This chapter provides an in-depth discussion of the income security and work incentives provided by unemployment benefits to jobseekers with previous periods of part-time and unstable dependent employment. It sheds light on the accessibility and adequacy of this key social protection tool, and on the work incentives affecting different types of workers, two key factors for its design and implementation. In particular, the chapter compares entitlements to unemployment benefits for workers with a range of typical employment trajectories, including alternating spells of dependent employment and unemployment. Issues, such as the extension of out‑of‑work support to individuals in “part-time” or “partial” unemployment, options for accumulating entitlement rights across different spells of employment, saving “unused” benefit entitlements for future out‑of‑work spells, and strategies for integrating in-work and out-of-work support, are all assessed.
OECD Employment Outlook 2020
2. Unemployment benefits and non‑standard dependent employment: Striking the balance between income security and work incentives
Abstract
In Brief
Key findings
This chapter provides an in-depth discussion of how unemployment benefits provide income security and work incentives for jobseekers whose recent work history included periods of non-standard dependent employment. Non-standard dependent employment refers to wage and salary workers working either on a part-time or on an unstable basis (i.e. involving frequent transitions between dependent employment and unemployment over a number of years, see Box 2.1 for details). Typically, entitlements to unemployment benefits depend on the characteristics of the job held before becoming unemployed (e.g. number of months employed, wage, hours worked). Since such characteristics refer to a period rather than just to the moment before becoming unemployed, this chapter assesses non‑standard dependent employment in terms of trajectories instead of the type of contract. This allows for a more nuanced and sophisticated analysis of the interactions between unemployment protection and non-standard employment.
Using longitudinal microdata, unemployment benefit legislation and simulations of entitlements under several policy-relevant employment trajectories, the main findings of this chapter are:
Workers with non-standard dependent employment trajectories make up a sizeable part of dependent employment (22% on average across 26 European OECD countries). On average across European OECD countries, 69% of employees who have been unemployed have histories of non-standard dependent employment (51% in the form of unstable employment and 18% as part-time employment).
Non-standard dependent employment, particularly part-time work, has increased in the last decade, especially among the young. In the decade leading up to 2016‑18, non-standard dependent employment rose by 5 percentage points on average among employees aged 20 to 29. Most of the rise was due to an increase in part-time employment. Yet, there is considerable variation across countries. In Spain, for example, the share of unstable dependent employment among youth increased by 8 percentage points, in spite of the contraction of fixed-term contracts (OECD, 2019[1]).
Women are more than two times more likely to work part time than men. On average, almost one quarter of women – often mothers – work part-time, although figures differ widely across OECD countries. Part-time employees tend to receive lower hourly wages in all OECD countries, have higher job insecurity and participate less in training. Yet, in a number of countries, many part-time employees work shorter hours by choice and as a way to achieve work-life balance.
Poverty rates for workers with non-standard dependent employment trajectories are higher than for workers with standard employment. Across European OECD countries, on average, 10% of workers in part-time employment and 19% of workers in unstable employment live in a household with an annual disposable income of less than 50% of the national median. For standard employees the figure is only 3%.
Unemployment benefit legislation directly affects the duration of benefit entitlements and the level of benefits payable for workers with standard and non-standard dependent employment trajectories. Required minimum employment durations, working hours or earnings are harder to meet for those in unstable or part-time employment. Some countries operate second-tier unemployment benefits with less demanding employment requirements, which are easier for jobseekers with non-standard employment trajectories to access.
Analysis of detailed rules of first- and second-tier unemployment benefits in 11 OECD countries (Australia, Austria, Canada, Finland, France, Japan, Latvia, Netherlands, Poland, Spain and the United Kingdom) finds that the same rules can affect workers with part-time and unstable employment differently from standard employees.
Benefit claimants with a history of part-time employment usually have the amount of their benefits calculated in line with the number of hours worked in their previous job. In Australia, Poland and the United Kingdom, however, jobseekers previously in part-time employment receive the same amount as those previously working full-time.
Many countries operate part-time unemployment schemes, which combine the entitlement to unemployment benefit with some casual part-time work. In most countries for which the analysis was undertaken, the amount of benefit is reduced in such cases, although there is a considerable variation in the rules among countries. In Latvia and Poland, jobseekers receive no unemployment benefit if they take up any amount of work.
Jobseekers can keep unused entitlements for future claims if they find full-time work before benefits expire, in all analysed countries but in Latvia and Poland. Countries use a range of rules concerning the possibility for jobseekers to keep and access unused entitlements. In Canada, the Netherlands and the United Kingdom, strict conditions determine whether entitlements are retained for subsequent unemployment spells. If the jobseeker with an unused entitlement accrues a new one from more recent employment, she must choose between them in Spain, while can utilise both in France – first using up the older entitlement and then claiming the new one.
Simulations based on a set of typical employment trajectories for four OECD countries (Australia, France, Latvia and Spain) show that benefit entitlements can vary substantially depending on the type of employment and the specific sequence of in-work and out-of-work spells. Significant differences exist even when workers have the same personal characteristics and same level of earnings and working hours over a given period. These gaps call into question the adequacy of benefit generosity for all types of workers.
In Australia and France, entitlements do not differ significantly between jobseekers with fairly stable employment records (working for a full year before becoming unemployed) and jobseekers with more unstable employment trajectories (working either every other month or every other semester, but for the same total number of hours). In France, an upcoming reform will reduce the benefit entitlements of workers with unstable employment records, as the calculation will take into account the months in which she has not worked. In Latvia and Spain, where contribution requirements are more demanding, jobseekers who previously had more unstable employment records are entitled to fewer months of benefits.
In all four countries, entitlements are lower for workers on part-time unemployment. In Australia, France and Spain, earnings from work reduce the benefit amount. In Latvia, benefits are suspended as soon as the jobseeker has some earnings from employment.
Unemployment benefits also create different financial incentives to work across countries and employment trajectories. On average, work incentives are weaker in France, followed by Australia, Spain and Latvia. In Latvia and Spain, however, in some cases work incentives are stronger because some employment trajectories do not lead to any entitlement to unemployment benefits.
To address the main issues of entitlement and access, on the one hand, and work incentives, on the other, for workers with non-standard employment trajectories, a number of policy tools could be considered:
Reductions in benefit payments when jobseekers have earnings from part-time unemployment schemes help to smooth income variations and to prevent benefit misuse. However, they also affect the neutrality of the system and reduce the incentives of using casual part-time work as a stepping stone to a better job. Earnings disregards and withdrawal rates can help balance these objectives. Earnings disregards are the amounts of earnings not considered when calculating benefits during periods of partial work. The withdrawal rate is the percentage of earnings (above any disregard) by which benefit is reduced in such cases.
Easy access to benefits and the possibility of retaining unused benefits for future out-of-work spells may encourage frequent periods of unstable employment or increase the overall duration of unemployment benefits. Waiting periods, moderately long employment requirements, reductions in case of frequent reclaiming, and limits on the accumulation of old and new entitlements can help to prevent such distortions. However, employment requirements, even if they provide work incentives, prevent benefit abuse and protect the system’s financial sustainability, they also penalise workers with unstable employment records. Customised extensions of reference periods for workers more prone to job instability (e.g. young workers, employees with temporary contracts) could alleviate these negative effects. Differentiated unemployment insurance contribution rates could also be used to create financial incentives for employers and employees to choose more stable employment contracts and discourage collusion.
Work incentives can be enhanced through improved co-ordination of in-work and out‑of‑work support. Potential measures include: extending to full-time workers with low wages the possibility to cumulate earnings and unemployment benefits; making use of in-work benefits; or integrating in- and out-of-work benefits.
During periods of economic downturns, unemployment benefit instruments can be adjusted to the changes in labour market circumstances. In the current COVID‑19 crisis, many countries extended maximum duration and generosity of unemployment benefits. A number of countries also adopted measures to facilitate benefit access, such as reducing minimum employment requirements or extending reference periods, which are likely to specifically increase benefit coverage of non-standard workers during the emergency.
Introduction
Non-standard forms of dependent employment (i.e. jobs that are part time or of short duration) represent a significant share of wage and salary workers. On average across OECD countries, part-time employment accounts for 16.5% of all employment – 2 percentage points more than two decades ago (OECD, 2020[2]). Temporary employment accounts for 11.7% of dependent employment and has remained somewhat stagnant in recent years. Among young workers, however, temporary employment makes up 25.7% of dependent employment – almost 2 percentage points more than 20 years ago (OECD, 2020[3]). Non‑standard forms of dependent employment have been associated with increasing labour market instability, underemployment and economic vulnerability, particularly among young people and those with less than tertiary education (OECD, 2019[1]).
As also stressed in the 2019 edition of the OECD Employment Outlook (OECD, 2019[1]), in many countries, social protection systems in general, and unemployment benefits in particular, have not yet fully adapted to the specific needs and circumstances of workers in non-standard forms of dependent employment. Analyses of out-of-work social benefits in six countries (France, Greece, Hungary, Italy, Spain and the United Kingdom) found that in Spain and Italy unstable employees were significantly less likely to receive benefits than standard employees (i.e. those with full-time jobs and open-ended contracts). This gap was also considerable in the United Kingdom. The social protection gap between standard and non-standard employees was even larger when assessed in terms of benefit generosity (benefit amount as a proportion of median income), especially in Greece, Italy and Spain (OECD, 2019[1]).
Due to their unusual work trajectories, jobseekers with previous non-standard dependent employment may not receive the type of support that unemployment benefits offer to jobseekers with standard employment careers. In fact, non-standard employees’ access to unemployment benefits tends to be more difficult than to other insurance programmes such as maternity and sickness benefits (Avlijas, 2019[4]). Non-standard dependent employment may also be one of the factors that explain why, in many OECD countries, only a small share of jobseekers receive unemployment benefits – fewer than one‑third on average across 32 countries (OECD, 2018[5])).
At the same time, depending on their design, unemployment benefits may contribute to job instability. For example, they may provide incentives for alternating between short periods of employment and unemployment (Boeri, Cahuc and Zylberberg, 2015[6]). In particular, this can arise in the absence of waiting periods, overly short qualification periods for contributory benefits and ill-designed partial unemployment insurance schemes (Kyyrä, 2010[7]; Le Barbanchon, 2016[8]; Fontaine and Malherbet, 2016[9]). Results in OECD (2019[1]) illustrate that, in some circumstances, entitlements for those with patchy work histories may be equally or more generous than for those with more stable employment records.
This chapter provides an in-depth review of the key policy mechanisms that affect how unemployment benefits strike a balance between income security and financial work incentives for non-standard employees. By providing comparable cross-country estimates of protection for jobseekers with different employment trajectories, the chapter seeks to provide input into the policy debate on the accessibility, adequacy and effectiveness of unemployment benefits for different types of employees.
Typically, entitlements to unemployment benefits depend on the characteristics (e.g. number of months employed, wage, hours worked) of the job held before becoming unemployed. Since such characteristics refer to a period rather than just to the moment before becoming unemployed, this chapter assesses non-standard dependent employment in terms of trajectories instead of the type of contract (Box 2.1 explains in detail how employment trajectories are defined and measured). Thus, employment is observed and characterised over a continuous period, instead of at a single point in time. This approach provides a more refined assessment of job instability and the possibility to measure the unemployment risk of each type of employment. In turn, this allows for a more nuanced and sophisticated analysis of the interactions between unemployment protection and non-standard employment.
The chapter is organised as follows. Section 2.1 assesses the scale and development of unstable and part-time dependent employment and the characteristics of workers with such trajectories, comparing them with workers with stable full-time dependent employment trajectories. Section 2.2 takes stock of the legal provisions on unemployment benefits that may give rise to an uneven treatment of standard and non-standard employees. Section 2.3 sheds light on the provision and level of unemployment benefits in a number of non-standard dependent employment trajectories, by making use of simulation techniques. Furthermore, it computes indicators of income security and financial work incentives for a range of policy-relevant scenarios of standard, unstable or part-time employment. Section 2.4 discusses desirable features in the design of unemployment benefits to provide reliable and effective support to workers with non-standard employment.
Box 2.1. Trajectories of non-standard dependent employment: definitions and measures
“Non-standard” employment is an umbrella term that typically covers all temporary, part-time and self-employment arrangements, i.e. everything deviating from the “standard” of full-time, open-ended employment with a single employer (OECD, 2014[10]; OECD, 2018[5]). This chapter analyses a slightly different employment group: “non-standard dependent employment”.
Definitions
Non-standard dependent employment refers to wage or salary workers who experience periods of part-time or unstable work. Self-employed workers and persons not in the labour force are not included. The OECD defines part-time employment as people in employment who usually work less than 30 hours per week in their main job (OECD, 2020[2]). Unstable employment is defined here as a situation characterised by frequent transitions between employment and unemployment over a number of years.1
Measurement for the empirical analysis
The empirical analysis in this chapter is based on panel data of the European Union Statistics on Income and Living Conditions (EU-SILC). Employment trajectories are measured through a continuous month-to-month observation of the employment statuses of individuals over a period of 36 months.
The definition and measurement of unstable employment used here differs from previous OECD work, which focussed on transitions from one year to the next (OECD, 2019[1]). In the context of the present chapter, annual transitions may conceal one (or several) transitions into and out of employment that occur over the course of the reference year.2 By measuring monthly transitions, the definition used here is more likely to capture transitions that involve short periods of unemployment between jobs.
Workers in part-time dependent employment are people reporting to work for a wage or salary on a part-time basis over most of the period of 36 months. The definition is self-reported and implausible information is detected and corrected as far as possible (Eurostat, 2018[11]).
Workers in unstable dependent employment are people reporting to work for a wage or salary with at least three transitions between (full-time or part-time) dependent employment and unemployment over the 36‑month period. The rationale for requiring a minimum of three transitions is to ensure that the person experienced at least two spells of unemployment during the 36‑month period. Workers reporting at least three transitions are classified into unstable employment even if part-time work is the most frequent employment status.3
Standard employment refers to workers whose dependent employment is stable (i.e. not unstable) and work mainly on a full-time basis within three years.
Sample
The sample pools together rolling panel samples ending on the years 2016, 2017 and 2018 (and the years 2006, 2007 and 2008) to minimise year-to-year variability in small subsamples. The sample is restricted to individuals aged between 20 and 59 years old who declared to be in dependent employment for at least one month in each of the three observed years. People reporting some months not in the labour force (e.g. students or pensioners) are included only if they were active (i.e. employed or unemployed) in most of the three‑year period and were employed at least one month in each calendar year. Individuals experiencing self-employment are only part of the sample if self-employment is not the dominant status and there is no transition from self-employment to unemployment.
1. Employees with unstable jobs are closely associated with employees with temporary contracts (see Section 2.1.3). However, instead of being defined based on the type of contract they are identified based on the actual instability of their employment trajectories. Thus, this definition may include employees with permanent or long-term contracts who have frequent periods of unemployment and may exclude employees with temporary contracts who remain employed either with the same or with a new employer. It is possible that workers with permanent contracts have different underlying reasons and labour market pressures for changing jobs from workers with temporary contracts. However, regarding entitlements to unemployment benefit, rules are more sensitive to work tenure than the type of contract. Furthermore, across OECD countries, there is wide variation in the use of permanent and temporary contracts, partly reflecting differences in employment protection legislation (see Chapter 3).
2. On average across analysed countries, 1.4% of employees in the 2016‑18 sample have employment trajectories that are both mostly part time and unstable (i.e. with many transitions).
3. On average across analysed countries, 1.4% of employees in the 2016‑18 sample have employment trajectories that are both mostly part time and unstable (i.e. with many transitions).
2.1. Trajectories of non-standard dependent employment
According to OECD data for 2018, part-time employment accounts for 16.5% of all employment – 2 percentage points more than two decades ago (OECD, 2020[2]). Temporary employment accounts for 11.7% of dependent employment and has remained somewhat stagnant. Among young workers, however, it makes up 25.7% of dependent employment – almost 2 percentage points more than 20 years ago and 8 percentage points more than in 1980 (OECD, 2020[3])
While some non-standard jobs, particularly for high-skilled professionals, may pay high and stable earnings and provide good working conditions, there is an association between non-standard work and poorer job quality, particularly for workers with low and middle skills. Wages tend to be lower, employment less protected, access to employer and social benefits reduced, safety and health risks greater, investments in lifelong learning lower, and bargaining power weaker (OECD, 2014[10]; OECD, 2019[1]).
Non-standard jobs are also associated with higher job instability. In a context of job polarisation (see Chapter 4), some workers (particularly those in middle-skill occupations) may not enjoy the positive aspects of low job tenure, like upward transitions when former jobs disappear. Past OECD research (OECD, 2013[12]) indicates that workers who lose their job involuntarily experience a fall in job quality after re-employment. Whereas income losses affect most previously displaced workers, the magnitude of this effect differs by country (lower in Northern European countries, higher in others), age (higher and more persistent for older workers), skill level (higher for low-skilled workers) and gender (higher for men). Furthermore, displaced workers are more likely to be re-employed in temporary or part-time jobs.
Evidence from the 2019 edition of the OECD Employment Outlook (OECD, 2019[1]) shows that job instability has increased in most OECD countries once changes in the demographic composition of the workforce are taken into account. The average job tenure decreased across the OECD by 4.9% (or around five months) between 2006 and 2017. The largest declines in job stability occurred for low-educated workers (i.e. those without an upper secondary qualification).
This section assesses the incidence, characteristics and trends of workers with trajectories of unstable and part-time dependent employment (see Box 2.1 for detailed definitions), and compares them to those with stable full-time dependent employment histories.
2.1.1. How prevalent are non-standard employment trajectories?
Workers with non-standard dependent employment trajectories make up a sizeable part of dependent employment. On average across 26 European OECD countries for which data is available, non-standard dependent employment accounts for 22% of employees, part-time employment comprises 16% and unstable employment makes up 6%. Among employees who experienced spells of unemployment, past unstable employment accounts for 51%, standard employment for 30% and part-time employment for 18%.
The shares of employees with trajectories of unstable and part-time dependent employment vary considerably across countries. Unstable employment is lower than 5% of dependent employment in Belgium, Czech Republic, Denmark, Estonia, Germany, Luxembourg, Norway, Slovak Republic and the United Kingdom, while it amounts to more than 10% in Finland, Greece and Spain. Part-time employment is low in Eastern European countries, Portugal and Finland, above 25% in Belgium, Germany and Switzerland, and highest in the Netherlands (41%).
In most countries, previous patterns of non-standard dependent employment account for the bulk of workers who experience unemployed, with unstable employment being the largest group. Workers with past unstable employment make up at least half of the unemployed in Austria, Finland, Greece, Iceland, Hungary, Spain, Sweden and the United Kingdom. The shares are lowest in Czech Republic and Slovak Republic. The share of unemployed with previous part-time employment is generally low (less than 20%), except in Germany, Ireland and the Netherlands, where they make up at least 30%, and, to some extent, in Belgium, France, Italy and Switzerland.
Figure 2.1. Distribution of employment trajectories across employees and unemployed
Employment trajectory types, 2016‑18 (in percentage)
Note: Data is from the longitudinal panel of the EU Statistics on Income and Living Conditions (EU-SILC). The sample pools together rolling panel samples for the years 2016, 2017 and 2018 to minimise year-to-year variability in small subsamples. The sample is restricted to employees between 20 and 59 years old. People not in the labour force (e.g. students or pensioners) are excluded. OECD: weighted average of listed countries.
1. Employment trajectories are measured through a continuous month-to-month observation of the employment statuses of individuals over a period. Unstable dependent employment comprises individuals with at least three transitions between employment and unemployment in 3 years. Standard employment refers to individuals whose employment is stable (i.e. not unstable) and work mainly on a full-time basis within 3 years. Part-time employment refers to individuals whose employment is stable and work mainly on a part-time basis within 3 years.
2. The unemployment subsample is defined as employees who have been unemployed for at least one month in the year. Employment trajectory types in this case refer to employment trajectories before entering unemployment. The unemployment status is based on individual self-classification of main activity status stated by respondent. This definition may not satisfy the ILO criteria for unemployment, which require active job search and immediate availability.
Source: Longitudinal EU-SILC.
2.1.2. More young workers are in non-standard dependent employment than a decade ago
Non-standard dependent employment, particularly part-time work, has slightly increased in the last decade, especially among the young. However, on average across European OECD countries between 2006‑08 and 2016‑18, the rise has been small and restricted to part-time employment (Figure 2.2, Panel A). These patterns are partly confounded, however, by changes in the demographic composition of the labour force. Population ageing and pension reforms (restricting early retirement and increasing the statutory retirement age) have increased the proportion of older workers1, whose jobs are usually stable and full-time (OECD, 2019[1]).
Focusing on young employees (aged 20 to 29), non-standard dependent employment rose by 6 percentage points on average between 2006‑08 and 2016‑18 (Figure 2.2, Panel B). Again, most of the rise took place among part-time employment, with unstable employment rising by just 1 percentage point. Young unstable employment increased considerably in Denmark, Greece, Ireland, Slovenia and Spain. In contrast, the share of young employees with unstable employment trajectories fell in the Netherlands and Norway, where it was offset, in part, by increases in part-time employment. In fact, part-time employment among youth increased in 20 out of 23 countries. The rise was notably high (more than 10 percentage points) in Greece, Italy, Ireland and Spain, while only in Denmark the share of part-time jobs among young employees fell significantly. These changes may be affected, in part, by the economic cycle, as the reference period (2006‑08) may partly reflect labour market consequences of the abrupt downturn associated with the Great Recession.
2.1.3. Young workers on temporary jobs are more likely to have unstable employment trajectories
Unstable employment trajectories are more prominent among workers who are single and young, have less than upper secondary education, work in jobs with temporary contracts, are in low-skill occupations, and earn less than the median monthly wage. The results in Table 2.1 indicate that workers on temporary contracts are more likely to experience unstable employment. On average across countries shown, a person with a temporary contract is 17% more likely to be in unstable employment than an employee with a permanent contract. The association between temporary contracts and unstable employment is particularly strong in Austria, Estonia, Finland, Greece and Spain.
Most workers with unstable employment have a temporary contract – 64% on average across European OECD countries (Figure 2.3), but cross-country variation is considerable. Temporary contracts account for the vast majority of workers with unstable employment in Poland and Southern European countries. On the other hand, in the Baltic countries, Austria, Denmark, Iceland, Ireland, Norway, Switzerland and the United Kingdom, most workers with unstable employment have permanent contracts (fewer than 30% have temporary contracts), reflecting, in part, the low incidence of temporary contracts in employment and the lighter protection of open-ended contracts against dismissal (see Chapter 3).
In most countries, unstable employment is also more likely among jobs in low- and middle-skilled occupations, particularly in Austria, the Netherlands, Spain and Eastern European countries (Table 2.1). Unstable employment is also more frequent among young people, especially in Northern European countries as well as Hungary, France and Switzerland. In many countries, people with less than upper secondary education are more likely to be in unstable employment (especially in Greece, Hungary, Latvia and Poland), but in Denmark and Germany, the estimates suggest the opposite. Unstable employment is more frequent among women than men in Belgium, Finland and Sweden; while the opposite holds in Austria, Denmark and Germany. In most countries, people in couples (i.e. married or in partnership) are less likely to be in unstable employment; this is particularly the case in the Denmark, Hungary and Lithuania. The association between having children and unstable employment is ambiguous; it is positive and significant in Germany, Luxemburg, Norway and Portugal but negative in the Czech Republic, the Netherlands and Poland. Higher monthly wages (normalised to within-country median wage) are negatively related to unstable employment in most countries, particularly in Austria, Denmark, Finland and Luxembourg; however, these results must be treated with caution as they may be biased by the wage data available.2
Figure 2.2. More workers experience trajectories of non-standard dependent employment
Percentage point change in the share of employees with standard, part-time and unstable employment trajectories, 2006‑08 to 2016‑18
Note: Data is from the longitudinal panel of the EU Statistics on Income and Living Conditions (EU-SILC). To minimise year-to-year variability in small subsamples, the sample pools together rolling panel samples. The reference sample includes panels from 2006‑08 and the recent sample panels from 2016‑18. The sample is restricted to employees between 20 and 59 years old. People not in the labour force (e.g. students or pensioners) are excluded. OECD: weighted average of listed countries. Employment trajectories: unstable dependent employment comprises those individuals with at least three transitions between employment and unemployment in 3 years. Standard employment refers to those individuals whose employment is stable (i.e. not unstable) and work mainly on a full-time basis within 3 years. Part-time employment refers to those individuals whose employment is stable and work mainly on a part-time basis within 3 years.
Source: Longitudinal EU-SILC.
Table 2.1. Incidence of unstable employment, by socio-demographic characteristics and countries
Note: Marginal effects from probit regressions. ***, ** and * denote statistical significance at the 1%, 5% and 10% significance level.
The dependent variable “unstable employment” equals 1 if employment is unstable, i.e. there are at least 3 transitions between employment and unemployment, and 0 if employment is not unstable. Estimated marginal effects indicate the estimated change in the probability of being in unstable employment associated with an increase of a given continuous independent variable, and are calculated for an infinitesimal change of that variable from the sample average. Marginal effects for categorical variables refer to a discrete change from the base level. While unstable employment is based on month-to-month information (Box 2.1), the characteristics used are observed in the data only once a year. Data from the longitudinal panel of the EU Statistics on Income and Living Conditions (EU-SILC). The sample pools together longitudinal samples ending with the years 2016, 2017 and 2018 to minimise year-to-year variability in small subsamples. The sample is restricted to employees between 20 and 59 years old; people not in the labour force (e.g. students or pensioners) are excluded. All countries: sample includes observations for all listed countries.
Source: Longitudinal EU-SILC.
Figure 2.3. Most workers with unstable employment have temporary contracts
Share of temporary contracts by type of employment trajectories, 2016‑18, in percentage
Note: Data from the longitudinal panel of the EU Statistics on Income and Living Conditions (EU-SILC). The sample pools together rolling panel samples ending with the years 2016, 2017 and 2018 to minimise year-to-year variability in small subsamples. The sample is restricted to employees between 20 and 59 years old; people not in the labour force (e.g. students or pensioners) are excluded. OECD: weighted average of listed countries. Employment trajectories: measured through continuous month-to-month observation of the employment statuses of individuals over a period of 3 years. Unstable dependent employment comprises those individuals with at least three transitions between employment and unemployment in 3 years. Standard employment refers to those individuals whose employment is stable (i.e. not unstable) and work mainly on a full-time basis within 3 years. Part-time employment refers to those individuals whose employment is stable and work mainly on a part-time basis within 3 years. Temporary contract: a contract is classified as temporary if the termination of the job is determined by objective conditions such as a specific contract end date, completion of an assignment or return of another employee who is temporarily replaced.
Source: Longitudinal EU Statistics on Income and Living Conditions (EU-SILC).
2.1.4. Women are more likely than men to work part-time
Part-time work can be an option for employees who need to reduce their working hours on a permanent basis, though it widely comes at the price of reduced earnings and curtailed career prospects (OECD, 2017[13]; 2018[5]). Almost one quarter of women – often mothers – work part-time in OECD countries, and are more than two times more likely than men to work part-time (Figure 2.4).
Beyond average figures, female part-time employment differs widely among OECD countries (Figure 2.4), ranging from less than 15% in Eastern European countries and Portugal to almost 60% in the Netherlands. Women account for the majority of part-time jobs in all countries. In Austria, Belgium, Luxembourg and Switzerland, about four‑in‑five part-time jobs are performed by women.
There is no definitive conclusion why the share of part-time work differs so widely among OECD countries. The institutional framework may play a role to explain cross-country differences, but there seems to be no clear relationship between the generosity of protection for part-time workers and the incidence of part-time work (OECD, 2010[14]). Childcare policies can also play a role in incentivising women to take up part-time jobs. In particular, several studies find a positive effect of reduced childcare costs on maternal labour supply (Baker, Gruber and Milligan, 2015[15]; Berlinski and Galiani, 2007[16]; Carta and Rizzica, 2018[17])
Part-time employees tend to receive lower hourly wages in all OECD countries (OECD, 2010[14]; 2018[5]). Individual characteristics explain part of the part-time pay penalty, but occupational profile differences are more important in accounting for the wage differentials between part-time and full-time workers (Manning and Petrongolo, 2008[18]). Job insecurity is higher among part-time employees and more so for men than for women. Part-time workers also take part in less training and tend to be less optimistic about their career prospects. Yet, workers, especially women, use part-time work as the primary way to achieve work-life balance (Mas and Pallais, 2019[19]).
Despite higher job insecurity and lower pay, in most countries part-time employees work shorter hours by choice, even though this choice is often dictated by external constraints, such as uneven family responsibilities within couples (OECD, 2018[5]). In 2018, only 5.2% of female employment was involuntary part-time work (Figure 2.4), i.e. working fewer hours than desired because they could not find a full-time job. Evidence from the Netherlands, the country with the highest part-time employment share, shows that women who work part-time display high job satisfaction rates and do not aim to work more hours (Booth and van Ours, 2013[20]). There are exceptions, however. In Italy, Chile, France and Spain, the countries with the greatest share of involuntary part-time work, the majority of part-time women would have liked to work more hours (OECD, 2019[1]).
Figure 2.4. Part-time work is more frequent among women
Incidence and share of female part-time and involuntary part-time employment, 2018, in percentage
Note: Part-time employment as a proportion of total employment. “Part-time” here refers to persons who usually work for less than 30 hours per week in their main job. For Japan and Korea, part-time employment is based on actual rather than usual weekly working hours. Population of persons aged 15 or more, except for Italy, Spain and United Kingdom where the lower age limit is 16. For Ireland and Portugal, involuntary part-time employment refers to 2017. OECD unweighted average for all countries but Korea and Mexico, where data on involuntary part-time employment is missing.
Source: OECD Employment Database http://www.oecd.org/employment/emp/onlineoecdemploymentdatabase.htm.
2.1.5. Poverty rates are higher among workers with non-standard employment
Workers with trajectories of non-standard dependent employment have higher income poverty rates (i.e. live in a household with an annual disposable income of less than 50% of the national median). On average across 24 European OECD countries, the share of workers living in households with income below the poverty line is 10% among those with trajectories of part-time employment and 19% among workers in unstable employment; in contrast, the poverty rate among standard workers is 3% (Figure 2.5).
There is considerable variation between different countries. Among workers with trajectories of part-time employment, the poverty rate is below 5% in Belgium and the Netherlands, and around or above 20% in Greece, Latvia, Portugal and Spain. Further, in the Baltic countries and Sweden, more than one‑in‑four workers with unstable employment trajectories live in a household with income below the poverty threshold. Among standard employees, the poverty rates are lower and more homogenous across countries, ranging from less than 1%, in Finland, Ireland and the Netherlands, to slightly above 5% in Estonia and Luxembourg.
Non-standard dependent employment may be a driving factor increasing the risk of falling into income poverty for several reasons. Earnings potential may be reduced due to lower work intensity, and lower current and future wages.
Non-standard employees may work for fewer hours in part-time and unstable employment.3 Workers with trajectories of non-standard dependent employment may also undergo a penalty in terms of hourly wages. In comparison to standard employees, workers with part-time and temporary jobs experience not only wage penalties in their current earnings (lower hourly wages) but are also less likely to participate in training and be promoted, which impacts future earnings (OECD, 2010[21]; OECD, 2014[10]).4 OECD evidence also suggests that personal characteristics (e.g. age, education and work experience) play a minor role, while job characteristics (e.g. occupation, industry, firm size and contract) explain some of such penalties (OECD, 2010[21]).
Household characteristics can also play an important role in determining the poverty risk of non-standard employees. Evidence on part-time employees shows that they are more likely to be income poor if their job is the main source of household earnings than if they cohabit with standard employees with higher earning (OECD, 2015[22]).
Workers with part-time and unstable employment may also experience penalties in terms of access to social benefits and protections. In many countries, social protection systems in general, and unemployment benefits in particular, have not yet fully adapted to the specific needs and circumstances of workers in non-standard forms of dependent employment. Analyses of out-of-work social benefits in six countries (France, Greece, Hungary, Italy, Spain and the United Kingdom) found that in Spain and Italy unstable employees were less likely to have access to benefits than standard employees (OECD, 2019[1]). This gap was also considerable in the United Kingdom, although not statistically significant. The social protection gap between standard and non-standard employees was larger when assessed in terms of benefit generosity (benefit amount as a proportion of median income), especially in Greece, Italy and Spain (Fernández, Immervoll and Pacifico, forthcoming[23]).
Due to their unusual work trajectories, jobseekers with previous non-standard dependent employment may not receive the type of support that unemployment benefits typically provide to jobseekers with standard employment. Unemployment benefits, in particular, tend to be more difficult to access for non-standard employees than other insurance programmes, such as maternity and sickness benefits (Avlijas, 2019[4]). In fact, low coverage by unemployment benefits – fewer than one‑in‑three jobseekers receive unemployment benefits on average across countries (OECD, 2018[5]) – may be associated with non-standard dependent employment. The condition for access to unemployment benefits in OECD countries are examined in the next section.
Figure 2.5. Poverty rate by type of dependent employment
Share of employees with household disposable income below the poverty line by type of employment trajectory, 2016‑18 (in percentage)
Note: Data is from the longitudinal panel of the EU Statistics on Income and Living Conditions (EU-SILC). To minimise year-to-year variability in small subsamples, the sample pools together rolling panel samples ending with the years 2016, 2017 and 2018. The sample is restricted to employees between 20 and 59 years old. People not in the labour force (e.g. students or pensioners) are excluded. OECD: weighted average of listed countries. Poverty: defined as household disposable income (adjusted for household size) below 50% of the median. Household disposable income includes, from all household members, all gross personal income components (e.g. earnings, social transfers, income from rent, regular inter-household transfers received, and income from capital) minus taxes on wealth, regular inter-household transfers paid and income tax and social insurance contributions. Employment trajectories: unstable dependent employment comprises those individuals with at least 3 transitions between employment and unemployment in 3 years. Standard employment refers to those individuals whose employment is stable (i.e. not unstable) and work mainly on a full-time basis within 3 years. Part-time employment refers to those individuals whose employment is stable and work mainly on a part-time basis within 3 years.
Source: Longitudinal EU-SILC.
2.2. Are unemployment benefit rules adapted to non-standard employment?
Requirements related to minimum time in employment or social contributions are harder to meet for those in unstable or part-time employment. With frequent job changes and job losses, unstable workers tend to have comparatively short employment tenure. Depending on the country, differences between workers with standard and non-standard employment might go beyond a mere pro-rata equivalence of entitlement and result in less favourable entitlement to unemployment insurance such as smaller benefit amount, shorter duration or restricted access. Furthermore, already acquired entitlements may be lost during a change in employment status or job (e.g. following a transition from dependent employment to self-employment when entitlements differ across employment statuses or between jobs if they are tied to a specific employment relationship). To deal with these gaps, several countries have special measures including exemptions for specific contractual arrangements, such as casual employment and seasonal work (OECD, 2019[1]).
In many OECD countries, some part-time work is compatible with the receipt of unemployment benefits (“part-time unemployment benefit”) and may or may not open new entitlements. Some countries offer partial benefits to workers whose working hours have been reduced (“short-time unemployment benefit”) – see e.g. Cahuc (2018[24]). In contributory systems, out-of-work support for unstable workers often includes provisions that allow for contribution periods or unused entitlements to be carried forward to future claim periods.
2.2.1. Main unemployment benefit rules
This section describes the main unemployment benefit rules focusing on their possible specific adaptation – or lack of it – to the special circumstances of workers with part-time or unstable employment. Formally, workers with part-time and unstable employment have equal access to unemployment benefits as workers with standard employment. In practice, however, if eligibility criteria are not adapted to their special circumstances their access and protection level may be considerably different and, in some cases, inadequate. This is particularly the case in contributory systems, which require meeting a minimum amount of time in employment, of hours worked and/or of earnings received.
Most unemployment benefits in OECD countries are contributory and eligibility is conditional on employment requirements (Figure 2.6). Employment requirements of first-tier unemployment benefits vary considerably across OECD countries. On average, the minimum time in employment is about 12 months and ranges from three months, in Iceland and Italy, to 24 months in the Slovak Republic. The reference period is on average about 24 months, and varies between nine months in the Netherlands and six years in Spain. Interestingly, the country with the longest minimum time in employment, the Slovak Republic, permits a longer assessment period for workers with temporary contracts.
Employment requirements also apply to some second-tier unemployment benefits. Usually, both minimum time in employment and reference periods are shorter than first-tier benefits. In France, however, access to the second-tier unemployment benefit requires at least five years of employment in the last ten years, typically targeting the long-term unemployed rather than non-standard employees.
Some countries have restrictions on the minimum amount of contributions or of hours worked, which may hinder access of part-time employees, who are less likely to fulfil those criteria. Poland, for example, requires monthly contributions to be based on earnings that are equal to, or greater than, the national minimum wage. In the United Kingdom, a minimum earnings level is also required for contributions to count towards unemployment insurance entitlements – the minimum earnings level is approximately 15% of average full-time earnings in the United Kingdom (OECD, 2019[1]). In Finland, contributions must be based on at least 18 hours of work per week. In Australia, the Jobseeker Payment provides social protection for jobseekers through means-tested payments that are independent of past employment.
The financing of contributory unemployment benefits usually stems, at least in part, from social security contributions paid by employees and/or employers. In most countries, social security contributions include “unemployment insurance contribution rates” which correspond to the part of contributions earmarked to unemployment protection. Some countries differentiate unemployment insurance rates to create financial incentives for employers and employees to choose more stable employment contracts and discourage collusion. Box 2.2 shows unemployment insurance rules in a subsample of OECD countries.
There is considerable variation across countries in the design of unemployment benefit systems, and hence, the extent to which they support incomes during joblessness and facilitate job search. Figure 2.7 summarises some of the key institutional details of first-tier and second-tier unemployment benefits in all OECD countries, except Colombia. Second-tier unemployment benefits protect jobseekers who are not (or no longer) eligible to the first-tier contributory unemployment benefits.5 Special unemployment programmes such as training allowances, additional “lower-tier” unemployment benefits6, social assistance benefits and other programmes that are not exclusively targeted at jobseekers (e.g. family or sickness benefits) are not considered.
Figure 2.6. Employment requirements in unemployment benefits
Minimum time in employment and reference period, number of months at 1st of January 2020
Note: Minimum time in employment indicates the minimum contribution period to be eligible for receiving first-tier unemployment benefit payments. Reference period is the period to assess whether the respective employment condition is fulfilled. First-tier and second-tier unemployment benefits are identified by the country acronym plus the number 1 and the number 2, respectively. Second tier benefits are only included if claim is conditional on fulfilling an employment condition. In countries without a contributory second-tier unemployment benefits, first-tier benefits are identified only by the country acronym. Countries where first-tier benefits are not conditional on fulfilling an employment condition are not shown.
(1) Minimum earnings requirement of NOK 149 787 in the past 12 months or alternatively NOK 299 574 in the past three years. (2) At least 30 days of employment in the 12 months prior to the start of the unemployment spell. (3) Or 200 days in last two years. (4) Earnings condition also applies. (5) Assuming 40‑hour workweek. (6) six months in any one of the past two years. (7) The first-tier benefit requires claimants to meet a minimum membership period in an unemployment insurance fund (A-kassa) of 12 months. (8) Or 26 weekly contributions in each of previous two years. The claimant must also have made 104 weekly contributions in the whole career. (9) First unemployment benefits claim: 12 months within two years in general, and 6.5 months within one year for people under 25. For all subsequent UB claims, seven months within one year or 12 months within two years. (10) The claimant must have 12 months of contributions since the previous unemployment spell or the last time a withdrawal from the individual account was made; the last three contributions must be continuous and with same employer. (11) The claimant must have held the labour contract for the last 120 days before its termination. (12) The last three monthly contributions before unemployment must be continuous and with same employer.
Source: OECD tax-benefit model and policy database (http://oe.cd/TaxBEN), Avilijas (2019[4]), and information collected and compiled by the Employment, Labour and Social Affairs Directorate of the OECD.
Most unemployment benefits are contributory and not means-tested, especially if they are first-tier benefits. Out of the 47 unemployment benefits, 37 are contributory and ten are non-contributory (Figure 2.7, Panel A). Only in Australia and New Zealand, first-tier unemployment benefits are not contributory and means-tested. Most second-tier unemployment benefits are not contributory and means tested. Only in Austria and Chile, second-tier unemployment benefits are contributory and not means test. Conversely, in Estonia, France, Portugal and Spain, second-tier unemployment benefits are contributory and means tested.
Typically, second-tier unemployment benefits are subject to strict income and asset testing, which may hamper access for jobseekers in households with other income sources, such as earnings from a spouse. Among OECD countries with a second-tier unemployment benefit, only in Austria and Chile the benefit is not means-tested (Figure 2.7, Panel B). In Austria, France and Greece, the second-tier unemployment benefits are available only for people who ran out of the first-tier benefit – thus, they are not available to jobseekers who did not qualify for the primary benefit in the first place.
Box 2.2. Unemployment insurance contribution rates of employees and employers
In some countries, social security contributions for unemployment insurance vary for specific types of employment, including part-time and temporary contracts. Total contribution rates, i.e. both employer and employee contributions, vary from 0.9% of pre-tax wages in Japan to 8.3% for temporary employees in Spain. In most countries with contributory unemployment insurance schemes, the lion’s share of total contributions is borne by the employer. The employer-to-employee contribution ratio ranges from 50:50 in Finland to 100:0 in France1, the Netherlands and Poland. Some countries apply different contribution rates according to occupation (Japan), type of dependent employment (Spain and Netherlands) and wages (Austria and the United Kingdom).
Different contribution rates for temporary and permanent contracts in Spain and the Netherlands may significantly alter employer and employee incentives to favour a certain type of contract. In France, a new experience rating system according to firm-specific job separation rates will modify employer contributions rates from January 2021 onwards.2 Employers will pay a reduced rate if firm separation rates are below the median sectoral separation rates. The contribution floor of the new system is 3% and the ceiling is 5.05%. A recent study suggests, however, that, by increasing labour cost for marginal works, taxing temporary jobs of short duration […] increases the share of open-ended contracts but reduces the mean duration of jobs and decreases job creation, employment and welfare of unemployed workers (Cahuc et al., 2019[25]).
Similarly, discontinuities in the contribution schedule, as in Austria and the United Kingdom, may lead to coordinated behaviour of employers and employees to determine wages or adapt labour market participation. Results from empirical studies exploiting discontinuities in social security contribution ceilings for employers indicate that labour supply and wage responses are small to negligible (Saez, Matsaganis and Tsakloglou, 2012[26]; Saez, Schoefer and Seim, 2019[27]). At the lower end of the wage distribution, however, reduced taxes or social security contributions for employees can significantly alter labour supply (Meyer and Rosenbaum, 2001[28]; Chetty, Friedman and Saez, 2013[29]). Considering this, sharp discontinuities in social security contributions at low earnings levels in Austria and the United Kingdom may affect both the decision of whether to work (extensive margin) and how many hours to work (intensive margin).
1. In France, the contributions paid on wages are not the only source of funding. Unemployment protection is also funded by a general tax of 1.47% on market incomes.
2. France implemented different contribution rates by contract types in 2013 but abolished them in 2017. Since January 2020, a fixed rate of EUR 10 applies on contracts based on “CDD d’usage”– this is a specific flexible temporary contract authorised only in certain sectors of activity. In addition, for certain activity sectors (dockers and entertainment workers), instead of this flat-rate tax, an increased rate of employer contributions to unemployment insurance of 0.05 points is applied to “CDD d’usage” fixed-term contracts shorter than three months.
In some countries, second-tier unemployment benefits facilitate access to non-standard employees through less demanding employment requirements. In Spain, the second-tier unemployment benefit requires a shorter minimum time in employment (six months in general, three months in case of having family responsibilities). In the United Kingdom and Finland, the second-tier unemployment benefits do not depend on past employment.
Generally, the amount of first-tier contributory unemployment benefits fluctuate between a minimum and maximum amount, while second-tier, non-contributory and means-tested unemployment benefits have fixed amounts. Benefit amounts are usually considerably higher in first-tier than second-tier unemployment benefits, but the difference depends on whether the entitlement to the first-tier benefit is closer to the minimum or maximum limit. Panel A in Figure 2.7 indicates the benefit amount for a recipient previously earning a salary equivalent to two‑thirds of the average wage in the country.7 Under these circumstances, in most countries, the amounts of first-tier unemployment benefits exceed those from second-tier benefits by a significant margin (at least 10 percentage points of the average wage).
In some countries, however, the benefit amounts of first- and second-tier unemployment benefits are similar (Figure 2.7, Panel A). In Austria, where access to the second-tier benefit is conditional on exhausting first-tier benefit, the amount of the second-tier benefit is just slightly lower than that of the first-tier benefit, though it is means-tested and other household incomes reduce the entitlement. In Ireland and the United Kingdom, the benefit amounts of first-tier and second-tier benefits are the same. This may help explain why only 14% of the unemployment benefit claimants in the United Kingdom receive the first-tier benefit (Bradshaw and Bennett, 2017[30]; Avlijas, 2019[4]).
In Estonia, Finland, Greece, Portugal, Spain and Sweden, the amount of the flat-rate means-tested second-tier benefit is similar to the minimum amount of the earnings-related first-tier benefit (Figure 2.7, Panel A). In the case of Finland, not only the amounts are the same but also the means test applied to the second-tier benefit has limited impact on benefit amounts as many income sources are fully disregarded (e.g. earnings of the spouse, social assistance and housing allowance). In contrast to other earnings-related social security systems, in Finland, the first-tier benefit does not have an upper ceiling for the benefit level. This feature aims to incentivise high-wage earners to contribute to the system.
The maximum duration of first-tier unemployment benefits tends to be shorter than of second-tier benefits. In most countries, the maximum duration of first-tier unemployment benefits is up to 36 months, with 24 months being the most frequent limit (Figure 2.7, Panel B). In the non-contributory and means-tested first-tier unemployment benefits of Australia and New Zealand there is no maximum duration limit. Similarly, no maximum duration limit applies in Belgium, where first-tier unemployment benefits are contributory and not means tested. In several countries, the duration of second-tier unemployment benefits is unlimited. In most cases, these benefits with unlimited duration are means tested. In Austria, however, the benefit is not means tested on household income, although it is suspended if the claimant earns above the minimum base of social contributions.8 Among second-tier unemployment benefits with limited duration, the maximum duration ranges from nine months in Estonia to 24 months in Portugal.
2.2.2. Detailed unemployment benefit rules
This section describes detailed rules of first- and second-tier unemployment benefits that can affect workers with part-time and unstable employment differently from standard employees. The data is based on a tailored questionnaire, which was submitted to 11 OECD member countries (Australia, Austria, Canada, Finland, France, Japan, Latvia, the Netherlands, Poland, Spain and the United Kingdom). This country selection aims to illustrate unemployment benefits in European and non-European countries, based on contributory and non-contributory systems, using and not using means testing, and with different levels of legal complexity and access conditions. The detailed information on unemployment benefit legislation is summarised in Table 2.2.
In all selected countries, benefit claimants with previous part-time employment are subject to the same access conditions as those who previously worked full-time. The calculation of benefit amounts may differ however, particularly in the case of first-tier benefits. In most countries (Austria, Canada, France, Japan, Latvia, the Netherlands and Spain), the amount of first-tier benefits is calculated pro-rata, i.e. in proportion to the wage in the previous job. In Australia, Poland and the United Kingdom, where the benefit amount is a flat rate, jobseekers previously in part-time employment receive the same amount as those previously working full-time. Finland has a hybrid system combining a fixed component with a supplement that is proportional to earnings. Second-tier benefit amounts are also independent of previous hours worked in Finland, France, and the United Kingdom. In contrast, in Austria and Spain, the amounts of second-tier benefits are computed in proportion to the previous number of hours worked.
Figure 2.7. Main characteristics of unemployment benefits, 2020
Note: Panel A shows the interval between minimum and maximum benefit as a percentage of the average wage and the amount the jobseeker is entitled to when previous-job earnings were exactly two‑thirds of the average wage. Panel B shows maximum duration in months. First-tier and second-tier unemployment benefits are identified by the country acronym plus the number 1 and the number 2, respectively. In countries without second-tier unemployment benefits, first-tier benefits are identified only by the country acronym.
Benefit rules and calculations assume a 40‑year old individual in a single household without children and not eligible to any supplement. For other household types and age groups, see annual OECD tax-benefit country reports. In some countries, benefit minimums may be lower than stated for former part-time employees (not considered).
Unemployment benefits as of 1st January, 2020. Benefit amounts for jobseeker previously earning 2/3 of average wage estimated by combining results from the OECD tax-benefit web calculator for 2018 and information collected and compiled by the Employment, Labour and Social Affairs Directorate of the OECD for 2020. Estimations of the benefit amount at 2/3 of average wage additionally assume that claimants are in the second benefit month and have a long contribution record (22 years).
Average wage: gross annual value for a full-time worker. Wages for 2020 are preliminary wage estimates based on projected average wage data, calculated by the Secretariat. All estimations use annualised benefit amounts.
[1] Benefits may be extended up to a maximum of 48 months if taking part in certain active labour market policies.
[2] Information on the USA reflects the situation of the Michigan unemployment benefit scheme.
[3] Claimants deemed to be difficult to re-employ may receive up to 360 days of unemployment benefits.
[4] The maximum benefit duration is determined by the regional unemployment rate. The estimated maximum duration shown here is based on the national unemployment rate for December 2019.
Source: OECD tax-benefit model and policy database (www.oecd.org/social/benefits-and-wages.htm) and information collected and compiled by the Employment, Labour and Social Affairs Directorate of the OECD.
Many countries use part-time unemployment benefit schemes, which enable claimants to keep part of their unemployment benefits while earning low wages from a job. These schemes aim at people who have lost a full-time and found a part-time job or have lost a secondary job, and are seeking a new job in order to work more hours (Cahuc, 2018[24]). Typically, jobs are restricted to a temporary basis or subject to specific limits regarding working hours and/or earnings.
Except for Latvia and Poland, all analysed countries have schemes of part-time unemployment, although specific conditions apply. Most countries set an implicit earnings limit on part-time unemployment by phasing out the amount of the benefit by reducing it in line with earned wages. Some unemployment benefits operate, however, explicit limits based on earnings, hours or duration. Explicit earning limits are in place in Austria (lower than the “marginal income limit”), second-tier benefit in Japan (80 000 JPY) and the Netherlands (87.5% of the benefit amount). Limits on the number of hours worked are imposed in Finland (80% of full-time), first-tier benefit in Japan (20 hours per week) and Spain (70% of full-time). In France, the second-tier unemployment benefit sets a duration limit, as work is only compatible with the benefit for up to three months.
Some countries encourage unemployment benefit recipients to take up part-time work using financial incentives that allow recipients to “keep” some of the additional income received in wages by not reducing the benefit amount by the same magnitude. First- and second-tier unemployment benefits in Austria and second-tier benefit in France do not reduce the amount of benefits with wages earned, although both apply limits, as seen above. In Australia, Canada, Finland, France (first-tier benefit) and the Netherlands, benefits apply ‘withdrawal rates’, which reduce the magnitude of benefits at a slower pace than wages. Furthermore, in Australia, Finland and the United Kingdom, benefits apply ‘earning disregards’, which ignore part of the earnings amount to be deducted from the benefit. Only Spain and first-tier benefit in Japan do not provide any financial incentive.
Putting together the limits and incentives to part-time unemployment, Table 2.2 presents the maximum possible sum of unemployment benefits and earnings (SUBE), expressed as a proportion of the average wage in the country. SUBE ranges from 9% in the United Kingdom to 70% in Canada. The sum of unemployment benefits and earnings can also reach levels from 40% of the average wage in Austria, Finland, France, Japan, the Netherlands and Spain (first-tier benefit only).
Workers experiencing unstable employment, with repeated transitions between work and unemployment, may not satisfy all unemployment benefit conditions when applying for a subsequent time. To facilitate their access, some countries make use of ‘recharging’ rules, with specific access conditions for jobseekers who are not applying to unemployment benefit for the first time. In Austria, for example, people who received unemployment benefits before are favoured by an alternative employment requirement (28 weeks of employment in the past 12 months) besides the one available for first-time claimants (52 weeks in the past 24 months).
Rules easing repeated benefit claims may produce incentives for workers and employers to adjust their behaviour. Unemployment benefit entitlements readily available after the end of temporary contracts can trigger “carousel effects” (i.e. repeated movements in and out of unemployment) because benefit claims are not restricted to situations where the firm has an objective economic reason for layoffs (OECD, 2002[31]).
Some countries reduce the amount of first-tier unemployment benefits after some period to encourage job search (France, Latvia, the Netherlands and Spain).9 However, such reductions may also incentivise workers with intermittent employment to plan strategically their employment spells in a way to maximise the amount of unemployment benefit payments (Fontaine and Malherbet, 2016[9]; Kyyrä, 2010[7]; Le Barbanchon, 2016[8]).
In some countries, the jobseeker can keep unused entitlements if she finds full-time work before exhausting the benefit duration to which she is entitled. Workers with unstable employment may then access such unused entitlements in a later unemployment spell. The rules for saving unused entitlements diverge significantly across countries and benefit programmes. Unused entitlements to unemployment benefits are completely lost in Poland and Latvia,10 so that workers need to rebuild their benefit rights from scratch whenever they start a new employment spell. In Canada, the Netherlands and the United Kingdom, strict conditions determine whether entitlements are kept for subsequent unemployment spells.11 In Spain, “old” (unused) entitlements are kept, but cannot be accumulated with “new” entitlements. If, due to a recent period in employment, jobseekers are eligible to “new entitlements”, they must choose between the old and the new entitlement, and discard the other. In France’s first-tier unemployment benefit rules, jobseekers can accumulate old and new entitlements, as long they first finish the old one before claiming the new one. Otherwise, jobseekers have the droit d’option, which allows them to start receiving the new entitlement, while discarding the old one. In some countries, such as France and Japan, unused entitlements can be (partly) paid out as an in-work benefit or re-employment allowance/bonus.
About half of the countries shown use benefit waiting periods to incentivise job search during an unemployment spell. However, these provisions may also make support difficult to access for those with unstable employment. Waiting periods typically last for a week and apply to all applicants, but individual circumstances can extend them to one month or more in Australia or (for second-tier benefits) in Spain. In some countries, waiting periods are waived for those with unstable employment, i.e. if the last benefit payment was received a short time before (12 weeks in the United Kingdom and 12 months in France).
Table 2.2. Summary of unemployment benefits rules that impact differently on non-standard employees
Selected countries, 2020
Note: All benefit rules and calculations assume a single household without children. For other household types, see annual OECD tax-benefit country reports.
* SUBE: Sum of unemployment benefits and part-time employment earnings. This indicator is estimated assuming a single individual without children, aged 40 years-old, eligible for the unemployment benefit and, if previous earnings are required, previously earning two‑thirds of the average wage in the country.
[1] From 20 March 2020, Newstart Allowance was replaced by the JobSeeker Payment. JobSeeker Payment has the same payment rates and indexation arrangements as Newstart Allowance https://www.dss.gov.au/about-the-department/benefits-payments/working-age-payments.
[2] Months worked do not count for future benefit eligibility if earnings are below EUR 461.
[3] In case of multiple part-time jobs, only contributions of lost jobs are considered for eligibility and duration;
[4] Part-time work during benefit payment periods do not count for fulfilling renewed employment conditions.
[5] GBP 20 per week in some special cases.
Source: OECD tax-benefit model and policy database and tailored questionnaire to national authorities.
2.3. Unemployment benefits: Income protection and financial work incentives
Do unemployment benefits provide adequate income protection and strong work incentives to jobseekers who were in non-standard dependent employment? How different are indicators of income protection and work incentives for jobseekers with a history of standard employment? What are the circumstances and policies that drive non-standard employees to be treated differently than standard employees? Are there particular systems (e.g. contribution-based or means-tested based) or policy mechanisms that make unemployment benefits better equipped to protect non-standard employees?
In order to address these questions, this section develops new indicators of the impact of unemployment benefits on income protection and financial work incentives for non-standard employees. The indicators are obtained by simulating the unemployment benefit rules on employment scenarios that characterise workers with standard and non-standard dependent employment, using definitions that are comparable across countries. The simulations compute unemployment benefit entitlements for people who worked the same total number of hours and earned the same amount of wages over a period of several years, but through different employment trajectories.
The simulations are similar in spirit to model calculations, such as those based on the OECD tax-benefit model (TaxBEN), that are commonly used to compare benefit replacement rates and work incentives. However, TaxBEN currently does not cover benefit provisions for unstable employment. The simulations focus on the first-tier and second-tier unemployment benefits described in Table 2.2. Lower-tier benefits, such as social assistance, as well as in-work support and tax provisions also shape the income consequences of these different work patterns. Accounting for the full range of tax-benefit policy levers would require an extension of the TaxBEN model to cover unstable employment, which is left to future work. Box 2.3 provides a detailed description of each scenario as well as additional assumptions.
The simulations were carried out for four OECD countries out of the 11 whose unemployment benefits were described in Table 2.2: Australia, France, Latvia and Spain. While the long-term objective would be to include all OECD countries, the simulations for these four countries provides a pilot for future extensions. These specific countries were selected to deliver diversity in terms of geography (European and non‑European countries), benefit system (contributory and non-contributory), legal complexity and access conditions. Despite such diversity, given the small number of countries, the evidence obtained with simulations aims to illustrate and highlight policy issues rather than to be representative of all OECD countries.
2.3.1. Income protection favours some forms of non-standard employment trajectories over others
Unemployment benefits provide varying degrees of income protection depending on the type of employment trajectory. Such differences are observed even when workers have the same personal characteristics, have earned the same amount of wages and have been in and out of work the same number of hours, over a given period.
In Australia, France, Latvia and Spain, workers with a standard employment trajectory receive at least as much income protection as any worker with a non-standard employment trajectory. Figure 2.9 to Figure 2.12 illustrate the unemployment benefit entitlements in each month and for each scenario described in Box 2.3. As a summary for these figures, Figure 2.13 decomposes income protection into two indicators: receipt and level. Receipt measures the number of months that unemployment benefits are received. Level assesses the average amount of unemployment benefit received over the analysed seven‑year period (accounting only the months in which the benefit is effectively paid).
Box 2.3. Simulation scenarios
Results presented in the following sections are drawn from four different simulation scenarios, each corresponding to a distinct employment trajectory over a seven‑year period. In all scenarios, the employment status is the same in the first two years (full-time work) and in the last year (unemployment during the entire year, without any earnings). Employment patterns in years 3 to 6 differ across scenarios.
This relatively long period of analysis allows the assessment of a range of different circumstances. The first three years correspond to a worker who transitions from full-time stable work to unemployment or non-standard employment. This trajectory is akin to that of a displaced worker (OECD, 2018[5]; Farber, 2017[32]), who is struggling to secure a stable job after several years of continuous full-time employment.1 The next three‑year period (years 4 to 6) is illustrative for workers who are caught in a long cycle of non-standard employment. The seventh year of full unemployment permits measuring the entitlements that are left after an extended period of non-standard employment.
In all the simulated scenarios, over the whole period, the person works a total number of hours that is equivalent to 48 full-time months: 24 months during the first two years and 24 months during the following four years. The person is unemployed a total number of hours that is equivalent to 36 full-time months: 24 months between the third and sixth years, and 12 months in the seventh year. In all instances, the person works at an hourly wage that is equivalent to two‑thirds (67%) of the average wage (AW)2.
The simulations assume full take-up of unemployment benefits (de jure entitlements). Entitled jobseekers claim benefits from the first month they are eligible, thus maximising their receipt in the short-term; in some circumstances, jobseekers may receive larger benefit amounts by strategically postponing a claim. In all scenarios, the jobseeker is 45 years old, lives alone and has no children. Fixing these characteristics boosts cross-country comparability, by avoiding specific age- and family related unemployment benefit rules that change from country to country.
The simulated scenarios are illustrated in Figure 2.8. The horizontal axis represents each of the 84 months of the seven years that are assessed. The vertical axis represents the amount of wages and unemployment benefits, expressed as a percentage of the average wage – hence, in the first two years the wage is 67% (two‑thirds) of the average wage. UB_1 represents the amount of the first-tier unemployment benefit and UB_2 represents the amount of the second-tier unemployment benefit.
The characteristics of each simulation scenario – particularly each employment patterns, which is different between years 3 to 6 – are explained as follows:
A. Standard: illustrates a benchmark employment trajectory (further referred to as a “standard employment trajectory”), which alternates full-time dependent employment and full unemployment every 12 months.
B. Unstable 6 months: illustrates an “unstable employment trajectory with changes every semester”, with job changes between full-time dependent employment and full unemployment every 6 months.
C. Unstable 1 month, illustrates an “unstable employment trajectory with changes every month”, with job changes between full-time dependent employment and full unemployment every month.
D. Part-time: illustrates a “part-time employment trajectory”, which consists of recurrent one‑year periods seeking for a full-time job, with the first month in full unemployment and the following 11 months at part-time employment (i.e. working at 54.5% of full-time).
Figure 2.8. Simulation scenarios
1. The US Bureau of Labor Statistics defines displaced workers to be “persons 20 years and over who lost or left jobs because their plant or company closed or moved, there was insufficient work for them to do, or their position or shift was abolished” (Bureau of Labor Statistics, 2020[33]). According to previous OECD work, between 1% and 7% of the workforce is displaced annually, implying a significant probability that a typical worker will experience displacement one or more times during her working life. A considerable number of displaced workers find a suitable new job rapidly, but the majority experience significant losses of income and potentially would benefit from re-employment assistance and income support (OECD, 2018[5]).
2. AW is the annual average wage among full-time employees in the non-agricultural business sector – sectors B to N (inclusive) of the International Standard Industrial Classification of All Economic Activities (ISIC rev. 4) (United Nations Statistical Division, 2008[34]). More details on the calculation of the AW measure for OECD countries are available in the methodological section of the OECD Taxing Wages publication (OECD, 2019[35]).
In Australia and France, in each month that the worker is fully unemployed she receives some unemployment benefit, in all simulated scenarios. In France, workers with unstable employment trajectories (scenarios B and C) receive unemployment benefits for the same number of months (36 months) as those with standard employment (scenario A). This is not the case in Australia, Latvia and Spain.
In Australia, workers with a standard employment trajectory (scenario A) receive unemployment benefit for 38 months, workers who switch between employment and unemployment every six months (scenario B) receive unemployment benefit for 40 months, and those who switch every month (scenario C) receive it for 60 months. The entitlement, means test and payment of unemployment benefit in Australia is assessed every fortnight based on the income from the previous fortnight. Because of that, the benefit payment shifts by a fortnight and may be received in a month when the person is employed (see the spikes in Figure 2.9). This explains why in scenario C the unemployment benefit is received in more months than in scenarios A and B. If the number of payments were measured in fortnights, however, the number of times across scenarios would be the same.
In Latvia, workers with unstable employment trajectories (scenarios B and C) receive unemployment benefits for 6 months or less, while those with standard employment (scenario A) receive it for 24 months. The unemployment benefit in Latvia requires a relatively long minimum time in employment (12 months), while Latvia has the shortest reference period (16 months) out of the four countries (Figure 2.6). Meeting these entitlement conditions is hard for those with unstable employment patterns. Latvia also has no second-tier unemployment benefit as an additional protection to those with shorter contribution records.
In Spain, workers with unstable employment trajectories (scenarios B and C) receive the first-tier unemployment benefit for half the time received by standard workers (8 instead of 16 months). As in Latvia, this is partly because the first-tier benefit requires a relatively long minimum time in employment (12 months). In addition, after six months of employment, jobseekers can claim the second-tier unemployment benefit. After six months of unstable work, workers may prefer to claim the lower-paid second-tier benefit than to wait until they are able to accumulate 12 months of work; months used to claim the second-tier unemployment benefit cannot be used to claim the first-tier benefit (see Table 2.2).
Across the four analysed countries, unemployment benefits are more generous under contributory than non-contributory systems. In Australia, where unemployment protection is non-contributory, the benefit level for workers with standard or unstable employment trajectories ranges from 10% to 15% of AW. Benefit levels are considerably higher in France, Latvia and Spain, where their systems are contributory.
In France, benefit levels are the same for workers with stable and unstable employment trajectories. In Australia and Spain, benefit levels are slightly higher for workers with stable employment trajectories, because they are less subject to reductions due to waiting periods12 or are often eligible for new benefits with higher amounts. Conversely, in Latvia, benefit levels are higher for unstable workers, because of shorter benefit durations that are less affected by the reduction in the benefit rate as the unemployment spell lengthens (Table 2.2).13
Workers with employment trajectories that include periods combining unemployment and part-time work within a month (scenario D) generally receive unemployment benefits for fewer months than workers with standard employment (scenario A). In Latvia, access is restricted because the unemployment benefit is incompatible with any type of employment (see Table 2.2). In Australia, Spain and France, the unemployment benefit is compatible with some work, but in some cases the eligibility depends on the wage received while in employment (in the simulations presented here the wage is 33.3% AW). In Australia, the unemployment benefit is completely depleted by the means test if earnings are above an amount equivalent to 23% AW (see Table 2.2). In Spain, unemployment benefits are reduced in proportion to the number of hours worked. In addition, the second-tier benefit also applies a means test that makes benefit receipt incompatible with earnings higher than an amount equivalent to 29% AW (see Table 2.2). In France, the first-tier benefit is reduced with a 70% withdrawal rate; for the second-tier benefit, work and benefits are compatible for up to three months, after which the benefit is interrupted if the professional activity continues.
Figure 2.9. Unemployment benefits in Australia
Simulations under different employment scenarios
Note: AW: average wage. UB_1: entitled amount of first-tier unemployment benefit. UB_2: amount of second-tier unemployment benefit. The horizontal axis represents each of the 84 months of the seven years that are assessed. The vertical axis represents the amount of wages and unemployment benefits, expressed as a percentage of the AW. All cases consider two initial years of full-time employment paid at 67% of AW. Different cases: A. Alternating periods of full-time dependent employment and full unemployment every 12 months. B. Unstable 6 months and C. Unstable 1 month: situations with job changes between full-time dependent employment and full unemployment every 6 months and every other month, respectively. D. Part-time: “part-time employment trajectories”, which consists of recurrent one‑year spells seeking for a full-time job, with the first month in full unemployment and the following 11 months at partial employment (working at 54.5% of full-time hours in working periods). See Box 2.3 for more details about assumptions and simulation scenarios.
Source: Secretariat calculations based on unemployment benefit legislation extracted from OECD tax-benefit model and tailored questionnaire to national authorities.
Figure 2.10. Unemployment benefits in France
Simulations under different employment scenarios
Note: AW: average wage. UB_1: entitled amount of first-tier unemployment benefit. UB_2: amount of second-tier unemployment benefit. The horizontal axis represents each of the 84 months of the seven years that are assessed. The vertical axis represents the amount of wages and unemployment benefits, expressed as a percentage of the AW. All cases consider two initial years of full-time employment paid at 67% of AW. Different cases: A. Alternating periods of full-time dependent employment and full unemployment every 12 months. B. Unstable 6 months and C. Unstable 1 month: situations with job changes between full-time dependent employment and full unemployment every 6 months and every other month, respectively. D. Part-time: “part-time employment trajectory”, which consists of recurrent one‑year spells seeking for a full-time job, with the first month in full unemployment and the following 11 months at partial employment (working at 54.5% of full-time hours in working periods). See Box 2.3 for more details about assumptions and simulation scenarios.
Source: Secretariat calculations based on unemployment benefit legislation extracted from OECD tax-benefit model and tailored questionnaire to national authorities.
Figure 2.11. Unemployment benefits in Latvia
Simulations under different employment scenarios
Note: AW: average wage. UB_1: entitled amount of first-tier unemployment benefit. UB_2: amount of second-tier unemployment benefit. The horizontal axis represents each of the 84 months of the seven years that are assessed. The vertical axis represents the amount of wages and unemployment benefits, expressed as a percentage of the AW. All cases consider two initial years of full-time employment paid at 67% of AW. Different cases: A. Alternating periods of full-time dependent employment and full unemployment every 12 months. B. Unstable 6 months and C. Unstable 1 month: situations with job changes between full-time dependent employment and full unemployment every 6 months and every other month, respectively. D. Part-time: “part-time employment trajectory”, which consists of recurrent one‑year spells seeking for a full-time job, with the first month in full unemployment and the following 11 months at partial employment (working at 54.5% of full-time hours in working periods). See Box 2.3 for more details about assumptions and simulation scenarios.
Source: Secretariat calculations based on unemployment benefit legislation extracted from OECD tax-benefit model and tailored questionnaire to national authorities.
Figure 2.12. Unemployment benefits in Spain
Simulations under different employment scenarios
Note: AW: average wage. UB_1: entitled amount of first-tier unemployment benefit. UB_2: amount of second-tier unemployment benefit. The horizontal axis represents each of the 84 months of the seven years that are assessed. The vertical axis represents the amount of wages and unemployment benefits, expressed as a percentage of the AW. All cases consider two initial years of full-time employment paid at 67% of AW. Different cases: A. Alternating periods of full-time dependent employment and full unemployment every 12 months. B. Unstable 6 months and C. Unstable 1 month: situations with job changes between full-time dependent employment and full unemployment every 6 months and every other month, respectively. D. Part-time: “part-time employment trajectory”, which consists of recurrent one‑year spells seeking for a full-time job, with the first month in full unemployment and the following 11 months at partial employment (working at 54.5% of full-time hours in working periods). See Box 2.3 for more details about assumptions and simulation scenarios.
Source: Secretariat calculations based on unemployment benefit legislation extracted from OECD tax-benefit model and tailored questionnaire to national authorities.
Unemployment benefit protection is also less generous for workers with employment trajectories that include periods combining unemployment and part-time work (scenario D) than for a standard worker (scenario A). This is because entitlements are reduced with the wage earned from concomitant work. Usually, unemployment benefit entitlements are the same independently of how workers distributed their working hours within a month (e.g. worked part-time every day or full-time some days of the month). In France, however, the unemployment benefit differs depending on the distribution of working hours within the month (see Box 2.4).
Figure 2.13. Non-standard dependent workers receive fewer unemployment benefits
Note: AW: average wage. UB_1: entitled amount of first-tier unemployment benefit. UB_2: amount of second-tier unemployment benefit. The vertical axis represents, in the first panel, the number of months that unemployment benefits are received, and in the second panel, the amount of unemployment benefits, expressed as a percentage of the AW. All cases consider two initial years of full-time employment paid at 67% of AW. Different cases: A. Alternating periods of full-time dependent employment and full unemployment every 12 months. B. Unstable 6 months and C. Unstable 1 month: situations with job changes between full-time dependent employment and full unemployment every 6 months and every other month, respectively. D. Part-time: “part-time employment trajectory”, which consists of recurrent one‑year spells seeking for a full-time job, with the first month in full unemployment and the following 11 months at partial employment (working at 54.5% of full-time hours in working periods). See Box 2.3 for more details about assumptions and simulation scenarios.
Source: Secretariat calculations based on unemployment benefit legislation extracted from OECD tax-benefit model and tailored questionnaire to national authorities.
Box 2.4. The 2020 reform of the “reference wage” in France
Until 2020, France is the only analysed country where the amounts of unemployment benefits differed depending on the distribution of working hours within a calendar month. Despite working the same number of hours each month, a person working full-time some days per month would receive a larger amount of unemployment benefit than if working part-time for the entire month (Figure 2.14).
The cause of this difference was the formula used to compute the “reference wage” of the first-tier unemployment benefit (Allocation d’aide au retour à l’emploi). The reference wage is an indicator that summarises the wage that the person earned before becoming unemployed and to be replaced, to some extent, by the unemployment benefit. In most unemployment insurance systems, the reference wage is based on the average of previous monthly wages, whether the person worked every day over the reference period or not. In France, however, the reference wage is calculated based on daily wages on working days (i.e. days not worked do not count in the calculation). Because of that, the reference wage of a person working part-time for a whole month (scenario ‘Part-time’) is lower than the reference wage of a person working full-time for half of the month (scenario ‘Part-month’).
A reform of the French first-tier unemployment benefit, originally planned for April 2020 and later postponed to September 2020, eliminates this feature. The new reference wage will be calculated using monthly wages over a 12‑month period, irrespective of the number of months and days worked.
Figure 2.14. The distribution of working hours affects unemployment benefit entitlements in France
Note: AW: average wage. UB_1: entitled amount of first-tier unemployment benefit. UB_2: amount of second-tier unemployment benefit. The horizontal axis represents each of the 84 months of the seven years that are assessed. The vertical axis represents the amount of wages and unemployment benefits, expressed as a percentage of the AW. All cases consider two initial years of full-time employment paid at 67% of AW. Different cases: Part-time and Part-month, which both consist of recurrent one‑year spells seeking for a full-time job, with the first month in full unemployment and the following 11 months in partial employment (working at 54.5% of full-time hours in working periods). The difference between the Part-time and Part-month scenarios regards the distribution of working hours within the month. In Part-time, the person works every day for 54.5% of the full working time. In Part-month, the person works on a full-time basis 54.5% of the days in the month.
Source: Secretariat calculations based on unemployment benefit legislation extracted from OECD tax-benefit model and tailored questionnaire to national authorities.
2.3.2. Financial work incentives depend on current and previous employment trajectories
Unemployment benefits aim to provide adequate income protection to jobseekers and their families, while also maintaining work incentives. Improving work incentives and facilitating the return to self-sufficiency is important because the risk of long-term poverty is much higher for jobless individuals on benefits than for fully employed people. Moreover, the cost of safety nets to budget-constrained governments reinforces the need to induce people – who can – to get back into work (OECD, 2005[36]).
When the unemployment benefit amount received by jobseekers is high relative to the wage they can obtain from work, jobseekers may find themselves in an “unemployment trap” – i.e. discouraged from searching for a new job (OECD, 2005[36]). In the literature (OECD, 2007[37]; OECD, 2020[38]), the financial incentive of a jobseeker to pick up a new job is typically measured through replacement rates (RR). This sub-section uses a variation of the indicator of replacement rates (adjusted replacement rates, ARR), which is adjusted to account for the specificities of the simulated scenarios (Box 2.5).
Figure 2.16 presents the ARR for the final year of the 16 scenarios depicted in Figure 2.15, in Australia, France, Latvia and Spain. The ARR is an indicator of financial work disincentives – the higher the ARR, the lower the financial incentives to work more. The ARR expresses the amount of total income (wages and benefits) that a worker would get relative to the full-time wage (Box 2.5).
Since workers with non-standard employment earn wages that are equivalent to 50% of a full-time wage, their ARR is at least 50%. ARR for workers in full unemployment provide a benchmark to assess the incentives for a worker not to work at all.
Work incentives vary considerably across countries, current and previous employment trajectories. On average, the ARR is higher in France, followed by Australia, Spain and Latvia. Across current non-standard employment trajectories, the ARR is higher among workers with unstable employment. Across previous employment trajectories, ARR are unsurprisingly higher among workers with previous standard employment, except for Australia.
In Australia, a worker in part-time employment would get an income equivalent to 53% of a full-time wage. Hence, her income would increase by 47% of a full-time wage if she worked full time. If she did not work at all, her income would decrease by 18% of a full-time wage, as the ARR in full unemployment is 35%. Conversely, the ARR for a worker with current unstable employment would be 68%, full employment would increase income by 32% of a full-time wage, and full unemployment would reduce income by 33% of a full‑time wage.
In France, workers with unstable employment have lower incentives to take up full employment. A worker in unstable employment would get a benefit amount equivalent to 78% of a full-time wage but working only half of the time. If she did not work at all, her income would decrease by 21% of a full-time wage, as the ARR in full unemployment is 57%. In contrast, the ARR for a worker in part-time employment is 64%. If she did not work at all, her income would decrease by only 7% of a full-time wage. Previous unstable employment trajectories produce the same work incentives as previous stable employment. For previous part-time employment, however, adjusted replacement rates are lower – 18 percentage points lower for full-time unemployed and around 9 percentage points lower for workers in current non-standard employment.
Box 2.5. Measuring work incentives across employment trajectories
Adjusted Replacement Rates (ARR)
The financial incentives of a jobseeker to pick up a new job are typically measured through a replacement rate (RR). A replacement rate is the ratio between out-of-work and in-work income (OECD, 2007[37]; OECD, 2020[38]).
where yow denotes income received while out of work and yiw denotes income received while in work.
Usually, RR is calculated to analyse the effects of a transition between employment and unemployment. In the case of a transition from employment to unemployment, yiw would denote the income before the transition and yow would denote the income after the transition. Generally, income is assessed based on the pay period (e.g. week, fortnight or month).
The adjusted replacement rate (ARR) used in this sub-section uses a variation of the definition above. Instead of analysing transitions, it compares the effects of counterfactual scenarios. Two scenarios are compared, full employment (which would be similar to iw) and not full employment (which would be similar to ow). “Full employment” refers to full-time full-year employment. A person “not in full employment” works reduces the number of hours or days, or does not work at all (i.e. fully unemployed). Finally, incomes are assessed over a year instead of the pay period. So, ARR could be express as:
where Y denotes annual income, YnFE denotes income received if not in full employment, and YFE denotes income received if in full employment.
Current and previous employment trajectories
In the absence of taxes or other social benefits, the amount of YFE depends exclusively on wages, while YnFE depends on the amount of wages and unemployment benefits. The amounts of unemployment benefits depend on current and on previous employment circumstances.
Figure 2.15 presents 16 scenarios that combine different current and previous employment trajectories. Each scenario includes seven years. The first six years correspond to the “previous trajectory”. The seventh year corresponds to the “current trajectory”, as this is the year in which ARR is computed in Figure 2.16.
The employment patterns in the first six years are the same as presented in Box 2.1: “A. Standard”, “B. Unstable 6 months”, “C. Unstable 1 month” and “D. Part-time”. The seventh year repeats the employment pattern of the third year of each of the four scenarios: a. full unemployment, b. 6 months in unemployment followed by 6 months in employment, c. alternating between unemployment and employment, and d. one month unemployed followed by 11 months working at 54.4% of the time.
While this framework has the advantage of accounting for changes across time and counterfactual scenarios, it also has some limitations:
1) Taxes and other transfers are not taken into account. The ARR is computed as a “gross rate”, rather than a “net rate”, as often done in work incentive analysis (OECD, 2007[37]; OECD, 2020[38]).
2) Work incentives are measured in one specific period (the seventh year of each scenario). Therefore, the circumstances analysed are specific and cannot be held as representative of what would happen at different points in the trajectory.
3) An extremely low ARR may indicate the absence of unemployment protection rather than adequate work incentives. In these cases, constraints are more likely to be on labour demand.
Figure 2.15. Scenarios for computing adjusted replacement rates
Figure 2.16. Work incentives depending on current and previous employment trajectories
Adjusted replacement rates, based on current and previous employment trajectories (in percentage)
Note: The vertical axis represents adjusted replacement rates (ARR), which expresses the amount of total income (wages and benefits) that a worker would get relative to the full-time wage. ARR is decomposed in three parts: “wage”, representing worker’s wage relative to the full-time wage; “UB_1”, representing the entitled amount of first-tier unemployment benefit relative to the full-time wage, and “UB_2”, representing the amount of second-tier unemployment benefit relative to the full-time wage.All scenarios combine different current and previous employment trajectories. Each scenario includes seven years. The first six years correspond to the “previous trajectory”. The seventh year corresponds to the “current trajectory”, as this is the year in which ARR is computed. Previous trajectories consider two initial years of full-time employment paid at 67% of AW, followed four years with one of the following employment patterns: A. Alternating periods of full-time dependent employment and full unemployment every 12 months. B. Unstable 6 months and C. Unstable 1 month: situations with job changes between full-time dependent employment and full unemployment every 6 months and every other month, respectively. D. Part-time: “part-time employment trajectory”, which consists of recurrent one‑year spells seeking for a full-time job, with the first month in full unemployment and the following 11 months at partial employment (working at 54.5% of full-time hours in working periods). The seventh year repeats the employment pattern of the third year of each of the four scenarios: a. full unemployment, b. 6 months in unemployment followed by 6 months in employment, c. alternating between unemployment and employment, and d. one month unemployed followed by 11 months working at 54.4% of the time. See Box 2.5 for more details about assumptions and simulation scenarios.
Source: Secretariat calculations based on unemployment benefit legislation extracted from OECD tax-benefit model and tailored questionnaire to national authorities.
In Latvia, outcomes are to a large extent conditioned by the previous employment trajectory. In the scenarios considered here, workers with previous unstable employment trajectories would not be entitled to any unemployment benefit, because minimum contribution conditions requires 12 contributions in the last 16 months (Figure 2.6). Independently of previous employment trajectories, the worker would also not receive unemployment benefit if working on a part-time basis, as the Latvian benefit is incompatible with any work (in scenario D, the small part of ARR related to benefits is due to the month in full unemployment). Workers with current unstable and previous stable employment would receive benefits with sizeable replacement rates. A worker switching every month between unemployment and employment would get an amount equivalent to 79% of a full-time wage. If she did not work at all, her income would decrease by 52% of a full-time wage, as the ARR in full unemployment is 27%. The adjusted replacement rates are slightly lower for workers current switching every six months between unemployment and employment, and for workers who previously were in part-time employment.
In Spain, previous employment trajectories also play an important role in determining the amount of future unemployment benefits. Workers without previous standard employment trajectories would not be entitled to first-tier unemployment benefit, because minimum contribution conditions and means testing rules (as discussed in Section 2.3.1). They would be entitled, however, to a few months of the second-tier benefit. Workers with current and previous part-time employment trajectories would not be entitled to any unemployment benefit. Conversely, a worker with current unstable and previous stable employment would receive benefits with sizeable replacement rates (73%). If she did not work at all, her income would decrease by 37% of a full-time wage, as the ARR in full unemployment is 36%. The adjusted replacement rate is lower for a worker with current part-time and previous stable employment (64%).
2.4. Policy issues: striking the right balance
This section discusses some key features of unemployment benefit systems and considers potential policy options to improve their reliability and effectiveness to support workers with trajectories of non-standard employment. Following the evidence provided in previous sections, the discussion addresses benefit neutrality, work incentives, activation and protection during an economic downturn. Table 2.3 recaps the arguments below with a summary of unemployment benefit instruments in terms of their objectives, trade-offs and effects on workers with non-standard dependent employment trajectories, and possible alternatives of adjustments in order to strike the right balance between conflicting objectives.
2.4.1. Benefit neutrality
In general, for the countries reviewed in this chapter, unemployment benefits are not neutral with respect to the distribution of time between dependent employment and unemployment before the benefit is claimed. Evidence gathered in section 2.3 reveals considerable differences concerning the extent of protection that unemployment benefits provide to workers who, over the medium run, have worked the same amount of time.
Strict employment requirements tend to penalise workers with unstable employment trajectories, in particular when the minimum time in employment is relatively long and the reference period is short. Unstable employment prolongs the period in which workers need to work in order to fulfil the minimum requirement, thus hindering their ability to claim the benefit. While employment requirements are important features of unemployment benefits – they provide work incentives, prevent benefit abuse and protect the financial sustainability of the system – some adjustments can contribute to reduce such a penalty on instability. The difference between the length of the reference period and the minimum time in employment can be designed to be sufficiently long in order to account for unstable employment or specifically extended for workers more prone to job instability, as in Austria (young workers) and the Slovak Republic (temporary contracts).
The reduction of benefit amounts with wages from part-time unemployment penalises jobseekers able to carry out some casual part-time work while searching for an adequate full-time job. In some countries, the penalty is absolute as unemployment benefits are incompatible with any work. Reducing the amount of benefits in line with wages is justified by the need to smooth income and avoid the misuse of unemployment benefits as a sort of wage subsidy. On the other hand, besides neutrality, part-time unemployment may produce desirable outcomes such as improving jobseekers’ employability through a “stepping-stone effect” (Kyyrä, Arranz and García-Serrano, 2019[39]). Financial incentives, such as earnings disregards and lower withdrawal rates, encourage unemployment benefit recipients to take up part-time work by allowing them to keep some of the additional income received in wages, and can be effective instruments for balancing objectives. Yet, part-time unemployment benefits may also produce “lock-in effects”, by reducing the exit rate from part-time to full-time employment. Thus, like in other activation policies, it is important to use “mixed strategies” that profile and monitor the effect of causal part-time jobs on the long-term employability of jobseekers (OECD, 2018[40]).
2.4.2. Work incentives
The potential adverse impact of unemployment insurance systems on work incentives is a main concern in many countries, particularly for those who have been out of work for a long time. Furthermore, ill-designed unemployment benefits may incentivise unstable employment.
Soft access conditions may generate a carousel of shifts between periods in short-tenure jobs and periods in unemployment while receiving generous benefit amounts. Waiting periods, moderately long requirements of minimum time in employment and reductions in case of frequent reclaiming can contribute to preventing such distortion. Differentiated unemployment insurance contribution rates may also be used to create financial incentives for employers and employees to choose more stable employment contracts and discourage collusion (OECD, 2004[41]; OECD, 2019[1]).
Keeping unused benefits for future use may encourage frequent periods of unstable employment (as discussed above) and incentivise over-extending the duration of unemployment benefits, thus promoting long-term unemployment. Limits on the possibility of keeping and accumulating previous and new entitlements can regulate the trade-off between work incentives and acquired rights.
The coordination between in-work and out-of-work support can produce important back-to-work incentives. One option could be to extend to full-time workers with low earnings the possibility to cumulate unemployment benefits and income from work as is the case for part-time unemployment (Hijzen and Salvatori, 2020[42]). Another possibility would be introducing in-work benefits, like the Earned Income Tax Credit in the United States or the Prime d’activité in France (Carcillo et al., 2019[43]). Alternatively, in-work and out-of-work benefits could be integrated in a single comprehensive strategy (e.g. the United Kingdom’s Universal Credit).
Table 2.3. Striking the right balance by adjusting unemployment benefit’s tools
Unemployment benefit’s tools, objectives, trade-offs and adjustments
2.4.3. Activation
Unemployment benefit’s activation requirements also play a role in strengthening incentives to look for, prepare for, and accept employment. Evidence suggests that unemployment benefit programmes featuring job-search monitoring and sanctions yield positive employment effects, while overly demanding eligibility criteria can exclude some intended recipients (Immervoll and Knotz, 2018[44]). Exceedingly strict sanctions and excessive reliance on job-search incentives could be counterproductive and increase the unemployment exit rate into unstable jobs. Profiling plays a crucial role. Workers in need of more intensive activation services (such as training) should be identified and supported (OECD, 2018[40]).
According to the 2018 OECD Jobs Strategy, unemployment benefits can play a pivotal role in the success of activation strategies through the rigorous enforcement of a “mutual obligations framework” (OECD, 2018[40]). Within a mutual obligations framework, governments have the duty to provide jobseekers with benefits and effective services to enable them to find work and, in turn, beneficiaries have to take active steps to find work or improve their employability. The threat of potential sanctions in terms of benefit withdrawal significantly increases the financial incentive for seeking and taking up gainful employment as well as seriously participating in active programmes. Thus, unemployment benefit recipients are referred to employment services.
Unemployment benefits are key to connect jobless people to active labour market programmes, by referring them to employment services, which provide job-search assistance or interventions to increase their employability. In the absence of unemployment (or social assistance) benefits, it is often difficult to reach out to those facing multiple barriers to employment, who risk being left behind. Similarly, in the absence of effective active labour market policies, there is a risk that unemployment benefits reduce work incentives and deepen labour market exclusion. Passive and active policies should therefore be conceived together rather than in isolation (OECD, 2018[40]).
Increasing the access of workers in non-standard dependent employment to unemployment insurance is, therefore, a promising avenue for promoting labour market security and inclusiveness, provided it is carried out together with the rigorous enforcement of a “mutual-obligations” framework to preserve work incentives.
2.4.4. Economic shock protection
Unemployment benefits play an important role as key instruments of social protection to provide income support and employment opportunities in all phases of the economic cycle. This role is stronger during and after an economic shock. By helping individuals and families cope with the consequences of economic shocks, unemployment benefits contribute to prevent temporary economic problems from turning into long‑term disadvantage for workers and a social crisis for society (OECD, 2014[45]).
Unemployment benefits need to be sufficiently responsive to changes in circumstances, following an economic shock. In countries where access to, and duration of, unemployment benefits is restricted, there may be a case for, temporarily, loosening employment and other requirements and extending the maximum duration of the benefit (e.g. in the Canada and the United States). In a context of fewer job opportunities, the emphasis of active labour market policies may temporarily shift from work-first to train‑first, in particular for hard-to-place jobseekers (OECD, 2018[40]).
Complementing the overall discussion in Chapter 1 on the challenges coming from the COVID‑19 crisis, Box 2.6 discusses the measures introduced by countries that have revised unemployment benefit elements in order to extend unemployment protection of non-standard workers during the crisis.
2.5. Concluding remarks
This chapter provides an in-depth review of the key policy mechanisms that affect how unemployment benefits strike a balance between income security and financial work incentives for those with non‑standard dependent employment trajectories. It presents new evidence on the scale and development of unstable and part-time dependent employment and the characteristics of workers with such trajectories. In particular, it shows that workers who experienced non-standard dependent employment trajectories make up for a considerable part of dependent employment and the majority of the unemployed; that their share of employees has increased, particularly among young workers; and that their poverty rates are several times higher than those of workers in standard employment.
The chapter also takes stock of and analysed the legal unemployment benefit provisions that apply to non‑standard employees, and how these might give rise to an uneven treatment of standard and non‑standard employees. Unemployment benefit rules can result in considerable differences between workers with standard and non-standard dependent employment trajectories in terms of the duration of benefit entitlements and the level of benefits paid. Some requirements are harder to meet for those in unstable or part-time employment.
Box 2.6. COVID‑19 crisis: Non-standard dependent employment and extraordinary unemployment benefit measures
The coronavirus crisis has hit hard, with unprecedented job losses around the world. Workers in non-standard dependent employment are particularly vulnerable against the risks of job and income loss. Unemployment benefits and related income support are crucial for cushioning these losses. However, as discussed in the 2019 edition of the OECD Employment Outlook (OECD, 2019[1]) and in this chapter, workers in non-standard dependent employment are less likely to receive adequate unemployment support.
Many OECD countries have taken rapid action by adjusting their unemployment benefit systems to the new labour market circumstances. In line with the OECD Jobs Strategy (OECD, 2018[40]), most measures adjusted existing benefits by loosening requirements in order to facilitate access and revised entitlements to extend income support and duration. Some countries have introduced new benefits. Table 2.4 summarises the extraordinary unemployment benefit measures, with potential impact on non-standard dependent employment, which were introduced during the COVID‑19 crisis between March and May 2020. Chapter 1 provides a wider analysis of employment- and social-policy responses to the COVID‑19 crisis.
Facilitating access: six OECD countries facilitated access to unemployment benefits. France, Israel, Spain and Sweden reduced or entirely waived minimum employment requirements to workers. In Norway, the minimum previous income to be eligibility to unemployment benefits was reduced. The length of the reference period to assess employment requirements was extended in Switzerland (doubled) and in France (by disregarding the confinement period for calculating the entitlements of new claimants). All these measures are, in particular, expected to facilitate access for workers with unstable trajectories, who often find it difficult to meet minimum employment requirements.
Extend benefit duration: 13 OECD countries have extended the maximum possible duration of unemployment benefit payments. Some countries triggered automatic extensions. Other countries extended the expiring benefit claims until the end of the health crisis (Luxembourg, Portugal and Spain) or for a specific period (by up to one month in the Slovak Republic, two months in Greece, three months in Germany, Norway until the end of June, and four months in Switzerland). Denmark and France froze benefit entitlements for the duration of the health crisis. In the United States, the Federal Government extended the maximum unemployment benefit duration to nine months. Since, in some countries, non-standard employees are entitled to fewer months of unemployment benefits (Figure 2.13), these measures are likely to be particularly important for them.
Raising unemployment benefit generosity: seven OECD countries temporarily increased benefit levels. Australia introduced, for six months, a “Coronavirus Supplement” of AUD 550 per fortnight to recipients of income support payments, including the unemployment benefit “Jobseeker Payment”. Austria increased the benefit amount of the second-tier unemployment benefit (Notstandshilfe) to the same level as the first-tier benefit (Arbeitslosengeld). Sweden raised, until the end of 2020, the unemployment benefit floor (by about 30%) and ceiling (by about 40%) for 100 days. New Zealand increased the amount of “Jobseeker Support” by NZD 25 per week. Norway increased the unemployment benefit from 62.4% to 80% of the calculation basis up to an income of NOK 24 966 per month. The United Kingdom raised the allowance of the second-tier unemployment benefit (Universal Credit) by GBP 20 per week. The United States raised the unemployment benefits by USD 600 per week for all recipients for a maximum period of four months, for up to four months. While all these measures are not specifically targeted to non-standard employees, to the extent that those with “naturally” unstable jobs are likely to bear the brunt of the employment adjustment, they are among the categories potentially more positively affected by these measures. Moreover, non-standard employees are likely to benefit the most from the increase in generosity of second-tier benefits to the extent that the risk of poverty is greater for non-standard employees.
New unemployment benefits: six OECD countries introduced new temporary non-contributory benefits, which have clearly a greater impact for people with unstable trajectories. In Canada, the “Canada Emergency Response Benefit” provides CAD 500 a week for up to 16 weeks, to people who have involuntarily lost their job during the COVID‑19 crisis, receive the first-tier unemployment benefit (Employment Insurance) or recently exhausted its claim. In Ireland, the new “COVID‑19 Pandemic Unemployment Payment” supports individuals who lost their job during the COVID‑19 crisis. Claimants receive EUR 350 per week until at least 8 June. The payment is fully compatible with other unemployment benefits. Latvia introduced a new temporary unemployment assistance benefit for unemployment insurance benefit recipients whose entitlements expire. Benefits are payable for up to four months, at the level equal to the unemployment insurance benefits paid during the 8th and 9th month (max. EUR 180 per month). New Zealand created the “COVID-19 Income Relief Payment” for people who lost their jobs from 1 March 2020 to 30 October 2020 due to COVID-19. The payment amount depends on previous hours worked, is means-tested against partner’s earnings, is incompatible with receiving other main benefit (but people are allowed to choose) and is exempt from taxes. Slovenia introduced a temporary unemployment benefit (Začasno denarno nadomestilo plače zaradi izgube zaposlitve v obdobju epidemije) for those who lost their jobs in the crisis. Claimants receive EUR 513.64 per month for the duration of the temporary restrictions; the benefit is not compatible with the first-tier unemployment benefit.
Table 2.4. COVID‑19 extraordinary unemployment benefit measures with potential impact on non-standard employees
Note: * Information for the United States refers to the federal level.
[1] To recipients of income support payments, including the unemployment benefit “Jobseeker Payment”. The supplement will be paid for 6 months.
[2] For newly unemployed and people who receive Employment Insurance or recently exhausted their claim. CERB provides USD 500 a week for up to 16 weeks. Newly unemployed must have involuntarily lost their job in the COVID‑19 crisis. Claimants need to have earned at least CAD 5 000 in 2019 and may not have earned more than CAD 1 000 for the four‑week period of the benefit claim.
[3] Maximum duration extension triggered as the national unemployment rate is 1 percentage point above its 4‑year mean.
[4] Claimants receive EUR 350 per week until at least June, 08. Fully compatible with 1st and 2nd tier benefits.
[5] Claimants who only fulfil the shortened employment condition receive half the number of benefit days.
[6] The Income Relief Payment is paid for up to 12 weeks. The payment is NZD 490 per week if the person was previously working 30 hours or more a week, and NZD 250 per week if she was previously working 15 hours to 29 hours a week. The payment is means-tested against partner’s earnings. A person cannot get the benefit if the partner earns NZD 2 000 or more per week (before tax). It is incompatible with receiving other main benefit (but people are allowed to switch to it if they meet the criteria for who can get it). The payment is tax exempt.
[7] Minimum income reduced from NOK 149 787 to 74 894 over the past 12 months, or from NOK 299 574 to 224 680 over the past 36 months.
[8] Začasno denarno nadomestilo plače zaradi izgube zaposlitve v obdobju epidemije is available for those who lost their jobs in the crisis for economic motives or due to the end of a temporary contract. Claimants receive EUR 513.64 per month for the duration of the temporary restrictions. The benefit is not compatible with the first-tier unemployment benefit.
[9] Shortening of the required membership period in the unemployment insurance fund from 12 to 3 months (1st tier); reduction of the employment requirement from at least 80 hours per months in last 6 months or 480 hours in last 6 months and at least 50 hours per month to 60 hours per months in last 6 months or 420 hours in last 6 months and at least 40 hours per month (1st and 2nd tier).
[10] minimum benefit amount from SEK 365 to 510 per day and maximum benefit amount from SEK 910 to 1 200 per day for the first 100 days (1st tier). All rules until end of 2020.
Source: Chapter 1 and OECD (2020[46]), “Supporting people and companies to deal with the COVID‑19 virus: Options for an immediate employment and social-policy response”, ELS Policy Brief on the Policy Response to the COVID‑19 Crisis, OECD, Paris, http://oe.cd/covid19briefsocial.
Simulations of some policy-relevant employment trajectories for four OECD countries (Australia, France, Latvia and Spain) show that unemployment benefits provide different levels of support depending on the employment trajectory. Such differences are evident even when workers have the same personal characteristics, have earned the same amount of wages and have been in and out of work the same number of hours over a given period. In other words, the systems are not neutral in terms of social protection and incentives. Jobseekers with unstable employment trajectories (working either every other month or every other semester) are entitled to the same number of months and the same amount of unemployment benefits in Australia and France, but considerably less in Latvia and Spain, where contribution conditions are more demanding. In all of the four countries, workers with trajectories that include periods combining unemployment and partial work are entitled to lower unemployment benefit than workers with standard employment trajectories.
The chapter shows that there are several policy tools that can be deployed to design unemployment benefit systems that strikes the right balance between the sometimes conflicting objectives of income security and work incentives for those with unstable employment histories. For example:
Customised extensions of employment reference periods could be used for groups who are more prone to job instability, such as young workers and those on temporary contracts.
Earnings disregards and withdrawal rates applied to the benefits of jobseekers with earnings from casual or part-time work can be calibrated to enable such work to be used as a stepping stone to a better job.
Waiting periods, moderate minimum employment duration requirements, reductions in case of frequent reclaiming and limits on the accumulation of old and new entitlements can be used to reduce incentives to choose an unstable employment trajectory.
Differentiated unemployment insurance contribution rates could create financial incentives for employers and employees to choose more stable employment contracts and deter them from colluding in unstable arrangements that reduce workers’ pay and maximise benefit entitlements.
Finally, the integration or at least improved co-ordination of in-work and out-of-work benefits can enhance work incentives. Activation measures can also strengthen incentives to seek and accept stable employment.
References
[4] Avlijas, S. (2019), “The Dynamism of the New Economy: Non-Standard Employment and Access to Social Security in EU-28”, SSRN Electronic Journal, http://dx.doi.org/10.2139/ssrn.3354463.
[15] Baker, M., J. Gruber and K. Milligan (2015), “Non-cognitive deficits and young adult outcomes: the long-run impacts of a universal child care program”, NBER Working Paper, No. 21571, National Bureau of Economic Research, http://www.nber.org/papers/w21571.
[16] Berlinski, S. and S. Galiani (2007), “The effect of a large expansion of pre-primary school facilities on preschool attendance and maternal employment”, Labour Economics, Vol. 14/3, pp. 665-680, http://dx.doi.org/10.1016/j.labeco.2007.01.003.
[6] Boeri, T., P. Cahuc and A. Zylberberg (2015), “The Costs of Flexibility-Enhancing Structural Reforms: A Literature Review”, OECD Economics Department Working Papers, No. 1264, OECD Publishing, Paris, http://dx.doi.org/10.1787/5jrs558c5r5f-en.
[20] Booth, A. and J. van Ours (2013), “Part-time jobs: What women want?”, Journal of Population Economics, Vol. 26/1, pp. 263-283, http://dx.doi.org/10.1007/s00148-012-0417-9.
[30] Bradshaw, J. and F. Bennett (2017), ESPN Thematic Report on Access to social protection of people working as self-employed or on non-standard contracts United Kingdom, https://ora.ox.ac.uk/objects/uuid:f6d86772-fba0-4088-858c-9e96a9b1fcbf.
[33] Bureau of Labor Statistics (2020), Glossary : U.S. Bureau of Labor Statistics, https://www.bls.gov/bls/glossary.htm#D (accessed on 23 April 2020).
[24] Cahuc, P. (2018), “Wage Insurance, Part-Time Unemployment Insurance and Short-Time Work in the XXI Century”, IZA Discussion Papers, No. 12045, IZA, http://ftp.iza.org/dp12045.pdf.
[25] Cahuc, P. et al. (2019), “Taxation of Temporary Jobs: Good Intentions with Bad Outcomes?”, The Economic Journal, Vol. 130/626, pp. 422-445, http://dx.doi.org/10.1093/ej/uez062.
[43] Carcillo, S. et al. (2019), “Assessing recent reforms and policy directions in France: Implementing the OECD Jobs Strategy”, OECD Social, Employment and Migration Working Papers, No. 227, OECD Publishing, Paris, https://dx.doi.org/10.1787/657a0b54-en.
[17] Carta, F. and L. Rizzica (2018), “Early kindergarten, maternal labor supply and children’s outcomes: Evidence from Italy”, Journal of Public Economics, Vol. 158, pp. 79-102, http://dx.doi.org/10.1016/j.jpubeco.2017.12.012.
[29] Chetty, R., J. Friedman and E. Saez (2013), “Using differences in knowledge across neighborhoods to uncover the impacts of the EITC on earnings”, American Economic Review, Vol. 103/7, pp. 2683-2721, http://dx.doi.org/10.1257/aer.103.7.2683.
[11] Eurostat (2018), Methodological Guidelines and Description of EU-SILC Target Variables. (Version August 2017).
[32] Farber, H. (2017), “Employment, Hours, and Earnings Consequences of Job Loss: US Evidence from the Displaced Workers Survey”, Journal of Labor Economics, Vol. 35/S1, pp. S235-S272, http://dx.doi.org/10.1086/692353.
[23] Fernández, R., H. Immervoll and D. Pacifico (forthcoming), “Beyond repair? Anatomy of income support for standard and non-standard workers in OECD countries”, OECD Social, Employment and Migration Working Papers, OECD Publishing, Paris.
[9] Fontaine, F. and F. Malherbet (2016), CDD vs CDI : les effets d’un dualisme contractuel, Presses de Sciences Po, Paris.
[42] Hijzen, A. and A. Salvatori (2020), “Designing fair and work-oriented unemployment benefits: The case of Belgium”, OECD Social, Employment and Migration Working Papers, No. 237, OECD Publishing, Paris, https://dx.doi.org/10.1787/ac17d171-en.
[44] Immervoll, H. and C. Knotz (2018), “How demanding are activation requirements for jobseekers”, OECD Social, Employment and Migration Working Papers, No. 215, OECD Publishing, Paris, https://dx.doi.org/10.1787/2bdfecca-en.
[7] Kyyrä, T. (2010), “Partial unemployment insurance benefits and the transition rate to regular work”, European Economic Review, Vol. 54/7, pp. 911-930, http://dx.doi.org/10.1016/j.euroecorev.2010.02.005.
[39] Kyyrä, T., J. Arranz and C. García-Serrano (2019), “Does subsidized part-time employment help unemployed workers to find full-time employment?”, Labour Economics, Vol. 56, pp. 68-83, http://dx.doi.org/10.1016/j.labeco.2018.12.001.
[8] Le Barbanchon, T. (2016), “Optimal Partial Unemployment Insurance: Evidence from Bunching in the U.S”, undefined, https://www.semanticscholar.org/paper/Optimal-Partial-Unemployment-Insurance%3A-Evidence-in-Barbanchon/1ba1750d1779a5ff1c9b867c08413ac5efc393bc (accessed on 12 May 2020).
[18] Manning, A. and B. Petrongolo (2008), “The Part‐Time Pay Penalty for Women in Britain”, The Economic Journal, Vol. 118/526, pp. F28-F51, http://dx.doi.org/10.1111/j.1468-0297.2007.02115.x.
[19] Mas, A. and A. Pallais (2019), “Alternative Work Arrangements”, IRS Working Papers, No. 634, http://arks.princeton.edu/ark:/88435/dsp01bk128d78n (accessed on 7 April 2020).
[28] Meyer, B. and D. Rosenbaum (2001), “Welfare, the Earned Income Tax Credit, and the Labor Supply of Single Mothers”, The Quarterly Journal of Economics, Vol. 116/3, pp. 1063-1114, http://dx.doi.org/10.1162/00335530152466313.
[2] OECD (2020), Part-time employment rate (indicator), https://dx.doi.org/10.1787/f2ad596c-en (accessed on 12 March 2020).
[46] OECD (2020), “Supporting people and companies to deal with the Covid-19 virus: Options for an immediate employment and social-policy response”, OECD Policy Responses to Coronavirus (COVID-19), OECD Publishing, Paris, http://www.oecd.org/coronavirus/policy-responses/supporting-people-and-companies-to-deal-with-the-covid-19-virus-options-for-an-immediate-employment-and-social-policy-response-d33dffe6/.
[38] OECD (2020), TaxBEN: The OECD tax-benefit simulation model - Methods and policy applications, OECD, https://taxben.oecd.org/tax-ben-resources/OECD_TaxBEN_methodology.pdf.
[3] OECD (2020), Temporary employment (indicator), https://dx.doi.org/10.1787/75589b8a-en (accessed on 12 March 2020).
[1] OECD (2019), OECD Employment Outlook 2019: The Future of Work, OECD Publishing, Paris, https://dx.doi.org/10.1787/9ee00155-en.
[35] OECD (2019), Taxing Wages 2019, OECD Publishing, Paris, https://dx.doi.org/10.1787/tax_wages-2019-en.
[40] OECD (2018), Good Jobs for All in a Changing World of Work: The OECD Jobs Strategy, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264308817-en.
[5] OECD (2018), OECD Employment Outlook 2018, OECD Publishing, Paris, https://dx.doi.org/10.1787/empl_outlook-2018-en.
[13] OECD (2017), The Pursuit of Gender Equality: An Uphill Battle, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264281318-en.
[22] OECD (2015), In It Together: Why Less Inequality Benefits All, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264235120-en.
[10] OECD (2014), OECD Employment Outlook 2014, OECD Publishing, Paris, https://dx.doi.org/10.1787/empl_outlook-2014-en.
[45] OECD (2014), Society at a Glance 2014: OECD Social Indicators, OECD Publishing, Paris, https://dx.doi.org/10.1787/soc_glance-2014-en.
[12] OECD (2013), “Back to work: Re-employment, earnings and skill use after job displacement”, in OECD Employment Outlook 2013, OECD Publishing, Paris, https://dx.doi.org/10.1787/empl_outlook-2013-8-en.
[14] OECD (2010), “How Good is Part-Time Work?”, in OECD Employment Outlook 2010: Moving beyond the Jobs Crisis, OECD Publishing, Paris, https://dx.doi.org/10.1787/empl_outlook-2010-5-en.
[21] OECD (2010), OECD Employment Outlook 2010: Moving beyond the Jobs Crisis, OECD Publishing, Paris, https://dx.doi.org/10.1787/empl_outlook-2010-en.
[37] OECD (2007), Benefits and Wages 2007: OECD Indicators, OECD Publishing, Paris, https://dx.doi.org/10.1787/ben_wages-2007-en.
[36] OECD (2005), OECD Employment Outlook 2005, OECD Publishing, Paris, https://dx.doi.org/10.1787/empl_outlook-2005-en.
[41] OECD (2004), OECD Employment Outlook 2004, OECD Publishing, Paris, https://dx.doi.org/10.1787/empl_outlook-2004-en.
[31] OECD (2002), OECD Employment Outlook 2002, OECD Publishing, Paris, https://dx.doi.org/10.1787/empl_outlook-2002-en.
[26] Saez, E., M. Matsaganis and P. Tsakloglou (2012), “Earnings Determination and Taxes: Evidence From a Cohort-Based Payroll Tax Reform in Greece”, The Quarterly Journal of Economics, Vol. 127/1, pp. 493-533, http://dx.doi.org/10.1093/qje/qjr052.
[27] Saez, E., B. Schoefer and D. Seim (2019), “Payroll taxes, firm behavior, and rent sharing: Evidence from a young workers’ tax cut in Sweden”, American Economic Review, Vol. 109/5, pp. 1717-1763, http://dx.doi.org/10.1257/aer.20171937.
[34] United Nations Statistical Division (2008), International Standard Industrial Classification of All Economic Activities (ISIC), United Nations Publications.
Notes
← 1. The impact of the demographic change on the incidence of standard and non-standard employment is mainly due to an “age effect” rather than a “cohort effect”. Even among young workers, the incidence of standard employment increases as a cohort ages.
← 2. Average monthly wages were computed dividing the annual wage of the current year by the number of months in work. Since people experiencing many transitions are unlikely to have changed status at the first or last calendar day of each month, each month in work is likely to correspond to fewer hours of work than, say, a standard employee staying in the same job the whole year. This is particularly the case for those that are more employable and therefore are less likely to spend most of the month out of work. Therefore, the prevalence of unstable employment among below-median wages could in part driven by the underestimation of their wages.
← 3. Previous OECD evidence suggests, however, that the average number of hours worked by part-timers does not explain much of the cross-country difference in poverty penalties (OECD, 2010[21]).
← 4. Results from Table 2.1 also suggest that workers with unstable employment have lower wages, although these results control for the number of months and not for the number of hours worked.
← 5. Jobseekers who are not entitled to first- or second-tier unemployment benefits may be protected by safety-net programmes, such as social assistance benefits, with neither employment nor employment-availability requirements.
← 6. Some countries have unemployment benefits available for jobseekers not entitled to first-year and second-tier unemployment benefits. In Spain, for example, Active Integration Income (Renta Activa de Inserción) and Extraordinary unemployment allowance (Subsidio extraordinario por desempleo) are aimed at jobseekers in long-term unemployment who have extinguished their access to unemployment insurance and unemployment assistance benefits.
← 7. All benefit amounts assume a single individual without children and not eligible to any supplement. For other family circumstances, see annual OECD tax-benefit country reports.
← 8. In 2020, the minimum base of social contributions in Austria was 461 EUR per month.
← 9. In France, the benefit amount is reduced only for jobseekers with previous high wages (i.e. those who earned more than EUR 4 500 per month on average before unemployment).
← 10. In Latvia, one can suspend the benefit for up to two months of work without losing entitlement.
← 11. In Canada, entitlements are only retained if they were suspended due to high earnings from partial work. In the Netherlands, entitlements are only kept if the new job provides the same amount of working hours as the job prior to unemployment or if earnings from the new job are below 87.5% of the unemployment benefit level. In the United Kingdom, entitlements are kept for 12 weeks.
← 12. In Figure 2.9, Figure 2.10 and Figure 2.12 (UB_2), waiting periods can be identified as the first and lower amount of benefit received over a period. Waiting periods shift the time of receipt, but not the potential duration and potential total amount of payments, which remain the same. Yet, in comparison to the case without waiting period, the effective duration and effective total amount may be lower if the jobseeker finds a job before exhausting the benefit.
← 13. In Figure 2.11 and Figure 2.12 (UB_1), reductions due to the length of the receipt periods can be identified as the lower amounts once the benefit has been received for a few months.