Chapter 1. Ensuring social protection for non-standard workers

This chapter sets out by presenting the main challenges of covering non-standard workers in contributory social protection systems. It then discusses the advantages and pitfalls of two basic ways in which social protection systems could adapt to these challenges: by tying entitlements to individual workers rather than employment relationships or by doing the opposite and untying benefits from contributions. The chapter offers theoretical considerations and practical country experiences in offering voluntary social protection to non-standard workers, and looks at how social security contributions themselves can be a driver of non-standard work. It then presents two examples of special schemes for non-standard workers, and discusses the emerging challenge of improving the social protection and job quality of platform workers. Finally, it draws policy lessons on improving the social protection of non-standard workers and enhancing the income security of the increasing number of on-demand and flexible hours workers.


The statistical data for Israel are supplied by and under the responsibility of the relevant Israeli authorities. The use of such data by the OECD is without prejudice to the status of the Golan Heights, East Jerusalem and Israeli settlements in the West Bank under the terms of international law.

1.1. Introduction

Social protection systems are often still designed with the archetype of full-time, permanent work for a single employer in mind. Deviations from this pattern – be it self-employment or the combination of different income sources – can lead to coverage gaps or loss of accumulated entitlements.

This is not a marginal issue. Across the OECD on average, 16% of all workers are self-employed (Figure 1.1), and a further 13% of all dependent employees are on temporary employment contracts. Temporary workers often struggle to accumulate minimum contribution periods, and the self-employed are often only covered by the most basic of benefits – only 6 out of 28 European Union member states insure the self-employed in the same way as standard employees (Spasova et al., 2017[1]).

Box 1.1. Key terms

Standard workers

Regular, open-ended dependent employment with a single employer. The definition often also specifies full-time hours; however, since social protection systems typically treat part-time workers like standard workers as long as they meet the minimum income requirements, part-time workers are included in this definition throughout this chapter.

Temporary workers

Workers on a fixed-term contract. Since social protection systems typically do not differentiate between standard- and temporary workers as long as they achieve the minimum contribution periods, this chapter concentrates on the self-employed, and those on the border between self- and dependent employment, as well as those on flexible or on-call contracts.

Flexible or on-call contracts

These are employment contracts that do not require the employer to offer a minimum of working hours, while the employee is not required to accept hours offered; however, times of availability for the worker may be agreed.

Platform workers

Are workers who offer and find their work through online labour platforms.

Para-subordinate workers

In Italy, para-subordinate workers are self-employed, but highly dependent on one or very few clients. They are mandatorily enrolled in a special public fund called Gestione Separata.

Independent contractors

Independent contracts (freie Dienstverträge) is a form of labour contract between self- and dependent employment. Independent contractors have no or little “personal dependence” on their employer and control their own working time and workflow, but they are contracted for their time and effort (that is, they do not carry entrepreneurial risk).

Figure 1.1. Non-standard work is widespread in some countries

Note: There is no data on temporary employment for Israel, New Zealand or the United States, no data on self-employment for Iceland and Luxembourg; no data on France before 2003, on Lithuania and Mexico before 2005 and on Switzerland before 2010. Data on self-employment for Austria refer to 1999, for Spain to 1998, for Turkey and the United Kingdom to 2000, for the Slovak Republic to 2000 and 2015, for Latvia to 1998 and 2014

Source: OECD (2018), “Labour Force Statistics: Summary tables” and OECD Employment and Labour Market Statistics (database).

Rising numbers of non-standard workers also threaten to erode the contribution base of social protection systems. If tax and social protection systems are designed such that only some categories of workers are covered by social protection – and liable to pay social contributions – while others are not, firms have an incentive to shift work onto those workers who enjoy the least protection. But also low-risk workers (e.g. younger and better-educated workers) may self-select into self-employment and other non-typical or new employment forms. This undermines risk pooling that is crucial to any form of insurance.

Global trends such as globalisation and digitalisation are likely to make the divergence between the ideal of dependent, full-time employment and labour market realities more salient. New technologies make it easier and cheaper to offer and find work online, and online intermediaries (work platforms) have experienced spectacular growth in recent years, although they still account for a very small share of workers in OECD countries (e.g., Katz and Krueger (2016[2]), Pesole et al. (2018[3])).

These new technologies lower transaction costs, allowing firms to outsource more and more, pushing the boundaries of the firm (OECD, 2016[4]). This offers valuable flexibility to many who have previously been excluded from the labour market, such as those with caring responsibilities, or those in remote areas.

But it also exposes workers to new risks: since these gig or crowd workers are hired for specific tasks only, they can be instantly dismissed if demand drops. They cannot rely on the insurance function of the standard employment contract, as the first line of defence against demand fluctuations (Parsons, 1986[5]). At the same time, they have less access to income replacement payments than other workers. While firms may thus have a dual cost advantage of avoiding both social contributions as well as having to pay employees during down-times, workers incur a double risk.

Shifting low-paid work onto independent contractors, regardless of whether it is through online platforms or traditional markets, also threatens the effectiveness of minimum wage floors. This applies to flexible hours contracts as well: while minimum wages guarantee minimum hourly payments, the lack of fixed working hours introduces income insecurity on the working time margin. Firms may use flexible forms of employment to circumvent minimum wage increases: Datta et al. (forthcoming[6]) show that while the hourly wage rate of UK domiciliary care workers paid at the minimum wage went up by 7.5% following the introduction of the National Living Wage in 2016, their probability of being on a zero hours contract increased by 4.7%. Thus, workers may suffer from decreasing incomes despite minimum wage increases.

New technologies and the new forms of work they create bring the incomplete social protection of non-standard workers to the forefront of the international policy debate (see, for example, the European Commission’s initiative on Access to social protection (European Commission, 2018[7]; European Commission, 2018[8])). It is hardly a new phenomenon, however: across the OECD on average, self-employment as a share of total employment decreased by 4 percentage points over the past 20 years, although trends have been very uneven across countries (see Figure 1.1). Recent US-census data show that the share of workers who are independent contractors decreased slightly between 2005 and 2017 (from 7.4% to 6.9% of total employment), while the share of on-call and agency workers stagnated over this period (BLS, 2005[9]; BLS, 2018[10]).

But labour force surveys likely underestimate the actual extent of self- and contingent employment. Abraham et al. (2017[11]) show that nearly two-thirds of all workers who have self-employment income in US tax data do not report having self-employment income in labour force surveys – half of them have incomes as both self-employed and dependent employees, and are thus likely omit their second job, while the other half report being employed, indicating misclassification – and this gap is widening. As a consequence, levels and trends in self-employment appear much lower in survey data estimates than indicated in the tax data.

As non-standard work is not generally new to OECD economies, there are lessons to be learned from strategies that countries already employ to provide social protection to such workers. In fact, as Prassl (2018[12]) argues, even gig work, while enabled by new technologies, shares many traits with historical on-demand work such as 19th century dock labour. Policy solutions that were and are applied to non-standard workers therefore deserve attention as possible remedies to current challenges.

The diversity of social protection systems across the OECD means that some policy solutions that are currently discussed have already been implemented in practice. For example, untying social protection from the employment relationship, and instead offering benefits on a means-tested basis, is often brought forward as a solution to the problem of closing coverage gaps. Australia’s largely general revenue-financed social protection system (supplemented by ‘Pay-As-You-Go’ financed and income tax subsidised compulsory superannuation) provides an example of such a system. Similarly, some countries extend coverage to non-standard workers by establishing voluntary schemes; the Swedish unemployment insurance system offers lessons on how such schemes work in practice.

This volume collects seven case studies that shed light on different aspects of the social protection of non-standard workers. The case studies analyse

  • the implications of a general revenue-financed social protection system on non-standard workers (Chapter 2 on Australia);

  • voluntary social protection schemes for self-employed workers (Chapter 8 on Sweden);

  • the effects of differing social protection coverage of standard and non-standard workers on the incidence of non-standard employment in the Netherlands (Chapter 7), Italy (Chapter 6) and Austria (Chapter 3), including reforms to increase the coverage of non-standard workers on para-subordinate workers in Italy the integration of independent contractors into the Austrian social protection system;

  • special schemes for non-standard workers, specifically the French Régime Social des Indépendants (RSI) in France (Chapter 4), and programmes that provide social protection to special subgroups in the creative industries (the German artists’ insurance for artists and writers (Chapter 5) and the intermittent du spectacle scheme for performing artists and some stage technicians in France (Chapter 4).

This chapter brings together some key policy insights from these case studies, as well as recent policy initiatives across the OECD. It also looks at the special challenge of providing social protection to platform or gig-workers and offers policy options to increase the income security of on-demand and flexible hours workers.

The chapter sets out the main theoretical and practical considerations that complicate social protection for non-standard workers (Section 1.2), and discusses the advantages and pitfalls of two basic ways to reform social protection systems to facilitate the inclusion of non-standard workers (Section 1.3). It goes on to synthesise policy lessons from voluntary social protection schemes (Section 1.4), the incorporation of non-standard workers into the standard social protection system (Section 1.5) and discusses two country experiences of special schemes for artists and workers in the entertainment industry (Section 1.6). Section 1.7 discusses the special situation of workers in the platform economy. Section 1.8 offers policy lessons on improving social protection for non-standard workers.

1.2. Independent and contingent workers in contributory social protection systems

Independent and contingent workers do not easily fit into the framework of contributory social protection systems: who should be liable for their employer contributions, and how should contributions and benefit entitlements be calculated with highly fluctuating earnings? Providing unemployment insurance for the self-employed, in particular, also raises significant moral hazard problems.

First, in contributory systems, employers and employees share the contribution burden – who should be liable for the employer contributions if an employer is not easily identified? This is the so-called double contribution issue.

Requiring the self-employed to pay both employer and employee contributions is the straightforward solution to this problem. This would effectively force them to raise their prices, bringing their labour cost more in line with that of dependent employees. In the United States, self-employed workers pay both employer and employee contributions to social insurance; the difference in the non-wage labour cost between dependent employees and self-employed workers at the average wage is only 1.6 percentage points, compared to 30 percentage points in the Netherlands, and mainly due to employer contributions to unemployment and work-injury insurance that do not cover the self-employed (OECD, forthcoming[13]; SSA and ISSA, 2018[14]).

Not all self-employed workers have the bargaining power to shift these costs onto consumers however. Self-employed earnings are typically dispersed, with a high share of low earners (see the articles in this publication for non-standard workers in selected OECD economies, and Berg (2016[15]) for crowd workers). For example, one in-four self-employed workers in France earns less than EUR 12 000 per year (Chapter 4). Minimum wages typically do not apply to them, and their scope for collective bargaining is often limited by competition law (BMAS, 2017[16]). Raising labour costs for the self-employed also comes at the risk of pushing economic activities into the informal economy, especially for peer-to-peer transactions that are difficult to monitor for tax authorities (OECD, 2018[17]).

Charging clients directly is a possible approach to address this challenge, but this is not straightforward from an administrative perspective. The German artists’ insurance scheme is an example of this approach (Section 1.6.1): firms and public entities who contract artists, writers or journalists pay a fixed contribution to a special fund administering social protection for artists and writers. The German pension fund monitors these contributions in the course of their regular social security compliance inspections; a significant advantage considering that monitoring of individual self-employed earnings can be difficult for tax authorities, compared to taxes on employee earnings which are often withheld at the source (OECD, 2018[17]).

The government could also heavily subsidise schemes for the self-employed, which raises concerns about equal treatment and may create adverse incentives for both employers and employees. Where they exist, such schemes are therefore often limited to occupations that are thought to create special value for the public, such as the arts. The German government funds half of employer contributions in the artists’ insurance scheme (Chapter 5); in the French intermittents du spectacle scheme for performing artists and related occupations contributions cover less than 20% of total benefit expenditure (Chapter 4).

Second, the self-employed often have fluctuating earnings – because they are paid at irregular intervals, because there are time-lags between work and payment, or because demand for their services is erratic (ISSA, 2012[18]). Thus, contributions are difficult to calculate – even if they are annualised, contributors might struggle to pay in bad years; contributors also have some control over the timing of payments, and could time them to circumvent means-tests.

Third, the self-employed do not meet several conditions that typically limit moral hazard in unemployment insurance: fluctuations in demand are hard to distinguish from voluntary idleness. Income fluctuations complicate the calculation of benefit entitlements, even if income is annualised. There is no employer to confirm a layoff, and job search efforts are even more difficult to monitor than for dependent employees. Also, because downward wage rigidity does not apply to the self-employed, they are more likely to have lower present than past earnings, and therefore a stronger incentive to become and remain unemployed (Chapter 8).

As a consequence, unemployment is generally the least-covered risk for self-employed workers: only 8 of all 28 European Union member states fully cover self-employed workers for unemployment insurance, and nine do not offer any form (even partial or voluntary) insurance. In contrast, self-employed women are covered for maternity benefits in all but 6 European Union member states (Avlijas, 2018[19]).

Where the self-employed are covered by unemployment insurance, they often face more stringent eligibility conditions:

  • In Sweden, self-employed workers have to close down their business before claiming benefits. Because setting up a business has a significant administrative cost, this is an expensive check, which seems to work: over the 2004-2016 period, the average unemployment rate among insured self-employed workers was 4%, compared to 7% among dependent employees (Chapter 8).

  • In Austria, self-employed workers have six months to decide whether to opt-into voluntary unemployment insurance upon starting their business – this decision is binding for eight years (Chapter 3).1 This check is designed to prevent those whose business is winding down to opt in just before collecting benefits. It has the drawback, however, that it asks start-ups to commit to a long-term fixed cost just as their finances are the tightest. In 2015, only 0.3% of all eligible self-employed persons opted into this insurance.

  • In Belgium, self-employed workers who have been declared bankrupt, are in a collective debt settlement, or who have been forced to interrupt their business activities, as well as self-employed workers in economic difficulties who cease all their self-employed activities may (under certain conditions) be entitled to a monthly benefit and health care without paying social contributions.

Box 1.2. Improving unemployment benefit access to the self-employed

The Danish unemployment insurance reform

In 2018, Denmark implemented a reform designed to make unemployment benefits more accessible to the self-employed and other non-standard workers. Before the reform, self-employed applicants had to produce documentation not only on earnings, but also revenue and tax declarations, proof of orders etc., while employees only needed to prove that they met the minimum earnings requirements. Benefit entitlement was therefore less predictable for the self-employed than for standard workers. Also, enrolees could only be insured as either dependent employees or self-employed, which made it harder for those combining dependent and self-employment to meet the minimum earnings requirements

The reform intends to harmonise benefit receipt rules: eligibly will solely be based on reaching a minimum (taxable) income over a three-year period and will not be conditioned on the type of employment. This should make eligibility more predictable for workers, as they can verify that they reached the required earnings-threshold on their tax return. As all income from work will be considered together, the reform should also improve eligibility for those who combine income from various sources. It also aims to simplify the administrative process of proving that a company has in fact closed down. To avoid that the self-employed continue working while receiving benefits, it also introduces a six-month “job search” period, during which benefit recipients have to look for dependent employment and are not allowed to start their own business. The implementation and effectiveness of this reform needs to be followed closely and could provide interesting policy lessons to other countries.

Source: Report from the working group of self-employed persons in the unemployment insurance system (Arbejdsgruppen om selvstændige i dagpengesystemet, (2017[20])); Kvist (2017[21])).

1.3. Potential avenues for reform

Social protection systems could adapt to these challenges in two basic ways: tie entitlements to individual workers rather than to specific employment relationships or do the opposite and untie benefits from contributions. This section briefly discusses each of these options.

1.3.1. Individualisation of social protection

This approach ties social protection entitlements to individuals, not employment relationships, by recording all social protection contributions made by workers themselves, employers or the state on their behalf in one account. As the timing and the provenance of contributions are irrelevant, such individual activity accounts would solve the problems of high earnings variability as well as of combining incomes from different sources. Recording all contributions in one place would also preserve entitlements during job changes and career breaks, supporting increasingly uneven employment patterns and labour market flexibility. This is why this idea, while not new, has been gaining popularity in recent years, especially among advocates for workers in new and emerging employment types such as “gig work” and micro-entrepreneurship (e.g. Etsy (2016[22])). In theory, individual activity accounts could accommodate contract work and short-time, contingent employment and collect the entitlements of multiple job holders in one place.

Several OECD countries are currently planning to introduce such “individual activity accounts”, which also allow beneficiaries to withdraw funds for causes not previously insured by social protection, such as education and vocational training, or starting a business. Depending on the specific model, individuals might also use them to take time out for caring responsibilities, or to retire early.

As individual accounts collect individual contributions for individual use, in their purest form, they do not incorporate risk-sharing, which is fundamental to any insurance. Thus, they would be unable to protect even high-earning individuals against catastrophic risks such as disability. Any other, implicit redistribution in the system – such as from those with very stable jobs to those who become unemployed frequently, or between those whose work carries health risks and those who can work healthily until retirement – has to be made explicit.

Individual activity accounts per se do not solve the double contribution problem (see above), and as such, meaningful benefits for the self-employed will remain elusive unless ways are found to levy social contributions from customers. Also, without substantial subsidies, they would be worthless to many low-income and part-time workers – though governments could of course decide to pay into accounts directly, e.g. by giving “starting endowments” to young people.

Making social protection entitlements more fungible for beneficiaries is moreover not without risk, as myopia can lead individuals to spend their entitlements too early, leaving them poor in old age. The experience of the Dutch Life Course Savings Scheme shows that many chose to use their funds to retire early instead of using them for further training or caring for family members (Delsen and Smits, 2014[23]).

1.3.2. Making social protection more universal

Untying social protection from the employment relationship – that is, granting individual entitlements to tax-financed benefits based on need rather than on earnings or contributions – would extend coverage to non-standard workers and get around the problem of tracking entitlements across jobs and over the lifecycle. Some benefits – such as health insurance and maternity or parental leave – are already universal in a number of OECD countries, and most countries have social assistance benefits of last resort, that provide basic assistance to those in need who are not entitled to any other benefits.

Depending on how exactly means-tests are implemented, non-standard workers may be more, not less likely to receive them, because they have less stable careers, are more likely to work part-time, and because their median earnings tend to be lower (Chapter 2). In the Netherlands, for example, the inflow rate into social assistance is nine times higher for non-standard workers than for dependent employees (Chapter 7). In Austria, on the other hand, only 0.2% of social assistance recipients are self-employed, while they make up about 12% of total employment. This is likely due to the fact that the means-test generally asks for business assets to be liquidated (Chapter 3).

However, in making entitlements more universal, policy makers should consider the risk of crowding out employer contributions. Australia and New Zealand are examples of general revenue-financed social protection systems that do not condition eligibility on previous contributions. In Australia, benefits are income- and assets-tested, although targeting has been relaxed incrementally over the past 50 years. The Australian story is complicated, however, by the social protection system’s interaction with a system of workplace benefits and entitlements, including paid holidays and sick and carer’s leave, that casual workers (about a quarter of all workers) as well as independent contractors (a further 9% of total employment) are not entitled to. Casual workers do not receive notice of termination or redundancy pay, which leaves them less time to search for other work. As a consequence, they end up unemployed more frequently than standard employees (although they are also more likely to work in high-unemployment sectors). This lack of workplace entitlements contributes to their more frequent receipt of general revenue -financed benefits and is the biggest source of cost-savings for firms in choosing this employment form (Chapter 2). In decoupling entitlements from jobs, policy makers have to think about how to ensure that employers continue to contribute, and that the government does not just take over parts of normal compensation from employers.

While means-testing benefits solves the problem of coverage gaps, it does not remove the need for tracking self-employment income, but rather makes it more salient: the problem of highly fluctuating earnings among the self-employed makes overpayments more likely, which is arguably an issue in a general revenue-financed system where benefits are not balanced out by contributions. This problem has no straightforward technical solution.

Abolishing means-testing altogether (that is, moving in the direction of a basic income) would remove all compliance issues and could easily incorporate non-standard workers. But it would be a budgetary challenge: replacing all existing working-age benefits by a flat-rate amount (without raising general revenue) in a budget-neutral way would lead to benefit levels below the poverty line in all OECD countries. It would also imply sometimes significant losses for disadvantaged groups such as the disabled, as benefits would no longer be targeted (OECD, 2017[24]). This raises substantial fairness concerns, as some individuals, such as the disabled, have greater needs than others (Piachaud, 2016[25]).While a basic income scheme could incorporate top-ups for specific groups, this would undermine the appeal of simpler (and cheaper) administration and complete predictability of basic income.

A basic income can also reduce work incentives, although individual labour supply is determined through a complex interplay of individual preferences, labour market characteristics and institutional and societal factors, leading to income and substitution effects, and thus depends on a range of factors, including the overall design of the tax system including basic-income payments, local labour demand, and the availability of child care. Thus, the effects of basic income schemes have to be assessed empirically.

Experiments with basic income schemes are currently under way in several OECD countries. Finland started paying randomly selected unemployment benefit recipients a basic income in January 2017. Payments are not conditional on work and participants are not required to seek employment. With a trial run of two years, the pilot will end in 2018, and first results of the experiment will be published in late 2019 to early 2020. The Finnish government has recently rejected the proposal to expand the experiment to a sample of employees (Peter, 2018[26]). Ontario, Canada, started a randomised experiment, the Ontario Basic Income Pilot in 2017. Exclusively targeted at low-income households, the pilot was supposed to run for three years. Payments were not unconditional, but were withdrawn against labour income at a rate of 50%. The Canadian government has however decided to end the programme prematurely as early as autumn 2018 (Kassam, 2018[27]).

1.4. Offering voluntary protection to non-standard workers

In extending social security coverage to non-standard workers, several countries opt for voluntary schemes: e.g. Austria introduced a voluntary unemployment insurance option for the self-employed in 2009 (Chapter 3), and Spain made its protección por cese de actividad de los trabajadores autónomos (unemployment benefit for the self-employed) voluntary in 2014 (Moral-Arce, Martí-Román and Martí-Román, 2018[28]).

Voluntary insurance schemes risk adverse selection of members: those self-employed workers with the highest risk have the biggest incentive to join. If the scheme is entirely self-funded, this can lead to a vicious circle of contribution hikes and low-risk members leaving.

When an experience rating was introduced in the Swedish unemployment insurance scheme in 2007/08, raising average premiums by 300%, membership in the voluntary Unemployment Insurance Funds dropped by around 10 percentage points. The two groups of workers who generate the lowest overall unemployment insurance expenses were most likely to leave the funds: workers over the age of 60 who have the lowest unemployment risk of all age-groups, and those under the age of 25 who, despite a high unemployment risk, have very low unemployment durations and low earnings (Chapter 8).

Partly general revenue-funded schemes can result in high public subsidies. Since 2010, self-employed workers in Canada have the option of voluntarily contributing to the Special Benefits for Self-employed Workers (SBSE) scheme to gain access to maternity and parental benefits, sickness benefits and care benefits for ill family members. Benefit entitlements and contribution rates are equal to those of dependent employees; that is, employer contributions are covered by a public subsidy. An evaluation of this programme (Employment and Social Development Canada, 2016[29]) found strong indications of adverse selection: Over three quarters of claims were for maternity and parental benefits, and two-thirds of opt-ins were women (who represent only 43% of all self-employed workers), while two-thirds were between the ages of 25 and 44 (compared to just one third of all self-employed). Opt-ins also had significantly lower incomes than other self-employed workers. In 2011, the first year benefits were paid out, premiums covered less than one-third of benefit payments.

In Austria, self-employed workers can opt into an income replacement programme in case of short-term illness, and about 8% of all eligible self-employed do. In 2016, nearly half of those who were covered received a benefit, and the average benefit duration was 22 days, nearly twice the average duration of sick-leave among (compulsory insured) dependent employees. In response to the deficits this scheme was running, the minimum benefit was cut significantly in 2017 (Chapter 3).

Thus, for voluntary schemes to be financially viable, they have to reach very high coverage rates, which can be tricky to achieve as workers seem to have a very low willingness to pay for social protection: the increase in Swedish UI fund premiums, while high in relative terms, only amounted to about 1% of median net-wages in 2008; at the same time, a tax credit was introduced that raised net earnings by about 5% for the median worker, so most workers could have paid the increased premiums and still experienced an increase in take-home pay. Still, one in eight members left the Unemployment Insurance Funds (Chapter 8). While the reform has since been largely reversed, membership rates have not yet recovered to their pre-reform levels.

A recent survey among European non-standard workers also found a low willingness to pay for social protection: only one in five non-standard workers who are presently not covered by social protection would be willing to pay 5% or more of their income for unemployment insurance; only a third would be willing to pay over 5% to be covered by old-age benefits (European Comission, 2018[30]). In designing effective voluntary contribution schemes, policy makers will thus have to accept substantial public subsidies if they want to achieve high coverage rates and avoid adverse selection (Chapter 8).

1.5. Social security contributions as a driver of non-standard work

Incomplete social protection coverage of non-standard workers can also increase non-standard work as employers seek to minimise non-wage labour costs. This is particularly the case for work arrangements that sit on the border between self- and dependent employment, and can therefore serve as substitutes for standard employees.

In the Netherlands, for example, the total “payment wedge” (including personal income tax, social security contributions and other compulsory payments) between hiring a dependent employee and an independent contractor at the average wage is 30%, 21pp of which are employer social security contributions (OECD, forthcoming[13]). This cost difference may be claimed by the employer or the employee, depending on their respective bargaining power on the labour market. In the Netherlands, research indicates that for low-wage employees, this payment wedge is almost entirely captured by employers, while at the upper end of the wage distribution, workers do capture part of it (Ministry of Finance (2015[31]) and Chapter 7).

In Italy, para-subordinate workers are self-employed, but highly dependent on one or very few clients. They used to pay lower pension contribution rates and were not covered for unemployment or sickness benefits. This resulted in significantly lower non-wage labour costs, and rising numbers of para-subordinate workers: in 2007, para-subordinates made up over 11% of all dependent employment. Recent labour market entrants were especially likely to work as para-subordinates: almost a third of new university graduates started their careers this way in 2011. As pensions directly depend on lifetime contributions, lower pension contribution rates imply lower pension entitlements.

In response to the growing number of para-subordinate workers, Italy gradually increased their social security contribution rates (and thus welfare guarantees) from the late 1990s, and accelerated this process from 2007 to an annual increase of one percentage point, until they reached the contribution rate of employees. As the total social security contribution rate increased by seven over a decade, the number of para-subordinate workers more than halved (Figure 1.2, Panel A). The onset of the decline in the number of para-subordinate collaborators predated other measures, such as the tightening of the regulation of para-subordinate arrangements (2012) and the abolition of some types of para-subordinate arrangements (2015), supporting the idea that the widespread use of para-subordinate collaborations was mainly driven by their lower cost. Further measures to persuade employers to hire workers on permanent contracts – such as a significant reduction in employer social security contributions for new hires through the 2015 Jobs Act – further contributed to the decline of para-subordinate contracts (Chapter 6).

Similarly, in Austria, independent contractors (freie Dienstnehmer) control their own working time and workflow but are contracted for their time and effort. Concerns that employers might use this form of employment to evade the compulsory social protection system drove their gradual integration into the system: since 2008, independent contractors are liable for the same (employer and employee) social security contributions as standard employees.2 While their number had been steadily growing until early 2007, it began to fall following the reform’s announcement, and was at its all-time low in 2016 (Figure 1.2, Panel B, Chapter 3).

The fall in the number of independent contractors was mainly driven by the falling number of workers taking up jobs as independent contractors, rather than by the dissolution of existing independent contracts – which is why their number continues to fall six years following the reform.

However, the likelihood of being in standard employment within one month of transitioning out of an independent contract increased from 11% before the reform’s announcement in 2006 to 13% after the reform’s implementation. In addition, the share of workers who transitioned out of the labour force upon exiting an independent contract dropped significantly, in line with the fact that prior to the reform, independent contractors were not covered by unemployment insurance. Unemployment insurance coverage proved important for former independent contractors during the Financial Crisis (Hofer, Hyee and Titelbach, 2018[32]).

Figure 1.2. Social protection coverage influences the use of non-standard work

Note: Exclusive para-subordinate collaborators collaborate on short- or long-term projects and are neither retired nor do they have income from dependent employment. They are the biggest group of para-subordinates, making up roughly 60% of all para-subordinates in 2008 and 40% in 2016.

Source: Panel A: Adapted from Chapter 6. Panel B: Adapted from Chatper 3, contribution rates from Hauptverband der Österreichischen Sozialversicherungsträger.

1.6. Special schemes for non-standard workers

Where self-employed workers are covered under a completely separate social protection regime, the reasons are often more historic than pragmatic. In France, for example, the self-employed initially opposed their integration into the general social insurance system. Instead, occupational schemes were established, the scope of which varies across professions. This led to very unequal coverage of dependent employees and the self-employed: while compulsory, occupational retirement schemes have existed since the 1960s, self-employed workers have only been entitled to sickness benefits since 2016. They remain uncovered for some risks including occupational accidents and unemployment, although voluntary options have been introduced in recent years (Chapter 4). This interplay of several autonomous schemes and a complex system of contribution rates and entitlements obscure the relationship between gross and net wages and hinder the mobility of workers across jobs and occupations.

To lower administration costs and improve service delivery, several independent schemes were merged into the Régime social des indépendants (RSI). The RSI remained autonomous in that it was run by elected representatives of its members. It does not have the administrative capacity to collect contributions itsself, but delegates this to URSSAF, a network of private bodies that collect social contributions for dependent employees.3 This requires intense coordination between the two bodies, which could not be achieved, resulting in insufficient collection of contributions, erroneous overcharging of members and delayed payments. These major malfunctions led to the dissolution of the RSI; it is set to be completely absorbed by the general social security system by 2020 (Chapter 4).

Some countries also have special, more generous schemes for artists and other creatives who often have unstable employment patterns and may thus struggle to accumulate the contribution periods necessary to receive social protection benefits. This section discusses two of these schemes: the German artists’ insurance, which charges consumers employer contributions, and the French intermittents du spectacle scheme.

1.6.1. Customers contribute to social protection: the German artists’ insurance

The German writers’ and artists’ insurance scheme directly addresses the double contribution issue by levying social security contributions upon customers of artistic services. Qualifying writers and artists only pay employee social security contributions that make up half of the scheme’s total budget. A public subsidy, justified by private households’ consumption of art and writing, covers 20% of the overall cost. Institutions that rely on the services of artists and writers (e.g. publishers, theatres, libraries or private companies) cover the remaining 30%, proportional to their use of artists’ and writers’ services – the Ministry of Employment and Social Affairs calculates the contribution rate annually, depending on the number of insured artists and the artist and writer fees reported by institutions.

Membership is mandatory but low-earning artists and those with high incomes or private insurance can be exempted. Customer contributions are levied on all expenditure on services commissioned from artists and writers, however, in order to not create distortions on the market for artistic work. This also mitigates some of the cost savings of contracting instead of hiring artists and writers.

Compliance and administrative costs seem comparable to those in the general system. Companies are required by law to declare their expenditure on artistic and writing services. The German Pension Fund is in charge of company inspections for the artists’ insurance, which it carries out in the course of every regular social security compliance inspection since 2015. This keeps administrative costs low and led to a jump in contributing companies by 50 000 within two years. Employers’ associations claim that costs for administering fund contributions can run as high as 70-100% of overall contributions, but it is not clear whether these estimates are realistic, or how they compare to administering social security contributions for regular employees.

Fund membership has continually risen over the past twenty years and stands at 0.5% of total employment in Germany. In line with an increase in Fund membership, the public subsidy has almost quadrupled. While there are no data on how many self-employed artists’ and writers would qualify for the scheme if they applied – admittance is based on reaching a minimum required income in a listed profession – labour force survey estimates based on the number of self-employed artists indicate a coverage rate of about one-third.

While the artists’ insurance scheme seems effective in securing customers’ contributions to artists’ social protection, providing artists with adequate pension entitlements remains challenging. Declared earnings are very low: in 2016, on average, male artists earned just 54% and female artists just 40% of overall average gross earnings in Germany. Combined with the unstable career patterns of many artists and writers, pension contributions on these incomes do not give rise to pensions sufficient to prevent old-age poverty.

Most self-employed artists and writers combine their artistic work with other income, and those at the low end of the income distribution seem to self-select into the fund. A high share of enrolees report incomes just above the minimum threshold for eligibility, which may indicate that they mainly join the fund to gain access to health and long-term care insurance (even minimum contributions guarantee full health- and long-term care coverage, while pensions depend on earnings).

Hence, many insured artists enjoy full access to health insurance while paying only minimum contributions, which leads to a concentration of bad risks in the fund. Low declared earnings may be a result of weak demand and low prices for art-work, but it might also be the result of under-declaration of earnings. Incentives to declare income above the minimum contribution thresholds are weak if individuals either have other, private pension insurance, or their true earnings are still so low that they cannot expect a pension above the poverty line. Many fund members might therefore have to rely on general revenue-financed social assistance in retirement age (Chapter 5).

1.6.2. Intermittents du spectacle

In France, artists and technicians on fixed-term contracts in the entertainment industry benefit from shorter contribution periods for eligibility to the unemployment insurance system under intermittents du spectacle scheme. This special scheme was set up to accommodate the fact that performers and stage and film technicians are hired for productions on a short-term basis and would otherwise struggle to qualify for benefits

The short contribution periods make this scheme very attractive: to qualify for benefits, only 507 hours of work over the past 12 months are required, and spells of employment increase the benefit receipt duration. Thus, claimants can receive the benefit indefinitely, if they work at least two to three months per year.

The number of qualifying workers rose from 50 000 to over a quarter of a million from 1980 to 2015, and about 40% of them claim unemployment benefits. Many claimants become unemployed once they reach the minimum contribution period to qualify for the benefit, and many cycle back and forth between unemployment benefit receipt and working for the same employer, suggesting that some companies adapt their workforce management to the regulations of the scheme (Cour des Comptes, 2012[33]). In 2012, expenditures on benefits exceeded contributions by a factor of more than five. The net cost of the scheme – defined as expenditure on benefits minus contributions to the scheme – was roughly EUR 1 billion in 2012, for a scheme covering about 250 000 workers (Chapter 4). In comparison, that same year the deficit of the general unemployment system in France, covering around 24 million dependent employees, was EUR 2.7 billion (OECD, 2018[34]).

1.7. Social protection of platform workers

Online labour platforms experienced spectacular growth in recent years. According to data provided by the Oxford Internet Institute, the five largest online facilitators of work carried out exclusively online (from low-skilled “click work” to high-skilled freelance jobs) grew by over a third between May 2016 and April 2018 (Oxford Internet Institute, 2018[35]). Standard Labour Force surveys often do not separately capture work facilitated through online platforms, and might underestimate the true incidence where they do, given that platform work is often in the informal economy (OECD, 2018[36]). A recent dedicated survey of internet users in 15 European Union member states suggests that 8% of the working age population of these countries performs work over an online platform at least once a month, and about 2% have platform work as their main source of income (Pesole et al., 2018[3]).

Online platforms have the potential to enable individuals to compete with firms and opt for entrepreneurship instead of employment. They lower transaction costs by providing consumers with information on service quality and monitoring and enforcing transactions, making it less risky for consumers to choose an individual instead of an established firm (Choudary, 2018[37]; OECD, 2016[4]). They also offer individuals access to a network of potential consumers as well as tools to facilitate the sale of goods or services, reducing barriers to market entry. Through online platforms, individuals can achieve a reputation and client base that would have required the resources of a firm in traditional markets.

In lowering the barriers to self-employment, online platforms can disseminate its advantages – flexibility in working time and place, autonomy in the organisation of work – to those who were previously excluded from the labour market, including parents and others with caring responsibilities, or individuals living in areas with weak local labour markets. Platforms can also offer both employed and non-employed individuals an easy way to smooth temporary income shocks.

However, some platforms go beyond a mere “facilitator” or “marketplace” role in determining prices, working times, work attire or details of service provision such as how costumers should be greeted (Choudary, 2018[37]). These controls and regulations guarantee that costumers receive a consistent and standardised service, but undermine the flexibility and autonomy associated with genuine self-employment. Platforms also limit workers’ entrepreneurial decision making in more subtle ways, e.g. by withholding key information, exemplified by rideshare app Uber’s practice of not letting drivers know the rider’s destination before accepting a fare, meaning the driver cannot judge the profitability of the ride (Prassl, 2018[12]).

Thus, gig workers may end up enjoying few of the advantages of self-employment, but suffer many of its drawbacks, including the risk of demand fluctuations and unpaid down- or waiting times, and the lack of employee benefits such as paid vacation, sick-leave or severance pay. As with other self-employed workers, employment protection and minimum wage regulations do not apply to them, and they are often not or not as well covered by social protection.

Some platforms intervene in gig-workers’ price setting, working time and work organisation to such an extent that they have been found to be the de facto employers by national courts. Many jurisdictions have “primacy of facts” principles that allow workers to be classified as employees if they meet minimum standards of dependency on and accountability; examples include FedEx in the United States or Uber in the United Kingdom, and several other cases are pending in other OECD countries (Prassl, 2018[12]). In cases of straightforward misclassification of dependent employees as independent contractors, labour law (when properly monitored and enforced) may thus be sufficient to ensure the adequate protection of workers.

Policy makers who want to improve the employment quality and social protection of gig-workers have to look carefully at the individual case of both worker and platform. Platform workers are a very heterogeneous group. While the majority seem to use platforms to top-up income from other sources, some do rely on platforms for most of their income: a recent survey of platform workers in seven European countries showed that platform work was the only source of income for about 10% of platform workers (Huws, Spencer and Holts, 2017[38]). Thus, while most platform workers are likely to be covered by social protection through a first job or a spouse, a significant minority may be at risk of being unprotected in the event of unemployment, illness or disability.

Similarly, while flexibility in working time and location is the main reason for choosing platform work for many,4 there is a significant minority who work for platforms because no other work is available – in a 2018 survey, nearly one in five Italian and one in ten UK gig-workers stated that platform work was their only option (Boeri, Giupponi, Krueger, & Machin, 2018). Workers who perform gig-work as a last resort, as well as low-skilled workers performing tasks that are highly substitutable for platforms, are more likely to receive low pay and suffer from social protection gaps.

In contrast, there is no obvious difference in the need for social protection between self-employed workers who operate on traditional markets, and those who operate on platforms, but retain entrepreneurial control over their work, such as designers or translators who market their services through online platforms.5

What does distinguish labour platforms from conventional markets is that all transactions are digital and hence completely traceable. This raises the potential for increasing social protection coverage and tax compliance by shifting activities from the informal to the formal economy.

Payments to individual contractors can be difficult to monitor for tax authorities, and underpayments are often not pursued given the low amounts at stake (Prassl, 2018[12]). Unlocking platform payment data for the purpose of the collection of social security contributions and taxes can also lower the administrative burden on individuals, which can be substantial, especially when they derive only occasional or low incomes from self-employment. For example, Indonesia has introduced a compulsory accident insurance scheme for motor-taxies hailed through an online app – a portion of the fare is automatically deducted to insure both driver and passenger for the duration of the trip (ILO and OECD, 2018[39]).

Countries may require platforms to share payment and identification data for tax compliance purposes – e.g. Finland has introduced legislation to access data from peer-to-peer and crowdfunding platforms located in Finland. Requiring platforms to withhold taxes is also a possible tool to increase compliance. However, if the platform is located in a different jurisdiction than the worker, domestic legislation might not be enough to ensure compliance by the platform. Countries may still reach agreements directly with platforms, but data transmission might require user consent because of data protection legislation (OECD, 2018[17]).

Estonia, for example, already collaborates with ridesharing apps Uber and Taxify for income tax purposes: drivers can permit the platforms to transmit income and deduction data to the tax authorities, who pre-fill a tax return for self-employment income to ease the administrative burden on drivers (Loite, 2016[40]). In 2017, drivers declared nearly half a million euros in earnings, over five times as much as in 2016, the first year that earnings information was directly submitted by platforms (Tax and Customs Board, Republic of Estonia, 2017[41]). While such measures are likely to increase compliance by reducing compliance costs for drivers, their reach is limited by the need to obtain drivers consent to share their data.

Thus, there is a strong case for international co-ordination. Through the OECD’s Forum on Tax Administration, 50 tax administrations are currently exploring which data on platform users would be required to match payments to tax records, and how and with which periodicity such data should be transmitted. A common and standardised solution, underpinned by an agreement on information exchange between tax authorities, would reduce burdens both for national tax authorities and platforms, who would not have to handle individual requests by different tax administrations, in different formats and periodicities. Results of these consultations are expected in 2018 (OECD, 2018[17]).

Some countries have developed dedicated schemes for the administration of social security contributions of platform workers. In Sweden, gig-workers are legally required to register with an umbrella company, a private entity who nominally acts as the employer of the gig-worker in administering their payroll tax and social security payments in exchange for a fee. They are thus covered by the general public social protection system, but they are excluded from important components of the social protection system that are part of collective agreements, including additional pension, sickness and accident insurance covers; and they do not have access to union-run unemployment insurance.

These are important limitations: about 25% of all currently paid pensions in Sweden come from occupational schemes as part of collective agreements. This share is expected to grow, as public pensions grow slower than labour earnings. Also, the threshold for pensionable income is indexed to the average wage, and therefore, the net replacement rate for higher-earners falls as wage inequality increases in Sweden. The same effect lowers the net replacement rates for above-average earners in the public sickness and accident insurance, increasing the importance of collectively bargained benefits.

Platform workers also struggle to access the main pillar of the Swedish unemployment insurance, provided by Unemployment Insurance Funds: while they may join a fund for standard workers, in practice, funds have been reluctant to pay out benefits to platform workers, because they have no employer to confirm that they have been laid off (Chapter 8).

In Belgium offers a simplified tax treatment to platform workers who only earn “occasional” income (below EUR 6 130 per year in 2018) from platform work is exempt from social security contributions and income tax. This exemption only applies to services provided to individuals (not firms) and excludes the provision of goods, letting of property and movables (including car-sharing). It is limited to a number of platforms based in the European Economic Area who fully disclose their workers and their remuneration.

Platforms themselves could also provide social protection to workers – e.g. Uber will offer basic medical, accident, maternity and paternity insurance to regular drivers in some European countries starting in mid-2018 (Uber, 2018). Platforms have been reluctant to participate in workers’ social protection, however, to avoid the appearance of an employment relationship.

1.8. Policy lessons

This section summarises some key policy insights from the country studies collected in this publication, and suggests policy options for increasing the income security of on-call workers or those on flexible hours contracts.

1.8.1. Social security contributions should be harmonised across forms of employment as much as possible

Including workers that sit on the border between dependent and independent forms of work in the standard social protection scheme closes coverage gaps and helps ensure that social protection systems cover those who are most at risk.

It can also curb the extent of non-standard employment, and thus limit the erosion of the contribution base of social protection systems, as shown by the policy reform experiences in Italy and Austria. Raising non-wage labour costs, of course, comes at the risk of decreasing employment, in the same way as this can be the case for standard workers. If certain forms of employment are subject to lower non-wage labour costs, this should be a deliberate policy choice.

1.8.2. Voluntary schemes do not seem to work well for non-standard workers

Any insurance depends on risk sharing among members. With a voluntary insurance scheme, those who have the highest risk have the greatest incentive to join. Unless a voluntary scheme achieves a very high coverage rate, this adverse selection either leads to a downward spiral of rising premia and falling coverage, or to additional costs in the system. High coverage rates, in turn, may require public subsidies, as the willingness to pay voluntarily for social protection appears to be low in some cases, as seen, for example, in Sweden where voluntary unemployment insurance coverage fell after a moderate rise in contribution rates in 2007/08.

Adverse selection in voluntary schemes is not limited to unemployment insurance but can be observed for sickness benefits (voluntary sickness benefits for self-employed workers in Austria) and even maternity and carers’ benefits (evidenced by the Canadian Special Benefits for Self-employed Workers scheme).

1.8.3. Making entitlements portable supports mobility across jobs and forms of employment

Untying entitlements from specific relationships with employers, and tying them to individual contributions instead, not only makes it easier for workers to switch jobs, but it also makes it easier for them to switch between self- and dependent employment. Furthermore, it facilitates the harmonisation of entitlements across contractual arrangements. Individualised forms of social protection, however, can only offer protection if sufficient contributions are paid by or on behalf of the beneficiary.

Austria offers an interesting policy lesson in this respect when, in 2003, its severance pay scheme was replaced by company-based pension accounts. While the old severance pay entitlements benefitted only laid-off employees, all dependent employees now have a company-based pension account, which is portable across jobs. This measure increased job mobility for workers in distressed firms (where a plant closure or mass layoff will take place in the near future, Kettemann et al. (2016[42])). As the pension accounts are tied to the individual, and employers contribute a fixed rate of individual earnings, it was easy to extend the programme to independent contractors in 2008.

1.8.4. Increasing income security for those working flexible hours

Independent contractors – whether they do work mediated by online platforms or not – as well as workers on on-call or flexible hours contracts lack the income security provided by regular employment relationships while enabling firms to cheaply adjust to demand fluctuations. One way to redress this imbalance is to introduce a wage premium for flexible work as a compensation for assuming part of the entrepreneurial risk. The idea of requiring employers to pay higher rates to those who assume part of the entrepreneurial risk has been gaining traction both in the context of platform work (Chapter 8) as well as flexible hours work contracts (Taylor, 2017[43]).

In Australia, casual workers, who may be terminated on short notice and do not receive some benefits like paid holidays or paid sick leave that are provided to permanent employees, are already entitled to a casual loading of 25% for each hour worked compared to a worker doing the same job on an ongoing basis (Chapter 2).6

On-call or other flexible workers would also benefit from minimum hours guarantees: in Australia and the Netherlands, for example, employers already have to pay on-call workers at least three hours of work each time they call on the employee (Chapter 7).

Minimum earnings floors may also be applied to independent contractors: the New York Taxi and Limousine Commission is currently considering introducing a minimum driver pay standard for drivers for ride-hailing apps. The proposal would set an earnings floor at roughly 115% of the minimum wage and increase driver earnings after expenses by 22.5% on average (Parrott and Reich, 2018[44]). The Netherlands plans to introduce a payment floor of EUR 15-18 per hour for independent contractors who are paid below 125% of the statutory minimum wage or the lowest wage in the relevant collective agreement (Government of the Netherlands (2017[45]) and Chapter 7).


[11] Abraham, K. et al. (2017), Measuring the Gig Economy: Current Knowledge and Open Issues, U.S Census Bureau, (accessed on 10 December 2017).

[20] Arbejdsgruppen om selvstændige i dagpengesystemet (2017), Et nyt dagpengesystem for fremtidens arbejdsmarked. Report from the working group on self-employed persons in the unemployment benefit system,

[19] Avlijas, S. (2018), “The dynamism of the new economy: Non-standard employment and access to social security”, Research note.

[15] Berg, J. (2016), “Income security in the on-demand economy: Findings and policy lessons from a survey of crowdworkers”, Conditions of Work and Employment Series, No. 74, ILO, Geneva, (accessed on 12 December 2017).

[10] BLS (2018), Contigent and alternative employment arrangements - May 2017, (accessed on 11 June 2018).

[9] BLS (2005), Contingent and alternative employment arrangements, February 2005, (accessed on 11 June 2018).

[16] BMAS (2017), White paper work 4.0, German Federal Ministry of Labour and Social Affairs,

[37] ILO (ed.) (2018), The architecture of digital labour platforms: Policy recommendations on platform design for worker well-being,

[33] Cour des Comptes (2012), Le régime des intermittents du spectacle: la persistance d'une dérive massive, Cour des comptes.

[6] Datta, N., G. Giupponi and S. Machin (forthcoming), “Zero Hours Contracts and Labour Market Policy”, mimeo.

[23] Delsen, L. and J. Smits (2014), “The Rise and Fall of the Dutch Savings Schemes”, SSRN Electronic Journal,

[29] Employment and Social Development Canada (2016), Evaluation of the Employment Insurance Special Benefits for Self-employed Workers,

[46] ESIP (2017), Statutory Pension Insurance of Digital Platform Workers - a comparative perspective, ESIP,

[22] Etsy (2016), Economic Security for the Gig Economy,

[30] European Comission (2018), Behavioural study on the effects of an extension of access to social protection for people in all forms of employment, European Commission,

[8] European Commission (2018), Commission Staff Working Document, Impact Assessment accompanying the Proposal for a Council Recommendation and Access to social protection for workers and the self-employed, SWD(2018)71,

[7] European Commission (2018), Proposal for a Council Recommendation and Access to social protection for workers and the self-employed, COM(2018)132.

[45] Government of the Netherlands (2017), Confidence in the Future - 2017-2021 coalition agreement,

[32] Hofer, H., R. Hyee and G. Titelbach (2018), Where did all the independent contractors go?.

[38] Huws, U., N. Spencer and K. Holts (2017), Work in the European Gig Economy, FEPS,

[39] ILO and OECD (2018), Promoting adequate social protection and social security coverage for all workers, including those in non-standard forms of employment, Paper prepared for the G20 (Argentinian presidency).

[18] ISSA (2012), Handbook on the extention of social security coverage to the self-employed,

[27] Kassam, A. (2018), “Ontario's new Conservative government to end basic income experiment”, The Guardian,

[2] Katz, L. and A. Krueger (2016), “he Rise and Nature of Alternative Work Arrangements in the United States, 1995-2015”, NBER Working Papers, p. No. 22667.

[42] Kettemann, A., F. Kramarz and J. Zweimüller (2016), “Beyond Severance Pay: Labor market responses to the introduction of occupational pensions in Austria”, Mimeo, University of Zurich,

[21] Kvist, J. (2017), “Denmark: a new unemployment insurance scheme for the future labour market”, ESPN Flash Report 2017/45,

[40] Commission, E. (ed.) (2016), e-stonian look at: collaborative economy, (accessed on 1 July 2018).

[31] Ministry of Finance (2015), IBO Zelfstandigen zonder personeel, Ministry of Finance, Netherlands.

[28] Moral-Arce, I., J. Martí-Román and Á. Martí-Román (2018), “Cessation of activity benefit of Spanish self-employed workers: a heterogeneous impact evaluation”, MPRA Paper No. 85111,

[36] OECD (2018), “Policy brief on the future of work”, in Online work in OECD countries,

[34] OECD (2018), Social Expenditure Database (SOCX),

[17] OECD (2018), Tax Challenges Arising from Digitalisation – Interim Report 2018: Inclusive Framework on BEPS, OECD/G20 Base Erosion and Profit Shifting Project, OECD, Paris,

[24] OECD (2017), “Basic income as a policy option: Can it add up?”, Policy brief on the future of work , (accessed on 14 September 2017).

[4] OECD (2016), “New Forms of Work in the Digital Economy”, OECD Digital Economy Papers, No. 260, OECD Publishing, Paris,

[13] OECD (forthcoming), The tax treatment of non-standard work, OECD Publishing, Paris.

[35] Oxford Internet Institute (2018), The Online Labour Index,

[44] Parrott, J. and M. Reich (2018), An Earnings Standard for New York City's App-based drivers: Economic Analysis and Policy Assessment. Report for the New York City Taxi and Limousine Commission, The New School, Center for New York City Affairs,

[5] Parsons, D. (1986), “The employment relationship: Job attachment, work effort, and the nature of contracts”, in Ashenfelter, O. and R. Layard (eds.), Handbook of Labor Economics, Elsevier, North-Holland,

[3] Pesole, A. et al. (2018), Platform workers in Europe, Publications Office of the European Union.

[26] Peter, L. (2018), No plans to expand Finland basic income trial,

[25] Piachaud, D. (2016), “Citizen's Income: Rights and Wrongs”, Case, No. 200, Centre for the Analysis of Social Exclusion (CASE), London,

[12] Prassl, J. (2018), Humans as a service - The Promise and Perils of work in the gig economy, Oxford University Press,

[1] Spasova, S. et al. (2017), Access to social protection for people working on non-standard contracts and as self-employed in Europe, European Commission, Brussels,

[14] SSA and ISSA (2018), United States of America, Social Security Administration,

[41] Tax and Customs Board, Republic of Estonia (2017), “Income declared trough Uber and Taxify has overwhelmingly increased”, Press release, (accessed on 1 July 2018).

[43] Taylor, M. (2017), Good work: the Taylor review of modern working practices, Her Majesty's Government,


← 1. Those who were already self-employed upon the introduction of the programme in 2009 had one year to decide whether to opt-into the programme.

← 2. They have to contribute to pension, health and accident insurance since the late 1990s; in 2008 they were incorporated into the unemployment insurance system, the insolvency remuneration fund, the chamber of labour and the severance pay scheme Abfertigung neu (a form of portable employer-funded pension account).

← 3. The URSSAF (Unions de recouvrement des cotisations de sécurité sociale et d’allocations familiales) are private bodies collecting employees’ and employers’ contributions to the general social security system, see Chapter 4.

← 4. In a 2015 survey, around 30% of gig-workers performing work exclusively online on the Amazon Mechanical Turk and Crowdflower platforms named a preference to work from home, or only being able to work from home as the main reason for choosing this type of work (Berg, 2016[15]).

← 5. Indeed, most social protection systems do not seem to make such a difference: a survey conducted by the European Social Insurance Platform, an association of European statutory social insurance organisations (ESIP, 2017[46]) found that in all twelve participating European Union and EEA member countries, workers providing “virtual tasks” (graphic design, translation etc.) would be considered as self-employed. While there are differences in statutory coverage by the pension insurance across the surveyed countries, they are the same as for other self-employed persons.

← 6. Empirical evidence suggests, however, that casual workers do not in fact receive higher wages and might even suffer a wage penalty compared to similar permanent workers (see Chapter 2 for an overview of recent evidence). Thus, monitoring is crucial (as in the case of minimum wages).

End of the section – Back to iLibrary publication page