COMMISSION STAFF WORKING DOCUMENT IMPACT ASSESSMENT Accompanying the document Proposal for a Council recommendation on access to social protection for workers and the self-employed - Hoofdinhoud

Council of the European Union

Brussels, 21 March 2018 (OR. en)

7416/18

Interinstitutional File: ADD 1

2018/0059 (NLE) i

SOC 164 EMPL 124

COVER NOTE

From: Secretary-General of the European Commission, signed by Mr Jordi AYET PUIGARNAU, Director

date of receipt: 14 March 2018

To: Mr Jeppe TRANHOLM-MIKKELSEN, Secretary-General of the Council of the European Union

No. Cion doc.: SWD(2018) 70 final

Subject: COMMISSION STAFF WORKING DOCUMENT IMPACT ASSESSMENT Accompanying the document Proposal for a Council recommendation on access to social protection for workers and the self-employed

Delegations will find attached document SWD(2018) 70 final.

Encl.: SWD(2018) 70 final

EUROPEAN COMMISSION

Strasbourg, 13.3.2018 SWD(2018) 70 final

COMMISSION STAFF WORKING DOCUMENT

IMPACT ASSESSMENT

Accompanying the document

Proposal for a Council recommendation

on access to social protection for workers and the self-employed

{COM(2018) 132 final i} - {SWD(2018) 71 final} Table of contents

• 1. 

  INTRODUCTION: POLITICAL AND LEGAL CONTEXT ........................................................................... 1

• 2. 

  PROBLEM DEFINITION ..................................................................................................................... 2

2.1. A FIRST LOOK AT HOW MANY PEOPLE ARE POTENTIALLY NOT COVERED BY SOCIAL SECURITY ............................... 3

2.2. P ROBLEM DRIVERS .............................................................................................................................. 6

2.3. C ONSEQUENCES ............................................................................................................................... 12

• 3. 

  WHY SHOULD THE EU ACT? ........................................................................................................... 16

3.1. L EGAL BASIS .................................................................................................................................... 16

3.2. R ESULTS OF THE SOCIAL PARTNERS CONSULTATIONS AND OF THE PUBLIC CONSULTATION ................................ 16

3.3. S UBSIDIARITY : N ECESSITY AND VALUE - ADDED OF EU ACTION .................................................................... 17

• 4. 

  OBJECTIVES: WHAT IS TO BE ACHIEVED? ....................................................................................... 18

4.1. G ENERAL OBJECTIVES ........................................................................................................................ 18

4.2. S PECIFIC OBJECTIVES ......................................................................................................................... 19

• 5. 

  WHAT ARE THE AVAILABLE POLICY MEASURES? ............................................................................ 20

5.1. B ASELINE SCENARIO .......................................................................................................................... 20

5.2. M EASURES DISCARDED AT AN EARLY STAGE ........................................................................................... 23

5.3. D ESCRIPTION OF THE POLICY MEASURES ................................................................................................ 24

5.4. I MPACTS OF THE POLICY MEASURES ...................................................................................................... 28

• 6. 

  WHAT ARE THE POLICY PACKAGES AND THEIR IMPACTS? ............................................................. 36

6.1. D ESCRIPTION OF ALTERNATIVE POLICY PACKAGES .................................................................................... 36

6.2. I MPACTS OF THE POLICY PACKAGES ...................................................................................................... 38

6.3. C OMPARING THE POLICY PACKAGES ...................................................................................................... 39

• 7. 

  PREFERRED PACKAGE .................................................................................................................... 41

7.1. P REFERRED POLICY OPTION ................................................................................................................ 41

7.2. O VERALL IMPACTS ............................................................................................................................ 41

• 8. 

  THE CHOICE OF LEGAL INSTRUMENT ............................................................................................. 43

8.1. L EGAL OPTIONS ................................................................................................................................ 43

8.2. P REFERRED INSTRUMENT ................................................................................................................... 44

• 9. 

  HOW WILL ACTUAL IMPACTS BE MONITORED AND EVALUATED? ................................................. 46

9.1. M ONITORING .................................................................................................................................. 46

9.2. E VALUATION ................................................................................................................................... 47

ANNEX 1: PROCEDURAL INFORMATION ................................................................................................. 48

ANNEX 2: GLOSSARY AND TERMINOLOGY ............................................................................................. 50

ANNEX 3: FURTHER INFORMATION ON THE ORGANISATION OF SOCIAL PROTECTION SYSTEMS ........... 55

ANNEX 4: FURTHER INFORMATION ON NON-STANDARD EMPLOYMENT AND SELF-EMPLOYMENT ....... 62

ANNEX 5: VISUALISATION OF THE PROBLEM TREE ................................................................................. 74 ANNEX 6: INTERNAL DRIVERS ................................................................................................................ 75

ANNEX 7: OVERVIEW OF RECENT REFORMS IN MEMBER STATES........................................................... 93

ANNEX 8: THE EXISTING EU INSTRUMENTS .......................................................................................... 105

ANNEX 9: ANALYTICAL METHODS ........................................................................................................ 109

ANNEX 10: FURTHER INFORMATION ON IMPACTS OF POLICY MEASURES AND POLICY PACKAGES...... 115

ANNEX 11: STAKEHOLDER CONSULTATION .......................................................................................... 129

ANNEX 12: WHO IS AFFECTED AND HOW? ........................................................................................... 142

ANNEX 14: MONITORING INDICATORS ................................................................................................ 146

ANNEX 15: SUMMARY OF CASE STUDIES ............................................................................................. 149

ANNEX 16: BEHAVIOURAL STUDY ........................................................................................................ 157

• 1. 

  I NTRODUCTION : P OLITICAL AND LEGAL CONTEXT

Globalisation, technological developments and demographic ageing have driven changes in the European labour markets for the last two decades and will continue to do so in the future. Employment will increasingly be more diverse and careers will be less and less linear. As labour markets evolve, social protection systems too need reforms in order to make our social model future-proof making the most of the future world of work.

Offering access to social protection is crucial for the economic and social safety of the workforce and the well-functioning labour markets that create jobs and growth. Yet there is a growing number of people who, due to their employment status, are left without sufficient access to social security.

The Commission Reflection Papers on harnessing globalisation 1 and on the social dimension 2 recall that national social protection systems in the EU differ due to different national traditions, political preferences and budgets but they face similar, transformative challenges. They underline the importance of social policies for a better distribution of the benefits of globalisation so people are confident that governments will protect and empower them in the adaptation to change. They also stress the importance of delivering new protections to help individuals use their full capacities in fast-changing labour markets.

During the public consultation 3 on the European Pillar of Social Rights many stakeholders expressed concerns regarding the obstacles faced by workers in non-standard forms of employment and by the self-employed in accessing social protection such as unemployment benefits, sickness benefits, disability benefits and pensions.

In his State of the Union speech in September 2017, President Juncker further stressed that a social fragmentation and social dumping are to be avoided in Europe. He underlined that, whilst national social systems will still remain diverse and separate for a long time, the EU should strive for a European Social standards Union and for a common understanding of social fairness in the single market.

Solemnly proclaimed in November 2017, the European Pillar of Social Rights 4 is the EU's political commitment to address these challenges and to uphold citizens' social rights in times of structural changes, including by stressing that "regardless of the type and duration of their employment relationship, workers, and, under comparable conditions, the self-employed have the right to adequate social protection" (principle 12). The Commission's initiative on "Access to social protection for workers and the self-employed" 5 is a key step to make principle 12 of the Pillar a reality on the ground. The initiative (2017/EMPL/003) is contained in the

Commission Work Programme 2017 6 and 2018 7 and is a key component of the Social Fairness Package 8 to be adopted by the Commission on 13 March 2018.

The objective is to support people in non-standard forms of employment and self-employment who, due to their employment statuses, are not sufficiently covered by social security schemes and thus are exposed to higher economic uncertainty. By focusing on the level playing field

1 COM/2017/0240 final, available online.

2 COM/2017/0206 final, available online.

3 SWD(2017) 206, available online .

4 COM(2017)250 i, available online .

5 COM(2017)251 i

6 COM/2016/710 final, available online.

7 COM(2017) 650 final i, available online.

8 Together with a proposal for a Regulation of the European Parliament and the Council establishing a European Labour Authority (COM(2018)131 i) and a communication on the monitoring of the implementation of the European Pillar of Social Rights (COM(2018)130 i)

for economic agents at national level, the initiative will also aim at avoiding short-term competitions in the internal single market, complementing the coordination of social security across borders 9 already set at EU level and ensuring long-term gains stemming from the economic benefits throughout the Union.

While acknowledging that many social protection schemes aim at protecting all citizens - whether or not active in the labour market - the initiative is expected to concentrate primarily on social security schemes which are governed, financed and designed in function of the labour market dynamics and leaves out of scope social assistance and other social protection schemes which are based on residence/citizenship status. The initiative focuses in particular on social security areas which are mostly impacted by the changing nature of work relationships, namely unemployment benefits; sickness benefits; maternity and equivalent paternity benefits; invalidity benefits; old-age benefits as well as benefits in respect of accidents at work and occupational diseases.

This Impact Assessment describes the current gaps in social security for people in nonstandard forms of employment and self-employment across the Member States and discusses different policy options and impacts for EU action. It is important to note that the quantification of impacts described in this document does not capture the entire scope of the problem analysed. The limited nature of the quantification made stems from technical limitations of the statistical data and of the simulation models available. It also reflects the uncertainty related to the decisions on design and financing of social security systems. Since these decisions are to be left to the Member States, according to the principle of subsidiarity underlined in Article 153 (4) TFEU, the choices to be made are yet unknown and therefore a full measurement of their fiscal and economic impacts is not possible.

In light of the above, this document presents a proportionate assessment of impacts. The impacts are primarily described through a qualitative assessment and, where possible, a quantification of impacts is outlined.

Procedural information regarding the preparation of this Impact Assessment can be found in Annex 1. Annex 2 provides the glossary and definitions of key terms used in this document.

• 2. 

  P ROBLEM DEFINITION

Social security serves to protect people against the financial implications of social risks, such as ill health, old age or job loss, and contributes to preventing and alleviating poverty and social exclusion. It also functions as a buffer during economic downturns by supporting domestic demand. Social security can take the form of in-cash benefits which replace or supplement work income (unemployment benefits, pensions, sickness benefits, child benefits) and/or in-kind benefits which provide support through the supply of goods and of services (healthcare, child care, training or job search assistance). Further information on the basic characteristics of social security systems in the EU can be found in Annex 3.

Well-designed social security systems are an important lever of competitiveness and sustainable growth as they facilitate labour market participation and support human capital investment in a world of rapid and growing labour market transformations. Adequate social security contributes to a better allocation of the labour force towards emerging and dynamic sectors of the economy by providing workers opportunities and incentives to participate in

9 COM(2016) 815 final i, available online

education and training and devote time to search for a job matching their skills. It helps people feel secure in order to fully embrace the opportunities created by new forms and patterns of work, new working relationships with employers, growing career transitions as well as self-employment. Social security systems thereby fulfil important functions for individuals, labour markets as well as at macroeconomic level, where they act as automatic stabilisers dampening welfare-reducing economic cycles.

This initiative addresses the problem that a significant and increasing share of people in the EU's labour market has insufficient social security coverage due to their employment status, in particular non-standard workers and the self-employed.

This development impedes social security systems from fully fulfilling their multiple functions outlined above. Such gaps increase risks to the welfare of the affected individuals and their families who endure greater economic uncertainty. The accumulated effects of such disparities in entitlements are likely to give rise to new inter- and intra-generational inequalities between those who have or manage to gain employment on contracts with full social rights and those who do not. Women may also be more disadvantaged by social security systems still geared towards linear full-time employment as they are more exposed to non-standard employment, particularly to part-time work, but also to temporary work and to fragmented careers due to interruptions related to children or caring for dependent relatives.

Negative consequences are also felt at the aggregate level, including in terms of higher labour market segmentation and unfair competition between firms employing staff on non-standard contracts exempted from social security contributions and those not doing so. Further, there can be effects on the financing base of social security systems, and labour market dynamism to the extent that mobility and transitions between different employment forms are hindered by differences in social security coverage. In the long run, it is the social and economic sustainability of national social security systems that is at stake. Gaps in access to social security for growing groups of the workforce will lead them to take recourse to safety nets of last resort in the case of social risk while the number of people contributing to social security systems will be proportionately smaller.

2.1. A first look at how many people are potentially not covered by social security

Social security systems were primarily developed for and remain geared towards workers in

‘standard employment relations’, implying a long-term, full-time work relationship 10 .

Recently developments towards social security systems with more incentives to work have taken place, but the systems were not tailored to the specificities of people in different forms of employment, creating deficiencies in the way they are protected.

This potentially concerns a large share of people in the EU's labour market since the share of people in non-standard employment relationships and dependent self-employment has been increasing 11 . In 2016, 14% of the total 219 million people employed in the EU were selfemployed, 8% were full-time temporary employees, 4% were part-time temporary employees, 13% were part-time permanent employees and 60% were employees with a full-time permanent contract (Figure 1). Altogether, people in all categories of non-standard employment and self-employment constitute around 39% of the employed population.

10 Eurofound (2016), Sixth European Working Conditions Survey – Overview report.

11 European Commission, White Paper on the Future of Europe, COM(2017)2025 i, 1 March 2017.

In addition to the increasing diversity in employment forms, careers are also becoming less

linear. In the space of a generation, the average European worker has gone from having a job

for life to having more than ten in a career 12 . People are frequently moving between different

employment statuses and/or combining salaried employment and self-employment. Moreover,

transitions from school to work are taking more time than before, and many young people

accumulate several traineeships in the process, which in many Member States’ social security

systems implies delaying the start of the payments of contribution and thus has an impact on

the building up of entitlements. In general, evidence shows that the crisis has led to rising

involuntary temporary and part-time work and to increased perceptions of job insecurity

overall 13 , which may have been accentuated by the existing deficiencies in protection (further

information on the levels and the observed trend increase in non-standard employment and on

self-employment are presented in Annex 4).

Figure 1: Extent of different types of employment relationship in the EU28 in 2016

Source: Eurostat

Matsaganis et al. (2015) 14 estimated the risk for non-standard workers and the self-employed of not being covered 15 by three social benefit schemes: sickness, unemployment and maternity benefits. Their results show 16 a significant risk of not being covered by benefits for nonstandard

workers and the self-employed in relation to these three benefit schemes.

As shown in Figure 2, two million people employed on non-standard contracts in the EU

could be at risk of not being covered by sickness benefits, with a significant variation across

Member States. In particular, this risk concerns 5.1% of temporary full-time workers and

9.7% of temporary part-time workers in the EU. Concerning the risk not to be covered by

unemployment benefits, eight million people in non-standard contracts are at risk: 31.9% of

temporary full-time workers and 38.7% of temporary part-timers are at risk, against 0.1% of

the standard workers (Figure 3). When it comes to maternity benefits, the proportion of nonstandard

workers at risk of not being covered is 1.8% for permanent part-timers, 8.5% for

12 European Commission, SWD(2016)51 Key economic, employment and social trends behind the European Pillar of Social Rights.

13 EP (2016): Precarious Employment in Europe: Patterns, trends and policy strategies. Luxembourg. Publications Office of the European Union. Henceforth, this is referred to as P (2016).

14 Matsaganis et al. (2015): Non-standard employment and access to social security benefits. Luxembourg. Publications Office of the

European Union. Henceforth, this is referred to as Matsaganis et al. (2015).

15 A detailed explanation of the methodology and the assumptions of the approach can be found in Annex 9.

16 In view of the methodological approach, the estimates could be regarded as an upper bound for two reasons: Firstly, as explained in Annex 9, based on the Labour Force Survey, the analysis cannot precisely identify the numbers of non-standard workers and the self-employed who are unable to meet the requirements for eligibility at the present moment. Secondly, the research does not take into account access of all categories of the self-employed when access to benefit schemes is voluntary — e.g. the study assumes that ‘the self-employed with employees

belong to such schemes whereas those without employees do not’ (Matsaganis et al. 2015: 17).

temporary full-time workers and 12.7% for temporary part-timers. As will be shown later, in

general, the existing rules do allow non-standard employees to adhere to the corresponding

social security schemes ("formal coverage"), although some specific categories may be

excluded. The estimated figures thus also reflect the fact that despite being members, nonstandard

employees may often not be able to build up and access benefits ("effective

coverage").

Figure 2: Distribution of people in employment aged 15-64 at risk of not being entitled

to sickness benefits by type of employment, 2014 (% of the total in employment)

Source: Matsaganis et al. 2015

Figure 3: Distribution of people in employment aged 15-64 at risk of not being entitled

to unemployment benefits by type of employment, 2014 (% of total in employment)

Source: Matsaganis et al. 2015

Among the self-employed in the EU, about 12 million are estimated not to be eligible for

sickness benefits, again with a wide variation between countries (Figure 2). The risk of not

being covered by unemployment benefits for the self-employed is as high as 54.5%

(corresponding to ca. 17 million), again with significant cross-country differences (Figure 3).

In both schemes the variation across Member States relates to the fact that insurance is

mandatory in some Member States, voluntary in others and non-existent in another group of

countries. The proportion of the self-employed who are estimated not to be covered by

maternity or paternity benefits is much higher: 46.1%.

These figures illustrate the problem for three branches of social security. A qualitative mapping 17 by Member State, by social security branch and by employment category carried out for the purposes of this Impact Assessment (see Section 2.2.2 and Annex 6) shows that similar problems also exist for other branches of social security even if data limitations prevent the presentation of corresponding quantitative estimates.

Complementary evidence is available from a survey commissioned for this Impact

Assessment. 18 Considering the self-reported rates of coverage by key social security branches,

large gaps emerge between standard employees on the one hand and the self-employed and non-standard workers on the other (see Figure A.28 in Annex 16). These gaps are particularly pronounced when considering unemployment benefits and sickness benefits. Compared to standard employees, the incidence of self-reported benefit coverage in case of need is 50 percentage points (pps) lower for the self-employed and 29 pps lower for non-standard employees regarding unemployment benefits. For sickness benefits, self-reported benefit coverage is 35 pps for the self-employed and 33 pps for non-standard employees.

Differences are less pronounced in the case of old-age benefits between standard employees and the self-employed non-entrepreneurs (6 pps), while there is a difference of 22 pps between standard and non-standard employees. Similar gaps were registered in case of maternity, invalidity and accidents at works and occupational injuries.

It is important to notice that such gaps in self-reported coverage may reflect existing real limitations, but also reflect a lack of transparent information allowing people to be aware of benefits they are entitled to.

2.2. Problem drivers

The gaps of access in social security affecting non-standard workers and the self-employed are the result of external drivers to social security (like changes in labour markets, technological progress, globalisation and demographic change) as well as intrinsic ones. They are now reviewed according to the structure outlined in the problem tree shown in Annex 5.

2.2.1. External drivers

Several factors are shaping developments in the EU labour market(s) contributing to an increasing share of people being outside standard employment and thus potentially affected by gaps in social security coverage.

Intensifying globalisation has changed production processes and increased competition for

firms 19 . Resulting needs to increase cost efficiency have, among others, been addressed by

firms by lowering their labour costs – by offshoring and also by containing domestic labour costs and making it more flexible. As part of these efforts, employers may call in selfemployed workers or hire temporary staff as well as people in new forms of work. Such flexible forms of labour can at the same time be more compatible with the above-mentioned requirements of production in a global economy. Moreover, there is an ongoing shift in the

17 See ESPN (2017): Access to social protection for people in non-standard employment and in self-employment, Synthesis report. Luxembourg: Publications Office of the European Union. Henceforth, this is referred to as ESPN (2017).

18 Open evidence (forthcoming): Behavioural Study on the Effects of an Extension of the Access to Social Protection for People in All Forms

of Employment. Brussels: European Commission, Henceforth, this is referred to as Open Evidence (forthcoming).

19 Bang K.E., Markeset T. (2012): Identifying the Drivers of Economic Globalization and the Effects on Companies’ Competitive Situation. EC: Globalisation and the EU economy, available online .

sectoral structure of the EU economy from manufacturing towards an increasing role of services. Since non-standard employment is more prevalent in the services sectors, this shift

also implies an increase in this kind of employment 20 , including non-standard jobs in some high-skills service areas 21 .

The increasing recourse to existing forms of non-standard work as well as the emergence of new ones is also closely linked to recent technological developments transforming the world of work in unprecedented ways. New IT technologies, automation and digitalisation are changing the production and delivery of goods and services while technologies such as increased robotisation of standardisable tasks or the development of digital platforms are also

having important impacts on the organisation of work 22 . These developments are giving rise

to such phenomena as platform work, crowdsourcing and the collaborative economy. In conjunction with the above-mentioned changes related to globalisation, these technological transformations also resulted in higher use of traditional non-standard employment and selfemployment and to the development of new atypical working arrangements 23 , such as workers with no employment contract (casual workers; "zero hours" contracts), workers who report working a very small number of hours (< 8 hours a week) or who hold a temporary contract for less than 6 months.

At the same time, an increase in non-standard forms of work may result from policies in response to population ageing, which is weighing on the size of labour forces and putting financial pressure on social security systems, both on the expenditure side (healthcare, pensions) and on the financing side (shrinking workforce) 24 . Offering flexible working practices 25 may play a key role in encouraging more people to enter the workforce (including women, low-qualified people and migrants), and may imply increasing the number of nonstandard workers and expanding the diversity of new forms of employment.

The additional pressures of the recent economic crisis may have further contributed to the increasing use of new arrangements instead of traditional forms of work and contracts. While such new forms of work in relation to the developments outlined above respond to flexibility needs of (public and private) employers and workers, and have supported the integration of people in the labour market who otherwise risked to stay in unemployment, increases in labour flexibility may be associated with higher levels of job insecurity, as reflected in the emergence of unstable and informal jobs, with irregular salaries and reduced social security. Effective welfare policies are thus needed to accompany the changes to avoid larger

inequalities within countries and heightened fragility of employment 26 .

Changes in labour markets institutions can mediate the effects of globalisation and technological progress on work and employment relations 27 . The ongoing rise of new working arrangements poses some challenges to the existing framework of employment legislation and, more broadly, of welfare policy – from minimum wages to benefits and pensions – which

20 Greve (2017): Technology and the Future of Work: The impact on labour markets and welfare states. Edward Elgar Publishing.

21 Eichhorst and Marx (2015): Non-standard employment in post-industrial labour markets: an occupational perspective. Edward Elgar Publishing.

22 While less than 5% of all occupations can be automated entirely through the use of demonstrated technologies, about 60% of all occupations have at least 30% of constituent activities that could be automated. Tasks can sometimes be replaced by smart machines or be split up, making it possible to outsource tasks to freelancers online.

23 Broughton et al. (2010): Flexible forms of work: 'Very atypical' contractual arrangements. European Working Conditions Observatory. 24 See European Commission (2015), The 2015 Ageing Report, European Economy, 3.

25 Di Tella, MacCulloch (2002): The consequences of labour market flexibility: Panel evidence based on survey data, European Economic Review 49.

26 Bluth (2017): Globalisation and the welfare state. Focus paper, Bertelsmann Stiftung.

27 Gonos (1997): The Contest over “Employer” Status in the Postwar U.S.: The Case of Temporary Help Firms. Law & Society Review 31: 81– 110.

remains structured around the concept of standard employment contract. Furthermore, the employment/self-employment binary approach of employment legislation is no longer clear in labour markets. New forms of employment have blurred the boundaries between the two groups creating legal loopholes and leaving many people in the labour market unprotected.

The observed decrease in collective bargaining coverage 28 implies a loss of protection of working conditions for a growing number of workers 29 . Moreover, the instability of

employment relationships, the fear of reprisals and the limited opportunity to join a union could sometimes hamper the "de facto" exercise of collective rights which are often being

defined with reference to an open-ended contract model 30 , affecting especially non-standard

employees. These developments thus imply an additional, complementary lack of protection.

2.2.2. Internal drivers

Alongside the external drivers, there are factors intrinsic to the social security systems which are root causes for the lack of access for non-standard workers and the self-employed. They regard the rules governing formal coverage, effective coverage, transferability and transparency. A summary overview of these drivers is presented in this section; Annex 6 provides more detail and background.

It should be noted that, although principle 12 of the European Pillar of Social Rights refers in general to social protection as an instrument to support fair and well-functioning labour markets and welfare systems, the internal drivers presented here relate only to a subset of social security branches.

A comprehensive mapping 31 of gaps in access for all branches – unemployment benefits, sickness benefits, old age benefits, benefits related to accidents and injuries at work, invalidity benefits, healthcare, family benefits, maternity/paternity benefits and long-term care - has been conducted. The mapping clearly showed that not all social protection areas are affected by labour market evolutions. 32

The mapping showed that the areas where the type of work contract or employment status has a strong influence on the social protection coverage are unemployment benefits, sickness benefits, maternity and equivalent paternity benefits, invalidity benefits, old-age benefits, benefits in respect of accidents at work and occupational diseases; the analysis on these branches will thus be presented in more detail below.

The common element among these branches is that the benefits they provide have a function of income replacement in case the associated risk materialises and interrupts the person's work. For instance, sickness benefits provide income replacement in the case of prolonged illness. While healthcare also involves some cash benefits, these relate to the provision of services rather than to the replacement of income, and it has not been possible to identify a strong impact of a person's labour market status on the corresponding coverage. This branch is thus not included it in the scope of this initiative.

28 See European Commission (2015): Employment and social development in Europe, Publications Office of the European Union, Luxembourg.

29 Koukiadaki et al. (2016): Joint regulation and labour market policy in Europe during the crisis: a seven-country comparison. Joint regulation and labour market policy in Europe during the crisis, Brussels, ETUI, 7-134.

30 Gyulavári et al. (2008): The impact of new forms of labour on industrial relations and the evolution of labour law in the European Union, p. 72, European Parliament.

31 See ESPN (2017).

32 Going beyond the social security branches, for the same reason social assistance, a key part of social protection, is considered out of scope for this initiative, as mentioned in Chapter 1.

Another important point to note concerns the personal scope of the initiative. The initiative supports workers (as defined in Annex 2 33 ) which do not have a full-time open-ended contract, and the self-employed. Both groups show a high level of heterogeneity and the current forms of non-standard work and self-employment are described more in detail in Annex 2 and Annex 4. By virtue of the residual definition of non-standard workers given above, all current forms of non-standard employment, as well as of self-employment, are included in the scope of the initiative, without further sub-classifications. This also allows the initiative to be future-proof, since labour markets are evolving fast and technologies are giving rise to many different new forms of employment not yet known.

Formal coverage

Formal coverage means that a person in employment or self-employment is entitled to participate in social security systems. It is a prerequisite for benefitting from the protection granted by the scheme in case of need. Limited formal coverage means that the rules governing access to social security schemes do not allow individuals from specific groups to become members. Thus individuals are neither covered on a mandatory basis nor can they opt to join corresponding schemes on a voluntary basis.

Overall, people in non-standard employment usually have the same formal coverage by social benefit schemes as those in standard employment. However, in certain Member States, there are important exceptions relating to specific categories of non-standard workers, notably casual and seasonal workers, apprentices or trainees, on-demand workers, and those on temporary agency contracts as well as people working on certain contracts defined at national level, such as civil law contracts in Poland (more than half a million people) or mini-jobs in Germany (7.6 million) (see Table A.2 in Annex 6). Such coverage gaps are wide-spread geographically, they have been identified in fifteen Member States for unemployment benefits, twelve Member States for sickness benefits, twelve Member States for maternity benefits, seven Member States for accident at works and occupational injuries and thirteen Member States for old-age and survivors' benefits.

The situation is more critical for the self-employed. There are fundamental gaps in formal coverage regarding three core branches, where the self-employed as a group are excluded from membership. In particular, the self-employed, or important subgroups, do not have any access to unemployment benefits in eleven Member States (see Table A.3 in Annex 6). They cannot be insured against accidents at work and occupational injury through social security schemes in ten Member States and they are not covered by sickness benefits in three.

In other countries, the self-employed, or at least some categories, have formal coverage on a voluntary basis for unemployment insurance, sickness benefits and accidents at work in eleven, seventeen, nine countries, respectively. It is worth noting however that rates of enrolment in voluntary schemes are low (less than 1% to 20 %) in most Member States where this possibility exists. The low adherence of self-employed may be related to alternative individual coping strategies, affordability and financial disincentives, to unfavourable system design, lack of awareness or individual myopic behaviour. The latter may affect the individual decisions on intertemporal income distribution. This may be related to erroneous foresight, or to an unfounded fear that what has been promised for tomorrow might not materialise.

33 The definition is in line with the one in the recent proposal for a Directive on transparent and predictable working conditions in the EU

(COM(2017) 797 final i, available online). The link of the current initiative with the proposal is further discussed in Chapter 5.

Effective coverage and transferability

Even when non-standard workers and the self-employed have formal coverage, they may experience gaps of effective coverage, that is, rules of social security schemes applicable to non-standard workers and the self-employed, compared to those applicable to standard employees, embed obstacles preventing them from accruing and taking up adequate entitlements in case of need. This may be because the rules are less favourable for the selfemployed or non-standard workers or because identical rules for all three groups are more difficult to meet for people outside standard employment. For instance, given that temporary

employees transition more frequently into unemployment than standard employees 34 , it can be

relatively more difficult for them to accumulate a given qualifying period within the corresponding reference period to qualify for unemployment benefits.

According to the expert mapping 35 , the rules affecting effective coverage of the self-employed and of non-standard employees relate to time thresholds for contributions and entitlements (minimum qualifying period, minimum working period, waiting period or duration of benefits), to the way the income base is assessed for the self-employed or to the level of adequacy of the benefits (Annex 6). Overall, obstacles to effective coverage have been identified for at least nineteen Member States and in relation to six social security branches, notably unemployment, sickness, maternity, accidents at work and occupational injuries, invalidity and old age and survivors' benefits. Minimum qualifying periods and are particularly problematic for non-standard workers, while the self-employed mostly face problems related to longer waiting periods and shorter duration of benefits.

Regarding non-standard workers, minimum qualifying periods and minimum working periods in unemployment benefit schemes have been identified as problematic in six and four Member States respectively. Minimum qualifying periods are also challenging in the case of sickness benefits in five Member States. Waiting periods are unfavourable in six Member States in the case of maternity benefits (see Table A.6 in Annex 6).

Regarding the self-employed, they face longer qualifying periods for unemployment benefits in four Member States and are subject to longer waiting periods before receiving sickness benefits in eight Member States (see Table A.7 in Annex 6). The methods for calculating the income base or the reference income may also be unfavourable to them because they consider long previous periods of earnings or on future earnings instead of current earnings, not taking into account the fluctuations characterising incomes from self-employment. Moreover, in some Member States many self-employed may receive lower benefits compared to an employee with the same earnings because benefits are set at different levels according to the form of employment (e.g. pensions for the self-employed in Italy or sickness benefits for the self-employed in Belgium).

Transferability may refer to the possibility of transferring accumulated entitlements to another scheme following professional transitions between sectors or employment forms. 36 In this case a lack of transferability can hinder the building up of adequate entitlements over the course of a career characterised by frequent transitions as in the case of old age pensions.

34 See European Commission (2017): Employment and Social Developments in Europe review, Luxembourg: Publication Office of the

European Commission.

35 See ESPN (2017).

36 This implies that entitlements can only be transferred when the individual has formal coverage (and does choose to be enrolled in the case of voluntary coverage) before and after the transition.

Transferability can also refer to making qualifying periods in the previous employment status (or in concomitant employment statuses) count towards the qualifying periods in the new status. This is because the entitlement to benefits in the new scheme depends on contribution history but the entitlements in the old scheme cannot be transferred. In this second case, gaps in transferability may impede individuals for example of transitioning between employment and self-employment when these two statuses are covered by different schemes (see Table A.8 in Annex 6). In view of careers that are increasingly characterised by frequent changes between jobs and employment statuses, this regulatory failure is of rising importance. Obstacles to transferability can be due to lack of regulation in this area (four Member States), to extremely high administrative costs (one Member State) or to different rules governing different schemes (two Member States) which prevent people from transferring accumulated rights.

In the case of voluntary schemes, the transfer of entitlements may be influenced by behaviour of individuals. A number of factors can drive the decision to adhere or not to a voluntary scheme, including whether coverage is already available through the first job, the family context and private wealth accumulation which provide alternative means of coping with risks, as well as the individual's assessment of future risks facing them.

In some Member States the time profile of replacement rates in unemployment insurance depends on the contribution history and shorter contribution histories are likely to characterise workers on temporary contracts as compared to full-time workers. Here again behaviour of individuals matter. Particularly in the case of the self-employed, where the rules for declaration of income are less stringent, or where the individual has the option to (legally) choose what level of income to be assessed upon, they may report low incomes, leading to low entitlements. The low declaration of income due to myopic behaviour has been identified as an obstacle to the adequate protection of the self-employed in at least half a dozen of

Member States 37 .

Transparency

Finally, there are gaps in the provision of transparent information to the general public, potential scheme members, schemes members and beneficiaries about the scheme's rules or about the individual obligations and entitlements. Transparency of information is ever more important due to the growing numbers of transitions between different employment status, contracts forms and social security schemes. A lack of simple and transparent information lowers the level of people's awareness on their rights and obligations towards social security and stops them from taking informed decisions. Indeed, a significant proportion of the population (see also Annex 6), especially among the self-employed and non-standard workers,

is unaware of their social security coverage (up to 20% for some categories) 38 . If generic

information about social security schemes is only absent in five Member States, personalised information, including an overview of rights and obligations and the possibility to obtain information through online simulation tools, is only available in about half of the Member States (see Table A.10 in Annex 6).

Despite the general acknowledgement of the challenges at hand, corroborated by the majority of stakeholders during the consultations organised in the framework of this Impact Assessment, national reform efforts are partial and uneven across the EU Member States. In

37 In Spain, 86 % of self-employed choose the minimum contribution base (see Case study on economically dependent self-employed (TRADE) and box A.6.

38 OpenEvidence (forthcoming).

recent years, some Member States started reforms to address some of the deficiencies related to social security coverage for non-standard worker and for the self-employed (see Annex 7 for a full overview of the national reforms). These reforms have mostly concerned the extension of social security coverage to people in non-standard employment and in selfemployment, in particular concerning unemployment benefits coverage (e.g. France, Lithuania, Poland, Portugal and Sweden). Only a few reforms have addressed the problems of effective coverage (e.g. Spain), transferability (e.g. France and Ireland) and transparency. Moreover, although a number of countries have developed specific guidance and start-up programmes for the entrepreneurial self-employed, this has rarely also led to a redesign of social security taking into account their special needs.

The results of the implementation of already undertaken national reforms can only be evaluated in some years. However, the reforms seem partial in their scope (both regarding the branches of social security and groups of people in the labour market) and the efforts appear to be uneven across the Member States. The identified gaps in formal and effective coverage and the lack of transparency and transferability of accumulated rights and together with behavioural biases (and their influence on the voluntary take-up of social security schemes), therefore leave a large number of people in the labour market uncovered by social security.

2.3. Consequences

With an increasing number of people outside the scope of the social security systems, their ability to insure people against the financial implication of social risks, to prevent poverty, to support labour market participation and act as an automatic macroeconomic stabiliser is significantly reduced. Overall, a differential coverage of social security, based on contract and

status of people in employment, has previously been criticised and documented 39 as both

unfair and inefficient. It has consequences on those directly affected by the gaps but also on companies, on the economy and on society as a whole.

2.3.1. Consequences for non-standard workers and the self-employed

Insufficient coverage of social security significantly raises the risks to the welfare of the affected and their families. People who earn their living from non-standard employment or self-employment and do not have sufficient social security coverage are exposed to greater economic uncertainty due to the individualisation of risk. At the same time, some of these groups may be less able to deal with these risks on an individual basis. Together, this may lead to more precarious lives. Indeed non-standard workers and the self-employed are exposed to significantly higher poverty risk in most EU countries.

Non-standard employees' rate of being at-risk-of-poverty (AROP) is higher in all Member States (except in Malta) than the one for workers with a permanent and full-time job. On average in the EU28, around 15% of temporary and part-time workers are at risk of poverty, more than double the share observed for standard employees.

Figure 4: In-work at-risk-of-poverty rate employees according to their status, by countries, 2015, %

39 Matsaganis et al. (2016) and ESPN (2017). See European Economic and Social Committee (EESC) (2010): New trends in self-employed work: the specific case of economically dependent self-employed work, available online . EESC (2013): Abuse of the status of self-employed, available online . EESC (2016): The changing nature of employment relationships, available online . European Parliament (2014): Social protection for all, including self-employed workers, available online .

Source: Eurostat, Employees with a temporary job; Employees at part-time.

Note: countries are ranked according to the growing poverty risk rate of workers with permanent jobs.

The significant variation that exists across Member States (see figure 5) reflects differences in both protection through the tax-benefit system and in income. Indeed, non-standard workers are estimated to have lower wages with respect to standard employees. For instance, it has been estimated that, controlling for workers characteristics, there remains a significant wage gap disadvantaging temporary employees relative to standard employees of 13% to 21%

depending on the Member State 40 . The lack of social security coverage can thus compound

disadvantages stemming from lower levels of income and unfavourable social security rules.

Similarly, the self-employed face a higher risk of poverty 41 than standard employees; their AROP exceeds 20% in the EU28, nearly three times the risk of standard employees. As reflected in Figure 5, observed ratios are much more pronounced in some Member States, with a maximum of 10 reached in Romania. Moreover, there is a negative correlation (- 0.6), between the at-risk-of-poverty of the self-employed and that of the employees 42 . This means that, in general, in Member States where employees are relatively well-protected against the risk of poverty, this is far less so for the self-employed. Where also the employees face a high risk, differences between both groups are less marked. When analysing the role of social

security in alleviating poverty of both groups, it is noticeable that (see Figure 6 43 ) social

security diminishes the poverty rate of both people in employment and self-employment in the majority of the Member States. However, across most Member States the reduction is visibly more pronounced for employees than for people in self-employment, which partly reflects the

identified gaps in access for the latter group 44 .

Figure 5: In-work at-risk-of-poverty rate Figure 6: Percentage change in at-risk-ofemployees vs. self-employed, by poverty rate, from pre-benefit to disposable countries, 2015, % (%) income

40 See European Commission (2017): Labour Market and Wage Developments in Europe Annual Review. The gap is estimated to be more pronounced in Member States with a high share of temporary workers.

41 Measuring poverty among the self-employed is complex. While income poverty levels are quite high among the self-employed, material deprivation rates are generally much lower and this difference is higher for the self-employed than for the employees. This can be due to the fact that the self-employed can more often draw on assets accumulated over the life cycle or on business assets they control. The selfemployed constitute a very mixed segment of the workforce and within-group inequality is quite significant. Own-account workers emerge as being particularly at-risk of poverty substantiating worries about the rise of this form of self-employment.

42 See ESPN (2017), fig. 6, p. 25.

43 The figure shows the difference between at risk of poverty rates calculated on the basis of market income adjusted for taxes, but not yet taking into account benefits, versus those calculated on the basis of disposable income (post-benefits).

44 In addition to differences in the effect of social security benefits, results may inter alia also reflect differences in household composition between both groups.

Source: Eurostat, EU SILC data for 2016 Source: Eurostat, EU SILC data for 2016.

The gaps in coverage of non-standard employment contribute to a lack of fairness within

society, widening divides between generation, gender or social groups, since young people,

the foreign-born and women are less frequently hired on full-time, open-ended contracts. The

differences in coverage may also impede labour market mobility of these groups, which

makes it even more difficult to close gaps between groups in the long-term.

2.3.2. Consequences for businesses and competitiveness

Differences in treatment related to social security create bias in the level playing field relating

to companies’ hiring choices: firms may overuse employment contracts exempted from social

security contribution. They transfer the risks and the costs of protection through social

security systems to third parties (workers, dependant/bogus self-employed and/or public

budget). Such practices contribute to maximise profits in the short-term, but they also have

costs in the long run: they demotivate workers, lead to higher turnover, have reputational

effects, and slowdown productivity gains and innovation at the company level. They also

create unfair competition with companies that continue to contribute to social security

systems and they threaten to erode the economic basis for standard employment as such jobs

will be subject to competition from jobs exempted from the cost of social contributions.

During discussions on globalisation, business and political circles have expressed concerns

that high employer contributions to social security make national economies less competitive.

However, several studies have found a positive correlation between the openness and size of

the welfare state, and social expenditure and competitiveness 45 . In particular they found that

social spending increases competitiveness and show that reverse causality (i.e. that higher

competitiveness leads to larger welfare states) is weak. They also find that inward FDI flows

in OECD countries are positively affected by countries’ expenditure on social welfare.

2.3.3. Consequences for the labour market, public budgets, the economy and society

From an employment perspective, deficiencies in social security coverage increase labour

market segmentation 46 , which can be associated with higher levels of unemployment (mainly

45 Epifani and Gancia (2009): Openness, Government Size and the Terms of Trade, Review of Economic Studies, 2009, vol. 76, issue 2, 629-

668.

46 Labour market segmentation indicate the persistent divide in job security and working conditions between workers with similar characteristics but holding different types of contracts.

for vulnerable groups) and lower quality of skills matching 47 . Moreover, the absence of social security make it less attractive for people to take non-standard jobs or move between and combine dependent employment and self-employment, preventing labour reallocation and hampering labour market's ability to respond to shocks as well as making segmented labour markets self-perpetuating, ultimately with negative implications on (total factor) productivity

growth 48 .

Moreover, differences in social security coverage distort the playing field between people who pay no or lower social security contributions and those with full, obligatory contributions. It creates legal possibilities to avoid paying contributions and to underbid, leading to a race to the bottom and downward social convergence.

Insufficient protection through social security systems for certain groups may contribute to higher spending for public budgets in other areas. In the short term an increasing share of people in non-standard work or self-employment without social security coverage does not necessarily save public money and may even lead to rising costs for (generally tax-financed) safety nets of last resort that they may then rely upon. This is because, on the one hand, even if not paying social contributions and receiving social benefits, they are entitled to social assistance, minimum income or universal minimum healthcare benefit package. Increased reliance on the latter can have an additional efficiency cost including through delays in care seeking and overuse of costly emergency services instead of primary or integrated care services.

In the longer run, the absence of benefits to facilitate social and economic inclusion may make long-term unemployment and mental health problems more likely 49 . Therefore, if the longer-term welfare of people in non-standard work and self-employment is not secured, this will imply a social cost to be borne by these individuals and by the society, with higher costs to public budgets 50 .

Importantly, the main fiscal basis of social insurance remains, in many countries, standard work. An increasing share of people outside the systems may thus impair the fiscal sustainability of social security systems, an effect comparable in nature to the corresponding detrimental effects of population ageing and resulting increases in the dependency ratio 51 on their sustainability 52 . Annex 3 provides further information on the financing of social security

systems and the challenges to their sustainability relating to the problem under review.

47 Brodolini (forthcoming).

48 OECD (2012): OECD Employment Outlook 2012, OECD Publishing, Paris.

49 Eurofound 2013: Access to social benefits: reducing non-take-up. Moreover, unemployment insurance mitigated adverse health effects during the financial crisis. Unemployment benefits may provide alleviation for people in vulnerable situations, allowing them access to basic services, with the potential to prevent or reduce escalation of problems (Ferrarini et al (2014): Decomposing the effect of social policies on population health and inequalities: an empirical example of unemployment benefits. Scand J Public Health. 2014 Nov; 42(7):635-42.).

50 Increasing the coverage of non-standard workers could have a positive net effect. E.g. the mini-job and midi-regulations engender fiscal costs (social security contributions and taxes) ranging between EUR 600 and 800 million per year (see Case study – gaps in access to social protection for mini-jobbers in Germany) and the estimated loss of the public social insurance systems in Romania due to the replacing of regular employment contracts by day-labour is around 1,824 million RON in 2017 (see Case study – gaps in access to social protection for casual workers in Romania).

51 Sachverständigenrat zur Begutachtung der gesamtwirtschaftlichen Entwicklung (2006), Jahresgutachten 2006/2007, available online , p. 262.

52 As the new forms of work may push down the overall wage level (see European Commission (2017): Labour Market and Wage Developments), the fiscal basis for insurance could be further eroded.

• 3. 

  W HY SHOULD THE EU ACT ?

3.1. Legal basis

The initiative "Access to social protection for workers and the self-employed" will support the Union's aims recognised in Art. 3 of the Treaty on the EU: to promote the well-being of its peoples, the sustainable development of Europe aiming at full employment and social progress, but also the aim to promote social justice and protection, equality between women and men and solidarity between generations.

Moreover, it will respond to Art. 9 TFEU, which states that "In defining and implementing its policies and activities, the Union shall take into account requirements linked to the promotion of a high level of employment, the guarantee of adequate social protection, the fight against social exclusion, and a high level of education, training and protection of human health". In addition, Art. 151(1) TFEU states that the Union and the Member States shall have as their objectives the promotion of "proper social protection".

Furthermore, the initiative will promote the aims of the Charter of the Fundamental Rights of the European Union as reflected in its dedicated article on Social security and social assistance (Art. 34).

Article 153(1)(c) and (2) TFEU enables the EU legislator to adopt measures in the field of social security and social protection of workers as long as it (i) avoids imposing administrative, financial and legal constraints in a way which would hold back the creation and development of small and medium sized undertakings, ii) does not affect the right of Member States to define fundamental principles of their social security systems; (iii) does not significantly affect the financial equilibrium of these systems. This legal basis can be used to address the challenges of access to social security and social protection for persons in nonstandard employment. The Council is requested to act in unanimity.

Article 352 TFEU can provide the legal basis of an initiative related to the challenges of access to social security for people in different forms of self-employment. In this case the Council is also requested to act in unanimity.

The existing EU acquis related to the rights and obligations in relation to social protection is implemented through a variety of legislative and non-legislative provisions. However, none of these fully address the rights and obligations of people in non-standard employment and forms of self-employment. For an overview of available EU instruments, see Annex 8.

3.2. Results of the social partners consultations and of the public consultation

The European Pillar of Social Rights was jointly proclaimed by the European Parliament, the Council and the Commission on 17 November 2017. The consultations undertaken in the context of the Social Pillar had already seen many stakeholders express concerns regarding the obstacles faced by workers in non-standard forms of employment and by self-employed in accessing social protection such as unemployment benefits, sickness benefits, disability benefits, health insurance and pensions. Relevant contributions from other EU institutions, notably the European Economic and Social Committee, the Committee of Regions and European Parliament to the Consultation also highlighted issues in this area and, in the case of the European Parliament, called on the EU to act. 53

Article 155 TFEU requires that labour and management are consulted on a range of issues concerning employment and social affairs set out in Article 153 of the Treaty, including the area of social security and social protection. Consultation of representative social partner organisations took place prior to the presentation of the proposal regarding access to social security of non-standard workers in accordance with Article 154 TFEU. The procedure of Article 154(2) TFEU is not as such applicable to Union action to address the challenges related to the self-employed on the basis of Article 352 TFEU. It was therefore on a voluntary basis that the Commission invited the social partners to share their views as regards the selfemployed.

There was no agreement among social partners to enter into direct negotiations on any of the issues raised during the consultations While both sides broadly agreed with the challenges presented by the Commission during the consultations, they differ on the opportunity for EU action. On the side of labour, trade unions expressed support for a legislative action setting minimum standards at EU level. On the side of management, representatives of employers have underlined that the action should be left to Member States, given their responsibility on the design of social security systems. During the second phase of the consultation, all social partners agreed on higher transparency. Whereas trade unions call for full transferability, employers would like to limit transferability of minimum social protection rights and according to them there should not be an additional burden or cost of transferability.

All trade unions favour mandatory formal coverage, equalised to those of standard workers but national traditions should be taken into account. Employers referred to subsidiarity and expressed different views on the options. Some agree that social protection should be mandatory but employees should have the right to choose the form of coverage (public or private) and the freedom of choice for the self-employed is particularly highlighted. Other employers favour voluntary formal coverage in order to take into account the diversity of forms of employment of the heterogeneity within the self-employed.

The detailed outcomes of the consultation are presented in Annex 11.

In order to further support the preparation of the Commission's initiative, the Commission launched an open public consultation on 20 November until 15 January 2018, available in all languages. Targeted hearings with representatives of the self-employed, social insurance organisations and civil society and the Social Protection Committee (composed by representatives of Member States) were also organised in December 2017 (see also Annex 11 for further details).

3.3. Subsidiarity: Necessity and value-added of EU action

The problem of insufficient access to social security for a growing number of people in the labour market and its negative consequences on social cohesion, labour market dynamism and sustainable growth are wide-spread across Member States.

In principle Member States could take measures to improve the situation and several have indeed adopted reforms or are carrying national debates on the subject. In the Member States

53 See Annex 11 for more details on the position of these institutions as well as on the overall results of relevance of the consultations relating to the European Pillar of Social Rights.

where reforms have been implemented, those reforms are steps in the right direction, but the evidence is that they are not comprehensive and systematic, i.e. they are not closing gaps in access for all groups and all branches of social security. Nor do current debates imply that all Member States necessarily plan to take action. Individual Member States may hesitate to regulate in this area on their own, as they may fear that such action could put their own companies at a disadvantage relative to companies from other Member States.

Action at EU level will support the implementation of principle 12 of the European Pillar of Social Rights so that all people in the labour market, regardless of their employment status, are protected against social risks such as unemployment or illness. Through a careful choice of the level of detail of the provisions, such EU action fully respects subsidiarity, irrespective of the specific legal instrument chosen. EU action can be a springboard for further national reforms and ensure that progress is not partial or uneven but is instead secured across groups, social security branches and Member States, creating a level playing field for economic agents in the internal market. The EU action can avoid short-term distortions of competition and ensure that all Member States move in the same direction at the same time promoting upward convergence, to the benefit the EU economy as a whole. In this respect, it is also worth noting that a special Eurobarometer survey on the Future of Europe shows strongly increasing support for more European level decision-making dealing with health and social

security issues 54 .

EU action can also diminish the obstacles that the problem poses to the pursuit of the Union's objectives. Differences between employment statuses in access to social security can be detrimental the EU's fundamental objective of promoting full employment by hampering occupational labour mobility including towards higher productivity sectors. It can discourage workers from undertaking a certain type of employment or deter people from taking on the risks associated with self-employment, which thus hinder entrepreneurship and innovation. The problem can also be an obstacle to the achievement of the Union's objectives in relation to quality of jobs and poverty.

By enhancing convergence towards robust labour market institutions and social infrastructure, EU action can facilitate resilience, social cohesion and macro-economic adjustment within the euro area and beyond. As highlighted in the Five Presidents' Report on Completing Europe's Economic and Monetary Union 55 , in a single currency area, there is a need to build up the shock absorption capacity and labour market adjustment mechanisms of Member States.

• 4. 

  O BJECTIVES : W HAT IS TO BE ACHIEVED ?

4.1. General objectives

The general objective of the initiative "Access to social protection for workers and the selfemployed" is to support all people who, due to their employment statuses, are not sufficiently covered by social security schemes and thus are exposed to higher economic uncertainty.

The initiative is a concrete way to fulfil the EU's objective to promote social justice and protection, as stated in Article 3 of the Treaty on the European Union. It further aims at implementing principle 12 of the European Pillar of Social Rights that states that 'regardless of the type and duration of their employment relationship, workers, and, under comparable

54 Special Eurobarometer 451: The Future of Europe. Field work: October 2016. Publication: December 2016, available online .

55 "Completing Europe's Economic and Monetary Union", Report by Jean-Claude Juncker, in cooperation with Donald Tusk, Jeroen Dijsselbloem, Mario Draghi and Martin Schulz, June 2015

conditions, the self-employed have the right to adequate social protection'. It also contributes to the implementation of other Pillar principles such as those relating to "secure and adaptable employment", "unemployment benefits" and "old-age income and pensions".

It will support Member States in adapting social security to present and future changes in labour markets. It will contribute to improving labour market dynamism by reducing labour market asymmetries in social security obligations and rights for different forms of employment. It will facilitate transitions between jobs, sectors and employment status by ensuring better preservation and transferability of acquired rights.

Its ambition is to foster sustainable growth as well as increase fairness, competitiveness and social justice by improving adequate access to social security for those without it, including many categories of people in non-standard employment and the self-employed, thereby addressing long-standing and new gaps and inequities in access. At the same time, ensuring that more people in the labour market contribute to the financing of social security systems would enhance their financial sustainability, contributing to intergenerational fairness and promoting social cohesion in the longer run.

4.2. Specific objectives

In order to reach the previous general objective, the initiative will have the following specific objectives:

• - 

  Closing formal coverage gaps: ensuring that legislation and collective agreement allow everyone in employment or self-employment to participate in the social security schemes.

• - 

  Addressing the problem of effective and adequate coverage and facilitating transferability of social security rights: ensuring that rules governing contributions and entitlements ensure the adequacy of protection and do not limit the accrual of entitlements and the access to benefits on the basis of employment status or type of employment relationship and do not distort the playing field between employment statuses or type of employment relationship. The initiative aims at preserving social security rights when workers or the self-employed change jobs, sectors of activity, type of employment relationship, or employment status.

• - 

  Increasing transparency on corresponding systems and rights: ensuring access to user-friendly information on rights and obligations to social security, irrespective of the employment relationship or employment status.

Formal coverage is a necessary, but not a sufficient condition to ensure adequate protection through social security systems. A person may be a member of a social security scheme without de facto being able to build and take up adequate entitlements to benefits and services, notably if the rules governing contributions and entitlements disadvantage some types of employment, employment statuses or career history. The rules should, on the contrary, promote social justice and a level playing field between employment statuses and types of employment relationship. Any difference between employment statuses or types of employment relationship regarding the rules governing entitlements should reflect the situation of beneficiaries and be justified by policy objectives. This objective concerns in particular the rules governing time thresholds, i.e. minimum qualifying period, minimum working period, waiting period and the duration of benefits. It also concerns the rules determining the assessment of the income assessment base.

If a person changes employment status, type of employment or economic sectors or if a person combines different jobs, the transferability and preservation of rights between the scheme available under the new status and the previous scheme covering the same risk should be facilitated. The preservation of rights should also be facilitated for persons combining different jobs and/or employment status at the same time. This is key in careers that are characterised by several job contracts and transitions. If rights are transferable, people will be more confident to make transitions, the labour market can be more dynamic and the allocation of resources more efficient.

Even when they have formal and de facto coverage, people may not benefit from social security if they do not have sufficient knowledge about their rights. If the rules governing social security schemes are transparent and if information about individual entitlements and obligations are easily accessible, individuals will be in a better positon to exercise their rights to social security.

Although these objectives are valid across all branches of social security, the issues have been found to be more acute in the branches that are more closely related to employment status, (unemployment benefits; sickness benefits; maternity and equivalent paternity benefits; invalidity benefits; old-age benefits; benefits in respect of accidents at work and occupational diseases).

• 5. 

  W HAT ARE THE AVAILABLE POLICY M EASURES ?

5.1. Baseline scenario

Under the baseline scenario the existing coverage gaps in social security would continue to exist across Member States and can be expected to lead to insufficient social security coverage for an increasing number of citizens.

On the one hand, external drivers like globalisation, transition towards a service economy, technological developments, demographic ageing outlined in section 2.2.1 are expected to persist. They will continue to rapidly change the European labour markets and make employment opportunities steadily more diverse. This implies that overall, employment on both old and new forms of non-standard contracts and in new forms of self-employment will likely constitute a steadily rising share of job-opportunities in the future. If observed recent trends continue, the share of traditional non-standard work in employment could rise from 25% in 2016 to 29% by 2030, while self-employed would remain broadly constant at 4 % (14% in 2016) 56 . For many individuals, most notably the young and the high-skilled, new forms of employment or own-account self-employment will represent an opportunity to be integrated in the labour market, expand choices and find a better work-life balance.

On the other hand, several Member States have taken steps in the recent past (see also Annex 7) to address some parts of the problem and change the internal drivers but overall progress is uneven across Member States, social security branches and targeted groups. Existing policies and planned reforms appear insufficient to prevent gaps in formal coverage, effective coverage, transferability and transparency in the absence of further EU action.

56 Source: Own calculations based on EU-LFS.

Over time an increasing share of the working population would thus be left without effective coverage under the baseline scenario.

There are a few relevant initiatives at EU level which address parts of the problem, yet a more comprehensive approach is lacking.

The existing EU guidance in the area of access to social protection 57 , particularly in the context of the European Semester and the Social Open Method of Coordination, will continue to provide recommendations for Member States' in this field. These efforts have yielded results in the past 58 by reinforcing cooperation between Member States and helping identify common objectives, but it is less likely that they could trigger the needed accelerated changes to social security systems through comprehensive action in the remaining areas, given their general nature and Member States' reservations that unilateral steps in some areas would entail risks for their competitiveness.

With the European Pillar of Social Rights, Member States and the EU institutions politically committed to strengthen the social acquis and deliver more effective rights to citizens. This political commitment should encourage Member States' actions and some other recent initiatives taken in the context of the Pillar can be expected to contribute only partially to solving the problem (see below e.g. for the Commission's proposal for a Directive on transparent and predictable working conditions in the European Union regarding transparency).

Regarding formal and effective coverage, the Directive 2010/41 i/EU 59 on equal treatment

between men and women engaged in self-employed activity foresees the possibility to grant women (including spouses of the self-employed) the right to be entitled to maternity leave and benefits for at least 14 weeks. Since this Directive leaves to Member States decision if the right to such benefits should be granted on a mandatory or voluntary basis, levels of benefits and eligibility conditions vary widely among Member States. An independent expert evaluation concluded that the Directive was unsuccessful in delivering improved maternity

and social benefits to the self-employed. 60 The Commission's proposal for a Directive on Work-Life Balance for Parents and Carers 61 provides for coverage by paternity and parental benefit and carers' benefit for all workers with an employment contract or relationship, including non-standard workers to whom the definition applies.

Regarding transferability, there is no requirement in EU law to ensure the transferability of occupational pension rights but the Directive 2014/50 i/EU 62 on the portability of supplementary pension rights encourages Member States to improve it and in fact, all Member States have opted to extend the same legal treatment to all workers (the Directive does not apply to self-employed persons) who leave an occupational pension scheme.

57 Council Recommendation of 27 July 1992 on the convergence of social protection objectives and policies (92/442/EEC), OJ L 245, 26.8.1992, p. 49–52, Commission Recommendation of 3 October 2008 on the active inclusion of people excluded from the labour market (C(2008) 5737), OJ L 307, 18.11.2008, p. 11–14, Commission Communication Towards Social Investment for Growth and Cohesion – including implementing the European Social Fund 2014-2020, COM (2013)83 i, A renewed commitment to social Europe: Reinforcing the Open Method of Coordination for Social Protection and Social Inclusion’, COM(2008) 418 final, Council Decision (EU) 2015/1848 of 5 October 2015 on guidelines for the employment policies of the Member States for 2015, OJ L 268, 15.10.2015, Country-specific recommendations .

58 See Annex 7 for a review of reforms undertaken by Member States and Annex 8 for an overview of relevant EU level instruments.

59 Available online .

60 Barnard C. and Blackham A. (2015), The implementation of Directive 2010/41 i on the application of the principle of equal treatment between men and women engaged in an activity in a self-employed capacity, report of the European network of legal experts in the field of gender equality, commissioned by the Directorate-General Justice of the European Commissioned

61 COM(2017)253 i/884003.

62 See Annex 8 for more details.

Moreover, several recent EU initiatives will contribute to addressing the problem of transparency. First, this concerns old-age benefits in the context of occupational pensions. This will be addressed by the Directive on minimum requirements for enhancing worker mobility between Member States by improving the acquisition and preservation of supplementary pension rights 63 (Portability directive) and by the recast Directive on the activities and supervision of institutions for occupational retirement provision (IORP

directive) 64 . Both directives improve the provision of information on the pension schemes.

The Portability directive also covers the protection of pension rights upon termination of

employment 65 . The two directives should be transposed into national law by 21 May 2018 and

13 January 2019 respectively.

Second, the lack of transparency for other social security areas for workers will be partially addressed by the Commission's proposal for a Directive on transparent and predictable

working conditions in the European Union 66 , but not for the self-employed. According to the

proposal, workers will receive information on the social security institution(s) receiving the social contributions attached to the employment relationship and any protection relating to social security provided by the employer.

This legislative proposal also allows clarifying the boundaries between workers and selfemployed bringing clarity on where the status of self-employed is unduly used (bogus selfemployment). In relation to access to social security, such clarity helps ensure that those people erroneously considered self-employed can now have access to the available protection as workers (which is generally higher). However the personal and material scope of the two initiatives is different as the proposed directive does not cover the self-employed. Also, the proposed directive does not determine whether workers should have access to social security. It limits itself to secure that information on acquired rights is available to employees.

It could be argued that even without public intervention at EU or Member State level, the setup of effective, efficient and accessible – in terms of prices/premiums - social security systems by social partners or professional organisation may solve the identified access problems. However, it is unlikely that such initiatives would lead to a comprehensive solution. Occupational schemes for pensions are indeed wide-spread in a number of Member States (including BE, DE, DK, FR, IE, NL, SE, SI and the UK), occupational funds may also play an important role in health insurance (e.g. IT) 67 and social partners may be involved in unemployment insurance schemes (e.g. DK, FR). However, there is no evidence of a trend increase in such schemes despite the existing problems. The challenges that have to be overcome are among others the need to mobilise a sufficient number of members to pool risks, while there has been a general decline in most Member States in the organisation of workers in trade unions and in trade union density.

Summing up, it is unlikely that the problem of insufficient access to social security will be comprehensively addressed by Member State action only with the support of the existing EU

63 Directive 2014/50 i/EU of the European Parliament and of the Council of 16 April 2014 on minimum requirements for enhancing worker mobility between Member States by improving the acquisition and preservation of supplementary pension rights. Article 6 requires Member States to ensure that active scheme members can obtain, on request, information on how a termination of employment would affect their supplementary pension rights.

64 Directive (EU) 2016/2341 of the European Parliament and of the Council of 14 December 2016 on the activities and supervision of institutions for occupational retirement provision (IORPs).

65 Article 37 (1,j) of Directive 2016/2341 i on general information on the pension scheme has provision on transferability. In case a member has the right to transfer pension rights, further information about the arrangements relating to such a transfer must be provided..

66 COM(2017) 797 final i, available online

67 ESPN (2017) synthesis report p. 33.

initiatives. As a consequence, it can be expected that the share of people in the EU who do not have access to social security will increase significantly under the baseline scenario.

5.2. Measures discarded at an early stage

When considering the policy measures, a number of approaches have been discarded:

Labour inspections: A non-legislative option to improve formal coverage that was discarded is increasing and improving labour inspections. Doing so could reduce the incentives of companies to replace employees with independent contractors not covered by social security

who are in reality disguised employees to reduce the cost of labour 68 . However, that could

potentially solve only a small part of the problem of lack of formal coverage and not solve the problem of lack of coverage for the majority of the self-employed 69 .

Universal basic income: In the face of rising non-standard employment and self-employment, one popular option could be delinking social insurance from employment by providing a universal social assistance benefit regardless of income, contribution history and employment status. Firstly, the focus of the current analysis is on the effects of employment status or contract type on the available protection through social security schemes. Coverage by social assistance and similar schemes is primarily based on criteria regarding residence and income, and thus they are out of the scope of the current initiative. Secondly, while this option may be feasible for countries that are in the process of setting up their social protection system, it seems to be more difficult to implement in countries that have already a well-established welfare system (as is the case for the EU Member States) 70 due primarily to strongly rooted traditions and preferences, the significant cost of switching from the old to the new model,

and the extra cost associated with providing a comparably adequate protection for all 71 . Such

an option would therefore very likely fail to garner the necessary political support for adoption and/or implementation. Therefore, the problem of insufficient access to social security is best addressed in the framework of the existing social security systems in the Member States.

Individual protection arrangements outside social security schemes: One option mentioned by stakeholders would be to restrict legislative action to demanding that all groups, notably including the self-employed, prove having some form of protection or insurance against the risks covered by social security systems, leaving to them the choice whether to arrange this via social security schemes or through purely individual private arrangements. First, private purely market-based arrangements are not available regarding all the risks under consideration (e.g. regarding unemployment), Second, where they exist, the solidarity manifested in collective schemes could be undermined, including by significant self-selection 72 where the corresponding potentially substantial costs could possibly weaken their financial

68 An increase in the gap between the level of employees and self-employed contributions by one percentage point increases the likelihood of being a solo self-employed by 2.4% (slightly less for women, those in manufacturing electricity and wholesale) according to the Labour Market and Wage Developments in Europe 2017.

69 Eurofound (2017), "Exploring self-employment in the European Union", Publications Office of the European Union, Luxembourg.

70 Digital dividends, World Development Report 2016. Nevertheless, there are still some EU Member States (Bulgaria, Estonia, Hungary, Latvia, Lithuania and Spain) for which addressing shortcomings in the coverage and adequacy of minimum income schemes remains high on the policy agenda according to the 2017 Country-Specific Recommendations (CSRs) (Labour Market and Wage Developments in Europe 2017).

71 Basic income as a policy option: Technical Background Note Illustrating costs and distributional implications for selected countries, OECD, 2017.

72 Whereby they would presumably attract mainly the "bad" risks, whereas individuals with low risk characteristics would be better off buying private insurance.

sustainability. Possible measures and options therefore refer only to social protection concluded through an employment relationship which is coherent with the EU approach used in the Directive 2014/50 i/EU and Directive 2006/54/EC i.

Benefits in kind: While benefits in kind account form an important part of the protection offered through some branches of social security, notably including health care, they are out of the scope of the current initiative. Regarding healthcare, an initial mapping did not provide sufficient evidence of significant differences in access related to labour market status. Similarly for other areas where benefits in kind are of key importance, the labour market situation is not expected to play a key role for access.

5.3. Description of the policy measures

The objective to improve access to social security can be achieved by a combination of measures relating to three areas:

 Closing formal coverage gaps:

 Addressing the problem of effective coverage and facilitating transferability of social security rights

 Increasing transparency on corresponding systems and rights

The first dimension allows people to be member of relevant social security schemes. It is thus a prerequisite for the second dimension. We discuss measures for each of these areas. The policy measures retained are those that provide the alternatives that are most feasible (legally, technically and politically), coherent with other EU instruments, effective, relevant and proportional to tackle the problem drivers detected in section 2. In section 6.1, option packages will be built from the individual measures regarding the defined specific objectives.

As explained in Section 2.2.2, the personal scope (same or different measures to apply for workers and self-employed) considered across policy measures and options will be left general enough to allow the initiative to take into account new forms of work that could arise in the future. In terms of material scope, the measures would apply to the branches of social security which have been identified as most relevant of the Commission's initiative, namely unemployment benefits; sickness benefits; maternity and equivalent paternity benefits; invalidity benefits; old-age benefits as well as benefits in respect of accidents at work and occupational diseases.

The measures aim at eliminating or reducing obstacles which impair social security systems from providing people with adequate protection regardless of their employment status, while respecting Member States’ competences in design their social security systems.

Dimensions of the policy options left to Member States’ discretion in view of subsidiarity and proportionality and in respect of article 153(4) TFEU include i) the level of protection provided, ii) whether to extend coverage under existing schemes or by the creation of new schemes, iii) the way to finance the protection, which includes important design issues such as how to address the split of social contributions between employees and employers in the case of the self-employed and iv) combinations of schemes to be used (public, occupational or private). These issues are beyond the scope of the current initiative. Member States are best placed to decide about these dimensions in ensuring effective protection in the most efficient way given the diversity of existing social security systems across the Member.

5.3.1. Measures to extend formal coverage

Formal coverage can be addressed by granting access to social security schemes to previously excluded groups on a mandatory or voluntary basis. These alternatives may reflect different levels of ambition and may strongly influence the number of additional people that would be covered by social security as a result of this initiative.

• i) 

  mandatory coverage: extending coverage to groups of people previously excluded by virtue of their employment status or contract type 73 but also reviewing minimum income levels below which individuals in some Member States currently do not have access or only have access on a voluntary basis regarding certain branches of social security.

• ii) 

  voluntary coverage: previously uncovered groups would have the possibility to have join social security schemes, but the final decision on whether to use this option would be left to the individual.

The approach to coverage could be the same (either mandatory or voluntary) for all groups without access to social security (i.e. non-standard workers and self-employed) or could be differentiated between non-standard workers (mandatory coverage) and self-employed (voluntary) to reflect the differences of both groups regarding the allocation of risks, discretion over their work, decision-making power, financial autonomy and responsibility.

Combinations of the two alternatives could also be considered. Decisions on coverage could be linked to decisions on the level of protection (which is to be decided by Member States). In particular, a basic level of protection can be mandated and supplemented with additional coverage on a voluntary basis 74 . Such calibration of the mandatory element can also be motivated by the observation that uncovered individuals in most Member States do have access to non-contributory safety nets of last resort. Mandating a basic level of coverage close to the one provided by basic safety nets can thus also contribute to avoiding free riding behaviour of uncovered individuals and the corresponding costs to public budgets. Decisions on additional coverage, which are not associated with the same possible externality, are left to the individual 75 .

In their replies to the second phase social partners' consultation (see also Annex 11), all trade unions favour mandatory formal coverage, equalised to those of standard workers but national traditions should be taken into account. Employers expressed different views on these options. Some agree that social protection should be mandatory but employees should have the right to choose the form of coverage (public or private) and they highlighted in particular the freedom of choice for the self-employed. Others favour voluntary formal coverage in order to take into account the diversity of forms of employment of the heterogeneity within the self-employed. Many employers referred to subsidiarity that should prevail.

Regarding the open public consultation, around 2/3 of the respondents consider that social protection rights and obligations should be mandatory for every kind of job, irrespective of

73 It is proposed to exclude the following groups :Family workers, for whom granting access in terms of obligations and benefits seems unfeasible in view of the unpaid nature of their work, and, farmers.

74 This is the case for instance in the two-tiered unemployment insurance scheme in Sweden. A universal flat-rate benefit is provided by the state. Coverage by state-subsidised earnings-related compensation going beyond this minimum level is optional.

75 According to the ILO 75 , voluntary coverage may be considered a suitable supplement to mandatory schemes in ensuring effective protection and the global social efficiency of national systems only when the following conditions are fulfilled: i) they are controlled by the State or managed by representatives of the social partners; ii) also applicable to low-income workers and iii) meeting to the general requirements of compulsory social insurance schemes (inter alia collective financing, periodicity and predictability of benefits, legal protection, financial guarantees etc.).

the type of contract. Only a small share of respondents think that social protection rights and obligations should be voluntary (5.9 % for non-standard work and 10.1 % for self-employed).

5.3.2. Measures to promote effective coverage and transferability

Measures to address problems of effective coverage relate to the rules governing time

thresholds (minimum qualifying periods, the minimum working periods, waiting periods 76 )

and to rules governing transferability. The latter concerns the two cases outlined earlier in Section 2.2.2: i) the build-up of entitlements over the course of a career characterised by transitions; ii) the preservation and transfer of entitlements from one scheme to a new one where there are minimum time thresholds to be respected.

During the second phase social partners consultation, views on effective coverage and transferability were mixed. Trade unions called for guaranteeing full portability by tying social protection entitlements to individuals through an equalised calculation and aggregation but without leading to an individualisation of social protection. On the other hand, some employers agree with the principle of transferability of rights but call for an impact assessment before creating legally binding realities or argue that subsidiarity prevails and that a mapping of existing transferable systems should be favoured. Others recommend the EU to limit itself to general provisions or would like to limit transferability to minimum social protection rights. The administrative cost of transferability should be kept at a minimum and no additional costs for the new employer to take on the provision of the transferred rights should occur.

Rules governing time thresholds

In schemes where minimum qualifying periods, minimum working periods, or waiting periods exist and are stricter for non-standard workers and the self-employed than for standard employees, or are the same as for standard employees, it can be more difficult for the two groups to fulfil them. Policy measures can adjust the minimum periods, so they are not longer than justified by policy objectives, including avoiding abuse. 77

Time thresholds make the conditions for take-up of benefits more stringent and are used to minimise abuse, protecting the interest of the members of the scheme. The total elimination of time thresholds is therefore unrealistic because it would eliminate all disincentives to abuse embedded in the system. This hypothesis is therefore disregarded in our analysis below.

Modifications regarding qualifying periods, minimum working periods and waiting periods could be made for all insured people. Alternatively, they could be adjusted only for nonstandard workers and/or the self-employed in the specific cases where the time thresholds have been identified as obstacles to reach the same social protection outcomes for everyone.

76 For definitions refer to Annex 2.

77 This qualification is important as these features may play an important role for ensuring the incentive compatibility of social security systems, see Box A.4 in Annex 6 for a discussion.

Rules governing transferability

• i) 

  Ensuring preservation, transferability and aggregation

Preservation of entitlements can support the build-up of entitlements over the course of a career where transferring is not possible (because for instance the person will not have

coverage by another scheme after changing job or sector or employment status 78 ). The

method of preservation may vary depending on the nature of the scheme. For instance, the value of pension rights may be adjusted in accordance with: inflation rate or salary levels (typically in a defined benefit scheme); return from investments derived by the scheme (typically in a defined contribution scheme). Preservation of entitlements can also mean

that individuals also have the possibility of continue contributing to the former scheme 79 .

In cases where transferability is possible, then mandatory transferability ensures that all workers or self-employed leaving a scheme have the possibility to take their acquired rights with them. Conditions should also be set so that this transfer does not reduce substantially the entitlements (for instance due to unfavourable calculation of the transfer

value or high administrative costs).

Aggregation of contributions period records within a given social security branch can support smooth transitions in the labour market. This implies that the past periods of employment or self-employment prior to changing employment status are taken into

account as if the periods were completed in the current status 80 . Such aggregation should

be made possible up to a limit justified by policy objectives, bearing in mind the need to preserve incentives for the functioning of the system.

• ii) 

  Personal social security accounts

This refers to disconnecting rights to protection through social security systems from professional status and connecting them to individuals, thereby providing better protection for people who change employers, jobs, sectors, employment status and even countries. Personal accounts can be set on the basis of existing schemes with a high coordination at national level among providers of all existing schemes or can be triggered through a fundamental revision of the social security system, implying a major reform effort to scale down the number of schemes related to specific employment statuses and occupational categories and to 'homogenize' their rules. Currently, only Ireland and Latvia have implemented personal accounts, while France is working to introduce them in its system 81 .

78 For instance, in the area of occupational pensions Member States are encouraged to extend the standards of Directive 2014/50 i/EU to selfemployed people who change jobs within the country: When leaving a pension scheme, a person is entitled to keep his vested pension rights in the scheme, unless he agrees to receive them as a capital payment. The pension rights of the former member must be preserved fairly compared to the rights of current members.

79 This is for instance the case in the Netherlands regarding occupational old-age pensions for former employees that become self-employed 80 For instance, in Luxemburg, the self-employed must have completed at least two years of compulsory insurance as a self-employed person

to qualify for unemployment benefits. However for the purpose of calculating this qualifying period, periods of insurance completed as an employed person can be aggregated provided the person has carried out activities as a self-employed person for at least six months before the submission of the request for compensation. See: ESPN (2017).

81 The success and the aim of those personal accounts differ among those countries: in Latvia there are problems in terms of redistribution, which is let at the minimum while in Ireland, major critiques have been in terms of privacy violation.

5.3.3. Increasing transparency on social security rights and obligations

During the second phase social partners' consultation, all social partners agreed with the need for higher transparency but they did not provide specific suggestions how this could be achieved.

Regarding measures to increase transparency, they could explicitly define what type of information would have to be provided, whether generic and/or personalised. In both cases they are expected to be free-of charge for the beneficiary.

• i) 

  Generic information on rights and obligations (broken down by employment status and

contract type) could be provided, where not already done 82 , for instance via dedicated

webpages relative to each relevant social security branch. The information provided should cover differences regarding the relevant groups, notably as regards standard employees vs. relevant categories of non-standard employees and the self-employed. The information

should be easy to access and understand, and updated on a regular basis.

• ii) 

  Personalised information could be provided to individuals of their rights through dedicated webpages that could offer the possibility to calculate benefits (simulation tools) based on input on personal information such as income and employment situation. Other means to provide such personalised information would be via personal accounts where the personal information such as income and employment is already available and the calculation of benefits does not require any specific knowledge by the individual. Dedicated tools should be easy to find and use.

5.4. Impacts of the policy measures

The assessment of impacts of policy measures 83 covers how the measures are likely to meet

the defined objectives (effectiveness) and the key types of impacts, including economic and social impacts as well as those on national authorities. No environmental impact nor impact on fundamental rights has been identified.

The assessment of impacts combines qualitative arguments with quantitative results to the extent possible. A number of limitations need to be highlighted (see Annex 9 for other methodological considerations).

Firstly, the wide scope of the initiative addressing simultaneously different branches of social security in conjunction with the limitations in data available seriously constrains the possibility to provide a comprehensive picture. Available data only allow quantifications regarding some social security branches, and simulations are limited to the self-employed for the same reason. Furthermore there exists a certain asymmetry regarding the possibility to quantify costs versus – partially non-monetary – benefits. For instance, the positive impacts of reduced insecurity are hard to quantify, and neither do models allow combining microeconomic benefits of increased social security coverage with the more macroeconomic benefits of stronger economic stabilisation over the business cycle or potentially increased labour market transitions and dynamism.

82 See Annex 6 table A.10 and box A.8 for more details.

83 The joint impacts of possible policy packages will be explored in Chapter 6.2.

Secondly, as explained earlier, subsidiarity implies that a number of key policy choices on how to improve access to social security are left to Member States, including decisions on the level of protection embodied in nations' social security systems, financing choices as well as the type of schemes through which coverage is provided. These choices fundamentally determine the qualitative and quantitative impacts that can be expected, impeding a more general or comprehensive illustration in this analysis.

Finally the evolving nature of the problem, as the rapid and fundamental changes in the labour market in the context of the new world of work described above, may also imply changes in the impacts. At the same time, it should also be noted that the choice of (non-binding vs. binding) legal instrument (see Chapter 8) may influence the degree of certainty with which the impacts materialise. The assessment of impacts is thus conditional on action being taken in response to the Initiative and could thus be regarded as an upper bound in this respect in case of a non-binding legal instrument.

5.4.1. Impacts of the measures extending formal coverage

Effectiveness

Mandatory coverage has a very high effectiveness as, by definition, it implies coverage for all previously excluded. Mandatory coverage is a key means to implement the principle of solidarity, enabling also the most vulnerable categories to have access to social security schemes 84 . Moreover, mandatory coverage below a minimum level of benefit entitlement - corresponding to the one available from social safety nets - , can be justified by a need to limit that people would fall back non-contributory social safety nets.

Voluntary coverage can have a good effectiveness. Based on available evidence on enrolment in voluntary systems 85 it can be expected that a significant number of people would choose not to have coverage. At the same time, such an approach would respect the freedom of choice and reflects the heterogeneity of the self-employed, which were highlighted by key stakeholders during the consultation.

Choices on the financing of schemes - in the remit of Member States - can impact the degree of effectiveness, especially when access is granted on a voluntary basis. Indeed the way financing is set up can entail powerful (dis)incentives to participate. There are trade-offs between encouraging higher participation through public subsidies and public finance considerations. Financing also affects the degree of redistribution of income through the social security systems. For a given rate of contribution, incentives to enrol are higher for people who face the largest corresponding risks and are therefore most likely to benefit from the system, i.e. also generate a higher cost for it. Subsidies have successfully been used in some Member States to encourage broader membership and thereby reduce problems of adverse selection; however, such measures may also entail significant costs 86 . Alternatively, Member States could compel providers of social security schemes to offer their members

84 The principle of solidarity justifies the existence of social security schemes in addition to individual protection measures (Gruat (1998): Adequacy and social security principles in pension reform, Ageing working papers, Maintaining Prosperity In An Ageing Society: the OECD study on the policy implications of ageing, Working paper AWP 3.1. Available online ).

85 See Annex 10, Section 1 for a discussion.

86 See e.g. Landais et al (2017): Risk-Based Selection in Unemployment Insurance: Evidence and Implications, CEPR Discussion Paper No. DP12364.

identical obligatory benefits which do not depend on the amount of the contributions and by equalising risks and costs between different funds.

Where individuals are enrolled by default and need to take action to leave the system (through opt-out clauses), usually leads to higher coverage than opt-in measures (see Annex 10, Section 1).

Direct costs and other economic impacts

The direct cost of providing formal coverage depends on the number of people newly covered and the per-capita cost of coverage. The first and likely dominant component is clearly higher under mandatory coverage than under voluntary coverage, as not everyone will choose to be covered under the latter option. Differences in the per capita cost of coverage may arise due to problems of adverse selection, i.e. self-selection, under voluntary access. Resulting budgetary costs could be accompanied by savings from lower expenditure for social assistance and other safety nets that previously uncovered individuals would fall back on.

In addition to self-selection effects, moral hazard 87 could also influence the cost of extending coverage, notably regarding unemployment insurance for the self-employed, who have more control over the relevant risk than employees do in an employment relationship. While this is correct, the successful establishment of social security schemes protecting the self-employed in case of unemployment in the majority of Member States (the mapping in Chapter 2.2.2 only established corresponding formal coverage gaps in nine Member States) implies the potential for learning from peers about the necessary control mechanisms and rules which minimise this kind of behaviour. A recent study of the Danish unemployment insurance for

the self-employed indeed finds a limited risk of moral hazard 88 .

Regarding formal coverage of the self-employed, simulations 89 (see also Annex 10 for all

details) were undertaken for the extension of existing schemes of unemployment benefits for employees to the self-employed under the same rules in five Member States 90 . Results suggest that the cost of extending coverage in terms of the corresponding total expenditure for benefits could be in a range from below 0.1% to 0.4% of GDP annually, depending inter alia on the number of self-employed in the country, the unemployment risk they face, as well as on the generosity of the system in place. Comparing the simulated cost of a similarly generous unemployment insurance system for the self-employed across all Member States results in similar limited costs, albeit with higher cross-country variation.

Expressing the costs as a share of gross labour income of the self-employed, to be interpreted as contribution rates to be paid by them in a revenue-neutral scenario, results in contribution rates between less than 0.1% and up to 1.7% of gross labour income, in line or lower than contribution rates observed in many existing systems in Member States. Simulating costs for

87 This refers to the risk that once covered by insurance for a specific risk, the insurance taker modifies his behaviour in a way that makes the materialisation of the risk more likely, increasing the cost for the insurance provider.

88 The study finds that self-employed men in Denmark who are covered by unemployment insurance are more likely to become unemployed than those who are not covered. But any increased risk of the failure of their businesses that is created by the safety net of insurance is small compared with the overall risk of business failure; the bulk of the difference between the covered and the uncovered group is not related to differences in behaviour but to heterogeneity or sorting. See M. Ejrnæs and S. Hochguertel (2013), Is Business Failure Due to Lack of

Effort? Evidence from a Large Administrative Sample’, Economic Journal. 123, no. 571, pp. 791-830 for details.

89 The simulations do not consider second-round effects resulting from changes in behaviour resulting from the extension of coverage and

should therefore be considered conservative

90 The simulations can thus be regarded as relating to a mandatory approach as all self-employed fulfilling the relevant eligibility conditions are covered; in this sense, the simulations provide an upper bound of the impacts.

extending sickness benefits also results in limited costs in most Member States, with a significant degree of cross-country variation.

At the same time, depending on Member States' financing decisions, extending formal coverage can also avoid that the financing base of social security systems is skewed because an increasing group of people on the labour market is outside the system and does therefore not pay contributions.

Similarly, other choices about the nature of schemes may have implications for their cost. For instance, under voluntary coverage giving a choice between possibly more redistributive public and more individualised occupational schemes may be costly for the public scheme due to self-selection issues. The creation of new, separate schemes specifically designed for those not currently covered can weaken the solidarity ties between the insiders (in the old schemes) and outsiders (in the new schemes).

At the macro level, benefits (not simulated) from the extension of coverage can also be expected by strengthening the well-established key role of automatic stabilisers of public social insurance schemes', dampening business cycle fluctuations 91 . These help limit the loss of economic efficiency resulting from volatility in the economy, contributing to enhanced economic performance to the extent that large output fluctuations can, notably in the absence of wage flexibility, result in a trend increase in unemployment (hysteresis effects) and erode human capital.

A noticeable reduction in unfair competition between firms employing staff on non-standard contracts exempted from social security contributions and competitors not reducing costs in this manner could be expected from the extension of formal coverage. Such effects would be very pronounced under mandatory formal coverage for all groups while a voluntary approach would reduce but not eliminate possibilities for such arbitrage via contract forms.

A differentiated approach extending coverage on a mandatory basis to non-standard employees and on a voluntary basis to the self-employed could incentivise a movement towards the latter. In this respect, it is important to note that the enforcement would be facilitated by the recent proposal for a Directive on transparent and predictable working conditions, which allows to clarify the boundaries between workers and self-employed brings clarity on where the status of self-employed is unduly used (bogus self-employment). In relation to access to social security, such clarity will help ensure that those people erroneously considered self-employed can now have access to social security as workers.

Depending on financing decisions taken by Member States, an increase in labour cost associated with the extension of coverage could lead to a reduction in labour demand for the concerned groups, while work incentives for people paying contribution rates can also be reduced. In conjunction, both effects could imply a reduction in employment.

Finally, positive effects on labour market dynamism of removing the lack of social security coverage as an obstacle to transitions between self-employment, non-standard employment

91 Looking at the recent economic crisis, the importance of automatic stabilizers and in particular of social protection systems in this regard has been highlighted by recent research: "Social transfers, in particular the rather generous systems of unemployment insurance in Europe, play a key role for the stabilization of disposable incomes and explain a large part of the difference in automatic stabilizers between Europe and the US", see M. Dolls, C. Fuest and A. Peichl (2010), Social Protection as an Automatic Stabilizer, IZA Policy Paper No. 18 September 2010, as well as M. Dolls, C. Fuest and A. Peichl (2012), Automatic stabilizers and economic crisis: US vs. Europe. Journal of Public

Economics 96(3), pp. 279.294.

and standard employment could reasonably be expected. Labour market segmentation would be reduced.

Social impacts

Key expected impacts stemming from the improved protection would be an increase in the mutualisation of risk, increase of income security resulting in less precariousness and poverty. Given the higher level of effectiveness of mandatory coverage, this approach is also expected to bring higher social impacts.

The above-mentioned simulations regarding unemployment insurance show that there are noticeable reductions in the risk-of-poverty for the previously uncovered group of selfemployed (for more details, see Annex 10). Coverage reduces the poverty risk among all selfemployed noticeably (by up to 13 pps in the simulations) and softens especially the impact of job loss on poverty at household level (by up to 25 pps for household with former selfemployed unemployed), even if the corresponding effects at the aggregate level are small.

It should be noted that to the extent that women currently are more disadvantaged by social security systems still geared towards linear full-time employment, the improved protection through the extension of mandatory formal coverage but also better effective coverage and transferability for non-standard employees would be especially beneficial for them.

5.4.2. Impacts of measures promoting effective coverage and transferability

Adjusting time thresholds

Effectiveness:

Time thresholds make the conditions for take-up of benefits more stringent thereby limiting inter alia fraud and budgetary costs. Nonetheless the time thresholds can be adjusted to a minimum justified by policy objectives while preserving features in the systems that avoid abuse 92 . This adjustment can significantly reduce the barriers for non-standard workers and the self-employed to the take up of benefits in case of need.

Direct cost and other economic impacts:

This measure would have a direct cost corresponding to the payment of benefits to persons who were previously unable to meet the time requirements. As this relates to persons who were already formally covered by a scheme and thus contributing to it, no increase in revenue would offset these benefit payments. At unchanged contribution rates, it can therefore result in higher costs for the provider of the scheme, notably including an impact on public budgets in the case of statutory schemes. Budgetary costs could be partly offset by the savings resulting from lower outlays on social assistance and from higher consumption-related tax

revenue resulting from the improved effective coverage 93 .

As an example, the effects of relaxing qualifying periods regarding unemployment insurance (for all potential unemployed) have been quantified in dedicated microsimulations (see Annex 10, Section 3). In the simulations, the net costs for the unemployment benefit, going up by

92 See also Box A.4 in Annex 6.

93 See Annex 10 Section 3 for details.

0.13% of GDP. The initial budgetary costs are reduced due to increased tax revenues related to household consumption and positive effect on the welfare of individuals.

As regards other economic impacts, carefully choosing the extent to which time thresholds are adjusted should limit – but would not avoid - further indirect costs potentially resulting from behavioural changes in response to the adjusted requirements. Such costs could notably stem from reduced incentives to work associated with less stringent and more generous unemployment insurance.

Social impacts:

The measure is expected to have social impacts similar to those observed for formal coverage, resulting in a reduction in the individualisation of risk, income uncertainty, precariousness and notably a lower risk of poverty. The above mentioned simulations show a positive impact on disposable household incomes on the lowest quintiles of the income distribution and would significantly decrease the poverty risk for the group of the unemployed (see Annex 10 for details).

Ensuring preservation, transferability and aggregation of entitlements

Effectiveness

Preservation of rights can be expected to be moderately effective since it will at least allow individuals not to lose their accumulated entitlements and to preserve their value but the individual would have to rebuild his entitlements in the new schemes and re-start a contribution history. Here there is an interaction with the time thresholds discussed previously. Regardless of past contributions, the individual would have to wait before taking up benefits in the new scheme

The aggregation of past contribution records would have a good degree of effectiveness in ensuring that people are covered in the short term following transitions.

Mandatory transferability would have a high degree of effectiveness as it would allow individuals not to lose their previous accumulated entitlements when changing job or employment status in a way which entails the transition to a different social security scheme and at the same time enable them to continue contributing to social security. There is an interaction with the measures regarding effective coverage discussed above, to the extent that adjusting time thresholds (e.g. waiting or qualifying periods relating to unemployment or sickness insurance) reduces the need to transfer elements related to the past membership into another scheme. The lower the time thresholds are, the lower are the barriers to transferability across schemes.

The introduction of personal accounts, disconnecting rights to protection through social security systems from professional status and connecting them to individuals, would ensure coverage while reducing the need for transferability (and its related costs) between schemes that are tailored to specific employment statuses. In practice, entitlements would be attached to the individual regardless of the number of labour market transitions and coverage would not be impaired by the need to start a new contribution history at a change of a job for example.

Direct costs and other economic impacts

Preservation of entitlements is not expected to entail particular direct costs.

Aggregation of past contribution periods would impact benefit payouts between schemes in situations where people changing employment status are covered by different and financially independent social security schemes before and after the change. Setting up some protection of the financial interests of the "recipient" schemes would thus be necessary. This would relate to cases where most of the minimum period is accumulated – and the corresponding contributions paid- in the scheme under the old employment status, but the risk materialises soon under the new employment status, with the corresponding costs accruing to the associated "recipient" scheme. This would imply either requesting a certain minimum length of attachment to the new scheme i.e. setting a time threshold 94 , extending the coverage of the schemes by a few months after the person stops contributing or setting up a coordination mechanism for an allocation of contributions and costs across schemes. While the first alternative runs counter to the underlying objective of ensuring coverage also in the initial phase after a change of employment status, the second and third may entail larger costs associated to the administrative complexity.

Additional indirect costs related to the aggregation of past membership periods in other statuses could result from potential behavioural effects. As discussed earlier, minimum periods are in place to align people's behaviour with existing systems and avoid abuse. Policy objectives may thus set a limit to transferring past periods, like shortening minimum periods, to preserve the right incentives to the insured.

Indirect positive effects on the labour market can be expected from mandatory transferability. Transferability can be expected to facilitate labour mobility including between sectors as well as transitions between employment and self-employment. This would contribute to increased labour market dynamism and a better allocation of resources.

The introduction of personal accounts are expected to have the same positive effects on labour market dynamism as an approach securing accumulation transferability and aggregation but it would also entail increase the costs for public authorities and insurance providers as it would imply important changes in the overall coordination of existing social security systems. Estimating the costs of such a system change is not possible with available tools.

For all options measures above, it is important to consider the starting point of Member States. In countries with few or rather homogeneous schemes, the preservation aggregation and transferability of entitlements is likely to entail less administrative costs needed for the coordination between schemes than in countries with a significant number of diverse schemes and a rather complex administrative environment. The more homogeneous the conditions d the schemes are, the faster the transferability of entitlements is likely to be, re-assuring individuals and supporting their dynamism in the labour market.

Social impacts

Key social impacts should be qualitatively similar to the ones regarding measures on effective coverage, notably including reductions in precariousness and the risk of poverty since they will facilitate protection for those who alternate and/or combine different employment and

94 E.g. under regulation 883/2004 i this is working at least 3 months in the context of cross-border mobility.

self-employment statuses. Their importance can be expected to increase to the extent that careers are increasingly characterised by frequent transitions, with positive consequences on employment levels, competitiveness and productivity.

An important negative social impact of connecting the rights to individuals via the introduction of personal accounts could be a reduction in the solidarity and redistribution function of protection provided through social security systems and an associated increase in the individualisation of risks 95 .

5.4.3. Impacts of measures promoting transparency

Effectiveness

The effectiveness of measures would depend on the degree of specificity of the information provided, on the frequency at which it is available and on the fact whether individuals need to take action to obtain the information.

The provision of generic information alone would be moderately effective as it would not reflect the specificities of individuals' situation and action by individuals would be required to access the information. Personalised information is expected to be more effective. The frequency at which this information is provided would affect its effectiveness. If provided via online simulation tools or personal accounts, the information could be accessed at all times. The provision of personalised information could moreover have a strong impact on encouraging individuals to provide for their social protection, especially if the protection is not mandatory. It may thus help achieve the objective of coverage especially in voluntary schemes.

Direct costs and other economic impacts

The provision of generic information is likely to entail low direct costs.

The cost of providing personalised information through dedicated websites and tools would consist of a one-off cost and further costs on an ongoing basis for the necessary updating and maintenance. The costs of providing personalised information by default have been assessed to support the proposal on the activities and supervision of institutions for occupational retirement provision. The one-off cost is expected to amount, on average, to around 7€ per

member 96 .

The costs of personalised information could be reduced by using digital solutions instead of paper. However, this should not be to the detriment of vulnerable groups that might lack access to such solutions and should take into account the needs of people with disabilities.

95 Reported by ESPN experts for the case of Latvia.

96 The cost estimated relates to the burden of providing an annual pension benefit statement to members of schemes. According to the

impact assessment, the total costs primarily consist of a one-off implementation cost in the short term because, once the measure is implemented, the cost of regularly producing the pension benefit statement is likely to be immaterial. This cost is judged to be considerably lower than the expected benefits for the employees over an entire 40-year accumulation period. The impact assessment also noted that employers are likely to pass on the burden to employees, unless social dialogue results in some degree of cost sharing.

Importantly costs incurred in providing transparent information via new technologies are strongly influenced by the departing positions of Member States. In the maturity of IT environments and e-government solutions.

Personalised information is also expected to generate economic benefits as more people would be encouraged to save and the reliance on safety nets is expected to decrease.

Social impacts

An increase in transparency of social security rights and obligations is likely to entail positive social impacts. Awareness on the conditions to take up benefit is likely to increase. The perception of insecurity vis-a-vis the real protection provided by social security (see Annex 16 for more evidence on perceptions) is likely to decrease, with positive impacts on confidence indicators and higher perception of social fairness.

Increased transparency will also bring more accountability of the organisations administering social security schemes towards their contributors, here again having positive impacts on the general perception of social fairness.

• 6. 

  W HAT ARE THE P OLICY PACKAGES AND THEIR IMPACTS ?

6.1. Description of alternative policy packages

Policy measures described in section 5.3 are now grouped in alternative policy packages. To address all objectives of the initiative, each policy package must necessarily include measures relating to all the areas, notably formal coverage, effective coverage & transferability as well as transparency. Measures are grouped together in a way that allows them to be coherent and mutually reinforcing in achieving the objectives of the initiative.

Only working in combination can the three objectives promote adequate protection. As previously pointed out, formal coverage is only a prerequisite for benefitting from the protection granted by social security schemes in case of need. If not accompanied by measures ensuring the build up and take up of adequate entitlements, adequate and effective protection will not be achieved. Likewise, omitting measures on transferability would imply a lack of protection in the presence of transitions between employment statuses. The need for transparency is the highest when an individual has over time access to several schemes with differentiated rights, some mandatory others voluntary, due to a high number of employment transitions. Importantly, the lack of transparent information may hinder people's ability to be aware of their rights and thus benefit from the protection they may grant.

Package A is the baseline in the absence of new EU action.

Packages B and C differ with regard to the ambition relating to the formal coverage of the previously uncovered groups, where package C entails formal coverage for all and package B provides for a differentiated approach (mandatory coverage for non-standard employees and voluntary coverage for the self-employed), also in reaction to the emphasis stakeholders put on the specific characteristics and heterogeneity of the group of the self-employed.

It should be noted that establishing mandatory coverage for the self-employed is a realistic option. Based on the mapping presented in Chapter 2.2.2, there are 16 Member States where social security coverage is mandatory for the self-employed for at least some of the branches under consideration. For each one of these branches, there exist relevant examples of existing mandatory schemes for this group in at least a few Member States. The suggested option is thus in line with realities in the Member States.

The choice of measure for formal coverage has implications for the choices regarding the other measures in the packages. This is because – as mentioned above - formal coverage is a prerequisite for the other dimensions, notably for effective coverage and transferability. It is for this reason that only package C, the more ambitious package regarding formal coverage, maximising the number of people for whom effective coverage and transferability may become relevant issues in the first place, also includes the provision of personal accounts to facilitate transferability. Personal accounts tie rights to the individual rather than his employment status, requiring comprehensive changes to the set-up of current systems, and can thus be considered the more ambitious measure in this respect. Indeed individuals will only benefit from the introduction of personal accounts if they do have social security rights in the sense of being formally covered. Thus for package B which only extends formal coverage on a voluntary basis to the quantitatively important group of the self-employed, the introduction of personal accounts and the corresponding necessary changes to the set-up of social security systems would seem disproportionate. Package B thus aims to facilitate transferability in a more proportionate way, via changes to rules in the existing systems, notably establishing mandatory transferability of accumulated entitlements and allowing the aggregation of periods upon changes between schemes.

At the same time, it should be noted that regarding effective coverage, both packages propose the same measure – adjusting (to the extent compatible with policy objectives) time thresholds that have been identified as obstacles in this respect. Precise levels or thresholds for such adjustment cannot be fixed at EU level, limiting the scope for differentiation between the two packages. Firstly, such levels would depend on the precise set-up of the system and interactions with other parts of the tax-benefit system. Secondly and relatedly, this is due to subsidiarity regarding the system design.

Finally, it is logical that higher ambition regarding measures to improve formal and effective coverage as well as transferability reduces the needs for transparency – the less differentiated the corresponding rights are, the less need there is for personalised information. For this reason, package B (which entails a higher degree of differentiation in rights) involves requirements for the provision of generic and personalised information on social security rights and obligations. In package C, no measures to provide personalised information would need to be taken separately – but indeed such information will be readily available as a byproduct of the introduction of personal accounts, without further efforts.

Package B Package C

 Mandatory for all workers in non-standard  Mandatory for all workers in nonforms of employment not yet covered and standard forms of employment and voluntary for the self-employed. self-employed

 Adjustment of time thresholds; mandatory  Adjustment of time thresholds; transferability of accumulated entitlements; entitlements disconnected from type aggregation of periods of employment via personalised

 Provision of generic and personalised accounts information  Provision of generic information

Table 1: Mapping of policy measures and policy packages

Packages

Objectives Measures B C

X (for non X (for non-standard i) Mandatory standard workers and for the

Formal coverage workers) self-employed)

• ii) 

  Voluntary X (for selfemployed)

  • i) 

    adjustment of time thresholds X

Effective coverage +transferability+aggregation

• & 

  transferability ii) adjustment of time thresholds +personal X

  accounts

  • i) 

    provision of generic information X X Transparency

  • ii) 

    provision of personalised information X X 1

1 No dedicated measures are required for this, as personalised information would become available as a by-product of the introduction of personal accounts.

6.2. Impacts of the policy packages

The following assessment does not repeat the assessment of the individual measures undertaken in Section 5.4. Instead, it describes the joint impacts, notably highlighting interactions between the different measures grouped together both in terms of synergies and trade-offs.

Package A / Baseline Scenario: This option does not address the problems, but represents the baseline, as set out in section 2.4, with which other options can be compared.

Package B: The package would lead to high effectiveness regarding protection and thus pronounced positive social impacts while entailing corresponding costs for public authorities and for employers. The differentiated approach fully respects the freedom of choice of the self-employed, and thus fully takes into account the nature of self-employment while reflecting the significant heterogeneity among the concerned group. The provision of generic and personalised information improving transparency could stimulate higher voluntary enrolment, at a relatively small cost for public authorities. The package could make a noticeable contribution to supporting labour market transitions and dynamics as it helps ensure the build-up of entitlements along careers associated with changes between social security schemes while also providing for coverage in the short term following such changes. This may help improve the allocation of resources.

Package C:

This package has a very high effectiveness by ensuring all on the labour market have formal coverage. The introduction of personal accounts usefully complements this by also providing protection following professional transitions, again with very high expected effectiveness. In thus reducing the differences in protection between groups, the package limits the needs for the provision of information, although the introduction of personal accounts would also entail the provision of personalised information as a by-product. As regards effective coverage, it is identical to Package C.

Table A.14 in Annex 10 summarises the key impacts of the packages in a comparative

manner.

6.3. Comparing the policy packages

The table below summarises the results of comparing the baseline scenario to the two alternative packages B and C. The comparison is based on elements discussed in more detail for each policy package in the sections above. A seven-stage qualitative grading system is proposed on the basis of the analysis of impacts in previous sections as follow:

No impact (0) strong positive impact / very good assessment (+++), small cost (-), substantial positive impact / good assessment (++), significant cost (--), small positive impact / satisfactory assessment (+), high cost (---).

Criteria for comparing policy packages A B C Effectiveness

• 1. 

  Formal coverage 0 ++ +++

• 2. 

  Effective coverage and transferability 0 +++ +++

• 3. 

  Transparency 0 ++ +++

Social Impacts 0 ++ +++

Cost 0 -- ---

Subsidiarity 0 ++ +

Proportionality 0 ++ +

Stakeholders' views 0 ++ +

Coherence

• 1. 

  Social Pillar objectives 0 +++ +++

• 2. 

  With other EU objectives 0 ++ +++

Both packages B and C achieve an overall good level of effectiveness:

 Formal coverage: There will be a substantial increase in the number of previously uncovered people in the case of Package B. The extent of the increase in the number of self-employed covered by social security depends on how many people enrol in voluntary schemes, while non-standard workers will be mandatorily protected. The increase in the number of people protected is expected to be strong in the case of Package C, since everyone will be covered on a mandatory basis.

 Effective coverage and transferability: the two packages have the same requirements for what concerns the adjustment of the time thresholds which determine the qualification for benefits and are thus expected to have the same substantial level of effectiveness in this respect. The creation of personal accounts in Package C would effectively solve all problems of transferability by de-connecting rights from employment status and linking it to individuals. Establishing rules facilitating transferability in the existing system as per Package B is also expected to have a good effectiveness in this respect.

 Transparency: The provision of personalised information as per Package B would have a substantial effectiveness in increasing transparency. In Package C, the creation of personal accounts would entail the provision of personalised information available at all times as a by-product without further action, achieving a strong degree of effectiveness.

Social impacts are expected to be strongly positive in Package C. Mandatory coverage together with full transferability through personal accounts and aggregation of contribution periods strengthen the ability of people to effectively access social security benefits in case of need, contributing to the corresponding expected reduction of poverty among people in nonstandard employment and among the self-employed. Positive impacts for the self-employed under Package B depend on enrolment rates; overall positive impacts could be substantially positive.

The direct costs of the packages comprise the cost of providing the benefits as well as costs related to improving transferability and transparency. As previously explained, no comprehensive quantification of the cost of the Packages is available. However, it can be expected that the corresponding cost is high in Package C entailing mandatory coverage for both groups and the creation of personal accounts, which would require important costs to be set up. For Package B, the cost associated to extending coverage is expected to be lower because through a voluntary approach for the self-employed, the number of individuals to be covered will be also lower.

Beyond the socio-economic criteria discussed above regarding the impacts of the packages, there are criteria of political nature against which the packages are assessed. Firstly, considerations on subsidiarity and proportionality are to be made. While both Packages respect subsidiarity since decisions of financing and design are left to the Member States, the mandatory approach of Package C for both groups leaves a low degree of flexibility in tailoring choices of formal coverage to the specificities of the self-employed. Given the special characteristics of and heterogeneity among the self-employed, a voluntary approach suggested in Package B appears more proportionate than Package C.

Secondly, the stakeholders' feedback gathered in the preparation of this impact assessment does not indicate a clear consensus among stakeholders, and in particular among representatives of the self-employed, for the strictest approach to formal coverage of the selfemployed. Indeed views are mixed and several stakeholders recalled the importance to leave freedom of choice to the self-employed, given their heterogeneity and the current limited knowledge base capacity to fully assess how the problem is likely to evolve. Measures proposed in Package B are thus assessed as representing a good compromise between the different views expressed. As Package C does not take into consideration the arguments presented by some employers and by some of the self-employed representatives; it is assessed as satisfactory in this respect.

Finally, the analysis of coherence with other EU policies takes into account in particular the European Pillar of Social Rights. The two packages, which aim at a comprehensive improvement of coverage and also address comprehensively the issue of effective coverage and transferability correspond closely to the Pillar's ambition. Coherence with the EU2020 objectives related to poverty is very strong under mandatory coverage for all (Package C), where a high reduction in poverty can be expected and coherence is thus assessed as good regarding Package B.

• 7. 

  P REFERRED P ACKAGE

7.1. Preferred policy option

The preferred policy option is policy package B. This package is the one that best balances achieving the policy objectives with the related costs, with proportionality and with political considerations notably including stakeholder feedback and subsidiarity.

By striking a balance between views expressed in the formal social partner consultation and taking into consideration the specificities and heterogeneity among the self-employed, the preferred option represents a realistic and proportionate set of measures appropriate to contributing to a realisation of the ambitions of the European Pillar of Social Rights.

It is important to notice that voluntary schemes can be set up with different mechanisms to nudge individuals to adhere to the schemes. Opt-out clauses imply that the individual is by default included in the scheme but he/she can opt to move out of the scheme. Opt-in clauses imply that the individual is by default outside of the scheme but he/she can opt to adhere to the scheme. In both cases the individual has the freedom to choose whether or not he wants to be a member of a given scheme. Voluntary schemes with opt-out options are usually more effective in providing protection (see Section 1 in Annex 10).

Given the nature of self-employed work, a mixed approach may be considered. For those risks that are more difficult to evaluate due to asymmetry of information and in absence of a counterpart that verifies the information - as the employers for the dependent workers - opt-in clauses might be the most appropriate. These could apply to schemes such as unemployment and accidents at work and occupational injuries. On the contrary, an approach that nudges more effectively self-employed to be covered may be considered for schemes where the asymmetry of information is less problematic. Opt-out clauses may be considered for risks such as pensions, invalidity sickness and maternity.

7.2. Overall impacts

On workers and the self-employed

There will be an increase in the number of people protected, which will decrease the number of non-standard workers at risk of poverty and to some extent of the self-employed (depending on how many will enrol in voluntary schemes). There could also be a general moderate reduction in income inequality via impacts on the self-employed and the nonstandard workers, although the improvement would depend on how many self-employed will enrol in voluntary schemes and on the hiring preferences of the employers.

Depending on Member States' financing choices, if part or all of the cost of extending coverage is left to the newly-covered groups, there would, as is usual in insurance schemes, be some income redistribution between those only bearing part of the cost and those for whom the risks and associated risks materialise. 97 At the same time, non-standard workers could experience a reduction in their disposable income if the number of the self-employed that enrols in voluntary scheme is low. A potential push towards (bogus) self-employment to avoid the payment of social security contributions would be limited by the clarity on the boundaries between workers and self-employed offered by the Commission's proposal for a Directive on transparent and predictable working conditions, which would help enforcement in this respect. Again depending on Member States' financing choices, there could be a limited loss in employment for affected groups in reaction to a potential increase in labour cost

Improved effective coverage and transferability will likewise result in a better protection of people in non-standard work and self-employment, including after transitions in the labour market which entail a change in social security scheme. Moreover, people in non-standard employment and self-employment will be able to maintain and use their accumulated entitlements.

Receiving personalised information will help them to know which rights they are entitled to. and will allow them to better prepare for the risk. Transparent information will also stimulate the participation of the self-employed in voluntary schemes or their choice to increase contribution rates in schemes allow to do it.

On companies

Depending on financing choices, companies are expected to experience an increase in the cost of managing fluctuations in production, in line with some labour cost increase related to nonstandard employment and the self-employment. The increase in cost should be reduced in relation to non-standard employment because most categories of non-standard workers are already covered by social security schemes, so that the share of those to be mandatorily covered is small compared to the totality of the workforce. The cost increase due to hiring services of the self-employed will depend on how many self-employed will decide to enrol in voluntary schemes and on the extent to which the self-employed will pass on their cost through higher prices.

On the other hand, companies could benefit from an increase in productivity of individuals in the newly protected forms of employment, from reduced unfair competition and from the positive effects of social security systems on labour markets.

Modification of effective coverage, transferability and transparency do not impact on companies at large.

On Member States and insurance providers

The exact financial impact will depend on i) how many unprotected people are present in the Member States and ii) how the new schemes will be designed or the old schemes extended, combining public and non-public schemes, as well as on financing choices. For instance, if Member States will finance the schemes through taxes, public expenditure would increase,

97 This implies a decrease in disposable income for those newly covered individuals for whom the risks do not materialise.

but this could be partially offset by reduced expenditure on last resort safety nets (e.g. social assistance) via a reduction in the number of people relying on them.

Adjusting thresholds to improve effective coverage could result in increased expenditure for (public or other) insurance providers, without corresponding changes on the revenue side. Establishing a set of rules for transferability and the protection of financial interests of schemes will also entail some costs and some redistribution between the schemes which are losing participants and those who are gaining. On the other hand, increasing the number of people protected and assuring that they are effectively protected in case of transitions can improve dynamism of national labour markets, the allocation of labour resources and the competitiveness of the economies. This can be expected to lead to lower recourse to benefits and thus have a positive impact on the financial sustainability of the insurance providers.

Concerning the provision of generic and personalised information, there would be an initial cost to set up the system but it would be followed by a limited cost for maintenance. These costs are expected to be offset by the positive impacts that transparency can bring on accountability and good governance of social security schemes. Furthermore, this could lead to improved social perceptions and sense of security, contributing to higher levels of consumer confidence, higher savings, more labour market dynamism and higher productivity.

• 8. 

  T HE CHOICE OF LEGAL INSTRUMENT

8.1. Legal options

The legal instruments considered for the initiative are a Council Recommendation and a Directive; there could also be a combination of the two. The possibility of an increased focus on the issue of social protection in the framework of existing tools, notably the European Semester and the Open Method of Coordination, is also considered. A careful choice of the level of detail of the provisions would ensure full respect of subsidiarity, irrespective of the specific instrument chosen. Subsidiarity is thus not an argument for determining the choice between the different legal options since it is compatible with the considered legal instruments.

Article 153 (2) TFEU provides the possibility of adopting a Directive in the area of 'social security and social protection of workers' to set binding minimum requirements for gradual implementation in the Member States. Article 352 TFEU provides the legal basis of an initiative related to self-employment.

A directive could set binding minimum requirements, reflecting the ambition of the corresponding principle of the European Pillar of Social Rights and de facto establishing rights. A directive would bring a higher predictability of implementation while leaving some discretion to Member States to transpose according to their national specificities. Given its binding character, a directive would have a high degree of effectiveness in reaching the objectives of the initiative. Taking into account the national competences in the area of social security as foreseen in Article 153 (2) TFEU, a directive could establish clear policy goals but would not be able to provide detailed guidance on the specific design and financing of social security systems.

Alternatively, the Commission may adopt a proposal for a Council Recommendation providing guidance to Member States on how to achieve the policy goals set. Although such guidance would not be binding, it can nonetheless provide a basis towards a level playing field between Member States on the way they will address the gaps of access to social security. A Council Recommendation would be based on Article 292 TFEU in combination with Articles 153(1)(c) and 153(2), third subparagraph, as well as with Article 352 TFEU.

While a Council Recommendation may have a somewhat lower degree of effectiveness - as it would not give the EU capacity to legally sanction a Member State nor give citizens the possibility of appealing to court in the case of no implementation - the non-binding character of a Council Recommendation would also allow the EU to leave maximum flexibility to Member States on the way to achieve the goals. Reporting commitments can be included in the Recommendation itself and policy coordination instruments such as the European Semester can be used to promote progress.

An alternative to a stand-alone Recommendation or Directive would be an increased focus on access to social security in the context of the existing policy framework and instruments in the field of social protection, namely the European Semester, its country-specific recommendations (CSRs) and the Open Method of Coordination (OMC) on Social Protection and Social Inclusion. These instruments have already been used to stimulate dialogue at EU level on how to improve the access to social security for non-standard workers and the selfemployed. However, the topic has not been covered in a systematic way in the European Semester – where it competes with a number of other key priorities - or on a regular basis in the OMC and guidance is usually limited and expressed in general terms. Consequently, only one country, the Netherlands, received a Country Specific Recommendation (CSR) on access

to social protection for non-standard workers in 2017 98 and only few peer reviews were

organised on this topic within the Social Protection Committee. This is partly due to the fact that the common objectives and indicators on access to social security gave little attention on how such access has evolved in function of changes in labour law and types of employment.

8.2. Preferred instrument

The preferred instrument is a Council Recommendation as it is regarded as the most effective and proportionate instrument at this point in time.

The consultations carried out in 2017, with social partners, relevant stakeholders and general public (see also Annex 11 for more details) indicated that there is no general consensus on the best instrument to be used. Some stakeholders underlined the stronger effectiveness of a directive; many others, including many Member States, showed little support for a legislative proposal, taking into account the different national points of departure, traditions, political preferences and budgets. Although a directive would likely to be more effective from a strictly economic point of view in reaching the objectives set out for this initiative, it is

98 “Tackle remaining barriers to hiring staff on permanent contracts. Address the high increase in the self-employed without employees, including by reducing tax distortions favouring self-employment, without compromising entrepreneurship, and by promoting access of the self-employed to affordable social protection. Based on the broad preparatory process already launched, make the second pillar of the pension system more transparent, inter-generationally fairer and more resilient to shocks”, Council Recommendation on the 2017 National Reform Programme of the Netherlands and delivering a Council opinion on the 2017 Stability Programme of the Netherlands, 9306/17, available online .

doubtful that it would gather support from all Member States, making it the less effective instrument overall.

Furthermore, the problem identified is of an evolving nature. The debate on access to social protection and the future of work is ongoing in several Member States and it will be influenced by ongoing reforms undertaken at national level. Certain aspects of it will also be influenced by the potential impact of some ongoing initiatives at EU level (including the Commission's proposal for a Directive of the European Parliament and of the Council on transparent and predictable working conditions in the European Union or the directive on pensions portability). The problem is however too important to postpone action waiting for the full materialisation of national reforms which, on their own and without general common guidance, may lead to uneven progress across groups (the self-employed vs. non-standard employees), social security branches and Member States.

Finally, the current statistical limitations make it difficult to fully monitor the evolution of the problem and such limitations – compounded with a number of choices left to the Member States such as the level and way of financing the action - have also strongly influenced the capacity to fully quantify the impacts of an EU action..

For the reasons explained above – the evolving nature of the problem, the ongoing reforms in some Member States and the strong statistical limitations – a Council Recommendation is a more proportionate approach than a Directive (or a Directive for certain elements of the proposal in combination with a Recommendation for the rest) which would impose binding outcomes.

The key added value of a Recommendation at this stage is to create momentum supporting and complementing national debates and reforms and guiding Member States' efforts in this. The Recommendation allows the EU level and the Member States to work further to understand the phenomena, its implications at national level and to stimulate and guide the reforms at national level. Importantly it will ensure that progress is not partial or uneven across target groups and social security branches. A focused approach is needed to give the issue the necessary political visibility, raise awareness and build momentum. This would help ensure that all Member States move in the same direction at the same time, thus promoting upward convergence to the benefit the EU economy as a whole. Indeed efforts are needed from all Member States in a coherent way and also in a somewhat simultaneous fashion to dispel doubts about possible first mover disadvantages.

A Council Recommendation can also contribute to create consensus on the best reform options, which is, for the time being, lacking at national and European debate, and is however necessary to design and implement reforms. The stakeholders' consultations (see below) reveal substantial divergence on the best instruments, as well as the direction of reforms.

Several examples from the recent past show how Council Recommendations can be effective instruments in achieving progress in the social field, notably the Youth Guarantee 99 and the Recommendation on Long Term Unemployment 100 . In particular, the Youth Guarantee, together with the Youth Employment Initiative, was instrumental in encouraging Member States' to step-up their efforts in this field. The Council recognised "the work done by

99 Council recommendation of 22 April 2013 on establishing a Youth Guarantee, 2013/C 120/01

100 Council recommendation of 15 February 2016 on the integration of the long-term unemployed into the labour market, 2016/C 67/01

Member States to mobilise additional financial resources, to strengthen and modernise their labour market institutions, and to support partnerships between the employment, education and youth sectors, as well as to engage with social partners, in line with national practices, and other stakeholders, in order to better deliver the Youth Guarantee." 101

The Recommendation could usefully be embedded in the existing processes such as the European Semester or the Social Open Method of Coordination, where Member States are already intensively involved. By stepping-up existing efforts to improve the knowledge-base and by providing specific guidance, the Recommendation would complement the analysis and guidance provided in the European Semester through country reports and country specific recommendations (CSRs). For countries with CSRs on this topic or related areas, the recommendation provides further guidance supporting their implementation effort. For countries without a CSR, it provides an EU level impetus for improvement of the coverage of the self-employed and non-standard workers. A Recommendation can also provide a coherent framework with a set of key principles, which could guide further exchanges, within the Social OMC, on how to improve the coverage and adequacy of protection through social security schemes, in practical terms.

• 9. 

  H OW WILL ACTUAL IMPACTS BE MONITORED AND EVALUATED ?

9.1. Monitoring

Member States will be invited to collect and publish reliable statistics, building on existing statistics where possible but acknowledging the existing gaps, broken down by employment status, gender, age and citizenships and to report on the corresponding measures taken at national level on a regular basis in the framework of monitoring of the European Pillar of Social Rights, of the multilateral surveillance and, in particular, of the European Semester.

Moreover reinforced cooperation will be established with Eurostat to create indicators appropriate for this area. Although long-time series would not be available, it should be possible to track progress towards formal coverage, effective coverage and transparency since the publication of the Recommendation. Work on improving indicators has already started and could be available at the same time as the Member States start implementing the Recommendation.

At European level, monitoring of the implementation of the Council Recommendation in the Member States could be built on the reporting arrangements within the European Semester, in the same vein as done for other Council recommendations. In the current Semester, the Joint Employment Report already includes - for the first time - a short analysis of the gaps in protection through social security systems related to the employment status of people in the labour market. For the Member States where the gaps are more acute, dedicated country analysis will also be subject of multilateral and thematic reviews held by the Social Protection Committee.

On the basis of a commonly agreed list of indicators, the Commission will report to other relevant EU institutions, including the European Parliament, Council, and the European Economic and Social Committee on progress in the implementation of the Recommendation.

101 See Council Conclusions on the implementation of the Youth Guarantee and the Youth Employment Initiative (2016)

In line with the Five Presidents' report 102 , the Commission will also propose to further explore, in the context of the Social Protection Committee, the feasibility of establishing a benchmarking framework for access to social security (see annex 14), pursuing the ongoing effort to establish benchmarking frameworks in areas that affect the sustainability of EMU.

The benchmarking and monitoring framework will aim at evaluating progress made by Member States in decreasing the share of non-standard workers and the self-employed who have no formal coverage and no effective coverage, by social security branch. It will evaluate whether Member States have adjusted time thresholds to ensure better coverage. It will also check whether legal frameworks enable transferability of entitlements and aggregation of periods across schemes. It will examine whether generic and personalised information on social security entitlements is available. Outcome impacts such as the poverty reduction effects of social benefits for the self-employed and non-standard employees compared to standard employees will also be evaluated.

The table A.15 in Annex 14 presents the operational objectives of the preferred option and the corresponding indicators to measure progress with regard to these objectives.

9.2. Evaluation

The Commission will proceed to a review of the actions taken in response to this Recommendation and an evaluation of the impact of the Recommendation in consultation with the Member States and social partners at EU level in line with a review provision in the text of the Recommendation.

The evaluation will include an assessment of whether the operational objectives of the Recommendation have been reached. The benchmark against which progress will be measured is the baseline situation defined in this Impact Assessment.

A mapping of existing gaps of social security has been done for the purposes of this Impact Assessment and will be repeated in some years. The comparison of the two mappings – exante and ex-post - will allow monitoring the evolution of the social security schemes, their coverage and their eligibility conditions., Targets can thus be established with reference to a reduction in the number of groups and issues identified as problematic across branches and Member States.

102 The Five Presidents' Report: Completing Europe's Economic and Monetary Union, June 2015

A NNEX 1: P ROCEDURAL INFORMATION

• 1. 

  Lead DG, D E cide Planning/CWP references

The lead DG is DG Employment, Social Affairs and Inclusion

Agenda planning/Work Programme reference 2017/EMPL/003.

Social Fairness Package: Initiative no. 8 in the Commission Work Programme 2018 103 .

• 2. 

  Organisation and timing

The Impact Assessment was assessed by the Interservice Steering Group on 11/12/2017 (present DGs SG, ECFIN, GROW, TAXUD) and on 12/01/2018 (present DGs SG, SANTE, GROW).

The Analytical Document accompanying the second phase consultation of social partners on which the Impact Assessment is based, together with the second stage consultation document, was assessed by the Interservice Steering Group on 11/07/2017 (present DGs SG, GROW, HOME, JUST, SANTE, SJ, TAXUD) and on 13/10/2017 (present DGs SG, GROW, SANTE) and adopted following ISC (DGs consulted AGRI, BUDG, CNECT, COMM, COMP, EAC, ECFIN, ENER, EPSC, ESTAT, FISMA, GROW, HOME, JUST, MARE, MOVE, REGIO, RTD, SANTE, SG, SJ, TAXUD, TRADE).

The first stage consultation document, together with the TOR of the impact assessment study were assessed by the Interservice Steering Group on 27/03/2017 (present DGs SG, ESTAT, GROW, HOME, JUST, SANTE, SJ, TAXUD) and adopted following ISC (DGs consulted AGRI, BUDG, COMM, COMP, DEVCO, EAC, ECFIN, ECHO, ESTAT, HOME, HR, CNECT, JUST, MARE, MOVE, REGIO, RTD, SG, SJ, TAXUD, TRADE, GROW, SANTE, FISMA).

• 3. 

  Consultation of the RSB

The Impact Assessment report was reviewed by the RSB and discussed with the author DG in a

meeting on 7 February 2018. On 9 February 2018 the RSB issued a positive opinion with reservations.

The revisions introduced in response to the RSB opinion are summarised in the table below:

RSB main reservations Changes done in the Impact Assessment

• (1) 

  The report does not sufficiently substantiate The dedicated Chapter on the choice of options now

the rationale for the choice of a first discusses in more detail the three possible options recommendation (Recommendation, Directive or more focus on access to social security in the context of existing instruments, notably European Semester and OMC) in Chapter 8.1. Chapter 8.1 (as well as Chapter 3.3) also emphasises that subsidiarity did not drive the choice of instruments but would be ensured for each of the possible instruments through an adequate choice of detail of the

provisions

The choice of preferred legal option is then further substantiated in Section 8.2, stressing in particular proportionality in view of ongoing reforms, the need to

103 COM(2017) 650 final i ANNEX 1, ANNEX to the Communication from the Commission to the European Parliament, the Council, the

European Economic and Social Committee and the Committee of the Regions Commission Work Programme 2018, An agenda for a more united, stronger and more democratic Europe.

guide reforms to ensure that progress is not uneven but convergence is achieved, as well as the results of consultations and Member States ownership as a key determining factor for the effectiveness of the

instrument.

• (2) 

  The report does not properly explain and The analytical steps that led to the narrowing of the

justify the limited material scope of the scope have been described in Chapter 2.2.2, where also initiative; it does not sufficiently clarify the the personal scope was further clarified. The link of coverage of the personal scope. the initiative with Art. 9 TFEU and Article 34 of the

Charter of Fundamental Rights is established in Section 3.1 which discusses the legal basis. The link to the European Pillar of Social Rights is further strengthened and relevant results from the consultation undertaken in its context are summarised in Chapter 3.2 and Annex 11. The link to the proposed Directive on transparent and predictable working conditions is further clarified in the discussion of the baseline (Chapter 5.1) and of impacts of measures on formal

coverage (Chapter 5.4.1).

• (3) 

  The report does not systematically back Reference has been made to the evidence gathered in claims on options and their impacts with presenting the rationale for the reduction in scope of robust evidence. the initiative (Chapter 2.2.2) and the presentation of

the packages (Chapter 6.1) better explains and backs the rationale for the combination of measures into packages. The higher incidence of non-standard work among women and the corresponding impact of the initiative on them are highlighted (Introduction to Chapter 2, Annex 4, Chapter 5.4.1) to underpin the gender dimension of the initiative. The effect of the choice of instrument on the assessment of impacts has been further emphasised and a more detailed discussion of macroeconomic effects has been added in Chapter 5. The monitoring framework (Chapter 9 and Annex 14) further details how progress could be

monitored despite the existing data gaps.

• 4. 

  Evidence, sources and quality

The following expert advice has fed into the Impact Assessment:

• - 

  From the Report of the public consultation "Accompanying the document establishing a European Pillar of Social Rights (SWD(2017)206)".

• - 

  From the commissioned study "Access to social protection for all forms of employment: Assessing the options for a possible EU initiative" by Fondazione Giacomo Brodolini.

• - 

  From the commissioned study "Behavioural study on the effects of an extension of the access to social protection for people in all forms of employment" by Open Evidence.

• - 

  From six commissioned case studies on special forms of employment (Mini-jobs in Germany, civil law contracts in Poland, self-employed without employees in the Netherlands, casual worker in Romania, economically dependent self-employed in Spain, project workers on continuous collaboration projects in Italy).

• - 

  From literature review as referred to in the footnotes.

A NNEX 2: G LOSSARY AND T ERMINOLOGY

• 1. 

  Glossary

Term or acronym Meaning or definition

CJEU Court of Justice of the European Union

CWP Commission's Work Programme

EMU Economic and Monetary Union

ESDE Employment and Social Developments in Europe Report

ESPN European Social Policy Network

Eurofound European Foundation for the Improvement of Living and Working

Conditions

EU-SILC European Union Survey on Income and Living Conditions

IA Impact Assessment

ICT Information and communications technology

ILO International Labour Organisation

LFS European Union Labour Force Survey

OECD Organisation for Economic Co-operation and Development

RSB Regulatory Scrutiny Board

SMEs Small and medium-sized enterprises

SWD Staff Working Document

TFEU Treaty on the Functioning of the European Union

• 2. 

  Definitions

• 1. 

  Social security

Social security covers all measures providing benefits, whether in cash or in kind, to secure protection, inter alia:

 from lack of work-related income (or insufficient income) caused by sickness, disability, maternity, accidents at work or occupational diseases, unemployment, old age;

 lack of (affordable) access to health care;  insufficient family support, particularly for children and adult dependants;  general poverty and social exclusion 104 .

This initiative focuses on protection from lack of work-related income (or insufficient income) caused by sickness, disability, maternity or paternity, accidents at work or occupational diseases, unemployment or old age.

Formal coverage: a group can be identified as formally covered in a specific social security branch (e.g. old age, unemployment protection, maternity protection) if the existing legislation or collective agreement sets out that this group is entitled to participate in the scheme.

Effective coverage: a group can be identified as effectively covered in a specific social security branch if the individuals in this group have an opportunity to accrue benefits and can access to those benefits in case of need (currently or in the future).

A social security scheme is a distinct framework of rules to provide benefits to entitled beneficiaries. Such rules would specify the personal scope of the programme, entitlement conditions, the type of benefits, benefit amounts, periodicity and other benefit characteristics, as well as the financing (contributions, general taxation, other sources), governance and administration of the programme.

A benefit is a transfer in cash or in kind made by a public or private entity to someone entitled to receive it as part of a social security scheme.

Duration of benefits refers to the period of time during which an insured or the insured's dependents can avail of the benefits of the social security scheme.

Qualifying period means the period of membership of a scheme, required under national law or by the rules of a scheme, in order to trigger entitlements to the accumulated social security rights.

Minimum working period means that the person is required to have worked a minimum of hours/months/years before being eligible to receive the benefit in case of need.

Reference period is the period within which the qualifying period must be completed (for instance, 12 months within 24 months).

Waiting period: in some states benefits are not paid immediately upon the occurrence of a contingency, in other words the claimant must wait for a fixed period after the beginning of their unemployment, sickness, etc., before they are paid any social security benefit.

Transferability refers to the possibility i) to transfer accumulated entitlements to another scheme and ii) to make qualifying periods in a previous employment status (or in concomitant employment statuses) count towards the qualifying periods in the new status.

104 Regulation (EC) No 883/2004 — on the coordination of social security systems listed ten social security branches: sickness; maternity and paternity, old-age pensions, pre-retirement and invalidity pensions; survivors' benefits and death grants; unemployment, family benefits and accidents at work and occupational illness.

Transparency refers to the provision of information to the general public, potential scheme members, schemes members and beneficiaries about the scheme's rules or about the individual obligations and entitlements.

• 2. 

  Employment statuses and types of employment relationship

A "worker" means a natural person who for a certain period of time performs services for and under the direction of another person in return for remuneration.

Self-employment refers to employment in which persons pursue a gainful activity for their own account 105 .

Employment status refers to the classification of persons as either working in the framework of an employment relationship (worker) or working on their own account (self-employed).

Type of employment relationship refers to the various types of relationships that can differ regarding the duration of employment, the number of working hours or other terms of the employment relationship.

Non-standard employment is described by the ILO 106 as employment arrangements which deviate from the 'standard employment relationship' (full time, of indefinite length, part of a subordinate relationship between an employer and an employee). It includes temporary employment (fixed term, casual, seasonal work), part-time and on-call work, multi-party employment relationship (subcontracted labour, temporary agency work) and disguised employment (misclassified self-employment but also dependent self-employment).

New forms of employment 107 include employee sharing, job sharing, interim management, casual work (including intermittent work and on-call work), ICT-based mobile work, voucher-based work, portfolio work, crowd employment, collaborative employment. These forms include elements of non-conventional workplaces, support of ICT, different employment relationship organisation, different work patterns, networking. Eurofound underlines that there is currently no shared understanding of what constitutes ‘new forms of employment’.

Some specific forms of work are also referred to:

Casual work: according to Eurofound's definition 108 , casual work is a type of work where the employment is not stable and continuous, and the employer is not obliged to regularly provide the worker with work, but has the flexibility of calling them in on demand; it is work which is irregular or intermittent with no expectation of continuous employment. Workers’ prospects of getting such work depend on fluctuations in the employer's workload.

Eurofound divides casual work into two main categories:

• - 

  On-demand work (including zero-hours contracts)

• - 

  Intermittent work

105 This definition is meant to be used in the context of the Impact Assessment and does not have any legal consequence.

106 ILO (2016): Non-standard employment around the world: Understanding challenges, shaping prospects. International Labour Office – Geneva.

107 Eurofound (2015): New forms of employment, Publications Office of the European Union, Luxembourg.

108 Eurofound (2015), page 46.

The first, on-call work, involves a continuous employment relationship maintained between an employer and an employee, with the option for the employer to call the employee in as and when needed. In some countries employers are obliged to indicate a minimum number of hours, while in others they are not obliged to ever call the worker (i.e. zero-hours contracts). The second category of casual work, intermittent work, involves an employer approaching workers on a regular or irregular basis to conduct a specific task, often related to an individual project or seasonal work.

Crowd employment/platform work: “A European agenda for the collaborative economy” 109 defines the concept of collaborative economy, the presence of an online platform being a necessary element of the definition. It does not, however, refer to platform work, which is not formally defined at EU level. Platform work is carried out by service providers who can be professional or not; they can be self-employed persons or workers.

Eurofound (2015) defines crowd employment as an employment form that uses an online platform to enable organisations or individuals to access an indefinite and unknown group of other organisations or individuals to solve specific problems or to provide specific services or products in exchange for payment.

Crowd workers are very rarely, if ever, employed by the platform. Instead, they predominantly operate on a self-employed basis as independent contractors and outside the scope of employment legislation. Eurofound notes that crowd employment platforms have to follow general legal frameworks such as commercial codes, civil codes, consumer protection acts and data protection legislation, but there are no legal or collectively agreed frameworks specifically addressing crowd employment in Europe 110 .

Domestic work: ILO defines “domestic work” in Article 1 of the Domestic Workers Convention, 2011 (No. 189):

(a) the term “domestic work” means work performed in or for a household or

households;

(b) the term “domestic worker” means any person engaged in domestic work within an

employment relationship;

(c) a person who performs domestic work only occasionally or sporadically and not on an occupational basis is not a domestic worker.

'Temporary Agency Work' is a form of work where the worker has a contract of employment or an employment relationship with a temporary-work agency with a view to be

assigned to a user undertaking to work temporarily under its supervision and direction 111 .

Voucher-based work is 'a form of employment where an employer acquires a voucher from a third party (generally a governmental authority) to be used as payment for a service from a worker, rather than cash' 112 . This type of employment has been introduced in several Member

109 COM(2016) 356 final i.

110 Eurofound (2015): New forms of employment, Publications Office of the European Union, Luxembourg.

111 Directive 2008/104/EC i, Article 3 (1) (c).

112 Eurofound (2015): New forms of employment, Publications Office of the European Union, Luxembourg, p.82.

States as a way to better regulate domestic work and improve protection of domestic workers, or tackle the problem of undeclared work in agriculture.

Traineeship is a limited period of work practice, whether paid or not, which includes learning and training components, undertaken in order to gain practical and professional experience with a view to improving employability and facilitating transition to regular employment 113 .

‘Bogus’ or ‘false’ self-employment refers to the phenomenon of workers who would usually meet the legal definition of a worker (see above) but instead are registered as self-employed. In this case, the subordinate employment relation is disguised as autonomous work, usually for fiscal reasons, or in order to avoid the payment of social security contributions and thereby reduce labour costs, or to circumvent labour legislation and protection, such as the provisions on dismissals. This can be characterised as an employer abuse of contractual relations, and these workers are also, compared to employed workers, at a higher risk of precariousness due to a lack of social security and pension coverage and a lack of access to some employment

rights 114 .

Dependent self-employed are responsible for their own work and therefore formally do not work under the authority and responsibility of the main contractor even if de facto they are economically dependent from one employer/customer.

Economic sectors are sectors of the economy, grouped according to the kind of products or services they produced or according to ownership.

113 According to the EU Quality Framework for Traineeships.

114 EP (2016).

A NNEX 3: F URTHER INFORMATION ON THE ORGANISATION OF SOCIAL PROTECTION

SYSTEMS

Social protection schemes may differ in a number of dimensions. There is no single classification available and recognised internationally. However, the European system of integrated social protection statistics, abbreviated as ESSPRO and developed by Eurostat defines a number of characteristics which can be freely combined to produce different groupings of schemes as required by analysis.

The categories are defined on the basis of (i) which body takes the essential decisions, (ii) the existence or absence of a legal obligation, (iii) the way entitlements are established, (iv) the scope of the scheme, (v) the level of protection provided. The classification is part of ESSPROS's regulation.

Decision-making

In total, Eurostat registered 1082 social protection schemes in Europe (including Switzerland, Iceland, Serbia and Turkey). Around 70 of them are not controlled by the central government or local authority. Among these, you can find many combinations: occupational or general schemes, compulsory or not compulsory, etc.

Box A.1: Examples of private insurance schemes

Belgium: The Voluntary Supplementary Pension Scheme for the Self-Employed allows selfemployed to save for a supplementary pension. Some 46% of the self-employed exercising a main activity are covered by this non-compulsory private insurance scheme.

Germany: The self-employed (with the exception of farmers and artists) and the well-paid salaried workers (2016: above €4,676 a month, adjusted annually) can choose between the private health insurance (PHI) system and a voluntary membership in the statutory health system (SHI) system. 8.8 million (11 per cent) people in Germany were qualified to hold a private health insurance (PHI).

Latvia: Private insurance companies offer health insurance policies – but mainly to legal entities with the sufficient number of employees. Until recently, private persons had no opportunity to take out a health insurance policy – only companies could do it, now some of insurance companies are selling policies also to private persons, but prices are much higher in that case and few people can afford it. Many large and medium-sized companies are buying health insurance policies for their employees, but for small companies it is often not affordable because of high prices.

Typologies and financing of social protection schemes

Non-contributory schemes are social transfer schemes which normally require no direct contribution from beneficiaries or their employers as a condition of entitlement to receive relevant benefits. They include a broad range of schemes including universal schemes for all residents, categorical schemes covering specific categories of residents and means-tested social assistance schemes. They are usually financed through tax or other state revenues.

Universal schemes, in particular, are non-contributory transfer schemes, which cover all residents, and provide benefits for all, whether working or not and irrespective of income. Often the only condition attached to the receipt of the benefit is that the person must be a long-term resident or a citizen of the country.

Social insurance – like all forms of insurance – is a mechanism for (financial) risk pooling. It represents one of the most important instruments through which social security schemes operate, as it has done since the outset of national systems of social security, the earliest being the system introduced by Chancellor Bismarck in Germany in the 1880s. The insurance principle operates on the basis that a group of individuals (or perhaps their families), who are vulnerable to a specific risk (for example, disabling injury, or sickness) pay contributions into a pool or fund which is then used to pay benefits according to specified rules so as to indemnify those participants who suffer from the occurrence of the relevant risk.

Social insurance is financed by contributions, which are normally shared between employers and employees, with, perhaps, state participation in the form of a supplementary contribution or other subsidy from general revenue. Participation is compulsory, with few exceptions. Contributions are accumulated in special funds, out of which benefits are paid. A person's right to benefit is secured by his contribution record without any means-test. Within floor and caps contribution and benefit rates are often earnings-related 115 .

115 Available online.

Occupational social security schemes provide workers, whether employees or selfemployed, in an undertaking or group of undertakings, area of economic activity, occupational sector or group of sectors with benefits intended to supplement the benefits provided by statutory social security schemes or to replace them, whether membership of such schemes is compulsory or optional 116 .

Financing of social protection

There is a wide variety in financing structures of social protection systems across Europe. While on average, around 55% of the financing is relying on social contributions (employers and persons protected), the share of total social contributions in total resources actually varies from around 20% in DK to 80% in EE (Figure A.1). This diversity reflects the variety of the provision of social protection in Europe, such as typically whether social protection systems are more insurance based (and thus more financed by social contributions) or universal in nature (and then more financed by general taxation), but also other dimensions, such as the relative share in expenditure of the different main risks (for instance pensions are generally more financed based on social contributions and health expenditure more based on general taxation, see para below) or the relative importance of occupational versus publicly managed schemes.

Some estimates of the relative shares of social contributions and general taxes to the financing depends by type of social protection area considered can be derived 117 . In general pensions and family expenditure rely more heavily on social contributions (on average around 75% of total financing) as well as unemployment benefits (around 60% on average), while health care systems rely less on social contributions (on average 40%) and more on general taxation.

Trends in social security financing show that general taxes increasingly complement social contributions (Figure A.2a shows that this continued in terms of structure since the crisis hit in 2008, though to a lesser extent). The share of general taxation (and other sources) in total resources increased in the latest decades and is now close to 45% (against 35% in 1995, see Figure A.3b). It can be noticed that this increase in the share of general taxation corresponds mostly to a decline in the share of employers’ and employees’ contributions, while the share of social contributions from the self-employed or other persons covered remained broadly constant. This observed trend of decline in the shares of employers’ and employees’ contributions combined with a growing number of persons in self-employment poses a double challenge to the sustainability of social security systems, particularly Bismarckian ones 118 .

Figure A.1: Structure of social protection financing (2014)

116 Council Directive 96/97/EC i. Available online.

117 See for instance Social Protection Committee (2015): Social protection systems in the EU: financing arrangements and the effectiveness and efficiency of resource allocation.

118 See DICE Database (2008): Bismarck versus Beveridge: A Comparison of Social Insurance Systems in Europe. Ifo Institute, Munich

(%GDP) (% total resources)

Source: ESSPROS

Figure A.2: Trends in social security financing structures

• a. 

  b.

Source: ESSPROS

The structure of social benefits expenditure

At EU level, social expenditure on old age pensions and healthcare has grown over recent

years as a percentage of GDP, while the expenditure devoted to family and unemployment

benefits has tended to either remain static or decrease. In 2014, overall social protection

expenditure (excluding administrative and other expenditures) accounted for 28% of GDP on

average across the EU, ranging from around 14.5% in Latvia and Lithuania to over 30% in

Denmark, France and Sweden (Figure A.3).

While there are clear variations across Member States in the actual composition of expenditure (Figure A.4), in all countries pensions (old age and survivors) and healthcare (sickness/healthcare and disability) account for by far the largest shares of social protection expenditure, with these averaging 44% and 35% respectively (or 13% and 10% of GDP) at EU level. In general, pensions account for the largest share of expenditure in almost all countries, with this being especially the case in Cyprus, Italy, Poland, and above all Greece. The other branches of social protection account for much more limited shares of social expenditure – at EU level family/child benefits account for 8% of total expenditure (or 2.4% of GDP), unemployment 5% (or 1.4% of GDP) and housing 2% (0.6% of GDP) although notable variations exist. For example, in Belgium, Finland, Ireland and Spain expenditure on unemployment benefits is well above the EU average, while in Denmark, Finland, Germany, Luxembourg and Sweden this is more the case for family/child benefits, and in the UK there is markedly more expenditure on housing benefits.

Generally speaking, the level of social contributions from the self-employed is significantly lower than that of employees, taking into account both employees and employers contributions. However, it is similar to the level of employees' contributions only (Figure A.5), which can typically relate to the design of social security provision (with less extensive or voluntary settings for the self-employed) 119 .

Figure A.3: Structure of social protection expenditure as % of GDP, 2014 (excluding administrative and other expenditures)

Source: ESSPROS

Figure A.4: Structure of social protection expenditure as % of total expenditure, 2014

119 This is, in particular, the case of some Member States (Estonia, Hungary, Latvia, the Netherlands, Romania, Poland and to a lesser extent Cyprus, Germany, Greece, Croatia, Portugal, Sweden and the UK). These comparisons need, however, to be considered with caution as in particular they do not account for possible differences in average earnings between employees and between the self-employed and employees (for instance, if self-employed earnings would be lower, the estimated apparent contribution rates would be higher), nor do they take into account possible additional individual savings in particular for the self-employed.

Source: ESSPROS

The role of social dialogue in easing transitions

Social partners play a key role in the design of social protection schemes that derive from collective bargaining. These schemes often relate to a given economic sector, reflecting the structure of sectoral dialogue. In other instances, there are initiatives with a cross-sectoral dimension. In either case initiatives from social partners no longer consider solely the social protection benefits inherent to a given contract and take a wider perspective of securing transitions and whole professional careers.

For example, a variety of joint approaches have been used to address issues related to extending social protection to Media, Arts and Entertainment sector workers, mostly as an outcome of social dialogue. Some trade unions offer social protection coverage to freelance members. In some countries, specific schemes have been developed to provide coverage to Media, Arts and Entertainment workers – for example, the social security insurance fund for

artists and writers in Germany 120 , which covers self-employed and freelance artists and

writers, and the unemployment benefit for intermittent entertainment workers in France.

Figure A.5: Estimates of the relative apparent contributions of self-employed vs. employees in the EU (2014)

120 Available online .

Source: ESSPROS

Note: Estimates of GDP per employment are derived from the national accounts for GDP levels and from LFS for employment levels; the number of self-employed is also taken from the LFS.

Since 1969 in France, performing artists working on a short-term basis (intermittent du

spectacle) benefit from a rebuttable presumption of employment enshrined in the French

Labour Code. This provision applies to performing artists and to technicians who regularly

work on a short-term basis. The mechanism of the Guichet Unique du Spectacle Occasionnel

(GUSO) is designed to facilitate the protection of short-term workers in France and their

enjoyment of this status. This mechanism and the vocational training fund that it facilitates are

an outcome of the strong social dialogue between employers and workers on behalf of the

intermittent du spectacle. The Swedish 'Teateralliansen' (there is also a similar 'dansalliansen'

which caters for dancers) is a structure specifically created to increase the social protection of

freelance actors 121 .

Another case in point is the bipartite funds that offer extra social protection to agency

workers. These funds are financed through employers' contributions and governed through a

bipartite model. The funds are cross-sectoral and allow agency workers to benefit from their

protection regardless of the sector in which they perform their assignments or between two

assignments.

121 Available online .

A NNEX 4: F URTHER INFORMATION ON NON - STANDARD EMPLOYMENT AND SELF -

EMPLOYMENT

• 1. 

  Self-employment: decline in traditional forms amid growth in new forms

1.1. The self-employed: group size, characteristics and heterogeneity

There were 32.7 million self-employed in the EU-28 in 2016, accounting for 15% of total

employment. They were mainly concentrated in the following sectors: agriculture, forestry

and fishing (14.1%), wholesale and retail trade (15.8%), construction (12.9%) and

professional, scientific and technical activities (12%) 122 . Southern and Eastern European

countries have the greatest percentage of self-employed people, with Greece topping the

charts (30%). At the other end of the spectrum, two central and Northern European countries

have the lowest proportion of self-employed, i.e. Denmark (8.3%) and Luxembourg (9%).

The Netherlands and Ireland have self-employment rates above the European average (Figure

A.6) 123 .

Figure A.6: Share of self-employed in total employment, by countries, 2016

Source: Eurostat, Employment by sex, age and professional status

Among the self-employed it is possible to distinguish between the self-employed with and

without employees. In 2016, 71.5% of the self-employed were self-employed without

employees 124 ; 28.5% of the self-employed without employees were registered as managers,

professionals or associate professionals, while around 21.4% were agricultural or elementary

workers, 16.2% were reported to be sales or service workers and 21% skilled or semi-skilled

manual workers such as craft and related trade workers. The occupations of the self-employed

vary considerably across countries. For example, in the Sweden, Belgium, Germany and

122 There is a substantial variation across Europe. The share of self-employment is very high in agriculture, forestry and fishing in countries like Romania (65.1% of the total number of self-employed) and Poland (38.9%), and rather low in countries like the United Kingdom (3.0%), Slovakia (3.7%) and the Czech Republic (4.4%). A large proportion of self-employed workers can be found in the wholesale and retail industry in Malta (31.3%), Bulgaria (26%), Spain (24%) and Italy (22.8%). The United Kingdom, Slovakia and Ireland have the largest proportion of self-employed in the construction sector (between 15.5 and 25.2%).

123 See ESPN (2017).

124 EU LSFA.

Luxembourg, well over 50% of the self-employed without employees worked as managers,

professionals or associate professionals in 2016, while in Romania they were just 3%.

The overall (marginal) decrease of the self-employed between 2008 and 2016 becomes an

increase, if the primary sector is excluded from the analysis (Figure A.7).

Figure A.7: Share of self-employed in total employment in the EU28, 2008 and 2016

Source: Eurostat

Moreover, it is possible to note that there has been a change in the socio-economic structure

of the self-employed population (Figure A.8). A high decrease can be noticed among the

managers, and the occupations related to agriculture and manufacturing, while there has been

an evident increase among the self-employed in services and sales and among the

professionals.

Figure A.8: Share of self-employed in total employment by occupation in the EU28, 2008

and 2016

Source: Eurostat

As also expressed by stakeholders during consultations, it is very important to go further indepth in the analysis to identify and take into account the diversity among the group of the self-employed (going beyond the traditional split between the self-employed with/without employees) in order to assess to which extent granting them access to social security responds to an actual need 125 .

A study by Eurofound 126 sheds light on the heterogeneity of the self-employed using survey evidence exploited with the help of advanced econometric methods 127 .

In particular, the self-employed can be clustered into five different groups ('employers 128 ', 'stable own account workers' 129 , 'vulnerable' 130 , 'concealed' 131 , 'small traders and farmers' 132 -

see also Figure A.9) in line with their characteristics along a pre-selected list of key dimensions relevant to their employment and work situation. The names of the groups were chosen to illustrate the most important aspect of the nature of each group. They are not commonly agreed or formal terms.

Of those five groups, two groups – the ‘employers’ and the ‘stable own-account workers’, representing about half of the self-employed or just over 16 million people - generally have more favourable characteristics. These groups are generally more independent with more discretion over their work, have bigger and economically more viable and sustainable activities and are more likely to be self-employed out of opportunity. For two other groups, ‘vulnerable’ and ‘concealed’, together a group of eight million people - or roughly the population size of Austria - the opposite is the case: they are generally more dependent and have less autonomy over their work. Also, these groups are in more precarious situations with lower levels of income and job security. They experience overall rather unfavourable working conditions. For the final group, ‘small traders and farmers’, which amount to also around eight million people, the picture is more mixed with a combination of both favourable and unfavourable conditions.

Figure A.9: Size of five groups of self-employed within the EU28 workforce (% of total employment)

125 For instance, self-employed who are business starters or new entrepreneurs may face particular difficulties, see SME Action programme" finalised by the SME Envoy Network in 2017, http://ec.europa.eu/growth/smes/business-friendly-environment/small-business-act/smeenvoys_en.

126 Eurofound (2017), Exploring self-employment in the European Union. Luxembourg : Publications Office of the European Union.

127 Latent class cluster analysis looks for patterns regarding a list of indicators listed and by showing how each of the clusters score on these indicators, reveals the defining characteristics of each cluster.

128 The self-employed in this cluster are characterised by working on multiple sites and employing multiple employees, often more than eight. They have high discretion over their professional situation, both in terms of laying off employees and of taking time off for personal or family matters.

129 The size of their economic activity is limited to one site only (98%) and 99% do not employ any staff. The job is relatively sustainable and not precarious, given this cluster’s low representation in the lowest income quintile and the average level (compared with that of other self-employed workers) of financial security in the case of sickness. Most of the self-employed in this category have high discretion over their professional life: they would have the authority to dismiss personnel (if applicable) and almost 90% are able to take time off at short notice for private matters.

130 The self-employed in this cluster almost exclusively work on one site only and a majority (76%) work alone, while 24% have between one and eight co-workers. The economic sustainability of these jobs is generally low: four out of five self-employed in this cluster are in the lowest and the second lowest income quintiles, and they are generally not paid a fixed fee on a weekly or monthly basis. Only a minority (28%) has the authority to dismiss employees, but taking time off at short notice is easy for 84%.

131 The majority of workers in this cluster (65%) is paid on a weekly or monthly basis and work with co-workers, while the majority (63%) work five days a week or fewer. This cluster is most strongly characterized by high levels of operational dependence or low autonomy at work. In general, they have limited discretion over their work situation: only 15% have the authority to dismiss other workers, and for 33%, it is difficult to take time off at short notice).

132 Self-employed people in this cluster are predominantly working without employees or with a limited number of employees and on one site only. They are generally economically independent because they have several clients and most of them find it easy to attract new customers. They have discretion over their work situation, they find it relatively difficult to take an hour or two off for personal or family reasons. They are overrepresented in agriculture and commerce.

Source: Eurofound (2017)

There is strong heterogeneity between the groups along a number of dimensions of high relevance for the present initiative. For example, nearly all the self-employed in the cluster ‘employers’ did not become self-employed out of necessity whereas only around 60% of the self-employed in the cluster ‘vulnerable’ are not self-employed out of necessity. The degree of being comfortable with the uncertainty and risks associated with the status of selfemployment also differs markedly across the groups. While more than 90% of the ‘employers’ and close to 100% of the ‘stable own-account workers’ like being boss and taking decisions, the same holds true for less than three quarters of the ‘vulnerable’ and only about 55% of the ‘concealed’.

The clusters differ substantially with respect to their autonomy: Three quarters of ‘stable ownaccount workers’ have the authority to hire or dismiss employees while the shares for ‘employers’ and ‘small traders and farmers’ even exceed 90%. This only holds for below 20% of the ‘concealed’ and around 30% of the ‘vulnerable’ self-employed. Likewise, more than 90% of the ‘stable own-account workers’, ‘small traders and farmers’ and ‘employers’ have more than one client, compared to less than half of all ‘vulnerable’ self-employed.

There is also large diversity regarding the position in the income distribution and financial vulnerability. More than 60% of the ‘vulnerable’ are in the lowest income quintile, while this holds for only around 10% of the cluster ‘employers’ and 20% of the ‘small traders and farmers’. Close to 60% of ‘employers’ consider themselves financially secure in case of sickness. The other groups are more vulnerable, with only around 20% of ‘small traders and farmers’, 20%-30% of the clusters ‘concealed’ and ‘vulnerable’ and below 40% of ‘stable own-account workers’ being in the same situation. Looking at income levels alone does not give a sufficient picture of vulnerability in this respect. For instance among the ‘small traders and farmers’, almost two-thirds would be financially insecure in case of sickness, although incomes in this group tend to be at the higher end.

1.2. New forms of self-employment: dependent self-employment and bogus self-employment

In line with the European Court of Justice rulings , the status of worker is determined by 133 three elements: existence of remuneration in return for services; existence of work in a relationship of subordination where the employed determines the choice of activity, remuneration and working conditions; and the existence of effective and genuine activity of economic value. Out of these three elements, the relationship of subordination allows to draw a distinction between workers and self-employed persons and the most specific way of defining self-employment is the absence of subordination between the employer and the employees 134 .

Given the characteristics associated with entrepreneurship – decision-making power regarding the business, financial autonomy and responsibility – the ‘self-employed with employees’ are considered as being by nature self-employed. The category of ‘self-employed without employees’ is more diverse and, in some ways, ambiguous 135 . The group of economically dependent workers belongs to this category. To be categorised as an independent selfemployed without employees, an individual should work as an autonomous subcontractor, who owns his/her means of production and works independently in order to accumulate

capital and serve several clients 136 .

Thus, there are emerging work arrangements which are “in a grey area”, meaning that they do not easily fit into the traditional binary distinction between "employees" and the independent "self-employed”. In fact, there is an increase in formally self-employed work relationships that are marked by both economic and personal dependence. Given that these workers have only one (main) employer, they generate the whole, or at least a substantial part, of their income from this work relationship. Economic dependence may also mean that the worker takes (part of) the entrepreneurial risk without the trade-off in terms of employment security. Furthermore, due to the demands of their 'main' employer, the entrepreneurial possibilities associated with self-employment are also rather limited or absent in their case. These characteristics form the category of the dependent self-employed who work with only one client, under similar conditions as employees, but without the rights and protection that employees are entitled to under labour law, social protection rules and collective agreements. Dependent self-employment also means that both organisational boundaries and the boundaries set by labour and social security law between employment and self-employment become blurred 137 . Indeed, the distinction between the dependent self-employed and employees has important labour law, social protection and fiscal consequences:

 dependent self-employed are responsible for their own work and therefore formally do not work under the authority and responsibility of the main contractor even if de facto they are economically dependent on one employer/customer;

 access to social protection (e.g. in the case of temporary employment, occupational

accidents, early retirement etc.) is more restricted for the dependent self-employed 138 .

133 Case C- 268/99, Case C-3/87, Case C-202/90, Case C-202/90. See also, to that effect, judgments in FNV Kunsten Informatie en Media, C- 413/13, EU:C:2014:2411, paragraph 35 and Allonby, C 256/01, EU:C:2004:18, paragraph 71.

134 IZA (2013): Social protection rights of economically dependent self-employed workers. European Union.

135 Eurofound (2013): Self-employed or not self-employed? Working conditions of 'economically dependent workers'. Dublin: Author.

136 Thörnquist (2015): False Self-Employment and Other Precarious Forms of Employment in the ‘Grey Area’ of the Labour Market. International Journal of Comparative Labour Law and Industrial Relations, 2015, 31(4), 411-429. Henceforth, this is referred to as Thörnquist (2015).

137 FIA (2015): Reaching out to Atypical Workers. Available online .

138 Available online.

 some working conditions (wages, working time, rest periods etc.) are governed by collective agreements or by specific legislative, administrative and regulatory

provisions that are not applicable to the dependent self-employed 139 .

Therefore, the dependent self-employed are neither clearly separated from nor integrated with the employer they contract with. Moreover, despite the lack of data which would allow to easily distinguish between the traditional self-employed and the dependent self-employed, it is estimated that approximately 15% of all self-employed are economically dependent workers in the EU 140 . Moreover, there is evidence that the biggest difference between dependent self-employed and other groups of self-employed lies in their financial situation. Their households have a correspondingly higher level of difficulty making ends meet and they have the lowest level of job security, and the fewest opportunities for career advancement 141 . A Eurobarometer survey 142 shows that 49% of respondents who had started up a business or were currently taking steps to start one, stated that they were doing so because they saw an opportunity and 29% were doing so out of necessity, i.e. the possibilities for regular employment were absent 143 .

Moreover, among the dependent self-employed, another group with the only choice of

working on their own account because of a lack of job opportunities 144 can be identified: that of the 'bogus self-employed' 145 .

‘Bogus’ or ‘false’ self-employment refers to the phenomenon of workers who would usually

meet the legal definition of an employee but instead are registered as self-employed 146 . In this

case, the subordinate employment relation is disguised as autonomous work, usually for fiscal reasons, or in order to avoid the payment of social security contributions and thereby reduce labour costs, or to circumvent labour legislation and protection, such as the provisions on

dismissals 147 .

The OECD has described the bogus (or ‘false’) self-employment as consisting of ‘people whose conditions of employment are similar to those of employees, who have no employees themselves, and who declare themselves (or are declared) as self-employed simply to reduce tax liabilities, or employers’ responsibilities’. This category is particularly used in labourintensive industries, such as construction, transport and the service sector 148 .

There are several factors leading to 'bogus' self-employment. One is that in almost all countries few formalities have to be fulfilled to set up as self-employed and the forms of bogus self-employment have gradually become more sophisticated 149 . Relatively restrictive national labour markets (i.e. highly regulated ones) and high non-wage labour costs are also

139 Available online.

140 Eurofound (2017): Exploring self-employment in the European Union, Publications Office of the European Union, Luxembourg.

141 Eurofound (2013).

142 Eurobarometer (2012): Entrepreneurship in the EU and beyond.

143 Bögenhold, and Klinglmair (2017): One-person enterprises and the phenomenon of hybrid self-employment: evidence from an empirical study. Empirica (2017) 44: 383..

144 Studies show a negative correlation between the level of social benefit available to unemployed individuals and the concentration of lowskilled self-employed, fortifying the idea that a certain kind of self-employment is due more to necessity than to personal free choice (CESifo DICE Report, 2016).

145 Fondeville et al. (2015): Recent changes in self-employment and entrepreneurship across the EU. Research note 6/2015. Publications Office of the European Union, Luxembourg.

146 Hatfield (2015): Self-employment in Europe. IPPR.

147 FIA (2015).

148 Eurofound (2016): Exploring the fraudulent contracting of work in the European Union. Dublin: Author.

149 Jorens, and Lhernould (2010): Europe of the self-employed: Self-employed between economic freedom and social constraints. In The Europe of the Self-employed, Proceedings.

factors creating 'bogus' self-employment 150 . From the point of view of the employers, it is a way to transfer costs, risks and responsibilities to the workers and to circumvent collective

agreements, labour laws, payroll taxes and social security contributions 151 . On the other hand,

while, in general, motives for people accepting a working engagement as dependent selfemployed could include time flexibility and increased autonomy, together with further advantages in the form of tax benefits and lower social contributions, in the case of 'bogus' self-employment, reasons seem to be related only to their economic situation, since the characteristics of autonomy and flexibility are not maintained.

• 2. 

  Non-standard work: temporary, part-time and new forms of employment

Overall, the share of non-standard work in employment has increased over time. Using data for the EU28 and a longer available time series for the EU15, Figure A.10 shows that this has been the case especially for younger workers aged 25-39: in 1995, 23% of younger employees in the EU-15 had non-standard contracts. This proportion had increased to 32% for the same age group by 2016, implying that there has been a clear intergenerational component to this development.

Figure A.10: Employees by type of contract and age, 1995, 2005 and 2016

40 Temporary Part-time

Temporary Full-time

35 Permanent Part-time

) 30 5 e (

%

4 3 4

typ

ct 25 2 3 5 3

n tr

a 11

20 3

2

11 5 2

12 5 b y co

e s 15 8

4 11

5

p loye

E m 10 19 21

14 16 15 16

17

5 12 12

13

0

1995 2005 2016 1995 2005 2016 2005 2016 2005 2016

25-39… 40-64… 25-39… 40-64…

Source: European Commission, ESDE 2017, p. 77.

2.1. Temporary employment

In 2016, the proportion of employees aged 15–64 in the EU-28 with a fixed-term employment contract was 14.2% (Figure A.11). More than one in five employees in Poland (27.5%) and Spain (26.1%) had a temporary contract. Among the remaining Member States, the share of employees working on a contract of limited duration ranged from 22.3% in Portugal, down to 2% in Lithuania and 1.4% in Romania.

150 European Parliament (2013): Social protection rights of economically dependent self-employed workers. Publications Office of the European Union, Luxembourg.

151 Thörnquist (2015).

Figure A.11: Share of temporary workers in total employment (age 15-64), 2016

Sources: Eurostat

Both the number of temporary contracts and the temporary employment rate (number of temporary as percentage of all employees) show a strong cyclical (and seasonal) pattern. Both decline at the onset of the recession and tend to increase with recovery (and pick up the

“double dip” around 2010-11), as many new hires are on temporary contracts 152 . Indeed, over

the last three years (2014-2016), as countries emerged from the crisis, the growth in the number of permanent employees has exceeded the increase in the number of temporary employees. Thus, in the year preceding the third quarter of 2016, the number of employees with a permanent contract increased by 2.8 million (1.8%) in the EU-28, while the number of employees with a temporary contract increased by 420,000 (1.6%). Similarly, full-time employment has increased faster than part-time employment for nearly three years. In the years up to the third quarter of 2016, the number of full-time workers increased by 2.8 million (1.6%) while the number of part-time workers increased by 370,000 (0.9%).

Denmark, Netherlands, Slovenia, Malta, Croatia and Poland show some appreciable increase in the temporary rate since the onset of the recession in 2008. Spain, Germany, Latvia and Lithuania show a significant post-recession decrease. Given Spain’s large share of temporary contracts in Europe, the decline there has held down the European rate over this period, and vice-versa for Poland.

In a more long-term perspective, a clear increase in the share of workers on temporary contracts has been observed especially for younger workers (aged 25-39), see Figure A.12. At the same time, it is worth noting that women are more likely to work on temporary contracts than men, although the gap has been shrinking (Figure A.12).

Figure A.12: Share of temporary workers in total employment (by age and gender,

1995-2001 (EU-15) and 2002-2016 (EU-28)

Source: European Commission (2017 ESDE), p. 79.

Among temporary employment, agency work is experienced as one of the more insecure of contractual forms 153 . This is characterised by a particular “triangular” relationship, where workers are hired by an entity (the temporary work or employment agency) and then re-hired out or assigned to perform their work at (and under the supervision of) a user firm. In 1999, when this form of employment had just become legally permissible in most of Europe 154 , agency workers accounted for 1.2% of all employment in Europe. In 2015, it accounted for

1.9% 155 . The highest rate was in the UK (3.8%), followed by the Netherlands (3.0%) and

France (2.1%). The lowest rate of 0.1% was in Greece. These figures are rather similar to those from 1999, with the exception of Germany where the agency worker rate increased from 0.7% to 2.4% of total employment.

The motives for employees to take on a temporary contract are primarily non-voluntary. Given the option of two identical jobs, one open-ended and the other with a temporary

152 Eurofound (2017): Aspects of non-standard employment in Europe, Eurofound, Dublin.

153 European Working Conditions Survey (2016).

154 Eurofound (2004): Temporary agency work in the European Union. Available online .

155 World Employment Confederation (2017): Enabling work, adaptation, security and prosperity.

duration, the Labour Force Survey shows that only around 12% report that they did not want a

permanent job 156 .

2.2. Part-time employment

Part-time work has grown for decades. This is due to the changing structure of the economy

towards services and preferences for part-time work, not least related to the increase in female

labour force participation. The proportion of the EU-28 workforce in the 15–64 age group

reporting that their main job was part-time increased steadily: from 17.5% in 2007 to 19.5%

by 2016, representing the type of non-standard employment that has undoubtedly grown since

the onset of the recession. Over the period from 2007 and 2015, there was a widespread

increase in the share of part-time workers, though it declined in Croatia, Lithuania, Romania

and Poland 157 .

Figure A.13: Share of part-time workers in total employment (age 15-64), 2016

Source: Eurostat

As shown in Figure A.13, by far the highest proportion of part-time workers in 2016 was

found in the Netherlands (almost 50%), followed by Austria, Germany, Denmark, the United

Kingdom, Belgium, Sweden and Ireland. By contrast, part-time employment is relatively

uncommon in Bulgaria (2%) as well as in the Czech Republic, Slovakia, Hungary and Croatia

(between 4.8% and 5.8%). In these countries, non-standard employment is rare, and people

may consider it to be an involuntary status.

The incidence of part-time work differs significantly between men and women. Almost one

third (32%) of women aged 15–64 who were employed in the EU-28, worked on a part-time

basis in 2015, a much higher proportion than the corresponding share for men (8.9%). 76.5%

of women employed in the Netherlands worked on a part-time basis in 2016, by far the

highest rate among EU Member States 158 .

156 One should treat this figure with some skepticism, as interview questions on motives can be open to many interpretations. Also as noted in 'Recent Developments in Temporary employment: employment growth, wages and transitions' (Eurofound 2015), responses vary significantly over time and in several cases they are obviously unreliable.

157 EU LFS

158 EU LFS

In the most recent years, as the labour market improves, the part-time employment rate has declined only marginally, thus indicating that not all part time is voluntary. Self-reported involuntary part-time work (defined in relation to the response in the LFS survey “Person could not find a full-time job”) increased from 22.4% of all part-time work to 29.1% between 2007 and 2015 159 . There is a very strong association between national labour market performance and changes in the share of involuntary part-time work. Where the unemployment rate has increased most, there is a strong likelihood that the involuntary parttime share has also grown strongly. Greece, Spain and Cyprus are illustrative examples. The corollary is also true. The Member State with the most improved unemployment record over the period, Germany, is also the one with the sharpest decline in involuntary part-time share (Figure A.14) 160 .

Figure A.14: Changes in involuntary part-time employment share and in unemployment rate 2008-2015.

Source: LFS as calculated in Eurofound (2017)

Changes concern also the working hours. While women account for the majority of very low hours' part-time work, the male share has grown relatively faster since 1996 (from 1.5% to 2.8% of all male workers) compared to the female share (from 6.1% to 6.6% of all female workers). Highest shares were recorded amongst both young workers (less than 24 years of age) and workers above 64 years of age and it was in these groups, in particular the younger group, that the greatest growth in the share of very short hour part-time work was recorded.

2.3. New forms of work

As in the case of the self-employed the available data do not allow to capture the variety of non-standard forms of employment present in the labour market, and, in particular, the new forms of work.

Eurofound (2015) 161 distinguishes three categories of employer-employee relationship: relationships between employers and employees involving either multiple employers for each employee, one employer and multiple employees or even multiple employer-multiple

159 EU LFS.

160 Eurofound (2017).

161 Eurofound (2015): New forms of employment. Dublin.

employee relationships; provision of work on a discontinuous/intermittent basis or for very limited periods of time; and networking and cooperation agreements involving self-employed persons, especially freelancers.

On the basis of this broad categorization, the following new forms of non-standard employment were identified by Eurofound:

 employee sharing, where a single worker is jointly hired by a group of employers to meet the HR needs of various companies, resulting in permanent full-time employment for the worker.

 job sharing, where an employer hires two or more workers to jointly fill a specific job, combining two or more part-time jobs into a full-time position;

 interim management, in which highly skilled experts are hired temporarily for a specific project or to solve a specific problem, thereby integrating external management capacities in the work organisation;

 casual work, where an employer is not obliged to provide work regularly to the employee, but has the flexibility of calling them in on demand;

 ICT-based mobile work, where workers can do their job from any place at any time, supported by modern technologies;

 voucher-based work, where the employment relationship is based on payment for services with a voucher purchased from an authorised organisation that covers both pay and social security contributions;

 portfolio work, where a self-employed works for a large number of clients, doing small-scale jobs for each of them;

 crowd employment, where an online platform matches employers and workers, often with larger tasks being split up and divided among a ‘virtual cloud’ of workers;

 collaborative employment, where freelancers, the self-employed or micro enterprises cooperate in some way to overcome limitations of size and professional isolation.

In general, crowd work seems to be used as a supplement to total income. A survey conducted in the UK showed that in 2016 platform workers amounted to 4% of the employees, 25% of whom reported it as main job 162 . Another survey 163 implemented in the UK, Sweden, Germany, Austria and the Netherlands in the first two quarters of 2016 showed that between 5% and 9% of the online population were engaged in some type of crowd work 164 . It is usually a small supplement to total income 165 .

Furthermore, together with these new forms of employment, a variety of new forms of contracts has been recognised in Member States:

 'Civil law contracts for a specific task' there are different legal forms from a contract of employment and they are governed by the provisions of the Civil Code. They are particularly prevalent in Poland and commonly used when the employer does not plan

162 CIPD (2017): To gig or not to gig? Stories from the modern economy. A nationally representative sample of 5,019 UK employed persons aged 18 to 70 was asked about the use of on-line platforms in the previous 12 months in December 2016.

163 Huws et al. (2016): Crowd work in Europe: preliminary results from a survey in the UK, Sweden, Austria and the Netherlands.

164 Although the sample has been weighted ex post to be representative of the national population as regards age, gender, region and working status, the results “cannot be generalized with complete confidence” to the entire population.

165 For 45% (from 58% in Austria to 33% in Sweden) it made up only 10% or less of all income.

to employ a worker under an employment contract, but requires the performance of

certain activities or for some specified work to be completed.

 Mini-job is a concept coined in Germany, which refers to employment contracts for earnings of currently up to EUR 450 per month with reduced income tax and social security contribution rates and entitlements.

 'On–demand work' involves a continuous employment relationship maintained between an employer and an employee where the employer has the option of calling in the employee as and when needed. There are employment contracts that indicate the minimum and maximum number of working hours, as well as so-called ‘zero-hours contracts’ that specify no minimum number of working hours, and the employer is not obliged to ever call in the worker.

A NNEX 5: V ISUALISATION OF THE P ROBLEM T REE A NNEX 6: I NTERNAL D RIVERS

This annex illustrates in detail the internal drivers described in section 2.2.2. The information in this section are based on ESPN synthesis and national reports on access to social protection for people in non-standard employment and in self-employment (2017), as well as on the report from Brodolini (2018, forthcoming) and on the report from Open evidence (2018, forthcoming). The information included reflects the information available at this stage. Further work with Member States will be undertaken to refine the picture. Member States where gaps were identified by experts are mentioned in the respective tables. However, there may be cases where gaps exist but have not been identified yet. Gaps in formal coverage and effective coverage are analysed for non-standard workers and for the self-employed. The annex indicates the branches in which major gaps of coverage by social security systems have been identified for non-standard workers and for the self-employed, and it shows as well the branches for which voluntary schemes are in place for the two groups.

The problems of transferability, transparency and behavioural biases are described for the two groups together. Regarding effective coverage, four criteria with a time dimension have been identified as obstacles on the basis of the information provided by the ESPN experts: long waiting periods, minimum qualifying periods, duration of benefits and minimum working period. Moreover, Member States in which transferability of accumulated social security rights across different social security schemes is not possible have been identified, as well as Member States in which there a lack of information/transparency regarding social security schemes.

Finally, problems of social security coverage have been also explored in relation to low enrolment rates observed in case of voluntary schemes.

• 1. 

  Regulatory biases

This annex explored gaps in the following branches of social security:

(a) unemployment benefits;

(b) sickness benefits;

(c) maternity and equivalent paternity benefits;

(d) invalidity benefits; (e) old-age benefits;

(f) benefits in respect of accidents at work and occupational diseases.

An initial, broader mapping 166 had shown that problems are concentrated in these areas.

1.1. Gaps in formal coverage

For the purpose of this Impact Assessment, gaps in formal coverage are identified in the cases in which people in non-standard employment or self-employment are not covered mandatorily by schemes in a given social security branch and do not even have the possibility to voluntarily adhere to them. Other than these regulatory deficiencies, in the case of voluntary coverage, behavioural choices may lead to people excluding themselves from formal coverage, which may be problematic. The sections below discuss how behavioural biases may play a role in such decisions and provide overview tables on the situation of formal and

166 See ESPN (2017).

effective coverage across the Members States and across all branches of social security. Indeed the extent of gaps in formal coverage differs significantly between the various types of non-standard workers and the self-employed, as well as between branches of social security.

Non-standard workers are usually formally covered by social security in the same way as standard workers. However, in some Member States specific categories of non-standard employees are not formally covered for some or for all branches of social security. The main groups for whom this is the case in a significant number of countries are casual and seasonal workers, as well as trainees and apprentices. Furthermore, there are a number of national categories to which restrictions apply in the concerned Member State, such as 'marginal parttimers' and 'marginal freelancers' in Austria, 'mini-jobbers' in Germany and 'civil law contracts for a specific task' in Poland. Table A.2 gives a detailed overview of access gaps in

Member States by social security branches and by category of non-standard work 167 .

Table A.2 - Lack of formal coverage to social security for people in non-standard employment

Social policy area/types Casual Seasonal National

of employment workers workers specificities Freelance Apprentices Trainees

Vocational trainees

b

Unemployment RO, HU, BG, RO, AT a , CZ , BE, EL, HR, EL, FR,

benefits MT, LT LV, HU, DE

c , PL d , MT, NL, PL IT, LT,

MT, LT SK e MT, NL, PL, RO

Sickness benefit HU, LT, HU, LT,

BE, HU, NL, DK, FR, DK, EL,

b

LV, RO LV, RO CZ , SI

d PL HU, LT, FR, HU, PL NL, PL

b

Maternity benefit LT, RO BG, LT,

CZ , BE, MT FR, HU, EL, FR,

LV, RO PL

d

, BG, FR IT, LT HU, IT UK h

Accident and

occupational RO, HR, BG, LT, b , ES f injuries LT LV, RO

CZ

CZ b , BE, HR, MT EL, FR,

Old age/survivors' HU

g , HU, IT,

pensions MT, LT

BG, HU,

RO, LT LU

g

, LT, MT MT h ,

PL d Invalidity HU, LT HU, LT AT a , PL d Note: a) Marginal part-timers; b) Agreement to perform a job; c) Mini-jobbers; d) civil law contracts; e) employees on "work agreement" with irregular income; f) domestic workers; g) on call jobs; h) temporary agency work.

Estimates for the size of the identified groups in the concerned Member States are not available in some cases. See Box A.2 for available figures. As regards the situation of the platform workers, recent research have tried to estimate the size of the platform economy and the extent to which it provides access to social security (see Box A.3).

Box A.2 – Extent of some new forms of employment facing gaps in formal coverage 168

Form of Extent

167 Information in the table is based on ESPN (2017).

168 The information is taken from the case studies commissioned for this study, see Annex 15.

employment

Mini-jobber Among the 7.6 million mini-jobbers, 4.9 million have mini-jobs as a main job, the remaining (DE) ones as a second job.

Mini-jobbers are a heterogeneous group. Pensioners, housewives and students and pupils are important sub-groups among the mini-jobbers having a main job as main occupation, see graph below.

Civil law Due to lack of data on civil law contracts, it is difficult to assess precisely the number of contracts (PL) people working under civil law contracts. According to LFS data, in 2016 there were 510,000

people working under civil law contracts (in their main job) which is equivalent to 3.1% of total employment in Poland. Among them, 429,000 people worked solely under a contract of

mandate, 36,000 people worked solely under a contract to perform specified work, 22,000 people worked under another type of a civil law contract (such as a managerial contract), and 22,000 combined various types of civil law contracts. Data from personal income tax records

published by the Ministry of Finance show that, in the last few years, approximately 1 million people has been earning income only on the basis of civil law contracts. The tax records also show that the number of such people has doubled since the early 2000s.

Casual workers As compared to 2011, when ca. 170.5 thousand day-labourers were registered, by 2016 their (RO) number increased 4.7 times, reaching 803.6 thousand persons. Their share in the total labour

force expanded from 1.2% in 2011 to 6.7% in 2016. The number of legal entities using daylabour also increased from 4.8 thousand in 2011 to 24.1 thousand in 2015.

Box A.3 – Social protection of platform workers 169

According to a recent survey of 1200 platform workers, lack of access to social protection is a major concern for

them. Nearly 70% of all platform workers report not having access to schemes that cover maternity, childcare, and housing benefits. 63% do not have access to unemployment benefits, whereas a significant proportion of platforms workers had been unemployed in the past five years. In the same vein, around half of platform workers do not have access to old age schemes (56%), to disability schemes (60%) and to sickness benefits (47%).

Lack of social protection was found to be especially severe among platform-dependent workers, i.e. those who

got more than half (50%) of their personal earnings through completing online micro tasks. They represented a quarter (25%) of the sample. However, gaps in social protection coverage were also significant among platform workers who have another form of employment and source of income: they were also significantly at risk of not having access to either unemployment benefit (61.2%) or sickness benefit (46.1%). These findings tend to

suggest that the corresponding individuals' "other job" was also a non-standard form of employment, offering limited access to social protection.

169 Forde et al. (2017): The Social Protection of Workers in the Platform Economy, report requested by the European Parliament's Committee on Employment and Social Affairs.

A key issue affecting the access of platform workers to social security is their classification as workers, selfemployed or other hybrid legal statuses 170 .

The self-employed as a group lack formal coverage regarding key branches of social security, notably regarding unemployment benefits, accidents and occupational injuries as well as sickness benefit in a number of Member States. In some Member States, different rules apply for certain categories of self-employed (e.g. in Germany, Greece, Finland, France and Italy) Notably, the self-employed generally have no mandatory formal coverage and cannot opt-into

voluntary schemes either for unemployment benefits in eleven Member States 171 , although

some specific categories of self-employed can be covered on a voluntary basis in some of

them 172 (Table A.3). The self-employed are also not formally covered by sickness benefits in three Member States 173 . Finally, the self-employed are also not covered in case of accidents

and occupational injuries in ten Member States, although some specific categories of selfemployed can be mandatorily covered 174 . It is also worth noting that in many branches and Member States, the self-employed have voluntary rather than mandatory formal coverage (see next section), which may lead to only a limited proportion of them being members of such schemes.

Table A.3 - Lack of formal social security coverage for the self-employed

Social security branch Member State

Unemployment benefits BE

a , BG, CY, DE, FR, IE, IT, LV, MT b , NL, UK b

Sickness benefit EL a ,IE b a , IT

Accident and occupational injuries BE, BG, CY, CZ, IE, LT, LV, NL, SK, UK

Note: The table reports in which branches and in which Member States at least one sub-group of the self-employed is excluded from formal coverage in the sense that they have no mandatory coverage and cannot opt-into voluntary schemes either. a) Only one or more sub-groups of the self-employed are not formally covered. b) In these Member States only means-tested benefits are available to the self-employed while they are excluded from contributory schemes.

Voluntary coverage

170 Forde et al. (2017): The Social Protection of Workers in the Platform Economy, report requested by the European Parliament's Committee on Employment and Social Affairs.

171 Malta and UK are included in this list since here the self-employed are only covered by means-tested non-contributory benefits, while they are excluded from contributory schemes.

172 This is the case of dependent self-employed and liberal professions in Italy. It should be noted that the Belgian situation is rather specific: while there is no possibility to be covered by unemployment insurance, there is a compulsory social insurance in the case of bankruptcy, the amount of which may be higher than the unemployment benefit paid to former salaried workers. In Germany, anyone who provides home nursing care for a dependent or is self-employed for at least 15 hours per week may make voluntary unemployment insurance contributions Moreover, there are tax-funded unemployment benefit II (basic income support for job seekers) applies to all persons of working age who are capable of work and who are not able to cover their living costs from their own resources

173 Ireland is included in the list because the self-employed are only covered by means-tested non-contributory benefits, while they are excluded from contributory schemes.

174 In the Netherlands, only the dependent self-employed are covered by accidents at work and occupational injuries benefits; in Spain, insurance for accidents at work or occupational diseases is voluntary for the self-employed and compulsory only for ‘dependent selfemployed' and mobile traders; in the Czech Republic people working under 'agreement to perform work/job' are not entitled or are only partially covered in case of accidents and occupational injuries; in Greece liberal professions and those providing services paid by receipt of business expenditure).

In several Member States, adherence to certain social security schemes is not mandatory for the self-employed and some categories of non-standard workers, but is available on a voluntary basis (see Tables A.4 and A.5). Voluntary schemes can be supplementary to the existing mandatory schemes, or they can apply only to categories of workers which are not covered by the main schemes.

Table A.4: Voluntary social security schemes for people in non-standard employment

Social policy area/types of Opt-in Opt-out employment

Unemployment benefits LV a , PT a , SE a , DK, FI a

Sickness benefit AT b , PL c

Maternity benefit AT b c , PL

Accident and occupational injuries PT a

Old age/survivors' pensions AT b , LT a , PT a , RO c DE b , NL c

Invalidity AT b , PT a , RO c

Note: a) supplementary schemes; b) if income below a certain threshold; c) For specific categories of non-standard employees.

Table A.5: Voluntary social security schemes for the self-employed

Social policy area/types of Opt-in Opt-out c employment

c

Unemployment benefits AT, DE , DK, ES, FI

a , FR c ,

RO c , SE a , SK PT, RO

AT c , BE a , BG, CZ, DK, Sickness benefit EE c , IE, IT, LU c a , FI , NL, AT b , PT, RO b , SK b , UK b

PL

Maternity benefit AT

c

, BG, CZ, DK a , LT, PL,

RO c AT

b , PT

c

Accident and occupational injuries AT , DE, DK, ES

c , FI, FR, LT, PT, RO

a

Old age/survivors' pensions AT

c

, BE, DE c , DK, EL , FI a , AT b , IE b , FI, PT, RO b , SK b , LU c , NL UK b

Invalidity AT c , DE c , NL AT b b , IE, PT, RO , SK b , UK b Note: a) additional schemes; b) if income below a certain threshold, c) for specific categories of the self-employed.

For Member States where voluntary schemes exist for all the self-employed or some groups, a low rate of enrolment can generally be observed. For instance, 5% of the self-employed are estimated to be covered by sickness benefits in Austria, 15.37% in Czech Republic, and almost no self-employed in Romania. As for old-age and survivors' pension schemes, 46% of self-employed are estimated to be covered in Belgium, only 650 self-employed in the Netherlands and 20% of the self-employed in Romania. Voluntary schemes for unemployment are also estimated to be under used: in Austria only 0.02% of the selfemployed is covered; in Spain less than one quarter; in Finland 20% of the self-employed without employees and 10% of the self-employed with employees; in Ireland 10037 selfemployed enrolled in a voluntary scheme 175 ; and in Romania only less than 1% is enrolled. Voluntary schemes for invalidity also result in low formal coverage: only 25% of selfemployed were estimated to be covered in 2013 in the Netherlands (against 30% in 2006) and in Romania only 20% 176 .

1.2. Gaps in effective coverage and transferability

People in non-standard employment or in self-employment may be mandatorily formally covered or have the legal possibility to opt for access without, however, achieving effective coverage in the sense of de facto being able to build and take up adequate entitlements to benefits.

This may be due to rules governing the contributions and entitlements which are more difficult to meet for people in non-standard employment or self-employment. This may be the case because the requirements are stricter for these groups than for standard employees. Additionally, it may in some cases also be more difficult for them to fulfil identical requirements to those for standard employees. For instance, temporary employees are more frequently affected by unemployment, making it more difficult for them to accumulate necessary qualifying periods for unemployment insurance 177 .

Based on a more comprehensive mapping exercise, the criteria most frequently used by experts and thus used to identify gaps in effective coverage are:

• - 

  Long waiting period, i.e. particularly long length of time between the occurrence of the risk and the receipt of the benefit.

• - 

  Minimum qualifying period, i.e. the minimum period that people in non-standard employment and self-employed have to contribute to a scheme in order to start receiving the benefit.

• - 

  Duration of benefits, i.e. the length of time during which the individual receive the benefit, in case this differ from that of standard workers.

• - 

  Minimum working period, i.e. if the person is required to have worked a minimum of hours/months/years in order to be eligible to receive the benefit in case of need. This criterion is stricter than the minimum qualifying period, since a person can be, for instance, unemployed, but have the possibility to still contribute to the system, while when a minimum working period is required the person not employed nor in self-employment is not eligible for the benefit.

In addressing these issues, it is important to recognise that some of these criteria play a key role in the design of social security systems (see Box A.4 below).

Another problem affecting effective coverage relates to the income assessment base. This can be problematic in the case of high heterogeneity of rules among schemes which can determine a lower level of entitlements.

175 This data for Ireland refers to 2015, and it has to be noticed an improvement with respect to 2008 when only 2547 self-employed where adhering to a voluntary scheme for unemployment.

176 ESPN (2017): National reports and synthesis report.

177 European Commission (2017): Employment and Social Developments in Europe review, Luxembourg: Publication Office of the European Commission, p. 73.

Finally, a lack of transferability when professional mobility (including between employment and self-employment) implies changing social security schemes can also determine gaps in effective coverage, as it does not allow the person to preserve the accumulated entitlements or make already fulfilled time thresholds count in the new scheme.

Box A.4 - Effective coverage and incentive compatibility of social security schemes

The analysis above has identified several features of social security systems and their rules which may prevent non-standard employees and the self-employed from having effective coverage, notably including minimum time requirements such as waiting periods and qualifying periods.

It is important to highlight that these features may fulfil an important role in the design of social security systems. They may help dealing with the information asymmetries between insurance clients and providers that complicate or impede the provision of such insurance in private markets 178 . Indeed the economic rationale for providing social insurance is closely linked to the existence of such information problems.

For instance, a lack of information about characteristics of a possible insurance taker which influence the ex-ante cost of providing insurance may hamper the functioning of insurance markets (adverse selection). State intervention in the form of mandatory membership is one solution to such problems. Similarly, a lack of information about the insured person's behaviour after taking out insurance which influences the cost of providing the insurance may limit the provision of such insurance (moral hazard). For instance, unemployment insurance may not be provided in private markets when insurers cannot observe the efforts of the unemployed to re-enter employment.

Some of the obstacles identified with regard to effective coverage may in fact be crucial in ensuring that (prospective and actual) members of social security schemes have incentives to comply by the system's rules (including truthfully revealing their private information). For instance, setting a qualifying period in unemployment insurance is a standard means of preventing self-selection. Both the insurance provider and the insurance taker know relatively little about what might happen in the future (after the qualifying period), whereas over a shorter time horizon, the insurance taker has an information advantage about issues influencing his risk and thereby the potential cost for the insurance, including e.g. the financial status of the firm.

EU policies to address these obstacles cannot specify levels for standards or thresholds. Such detailed provisions are in the competence of Member States and will inter alia depend on the national institutions and broader cultural and socio-economic context. Rather, the aim is to identify unduly high obstacles which are not justified by policy objectives, notably including avoiding abuse of the system. This entails scrutinising existing differences in rules between standard employees on the one side and people in non-standard employment and self-employed on the other. Likewise, where the same rules applied to all groups lead to poorer outcomes for people outside standard employment, there might be scope to better tailor the rules to the situation of the specific groups while remaining compatible with policy objectives.

In this context, it is interesting to note that the Council of Europe's European Code of Social security 179 mentions thresholds should not be more demanding than necessary to avoid abuse. It does not define what levels would be compatible with this. Instead, in the context of corresponding reporting the supervision mechanism of the Code would be allowed to determine whether or not certain national stipulations are excessive.

Rules related to time thresholds

As shown in table A.6, despite being formally covered, people in non-standard employment can lack effective coverage because the same rules and conditions as the standard employees apply also to people in non-standard employment, without taking into consideration the specificities of non-standard contracts and disadvantaging people working under this employment relationship. Long waiting periods to receive sickness benefits have been identified as an issue in two Member States and to maternity benefits in five Member States. A minimum qualifying period has been identified as an issue in case of unemployment in six

178 See for instance N. Barr (2012): The economics of the welfare state. Oxford University Press, for more details.

179 See online .

Member States, in three countries concerning old-age, in six countries concerning sickness benefits, in one in case of maternity and in two concerning invalidity. Minimum working periods have been identified as an issue in four Member States for unemployment benefits, in two for old-age benefits, in one for sickness benefits, maternity benefits and invalidity. Finally, the duration of unemployment benefits is shorter than for standard employees in one Member State.

Table A.6: Time thresholds limiting effective coverage for the non-standard workers

Social policy Minimum

area/types of Waiting period qualifying Duration of Minim working employment period benefits period

Unemployment SK, CZ, PL, EE,

benefits HU, ES BG EL, PT, SE, HU

Sickness benefit BE, EE IT, PL, AT, CZ, BG PT

Maternity benefit AT, CZ, HU, MT, HR, IT BG PT

Accident and occupational injuries

Old age/survivors'

pensions IT, NL, PT NL, IT

Invalidity CZ, PT CZ

Effective coverage can be problematic also for the self-employed despite being formally covered, because the eligibility conditions or rules are stricter than for standard workers. Waiting periods can be longer for the self-employed than for standard employees in case of unemployment in one Member State, and in eight Member States for what concerns sickness benefits. Or, the same waiting period can apply to self-employed and dependent employees. However, the latter are entitled to sick pay (i.e. they continue to receive their salary by the employers), which is not applicable to the self-employed for the nature of their working relationship, creating an effective difference in treatments. A minimum qualifying period in order to receive unemployment benefits has been identified as an issue in four Member States, in two for old-age and survivors' pensions and in one for invalidity benefits. Minimum working periods have been identified as an issue in three Member States concerning unemployment benefits, in two for old-age and survivors' pensions, in one for maternity and for invalidity and in one for accidents at work and occupational injuries. The duration of benefits is shorter for the self-employed with respect to standard employees in one Member State for unemployment, in two Member States for sickness benefits, in five Member States for maternity and in one for accidents at work and occupation injuries (see Table A.7).

Table A.7 - Time thresholds limiting effective coverage for the self-employed

Social policy Minimum

area/types of Waiting period qualifying Duration of Minimum employment period benefits working period

Unemployment

benefits PL EL, FI, LU, SK EL, BG EE, PT, SE

Sickness benefit PT, SI, BE, EE, FR, HR, PL, SE AT, PT

Maternity benefit EE, EL, FR, BE, MT DK

Accident and

occupational injuries EL EL

Old age/survivors'

pensions RO, UK RO, UK

Invalidity DE DE

1.3. Gaps in transferability

The risk of gaps in effective coverage for people in non-standard employment or selfemployment does not only depend on the difficulties to fulfil eligibility criteria, but also on problems related to the transferability of social rights across schemes. People in non-standard employment and in self-employment can have problems in transferring their rights from a scheme to another, especially in Member States with highly fragmented social security schemes.

For our purpose, transferability relates solely to transfers across employment statuses and sectors, not across risk areas. This means that transferability is to be considered in situations where a person changes employment status (notably between dependent employment and selfemployment) or economic sector and is covered by a social security scheme (scheme A) before and by a different one (scheme B) after the change, and where eligibility for a corresponding benefit (in scheme B) or the level thereof depends on past contribution history (length and/or contribution levels).

Lack of transferability is an issue for those branches of social security for which a minimum qualifying periods is required (this is especially the case of old-age pensions' schemes – see box A.5). In these cases, transferability can formally not be foreseen (i.e. there are no rules or laws which foresee this possibility), or there can be a problem related to effective transferability if the conditions are unfavourable (e.g. particularly high administration burden or costs). For example, in Italy, it is possible to accumulate the pension contributions made to different funds for receiving a single pension benefit. It is also possible to transfer the contributions made through the working life to a single fund, but this may be particularly expensive for the claimants of the benefit. In France, a note of the Pension Advisory

Council 180 has shown that, even if in principle supplementary pension schemes are

transferable, in practice, transferability only operates for contracts that obey to the same fiscal and social contributions treatment.

Table A.8: Difficulties in transferability of social security rights

Social policy area/types of Not foreseen by the High costs Others employment law

Unemployment benefits EL, PT IT

Sickness benefit AT, PL, PT IT

180 See the report available online .

Maternity benefit AT, PL, PT IT

Accident and occupational PL, PT IT injuries

Old age/survivors' IT FR pensions

Invalidity PL IT

Note: the information have been collected by MISSOC experts for 21 out 28 EU Member States. Belgium, Bulgaria, Denmark, France, Romania, Sweden and Slovakia are not included.

Box A.5 - Transferability of pension rights

Old-age pensions are accrued over several decades, therefore being able to retain and follow their pension rights through multiple labour market transitions can be both challenging and crucial for individuals' social protection in old age. Transferability and transparency of pension rights are important elements of improving pension outcomes for persons with diverse careers.

Statutory pension systems can streamline pension insurance throughout career changes by crediting all contributions paid by/for an individual to a single account, irrespective of the economic status or the type or duration of contract under which they were paid. This is already the case in the majority of Member States, though some public pension systems retain parallel regimes for different economic statuses.

Occupational pension rights tend to be more fragmented than public pensions, as individuals may be required to join a new scheme each time they change jobs. EU law does not require the transferability of occupational pension rights; however encourages Member States to improve it 181 . Following a Call for Advice from the Commission, EIOPA aimed to gather information on the transfers of assets within the countries and across borders 182 . The available information suggests that at least in several Member States 183 domestic transfers of pension rights between occupational pension funds are a relatively widespread practice 184 .

Wherever a transfer is not possible, the pension rights of former workers should be preserved, e.g. through indexation, on par with the rights of active scheme members, as provided for in the Supplementary Pension Rights directive 185 .

1.4. Adequacy of the benefits

While lack of adequacy of benefits may be problematic for all forms of employment, it can be even more challenging for people in non-standard employment or in self-employment, which can "experience higher […] adequacy risk" 186 . This may importantly reflect income differences between groups, addressing which is beyond the scope of the current initiative, but can also be related to i) for the self-employed, the income assessment base taken into account for the calculation of benefits; ii) the length of contributory history; and iii) different rules for the calculation of benefits for different forms of employment (e.g. benefits can be defined as a lump-sum rather than in relation to earnings, as for salaried employment). Finally, differences in adequacy between groups could arise if at progressivity and reductions in social contributions for low-income groups do not apply equally to workers and the self-employed.

181 Directive 2014/50 i/EU of the European Parliament and of the Council of 16 April 2014 on minimum requirements for enhancing worker mobility between Member States by improving the acquisition and preservation of supplementary pension rights, recital 20.

182 Final Report on Good Practices on individual transfers of occupational pension rights, EIOPA, 2015.

183 AT, BE, DK, ES, NL, PL, PT.

184 During the EIOPA mapping exercise, it became apparent that quantitative data on transfers were not readily available in a large number of Member States.

185 Directive 2014/50 i/EU directly applies only to outgoing workers who change jobs across borders, however, all Member States have opted to extend the same legal treatment to all workers who leave an occupational pension scheme. The Directive does not apply to self-employed persons.

186 For a review of the case studies, see Annex 15.

Income assessment base

Income assessment bases play a salient role regarding the accrual of entitlements to benefits for the self-employed. On the one hand, if the self-employed person pays social contributions on a very low income assessment base, s/he will receive a low level of benefits. On the other hand, if the reference base is too high, s/he may not be able to pay the required contributions. These issues may be related to the way of calculating the income base or reference income periods. Moreover, requirements for advanced payments may also hinder the self-employed person’s ability to build up entitlements. Income assessment bases calculated on long previous periods of earnings may not take into account the fluctuations characterising incomes from self-employment. This kind of estimation-based system may hinder the ability of the selfemployed to pay social contributions if their current income decreases below the estimated one. At the same time, the freedom to choose (within certain limits) the level of contributions in some schemes can lead to low adequacy of benefits (see Box A.6 for an example).

Box A.6 – Dependent self-employment – an example from Spain 187

In Spain, the figure of economically dependent self-employment (trabajador autónomo económicamente dependiente, TRADE) was regulated in 2007 in the Self-employed Workers’ Statute (Ley del Estatuto del Trabajador Autónomo, LETA). A self-employed worker is considered to be economically dependent if the income from a single client accounts for at least 75% of his/her total income 188 . However, the self-employed in Spain, including the category of dependent self-employed face adequacy problems as the amount of all cash benefits, in particular for pensions, that TRADEs are entitled to is determined by the contribution base and the monthly contribution. In this sense, attention should be drawn to the fact that, as opposed to employees, selfemployed and TRADE can choose the contributions they want to pay to the Social Security. Each selfemployed worker determines their contribution base and the monthly contributions.

Overall, figures on the contribution base chosen by self-employed show that most of them are selecting the minimum base in order to pay lower contributions. According to Social Security Statistics on Self-employed (2017, 2nd quarter), 86% of the self-employed workers chose the minimum contribution base. For the selfemployed with employees, the percentage of workers choosing the minimum contribution base is only a bit lower (81%).

For TRADE, the percentage of workers choosing the minimum contribution base is even higher (91% in the 3 rd quarter of 2017. Self- Employed Workers ‘RETA scheme’).

As social security is closely linked to the contribution bases, these lower contributions paid by self-employed, including TRADE, leads to lower levels of protection and, especially, lower pensions, with an increased risk of poverty rates.

According to the Social Security Statistics, in September 2017 self-employed received an average monthly pension of €711.13 compared to 1,211.57 for employees under the General Social Security Scheme. Thus, selfemployed pensioners receive, on average, €500 less than employees.

Contributory history

Contribution histories can matter not only for effective coverage (via minimum requirements for their length, see the section above) but also for the adequacy of benefits in some social security branches. For instance, in some Member States the time profile of replacement rates in unemployment insurance depends on the contribution history. At the same time, the latter is

187 The information in this box is based on the "Case study on economically dependent self-employment (TRADE)". For a review of the case studies, see Annex 15.

188 Article 11, paragraph 1 of the Self-employed Workers’ Statute.

very much linked to the form of employment of the individual (e.g. full-time vs part-time or temporary vs open-ended contracts).

Calculations of replacement rates for individuals with the OECD tax benefit model for unemployment benefits as well as housing and family benefits illustrate the major role of the contribution history in determining the amount of benefits and their evolution over time in some Member States (see Table A.9 below). For instance in Belgium an individual who

contributed one year will have a total resulting replacement rate at the 18 th month of

unemployment which is 24.8 percentage points lower when compared to an individual who contributed five years. In some Member States the gap is large already after two months of receiving the benefits (i.e. Bulgaria, Lithuania, Hungary, Romania and Slovakia). Shorter contribution histories are likely to characterise workers on temporary contracts, to the extent that they also face more frequent interruptions by unemployment spells.

Table A.9: Replacement rate for unemployment, housing and social assistant benefits. Comparison 5 vs 1 year of contribution, 2016

5y 1y

2nd 6th 12th 18th 2nd 6th 12th 18th month month month month month month month month

BE 87.8% 87.8% 81.8% 81.8% BE 87.8% 87.8% 81.8% 57.0%

BG 76.5% 76.5% 15.2% 15.2% BG 30.6% 15.2% 15.2% 15.2%

CZ 77.2% 56.3% 48.6% 48.6% CZ 77.2% 56.3% 48.6% 48.6%

DK 84.5% 84.5% 84.5% 84.5% DK 84.5% 84.5% 84.5% 84.5%

DE 59.0% 59.0% 59.0% 41.3% DE 59.0% 59.0% 41.3% 41.3%

EE 61.3% 49.0% 40.6% 40.6% EE 61.3% 49.0% 40.6% 40.6%

IE 72.3% 72.3% 72.3% 73.2% IE 72.3% 72.3% 72.3% 73.2%

EL 40.1% 40.1% 40.1% 22.3% EL 40.1% 40.1% 7.8% 22.3%

ES 77.3% 77.3% 54.4% 54.4% ES 77.3% 30.3% 30.3% 30.3%

FR 68.2% 68.2% 68.2% 68.2% FR 68.2% 68.2% 68.2% 47.5%

HR 73.4% 36.7% 31.5% 31.5% HR 73.4% 31.5% 31.5% 31.5%

IT 73.5% 69.8% 56.0% 42.2% IT 73.5% 69.8% 56.0% 42.2%

LV 68.8% 51.6% 36.9% 36.9% LV 68.8% 51.6% 36.9% 36.9%

LT 74.3% 49.3% 24.4% 19.5% LT 24.4% 24.4% 24.4% 19.5%

LU 84.0% 84.0% 84.0% 61.9% LU 84.0% 84.0% 84.0% 61.9%

HU 67.7% 18.5% 18.5% 18.5% HU 18.5% 18.5% 18.5% 18.5%

MT 53.2% 53.2% 53.7% 53.7% MT 53.2% 53.2% 53.7% 53.7%

NL 71.2% 66.4% 66.4% 66.4% NL 71.2% 66.4% 66.4% 66.4%

AT 55.0% 55.0% 51.9% 51.9% AT 55.0% 51.9% 51.9% 51.9%

PL 56.9% 52.1% 52.1% 33.4% PL 56.9% 52.1% 52.1% 33.4%

PT 75.0% 75.0% 67.5% 43.5% PT 75.0% 75.0% 23.5% 23.5%

RO 34.0% 34.0% 10.6% 10.6% RO 28.1% 28.1% 10.6% 10.6%

SI 82.5% 68.6% 47.3% 47.3% SI 82.5% 47.3% 47.3% 47.3%

SK 62.5% 62.5% 24.1% 24.1% SK 24.1% 24.1% 24.1% 24.1%

FI 71.6% 68.0% 68.0% 68.0% FI 71.6% 68.0% 68.0% 68.0%

SE 73.5% 65.0% 65.0% 61.1% SE 73.5% 65.0% 65.0% 61.1%

UK 51.6% 51.6% 51.6% 51.6% UK 51.6% 51.6% 51.6% 51.6%

Source: European Commission, elaborations on the basis of the OECD tax benefit model.

Box A.7 – Adequacy of pensions for the self-employed

The self-employed are characterised by high income inequality and high income poverty risk during working life (three times that of employees) but also relatively higher assets 189 . SHARE survey data 190 allow measuring the

impact of self-employment on retirement income and living standards by comparing the situation of retired selfemployed

191 to that of retired employees.

Retired self-employed have lower incomes than retired employees in almost all of the 19 countries where a comparison is available 192 (Figure A.15) 193 . The income gap, together with higher income inequality 194 , leads to

substantially higher monetary poverty for retired self-employed. Their at-risk-of-poverty (AROP) rate is around

twice as high as that of retired employees on average 195 .

Figure A.15: Median equivalised disposable income by former work type (a), and AROP (b) by former

work type, 2017

Source: SHARE (provisional data); EU* does not include Ireland, the Netherlands and the UK. Sample sizes in Bulgaria, Latvia, Lithuania, Portugal, Romania and Slovakia do not allow distinguishing between retirees.

Different definitions of benefits for different forms of employment

In some Member States many self-employed may receive lower benefits compared to an

employee with the same earnings because benefits are set at different levels according to the

form of employment.

Cases identified by experts 196 include that in Belgium, the amounts of sickness benefits paid

to the self-employed are equal to the minimum amounts paid to white-collar employees after

189 See also Section 2.3.1.

190 Survey of Health, Ageing and Retirement in Europe (SHARE). Data source: SHARE wave 7, 2017.

191 For the purposes of this comparison, "retired self-employed" are defined as retirees who have spent 50% or more of their working careers

as self-employed.

192 Comparison available for AT, BE, HR, CY, CZ, DK, EE, FI, FR, DE, EL, HU, IT, LU, MT, PL, SI, ES, SE. Sample sizes in BG, LV, LT, PT, RO, SK do not allow for an income comparison. Survey excluded IE, NL, UK.

193 The highest income gaps can be observed in Luxembourg, Denmark and France, and in eight out of the 19 countries the gap is over 20 percent. Only in Hungary and Estonia retired self-employed enjoy (slightly) higher incomes than retired employees.

194 The S80/S20 ratio is almost twice as high among retired self-employed as among retired employees.

195 The largest gaps are in Denmark, Poland and Malta. Only in the Czech Republic and Estonia the self-employed have lower AROP rates. 196 See Annex 6. The tables are based on the information provided by ESPN and MISSOC experts.

six months of sickness. As a result, the adequacy of sickness benefits in cash is lower for the self-employed than for standard workers. Moreover, projections show a large decline in replacement rates for the self-employed by 2040. In 2010, a self-employed worker had a gross replacement rate of 77.2% and the rate for a salaried worker was 73.7%. In 2040, the figures would be respectively 47.3% for self-employed and 61.7% for employees 197 .

In Denmark, the self-employed receive lower unemployment benefit than people in salaried employment, with a shorter period of receipt of benefits. In Finland, on average, an employee earning €2,500 would get a monthly unemployment benefit of €1,500 (contribution-based), whereas the basic monthly allowance for the self-employed, with the same earnings, would be €697.

In Finland, compensation in the case of a work accident would be about €2,500 for the employee and €1,750 for the self-employed, or if the person is not eligible for income-related benefits, the universal minimum allowance would be approximately €622 a month.

1.5. Lack of transparency of entitlements and user-friendly information related to social security

A further aspect that should be taken into account when dealing with the problem of lack of social security coverage for people in non-standard employment and in self-employment concerns the issue of transparency and complexity of administrative procedures. The lack of easily accessible and clear information about rights and obligations relating to social security can hamper a clear understanding of working conditions and future risks. This may be of particular relevance for people in non-standard employment and self-employment who may face more frequent transitions and/or be covered by different set of rules. This situation is often produced by the complexity of social security schemes and the lack of easily accessible information systems, and is potentially compounded by the limited representation of the selfemployed and non-standard workers in trade unions. A recent study identifies as causes of non-take-up of social benefits the inaccessibility and complexity of the procedures, the slow assessment, and the unawareness of people concerning benefits, entitlements and application procedure 198 .

Table A.10: Lack of information regarding social security schemes

Social policy area/types of

employment Generic information

Personalised information (based on a case)

Overview of rights Overview of rights Simulation (with and National and parameters you

responsibilities Institution locator responsibilities enter or select online)

BE, BG, HR, CY, BE, BG, HR, CY, Unemployment CZ, DK, FI, EL, EE, FR, EL, IT, LV,

benefits HR HU, IT, LT, MT, LU, MT, PL, PT,

SK, ES, SE, UK RO, SI, SE

Sickness benefit, AT, BG, HR, CY, BE, BG, HR, CY, Health, maternity, UK HR, DK, UK CZ, DK, FI, EL, EE, DK, EL, HU,

sickness HU, IT, LT, MT, IE, IT, LV, LU, NL, SK, ES, UK MT, PL, PT, RO,

197 ESPN (2017).

198 Eurofound (2015), Access to social benefits: Reducing non-take-up, Publications Office of the European Union, Luxembourg.

ES, UK AT, BG, HR, CY,

Old age/survivors' BE, BG, HR, CY, pensions EL, MT EL, MT

CZ, FI, DE, EL,

HU, LT, LU, MT, EE, IT, LV, LU,

SK, ES, SE, UK MT, PT, SE

Table A.10 above 199 provides a mapping about the availability of information regarding key

social security systems in the Member States. Generic information about social security schemes is provided in most Member States, for instance via dedicated websites, although the information may not always be available broken down by employment status or contract. By contrast, the table reflects a wide-spread lack of transparency regarding personalised information, both regarding an overview of rights and obligations and the possibility to obtain information through online simulation tools. Such information is only available in about half of the Member States. Additionally, Box A.8 provides a more detailed overview regarding transparency in relation to pension rights.

Box A.8 - Transparency of pension rights

While all Member States have put in place legislative provisions that ensure the right to information about the rules governing pensions and, increasingly, the state of individual entitlements, the approach and level of provided information varies significantly from country to country.

Better transparency and easy ways to access information about entitlements earned in different economic statuses can help individuals manage their careers and build up adequate pension rights more easily. All Member States have put in place legislation granting the right to information about pension rules and, increasingly, the state of individual entitlements. While some make it available from the early stages of working life, others only start providing it closer to pensionable age. Some Member States 200 pro-actively circulate annual pension statements to all socially insured persons, while others provide it less often, and some do it only upon request 201 . In most Member States, socially insured persons are entitled to individual information about the prospective pension amount, while some 202 communicate two or more scenarios of the prospective retirement income.

EU legislation on occupational pensions 203 provides that both current and former scheme members have the right to information about the value and treatment of their rights. However, to former scheme members information has to be provided only upon request. In practice, individuals may struggle to identify the pension fund associated with a former job, let alone keep track of their entitlements.

To help individuals follow their fragmented pension rights, about half of Member States have already set up pension tracking services, online platforms giving users one-stop access to information about their various pension rights, with more under way. The scope of schemes covered by tracking services varies, including some that bring together information about all pillars of the pension system 204 . Online platforms can also provide additional tools such as on-line pension calculators 205 , allowing interested users to project the eventual impact of different career developments on their future benefits.

A recent survey from Open Evidence shows the level of awareness of social security rights across people in different types of employment (Figure A.16) 206 . In general, the share of

199 The table is based on results from an unpublished Feasibility study on a European Mobility Portal on Social Security (EMPSS) commissioned by DG Employment.

200 DE, NL, PL, SE.

201 BG, HR, LU, LT, UK.

202 DE, FR, NL, PL, SE.

203 Directive 2014/50 i/EU and Directive (EU) 2016/2341 of the European Parliament and of the Council of 14 December 2016 on the activities and supervision of institutions for occupational retirement provision (IORPs).

204 DK, EE, SE.

205 FI, LT, SE, SI, SK.

206 OpenEvidence (forthcoming).

people unaware of their coverage is significantly higher among people which are not in

standard employment. In general, it is 7% among full-time open ended employees, but 18%

among own-account self-employed. Maternity benefits seems to be the ones for which people

are more aware despite the status and the type of employment.

Figure A.16: Share of respondents unaware of their social security coverage, by type of

employment

Source: Open Evidence

Invalidity and accidents at work and occupational injuries benefits seem to be the most

problematic schemes in terms of transparency for non-standard workers, followed by oldage

pensions' rights and unemployment benefits. 15% of temporary full-time workers are not

aware of whether they have coverage regarding sickness benefits, while the percentage is

lower in case of part-time (both open-ended and temporary). 16% of temporary part-time also

seems unaware of whether they have coverage with respect to unemployment benefits (while

the share both for open-ended part-time and temporary full-time workers is 13%).

Awareness concerning coverage regarding accidents at works and occupational injuries, and

invalidity benefits is scarce also among the self-employed (21% for the self-employed

without employees and 18% for the entrepreneurs are not aware of their entitlements). The

share of those declaring to be unaware of whether they are entitled to old-age pensions is also

high in the case of self-employed: 19% for the self-employed without employees and 17% for

the entrepreneurs. Lack of awareness is also detected concerning unemployment benefits

(again the percentage of self-employed that doesn't know if it is entitled to unemployment

benefits is higher for the self-employed without employees, 17%, than for the entrepreneurs,

14%).

• 2. 

  Behavioural biases

In the case of voluntary access to social security schemes, it is individuals' behaviour and choice that determine whether they have formal coverage or not. A number of factors can drive the decision not to adhere to systems, including whether coverage is already available through the first job in the case of voluntary access in the second job, the family context and private wealth accumulation which provide alternative means of coping with risks, as well as the individual's assessment of future risks facing them. This section explores how behavioural biases may contribute to low coverage rates observed in social security systems under voluntary access.

Aside possible financial disincentives influencing take up in voluntary schemes, individual preferences can also play a significant role in the way people adhere to existing social security schemes.

Individual's intertemporal distribution of consumption could be called myopic when the allocation of consumption to the present does not take sufficiently into account possible future risks. The tendency may be attributable to an erroneous foresight, or to a well-grounded optimism on the future, or to the fear that what has been promised for tomorrow might not

materialise 207 . Researchers estimate that correctives for inadequate savings due to myopia

could raise individuals’ lifetime welfare to an extent equivalent to almost an additional year’s

worth of income 208 . Already in 1958, Samuelson found that individual myopia may produce

suboptimal saving rates.

Myopic behaviour is increasingly identified by the academic literature to be a significant determinant of individual choices and is thus considered as a rationale for mandatory social

security. 209 Especially concerning protection against longevity risk, workers are forced to save

for retirement in order not to spend their accumulations without sufficient consideration for the future and thus end up in poverty.

Myopia can also limit the voluntary enrolment in social security schemes by people in nonstandard employment and by the self-employed. Myopia can worsen individual judgment about future social risks and protection needs. Without comprehensive and clear information, individuals could prefer to forgo or reduce contributions when this possibility is allowed 210 .

Some results of the recent survey funded by the Commission can be interpreted as indication that a significant share of individuals can have this kind of myopic preferences, not giving enough importance to future risks (Figure A.17; for a full overview of the survey results see Annex 16).

Figure A.17: Importance of getting prepared for retirement now by current main employment status

207 Pollock (2001): Environmental economics.

208 Laibson (1997): Golden Eggs and Hyperbolic Discounting. The Quarterly Journal of Economics 112 (2): 443–77.

209 Barr (2012): The Economics of the Welfare State. Oxford university press.

210 The decision not to adhere to a voluntary scheme can be also driven by financial reasons beyond individual preferences, in particular when non-standard employees and the self-employed with irregular and low revenues cannot afford to pay social contributions associated with enrolment in a voluntary scheme.

Source: Open Evidence 2017

Individual preferences and behaviour also play a role in the assessment of the income base:

the way in which the income assessment base is calculated can act as a disincentive to

properly contribute or to legally declare the income. Where the rules for declaration of income

are less stringent, or where the individual has the option to (legally) choose what level of

income to be assessed upon, individuals may present the same myopic attitudes and erroneous

foresight and lead them to under-report their present income. This under-reporting of income

and the corresponding underpayment of contributions may prevent effective coverage or

reduce adequacy (e.g. Bulgaria, Hungary, Estonia, Lithuania, Latvia, Romania, and

Slovakia) 211 . For example, in Romania, due to both high contributory rates and the method of

calculation of the assessment base only few self-employed contribute to social insurance.

Indeed, the self-employed represent the 17.6% of employed people, but 90% of them are not

covered for old-age pensions, invalidity, sickness or maternity benefits because of underinsurance

212 .

211 Under-declaration of income could be related to informal work (e.g. Bulgaria, Estonia, Hungary, Lithuania, Latvia, and Romania).

212 As mentioned, sickness and unemployment insurance are voluntary in Romania, while pension and health insurance are compulsory for

earnings exceeding the minimum insurable threshold; this represents 35% of the average gross salary at the national level per month (i.e.

592€).

A NNEX 7: O VERVIEW OF RECENT REFORMS IN M EMBER S TATES

Member Relevant Reforms State

Austria introduced several reforms aimed at increasing the coverage of public social insurance concerning self-employed and people in non-standard employment.

 Extension of social protection for non-standard workers

         In the second half of the 1990s a first wave of measures was decided to (partly) integrate so-called “new self-employed” and “freelance contracts” into social insurance and for people in marginal part-time employment the possibility to opt-in to health and pensions insurance got introduced. This option the facto creates a very low-cost opportunity to obtain health insurance plus times of pension insurance. At the same time regulations on marginal part-time employment got reformed so that income from such contracts, in the case of multiple jobholders, has to be added to other income from dependent employment or Freelance Contracts. And once the lower earnings limit is met, income from such contracts leads to coverage by health and pensions insurance, with related insurance contributions to be paid. Furthermore, as from January 2008, freelance contractors are covered by unemployment insurance and the severance pay

Austria scheme and since then they also have the right to receive sick-pay from the employer.

 Extension of social protection for self-employed

         Regarding the measures to foster the transition between unemployment and selfemployment, in Austria the level of support provided during the first months of selfemployment corresponds to that of the applicable unemployment benefit or unemployment assistance/subsistence allowance, including any family supplements.

         Paradigm shift has been identified in Austria where new self-employed have been included in pension, health and work accident schemes since 2000. The self-employed registered with the Chamber of Commerce (the vast majority of all the self-employed) are compulsorily insured. Since January 2009, the self-employed can opt-in to

unemployment insurance 213 . Recent reforms aim at addressing the issue of partial

access to entitlements to sickness benefits by extending social protection to the selfemployed. Further on, Austria has distinguished the 'dependant self-employment' by introducing specific provisions to extend some protection.

         In some countries, even if the self-employed are compulsorily covered by sickness protection, they are subject to much longer waiting periods than salaried workers. In order to tackle this issue, the self-employed in Austria have a right to opt into insurance covering this waiting period.

 Reducing legal conflicts about self-employed

         As a step taken to deal with dependent self-employment, in 2006 Belgium adopted new legislation on the nature of labour relations to ensure legal security and to fight the

Belgium phenomenon of bogus employment.

 Transferability of rights and entitlements

       Mechanisms allowing a person to include contribution periods of previous employment in the calculation of unemployment benefits.

213 However, given the rather strong “principle of equivalency” (i.e. the link between duration and amount of contributions paid and the level of social transfers from social insurance) within the Austrian system of social insurance outcomes in terms of social security and social inclusion for these groups tends to be problematic more often than for “normal” employees. The principle of equivalency quasi re-produces inequality of earned income, whereby self-employed and people in non-standard employment are more often affected by in-work poverty than normal employees. However, to the best of our knowledge, no detailed assessments are available about the concrete financial implications of these reforms. Integrating these forms of employment into social insurance made them more costly for employers, by this reducing the incentives to replace standard employment by non-standard forms of employment. However, marginal part-time employment and New Self-employment continued to rise irrespective of (partial) coverage by social insurance. In the case of Freelance Contracts new regulation, getting effective as from 1st January 2008, inter alia integrating them into unemployment insurance, appears to have had a mitigating effect, coming along with an overall reduction of such contracts. However, it may be the case that Freelance Contracts got increasingly substituted by New Self-employment, which continued to rise.

 Extension of social protection for self-employed

         When it comes to reforms aimed at extending social security for the self-employed, both 'parametric' and 'paradigmatic' reforms have been identified in Belgium. As changes in some parameters of the scheme, measures introduced include greater flexibility to take up maternity leave, and more favourable conditions for selfemployed carers of dependant persons. Recently, a number of measures have been adopted to improve the social status of self-employed persons as well as to promote female entrepreneurship. For instance, as of 1 January 2017, maternity leave has been expanded by another 4 weeks (12 weeks in total).

         Further on, the process of harmonisation started in the 2000s. Due to social policy measures focused on developing a fully-fledged social status for self-employed persons, there has been a net improvement in the adequacy of some benefits. For instance, the level of the minimum old-age pension for self-employed persons shows a strong increase over the last decade, reaching the equal level to the one for employees in August 2016. The self-employed have been included in compulsory state insurance for most social risks: healthcare and sickness benefits, long-term care, invalidity benefits, maternity benefits, family benefits, old-age pensions, survivors’ pension.

Regarding the 'bogus self-employment', in 2006 a new legislation on the nature of labour relations was adopted in order to ensure legal security and to fight effectively against the phenomenon. Moreover, concerning defining and assessing the reference income for social security contribution payments, the Belgian ESPN experts highlight a best practice: a balance between paid contributions and spending for the self-employed. In particular, the contributions paid by the self-employed amounted to some 5% of total social contributions received, and social security spending for the self-employed amounted to 6 to 7% (estimates in absolute terms show that the amount received corresponds to the amount paid) 214 .

Bulgaria No initiatives identified.

Cyprus No initiatives identified.

Czech No initiatives identified. Republic

 Reducing legal conflicts about self-employed

         Germany has introduced a special, hybrid legal category for dependent self-employed. Due to such legal status, dependent self-employment enjoys both more formal and actual protection. It is important to note that dependent self-employed have also been subject to compulsory insurance for old-age pension since 1999.

         Germany has set up the scheme to foster the transition between unemployment and self-employment by introducing the Enterprise Subsidy which can be made available to unemployed persons who still have at least 150 days’ entitlement to unemployment benefit on entering self-employment.

         It might be important to point out that white paper entitled ‘Work 4.0’ published by the Minister of Labour and Social affairs from 2016 contains a range of suggestions and proposals for the shaping of working conditions and the provision of adequate financial and social protection in this context of change. These proposals include old

Germany age provision for all groups of self-employed, shaping transitions between dependent

employment and self-employment. The following points should be mentioned: promotion of education, further training and qualification; shaping of the transitions between the phases of working-life and between dependent employment and selfemployment; step-by-step development of the unemployment insurance towards employment insurance; old-age provision for all groups of self-employed.

 Limiting abuse of atypical employment

         The Grand Coalition has taken steps to restrict the widespread abuse of fixed-term employees and temporary agency workers. In particular, policy makers are confronted with the task of combating and controlling bogus self-employment and undeclared work.

 Extension of social protection for self-employed

214 However, concerns are been expressed by Belgian experts on issues for the future financing of social security regarding this practice. Namely, the current contribution rates for the self-employed decrease as income rises, and there is an annual ceiling on contributions.

The coalition agreement of 07 February 2018 foresees to include the self-employed in the public old-age pension system on a voluntary basis, implemented via opt-out clauses, as well as an obligation for this group to make provisions for old-age. In Denmark, the general system of unemployment insurance is undergoing a major reform in 2017. In 2015 Unemployment Insurance Commission identified potentially problematic issues for self-employed, non-standard and entrepreneurs and has suggested establishing a better system for non-standard workers and better income insurance for entrepreneurs. Further on, the Commission saw a need to simplify and reduce the amount of documentation for the self-employed. One proposal considered by the Commission was to allow a company to close down temporarily for economic reasons or because of seasonal changes to then re-open when things have improved. A special task force was set up in March 2016 to come forward with recommendations on how to simplify and bring the scheme into line with the general unemployment insurance scheme by 2017.



Denmark Extension of social security for self-employed In September 2016, the Liberal government proposed to introduce mandatory pension

saving for persons without sufficient savings, though not with any particular view to the self-employed. The proposal was to introduce compulsory saving for these groups with insufficient savings (defined as less than six percent of gross income), starting with 0.25% of income in 2018, increased to 2% of income in 2025 215 . Mandatory minimum pension savings for everybody would be innovative and resembles compulsory fire insurance for home owners/occupiers and third party insurance for car drivers. The Ministry of Finance estimates that this measure would reduce the number of individuals with insufficient savings from (around) 750,000 to 707,000 by 2025. However, this was opposed by representatives for self-employed who preferred to decide between themselves when and how much to save. Due to lack of general political support for the reform-package of which it was part the proposal was not adopted and is not part of programme for the new coalition government between the Liberals, Conservatives and the Liberal Alliance announced on 27 November 2016.

 Reducing legal conflicts about self-employed

         Estonian employment policy and social protection system and corresponding legislation presume clear distinction between self-employed persons and employees, and passive income (i.e. dividends) and active income (i.e. wages, salaries). Enforcement of this distinction is also the major measure that controls different types of employment and extend social protection coverage. In this regard, it is important to point out the role of judiciary bodies in Estonia. As reported by the Estonian ESPN expert, the Supreme Court delivered three significant decisions (in 2014 and 2015) which supported the position that in some cases, service agreements between companies can be considered a fraudulent form of (self-) employment, for the purpose

Estonia of avoiding employment taxes, including social security contributions. The decisions

stipulated that service agreements between companies are considered employment relationships if one establishment works for another company subject to the management and control of the company. Subordination is a strong indication of employment relationship, since according to Employment Contracts Act a natural person (employee) does work for another person (employer) in subordination to the management and control of the employer and the employer pays to the employee remuneration for such work’. The court decision gives Estonian Tax and Customs Board (EMTA) the right to requalify the service agreement to employment contract in case an actual employment relationship between the parties has been identified. Labour taxes from the remunerations (that is the prerequisite for social protection coverage in some social risks like sickness) shall then be declared and paid.

 Transparency and administrative accessibility of social protection

         The procedure of changing the employment status is relatively simple, as are the procedures regarding the suspension of professional practice in the case of

Greece unemployment, and the reactivation of social insurance.

 Extension of social protection for the self-employed and non-standard workers –

closing the gap in coverage of social protection Greece has implemented paradigmatic reform in order to extend social security for the

215 The “income” referred to here is solely the income from work for the employees, the earned income for the self-employed and the income from social security for the social security claimants

self-employed. The gradual implementation, since May 2016, of the recent reform of the social protection system (Law 4387/16), is expected, among other things, to ease further the remaining gaps and differences in the provision of social protection benefits, between the ‘self-employed’ and the ‘employees’, as well as among the different types of ‘self-employed’ and those of ‘non-standard’ employment. In particular, the new pension system (as of 2017) brings together several social insurance funds into one unified pension fund. The establishment of this fund implies the application of uniform rules for contributions and benefits to all salaried workers and the self-employed. Additionally, income assessment base was calculated until end of 2016 as percentage of the so-called ‘imputed income’, which is a fictitious income (and not the real income). From 1st January 2017, the ‘self-employed’ contributions have been calculated as percentage of their net taxable income of the previous year, to be paid in 12 monthly instalments. Importantly, unemployment insurance has been extended to the self-employed 216 .

        Further on, since 2015 the conditions of retirement are the same for all employees and self-employed.

Social protection for 'dependent self-employed' In Greece, if the ‘dependent self-employed’ person has up to two employers per year, social insurance contributions will be paid according to an ‘employee’ status (i.e. onethird by the ‘self-employed’ individual and ‘two-thirds’ by the employer).

 Extension of social security for self-employed

         The new regulatory framework (Law 20/2007 on the Self-Employed Work Statute) now treats employment and self-employment equally in relation to social protection. Among the diverse measures provided for on this matter, Spanish experts highlighted the establishment of new incentives in the form of exemptions, reductions or subsidies for Social Security contributions.

        The incentives for self-employment improved since September 2015 (Law 31/2015, art. 31) by reducing the social contributions of new self-employed workers, especially those under 35 years old and women. Further on, unemployment insurance has been extended to the self-employed in 2010.

 Social protection for 'dependent self-employed'

         Economically dependent self-employed workers are defined as those self-employed workers whose economic activity is developed predominantly for a single client on which they depend for at least 75% of their income, do not employ third parties, have

Spain their own production infrastructure, organisational independence and their income

depends on the work carried out. Spain made it compulsory for the dependent selfemployed (‘economically dependent self-employed’ (TRADE)) to opt into accident at work insurance, while this is voluntary for the ‘independent’ self-employed.

 Extension of social protection for non-standard workers

          As a way of extending protection to non-standard workers, in 2013 (Royal Decree 11/2013) Spain introduced new and more favourable way of calculating the contribution periods for contributory social security benefits (excluding unemployment benefit) for part-timers. The calculation is no longer carried out through equivalence to a full working day but is calculated improving the proportionally by applying a general rate of part-time work. The 'Social Dialogue' was resumed in the second half of 2013 with the aim of improving the social protection of part-time workers. As a consequence of the "Agreement on improving the conditions of coverage of social protection for part-time workers" Social Security legislation was amended to include the part-time relief contract, the fixed-discontinuous contract, and the incorporation of contracts of these types in the special scheme for domestic employees 217 .

216 Yet, as far as the risk of unemployment, accident at work and temporary disability are concerned, there are still gaps of coverage for the ‘self-employed’ which have not been dealt with by the new Law. According to Greek experts, despite the various measures taken over the recent years, certain differences and inequalities between the ‘employees’ and the ‘self-employed’, as regards their social protection obligations and rights, still remain. Moreover, there are indications that the recent change of the contributions’ calculations basis has brought about an increase of the financial burden on social insurance contributions of some categories of the ‘self-employed’. Furthermore, the calculation of contributions as a percentage of the ‘imputed income’ had also brought about negative financial implications for a number of ‘self-employed’. As a result, a large percentage of the self-employed lost the entitlement to healthcare benefits. 217 One of the weaknesses of employment creation in Spain in the last three years, according to the consulted trade unions, is the abuse of part-time contracts and the growth of contracts of less than one week. This requires pursuing the path of reforms that improve the social

        Further on, apprentices and paid trainees have a unique monthly contribution quota that gives entitlement to all benefits, including unemployment.

 Extension of social security for the self-employed

         Finland has closed down an important issue of the option to self-report income for insurance purposes. Instead of self-declaration (which often leads to low income coverage) there is a trend towards using registered income as the basis for insurance fees.

 Social protection for the non-standard workers: family benefits

         In Finland, there is a specific measure concerning families which combine nonstandard work with homecare allowance 218 . There is also part-time home care allowance for those parents that have part-time employment. Furthermore, flexible care allowance can be paid to a parent caring for a child under 3 years of age and who works no more than 30 hours per week 219 .

 Measures discussed at the moment

Unemployment protection An expert group was set up to evaluate how to improve unemployment protection

among the self-employed and atypical workers. The expert group, in its report of December 2016, suggested that a new unemployment insurance form – combination insurance – should be implemented. The self-employed who combine paid labour and self-employment could insure both their wages and income from self-employment. Thus the non-coverage problems linked to the unemployed who have combined parttime work with part-time self-employment could be eliminated. The government is discussing the proposal and will possibly, in the course of 2017, make changes in legislation, which would be implemented in 2018.

Finland  Non-standard workers – proposals concerning zero-hour contracts and seasonal

workers A citizens initiative in 2015 followed a report made for the Ministry of Economic

Affairs and Employment and proposed the prohibition of zero-hour work contracts, as well as the introduction of mandatory minimum working hours. The citizen's initiative was turned down and more research has been called for. There have also been views presented that the Employment Contracts Act can already be interpreted so that zerohour contracts can be declared illegal since to fulfil the criteria for an employment contract, an employer must offer at least some work. A law proposal amending the working time legislation and conditions for zero-hour contracts was submitted to the Finnish Parliament in July 2017. The proposal would limit the use of contracts to situations where the employer can justify that the demand for labour varies. The employer would also have to provide the employee an estimate of the amount of work offered. The proposal also aims to ensure that the employee would be entitled to sick pay, as well as pay during the resignation period.

        Further on, the same Ministry ordered a report on the employment status of seasonal workers. The report proposed that all foreign wild berry pickers should be officially employed. The proposal was found to be difficult to coordinate with the existing definition of employment where it is sometimes hard to make a clear distinction between dependent employment, different forms of self-employment and other nonstandard work. However, an information desk for foreign berry pickers was set up by the Ministry on 21st July 2014.

 Basic income experiment

         Aim of the experiment is to see if basic income would improve economic security and social protection among the precarious workforce, be it self-employed or people in non-standard employment contracts. The experiment started on 1st January 2017 with 2000 randomly selected unemployed persons receiving benefits from the Social

protection of this type of contract, as well as fixed-discontinuous contracts in matters of unemployment and duration of contributions to access a pension.

218 Those parents who do not use municipal care are paid a compensation called homecare allowance. Some families are able to combine odd hours of paid work and childcare by working one after another. 219 The flexible care allowance is payable at two rates depending on the parent's total working time.

Insurance Institution of Finland (Kela). The experiment will also show if unconditional benefit will enhance possibilities within self-employment or increase the incentives to

accept work. In France a significant standardisation effort has been made, starting from the 1970s, leading to greater similarity between social protection for self-employed people and employees covered under the general scheme. As a result, today the health insurance schemes provide identical coverage of healthcare costs, as do the pension insurance schemes, with the exception of liberal professions. In addition, since 1945, self-employed people have been covered by the general family benefits scheme.

 Addressing the gap in coverage: Personal accounts

         As a solution to ease transitions between labour market statuses France introduced reform extending compulsory insurance by standardising many schemes in order to be more accessible for self-employed and non-standard workers. Since January 2017 France is implementing the personal activity account which integrates different types of earnings into one unique account. In particular, it aims to better manage pensions, entitlement to life-long learning services and the time saving scheme. It combines three already-existing accounts to accumulate points for job-related training and education, days off not taken and strenuous work. It aims at disconnecting rights to social protection from professional status and connects them to individuals: moving from social rights based on labour market status to individual rights transferable from one status to another. The main idea is to make benefits no longer contingent on a person’s labour market status. Instead, every worker would have an account where their social benefits would be kept in the form of ‘points’, regardless of labour market status. This would make it easier portability and transferability of rights with the idea of making professional career paths more secure. This will simplify the process and will improve

France knowledge of rights (for instance, mobile access). Each person will be able to simulate

different scenarios of future contributions and benefit amounts. For example, if a person does not want to take some days off he/she is entitled to, they can be converted into points for additional training. This is applicable from 2017 for salaried workers

and from 2018 for the self-employed.

 Extension of social protection for self-employed

         Madelin contracts are a way for self-employed workers to improve their social protection on their own initiative. To encourage them to do so, contributions payable under these contracts can be deducted from their taxable income. Greater convergence between social protection for the self-employed and employed workers is also facilitated by the use of “Madelin” contracts established by a law of 1994. However, despite some success, this measure is only financially attractive to beneficiaries who come into a high tax bracket.

         In March 2018, the Government announced a reform that would extend access to unemployment benefits to the self-employed.

 Reducing legal conflicts about self-employed: judicial role

         In France, the role of labour courts is very important in reassigning labour statuses, especially in the case of ‘dependent’ and ‘bogus’ self-employment. A judge could potentially redefine auto-entrepreneur activity (self-employment) as a salaried work contract in cases where a position of permanent legal subordination is established. In such cases, the auto-entrepreneur is considered as a salaried worker and thus falls under the general social security scheme.

The Croatian unemployment scheme is compulsory social insurance covering all employees. Since 2014, it covers also the self-employed. Persons receiving “other income” 220 as well as persons working based on student contracts or occupational training

Croatia contracts are not covered. However, at the same time, their work does not fall under the obligation to pay the contribution for unemployment insurance.

 Extension of social protection for non-standard workers

220 E.g. persons working based on a contract for service, or author’s contract.

         Tax reform which came into effect in January 2017 includes extension of social protection and the obligation to pay social security contributions to non-standard workers, through the broad suggestion, with some exceptions, to treat different kinds of income similarly for tax and contribution purposes. However, the proposal has raised concerns that a significant proportion of artists and creative workers, who exist on incomes barely above the poverty line, will see the net value of these incomes reduced considerably.

Hungary No initiatives identified.

 Extension of invalidity pensions and the treatment benefit (healthcare) to the selfemployed

The reform came into place from 2017, for which the created Advisory Group was

guided by the findings of the 2010 Actuarial Review of the Social Insurance Fund. In particular, Advisory Group increased the level of contribution for Invalidity Pension for 1.5 percentage points extra after assessing that it is financially achievable.

        The Advisory Group also recommended that the Department of Social Protection should consider how the methodology for assessing means for Jobseeker’s Allowance could be simplified, with a view towards ensuring that the principle of equity in assessment of incomes of households in similar income circumstances is observed in a consistent manner, while bearing in mind technical factors in and differences between

Ireland assessing employed and self-employed incomes. Other recommendations related to

ensuring that self-employed people are aware of the need to maintain their social insurance record and the benefits of making contributions on a voluntary basis.

        The Advisory Group also considered the issue of ‘disguised self-employment’ and pointed out that it may have resulted in the tendency to wrongly classify some employees as self-employed, which itself would lead to new social risks. To mitigate these risks, the Advisory Group recommended that the Department of Social Protection and the Revenue Commissioners should ensure greater compliance with the Code of Practice for Determining Employment or Self- Employment Status of Individuals. To this end, there should be regular reporting on the levels of reclassification.

         As a follow-up to the report, a survey is being conducted by the Department of Social Protection of the preferences of some 20,000 self-employed for reform of the system and their willingness to pay increased contributions 221 .

 Social protection for 'dependent self-employed'

          A specific new scheme was introduced in 2015 (Bill no. 22) for ‘dependant selfemployed’ persons working on continuous collaboration contracts who lose their job. In particular, they must have paid contributions for at least one month during the previous year or have a contract with duration of at least a month. DIS-COLL 222 is provided for a number of months, and is equal to half of the months of contributions paid in the previous year. The maximum duration of the scheme is six months. No pension contribution on behalf of the unemployed person is paid when receiving the DIS-COLL. The amount of the benefit is equal to 75% of the monthly average income. The amount cannot exceed €1,300 and the benefit is reduced by 3% from the fourth

Italy month on.

 Extension of social protection for the self-employed and non-standard workers:

parental leave, maternity and sickness allowances ‘Jobs Act’ includes new measures for parental leave, maternity allowances and

sickness allowances for non-entrepreneurial self-employed workers and it is currently under discussion.

         Policy measures were also introduced over time to extend social protection to the selfemployed and atypical workers through maternity and paternity leave and vouchers for babysitting services. Also for maternity/paternity benefits some occupational welfare schemes provide coverage for self-employed workers and atypical ones 223 . Atypical workers and self-employed persons have the same rights and opportunities to be

221 According to the Irish experts, there is no evidence available on the conceivable consequences for the labour market of these measures. Nor has there been much consideration of this. Nor in the Irish context is there any consideration of developing personal accounts. All relevant reforms tend to take a ‘within the system’ approach.

222 The DIS-COLL scheme covers specific atypical workers profiles from the unemployment risk.

223 For example Ebitemp, the Bilateral Agency for atypical workers, provides a contribution of around €2,250 for all those female workers with atypical contracts, who have labour contracts ending during the first 180 days of pregnancy and are not covered by maternity benefits by the National Institute of Social Insurance (INPS). Moreover Ebitemp also provides a contribution of €1,000 for child care costs in nurseries for those female workers who have worked for at least three months in the previous twelve ones.

covered for health care provision, maternity cash benefits and accidents at work and occupational injuries benefits. For these types of social risks, the system seems to work relatively along similar lines as for employees with open-ended contracts.

 Extension of social security for the self-employed and non-standard workers

        The contributory social protection (social security) schemes are in transition in Lithuania since beginning of 2017. Main trend of reform is further expansion of coverage. More types of self-employment will be covered from social risks like sickness and maternity.

        The ‘Social Model’ reform aims to include more people working as self-employed or on non-standard contracts in the State Social Insurance Fund system. The coverage of the self-employed is becoming much closer to that of salaried workers. Moreover, unemployment insurance has been extended to the self-employed. Furthermore, people working on non-standard contracts are fully covered by social protection against all

Lithuania social risks because they are insured by social security on the same legal basis as full time employees. Also, they are covered by universal and means-testing social

protection schemes on the basis of residency without any exception related to their economic status. Only persons engaged in individual activities under a business certificate are still covered only by pensions’ (retirement, disability, orphans’ and widows’) insurance. However, they will be insured not only for basic pension, but for the earnings related part as well.

         As of 2017, the new Law on State Social Insurance increases the involvement of selfemployed in the system of social insurance and self-employed will have the right to sickness benefit. According to the data of the State Social Insurance Fund, due to the envisaged expansion of social insurance coverage of self-employed, rising contributions of unemployment social insurance and bonus taxation will produce an extra budget income of €70m.

 Extension social security coverage for self-employed

         Expert report from Luxembourg identified an initiative aimed at tackling the issue of the self-employed who are compulsorily covered by sickness protection, but are subject to much longer waiting periods than salaried workers. Therefore, Luxembourg offers the opportunity for self-employed to voluntarily join the Mutual Employers’ Insurance Fund without a contributory period being required. Additionally, Long-Term Care Insurance scheme is compulsory for all residents.

Luxembourg

 Transferability of rights and entitlements

         In Luxembourg, the unemployment scheme facilitates transition between labour market statuses. The self-employed have to complete at least two years of compulsory pension insurance (instead of 26 weeks for contractual workers). However, periods of insurance completed as an employed person can be aggregated to this period of two years, provided the person has carried out activities as a self-employed person for at least six months before the submission of the request for compensation.

 Extension of social security for the self-employed

         All self-employed matching the micro-enterprise criteria 224 have an opportunity to

register as a Micro-enterprise taxpayer to decrease the overall tax burden. The creation of the Microenterprise Tax status aims at reducing unemployment, encouraging citizens to start their own business and simplifying tax payment procedures. However, according to Latvian experts, this regime has led to inadequately low contributions to social insurance budget and weak social protection of this group of employees.

Latvia

 Transferability of rights and entitlements

         The social insurance system is fully individualised, as each person’s contributions are registered on a separate account. Personal accounts (individual social security accounts) make it possible to attach acquired rights to the individual, rather than the work contract. When moving to a less-regulated work relationship, for instance from a permanent job to self-employment, people may begin accumulating fewer rights, but they would not lose the rights previously accumulated. However, the Latvian expert points to some drawbacks, as the individual social security account keeps redistribution as low as possible.

224 As reported by the Latvian ESPN expert, one specific category is composed of micro-enterprises. In this country this is a company: (i) that employs up to 5 employees, (ii) the annual turnover of which does not exceed €100,000, (iii) where all shareholders are natural persons and (iv) where the monthly salary of any employee may not exceed €720.

 Reducing legal conflicts about self-employed

         Legislative measures were taken in 2012, whereby clear criteria were established to

Malta determine whether an employment relationship is that of a self-employed or of an employee. Despite this legal provision, in practice it is known that many employers

and workers respectively prefer to recruit and to be registered as self-employed even when they meet the required minimum of five conditions from the eight specified in the Legal Notice, and should thus be considered employees.

 Transferability of rights and entitlements

        To deal with the issue of transferability of occupational pensions entitlements, the Netherlands has established specific collective schemes for the self-employed, facilitating transferability.

 Social protection for 'dependent self-employed'

         Only those self-employed who work at the premises of/under supervision of an employer (‘dependent self-employed) are compulsory covered by occupational and work injury schemes.

 Extension of social protection for non-standard workers

         The Netherlands introduced an initiative of re-defining the number of temporary jobs by transforming them more quickly into permanent employment. Namely, from 2015 (Work and Security Act), new rules are aimed at getting earlier and easier steady work

The for flexible workers. New rules refer to the probationary period (should not be with Netherlands contracts shorter than 6 months), the non-competition rule (not allowed if there is no

permanent contract) and the number of successive temporary contracts (a fourth extension of more than two years on a temporary contract automatically transforms it into a permanent position) 225 .

 Reducing legal conflicts about self-employed

        The only relevant recent change in legislation concerns the Declaration of Independent Contractor Status (VAR). This is a statement for the self-employed under which the tax authorities will assess whether the agreement between the self-employed person and the client is actually not an employer-employee relationship. This is to counteract any improper advantage of the self-employed by other principal (i.e. bogus selfemployment). In this case, payroll and social security taxes would be imposed on the employer. In the case of VAR the self-employed person was always responsible for the accuracy of the data submitted. If it turned out that the information provided about the relationship between the self-employed person and the client was wrong, both could be liable for additional tax assessments.

 Extension of social protection for the self-employed and non-standard workers

        The changes in the legislation, introduced in 2015 and 2016 closed a gap in the coverage of social protection between salaried workers and those employed based on the non-standard contracts.

         Poland introduced changes in maternity benefit entitlement. Particularly, the maternity allowance is paid at 100% of the assessment base only in case of contributing for at least 12 months prior to the childbirth. If the contribution period is shorter, maternity allowance is set at the minimum level, and increases by 1/12 for every month of the contribution payment. The rationale behind this mechanism is to prevent claiming very

Poland high benefits, with the short contributing period. This mechanism is effective from

January 2016, with the amendment to the Act on cash social insurance benefits in the case of sickness and maternity.

        Provision of maternity/paternity benefits as well as family benefits is in fact universal and does not favour employees on regular contracts over the self-employed, those working on non-standard contracts or other groups. In these areas, improvements are relatively recent. The most spectacular one came in 2015, with the establishment of the new parental benefit, effective from January 2016. Moreover, since 2017 the new minimum income provision covers people working under commission contracts. It introduces a minimum hourly rate for such contracts. This limit will be increased every year, following the growth of the minimum wage.

225 However, there is great doubt among stakeholders whether this law will improve the position of flexible workers. It might even leads to a worse situation. In this context it is interesting that the law provides that the unemployed have to accept any available work after six months, not only suitable work.

 Limiting abuse of atypical employment

         In 2016, new legal provisions came into force aimed to reduce the share of nonstandard work contracts. New rules when entering into fixed-term contracts were provisioned to reduce the asymmetry in terms of hiring and firing workers employed on permanent contracts and other types of contracts. Also at the beginning of 2016, rules of social insurance coverage of workers under civil law commission contracts were aligned with the rules applicable in the provision of employment under contracts of employment based on the Labour code. The new principle of the minimum base for social insurance equal to minimum wage was introduced. As a result, if an employee is covered by more than one contract with the remuneration lower than minimum, the social insurance contribution is paid on the basis of all relevant contracts, up to the required minimum. This provision reduced possibility to hire non-standard employees on very low base contract covered by social insurance and combine it with another one with higher base, not covered by social insurance.

 Extension of social protection for the self-employed: paradigmatic reform

         Law 20/2012 extended unemployment benefits to cover self-employed workers in a situation of economic dependence i.e. at least 80% of their activity is conducted on behalf of a contracting entity (provision of services contract). In 2013, unemployment benefits have been again extended to cover self-employed workers who earn their income through a registered business of their own or possess an individual commercial establishment, as well as their spouses if performing a professional activity together with them on a regular and permanent basis.

         In parallel, there are declared intentions of extending the mechanisms already in place regarding bogus self-employment to other bogus situations such as bogus traineeships and to include mechanisms for the protection of precarious workers against unemployment.

 Social protection for 'dependent self-employed'

         In Portugal, the dependent self-employed enjoy better legal protection than the other categories of self-employed. Since 2012 they have been entitled to unemployment benefits. However, the eligibility conditions for these people are different than for salaried employees 226 .

Portugal

 Reducing legal conflicts about self-employed

         One instrument of the ongoing reforms is the integration of the taxable income of people who work under different statuses or work for different employers. According to the ESPN experts, such integration could partly solve the problem of the boundaries between different employment statuses and could reduce the legal conflicts about dependent self-employment. In Portugal dependent self-employment is formally defined, drawing a distinction between income from ‘dependent work’ and ‘independent work’: Law 55-A/2010 set different conditions for the cases when a particular collective entity benefits from ‘at least 80% of the total value’ of the selfemployed worker’s yearly activity. In this case, the worker is considered to be economically dependent on the collective entity, and the latter is thus required to pay a portion of the worker’s contribution to the social security scheme.

         Further on, Law 63/2013 set new mechanisms to tackle the fraudulent use of selfemployment contracts in subordinate employment relationships. The new law established as a duty of labour authorities to open an enquiry procedure whenever they were informed of a case of self-employment with some features of subordinate employment. It also simplified the administrative and juridical procedures, especially through the definition of short periods for each step to be completed and the clarification of the roles and responsibilities of each party in the process.

 Social protection for 'dependent self-employed'

         In Romania, dependent self-employed workers (with one client) benefit from

Romania compulsory pension and health insurance (while for the ‘independent’ self-employed, this is conditional on a certain income), and the level of social contributions paid by the

employer is equivalent to that for a salaried worker. The beneficiary of the work is required to pay contributions equivalent to those of an employer. Yet, compared to a salaried worker, the dependent self-employed do not have mandatory insurance against

226 Namely, 720 days in the previous 48 months compared to the 360 days in the previous 24 months applicable to most salaried workers.

unemployment, and their job stability is not guaranteed. A recent Government Commission on Sustainable Labour Markets and Social Policy proposed a number of reforms that would substantially improve adequacy of social protection for large groups of the self-employed and people with atypical forms of employment. Particular reforms were suggested in unemployment benefits and sickness benefits, including both relaxed and less stringent qualification criteria in unemployment benefits (either for all or for certain particularly vulnerable groups in the labour market), reforms in unemployment benefits to encourage self-employed to re-activate their business, and changes in the calculation of entitlements for sickness benefits (will also have consequences for related programs such as parental leave benefits). However, whether any of these reforms will be implemented is far from certain.

Sweden

 Transferability of rights and entitlements

         Sweden is among countries which have implemented mechanisms allowing a person to include contribution periods from previous employment in the calculation of unemployment benefits. Namely, unemployment benefit for the self-employed is based on the average income according to the most recent tax statement, or on the average income from the last two tax statements, depending on which alternative is most favourable to the self-employed. For those who have been self-employed for less than 24 months, special rules apply whereby compensation can be based on earnings before the start-up of the business.

 Reducing legal conflicts about self-employed

         Slovenia is the only Member State that offers unambiguous definition applicable drawing a clear-cut distinction between ‘genuinely’ self-employed, ‘dependent’ selfemployed and ‘bogus self-employed’. In Slovenia, a self-employed person is considered to be an economically dependent person when she/he performs work on the basis of a civil contract in person, for payment, independently and for a longer period in circumstances of economic dependency, and does not employ any workers. Economic dependency means that the person obtains at least 80% of his or her annual income from the same contracting authority.

Slovenia 

Extending social protection for non-standard workers In Slovenia, employers’ contributions (since 2013) and employees’ contributions

(since 2014) to pension and disability insurance have been paid on work under civil law contracts. Similarly, since 2015 social security contributions have been paid on student work, which has eliminated the main factor that made it a precarious form of temporary and occasional work. This new regulation on student work has improved their social security rights. Since 2014, those workers on non-standard contracts in the cultural sector who are engaged in artistic creation are entitled to the daily amount of sick pay if unable to work for at least 31 days. This measure is aimed at encouraging independent work in the cultural sector, improving the position of workers and creating

more favourable conditions for their work.

 Reducing legal conflicts about self-employed

         In Slovakia, the definition of dependent work was tightened in 2013 and 2015 to prevent labour-law relations being replaced by other forms of contractual relations and to support the transition of the bogus self-employed to standard employment relationships.

 Extension of social security for the self-employed

         Slovakia is among countries which implemented parametric reforms aimed at

Slovakia extending the coverage of the self-employed, namely by increasing the minimum

insurance base.

 Extension of social protection for the non-standard workers

         Gaps in coverage and a level of social protection between employees in standard forms of employment and employees relying on non-standard forms were seen - especially by representatives of trade unions - as inadequacy of social protection system, caused by interest to cut off labour costs. Situation of persons on work agreements may serve as an example. Until 2013, work agreements were exempted from social and health insurance payments and enjoyed much less employment protection. Employing a person on work agreement was much cheaper than a person on a standard contract. Discussions resulted into reforms of some parts of the social protection system which harmonised status of non-standard workers with those in standard employment. In 2013 and 2015 new legislation increased the protection of people on work agreements and harmonised the social contributions regime for these contracts with the regime for standard contracts.

 Current efforts

The Government appointed the Taylor Commission in 2016 to review workers’ rights and practices and the Work and Pensions Select Committee is now setting up an

United inquiry into self-employment and the gig economy, with a particular focus on whether

Kingdom the UK welfare system adequately supports these workers. Relevant debate gaining pace is on the funding side: this concerns the potential for

closer alignment, or even merging, of income tax and NICs. This would have implications for self-employed people in particular, but also – because of the growing gap between the lower earnings limit for social security contributions and the personal

tax threshold – for part-time and low-paid employees.

A NNEX 8: T HE EXISTING EU INSTRUMENTS

The EU acquis related to the rights and obligations in relation to social security and protection is implemented through a variety of legislative and non-legislative provisions. However, none of these fully address the rights and obligations of people in non-standard employment and forms of self-employment.

The Open Method of Coordination (OMC), the European Semester and benchmarking

The Recommendation 92/442/EEC 227 on 'Convergence in social protection objectives and policies' identified commonly held objectives in the area of social protection and many initiatives in this field which followed can be seen as continuous process of implementation, monitoring and follow-up of this Recommendation, leading to the Social Protection Committee and the Social OMC. Together with the Recommendation on "common criteria concerning sufficient resources and social assistance in social protection systems" from the same year 228 , the 92 Recommendation had great impact on development and convergence of social protection systems in decades which followed.

The Recommendation was successful in establishing the firm bases for the Treaty of Amsterdam in 1999 where objectives started to be transformed into obligations , led to the 229 establishment of the High Level Working Party on Social Protection which was replaced by newly established Social Protection Committee in 2000, as the first body with the specific purpose of monitoring developments, promoting exchanges of information and good practice, and preparing an annual report on social protection. It offered also a basis for establishment of the Social OMC, particularly because of the resemblance of the methods of implementation.

One of the reasons why it was so influential is the success in setting up a strong dialogue between Member States and identifying common issues in social protection systems. In particular, the debate established a consensus among Member States and the EU institutions that social protection systems need to be modernised . The reason why the Recommendation 230 was so effective in raising the interest among Member States to exchange practice and actively participate in such debate might be its 'inviting', non-obligatory nature of the Recommendation and its aim of keeping the important balance between providing guiding principles in development of national systems and acknowledging that Member States remain free in determining how their systems should be organised and financed.

While the Recommendation gave a detailed rendition of many aspects of social protection and did so by making reference to ‘workers’, it did not mention explicitly people that fall outside the standard worker category. When it comes to the self-employed, the recommendation merely states that Member States should 'examine the possibility of introducing and/or developing appropriate social protection for self-employed persons.'

Later on, the common objectives and indicators on coverage of adequate and sustainable pensions, health care and long-term care set out in the context of the Social Open Method of

227 Council Recommendation of 27 July 1992 on the convergence of social protection objectives and policies (92/442/EEC), OJ L 245, 26.8.1992, p. 49–52.

228 Council Recommendation of 24 June 1992 on common criteria concerning sufficient resources and social assistance in social protection systems.

229 European Social Observatory, Social Policy at the EU Level: from the Anti-Poverty Programmes to Europe 2020, Background Paper, 2012, VC/2012/0658, p.11.

230 Modernising and improving social protection in the European Union, Communication from the Commission, COM (97) 102 final i, 12 March 1997.

Coordination 231 also gave little attention on how such coverage functioned in the context of labour law and types of employment.

Finally, 'Europe 2020' built on all of these steps when, in connection with the European Semester, it became the EU policy framework for policy coordination in the field of social protection with concrete and defined aims, implementation methods and monitoring systems.

EU Labour law: directives on non-standard forms of employment, on protection of employees in

case of employers' insolvency and on working conditions

Over the years, an EU legal framework for protection of rights in some areas of non-standard employment has been developed through the Directive on Temporary Agency Work 232 and through negotiated agreements between social partners and implemented via the Directives on Part-Time Work and Fixed-Term Work 233 . But as the Directives apply to employment conditions, their impact on social protection is rather limited. While social protection closely linked to employment as part of pay, such as occupational pensions 234 , maternity pay and sickness pay is covered in these Directives the Court of Justice has made clear in its case-law that the Directives do not apply to statutory social security schemes . In the preambles to the 235 Agreements the social partners call on Member States to ensure that social protection arrangements are adapted to fit with evolving flexible forms of work. These preambles are not legally binding and, while Member States on several occasions have stated their commitment to make social protection more employment friendly, a recent review of the extent to which this has happened suggests that in several Member States there is ample room for improvement 236 .

The Directive 2008/94/EC i on protection of employees in the event of employer insolvency covers protection of social security benefits under national statutory social security and supplementary occupational or inter-occupational pension schemes.

The Directives on work in specific sectors also provisions related to social protection. For

instance, article 31 of Directive 2017/159 i on the Work in Fishing Convention 237 establishes

the right of fishermen to compensation in the event of injury due to occupational accident or disease. Similarly, the current maritime proposal includes a provision on social protection of seafarers on the compensation in case of death or long-term disability for occupational reasons.

Directives on equality between women and men

The EU legal framework on equality between women and men guarantees a number of rights in the field of social protection. The Directive 1979/7/EEC i on the progressive implementation of the principle of equal treatment for men and women in matters of social security covers statutory schemes which provide protection against the risks of sickness, invalidity, old age,

231 COM(2008) 418 i , available online . For a presentation of the Social OMC see online .

232 Temporary Agency Work, available online .

233 Fixed-Term Work Directive 1999/70/EC i, available online ; Part-time Work Directive 97/81/EC i, available online .

234 See Joined Cases C-396 and 395/08 Bruno and others at para.41, available online .

235 Ibid and Case C-385/11 Moreno v INSS, available online .

236 ESPN (2017), Thematic Reports on Access to social protection of people working as self-employed or on non-standard contracts.

237 Council Directive 93/103/EC i concerning the minimum safety and health requirements for work on board fishing vessels (thirteenth individual directive within the meaning of Article 16(1) of Directive 89/391/EEC i).

accidents at work and occupational diseases and unemployment, as well as social assistance, in so far as it is intended to supplement or replace the statutory schemes. Occupational pensions are covered by the Directive 2006/54/EC i on the implementation of the principle of equal opportunities and equal treatment of men and women in matters of employment and occupation (recast). The EU legal framework also guarantees a number of rights in the field of maternity leave and parental leave.

Directive 2010/41 i/EU on equal treatment between men and women engaged in self-employed

activity

At EU level, one step towards improving the protection of individuals engaged in selfemployment and the coverage of social protection for spouses of the self-employed workers has already been made. The Directive 2010/41 i/EU 238 on equal treatment between men and women engaged in self-employed activity grants women (including spouses of the selfemployed) the right to be entitled to maternity leave and benefits for at least 14 weeks; levels of benefits and eligibility conditions vary widely among Member States. Furthermore, where a system for social protection for self-employed workers exists in a Member State, it requires providing coverage also to spouses of self-employed workers engaged in their activities.

Social security coordination

Ever since 1959, the social security systems of the Member State have been coordinated so that persons moving within the Union could avoid losing their social security protection. However, Member States remain competent to determine the conditions for becoming affiliated to the social security system, provided that these conditions are not discriminatory.

The rules on social security coordination apply to national legislation on sickness, maternity and equivalent paternity benefits, old-age pensions, pre-retirement and invalidity benefits, survivors’ benefits and death grants, unemployment benefits, family benefits, benefits in respect of accidents at work and occupational diseases . 239

Self-employed persons in the social acquis

The notion of "self-employed person" is not defined at EU level. Depending on the objectives pursued by different EU instruments, self-employed persons are covered or not.

Some directives concern only the employed person as they treat issues linked to employment contracts and are not relevant for self-employed persons (e.g. Directives 89/391/EEC i, 91/533/EEC, 1999/70/EC i, 96/71/EC i). Other EU law provisions apply to both categories of persons, either by applying the same rules to self-employed as to employed persons (especially the current acquis implementing the principle of equal treatment, e.g. Directives 2000/78/EC i, 2006/54/EC, 1979/7/EEC i), or by setting different rules for self-employed persons (EU rules on social security coordination 240 ).

238 Available online .

239 See most notably Regulation (EC) 883/2004 i on the coordination of social security systems, which was adopted on the basis of Article 42 EC (now Article 48 TFEU) and Article 308 EC (now Article 352 TFEU) and its implementing Regulation (EC) 987/2009 i. 240 Regulations 883/2004 i and 987/2009 apply both to employed and self-employed persons. The coordination rules differ depending on the status of the person. A person is employed or self-employed depending on the national legislation under which the activity is performed. If a person is performing both employed activities in one Member State and self-employed activities in another, then priority in determining the

Directives related to supplementary and occupational pensions

The Directive 2014/50 i/EU on the acquisition and preservation of supplementary pension rights lays down rules aimed at facilitating the exercise of the right of workers to freedom of movement between Member States by reducing the obstacles created by certain rules concerning supplementary pension schemes linked to an employment relationship. The Directive (EU) 2016/2341 on the activities and supervision of institutions for occupational retirement provision requires that every institution provides both general and personalised information to the beneficiaries.

applicable legislation is given to the State of employment, knowing that the social security protection of employed persons is more developed comparing to self-employed persons.

A NNEX 9: A NALYTICAL METHODS

The estimation of social and economic impacts include several sets of microeconomic simulations, to quantify the impact of the measures, as well as more qualitative analysis, allowing to better understand behavioural reactions and attitudes (through a survey) and country specific context (through case studies).

• 1. 

  Evidence, sources and quality

The following expert advice has fed into the Impact Assessment:

• - 

  From the Report of the public consultation "Accompanying the document establishing a European Pillar of Social Rights (SWD(2017)206)".

  • - 

    From the commissioned study "Access to social protection for all forms of employment: Assessing the options for a possible EU initiative" by Fondazione Giacomo Brodolini.

• - 

  From the commissioned study "Behavioural study on the effects of an extension of the access to social protection for people in all forms of employment" by Open Evidence.

• - 

  From six commissioned case studies on special forms of employment (Mini-jobs in Germany, civil law contracts in Poland, self-employed without employees in the Netherlands, casual worker in Romania, economically dependent self-employed in Spain, project workers on continuous collaboration projects in Italy)

• - 

  From literature review as referred to in footnotes.

• 2. 

  Quantification of impacts - simulations on the basis of the EU-SILC database

The microeconomic simulations were carried out by using the latest available EU-SILC crosssection and panel data. Two complementary approaches were used: simulations in EUROMOD, the EU's tax-benefit microsimulation model, and ad-hoc simulations outside this modelling framework, respectively.

2.1 EUROMOD simulations

Simulations based on EUROMOD, the EU's tax-benefit microsimulation model, were commissioned by DG Employment. Further details can be found in the above-mentioned corresponding study, Fondazione Brodolini (forthcoming). EUROMOD was used to simulate the fiscal and distributional impacts of extending formal coverage by unemployment insurance to the self-employed in five Member States where they are currently excluded from the corresponding schemes, namely Greece, Croatia, Italy, Latvia and the Netherlands. Simulations were based on EU-SILC data for 2016. Available data did not allow performing similar simulations for non-standard workers since the data set does not distinguish between standard workers and non-standard workers.

The simulated unemployment shock was defined using historical data. Namely, the average transition rate from self-employment to unemployment or inactivity for 2007-2015 from EU- SILC data was considered, focussing on those self-employed workers (in year t-1) who become unemployed or inactive as a result of the unemployment shock (in year t). The selfemployed are defined as individuals with positive self-employment income and no (dependent) employment income. From that sample, those subject to the unemployment shock are randomly selected. Table A.11 below reports on the unemployment risk underlying the simulated unemployment shocks.

Two scenarios were simulated:

• - 

  Scenario 1 ("earnings-related" scenario in the main text) assumes that recipients qualify for the unemployment benefit currently in place for employees if entitled, given their characteristics, and for as many months as dictated by the programme rules currently in place.

• - 

  Scenario 2 ("flat-rate" scenario in the main text) assumes that all formerly self-employed workers (in year t-1) who are either unemployed or inactive (in year t) qualify for a flat-rate unemployment benefit at 60% of median income, paid for at least as many months (in year t) as they had been self-employed (in year t-1).

In both scenarios, the rules of the systems in place are also extended on the financing side: self-employed workers pay the lowest unemployment insurance contribution rate currently applying in the Member State's existing system to employees and their employers. The corresponding information was retrieved from the MISSOC database. Table A.11 below presents the resulting contribution rates.

Reported results compare the outcomes in each scenario to the baseline. This is defined here as the counterfactual situation that might prevail after the unemployment shock and under the current state of affairs of no unemployment insurance coverage (and no unemployment benefit) for self-employed workers.

Table A.11: Unemployment risks and contribution rates 241

Unemployment Contribution risk Rate

Italy 9.0% 1.9%

The Netherlands 4.2% 2.07%

Greece 6.2% 1.93%

Latvia 11.7% 1.9%

Croatia 12.9% 1.7%

Distributional results relate to percentage point changes in the at-risk-of-poverty rate for several types of households, with respect to a poverty threshold that is kept constant at 60% of median equivalised household disposable income in 2016.

2.1 Ad hoc simulations for all Member States

The aim of the ad-hoc simulations is presenting some estimates about the budgetary costs that would be due to the provision of comparable unemployment benefits (henceforth UB) and sickness benefits for self-employed workers (henceforth SE) across all EU28 Member States. 242

The reported simulations are based on hypothetical scenarios where the same type of (earnings-related or flat-rate) UB scheme is introduced in all EU-28 countries, regardless of the design of existing schemes. Based on the estimated number of self-employed that would be eligible for benefits according to the evidence recorded in EU-SILC about the SE risks to

241 The information is presented in the main body of Brodolini (2018, forthcoming) for HR, IT and LV, and in the annex (scenarios 5 and 6) for GR and NL.

242 While data availability in EU SILC does not allow corresponding calculations for the other branches of social security, qualitative considerations are added in the main text to generalise the obtained results.

become unemployed, corresponding benefit payments to them are simulated. Aggregating these payments provides an estimate of resources needed.

To get a magnitude of the estimated costs of these reforms, the total “gross” spending for the new simulated UB scheme for SE is expressed as a share of annual gross labour incomes of SE in the reported results, which can be interpreted as a contribution rate . 243

This rate has to be interpreted as a gross rate in the sense that possible spending already existing for UB paid to SE is not taken into account and, thus, not subtracted from the simulated cost of the new UB scheme. To infer a “net” rate, the ratio between the unemployment benefit already paid to the self-employed and the corresponding wage bill should be subtracted from the “gross” rate.

It should be noted that the simulations reflect a “partial equilibrium” setting. This means that the model does not allow to capture behavioural reactions the introduction of benefits nor does it capture secondary effects due to the interplay between a possible change in contribution rates and received benefits or the other components of the public budget (in terms, for instance, of personal income tax, tax deductions or other means tested benefits).

Two types of benefits are simulated:

• 1. 

  A replacement rate scenario: The unemployment benefit is paid according to the number of months spent in unemployment or inactivity for reasons different from retirement during 2014 for a maximum of 6 months. The amount of the benefit is equal to 50% of previous monthly gross labour income with a maximum monthly amount equal to 60% of national median gross labour income (from employment and self-employment).

• 2. 

  A minimum level scenario: The unemployment benefit is paid according to the number of months spent in unemployment or inactivity for reasons different from retirement during 2014 for a maximum of 6 months. The amount of the benefit is equal to 30% of the national median gross labour income (from employment and selfemployment), independently on previous individual monthly gross labour income as a self-employed.

Therefore, scenario a) is more consistent with the rules about unemployment benefits that apply to employees in most countries, while scenario b) is more consistent with the adoption of a minimum level of mandatory social security.

Sensitivity analysis was undertaken where the benefit was paid for 4 and 8 months instead of six or where the benefit level varied. Notably, the earnings-related benefit was set at 40% or 60% of previous monthly gross labour income instead of 50%. The flat-rate benefit was set at 20% and 40% of the national median gross labour income instead of 30%.

Estimates have been carried out by using a sample built merging 7 cross-sections of EU-SILC (2009-2015 cross-sections, i.e. covering the period 2008-2014), that, following harmonised definitions of the relevant variables, record workers’ self-reported statuses (i.e., employee, self-employed, unemployed or the various inactivity conditions) in each month of a certain year and also record annual gross labour incomes from employment and self-employment.

In the simulations, for each year of the period 2008-2014 recorded in EU-SILC 2009-2015 cross-sections, the focus is on those who were self-employed (both full-timers and part

243 The study also reports on contribution rates for all in employment and self-employment, where the cost is divided by the annual gross labour incomes for SE and employees.

timers) at January of each year and, being tracked their working status since February up until to December of that year in EU-SILC, measure the cost to cover these workers’ subgroups against unemployment risks is measured. To additionally consider the annual cost for paying unemployment benefits to those who have been self-employed in the previous and are still entitled to receive an unemployment benefit at the beginning of the next year, the results reported here add to the annual cost the spending for those who are still unemployed at the end of the year and have not exhausted the total period of receipt of UB, thus assuming they will not be again employed before the period of receipt of UB expires 244 .

The results reported consider as becoming unemployed those self-employed in the data who move into unemployment or inactivity for reasons other than retirement throughout the year. This broad definition is chosen in response to a possible underreporting of unemployment among the self-employed in EU-SILC data. It may allow to reduce a possible underreporting of unemployment risks for SE due to the fact that, being absent UB schemes for them in many countries, those who stop to work can answer in the SILC interviews to have become an inactive rather than an unemployed merely because they are discouraged, even if they would like to find a new job. Actually, for many not retired individuals, when they answer about their self-reported employment status the difference between the statuses of unemployed or inactive is rather slight and may depend on an individual feeling rather than on an objective status. And this holds true especially for SE when – being almost always absent UB schemes – a sort of official definition of unemployment, that makes the individual entitled to receive an UB, is not perceived by the them.

The report contains more details on the steps of the simulations, where each step has been run at the country level and pooling together the 7 cross-sectional waves, thus providing an average cost over a 7-year period.

Ad-hoc simulations on the extension of sickness benefits to the self-employed:

Differently from unemployment benefits, which can be simulated using information on periods spent in unemployment or inactivity, there is no information in SILC on which selfemployed are more likely to receive a sickness benefit, if this benefit were provided. The contractor therefore simulated possible sickness benefit for the self-employed imputing benefits according to their characteristics, exploiting knowledge on employees who receive such benefits. More details can be found in the report, section 3.2.7.

• 3. 

  Qualitative assessment

3.1. Comprehensive literature review and qualitative evaluation

The qualitative assessment of impacts is based on comprehensive literature surveys and a country expert questionnaire in 16 Member States (Denmark, the Netherlands, Ireland, Belgium, Germany, Latvia, Romania, Italy, France, Finland, Sweden, Spain, Greece, Poland and Slovakia). The questionnaire focuses on the most likely economic and social impacts of policy options through a qualitative evaluation made by national experts with reference to those social provisions where there is a lack of formal access for workers in non-standard forms of employment and self-employed or where the de facto coverage is highly limited.

More generally, as highlighted above, the use of quantitative tools to assess the impacts of the initiative implies a certain asymmetry given that available models capture more easily the

244 In addition to these "upper bound" results, , the study also reports "lower bound" results focusing only on those who were self-employed in January and then become unemployed, which can be regarded as a measure of the first-year cost of the reform scenario.

costs than the benefits associated with the initiative. Special care will therefore be taken to highlight in a qualitative manner the benefits associated with the extension of social security to the previously unprotected groups in the labour market (in terms of changes in income distribution, labour market dynamism, sustainability of the welfare state). To give a complete picture, the qualitative analysis also points to additional negative effects not captured by the available quantitative models. For instance, there could be a reduction in labour demand, due to an increase in labour cost.

3.2 Behavioural survey

Furthermore, a behavioural survey based on an overall sample of 8000 people was conducted in 10 Member States (Germany, France, The Netherlands, Italy, Spain, Slovakia, Romania, Sweden, Poland and Belgium) to find out whether an extension of access to social protection would affect, and how, the behaviour of workers in non-standard forms of employment and self-employed. Respondents were recruited through an online panel so the sample is representative of the Internet population, but not representative of the population in general. The survey combines traditional online survey research and online experiment. It explores attitudes towards the existing national social protection systems (effectiveness in protecting people in different situations) and people's preferences (degree of risk aversion strength of preference for having social protection coverage as well as willingness to pay for it) as well as behavioural implications (whether availability of social protection would make selfemployment more attractive, importance of transferability of rights). Socioeconomic information about the respondents was also collected, so as to control to specific characteristics of the individuals. Internet was used as a methodological tool for data gathering, using online panels in the selected countries. Annex 16 presents the main results of the survey.

3.3 Case studies

Finally, six case studies were conducted by external contractors to study relevant specific contract types in particular Member States:

 DE: Mini-jobs  NL: self-employed without employees  PL: civil law contracts  RO: casual workers  ES: economically dependent self-employed  IT: workers on continuous projects (co.co.co.)

In the first place, the studies are used to describe the nature and the extent to which special forms of employment in the six concerned Member States are being used as a 'new form of non-standard work'. They also illustrate gaps in the de facto ability of people in these forms of employment to acquire adequate protection through social security systems.

In a second step, the case studies inform about the possible existence of derived rights (e.g. free health insurance for spouses, survivors pensions) social protection of people working in these special forms of employment and the take-up of social protection benefits at household level (e.g. widow pension). They also highlight the extent to which gaps in access to statutory social protection are compensated by other means (e.g. private health, pension or life insurance, position in the wealth distribution).

Finally, the case studies look at the costs of these forms of employment and their estimated potential benefits.

A NNEX 10: F URTHER INFORMATION ON IMPACTS OF POLICY MEASURES AND POLICY

PACKAGES

This Annex complements Chapters 5.4 and 6.2 on the impacts of the proposed policy measures and packages, respectively. It comprises four sections. Section 1 presents evidence on the impact of opt-out and opt-in clauses for enrolment in voluntary schemes. Section 2 provides details on the available quantification of the impact of measures to extend formal coverage and puts them in perspective. Finally, Section 3 presents an exemplary quantification of possible impacts of relaxing qualifying periods as a measure addressing effective coverage. Finally, Section 4 presents an overview table summarising main impacts of the policy packages.

• 1. 

  Voluntary coverage under opt-in vs. opt-out clauses

Voluntary coverage by social security schemes can be implemented via opt-in clauses or optout clauses. Opt-in clauses require potential beneficiaries to take action to choose coverage whereas opt-out designs imply being covered is the default option, with action needed to avoid being covered. Experiments and observational studies show that making an option a default increases the likelihood that it is chosen (default effect). In the context of insurance, there is evidence that when consumers are faced with car insurance options that are very similar, except for the possibility to opt-in vs. opt-out of additional coverage, the former design option results in much lower levels of participation than opting-out (20% vs. 75% in a seminal study) 245 . In a similar vein, a large empirical literature shows the importance of default options in shaping decisions about retirement savings including savings plan

participation and levels of contributions 246 . The following empirical evidence from Member

States underlines the point regarding social security coverage:

In the UK pension system, flat-rate state pensions providing basic old-age income are complemented by voluntary contribution-based workplace pensions, set up or facilitated by the employer for their employees. To reverse the observed long-term trend decline in coverage, from 2012 the UK progressively switched to an opt-out system from the previous opt-in one. Workplace pension scheme membership subsequently increased from 50% in 2013 to 64% in 2015. The employee opt-out rate was 8% to 11% among larger sized employers and reached 17% in smaller firms.

In Germany, non-standard employees on a mini-job contract have voluntary access to the public pension system. In 2013, the design was changed from an opt-in to an opt-out option. The default conversion resulted in a tripling of the share of mini-jobbers covered by the

pension system, albeit to a still low level (from below 6% to 18%) 247 .

As these two examples demonstrate, setting defaults can be an effective way of influencing behaviour also in a context of providing social security coverage.

245 See Johnson, Hershey, Meszaros, and Kunreuther (1993), Framing, Probability Distortions, and Insurance Decisions. Journal of Risk and Uncertainty, Vol. 7, No. 1, 1993.

246 See Beshears, J.; Choi, J. J.; Laibson, D.; Madrian, B. C. (2009). "The Importance of Default Options for Retirement Saving Outcomes". Social Security Policy in a Changing Environment.

247 The share of mini-jobbers not working in private households covered on a voluntary basis more than doubled between December 2012 and March 2013 (from 5.6% to 11.2%) and has since reached 18%. See RWI, Nachfolgestudie zur Analyse der geringfügigen Beschäftigungsverhältnisse (Minijobs) sowie den Auswirkungen des gesetzlichen Mindestlohns Endbericht Gutachten im Auftrag des Ministeriums für Arbeit, Integration und Soziales des Landes Nordrhein-Westfalen, November 2016, p. 21.

• 2. 

  Micro-simulations for the quantification of the impact of measures to extend formal coverage

Micro-simulations have been commissioned to Fondazione Brodolini 248 for the purposes of

this Impact Assessment in order to estimate the economic costs and benefits of extending unemployment and sickness benefits to the self-employed. An introduction to these simulations is provided in Box A.9 (for further methodological details please see Annex 9) before their set-up and results are presented and discussed in more detail.

Box A.9: Simulations of impacts on the basis of EU SILC data

All commissioned simulations focus on a change in the number of people to be covered and thus aim at illustrating the potential impact of extending formal coverage based on micro data from the EU-SILC database. As per the gaps identified in Section 2.2.2, this is particularly relevant for the self-employed, while available data do not allow undertaking such simulations for non-standard employee. The simulations therefore focus exclusively on extending coverage to the self-employed. At the same time, the simulations do not consider a change in the eligibility conditions (change in the time thresholds, for example) and thus do not shed light on the potential impact of addressing effective coverage. Given data limitations, the simulations focus on unemployment benefits, and to some extent on sickness benefits, while undertaking simulations for the other branches of social security of relevance for the initiative has not been possible.

The microeconomic simulations were carried out using the latest available EU-SILC crosssection and panel data. Two complementary approaches were used: simulations in EUROMOD, the EU's tax-benefit microsimulation model, and ad-hoc simulations outside this modelling framework.

EUROMOD has the key advantage of embodying the set of existing rules of each Member State's tax-benefit system, including eligibility criteria, financing features and rules for the calculation of benefits of existing unemployment benefit systems. It thus allows simulating the budgetary and distributional effects of extending coverage by the specific unemployment benefit system in place in a given Member State to the self-employed, while taking into account interactions with other elements of the tax-benefit system. Such simulations were undertaken for five Member States where the self-employed currently are not covered by unemployment benefits, as these may arguably be among the Member States where impacts of extending coverage may be highest.

As pointed out, the results of this first exercise reflect and are driven by the existing systems in place in specific Member States. Key decisions on how to cover the self-employed are in the competence of Member States, notably including the choice of whether to extend existing schemes or establish separate ones, and importantly the associated level of protection provided. These decisions are of key importance for the impacts. Therefore, in a complementary approach a set of ad-hoc simulations were undertaken investigating hypothetical scenarios where the same type of (earnings-related or flat-rate) unemployment benefit scheme for the self-employed is introduced in all EU28 Member States, regardless of the design of existing schemes. The effects of granting a relatively modest standard level of

248 Brodolini (forthcoming).

protection to the self-employed can thus be compared across all Member States.

For both sets of simulations, the cost of providing the benefits (in terms of total benefit payout) is computed 249 and to account for key social impacts, the distributional effects of the benefits and corresponding contributions are simulated. Again, impacts will not only vary depending on the chosen benefit levels but also hinge on financing choices, which the Recommendation leaves to Member States. At the same time, it should be noted that the simulations assume that coverage is extended to all self-employed (corresponding to mandatory coverage). In this sense simulated impacts represent an upper bound relative to a voluntary extension of coverage.

Some key limitations of the simulations should be noted. First, they do not capture second round effects resulting from behavioural changes in response to the extension of coverage. They thus do not reflect possible changes in labour demand and supply nor the increased labour market dynamism that may result from removing the lack of social security coverage as an obstacle to transitions between self-employment and different types of employment. Second, neither are the simulations able to capture additional macroeconomic benefits of the provision of unemployment benefits, such as their well-established role as automatic stabilisers of the economic cycle.

• a) 

  Extending unemployment insurance

Simulating the impact on public budgets of an extension of existing unemployment benefit systems

The simulations show the fiscal and distributional effects of extending unemployment insurance to the self-employed in a situation in which some reforms led to a homogenisation in a given Member State between employees and self-employed of rights to unemployment benefits and corresponding contributions. The exercise was carried out using the EU's taxbenefit

model EUROMOD for five Member States (Croatia, Greece, Italy, Greece 250 , Latvia

and the Netherlands) where the self-employed are currently generally excluded from coverage, since these Member States may be facing the most pronounced impacts.

The key scenario considered a simulation where coverage was extended to the self-employed under the same rules for earnings-based benefit entitlement in place for employees in each Member State - "earnings-related scenario". Rules of the existing systems regarding eligibility as well as the calculation of benefits are thus applied to the self-employed who become unemployed 251 . Results suggest that, comparing with a baseline situation where the same unemployment risk prevails but no unemployment insurance is in place, extending unemployment insurance could lead to additional total gross expenditure for benefits ranging from around 0.1% (NL, LV, FI) and 0.2% (GR) to 0.4% (HR, IT) of GDP annually,

249 It is expressed as a percentage of GDP in the EUROMOD simulations, which are based on a single wave of EU-SILC data, whereas in the ad-hoc simulations, which show average effects derived from simulations based on six data waves, show them as a percentage of the total labour income of the self-employed, which can be interpreted as a contribution rate that finances the expenditure. 250 Note that in Greece an unemployment insurance scheme for self-employed workers is nominally in force since 2013. Nevertheless, because of stringent eligibility conditions, take-up of that scheme is extremely low: in 2016, when the total number of unemployed workers was 1,126,066 (of whom many had formerly been self-employed), the number of recipients of unemployment benefit for the self-employed was a mere 4,125. See Matsaganis M. (forthcoming) "Income support policies and labour market reform under austerity in Greece" in: “Labour market policies in the era of pervasive austerity: a European perspective” (edited by S. Theodoropoulou). Bristol: Policy Press. 251 To assess the effects of extending unemployment insurance and unemployment benefit to the self-employed, an unemployment shock was simulated. This does not imply that this is a crisis scenario – the shock is based on average probability of becoming unemployed for the selfemployed in 2007-2015 EU-SILC data. It is rather a modelling strategy which implies randomly assigning the status of unemployment among the group of the self-employed.

depending inter alia on the number of self-employed in the country, the unemployment risk they face 252 as well as on the generosity of the system in place (see Table A.12, column "Gross Spending / Earnings-related scenario").

As a first sensitivity analysis, to study the effects across these Member States of extending access to an insurance of similar generosity, a second scenario - "flat rate scenario” - is considered, where a flat-rate benefit is paid to the previously self-employed who have become unemployed, independently of previous labour income as a self-employed. The simulations considered a flat-rate benefit corresponding to 60% of national median disposable income 253 to the self-employed. The overall range of gross expenditure remains broadly unchanged at 0.1% - 0.4% of GDP, although for individual Member States (IT, LV), resulting expenditure was higher than in the scenario based on actual (less generous) systems in place 254 (see Table A.12, column "Gross Spending / Flat-rate scenario").

Table A.12: Simulated gross and net spending on unemployment benefits for the selfemployed in % of GDP 255

Gross spending Net spending Earnings-rel. Flat-rate Earnings-rel. Flat-rate Italy 0.25% 0.35% 0.13% 0.23% The Netherlands 0.15% 0.13% 0.06% 0.04% Greece 0.21% 0.20% 0.00% 0.00% Latvia 0.06% 0.13% 0.04% 0.12% Croatia 0.37% 0.38% 0.32% 0.33%

These simulations for the two scenarios assume no behavioural reactions by those affected by the changes. The estimated impacts should therefore be considered as conservative; for instance increased underreporting or decreases in employment could lead to higher

expenditure 256 .

Depending on financing decisions by Member States, the simulated expenditure would only partially translate into costs for the public budget. For example, if the self-employed were requested to pay social contributions rates corresponding to the lowest level currently levied on employees, parts of the public expenditure on benefits could be covered in this way. In 257 this example, the estimated net public cost would not exceed 0.3% of GDP (HR), 0.2% (IT), 0.1% of GDP (LV, NL). No positive cost remains for Greece (see Table A.12, columns "Net Spending") 258 .

252 E.g. the unemployment risk for self-employed was the highest in Croatia and lowest in the Netherlands.

253 For a single person household.

254 Differences in results were negligible for GR, HR, NL.

255 Differences relative to a baseline scenario without unemployment insurance for the self-employed.

256 No behavioural reactions by those affected by the changes mean that no jobs will be lost nor gross earnings will decrease as a result of the higher social contributions introduced under the reform. Furthermore, it assumes that compliance is not affected by the reform (i.e. the level of under-reporting of earnings does not change in response to the increase in social contributions).

257 The corresponding simulations are presented in the main body of Brodolini (forthcoming) for HR, IT and LV, and in the annex (scenarios 5 and 6) for GR and NL.

258 It should be noted that in addition to the social contributions there is a further effect on the revenue side of the public budget via income taxes. Social contributions are generally tax-deductible, while in the Netherlands received benefits are additionally taxed. The corresponding net effect on the total volume of income tax paid by households with self-employed members ranges from reductions by between 2 % and 2 ¾ % (GR, HR) to very limited losses by less than ½ % (IT, LV) to a small increase by up to ¾% (NL).

Simulating the cost of comparable unemployment benefits to the self-employed for all EU28

Member States

Complementary simulations were undertaken for all Member States for the introduction of a

comparable system of unemployment insurance to the self-employed, regardless of the

existing systems. So to facilitate the comparison of impacts previously self-employed

unemployed are assumed to receive a benefit calculated according to the same rules across all

Member States. Based on the estimated number of self-employed that would be eligible for

benefits according to the evidence recorded in EU-SILC about the self-employed risk to

become unemployed, corresponding benefit payments to them are simulated. Aggregating

these payments provides an estimate of resources needed.

Notably, two sets of simulations were run involving two different options for coverage: in the

first one, benefits amount to 50% of previous earnings ("replacement rate scenario"), in the

second one there is a modest flat-rate benefit amounting to 30% of national gross median

income ("minimum level scenario") 259 .

To get a magnitude of the estimated costs of these reforms, the total gross spending for the

new simulated unemployment benefit scheme for the self-employed is expressed as a share of

annual gross labour incomes of the self-employed, which can be interpreted as a contribution

rate. 260 It is a gross rate and presents an upper bound in the sense that possible spending

already existing for unemployment benefits paid to self-employed in Member States where

these are covered is not netted out from the estimate 261 .

Figure A.18: Gross contribution rate to finance the unemployment benefit scheme for

self-employed - Replacement rate scenario

Notes: To interpret the figure, note that the red triangles represent the contribution rate in the central scenario where the

benefit amounts to 50% of previous earnings and is paid for a maximum of six months. The other data points show the results

259 More precisely, the simulated earnings-related benefit amounts to 50% of previous monthly gross labour income, capped at 60% of national median gross labour income from employment and self-employment. In both scenarios the benefit is paid for a maximum of 6 months.

260 This contribution rate would finance the new benefit in a revenue-neutral scenario. At the same time, it should be noted that while the cost of extending coverage is expressed as a contribution rate, i.e. relative to the self-employeds' gross labour income, this does not imply that such a system would indeed need to be financed via contributions. They could alternatively also be financed via general taxation.

261 This is an important caveat – Table 29A in the Annex of Brodolini (forthcoming) provides information on the corresponding amounts

recorded in EU-SILC tentatively attributed to the self-employed. The net rates would unsurprisingly be low in the Member States where the self-employed already have coverage; indeed in some existing spending exceeds the simulated gross rates shown in Figure 22 (e.g. in Finland or Estonia, which show the highest gross rates).

of corresponding sensitivity analysis: they are the resulting contribution rates in scenarios where lower (squares) or higher (triangles) values were assumed for the replacement rate (40% and 60%, respectively) or for the maximum duration of the benefit (four and eight months, respectively). For Member States with a low cost in the central scenario the variation is rather limited (e.g. AT) whereas for those where the estimated cost is more significant, the variations in generosity also imply a more substantial variation in cost and thereby in the simulated contribution rate (e.g. CY).

Source: Elaborations of EU-SILC 2009-2015 cross sections

Under the replacement rate scenario, the estimated contribution rates would be rather low in most cases but there is large variation across Member States. The maximum rate is just below 1.2%, while the median rate is 0.25% and the mean one 0.37%. Compared to actual contribution rates in existing unemployment insurance systems, these rates are rather limited. In addition to the central scenario, Figure A.18 also shows the results of sensitivity analysis varying the generosity of the benefit paid to the self-employed 262 . While for most Member States resulting variation is limited 263 , ranges are quite large in the Member States with the highest simulated contribution rates.

The minimum level scenario featuring flat-rate unemployment benefits for the self-employed yields similar results for contribution rates, with a maximum of 1.7% (EE), and a median of 0.28% (mean: 0.44%, see figure A.19); again, there is significant variation of rates across

Member States 264 . Again, sensitivity analysis results in large within-country variation only for a small number of Member States 265 .

The low gross contribution rates resulting from extending coverage in both scenarios may partly reflect possible underreporting of unemployment regarding the self-employed in the underlying dataset 266 .

Figure A.19: Gross contribution rate to finance the unemployment benefit scheme for the self-employed - Minimum level scenario

262 See notes to the figure.

263 The mean is only 0.27% in the low duration scenario and reaches its highest value, 0.47%, in the high duration scenario.

264 While overall differences between the two scenarios are small, whether for a specific Member State the earnings-related or the flat rate system would be more costly depends on the characteristics of the self-employed who lose their job: Low-income self-employed benefit more from a flat-rate system than an earning-related one. Thus in Member States where unemployment hits primarily the low-income selfemployed, a flat rate system is more costly.

265 In the sensitivity analysis, the benefit duration was varied (four or eight months, respectively, instead of six) or the flat rate was modified (set at 20% or 40% of the median wage, respectively, instead of 30%).

266 See Annex 9 for more details on the strategy to address such underreporting.

Source: Elaborations of EU-SILC 2009-2015 cross sections

• b) 

  Extending sickness benefits

In the case of sickness benefits simulations using EUROMOD were not possible due to the

current technical limitations of the model. Neither an alternative model was found to be able

to capture the impact on public budgets of an extension of sickness benefits.

Ad-hoc simulations using EU SILC data, analogous to the ones done for unemployment

benefits have been run to quantify the cost (in terms of contribution rates) of extending access

to sickness benefits to the self-employed. Differently from unemployment benefits that can be

simulated using information on periods spent in unemployment or inactivity, there is no

information in SILC on which self-employed are more likely to receive a sickness benefit, if

this benefit were provided. The simulation therefore imputed benefits according to their

characteristics, exploiting knowledge on employees who receive such benefits.

Based on this approach, the simulations computed the total additional spending for sickness

benefits at the Member State level expressed as a share of the self-employed wage bill (if only

the self-employed would bear the cost of this additional spending) to derive corresponding

necessary contribution rates.

The bars in Figure A.20 represent increases in social security contributions to be paid by the

self-employed if they alone contribute to the financing of the additional sickness benefits.

While rates remain below ¼% in 17 Member States and even below 0.1% in eight countries,

in countries where spending on sickness benefits to employees in the existing systems is high

in relation to the wage bill , the rates can be much higher, reaching up to 3.0% in case of EE,

2.6% for SE, and 2.1% for SI.

Figure A.20: Contribution rate to finance the sickness benefit scheme for the selfemployed

3.50%

3.00% 3.00%

2.56%

2.50%

2.06%

2.00%

1.72%

1.48% 1.50%

1.00%

0.79%

0.63%

0.50% 0.43%

0.36% 0.35%

0.21% 0.21% 0.21%

0.15% 0.14% 0.19% 0.18%

0.08% 0.11% 0.10% 0.08% 0.09% 0.03% 0.03% 0.04% 0.04% 0.04%

0.00% AT BE BG CY CZ DE DK EE EL ES FI FR HR HU IE IT LT LU LV MT NL PL PT RO SE SI SK UK

Source: elaborations on EU-SILC 2009-2015 cross sections

Available data did not allow for simulations regarding other social security schemes covered under this initiative. Nonetheless, according to available data, the EU28 average spending on disability benefits and on survivors' benefits, respectively, is of broadly similar dimensions as

the one for unemployment benefits, 267 which could indicate that this could also broadly hold for related expenditure resulting from extending coverage 268 . Pension expenditure is

significantly higher.

It is worth underlining that the previous simulations do not consider second-round effects. Namely significant indirect costs can arise from possible behavioural changes associated with extending coverage. The associated higher labour cost can reduce labour demand (especially for low-skilled work), while work incentives for people paying contribution rates can also be reduced. In conjunction, both effects lead to employment losses.

Indirect labour market effects also include those on overall incentives to become selfemployed 269 . More generous social security benefits for employees than self-employed increase the opportunity costs of being self-employed. The resulting negative incentive for self-employment would be reduced by reducing (formal and effective) coverage gaps for the self-employed. Conversely, social security contributions for self-employed workers may dampen incentives for entrepreneurship through their effect on disposable income, highlighting the importance of financing choices by Member States. In this respect, it is worth

267 According to ESSPROS data, unemployment spending amounted to 1.4% of GDP in 2014, compared to 2.0% for disability, 1.6% for survivors and 2.4% for family/children. Combined spending on healthcare and sickness amounted to 8.1% of GDP, where no breakdown is available separating healthcare (which can be expected to correspond to the bulk of the expenditure) from the more limited expenditure on sickness benefits which is of relevance in the current context. Spending related to old age amounted to 11.1% of GDP. 268 This would inter alia assume that previously uncovered groups have a broadly similar risk profile to existing scheme members. No evidence underpinning a contrary assumption was found.

269 The following paragraph is partly based on the "Literature review on taxation, entrepreneurship and collaborative economy", TAXUD/2016/DE/315.

noting the results presented above whereby introducing for instance a modest unemployment benefit system for the self-employed could require only very modest contribution rates. Finally, social security may have a positive effect on entrepreneurial activity by creating a safety net in case of business failure. The overall empirical evidence on the net effects is mixed.

Social Impacts

An increase in coverage is expected to have significant favourable social impacts. The protection of previously uncovered groups against risks is expected to increase and thereby reduce uncertainty, precariousness and at risk of poverty. These impacts can be expected to be more pronounced under mandatory coverage given that more people would be affected by this measure.

The above-mentioned EUROMOD simulations relating to coverage of the self-employed by unemployment insurance in five Member States have also been used to estimate the distributional impact of extending coverage of unemployment benefits (see Table A.13 below).

In both scenarios simulated with EUROMOD – the "earnings-related scenario" and the "flatrate scenario", the distributional effects would be modest if assessed from the perspective of the population as a whole. Small reductions in the at-risk-of-poverty rates could notably be observed. However, among households with income from self-employment and no income from dependent employment, the extension of unemployment benefits to the self-employed would generally significantly reduce the incidence of being at-risk of poverty. The reduction would exceed 13 percentage points in one scenario for Croatia, the Member State with the highest unemployment risk in the simulations, while estimated decreases are in the range of 1 to 2 ½ percentage points in most other countries and scenarios 270,271 .

For those self-employed who would be affected by the unemployment shock, the social benefit would be far more substantial: it would significantly soften the impact of job loss on poverty at household level in all five Member States. Their poverty rates would decrease by a minimum of four percentage points, but reaching reductions of 15 – 25 percentage points in the other scenarios and Member States.

Table A.13: Simulated reduction in at risk of poverty rates due to extension of unemployment benefits to the self-employed 272

270 For the more restricted group of households whose main earner is self-employed, reporting that more of 50% of his earnings are derived from self-employment and where the latter earnings exceed the total personal income of other individual within the same household, effects are less pronounced and more mixed. Their at risk-of-poverty rate sees a small decrease of between 0.7 percentage point (pp) and 2.2. pps in Latvia and Croatia, is hardly affected in the Netherlands and increases slightly in Greece and Italy (by up to 0.8 pp). The less favourable results (compared to the overall group self-employed) for instance in the case of Italy may inter alia reflect that the risk of these selfemployed of becoming unemployed is lower than on average in the group.

271 The above-mentioned ad-hoc simulations looking at the extension of unemployment benefits to the self-employed under two scenarios with revenue-neutral contribution rates for all EU Member States yield limited distributional impacts at the aggregate level. This is accompanied by a small but noticeable reduction in poverty for the self-employed in many Member States and scenarios. For many Member States, the reduction in poverty is more pronounced in the earnings-related than in the flat-rate scenarios. The limited quantitative size of results reflects the limited number of self-employed moving to unemployment and the low generosity of the simulated schemes, as well as the cost neutrality assumptions where the self-employed pay the entire cost of the benefit extended to them. Complementary scenarios where the cost of extending the benefit was spread among all in the labour market (again expressed as a contribution rate, the ratio of the cost relative to all labour income from employment and self-employment) still generate small distributional results at the aggregate level, but in many cases a somewhat bigger (albeit still limited) decrease in the poverty risk of the self-employed. See Tables 41-46 in Brodolini (forthcoming) and the corresponding annex with further sensitivity analysis.

272 The corresponding results are presented in the main body of Brodolini (forthcoming) for HR, IT and LV, and in the annex (scenarios 5 and 6) for GR and NL.

Absolute difference relative to baseline

All Self-employed Previously selfemployed unemployed

Earnings Flat Flatrel.

rate Earnings-rel.

Flatrate

Earnings-rel. rate

Italy -0.3 -0.5 -1.3 -2.5 -14.8 -23.4 The Netherlands -0.2 -0.2 -1.8 -1.7 -22.7 -21.9 Greece 273 -0.1 -0.1 -0.6 -0.4 -19.1 -16.9

Latvia -0.1 -0.3 -1.1 -5.9 -3.6 -22.2 Croatia -1.2 -1.1 -12.1 -13.2 -22.9 -24.7

Simulations done to extending formal coverage of unemployment insurance to self-employed

in the case of Latvia 274 shed also light on the social impacts of effective coverage. In the

earnings-related scenario recreated in these simulations, existing time thresholds are extended to the self-employed previously not covered in the scheme. This scenario results in relatively low effective coverage since only 65.1% of previously self-employed unemployed qualify for the benefit. In an alternative flat-rate scenario, effective coverage is 100% by design as the benefit is paid to all previously self-employed unemployed. The simulations result in reduction in the at-risk-of-poverty rate for the households depending on self-employed income 275 much more pronounced (by 4.6 pp) in the second scenario, reflecting differences in both effective coverage and in benefit generosity 276 .

• 3. 

  Quantifying effectiveness, budgetary and distributional impacts of relaxing qualifying periods for unemployment benefit schemes

The effects of adjusting, i.e. relaxing, qualifying periods (for all potential unemployed) 277

have been quantified in dedicated microsimulations by the European Commission's Joint Research Centre (JRC) 278 . In a first scenario, qualifying periods are reduced to 40% of the reference period in 15 Member States 279 . This results in increased effective coverage, by

between as little as 1 pp (in AT, LU and the UK) to up to 13-17 pps. in Latvia, Portugal and

Belgium. The positive impact on disposable household incomes would be strongly focused on

the lowest quintiles of the income distribution, where it exceeds 1% in two Member States (PT and LV). A small decline in the overall at-risk-of-poverty rate would be accompanied by

a more significant reduction for the group of the unemployed, with the most pronounced fall

resulting in Austria, Cyprus and Latvia (by around 1 pp).

On the fiscal side, the policy measure increases the net costs 280 for the unemployment benefit

by between 0.02% and 0.13% of GDP, depending on the Member State. Such a change

273 See footnote 233 above.

274 Results of the above-mentioned EUROMOD simulations.

275 I.e. households with income from self-employment and no income from dependent work.

276 The reduction is jointly driven by the differences in effective coverage and in benefit generosity - for Latvia, the flat rate scenario also has a significantly more generous benefit compared to the average benefit paid in the earnings-related scenario. 277 Changing minimum eligibility conditions by reducing the qualifying period to 50% or 40% of the reference period. 278 ZEW and JRC (2017): Convergence of unemployment benefits schemes in Europe: Evidence at the Micro Level. Study finished in January 2017.

279 AT, BE, BG, CY, IT, LU, LV, NL, PL, PT, RO, SE, SK and the UK - for the Member States where eligibility criteria were already less strict, no changes were simulated.

280 Net costs are calculated as potential gross costs of simulated addition or changed unemployment benefits minus personal income taxes or/and social contribution paid (for the countries, where unemployment benefits are taxed) and minus foregone social assistance payments.

positively affects disposable household income especially for the poorer households which

have a larger propensity to consume. Thus, consumption related tax revenues are also expected to increase. Because of the increased consumption tax revenues, the net fiscal cost of the reforms could be reduced by around 10-15% for most of the countries 281 . In a similar vein,

the simulated net cost takes into account significant simulated savings on social assistance

spending resulting from the policy measure.

At the same time, the study also illustrates the trade-offs involved in addressing effective coverage by adjusting time thresholds as the policy measure would generally reduce the

incentives to take up a job 282 .

Note that estimates of net costs are based on approximations (e.g., the average tax rates by income quintiles of unemployment benefit recipients and ceased social assistance benefits). The social contribution side to offset the costs of the reform is left unchanged. 281 Relatively smaller increased spending is expected in Lithuania (5% of the estimated net costs), followed by Spain and the UK (9% on average).The change in consumption tax revenues due to the reform will depend on the propensity to consume of each income group on the impact of convergence at the bottom of the income distribution and the existing tax burden on consumption in each Member State. 282 The unemployment trap indicator measures the short-term financial incentive for an unemployed person receiving unemployment benefit to take up a job. It measures what part of an increase in gross income due to the changed employment status is taxed away because of direct taxes or withdrawal of social benefits. The higher the index, the lower the initiative to move from unemployment to paid employment. The unemployment trap indicator for Portugal and Belgium would increase by around 5-9 points for both sexes (more strongly for women), from an already high level. Significant increases (by 6-7 point) would be observed also for Latvia and Bulgaria.

Table A.14: Impacts of packages

PACKAGES IMPACTS

ECONOMIC SOCIAL PUBLIC AUTHORITIES

• - 

  Impact on business (SME included) - Employment (mobility, transitions, skills, - Budgetary consequences

-Operating (labour costs) and conduct of labour market segmentation) - Public authorities organisation

business - Working conditions

• - 

  Competitiveness and unfair competition - Social protection (income insecurity, poverty, precariousness, inequality)

  COSTS BENEFITS COSTS BENEFITS COSTS BENEFITS

PACKAGE A  Unfair competition  Social dumping  Negative impacts  No extra cost

(BASELINE)  due to

Absence of level playing 

on the sustainability

Individualisation of of social security reorganisation field below employment social risk systems by of the social and contract forms distorts weakening the security

competition and factor  Increase of revenue base systems

allocation precariousness

 Restrained mobility and  Higher poverty risk employment transitions for increasing share of 

people in the labour Possible increase  Decreasing stabilisation market in cost of last resort

role of social security safety nets (social 

Increase of intra and assistance, inter-generational minimum pensions inequality etc.)

PACKAGE B  Some increase in the cost  Some improvement  Depending on  More pronounced reduction  Noticeable cost of  Possibly

of managing fluctuations in of labour market financing choices, of number of self-employment extending coverage some reduction production in line with dynamism (depending some redistribution and non-standard workers at (to be borne by of the some cost increase of NSE on the rate of opting-in) within the newly risk of poverty, when the social public authorities, expenditure on and SE. covered, as affected risk materialises (will depend i.e. taxpayers, or the last resort

unequally by benefits on the rate of opting-in), concerned groups

 Some improvement of vs. contributions, including after labour market via contributions, safety nets. playing field between taxes or lower wages transitions depending on employment and in the case of NSE if financing choices) contract forms with employers pass on the and some cost for

corresponding effects cost of their Partial de-segmentation of improving on competition and contributions labour market, reducing transparency

factor allocation somewhat inequality, by incentivising the wider use of

 Possible trade-off standard employment.  Limited increase

 Some improvement of between NSE and SE  stabilisation function of labour demand if the Improved transferability and

in social security

social protection rate of self-employed transparency as well as

contributions.

upholding aggregate enrolling is low reduction in coverage demand by providing differences strongly facilitate  Possible financial counter-cyclical labour market transitions and burden to introduce

spending during  thereby, improve matching of Slightly higher risk administrative economic downturns people with jobs. of some reduction in reforms.

employment through

possible adverse  Partial removal of obstacles

effects on labour that may discourage employees demand and supply from taking on the risks associated with selfemployment, improving factor allocation

PACKAGE C  More pronounced  Marked improvement  Limits to the freedom  Marked reduction of  High cost of  Increase in

increase in the cost of of labour market of choice and possibly number of self-employment extending coverage social security

managing fluctuations in dynamism (depending inefficient and non-standard workers at (to be borne by contributions.

production in line with on the rate of opting-in) overinsurance of the risk of poverty, when the public authorities,

more pronounced cost self-employed through social risk materialises, i.e. taxpayers, or the

increase of NSE.  Marked improvement mandatory social including after labour market concerned groups

of playing field between security schemes. transitions via contributions,  Possibly employment and depending on more

contract forms with

corresponding strong 

financing choices), pronounced

De-segmentation of labour and high cost of reduction of

effects on competition  Depending on market, reducing inequality, changes in existing the expenditure

and factor allocation financing choices, more pronounced by incentivising the wider use system to move to on last resort

redistribution within the of standard employment. personal accounts safety nets.

 Marked improvement newly covered, as 

affected unequally by Possible financial

of stabilisation function benefits vs.  Improved transferability burden to introduce of social protection contributions, , taxes or and sharp reduction in administrative upholding aggregate lower wages in the case coverage differences strongly reforms. demand by providing of NSE if employers facilitate labour market counter-cyclical pass on the cost of their transitions and thereby, spending during contributions improve matching of people economic downturns. with jobs.

 Some risk of some

reduction in  More pronounced removal

employment through of obstacles that may possible adverse effects discourage employees from on labour demand and taking on the risks associated

supply with self-employment, improving factor allocation

A NNEX 11: S TAKEHOLDER CONSULTATION

Context - the European Pillar of Social Rights

On 8 March 2016, the European Commission adopted a Communication putting forward a

first, preliminary outline of what should become the European Pillar of Social Rights 283 . The Communication set out the rationale behind the initiative and its role, scope and nature 284 .

On this basis, the Commission launched from March until 31 December 2016 a broad public consultation to gather feedback on the proposed outline to feed into its final proposal. The consultation aimed at discussing existing social rights, the changing realities of the world of work and societies, and the role of the Pillar as part of the social dimension of the Economic and Monetary Union. A European conference took place on 23 January 2017 to wrap up this

consultation process 285 . How to ensure social protection for people in all forms of

employment was a central topic in the consultation process.

Stakeholder stated that "adequate and sustainable social protection should cover all people regardless of their employment status, with due consideration for the most vulnerable, based on an integrated approach of benefit and service delivery, including accessible and sustainable healthcare that takes into account the national circumstances, and quality long-term care based in particular on homecare and community-based services, as well as adequate housing and support to the homeless towards their social reintegration. Adequate unemployment benefits of reasonable duration as well as adequate minimum income schemes should provide a resilient social protection floor capable to labour market reintegration. Pension systems should provide adequate protection against old-age poverty while assuring sustainability with a view to protect the future generation and provide the necessary resources for a childhood

without poverty". 286

Other EU institutions also contributed to the relevant debate. Whereas the European Economic and Social Committee 287 , ("considers it imperative that the debate launched by the

European Commission on developing a European Pillar of Social Rights also encompass the situation of workers engaged in the new forms of employment and, above all, look at how to recognise their status and ensure they have adequate access to social security and social protection systems") and the Committee of Regions 288 ("believes that the new employment forms or the introduction of new common EU minimum standards must always allow for appropriate standards of social protection") brought relevant elements into the debate, the

European Parliament 289 went further and called for a "recommendation enabling all people in all employment forms and self-employment".

283 COM (2016) 127 final i.

284 It was accompanied by two Staff Working Documents: the first one describing key economic, labour market and societal trends on which the Pillar builds and which it should contribute to address; the second one recalling the most relevant legal acquis at EU level. Commission

Staff Working Documents "Key economic, employment and social trends behind the European Pillar of Social Rights" (SWD (2016) 51), and "The EU social acquis" (SWD (2016) 50) of 8 March 2016.

285 Conference website .

286 SWD(2017) 206 final, available online .

287 EESC SOC/542, available online

288 COR opinion 2868/2016, available online .

289 EP resolution 2016/2095(INI) i, available online .

The access to social protecion initiative

Several stakeholder consultations have been performed to inform this initiative. This includes a two-stage consultation procedure of European social partners as stipulated in Article 154

TFEU and an Open Public Consultation.

• 1. 

  Results of the inception impact assessment

The Commission received nine items of feedback to the inception impact assessment from individuals/entities in the Netherlands (6), Belgium, Germany and the UK. The feedback from people/associations in the Netherlands stress subsidiarity and that self-employed would like to be responsible for their decisions they take, including for insuring risks. The contribution from a business organisation in the UK considers voluntary schemes easier to implement across Member States in the short- to medium-term (2-5 years), but require implementation at the national level and could be difficult for governments to agree. The feedback of

EuroHealthNet focus on the great role Social Protection has in the goal of reducing health inequalities and the umbrella associations of the German Social Insurance (DSV) supports the discussion on access to social protection and to make social protection systems future-proof in a changing world of work.

• 2. 

  Results of the first phase Social Partners consultation

The first phase of social partner consultation was launched on 29 April and closed on 23 June 2017.

The trade unions that replied to the consultation were the European Trade Union Confederation (ETUC), Eurocadres and the European Confederation of Executives and

Managerial Staff (CEC), the European Confederation of Independent Trade Unions (CESI), the European Arts and Entertainment Alliance (EAEA) and the European Federation of

Journalists (EFJ). It should be noted that ETUC's reply also took into account the view of 10 ETUC sectorial trade union organisations.

On the side of the employers, replies to the consultation were sent by BusinessEurope, the European Centre of Employers and Enterprises providing Public Services (CEEP),

EuroCommerce, the European Association of craft, small and medium-sized enterprises (UEAPME), the Council of European Employers of the Metal, Engineering and Technology

Based Industries (CEEMET), the Council of European Municipalities and Regions (CEMR), the European Chemical Employers Group (ECEG), the Hotels, Restaurants and Cafés in

Europe (HOTREC), the Retail, Wholesale and International Trade Representation to the EU (EuroCommerce), the Employers' Group of the Committee of Agricultural Organisations in the European Union (GEOPA-COPA) and the World Employment Confederation (WEC).

Identification of issues related to access to social protection

The social partners agreed, largely, that there are problems related to access to social protection for workers in non-standard forms of employment and for the self-employed.

However, employers stressed that grouping very different types of employment forms together under the heading of 'non-standard' is not appropriate as it ignores the diversity between these different forms and the need and desires of those working under them, also in terms of access to social protection (BusinessEurope). WEC highlighted in addition the heterogeneity within the self-employed. UEAPME called for a solid mapping exercise of the national policies, the availability and offer of tailored social protection schemes, and the takeup broken down by different groups. CEEP highlight the different labour market and social models, which explain different levels of protection.

On the other hand, trade unions found that the identification of issues goes in the right direction, but could, nevertheless be improved upon by specifying that new forms of work entail poor job quality and insecure conditions, which lead more to a lack of protection against life risks than to real job opportunities (ETUC, Eurocadres). The issues could be extended to a right to collective bargaining, fair remuneration for all and the right to freedom of association for all workers (EFJ). Trade unions further underlined that consistency should be ensured with proposals about information on the social security system in the Written

Statement Directive. (ETUC, Eurocadres).

Most relevant branches of social protection and employment services

Trade unions consider all branches to be equally important and promote an integrated and holistic approach to social protection, notwithstanding the points above about the distinctions between the policy fields, including all branches covered by the ILO convention 102 (and the

Regulation on social security coordination). If a prioritisation has to be made, CEC would favour those who are not universal. Access to related employment services should be granted to all.

The picture at the employers' side is mixed. BusinessEurope considers that the question presupposes that there will be an EU initiative in this field (which the organization is not in favour of). EuroCommerce questions the feasibility of the classification of social protection strands, while COPA prioritises accident at work and occupational diseases. HOTREC refers to social protection as a national level competence, but stresses that skills, education and training should be easily accessible to all citizens.

Personal scope of an EU initiative

In general, trade unions are in favour to have a wide personal scope and to include all workers in non-standard forms of employment as well as the self-employed in an EU initiative.

However, some call in the first place for the definition of principles to set a common understanding at EU-level of the legal nature of the different forms of employment (CEC,

CESI).

Employers are rather reluctant to reply the question on the personal scope of the EU initiative because they are not in favour of an EU initiative (BusinessEurope) or they refer to the principle of subsidiarity (HOTREC) or they consider that the personal scope would depend on the nature of the respective initiative (EuroCommerce). CEMR emphasizes on the need for more clarity in the definition of the legal nature of the different forms of employment whereas

COPA would include all workers in non-standard employment in an EU legal initiative but would include the self-employed only in a recommendation.

EU legislation and EU level instruments

BusinessEurope, UEAPME and EuroCommerce do not consider that changes to EU legislation in this field are needed or appropriate. The open method of coordination and the

European Semester process, including benchmarks, would be the right tools for mutual learning and exchanges of good practice. The aim should be to improve national policy responses by learning from relevant other national practices. Other employers' organisations draw the attention to sectorial social dialogue (COPA), Member State competences and the principles of subsidiarity and proportionality (CEEMET, ECEG, EuroCommerce, WEC and

CEEP).

The trade union side considers that improvements should be made to EU legislation. ETUC and CESI see the need for better enforcement of existing legislation, especially at national level. CEC mentions that the need for reducing the administrative burden that can be associated with the provision of information concerning an individual's employment situation and ensuring an effective transferability of all entitlements.

Willingness to enter into negotiations

All of the trade unions have indicated their willingness to enter into negotiations. However, in the event that the EU social partners do not agree to negotiate, or in case negotiations do not lead to a successful outcome, trade unions urge the Commission to come up with a legislative proposal.

Employers are not considering initiating a dialogue under Article 155 TFEU because they do not see the need for EU level action, apart from peer learning and exchange of practices

(BusinessEurope) or because subsidiarity prevails in this topic (HOTREC, UAPME) and it can be better dealt with by national governments (EuroCommerce) or in sectorial social dialogue (COPA). CEEP underlines that valid indicators have to be defined in the first place.

Scope of the consultation

ETUC missed clear proposals on the creation of quality jobs – with agreed indicators to track them, including for access to social protection. Therefore, ETUC considered the identified key aspects of insufficient access (gap in access to social protection, lack of transferability of rights as well as lack of transparency about their social protection entitlements) as important challenges to be tackled, but would put the emphasis on challenging divergent social rights of people with respect to social protection and employment services.

Additional consultation of social partners

DG Employment and Social Affairs has conducted bilateral meetings with all 16 social partners who submitted a contribution to the first phase social partner consultation. These meetings took place between 14 July and 5 September. The purpose of the bilateral meetings was to focus on technical clarifications and gathering additional information where ever possible. The information received is presented in the relevant chapters of this document.

• 3. 

  Results of the second phase Social Partners consultation

The second phase of social partner consultation was launched on 20 November and closed on 5 January 2018.

The trade unions that replied to the consultation were the European Trade Union Confederation (ETUC), Eurocadres and the European Confederation of Executives and

Managerial Staff (CEC), the European Confederation of Independent Trade Unions (CESI), the European Arts and Entertainment Alliance (EAEA) and UNI global union europa and the

World Employment Confederation (WEC).

On the side of the employers, replies to the consultation were sent by BusinessEurope, the European Centre of Employers and Enterprises providing Public Services (CEEP),

EuroCommerce, the European Association of craft, small and medium-sized enterprises (UEAPME), the Council of European Employers of the Metal, Engineering and Technology

Based Industries (CEEMET), the European Chemical Employers Group (ECEG), and the Hotels, Restaurants and Cafés in Europe (HOTREC).

The objectives of coverage, transferability and transparency

Trade unions share the initiative's objectives of coverage, transferability and transparency. Formal and effective coverage is of utmost importance for ETUC and access to adequate social protection should be added. Contributions and benefits (for self-employed and nonstandard workers) should be as equal as possible to those for standard work contracts (CESI).

Tying social protection rights to individuals should not lead to an individualisation of social protection and the collective dimension must remain (EAEA). Full transferability and accumulation of these rights, benefits and entitlements should be ensured, no matter the form or duration of employment (UNI Europa). The importance of access to training and employment services was also highlighted (Eurocadres). All trade unions agree with the need for higher transparency.

The picture at the employers' side is rather mixed. Most employers' organisations agree with the three objectives to a large extent but highlight the Member States' competence in this field

(BusinessEurope, EuroCommerce, CEEMET, CEEP and UAPME).

ECEG does not share the view that non-standard workers and self-employed have limited access to social protection and sees no need to explore a common definition of a worker at EU level.

HOTREC agrees on the need of increased transparency but argues that the subsidiarity principle prevails and favours therefore a mapping of transferable systems at national level.

The options of mandatory or voluntary formal coverage

All trade unions favour mandatory formal coverage, equalised to those of standard workers (ETUC, EAEA) but national traditions should be taken into account (CEC).

Employers expressed different views on these options. ECEG agrees that social protection should be mandatory but employees should have the right to choose the form of coverage

(public or private). BusinessEurope highlighted the freedom of choice for the self-employed. CEEP and EuroCommerce favour voluntary formal coverage in order to take into account the diversity of forms of employment of the heterogeneity within the self-employed. HOTREC,

CEEMET and UAPME referred to subsidiarity that should prevail.

The appropriate action to ensure effective coverage

Trade unions and employers have opposite views on the appropriate action. Trade unions favour extending and adapting existing schemes to people in all forms of employment

(ETUC), by going as far as possible under the Treaties (CESI). An EU directive with binding minimum principles and requirements is mentioned (Uni europa, eaea) but it should not undermine existing EU acquis, nor lower existing national standards, nor lead to downward harmonisation of rights.

On the other hand, employers' organisations are opposed to legally binding actions in this field as they argue that the subsidiarity principle prevails (ECEG, HOTREC) but would welcome the use of the OMC, the European Semester, an exchange of best practices, benchmarking and the social scoreboard (BusinessEurope). Some employers' organisations

(CEEP and EuroCommerce) consider a Council Recommendation to be suitable.

The minimum requirements appropriate to ensure transferability and transparency

According to trade unions, labour market trends call for a certain common minimum standard in the field of social protection (CESI). Full portability should be guaranteed by tying social protection entitlements to individuals through an equalised calculation and aggregation but without leading to an individualisation of social protection (ETUC).

Some employers agree with the principle of transferability of rights but call for an impact assessment before creating legally binding realities (ECEG) or argue that subsidiarity prevails and that a mapping of existing transferable systems should be favoured (HOTREC).

Others recommend the EU to limit itself to general provisions (UAPME) or would like to limit transferability to minimum social protection rights (EuroCommerce). The administrative cost of transferability should be kept at a minimum and no additional costs for the new employer to take on the provision of the transferred rights should occur (CEEP).

Willingness to enter into negotiations

All of the trade unions have indicated their willingness to enter into negotiations. However, ETUC was convinced that the conditions for formal negotiations no longer exist and other trade unions urge the Commission to come up with a legislative proposal before the 2019

European elections. Employers are not considering initiating a dialogue under Article 155 TFEU because they do not see the need for EU level action.

• 4. 

  Results of the public consultation

The open public consultation (OPC) was open from 20 November 2017 until 15 January 2018. 119 respondents replied to the OPC of which 62 organisations, 7 administrations, 37 citizens and 13 others (e.g. research institutes, networks). Most replies came from countries with well-developed social protection systems like Germany (18), Belgium (16), France (15) and Sweden (14). 15 position papers were received, mainly from Germany (5), Belgium (4),

France (2) and the UK (2), and covering regional and central governments, social protection providers, cooperatives, organisations of self-employed, a trade union, a n on-governmental organisation and a company.

Challenges

"Do you agree with the identification of the challenges outlined in the background document?"

Around 2/3 of the respondents agreed with the challenges identified by the Commission. The highest agreement was on gaps in effective coverage, followed by regulatory complexity.

I agree (replies out of 119 replies) Gaps in formal coverage 66

Gaps in effective coverage 78

Insufficient transferability 65

Insufficient transparency 69

Regulatory complexity 75

There are other challenges 62

Principles

"Do you think that the following general policy principles should be pursued by a possible EU initiative?"

The great majority of the respondents largely agreed that the principles should be pursued by a possible EU action include the provision of adequate social protection to all workers regardless of their employment relationship, tying rights to individuals as they work and making rights transferable, making information transparent and simplifying administrative requirements. Only 14 respondents did agree with the option that 'no action was required'

I agree (replies out

of 119 replies)

Regardless of the type and duration of their employment relationship, 85 workers, and, under comparable conditions, the self-employed, have the right to adequate social protection

Tying rights to individuals as they work (and not to the contract) and 72 making rights transferable

Making rights and related information transparent 86

Simplifying administrative requirements 82

No action required 14

Options

In relation to the options listed, 69.7 % of the respondents said that social protection rights and obligations should be mandatory for every kind of job, irrespective of the type of contract and 56.3 % of the respondents consider that the mandatory protection and contributions of non-standard workers should be aligned to the level of standard workers

64.7 % or the respondents think that social protection rights and obligations should be mandatory for all self-employed and 47.1 % of the respondents think that mandatory protection and contributions of the self-employed should be aligned to the level of standard workers.

Finally 52.94 % of the respondents think that a single social protection scheme covering all people in employment would be the most appropriate way to ensure effective coverage. Even 54.6 % of the respondents consider a single social protection scheme appropriate for selfemployed.

Instruments

Almost three quarters of the respondents (72.3 %) highlight the need for an EU level action.54.6 % of the respondents consider the introduction of a new EU legislation (e.g. a directive) as highly effective. Only 4.2 % of the respondents think that soft EU legislation (e.g. a Council recommendation) would be highly effective.

Impacts

According to the large majority of the respondents, the impact of making social protection rights mandatory to all forms of employment would be positive for the European society, the labour market, the workers, public finances and the economy

"According to your assessment, the impact of making social protection rights mandatory in all forms of employment would be positive, neutral or negative for: The economy"

Positive (replies out of 119 replies)

Competitiveness 73

Resilience and adaptability 72

Economic growth 72

but rather neutral on small and medium sized enterprises (SMEs)

"According to your assessment, the impact of making social protection rights mandatory in all forms of employment would be positive, neutral or negative for: Small and medium sized enterprises"

Neutral (replies out of 119 replies)

Cost 51

Competition 45

On the other hand, the impact of making social protection rights voluntary where gaps currently exist would have a neutral/negative impact on the European society, the labour market, the workers, public finances and the economy.

"According to your assessment, the impact of making social protection rights voluntary where gaps currently exist would be: The economy"

Negative (replies out of Neutral (replies out of 119 119 replies replies)

Competitiveness 45 43

Resilience and adaptability 49 40

Economic growth 49 41

but rather neutral on small and medium sized enterprises (SMEs)

"According to your assessment, the impact of making social protection rights voluntary where gaps currently exist would be: Small and medium sized enterprises"

Neutral (replies out of 119 replies)

Cost 55

Competition 62 Position papers

18 respondents sent position papers of which 4 (local, regional and central) governments, 4 direct selling organisations, 2 social insurance organisations and 2 civil society organisations. Most replies were received from Germany (7) followed by Belgium (5). In general, the senders agreed with the challenges but called for the respect of the subsidiarity principle.

On the options the pictures was mixed. Whereas some are in favour of voluntary schemes (Deliveroo, Scottish government, Federation – direct selling), others defend the idea of mandatory coverage (Smart).

The senders mentioned as instruments to be used, a Council Recommendation (Scottish government), a better use of the open method of coordination (German central government, Federation – direct selling), the European Semester (Dutch Counsellor for Employment, REIF) and the exchange of best practices (Franchiseverband, CDH, BDD). Others favour binding instruments such as a Framework Directive on Minimum Income (Save the children and Social platform) or an adequate legal framework for cooperatives of independent workers and salaried entrepreneurs (CECOP).

As regards financing, some proposed a broadening of the financing base (Smart) whereas others defend a fiscal neutral approach (County of Vienna). A common point for the direct selling sector in Germany was the need to adapt social protection thresholds to real income and not to a fictive income (BDD, CDH, Franchiseverband).

Targeted Hearings

· Exchange of views with Social Protection Committee (1 December2017)

Delegations stressed the importance of the topic, and highlighted its complexity: solutions might differ for non-standard workers and self-employed (DE, MT, NL, HU, DK) and by social protection branches (DK). A couple of delegations stressed that the issues are not so prominent in their country where social protection system is not based on the type of employment relationship (FI, HU). One country stressed that the financing of social protection is a key issue wherever social protection is financed by salary-based social contributions (FR). Some countries are already in the process of reforming their social protection systems (IT, PT, UK). Delegations were in favour of encouraging mutual learning activities on this topic; one country (IT) supported the idea of a strong legal instrument; two countries (AT, SE) were open to such idea, while ten countries (DE, DK, MT, NL, PT, HU, FI, LT, UK and IE) expressed doubts about the possibility to adopt a legal instruments while respecting subsidiarity in this field.

· Hearing with representatives of self-employed (14 December 2017)

Out of the 24 organisations invited, only seven attended the hearing. All participants agreed with the Commission's analysis but have underlined some specific issues. In particular, several drew the attention to the trend of people combining several jobs at the same time but the phenomena is still difficult to quantify (SMART, UBER and Faircrowdwork/IG Metall). SMART recalled that traditional self-employed had no need for social protection as they had patrimonial assets at their disposal. The situation is very different for the "new self-employed" with have production tools like a laptop and a smartphone but not a high income. Faircrowdwork/IG Metall stressed that there are 2.2 million self-employed in Germany without access to pension which will most probably increase old-age poverty in the future. FEANI focused on the particular situation of mobile workers in the EU and the difficulties to calculate a pension if a worker has worked in several European countries. There is a need for an overarching platform. According to FEANI, job polarisation is a major trend with the risk of the disappearance of the middle-class, the backbone of social protection systems.

Regarding the organisation of social protection and the need for EU level action, views were divergent. Some participants favoured a system where all people have the same access to social protection (SMART, Faircrowdwork/IG Metall), without the possibility to opt-in or opt-out and with the obligation for all to contribute to this system (Faircrowdwork/IG Metall). They have also favoured broadening the financing basis (SMART, Faircrowdwork/IG Metall, PGEU) so that other sources of income, beyond labour, contribute. They also favoured the establishment of EU mandatory minimum standards (SMART) or EU focus on some social security branches like sick pay, maternity leave, unemployment benefit and pensions for the self-employed (FEANI). One participant underlined that self-employed should have the right to collective bargaining (Faircrowdwork/IG Metall). Another participant (UBER) underlined that everybody should have access to an affordable and reliable safety net. Benefits should be attached to workers and be made portable

On the other side, some participants recalled the subsidiarity and proportionality principles and that social protection of the self-employed should be voluntary (Direct Selling Europe, PGEU). They also indicated the European Semester as the most appropriate instrument to address the challenges at hand (UBER, Direct Selling Europe)

· Hearing with civil society (15 December 2017)

13 organisations attended the hearing. Participants agreed with the challenges identified by the Commission but underlined the need to add analysis in relation to some groups: the gender pension gap for women (European women's lobby), specific needs of disabled people (Disability Forum) and of migrants (PECON). Eurohealthnet highlighted that some selfemployed are not properly protected because they underestimate their risk or can simply not afford it.

Some recalled that the same challenges had been identified in the1992 recommendation on the convergence of social protection objectives and policies (CARITAS) and the difference compared to the past is that whilst in the past the challenge related to combat the informal economy and tax evasion, today there are legal ways of circumventing social contributions (CARITAS, EESC). EAPN questioned the coherence of Commission messages and named the joint employment report and the general message to reduce social contributions in order to make work cheaper.

Most participants advocated for universality and adequacy of social protection (CARITAS, European network for independent living, COFACE, Eurodiaconia, Social platform, EAPN, Eurocities, Eurochild ), the need to reduce conditionality criteria to reduce discrimination and unfair treatment (social platform), the need to reduce administrative barriers and complexity (Eurodiaconia, PECON, Disability Forum) to foster transparency of rights (Eurohealthnet). Adequacy should be defined as a combination of an income above the poverty line and the reference budgets (a basket of goods and service which covers basic needs) which takes into account the cost of living (EAPN, Eurodiaconia). Several argued for a rights-based approach (EAPN, Eurodiaconia, Social Platform) and for rights that are tied with the individuals (Eurodiaconia). Social Platform warned that social security is not a commodity and privatisation of social insurance will increase inequalities. Several participants referred to the need of addressing taxation to ensure the financing of social security, for example through a financial transaction tax (European women's lobby) or ensuring a buffer from general taxation (EESC).

Regarding EU action, several participants indicated their support to a directive ensuring minimum standards (Eurohealthnet, EESC, Social Platform) and recalled the importance of enforcement (COFACE, EESC). The European women's lobby thinks that the Commission should come up with a larger social protection strategy in order to complement the European employment strategy. Eurocities called on the mainstreaming of social protection issues in European legislation - e.g. public procurement.

· Hearing with self-governing bodies and private insurers (15 December 2017)

Most participants agreed with the challenges at EU level identified by the Commission (ESIP, DSV, French Social Insurance, Austrian Social Insurance) but some underlined that there are no problems in some countries (Germany - DSV) and others referred to ongoing national reforms to address the gaps on social protection (French Social Insurance) or regulate the dependent self-employed (Austrian Social Insurance).The Dutch Pension Fund calculates a huge pension gap in the future for self-employed in the Netherlands). Several have underlined the fact that people under-estimate the risk (Zurich Insurance, Belgium Invalidity Insurance) due to lack of awareness. The Belgium Invalidity Insurance sees a risk of having low cost private schemes (through platform companies) versus more expensive public schemes and the need to break barriers between private and professional risks for self-employed (distinction becoming more and more difficult).Private insurers draw the attention to the fact of people underestimating risks due to low awareness.

Some participants referred to their own ongoing research to map the situation across the EU for all branches of social security (ESIP) and on the income protection gap (Zurich Insurance).

In general self-governing bodies are in favour of a universal system integrating self-employed in the general scheme to guarantee transparency and transferability (Belgian pension fund, French Social Insurance) since different schemes for professional sectors and at regional level lead to complexity. The Dutch Pension Fund sees the need to review the occupational pension system in order to cover all forms of employment. Together with Zurich Insurance, they find auto-enrolment with opt-out as an interesting avenue for self-employed. The Belgian Pension Fund highlighted that social protection can be complemented by private insurances.

Some self-governing bodies would be open to the establishment of EU minimum standards (Belgian pension fund, Austrian social insurance) if national systems are duly taken into account. Others however (ESIP, DSV, Dutch Pension Fund) underlined the subsidiarity principle which would limit EU action to a better use of the European Semester, the OMC and an exchange of best practices. ESIP considers that some more reflection would be necessary regarding the abolishment of thresholds and ceilings and their impact on accrued rights.

Some noted a shift from contributions to taxation in terms of financing sources (Austrian Social Insurance, Belgian Pension Fund).The French Social Insurance defended that all kinds of revenue, including coming from a side-job, should contribute to social protection and that platforms should have the possibility to declare platform workers.

A NNEX 12: W HO IS AFFECTED AND HOW ?

Key recommendation to Member States

Public administration will have to ensure the following actions:

• - 

  review their legal framework in order to ensure the inclusion of categories of non-standard workers

actually not covered.

• - 

  review their legal framework in order to ensure that time requirements and the duration of benefits are designed so to not disadvantaging people in non-standard employment and in self-employment.

• - 

  review their legal framework in order to ensure the transferability of accumulated entitlements across

  different schemes.

  • - 

    review their legal framework in order to set up voluntary schemes for the self-employed in case they

  do not exist.

  • - 

    Set up programs that ensure accessible and transparent information on a generalized and personalised basis.

• I. 

  Overview of Benefits (total for all provisions) – Preferred Option

  Description Amount Comments

Direct benefits

More pronounced reduction in Not possible to quantify because it depends

inequality through the deon companies hiring choices, on the way segmentation of the labour schemes are financed and on the take-up of

market voluntary schemes by the self-employed

More pronounced reduction of Not possible to quantify because it depends

number of self-employment and on the taking up of voluntary schemes by the non-standard workers at risk of self-employed

poverty, when the social risk materialises (will depend on the rate of opting-in), including after labour market transitions

Improved transferability and Not possible to quantify because it depends

transparency as well as on how many people are not aware of their reduction in coverage rights and on the amount of schemes which

differences strongly facilitate

labour market transitions and do not allow to transfer, preserve and

thereby, improve matching of aggregate rights.

people with jobs.

Indirect benefits

Some improvement of Not possible to quantify stabilisation function of social protection upholding aggregate demand by providing counter-cyclical spending during economic downturns

Some improvement of playing Not possible to quantify

field between employment and contract forms with corresponding effects on competition and factor allocation

Partial de-segmentation of Not possible to quantify

labour market by incentivising the wider use of standard employment

Higher taxable basis if the Not possible to quantify

adjustments in social security contributions steer a movement from forms of employment which have a simplified taxation system to standard employment and more highly taxed employment contracts

Possibly some reduction of the Not possible to quantify

expenditure on last resort safety nets.

Partial removal of obstacles that Not possible to quantify

may discourage employees from taking on the risks associated with selfemployment, improving factor allocation

II. Overview of costs – Preferred option

Citizens/Consumers Businesses Administrations

One-off Recurrent One-off Recurrent One-off Recurrent

None Limited decrease None Possible Possible None

in disposable increase of financial

income for a more labour cost burden to

sizeable group introduce (previously administrativ uncovered and for e reforms. whom the risks do not materialise) – through

Formal contributions,

coverage Direct costs taxes or lower

wages in the case of NSE if employers pass on the cost of their contributions (Depending on the financing of social security schemes).

Indirect costs None Slightly higher None None None None

risk of some reduction in employment through possible adverse effects on labour demand and supply.

Effective None None None None Possible None

coverage financial

and Direct costs burden to

transfera introduce

bility administrativ e reforms.

Indirect costs None None None None None None

Transpar Direct costs None None None None Possible None ency financial

burden to introduce administrativ e reforms.

Indirect costs None None None None None None

The SME Test – Summary of results

• (1) 

  Preliminary assessment of businesses likely to be affected

The initiative will apply to all employers. It will, however, particularly affect all

employers of non-standard employees and, to less extent, companies working with selfemployed.

The data used are from the ESDE 2015 and allow determining to which extent the

companies use these typologies of contracts. The self-employed with and without

employees work mainly with micro enterprises (1-9 employees), and much less with

SMEs and large companies. On the other hand, non-standard employees are particularly

used by the public administration (28%) and SMEs and to a less extent also by large and micro enterprises.

• (2) 

  Consultation with SMEs representatives

SME's representatives have been consulted through the two social partners' consultations (See

and also through dedicated hearings. The results of these consultations have been Annexes11

presented in Annex 10. Moreover, a survey has been carried out by Open Evidence and 16)

interviewing a total of 612 employers across 10 Member States, the results of which are

explained in point 3 below.

• (3) 

  Measurement of the impact on SMEs It is firstly important to notice that only measures related to formal coverage can have an (See

impact on employers since they can affect the administrative burden and the labour cost. Annex 16)

The policy measures concerning effective coverage, transferability and transparency do

not affect companies.

Secondly, the questions in the survey referred to a mandatory extension of coverage both for non-standard workers and for the self-employed. Therefore, the results presented

below can be considered as an upper bound with respect to the effects of the initiative,

which requires mandatory coverage only for the non-standard workers.

The survey asked what would be the impact of equalizing social security across all forms

of employment on employers' administrative burden. To this question 21% answered that it would increase to a large extent the time you need to fulfil the bureaucratic

requirement, 31% said that it would increase to some extent and the 48% answered that it

would not make a big difference.

Moreover, when they were asked how the extension of coverage would influence their

hiring preferences, 9% replied that they would rely on less extent on the services of selfemployed

 because they would become relatively more costly, they answered that they

would still resort to temporary contracts (11%) and to the self-employed (22%) to retain flexibility and not incur in fixed costs, while 21% answered that they would increase open

ended employment contracts if the cost of labour are equalised and the 49% answered that

they would not change their hiring behaviours.

• 4) 

  Assess alternative options and mitigating measures

No alternative options neither mitigating measures

A NNEX 14: M ONITORING INDICATORS

On the basis of the monitoring indicators below, specific targets could be agreed with

Member States. The targets would express, in quantifiable terms, the objectives of the

Recommendation that would be to close the gaps identified in the mapping presented in the

IA. For instance, targets could be specified in terms of formal coverage as follows: the number of groups that do not have formal coverage in a given branch in a given Member

State should be zero within three years after the publication of the Recommendation, or the

number of Member States where there are groups that lack coverage in a given branch should

be zero within the same period (it reached nine for self-employed regarding unemployment

insurance in 2017, for instance, see tables A.2 to A.5). As regards effective coverage, targets can be formulated on the basis of data on the number of countries where time thresholds are

limiting effective coverage for non-standard workers and the self-employed (table A.6 and

A.7). The same goes for transferability (table A.8). As regards transparency, the target could

be: all MS should provide individualised information within three years after the publication

of the recommendation (currently, almost half MS do not provide such information, see table A10).

Table A.15: operational objectives and monitoring strategy

Specific objective Operational Indicators* Source objective

Formal coverage Decrease in share of nona) Categories of nona) MISSOC standard employees who standard employees who

have no formal coverage have no formal coverage,

by social security branch, b) EU-LFS, where by country (allowing exnecessary complemented ante/ex-post evaluation, by national sources.

on the basis of the mapping present in Annex 6)

• b) 

  Number of nonstandard employees in these categories

Decrease in share of selfa) Categories of selfa) MISSOC provides employed who have no employed who have no information on access to formal coverage formal coverage, by social security of selfsocial security branch, by employed by branch. country (allowing exante/ex-post evaluation,

on the basis of the b) LFS can provide mapping present in estimation on the number Annex 6 and a later of self-employed in each update) country and per

occupations or economic sector.

• b) 

  Number of selfemployed in these categories

Effective coverage Ensure that rules Length of qualifying MISSOC provides governing contributions period, waiting period (if information on qualifying and entitlements do not relevant) by employment period, working period, restrain the possibility to status, by branch and by waiting period,

accumulate and access country calculation and duration adequate entitlements on of benefits and material

the basis of type of Replacement rate; and scope by employment employment relationship duration of benefits by status and by branch

and employment statuses. employment status, by branch and by country

Comparing the first mapping presented in Annex 6 of the IA to a later update would allow an ex-ante/ex-post evaluation.

Ensure that the Social contributions as MISSOC provides contributions to social percentage of income, per information on applicable security are proportionate employment status and rates, allowing to check to the contributory per income level, whether exemptions or capacity of workers and comparing by lowered rates on the basis self-employed (including employment status of income apply to all ensuring progressivity) employment statuses and ensure a level-playing

field EU-SILC+EUROMOD

Enable transferability of Number of Member Not available in MISSOC entitlements States and social security on a regular basis branches where legal frameworks enable transferability of entitlements

Enable aggregation of Number of Member Not available in MISSOC periods across schemes States and social security on a regular basis branches where periods can be aggregated across schemes

Transparency Increase in number of Number of Member Specific ad-hoc study, not Member States and social States and social security available in MISSOC on security branches where branches where generic a regular basis

generic information is information is available

available (allowing ex-ante/ex-post

evaluation, on the basis of

the mapping present in

Annex 6 and a later

update)

Increase in number of Number of Member Specific ad-hoc study, not Member States and social States and social security available in MISSOC on security branches where branches where a regular basis personalised information personalised information is available is available (allowing exante/ex-post evaluation, on the basis of the mapping present in Annex 6 and a later update)

Outcome indicator: guarantee a decent standard of Poverty reduction effect EU-SILC living of social benefits for selfemployed,

atypical employees and standard employees, per country (allowing ex-ante/ex-post evaluation, on the basis of statistics presented in this IA)

• * 

  The above table provides a tentative list of indicators. The feasibility of collecting such indicators will be further analysed (in terms of data availability, sensitivity, sample size issues, etc.).

  A NNEX 15: S UMMARY OF CASE STUDIES

In the context of a changing world of work, many new forms of non-standard work developed in EU Member States. These new forms of work can have lower levels of protection through social security systems. However, some of the gaps can be compensated by derived rights at household level or by private insurances. The Commission decided to have a closer look at six forms of employment in six Member States:

 Mini-jobs in Germany  Civil law contracts in Poland  Self-employed without employees in the Netherlands  Casual worker in Romania  Economically dependent self-employed in Spain  Project workers on continuous collaboration projects in Italy

• 1. 

  Mini-jobs in Germany

Mini-jobs have become a key feature of the German labour market, as one fifth of dependent employed hold a mini-job. Among the 7.6 million mini-jobbers, 4.9 million have mini-jobs as a main job, the remaining ones as a second job. The by far most important form of mini-jobs (geringfügige Beschäftigung) is regulated through a maximum monthly net income from work (accumulation of contracts is possible in case mini-jobs are the only income source from dependent work) up to the defined income ceiling, sine 2013 set at EUR 450 per month or EUR 5 400/year. In the long term, the share of persons with a mini-job as a second job on all mini-jobbers increased constantly. Mini-jobs, in particular if they are main jobs, are typically carried out in sectors like retail trade, repair services, accommodation and restaurant, health, social and household services as well as administrative support. Nearly half of mini-jobbers worked in unskilled or semi-skilled activities as compared of a fifth among standard dependent employed.

For mini-jobs social security contributions level differ depending on the activity: they are lower for employers in the private household sector than for those in the commercial sector. The sum of employer and employee contributions for mini-jobs is lower than for standard dependent employed. Employees have the option to opt out of paying their contributions to the pension system. Contributions to the health care system are solidarity payments that increase the budget of the statutory health insurance without creating health coverage for the employed mini-jobber. In addition, there are small contributions to the insolvency insurance, accidence insurance and a sickness/maternity risk lump sum. No contributions are made to the unemployment insurance and the statutory old-age care insurance schemes. The employer of mini-jobbers transfers in general 2% of income taxes based on the gross wage. This means that this part of earned income is not taxed according to individual tax rate based on the totality of income and will consequently not be added up to other income in the individual income tax declaration. This represents saved taxes for most mini-jobbers, in particular if their income from the main job or of the spouse is high.

Specific rules regarding access to employment (early retirees), opt-out rules for pension contributions and the prevalence of derived social protection rights, as well as the income tax system in general and the taxation system for married couples in particular, set different incentives for working as a mini-jobber. For employers the registration of mini-jobs follows eased administrative procedures. The composition of mini-jobber reflects the different incentives.

More than one third of mini-jobber have acquired rights from the social security system from their main dependent employment. In 2014, more than a sixth of mini-jobbers were retired and 19% were in education and had usually derived rights to the health care system. About 11% were unemployed and most of them will also have derived rights to health care. A fourth of mini-jobbers were housewives. The vast majority of them have derived rights to the statutory health care system and to the pensions of their spouses. Overall, only a fifth of minijobbers acquire own pension rights, as the others make use of the opt-out regulation. About 4% of mini-jobber in the commercial sector are not covered by the statutory health insurance through derived rights and will need to get private health insurance. These will have a strong incentive to get a job subject to social security payments. One group more often affected by a social security gap is for example divorced women. They have no derived rights to access health insurance. The risk of getting welfare net transfer at older ages is higher for divorced people, as a third of them would claim it, while widows have in general not a higher probability of getting this minimum income benefit. A fifth of widows getting the minimum income have neither own nor derived pension rights. The main underlying reason for getting low pensions and being at risk of old age poverty is linked to broken working biographies and a low number of hours worked, in particular among singles.

Other gaps may result from the fact that mini-jobbers are often not well informed about their rights, e.g. to paid holidays or rest time, accident or sick pay. Estimations of costs of minijobs in terms of lost income to the pension and health insurance scheme range between EUR 1.1 and 1.6 bn per year. Net fiscal costs (including social security contributions, taxes and transfers) vary between 0.6 billion and 0.8 billion according to a study of IZA but may be significantly higher according to others, depending on assumptions made about the supply and demand effects as well as the income distribution. It has been argued that on cases minijobs have led to net employment creation the tax and contributions to the social security system losses are justified. However, also the costs of the net employment creation would need to be taken into consideration. Overall, deadweight effects as well as substitutions effects are likely to be large. Furthermore, major uncertainties remain about the assessment of labour supply and labour demand effects of mini-jobs. One important potential effect of the mini-job regulation is the limitation of the number of hours worked. Effects of a restricted labour supply in times of labour shortages, and thus foregone growth potentials would need to be taken into account.

Regulations on mini-jobs have existed for many decades. In the 1960s and 1970s the main objective was to enhance labour supply mainly of those who were participating in the labour market and to expand the number of hours of those in employment in order to overcome labour shortages. Mini-jobs regulations were reformed in 2003 in the context of the major labour market reforms, the so-called “Hartz reforms”. Mainly, the objectives have changed in the context of high unemployment. Current reform discussions are marked by a changed labour market context of labour shortages. The discussion about mini-jobs is also influenced by major societal trends such as increased female employment, improved child-care facilities, more single parents, divorces etc. These trends put a question mark on the initial intention of mini-jobs.

• 2. 

  Civil law contracts in Poland

The civil law contracts are a special category of atypical contracts used to contract labour in Poland. There are two main types of civil law contracts: the contract to perform specified work (umowa o dzieło) and the contract of mandate (umowa zlecenie). The latter is more popular than the former. The number of workers with civil law contracts in Poland has increased substantially since the early 2000s. The incidence of other forms of temporary work – fixed-term employment contracts, temporary agency work – has also risen, but civil law contracts imply an especially acute form of labour market segmentation because they’re much less regulated than other temporary contracts. This applies both to protection against dismissals and worker rights such as paid leave or the right to join trade unions, and to social security coverage.

Due to lack of data on civil law contracts, it is difficult to assess precisely the number of people working under civil law contracts. According to the LFS data, in 2016 there were 510,000 people working under civil law contracts (in their main job) which is equivalent to 3.1% of total employment in Poland. Among them, 429,000 people worked solely under a contract of mandate, 36,000 people worked solely under a contract to perform specified work, 22,000 people worked under another type of a civil law contract (such as a managerial contract), and 22,000 combined various types of civil law contracts. Data from personal income tax records published by the Ministry of Finance show that in the last few years approx. 1 million people has been earning income only on the basis of civil law contracts. The tax records also show that the number of such people has doubled since the early 2000s.

Individuals working under civil law contracts tend to be less educated than individuals working under employment contracts. The incidence of civil law contracts is the highest among young workers but it is also noticeable among workers aged over 50. Women constitute majority of workers under civil law contracts, but once the effect of education, age and occupation is taken into account, women are not more likely than men to work under civil law contracts. The highest incidence of civil law employment is found among low skilled occupations: elementary occupations, and services and sales workers. The sectors which record the highest incidence of civil law contracts are accommodation and food services, and business support services which include temporary work agencies. The prevalence of these contracts is the highest among small firms (with less than nine workers). Besides employment in call centres, the expansion of civil law employment cannot be associated with the expansion of digital economy. Workers under contract of mandate earn less than workers under employment contract. It is due to lower hourly wage and larger incidence of part time employment. The median hourly wage of individuals working under contract of mandate is by 25% lower than the median hourly wage of individuals on employment contract.

Individuals working under contract of mandate can be either exempted from social security and public health insurance contributions in some circumstances (if they are students aged up to 26 years), or can pay lower contributions than they would pay if they earned the same net income under an employment contract. Individuals working under contract to perform specified work are exempted from all social security and public health insurance contributions – this group might not be covered at all. However, more than half of them have probably access thanks to other means, mainly because of an employed spouse. After a raft of changes which have been introduced since 2015 to improve access of workers with civil law contracts to unemployment benefits and maternity benefits. The social security contributions which pertain to civil law contracts have also been increased, although they are not yet identical to contributions pertaining to employment contracts. The key challenge related to social security coverage is in expected low retirement pensions. In the Polish defined-contribution system, low contributions of workers with civil contracts will translate into low pensions even if these workers spend only a share of their careers working under civil law contracts. Estimates show that the pension gap will be about 17%. The recent changes in contribution rules have closed it only to some extent. Many of these workers will be receive minimum pensions which will have to be subsidised from general taxation. Lower contributions paid by civil law contract workers aggravate the deficit in the pension system. On the other hand, individuals working under civil law contracts contribute more in the sickness and accident insurance contributions than they receive in sickness, carer and rehabilitation benefits, and maternity allowances. In aggregate terms, they cross-subsidise other groups, in particular the self-employed.

• 3. 

  Self-employed without employees in the Netherlands

Coverage and barriers for SEWE

As stated, all Dutch citizens residing in the Netherlands between the ages of 15 and the official retirement age are entitled to the state pension. However, most self-employed workers are not covered by the second pillar, and participation in the third pillar is voluntary. Hence, the pension coverage for self-employed workers depends on the person.

A recent study conducted every two years shows the pension arrangements made by SEWE, as shown in the figure below (Lautenbach et al., 2017). The two most frequently used forms of pension arrangements are firstly savings and investments, and secondly home ownership. More formal arrangements such as pension funds and fiscal old-age reserves or annuities are slightly less frequently used, but are still quite common. Around a quarter of SEWE have made no pension arrangements at all. Their main reason for this (in their own words) is that they cannot afford to make pension arrangements (as shown in Figure A.22).

Figure A.21: Pension arrangements made by SEWE (multiple answers possible)

Source: Lautenbach et al., 2017

Figure A.22: Reasons for SEWE not making any pension arrangements

Source: Lautenbach et al., 2017

Pension funds are free to choose whether or not they allow salaried workers who are transitioning into SEWE to continue their pension scheme, and they can formulate their own conditions and requirements. A recent study (Bureau Bartels, 2016) commissioned by the Dutch Ministry of Social Affairs and Employment shows that 85 percent of pension funds provide the possibility of voluntary continuation of the pension scheme for the self-employed. However, very little use is made of this by self-employed workers (only approximately 650 persons nationally). This limited usage does not appear to depend on the specific requirements or conditions set by the individual pension funds. More important obstacles are firstly, the legal requirement that the pension has to be continued immediately after the end of the prior period in employment, and secondly, the required level of the pension contributions. 290 Also, this instrument can only provide a solution for workers transitioning from an employment contract to self-employment. Therefore, the State Secretary of Social Affairs concludes that this instrument is not effective in promoting pension accumulation among the selfemployed

291 .

The problem of SEWE not making pension arrangements has been identified by the Dutch government in 2016, which has been reviewing possible ways to increase the pension

coverage for SEWE, for instance by introducing incentives 292 . However, the recently formed government has not mentioned this in their coalition agreement 293 .

• 4. 

  Casual workers in Romania

Casual work in the form of contractual day-labour has been legally endorsed in Romania by the Law 52/2011, which followed shortly the revised labour code (Law 40/2011) that significantly deregulated the labour market and limited trade unionism. In its initial version, the law represented a sharp departure from the principles of social insurance and historically established labour rights such as the right to a minimum national gross wage or collective labour contracts. While the subsequent revisions of the law, especially in 2014 and 2015, introduced some more responsibilities for the beneficiaries of day labour, the exemption from the payment of social insurance contributions (pensions, health care, unemployment, and paid sick and maternity leaves) had been maintained, and insurance remained voluntary and fully covered by the day-labourers.

290 The pension premium for self-employed workers who opt for voluntary continuation of their pension scheme is equal to the sum of the employee’s contribution and the employer’s contribution.

291 Kamerbrief Klijnsma 11 juli 2016 ‘Vrijwillige voortzetting pensioenfonds voor zzp’ers’.

292 Kamerbrief Klijnsma 11 juli 2016 ‘Vrijwillige voortzetting pensioenfonds voor zzp’ers’.

293 Regeerakkoord ‘Vertrouwen in de toekomst 2017 – 2021’

The rationale to grant the legal possibility of using day-labour consisted of tackling undeclared work, especially in the large agricultural sector that extensively employed seasonal workers. The new law granted the possibility of employment on a daily basis, with the minimum obligation to pay the flat-rate income tax (16%) and to provide for the hospitalization/funeral costs in case of workplace accidents. Minors aged 15 can work as daylabourers only with the written approval of their parents. A maximum of 90 days/year can be worked as day-labour for the same beneficiary, with the exception of agriculture, where the threshold was extended to 100 days/year. The domains where day-labour can be performed were set by the law and the nature of work restricted to unqualified labour. The leasing of day-labourers for a third party was prohibited. Economic agents that utilize day-labour should submit monthly a register to the Labour Inspection, signed by each labourer.

As compared to 2011, when ca. 170.5 thousand day-labourers were registered, by 2016 their number increased 4.7 times, reaching 803.6 thousand persons. Their share in the total labour force expanded from 1.2% in 2011 to 6.7% in 2016. The number of legal entities using daylabour also increased from 4.8 thousands in 2011 to 24.1 thousands in 2015. According to the statistics of the Labour Inspection, 80% of day-labour is performed in agriculture (including vineries, orchards and animal husbandry), forestry and fisheries. This sector typically employs persons with lower levels of education, who at times combine seasonal work abroad with casual work in Romania. A smaller share of day-labour was registered in the case of organizing events or leisure activities, more frequent in urban areas and typically employing students.

Given that “day-labourer” does not constitute a statistical category for occupied persons in the reports of the National Institute for Statistics, there is a lack of publicly available information and reliable micro-data that could allow the demographic profiling of day-labourers, as well as estimations regarding their uptake of various social protection benefits and services. However, given the unqualified nature of work, prevalence in the agricultural sector, and low incomes, most probably they can hardly afford paying voluntarily all social insurance contributions.

For 2017, the estimated loss of the public social insurance systems due to the replacing of regular employment contracts at the minimum wage with day-labour is around 1,824 million Ron.

Since 2015 it has become legally possible to cumulate income from day-labour with social assistance benefits granted under the law on the Guaranteed Minimum Income (Law 416/2001, GMI), and, as a collateral provision under GMI, to benefit from public health care insurance covered from the national budget. Consequently, in the particular case of daylabourers who lack imputable assets that would disqualify them from GMI, and who are willing to perform the required community work, there is a cross-subsidizing of labour. The main beneficiaries of this cross-subsidizing are the economic agents from the agricultural sector, as they employ the largest share of day-labour.

• 5. 

  Economically dependent self-employed in Spain

In Spain, the figure of economically dependent self-employment (trabajador autónomo económicamente dependiente, TRADE) was regulated in 2007 in the Self-employed Workers’ Statute (Ley del Estatuto del Trabajador Autónomo, LETA). The most important criterion used to establish whether a worker belongs thereto is of an economic nature, measured as the percentage of income earned from a single client. A self-employed worker is considered to be economically dependent if the income from a single client accounts for at least 75% of his/her total income (Article 11, paragraph 1 of the Self-employed Workers’ Statute). Self-Employed Workers’ Statute conferred TRADE status higher social protection compared to genuine selfemployment.

In 2017 (2nd quarter), there were 10,530 TRADEs registered. The number of registered TRADE increased significantly in the first years after the LETA was passed. However, after the initial increase, since 2015, the number of TRADE registered has barely increased and it accounts for a very low proportion of total self-employed without employees (0.7%). Overall, social partners, self-employed organisations and experts find that the existing legal regulation of TRADE has failed to attract self-employed because it does not respond to the needs of this specific collective nor to the needs of employers.

Difficulties for TRADE of acquiring their own social protection rights

Social protection rights of TRADE are almost identical to those of normal self-employed. Regarding in-kind benefits associated to healthcare, self-employed, included TRADEs, are entitled to the same rights as employees. As far as cash benefits are concerned, TRADEs are obliged to contribute to insurance for most of the contingencies that apply to employees (pensions, labour accidents, etc.). The most important in-cash benefits from which TRADE are excluded are unemployment benefits and Wage Insolvency Fund (Fondo de Garantía Salarial, FOGASA).

The amount of all cash benefits that TRADEs are entitled to are determined by the contribution base and the monthly contribution. As opposed to employees, self-employed and TRADE can choose the contributions they want to pay to the Social Security. Most of them (81% of self-employed and 91% of TRADE in 2017) are selecting the minimum base in order to pay lower contributions. As social protection is closely linked to the contribution bases, this leads to lower levels of protection and, especially, lower pensions (see below), with an increased risk of poverty and social exclusion.

Possibilities of gaining access through derived social protection rights

In Spain, the social protection system does not contain many elements which connect the social protection of an individual to his or her specific family situation. Main social security derived rights fall under the juridical category of ‘in-cash benefits for dead and survival’ (prestaciones por muerte y supervivencia). Self-employed and TRADE are entitled to the same rights as employees, as far as they contribute to the insurance against temporary incapacity which is mandatory for all self-employed workers.

Coverage and access by other means

Some private funds such as health funds have been incentivised for self-employed through discounts in their direct income taxes. Regarding pensions and dead insurances, some studies (UNESPA, 2016) find that the share of self-employed having private insurances is very close to the share of employees having one.

Cost-benefit of cross-subsiding TRADE

Measures favouring transition from unemployment to TRADE are cost-effective as they could contribute to save social expenditure related to unemployment benefits. However, a political strategy favouring these transitions should also consider that there is the facto a considerable gap in terms social rights (no unemployment benefits and lower retirement pensions) and a considerable gap in terms of employment rights.

• 6. 

  Project workers on continuous collaboration projects in Italy

Project-based continuous collaboration contract (Contratto di collaborazione a progetto, or Co.co.pro.) is a non-standard form of employment contract established in Italy in 2003 that gradually replaced the earlier coordinated and continuous collaboration contract (Contratto di collaborazione coordinata e continuativa, or Co.co.co). In 2015 the former was abolished to be replaced through the reintroduction of its predecessor.

Both kinds of contractual arrangements can be included in the broader class of “parasubordinate” forms of employment, whereby workers find their condition to be somewhat in between that of an employee and that of self-employed collaborator. These arrangements came to fulfil specific employer requirements that could not be met by previous forms of subordinate work. However, the widespread use of these contracts has often simply masked standard employment relationships while allowing the employer to benefit from lower gross costs, both in terms of the explicit costs of contribution rates and the implicit costs of dismissals. Such economic incentives have mostly been eliminated by various reforms of the Italian labour market. Such reforms have gradually increased the contribution rates by aligning these to the levels of standard employment contracts and by reducing the limits to the use of fixed term employment contracts. As a result, in recent years the number of Co.co.pro. workers, i.e. project-based continuous collaboration contract workers, has remarkably decreased, from 692,078 in 2011 to 379,299 in 2015. At the same time, from the point of view of access to the welfare system, such reforms have progressively been aligning the conditions of these workers to those employed with standard contracts, thus providing them with equivalent pension schemes and a similar level of access to healthcare, sickness, family benefits and occupational injury benefits. From this perspective, the key remaining statutory difference concerns unemployment benefits, which are less generous and have a shorter duration.

However, de facto gaps in access to the welfare system still persist, due to the different labour market outcomes of these workers, whose conditions are often characterized by lower wages, fewer yearly working hours and a higher risk of unemployment, and also to the higher weight of the weaker segments of the Italian labour supply, i.e. women and the young.

Thus, policies aimed at reducing such gaps would benefit from those instruments typically addressed at low-income workers. In this sense, it is worth to observe that Italy lacks a comprehensive policy on social exclusion and income support.

Moreover, since the Italian pension system follows an NDC approach, lower contribution rates involve a higher risk of poverty at the moment of retirement and require some adjustment of the guaranteed minimum pension scheme.

A NNEX 16: B EHAVIOURAL STUDY

Behavioural economics and more generally behavioural sciences study heuristics (mental shortcuts) and biases and provide a more realistic picture of our attitudes, decision making processes, choices and behaviour than the one proposed in the utility maximising view of human nature proposed by neo-classical economics and rational choice theory. There are a number of heuristics and biases, empirically documented in other field indirectly relevant to the object of this study. The conventional decision theory, assuming that agents are perfectly rational with not cognitive issues and stable preferences, fails to explain many critical decisions related to social protection, such under-optimal saving for pension or misperception of the likelihood of events such as losing one’s job or the need to use health services.

The analysis below summarises the results of the survey carried on by Open Evidence on 8000 individuals in 10 Member States (Germany, France, Italy, the Netherlands, Poland, Portugal, Romania, Slovakia, Spain and Sweden). It is important to note that the survey is not representative of the population, as explained in Annex 9.

Equal and Adequate Protection

 Social protection considered inadequate. Overall in the sample social protection in

terms of the six types of benefits considered in this survey (unemployment, old-age, maternity/paternity, sickness, invalidity, accidents/occupational diseases) is considered inadequate by 60% of respondents. Such appraisal differs markedly by

employment status, for instance: taking the self-employed and those employed fulltime on an open ended contract the appraisals are specular: 73% of the former consider is not adequate at all (25%) or not very adequate (48%); whereas for the latter this percentage goes down to 42% with as much as 58% considering it very adequate (7%) or fairly adequate (51%);

 For none of the six types of benefits coverage reach 50% of the sample. Moreover,

coverage differs widely by forms of employment, for instance: 72% in open ended full time employment have access to unemployment benefits but only 22% among the selfemployed and 43% among the part-time employed with temporary contract;

 Unemployed with very little coverage. The unemployed show low level of coverage

for the five type of benefits other than unemployment one. For instance, coverage for old-age ranges from only 10% to 23% depending on the forms of employment in

which they worked before becoming unemployed;

Transferability

 Transfer rights perceived as complex. As many as 68% of respondents who are

employee think that transferring social rights to become self-employed would difficult and another 13% lack information about this aspect;

System transparency

 Unawareness about coverage. Between 13% and 17% of the all sample do not know

whether or not they have access to the six type of benefits considered in this survey. Lack of awareness is higher for individuals in less consolidated forms of employment, which suggest the systems result less transparent to those individuals who would need

information the most;

 Difficulty to apply for benefits. Between 50% and 65% of respondents, depending on

the kind of benefits, consider applying for them to be difficult or very difficult;

 The relation between transparency and the other two dimension. Various other

indirect measures seem to confirm that social protection systems are not very transparent. This may also explain why, despite considering social protection not adequate, respondents do not perceive personal accounts and transferability of rights as important as they might. Lack of information and transparency function as a

mediator creating doubts and uncertainty about changing employment status.

Willingness to pay for, and ranking of, benefits

 Willingness to pay for voluntary schemes. Willingness to pay for voluntary schemes

is not very high (unless the contribution does not exceeds 5% of annual gross income) for all six types of benefits;

 Ranking the importance of different benefits. Unemployment and old-age benefits

stand out as the most important, whereas maternity/paternity benefits are by far the least important, despite the fact that in our sample there a sizeable proportion of respondents with children. This result may be interpreted as one of the sign of the current crisis that pushes individuals to be concerned about the most basic needs and to forget less pressing but symbolically important welfare entitlements; one could see in such findings further confirmation that the Great Recession is bringing back Europe from the post-materialist orientation consolidated in the 1990s to a more materialist

outlook;

16.1 Is social protection adequate?

When asked respondents whether they consider that in general the social protection they have access to provides adequate support for various situations (i.e. becoming unemployed, covering old age pension, maternity/paternity leave, sickness leave, disability leave – Figure A.23), 60% of them find it not very adequate (44%) or not at all adequate (16%).

Figure A.23. Social protection adequate support

n=8000

Differences in perception can be spotted across employment statuses (Figure A.24). Taking, for instance, the self-employed and those employed full-time on an open ended contract the appraisals are specular: 73% of the former consider is not adequate at all (25%) or not very adequate (48%); whereas for the latter this percentage goes down to 42% with as much as 58% considering it very adequate (7%) or fairly adequate (51%).

Figure A.24. Social protection adequate support by employment status

n=8000 (p<0.001)

Another interesting fact is the perception individuals have about their entitlements (Figure A.25). First, for all the social security branches considered, less than 50% of the respondents report being covered. Second, coverage seems relatively higher for sickness and unemployment benefits compared to the other benefits, and particularly low appears coverage for maternity/paternity benefits where as many as 61% report not being covered; also high are the percentages of those saying they are not covered for old-age (52%) and invalidity (49%).

Figure A.25. In case of need, would you be entitled through your main job to…

n=8000 (Unemployment benefits n=6000)

When breaking down by employment status, differences in perception of coverage are particularly evident in case of unemployment benefits (with 22% of self-employed declaring to be covered, against the 72% of the standard employees), sickness benefits (36% of coverage in case of the self-employed and 71% in case of standard employees), accidents at work (30% for the self-employed against 71% covered among the standard employees) and invalidity benefits (31% against 51%).

Figure A.26: Self-declared coverage by benefits and employment status

Unemployment benefits

Self-employed 22% 60% 18%

Entrepreneur 35% 51% 14%

Part-time, temporary 43% 42% 16%

Full-time, temporary 55% 31% 14%

Part-time, open-ended 53% 33% 14%

Full-time, open-ended 72% 21% 7%

Yes No Don't know

When coming to transferability issues, both self-employed and employees consider as difficult to transfer the acquired rights in case of changing employment status. Indeed, 68% of respondents who are employees consider that transferring their social rights if they become self-employed would not be easy (28% not easy at all and 40% not very easy – figure A.27).

Figure A.27. Ease of transfer social rights from dependent employment to selfemployment

n=4048

Figure A.286 Ease of transfer social rights from self-employment to dependent employment

n=1454 Transparency is also an issue. Lack of awareness due to unclear or difficult to access information is not excessively high but still sizeable and seems to be higher for old-age, invalidity, and accidents/occupational benefits (figure A.29).

Figure A.29. In case of need, would you be entitled through your main job to…

Difficulties are also expressed concerning the procedures to apply for a certain benefits. Only applying for maternity/paternity 294 , unemployed benefits, and sickness benefits is considered very easy or fairly easy by about 50% of respondents (figure A.30).

Figure A.30. How easy do you think it would be to make an application for the following issues?

294 For maternity/paternity benefits, women considered the application relatively easier as compared to men.

n=8000

The most difficult application is deemed that for invalidity benefits (a total of 74% consider it from not very easy to difficult), followed by accidental/occupational disease benefits (a total of 68% consider it from not very easy to difficult), and old-age benefit (a total of 63%

consider it from not very easy to difficult) 295 .

16.2 Focus on the employers for the SMEs test

Employers results to be distributed across all ages (4% between 18-24 years old, 20% between 25 and 34, 29% among 35-44, 26% and 22% between 55 and 66 years old) and mostly male (62%). Concerning their education levels, 7% results to have until 11 years of instructions, 28% have a high school diploma, 10% attended university for some years and 55% results to have a university degree or a post-graduate degree. Moreover, 17% declared to be in the first quartile of the income distribution, 13% in second, 17% in the third and 41% in the fourth (11% didn't reply).

295 This question can be taken as a more accurate measure on the level of system transparency with respect to the previous one since it forces respondents to think about doing something in practice rather than simply stating whether or not they are covered.

Figure A.31: employment status by age, gender, education and individual income distribution

When asked how an extension of mandatory coverage to the non-standard employees and the self-employed would affect their hiring policies and their administrative burdens, the employers answered as shown in figures A.32 and A.33.

Figure A.32: Hiring preference in case of extension of mandatory social protection coverage to all forms of employment

Figure A.33: impacts on employers' administrative burdens in case of an extension of mandatory social protection coverage to all forms of employment

16.3 Focus on platforms workers

Forms of work and self-employment related to online economy are also of interest for this initiative. However, there is still lack of information about the characteristics of those working on platforms. For this reason the survey also focused on platform economy workers. In total, 28% of the total sample is engaged in such activities, 57% doing work for business or individuals and 48% selling goods and 12% renting space (Figure A.34). The main motivation for engaging in online platforms activities is the possibility of working from home (55%), followed by flexibility (34%), whereas only 11% apparently engage in such activities because

they were unable to find a regular job 296 .

296 When looking at the distribution of motivations among countries, it is possible to detect certain heterogeneity. Please refer to Figure 29 of the OpenEvidence Report (2018, forthcoming).

Figure A.34: Share of people participating in the online economy by using platforms to help generate income, main activities and reason to work for platform economy?

Figure A.35: Participation in the online economy by employment status

n=7502 (p<0.001)

When coming to the kind of employment of those working on platforms economy, figure A.35 shows that the self-employed are more represented among those doing work through online labour platforms, whereas very few unemployed are engaged in such activities and

part-timers are engaged in them less than one could expect 297 . On the other side, the

proportion of employees with an open ended contract also using these platforms to generate income is 29%.

297 This evidence seems not to support the idea that the sharing economy can help bringing the unemployed and under-employed back to work.

Figure A.36 Proportion of income derived through online platforms activities

n=1893

Finally, when coming to the part of the total income generated through online platforms, 15% generate revenues between 50% and 100% of their income (summing up 75% to 99%, 50% to 75). Another 15% generate revenues between 25% and 50% of their income, 39% between 10% and 50%, and 32% only up to 10% of their income (figure A.36).

Met de Parlementaire Monitor volgt u alle parlementaire dossiers die voor u van belang zijn en bent u op de hoogte van alles wat er speelt in die dossiers. Helaas kunnen wij geen nieuwe gebruikers aansluiten, deze dienst zal over enige tijd de werkzaamheden staken.