the document

Proposal for a
Regulation of the European Parliament and of the Council

amending Regulation (EC) No 883/2004 on the coordination of social security systems and regulation (EC) No 987/2009 laying down the procedure for implementing Regulation (EC) No 883/2004

(text with relevance for the EEA and Switzerland)

{COM(2016) 815 final}
{SWD(2016) 461 final}

Table of Contents

1. ANNEXE IV: ANALYTICAL MODELS USED………………………….........................3

2.annex v: Impact assessment study lONG tERm cARE AND

uNEMPLOYMENT BENEFITS…………….17

ANNEXE IV – Analytical models used in preparing the impact assessment

Several studies, using different analytical models and methodologies, have been used to prepare the impact assessments 1 . When reliable quantitative information on the totality of impacts of the proposed initiative was not available, the analysis has been mainly based on a qualitative assessment of cross-checked criteria. Since options on the coordination of long-term care benefits, coordination of unemployment benefits for frontier workers and export of unemployment benefits had been assessed in 2013-2014, an update with more recent and newly available data has been conducted in 2015 2 .

A general point in terms of data limitations is that some statistical treatment are based on citizenship (Labour force survey) and therefore identify EU mobile citizens/workers (those living/working in another country than their country of citizenship) – while other data (administrative data collection) are based on headcounts of case where citizenship is not collected and that therefore constitutes a broader definition of mobility , i.e. includes not only EU mobile citizens/workers but also nationals returning to their country of citizenship as well as third-country nationals moving between EU Member States. This means that, at least for the treatments based on the administrative data collection, the Impact assessment adopts a broad definition of mobility and does take into account that not only EU mobile citizens benefitting from coordination but also other groups, which also means that there is therefore no underestimation of the phenomenon when using those data.

Hereunder a detailed list of the analyticial models and methodologies which have been used by the different studies.

1.Coordination of Long-Term Care benefits

HIVA

The fact there is no specific coordination regime and a common definition, made it difficult to collect data on LTC. Member States did not explicit collect data on LTC and had no common understanding of LTC benefits. Administrative data on LTC are only available in specific forms dealing with the coordination rules of the sickness chapter. The number of those insured for health care living in another country than the competent country – which sometimes includes long-term care or to which LTC-insurance is closely linked – can be calculated based on the number of PD S1. However, no data are currently collected on the number of PD S1 within the framework of the Administrative Commission. The number of PD S1 was estimated by the sum of 3 categories:

• Cross-border workers (and their family members);

• Retired former cross-border workers (and their family members);

• Other mobile pensioners (and their family members).

Firstly, by way of using the LFS, the number of cross-border workers were estimated. Secondly, we assumed in the calculation model that 20% of the cross-border workers will have an insured family member. Thirdly, to estimate the total number of retired former cross-border workers, we applied the percentage of cross-border workers on the labour market to the number of pensioners in 2013 and this by individual (former) working Member State and assumed that 1 in 5 retired cross-border workers always had worked in the same Member State of employment. Fourthly, an estimation of the number of migrant pensioners was calculated by using the LFS. Finally, we assumed in the calculation model that 25% of the pensioners will have also an insured family member. The sum of all these categories results in an estimate of the number of PD S1. As next step we have estimated the cross-border expenditure on long-term care in kind and in cash based on figures from the 2015 Ageing Report. Here we have applied the average LTC benefits in cash and in kind per dependent user. It is as mobile citizens (workers, pensioners, their family members) are using this system of LTC as if they were nationals. This involves a ‘potential’ overestimation of the number of users of cross-border LTC benefits and the related expenditure due to fact some Member States consider their LTC benefit as not exportable. At the same time these estimates assume a complete ‘take-up’ of rights by mobile citizens which will not be the case in the baseline scenario.

Deloitte

The data collection and analysis are focused on a representative sample of 14 Member States 3 . The sample of Member States covers seven of the eight welfare state models 4 (given their similarities, the “new” Bismarck-oriented Mediterranean Member States –Malta and Cyprus- not included). The selection of the countries is primarily based on relevant mobility patterns and a balanced coverage in terms of types of the social security systems, more notably in the area of long-term care benefits.

Two types of data sources were used during this study:    secondary data (available literature and reports at EU and MS level, particularly the trESS reports; replies to the online public EC Consultation on the need to revise of the current rules; available statistical data with regard to mobility patterns and the use of LTC benefits in cross-border cases) and primary data, collected through interviews and a consultation of the stakeholders (findings from strategic interviews with DG EMPL officials; findings from interviews with stakeholders at EU level, e.g. European umbrella organisations; findings form interviews with key stakeholders at national level (health insurers, healthcare providers); replies to the EU-wide web-based survey among responsible public authorities; new, generated statistical data with regard to mobility patterns and the use of LTC benefits in cross-border cases; findings from the 13 workshops/group interviews and 8 phone interviews on the administrative costs and administrative burden related to the policy options).

2.Unemployment benefits: coordination of unemployment benefits for frontier workers

HIVA

Based on Labour force Survey (LFS) data for 2013 and 2014, an estimation of the number of cross-border workers has been made. In the further analysis we considered all workers who worked in another country than the country of residence as cross-border workers. Workers who worked in a neighbouring country are considered as frontier workers. This is different from the legal definition. National unemployment rates from Eurostat were applied to the number of cross-border workers in order to estimate the number of unemployed cross-border workers.  The unemployment rates of the country of last activity and not of the country of residence have been applied on the number of cross-border workers. In order to estimate the budgetary impact of the baseline scenario, the estimated number of unemployed cross-border workers are multiplied by the annual unemployment benefit per unemployed by taking into account the annual average duration of the payment of the unemployment benefit (on the basis of ESSPROS, Eurostat figures and the LFS). 

Deloitte

The data collection and analysis are focused on a representative sample of 14 Member States 5 . The sample of Member States covers seven of the eight welfare state models 6 (given their similarities, the “new” Bismarck-oriented Mediterranean Member States –Malta and Cyprus- not included). The selection of the countries is primarily based on relevant mobility patterns and a balanced coverage in terms of types of the social security systems, more notably in the area of unemployment benefits.

Two types of data sources were used during this study: secondary data (available literature and reports at EU and MS level, particularly the trESS reports; replies to the online public EC Consultation on the need to revise of the current rules; available statistical data with regard to mobility patterns and the use of unemployment benefits in cross-border cases) and primary data, collected through interviews and a consultation of the stakeholders (findings from strategic interviews with DG EMPL officials; findings from interviews with stakeholders at EU level, e.g. European umbrella organisations; findings form interviews with key stakeholders at national level (public employment services); replies to the EU-wide web-based survey among responsible public authorities; new, generated statistical data with regard to mobility patterns and the use of unemployment benefits in cross-border cases; findings from the 13 workshops/group interviews and 8 phone interviews on the administrative costs and administrative burden related to the policy options).

3.Unemployment benefits: export of unemployment benefits

HIVA

Figures for all EU-Member States on the export of unemployment benefits have become available via the administrative PD U2 Questionnaire launched within the framework of the Administrative Commission (for 2013). Additional data available for Belgium has been used to describe the impact of the prolongation period on finding a job abroad.  Finally, figures of Eurostat (based on the LFS) were used to calculate the average duration of the unemployment period.

4.Unemployment benefits: aggregation of unemployment benefits

HIVA (financial impact for Member States)

Based on the data from the administrative questionnaire on the aggregation of periods for unemployment the budgetary impact of the current rules and the different alternative options could be calculated. Member States had to provide a breakdown by Member State of origin and a breakdown by length of insurance. The reported cases have been multiplied by the annual average expenditure per unemployed person (also by taking into account the annual average duration of the payment of the unemployment benefit) in order to estimate the public unemployment spending. Option 4 (change of the calculation method) required more detailed information about the unemployed recent migrant worker’s salary. No information on the salary earned in the competent Member State as well as in the Member State of origin was collected via the administrative questionnaire.  Therefore, wage data published by Eurostat have been used.

Brodolini (costs for public administrations)

The methodology for assessing the administrative costs and burden for the national administrations in the Member States handling cases of aggregation of unemployment benefits takes outset in the definition of administrative costs and burden provided by the EC Impact Assessment Guidelines. While doing this, we have adopted a broad definition of administrative information obligations – i.e. we have considered the costs of administrative tasks such as the determination of Member State of competence, the calculation of benefits, and the reimbursement and recovery of benefits in between Member States. In addition, we have looked into the administrative burden for the mobile EU-workers.

The calculations of the administrative costs for the national administrations are as illustrated in the below figure in principle simple. Firstly, the number of cases in the left-hand side of the figure are the annual cases currently registered in six case study Members States (Germany, Denmark, the Netherlands, Poland, Romania, and the UK) – information that has been provided by the HIVA KU Leuven Research Institute for Work and Society via a data gathering exercise. In our main calculations, we assume that the number of cases does not change as a result of revisions to Regulation (EC) No 883/2004. However, we do to illustrate the sensitivity of this assumption also show scenarios where we assume a change in the number of cases based on the results of Task 3 and 4, respectively, of this study.

Secondly, we calculate the unit costs for the national administration – i.e. the current administrative costs per case as well as the changed administrative costs per case as a result of the EU revisions. Such administrative costs per case clearly differ between cases, i.e. there will be easy cases only involving standard administrative tasks while other more complex cases will involve additional tasks. Hence, we estimate in practice average costs.

Calculation of administrative costs for national administrations

Most study resources were allocated to the estimation of the man-hours needed to carry out the administrative tasks. Such information is not available from official sources, and so we have gathered the information through interviews with national administrations in the six case study Member States. For this purpose, we identified the national administrations to interview, and we developed interview guides that was applied by our Member States experts/interviewers in the six case study Member States.

It must in this context be acknowledged that most of the man-hour estimates are based on a few interviews only. This is partly because there is often only a few national administrations that handle cases in each of the Member States, and partly because of limited interview resources. Hence, the premise for this analysis and so its results is the uncertainty associated with few assessments that, furthermore, mostly are based on the subjective assessments of the interviewees rather than on actual registrations of time spent on different administrative tasks.

Brodolini (secondary impacts)

The simulations used in the study aim at translating the implications of the proposed revisions into expected income changes. Assuming a connection between income differentials and the propensity to move among EU countries, a variation in mobility flows and in the dimension of the target population is estimated. Finally, taking into account the variation in the latter, the different level of benefits and the redistribution of competences between sending and receiving countries envisioned by the policy options, we give an estimate of the variation in expenditure for the aggregation of periods or salaries for unemployment benefits for the selected countries.

The variation both in mobility and in expenditure, are estimated in relation to the status quo scenario. This scenario is built on a virtual ‘baseline year’, with a defined level of mobility flows and a defined level of expenditure. The mobility flow is computed as the averages of the most recent values of annual bilateral mobility flows among the selected countries (7 MS for task 3 and 8 MS for task 4). The number of potential cases of aggregation of periods or salaries for unemployment benefits (the ‘target population’ of task 4) is computed by multiplying the status quo mobility flow registered from a given sending country to a given receiving country with the ratio of unemployed non-national EU28+EFTA individuals on the total number of non-national EU28+EFTA individuals in the given receiving country. Then, the aggregate expenditure is estimated by multiply the target population identified for the average values of the benefits in each selected MS.

It is assumed that the income differences observed in the status quo scenario – along with all other factors that the policy options do not have an impact on – generate the mobility observed in the status quo period both in terms of general mobility flows and in terms of dimension of target population. The influence of unemployment benefits on mobility is evaluated by calculating the change attributable to the various policy options on the amounts unemployment benefits paid in relation to the family income in the country of origin or in the country of previous residence.

Once the income change is calculated, we convert that change into a potential mobility effect, using a theoretical assumptions on the elasticity of migration flows to income changes, which relates elasticity to income differentials for each bilateral combination of countries. As a measure for earnings, we consider the net earnings (PPS-adjusted) of a single individuals without children at 100% of the average wage. Once elasticity is computed, the percentage of income change is translated into a mobility percentage change. The percentage of mobility change is then translated into a change in mobility flows and in potential target population by relating the calculated percentage change to the absolute number of mobile citizens in the status quo case, according to the policy options proposed.

In order to define the expenditure variations, we compute the present level of expenditure by multiplying the number of present potential cases of aggregation of periods or salaries for unemployment benefits by the average amount of unemployment benefits. We then observe the percentage variation of expenditure – total and related to each bilateral relation.

FreSsco (legal analysis)

FreSsco analytical impact assessment reports are based on the description of a specific problem of social security coordination and/or free movement of workers which arises in the application of EU law and which should be subjected to a legal analysis.

Immediately after receiving and confirming the mandate, thorough consideration is given to the selection of the team of experts entrusted with the analysis of the specific legal problem, one of whom is selected as the team coordinator. It is observed that their expertise is to the highest possible extent tailored to the subject matter of the analysis. In addition, two meetings between the experts are organised. One is arranged at the beginning of the task, in order to get acquainted with the task and allocate the responsibilities, and another one before the final results are presented, in order to come to an agreement with the final text. If the deadline, set by the mandate, is too pressing, only one meeting is organised. Nevertheless, in both cases, i.e. one or two meetings, intensive communication between the experts is guaranteed and facilitated during the analysis, mainly by the internal coordinator and the FreSsco management. Only by applying the dialectical method of testing various options, and confirming or rejecting ideas via discourse between the experts, can the best solutions be found. It goes without saying that better and faster results might be achieved when discussions occur inter preasentes and not inter absentes.

Regarding the report itself, first the existing application of EU law to a specific cross-border situation is presented. To this end, the existing legislative and non-legislative documents as well as decisions of the Court of Justice of the European Union (CJEU) and other bodies (like the Administrative Commission for the Coordination of Social Security Systems) are taken into account. The historical method might be applied in order to discover the reasons for the present legislative solutions.

In the next step of the analysis several options – either enumerated in the mandate and/or discovered during the analysis – which might present a solution to a given problem are tested. Their strengths, weaknesses, opportunities and threats are stressed. In some of the reports, for each of the possible solutions specific criteria are taken into account, such as clarity, simplicity, protection of rights, administrative burden and implementation arrangements, risk of fraud and abuse and potential financial implications.

To this end, not only the relevant EU law is taken into consideration, but other sources as well. These might include relevant literature and academic studies, reports and selected key policy documents of the EU and possibly of other international organisations.

Next to this comparison on an EU (and international) level, a vertical comparative method is applied in order to present the problems and test the solutions in (some or all, depending on the mandate) Member States. A horizontal comparative method at the national level is enabled by mapping the situation in several (or all) Member States.

Logical, grammatical and teleological methods of legal interpretation are relied upon as well. Conclusions are as a rule drawn from the descriptive-analytical method of the research, which might prove useful for selecting the best policy options at national and EU level and for finding the most appropriate normative solutions de lege ferenda.

5.Export of family benefits

HIVA (financial impact for Member States)

Data of the administrative questionnaire has been used to report figures on the baseline scenario. Member States provided data on their export of family benefits and provided sometimes even more detailed data by the primarily or secondarily competence of the reporting Member State. By making use of the price level indices for consumer goods and services reported by EUROSTAT a correction coefficient between the exporting Member State and the Member State of residence of the child(ren) could be calculated in order to make an adjustment of the amount of exported family benefits to the cost of living in the Member State of residence of the child(ren). The scope of the administrative questionnaire was limited to the number of households and children who received a child benefit from a competent exporting Member State. However, for the calculation of one option (Option 3: Reverse order of competence) the complete reference group affected by the coordination of family benefits had to be taken into account. Also, more information on the average amount of the family benefit on the basis of ESSPROS was required as not all Member States had answered the administrative questionnaire.

Brodolini (costs for public administrations)

The methodology for assessing the administrative costs and burden for the national administrations in the Member States handling cases of export of family benefits takes outset in the definition of administrative costs and burden provided by the EC Impact Assessment Guidelines. While doing this, we have adopted a broad definition of administrative information obligations – i.e. we have considered the costs of administrative tasks such as the determination of Member State of competence, the calculation of benefits, and the reimbursement and recovery of benefits in between Member States. In addition, we have looked into the administrative burden for the mobile EU-citizens and their families.

The calculations of the administrative costs for the national administrations are as illustrated in the below figure in principle simple. Firstly, the number of cases in the left-hand side of the figure are the annual cases currently registered in six case study Members States (Germany, Denmark, the Netherlands, Poland, Romania, and the UK) – information that has been provided by the HIVA KU Leuven Research Institute for Work and Society via a data gathering exercise. In our main calculations, we assume that the number of cases does not change as a result of revisions to Regulation (EC) No 883/2004. However, we do to illustrate the sensitivity of this assumption also show scenarios where we assume a change in the number of cases based on the results of Task 3 and 4, respectively, of this study.

Secondly, we calculate the unit costs for the national administration – i.e. the current administrative costs per case as well as the changed administrative costs per case as a result of the EU revisions. Such administrative costs per case clearly differ between cases, i.e. there will be easy cases only involving standard administrative tasks while other more complex cases will involve additional tasks. Hence, we estimate in practice average costs.

Calculation of administrative costs for national administrations

Most study resources were allocated to the estimation of the man-hours needed to carry out the administrative tasks. Such information is not available from official sources, and so we have gathered the information through interviews with national administrations in the six case study Member States. For this purpose, we identified the national administrations to interview, and we developed interview guides that was applied by our Member States experts/interviewers in the six case study Member States.

It must in this context be acknowledged that most of the man-hour estimates are based on a few interviews only. This is partly because there is often only a few national administrations that handle cases in each of the Member States, and partly because of limited interview resources. Hence, the premise for this analysis and so its results is the uncertainty associated with few assessments that, furthermore, mostly are based on the subjective assessments of the interviewees rather than on actual registrations of time spent on different administrative tasks.

Brodolini (secondary impacts)

The simulations used in the study aim at translating the implications of the proposed revisions into expected income changes. Assuming a connection between income differentials and the propensity to move among EU countries, a variation in mobility flows and in the dimension of the target population is estimated. Finally, taking into account the variation in the latter, the different level of benefits and the redistribution of competences between sending and receiving countries envisioned by the policy options, we give an estimate of the variation in expenditure for the export of family benefits for the selected countries.

The variation both in mobility and in expenditure, are estimated in relation to the status quo scenario. This scenario is built on a virtual ‘baseline year’, with a defined level of mobility flows and a defined level of expenditure. The mobility flow is computed as the averages of the most recent values of annual bilateral mobility flows among the selected countries (7 MS for task 3 and 8 MS for task 4). In the definition of the tartget population interested by the proposed revisions we made use of the HIVA study (Pacolet et al., 2015), which allows to identify, for each selected country, the total number of persons entitled to the export of child benefits. Then, the aggregate expenditure is estimated by multiply the target population identified for the average values of the benefits in each selected MS.

It is assumed that the income differences observed in the status quo scenario – along with all other factors that the policy options do not have an impact on – generate the mobility observed in the status quo period both in terms of general mobility flows and in terms of dimension of target population. The influence of family benefits on mobility is evaluated by calculating the change attributable to the various policy options on the amounts of family benefits paid in relation to the family income in the country of origin or in the country of previous residence.

Once the income change is calculated, we convert that change into a potential mobility effect, using a theoretical assumptions on the elasticity of migration flows to income changes, which relates elasticity to income differentials for each bilateral combination of countries. As a measure for earnings, we consider the net earnings (PPS-adjusted) of a one-earner married couple with two children at 100% of average wage. Once elasticity is computed, the percentage of income change is translated into a mobility percentage change. The percentage of mobility change is then translated into a change in mobility flows and in potential target population by relating the calculated percentage change to the absolute number of mobile citizens in the status quo case, according to the policy options proposed.

In order to define the expenditure variations, we compute the present level of expenditure by multiplying the number of present potential cases. We then observe the percentage variation of expenditure – total and related to each bilateral relation.

FreSsco (legal analysis)

FreSsco analytical impact assessment reports are based on the description of a specific problem of social security coordination and/or free movement of workers which arises in the application of EU law and which should be subjected to a legal analysis.

Immediately after receiving and confirming the mandate, thorough consideration is given to the selection of the team of experts entrusted with the analysis of the specific legal problem, one of whom is selected as the team coordinator. It is observed that their expertise is to the highest possible extent tailored to the subject matter of the analysis. In addition, two meetings between the experts are organised. One is arranged at the beginning of the task, in order to get acquainted with the task and allocate the responsibilities, and another one before the final results are presented, in order to come to an agreement with the final text. If the deadline, set by the mandate, is too pressing, only one meeting is organised. Nevertheless, in both cases, i.e. one or two meetings, intensive communication between the experts is guaranteed and facilitated during the analysis, mainly by the internal coordinator and the FreSsco management. Only by applying the dialectical method of testing various options, and confirming or rejecting ideas via discourse between the experts, can the best solutions be found. It goes without saying that better and faster results might be achieved when discussions occur inter preasentes and not inter absentes.

Regarding the report itself, first the existing application of EU law to a specific cross-border situation is presented. To this end, the existing legislative and non-legislative documents as well as decisions of the Court of Justice of the European Union (CJEU) and other bodies (like the Administrative Commission for the Coordination of Social Security Systems) are taken into account. The historical method might be applied in order to discover the reasons for the present legislative solutions.

In the next step of the analysis several options – either enumerated in the mandate and/or discovered during the analysis – which might present a solution to a given problem are tested. Their strengths, weaknesses, opportunities and threats are stressed. In some of the reports, for each of the possible solutions specific criteria are taken into account, such as clarity, simplicity, protection of rights, administrative burden and implementation arrangements, risk of fraud and abuse and potential financial implications.

To this end, not only the relevant EU law is taken into consideration, but other sources as well. These might include relevant literature and academic studies, reports and selected key policy documents of the EU and possibly of other international organisations.

Next to this comparison on an EU (and international) level, a vertical comparative method is applied in order to present the problems and test the solutions in (some or all, depending on the mandate) Member States. A horizontal comparative method at the national level is enabled by mapping the situation in several (or all) Member States.

Logical, grammatical and teleological methods of legal interpretation are relied upon as well. Conclusions are as a rule drawn from the descriptive-analytical method of the research, which might prove useful for selecting the best policy options at national and EU level and for finding the most appropriate normative solutions de lege ferenda.

6.Access for economically inactive mobile EU citizens and jobseekers to certain social benefits

FreSsco (legal analysis focussed on access to SNCBs)

FreSsco analytical impact assessment reports are based on the description of a specific problem of social security coordination and/or free movement of workers which arises in the application of EU law and which should be subjected to a legal analysis.

Immediately after receiving and confirming the mandate, thorough consideration is given to the selection of the team of experts entrusted with the analysis of the specific legal problem, one of whom is selected as the team coordinator. It is observed that their expertise is to the highest possible extent tailored to the subject matter of the analysis. In addition, two meetings between the experts are organised. One is arranged at the beginning of the task, in order to get acquainted with the task and allocate the responsibilities, and another one before the final results are presented, in order to come to an agreement with the final text. If the deadline, set by the mandate, is too pressing, only one meeting is organised. Nevertheless, in both cases, i.e. one or two meetings, intensive communication between the experts is guaranteed and facilitated during the analysis, mainly by the internal coordinator and the FreSsco management. Only by applying the dialectical method of testing various options, and confirming or rejecting ideas via discourse between the experts, can the best solutions be found. It goes without saying that better and faster results might be achieved when discussions occur inter preasentes and not inter absentes.

Regarding the report itself, first the existing application of EU law to a specific cross-border situation is presented. To this end, the existing legislative and non-legislative documents as well as decisions of the Court of Justice of the European Union (CJEU) and other bodies (like the Administrative Commission for the Coordination of Social Security Systems) are taken into account. The historical method might be applied in order to discover the reasons for the present legislative solutions.

In the next step of the analysis several options – either enumerated in the mandate and/or discovered during the analysis – which might present a solution to a given problem are tested. Their strengths, weaknesses, opportunities and threats are stressed. In some of the reports, for each of the possible solutions specific criteria are taken into account, such as clarity, simplicity, protection of rights, administrative burden and implementation arrangements, risk of fraud and abuse and potential financial implications.

To this end, not only the relevant EU law is taken into consideration, but other sources as well. These might include relevant literature and academic studies, reports and selected key policy documents of the EU and possibly of other international organisations.

Next to this comparison on an EU (and international) level, a vertical comparative method is applied in order to present the problems and test the solutions in (some or all, depending on the mandate) Member States. A horizontal comparative method at the national level is enabled by mapping the situation in several (or all) Member States.

Logical, grammatical and teleological methods of legal interpretation are relied upon as well. Conclusions are as a rule drawn from the descriptive-analytical method of the research, which might prove useful for selecting the best policy options at national and EU level and for finding the most appropriate normative solutions de lege ferenda.

Access for economically inactive mobile EU citizens and jobseekers to certain social benefits

In relation to the option extending the derogation from the principle of equal treatment in relation to access to non-contributory family benefits, long-term care benefits and sickness benefits for economically inactive mobile EU citizens and jobseekers, as this was developed quite late in the process, data from the Labour Force Survey (LFS), ESSPROS and the 2015 Ageing report were used to estimate the affected population and economic impact (see Annex XXIV).

Annex VStudy for an impact assessment for revision of Regulations (EC) Nos 883/2004 and (EC) Nos 987/2009

Study for an impact assessment for revision of Regulations (EC) Nos 883/2004 and 987/2009 VT/2008/87Final report6 December 2013Richard Doherty, Benoît Vandresse, Steven Bulté, Miguel Bardaji Horno, Michael Ulrich * Jozef Pacolet, Frederic De Wispelaere*** Deloitte Consulting ** HIVA KU LeuvenStudy for an impact assessment for revision of Regulations (EC) Nos 883/2004 and 987/2009 VT/2008/87Final report6 December 2013Richard Doherty, Benoît Vandresse, Steven Bulté, Miguel Bardaji Horno, Michael Ulrich * Jozef Pacolet, Frederic De Wispelaere*** Deloitte Consulting ** HIVA KU Leuven 

COLOPHON

Title: Study for an impact assessment for revision of Regulations (EC) Nos 883/2004 and 987/2009

Authors: Richard Doherty (Deloitte Consulting), Benoît Vandresse (Deloitte Consulting), Steven Bulté (Deloitte Consulting), Miguel Bardaji Horno (Deloitte Consulting) , Michael Ulrich (Deloitte Consulting), Jozef Pacolet (HIVA KU Leuven), Frederic De Wispelaere (HIVA KU Leuven),

National experts: Roland Eisen, Hans-Christian Mager and Jürgen Faik (Johann Wolfgang Goethe Universität Frankfurt am Main,), Norman Wagner (Social Policy Unit in the Chamber of Labour, Austria), Jan Edling (Flexicurity, Sweden), Eigil Boll Hansen (KORA, Danish Institute for Local and Regional Government Research), Raphael Wittenberg (London School of Economics and Political Science), Catalin Ghinararu (National Research Institute for Labour and Social Protection Romani), Marek Radvansky (Institute of Economic Research (EU SAV), Slovakia), Jüri Kore (Tartu University), Gregorio Rodríguez Cabrero (University of Alcalá), Dorota Kawiorska (Krakow University of Economics), Franz Clément (Centre d'Etudes de Population, de Pauvreté et de Politiques Socio-Economiques, Luxembourg), Cristina-Maria Martin (Deloitte Romania), Filip Van Overmeiren (Laga Belgium).

Acknowledgements: The authors would like to thank DG EMPL for the research possibilities provided within this project and especially the department EMPL B4 and the Impact assessment steering group for their constructive, challenging and inspiring support for this work and the hospitality while discussing this and previous versions of the report.

Contents

1Introduction

1.1Purpose of the document

1.2Structure and content of the final report

2Study objectives

3Methodological approach

3.1In-depth analysis based on a sample of 14 Member States

3.2Data sources

3.2.1Strategic interviews with DG EMPL officials and other key EC officials

3.2.2Available literature

3.2.3Analysis of replies to the online EC public consultation

3.2.4EU-wide web-based survey among responsible public authorities and social security institutions

3.2.5Interviews with key stakeholders at EU level

3.2.6Country visits: data collection at national level

3.2.7Available and newly generated statistical data

4Baseline scenario

4.1Coordination of unemployment benefits

4.1.1Scale of cross-border mobility of workers in EU-27

4.1.2Aggregation of periods of employment/insurance/self-employment

4.1.3Export of unemployment benefits

4.1.4Estimated cros-border expenditure on unemployment benefits

4.1.5Reimbursement claims

4.1.6Estimated current administrative costs and burden

4.2Coordination of long-term care benefits

4.2.1Scale of cross-mobility of pensioners

4.2.2Number of persons insured for LTC

4.2.3Number of persons receiving LTC benefits in cash

4.2.4Number of persons receiving LTC benefits in kind

4.2.5Estimated number of PD S1 issued by category and estimated expenditure on LTC benefits

4.2.6Estimated cross-border expenditure on healthcare and LTC

4.2.7Current estimated administrative costs and burden

4.3Summary – Estimated current administrative cost - Baseline scenario

5Problem definition

5.1Introduction

5.2Coordination of LTC benefits

5.2.1Problems of classification of long-term care benefits under EU law

5.2.2Uneven application of coordination rules by Member States

5.2.3Gaps in social protection of mobile citizens

5.2.4Risk of double-payments for the same risk

5.2.5Instability of the current system of coordination of long-term care benefits

5.2.6Citizens not sufficiently understanding their rights under EU law

5.2.7Coordination of long-term care benefits: Problem tree

5.3Coordination of unemployment benefits

5.3.1Unawareness of rights of mobile workers among workers and public authorities

5.3.2EU rules not sufficiently taking into account changing job market conditions & Risk mobile workers enjoy less favourable conditions for reintegration into the labour market

5.3.3Unbalances distribution of financial burden

5.3.4Migrant workers receiving low returns on their contributions

5.3.5Risk of loss of rights to benefits

5.3.6Barriers to job-seeking abroad

5.3.7Uneven application of current coordination rules by Member States

5.3.8Coordination of unemployment benefits: problem tree

6Description of the policy objectives

7Impact assessment of the policy options

7.1Introduction

7.2Impact assessment criteria

7.3Scoring indicators

7.4Coordination of long-term care benefits

7.4.1Overview of possible options

7.4.2Option A: Baseline scenario

7.4.3Option B: Adoption of a safeguarding provision ensuring that a mobile person does not remain without any protection

7.4.4Option C1: Member State of residence provides long-term care benefits on the basis of its legislation

7.4.5Option C2: Competent Member State provides long-term care benefits to insured persons residing abroad (export)

7.4.6Overview of impact assessed

7.5Coordination of unemployment benefits

7.5.1Overview of possible options

7.5.2Option A: Baseline scenario

7.5.3Option B: Right of choice for frontier workers (B)

7.5.4Option C: Unemployment benefits for all workers to be provided by the country of the last activity regardless of the person’s residence

7.5.5Option D: Prolongation of the minimum period of export from 3 months to 6 months

7.5.6Option E: Clarification of the provision on the aggregation of periods of insurance, employment and self-employment

7.5.7Overview of the impact assessed

8Comparison of the options

8.1Coordination of unemployment benefits

8.1.1Overview

8.1.2Budgetary impacts of the different options and likely reimbursement claims

8.1.3Administrative cost and burden of rules related to the aggregation of periods of insurance or (self-)employment

8.2Coordination of LTC benefits

8.2.1Overview

8.2.2Budgetary impact of the different options

8.2.3Administrative cost and burden of cross-border LTC in kind or in cash

9Summary of the conclusions

9.1Introduction

9.2Current situation

9.3Current problems

9.3.1Coordination of LTC benefits

9.3.2Coordination of unemployment benefits

9.4Policy objectives

9.5Impact assessment of the options

9.5.1Coordination of LTC benefits

9.5.2Coordination of unemployment benefits

10Annexes

10.1Data collection: questionnaire

10.1.1Coordination of unemployment benefits: data collection

10.1.2Coordination of LTC benefits: data collection

10.2Labour Force survey and other survey data

10.3Detailed tables on unemployment benefits

10.3.1Estimated annual expenditure baseline scenario and different options

10.3.2Estimated reimbursement claims baseline scenario and different options

10.4Detailed tables on long-term care

10.4.1Estimated number of insured persons living in another country than the competent country

10.4.2Estimations based on data from the Audit Board

10.4.3Annual cross-border expenditure LTC benefits in kind and in cash

10.5Eurostat: inactivity rates

10.6List of national experts for the statistical data collection

10.7Reliability limits LFS

10.8Analysis of the administrative burden

10.9Analysis of online survey of national administrations and social security institutions

10.9.1Coordination of unemployment benefits

10.9.2Coordination of LTC benefits

10.10List of strategic interviews

10.11List of face-to-face interviews

List of figures

Figure 1: Welfare state models    

Figure 2: Applied methodology - LTC    

Figure 3: Applied methodology - UB    

Figure 4: Problem tree - coordination of LTC benefits    

Figure 5 - Problem tree coordination of unemployment benefits    

Figure 6: Options - Coordination of LTC benefits    

Figure 7: Coordination of unemployment benefits: considered policy options    

Figure 8: Current and future flow of documents applicable to the aggregation of periods    

Figure 9: Current and future flow of documents applicable to the export of UB    

Figure 10: Current and future flow of documents applicable to LTC    

List of tables

Table 1 : Overview of data sources    

Table 2: Overview of statistical data sources used: data availability and limitations    

Table 3: Synoptic overview of the scope of the cross border use of unemployment benefits and LTC benefits under scrutiny**    

Table 4: Number of cross-border workers and frontier workers, EU-27, Average of 2010 & 2011, in .000    

Table 5: Migrant workers (from 15 to 64 years) in 2011 by country of previous residence, EU-27    

Table 6: Estimated number of cross-border workers who will become unemployed, in .000    

Table 7: Estimated number of recent mobile workers who will become unemployed    

Table 8: Estimated number of received PD U1 in the EU-27: 2010, 2015 & 2020    

Table 9: Yearly estimated amount paid to unemployed persons who received a PD U1 2010, 2015 & 2020 (in EUR)    

Table 10: A projection of the number of export unemployment benefits (PD U2): 2010, 2015 & 2020    

Table 11: Number of jobseekers registered in EURES cv online and comparison with estimated number of PD U2 certificates issued, by country of residence, in .000    

Table 12: Estimated administrative Cost - PD U1 (‘received’), EU-27, EUR, 2010    

Table 13: Estimated administrative Cost – PD U1 (‘issued’), Poland and Belgium, EUR, 2013    

Table 14: Estimated administrative Cost –SED U004 ‘Salary Info’, SED U006 ‘Family Info’, Poland EUR, 2013    

Table 15: Estimated administrative Cost –PD U2 (‘issued’), EU-27, EUR, 2010    

Table 16: Estimated Administrative Cost – Competent employment service, SED U001, SED U012, Poland, EUR, 2013    

Table 17: Estimated Administrative Cost – Employment service of the Member State where the jobseeker has gone, PD U2 (‘process’), SED U007, SED U009, SED U010, PD U3 ‘issue’, SED U013, SED U028, Poland, EUR, 2013    

Table 18: Estimated Administrative Cost, Member State of Residence, SED U020, SED U025, Poland, 2013    

Table 19: Estimated Administrative cost – Competent Member State, SED U021, SED U022, SED 023, SED U024, Poland, EUR, 2013    

Table 20: Migrant pensioners (65 years or over) in 2011 by country of previous residence, EU-27    

Table 21: Estimated number of PD S1 issued by category of citizen, in .000    

Table 22: Estimated number of PD S1 issued/received, in .000    

Table 23: Entropy and Herfindahl indexes of concentration of cross-border insured persons by PD S1    

Table 24: Estimated LTC cross-border expenditure baseline scenario (in € billion), as % of total spending and projections 2020 and 2030    

Table 25: % cross-border users LTC in kind or in cash of total population and average amount per dependent person using LTC in kind or in cash (thousand €)    

Table 26: Estimated number of cross-border users benefiting from LTC (in thousand) and budget (in million €)    

Table 27: Estimated number of cross-border users from LTC in kind or in cash, projections 2020 and 2030 (in thousand)    

Table 28: Estimated LTC cross-border expenditure baseline scenario (in € .000), by country of residence*    

Table 29: Estimated claims of LTC benefits in kind introduced in 2011, by debtor country, in .000    

Table 30: Estimated administrative cost - PD S1 'issued', EU-27, EUR, 2013, in 000    

Table 31: Estimated administrative Cost – Competent Member State, E125, Poland, EUR, 2013    

Table 32: Estimated administrative Cost – Member State of residence, E125, S1/E100/E107/E001, S001, Poland, EUR, 2013    

Table 33: Baseline scenario – estimated administrative cost: PD U1 (in €), PD U2 (in €), PD S1 (in € 000)    

Table 34: General attitude towards need to change the current LTC coordination rules    

Table 35: Lack of common definition of LTC benefits under EU law - comparison of definitions    

Table 36: Diversity of benefits that can/should be considered as LTC benefits in the MS & Benefits situated under different social security branches    

Table 37: Diversity of LTC benefits system: in cash, kind or both    

Table 38: Impact assessment criteria    

Table 39: Coordination of LTC benefits: overview of the impacts assessed per option    

Table 40: Coordination of unemployment benefits: overview of impact assessed per option    

Table 41: Comparison of the effectiveness, efficiency and coherence of the options (baseline=1)    

Table 42: Total estimated yearly expenditure unemployment benefits for unemployed cross-border workers (in € 000)    

Table 43: Total estimated yearly expenditure unemployment benefits for unemployed cross-border workers paid by the country of residence (in € 000)    

Table 44: Total estimated yearly expenditure unemployment benefits for unemployed cross-border workers paid by the country of last activity (in € 000)    

Table 45: Total estimated yearly expenditure unemployment benefits for unemployed cross-border workers - % proportion country of residence vs. country of last activity    

Table 46: Estimated reimbursement claims (scenario 3 months) for the baseline scenario and the different options and the impact of the maximum boundary, in € 000    

Table 47: Summary estimated yearly expenditure on unemployment benefits and estimated reimbursement claims (in million €)    

Table 48: Estimated number of unemployed cross-border workers and country responsible for payment and reimbursement    

Table 49: Estimated administrative cost aggregation of periods of insurance of (self-)employment    

Table 50: Comparison of the effectiveness of the options (baseline scenario = 1)    

Table 51: Estimated number of cross-border users benefiting from LTC (in thousand) and budget (in million €), Option where competent country is providing LTC benefits    

Table 52: Estimated number of cross-border users benefiting from LTC (in thousand) and budget (in million €), Option where country of residence is providing LTC benefits    

Table 53: Estimated budgetary impact as debtor of baseline scenario and options where country of residence or competent country are providing LTC benefits, breakdown LTC benefits in cash and in kind, in .000 €    

Table 54: Estimated % difference with baseline scenario (=100%) of options where country of residence or competent country are providing LTC benefits, breakdown LTC benefits in cash and in kind, by debtor country    

Table 55: Estimated administrative cost and burden baseline scenario and options where country of residence or competent country are providing LTC benefits    

Table 56: Synoptic overview of the scope of the cross border use of unemployment benefits and LTC benefits under scrutiny**    

Table 57: Comparison of the effectiveness of the options (baseline scenario = 1)    

Table 58: Comparison of the effectiveness, efficiency and coherence of the options (baseline=1)    

Table 59: Number of unemployed persons that have been certified insurance or (self-)employment periods completed in another Member State taken – number of received forms (U1/E301) by Member State issuing a U1/E301 form    

Table 60: Number of unemployed persons that have been certified insurance or (self-)employment periods completed in another Member State taken – number of unique unemployed persons by Member State issuing a U1/E301 form    

Table 61: Number of unemployed persons that have been certified insurance or (self-) employment periods completed in another Member State – number of received forms (U1/E301) or number of unique unemployed persons by Member State issuing a U1/E301 form (summary)    

Table 62: Number of unemployed persons that have been certified insurance or (self-)employment periods completed in another Member State – estimation EU-27    

Table 63: Total yearly amount of unemployment benefits paid to unemployment persons who have certified periods from other Member States?    

Table 64: Number of issued U1/E301 forms    

Table 65: Number of unemployed persons who want to look for work in another Member State while continuing to receive unemployment benefits in the competent Member State – number of requests    

Table 66: Number of unemployed persons who want to look for work in another Member State while continuing to receive unemployment benefits in the competent Member State – number of requests of prolongation    

Table 67: Number of unemployed persons who want to look for work in another Member State while continuing to receive unemployment benefits in the competent Member State – success rate    

Table 68: Number of U2/E303 forms received    

Table 69: number of unemployed persons who went to look for work in another Member State while continuing to receive unemployment benefits in the competent Member State - estimation for the whole EU-27    

Table 70: Numbers of U3 forms issued and received    

Table 71: Reimbursement of competent Member State to institutions of place of residence - claims paid (as debtor - amount in EUR    

Table 72: Reimbursement of competent MS to institutions of place of residence - claims received (as creditor) - amount in EUR    

Table 73: Number of persons (insured persons and family members) insured for LTC benefits in kind and in cash who are living in another MS than the competent MS - status    

Table 74: Number of persons (insured persons and family members) insured for LTC benefits in kind and in cash who are living in another MS than the competent MS    

Table 75: Number of persons (insured persons and family members) receiving LTC benefits in cash who are living in another MS than the competent MS – in cash    

Table 76: Number of persons (insured persons and family members) receiving LTC benefits in kind who are living in another MS than the competent MS – reporting country=debtor – in kind    

Table 77: Number of persons (insured persons and family members) receiving LTC benefits in kind who are living in another MS than the competent MS – reporting country=creditor – in kind    

Table 78: Yearly cross-border expenditure related to LTC in cash? Reporting country = debtor – in cash    

Table 79: Yearly cross-border expenditure related to LTC in kind – reporting country = debtor – in kind    

Table 80: Yearly cross-border expenditure related to LTC in kind - reporting country=creditor - in kind    

Table 81: Number of cross-border workers (in .000), 2011    

Table 82: Number of cross-border workers (in .000), 2010    

Table 83: Average number of cross-border workers EU-27 (in 000), Average 2011 and 2010    

Table 84: Share in total number of incoming cross-border workers (as % of column total), by country of residence, top 3 of incoming MS, average of 2011 and 2010, EU-27    

Table 85: Share in total number of outgoing cross-border workers (as % of row total), by country of employment, top 3 of outgoing MS, average of 2011 and 2010, EU-27    

Table 86: PD A1 issued in reporting countries by destination, 2010, EU-27 (without SE as sending country)    

Table 87: PD A1 issued in reporting countries by destination (as % of row total), top 3 of sending countries, 2010, EU-27 (without SE as sending country)    

Table 88: PD A1 issued in reporting countries by destination (as % of column total), top 3 of destination countries, 2010, EU-27 (without SE as sending country)    

Table 89: Baseline scenario: frontier workers return to country of residence and other cross-border workers have right to choose, scenario: 50% register in country of residence and 50% register in country of employment, in € .000    

Table 90: Baseline scenario: frontier workers return to country of residence and other cross-border workers have right to choose, scenario: rational decision, in € .000    

Table 91: Right to choose: All cross-border workers have right to choose, scenario: 50% register in country of residence and 50% register in country of employment, in € .000    

Table 92: Right to choose: All cross-border workers have right to choose, scenario: rational decision, in € .000    

Table 93: Stay in country of employment: All cross-border workers stay in the country of last employment, in € .000    

Table 94: Return to country of residence: All cross-border workers return to their country of residence, in € .000    

Table 95: Baseline scenario: frontier workers return to country of residence and other cross-border workers have right to choose, scenario: 50% register in country of residence and 50% register in country of employment – reimbursement claim of 3 months, in € .00    

Table 96: Baseline scenario: frontier workers return to country of residence and other cross-border workers have right to choose, scenario: 50% register in country of residence and 50% register in country of employment – actual payment, in € .000    

Table 97: Baseline scenario: frontier workers return to country of residence and other cross-border workers have right to choose, scenario: rational decision – reimbursement claim of 3 months, in € .000    

Table 98: Baseline scenario: frontier workers return to country of residence and other cross-border workers have right to choose, scenario: rational decision – actual payment, in € .000    

Table 99: Baseline scenario: frontier workers return to country of residence and other cross-border workers have right to choose, 50% register in country of residence and 50% register in country of employment – reimbursement claim of 3 months, in € .000    

Table 100: Right to choose: All cross-border workers have right to choose, scenario: 50% register in country of residence and 50% register in country of employment – actual payment, in € .000    

Table 101: Right to choose: All cross-border workers have right to choose, scenario: rational decision – reimbursement claim of 3 months, in € .000    

Table 102: Right to choose: All cross-border workers have right to choose, scenario: rational decision – actual payment, in € .000    

Table 103: Return to country of residence: All cross-border workers return to their country of residence – reimbursement claim of 3 months, in € .000    

Table 104: Return to country of residence: All cross-border workers return to their country of residence – actual payment, in € .000    

Table 105: Number of insured incoming cross-border workers and their family members (20%), in 000, average 2010 and 2011    

Table 106: Number of insured retired cross-border workers and their family members (25%), in 000    

Table 107: Number of insured migrant pensioners (number of persons aged 60 and older at arrival and retired by country of birth), in 000, 2011    

Table 108: Number of insured migrant pensioners and their family members (25%), in 000, 2011    

Table 109: Share in total number of persons insured in another country than the country of residence (as % of row total), top 3 of competent MS    

Table 110: Share in total number of persons living in another country than the competent country (as % of column total), top 3 of MS of residence    

Table 111: Claims on health introduced in 2011 by creditor countries    

Table 112: Outstanding health claims on 31/12/2011 by creditor and debtor country as new health claims introduced in 2011, in € .000    

Table 113: Estimate of new claims introduced in 2011 (health including LTC) by debtor and creditor countries (based on outstanding claims), in €. 000    

Table 114: Cross-border expenditure LTC benefits in kind, based on LTC spending per capita of the country of residence, in € 000    

Table 115: Cross-border expenditure LTC benefits in kind, based on LTC spending per capita of the competent country, in € 000    

Table 116: Cross-border expenditure LTC benefits in cash, based on LTC spending per capita of the country of residence, in € 000    

Table 117: Cross-border expenditure LTC benefits in cash, based on LTC spending per capita of the competent country, in € 000    

Table 118: Cross-border expenditure LTC benefits in cash, based on LTC spending per capita of the country of residence, in € 000    

Table 119: Cross-border expenditure LTC benefits in kind and in cash, based on LTC spending per capita of the competent country, in € 000    

Table 120: Inactivity rates (inactive population as a percentage of the total population, from 20 to 64 years, quarterly figures    

Table 121: Statistical data collection: list of national experts    

Table 122: Reliability limits Labour Force Survey, 2012    

Table 123: Administrative burden – aggregation of periods – baseline scenario - Case: Belgium    

Table 124: Administrative burden – export of UB – baseline scenario - Case: Belgium    

Table 125: Administrative burden – LTC benefits in kind – baseline scenario - Case: Belgium    

Table 126: Administrative burden – LTC benefits in cash – baseline scenario - Case: Belgium    

Table 127: Profile of respondent – unemployment benefits survey - by MS    

Table 128: With which countries do you mostly exchange information (top-3)?    

Table 129: Opinions on likely impacts of the options - unemployment benefits    

Table 130: Opinions on impacts of options (export of unemployment benefits) - unemployment benefits    

Table 131: Export of unemployment benefits - current practices    

Table 132: Opinions on the need for physical presence while job seeking    

Table 133: Opinions on communication with other MS – unemployment benefits    

Table 134: Opinions on need to reform coordination rules - unemployment benefits    

Table 135: Profile of respondent – LTC benefits survey - by MS    

Table 136: options on competent MS for provision of LTC benefits    

Table 137: Opinions on impacts of the different LTC options    

Table 138: Opinions on fair burden sharing of the financial burden between Member States    

Table 139: Opinion on functioning of communication between MS - LTC    

Table 140: Opinions on the need to reform the current LTC coordination rules    

Table 141: Main problem regarding application of LTC coordination rules    

1Introduction

1.1Purpose of the document

This is Deloitte’s and HIVA’s report to the European Commission, DG EMPL with regard to the “Study for an impact assessment for revision of Regulations (EC) Nos 883/2004 and 987/2009”. 

1.2Structure and content of the final report

We have structured the final report as follows:

Chapter 1 includes a short introduction to the final report;

Chapter 2 outlines the objectives of this study;

Chapter 3 recalls the methodology that we have used for this study;

Chapter 4 presents the current situation with regard to the coordination of LTC and unemployment benefits. We focus specifically on the scale of potentially affected people and the most common mobility patterns

Chapter 5 contains a definition of the problems under the current rules that were identified for both the coordination of unemployment benefits and the coordination of long-term care benefits.

In chapter 6, we describe the policy objectives that the Commission aims to achieve with the coordination of LTC and unemployment benefits.

Chapter 7 aims to describe the likely qualitative and quantitative impacts of the various policy options in relation to the baseline scenario.

In Chapter 8, we provide an overview of the strengths and weaknesses of the various policy options according to three meta-criteria: efficiency, effectiveness and coherence.

Chapter 9 contains a summary of the main conclusions of this study (we have also produced this summary in a separate standalone document).

Chapter 10 contains the annexes attached to this final report.

2Study objectives

The following tasks were carried out by the research team:

The collection of socio-economic data for determining the scale of the identified problems and the baseline scenario in the area of coordination of social security schemes, in particular with regard to the coordination of long-term care and unemployment benefits. The study should provide the evidence-based description of the baseline scenario and focus on providing supporting data, providing input for the verification of the intervention logic.

A description of the qualitative and quantitative impacts of the policy options.

Comparison of the policy options.

Providing support in the process of stakeholders' consultation, namely to:

oevaluate results of the public consultation (online questionnaire launched online in all EU languages by the EC on 5/12/2013 and closed on 5/3/2013)

ocarry out and evaluate results of a consultation of organisations & networks operating in cross-border regions

ocarry out and evaluate results of a consultation of national institutions and administrations (e.g. employment services, health services).

3Methodological approach

3.1In-depth analysis based on a sample of 14 Member States

The data collection and analysis are focused on a representative sample of 14 Member States. The selection of the countries is primarily based on relevant mobility patterns and a balanced coverage in terms of types of the social security systems, more notably in the area of unemployment and long-term care benefits.

Existing research on the welfare state and the social protection for dependent older persons (long-term care) has shown that a characterisation along the lines of more Bismarck-oriented and more Beveridge oriented welfare states remains fruitful 7 .

Bismarck-oriented models are characterized by income-based contributions earmarked to specific funds destined to cover concrete social risks. They are usually tightly linked to the employment system (contributions made by the employer and/or the employee, and benefits limited to the contributor and his/her family; besides, the job employment sector often determines the corresponding insurance fund).

Beveridge-oriented models, on their side, are based on the idea of universal coverage against the main social risks, providing access to all the citizens. The Nordic countries and the UK represent this approach, though universal coverage became the guiding principle in all European countries during the second half of the 20th Century. Some of the “new” Member States, which mainly followed Beveridge-oriented models during the second half of the 20th Century, are shifting to a “Bismarck”-oriented model since the 90s.

The system of social protection in the EU is characterized by welfare pluralism. Member States aim at the same or similar goals, however making use of different implementation mechanisms and institutions. Although the differences between social security models are blurring and each country has its own specificities, eight groups or clusters of Member States can be distinguished following their Bismarck/Beveridge orientation. Four models can be differentiated among the “old” Member States (EU-15); similarly, four models can be identified among the “new” Member States.

In spite of the apparent convergence, welfare state regimes remain rooted in history are path depending that determines further directions for new branches of social protection, like long-term care. Those typologies of welfare state regimes seem to be clustered also geographically, and are characterised also by differences in the development of the welfare state (at a higher or lower level), coupled with higher or lower levels of economic development (GDP) and more or less oriented to in cash or in kind benefits. Those dimensions substantially influence the flows of funds related to the social protection of mobile citizens. The apparent geographic clustering seems also to be confirmed by labour market mobility 8 . Those typologies appear in similar form in most of the comparative studies of welfare state regimes. 9 Furthermore, research on unemployment systems has yielded similar clusters of countries 10 .

For these reasons, we have decided to use the same sample of countries for both areas (long-term care and unemployment). We selected from each of those regimes one or more countries to cover the welfare state pluralism in Europe. The distinction between old and new will disappear, but we retain it for this study because of the still relative recent occurrence of the enlargement and because of the difference in economic development between both at this point of time.

Figure 1: Welfare state models

Source: Pacolet J. and Coudron V., 2006 “State of the Welfare State in the EU Anno 1992: Ten years later and with ten new Member States. Welfare State trend spotting”, in Pacolet J., Ed. (2006). L'Etat de l'Etat-providence dans l'UE en 1992 et dix ans plus tard avec dix nouveaux Etats membres. (Pacolet, J., Ed.). La Revue Belge de Sécurité Sociale

The 14 countries retained, as indicated in the previous table, are: Austria, Belgium, Denmark, Estonia, France, Germany, Poland, Luxemburg, Netherlands, Romania, Spain, Slovakia, Sweden and United Kingdom.

With the aim of ensuring the representativeness and robustness of the impact analysis, which according to the Terms of Reference may be done by clusters of countries, our sample of Member States covers seven of the eight welfare state models (given their similarities, the “new” Bismarck-oriented Mediterranean Member States –Malta and Cyprus- not included).

The allocation of a country to a specific model (Beveridge or Bismarck-oriented) does not necessarily imply that this country will have the same problems in terms of the management of social security cases compared to the other countries allocated to this specific model. The administration and the legislation in this area could diverge even within a specific model.

We account as well, in the country selection, labour mobility trends. With the selection of those countries we include at least 80% of both the outgoing and ingoing cross-border working citizens in the EU-27. In terms of population, our sample accounts for around 75% of the EU-27 population (cf. from Table 83 until Table 88 ).

3.2Data sources

Two types of data sources were used during this study:

secondary data, in particular existing statistical data;

primary data, collected through interviews and a consultation of the stakeholders.

Table 1 : Overview of data sources

In the initial stage of the study, we conducted 5 face-to-face interviews with EC officials within DG EMPL and DG ECFIN. The interviews with DG EMPL officials served to acquire more detailed knowledge concerning the contextual environment of the study. DG ECFIN officials assisted us in identifying relevant data from the Labour Force Survey (LSF) and the Ageing Working Group that was used for this study. These interviews also led to the transfer of relevant documents to the contractor.

The list of conducted strategic interviews can be found in Annex 10.10

For this study, we made use of a wide ranging and comprehensive desk based research. It encompassed a comprehensive review of available literature on both topics (coordination of long-term care benefits and of unemployment benefit systems) and its implications for mobile workers at EU level. Review of available literature on the subject provided input into three main issues: the underlying causes of the problems that provide the focus for the regulations on the coordination of long-term care benefits and unemployment benefits at EU level;

the description of the problem situation and of the policy objectives;

the impact of the “baseline scenario” of continuity with the current situation.

Literature that was part of the scope of the review included:

Communications, resolutions and legislation at EU level, i.e. Regulation (EC) N° 883/2004 and Implementing Regulation N°987/2009 on the coordination of social security systems, 30.10.2009

Relevant Court of Justice rulings (cf. Chapter: Context);

Reports and studies at EU level, including the work conducted by the TrESS network;

Reports and studies at national level.

The European Commission launched on the 5th December 2012 a public consultation ( http://ec.europa.eu/social/main.jsp?langId=en&catId=699&consultId=12&furtherConsult=yes ) addressed to: 

All EU and non-EU citizens who are insured under the social security system of an EU or EEA/EFTA Member State.

Any public and private organisations with activities in an EU or EEA/EFTA Member State.

The objective of the consultation was to collect opinions and experience in the area of coordination of unemployment and long-term care benefits for persons who are in a cross-border situation. Furthermore, opinions on possible policy options and further ideas on how to remove existing problems or obstacles in these areas were gathered.

We were asked to analyse and exhaustively report the replies to the EC public consultation.

The consultation received 299 online (+3 email) replies across the EU and EEA/EFTA states. 199 were from individuals and 103 on behalf of an organisation or as specialists. In addition, three other stakeholders did not fill out the online questionnaire, but sent their opinions separately by e-mail.

By nationality, Spanish were the most numerous among individual respondents, accounting for 26.6% of the responses. No replies were received from Cyprus, Denmark, Iceland, Liechtenstein, Luxembourg, Malta and Switzerland.

The analysis of the public consultation replies is considered a stand-alone document, in line with the ToR. It was submitted to the European Commission before the finalisation of the final report.

In order to expand the scope of our analysis to all EU and EEA countries, we launched a web-based survey among the responsible national public authorities and other key actors with regard to both topics. The Commission sent an invitation to the members of the Administrative Commission for the Coordination of Social Security systems in December 2012. After several reminders and prolongation of the deadlines, we finally closed the survey on the 28 February 2013.

We received 81 complete replies to our survey:

No answers at all were received from the following countries: Bulgaria, Finland, Greece, Iceland, Liechtenstein and Norway.

59 respondents have only filled in the survey on unemployment benefits.

10 respondents have only filled in the survey on long-term care benefits.

12 respondents have filled in both surveys on unemployment and long-term care benefits.

In addition, we received 67 incomplete replies. The incomplete replies were not taken into account in the survey analysis (in the figures and tables) to avoid double counting. However, relevant and useful information, examples or arguments from these incomplete replies were taken into account in the impact analysis.

On top of the countries that have not participated at all in the survey, we have no answer from BE, CY, DK, FR, IE, PT, RO, CH and UK for the questions on LTC. Among these countries, the UK public officials expressed their intention of submitting a reply during the meetings held with them after the survey’s deadline. Belgian authorities also expressed in mid-March their intention to provide a reply.

On top of the countries that have not participated at all in the survey, we have no answer from SV for the questions on unemployment benefits.

Some countries from which we have received several completed questionnaire are overrepresented in the aggregated results. Therefore, the analysis of the aggregate results was complemented by the analysis of national replies when necessary.

The analysis of the on-line survey can be found in Annex 10.9 to the final report.

In order to complement the EC public consultation (and to address stakeholders that have not replied to the public consultation), a number of face-to-face interviews were conducted with stakeholder parties that are organised at EU level. We particularly addressed: We will particularly addressed:

networks operating in cross-border regions;

social partners organised at EU level;

civil society organisations at EU level.

The list of interviews with stakeholders at EU level can be found in Annex 10.10 to the final report.

In order to assess the impacts of the policy options, we organised country visits in the 14 countries of our sample. During these field visits, we met the key stakeholders and affected actors in the field of long-term care benefits and unemployment benefits. The field research was conducted by Deloitte and HIVA staff, supported by academic experts with a good knowledge of the national context. A country visit usually took 3 to 6 working days depending on the availabilities of the interviewees. Country visits consisted of interviews with responsible public authorities and other interested parties (depending on the specificities of the national context) and where feasible in the time-frame of our visits, workshops to be able to assess the administrative costs and administrative burden related to each of the policy options that are being considered. We have privileged face-to-face interviews or group interviews but after our visits in the countries we have had also the opportunity to come-back to some interviewees by phone in order to complete our information.

For both topics (LTC and unemployment benefits), 5-8 key stakeholders at national level were consulted. In total, we have conducted more than 100 relevant interviews with stakeholders in these 14 Member States.

Interviews with these people fed the problem definition and generated anecdotic evidence and examples and helped us to assess the likely impacts of the different policy options (for example, with regard to the impact on the level of social security coverage of affected persons, changes in administrative burden and costs, impact on the budgets of Member States, expected changes in the EU internal market, etc.).

The (type of) stakeholders differed from country to country and per topic. We consulted:

National, (regional or local) institutions and administrations, particularly the public employment services (coordination of unemployment benefits) and healthcare services (coordination of long-term care benefits).

Health insurers (coordination of long-term care benefits);

Social partners (coordination of unemployment benefits);

Civil society organisations operating in cross-border regions.

The face-to-face interviews were based on a semi-structured questionnaire that was prepared by the core team in advance.

Please find a list of interviews (per country) in Annex 10.10 to the final report.

In order to assess the administrative burden and administrative costs related to the baseline scenario and each of the different policy options, we aimed to organise a one-day workshop per topic (coordination of long-term care benefits and coordination of unemployment benefits) and per country. We used face-to-face interviews and group interviews, where we faced availability or practical difficulties or where we have been referred to only a limited number of key persons that could help us further.

For the assessment of the administrative burden and costs, we used the Standard Cost Model, as requested by the ToR (more detail cf. 4.1.6 – baseline scenario).

The Standard Cost Model (SCM) serves as a useful tool for assessing the administrative cost and administrative burden stemming from Information Obligations (IOs) imposed by a Regulation. We have applied the SCM for assessing the administrative cost and administrative burden of the management of a cross-border case of unemployment benefits.

In our approach, we measured the time (T) the national administrations and stakeholders spent on the information obligations (IO) imposed by the Regulations on cross-border cases. Once we have defined the time spent on each IO, we have multiplied the (average) time by the hourly tariff (W) of those involved in meeting the information obligation (administrative staff). For consistency and comparability with other SCM assessments of EU regulation, the tariff variable used in this study is based on hourly labour costs (plus overheads) per category of employment that has previously been used in recent SCM studies for DG EMPL 12 and our recent Impact Assessment studies we have conducted for the Commission. We have applied an average tariff/hour of 18€. The multiplication of the average tariff by the time spent on the IO results in an estimated cost per unit.  Then, we have multiplied the estimated cost per unit by the amount of cases per year (N). The following formula summarizes our approach:

N x W x T

The following data sources were exploited to obtain a detailed quantitative view on the baseline scenario of the coordination of unemployment benefits and LTC benefits:

Replies to our own questionnaire on LTC and unemployment benefits. The data collection was conducted in 14 Member States by national experts 13 ;

The EU Labour Force Survey (LFS);

The EC 2012 Ageing Report 14 ;

Other databases from Eurostat (e.g. Eurostat migration statistics 15 and ESSPROS 16 );

Other data sources (e.g. national reports, EU publications, Audit Board Report).

Replies to our own questionnaire on LTC and unemployment benefits by national experts

First, in order to describe and assess the baseline scenario, we deployed a network of national researchers in 14 Member States ( Table 121 ). Their task was to contact the responsible national administrations in order to collect statistical data with regard to a significant number of indicators in the areas of LTC and unemployment benefits. The reporting format for this data collection contained an Excel-file. A manual was provided to the national experts, including a short description of the objectives of the study and of the coordination legislation, more specific concerning the applicable rules on the coordination of unemployment and long-term care benefits. In the annex of the manual, a list with specific LTC benefits by Member State - defined by trESS 17 and based on the MISSOC-tables 18 - was provided. Also, possible contact persons sitting in the Administrative Commission or the Audit Board were provided. The questionnaire referred explicitly to the old E-forms and current Portable Documents (PD) 19 and Structured Electronic Documents (SED) 20 in order to obtain a similar understanding of the data needs. In the analysis below we refer frequently to these documents. Therefore, we first give a short definition/description of the Portable Documents we refer to: 21

PD U1 certificate: “The PD U1 certifies periods of insurance and employment or self-employment in another EU country that will be taken into account for the award of unemployment benefits”;

PD U2 certificate: “The PD U2 is the authorization you need to export your unemployment benefit if you are unemployed in an EU country and wish to move to another EU country to look for work”;

PD S1 certificate: “The S1 form allows you (and/or your family members) to register for healthcare if you live in an EU country but you are insured in a different one. The form is delivered per person (not per family)”;

PD P1 certificate: “The P1 form provides an overview of the decisions taken in your case by the various institutions in the EU countries from which you have claimed an old age, survivors or invalidity pension. The information on the P1 is intended to enable you to examine whether any of your pensions have been adversely affected by the interaction of decisions taken by two or more institutions”.

EU Labour Force Survey (LFS)

Second, we relied on data from the EU Labour Force Survey (LFS). The LFS is the main source of information with regard to the labour market situation and labour market trends in the European Union. While the LFS provides demographic data for all age groups, questions relating to labour market status are limited to persons in the age group of 15 years or older. The economically active population consists of employed and unemployed persons. Employed persons are persons aged 15 year and older, while unemployment persons are aged between 15 to 74 years. The LFS includes estimates of certain aspects of social protection for the whole population. It also contains some data on mobile population (e.g. cross-border workers, frontier workers, (recent) migrants working-age population, migrant pensioners, …) which can be used as a proxy for certain subcategories of the mobile population or as a benchmark for the total population and the total level of social expenditure, both at current point of time and in the future. The main advantage of the LFS is the data availability for all 27 MS and enables us to calculate proxies for indicators which could not been obtained by the questionnaire in 14 countries. At the same time these proxies could be cross-checked by the available administrative data for some MS collected by the questionnaire. This report presents different cross-tables based on the use of the LFS. We present them as detailed as possible. However, to guarantee reliability of the figures we only take into consideration the row and column totals of the cross-tables in our analysis. Row and column totals which can be assumed as reliable are marked with a *. 22 Nevertheless, the different cells in these cross-tables are important to estimate the baseline scenario and the different options (e.g. we need to select the cells which describe flows between neighbouring countries to define the number of frontier workers). Most of the cells will fall under the reliable limits. For that reason we refer especially to the reliability limits. The absolute numbers of those reliability limits are also given in Annex 10.7 . Cells or even row and column totals with numbers of persons below those figures are considered as not reliable. An additional part of the LFS is the so called 'ad hoc module’ which adds a set of questions to the questionnaire. Each year the subject of this module is different. In 2008 an ad hoc module about ‘the labour market situation of migrants and their immediate descendants’ was added to the core questionnaire. In 2014 this topic will be added again to the LFS which can be considered as an opportunity to analyse some of the specific defined variables in this module. 23  

EC 2012 Ageing Report

The 2012 Ageing Report was written by the European Commission (DG ECFIN) and the Economic Policy Committee (AWG) and presents projections of the budgetary impact of an ageing population in the EU-27 over the period 2010–2060. The report includes specific scenarios on the LTC public expenditure based on combined information from the System of Health Accounts (SHA) and ESSPROS (European system of integrated social protection statistics). The forecasted LTC public expenditure comprises both in-kind and cash benefits. The breakdown by type of LTC benefit (in cash or in kind), which is a crucial element for the calculation of the baseline scenario, is described in another publication of DG ECFIN. 24 Also the projections regarding employment and unemployment (e.g. amount paid to unemployed persons, number of unemployed persons) and other interesting general variables integrated in the Ageing Report will be used. The underlying assumptions and projection methodologies used in the Aging Report are described in a separate report of the EC. 25  

Eurostat data

Eurostat provides statistical information based on national administrative information or statistical surveys organized at national or EU level. Within the context of this project, we have mainly focused on the use of ESSPROS and the available migration statistics. ESSPROS contains data on the expenditure and receipts for all national social protection schemes and allows us to compare the total expenditure on unemployment benefits by a Member State (e.g. unemployment benefits paid to full unemployed persons) with unemployed benefits paid to persons who export their unemployment benefit (PD U2) or who needed to prove completed periods of insurance or (self-)employment in another Member State (PD U1). Also the Eurostat migration statistics are useful to estimate the number of migrants who moved abroad. We were particularly interested in the flows (yearly movements) and stocks (situation at a certain period of time) of recent migrants at working age (aged from 15 or 20 to 64) as well as in migrant pensioners. As mentioned above, also the Ageing Report is using ESSPROS regarding the expenditure on LTC. 26 But also the migration statistics collected by Eurostat, delivered by MS via administrative data or national surveys, will have their importance. Obtaining a reliable view on the number of migrant workers and pensioners will be important to assess the baseline scenarios of unemployment and LTC.

Other statistical data

Finally also other statistical information was considered. Yearly reports of national administrations and specific national reports discussing the export of unemployment and long-term care benefits delivered general or detailed information. At the same time, some recent reports published on EU level describe (labour) migration/mobility (E.g. Eurostat (2011), Indicators of Immigrant Integration; European Commission (2011), Mobility in Europe 2011).

Overview of the statistical data sources used

Table 2 provides an overview of the data sources that were used, with a particular attention to their limitations in terms of data availability. It gives a first impression of the strengths and weaknesses of the data. Within the context of this project data should be collected concerning the aggregation of periods of insurance or (self-)employment (PD U1/E301 form), reimbursement of unemployment benefits (SED U020-SED U025), export of unemployment benefits (PD U2/E303 form). Also, data on the number of PD P1 (Summary of pension decisions) could be interesting to determine the competent Member State. Not for all 14 selected MS administrative data was received by the questionnaire. Also, none of the Member States could response to all questions. More data input related to the application of the coordination rules for unemployment was obtained by the national experts compared to the application of the coordination rules for LTC benefits. Reasons for the lack of data for some MS, topics or questions are divers. The fact that administrative forms are received/issued decentralised OR are received on paper documents and not electronically will have a negative impact on the availability of administrative data. Due to the fact LTC is not considered as a specific social security branch in the coordination rules and is embedded within the rules of sickness benefits, MS do not explicit collect data on LTC. However, the PD S1 indicates if the holder is receiving a LTC benefit in cash 27 and a breakdown by nature of benefit is taken up in the SED S080 (claim of reimbursement). Today, still old E-forms are used which excludes the availability and use of information based on those ‘new’ forms. Also, some MS consider (some of) the LTC benefits in cash as not exportable within the current coordination rules. 28 The response to the questions is explicitly mentioned for each of the described topics below. MS who have answered the questionnaire are also taken up in the column titles (see Annex 10.1.2 ). Data collected by other administrative sources than by the national experts are inserted in footnotes below the tables (e.g. on the basis of national reports).

The LFS is confronted with a high non-response rate of migrants – and more particularly recently arrived migrants. 29 This will have a negative impact on the sample size and will cause a possible distortion of the composition of the migrant population. One of the limitations of using the LFS is that the calculated proxies do not completely correspond to the legal definitions. E.g. the number of frontier workers is calculated by taking into account the neighbouring countries and not the legal definition (return daily or at least once a week). Some ‘small’ adjustments could be made in the Labour Force Survey to guarantee useful data collection and more conformity with the legal definitions. Currently, migration statistics from the LFS are based on the ‘country of birth’ or the ‘nationality’. However, to obtain a reliable view on the migrant pensioner one should know the ‘competent Member State’ or the Member States which are paying an old-age benefit to these pensioners or at least the previous residence of the migrant pensioner (but also for migrant workers). This could be a priority in the ad hoc module of the Labour Force Survey of 2014 ‘Labour market situation of migrants and their immediate descendants’ or in a next module or on a permanent basis in the LFS. It could become a proxy for the number of pensioners who received a PD S1.

The 2012 Ageing Report is an important data source for the calculation of the baseline scenario of the coordination of unemployment and LTC benefits as well as for projections of both social security branches. However, these data will have its limitations compared to administrative data. First, no breakdown by type of unemployed person is available. In view of the reimbursement procedure it is necessary to select only cross-border workers and not migrant workers. Also, the Ageing Report made use of assumptions for the calculations of the chapter LTC. As reported in the methodology report (EC, 2011, p. 234) “The choice of methodology and various scenarios is heavily constrained by the availability, accessibility and quality of long-term care data”. 30 Nevertheless, a general agreement was reached of the underlying assumptions and projections for LTC in the Ageing Report and a number of sensitivity tests were carried out. The LTC projections in the Ageing Report are calculated for different scenarios (AWG 31 reference scenario, demographic scenario, high life expectancy scenario, base case scenario, constant disability scenario, shift 1% of dependents to formal scenario, coverage convergence scenario, cost convergence scenario, AWG risk scenario). The AWG reference scenario is used in this research report to calculate the projections on cross-border LTC spending. Thus this involves the assumptions used for this scenario. For calculating the LTC projections assumptions are made in the ageing report for:

“the future numbers of elderly people (through changes in the population projections used);

the future numbers of dependent elderly people (changes to the prevalence rates of dependency); 32

the balance between formal and informal care provision (assuming a given shift in demand

or exogenous changes in the availability of informal carers);

the balance between home care and institutional care within the formal care system;

the unit costs of care.”

In the Ageing Report, long-term care is defined 33 according to the System of Health Accounts (SHA), “as the sum of the following publicly-financed items:

Services of long-term nursing care (HC.3) (which is also called “the medical component of long-term care” or “long-term health care”, and includes both nursing care and personal care services), and

Social services of long-term care (HC.R.6.1), which is the “assistance services” part, relating primarily to assistance with IADL (instrumental activities of daily living) tasks". 34

If necessary, the SHA database will be supplemented with data from ESSPROS. The SHA-definition mainly represents in kind benefits. For that reason cash benefits from the ESSPROS functions ‘Disability’ and ‘Old-age’ are added to the SHA database.

The LTC cash benefits are projected separately from LTC benefits in kind – at home or in the institutions.

The migration statistics contain a number of interesting variables which could be used in the baseline scenarios. Especially variables ‘immigration by sex, age group and country of previous residence’ and ‘emigration by sex, age group and country of next usual residence’ should have an added value. By these variables the last country of residence is known, which is a much better proxy for the determination of the competent country than the variables ‘country of birth’ or ‘nationality’. One of the limitations of these variables is the lack of data for some of the MS as also a detailed breakdown by MS.

As already mentioned above, administrative data as national reports shall be added to the data received from the questionnaire. Next to it, interesting data on European level is collected. One of the tasks of the Audit Board of the Administrative Commission for the coordination of social security systems is ‘collect the necessary data and carry out the calculations required for establishing the annual statements of claims of each Member State’. 35 Information about the claims of reimbursement (based on actual expenditures or fixed amounts) of healthcare costs, which includes also LTC, is collected by the Audit Board. However, at this moment the Audit Board has only a detailed view on the outstanding claims between countries and not on the total yearly amount of claims received/issued by MS.

Some first conclusions can be made concerning the availability of information. The questionnaires launched within the framework of this study, and the collected data, show that important data is missing and that data should be collected more systematically. Particularly, the data collection on the number of insured persons (PD S1) by type of person (insured person, family member of insured person, pensioner and family member of pensioner) should be granted priority. Also, an analysis of the total yearly number and amount of claims for health care seems to be of the utmost importance. This should be taken up by the Audit Board of the Administrative Commission.

The launch of questionnaires related to the use of PD A1 (posted workers), EHIC (European Health Insurance Card – Health insurance), PD U2 (exportation of unemployment benefits) as well as the planned questionnaire on PD S2 (planned/scheduled care) by DG EMPL are important steps to collect more administrative information at EU level. Also, the launch of an electronic exchange system between national administrations could be seized as “an opportunity” to collect more data. Especially the detail of some of the structured electronic documents will be useful. E.g. the SED S080 (claim of reimbursement) offers the possibility to select data concerning the number of persons who received LTC benefits in kind and the related claim (however, only for actual expenditure). At this moment we have no information with regard to this point. For the benefits in cash a further analysis of the administrative data collected within the context of the Ageing Report 2012 might be useful.

Table 2: Overview of statistical data sources used: data availability and limitations

Source: own figure based on the data collection

Both figures below provide a first overview of the applied methodology to estimate/calculate the budgetary impact of the baseline scenario and the different options. It was the ambition to collect in 14 Member States administrative data from the competent institutions. Afterwards, the results would have been extrapolated to the EEA countries and Switzerland. However, the scale of this administrative data collection in terms of number of Member States which have responded and in terms of available data in these Member States was too limited to assess in detail the baseline scenario and the different options. As result, mainly data from the LFS, the Ageing Report and the Audit Board Report was exploited to estimate the number of involved persons and the budgetary impact. Both schemes are described in detail in chapter 4.

Figure 2: Applied methodology - LTC

Source: own figure

Figure 3: Applied methodology - UB

Source: own figure

4Baseline scenario

Before further elaborating the problems that we identified with regard to the current coordination rules, we would like to give some insights in the number of cross-border workers and pensioners in the EU-27 that are affected by the coordination rules, their main mobility patterns and their use of long-term care and unemployment benefits. Cross-border workers and pensioners – and their insured family members – are the two main groups which impact cross-border expenditure on unemployment and LTC benefits.

The table below summarises the statistics that are described and discussed in detail in the next sections.

Table 3: Synoptic overview of the scope of the cross border use of unemployment benefits and LTC benefits under scrutiny**

* Figure calculated in the interim report

** Figures described in detail in several chapters of this report

*** No data for BE, BG, HU, MT, NL, PL and RO

4.1Coordination of unemployment benefits

In the next section, we present a number of basic tables on the number of cross-border workers (incl. frontier workers) and recent migrant workers.

Based on Labour force Survey (LFS) data, an estimation of the number of cross-border workers can be made (based on the question ‘What is the name and address of the local unit of the enterprise where you work?’ and variables ‘COUNTRYW’ (country of place of work) and ‘COUNTRY’ (country of residence) in the database. These LFS data were also used in another, recent report on ‘cross-border commuting’ 36 . However, some interpretation problems appear. While legally a distinction should be made between the free movement of workers and of services, this distinction is not made by this question in the LFS. The applicable rules differ considerably between cross-border workers and posted workers. Cross-border workers will be insured in their country of employment while posted workers are still insured in their country of residence. For that reason we assumed that the LFS question covers both cross-border workers (within the rules of free movement of workers) and posted workers (within the rules of free movement of services). Ideally, the LFS should make this distinction to avoid possible interpretation problems. In the further analysis we considered all workers who work in another country than the country of residence as cross-border workers. Workers who work in a neighbouring country are considered as frontier workers (as also is assumed in the report ‘Mobility in Europe 2011’) (which is not equal to the legal definition of a ‘frontier worker’) 37 .

Table 4 gives an overview of the number of cross-border workers 38 and frontier workers 39 in the EU-27. The average of 2010 and 2011 is calculated to avoid outliers. To guarantee reliability of the figures we only took into consideration the row and column totals of the cross-tables. However, the details in the different cells are important to estimate the baseline scenario and the different options (e.g. we need to select the cells which describe flows between neighbouring countries to define the number of frontier workers) 40 . The national employment figures (living and working in the same country) are yellow coloured while the figures coloured in red are the neighbouring countries of a specific country.

The table shows that on average 1 million cross-border workers are employed in the EU27 or 0.5% of the total employed population, of which on average 701.000 frontier workers are employed in a neighbouring country. This implies that 68% of the cross-border workers can be assumed as frontier workers. These figures are similar to the results in the ‘Mobility in Europe 2011’ Report which reports “that just 5 people in 1,000 of those employed commute across borders between EU Member States” and “some 63% of cross-border commuters go to work in a bordering country” 41 .

In general, we observe an increase of the number of cross-border workers between EU-27 MS of 1.4% between 2010 and 2011 (see Table 81 and Table 82 ). The Report ‘Mobility in Europe 2011’ already observed this increase for earlier years (with a slowdown for more recent years).

Outgoing cross-border workers

In absolute figures, most outgoing cross-border workers live in Germany (165 thousand), France (160 thousand) and the Slovak Republic (117 thousand). However, in terms of share in national employment (see Table 83 ), the highest impact is observed in the Slovak Republic (5.0%), Estonia (3.1%) and Belgium (2.2%). For Belgium (97%), the Netherlands (95%) and Ireland (88%) most of the outgoing cross-border workers are employed in a neighbouring country. While for Romania (1%), Lithuania (5%) and Latvia (9%) few of the outgoing cross-border workers are employed in a neighbouring country. As Croatia joined the EU on 1 July 2013 it is interesting to look at the number of cross-border workers of this country. For 2011, 20.6 thousand outgoing cross-border workers from Croatia were counted primarily going to Italy and Germany (see Table 81 ).

Incoming cross-border workers

Most incoming cross-border workers are employed in Germany (186 thousand), Luxembourg (130 thousand) and the Netherlands (110 thousand). The highest impact of incoming cross-border workers on the national employment is identified for Luxembourg (37.4%), Austria (2.4%) and Belgium (1.4%). Most of the incoming cross-border workers in Luxembourg (99%), the Czech Republic (98%), Austria (94%) and the Slovak Republic (94%) are living in a neighbouring country. Only 6.6% of the incoming cross-border workers employed in Italy, 10% in Romania and 11% in the UK are living in a neighbouring country. A very popular country for incoming cross-border workers is Switzerland. In 2011, Switzerland employed 320 thousand incoming cross-border workers mainly coming from France (see Table 81 and Table 82 ).

As it is possible that the LFS data also include posted workers in the number of cross-border workers, the profile of the cross-table of cross-border workers from the LFS is compared with this from the number of PD A1 certificates issued (certifies which social security legislation applies to the holder of the form). This was done by calculating the row and column percentages and indicating the 3 MS with the highest percentages.

First, we compare the data of the outgoing cross-border workers with these from the outgoing posted workers (See Table 84 and Table 87 ). Germany (14 times), the Slovak Republic (8 times) and Poland (8 times) appear most frequently in the top-3 of ‘sending’ MS concerning cross-border workers. At the same time most important (in top 3) sending MS for posted workers are Germany (26 times), Poland (22 times) and France (20 times). However, the spread of sending countries is somewhat more diverse for cross-border workers compared to posted workers (more concentrated in Germany, Poland and France as sending countries).

Also the spread over MS receiving incoming cross-border workers was compared with this for MS receiving incoming posted workers (see Table 85 and Table 88 ). Germany (17 times) and the UK (12 times) are the most important working countries for cross-border workers while posted workers mostly are sent to Germany (18 times), France (12 times) and the Netherlands (10 times).

The relationship between the two variables can be calculated by the correlation coefficient. 42 We see a strong positive linear relationship (0.76) between the variables ‘incoming posted workers’ and ‘incoming cross-border workers’, which implies the same MS receive as well cross-border workers as well as posted workers, probably to be considered as attraction pools of mobile workers. Besides, there is also a strong positive linear relationship (0.71) between the ‘outgoing posted workers’ and ‘outgoing cross-border workers’.

The possible number of PD U1 issued is also influenced by recent migrant workers. Eurostat migration statistics are collecting the number of migrant workers (from 15 to 64 years) who have moved from one EU-country to another in 2011 ( Table 5 ). One of the limitations is the lack of figures for BE, BG, HU, MT, NL, PL and RO.

For those countries of which figures are available, we observe most migrant workers migrated to Germany (237 thousand in 2011), the UK (185 thousand in 2011) and Spain (54 thousand in 2011).

This table shows also the importance of Romania as emigration country. Also here we have calculated the relationship between ‘incoming migrant workers’ and ‘incoming cross-border workers’. We observe a strong positive linear relationship (0.70) between both variables, which implies that the same MS are dominant or less dominant. Within this group of recent migrant workers the unemployment rate of their current country of employment has been used to estimate the number of unemployed recent migrant workers.

Table 4: Number of cross-border workers and frontier workers, EU-27, Average of 2010 & 2011, in .000

* Row and column totals which lay above the reliability limits are indicated with a *. Please take notice that some of the figures mentioned in the row and column totals are not reliable. This will specifically be the case for the detailed cells. For an overview of the LFS reliability limits see Annex 10.7: http://circa.europa.eu/irc/dsis/employment/info/data/eu_lfs/LFS_MAIN/Related_documents/reliab_annual_average.htm  

** Red colour: frontier workers; yellow colour: national employment; ‘EU-27’= total employment

Source: own calculations based on LFS

Table 5: Migrant workers (from 15 to 64 years) in 2011 by country of previous residence, EU-27

* Total= sum of migrant workers of which the current country of residence is known

Source: Eurostat Migration Statistics, [migr_imm5prv]

Making the breakdown between unemployed cross-border workers and migrant workers is important for different reasons. First, it gives a first impression of the impact both categories have on the number of issued PD U1 certificates and the budgetary impact on the unemployment expenditure. Second, Member States can only claim a reimbursement for the unemployment benefits they have paid to cross-border workers.

National unemployment rates are applied on the number of cross-border workers. 43 The national unemployment rates of 2010 (from 20 to 64 years) defined in the 2012 Ageing Report were used. Also, the unemployment rates of the country of employment and not of the country of residence have been applied on the number of cross-border workers calculated by way of the LFS. This results in an estimation of 73.7 thousand unemployed cross-border workers of which 45.2 thousand frontier workers ( Table 6 ). These figures will be used to estimate the expenditure as well as the claimed reimbursement of the provision of unemployment benefits to cross-border workers taking into account the baseline scenario and the policy options.

The same unemployment rates are applied on recent migrant workers. As no data are available for BE, BG, HU, MT, NL, PL and RO no exhaustive view on the number of unemployed recent migrant workers can be obtained. When we extrapolate these figures for the EU-27, 128 thousand unemployed migrant workers are counted ( Table 7 ). The chance is rather high that this group will need a PD U1 certificate to prove periods of insurance or (self)-employment in another country.

By counting the estimated number unemployed cross-border workers ( 4.1.1.4 ) and migrant workers ( 4.1.1.5 ) together, we become a total result of 202 thousand unemployed persons who may need a PD U1 certificate to prove periods of insurance of (self)employment from another country. This seems rather a minimum estimation compared to our other estimation of unemployed persons who will issue/receive a PD U1 certificate, namely 341 thousand unemployed persons (see Table 8 ).

Table 6: Estimated number of cross-border workers who will become unemployed, in .000

* Row and column totals which lay above the reliability limits are indicated with a *. Please take notice that some of the figures mentioned in the row and column totals are not reliable. This will specifically be the case for the detailed cells. For an overview of the LFS reliability limits see Annex 10.7: http://circa.europa.eu/irc/dsis/employment/info/data/eu_lfs/LFS_MAIN/Related_documents/reliab_annual_average.htm  

** Red colour: frontier workers; yellow colour: national unemployment; ‘EU-27’= total unemployment

Source: Estimate based on LFS and 2012 Ageing Report

Table 7: Estimated number of recent mobile workers who will become unemployed

* The unemployment rate of 2010 (Ageing Report) was multiplied by the number of recent migrant workers (Eurostat migration statistics)

** Row lable: current residence and column lable: previous residence

Source Estimate based on Eurostat migration statistics and 2012 Ageing Report

When making the decision to grant the unemployment insurance benefit all completed unemployment insurance periods in other member countries shall be taken into consideration. The unemployment insurance periods completed in different countries are then to be aggregated. In the following paragraphs, we aim to assess the extent to which periods of insurance and (self-) employment in another EU Member State were taken into account when granting unemployment benefits. Therefore, the number of PD U1 or E301 forms received and issued was counted.

Ideally, a distinction should be made between frontier workers, ‘other’ cross-border workers and migrant workers:

Frontier workers return to their country of residence daily, or at least once a week.

‘Other’ cross-border workers return to their country of residence less than once a week.

Migrant workers worked and lived already in the competent MS before their unemployment but will prove completed periods of insurance or (self-)employment in another Member State.

For only some EU Member States, it was possible to make this distinction.

Furthermore, in our original questionnaire, we aimed to find data which would make it possible to distinguish between the number of PD U1/E301 forms received each year and the number of ‘unique persons’ behind these forms. This distinction might be relevant, as a single person can submit more than one form during a single year. However, in this analysis both questions are discussed together in the ambition to extrapolate to an EU-27 level.

Also, we did not ask specifically for the number of PD U1/E301 forms issued by each EU Member State(questionnaire was limited to 14 MS). Nevertheless, it was possible to obtain statistics for some other EU Member States and these are included in this report.

We compared the number of documents received and the number of documents issued. This for two specific reasons: firstly, it enabled us to fill gaps in the data on the PD U1/E301 forms received. Secondly, it allowed us to cross-check the data collected (e.g. whether the number of PD U1/E301 received forms from a specific EU Member State equals the number of PD U1/E301 forms issued by the other EU Member State).

Below, results from the questionnaire regarding the aggregation of periods are summarised. It will be mentioned explicitly in the text when another source than the questionnaire is used. Also, we will refer to the different tables added in Annex of this report. The number of received PD U1 certificates will be compared to the total number of unemployed persons. For this, data from the ‘unemployment – LFS adjusted series’ was used. An unemployed person is defined in the LFS as “a person 15 to 74 years of age (16 to 74 years in ES, IT and the UK) who was not employed during the reference week, had actively sought work during the past four weeks and was ready to begin working immediately or within two weeks” (Eurostat) 44 .

Number of PD U1/ E301 forms received

Data on the number of PD U1/ E301 forms received (or ‘unique’ persons involved) was collected for Belgium, Estonia, France, Poland, Romania, Slovak Republic, Sweden and UK. Only for four countries the distinction between frontier workers, cross-border workers and migrant workers could be made: France, Poland, Romania and Sweden. For the other EU Member States, it was not possible to obtain data.

In 2012, the competent institutions in Belgium took 2,400 PD U1 forms into account in granting an unemployment benefit (cf. Table 59 and Table 61 ). This is not the actual number of PD U1’s received by the Belgian competent institution; which shall be (much) higher. Only the PD U1’s which were used to grant an unemployment benefit because of a too short period of insurance or (self-)employment in Belgium are counted. Most of the PD U1 documents taken into account were for periods of insurance and (self-)employment completed in neighbouring countries, primarily the Netherlands (45.5%) and Luxembourg (31.9%), but also France (5.4%) and Germany (4.9%). The figure for Spain was also relatively important (4.9%). As the average number of unemployed in Belgium in 2012 was 369,000, this implies that only 0.7% of Belgium’s unemployed made use of the PD U1 document to receive an unemployment benefit.

Between 2010 and 2012, the number of PD U1/E301 forms received by Estonia increased by 38% (from 1,505 to 2,082) (cf. Table 59 and Table 61 ). Of these, 53.5% were issued by Finland (in 2012). The proportion of the unemployed in Estonia providing a PD U1 to receive an unemployment benefit in 2012 was 2.9%.

In 2011, 50,003 people were granted an unemployment benefit in France which took periods from other EU Member States into account (see CLEISS (2012), Rapport Statistique 2011) 45 . Of this group, 49,961 were in the cross-border worker category and 42 were migrant workers (however, based on article 61 Regulation (EC) No. 883/2004 and not art. 65) 46 . As in the case of Belgium, most of the PD U1/E301 forms received were issued by neighbouring countries, in particular Switzerland (48.2%), but also Luxembourg (20.7%), Germany (15.1%) and Belgium (14.8%). The percentage of the total average number of unemployed in France in 2011 providing a PD U1 form was 1.8%.

Poland received a PD U1 for 13,884 cross-border workers (77.8% of the total) and 3,980 migrant workers (22.2% of the total) in 2011 ( Table 59 , Table 60 and Table 61 ). The forms were primarily issued by the United Kingdom (42.3%), the Netherlands (18.6%), Ireland (10.2%) and Germany (9.2%). In 2011, 1.2% of Poland’s unemployed provided a PD U1 to receive an unemployment benefit. For Poland, as well the number of PD U1 documents received as the number of ‘unique persons’ concerned are known. One person received 1.1 PD U1 documents (19,432 forms compared to 17,481 ‘unique’ persons).

For Romania, data was available only on the PD U1 documents delivered by migrant workers. In 2012, 92 forms were received, of which 66% applied to Spain (Table 36).

It was estimated by the national competent institution that the United Kingdom received some 300 PD U1/E301 forms ( Table 59 and Table 61 ). Compared to the data from other Member States, this seems a rather low figure. It means that only 0.012% of the UK’s unemployed would have provided a PD U1 form.

In 2012, the Slovak Republic received 10,912 PD U1 forms ( Table 59 and Table 61 ). Most were issued by the Czech Republic (27.9%), United Kingdom (22.1%) and Hungary (15.9%). The Slovak Republic ranked with Estonia at the top end of the range of the percentage of the unemployed receiving unemployment benefit based on evidence from a PD U1, i.e. 2.9%.

Finally, Sweden received 2,202 PD U1/E301 forms in 2011 ( Table 60 and Table 61 ). Of these, 1,762 were from cross-border workers (80%) and 427 were from migrant workers (20%). The forms mainly came from Norway (42.8%) and Denmark (36.6%). Of the total number of unemployed in 2011, 0.6% had proven periods of insurance or (self-)employment in other Member States to confer unemployment rights in Sweden.

Number of PD U1/E301 forms issued

Data on the number of PD U1/E301 forms issued were collected for Belgium, Estonia, Luxembourg and Romania ( Table 64 ).

Belgium issued 11,522 PD U1 forms in 2011 ( Table 64 ). No breakdown by receiving EU Member State is available, but there is data by nationality. Of the total, 22.9% were issued to Belgian nationals. According to a recent report (Pacolet, et. al., 2012) 47 which sought to provide a detailed profile of French, Belgian and German frontier workers and their knowledge and use of, and satisfaction with social security benefits (especially sickness benefits), 33% of the frontier workers living in France and working in Belgium are born in Belgium. However, a PD U1 might also be issued for a migrant worker coming from Belgium but living and working in another Member State. Of the PD U1’s issued by the Belgian authorities, 61.2% were for French nationals.

Between 2010 and 2012, the number of PD U1 forms issued by Estonia increased by 45.3% (from 428 to 622 forms) ( Table 64 ). Of these, 73% were used to request an unemployment benefit in Finland.

The number of PD U1/E301 forms issued by Luxembourg is indicative of the popularity of this country for mobile workers ( Table 64 ). In 2012, Luxembourg issued 18,875 PD U1 forms, an increase of 5.9% compared to 2010. A high percentage of the forms were received by the French authorities (63.1%). An important number also went to Belgium (15.9%) and Germany (15.4%).

For Romania, we were only able to obtain the number of PD U1 forms issued for migrant workers. In 2012, this was 5,067 ( Table 64 ).

Estimated number of PD U1 to be received: mid-term (2015) and long-term projection (2020)

Data on the total number of PD U1 forms received was obtained from six EU Member States (Belgium, France, Estonia, Poland, Slovak Republic and Sweden) by way of the launched questionnaire. The proportion of unemployed (LFS definition ‘Average annual unemployed persons (from 14 to 75 years)’) with proven periods of insurance or (self-)employment from other Member States entitling them to unemployment rights is for this group of countries on average 1.6%.

Based on the number of received PD U1 forms by 6 Member States and data from the 2012 Ageing Report, we aimed to make projections of the baseline scenario for 2015 and 2020. For these calculations we will use the number of unemployed persons between 20 and 64 years calculated for 2010 and projected for further years (which differs from the definition used above for an unemployed person (from 14 to 75 years)). From the data we have collected in these Member States, we learned that in general 1.6% of the unemployed persons (from 20 to 64 years) received a PD U1 to prove a period of insurance of (self-)employment from another Member State. This percentage was applied to the total number of unemployed persons residing in Member States for which we found no data. By counting together the survey data together with these estimates, we estimate the total number PD U1 on 341 thousand issued certificates.

If we take into account the unemployment projections for 2015 and 2020 (between 20-64) – found in the 2012 Ageing Report and calculated by applying an unemployment rate to the labour force - the absolute number of issued PD U1 is expected to decrease to 324 thousand forms by 2015 and to 300 thousand forms in 2020. However, here we take only the evolution of the number of unemployed persons into account. Besides, also other variables can/will change in the future (e.g. share of PD U1 compared to total unemployment).

Table 8: Estimated number of received PD U1 in the EU-27: 2010, 2015 & 2020

Source: Estimate based on collected administrative data and 2012 Ageing Report

Estimated share of actual expenditure on unemployment benefits based on PD U1 in total expenditure on unemployment benefits (2010)

It is interesting to see how much is paid to those in unemployment with proven certified periods from other Member States and how much of global national expenditure on unemployment benefits this amount represents. We received data on this from only three Member States: Belgium, France and Sweden ( Table 63 ). More detail (i.e. a distinction between cross-border workers and migrant workers) was available for France and Sweden.

In Belgium, € 10.5 million was paid (in 2012) to those who needed a PD U1 to receive a Belgian unemployment benefit, or 0.2% of the total expenditure on full unemployment benefits ( Table 63 ). Of this, 47.8% was paid to unemployed who proved a period of insurance or (self-) employment in the Netherlands.

France paid € 516.8 million in 2011 to people in unemployment who aggregated their periods (mostly cross-border workers) or 1.8% of total expenditure ( Table 63 based on CLEISS – Rapport Statistique 2011) 48 . Of this, 59% of the amount was paid to those with a PD U1 issued by Switzerland.

Finally, Sweden paid € 22.6 million to this group in 2011. Of this, 90.4% was paid to former cross-border workers and 9.6% to migrant workers. Overall, this was 1.1% of total expenditure on full unemployment benefits. Of the total, 48.8% was paid to unemployed who proved insurance periods or (self-)employment in Denmark.

Estimated annual expenditure on unemployment benefits based on PD U1: projections for 2015 and 2020

The maximal annual amount paid to unemployed persons (if we assume that the unemployed did not find a job during the first year of unemployment) who received a PD U1 to prove an insurance period or (self) employment from another Member State can be counted by multiplying the number of estimated PD U1 with the unemployment benefit per unemployed person (in 2010 prices; projected in the 2012 Ageing Report). For 2010, a maximal cost of € 2.07 billion was found. The cost is expected to increase (in absolute figures) to € 2.19 billion in 2015 and to € 2.22 billion in 2020 (in prices of 2010), driven by an increase of the unemployment benefit per unemployed person. However, the breakdown by unemployment cross-border workers and migrant workers is not available. This will be needed to calculate the amount of reimbursement claims between Member States in the baseline scenario and in the other policy options.

Table 9: Yearly estimated amount paid to unemployed persons who received a PD U1 2010, 2015 & 2020 (in EUR)

Source: Estimate based on estimated number of PD U1 and 2012 Ageing Report

The PD U2 is the authorisation which an unemployed person needs to export his/her unemployment benefit if (s)he wishes to move to another EU country to look for work. The competent national institution is responsible for granting this authorisation.

Statistics on the extent to which requests to export unemployment benefits are accepted or refused are available for Sweden and Poland. In 2012, Sweden refused 37% of the requests for a PD U2 ( Table 65 ). In 2011, only 19% of the requests were refused. The Polish competent institution refused 23% of the requests (in 2011). No further information could be collected on the proportion of persons who requested a prolongation of the export period to 6 months (the acceptance/refusal ratio).

For ten EU Member States (Austria, Belgium, Denmark, Estonia, Luxembourg, the Netherlands, Poland, Slovak Republic, Sweden and United Kingdom), data were received on the number of PD U2 issued ( Table 65 ).

Austria issued 1,186 PD U2’s (in 2012) (detail by Member State is not available) ( Table 65 ). This is 0.63% of the average annual unemployed persons.

Belgium issued 1,081 PD U2’s (in 2011), of which most for France (46.3%) but also 16.9% for Spain ( Table 65 ).

Denmark provided 1,108 authorisations (in 2011) to seek a job abroad of which 18.9% for the United Kingdom, 12.5% for Germany and 12.4% for Spain ( Table 65 ).

Estonia issued 64 PD U2’s (in 2012), which was a decrease of 7.2% compared to 2011 (69 forms) ( Table 65 ).

Luxembourg issued 148 PD U2’s (in 2012), mainly to seek a job in France (32.4%) and Portugal (24.3%) ( Table 65 based on Report of ‘Le gouvernement du Grand-duché de Luxembourg. Agence pour le développement de l’emploi – Les activités de l’agence pour le développement de l’emploi en 2012’) 49 . The number of PD U2’s issued decreased by 16.1% in 2011 compared to 2010, but increased again in 2012 by 28.7% compared to 2011. In 2012, the number of authorisations granted in Luxembourg to look for a job in another Member State was 1.14% of the average annual number of unemployed persons. This is the highest percentage in our sample of described Member States.

Poland issued 118 PD U2’s (in 2011), mainly for Germany (34.7%) and the United Kingdom (33.9%) ( Table 65 ).

The competent institution in the Netherlands issued 637 PD U2’s in 2012, of which 18.2% both for Germany and for Austria ( Table 65 based on report UWV – Kwantitatieve informatie 2011).

Romania issued 11 PD U2’s in 2012 ( Table 65 ).

The Slovak Republic issued 79 certificates in 2012, mainly to look for a job in the Czech Republic, Germany and the United Kingdom ( Table 65 ).

In 2012, Sweden issued 264 PD U2/E303 forms ( Table 65 based on report IAF) 50 . This was 11% fewer than in 2011. Most of the forms were for those wanting to look for a job in the United Kingdom (15.4%) or Spain (14.7%). More detail about trends in the number of PD U2/E303 forms issued and the profile of the jobseeker is available for Sweden. There has been a strong decrease in the number of PD U2/E303 forms issued in last decade. E.g. in 2004, even 1,036 E303’s were issued. 87.4% of the unemployed who received a PD U2 form (between May 2010 and August 2011) actually went abroad to look for a job. These jobseekers were mainly aged between 30 and 39 (44%). More unemployed women (58%) than men (42%) went looking for a job abroad. 46% of those with a PD U2 who were not born in Sweden were seeking a job in their country of birth.

Finally, it was estimated by the competent national institution that the United Kingdom issued about 300 PD U2’s ( Table 65 ).

For the EU-27, an estimate of the number of PD U2’s issued can be made by using the statistics collected together with data from Eurostat (LFS definition ‘Average annual unemployed persons (from 14 to 75 years)’). We have the total number of PD U2’s issued for 10 EU Member States, collected by way of launching a questionnaire. This suggests that on average only 0.11% of the unemployed looked for a job abroad. Extrapolating this percentage to arrive at an estimate of the total number of PD U2’s issued in the EU-27 each year the resulting estimation is an annual issue of 23,700 PD U2’s.

Policies on refusal/approval of requests for a prolongation vary considerably between Member States. For example, Sweden refused all 35 requests for prolongation in 2011, while Poland in the same year accepted all 20 requests for prolongation ( Table 66 ).

The main aim of the possibility of exporting unemployment benefits under the current rules is that the unemployed person actually finds a job abroad. A proxy for the success rate could be calculated for the unemployed from Belgium, Poland and Sweden who have moved abroad (export PD U2) and for the unemployed from other Member States who have sought a job in the Netherlands (import PD U2) ( Table 67 ). Of the unemployed in Belgium who received a PD U2, 44% actually found a job abroad. The success rate for the unemployed people coming from Belgium in the two most popular Member States for looking for a job, namely France and Spain, was 46% and 36%. However, this should be considered as a broad definition of the success rate for Belgium. It is based on the number of people who are no longer registered within the Belgian National Employment Office after their period of export. This might be for different reasons: they have found work in the country of export, they have found work in Belgium, or they moved to some other country. The success rates for Poland and Sweden are 10% and 12%. For the incoming jobseekers in the Netherlands 51 , a success rate of 22.8% was obtained. So the rates are highly variable.

Figures were available for the number of PD U2’s received in five Member States (Estonia, France, Luxembourg, the Netherlands and Sweden) ( Table 68 ).

In 2012, Estonia received 41 PD U2’s, mainly from Ireland. This is an increase of 17.1% compared to 2011 (35 forms received).

For France, only the number of received E303 forms in 2011 is available, mainly issued by Switzerland and Norway ( Table 68 based on CLEISS – Rapport Statistique 2011) 52 . 

Luxembourg received 171 certificates for jobseekers in 2012, many of them coming from Portugal (based on Report of ‘Le gouvernement du Grand-duché de Luxembourg. Agence pour le développement de l’emploi – Les activités de l’agence pour le développement de l’emploi en 2012’) 53 . The number of PD U2 received by Luxembourg has increased in recent years, mainly due to a spectacular increase in the number of jobseekers coming from Portugal.

In 2012, the Netherlands received 483 PD U2’s, mostly from Spain.

Finally, Sweden welcomed 691 jobseekers from other Member States. The number of persons seeking a job in Sweden has fluctuated strongly over the years (lowest in 2000 and highest in 2004). 54  

Observing the row totals for the number of issued PD U2 forms ( Table 65 ), France, Germany, Spain and the United Kingdom seem to be the most popular Member States in which to look for a job.

Statistical data about the number of PD U2 issued was already obtained for 10 EU Member States. This administrative information could be compared by the number of unemployed persons between 20 and 64 years calculated for 2010 and projected for further years (which differs from the definition used above for an unemployed person (from 14 to 75 years)). On average 0.11% of the unemployed persons in the countries of which administrative information is available moved abroad to seek work and export their unemployment benefits. This percentage was applied to the number of unemployed persons (20-64) in the other 17 EU Member States.

Taking together both components (survey data and estimates) we estimate that 23.7 thousand unemployed persons have exported their unemployment benefits in 2010. Based on the projections of the 2012 Ageing report, the number of jobseekers moving abroad would decrease to 22.8 thousand unemployed persons in 2015 and 21.2 thousand unemployed persons in 2020.

Table 10: A projection of the number of export unemployment benefits (PD U2): 2010, 2015 & 2020

Source: Estimate based on collected administrative data (cf. Annex 10.1.1 – UB) and 2012 Ageing Report

Jobseekers can create a CV in EURES and make it available for registered employers and EURES advisers helping employers to find suitable candidates. Those people generally still reside in their origin country. In June 2013 about 1 million jobseekers were registered on the EURES Portal (= current stock). Most registered jobseekers live in Spain, Italy and Portugal. As this number is the current stock of registered jobseekers it is necessary to have also an overview of the yearly new registered jobseekers. This could be a proxy of the total number of (unemployed) jobseekers looking for a job abroad but not a good proxy of people looking for a job in another country in which they already established. Between June 2012 and June 2013 about 274 thousand new persons registered on the EURES Portal. We estimated that about 23.7 thousand unemployed persons receive a PD U2 which is 8.7% of the total number of ‘new’ registered jobseekers in EURES.

Table 11: Number of jobseekers registered in EURES cv online and comparison with estimated number of PD U2 certificates issued, by country of residence, in .000

Source: Own calculations based on estimate EU-27 and EC, EU Employment and Social Situation – Quarterly Review June 2013.

In order to estimate the budgetary impact of the baseline scenario, the estimated number of unemployed cross-border workers based on the LFS and the unemployment rates of the Ageing Report is multiplied by the annual unemployment benefit per unemployed person (in 2010 prices; projected in the 2012 Ageing Report). For each of the flows between Member States (in the different cells) the unemployment benefit of the country of last activity and the unemployment benefit of the country of residence was taken into consideration (see  Table 89 ).

Under current rules unemployed frontier workers must claim unemployment benefits in the country of residence while unemployed other cross-border workers can choose to claim unemployment benefits in the country of last employment or in the country of residence. Due to fact the other cross-border worker can choose, an assumption has to be made about how many of them return to the country of residence and how many stay in the country of last activity. We assume first that 50% of the other unemployed cross-border workers will return and 50% will stay. An alternative assumption could be based on the rational decision which Member State (country of last activity or country of residence) is paying the highest unemployment benefit.

The actual total yearly expenditure is estimated based on the unemployment benefit per unemployed person in prices 2010 (unemployment benefit spending in 2010 prices / (labour force * unemployment rate)) taken up in the 2012 Ageing Report as also the estimated unemployed cross-border workers. This yearly expenditure assumes that the unemployed person did not find a job during the first year of unemployment. 55 A breakdown between the expenditure by the country of residence ( Table 43 ) or the country of last activity ( Table 44 ) is made. We also refer to bilateral expenditure between countries.

The baseline scenario (option A) whereby frontier workers have to return to the country of residence and other cross-border workers can choose between the country of residence or the country of last activity involves a yearly expenditure of € 540.0 million of which € 392.4 million related to frontier workers when assuming that 50% of the other cross-border workers are returning to their country of residence ( Table 89 - baseline scenarioA1). This involves that 81% of the yearly expenditure on cross-border workers will be paid by the country of residence and 19% by the country of last activity. 56 The specific expenditure for frontier workers will be fully covered by the country of residence.

Assuming for the baseline scenario that other cross-border workers choose on the basis of the amount of the unemployment benefit (= rational decision) (baseline scenarioA2), involves a yearly expenditure of € 638.5 million or an increase of 18.2% compared to the first baseline scenario ( Table 90 ). 70% of this expenditure shall be paid by the country of residence while 30% by the country of last activity ( Table 90 ). It implies that in more than 50% of the cases it is more interesting to claim an unemployment benefit in the country of last activity.

The unemployment benefits paid by the country of residence and the country of last activity do not completely reflect the burden sharing of unemployment benefits. Also the amounts of reimbursement should be taken into account.

Claims can be made by the country of residence to the country of last activity for fully unemployed frontier workers but also for other cross-border workers who have decided to register with the competent institution in their country of residence. The country of last activity shall reimburse the unemployed benefits provided in the country of residence during the first three months or five months (when the unemployed person during the preceding 24 months, completed at least 12 months of (self)employment in the country of last activity). Data were collected on the number of claims received (as debtor) and the number of claims issued (as creditor) ( Table 71 and Table 72 ).

It was possible to describe the position as debtor (from request until receipt of reimbursement) in the case of six Member States (Belgium, Denmark, Germany, Poland, Romania and Slovak Republic). This means that we have data on the cross-border workers who had worked in these Member States but are receiving an unemployment benefit in their current country of residence.

In 2011, Belgium received reimbursement requests for an amount of € 11.3 million or 3,664 cases ( Table 71 ). France accounted for 84.6% of the cases and 73.8% of the amount requested. This is not surprising when we look at the number of PD U1 forms issued by Belgium.

Denmark received 1,637 reimbursement requests, of which 67.5% came from Germany ( Table 71 ).

There was no detailed information for Germany, but the comment was made by the competent institution (received by the national expert) that some 7,000 reimbursement requests had been received (in the second semester 2011 and first semester 2012) ( Table 71 ). 57  

Poland only received 84 claims, for an amount € 207,000, mainly issued by Germany ( Table 71 ). Further detail on the treatment of the reimbursement request is available for Poland. Poland partially accepted 88% of the reimbursement requests, 4% were not accepted and 7% were entirely accepted. This might due to the fact that the amount of reimbursement cannot be higher than the amount payable under the legislation of the country of last activity (see art. 65, 6 Regulation (EC) No. 883/2004). 58 E.g. when we compare the ‘Unemployment benefit per unemployed person in 2010 prices’ calculated in the 2012 Ageing Report we find that the yearly unemployment benefit in Germany (€ 8,919) is much higher than in Poland (€ 397). The impact of this legal boundary will be discussed further in this report.  

In 2011, Romania received reimbursement claims for an amount of € 49,167 ( Table 71 ). Most of the claims were sent by Germany. Romania partially accepted 54% of the reimbursement requests, while 45% were not accepted and only 1% entirely accepted. Also here will the maximum amount of reimbursement play an important role as the unemployment benefit paid in Romania will be in some cases lower than the unemployment benefits paid in other Member States.

Finally, in 2011 the Slovak Republic received reimbursement requests for an amount of € 102,000 ( Table 71 ).

The position as creditor can be described for the same six Member States ( Table 72 ). This means that we collected data on cross-border workers receiving unemployment benefits in these Member States after having worked in other Member States.

In 2011, Belgium sent 3,353 reimbursement requests for an amount of € 8.7 million ( Table 72 ). Most of these claims were sent to neighbouring countries (France, Germany, Netherlands and Luxembourg).

Denmark only issued eight reimbursement requests ( Table 72 ).

No detail was available for Germany. However, the relevant German institution sent yearly about 16,000 reimbursement requests (second semester 2011 – first semester 2012) ( Table 72 ). See for the German case the comments made as debtor (footnote). 

Poland issued 7,599 reimbursement requests in 2011 for an amount of € 4.7 million ( Table 72 ). Most of the Polish claims were received by the United Kingdom (41% of cases) and the Netherlands (23% of cases). The debtor country rejected 71% of these Polish claims, while 29% of the claims were entirely accepted. The rejection rate is mainly influenced by the decision of the United Kingdom to reject all claims (644) from Poland.

Romania sent a total amount of € 33,000 of reimbursement claims, mainly to Germany and Luxembourg. Of Romania’s claims, 76% were entirely accepted and 24% not accepted ( Table 72 ).

Finally, the Slovak Republic claimed an amount of € 2.4 million, of which 42.9% was sought from the Czech Republic ( Table 72 ).

The extent to which Member States experience labour mobility will have an impact on the number and amount of reimbursement claims.

For example: The number of incoming cross-border workers in Belgium is high, but the number of outgoing cross-border workers is even higher ( Table 83 ). This high degree of labour mobility results in high levels of reimbursement requests sent and received. The amount of reimbursement requested (€ 8.7 million in 2012) will to a large extent compensate for the unemployment benefits paid out by the Belgian competent institution to former cross-border workers (€ 10.5 million in 2012). Germany, Poland and the Slovak Republic have a high number of outgoing cross-border workers which results in a much higher number/amount of claims issued compared to claims received. At the same time, Denmark and Romania attract more cross-border workers, which results in a higher amount/more case of claims received. The reimbursement procedure mainly affects Luxembourg and Switzerland, which have a high number of incoming cross-border workers (see  Table 81 , Table 82 , and Table 83 ).

A breakdown between claims for 3 months or 5 months is not available in the data of the launched questionnaire. 59 None of the responding countries could make this distinction in our questionnaire. In order to make an estimate, we will assume 3 months of claims (minimum scenario). Another crucial element which we should take into account for the calculation is the fact that the amount of reimbursement by the country of last activity is limited to the maximum unemployment benefit the unemployment person would receive in case of unemployment in the country of last activity. 60 For this exercise, we have multiplied the estimated number of unemployed cross-border workers based on the LFS and the unemployment rates of the Ageing Report by the unemployment benefit per unemployed person for three months (in 2010 prices; projected in the 2012 Ageing Report). For that reason the figures concerning the estimated yearly expenditure on unemployment benefits for unemployed cross-border workers should be read together with the figures dealing with the estimated reimbursement claims to assess who is sharing the burden of unemployment.

For each of the flows between MS (in the different cells) the unemployment benefit of the country of last activity and the unemployment benefit of the country of residence is taken into consideration (see also Table 95 ). The unemployment benefit of the country of residence will be used to calculate the claim of reimbursement. Also, this claim will be compared with the actual reimbursement taking into account the rule that the reimbursement cannot be higher than the amount payable, in the case of unemployment, under the legislation of the country of last activity (=maximum amount). The actual reimbursement will be equal to the claim when the unemployment benefit in the country of residence is lower than or equal to the unemployment benefit in the country of last activity. The actual reimbursement will be equal to the unemployment benefit in the country of residence (=maximum amount) when the unemployment benefit in the country of residence is lower than the country of last activity ( Table 95 - Table 99 ).

The baseline scenario A1 whereby frontier workers have to return to the country of residence and other cross-border workers can choose between the country of residence or the country of last activity involves a claim of € 108.8 million of which € 98.1 million for frontier workers when we assume that 50% of the other cross-border workers return to their country of residence ( Table 95 ). However, these claims are based on the unemployment benefits paid by the country of residence and not on the maximum payable amount by the country of last activity. It implies for this baseline scenario that the reimbursement will be 24% lower than the possible actual claim (see baseline scenario A1a - Table 96 ).

If other cross-border workers are making a rational decision on the basis of the amount of the unemployment benefit (baseline scenario A2a - Table 97 & Table 98 ), the claim will increase with 3% compared to the baseline scenario A1a whereby 50% of the other cross-border workers are returning to their country of residence.

In order to allow the stakeholders to identify the time spent on the information obligations related to the Regulations, we have defined prior to our visits in the Member States a standard legal process stemming from the Regulations, in cooperation with the Commission.

During our first visits, we noticed several issues concerning this process:

National administrations have developed their own administrative processes for processing/handling documents related to cross-border cases for unemployment benefits and long-term care. These differ substantially between the Member States. As a result, the experts in the respective countries faced difficulties in plugging the suggested administrative processes into their national way of working (processing documents);

The legal process encompassed several sub-administrative processes and documents and therefore Information Obligations (IOs). The complexity of the different processes proved to be an obstacle in making precise estimations of the (estimated) time spent for each of the processes. The experts were often not able to provide robust data on the time spent per each of the steps defined by the legal process.

Moreover, as the Regulations impose “principles” of coordination more than specific information obligations in the sense of the SCM, and as the principles were already applied partly or integrally by the administrations or applied still differently, it proved to be impossible for the stakeholders to differentiate the specific administrative burden 61 created by the Regulations from the business-as-usual (the administrative tasks they would perform anyway in the absence of the Regulations).

Another consequence of the nature of the Regulations is that each national process is different, meaning that it results in different requirements, documents, times and complexity. It makes it impossible to standardize one process that fits all national specificities.

There are examples demonstrating the complexity of the processing of cross-border cases for unemployment benefits which can result in administrative cost and burden for Member States’ authorities:

The occasionally ‘blurry’ distinction between frontier workers and other cross-border workers, the distinction between wholly and partially unemployed frontier workers, the highly interpretable character of the criteria to determine the residence of a worker, the provisions on the aggregation of periods of insurance, employment and self-employment, and the reimbursement mechanism were mentioned as factors rendering the current coordination rules as complex;

While these regulatory distinctions intend to reflect the complexities of real situations and account for the actual differences between different types of cross-border workers and different types of national systems, the result is a striking variety of possible cases in which the interpretation of the rules carried out by each institution plays a significant role;

There are notable differences in the interpretation and application of the rules on the aggregation of periods and the extension of the period of export of unemployment benefits;

The classic distinction between frontier workers and other cross-border workers has become more problematic. Inter alia, the improvement and reduction in the cost of different means of transportation has allowed workers to cover ever larger differences to commute daily or weekly for work. The elements fixed in Article 11 of Regulation (EC) No 987/2009 are broad enough to prevent mobile workers to know with certainty their country of residence and hence the legal regime applicable to them in case of unemployment;

The reimbursement mechanism was often criticized, including claims considering that it should be made more transparent (Belgium) and that clear guidelines should be provided to each country (Luxembourg).

While the interviewees in certain countries defended that the current rules are sufficiently clear (e.g. the German Employment Services), the prevalent view was that the current coordination rules do not facilitate transparency and could be simplified. The burdensome character of the current rules was also criticized in countries which did not call for a revision of the coordination rules.

The diversity of opinions and practices in the application of certain aspects of the coordination rules is a testimony to the complexity of the rules and the lack of transparency they generate (since, given a similar situation in different regions, the similar outcome is not guaranteed). This complexity and incoherent understanding and application of the rules create a substantial (administrative) burden for the (national) administrations. This ‘burden’ is inherent to the management of cases where different understanding and national administrative processes apply; it goes beyond the definition of the administrative burden of the SCM where it is related to legal information obligations.

Around 40% of the participating public administrations reported that the EU rules create significant administrative costs and burden for national administrations. They consider the different types of forms/documents used per country, the varying requirements/understanding in terms of the information needed to fill out the documents, their mandatory or optional character and advance the procedures, and the different delays in the completion and transmission of documents as some of the most salient and recurrent problems. The reimbursement mechanism was repeatedly mentioned as a source of burden mainly due to the slow and ineffective communication between Member States.

“There are high administrative costs in what concerns to the reimbursement of the unemployment benefits that were paid. Moreover, we would highlight the delay on the treatment of the processes and the requests for payment that are denied. Because the EU rules create significant administrative costs and burdens for national administrations, EU law is not uniformly "understood" and applied by Member States and vice versa. … Paper SEDs are not always suitable for the exchange of information and not all MS use the same documents/forms. Reimbursement procedures create high administrative burden and important costs for both the MS of last activity and the MS of residence and the cost/benefit ratio is not effective, mainly for the MS of residence. The communications between institutions is slow and needs to be more effective.”

Several public officials expect the administrative burden to decrease in the next couple of years as a result of learning effects after the successful implementation and alignment of the rules. While the adoption of the Regulations took place ten years ago, it has taken time to fine-tune the implementation of the new rules and procedures. The lack of sound implementation of the new rules and procedures is particularly visible in a number of Member States. According to the online survey, 64% of the administrations stated that the communication (with other Member States) works well in general. However, there are problems with specific Member States. These reported problems are expected to be the main source of administrative costs.

Technological evolution could resolve some of the problems related to cooperation and communication. However, divergent interpretations of the rules and the information requirements for the completion of portable documents will continue to pose difficulties in the proper application of the Regulations.

In light of the limitations associated with the quantification of information obligations stemming from the application of the Regulations, we have adapted our approach for quantifying the IOs resulting from the Regulations and for assessing the (potential) impact of the policy options on the overall administrative process. In our analysis, we focused on a selected number of documents for which the stakeholders were able to provide robust information on a) the time spent to process/handle a document and b) the (approximate) number of cases.

We have collected useful information on the processing of documents related to cross-border cases for unemployment benefits on a) the estimated time and b) the number of cases in the following countries: Belgium, Poland, Luxembourg and Romania by means of a workshop. Other countries have provided a wealth of qualitative information which is useful for understanding the underlying problems related to the processing of the different documents and for assessing the (potential) impacts of the different options.

Despite the data limitations resulting from the problems discussed in this chapter, the assessment of the administrative cost (baseline scenario) for a number of key documents provides a robust basis for assessing the theoretical impact (positive or negative) of the different policy options on the administrative cost.

The number of PD U1/E301 documents received/issued provides insight into the extent to which periods of insurance and (self-) employment in another Member State were taken into account when granting unemployment benefits. For the purpose of the assessment of the administrative costs, we do not make any distinction between PD U1 documents and E301 documents (Member States are using either of the documents, depending on their national administrative processes). Both documents are treated interchangeably for the purpose of this exercise.

In the framework of this study, we have collected data on the number of PD U1 documents ‘issued’ and ‘received’. The following countries provided data on the total number of PD U1 documents ‘issued’: Belgium, Estonia, Luxembourg and Romania. With regards to the number of PD U1 forms ‘received’, we have collected data for Belgium, Estonia, France, Poland, Romania, Slovak Republic, Sweden and the UK. In addition, we have collected data on the aggregation of periods of employment/insurance/self-employment by means of a workshop in the following countries: Belgium, Poland and Romania (data provided for E301 only).

In order to assess the administrative costs for the EU-27 stemming from the processing of the PD U1 documents, we have carried out the following steps according to the Standard Cost Model (SCM) 63 :

Calculation of the unit cost per case:

The unit cost per case (processing/handling of a PD U1 form by the administrative staff – clerk level) provides insight into the total cost for processing one single PD U1 document (in a given Member State). It is based on the following formula:

Time (T) x Wage (W)

During the workshops and interviews in the Member States, we have collected data on the average standard time spent for processing/handling a PD U1 document for the following countries: Poland (5 minutes), Belgium (60 minutes) and Romania (363 minutes) 64 . As the data show, there are stark differences between the lowest time for processing data/information (Poland - 5 minutes) and countries where the processing time is relatively higher (Romania - 363 minutes). Belgium (60 minutes) ranges in the middle.

In Poland, for example, the process for handling PD U1 documents is automatized - Poland uses the portable documents efficiently (the administrative staff faces less administrative burden). According to the interviewees (national administration), the handling of the documents is reported to be less burdensome.

In Romania, on the other hand, the administration of E301 documents (note: not PD U1 in this case) is reported to be more burdensome. According to the interviewees, the administration of simple cases, with limited or no clarifications requested from the beneficiary or employer, may take minimum 1 hour of work in total for the person in charge 65 . The administration of complex cases, with a lot of missing, inadequate or incorrect information in the dossier, may request up to 8 hours of effort from the person in charge. In such cases, the respective civil servant assumes an active role in the completion of a correct dossier and starts giving phone calls, researching different taxes and employment data bases etc. 

Based on the interviewees’ responses for Poland, Belgium and Romania, it can be assumed that these three countries give good indications for calculating the average unit cost for processing/handling a PD U1 document: Poland (low administrative burden – 5 minutes), Belgium (average administrative burden – 60 minutes) and Romania (high administrative burden – 363 minutes).

For consistency and comparability with other SCM assessments of EU regulation, the tariff variable used in this study is based on hourly labour costs (plus overheads) per category of employment that has previously been used in recent SCM studies for DG EMPL 66 and our recent Impact Assessment studies we have conducted for the Commission. We have applied an average tariff/hour of EUR 18. It results in a rate per minute of EUR 0.3 (EUR 18/60 minutes).

The average unit cost for the EU-27 is EUR 42.8. It is calculated on the following basis: Time ((5 minutes (Poland) + 60 minutes (Belgium) + 363 minutes (Romania)) / 3) x Wage 0.3 = EUR 42.8 

1.Number of cases:

We have collected data for the number of PD U1 documents ‘received’ for the following countries: Belgium, Estonia, France, Poland, Slovak Republic and the UK. We have estimated the number of PD U1 documents for the other EU-27 countries on the basis of our own calculations based on collected administrative data and the 2012 Ageing Report (see section 4.1.2.1 for more detailed information on the number of PD U1/E301 forms ‘received’ and ‘issued’). We were able to calculate the estimated administrative cost for the EU-27 on the basis of this data. The total estimated number of PD U1 documents ‘received’ in the EU-27 in 2010 is around 340 000.

2.Calculation of the administrative cost (per Member State and for the EU-27)

We have calculated the administrative cost for processing PD U1 documents on the basis of this formula:

Time (T) x Wage (W) x Number (N)

The table below presents the total estimated administrative cost for processing PD U1 documents. The estimated total cost for the EU-27 in 2010 was EUR 14 604 326. Within the EU-27, the estimated total cost for processing PD U1 documents was highest (˃ EUR 1 million) in a number of the old Member States (in descending order): Spain, France, Germany and Italy. It was lowest (˂ EUR 100 000) in descending order in Sweden, Estonia, Slovenia, Cyprus, Luxembourg and Malta.

Table 12: Estimated administrative Cost - PD U1 (‘received’), EU-27, EUR, 2010

Source: Estimate based on collected administrative data and 2012 Ageing Report and data provided during the workshops on administrative burden (Belgium, Poland and Romania).

We have also calculated the average administrative cost for processing/handling a number of other documents, based on the data available. We were only able to produce the administrative cost for processing PD U1 documents (‘received’) for the EU-27 as we had data available for the EU-27 on the basis of our own calculations (there are no calculations for the other documents presented below).

We have calculated the administrative cost for ‘issuing’ a PD U1 document (‘issued’) for Poland and Belgium (based on the data we have collected during the workshops in the different countries). The total estimated cost for ‘issuing’ a PD U1 document in Poland is estimated at EUR 19 800. The amount is EUR 103 698 in Belgium. The table below presents the estimated cost for ‘issuing’ a PD U1 document for Poland and Belgium.

Table 13: Estimated administrative Cost – PD U1 (‘issued’), Poland and Belgium, EUR, 2013

Source: Deloitte, Workshop, Poland and Belgium

We have also estimated the cost for the following documents for Poland 67 :

SED U004 'Salary Info' (answer on SED U003);

SED U006 'Family Info' (answer on SED U005).

The table below presents the estimated cost for processing the above-mentioned documents in Poland. The total estimated cost for processing a SED U004 document ‘Salary Info’ in Poland is EUR 402. The cost for processing SED U006 documents ‘Family Info’ is estimated at EUR 825.

Table 14: Estimated administrative Cost –SED U004 ‘Salary Info’, SED U006 ‘Family Info’, Poland EUR, 2013

Source: Deloitte, workshop in Poland 

The PD U2 form is the authorisation which an unemployed person needs to export his/her unemployment benefit if (s)he wishes to move to another EU country to look for work. The competent national institution is responsible for granting this authorisation. There is a wide variety of practices in the EU-27 with regard to granting (and prolonging) authorisation to export unemployment benefit.

We have collected data on the number of PD U2 documents ‘issued’ for ten EU Member States: Austria, Belgium, Denmark, Estonia, Luxembourg, the Netherlands, Poland, Slovak Republic, Sweden and the UK. Taking together both components (survey data and own estimates) we estimate that 23.7 thousand unemployed persons have exported their unemployment benefits in 2010 (see section 4.1.3.3 for a detailed discussion on the calculation of the number of PD U2 ‘received’ and on the methodology for calculating missing data).

We have calculated the administrative cost for processing/handling a PD U2 document for the EU-27 using the following methodology:

1.Calculation of the unit cost per case:

The average unit cost per case is based on the data we have received from Poland (the only country for which we have received robust data on the time spent for processing a PD U2 document 69 ). The average unit cost per case that we found concerns the export of an unemployment benefit to 3 months 70 . Following the formula Time (T) x Wage (W), we have estimated an average unit cost per case (PD U2 ‘issued’) at EUR 4.5 71 .

The estimated unit cost should be treated with caution as it is based on one case only (Poland). As discussed in the section on the ‘aggregation of periods’, Poland seems to have an efficient (automatized) system for processing/handling PD documents (the processing of the documents is reported to be less burdensome). Therefore, it is to be expected, that the Polish example presents a rather positive picture on the overall time spent to process these documents. Other countries, such as Romania (which reported a much higher time spent for processing the PD U1 document) may report longer periods for processing/handling these types of documents. Due to data limitations, we have calculated the average unit cost on the basis of the Polish example.

2.Number of cases:

We have collected data on the number of PD U2 documents ‘issued’ by means of a questionnaire for the following countries: Austria, Belgium, Denmark, Estonia, Luxembourg, the Netherlands, Poland, Slovak Republic, Sweden and United Kingdom. In 2010, the total EU-27 number of PD U2 documents ‘issued’ is estimated at around 23 700.

3.Calculation of the administrative cost (per Member State and for the EU-27)

We have calculated the administrative cost for processing PD U2 documents (‘issued’) on the basis of this formula:

Time (T) x Wage (W) x Number (N)

The calculation includes the time spent on national administrative procedures supporting the processing of the SEDS and the time needed for processing the SED.

The table below presents the total estimated administrative cost for processing PD U2 documents. The estimated total cost for the EU-27 in 2010 was EUR 106 695. Within the EU-27, the estimated total cost for processing a PD U2 documents was highest (˃ EUR 10 000) in a number of the old Member States (in descending order): Spain, Germany and France. It was lowest (˂ EUR 500) in descending order in Slovenia, Slovak Republic, Estonia, Cyprus Malta and Romania.

Table 15: Estimated administrative Cost –PD U2 (‘issued’), EU-27, EUR, 2010

Source: Estimate based on collected administrative data and 2012 Ageing Report and data provided during the workshops on administrative burden (Poland).

We have also estimated the cost for the following documents for Poland 72 :

Competent employment service:

SED U011 'Effect to Entitlement - Export' (answer to SED U010);

SED U012 'Request for monthly follow-up'.

Employment service of the MS where jobseeker has gone:

Process PD U2;

SED U007 'Request Document on Export';

SED U009 'Notification Registration - Export';

SED U010 'Circumstances Affecting Entitlement - Export' (linked with U3 form);

Issue of PD U3 (linked with SED U010);

SED U013 'Monthly Follow-up' (answer on SED U013);

SED U028 'Request Entitlement to Export'.

The tables below present the total estimated administrative cost for processing the respective documents presented according to a) competent Member State and b) employment service of the Member State where the jobseeker has gone. The estimated unit cost per case is based on the data provided by Poland (T: time and W: wage (EUR 0.3)). Note that the unit cost per case differs from the one calculated for processing the PD U2 document in the documents presented below. We have not calculated the EU-27 average cost for all documents due to data limitations. Be aware that these costs occur separately, others are combined. There is no overview of the total number of flows. In the future this should be made possible by EESSI.

Table 16: Estimated Administrative Cost – Competent employment service, SED U001, SED U012, Poland, EUR, 2013

Source: Deloitte, workshop in Poland 

Table 17: Estimated Administrative Cost – Employment service of the Member State where the jobseeker has gone, PD U2 (‘process’), SED U007, SED U009, SED U010, PD U3 ‘issue’, SED U013, SED U028, Poland, EUR, 2013

Source: Deloitte, workshop in Poland

Claims for reimbursement can be made by the country of residence to the country of last activity for fully unemployed frontier workers but also for other cross-border workers who have decided to register with the competent institution in their country of residence. The country of last activity reimburses the unemployed benefits provided in the country of residence during the first three months or five months (when the unemployed person during the preceding 24 months, completed at least 12 months of (self)employment in the country of last activity). Reimbursement procedures are defined under art. 65(6) and (7) of Regulation (EC) No. 883/2004 and art. 70 of Regulation (EC) No. 987/2009.

1.Calculation of the unit cost per case:

The average unit cost per case is based on the data we have received from Poland (the only country for which we have robust data on the reimbursement claims. Following the formula Time (T) x Wage (W), we have estimated an average unit cost per case for each of the individual documents.

2.Number of cases:

We have collected data on the number of cases for Poland for a number of documents. There are no estimated data available for calculating the estimated total number of cases of reimbursement claims in the EU. For a detailed discussion on the number of claims received (as debtor) and the number of claims issued (as creditor) (see section 4.1.5 )..

3.Calculation of the administrative cost (Poland)

We have calculated the administrative cost for processing a number of documents related to reimbursement claims for Poland by applying the following formula: Time (T) x Wage (W) x Number (N).

Data were collected for the following documents:

Member State of residence:

SED U020 'Reimbursement Request';

SED U025 'Reimbursement Receipt/Closing notification'.

Competent Member State:

SED U021 'Reimbursement Full Acceptance' (possible answer to SED U020);

SED U022 'Reimbursement Non Acceptance' (possible answer to SED U020);

SED U023 'Reimbursement Partial Acceptance' (possible answer to SED U020);

SED U024 'Reimbursement Payment Notification'.

The table below presents the total estimated administrative cost for processing the following documents for Poland presented according to ‘Member State of residence’ and ‘Former working Member State’:

Table 18: Estimated Administrative Cost, Member State of Residence, SED U020, SED U025, Poland, 2013

Source: Deloitte, workshop in Poland 

Table 19: Estimated Administrative cost – Competent Member State, SED U021, SED U022, SED 023, SED U024, Poland, EUR, 2013

Source: Deloitte, workshop in Poland 

4.2Coordination of long-term care benefits

Pensioners and cross-border workers – and their insured family members – are the most important group impacting cross-border expenditure on LTC benefits.

The recent migrant pensioners (aged 65 and over) who have moved from one EU-country to another in 2011 are counted in the Eurostat migration statistics ( Table 20 ). Also for this group of recent migrant pensioners we are missing figures for BE, BG, HU, MT, NL, PL and RO. Especially the UK (11 thousand in 2011) and Spain (10 thousand in 2011) are important migration countries for pensioners.

Ideally, a table with the current stock of pensioners by previous country of residence should be available because this variable is probably the best proxy to estimate the number of pensioners insured in another country than the country of residence and to calculate the amount of LTC reimbursements between countries.

Table 20: Migrant pensioners (65 years or over) in 2011 by country of previous residence, EU-27

* Total= sum of migrant pensioners of which the current country of residence is known

Source: Eurostat Migration Statistics, [migr_imm5prv]

The number of those insured for health care living in another country than the competent country – which sometimes includes long-term care or to which LTC-insurance is closely linked – can be calculated based on the number of PD S1s - or E106 forms (insured person), E109 forms (family member of insured person) and E121 forms (pensioner and family member of pensioner).

The PD S1 form allows a person to register for healthcare in the country of residence when insured in a different one. The form is delivered per person (not per family). The distinction between the types of person insured for health care is still made by the PD S1. For that reason, the questionnaire (limited to 14 MS) asked for the number of insured persons, family member of insured persons, pensioners and family members of pensioners. Only two Member States (Belgium and Luxembourg) provided a breakdown of the number of PD S1 issued by type of person ( Table 73 ). More general data was obtained for two other Member States (the Netherlands and Slovak Republic). The questionnaire asked only for the number of PD S1s issued. However, the number of PD S1s received could also be given. For example, there are data on the number of PD S1’s received by Belgium (including by type of insured person) ( Table 74 ). For reason of this poor response to our administrative questionnaire we opted for an estimate of the PD S1. Two large categories can be distinguished (see also tables above). PD S1 forms will be issued to cross-border workers (and their family members) and mobile pensioners (and their family members).

A PD S1 is issued to incoming cross-border workers and in some cases also to their family members (e.g. when the partner is inactive). Based on our survey we observed the following for Luxembourg and Belgium.

As Luxembourg attracts a lot of cross-border workers, it issues a high number of PD S1 forms. As of end March 2011, 162,638 cross-border workers were insured in Luxembourg ( Table 73 ). They resided mainly in neighbouring countries: 48.1% in France, 24.9% in Belgium and 24.7% in Germany. In addition, 60,868 family members of cross-border workers working in Luxembourg were insured there, or 37.4% of the number of cross-border workers. It should be noted that more than one family member can be related to a cross-border worker (inactive partner and children).

For Belgium, only figures on incoming frontier workers were received. As of end-June 2012, there were 46,484 frontier workers in Belgium, of whom 79.1% resided in France ( Table 73 ). The number of family members of cross-border workers employed in Belgium and insured in this Member State was 512. A striking figure is the percentage of family members residing in Poland who are insured in Belgium (32.8% of the total number of family members of persons insured in Belgium).

Due to the lack of data, it is necessary to use other data sources in estimating the number of insured cross-border workers for the EU-27. By making use of LFS data, the number of cross-border workers can be estimated. 74 On average, one million cross-border workers live in one EU-27 Member State but work in another (average of 2010 and 2011). Cross tables provide insights into their country of residence and country of employment. We consider the row and column totals as the most reliable, in line with the reliability limits provided by Eurostat for the LFS. 75  The bilateral cells between countries need to be interpreted with even more care. All cells should be compared with these reliability limits. However, by taking two years together, the cells can be assumed to be relevant by avoiding outliers and by counting more observations. We will need them for further calculations. Most cross-border workers are employed in Germany (186,000) and Luxembourg (131,000). The country of employment will have to issue a PD S1 for all these cross-border workers living in another country.

A PD S1 can also be issued to pensioners (and their family members) who move to or live in another Member State than the competent Member State. Chapter I ‘Sickness, maternity and equivalent paternity benefits’, section II ‘Pensioners and member of their families’ in Regulation (EC) No. 883/2004 makes the distinction between ‘Right to benefits in kind under the legislation of the Member State of residence’ and ‘No right to benefits in kind under the legislation of the Member State of residence’.

First, we look at the Regulation when a pensioner is entitled to benefits in kind under the legislation of the Member State of residence. Art. 23 states that: “A person who receives a pension or pensions under the legislation of two or more Member States, of which one is the Member State of residence, and who is entitled to benefits in kind under the legislation of that Member State, shall, with the members of his family, receive such benefits in kind from and at the expense of the institution of the place of residence, as though he were a pensioner whose pension was payable solely under the legislation of that Member State.”

Second, we look at the Regulation when a pensioner is entitled to benefits in kind under the legislation of the Member State of residence. Art. 24, 2 makes a distinction between being only entitled to benefits in kind under the legislation of a single Member State and being entitled to benefits in kind under the legislation of two or more Member States: “(a) where the pensioner is entitled to benefits in kind under the legislation of a single Member State, the cost shall be borne by the competent institution of that Member State;(b) where the pensioner is entitled to benefits in kind under the legislation of two or more Member States, the cost thereof shall be borne by the competent institution of the Member State to whose legislation the person has been subject for the longest period of time; should the application of this rule result in several institutions being responsible for the cost of benefits, the cost shall be borne by the institution applying the legislation to which the pensioner was last subject.”

We do not have data on the relative shares of exclusive or mixed pensions. It is a share between 0 and 100% that could be used as minimal or maximal estimate. The average would be 50% of those two extreme hypotheses.

This legislation is important in making an estimation of the number of pensioners who received a PD S1 form from a specific Member State. First, we discuss the data from the Member States of which data has been received ( Table 73 ).

As of end of March 2011 (= number of insured persons living abroad at that moment), Luxembourg had issued 7,622 PD S1 to pensioners mainly moving to/residing in Belgium, France, and Germany ( Table 73 ). There were also, 2,798 family members of pensioners moving/residing in another Member State insured in Luxembourg.

Belgium had issued 28,166 PD S1’s (situation end of June 2012) to pensioners ( Table 73 ). Most of these forms were issued to pensioners moving to/residing in France (42.3%). More ‘surprising’ is the high number of PD S1’s issued to pensioners moving to/residing in Spain (22.5%) and Italy (15.7%). It confirms, yet for Belgium, the hypothesis of pensioners moving to Mediterranean countries and also of Italian migrant workers moving to their country of origin/birth.

If we want to estimate the total number of pensioners in the EU-27 who have received a PD S1 form, other data sources have to be used. The legislation to determine which Member State is competent has a strong impact on the calculations. The number of pensioners moving abroad can be estimated by the LFS. We already discussed the limitations of the LFS regarding the interpretation of the number of pensioners who moved abroad after their retirement.

There are data from five Member States (Austria, Belgium, Germany, Luxembourg and United Kingdom) on exported LTC cash benefits ( Table 75 and Table 78 ).

Before describing the data, some explanation is needed. Some of our national experts who collected the data did not receive any data in their Member State because the LTC benefit in cash was not considered to be exportable (e.g. APA in France 76 ; Romania 77 ). At the same time, no (detailed) data were available for some Member States. Moreover, not every Member State has an LTC benefit in cash. 78  

In 2012, 2,570 people exported Pflegegeld from Austria to another Member State ( Table 75 ) 79 . This was only 0.6% of the total number of people entitled to it (444,000 persons). The number exporting this LTC benefit decreased between 2010 and 2012 by 7.4%. A breakdown by Member State could be made for February 2012, where 70% of this LTC benefit in cash was exported to Germany.

Data were collected for three types of LTC cash benefits cash in Belgium ( Table 75 ). 27 people living abroad were entitled to the Flemish Care Insurance. The estimated cost is € 42,000. This is only 0.01% of the total number entitled to this LTC cash benefit (217,400 in 2011) ( Table 78 ). In addition, on average 30 people exported the Integration allowance OR the allowance for assistance to the elderly from Belgium to another Member State. Thus, export of these Belgian LTC cash benefits was very limited. The competent institution assumes that the Personal Assistance Budget is not exportable. 80 For that reason no data was provided by them. It was not possible to collect exact figures on how many people receive an allowance for children with disabilities supplementary to child benefit. However, an estimate could be made. There were in 2011 42,242 children of foreign cross-border workers entitled to a child benefit paid by Belgium. 81 Also, in Belgium, there are in total 2.1% children with disabilities. This percentage was applied to these 42,242 children. This results in an estimate of 900 children entitled to an allowance for children with disabilities supplementary to child benefit or a total cross-border expenditure of € 3 million.

In 2010, 1,875 people living abroad were assessed to receive in cash Pflegeversicherung (‘Pflegegeld’) from Germany ( Table 75 ). 82 This was a decrease of 9.7% compared to 2006. The assessment was mainly asked for by people living in Spain (30.3%) and Austria (27.8%). The competent institution estimates that on average 5,000 persons living abroad receive the Pflegeversicherung from Germany. 83 This accounts for an expenditure of € 3 million ( Table 78 ). This is 0.2% of the total number of people in the Pflegeversicherung (2.4 million persons in 2010) or 0.01% of expenditure (€ 21.5 billion). 84  

In 2011, Luxembourg’s LTC Insurance was exported to another Member State in 359 cases, especially to Germany (51.8%) ( Table 75 ) 85 . This number represents some 2.3% of beneficiaries of LTC in Luxembourg). This involves a total yearly cross-border expenditure of € 3.6 million ( Table 78 based on CNS, Décompte de l’assurance dépense de l’exercice 2011 86 ). The Luxembourg report on social protection 2011 observes “Toutefois, le nombre de personnes bénéficiant de prestations et résidant à l'étranger est très faible alors que la population protégée résidant à l'étranger est importante” (p. 148). 87  This illustrates that our further calculations on cross border LTC expenditures will be considered as theoretical and maximized.

Finally, total numbers were collected for three types of LTC cash benefits in the United Kingdom ( Table 75 and Table 78 ). In 2011, 4,210 people exported the Disability Living Allowance from the United Kingdom, a cross-border expenditure of £11 million. Again the impact is limited to 0.1% of the total beneficiary population (3.2 million persons) receiving this cash benefit or 0.1% of total expenditure (£ 12.6 billion). The Attendance Allowance is received by 1,090 persons living abroad at a cost of £ 3 million. Finally, the Carers Allowance is exported by 230 persons to another Member State at a cost of £1 million.

The relative importance of the number of cross-border users and related expenditure on these LTC benefits in cash is very limited compared to the total number of users and expenditure on LTC benefits in cash by a Member State. It mostly fluctuates between 0.01% and 0.6% of total number of users and expenditure.

The number of persons who received a LTC benefit in kind and the cost involved can be calculated via the Structured Electronic Document (SED) S080 (claim for reimbursement) (point 3.14 ‘Long-term care benefit’ amount AND /OR point 3.20 ‘Nature of benefits’ = long-term care). However, this is only for reimbursements determined on the basis of actual expenditure and not on the basis of fixed amounts. Some Member States receive only fixed amounts calculated on the basis of a formula defined in Regulation (EC) No. 883/2004. Even though Member States can use the SED-forms related to health care, they still use the old E-forms (E125 ‘Individual record of actual expenditure’, E126 ‘Rates for refund of benefits in kind’ and E127 ‘Individual record of monthly lump-sum payments’). There is a limitation in that LTC is not mentioned on the E 125 form, which makes it very difficult to calculate LTC benefits in kind.

None of the Member States could respond to our question asking for a calculation of the number of people receiving a LTC benefit in kind and the cost involved ( Table 76 and Table 77 ). 88 However, a proxy of the use and cost of the export of LTC benefits in kind was noted for Belgium. We saw that 28.3% of the claims received by Belgium via an E125 form were applicable to persons aged between 65-80 and 10% to persons aged older than 80. 27.9% of the claims received by an E125 form involve persons aged between 65-80 and 16.5% aged older than 80 ( Table 77 ). This pattern is also visible in the other direction: 10% of the E125 claims sent by Belgium to other Member States involve someone aged 65-80 (or 20.3% of the amount of claims issued) and 5% are for people aged older than 80 (or 20.5% of the amount of claims issued) ( Table 77 ).

Our questionnaire aimed to obtain an overview of the number of persons insured for LTC benefits in kind and in cash, the extent to which these people actually received benefits and the corresponding cost of the benefits provided (reimbursed by the competent Member State).

However, this exercise requires detailed information from the reporting Member States concerning the issue of PD S1 by status (insured person, pensioner, family member of the insured person and family member of the pensioner) and the claims (e.g. by SED S080 – only for actual expenditure in kind). We observed that many of the Member States still use the ‘old’ E-forms to communicate with other national administrations or with the citizens involved. For example, the E-106 form (certificate entitlement to sickness benefits in kind insured person), the E-109 (certificate entitlement to sickness benefits in kind family members insured person) and the E-121 form (certificate entitlement to sickness benefits in kind pensioner and family members of the pensioner) were replaced by the portable document (PD) S1 (and SED S072 ‘Entitlement document – residence’). Nevertheless, the total number of these three forms is still counted to calculate the number of persons insured to sickness benefits in kind, who reside in a Member State other than the competent Member State. This detail was only collected for two Member States in our sample. That is why other data sources needed to be exploited.

The number of persons who received LTC benefits in kind could be calculated by extracting data from the SED S080. In principle, the general method of reimbursement is the refund on the basis of actual expenditure and exchanged between the competent national authorities by the use of SED S080 (former E125) which is an individual claim for an actual expenditure. The receipt of LTC is explicitly mentioned in this form (point 3.14 ‘Long-term care benefit’ amount AND point 3.20 ‘nature of benefits’: Long-term care). This detail would deliver us valuable information about the number of persons who received long-term care. Only by a way of exemption, those Member States 89 whose legal or administrative structures are such that the use of reimbursement on the basis of actual expenditure is not appropriate, can reimburse benefits in kind on the basis of fixed amounts for some specific categories: family members who do not reside in the same Member State as the insured person and pensioners and members of their families. Each of these Member States has to calculate the monthly fixed amount (average costs) per person, which should be as close as possible to actual expenditure. The method of calculation is defined in the implementing Regulation (EC) 987/2009. For the exchange of the information needed for the refunds on the basis of fixed amounts SED S095 (former E127) is used. That form is an individual record of monthly lump sum payments. However, a long-term care category is not explicitly taken up by this form. A possible proxy is the selection of SEDs S095 forms handling a claim of a person aged 65+ (see point 2.5 ‘Lump sum category’ in this SED). However, today both forms (SED S080 and SED S095) are not used by Member States. The E125 form (actual expenditure) and E127 form (fixed amounts) do not mention the specific category ‘long-term care’. By this, it was not possible to count the number of persons who received long-term benefits in kind. Again, we had to look for alternative data. We estimated the probable number of PD S1 and will use it further to estimate the use of health care and LTC.

Three categories of PD S1 are identified:

Cross-border workers (and their family members);

Retired former cross border workers (and their family members);

Other mobile pensioners (and their family members).

First objective was to calculate the number of persons who are insured to sickness benefits in kind living in a Member State other than the competent Member State. To work as detailed as possible a distinction should be made between the different categories of insured persons. By the Labour Force Survey, the number of cross-border workers was calculated for 2010 and 2011. An average of both years was calculated to improve the representativeness of cell data and to exclude outliers. Calculations were made only for the EU-27 Member States. On average 1 million cross-border workers are employed in another EU-27 Member State than the Member State of residence, or 0.48% of the total working population. The working countries will issue a PD S1 to these incoming cross-border workers while the countries of residence will receive this PD S1. However, as already mentioned also posted workers can be included by the definition used in the LFS. These persons are still insured in the country of residence. The cross table illustrates which countries are mostly involved in this cross-border mobility of workers ( Table 81 ). Most of the PD S1 certificates have been issued by Germany (186.1 thousand forms) and Luxembourg (130 thousand forms) while most of the PD S1 have been received by Germany (165.4 thousand forms) and France (159.6 thousand forms). The accurateness of these survey figures can be verified by the administrative data we received from two Member States (Belgium and Luxembourg) ( Table 73 ). Luxembourg issued 162.6 thousand PD S1 forms to insured persons (situation end of March 2011), which is somewhat higher compared to the LFS estimation of PD S1 issued. For Belgium we know from the administrative survey that 46.5 thousand PD S1 certificates were issued to frontier workers (only cross-border workers coming from France, Germany, Luxembourg and the Netherlands). Based on data of the LFS, 50.1 thousand frontier workers should have received a PD S1 certificate from Belgium.

Second, the number of family members of insured persons who are insured in a Member State other than the competent Member State should be counted. They should be added as ‘dependent persons’. We assume that 20% of the insured persons will have an insured family member. This assumption is based on data we received on the PD S1 from Belgium and Luxembourg, the inactivity rates published by Eurostat based on the LFS and a study we recently published. 90  Based on our questionnaire, the number of insured family members in proportion to the number of insured persons was very low in Belgium (almost 1 in 100) but high in Luxembourg (1 in 3). On the basis of the results of the LFS, Eurostat publishes on a quarterly basis the inactivity rate (inactive population as a percentage of the total population) (see 10.5 ). The main reason for inactivity of young people is participation in education, while retirement is the main reason for older people. For that reason we only looked at the population between 20 and 64 years. On average 24% of the population between 20 and 64 years old is considered as inactive. In our study we observed 22.8% of the frontier workers are living together with a partner who is inactive. A household counts on average 1.2 children which assuming the same composition in active and non-active households will fall also under the social security system of the cross-border worker when the partner is inactive (or 27.4% of the frontier workers). 91 Together, this is about 50% of the number of frontier workers. For those reasons, the assumption that 20% of the insured cross-border workers have an insured family member is perhaps rather a conservative hypothesis. A more liberal assumption could be made (e.g. 40% of the cross-border have an insured family member). We keep the conservative assumption since other hypotheses could overestimate the number of issued PD S1 (see below). Generalizing this percentage of 20% to the whole population, an estimation of 206.5 thousand family members of insured persons who have received a PD S1 is obtained.

Third, the number of persons who live in another Member State than the competent Member State has to be estimated. To determine the competent Member State, Regulation (EC) No. 883/2004 makes a distinction between pensioners and member of their families who have right or NO right to benefits in kind under the legislation of the Member State of residence’. Three different scenarios appear (see EC (2011), The coordination of healthcare in Europe, page 17):

“the Member State of residence, if the person concerned is in receipt of a pension from that State entitling him/her to benefits in kind (even where the person concerned is in receipt of pensions from one or more Member States);

the Member State responsible for paying the pension entitling the person concerned to benefits in kind if s/he resides there, if the person concerned is not in receipt of a pension in his/her State of residence (even where s/he is in receipt of benefits in kind in this State by virtue solely of his/her residence);

the Member State responsible for paying a pension entitling the person concerned to benefits in kind if s/he resides there, to whose legislation the pensioner was subject for the longest period, if s/he is in receipt of pensions from several Member States other than the Member State where s/he resides”.

This is perhaps the most difficult exercise, since there is no correct variable available in the LFS which can be considered as a good proxy for the scenarios described above. For that reason, we defined two separate scenarios. First, we have estimated the number of retired cross-border workers. When the cross-border worker only worked abroad he/she will receive only a pension from that Member State. In that case the former working country will be the competent Member State. When he/she was also for a period of time employed in the country of residence, he/she will receive a mixed pension. In that case the country of residence will be the competent Member State. The average would be 50% of those two extreme hypotheses. In our estimation we assume that all cross-border workers receive only a pension from their former working country. Nevertheless, also other scenarios can be considered. To estimate the total number, we applied the percentage of cross-border workers on the labour market (total average of 0.48%) ( Table 81 ) to the number of pensioners in 2010 (figure from 2012 Aging Report- variable ‘Pensioners aged 65+’) and this by individual (former) working Member State. E.g. 2.42% of the employment in Austria is related to incoming cross-border workers. This percentage is applied to the 1.8 million pensioners in Austria which results in an estimation of 43.9 thousand persons receiving only a pension from Austria whereby Austria will also be the competent Member State and will issue a PD S1 since this retired cross-border worker lives in another Member State. This assumption results in an estimation of 419.5 thousand pensioners who were previously working abroad.

Finally, an estimation of the number of migrant pensioners in 2011 is calculated by using the LFS (= selection of ‘retired persons aged older than 60 at arrival’ of which country of birth= EU27 and country of residence=EU27). We have used ‘country of birth’ as a proxy of the competent Member State (an alternative is ‘Nationality’). This was the only best practical option in the ambition to determine the competent Member State. In total 190.5 thousand pensioners live in another EU Member State than their country of birth (no data available of the number of pensioners living in Germany and Romania). 92 We assume the country of birth is the competent Member State and will issue a PD S1. By the Eurostat migration statistics already figures on the number of recent migrant pensioners (aged 65 and over) who have moved from one EU-country to another in 2011 were obtained (see also table). However, for this group of recent migrant pensioners no data were available for BE, BG, HU, MT, NL, PL and RO as migration country. For the limited group of countries, already 44 thousand pensioners migrated to another Member State in 2011. It seems that the total group of 190.5 thousand pensioners (stock value) who moved after retirement to another Member State is a (small) underestimation from the real situation. 93 However, as the ‘popular’ Mediterranean countries are already inserted in the table of recent migrant pensioners the total number of recent migrant pensioners will not that much higher. In 2011, 10 thousand pensioners migrated to Spain. Based on the LFS a total number of 61.4 thousand retired persons older than 60 at arrival lives in Spain. 3.6 thousand pensioners moved to France in 2011 while based on the LFS a stock of 34.1 thousand migrant pensioners was obtained. The proportion between flow and stock seems for both countries more or less realistic.

Finally, we had to estimate the number of family members of pensioners who live in another Member State than the competent Member State. We assume that 25% of the pensioners will have also an insured family member. Based on the administrative data from the questionnaire, a percentage of 37% was calculated for Luxembourg and 20% for Belgium. This assumption results in an estimation of another 152.500 family members of pensioners who live in another Member State than the competent Member State.

By counting these different components together, we estimated a total number of about 2 million insured persons living in another Member State than the competent Member State (cf. 10.4.1 ). Some 60% is determined by the present cross-border workers which imply some 40% is related to mobile pensioners. In the future this share of retired former cross-border workers and mobile pensioners probably will increase. Most PD S1 certificates are issued by Germany (18.6% of total), UK (11% of total) and Luxembourg (10.5% of total) (table). Most of these forms were received by France (15.7% of total) and Germany (13.8%). Belgium seems also to receive a high number of PD S1 certificates (11.4% of total). However, the calculations for Belgium as destination country for migrants are probably not reliable due to problems with the variable ‘years of residence’. 94 Just to illustrate the ambitions of these estimates, the calculated figure for Luxembourg as competent state of PD S1 and so by definition also insured for LTC, of 207 thousand insured persons is 262 thousand in the administrative questionnaire.

However, these figures should always be considered as an estimation of the number of PD S1 certificates based on several assumptions. 95 As explained above, we are confronted with several data limitations that had a significant impact on our calculations. By adding some additional questions to the LFS and by becoming an exhaustive view on the number of migrating pensioners by the Eurostat migration statistics most of these data limitations would be solved which would result in a more reliable ‘proxy’.

Table 21: Estimated number of PD S1 issued by category of citizen, in .000

Source: Estimate based on data from LFS and 2012 Ageing Report

Table 22: Estimated number of PD S1 issued/received, in .000

Source: Estimate based on data from LFS and 2012 Ageing Report

The table above provides the estimated cross table for the stock of provided (by competent member state) and received (by country of residence) portable documents S1. We can again read the table row by row or column by column. Each column shows the total number persons where the state is competent for and in which countries those persons are resident. Each row shows the total number of persons residing in their country and with a foreign state competent for their social protection and how this is distributed over those countries. We immediately observe that in absolute figures those cells are dominated by the large countries, but on top of that it seems to be concentrated among a limited number of countries. For instance taking the first row and column we can see that of the 69 thousand PD S1 received for persons living in Austria, 52 thousand come from Germany, 3 thousand from Slovakia and 2.8 thousand from Italy. The top 3 of competent states for S1 counts for 84 % of the total. Reading the column for Austria, of the 177 thousand PD S1 issued some 65.2 thousand persons are living in Germany, 43.5 thousand living in Slovakia, and 36.4 thousand living in Hungary. These 3 largest countries count for 82 %. Those ‘concentration ratios’ of the share or the 3 largest countries (in concentration analysis it is called the C3) can be calculated for each country either by country of residence or by country of competence. We observe that especially Germany (20 times in top 3 of competent MS) and the UK (16 times in top 3 of competent MS) are the most ‘important’ competent MS (in % of residents insured in another MS than the competent MS). The picture of the most ‘important’ MS of residence (in % of persons living in another MS than the competent MS) is more diverse.

The share of the largest countries includes only the information about those large countries, without telling something about the share of the other countries. For that reason in the economic analysis of the distribution some alternative measures are in use, the Herfindahl index weighting the share mi with its own, what will over accentuate the large shares, but includes in any case all shares, and the Entropy index that weights each share with the logarithm of the inverse of this share, what reduces the importance in the concentration analysis of large countries. To make those concentration indexes intuitively more appealing a ‘number (here of countries) equivalent is calculated what stands for the number of countries of the same size that ends up in the same Herfindahl or entropy index than we will obtain in reality, but now of the same size. Those indexes are calculated as follows:

When mi is the share of country i in the total for EU 27 of a certain on pages n variable, then the Herfindahl index H= and its number equivalent NEH = 1/H; the entropy index

E = and its number equivalent NEE = antilog2 E.

In reality we look here at 27 countries but they are already of an unequal size of the population. We calculate that those countries are unequal according to the size of their population similar to only NEH of 10.4 in the EU 27 of the competent countries and a NEE in the EU 27 for the competent countries of 13.4. The respective NEH and NEE of the total population is 10.6 and 14.3, illustrating how dominant the total population figures are. Those figures illustrate further how the Herfindahl index weights the largest share more, resulting in a lower number of countries.

Those are the distributions that can be expected in proportion with the population. What is now the distribution for the 27 member states of their issued PD S1 as a competent country and distributed by countries of destination or residence and the number of PD S1 received as a country of residence and distributed over competent country. Both NEH and NEE are calculated where the number (here number of countries) equivalent of Herfindahl over accentuates the concentration. Those figures are provided in Table 23 . But first we have to look at the first two columns that tell us if either a country is more a country of residence or a competent country. Some are even both. Then we can see if the number equivalent is large or small. Looking for the lowest number we find NEH in the resident countries of the number of competent states of some 1.8 in Austria or 1.3 and 1.4 in respectively Ireland and Estonia illustrating that the number of competent countries the people living there is small. Those countries will also being identified with similar figures but somewhat higher for the number equivalent of entropy in the residence country for the number of competent countries. They tell us that the distribution of the number of PD S1 according to country of competence is as those persons are coming from one to two or three countries. Some other countries demonstrate much higher numbers, illustrating that the people come from much more countries.

The same analysis can be made for either Entropy or Herfindahl indexes in the country of competence for the number of countries of residence, illustrating over how many countries the PD S1 issued are spread. This implies that people for whom the country is competent are limited to a small number of countries they are residing in, or to a larger number. For each country their index as a country of residence or a country of competence can be compared, for instance illustrating that the country is competent for people coming from a large or small number of countries while the people residing there with a PD S1 come from a small or large number of competent countries. Different patterns can be observed. For instance Belgium and Austria have an opposite profile, with Belgium being competent for people residing in a limited number of countries, while hosting people with a larger number of competent states. Austria was hosting people with a smaller number of competent states, while it is competent for a larger number of countries.

Table 23: Entropy and Herfindahl indexes of concentration of cross-border insured persons by PD S1

Source: Own calculations based on data from LFS and 2012 Ageing Report

As next step we have estimated the cross-border expenditure on health care and long-term care based on figures from the 2012 Ageing Report (variables ‘Health care spending in 2010 prices per person’, ‘Population (million)’ and ‘Long-term care spending in 2010 prices (in billion Euros)’).

We calculated our estimates on average benefits for the total of the insured population. It is as mobile citizens (workers, pensioners, their family members) are using this system of LTC as if they were nationals. This involves a ‘potential’ overestimation of the number of users of cross-border LTC benefits and the related expenditure due to fact some MS consider their LTC benefit as not exportable. At the same time these estimates assume a complete ‘take-up’ of rights by mobile citizens which will not be the case in the baseline scenario.

To get as close as possible to the applicable rules in Regulation (EC) No. 883/2004 a distinction had to be made between LTC benefits in cash and in kind. In the 2012 Ageing Report this breakdown is not available. Nevertheless, in a paper of DG ECFIN the public expenditure on LTC as % of GDP by type of care was published for each of the 27 EU Member States. 96 The yearly health and LTC (by type) expenditure per capita could be calculated for each of the Member States. These amounts are applied to the total number of insured persons who received a PD S1 certificates to estimate the claims issued as creditor and received as debtor for health and LTC. 97

Following rules are kept in mind when constructing the calculation model for the baseline scenario:

LTC benefits in kind are provided according to the legislation of the Member State of residence and reimbursed by the competent Member State;

LTC benefits in cash are provided according to the legislation of the competent Member State.

For the baseline scenario on LTC the current budgetary impact is calculated as well as the number of insured persons ( Table 24 , Table 25 and Table 28 ).

The overall budget is estimated at € 994.7 million of which € 618.3 million (62% of total budget) is related to LTC benefits in kind and € 376.4 million (38% of total budget) is related to LTC benefits in cash ( Table 24 ).

Compared to total national expenditure of LTC (variable ‘Long-term care spending in 2010 prices’ – 2012 Ageing report) the share of the cross-border LTC expenditure is limited to 0.4% of total EU expenditure or 0.008% of total GDP of the EU-27 (variable ‘Long-term care spending as % of GDP’- 2012 Ageing Report). Those low but realistic percentages illustrate we are making estimates literally on the frontier of the borderline of those systems, what results in ‘marginal’ shares in relative terms, but nevertheless substantial in absolute terms.

In absolute figures Germany (€ 172.9 million), the Netherlands (€ 166.3 million) and Luxembourg (€ 119.4 million) are the most important debtor countries taking into account the total cross-border expenditure on LTC.

In % of total spending, 29.4% of national expenditure on LTC by Luxembourg is granted to cross-border workers and pensioners (and their family members). In reality we observed not the tenth of this figure, illustrating probably also the difference of the use of LTC by the rest of the population and by the cross-border mobile population (see comments again on the situation in Luxembourg). 98  

Nevertheless, by using this ‘real life information’ (as the LTC expenditures and the hypothesis that the cross-border mobile citizens are in a similar way entitled and using LTC as the rest of the population) this is not only an optimistic interpretation of the application of the coordination principles, but it reflects also the possible implications of changes in those systems. For instance a shift of a LTC system from in kind to in cash systems (‘consumer oriented’ payment systems) that is under discussion in many countries, will have an impact on the application of the coordination regulation. Countries less oriented to in cash benefits (for instance Luxembourg, but probably also other countries) and countries substantially oriented to in cash (as the Netherlands), and the impact of this in those estimated flows, clearly ‘pop up’ in our calculations.

Main debtor countries for LTC benefits in kind received in another member state than the competent Member State are again Germany (€ 120.1 million), Luxembourg (€ 104.1 million) and the Netherlands (€ 73.2 million).

The highest total amounts of LTC benefits in cash to cross-border workers and pensioners are paid by the Netherlands (€ 93.1 million), Germany (€ 52.8 million) and Austria (€ 49.8 million).

Based on the projections, the cross-border expenditure on LTC will increase to €1.3 billion in 2020 (0.009% of GDP) and €1.8 billion in 2030 (0.010% of GDP). An estimated number of 93 thousand mobile workers and pensioners residing in another Member State than the competent Member State would have receipt LTC in 2010. Projections let increase the number of recipients to 106 thousand persons in 2020 and 121 thousand persons in 2030 ( Table 24 and Table 27 ).

Table 24: Estimated LTC cross-border expenditure baseline scenario (in € billion), as % of total spending and projections 2020 and 2030

Source: Estimate based on data from LFS, 2012 Ageing Report and Lipszyck, B., Sail, E. & Xavier, A. (2012), Long-term care: need, use and expenditure in the EU-27, EU.

Based on the constructed table on the number of PD S1, as a total of cross-border workers, retired cross-border workers and other mobile pensioners, and for all their family members, we made estimates on the potential users of LTC. We make a difference between benefits in cash and benefits in kind. This is not only an important distinction in the LTC itself, but also in the coordination regulation. We apply on this total PD S1 the same percentages of use of LTC in cash or in kind as is the case in the total population of the EU 27. This is acceptable since the structure of this ‘S1 population’ is similar to the total population including an active population, retired persons and their family members. Those percentages of users are derived from the Ageing report 2012 (additional data was delivered by DG ECFIN, necessarily for making a distinction between LTC in kind, LTC in cash and informal LTC). Based on those figures and the total spending on long-term care in cash and in kind in the EU 27 Member States also the average spending per dependent person benefitting either in cash or in kind benefits is calculated. In Table 25 , we provide for the total population by each country the % of users in cash and in kind and the average amount.

Table 25: % cross-border users LTC in kind or in cash of total population and average amount per dependent person using LTC in kind or in cash (thousand €)

Source: Estimate based on data from LFS, 2012 Ageing Report and additional data delivered by DG ECFIN

We will apply in three scenarios (the baseline scenario is descripted in this chapter and two options are described in chapter 8) those figures to calculate the number of people either benefitting from an benefit in cash or in kind, and the total amount of LTC spending that this implies. For each pair of country of resident combined with a competent country either the percentage of use but also the spending per dependent person can be defined on the level of the country of residence or the country of competence. Both dimensions, % of use and amount per user matter. The level of development of a LTC system is a matter of the ‘breadth’ or the number of persons that might be eligible, and of ‘depth’ or the amount of spending per dependent person. By applying those parameters to the number of cross-border mobile persons, we treat them in the same way as the ‘national’ total population, as is the ambition on this EU Coordination regulation. The difference between the three scenarios is that we make the hypothesis that both the use % and the amount are based on the country of residence or the country of the competent state.

In the baseline scenario we estimate that some 48 thousand mobile citizens are using LTC in kind, defined on the usage rate and average spending per dependent person. This implies a total cost of € 618 million, spend in the country of residence and also to be reimbursed by the competent state. We further estimate the total users of in cash benefits at 45 thousand. The total spending in cash is € 376 million, directly paid by the competent state to the dependent person. In terms of the coordination regulation it is an export of the benefit in cash. In total some 93 thousand users of LTC are estimated or a total budget of € 995 million.

Table 26: Estimated number of cross-border users benefiting from LTC (in thousand) and budget (in million €)

Source: Estimate based on data from LFS, 2012 Ageing Report, Lipszyck, B., Sail, E. & Xavier, A. (2012), Long-term care: need, use and expenditure in the EU-27, EU and additional data delivered by DG ECFIN

Table 27: Estimated number of cross-border users from LTC in kind or in cash, projections 2020 and 2030 (in thousand)

Source: Estimate based on data from LFS, 2012 Ageing Report and additional data delivered by DG ECFIN

Most important creditor countries for the expenditure of LTC benefits in kind are France (€ 171.9 million), Belgium (€ 138.8 million) and Germany (€ 82.1 million) (table). The highest amounts of LTC benefits in cash are also paid to insured persons living in these countries but who are insured in another Member State.

Table 28: Estimated LTC cross-border expenditure baseline scenario (in € .000), by country of residence*

* Amounts are paid by the competent countries

Source: Estimate based on data from LFS, 2012 Ageing Report and Lipszyck, B., Sail, E. & Xavier, A. (2012), Long-term care: need, use and expenditure in the EU-27, EU

It is important that (some of) these estimates could be verified by way of existing administrative data. Important data on health are collected by the Audit Board which is attached to the Administrative Commission (Art. 74 Regulation (EC) No. 883/2004). One of the defined tasks of the Audit Board is “collect the necessary data and carry out the calculations required for establishing the annual statement of claims of each Member State”. It implies that a view on the number and the amount of claims for healthcare in kind, which includes LTC, should be obtained from this information. However, it is limited to benefits in kind and not benefits in cash, so even there is this official source not ‘exhaustive’ and are additional (probably national) sources and data collection at European level needed. There is also another limitation. At the moment, the Audit Board reporting is (also legally) oriented at outstanding stocks of mutual claims. The budgetary impact of those LTC expenditures is in terms of flows of yearly expenditures and yearly new claims for reimbursement of those expenditures. Unfortunately, at the moment there is no specific reporting of the annual bilateral new claims between countries but only of the outstanding claims. The total new claims by creditor country were however available for 2011 in the reporting of the Audit Board. We applied the bilateral distribution of those totals over the debtor countries based on the structure of the outstanding stock. So assumptions could be made to estimate the amount of claims of LTC benefits in cash. Data concerning the claims introduced on health care by the creditor countries (based on fixed and actual amounts) in 2011 (taking into account that these claims submitted in 2011 will deal about provided health care mainly from 1 or 2 years ago) have been used to estimate the amount of claims of LTC benefits in kind received or issued (cf. 10.4.2 ). What we already know from data available within the Audit Board are the outstanding claims from creditor countries divided over the debtor countries (cf. 10.4 ). This detailed breakdown by debtor country for the outstanding claims is also used for the newly introduced claims in 2011 (cf. 10.4 ). So we assume that the breakdown of the newly introduced claims by debtor country is similar to this of the outstanding claims. It implies also that debtor countries which have important delay’s in payment will influence this assumption. This assumption results in a detailed cross-table (debtor and creditor country) of the LTC claims introduced in 2011. On top of that an additional hypothesis needs to be made on the share of LTC benefits in kind compared to the total level of health and LTC expenditures (by using variables ‘Health care spending as % of GDP’, ‘Long-term care spending as % of GDP’ – 2012 Ageing Report and the breakdown of LTC by type – see Lipszyck, B., Sail, E. & Xavier, A. (2012)). This results in a cross-table of introduced claims on LTC benefits in kind whereby also the debtor countries are known (see Table 29 ). We arrive at a total amount of LTC reimbursement of € 592 million. This grand total, estimated on different sources with different hypotheses, is very similar to the estimate we made in the previous exercise (also some € 618 million, in the baseline scenario).

The three main creditor countries based on the administrative data from the Audit Board are France (€ 207.2 million), Belgium (€ 113.3 million) and Germany (€ 74 million). These are the same main creditor countries as obtained by way of the calculation model based on several assumptions (table above). At the same time also the amounts of reimbursement claims are more or less similar to each other. Nevertheless, we observe an important underestimation of the claims issued by Spain in the calculation model when observing the administrative data from the Audit Board.

But also the three main debtor countries are the same in both methodologies, namely Germany, Luxembourg and the Netherlands. Also the received claims are to a high extent similar for Germany and the Netherlands but differ somewhat for Luxembourg.

Finally, also data we received from Belgium on the total number of issued claims on health can be used. In 2011, Belgium issued an amount of € 336 million claims of which € 137.2 million claims for persons aged 65 and older (which can be considered as the main group of dependent persons and recipients of LTC in kind). This amount is very close to the estimated issued claims on LTC by Belgium, namely € 138.8 million.

For the estimated figures on total health spending, we observed that at least for the EU 27 total but even for the individual Member States, they seems to be good proxies for the figures we found in those administrative data of the Audit Board. This could be an indicator the methodology used is reliable, but also that the real expenditures converge with what is expected because this is a mature system of social protection and well settled ways of coordination. For few of the estimated values on LTC on even the totals by country of residence or competent state, we obtained in the survey administrative information. It was so also difficult to systematically cross check our estimates with this administrative information. There are further several reasons why our theoretical calculations based on the hypothesis that the cross border citizens might differ from official statistics. We observe that most of the times no separate statistics could be obtained, or were made. Other reason is that LTC is not always recognized as such in the national situation, or is not considered as falling under the coordination regulation, or finally there can be a lack of knowledge on those entitlements, leading to non-uptake. This can result in substantial differences between what could have been found, and what we estimate in theory.

For cross-border workers different rules apply when the insured person is worker or pensioner. Within the sickness chapter of Regulation (EC) No. 883/2004 specific rules are adopted for active cross-border workers and their family members (art. 17 and 18) which will also have an impact on LTC. Cross-border workers have the right of choice to receive sickness benefits in kind in the country of residence (as defined in art. 17 – in accordance with the legislation in the country of residence and reimbursed by the competent country) or in the competent country (as defined in art. 18,1 – in accordance with the legislation in the competent country). It implies that about 1 million insured cross-border workers (with a PD S1) have the right to choose between LTC benefits in kind in the country of residence OR in the competent country. This will have budgetary consequences for the competent country, especially when the cross-border worker is taking a rational decision taking into consideration the most advantageous, most extensive, most expensive ‘LTC insurance package’. However, we observed from a recent study (Pacolet, De Wispelaere & De Coninck, 2012) that the main reason to choose for a specific health care system is the familiarity with the health care provisions. Most of the time, this will be in the country of residence. Despite this right of choice, most cross-border workers choose to receive benefits in kind in their country of residence (Ibid.). This right of choice is also applicable to the family members. However, art. 18.2 states that family members of a frontier worker “shall be entitled to benefits in kind during their stay in the competent Member State, unless this Member State is listed in Annex III of Regulation (EC) No. 883/2004”. Restrictions appear for Denmark, Spain, Ireland, Netherlands, Finland, Sweden and United Kingdom. Retired cross-border workers lose this right to choose. However, Regulation (EC) No. 883/2004 introduces some new rules applicable to this specific group (art. 28). Retired frontier workers are entitled to benefits in kind in the country of last activity, insofar as this is a continuation of treatment begun in that Member State (art. 28.1). Also, retired frontier workers (and their family members) who have worked at least two years in the five years before their retirement as frontier worker will be entitled to benefits in kind in the country of last activity (art. 28.2). This only applies if the country of last activity and the competent country are both listed in Annex V of Regulation (EC) No. 883/2004: Belgium, Germany, Spain, Luxembourg, Austria and Portugal. The retired cross-border worker will need a PD S3 ‘Medical treatment for former cross-border workers in former country of work’ to receive benefits in kind the country of last activity. We did not make any hypothesis on this right of choice.

Table 29: Estimated claims of LTC benefits in kind introduced in 2011, by debtor country, in .000

Source: Estimate based on data from the Audit Board, 2012 Ageing Report and Lipszyck, B., Sail, E. & Xavier, A. (2012), Long-term care: need, use and expenditure in the EU-27.

Introduction

For long-term care benefits, we have applied the same approach as for the unemployment benefits due to the same limitations with regards a standard process including the same Information Obligations in the sense of the SCM methodology.

For long-term care, it appears that the situation is even more complex than for unemployment benefits, as it encompasses different cares that are not understood in the same way in the whole EU and that imply a fragmented landscape of responsible and implementing actors in some Member States (e.g. over 70 different bodies are involved in Germany, while each of the 17 regions of Spain also has a different system; in many Member States, local entities are a dominant actor, etc.). The different national specificities result in a large variety of situations which may have a significant impact on the administrative burden when dealing with cross-border cases for long-term care.

There a number examples demonstrating the complexity of the processing of cross-border cases for long-term care which can result in administrative cost and burden for Member States’ authorities:

According to the interviewees, there is legal uncertainty about which benefits should be coordinated under the Sickness Chapter. Some countries still do not consider the care (social assistance) they provide as being included in the Sickness Chapter;

In our survey to the national administrations, around 50% of national administrations that are opposed to changes to the current coordination rules state that the current rules need only to be better applied in practice and to be better explained. National administrations who are in favour of a change of the current rules say that the identified problems (legal uncertainty, complex regulation and uneven applications of the rules by Member States) will persist if no change occurs;

Not in all Member States (particularly not in Member States that generally are in favour of keeping the status quo such as Germany, Austria and the Netherlands) administrative burden was perceived a major concern by national administrations. One Danish interviewee gave the following argument to put the administrative burden into perspective:

“Before 2009, Germany did not ask reimbursement to Denmark for costs that it incurred by provision of LTC services to citizens that fell under the Danish system, based on a special agreement between both Member States. However, recently, Germany asked to reintroduce a reimbursement system again between both countries. The fact that Germany asked to reinstall a reimbursement system again shows that other aspects seem more important for Germany than administrative burden from reimbursement claims, for example the financial impact of LTC services provided by Germany.”

A German health insurance considered the reimbursement of LTC benefits to be slow and problematic from an administrative point of view:

“There are EUR 500 000 – 600.000 interest costs per year that my organisation has to bear because of non- or late payment. The reimbursement mechanism is not functioning well and needs a substantial revision of the rules. There is an outstanding amount of EUR 12-13 million in 2013. Late payments can be the result of checks/scrutiny of services rendered by the country of residence; however, the checks do not justify a waiting period of up to 18 months in some cases. The time needed for checks should be reduced substantially. The reimbursement mechanism could be made more efficient by using lump sum compensation mechanism. However, the views about its effectiveness are divided amongst the insuring companies.”

An Austrian representative of a health insurance fund confirms the long processing time of reimbursement:

“Particularly the reimbursement of LTC benefits in kind by the competent MS poses difficulties. Often, the information about the amount/costs of benefits in kind rendered by the Member State of residence reaches the competent Member State (which reimburses these costs) very late. Regularly, it takes 1-1.5 years to reimburse such claims. The rules stipulating information procedures should be more detailed (e.g. duty for monthly information provision of the value of the benefits in kind rendered by the MS of residence). The time-bound provision of information by all Member States is of key importance as to ensure an effective application of the reimbursement mechanism”.

In general, regardless if they believed that administrative burden from the current rules is a major problem or not, only a small minority of national administrations have a good view on the actual administrative burden or are able to support their arguments with quantitative data or a detailed description of the burden. The lack of concrete (quantitative) evidence adds to the difficulty for making a sound judgment about this issue.

In terms of substantiation of the administrative costs related to the current rules and considering the limitations of the application of the SCM methodology in this exercise, we present in the table below the estimated costs related to processing of the PD S1 document.

Despite the data limitations resulting from the problems discussed in this chapter, the assessment of the administrative cost (baseline scenario) for the PD S1 document provides a robust basis for assessing the theoretical impact (positive or negative) of the different policy options on the administrative cost.

The methodology for assessing the administrative cost is based on the following formula:

Time (T) x Wage (W) x Number (N)

The hourly rate is EUR 18 per hour. We provide an estimate for the total number of cases for processing PD S1 documents for the EU-27.

Estimated current administrative cost (Baseline Scenario)

The PD S1 form allows a person to register for healthcare in the country of residence. This form is delivered per person (not per family). The number of PD S1 forms issued provides insight into the number of people who (may) receive LTC benefits in another Member State. In the framework of this study, we have collected data on the number of PD S1 documents ‘issued’ by category of citizen and have estimated expenditure on LTC benefits. In addition, we have collected data on the number of PD S1 documents issued for Poland and Belgium by means of a workshop with experts in the respective countries. In this section, we use the data available to calculate the estimated administrative cost for processing a number of documents related to long-term care in a cross-border case.

In order to assess the administrative costs for the EU-27 stemming from the processing of the PD S1 documents, we have carried out the following steps according to the Standard Cost Model (SCM):

1.Calculation of the unit cost per case:

The unit cost per case (processing/handling of a PD S1 form by the administrative staff – clerk) provides insight into the total cost for processing one PD S1 document. It is based on the following formula:

Time (T) x Wage (W)

During the workshops in the Member States, we have collected data on the average standard time spent for processing/handling a PD S1 document. Robust data are available only for Poland. The estimated time for processing one PD S1 document in Poland is estimated at 60 minutes.

The hourly rate for processing the administrative tasks is EUR 18; this results in a rate per minute of EUR 0.3 (EUR 18/60 minutes). The average unit cost for the EU-27 per case of handling a PD S1 document is EUR 18 100 . It is calculated on the following basis: Time (60 minutes) x Wage (EUR 0.3).

Caution should be paid when interpreting this estimated unit cost as the result is based on an example of one country only (Poland) which seems to have a rather efficient way of processing PD documents (see also the discussion on the processing of PD U1 documents above). It can be expected that the time for processing a PD S1 document in the other Member States may differ (substantially). Due to data limitation, however, we have calculated the administrative cost on the basis of the Polish example.

2.Number of cases:

In our research, we have estimated data for the number of PD S1 documents ‘issued’ for the EU-27 countries on the basis of our own calculations based on data from LFS (for a detailed discussion on the estimated number of PD S1 issued by category of citizen, see section 4.2.5 in this report). The total estimated number of PD S1 documents ‘issued’ in the EU-27 is estimated at around 1 980 000.

3.Calculation of the administrative cost (per Member State and for the EU-27)

We have calculated the administrative cost for processing PD S1 documents on the basis of this formula:

Time (T) x Wage (W) x Number (N)

The table below presents the total estimated administrative cost for processing PD S1 documents. The estimated total cost for the EU-27 is EUR 35 632 000. Within the EU-27, the estimated total cost for processing PD S1 documents was highest (˃ EUR 3 000) in a number of the old Member States (in descending order): Germany, the UK, Luxembourg, the Netherlands, Austria and Italy. It was lowest (˂ EUR 100) in a number of the new Members States (in descending order): Cyprus, Bulgaria, Slovenia, Estonia, Malta, Lithuania and Latvia.

Table 30: Estimated administrative cost - PD S1 'issued', EU-27, EUR, 2013, in 000

Source: Own calculations based on data from LFS and workshop in Poland

We have also calculated the administrative cost for processing a number of other documents related to long-term care benefits for Poland by applying the following formula: Time (T) x Wage (W) x Number (N). Data were collected for the following documents:

Member State of residence:

Service of E125 forms.

Competent Member State:

Request for the issue of S1 document/ E100 series form (service of E107/E001 forms);

Registration of the S1 document;

Registration of the E100 -series form (part B);

Service of SED S001 documents;

Issuing E125 forms.

There are no data available for the EU-27 for these documents; a calculation of administrative cost for these documents is therefore not possible at this stage. We present the data only for Poland, where robust data are available. The table below presents the total estimated administrative cost for processing the documents for Poland presented according to ‘Member State of residence’ and ‘Former working Member State’:

Table 31: Estimated administrative Cost – Competent Member State, E125, Poland, EUR, 2013

Source: Deloitte, workshop in Poland 

Table 32: Estimated administrative Cost – Member State of residence, E125, S1/E100/E107/E001, S001, Poland, EUR, 2013

Source: Deloitte, workshop in Poland

4.3Summary – Estimated current administrative cost - Baseline scenario

The table below summarises the administrative cost for the EU-27 for the following documents for the baseline scenario: PD U1 ‘received’, PD U2” ‘issued’ and PD S1 ‘issued’.

Table 33: Baseline scenario – estimated administrative cost: PD U1 (in €), PD U2 (in €), PD S1 (in € 000)

Source: Own calculations based on collected administrative data and 2012 Ageing Report and data provided during the workshops on administrative burden (Poland, Belgium and Romania).

5Problem definition

5.1Introduction

The free movement of persons, one of the “four freedoms” offered by the internal market 101 , is one the most important principles of the EU and a fundamental right of EU citizens. The rights to move, to reside and to work freely within the territory of the Member States are enshrined in both the Treaties (Article 21, 45, 49 and 56 TFEU) and the Charter of Fundamental Rights (Preamble, considerations 15 and 45). Legally resident third country nationals can also enjoy certain rights to free movement within the territory of the Union 102 .

Preventing disadvantages in the social security rights of citizens when they move is necessary to make the right to free movement effective. As such, Article 48 TFEU states that “The European Parliament and the Council shall … adopt such measures in the field of social security as are necessary to provide freedom for workers”. To comply with such mandate, currently Regulation (EC) No 883/2004 and the Implementing Regulation (EC) No 987/2009 coordinate Member States’ social security schemes. Regulation (EC) No 883/2004 started to apply in May 2010, and replaced previous Regulation (EEC) No 1408/71.

The EU legislation does not replace the different national social security systems, but coordinates them in situations with an intra-EU cross border element. EU regulations coordinate the cross-border aspects of the social security systems of EU Member States plus Iceland, Liechtenstein, Norway and Switzerland.

Removing the obstacles to mobility for European citizens is on the priorities of the European Commission. It has been highlighted on several occasions under President Barroso’s tenure (including with occasion of the European Year of Workers Mobility 2006, the European Parliament report on citizenship in 2009, the Monti report on the Single Market, President Barroso’s political guidelines, Commission Work Programme 2013) that despite the important legal acquis in the area of free movement of workers, European citizens still face problems and obstacles when moving across borders within the EU. Labour mobility across Member States remains low 103 , as emphasized in the 2012 Annual Growth Survey 104 . In line with this priority, the Commission issued a policy communication in April 2012 (“Employment package”), in which it identified the EU’s biggest job potential areas and the most effective ways for Member States to create more jobs. Among other objectives, the employment package also aims to contribute to a genuine EU labour market.

In the same perspective, the Commission announced, in the EU Citizenship report 2013 "EU citizens: your rights, your future" 105 that it would propose a revision of the social security coordination regulation, looking in particular into extending the export of unemployment benefits for longer than the mandatory three months, to make it easier for citizens to look for a job in another EU country (action 1).

The EU regulatory instruments in the area of social security coordination constitute necessary elements for making intra-EU mobility a reality and for the smooth operation of the EU labour market. They date back to the 1950’s and have been amended on numerous occasions in order to take into account developments at EU level, changes at national level and rulings of the Court of Justice. Both the Member States and the Commission have the obligation to make sure the Regulations are fit to meet today’s needs and reflect the developments in national and EU legislation, the case-law of the Court of Justice and the socio-economic context.

Regulation (EC) No 883/2004 represents a step in the modernisation of social security coordination rules, adapting them to changing realities at EU and national levels. The legislative process for the modernised rules lasted over 10 years, a period during which important socio-economic changes took place in the EU and the Member States (including two enlargements) and where the Court of Justice delivered a number of important rulings. The regulation strengthened the principles of coordination and brought improvements in several social security branches. It did not, however, in the Commission’s view, lead to the expected results in the areas of coordination of unemployment and long-term care benefits.

To fully align coordination in these areas with developments at EU and national level, and meet the citizen’s needs, the Commission has set in motion the process for a possible partial revision of Regulations (EC) Nos 883/2004 and 987/2009. The initiative covers the area of coordination of unemployment and long-term care benefits. It is also linked to the overarching EU objectives as reflected in the “Europe 2020 – A strategy for smart, sustainable and inclusive growth” 106 , which calls for the EU to encourage mobility and for European citizens to make more sue of their freedom of movement, as well as to ensure that a vulnerable part of the EU population is enabled to live in dignity and is not expose to the risk of poverty when exercising the right of free movement.

The initiative for a partial revision of the coordination regulation includes the present preparatory study for an impact assessment. As a first step, the current problems with regard to the coordination of long-term care and unemployment benefits are described and assessed. Further background details are briefly presented for each area and then the existing problems are discussed in depth. A problem tree was drafted based on documentary analysis, survey data complemented by statistical data analysis and interviews with stakeholders in 14 Member States. The problem tree reflects the problems discussed in detail in the text of the report and their underlying causes (drivers). It also links them to the policy objectives, which are divided into operational, specific and general objectives.

Given the substantial and thorough work conducted by the trESS network on the topic, we refer to the studies conducted by the network on both topics for an in-depth legal analysis on existing and potential problems under the current coordination Regulation. The reports have been used as sources for this study.

5.2Coordination of LTC benefits

Long-term care benefits were not explicitly mentioned in Regulation (EEC) No 1408/71. This is still the case in Regulation (EC) No 883/2004, with the exception of one overlapping provision (art.34). The coordination regulation does hence not include a definition of “long-term care”, nor does it contain a specific chapter for the coordination of these benefits. Long-term care benefits do however fall within the material scope of the coordination Regulation. The Court of Justice ruled that, in the absence of a specific legal regime for their coordination, they must be regarded as “sickness benefits” within the meaning of the Regulation and coordinated as such.

Long-term care benefits are increasing in importance given the demographic changes in the EU (namely, the ageing of the population). Member States continue to develop special schemes for persons in need of care, and national legislative developments (including new types of benefits) abound but vary across Member States. Since the last decade, the European Union has been promoting access, high-quality and sustainable healthcare and long-term care in Member States. 107   108 On 9 September 2010, the European Parliament adopted a Resolution on “Long-term care for older people” calling for a development of the Social OMC. To take into account this development, the Commission adopted in 2013 the Social Investment Package, describing the national orientations in this field.

Given these developments, the number of rulings delivered by the Court of Justice in this area and the problems persisted in the area (as highlighted by the trESS network) justify analysing the possibility of revising the coordination rules.

The current system of coordination of long-term care benefits thus presents a series of problems to which the “no policy change” option would not provide any solution. They are described in detail below, based on the information gathered through the various data collection activities conducted for this study. The problems are summarised visually in a problem tree, included at the end of the section. The problem tree links the problems to objectives, and also depicts the drivers behind such problems. General information about problems and the stakeholders’ view on the need to change the rules is first presented.

40% of the organisations that participated in the EC public consultation declared to have noticed problems of application of EU law in this field. In Deloitte’s survey of national administrations, 60% of national administration has experienced problems in the application of the current coordination rules.

A general complexity of the EU Law regulation on the coordination of long-term care is widely regarded as root problem of the current system according to the interviews conducted in 14 Member States. This complexity results in and is at the same elicited by a number of more specific problems.

The most recurrent problem, mentioned by 24% of all type of organisations participating in the public consultation was that migrant workers are not sufficiently aware of their rights under EU law. The second problem most commonly mentioned was that national administrations do not apply EU law correctly. Then came the problems with the cross-border communication between institutions of Member States and that national administrations do not provide EU workers with sufficient and correct information.

Looking into specific problems and who raised issues with them in the consultation, the lack of awareness of migrant workers about their rights was mostly pointed out by social partners and trade unions and national administrations. The incorrect application of EU law by national administrations was pointed out mainly by national administrations. No national administration, social partner, trade union, non-governmental or civil society organisation considered that migrant workers are abusing the possibilities of EU law. Two respondents (one on behalf of a company and an unknown one) did so. However, none of them was able to elaborate on such claim.

Interviewed civil society organisations active in the area (EUROCARERS and the AGE Platform) stated problems related to the lack of effective protection of migrant workers. However, it must be noted that some of these claims were linked to the residence requirements in national legislation, which is beyond the scope of the (coordination) regulation. They also concerned the interpretation and application of the law at national level.

The general attitude towards the need (or lack thereof) of modifying the rules of the different stakeholders participating in the EC public consultation was the following:

Table 34: General attitude towards need to change the current LTC coordination rules

Each country has its own social security system, where the risk of reliance of care is covered in very different manners. Differences include the definition (or lack thereof) of a definition for long-term care, the branches of the social security system and/or public assistance schemes under which the risk of reliance on care is covered, and the specific type of benefits granted.

Lack of common definition of LTC benefits under EU law & Lack of common criteria to determine them

There is no agreed definition of LTC benefits at EU level. There is no agreement on the definition of the risk of reliance on care and to determine common criteria for LTC benefits.

The OECD has adopted its own definition of long-term care, but it has no binding legal status under EU law. 109

Member States might or might not have an official definition for long-term care. Among those that have definitions, these vary. The trESS Think Tank Report 2011 on the coordination of long-term care benefits (“Coordination of Long-term Care Benefits – current situation and future prospects”) included the following synaptic table regarding the definition of LTC in the different Member States:

Table 35: Lack of common definition of LTC benefits under EU law - comparison of definitions

The European Court of Justice established in the Molenaar case (C-160/96, Para.3) that a person is reliant on care if “a permanent need were to arise for those insured to resort, in large measure, to assistance from other persons in the performance of their daily routine (bodily hygiene, nutrition, moving around, housework, and so on)”. In various rulings, the ECJ has outlined elements of LTC benefits that are to be regarded as sickness benefits for the purposes of coordination. trESS compiled in an annex to their Think Tank Report 2011.

The national administrations interviewed (including quite clearly Austria, Belgium, Germany the Netherlands and Spain) considered the lack of a common definition or criteria as a major obstacle for an effective coordination of LTC benefits and a smooth implementation of the current rules. National administrations in many Member States expressed their preference for a common definition of LTC and a detailed list of LTC benefits per Member State. However, some Member States - Sweden, Finland en Belgium – expressed their concern that several benefits in their system could then be seen as LTC benefits, while they do not consider them to be subject to the current coordination rules.

Diversity of benefits that can/should be considered as LTC benefits in the MS & Benefits situated under different social security branches

The risk of reliance on care is covered under different branches of the social security system and/or public assistance schemes.

This further complicates coordination at EU level, as “national” benefits are often coordinated under different regimes (including sickness, old age, family protection, work accidents and occupational diseases, invalidity, survivors and social assistance); they can be in kind or in cash or even a mixture of both; they can be social security benefits or social assistance benefits, contributory or non-contributory; they are processed by a wide range of bodies and institutions, which complicates control. In addition, the LTC beneficiaries are diverse (and might vary across countries: workers, unemployed persons, pensioners, survivors, family members) and the development of LTC benefits is not homogeneous in the different Member States. Differences in national systems and no common understanding of the concept complicate the coordination of LTC benefits at EU level.

The insights gained during the country visits confirmed the diverse ways in which Member States cover the risk of reliance on care and how this, in the view of the public officials directly in charge of applying the coordination rules, complicates the coordination. While EU rules aim to coordinate, and not harmonise, the national social security schemes, a root problem is perceived to be in the different manner in which the Member States recognise, consider and deal with LTC benefits.

Firstly, Regulation (EC) No 883/2004 aims to coordinate social security schemes. However, the risk of reliance of care is not covered in all countries by branches of the social security system. It is, in some cases, covered by public or social assistance schemes. Secondly, following the ECJ’s case-law, LTC benefits are to be coordinated as sickness benefits. However, LTC benefits present certain distinctive characteristics that make them different from sickness benefits, and benefits addressing the reliance on care might be coordinated under different social security branches (including family, old age and invalidity) in different. Thirdly, while the ECJ has outlined elements of LTC benefits that must be regarded as sickness benefits for the purposes of coordination, but there is still no common definition or agreed exhaustive list of criteria that allow to identify benefits as LTC benefits (however, the Court has ruled on particular benefits of certain countries (as is the case in the UK) that must be considered as LTC benefits for the purposes of coordination). All this renders the precise identification of benefits, provided under any social security of public assistance scheme, that must be considered as LTC benefits for the purposes of coordination. It generates uncertainty and hampers the understanding and ownership of the rules by national administrations, which might hold different views on the benefits that fall under the scope of the coordination regulation.

Four basic positions were identified during the country visits:

-Specific, dedicated system of LTC;

-LTC benefits considered as sickness benefits;

-LTC as a complement to pensions;

-LTC as social assistance.

The trESS network attempted to provide a picture of the statutory organisation of LTC in all Member States, simplifying to a certain extent the characteristics of national schemes to group countries in six different categories. The following, included in their Think Tank Report 2011, summarises their work:

Table 36: Diversity of benefits that can/should be considered as LTC benefits in the MS & Benefits situated under different social security branches

* Some (or all) of the benefits of these states have been declared as “normal” sickness benefits for the purpose of the application of Regulation (EC) No. 8823/2004 by the ECJ or the EFTA Court.

The main issues have been identified through interviews in countries for which LTC benefits are not part of the Social Security system. This refers mainly to countries in which LTC is considered as social assistance, or assimilated to social assistance. Social assistance is explicitly excluded from the coordination Regulations (Art.3.5 Regulation (EC) No 882/2004). In these cases, the transposition of the Regulation into the national legislation, and the subsequent application of the coordination mechanisms, becomes highly problematic. Amongst the countries visited, this is the case in Estonia, Romania (where the benefits that would fall under LTC are part of the social assistance system), Spain (although a specific regulation for LTC exists in Spain, these benefits are not part of the Social Security system; they are provided as “Social services”, which can be assimilated to social assistance), Slovakia and France (there is no explicit definition of LTC benefits in the French legislation) 110 . This complicates the application of the coordination rules in these cases.

As an example from the visits conducted to these countries, the authorities from the National Health Insurance Institute of Romania (Casa Naţională de Asigurări de Sănătate) considered that, given the lack of definition of long-term care regulated at EU level and materials submitted by the Administrative Commission, the coordination of long-term care benefits does not fall under their competence. They referred us to the Ministry of Labour, Family and Social Protection, which in turn declined to have further discussions, indicating that “The Regulations specify very clearly that they do not cover social assistance, only social security. The benefits that would fall under long-term care benefits are of social assistance”. Similarly, the officials of the French Direction Générale de la Sécurité Sociale expressed the difficulties they are suffering to apply the coordination rules. In France, benefits are granted by territorial bodies, and are not linked to contributions. Citizens do not contribute during their lives to cover the risk of reliance on care, but are granted benefits if and when in need of them, based on their residence. Benefits in kind are provided directly in care centres or at the person’s home. Both the degree of reliance on care and the resources of the concerned person are assessed. However, it is not possible to identify the citizens who have worked and/or resided abroad. While there is no lack of willingness to comply with the coordination rules, there are practical difficulties. An internal procedure has been launched to try and solve the issues.

The branch of the Social Protection system under which the benefits fall in each country determines the method by which they are financed, which in turns impacts the willingness of Member States to pay for or reimburse certain benefits provided to mobile citizens. In the EU, contributory benefits (based on LTC contributions by citizens) and non-contributory benefits (based on residence) co-exist. Similarly, benefits are means-tested (that is, the resources of the person in need of care are assessed to determine his/her contribution to the costs of providing the benefits) in some countries, but not in others.

LTC benefits can be in kind or in cash. The benefits available in each country, the way they are provided, and the eligibility conditions, vary per country. In an attempt to shed some light on the diversity of benefits per country, trESS grouped them according to the type of benefits (in kind, in cash or both) and their organisation (provider, spending and benefits) per Member States. The following table is included in the Think Tank Report 2011:

Table 37: Diversity of LTC benefits system: in cash, kind or both

trESS updated this mapping in their Analytical Study 2012 “Legal impact assessment for the revision of Regulation 883/2004 with regard to the coordination of long-term care benefits. The initial mapping was complemented with the replies submitted by the Member States to a questionnaire sent to the Administrative Commission members (A.C. 18/12). The questionnaire included questions about the definition, mapping and description of the national schemes. The results are presented in annex to the report, with a detailed list of the benefits in each country that should be considered LTC benefits for the purposes of the coordination. However, it should be noted that this list has no official validity. It only constitutes a (highly elaborated) theoretical exercise. During the country visits, the list of LTC benefits per country prepared by trESS received mixed reviews.

In line with the general problem of classification of LTC benefits under EU law and the drivers outlined, the trESS Think Tank Report 2011 identified and discussed in more detail the following challenges:

-Differences in national systems – no common understanding;

-Huge variety of LTC systems;

-Social security or social assistance;

-Benefits with elements of LTC considered as benefits from other branches of social security by Member States.

Distinct character of LTC benefits

LTC benefits vary across countries. They can be in cash or in kind, which affects the determination of the competent country for providing them and for bearing their cost. Their variety across countries is compounded by the fact that certain Member States are introducing new mixed-types of benefits. In these cases, it becomes increasingly difficult to determine the nature (in cash or in kind) of the benefit.

The same or similar benefits might be considered in kind in one country and in cash in others (this was the case, as reported by the national experts involved in trESS Analytical Study 2012). Regardless of how the benefits are considered or recognised in domestic legislation, however, the ECJ case law should be applied with regard to the distinction between benefits in kind and in cash.

As outlined by trESS, traditional benefits in kind do not usually present any problems. However, benefits in cash are more problematic. In these cases, the circumstances under which the money is given must be considered. In case money is given to the person so that he decides the use to make of it, respecting his autonomy and being paid out irrespective of any bill to be presented, the benefit should be considered as a cash benefit. In any other case where the use of the money must be justified by bills on services purchased, the benefits should be considered in kind.

Whilst the distinction between benefits in kind and in cash is not particularly problematic for sickness benefits, it becomes more complicated for LTC benefits (which must be coordinated as sickness benefits). We refer to trESS Analytical Study 2012 for a more detailed analysis of the problems the distinction between benefits in kind and in cash can cause in the following situations:

-Loss, withdrawal or suspension of LTC benefits when the beneficiary changes his or her residence to another country;

-Accumulation of the benefits in cash of the competent Member State and the benefits in kind for the same purpose of the Member State of residence (Article 34 of Regulation (EC) No 883/2004);

-The application of Article 34 of Regulation (EC) No 882/2004 in relation to Member States which have opted for lump-sum reimbursements;

-General aspects of accumulation of cash benefits and/or benefits in kind also outside Article 34 of Regulation (EC) No 883/2004.

Beyond the pure distinction between benefits in kind and benefits in cash, there are other mismatches caused by applying the coordination rules devised for sickness benefits to LTC benefits. Sickness and LTC benefits differ in their aims, instruments and means. Several articles of the Regulation (including Articles 19, 21 and 28) are not designed for LTC benefits but should be applicable to them, even if their wordings do not refer to them. In terms of goals, while traditional sickness benefits in cash are mostly related to the loss of salary or income, LTC benefits in cash can have other goals (therefore making the provisions on the calculation of benefits in cash under Article 21 (2) to (4) of little importance to LTC benefits). These problems were also identified by trESS, which considered as a challenge the “mismatches of concepts as a result of the inclusion of LTC benefits in the chapter on sickness benefits”.

To complement the information provided about the problems of classification of LTC benefits under EU law with insights from the country visits, during these several national administrations (in Germany, Austria, Poland and the Netherlands) expressed to be in favour of clearly listing all LTC benefits (in kind and in cash) that are considered to be LTC benefits in other Member States or that should or could fall under the coordination Regulations. A German health insurance fund representative stated that “the lack of a clearly defined list of LTC benefits makes any coordination between Member States difficult and slow”. An interviewee from the Austrian national administration added: “The present list attached to the Regulation is helpful but insufficiently concrete and meaningful for categorising the different national LTC benefits (in kind and in cash). Therefore, a concrete ‘benefits catalogue’ should be drafted, listing the available benefits in EU/EFTA and Switzerland with precise definitions of the types of benefits, the level of the amount, the essential eligibility criteria and the competent administration.”

Similarly, these national administrations would also like to better define “social assistance”, as more benefits could be qualified as a LTC benefit than often believed. A representative of a Polish national administration stated with regard to social assistance: “Several Member States claim to only have social assistance benefits, which fall outside the scope of the Regulation. However, some of these might be qualified as LTC benefits”. A Danish interviewee proposed to use the decisions of the ECJ as input to a common definition of LTC and the listing of LTC benefits (in particular the Molenaar case 111 ).

The application of the coordination rules is uneven among countries. It appears that the exchange of portable documents, the recognition of entitlement to LTC benefits, and the claims for reimbursement (impacting, in the cases foreseen in Article 34 of the Coordination Regulation, the cash benefits perceived by the beneficiary) would vary significantly across countries. This was reported during the interviews conducted in the country visits. The biggest problem according to the national administrations that responded to the EC public consultation is the incorrect and uneven application of the EU rules by national administrations (32% of them see this as a major concern). This is confirmed by another opinion revealed by the EC public consultation: among the organisations considering that there is no need to change the rules (53.3% of the total), the majority (79.17%) was not entirely satisfied, considering that they should be better explained or better applied in practice.

As an illustration of the opinion of public authorities believing that the rules do not need to be changed, but should be better applied in practice, a German health insurance fund representative stated the following: “Local health insurance funds - often the entry point for citizens’ claims – are insufficiently informed about the current rules and regulations which apply according to the EU Regulation. Current problems result from a wrong application of the current rules, not from the legal text itself; the coordination rules as such are sufficient and considered as good. The current problems could be reduced substantially, if the current rules were implemented more thoroughly by the different stakeholders involved in providing long-term care services in the different Member States”.

The differences in the way in which Member States cover the risk reliance on care and the different social security branches and/or social assistance schemes under which they are granted can lead to an uneven application of the coordination rules. Member States hold their own opinions on which benefits provided under their legislation should be considered as LTC benefits falling under the scope of the coordination rules. They cling strongly to such opinions, as evidenced during the country visits conducted, including expressing still disagreement with ECJ’s rulings (such was the case in the United Kingdom for instance). European organisations consulted (EUROCARERS and the Age Platform) raised concerns about the incorrect application of the coordination rules in some Member States. trESS encountered similar issues, as reported in their Think Tank Report 2011: “Many Member States do not recognise some benefits as LTC benefits that at European level would be considered as such. In some cases those benefits are considered, at national level, as social assistance benefits and are therefore excluded from the scope of the Regulation. In other cases, the fact that the benefits are still considered (by the said Member States) as special non-contributory benefits of a mixed type prevents them from being exported. Both ways of interpretation seem contrary to the spirit of European law. Unfortunately, it is not easy to convince Member States to change their minds and behaviour.”

The current coordination rules make it possible to leave a person reliant on care completely devoid of protection in very specific cases. A person reliant on care may be left with no protection if he/she receives LTC benefits (only) in kind from a Member State and moves to a Member State that covers the LTC protection only by means of benefits in cash 112 . 

This risk of a gap in protection of mobile persons was regarded as a minor problem by the majority of the national administrations: only three organisations responding to the EC public consultation considered it as a major problem. National administrations mostly considered a loss of social protection as a theoretical possibility and they were not able to identify effective cases of a complete lack of protection.

Beyond this gap in protection resulting directly from the design of the coordination rules, problems in the protection of mobile citizens reliant on care can appear due to the uneven application of coordination. As outlined in the previous points, trESS perceived issues in this regard and European organisation active in the area consulted for this study also voiced their concerns, quoting actual cases during the interviews. These same organisations also alluded to the fact that the imposition of residence conditions to perceive long-term care benefits in kind, as is the case in several Member States, also results in an effective lack of protection; mobile citizens reliant on care would be less protected, or completely unprotected, during the time spanning from their change of residence to the completion of the requested period of residence.

On another hand, significant disparities in the level of development of LTC benefits across countries have been observed. This was not perceived as an issue per se by the Member States. The generosity of each national system will always be different, and must be assessed in the context of the whole national social protection system. Social protection systems are designed with the aim of being internally balanced, and tend to compensate the relative weaknesses in certain benefits with more generous provisions in other fields. To some extent, individual workers will always have to gauge the amount and generosity of benefits they would be entitled to in each country when making a decision about mobility. In this sense, it should be borne in mind that the Regulation aims for the coordination, and not harmonisation (this falls out of the EU competencies in the field) of the systems.

However, interviewees in Member States with more generous benefits did point at the risk of losing quality of care when moving to another Member State, due to the fact that LTC benefits are not developed homogeneously in all Member States. In some, a mature, high-quality LTC scheme exists, while in others LTC benefits hardly exist. This makes LTC benefits different from other social security benefits (where the issue is more about higher of lower benefits in different Member States).

The heterogeneous development of LTC benefits in all Member States was also considered a challenge by trESS. As was outlined in their Think Tank Report 2011, a complete scheme exists in some countries, while in others this type of benefits barely exists. Such a situation does not however arise in other branches of social security, where the development is much more common and standard (such as in sickness protection). The situation for LTC is different. Thus, according to trESS, “Any cross-border movement may not only involve a quantitative loss of rights, but also a qualitative loss or rights or simply the removal or elimination of a possible right”.

While art.34 of Regulation (EC) No. 883/2004 intends to avoid the overlapping in LTC benefits, the lack of an EU-wide definition (and, relatedly, of a well-suited coordination regime) of LTC benefits, creates loopholes in the application of the article. Its application is impossible in many cases (e.g., cases in which countries do not consider their benefits in kind as LTC benefits at the European level; or cases in which the benefits are coordinated under different chapters – e.g. pension, family benefits).

However, social tourism in the case of LTC was generally not seen a risk by interviewees. 113  

Although most of the interviewed national administrations considered this risk to be rather low, concerns were raised which did not lie for the most on the possibility of double-payments for the same risks, but rather on the lack of operability of the system and the lack of visibility of Member States on the costs incurred into and the reimbursements claims received or entitled to issue 114 . The differences in the concept of LTC benefits and their treatment across Member States lead to a lack of operability of the reimbursement and mechanism of deduction for the avoidance of double payments set up by Article 34 of the Coordination Regulation. Countries lack information on the LTC benefits cross-border workers are entitled to in other countries, and therefore cannot always adjust the level of benefits provided accordingly. The exchange of portable documents does not always take place.

Sometimes, the recognition of the entitlement to LTC is not asked (no EU forms are exchanged), conveying the idea that some beneficiaries are content with the benefits they receive from the competent Member State or might not be aware that they should notify a country of residence about the benefits that they receive. In other occasions, the benefits are provided to cross-border citizens just as they are provided to nationals, without a possibility to distinguish them (and no reimbursement between Member States takes place). As illustration of this fact, a Danish interviewee stated:

“How many foreigners live in Denmark and are awarded LTC benefits by Denmark? We do not know that because our public authorities do not ask for it. We do not make the distinction between cross-border or mobile citizens and Danish people. If they have the right to live here, their Danish municipality will provide them the benefits”.

Frequent complaints about the reimbursement mechanism were received during the interviews conducted in different Member States. Given the frequent criticism, and given that the reimbursement mechanism is directly linked to the avoidance of double payments, it appears that this aspect of the coordination bears watching.

Additionally, practical problems concerning the identification of beneficiaries were sporadically raised. It seems that in certain cases the information sent by the competent Member State to the Member State of residence (or vice-versa) would not be detailed enough to identify the person in question. Social Security numbers or even national ID numbers would be provided to the other country, for which such information would not be sufficient; in other cases, the simple indication of the name and last name, without any additional details, would not be enough to identify the person either. However, this is a cross-cutting challenge for the entire EU coordination system and not strictly limited to the coordination of long-term care benefits.

In the lack of a comprehensive and coherent coordination regime well suited to the particularities of LTC benefits, it is the CJEU which constantly has to make clarifications. The system can thus be altered at any moment by a CJEU ruling.

Legal uncertainty was identified as an issue. The authorities interviewed at national level, dealing on a daily basis with the coordination of LTC benefits, confirmed the problems derived from the lack of a common definition and treatment of LTC across countries, and the poor fit that the coordination as sickness benefits provides on certain aspects. To a certain extent, the perception that the major issues have already been addressed by the case law of the CJEU was shared. However, the general view amongst those interviewed was that indeed the system is unstable and prone to constant modifications.

Similarly, the level of complexity of the regulatory framework and the rulings of the CJEU render the issue not easily accessible for all the concerned staff public authorities. The changes introduced by the CJEU require regular updates and continuous formation in the Member States’ administrations.

The complexity of the legal framework, including the lack of common definition and/or classification of long-term care benefits at EU level, and the large disparities across countries’ regimes results in an uneven application of the coordination rules. This results in a non-transparent, uncertain system, where citizens are not able to determine beforehand their entitlements in each country. Our findings confirm the lack of upfront visibility mobile citizens have on their entitlements. In the EC public consultation survey, the statement “Migrants are not sufficiently aware about their rights under EU law” was supported as one of the three major problems of the coordination mechanism by 57% of individual respondents, 40% of trade unions/social partners’ representatives and 26% of national administrations.

Interviews during field visits suggest that certain broad differences can be perceived amongst mobile workers. The more highly educated, tech-savvy amongst them would be making use of the Internet (including questions submitted to national administrations) to inform themselves about their entitlements in different countries and conditions. However, others would be making their mobility decisions rather blindly as it comes to entitlement to LTC benefits.

It seems, in particular, that mobile workers would have limited knowledge about the LTC benefits in kind to which they would be entitled in a different country (those are determined based on the legislation of the Member State of residence), and how that would affect any other LTC benefits they currently perceive or would perceive in the competent Member State (notably, reductions in cash benefits due to the enjoyment of LTC benefits in kind in another country).

This lack of knowledge apparently affects public authorities as well. Except for specifically dedicated bodies or units, the amount of public employees potentially dealing with a LTC case is very high while the frequency of the cases is low. While the cases can often be transferred to units with more specific knowledge, it appears that some obscurity would persist, even within the administrations, as to the exact content and entitlements of particular workers in cross-border situations. In this regard, findings from the public consultation survey reveal that 10% of the national administrations and 13% of the social partners’ representatives say that national administrations provide insufficient and/or incorrect information to mobile persons. One civil society representative argued that “Staff working in state bodies as well as migrant workers do not understand the rules. They need to be simple so that everyone knows what the entitlement is in each country. Also simple so that they are applied correctly and fairly.”

Similarly, the wide range of bodies and institutions dealing with the claims appears to generate certain issues of communication and to complicate the processing of claims. A Belgian health insurance fund representative argued: “You do not know to whom in the other Member State you can address your question. The responsible local public authority is not always known. The contact details of the Zorgkassen (Care funds), for example, are not adopted in the European database (EURES).”

Systems also vary significantly in each country depending on the amount of competencies devolved to the regions and local entities. The size, population and number of bordering countries influence the number of cases. The distribution of competencies within each country impacts decisively the way the claims are dealt with (including the average length of the process), ultimately affecting the beneficiaries. Generally, it was gathered that individual civil servants, numbering in thousands in many countries, are the entry point for the claims. Except for dedicated ones, handling LTC claims from cross-border workers is only one amongst a wide range of responsibilities and sometimes health insurance funds only have 1-2 cross-border cases per year. This makes it difficult to instruct a large workforce on the particularities of these cases. The complexity of the rules adds to this difficulty. Therefore, detailed knowledge of the rules not only affects citizens, but also public authorities. One Danish national administration representative argued:

“Several municipalities in Denmark – the competent level to provide most LTC benefits in kind to the citizen - did not even know that they could claim reimbursement from other Member States for the LTC benefits that they had provided to foreign residents in Denmark” (foreign citizens insured in another Member State and resident in Denmark).

The following problem tree, based on the information collected through the different data collection activities conducted for this study, summarizes in a visual way the main problems the current rules for the coordination of long-term care benefits present. The problem tree links the problems to policy objectives, in the upper part, and to the drivers behind such problems, in the lower part. Objectives are divided in general, specific and operational, following the methodology of the EC Impact Assessment Guidelines. The relationships between the individual objectives at each of the three levels are also depicted.

Figure 4: Problem tree - coordination of LTC benefits

5.3Coordination of unemployment benefits

Unemployment benefits fall within the material scope of the coordination Regulation. Chapter 6 of Regulation (EC) No 883/2004 is dedicated to “Unemployment benefits”. Therefore, in contrast with LTC benefits, there are specific rules for the coordination of unemployment benefits. The coordination rules have not changed substantially in comparison to Regulation (EEC) No 1408/71. There were limited amendments to the provisions on the calculation of unemployment benefits (in the light of the case-law of the Court of Justice) and to the provisions on export of unemployment benefits (a possibility to extend the export period from three for six months, upon a decision by the competent institution, was included). Notably, a limited reimbursement mechanism between Member States was also introduced. It covers the unemployment benefits provided by the country of residence to unemployed frontier or cross-border workers.

The Council already agreed in 2001 on Parameters for the modernisation of the coordination Regulations with regard to unemployment benefits. 115 The modernisation should result in a simplification of the coordination rules on unemployment benefits to enhance their readability and transparency.

However, apart from the adaptions previously outlined, the rules did remain similar as in former Regulation (EEC) No 1408/71. The trESS on the right to actively search for a work in the EU (possibility to extend the export period) and the changes applied to the calculation of benefits can be seen as elements of modernisation of the rules. Given the accrued internal mobility in comparison with the period of the previous regulation (approved in 1971), so does the introduction of the reimbursement provisions. However, no major simplification of the rules took place. The introduction of the reimbursement mechanism actually added some complexity to the coordination rules. These changes reflected a compromise achieved in the search for consensus (unanimity was then needed to change the coordination rules). In addition, a new provision (Article 65a) was introduced 116 to address a legislative gap in order to ensure the protection of rights of self-employed workers, in line with Article 48 TFEU.

The discussions in the Administrative Commission and in the Council 117 have demonstrated that this compromise is complex and is subject to different interpretations and misunderstandings. Against this backdrop, the Commission made a declaration for launching a broader revision of the unemployment chapter of the coordination regulation at the EPSCO Council in December 2011.

We describe in the following text the main problems elicited via the data collection activities conducted for this study. The problems are summarised visually in a problem tree, included at the end of the section. The problem tree links the problems to objectives, and also depicts the drivers behind such problems.

To further facilitate the understanding of the problems, and to substantiate their analysis, we first provide some additional background information and discuss one of the main drivers behind the identified problems, i.e., the complexity of the current coordination rules. Discussing this driver, which involves elaborating on different elements of the current rules, is relevant because it has a horizontal effect on the remaining drivers identified. The complexity of the rules is both caused by and reflected in the diverse aspects and cases contemplated by the current coordination rules. The support for a revision of the rules, as well as the sentiment or opinion of different stakeholders on particular issues is also provided by reflecting the results of the EC’s public consultation and Deloitte’s survey to national administrations.

Complexity of current coordination rules

The coordination rules were generally perceived as complex during the interviews conducted with public authorities and other stakeholders. The results of the online survey launched by Deloitte and the EC public consultation also point to such complexity and to the effect that, it becomes difficult for migrant workers and national institutions to know with certainty which are the workers’ rights under EU law.

In the online survey conducted by Deloitte among public authorities, 19.6% pointed out directly that “EU rules are too complicated” as one of the main problems noticed regarding the application of EU law in the area of unemployment benefits. Other responses also elicit such complexity. More precisely, 72% of the participating institutions thought that Member States are not correctly or uniformly applying EU law, whilst 60% considered that migrant workers are not sufficiently aware of their rights. These problems - which stem directly from a lack of simple, clear rules - were the most often selected by public institutions. In addition, 47% saw a problem in the ineffective and burdensome communication, while 39.2% saw administrative burden as a problem. Both of these problems could also be considered a certain by-product or consequence of complex coordination rules.

44.4% of the representatives from public institutions participating in Deloitte’s survey considered that the current coordination rules need to be changed. An additional 19.4% thought that, although they do not need to be changed, they should be better applied in practice (incorrect or uneven application can also stem from the complexity or lack of clarity of the rules). 12.5% of the respondents to the survey considered that, while the coordination rules need to be changed, they should be better explained. Among the 44.4% of representatives from public institutions that considered that the coordination rules need to be changed there were also concerns related to shortcomings of the coordination rules in defining precisely certain procedures or concepts, leading to a degree of obscurity that renders the coordination complex in practice.

Similar results were obtained in the EC public consultation. Three problems which can be linked to the complexity of the rules were the three most frequently selected by the respondents. 38% of the national administrations, 41% of the social partners and trade unions, and 56% of the civil society and non-governmental organisations considered that the lack of awareness of migrant workers about their rights (which results to a certain extent from complexity of the rules) is one of the three main problems of the application of EU law in the area. It was the problem noticed most frequently by each of these groups of stakeholders. The second problem encountered most often by the stakeholders consulted was the slow and/or ineffective communication between institutions of Member States; it was selected by 38% of national administrations, 24% of social partners and trade unions and 33% of civil society organisations. Another problem related to the complexity of the coordination rules was the third most-frequently noted problem. It referred to the fact that national institutions do not provide migrant workers with correct and sufficient information (29% of national administrations, 24% of social partners and trade unions and 33% of the civil society and non-governmental organisations participating selected this problem). More directly, for 13% of the national administrations and 12% of the social partners and trade unions, the excessive complexity of EU law was one of the three main problems of application of EU law in the area. For 23.6% of national administrations, the incorrect application of EU law by national institutions or administrations is also a top-three concern.

While they did not consider it necessary to change the coordination rules, 18% of the national administrations participating in the public consultation believed that the rules should be better explained, as did 29% of the social partners and trade unions. Another 20% of the national administrations thought that the rules should be better applied in practice. Among those who believed in the need to change the coordination rules (27% of national administrations, 35% of social partners and trade unions and 66% of civil society and non-governmental organisations) there were also concerns related to concrete aspects that render the coordination rules complex.

According to the stakeholders, different elements contribute to the complexity and lack of transparency of EU coordination rules. These include:

The distinction between frontier workers and other cross-border workers & the determination of cross-border workers’ country of residence

The first element of complexity comes from the distinction between migrant workers, frontier workers and other cross-border workers. The distinction is difficult to apply in practice, and becomes blurry in certain cases. During the interviews carried out it was generally agreed that, while there are marked differences between “classic” frontier workers and other types of cross-border workers, the latter are becoming less and less common and the distinction between frontier workers and other cross-border workers is increasingly blurry.

Typical frontier workers are considered to maintain stronger links with their country of residence, and are perceived as less likely to leave it due, inter alia, to family links. However, it was pointed out that the distinction between frontier workers and other cross-border workers is often rather complicated 118 . Even more, the distinction between cross-border workers and cases of foreign workers considered as national workers (i.e. migrant workers), whereby migrant workers become resident in the country of activity, also appears to be blurry. Relevant differences were also reported among frontier workers. As mentioned during the interviews conducted in Germany, frontier workers range from workers who live barely some minutes or kilometres away from their place of work to others who cover distances over 500 km. This challenges the original conception that frontier workers are close (in terms of distance and commuting time) to their place of employment (and hence unemployed frontier workers to their country of last activity).

It can also create conflicts with national law. Thus, for example, in order to be considered an unemployed person (and hence be entitled to unemployment benefits), the individual concerned must be close (in time and geographical terms) to his/her district or area 119 . Frontier workers covering greater distances to go to work do not fall into this category. On a similar note, according to the ZAV-IPS Frankfurt, not all national authorities inform the migrant workers sufficiently or correctly where to claim unemployment benefits, due in part to the fact that there is a lack of clarity regarding the “real” and “false” 120 cross-border workers.

In accordance with the views gathered, the determination of the country of residence remains difficult in practice, as it is subject to a high degree of interpretation. The criteria for the determination of residence set out in Article 11 of Regulation (EC) No 987/2009 are, according to the interviewees, broad and cannot be applied straightforwardly. This generates some uncertainty. Public authorities do not often have sufficient information at their disposal to control the frequency of return home of frontier workers, and the distinction between frontier workers and other cross-border workers fades out in many cases. Furthermore, economic development and infrastructural improvements reduce progressively the importance of physical distance. In these circumstances, the distinction between frontier workers and cross-border workers established by the Regulation adds an element of complexity that is on occasion not fully coherent with reality. As an example, the British authorities made explicit allusions in the interviews conducted to the case of Polish workers, whom in many cases are apparently not considered as resident in the United Kingdom even after living and working there for years 121 . Polish authorities, on their side, stated that the assessment criteria of the habitual residence test for cross-border workers does not work well and results in administrative burden. The Federal Ministry of Labour, Social Affairs and Consumer Protection of Austria (BMASK) considered that there should be a clear definition of a cross-border worker in the Regulation.

The distinction between frontier workers (and, within those, between wholly and partially unemployed frontier workers 122 ) and cross-border workers is important because it determines where the unemployed workers can register and claim unemployment benefits.

While wholly unemployed frontier workers must register and ask for unemployment benefits in their country of residence, cross-border workers other than frontier workers have the choice between their country of residence and the country of last activity. On one hand, this distinction might be confusing for the workers, who may not be aware of their own status. It might also be confusing for the public authorities. Problems regarding the lack of awareness of migrant workers about their rights, as well as problems concerning the fact that national institutions or administrations do not provide migrant workers with correct and sufficient information have been reported (see above). On the other hand, the right of choice given to cross-border workers other than frontier workers introduces an element of uncertainty for public authorities. It requires communication efforts with other Member States to have information on the situation of particular workers.

Rules on aggregation of periods of insurance, employment and self-employment

A second element of complexity is provided by the rules on the aggregation of periods. We hereby refer to the evidence provided on administrative burden in the baseline scenario chapter. The coordination Regulation provides for the aggregation of periods of insurance, employment or self-employment conducted under the legislation of different Member States for the purposes of the acquisition, retention, recovery or duration of unemployment benefits. A specific provision is introduced for the cases in which the right to benefits is conditional on the completion of periods of insurance. In these cases, the periods of employment or self-employment completed under the legislation of another Member State shall not be taken into account unless such periods would have been considered to be periods of insurance had they been completed in accordance with the applicable legislation.

Therefore, the coordination rules (Art.61 (1) of Regulation (EC) No 883/2004) start by drawing a distinction between cases where the national legislation of the competent Member State makes the entitlement and the length of the unemployment benefits subject to the completion of periods of insurance on one hand and, on the other hand, cases where the legislation makes the entitlement conditional on the completion of periods of employment. By so doing, the regulation intends to reflect actual differences in national systems.

The conditions to consider certain periods completed under the national legislation as periods of insurance, employment and self-employment under national legislation vary across countries. Since the acquisition, retention, recovery or duration of unemployment benefits are linked to the different periods of each type completed, and the consideration of the same periods might vary across countries, this makes it complicated for mobile workers to know their rights. This is also the case for public authorities (therefore, workers cannot always solve their doubts by asking to the administration). This diversity of opinions of the national administrations is well documented in the annexes to AC 470/11.

In particular, the interviews carried out with public officials pointed towards particular issues with the periods of self-employment. The regulation of self-employment varies significantly across countries. In some of them, contributions to Social Security, including for the coverage of the risk of unemployment, is entirely voluntary. In others, there are no schemes for self-employed workers at all. The way unemployment benefits are financed also varies across countries. While in certain countries the funds come exclusively from contributions of self-employed workers (e.g. Spain and Denmark), in other countries there is contribution from the State’s budget (e.g. Luxembourg). This in turns impacts the consideration of periods of self-employment as periods of insurance. Self-employment periods might not be considered as periods of insurance, and therefore not generate the right to unemployment benefits. Concerns about the particularities of self-employed workers were particularly raised in Spain. In the view of the Spanish public authorities, art.61 (1), second paragraph, fails to account sufficiently for the complexity brought about by the significant differences across countries. Representatives from French authorities also criticised the coordination rules in this aspect, suggesting that the coordination regulation should clearly set out from the beginning a difference between those countries that provide unemployment benefits to self-employed workers and those that do not (they could be specified in an Annex, covering potentially conflicting situations). The rules should then be different for each group of countries. In Belgium a representative from the National Employment Office pointed out the different interpretations that Member States make of periods of insurance. He signalled, as an example, the classification of self-employed periods treated as insurance period. Two specific cases for Belgium were provided as examples:

-OSSO-assurance (overseas social security office) 123 , 124 : periods of employment in another country insured by an OSSO-assurance were considered as ‘Belgian periods’. 125 This is no longer the case. Now, it is necessary to have/prove a Belgian period of insurance to receive an unemployment benefit. In addition, foreign periods outside the EEA, not employed as a posted worker, will not be mentioned on the U1-form.

-Belgium does not have unemployment insurance for self-employed persons. Self-employed persons living in Belgium and working in another Member State will not receive a Belgian unemployment benefit. However, these periods of self-employment will be taken into account when the person was afterwards employed as cross-border worker/frontier worker OR accomplished a Belgian period of insurance OR is a migrant person who has accomplished also a Belgian period of insurance.

The Confederation of the German Trade Unions also reported occasional problems with the aggregation of periods. These occurred in cases of short-term employment contracts and in cases when the former employer could no longer be traced.

Hence, while the coordination Regulation should cover all different contingencies, it follows from the opinions gathered in several Member States that the wide variety of cases that can occur is not, allegedly, sufficiently addressed by the Regulation. Self-employment’s peculiar characteristics and the disparity of its regulation across countries add to the overall complexity of EU rules. Mobile workers, and even officials in national public institutions, are put in a difficult situation to find out with certainty and a priori the entitlements of migrant workers. Public authorities in many of the countries visited (including Austria, Denmark, France, Germany, Poland and Spain) generally supported a clarification of the rules on the aggregation of periods of insurance, employment and self-employment (therefore acknowledging the problems stemming from the current ones).

trESS identified this same issue in their Think Tank Report 2012. As stated in the report, “A major problem in the field of the application of the aggregation rules is that the interpretations of these rules differ significantly among the Member States’ institutions. Especially in cases where the national legislation makes benefit entitlement conditional on the completion of periods of insurance, different approaches are taken with regard to periods of employment or self-employment, which under the legislation under which they were completed are not regarded as periods conferring entitlement to unemployment benefits”. The report outlines different problems arising from the application of the aggregation rules, including the periods of (self-) employment to be taken into account and the periods to be taken into account with regard to voluntary schemes. Other practical problems are also mentioned.

Provisions on the export of unemployment benefits: minimum period of export of three months with a possible extension to six months (discretionary decision of the competent institution)

A third element contributing to the lack of transparency is the regime for the exportation of unemployment benefits (cf. Table 131 : Export of unemployment benefits - current practices ). The regulation foresees a minimum period of export of three months, with a possible extension to six months. However, there are not fixed criteria for the concession or refusal of the extension. The decision is taken by the competent institution, which applies its own set of criteria or, in some cases, has no pre-defined criteria at all. The competent institutions in the Member States are supposed to apply discretionary power, meaning (according to the ECJ) an assessment of the situation of each individual case (and not the adoption of a general policy decision on the prolongation or not of the export). This reduces, in practice, the transparency for the workers, who cannot know in advance the likely duration of their export period, which might in turn affect their mobility decision. In addition, some countries do not grant the export at all (such as the UK).

Certain interviewees (including British public officials, a representative from the Confederation of the German Trade Unions and certain EURES advisors) considered that the duration of the period needs to be clearly defined (be it three months, six months or other duration –they held different views-). A time-span (as is the case, with a span between three and six months) can easily lead, in their view, to a misinterpretation by the citizens, and is ineffective in practice.

The interplay between the current coordination rules and national rules can also create conflicts in practice. This was the case in Germany, where it was reported that it is not possible for recipients of Hartz IV benefits (Arbeitslosengeld 2) to export their unemployment benefits; it is only possible to export the “true” unemployment benefits (Arbeitslosengeld 1). This is, allegedly, often unclear to the citizens who seek advice (e.g. at EURES).

Reimbursement of unemployment benefits between Member States

Even though they do not impact EU citizens directly, the rules on reimbursement between Member States also add to overall complexity of the system and represent a source of administrative burden for Member States. Several claims were raised during the interviews with public officials pertaining the inefficiency and burdensome character of the reimbursement mechanism (they are described more in detail below).

Depending on the conditions, the Member State of last activity might have to reimburse the Member State of residence the benefits provided during the first three or five months to a frontier or cross-border worker.

The reimbursement provisions are a novelty of Regulation (EC) No 883/2004, as they were not included in Regulation (EEC) No 1408/71.

Communication between institutions of Member States

Communication between institutions of Member States is clearly perceived as an area with substantial margin for improvement. Under the current coordination rules, problems of ineffectiveness and derived burden are reported by a significant number of public authorities. They are also noticed by other stakeholders.

As previously reported, the results of both Deloitte’s online survey and the public consultation carried out by the Commission reveal shortcomings in the communication between institutions of Member States. In Deloitte’s survey, 47% of the public authorities saw a problem in the ineffective and burdensome communication. Only 10.6% of the respondents considered that the communication is fast, efficient and effective; over twice as many respondents, in contrast, indicated that cross-border communication is slow, burdensome and/or ineffective.

In the EC public consultation, 38% of national administrations, 24% of social partners and trade unions and 33% of civil society and non-governmental organisations ranked the slow and/or ineffective communication between Member States as one of the three main problems of application of EU law in this area. It was the second most-frequently selected by the respondents who replied on behalf of organisations.

The complexity of the current system is reflected in the use of different types of documents for exchanging similar information. As an example, according to representatives from the Polish government:

“There are too many documents that are used to confirm basically the same information (e.g. U002/U017/E301). Other countries use other forms such as PD U, PD S, PD F, etc. This may result in confusion among migrant workers and may lead to the loss or reduction of their entitlement to unemployment benefits. The documents need to be clear and be used universally across the EU.”

As has been discussed above, the complexity of the rules generates problems both for mobile workers and for public authorities to know with certainty the rights mobile workers are entitled to under EU law. The results of both Deloitte’s survey and the EC public consultation, as well as the interviews conducted at national level point towards a relatively low level of awareness or rights. While workers would actively look for information upon becoming unemployed, the complexity of the rules and the uncertainty that surrounds the interpretation and application of certain ones would persist as obstacles for a high level of awareness of rights.

The coordination rules try to ensure that migrant workers receive unemployment benefits in the conditions most favourable to the search for new employment. Under the current coordination rules, cross-border workers other than frontier workers can choose to register and apply for unemployment benefits in their country of residence or in their country of last activity. The onus is on them, then, to ponder the pros and cons of both systems and opt for the one they consider more beneficial given their circumstances. Wholly unemployed frontier workers, meanwhile, must register and claim unemployment benefits in their country of residence. As a supplementary step, they might also register as job seekers in the employment service of their country of last activity. However, if they do so, the compliance with the job-seeking obligations of the country of residence takes priority over the requirements of the country of last activity. As such, the rule mandating frontier workers to claim unemployment benefits in their country of residence is based on the assumption that they enjoy the most favourable conditions for seeking employment in their country of residence. Due to changing job market conditions, such assumption is nowadays questionable.

In interviews carried out for this project, particularly with public officials, as well as in our online survey, several interviewees and respondents manifested the view that unemployed frontier workers might not enjoy the most favourable conditions for seeking employment in their country of residence. Instead, they could have better chances of finding a job in their country of last activity. The qualification conditions that apply, the value attached by employers to experience in a particular market, and the importance of professional networks are some of the factors that might turn the country of last activity into a more attractive option for frontier workers. In this sense, a British public official argued that “as research shows, many of the opportunities to find a job are due to formal networks, not informal ones”. He argued that formal networks would be stronger for the frontier worker in his/her country of last employment, while informal networks would be stronger in his/her country of residence.

However, it was also noted during the interviews that it is not possible to generalise. Each case is different. The most relevant factors to determine the chances of finding a job are the educational background, qualifications and experience of the job seeker as well and the conditions of the job market in the area concerned. Demand for specific skills and the economic conditions of a specific area are paramount (as evidenced particularly in times of economic crisis, were unemployment rates vary strikingly across countries and regions). Therefore, it is difficult to determine, on a general basis, whether frontier workers enjoy a better chance of finding a job in their country of residence or in their country of last activity. At any rate, precisely because of the importance of individual circumstances and the swings in the economic prospects of each region over time, the assumption that unemployed frontier workers enjoy better conditions for seeking employment in their country of residence does not seem to hold.

The Court of Justice did indeed identify an exceptional application of the rule for other cross-border workers to unemployed frontier workers, who have better prospects of finding another job in the State of last activity and could therefore apply for unemployment benefits there (Case C-1/85 Miethe).

While the Court has recently stated that the provisions of the new Regulation (the Miethe case concerned the previous coordination regulation, dating from 1971) are not to be interpreted in the light of its earlier case-law 126 , it does not deny the existence of a certain type of frontier workers who, having maintained particularly close personal and business links with the country where they were last employed, have better chances of reintegrating into working life in that Member State.

Therefore, while not denying frontier workers the chance of receiving support from a public employment service to find a job, it appears that the current coordination rules do not necessarily reflect the current job market conditions for unemployed frontier workers.

It should be noted, however, that some opinions in favour of making the country of residence responsible for the provision of unemployment benefits to frontier workers were also voiced. They were partially based on the belief that the country of residence offers the best chance for frontier workers to find a job. This was the case in Germany. According to the German Employment Services, unemployed workers receive an optimal support to find employment when they are supported by the employment services in the country of residence; in such cases, there are no disadvantages for the jobseeker resulting from the participation in reintegration services, such as travel costs, uncertainty about training courses, advanced vocational training, etc. In addition, the German Federal Ministry of Labour and Social Affairs argued that when receiving unemployment benefits from his/her country of residence, the unemployed person receives benefits which are in line with the person’s direct environment (e.g. cost of living and wage level) and is offered equal treatment compared to “national” unemployed persons. This system also guarantees, in their opinion, an undivided competence that ensures a proper functioning of the chain of reintegration –granting of services and control.

The previously discussed (increasingly) blurry distinction between certain frontier workers and some other cross-border workers, as well as the differences between different types of frontier workers (some of them maintain particularly strong links with the country of last activity, which creates effective differences between them and other frontier workers who remain more attached to their country of residence), also reflect changes in the job market not fully accounted for by the current coordination rules.

In addition, public authorities interviewed in Slovakia pointed out that the current coordination regulation is focused on the problems of the “old” EU Member States. Regulation (EC) No 883/2004 was enacted on 29th April 2004, two days before ten new Member States officially joined the European Union (1st May 2004). These interviewees argued that the “old” Member States generally had (and have) higher living standards than the new entrants, while their social security systems also vary. These differences affect the actual results of the implementation of the coordination rules. One interviewee, for instance, pointed out that “the motivation for a Slovak worker to go abroad largely differs from the one from old EU Member States”.

Member State of residence bearing costs of unemployment benefits and other social security benefits without receiving contributions and with only limited coverage by reimbursement mechanism

The impact of the current coordination rules on the fair sharing of financial burden among Member States is different with regard to frontier workers and to other types of cross-border workers.

The Member State of residence must bear the costs of the unemployment benefits of frontier workers, even though it has not received any contribution from them. Furthermore, it also has to bear other social protection costs related to the unemployed worker and his family.

The unfair character of the rules for the Member State or residence, in terms of distribution of costs between Member States (considering the link between the reception of contributions from workers and the expenditure in unemployment benefits), was agreed by the vast majority of the public authorities interviewed.

In general terms, it was agreed that to provide for a fair sharing of burden, the country receiving contributions should be in charge of paying out the unemployment benefits. In other words, there should be a link between contributions and payments.

Cross-border workers other than frontier workers can choose to claim their unemployment benefits in their country of residence or in their country of last activity. For those who choose the country of residence, the same reasoning just explained for wholly unemployed frontier workers applies. For those who choose their country of last activity, the Member State that received contributions from the worker bears the costs of the unemployment benefits.

In the online survey conducted among public authorities, these considered clearly the best option, from the point of view of a fair sharing of the financial burden between Member States, that the country where the person last worked and paid social security contributions (even if he/she lived in another Member State) should provide the unemployment benefits (it received a score of 3.19 in a scale of 1 to 4). By comparison, the options stating that the benefits should be provided by the Member State in which the person last lived (even if he/she last worked and paid social security contributions in another Member State), and that the unemployed person should be allowed to choose to claim the benefit either in the Member State of last employment or in the Member State of residence, were both similarly scored (2.29 and 2.28, respectively) and clearly below the option making responsible the country of last activity (see Annex  10.9 ). 

As mentioned previously, a reimbursement mechanism exists. According to it, the Member State of last activity must reimburse the benefits provided to wholly unemployed frontier workers during the first three or five months (depending on the periods of employment or self-employment completed by the worker) to the country of residence. As long as the unemployment spell of the concerned worker lasts for longer than the period of reimbursement, the current coordination rules impose an unfair burden on the Member State of residence. In this situation, the Member State of residence has not received contributions from the unemployed worker, yet it has to cover his/her unemployment benefits and possibly also bear other social protection costs. Numerous public officials with direct experienced in the matter interviewed considered that the coverage of the reimbursement is not sufficient to compensate the expenditure in unemployment benefits (let alone other social security expenses) given the average duration of unemployment spells. Among those, the following quote was received from Portuguese civil servant:

“The reimbursement of 3-5 months by the competent Member State is considered as peanuts compared to the costs that we face in paying out unemployment benefits to Portuguese seasonal workers who have worked abroad. That is why we are so much in favour of a system where they would claim their unemployment benefit in the country of last employment” (Portugal has a significant amount of seasonal workers who work abroad during a certain period of time while retaining their habitual residence in Portugal. When they come back and become unemployed, Portugal has to pay the unemployment benefit of these people).

However, the duration of the reimbursement period was considered sufficient to provide for a fair distribution of the costs by the public authorities interviewed in some Member States, including notably Luxembourg and the United Kingdom. Public officials in the United Kingdom argued that around 50% of unemployed workers there found a job within 13 weeks of becoming unemployed, and the remaining ones do it before six months. Therefore, the coverage of the reimbursement mechanism seemed appropriate to them.

Other problems of unfair sharing of financial burden

Other aspects of the current coordination also stood out as factors causing an unfair distribution of financial burden. These can generate an uneven distribution of the burden in one direction or the other, that is, unfair for the country of last activity of unfair for the country of residence. These include:

-The lack of a minimum period of employment required for the determination of the competent Member State: the current coordination rules do not fix any minimum amount of time of employment to determine the Member State of last activity. Thus, periods of employment of barely a few days or two or three weeks are enough to consider a country as the country of last activity; as such, the country in question will be responsible for the provision or reimbursement of unemployment benefits. Cases of this kind were reported during the interviews (inter alia, by public authorities in Austria, Belgium, France, Germany, Luxembourg and Spain). This can generate an unfair burden on the concerned countries (for instance, a country might have to reimburse up to three months of unemployment benefits for a worker that has only worked in the country for a few days).

-The exclusiveness of the last salary for the determination of the benefits the unemployed worker is entitled to: currently the last salary of the migrant workers is used to calculate the benefits he/she is entitled to (following art.62 Regulation (EC) No 883/2004). It is possible that such last salary has been perceived only for a very short period of time (see previous point). Salary conditions might also vary abruptly between the country of residence and the country of last activity (and hence possibly between the unemployed person’s two latest remunerated activities). In these cases, the last salary might not be a good indicator to use as a basis for the calculation of the benefits, in the sense that it might lead to benefits disproportionate to the contributions paid by the worker over a representative period of time. The worker might end up in a winning or losing situation, and situations of unfair sharing of the burden (via reimbursements) might arise. It was alleged by numerous public officials that previous sources of income, over a more extended period of time, should be considered for the calculation of unemployment benefits.

Examples of the first of these problems gathered during the interview include:

“A lot of people exporting their unemployment benefit quickly find an interim job in the Netherlands. Then they become unemployed after a while which makes them entitled to Dutch unemployment benefits, if they live in the Netherlands”

“A Polish person with a U2 document moves to Belgium. If he finds an interim job for a month in Belgium, he can be entitled to a Belgian unemployment benefit if his periods of employment are aggregated. This is social tourism.”

As evidenced by the second example, concerns about the lack of a minimum period required to determine the country of last activity (and hence potentially competent for providing unemployment benefits) are linked to concerns about a certain misuse or abuse of rights.

TrESS discussed the aforementioned problems and expanded briefly on them when dealing with the problems brought about by the adoption of the system of reimbursement of between Member States. These problems relate directly to the distribution of the financial burden between Member States. The following problems were listed in the Think Tank Report 2012:

-The State of last employment or last activity may be obliged to reimburse some amounts although the person concerned would not have been entitled to benefits under its legislation. The application of a foreign legislation is binding, which may in principle collide with the apparent neutrality of coordination provisions.

-The State of residence ameliorates its situation with respect to the provisions of Article 71 of Regulation 1408/71, but in many cases the reimbursements do not fully cover (three to five months) the cost of the unemployment or other benefits awarded.

-The amounts reimbursed may not be proportional to the periods completed in the Member State of last employment or last activity. No minimal period of insurance is required for the start of the obligation; an insurance period of only one day may suffice.

-The State of residence is obliged to be in a position of creditor and needs to require the correspondent reimbursement from the debtor.

-The maximum amount provided in paragraph 6 of Article 65 can be a theoretical amount, taking into account that the real amount payable by the State of last employment or last activity could amount to zero. Member States frequently disagree on this maximum amount.

-The administrative procedure is very complicated and burdensome. Delays of reimbursement are common and frequent. For the State of residence, there always is the uncertainty whether and when it will receive the reimbursement. This objection weighs most heavily against the reimbursement approach.

In addition, trESS reported persistent problems regarding the joint interpretation of the reimbursements, as some Member States declared they may not follow Decision U4 of the Administrative Commission (the decision constituted nevertheless a positive step towards a joint interpretation of the reimbursement rules). Practice also shows, according to trESS assessment, that there is uncertainty as to how to handle claims for reimbursement, and institutions seem in general to lack the experience to properly assess reimbursement claims.

As additional information beyond the pure problems identified in this regard, the following table summarises the opinions of the participants in the EC public consultation and Deloitte’s online survey regarding the question of which Member State should provide the unemployment benefits (which is intimately linked to the sharing of the financial burden between Member States):

Wholly unemployed frontier workers must currently claim unemployment benefits in their Member State of residence. However, these workers paid contributions to Social Security, covering the contingency of unemployment (or contributed otherwise, such as via taxes paid to the Treasury, to the financing of such benefits), in their country of activity. Therefore, frontier workers receive unemployment benefits based on a different system than that to which they contributed. This leads to situations in which certain wholly unemployed frontier workers may lose out on benefits, and receive unemployment benefits below what their contribution should have enabled them to. In other words, they do not receive benefits proportional to their contribution. 127  

This existence of this problem was manifest, according to the public authorities interviewed. Some cases of glaring disproportion were reported during the country visits. 128

As trESS pointed out in the Think Tank Report 2012, problems might also arise when there is a switch in the applicable legislation. Once a person is qualified as a wholly unemployed frontier worker or a non-frontier worker who continues to habitually reside in the State of residence, the State of residence will be, in virtue of Art.65, the competent State for the provision of social security benefits. The State of residence becomes then responsible for the payment not only of unemployment benefits, but also of sickness and family benefits. This change of system might give rise to disputes, especially when applying the legislation of the State of last employment would have resulted in a more favourable outcome for the person concerned and/or his family members.

Cross-border workers other than frontier workers, on their side, have the right to choose to claim unemployment benefits either in their country of last activity or in their country of residence. Hence, barring any major problems in terms of information about their rights in each country, they are able to compare the total benefits they are entitled to in each country and therefore avoid problems of lack of return on their contributions. However, as explained earlier when commenting on the complexity of the current coordination rules, there are widely perceived issues among EU mobile workers regarding a lack of awareness of their benefits. This was noted by public administrations as one of the main problems related to the application of EU law in the only survey conducted by Deloitte, as well as in the EC public consultation. Other stakeholders also declared to have noticed this problem, while the majority of individual respondents (general public) declared not to know (27.4%) or only have a vague idea (29%) about their rights under EU law. While the interviews carried out with public officials directly involved in the matter during country visits (clearly in France and Romania) confirmed that, upon becoming unemployed, mobile workers become much more aware about their rights, it appears that they would still face challenges to know with the certainty their entitlements. Thus, Deloitte’s survey and the EC public consultation showed that there appear to be problems concerning the level of awareness of the coordination rules among public authorities, causing them on occasion to provide migrant workers with incorrect and/or insufficient information. In these cases, cross-border workers other than frontier workers could also end up missing a return on their contributions if failing to identify the Member State that would provide them with the higher unemployment benefits and/or the highest chances of reintegration in the labour market.

Using the “impact on social protection” as a proxy for return on contributions, the public authorities participating in Deloitte’s survey ranked the right of choice as the best option (a score of 2.79 in a scale of 1 to 4), followed by the “country of last activity” (2.67). The “country of residence” option (currently applied to wholly unemployed frontier workers) received a lesser score than both of them (2.53). These results are consistent with the arguments presented.

Additionally, issues with the aggregation of periods could also lead to the loss of entitlements. In these cases, workers would also receive unemployment benefits which are not commensurate with the contributions paid. No glaring issues of this kind were identified in the interviews carried out with public administrations.

The aforementioned informational challenges for EU citizens and administrations can result in a loss of benefits for mobile workers. The complexity of the coordination rules creates challenges for workers to know their rights, which can result in migrant workers being unprotected (being left without any unemployment benefits) because they requested unemployment benefits in the wrong Member State. As staff working in national administrations also face challenges in obtaining an in-depth knowledge about the migrant workers’ rights, they are not always able to advise workers so as to avoid situations of losses of benefits.

TrESS points directly at the complexity of the legal framework as a potential source of problems, and in particular to the fact that the applicable legislation for the granting of unemployment benefits varies depending on whether the person concerned is wholly or partially unemployed. Regulation 883/2004 does not define these concepts, and while decision U3 of the Administrative Commission provides criteria to determine whether a person is to be regarded as a partially or wholly unemployed person, the interpretation of the criteria can still create problems. The unemployed person might see his/her application rejected because of the different interpretation of the facts and criteria that apply.

Changes in the applicable legislation, discussed previously, can also lead to a comparative loss of benefits.

However, this loss of rights was not considered a major issue by the stakeholders interviewed. While the relative lack of knowledge about their rights was confirmed during the interviews, and the challenges for the public authorities themselves were echoed, the general perception of the public authorities was that the system was not failing to protect the workers. No cases were a total loss of rights had occurred due to informational barriers and lack of transparency of the system, were reported by any interviewed representatives of national administrations.

9.8% of the public institutions participating in our online survey considered as one of the three main problems of application of EU law in the area is that “EU rules do not provide for the effective protection of social rights of migrant workers”. These included the following claim from an Austrian civil servant: “Many migrant workers are unaware which institution could support them. As the regulation is quite unspecific (for example, how long do you need to work in a foreign country to be entitled to unemployment benefits), each country adds its own rules which makes it difficult for jobseekers. If a cross-border worker loses his job and moves from country A to country B: Country A is not responsible because he does not live there anymore, and country B is not because he did not work there before... A German public official, meanwhile, stated that “Not all national authorities inform the migrant workers sufficiently … Migrant workers from abroad ( Eastern European Countries) are not sufficiently informed about their rights - social insurances, employees entitlements, contracts etc. A lot of guest workers are not well informed about minimum wages or collective labour agreements and usually earn less than a domestic worker.”

Two representatives of national administrations (out of 55) and one representative from the social partners (out of 17) participating in the EC public consultation considered that “EU rules do not provide for an effective protection of social rights of migrant workers” as an important problem.

The divergent Member States’ interpretations of Article 61 of Regulation 883/2004 also can lead occasion to the loss of benefits by unemployed workers. A survey launched in the Administrative Commission showed that divergent interpretations of art.61 lead some Member States to an application of the rules that does not conform to EU law. These issues, confirmed in the European Report 2011 of the trESS network, can have a negative effect on the rights of migrant worker –who may not qualify for unemployment benefits as a result-.

Other issues in terms of incorrect application of the coordination rules can also lead to loss of rights to benefits. 29% of national administrations participating in the Deloitte online survey and 22% of those participating in the EC public consultation considered that, while they do not need to be changed, the coordination rules should be better applied in practice.

The following issues were reported in Poland: According to the Ministry of Labour and Social Policy, certain institutions of Member States of last activity refuse to grant unemployment benefits (and consequently to export those benefits) to cross-border workers other than frontier workers only because they are not residents of that Member State (this in breach of Art. 65(5) (b) in connection with Art. 65(2) third sentence of Regulation 883/2004). Certain institutions do not fulfil their obligations as stipulated in Art. 12(1) in connection with Art. 54(1) of Regulation 987/2009 and make the persons concerned acquire their insurance records (on PD U1) from other Member States on their own. In addition, certain institutions do not fulfil their obligations as stipulated by Art. 54(3) of Regulation (EC) No. 987/209 and make the persons concerned acquire certificated of their family situation (on the outdated forms E302) form Member States on their own. This applies especially to the UK.

The shortcomings of communication between the institutions of Member States can also cause a certain loss of rights to workers by unnecessarily depriving them temporarily of their rights. Only 10.6% of public authorities participating in Deloitte’s online survey considered that cross-border communication with other Member States is fast efficient, and effective. In contrast, in the opinion of 25.5% of the respondents the cross-border communication is slow, burdensome and/or ineffective. For the remaining 63.8%, it generally works well, but there are specific problems with certain Member States. During the country interviews, such bilateral issues between different countries were confirmed. The slow or ineffective communication can affect the unemployed people insofar as it can lead to delays in the processing of documents. Since the payment of unemployment benefits depends upon the completion of the administrative procedures, unemployed workers are directly affected by issues in cross-border communication between institutions. These issues were mentioned by several Member States representatives (for instance in Denmark, the Netherlands and France).

In Belgium, it was reported that the issue of the U1 document by the competent Member State can take six to eight weeks; in the meantime, the unemployment benefit cannot be paid to the unemployed person. In addition, it was stated that not all the necessary documents are provided by the institution or the unemployed person when they are requested (e.g., C4 form, pay slip) and that there are enormous differences across countries in the number of asked “verification documents” (e.g., contracts, pay slips, information about the unemployed person or information about the last employer) when the Belgian unemployment office request a U1 document. Certain countries (notably France) would request a large amount of documents.

Long delays in the exchange of documents were also reported in France.

In Romania, problems were noted in certain cases, notably concerning the missing feedback from requests made to other Member States. Due to this, it was reported that a case can be solved in a timespan ranging from one month and a half to an undetermined duration. In addition, according to the public officials’ interviews, approximately 30% of the forms requested to other Member States were not received or were received with delays, after following up on the initial request.

Another claim related to the loss of rights to benefits was made in the Netherlands: “The current rules are pretty difficult to apply if a frontier worker is made redundant by a company within the framework of a collective redundancy. In this case, he could benefit from social plan arrangements concluded between the company and trade unions. Many social plans provide topping-up mechanisms where the company supplements the unemployment benefit of his former employee. Supplements are calculated on the basis of the unemployment benefit system of the competent Member State. For frontier workers, who have to claim their unemployment benefit in the country of residence, this kind of topping-up mechanisms could generate unforeseen situations (in their benefit or in their disadvantage)”.

Member States’ practices regarding the extension of the period of export of unemployment benefits (from three to six months) vary significantly. As reported in the trESS network Think Tank Report 2012, on the Coordination of Unemployment benefits, some countries have developed their own criteria to assess the convenience of granting the extension, while others assess each individual case on the basis of the reasons mentioned in the request. In some Member States, granting the extension is exceptional; while others do not apply the extension at all (such is the case in Sweden, the Netherlands and the United Kingdom) (cf. p.121 and Table 39).

In those cases when the extension is never granted, or only granted on an exceptional basis, the practices of the competent Member States equates to a lack of facilitation of job-seeking abroad (and hence of labour mobility across the EU).

No set of objective, homogeneous criteria to decide on the granting of the extension across Member States exist. This creates comparative disadvantages between workers in different countries. Those who see their petition of extension directly rejected, without an assessment of the circumstances of their case, are denied of a chance to facilitate their job seeking abroad.

The workers who are denied the extension of the export do maintain the right to receive unemployment benefits for the remainder of their period of entitlement upon condition of their return to the competent Member State. While maintaining the protection of the workers, this presents an effective obstacle to job-seeking abroad.

According to their replies to the Deloitte online survey, public authorities correlated the duration of the period of export with the impact on labour mobility (the current rules received a score of 2.47 in a scale of 1 to 4, while the option of raising the minimum period of export to six months received a score of 2.54 and the option of making the export possible until the end of the person’s entitlement received a 2.72 score). In other words, the longer the possibility to export unemployment benefits, the longer the impact on labour mobility. Therefore, if the duration of the export of benefits is limited due to the different national practices, labour mobility is curtailed.

trESS also reported additional problems of a more practical nature that can hamper job-seeking abroad. These include delays due to the difficulty of finding the required information in time, and also the insistence of many institutions on receiving a confirmation of the jobseeker’s registration with the employment services in another MS. This is relevant because the benefits will only be paid after this confirmation. Jobseekers may also encounter difficulties to register with the employment services in another MS, as experience shows. There are also differences with regard to the monthly follow-up information, and there is no uniform approach concerning the use of the PD U2 and the SED S009. Some institutions still use E- forms without any reference to the new Regulations.

As additional information, the opinions of stakeholders participating in the EC public consultation and Deloitte’s online survey about the rules on the export of unemployment benefits are summarized as follows:

Public authorities pointed out problems of application of the current coordination rules consistently via the different data gathering activities conducted for the study. 72.5% of the participating institutions selected in the Deloitte online survey “EU law is not uniformly understood and applied by Member States” as one of the three main problems regarding the coordination of unemployment benefits, making it the most-frequently noted problem. On top of the 44.4% of public authorities considering, for various reasons, that the coordination rules need to be changed, 19.4% of the respondents thought that, while the rules do not need to be changed, they should be better applied in practice. Thus, per countries, whilst considering that the coordination rules do not need to be changed, representatives of public authorities from the Czech Republic, Denmark, Estonia, Malta, Portugal and Spain pointed out in the online survey that the rules should be better applied in practice. Meanwhile, authorities from Austria, Latvia, Lithuania, Slovakia and Switzerland stated that, whilst they do not need to be changed, the rules should be better explained.

Results were similar in the EC public consultation. “National institutions/administrations do not apply EU law correctly” was the fourth most-frequently selected problem. 19% of the total amount of respondents from national administrations participating considered this as an important problem (in contrast, it was only selected by one social partner and one civil society organisation). On top of the 27% of respondents from national administrations (and 35% of social partners and trade unions, as well as 66% of civil society and non-governmental organisations) that considered that the rules should be changed, 20% of respondents from national administrations considered that, although they did not need to be changed, the coordination rules should be better applied in practice (this option was again only selected by one social partner and one civil society organisation).

Considering specific rules where an uneven application has been noticed, the rules of the aggregation of periods (art.61 Regulation (EC) No 883/2004) are a clear example.

A survey launched in the Administrative Commission in 2011 showed the widely divergent interpretations of art.61 by Member States, some of which were not in accordance with EU law. These issues were confirmed in the European Report 2011 of the trESS network, and were also confirmed during the interviews conducted with involved public officials for this study.

Two aspects are particularly problematic. First, the consideration of periods of employment or self-employment, when the national legislation of the competent Member State makes the entitlement to benefits conditional on the completion of periods of insurance, and the periods of employment or self-employment are not regarded as periods conferring entitlement to unemployment benefits, is controversial. Second, the consideration of periods with regards to voluntary schemes.

Concerning the first problem, some institutions hold the view that they should not be obliged, by virtue of art.61, to take into account periods which would not qualify to generate entitlement to benefits under the legislation of the State under which these periods were completed (but would have been considered as a period of insurance under the national legislation of the competent Member State). Concerning the second problem, a similar problem arises. Only the national legislation of the competent Member State is considered (therefore, if the period of employment or self-employment completed abroad would have been considered as a period of insurance if it had been completed in the competent country, then it will be considered as such and will qualify for benefit entitlement); this even in cases where the unemployed worker did not insure himself voluntarily. In these cases, as with the first problem, some Member States consider that such period should not generate the right to unemployment benefits (and hence should not be aggregated) since it would have not done so in the country where it was completed.

An example of this type was brought up in Denmark:” If Germans work in Denmark and if they chose not to become member of the unemployment fund, Germany needs to take into account the periods these people who have worked in Denmark but these people never contributed to the system. It should be clear that when people move to other MS, they should not have rights for which they have not contributed anything. Also Czech republic complained about it”.

Another similar example was gathered in the Netherlands: “Aggregation of period of insurance/employment problems with Germany and United Kingdom. Germany did not want to accept the period of employment of people in the Netherlands, as under German legislation this would not have led to entitlement to unemployment benefit. However, the problems are solved now with Germany. With the UK, there problem still persists”.

The uneven application of art.64 concerning the extension of the period of export of unemployment benefits has also been documented, as discussed above.

Different national practices concerning cross-border communication between institutions of Member States were discussed during the country interviews with the public authorities. This concerns in the first place the use of different portable documents. As clearly pointed out by the Polish authorities, this includes the use of different documents that are used to confirm essentially the same information (e.g. U002/U017/E301). Member States use different documents (e.g., E- and U- documents for coordination unemployment benefits), paper-based or in electronic form, and exchange them differently (by postal mail or electronically). Secondly, it concerns the practices regarding the requests for specific information to fill out portable documents. During interviews, representatives in most Member States commented on how some countries require much more information than others do (for example France), which effectively delays procedures. Thirdly, as brought up mainly by public officials from Spain and the United Kingdom, there are different practices with regard to the claims for reimbursement. According to them, certain countries were much more prone to ask for the reimbursement of unemployment benefits that others, particularly with regard to relatively unclear cases. In this sense, certain public officials felt a certain lack of “institutional loyalty”, whereby treatment and relations with other countries would not be reciprocal.

TrESS identified similar “practical problems” concerning the application of Article 61. As presented in the Think Tank Report 2012, sometimes wrong SEDs are used. In other occasions, the SEDs contain incorrect information or do not contain information on the nature of the employment, or the recipient cannot be identified. Even if the correct forms are used, the different approaches of Member States can lead to confusion as to the periods to be filled in on the PD U1 and the SEDs. Thus, while some institutions fill in all periods, others only include periods that qualify for benefit entitlement based on the argument that, otherwise, periods that cannot lead to an unemployment benefit in the State where the periods were complete have to be taken into account by the competent State which is to pay for the benefit. Moreover, as reported by trESS, some institutions refuse to deliver the PD U1, while in practice institutions must often wait for months for the PD U1 or the SEDs.

The following problem tree presents in a visual way the main problems presented by the current rules on the coordination of unemployment benefits. The tree is based on the information gathered via the different data collection tools and activities employed for this study. The problem tree relates the problems to their underlying drivers (which are depicted below the problems), and to the policy objectives, that are defined based on such problems (the objectives are depicted above them). The relationships between drivers and problems, and between problems are objectives, are pictures using arrows. Following the EC Impact Assessment methodology, a difference is made between general, specific and operational objectives. The relationships between the objectives at each of three levels are also presented visually by using arrows.

6Description of the policy objectives

A central question in this section is: what should the EU rules for coordination of unemployment and long-term care benefits aim at? The objectives that are set should reflect the identified problems, and should be structured according to the chain of problems (drivers). An objective tree was developed in which a hierarchy of objectives (operational, specific, and general) was defined according to a causal chain effect:

General objectives: These refer generally to treaty and/or EU 2020 Strategy-based goals (and are therefore a link with the existing policy-setting).

Specific objectives: Relate to the specific domain and the nature of the intervention considered. The specific objectives correspond to result indicators, and defining these is crucial, as they set out what the Commission wants to achieve with the intervention.

Operational objectives: Concern deliverables or objectives of actions that the initiative is expected to achieve and are linked with the outputs of the intervention.

The objectives are to be coherent with the overall EU policy in the area, and should be SMART (Specific, Measurable, Achievable, Realistic and Time-Dependent) to the possible extent.

For both the coordination of long-term care benefits and unemployment benefits, the general objective is to ensure the protection of mobile citizens’ social security rights and to facilitate free movement. By ensuring the rights of mobile citizens within the EU (plus Iceland, Liechtenstein, Norway and Switzerland), obstacles to free movement are removed. Actual mobility is not an objective per se, but creating an enabling and conducive environment for mobility is. All obstacles to free movement must be removed so that it is ensured. This general objective is in line with Articles 21, 45 and 48 of the TFEU, which constitute the basis for the coordination of social security schemes.

This general objective and the specific objectives set out are consistent with the horizontal policies of the European Union, including the Europe 2020 strategy, whose flagship initiative “An agenda for new skills and jobs” seeks inter alia to respond to the need to facilitate and promote intra-EU labour mobility. Removing the obstacles to mobility for European citizens is on the priorities of the European Commission, but it has been highlighted by it and other European institutions over the last few years that European citizens still face problems when moving across borders in the EU (see, among others, the European Year of Workers Mobility 2006, the European Parliament report on citizenship in 2009, the Monti report on the Single Market, President Barroso’s political guidelines and the Commission Work Programme 2013). Labour mobility across Member States remains low, as emphasized in the 2012 Annual Growth Survey 130 . In line with this priority, the Commission issued a policy communication in April 2012 (“Employment package”), in which it identified the EU’s biggest job potential areas and the most effective ways for Member States to create more jobs. Among other objectives, the employment package also aims to contribute to a genuine EU labour market. Boosting labour mobility as a means of boosting the EU's economic growth, alleviating social suffering among EU citizens and opening opportunities for their personal and professional development were also the aims pursued by Action 1 of the EU Citizenship report 2013 "EU citizens: your rights, your future", in which the Commission announced its intention, cited above, to propose a revision of the social security coordination regulation, looking in particular into extending the export of unemployment benefits for longer than the mandatory three months, to make it easier for citizens to look for a job in another EU country. 

For both topics, while it might be complicated to define SMART operational objectives, there a number of indicators that can be used as proxies to measure the progress in achieving the objectives, such as the number of Court cases, the amount of complaints raised to the Commission, the number of infringement procedures opened against Member States and the results of surveys that can be launched ad hoc (including surveys to EU-funded networks and services that help EU citizens to resolve their doubts and problems when dealing with public institutions and EU rules, such as EURES advisors, the SOLVIT network or Europa Direct).

The order in which the objectives are presented does not reflect any particular level of importance.

The following objectives have been identified at the specific and operational level for the coordination of long-term care benefits. The specific objectives are outlined, and for each of them the operational objectives corresponding to it are listed and briefly discussed. Operational objectives might relate to several specific objectives. To the extent possible, the objectives intend to express the desired future state, towards which the policy options should contribute. The links of the objectives with the identified problems the current coordination rules present are also briefly introduced.

Improve protection of mobile citizens

Ensure EU rules do not leave any migrant citizen without protection: the current coordination rules leave a gap in protection. Citizens reliant on care moving from a competent country which only grants benefits in kind to a country (new country of residence) that grants only benefits in cash can be left without any protection. The coordination rules must aim to close this gap and guarantee the protection of all Member States’ citizens.

Make legal framework simple, transparent and easy to administer: beyond the gap in protection that can arise depending on the types of benefits granted in the countries involved in the mobility of the citizens reliant on care, the protection of mobile citizens can be hampered by other reasons. These include the uneven or incorrect application of the coordination rules. Failures to correctly and consistently identify the benefits provided under each national system that should be considered as LTC and hence coordinated, can also lead to losses of benefits (e.g. Member States considering certain LTC benefits as social assistance and not applying the coordination rules). Similarly, the coordination of LTC benefits under different social security branches in different Member States can also generate issues in terms of the protection of citizens. Public administrations can face obstacles to apply the rules given the complexity of the legal framework. For these reasons, a simpler, more transparent and easy to administer legal framework would contribute to ensuring the protection of mobile citizens in the EU.

Facilitate classification of LTC benefits under EU law

Common understanding of LTC benefits across Member States (common definition of LTC or common criteria to determine them): there is currently no common definition of LTC at EU level, nor a common set of criteria to determine which benefits provided under each national system should be considered LTC benefits and hence coordinated. The ECJ has filled some of this gap, but problems of classification persist. Regardless of the means employed, a common understating of what LTC is a pre-requisite to make a clear classification of LTC benefits under EU law possible.

Make legal framework simple, transparent and easy to administer: a general simplification of the legal framework and accrued transparency would contribute to facilitate the classification of LTC benefits under EU law. No specific coordination rules exist for LTC, which must in turn be coordinated as sickness benefits. This results in a series of mismatches given the different characteristics of both types of benefits. In this particular case, the operational and specific objective could be seen as the two sides of the same coin. A simpler, more transparent and easier to administer legal framework would facilitate the classification of LTC benefits under EU law. At the same time, however, facilitating such classification would be in itself a major step towards simplifying the legal framework.

Rules taking into account the specific characteristics of LTC benefits

Make legal framework simple, transparent and easy to administer: the complexity of the current coordination rules and the different problems derived from it stem in part from the lack of a specific coordination regime for LTC and the application instead of the rules for the coordination of sickness benefits. LTC benefits present however distinct characteristics. The application of the rules from the sickness hence fail to account for the specific characteristics of LTC benefits, giving place to certain shortcomings that should be overcome.

Improve legal certainty & Stable regime of coordination of LTC benefits

Ensure EU rules do not leave any migrant citizen without protection: creating (or modifying existing ones) provisions ensuring that no citizen is left without protection regardless of his/her circumstances would in itself contribute to increase legal certainty. Cases of lack or total lack of protection would be prevented, and hence it would be expected that less cases and proceedings would be brought before the Court of Justice, which would bring some stability to the coordination regime.

Make legal framework simple, transparent and easy to administer: a simpler, more transparent and easy to administer legal framework would enhance legal certainty and, as direct result, make the system of coordination of LTC benefits more stable.

Limit individual complaints and proceedings before the Court of Justice: setting limiting the complaints and proceedings before the Court of Justice as an operational objective facilitates the measurability of improvements in terms of legal certainty and stability. The current instability of the system of coordination, manifested by the amount and relevance of the ECJ’s rulings in the field, leads logically to the objective of limiting the amount of cases and procedures.

High awareness of migrant citizen’s LTC rights

Make legal framework simple, transparent and easy to administer: as evidenced by the interviews conducted during the project, the results of the EC public consultation and Deloitte’s online survey, and as also pointed out by trESS, the complexity and intricacies of the legal framework for the coordination of LTC benefits make it difficult for migrant citizens and public authorities to know with certainty the entitlements of migrant citizens to LTC rights. A simpler, clearer framework would make it easier to gain knowledge about LTC rights.

Improve information regarding LTC in cross-border situations among citizens and public authorities: beyond the difficulties brought about by the complexity of the rules, a way to raise the awareness about the LTC rights of migrant citizens is directly tackling information and communication activities. While not all aspects of the coordination might be entirely clear even for those actively looking for information about it, the knowledge of the average citizens about LTC rights in case of mobility within the EU could be increased.

Fair distribution of the financial burden between Member States

Ensure there are no double payments: the current diversity of benefits provided in the MS that should be considered as LTC benefits for the purposes of the coordination, the actual coordination of the benefits under different social security schemes, the consideration of LTC as social assistance by some MS (which do not apply the coordination rules), combined with the lack of a definition of LTC or common criteria at EU level, results in difficulties to apply the provisions (Art.34) to avoid the overlapping of benefits. As a result, the risk of double payments persists. To guarantee a fair distribution of the financial burden, it must be ensured that the cost of providing LTC benefits is borne by the competent MS and that no double payments are done for the same risk. This, beyond the pure distributional aspects, can also lead to higher overall costs and affect negatively the sustainability of the system.

The following objectives have been identified at the specific and operational level for the coordination of unemployment benefits. As has been done for long-term care, the specific objectives are outlined, and for each of them the operational objectives corresponding to it are listed and briefly discussed. Operational objectives might relate to several specific objectives. To the extent possible, the objectives intend to express the desired future state, towards which the policy options should contribute. The links of the objectives with the identified problems the current coordination rules present are also briefly introduced.

Consistent application of the coordination rules

Uniform application of rules on the aggregation of periods: a survey launched within the Administrative Commission in 2011, the findings of the trESS Think Tank Report 2012 and the results of the interviews with public authorities conducted for this project indicate that Member States apply the rules on aggregation of periods in widely different ways. As reported in the answers to the Administrative Commission survey, some would not be in accordance with EU law. Therefore, in order to ensure a consistent application of the coordination rules, the rules on the aggregation of periods must be applied uniformly.

Simplified legal framework

Make legal framework simple, transparent and easier to administer: as part of the general effort to render the legal framework for the coordination of unemployment benefits less complex, a simpler, more transparent and easy to administer legal framework should be achieved. These objectives reflect in reality the same purpose, only the formulation varies to provide a more operational nuance, where proxies for output indicators can be found and hence progress towards achieving the goal monitored. The currently uneven application of the coordination rules, as well as the low level of awareness of mobile unemployed workers about their rights and the information and knowledge problems faced by public authorities justify the need to simplify the legal framework.

Facilitate reintegration in the labour market & improved protection of job-seekers abroad

Ensure all unemployed workers the possibility to export their unemployment benefits and look for a job abroad: not all unemployed workers in the EU enjoy currently the same possibilities to export their unemployment benefits and look for a job abroad. The rules about the export of periods are applied unevenly by Member States, creating differences among unemployed workers. To maximise the possibilities for reintegration in the labour market and providing a fair and similar chance to unemployed workers across the Member States, the rules must ensure to all unemployed workers the possibility to export their unemployment benefits and look for a job abroad.

Unemployed mobile workers receiving return on their contributions

Ensure link between contributions and benefits, and contributions and payments (financial burden): under the current coordination rules, wholly unemployed frontier workers might receive a low return on their contributions due to the fact that they receive their unemployment benefits from a country (their country of residence) other than the one in which they paid contributions (their country of last activity). The lack of a required minimum period of employment (or self-employment) to determine the country of last activity might also lead certain unemployed cross-border workers to receive benefits disproportionately lower with regard to their contributions. A similar effect might arise due to the fact that the unemployment benefits are calculated based only on the salary perceived during the last activity. 131  

Fair distribution of the financial burden

Ensure link between contributions and benefits, and contributions and payments (financial burden): the country of residence must currently bear the costs of the unemployment benefits of wholly unemployed frontier workers and certain cross-border workers other than frontier workers without having received any contribution from them. Although a reimbursement mechanism exists, it has a limited coverage. It does not cover all the costs of the unemployment benefits provided by the country of residence in any case (the period to be reimbursed is limited to three or five months), and the country of residence must also bear other costs such as those of sickness and family benefits. Ensuring the link between contributions and payment is necessary to provide for a fair distribution of the financial burden between Member States.

High awareness of rights

Increasing awareness among citizens, employers, public authorities and other stakeholders about the rights of EU migrant workers: the results of the EC public consultation and Deloitte’s online survey point to a low level of awareness about the rights of unemployed mobile workers. The interviews conducted with public authorities in 14 Member States showed that it is also complicated for public officials, and not only potential beneficiaries of the benefits, to have a thorough and reliable knowledge of the rules. This was also pointed out by trESS. The complexity of the coordination rules, including the wide variety of cases contemplated in the legislation, contribute to render difficult higher degrees of awareness. In response to such situation, increasing the awareness of rights among citizens, employers, public authorities and other stakeholders about the rights of EU migrant workers is set as an objective.

Rules reflecting current job market conditions

Ensure all cross-border workers the chance to apply for unemployment benefits in the country giving them the best chances for reintegration: wholly unemployed frontier workers must now register and apply for unemployment benefits in their country of residence. While they might register as jobseekers in their country of last activity as a supplementary step, compliance with the requirements of the country of residence’s public employment service has priority. These rules were created under the assumption that wholly unemployed frontier workers enjoy the best chance for reintegration into the labour market in their country of residence. Given the changing market conditions, this might however not always be the case anymore. Higher chances of reintegration might exist in their country of last activity.

7Impact assessment of the policy options

7.1Introduction

Multiple decision paths can be chosen by decision-makers to tackle effectively the existing problems and to address the challenges related to social security coordination. Decision makers could analyse all the different options, assess any combination of options, and approach the final solution step by step. Some of these options are of horizontal nature and could be taken irrespective of the way of coordination chosen; others are clearly separated paths of coordination. This makes it very complex to take a clear-cut decision.

In this chapter we present a description of each option. We would like to note that we have not examined all possible options, but a selection taking into account the work of trESS on this issue and the mandate of the Commission services for this study.

Furthermore, this chapter focuses on the assessment of the likely impacts of the policy options. All policy options have been assessed according to the same assessment criteria.

7.1Impact assessment criteria

The following impact assessment criteria have been applied:

Table 38: Impact assessment criteria

Impact on social security coverage of the insured persons

An important question is whether the social security rights of the person concerned are well protected 132 . This means that we have checked if all benefits which could be claimed without a cross-border situation can also be granted in a cross-border situation or if the person loses entitlements and thus could in the worst case be left without any entitlements although the Member States involved know such benefits.

Furthermore, concerning the coordination of unemployment benefits, we assessed all possible options on the return that workers receive on their contributions to the social security benefit system. Moreover, we evaluated the options with regard to their potential to create less or more favourable conditions for reintegration in the labour market of the unemployed. An important aspect that has been evaluated is the increase or decrease in services and assistance offered by public employment services to mobile jobseekers in the EU.

Impact on administrative burden, transparency and complexity of the rules

We assessed the impact of each option on the likely administrative burden on national administrations and other institutions that are involved in handling cross-border cases. More specifically with regard to the following questions:

Do new processes or new information flows between Member States have to be set up?

Do national administrations or other institutions have to set up new implementing arrangements to put the coordination into practice?

Impact on public finance of the Member States

First, we have assessed the impact on public finance of Member States, based on the impact on the social security coverage of the insured persons linked to each option. We have paid attention to the financial impact for Member States both at an aggregated level and at individual Member State level (where possible).

Second, we have assessed the impact of each option on the extent to which it enhances fair sharing of the financial burden between Member States. We would like to note that this is difficult to evaluate. Fair burden sharing between Member States is strongly depending on the system the Member States apply. Examples of unfair burden sharing:

Member States which have to reimburse benefits which their national legislations do not know may see this as unfair.

Member States with insurance-based schemes could be seen as burdened if they have to grant/reimburse benefits for persons which are not insured there and thus also do not pay any contributions towards these schemes of these Member States.

Examples of Member States which are not to be regarded as unfairly burdened:

Member States which are obliged to grant all the benefits which they would have to grant already under national legislation (e.g. in residence based tax financed schemes for all residents, such as in Sweden) are not to be regarded as unfairly burdened.

Impact on risk of abuse or misuse of rights

The options were evaluated based on their risk of abuse or misuse of rights by the citizens or workers, often referred in the general and more colloquial terms of “benefit tourism” or “social tourism”. One has to be very careful with the use of these concepts. Abuse is to be understood as “an artificial conduct entered into solely with the purpose of obtaining the right of free movement and residence under Community law which, albeit formally observing of the conditions laid down by Community rules, does not comply with the purpose of those rules” 133 .

Potential impact on EU internal market

We have also assessed if the options better stimulate mobility of persons and intra-EU labour market mobility - both in the sense of geographic mobility (movement of workers between countries and regions) and job-to-job mobility (e.g. moving to another job) than the baseline scenario. We particularly paid attention to the removal of barriers to labour market mobility.

Potential impact on EU budget expenditure

Impact on fundamental rights of citizen

The chosen option has to be compatible with the EU Charter of Fundamental Rights. The Commission “Impact Assessment guidelines” provide a full-list of fundamental rights in annex 8.1. Fundamental rights that are potentially affected by social security coordination rules are:

Chapter II – Freedoms:

oFreedom to choose an occupation and right to engage in work

oFreedom to conduct a business

Chapter III – Equality

oEquality before the law

oNon-discrimination

oThe rights of the child

oThe rights of the elderly

oIntegration of persons with disabilities

Chapter IV – Solidarity

oFamily and professional life

oSocial security and social assistance

oHealth care

Chapter V – Citizen's Rights

oRight to good administration

oFreedom of movement and of residence

With regard to the impact of prolongation of the period of export, several factors were specifically considered:

The impact the reintegration of people in the labour market;

Risks of misuse or abuse of the EU rules: we had to examine if this option favours situations where the persons concerned could easily influence and manipulate their situation in such a way that they receive more benefits than they would otherwise be entitled to. This is especially so if two Member States grant benefits or provide specific services and if it cannot be excluded that they do not know about the other Member States granting such benefits (the person concerned does not report the benefits received);

Intra-EU labour market mobility;

Social security coverage, particularly with regard to the return on the workers’ contributions.

7.2Scoring indicators

A scoring system was developed to allow judgments on ‘performance’ of an option with regard to the impact assessment criteria described above. The baseline scenario - a continuation of the current rules - served as the benchmark, as the decision maker needs to know if an option is better or worse than the situation we are confronted with today:

A ‘+’ indicates that an option is better than the baseline scenario with regard to of the criterion in question.

A ‘-’ indicates that an option is worse than the baseline scenario.

We used a “+/-” when it was not clear which impact the option will have with regard to the criterion in question or in case evidence gives a mixed picture (both positive and negative impacts were identified).

We use “0” for the identified impacts in the baseline scenario.

We have chosen a simple and straightforward scoring system in order not to complicate the assessment of the impacts. trESS made a similar consideration regarding the value system 134 135 .

The different impact assessment criteria were not weighted.

7.3Coordination of long-term care benefits

The figure below presents an overview of the options that were assessed on their likely impacts for the coordination of the LTC benefits.

Figure 6: Options - Coordination of LTC benefits

The baseline scenario (OPTION A) implies no change of policy and a continuation of the current rules for coordination.

Long-term care benefits in kind continue to be provided according to the legislation of the Member State of residence (if they exist) and reimbursed by the competent Member State (= where the person is insured);

Long-term care benefits in cash (if they exist) continue to be provided by the competent Member State (= where the person is insured).

The baseline means to change nothing and keep the coordination rules as they stand today. This implies that outlined problems and challenges. The status quo is the benchmark for any other alternative option proposed.

Long-term care benefits in cash will be provided by the Member State of residence if the application of the rules under the Sickness Chapter would mean that a person does not receive any long-term care benefits. The amount of the benefits provided by Member State of residence would then be reimbursed by the competent Member State.

With this policy option, the safeguarding provision would ensure that all the mobile persons are protected: they would always be entitled to some LTC benefits. Whenever a person would not be entitled to any LTC benefits (neither in cash nor in kind) under the current system (meaning the application of the Sickness Chapter’s rules), the Member State of residence would provide benefits in cash in accordance with its legislation. The competent Member State would then reimburse to the Member State of residence the amount of the benefits provided.

This safeguarding option however does not guarantee that the level of protection for long-term care would not be reduced after the person's move.

Score: +

This option addresses the current problem of the gap in the protection of mobile persons, as it can overcome the complete loss of social security protection; for example if the competent Member State does not have any LTC benefits in cash (only benefits in kind) and the Member State of residence has only LTC benefits in cash. However, the option does not guarantee perfect social security coverage of mobile persons reliant on care and cannot ensure that the protection would remain at the same level as if the person remained in his/her competent Member State. Therefore, a differential supplement would have been needed.

One EU civil society representative made the remark that under this option, “mobile persons – a small minority in the total population - would be put at a disproportional advantage over the non-mobile persons in that country. In case the competent country does not provide any long-term benefits at all, those people can receive cash benefits in the country of residence and would thus be better-off than non-mobile persons in the competent Member State (who would not be entitled to any benefits)”.

This option is likely to increase the administrative burden compared to the baseline scenario, as it requires a complex and burdensome cooperation between national public authorities (worse than the baseline scenario). This would be particularly the case if the obligation of the Member State of residence to provide LTC benefits depends on the actual granting of LTC benefits by the competent Member States (and not on the question whether or not such benefits are included in the legislation of the competent Member States).

Transparency: -

This option would be less clear than the status quo. Many questions could arise. It has to be clarified what exactly is meant by the condition that the competent MS does not grant LTC benefits in cash. Does this mean that no LTC benefits in cash are provided under the national legislation or does it also cover situations in which this legislation knows such benefits but the person concerned does not fulfil the conditions for entitlement (e.g. because the relevant degree of need of care or the relevant age are not reached – a question similar to the question concerning the existing text for taking into account of child raising periods under Article 44 of Regulation (EC) No 987/2009)? For the MS of residence this would give rise to many uncertainties. In relation to some MSs (which are competent for the persons concerned) LTC benefits in cash would have to be granted, whereas not in relation to others (in the same situation - depending on whether the competent MS grants such benefits or not).

Complexity of the rules: -

This option does also not provide more simple rules compared to the baseline scenario. All situations have to be explained in detail under which in addition to the obligations of the competent Member State, the Member State of residence has to grants its LTC benefits in cash. Furthermore, for the MS of residence it could become very complex in case it also has benefits in kind. Which of these benefits should this MS grant by priority (both benefits will be reimbursed); could the competent MS refuse the reimbursement of the more expensive version? Options with reimbursement are usually more complex than without reimbursement.

It could also be questioned why this obligation only becomes applicable when the competent MS does not at all grant any LTC benefits in cash, but not if it grants an LTC benefit in cash which is lower than the benefit in the MS of residence.

Administrative burden on public authorities: -

This option does not tackle any of the major problems in the baseline scenario which have an impact on administrative burden on public authorities (e.g. uneven application of coordination rules, problems of classification of LTC benefits under EU law and insufficiently specified reimbursement mechanisms).

It has to be considered that– as regards the administrative burden of granting benefits – the Member State of residence is expected to face an additional administrative burden compared to the baseline scenario.

Furthermore, when opting for a system of reimbursement for benefits in cash provided by the country of residence (which can be a sub-option), the administrative burden on both Member States is likely to increase even more.

According to a German civil servant, “this option would lead to substantial, additional administrative burden. First, additional administrative steps are needed to ensure the implementation of the ‘safeguarding provision’. Based on this, the insured person needs to submit a request to receive the LTC benefits (enabled by the ‘safeguarding provision’). Following, the country of residence will need to request information about any LTC payments by the competent Member State. The competent Member State will then reply to this request by sharing information on the LTC benefits that were already paid to the person. As a next step, the Member State of residence will need to provide the long-term benefits in cash if the application of the Sickness Chapter would mean that the person does not receive any long-term care benefits. Finally, in case of a reimbursement system, the country of residence will need to prepare and send a claim for reimbursement to the competent Member State”.

Similar administrative burden problems can be expected as in the case of reimbursement between Member States for unemployment benefits to frontier workers (Article 65 (5) and (6) of Regulation (EC) No 883/2004). A national administration explained:

“The reimbursement mechanisms are not properly applied by the institutions/bodies concerned. The terms of reimbursement are not applied. In some cases, Member States do not fully reimburse the costs incurred. This presents a major problem which could be aggravated by the introduction of this option.”

One should however put these additional administrative costs into the right perspective, as option B will only need to be applied in a limited number of cases.

Impact on public finance of Member States: +/-

Under this option, the protection of the mobile persons would thus come first. Once it had been ensured, the issue of which Member State bears the cost, and the opportune reimbursement, would be resolved. National public authorities considered that although logically an increase in public expenditure can be expected, the impact on their budgets of this option would be negligible.

Fair burden sharing between Member States: -

A sub-option where the benefits provided are reimbursed could contribute to fair burden sharing. However, there is also a flip side of the coin. According to a legal assessment by the trESS network 136 , it could be said that it is an additional burden for the competent MS, which “might deliberately make a decision to compensate the risk of LTC only by way of benefits in kind. This MS would thus also have to reimburse benefits in cash”.

Score: +

With this policy option, the safeguarding provision would ensure that the mobile persons are better protected: they would not remain without any entitlements due to the coordination rules. In this manner, mobile persons would not see their mobility options restrained because of a lack of LTC coverage as an effect of their decision to move.

Score: +

By contributing to eliminate gaps in protection and ensuring that mobile citizens reliant on care can receive a minimum level of care, this policy option has a positive impact on different fundamental rights that it would to promote: non-discrimination (certain citizens would not suffer a total loss of benefits based on their country of origin), the rights of the elderly, the rights of the persons with disabilities, family and professional life and, in general, on the fundamental right to social security and social assistance.

By guaranteeing that mobile citizens would not be left without any protection, this policy option would also contribute positively to ensuring the right of freedom of movement and of residence. Thus, citizens would not see their mobility options based on a possible total loss of entitlements to long-term care benefits.

Score: +/-

No impact on the EU budget could be identified.

In Option C.I, the Member State of residence shall provide long-term care benefits on the basis of its legislation:

with a supplement from the competent Member State if benefits in Member State of residence are at lower level (C.I.1)

without a supplement from the competent Member State (C.I.2).

This policy option sets out clearly which Member State would provide LTC benefits. Contrary to the baseline scenario, where the distinction between benefits in cash and in kind is the starting point and an element of crucial importance (as it determines which Member State must provide the LTC benefits), this policy option would not make any distinction among the types of benefits. The starting point and critical element would be the Member State legally in charge of providing the LTC benefits (the Member State of residence or the competent Member State depending on the sub-option).This Member State would then provide all LCT benefits (be them in kind, in cash or mixed), with a system of reimbursement.

The competent Member State would always be responsible for the cost of the benefits.

In Option C1, the Member State of residence would provide the LTC benefits (again, regardless of their type) on the basis of its legislation, and the competent Member State would then reimburse them. Option C1 includes two sub-options itself: in the first one, the competent Member State would provide a supplement to the beneficiary in the event that the benefits in the Member State of residence were at a lower level (in this case -in line with the non-distinction amongst benefits in this policy option-, the comparison of the level of the amount of LTC benefits between the residence and the competent Member State would have to include all types of LTC benefits). In the second sub-option, there would be no compensation even if the benefits in the Member State of residence were lower than those in the competent Member State.

Attitude of consulted stakeholders

Score C1.1 and C.I.2: +

Both options would ensure that mobile persons are always protected at the same level as the persons who are insured in the Member State of their residence. C1.1 option (with supplement) would also, address the problem of the gap in the protection of mobile persons. Both sub-options give entitlements to all the benefits which are provided for the residents of a Member State. This can in itself be seen as positive compared to the baseline scenario (which is why both sub-options received a positive scoring). In our public authorities’ survey, national administrations highly scored both sub-options with regard to their impact on the social protection of insured people. Sub-option C.I.2 with no supplement received an average rating of 4.3 on a 1-to-6 scale; whereas the sub-option C.I.1 with supplements remarkably scored slightly lower (4.0).

Findings from the face-to-face interviews reveal that the sub-option, which includes a supplement by the competent Member State (C.I.1), is without any doubt seen as better than the baseline scenario and (in many cases) also than C.I.2 (without supplement). It seeks very effectively to avoid that rights are lost and the highest benefits are always safeguarded.

With regard to the sub-option without supplement by the competent Member State (C.I.2), interviewees expect that it will depend on the specific case whether there will be a positive or negative impact on social security coverage. Depending on the specificity of the long-term care benefits in the country of residence and its generosity, citizens will be better or worse off in terms of social security coverage.

With regard to this point, one trade union representative stated: “There is a large discrepancy between the social systems between the Member States; the insured persons would be disadvantaged should no supplements be granted”. It illustrates the overall opinion of the interviewees.

In addition, an Austrian national administration employee/representative mentioned, “Austrian citizens (who have paid social security contributions in Austria) would be worse-off if they moved to a Member State with a lower level of LTC services or to a country that does not grant any LTC benefits at all.”

C.I.1 - Supplement provided by the competent Member State

Transparency: +

This sub-option would bring slightly clearer rules than the baseline scenario. The MSs involved would always know that they have to grant benefits (the MS of residence always provide the whole range of benefits; the competent Member State provides a top-up on these benefits in case its benefits are higher). Furthermore, the person concerned would know where to request the benefits and what legislation applies. The division of tasks between MSs would be clear and would not depend that much on peculiarities of the national legislation.

Complexity of the rules: -

In terms of complexity, this sub-option would open simultaneous entitlements under the legislations of more than one MS. This is perceived as more complex than the baseline scenario. The provision of supplements is always complex, as the legislator has to draft complex rules concerning priority, the benefits to be included into the calculation of the supplement (benefits in cash or also benefits in kind, what ‘amount’ of the benefits in kind etc.) and rules for procedures to settle these supplements.

Administrative burden on national administrations: -

See C.I.2

This a complex option from an administrative point of view, which is regarded as slightly worse than the baseline scenario according to the public authorities’ survey results (it receives 3.6 average rating on a 1-to-6 scale compared to a 3.5 average rating for the baseline scenario; a lower score means less expected administrative burden).

This option necessitates an extensive exchange of information between national public authorities concerned. Especially if also benefits in kind are included, it could lead to different supplements every month. It would necessitate a lot of new business flows and SEDs under the potential future EESSI.

The complexity and the administrative burden of supplement system is generally the main reason for the low support for this option among national public authorities.

C.I.2. No supplement provided by the competent Member State

Transparency and complexity of the rules: -

This sub-option would bring clarity, simplicity and legal certainty, by setting out clearly which Member State would provide LTC benefits. Contrary to the baseline scenario, where the distinction between benefits in cash and in kind is the starting point and an element of crucial importance (as it determines which Member State must provide the LTC benefits), this policy option would not make any distinction among the types of benefits. The starting point and critical element would be the Member State legally in charge of providing the LTC benefits. In this option, beneficiaries are able to know in advance which legislation will systematically apply to them depending on the MS where they will reside. There will be no doubts even if it is not clear under the relevant legislation whether a certain benefit is a benefit in cash or in kind.

It was gathered during the interviews with public authorities carried out in certain Member States that this option would help to circumvent to a certain extent the issues caused by the problems of classification of LTC benefits under EU Law. Without tackling the underlying problem, this policy option would help to bring certain clarity.

Administrative burden on national administrations: -

At first sight, the administrative burden on Member States’ administration is expected to decrease, because of reduced complexity of the rules under this option (only one Member State provides the long-term care benefits and no distinction between benefits in kind and benefits in cash).

In Option C 1.2 all calculations for the in kind benefits remain the same. The in cash benefits are now however defined at the % of use and the level of spending of the country of residence. The particularities of our sources reveal that even more persons are using LTC in cash, but the average amount is much lower so that the total budget for in cash is reduced to € 192 million and the total becomes € 810 million (cf. 8.2.3 ). The administrative burden is even somewhat higher (more cases means more costs) and compared to a lower needed budgetary cost, the share of the administrative burden in the total budget is enlarged to 1.7%, highlighting the impact of probably estimated number of cases.

The provision of all types of LTC benefits by the Member State of residence would require setting up a procedure for the reimbursement of LTC benefits in cash. The current reimbursement procedure only covers benefits in kind. Based on our interviews with national administrations, it appears that setting a system for the settling of monetary amounts, however, would not entitle major problems, as no differentiation needs to be made between benefits in cash and kind.

Impact on public finance of Member States: +

For options C1, namely claims (as debtor and as creditor) based on LTC spending per capita of the country of residence (option C1), the budgetary impact of a change in the Regulation was estimated 137  (based on estimated number PD S1 forms and variables ‘Health care spending in 2010 prices per person’, ‘Population (million)’ and ‘Long-term care spending in 2010 prices (in billion Euros)’ in the Ageing Report). For both options, a distinction should be made between LTC benefits provided in kind and in cash.

This option, whereby the Member State of residence shall provide LTC benefits (in kind and in cash) on the basis of its legislation, implies a total expenditure of € 810.1 million or a decrease of 19% compared to the baseline scenario 138 .( Table 53 and Table 54 ). 

In this option the expenditure on LTC benefits in kind shall be the same as in the baseline scenario, namely € 618.3 million. The LTC benefits in cash shall be provided on the basis of the legislation of the country of residence and no longer on this of the competent country. 139 This switch results in an expenditure on LTC benefits in cash of € 191.9 million or a decrease of 49% compared to the baseline scenario ( Table 53 and Table 54 ).  .

On the level of Member States especially a positive impact (less spending) is observed for Italy (decrease of 49% of expenditure on LTC benefits), Austria (decrease of 47% of expenditure on LTC benefits), Denmark (decrease of 43% of expenditure on LTC benefits) and the Netherlands (decrease of 42% of expenditure on LTC benefits). Primarily, a negative impact (more spending) is observed for the Slovak Republic (increase of 106% of expenditure on LTC benefits), Romania (increase of 60% of expenditure on LTC benefits) and Bulgaria (increase of 51% of expenditure on LTC benefits). Countries in which we do not observe a crucial negative or positive impact are: Belgium (same expenditure as in the baseline scenario), Estonia (same expenditure as in the baseline scenario), Germany (decrease of 1% compared to the baseline scenario) and France (increase of 3% expenditure compared to the baseline scenario) ( Table 54 ).

The changes are not always as what is expected since they are sensitive to the real levels of spending. For instance the shift of the reimburse mechanism to the basis of spending at the level of residence implied an increase for Luxembourg (as competent country) because they were ‘less generous’ (or less oriented) to in cash benefits.

Fair sharing of the burden between Member States: -

Reimbursement is a step towards burden sharing, but in this option, the negative aspects compared to the baseline scenario outweigh the positive impacts. As also the trESS Analytical report on LTC remarked:

“It could be said that it is an additional financial burden for the competent Member State, which might deliberately make a decision to compensate the risk of LTC only by way of benefits in kind. Under this option, these Member States would thus also have to reimburse benefits in cash”. 

According to the results of our public authorities’ survey, this option also received the lowest average score with regard to its impact on fair burden sharing (3.39 on a 1-to-6 scale).

Score: -

This sub-option incorporates risks of misuse of benefits, as two Member State could be competent to grant benefits at the same time. Persons concerned could thus be tempted not to inform the competent MS of the receipt of benefits under the legislation of the Member State of residence (leading to double payments).

Furthermore, several national administrations in Member States with more generous LTC benefits warned that this option could lead to social tourism in the EU (Austria, Germany and the Netherlands). One German national administration representative stated:

“This option may lead to an increase of inflow of persons from ‘poorer’ Member States seeking LTC benefits in our country (the standard of LTC benefits in our country is relatively high compared to other Member States). Indirectly, this could also result in an increased pressure on the (supply of) nursing staff, in case there is a high inflow of mobile persons seeking long-term care services as a result of this option.” 

This argument was repeated by an Austrian national administration representative:

“In 2012, the Austrian Bund (federation) allocated EUR 2.6 billion for LTC; there were 430.000 recipients of benefits. The benefits ranged between EUR 154 - EUR 1655 per month. Approximately 5% of the Austrian population receives LTC benefits, which is relatively high compared to other EU countries. For instance, Germany has a lower share of population receiving LTC benefits. The threshold as of which people receive LTC benefits is relatively low in Austria. Under this option, a Romanian citizen, who has not paid any contributions into the Austrian social security system, would be entitled to receive LTC benefits from Austria. So, reimbursement between Member States is crucial”.

C.I.1 – No supplement provided by the competent Member State +/-

Should there be no supplement, the difference in the level of LTC benefits between the competent Member State and the Member State of residence would seem to be a factor in the decision of (potentially) mobile persons.

C.I.2. Supplement provided by the competent Member State: +

This option would prevent the situation where persons would not move due to the impossibility of receiving any LTC benefits in their new country of residence.

According to the public authorities’ survey, national administrations believe that this option is one of two best options to stimulate free movement of persons within the EU (3.94 average score on a 1-to-6 scale), with a higher average score than the sub-option with supplement.

Score: +

For both options by contributing to eliminate gaps in protection and ensuring that all mobile citizens reliant on care receive adequate levels of care, this policy option has a positive impact on different fundamental rights that it would to promote: non-discrimination (citizens would be less likely to suffer loss of benefits based on their country of origin), the rights of the elderly, the rights of the persons with disabilities, family and professional life and, in general, on the fundamental right to social security and social assistance.

By removing obstacles to mobility linked to the potential lack or loss of rights to long-term care and increasing the transparency of the rules (mobile citizens should have more visibility on their entitlements with this policy option than they do with the current coordination rules), this policy option would also contribute positively to ensuring the right of freedom of movement and of residence.